ZFG Suppl 57 3

Zeitschrift für Geomorphologie Vol. 57 (2013), Suppl.
3, 001 Editorial
1
Stuttgart, SeptemberRainfall-runoff
2013 -erosion relationships study
Quaternary Geomorphology in Tectonically Active Areas with

Emphasis in Fluvio-Coastal Processes
Quaternary Geomorphology is the study of the Holocene and Pleistocene Epochs, based on the
principle of ‘Uniformitarianism’, introduced by James Hutton. Every process that takes place on
the relief at present is the key for the study of the past. This is the milestone of Geomorphology.
Based on this principle, many studies have been accomplished, in order to satisfy man’s thirst for
knowledge and his need to understand his natural environment, his home.
Late Quaternary is the period of man’s domination. During this geologic time span, man
builds civilisation. Man’s need for understanding the natural processes carving Earth’s surface,
in order to preserve his built environment, was early understood. The necessity for this compre-
hension was major and remains in this level in tectonically active areas. They are the most active
geological regions on Earth. Although they pose a challenge to modern engineering, they are also
considered as the key to unlocking the geologic history that has been recorded in landforms.
In the framework of the 15th Joint Geomorphological Meeting (JGM) in Athens, Greece,
under the auspices of the Romanian, Italian, Belgian, French and Greek Associations of Geo-
morphologists, a number of studies were announced, oriented at Quaternary Geomorphology
in tectonically active areas. The most prominent contributions are documented in this volume
of Zeitschrift für Geomorphologie. Our sincere gratitude goes to the Editorial Committee and
especially to the Editor-in-Chief, for accepting to host these contributions in this special issue of
the journal as well as to all colleagues who accepted reviewing the following papers.
Niki Evelpidou
Assistant Prof. in Geomorphology
Kosmas Pavlopoulos
Professor in Geomorphology
President of the Working Group on Geoarchaeology (IAG)
© 2013 Gebrüder Borntraeger Verlagsbuchhandlung, Stuttgart, Germany www.borntraeger-cramer.de

DOI: 10.1127/0372-8854/2013/S-00140 0372-8854/13/S-00140 $ 0.25
2 Carla Sofia Santos Ferreira et al.
Zeitschrift für Geomorphologie Vol. 57 (2013), Suppl. 3, 003-028 Article
3
Hydrogeomorphological
Stuttgart, September 2013 Vulnerability in the Romanian Plain
Hydrogeomorphological Vulnerability in the Romanian Plain

Florina Grecu, Liliana Zaharia and Cristina Ghiţă
with 11 figures and 9 tables
Abstract. Based on geomorphometrical and morphological theories and concepts, we aim at diagnosing
the vulnerability to hydrogeomorphological hazards in the Romanian Plain. The study focuses on specific
geomorphic, climatic and hydrological factors influencing the channel and interfluve dynamics in four main
watersheds: Vedea (including its tributary Teleorman River), Călmăţui, Neajlov and Mostiştea. Each factor
is analyzed in order to identify the threshold values for the occurrence of the geomorphological and hydro-
logical processes that negatively impact the environment and society. At the same time, based on certain
specific parameters of the three discussed factors, we have accomplished a multi-criteria analysis that high-
lights the vulnerability to hydrogeomorphological processes of the investigated watersheds. The basic meth-
ods are the analytical and synthetic ones, based on graphical and cartographical models. We could mention
the use of graphical and cartographical data representation.
Key words: hydrogeomorphological vulnerability, critical values, meanders, channel, geomorphometry,

Romanian Plain
1 Introduction and objectives
The paper aims at establishing the critical values concerning the dynamics of some of the factors
that induce alterations with negative effects on the environment and on the society (Goudie 2000).
This is done by using modern concepts and methods with broad applicability in various fields
of knowledge, including the dynamics of hydrogeomorphological or hydromorphological systems
(the channels and interfluves that are part of a catchment) (Malavoi & Bravard 2010). From this
perspective, the Horton-Strahler morphometric method finds applicability in the study of fluvial
bed forms, slope topography and river basins (through the index of development of the drain-
age system, the indexes of channel morphometry and the indexes regarding the morphometry of
the minor landforms occurring on the interfluves). This analysis highlights the present state and
dynamics of the minor landforms as an effect of hazard phenomena. If morphometry has a pre-
cise quantitative content, modern morphological theories use a qualitative approach, thus allowing
the detection of some thresholds and critical values, which trigger the natural hazards. The most
important from this perspective is the catastrophe theory, which explains the discontinuities and
thresholds (Brunet 1970, Boutot 1996, Grecu 1999, 2003, Ielenicz 1998, Petrea 1998, 2005).
2 Study area
The investigated area is represented by some hydrogeomorphological systems within the plain
area. The Romanian Plain (52,699 km2 together with the Danube floodplain and accounting for
DOI: 10.1127/0372-8854/2013/S-00141 0372-8854/13/S-00141 $ 6.50
4 Florina Grecu et al.
one fifth of the Romanian territory) is situated in the central-southeastern part of Europe and in
the southern part of Romania, stretching along the Danube, which borders it on the south and
east. To the north, however, the plain is bordered by other physiographic units, namely the Getic
Plateau, the Curvature Subcarpathians and the Moldavian Plateau (Fig. 1). The contact with
these major landforms influences terrain dynamics and especially the morphology and evolu-
tion of river channels. The Romanian Plain overlaps the sedimentation basin located on the
inside of the Carpathian-Balkan arc. Actually, judging from its genesis, it turns out to be a flu-
vial-lacustrine Quaternary plain (Vâlsan 1916, Liteanu 1969, Mihăilescu 1924, Rădulescu
1956, Coteţ 1976, Posea 1997, 2002, Grecu 2010, Grecu et al. 2009b). The Danube River,
whose course came into being in different stages while the Pliocene-Pleistocene Lake receeded
to the east and northeast, has contributed to the genetic diversity of the Romanian Plain’s topo-
graphy and at the same time, has imposed the general orientation of the river network.
The Romanian Plain is crossed from north to south or from west to east by a number of
tectonic lines and areas subjected to uplift or subsidence movements (Fig. 2). The formation of
the Romanian Plain is influenced by the general uplifting affecting not only the Carpathians
and the Balkan Mountains, but also the platform units overlapping the sedimentary depres-
sion. The direction of the movements has been responsible for the increased erosion and the
Fig. 1. Geographical position of Romanian Plain (at east of Olt River) (A) and the analyzed watersheds (B).
Hydrogeomorphological Vulnerability in the Romanian Plain 5
Fig. 2. Recent vertical crustal movements map of Southeastern Romania (after Zugrăvescu et al. 1998).
Legend: 1 – seismic lines (mm/year); 2 – active crustal faults; 3 – areas affected by diapirism.
emersion of the present territory of the Romanian Plain. The basement is represented by the
Wallachian Platform, which dips to the north. The northeastern part is nothing else but an
extension of the basement and the specific geological structures of the Central and Eastern
Dobrudja (Fig. 2) (Badea 2009, Zugrăvescu et al. 1998). The present study investigates some
representative rivers (Vedea with Teleorman, Călmăţui and Neajlov with Câlniştea) flowing
through the central section of the Romanian Plain – a physiographic unit bordered on the
west by the Olt River and on the east by the Argeş River (Fig. 1, Table 1). The Mostiștea water-
course, located on the est side of the Argeş River is also considered in the morphometrical
analysis. The geological background, but mostly the hydrogeological features (the depth and
discharge of groundwater) of the studied catchments strongly influence the flowing regime.
Rivers are fed in moderate percentage by groundwaters, with significant variations from west
to east (from 15 % in the case of Mostiştea up to 20–25 % in the case of Vedea and Neajlov)
(Liteanu 1969). The loessoid deposits overlying the investigated catchments are sometimes
25 m deep, thus influencing the position of aquifers and their relationship with the main
drainage system.
Table 1. Morphometric features of the main rivers in the study area (data taken from the Aquaproiect 1992).
Nr. River L Hizv HGV P*med Ks S Hmed

crt. (km) (m) (m) (‰) (km2) (m)
1 Călmăţui 139 157 17 1 1.67 1413 104
2 Vedea 224 504 16 2 1.39 5,430 169
1
3 Cotmeana 93 597 134 5 1.88 498 306
1
4 Teleorman 169 390 29 2 1.27 1,427 148
5 Neajlov 186 315 43 1.5 1.47 3,720 123
2
7 Câlniştea 112 117 46 1.1 2.10 1,748 101
8 Mostiştea 98 91 13 0.8 1.3 1,758 60
L = length, Hizv = elevation at the source, HGV = elevation at the mouth, Pmed = mean channel slope,
Ks = sinuosity coefficient, Hmed = mean elevation of the catchment, S = area of the catchment.
*) The gradients have been computed by the authors as the ratio of the difference between elevation at the
source and elevation at the mouth to the river’s length.
1
Tributary of the Vedea River, 2 Tributary of the Neajlov River
3 Data and methodology
The analysis is based on several types of data:

• Cartographic data extracted from topographical maps (scales of 1:25000 and 1:50000,
1970, 1972, 1977, 1993), ortophotoplans (scale 1:5000, 2006) and thematic maps (geo-
logical, hydrogeological, pedological, climatic).
• Data from field observations and measurements (topographic surveys), interpreting the
long profiles, graphics and schemes.
• Climate (air temperature, precipitations) and hydrological data (liquid discharge and
suspended sediment load) on weather and hydrometrical stations in the study area; these
data were taken from different sources (that were mentioned in the text), and the ana-
lyzed periods are variable, between 1961 and 2007.
The methods used mainly include: morphometric method of relief using DEM and Horton-
Strahler hierarchization of the river network; the analysis of the cartographic data and aerial
images, performed using GIS and remote sensing techniques; statistical methods.
Field observations and mapping were widely used.
4 Analysis and results
4.1 Geomorphological context (critical values concerning the landforms and hazard vulner-
ability)
The Romanian Plain’s topography includes a variety of landforms created by the gradational
agents, and especially by the allochtonous and autochthonous rivers, which are responsible for
the sedimentation of fluvial-lacustrine deposits. The shaping has taken place in distinct stages
(beginning with the Upper Pliocene and the Quaternary) and due to the different geological
deposits, specific neotectonics and local hydrogeology; it has led to the formation of several
genetic types of plains: piedmontane plains, glacis plains, terrace plains, subsidence plains and
fluvial-lacustrine plains (Coteţ 1976, Posea 1997, 2002, Grecu 2010). The interfluves, terraces
and floodplains, having morphometric and morphological features imposed on the one hand by
the geological and hydrogeological conditions and on the other hand by the geographical setting,
stand out clearly in the topography.
4.1.1 The analysis of some morphometric terrain parameters and the identification of critical
values
The hypsometric steps have been established depending on the terrain specific features and the
area of each catchment. The most numerous steps (6) are found in the Neajlov catchment, while
in the Mostiştea basin there are only four. As for the Vedea and Călmăţui watersheds, we have
identified five characteristic steps. The highest one, above 500 m altitude, lies in the Vedea catch-
ment, while the lowest one, below 40 m, is in the Mostiştea basin. More than half of the area of
the central part of the Romanian Plain develops at elevations between 50 and 130 m. For every
catchment, the smallest areas belong to the upper step. Thus, in the Vedea catchment this step
develops between 500 to 610 m and has an area of approximately 80 km2, accounting for 2 % of
the territory. In the Neajlov catchment the step lies above 250 m, its area is 40.7 km2, while the
share is only 0.91 %, in the Calmățui catchment it lies above 150 m (Albu 2011) whereas in the
Mostiştea catchment it is found above 80 m altitude and has a share of 6 %. Hypsometric analysis
is very important, inasmuch as these steps include the main landforms of the river basins (flat
interfluves, terraces and floodplains).
The local relief. With the exception of the Neajlov catchment, where five value intervals
have been established, for the other three basins we have retained only four intervals. The high-
est values of the relative relief are found in the upper catchment of the Vedea River (more than
80 m/km2), while the lowest (10 m/km2) are common for all the catchments (Table 2).
The values above 80 m/km2 that are specific for the Vedea catchment show a strong deepen-
ing of the riverbed in relation to the initial topographic surface. The interval with values less than
10 m/km2 characterizes most of the catchments’ area. As a matter of fact, these are typical values
for the plain regions and they are explained by the presence of many small and shallow valleys.
On the scale of every study catchment, the share of this interval exceeds half the entire area:
Vedea – 64 %, Neajlov – 85 %, Călmăţui – 53 %, and Mostiştea – 61 %.
The drainage density was computed as the ratio of the total length of the streams to the total
area of the catchment. For the investigated territories, the values range from 0 to 6 or 7 km/km2.
The lowest ones are specific for the Mostiştea and Neajlov catchments (0 – 0.1 km/km2 and less
than 0.5 km/km2 respectively). It is worth noting that for all the river basins, the areas character-
ized by the minimum values (flat interfluves and tracts of land scarred by minor valleys) have the
highest shares (Mostiştea 73 %, Neajlov more than 60 %, Călmăţui 40 % and Vedea 37 %). Likewise,
one can see that the areas with the highest values of local relief lie along the main watercourses and
at the junctions, having shares of less than 20 %. In the Calmatui catchment, the drainage densities
higher than 3 km/km2 account for 17 % of the area, in the Neajlov catchment those higher than
2.5 km/km2 represent 14 %, in the Mostiştea catchment the values higher than 0.7 km/km2 (cover
10 %), whereas in the Vedea basin the densities above 4 km/km2 (account for 5 %).
Slope gradient intervals range from less than 1° to more than 15° (Table 2). The higher gra-
dients (between 5 and 15°) are scarcely recorded within the plain area, which explains the low
shares of the territories characterized by such values. However, in the case of the Vedea catch-
ment, the slope gradients higher than 15° hold significant shares in the piedmontane and transi-
tion sections (10 %). Usually, such areas are common for the terrace scarps and riverbanks. As far
as the quasi-horizontal surfaces are concerned (with gradients of 1 to 2°), these are found within
the flat interfluves, the top of the terraces and the floodplains. On the catchment level, the values
differ as follows: Vedea 40 %, Neajlov 70 %, Călmăţui 77 % and Mostiştea 94 %.
4.1.2 The analysis of the river long profiles
The mean channel gradient of the Vedea River is 2.1 m/km, but in its upper stretch, the gra-
dient is higher than 5 m/km. The stream deepens its valley (which is almost rectilinear) into
the permeable deposits represented by the Cândeşti gravels and the loessoid formations overly-
ing them, which facilitates water percolation. Consequently, the discharges are low and the flow
gets an intermittent character, even more because the water table lies very deep. As soon as the
river enters the plain, the slope gradient gets lower (for instance downstream the junction of
the Teleorman it is only 0.3 m/km) and therefore the accumulation, winding and lateral cutting
intensify. The appearance of the long profile and the oblong shape of the catchment influence
flood wave formation and propagation.
Judging from the shape of the long profiles, the rivers flowing through the plains do not
show a high curvature, but instead are approximately uniform, which suggests a recent genesis.
The most developed proves to be the Vedea River (Fig. 3), the concavity of which is higher in the
piedmontane sector, while the valleys of the Călmățui, Neajlov and Mostiştea streams (on their
plain stretches) show a gentle slope. The tilt angles of the long profiles differ slightly. For instance,
one can see gradients between 2 and 2.5 m/km in the upper stretches of the Călmăţui and Neajlov,
as well as in the middle section of the Vedea River, values imposed by the lithology in which the
gravel prevails. On the other hand, there are also gradients less than 2 m/km and even less than
1.5 m/km (Mostiştea), which are found in the lower stretches of all the streams. Usually, these are
imposed not only by the relatively flat and even topography, but also by the many ponds that can
be seen along the main streams (Mostiştea, Neajlov, Călmăţui) and their tributaries.
The nick points along the rivers flowing through the plain are found at the confluences, where
tributaries have laid down high amounts of sediments. Such thresholds occur at 280–290 m, 230 m,
100 m and 50 m along the Vedea, 140–160 m, 80 m, 60–80 m and 40–60 m along the Neajlov and
90–100 m, 80 m, 40 m and 30 m along the Călmăţui. As far as the Mostiştea is concerned, the
thresholds are almost inexistent, and those that do occur are very small and can be seen only near
the dams (inasmuch as the river has been turned into a chain of pools).
Table 2. Synthetic data’s catchment morphometric basins.
Drainage Surface Max Min. L river Hypsometry Local relief Drainage Slope Stream Draiange Horton-
basin (km2) alt. alt. (km) (m) (m/km2) density (0) Frequency Density Strahler
(m) (m) (km/km2) (N1/S) (∑L/S) (order)
max min max min max min max min
Vedea 5266 619 16 251 >500 <50 >80 <20 >4 <1 >15 <1 1.98 2.34 7
Neajlov 3720 258 43 186 >300 <50 >20 <5 >2,5 <1 >5 <1 0.57 1.32 7
Călmătui 1375 158 20.2 143 >120 <20 >50 <10 >3 <1 >15 <2 1.04 1.26 5
Mostiștea 1750 92 12.4 11.5 >80 <40 >20 <8 >1 <0.1 >12 <3 0.25 0.45 5
N1 – segments of 1st order stream in Horton-Strahler System, L – river lengh, S – drainage basin surface
Hydrogeomorphological Vulnerability in the Romanian Plain
9
Fig. 3. Long profiles of Vedea river (after Grecu et al. 2010 a).
4.1.3 The analysis of the drainage system morphometry
This analysis highlights the headward erosion and the vertical cutting, through the number and
length of the low order river segments (Grecu 1992). Excepting Mostiştea River, the studied riv-
ers originate from the alluvial deposits laid down by the big rivers as soon as they enter the plain
(at the contact with the Getic Piedmont and the Sub-Carpathians). The fact is also proved by the
high values of the confluence ratio for the low order channel segments (6.54 in the case of the
Călmăţui) (Albu 2011), an abnormal situation for a river flowing through the plain (Table 3).
According to the Horton (1945) – Strahler’s (1952) system of stream ordering, Neajlov
River, downstream the junction with the Câlniştea (its most important tributary, Cocoş et al.
2011), becomes a 6-order stream, although in its upper course, upstream the junction with the
Dâmbovnic, it is only a 4-order channel. In its turn, the Vedea River becomes a 7-order stream
immediately downstream the junction with the Teleorman, its main tributary. It is worth not-
ing that the Cotmeana, the most important tributary of the Vedea River in the upper course, is
a 6-order stream. At east of Argeş River, the Mostiştea autochthonous river is a 5-order stream.
The confluence ratios of the low order streams and the degree of development of the drainage
systems in terms of the stream length order and the stream average lengths give some clues on
the channels’ dynamics. The development index shows the degree of development of every catch-
ment with respect to every parameter (Grecu 2004, 2008). Thus, in its upper stretch, the 4-order
Neajlov is well developed as far as the number of stream segments is concerned (IN = 105 %), but
underdeveloped at the length of stream segments (IL = 59 %) and the mean lengths of stream seg-
ments (Il = 56 %). The values suggest the high drainage density of this plain sector, characterized
by a mean confluence ratio (RC) of 6.63, which points at an atypical tendency for the rivers flow-
ing through the plain (Grecu et al. 2009a).
Table 3. Synthetic geomorphometric data (Horton-Strahler system).
Index
Mostiştea
Călmăţui
Neajlov
basins
basins
Vedea
Catchements
Order 7 7 5 5
Rc 4.8 3.89 5.2 6.54
Ir (%) 63 27 57 98
RL 2.2 2.1 2.7 2.02
IL (%) 41 34 89 63
Rl 2.9 1.85 1.92 3.24
Il (%) 121 56 52 66
Rc – confluence ratio, Ir – realization index, RL - length ratio; IL- length of stream segments;
Il - mean lengths of stream segments
The undersized sum of the stream lengths shows the torrential character of the stream seg-
ments, most of them of first and second order, represented by gullies and small valleys, with inter-
mittent flow and mean lengths varying from one to four kilometers (Table 2).
4.1.4.The morphometric analysis of meanders – the lateral cutting
The analysis of river meanders is aimed at estimating the channel’s horizontal dynamics by lat-
eral erosion, a process controlled especially by lithology, slope gradient and some local factors
(the river discharge, although extremely important, has been left aside as it is not subject to this
study). The measuring of the several meanders’ parameters was accomplished on characteris-
tic stretches on the rivers Vedea (Poroschia and Ţigănești), Mostiştea (Tămădău – Gurbăneşti)
(Fig. 4A) and Câlniştea (Drăgănești Vlaşca – Cămineasca) (Fig. 4B).
The study section lies south of Alexandria, between Poroschia and Ţigăneşti settlements,
at an elevation of 50 m. Here, the channel slope gradient is very gentle (less than 1°). On this
stretch, the river has a sinuous length of 15.6 km and shows four meander loops. According to
the morphometric theories, rivers are considered “meandered” when their meandering coeffi-
cient is higher than 1.5 (Bravard & Petit 2000, Grecu & Palmentola 2003). In the case of the
Vedea River, the meandering coefficient is 1.26, which means the channel is not meandered, but
sinuous. This coefficient is computed as the ratio of the sinuous length of the river to its straight
length (12.4 km), hence 15.6/12.4 = 1.26. This value is rather close to 1.5, which suggests the
stream is in transition stage from a sinuosity course to a meandering one. Both the sinuosity and
the straight length of the loops decrease downstream, from 5.6 km to 4 km, which may point out
at a future stage when lateral erosion will be prevailing.
The Câlniştea is an atypical watercourse for the Romanian Plain because of its flowing
direction from west to east, which in its middle and lower courses seems to be imposed by
a fault (Vâlsan 1916, Coteţ 1976). At the same time, the river exhibits a sequence of tight
Fig. 4. The morphometry of the meanders of the Mostiştea (A) and Câlniştea (B) rivers (after Grecu et al.
2010b).
meanders (Fig. 4). The comparative and evolutionary analysis took into account the dynam-
ics of the Câlniştea channel based on the Szatmary map of 1856 and the topographic map of
1972. In 1856, the meanders’ length had a growing tendency, but by 1972 this tendency ceased
and the meanders’ length exhibited a relative steady trend (Văcăru 2010). As for the sinuous
length, there was an obvious decreasing tendency, which also explains the drop of sinuosity
coefficient.
The Mostiştea channel, which is artificially improved, is carved into loessoid deposits and
sand formations (The Sands of Mostiştea). In its middle course (the Tămădău-Gurbăneşti stretch),
the meanders are complex (Fig. 4 right). On this stretch, the meandering coefficient is 2 and the
slope gradient is very low, which points at an advanced stage of evolution in the soft loessoid
deposits. The meanders’ length increases from upstream to downstream, from 2.15 to 4.1 km,
with an average of 3.36 km. The amplitude of the loops increases in the same direction, the high-
est value being specific for the meander lying near Gurbăneşti. The same meander has the high-
est sinuosity coefficient (3.41), because it is very tight and complex, thus suggesting an advanced
development of the river.
4.2 Critical climatic values for defining land vulnerability to geomorphological hazards
4.2.1 Multiannual analysis and the identification of threshold values (precipitations)
The precipitation is a major meteorological element that triggers, maintains and reactivates some
geomorphological processes, thus acting as a shaping agent in the plain areas. This influence
can be highlighted by quantifying the values of selected characteristic variables (duration, fre-
quency and intensity), as well as by delimiting in time the individual and successive snapshots
that present a certain degree of susceptibility.
The mean multiannual amount of precipitation specific for the Vedea catchment drops on the
north-south direction, from 569.3 mm at Stolnici, in the piedmont area, to 520.3 mm at Alexan-
dria s.w. (according NMA). The maximum is recorded in August in the north (78.6.mm) and in
July in the south (65.9 mm), whereas the minimum values are specific for January and February
in the north (32.6 mm) and October in the south (28.2 mm).
The multiannual regime emphasizes considerable departures from the mean. Thus, the high-
est annual amounts may reach in the rainy years more than 700 mm, both at Stolnici w.s. (1966,
1971, 1972, 1979, 1980) and at Alexandria w.s. (1964, 1966, 1969, 1972), years when floods were
frequent and water table rose. The lowest amounts, below 400 mm, are specific for the dry years:
at Stolnici w.s. in 1990 and 2000 and at Alexandria w.s. in 1965, 1985, 1990, 1996 and 2000).
The precipitation shows a slight decreasing linear tendency both in the north and in the south
(Fig. 5).
According to the Angot index (K), the land vulnerability classes have values ranging from less
than 1 to more than 2.50 (Dragota 2006, Agenşia Naţională de Meteorologie 2008).
For the Vedea catchment, the Angot index values for the northern section of the catchment
(at Stolnici) are more than one all year round, with a maximum in July (2.5). In the lower sec-
tion, at Alexandria w.s., the maximum is reached in July (2.19), while in October the value is less
Fig. 5. Variability of annual quantities of precipitation and their trends (Stolnici and Alexandria w. s.) (1961–
2000). Red line – linear trend at Stolnici; mauve line – moving average on 5 years at Stolnici; black line –
moving average on 5 years at Alexandria.
Fig. 6. Variation of monthly rainfall index Angot – Vedea catchment (1961–2000).
than one (0.88) (Fig. 6). This index shows that in this area of the Romanian Plain the amount of
precipitation is higher than in its eastern part, where the values more than one are specific only
for the interval April – September. For instance, in the Moştistea catchment the maximum value
of the Angot index is 1.47 in July (Ghiţă & Dragotă 2010). The index also emphasizes that the
driest months belong to the cold season.
During the period 1961–2000, the driest years did not favor the triggering of geomorpho-
logical processes because the pluvial impact was extremely low. At Stolnici w.s., the values of the
Angot index for the years 1961, 1962, 1977, 1982, 1983, 1985, 1986, 1988, 1989, 1990, 1992, 1994
and 2000 did not exceed 1.5. The driest of them was 1992, when the index value was less than 1,
while the rest of the years recorded values higher than 1. At Alexandria weather station, dry years
are considered only those characterized by index values less than 1.2, as for instance 1961, 1963,
1965, 1982, 1985, 1988, 1989, 1990, 1992, 1996, 1999 and 2000. For the southern part of the study
area the most significant years are 1985, when all the months had values less than 1.2, and 1990,
1992 and 2000, when the averages kept below 1.
By taking into account the values of the Angot index, we were able to identify a series of
rainy years as well, when K was greater than 2, which means the land was highly vulnerable to
pluvial erosion. The Stolnici weather station, lying to the north, shows a longer series of rainy
years, namely 1966, 1971, 1972, 1979 and 1980, while the Alexandria weather station, lying to the
south, recorded only three such years, namely 1966, 1969 and 1972. It should be noted that 1966
and 1972 appear in the datasets of both weather stations as rainy years, which proves that at that
times the entire Vedea catchment was confronted with heavy rainfalls. To the north, at Stolnici
weather station, in 1972 only four months (May, August, September and October) recorded val-
ues higher than 2, while southwards, at Alexandria w.s. there were five such months (May, July,
August, September and October). In October 1972, the Angot index reached its maximum values
at both weather stations (9.2 at Stolnici and 5.61 at Alexandria w.s.), which is a sign that the entire
catchment was suffering serious flooding (Zaharia et. al. 2011).
On a monthly scale, of the 960 possible cases (the total number of months in the 37 investi-
gated years) the highest frequency (40 %) is hold by the very low vulnerability class (Table 4). In
the cold semester, the very low and low vulnerability values are prevalent, while the relative fre-
quencies of the high and very high classes are low. In the warm semester, the very high and high
vulnerability classes prevail, but the medium class has a significant relative frequency.
Vulnerability classes (from Angot index) (Cîrciumaru 2011)
In order to get a comprehensive picture regarding the degree of influence of the precipitation
amount on the hydromorphological processes the study has also taken into account the monthly
frequency of the vulnerability classes for the Mostiştea catchment, which lies in the eastern section
of the Romanian Plain and experiences different climatic conditions.
For the Vedea catchement one can see that 40 % of the cumulated precipitation does not have
a significant impact on the landscape (K < 0.99) in terms of triggering or reactivating the geomor-
phological processes. With irrelevant exceptions, this situation is specific for the winter months,
which in this area are generally dry (Table 4).
At the same time, 18 % of the monthly amounts of precipitation, which lead to Angot index
(K) values between 1.00 and 1.49, impose a low vulnerability to the land (Table 5). These precipi-
tations are distributed all year round, with some differences introduced by the interval of the year
when they occur. At Stolnici w.s, only January has 10 cases, whereas at Alexandria w.s. there are
15 cases in April and 10 in November. The occurrence probability below 20 % affects the entire
catchment all year round, while the values higher than 20 % are recorded only in January, April
and November.
The moderate vulnerability classes (III and IV), which show the considerable impact of pre-
cipitation on the landscape through the triggering of geomorphological processes have an annual
occurrence probability of 14 %. However, the role of precipitation must be weighted by taking
into account the properties of loess deposits that cover the south of the catchment, the hydrogeo-
logical and morphological properties, as well as the land use in the region. One should note that
in the warm season the K values are clustered.
Table 4. Frequency of vulnerability classes, monthly quantities of precipitation (at Stolnici, Alexandria and
Fundulea w.s.).
Absolute frequency Cumulative frequency Relative frequency

K index (no. of cases) (no. of cases) (%)
Vulnerability
values Mostiştea Mostiştea Mostiştea
clases
(Angot) Vedea (Ghiţa & Vedea (Ghiţă & Vedea (Ghiţă &
Dragotă 2010) Dragotă 2010) Dragotă 2010)
I – very low < 0.99 386 348 960 444 40 78.3
II – low 1.00 – 1.49 246 74 574 96 18 16.6
III – moderate 1.50 – 1.99 139 16 399 22 14 3.6
IV – high 2.00 – 2.49 101 5 260 6 11 1.12
V – very high > 2.50 159 1 159 1 17 0.22
The upper vulnerability classes, with monthly values greater than 2 or 2.50, which favor the
most the dynamics of the channel and slope processes, and especially the floods, are mainly spe-
cific for the warmest months of the year (the interval May–July). Higher amounts of precipitation
are recorded in June and July, both in the north (with 19 and 18 cases) and in the south (with 11
and 17 cases respectively). The occurrence probability of some intense geomorphological pro-
cesses is specific for May, June and July, when the frequency exceeds 30 %.
The highest annual amounts of precipitation define the very high vulnerability class (K > 2.50).
In an interval of 40 years, the decade 1961–1970, only in 1966 the K value reached 2.60. This hap-
pened in the north of the catchment, where the impact of precipitation on the land is obvious,
especially on the slopes with higher gradients, which favor the formation of gullies and torrents.
If we take into account that for the entire investigated period of time there were 8 cases with
moisture surplus (K > 2.00), representing a frequency of about 17 %, we may contend that at least
in the north of the Vedea catchment (where K > 2.5) geomorphological processes are to a great
extent the result of the high amounts of precipitation. The amplest of these processes are trig-
gered only in the rainy years and especially in the areas with maximum vulnerability.
Likewise, the extreme pluviometric events that are mirrored by the land vulnerability to ero-
sion have brought about significant changes in a short period of time. We can mention in this
respect the very high values of the monthly Angot index (K > 2.5) at Alexandria w.s., when the
abundant precipitation made the Vedea River overflow its banks. Such repeated events led to sig-
nificant changes of the river channels (lateral erosion, stream piracy and accumulation) and to
the weakening of the dams built across the Vedea River and its tributaries.
4.2.2 Values and critical thresholds – heavy precipitation and droughts
The analysis of the values of the climatic parameters has revealed a series of thresholds and criti-
cal points regarding the temperature and precipitation, which correspond to extreme processes
and phenomena: heavy precipitation, torrential rainfalls, droughts and extreme temperatures.
The critical values and thresholds of the climatic parameters influence a series of geomorphologi-
cal processes and phenomena through their direct action on the topography.
In order to get a clear picture of the climatic thresholds it is useful to present further the
decadal march of temperatures and the mean annual precipitation at the Stolnici, Alexandria and
Fundulea weather stations.
At Alexandria w.s., for the same 42 years of recordings, one can see the mean annual tempera-
ture never dropped below 9.5 °C, whereas the highest average did not exceed 13.4 °C (Table 5). As
for the mean amount of precipitation, the lowest values are less than 300 mm, while the highest
exceed 700 mm. The warmest year was 2000, when the mean temperature reached 13.4 °C, and
the rainiest was 1966, when the mean amount of precipitation was 788 mm.
Table 5. The decadal march of temperatures and annual precipitation on Alexandria weather station.
Decade No years 1 2 3 4 5 6 7 8 9 10
1961 °C 11.6 11.1 10.6 10.5 10.5 11.7 10.8 11.4 9.8 10.7
– 1970 mm/yr 445.5 521.1 403.0 701.0 368.3 787.7 479.5 441.9 744.8 681.1
1971 °C 10.6 10.6 10.0 10.9 12.7 10.1 10.9 10.0 10.8 9.5
– 1980 mm/y r
652.1 787.4 516.0 550.1 614.3 653.7 493.3 524.7 667.8 640.0
1981 °C 10.7 10.3 10.7 10.2 10.0 10.5 10.0 11.0 11.6 11.8
– 1990 mm/yr 643.9 415.6 465.6 603.0 336.6 469.2 466.2 409.0 414.1 345.8
1991 °C 10.3 11.5 10.9 12.6 11.0 10.5 10.3 11.9 12.7 13.4
– 2000 mm/y r
601.6 295.3 452.6 462.3 506.5 390.6 638.6 490.3 401.2 330.2
2001 °C 12.4 11.1 – – – – – – – –
– 2002 mm/y r
563.9 591.7 – – – – – – – –
4.2.3 The threshold analysis
The thermal thresholds for the Romanian Plain have been established based on the mean values
of temperature and precipitation, as well as on the geographical location of the weather stations
in the area. Thus, to the north, the minimum thermal threshold was set to less than 9.5 °C and the
maximum one to more than 11 °C; to the south, the minimum threshold was established below
10 °C and the maximum one to more than 12 °C. The lowest thermal thresholds in the Romanian
Plain were recorded in the north, at Stolnici weather station, in 1980 and 1985, when the averages
dropped to 9.2 °C. To the south, at Alexandria w.s., in 1980 the mean temperature fell to 9.5 °C,
while in 1985, at Fundulea w.s., the average dropped to 9.2 °C (the same value as that recorded
at Stolnici for the same year). The maximum thermal threshold was recorded at Alexandria w.s.
in the year 2000, when the mean temperature value amounted to 13.4 °C, but on the whole,
1994 was a very warm year for almost the entire Romanian Plain, inasmuch as the temperatures
reached 11.6 °C in the north, while in the south exceeded 12 °C at all weather stations.
The minimum pluviometric thresholds vary according to the geographical location (in the
north below 450 mm/yr and in the south below 400 mm/yr), whereas the maximum threshold
was set above 700 mm/yr. The pluviometric minimum was recorded in the north that is in the
higher part of the Romanian Plain, in 1992, when the amount of precipitation reached 262 mm.
The same year, the values dropped to less than 350 mm at Furculeşti w.s., less than 300 mm
at Alexandria w.s. and at about 400 mm at Fundulea w.s. With the exception of the historical
maximum value recorded in 2005, a year that will be analyzed separately, the maximum pluvio-
metric threshold was reached in 1966, at Stolnici w.s., when the mean amount of precipita-
tion reached 930 mm. The same year, but in the south, the amount of precipitation exceeded
700 mm.
4.3 Critical hydrological thresholds and their importance for hydrogeomorphological processes
and on the society
In the specific physical conditions of the central and eastern sector of the Romanian Plain, river
water flow is reduced. Many rivers are temporary and have intermittent flow. In Table 7 are
presented data on the hydrological parameters of the mean liquid flow for the studied rivers.
Downstream the confluence with the Teleorman River, the Vedea’s river flow increases to about
11 m3/s.
The most important flow is in March and February (15–20 % of total annual average vol-
ume), and the lowest in August and September (3–6 % of total annual average volume) (Fig. 6,
left). Under these conditions, at seasonal scale, the most important transit of water is in spring
(34–36 % of annual average volume). A feature of this area is the high flow in the winter (29–34 %),
and floods due to penetration of warm and wet air with Mediterranean origin, which favors the
snow melting (Fig. 7, right).
Table 6. Data regarding the mean flow of the main rivers flowing through the central section of the Romanian
Plain.
No. Hidrometric S Hmed Q q (l/ W h

River Period
crt. station (km2) (m) (m3s-1) s.km2) (103 m3) (mm)
1 Vedea Alexandria 3312 194 1965 – 2007 7.5 2.26 236.7 71.5
2 Cotmeana Ciobani 430 151 1965 – 2007 1.37 3.18 43.2 100
3 Teleorman Teleormanu 1295 156 1965 – 2007 3.23 2.49 101.9 78.7
5 Călmăţui de Crângu 891 113 1950 – 1994 1.26 1.41 39.8 44.6
Teleorman
8 Neajlov Călugăreni 1700 154 1965 – 2007 8.86 5.21 279.6 164
S = catchment’s area, Hmed = catchment’s mean elevation, Q = mean multiannual discharge, q = mean multi-
annual specific discharge, W = mean water volume, h = mean multiannual water layer.
The sources of the data concerning S, Hmed, and Q: National Administration “Romanian Waters”.
Table 7. Maximum discharges (in m3/s1) with different exceeding probabilities for the rivers Vedea (at
Alexandria), Teleorman (at Teleormanu), and Neajlov (at Călugăreni).
River/Hydrometric
Exceeding probabilities (return period)
station
0.1 % 0.5 % 1% 2% 5% 10 %
(1,000 years) (200 years) (100 years) (50 years) (20 years) (10 years)
Vedea/Alexandria 1,729 1,261 1,059 853 611 434
Teleorman/Teleorman 531 378 311 245 168 115
Neajlov/Călugăreni 978 706 586 466 326 225
Fig. 7. Monthly (left) and seasonal (right) variability of the mean flow of the main rivers flowing through the
central section of the Romanian Plain (in % of total annual average volume).
4.3.1 Extreme hydrological conditions
The variability of the river flow occurs between two extreme hydrological events: the floods (cor-
responding to the maximum flow) and the low water periods (defined by the minimum flow).
The ranges between the water levels or the liquid discharges corresponding to the two extreme
moments on the one hand and their rate of change on the other hand, influence the dynamics and
magnitude of the fluvial processes.
The maximum flow has a major influence on the hydrogeomorphological dynamics, being
responsible for the alterations, generally rapid, of the river channels, especially during the flood
events. Over the investigated period (1965–2007), the highest maximum annual discharges
were recorded in 1972, 1975 and 2005 (Fig. 7, left), when the highest floods with extremely seri-
ous socio-economic consequences occurred. The maximum discharges of the Vedea River (at
Alexandria) varied between 4.5 m3s-1 (in 1992) and 949 m3s-1 (in 1972), with a variation coefficient
(Cv) of 0.95. In the case of the Teleorman (at the station bearing the same name), the maximum
annual discharges oscillated between 2.5 m3s-1 (in 1990) and 278 m3s-1 (in 1972) with a Cv of 1.1.
For the entire period, one can see a general decreasing trend of the maximum annual discharges.
Fig. 8. The variability of the maximum annual (left) and monthly (right) discharges (Q) of the rivers Vedea
(at Alexandria station), Teleorman (at Teleormanu station) and Neajlov (at Călugăreni station) during the
period 1965–2007.
The Neajlov River recorded at Călugăreni maximum annual discharges between 430 m3s-1 (in
1973) and 8.6 m3s-1 (in 1993), with a Cv of 1.03.
The maximum flow (seen as a phase of the hydrological regime) occurs in late winter and at
the beginning of spring, being triggered by snow melting and rains. However, the highest maxi-
mum discharges were recorded on Vedea and Teleorman in October (1972) and July (1975), in
the aftermath of heavy rainfalls (Fig. 8, right).
The minimum flow – the hydrological drought
During the minimum flow periods, fluvial dynamics processes are less active because the ero-
sive force of the river and its transportation capacity are low. The rapid alternation of the low and
high discharges undermines the channel’s stability and leads to significant physical alterations.
At the scale of the Romanian Plain, the monthly averages of the minimum specific discharges
are approximately 0.5 l/s/km2 in the northern section and less than 0.1 l/s/km2 in south and east
(Institutul de geografie, Geografia României, 2005). The minimum flow of the rivers flowing
through this region is recorded in summer and early fall, but also in winter.
The three studied rivers (Vedea, Teleormanu and Neajlov), at the considered hydrometric sta-
tions (Alexandria, Teleorman, and respectively Călugăreni) never run dry during the investigated
period (1965–2007), due to the fact that in their lower courses (where the gauging stations lie) they
are richly fed by the groundwaters. The minimum annual discharges of the Vedea River ranged from
0.124 m3s-1 (in 1968) to 2.8 m3s-1 (in 1970), the variation coefficient being relatively low (0.47). In
the case of the Teleorman, the minimum annual discharges varied between 0.1 m3s-1 (in 1968) and
1.54 m3s-1 (in 2006) (Fig. 9, left), with a variation coefficient of 0.53. It is worth noting that on the
previously mentioned interval the minimum annual discharges show a general rising trend (Fig. 8,
left). The minimum annual discharge of the Neajlov River ranged from 0.14 m3s-1 (in 1968), to
6.0 m3s-1 (in 1981), with Cv = 0.45. More often than not, the minimum flow is experienced in sum-
mer and autumn (because of the water deficit and the high evapotranspiration) and in the winter
months (because of the freezing phenomena). On the Vedea and Teleorman rivers, the lowest mini-
mum monthly discharges (less than 0.2 m3s-1) were recorded in July and August 1968 (Fig 9, right).
Fig. 9. The variability of the minimum annual (left) and monthly (right) discharges (Q) of the rivers Vedea
(at Alexandria) and Teleorman (at Teleormanu) during the period 1965–2007.
4.3.2 Critical hydrological thresholds
As mentioned previously, river flow variates between two extreme phenomena, the floods (maxi-
mum flow) and the low water periods (minimum flow), both having obvious negative socio-
economic, hydrogeomorphological and ecological consequences. Apart from the extreme values,
hydrological analyses also rely on values and thresholds with certain significance for the opera-
tional activities required by the engineering works, the management of water resources and the
management of hydrological or geomorphological risks.
In the practice of hydrology, river flow analysis takes into account the discharges or volumes
(mean, maximum and minimum) with different exceedence probabilities, estimated through sta-
tistical methods. In the case of the minimum flow, threshold values with practical importance
for the water supply are the annual minimum discharges with exceeding probabilities of 95–97 %
for the entire year (Diaconu & Şerban 1994). An extremely important threshold for keeping
the river in a good ecological state is the dilution discharge, equivalent to the mean monthly
minimum discharge with exceeding probabilities of 95 % (Pişota & Zaharia 2002). When the
maximum discharge is considered, the threshold values are represented by the exceeding prob-
abilities of 0.1 %, 0.5 %, 1 %, 2 %, 5 % and 10 %. These values are taken into account in the sizing
of hydrotechnical engineering works, the management of water resources, the management of
flood risks etc. According to the European Directive 2007/60/CE on the assessment and manage-
ment of flood risks, the maps of flood prone areas are to be prepared according to three scenarios:
floods with a low probability (or extreme event), floods with a medium probability (likely return
period ≥ 1 %) and floods with a high probability.
In Romania, the management of flood risks relies on the threshold values corresponding to
the following flooding alert levels1: attention stage, flood stage and danger stage. The attention
1 The alert levels are defined according to the Regulations for the management of emergency situations
generated by flooding, dangerous meteorological phenomena and accidents at hydraulic structures
(http://www.rowater.ro/Continut%20Site/Dispozi%C5%A3ii%20generale.aspx).
stage (CA) is equivalent to the level when flood danger becomes a reality within a relatively short
time, but which allows that protection or evacuation measures be taken. The flood stage (CI) cor-
responds to the level at which water inundates the first buildings and roads, whereas the danger
stage (CP) is the level that requires extreme measures for the evacuation of people and com-
modities, restrictions in the use of bridges and roads and special attention in the exploitation of
hydrotechnical works.
4.3.3 Critical thresholds concerning the flow of the rivers Vedea, Teleorman and Neajlov
Because hydrological dynamics and the accompanying risks are strongly influenced by floods
and maximum flow we focused our study on their specific thresholds taking into account the
discharges specific to the attention level (Qa), flood level (Qi) and danger level (Qp). Besides
their meaning as alert thresholds in the management of flood risks, these values also have hydro-
geomorphological significance (especially the flood and danger levels), during which river carry
large amounts of sediments. Under the circumstances, the erosive power grows and the channels
and floodplains suffer significant alterations. Based on the threshold values mentioned above,
we have analyzed the frequency of the most important floods that occurred on the Vedea (at
Alexandria), Teleorman (at Teleormanu) and Neajlov (at Călugăreni). The analysis is deemed to
be statistical pertinent because the datasets concerning the maximum monthly and annual dis-
charges for the period 1965–2007 were long enough.
On the Vedea River, at Alexandria station, the threshold values of the analyzed discharges2
were the following: Qa = 145 m3s-1, Qi = 220 m3s-1 and Qp = 462 m3s-1. Over the period 1965–
2007, 40 floods with discharges higher than Qa were identified, of which almost 50 % (18) topped
the Qi. Four floods exceeded the discharge corresponding to the danger stage (Qp), as follows: in
October 1972 (949 m3s-1), July 2005 (834 m3s-1), July 1975 (544 m3s-1) and July 1970 (463 m3s-1).
Most frequently, the floods occurred in March and February (Fig. 10, top left) and had a mixed
genesis (pluvial and nival).
In the case of the Teleorman, at Teleormanu gauging station, the threshold values for the
water discharges considered in the analysis were the following: Qa = 72 m3s-1, Qi = 106 m3s-
1
and Qp = 200 m3s-1. In the period 1965–2007, 12 floods with discharges higher than Qa were
identified, of which eight were higher than Qi and only one was superior to the equivalent dis-
charge of the danger stage (278 m3s-1, in October 1972). More often than not, 42 % of the floods
with discharges higher than Qa and 50 % of those with discharges higher than Qi occurred in
February (Fig. 10, top right).
On the Neajlov River, at Călugăreni, the threshold values of the analyzed discharges were:
Qa = 60,8 m3s-1, Qi = 110 m3s-1 and Qp = 212 m3s-1. Over the period 1965–2007, 48 floods with
discharges higher than Qa were counted, of which almost a third (15) topped the Qi. Six floods
exceeded the discharge corresponding to the danger stage (Qp): in March 1973 (430 m3s-1), March
2 The discharges corresponding to the attention, flood and danger stages for the considered rivers
were provided by the “Apele Române” National Administration, the Argeş – Vedea Catchment
Administration.
Fig. 10. The absolute monthly frequency of the discharges higher than the attention stage (Qa), the flood
stage (Qi) and the danger stage (Qp) for the rivers Vedea at Alexandria (top left), Teleorman at Teleormanu
station (top right) and Neajlov at Călugăreni (down left) during the period 1965–2007.
1984 (392 m3s-1), July 1975 (385 m3s-1), July 1970 (340 m3s-1), Octobre 1972 (301 m3s-1) and in April
1997 (225 m3s-1). Most frequently, the floods occurred in March and February (Fig. 9, down left).
Based on the datasets containing the maximum annual discharges recorded at the analyzed
gauging stations lying on the Vedea, Teleleorman and Neajlov rivers, we could estimate the maxi-
mum discharges with exceeding probabilities of 0.1 %, 0.5 %, 1 %, 2 %, 5 % and 10 % (Table 7).
The estimates were made using the Pearson III binomial distribution, considering the asymmetry
coefficient (Cs) of the maximum discharges series as being 3.5 times higher than the variation
coefficient (Cv) (in accordance with the recommendations contained in the Instrucţiuni pentru
calculul debitelor maxime în bazine mari, 1997).
By comparing the maximum recorded discharges with those estimated on statistical bases
one can see that the maximum discharges recorded on the Vedea River during the floods of 1972
and 2005 (949 m3s-1 and 834 m3s-1 respectively) correspond to an exceeding probability of about
2 %. The discharge equivalent to the danger stage (462 m3s-1) is approximately the same with that
of the decadal flood (434 m3s-1).
In the case of the Teleorman River, the highest discharge ever recorded in the study period is
278 m3s-1 and has an exceeding probability of about 1.5 %. The discharge equivalent to the flood
stage (106 m3s-1) has an exceeding probability of 10 %, while the discharge corresponding to the
attention stage (72m3/s) has an exceeding probability higher than 10 %.
The highest discharge on Neajlov River recorded at Călugăreni in March 1973 (430 m3s-1), is
close to the maximum discharge with exceeding probability of 2 %.
5 Vulnerability to hydrogeomorphological processes – synthetic map
5.1 Selecting the critical values
From the geomorphometric standpoint, we analyzed the following parameters: hypsometry,

vertical relief-dissection, drainage density, slope angle and the frequency of 1st order streams
(Table 2). For all these parameters, we established some value ranges depending on the variation
gaps and we assigned them a certain score.
As far as the climatic factor is concerned, for the analysis of the Vedea and Mostiştea water-
sheds, which are situated in slightly different conditions, we employed the Angot analysis (Table 4).
In the case of the Vedea watershed, the results point at a moderate to high vulnerability, whereas
in the case of the Mostiştea the vulnerability to precipitation is extremely low in these areas the
amount of precipitation is very similar to the annual one.
For the multi-criteria analysis of the hydrological factor, we took into account two criteria:
1. The number of floods exceeding the threshold value that corresponds to the attention
stage:
• Vedea at Alexandria: 40 floods higher than the discharge corresponding to the atten-
tion stage.
• Teleorman at Teleormanu station: 12 floods exceeding the discharge corresponding
to the attention stage.
• Neajlov at Călugăreni: 48 floods higher than the discharge corresponding to the
attention stage.
2. The threshold values represented by the maximum discharge with a return period of 10 %:
5 on the Vedea at Alexandria; 7 on the Teleorman and Teleorman station and 6 on the
Nejlov at Călugăreni.
5.2 Diagnosis and map representation
The geomorphometric, climatic and hydrological multi-criteria analysis allowed us to establish

some intervals of vulnerability as follows: sTable lands, low vulnerability, medium vulnerability,
high vulnerability, very high vulnerability (Tables 8, 9).
The map shows the classes of vulnerability to hydrogeomorphological hazards for the entire
watershed, on which we overlapped the high and very high vulnerability for specific types of
hazards (Fig. 11). The central unit of the Romanian Plain has therefore a high and medium vul-
nerability to hydrogeomorphological hazards with respect to the previously mentioned criteria,
while the Mostiştea watershed, lying in the eastern section, has a low vulnerability.
Table 8. The classes of vulnerability to hydrogeomorphological hazards depending on geomorphometric criteria.
Class (score) Vulnerability diagnosis Watershed
6 points stability 1 point/criterion
6 – 10 small Mostiştea
10.1 – 15 medium Neajlov
15.1 – 20 high Călmăţui
more than 20 very high Vedea
Table 9. Synthetic data for diagnosing the vulnerability to hydrogeomorphological hazards.
Neajlov Mostiştea Călmăţui

Factors Criteria Vedea watershed
watershed watershed watershed
Geomorpho- Geomorpho- 28 points 12 points 9 points 18 points
logical metric point
Hazards Channel erosion, Lateral erosion, Erosion and Headward
accumulation, meandering- accumulation erosion,
Torrentiality Calnistea Lacustrine torrentiality
abrasion
Climatic Precipitation Medium Medium Low and Medium
Angot indexes vulnerability vulnerability very low vulnerability
vulnerability
Hydrological Number of 40 + 12 48 Lacustrine

floods abrasion,
Probability 5+7 6 cavingins
10 %
Anthropo- River/lakes Dams/lakes Dams/lakes River improved Dams/lakes
genic improvements through
damming
Total points/ 73 + 19 66 8 18
vulnerability High and very Medium general Low general Medium – low
diagnosis high general vulnerability vulnerability general
vulnerability vulnerability
High vulnerabil- High
ity to floods vulnerability High
to abrasion vulnerability to
torrentiality
Fig. 11. The hydrogeomorphological vulnerability in the Romanian Plain – Synthetic map.
6 Conclusions
Due to the favorable natural setting, the Romanian Plain is a densely populated area with well-
developed economic activities, especially agricultural. Most of the settlements and transport
infrastructure (roads and railroads) are found along the watercourses. Under the circumstances,
the Romanian Plain is vulnerable to hydrogeomorphological processes. The study investigates the
specific features of three factors controlling the topography dynamics, namely geomorphology,
climate and hydrology. The study focuses on four watersheds lying in the central and eastern sec-
tions of the Romanian Plain: Vedea (including its main tributary, the Teleorman River), Călmăţui,
Neajlov and Mostiştea. For each of them, we have analyzed the geomorphological, climatic and
hydrological background and we highlighted for each specific parameter the critical values that
can act as thresholds for the occurrence of certain hydrogeomorphological processes that nega-
tively impact the environment and the human communities. Based on the multi-criteria analysis,
we have taken into account some ranges of values for the geomorphological, climatic and hydro-
logical parameters and we estimated the vulnerability to hydrogeomorphological processes. The
watersheds lying in the Central Section, between the rivers Olt and Argeş (Vedea and Neajlov)
have a high and medium vulnerability to hydrogeomorphological hazards, while the Călmăţui
and Mostiştea watersheds have a low vulnerability.
This study, which makes a diagnosis of the vulnerability to hydrogeomorphological hazards,
may serve as an information support for the regional and local authorities in their effort to miti-
gate and manage the risks induced by such hazards.
Acknowledgements
We wish to thank our colleagues who revised the text for their valuable opinions and use-
ful suggestions offered with the purpose to improve the content and help the publishing of this
material.
This work was supported by CNCSIS – UEFISCDI, project number PNII – IDEI code
1954/2008. ‟The hydrogeomorphological system in the concepts of geomorphometry and of
modern morphological theories. Applications to hazard and risk diagnosis in areas of Romanian
Plain” (Director, Prof. univ. dr. Grecu Florina).
References
ALBU (DINU) (2011): Călmăţui (Teleorman) Hydrographic Basin – Morphometric Analisis Elements. –
Rev. geomorf. 13: 123–130.
Aquaproiect (1992): Atlasul Cadastrului Apelor din Romania, Partea I-a şi a II-a, Bucureşti.
A.N.C.P.I. (2006): Ortofotoplanuri, scara 1:5 000.
A.N.A.R. – D.A.A.V. (2008): Proiectul planului de management al spaţiului hidrografic Argeş. – Vedea:
223–234.
Administraţia Naţională De Meteorologie (2008): Clima României. – Ed. Acad Române, Bucureşti,
366.
Badea, L. (2007): Relieful României şi neotectonica. – Stud. Geomorf., Univ. Craiova, 160.
Bravard, J. P., Petit, F. (2010): Les cours d’eau. Dynamique du système fluvial – Armand Colin, 222.
Brunét, R. (1970): Les phénoménes de discontinuité en géographie. – Mem. doc. 7: 117, Paris.
Boutot, A. (1996): Inventarea formelor. – Edit. Nemira, Bucureti, 155.
Cîrciumaru, E. (2011): Geneza, evoluţia şi dinamica actuală a bazinului morfohidrografic Vedea. – Teză de
doctorat (Coord. Prof. dr. Florina Grecu), 360 pp., Univ. Bucureşti.
Cocoş, A., Cocoş, O. & Sârbu, I. (2011): Coping with water scarcity: the case of the Câlniştea catchement
(România). – Environ. Earth Sci., DOI 10.1007/s12665-011-1513-y.
Coteţ, P. (1976): Câmpia Română. – Stud. geomorf. integr., edit. Ceres, Bucureşti, 256.
Diaconu, C. & Şerban, P. (1994): Sinteze şi regionalizări hidrologice. – Edit. Tehnică, Bucureşti, 388.
Direcţiă Topografică Militară (1970, 1972, 1977, 1993): Harţi topografice, scara 1:100 000, 1:50 000,
1:25 000, Bucureşti.
Dragota, C. (2006), Precipitaţiile excedentare în Romania. – Edit. Academiei, Bucureşti, 176.
Ghiţă, C. & Dragotă, C. S. (2010): Susceptibilitatea terenului la procese geomorfologice actuale în funcţie
de particularităţile regimului pluviometric din bazinul hidrografic Mostiştea. – Comun. Geogr. XIII:
41–47, Bucureşti.
Goudie, S.A. (2000): The Human Impact on the Natural Environment. – 5th ed., 511, Cambridge.
Grecu, F. (1992): Bazinul Hârtibaciului. Elemente de morfohidrografie – Edit. Academiei, Bucureşti, 168.
Grecu, F. (1999), Teoriile morfologice – semnificaţia reliefogenă. – Comun. Geogr. III: 25–34, edit. Univ.
Bucureşti.
Grecu, F. (2004): Quantification of some elements of drainage basins in Romania. – Geogr. Fisica Dinam.
Quatern. 27: 29–36.
Grecu, F. (2008): Index of Morphohydrographic basin completion by perimetres and area. Case study in
Romania. – Geogr. Fisica Dinam. Quatern. 31: 37–45.
Grecu, F. (2010): Geografia câmpiilor României. – Edit. Univ., Bucureşti, 256.
Grecu, F. & Palmentola, G. (2003): Geomorfologie dinamică. – Edit. Tehnică, Bucureşti.
Grecu, F., Ghiţă, C., Cîrciumaru, E. & Văcaru, L. (2009a): Diagnostic des aléas detérmines climatiques
dans les systèmes hydrogéomorphologiques de la Plaine Roumaine. – Extrêmes climatiques: genèse,
modélisation et impacts, XXII Colloque de l’Association Internationale de Climatologie, 1–5 sept.,
Cluj-Napoca, Geographia Tehnica, 229–234.
Grecu, F., Sacrieru, R., Ghiţă, C. & Văcaru, L. (2009b): Geomorphological landmarks of the eastern
Romanian Plain Holocene evolution. – Z. Geomorph. N.F. Suppl. 53 (1): 99–110.
Grecu, F., Zaharia, L., Ghiţă, C. & Văcaru, L. (2010a): The dynamic factors of hydrogeomorphic vulner-
ability in the central sector of the Romanian plain. – Metalurgia Int. XV, Spec. iss. 9: 139–148.
Grecu, F., Ghiţă, C. & Săcrieru, R. (2010b): Relation between tectonics and meandering of river channels
in the Romanian Plain. Preliminary observation. – Rev. Geomorf. 12: 97–104, Univ. Bucuresti.
Horton, R. E. (1945): Erosional development of streams and their drainage basins: Hydrophysical approach
to quantitative morphology. – Geol. Soc. Amer. Bull. 56 (3): 275–370.
Ielenicz, M. (1998): Le concept de discontinuité en géographie. – Rev. roum. géogr. 42: 169–176.
I.N.H.G.A. (2006): Caracterizarea hidrologică a anului 2005: 22.
I.N.M.H. (1997): Instrucţiuni pentru calculul debitelor maxime în bazine mari, Bucureşti, 53.
Institutul Geologic (1967): Harta geologică a R. S. România, Scara 1:200,000, foile: Giurgiu, Turnu
Măgurele, Videle, Neajlov şi Bucureşti.
Institutul De Geografie Al Academiei Române (2005): Geografia României (Câmpia Română,
Dunărea, Podişul Dobrogei, Litoralul românesc al Mării Negre şi Platforma Continentală). – Academiei
Bucureşti V: 967.
Liteanu, E. (1969): Harta hidrogeologică, scara 1:1,000,000, Institutul geologic, Comitetul de stat al
Geologiei, Bucureşti.
Malavoi, J. R. & Bravard, J. P. (2010): Elements d’hydromorphologie fluviale. – Onema, 224.
Mihailescu, V. (1924): Vlăsia şi Mostiştea – extras din BSRG, anul XLIII: 192, Bucuresti.
Mustăţea A. (2005): Viituri excepţionale pe teritoriul României. – I.N.H.G.A., 409, Bucureşti.
Petrea, D. (1998): Pragurile de substanţă, energie şi informaţie în sistemele geomorfologice. – Univ.
Oradea, 224.
Petrea, D. (2005): Obiect, metoda şi cunoaştere geografică. – Univ. Oradea, 230.
Pişota, I. & Zaharia, L. (2002): Hidrologie. – Univ. Bucureşti, 445.
Posea, Gr. (1997): Relieful şi evoluţia paleogeografică Câmpiei Române. – Ghidul excursiilor celei de-a XV-a
Conferinţe Naţionale pentru Ştiinţa solului, 26–30 august, 1997, Bucureşti, 19–31.
Posea, Gr. (2002): Geomorfologia României. – Fundaţiei România de Mâine, 444, Bucureşti.
Rădulescu, I. (1956): Observaţii geomorfologice în Câmpia piemontană Piteşti. – Anal. Univ. Bucureşti,
Seria Ştiintele Naturii, Tehnică, 177–206.
Strahler, A. N. (1952): Dynamic Basin of Geomorphology. – Geol. Soc. Amer. Bull. 63.
Zaharia, L., Grecu, F., Ioana-Toroimac, G. & Chirilă, G. (2011): Sediment Transport And River Chan-
nel Dynamics In Romania. Variability and Control Factors in Sediment transport in aquatic environ-
ments. – In: Manning, A. J. (ed.): Techweb.org, 293–316.
Zugrăvescu, D., Polonic, G., Horomnea, M. & Dragomir, V. (1998): Recent vertical crustal movements
on the romanian territory, major tectonic compartinents and their relative dynamics. – Rev. roum.
Géophys. 42: 3–14, Bucarest.
Văcaru, L. (2010), Bazinul hidrografic Neajlov. – Studiu de geomorfologie dinamică, teză doct. (coord.
Prof. Dr. Florina Grecu), Univ. Bucuresti, 336.
Vâlsan, G. (1916): Câmpia Română. – Bul. Soc. Regale Romane Geogr. XXXVI/1915: 313–568,
Bucureşti.
*** European Directive 2007/60/CE on the assessment and management of flood risks, 23 october 2007,
Official Journal of European Union.
*** Szatmary austrian Map, 1856.
Address of the authors:

Prof. dr. Florina Grecu and dr. Cristina Ghita, University of Bucarest, Faculty of Geography, Department
of Geomorphology, Pedology and Geomatics, 1 Bd Nicolae Balcescu, Sect. 1, 010074 Bucarest, Romania,
e-mail: florinagrecu@yahoo.com / grecu@geo.unibuc.ro.
Prof. dr. Liliana Zaharia, University of Bucarest, Faculty of Geography, Department of Hydrology and
Meteorology.
29
Morphological
Stuttgart, September 2013 analysis and related volcanic features
Morphological analysis and related volcanic features of

the Kolumbo submarine volcanic chain
(NE of Santorini Island, Aegean Volcanic Arc)
Paraskevi Nomikou*, Steven Carey, Katherine L. Croff Bell, Dimitros Papanikolaou, Konstan-
tina Bejelou, Matina Alexandri, Kat Cantner, Julie Fero Martin
with 19 figures
Abstract. Kolumbo submarine volcano is located in a small, elongated, rifted basin northeast of Santorini
Island, Greece, and has been the site of recent submarine volcanism in the central Hellenic Volcanic Arc. It
is the largest of a chain of nineteen volcanic cones occurring within this small rift zone and its most known
eruption in 1650 A.D. had a serious impact on Santorini and the surrounding islands. According to previous
studies, a range of ages is suggested for the activity along this volcanic line since many of the smaller volcanic
cones seem to have been built above the present seafloor, while others are partly buried by Quaternary sedi-
ments. The ROVs Hercules and Argus of O.E.T. (Ocean Exploration Trust) were used to explore the slopes,
summits and craters of 17 of the 19 submarine volcanic centres identified on multibeam map of the area
with E/V Nautilus (NA007) in August 2010. In this paper we present some of the most interesting submarine
morphological features along the Kolumbo Volcanic Chain.
Keywords: Kolumbo Volcano, volcanic morphology, ROV dives, swath bathymetry
1 Introduction
The volcanoes of the Aegean volcanic arc occur both onshore, at the areas of Methana, Poros, Milos,
Santorini, Kos and Nisyros and offshore (Nomikou et al. 2013). Submarine volcanoes are: Pausa-
nias in the western Saronikos Gulf (Pavlakis et. al. 1990), Kolumbo volcano and several smaller
cones northeast of Santorini (Nomikou et al. 2012a) and at the area around Nisyros (Papanikolaou
& Nomikou 2001). Santorini volcanic group comprises three distinct volcanic structures occurring
along a NE-SW direction. Christianna at the southwestern part of the group, Santorini occupies the
middle part (including the islands of Thera, Therasia, Nea Kameni, Palea Kameni and Aspronisi,
arranged in a circular shape) and Koloumbo volcanic chain extends towards north-eastern (Fig. 1).
Multibeam mapping aboard R/V Aegaeo of HCMR, conducted in 2001 and 2006 at the area
of Anydhros basin, revealed the existence of approximately 20 submarine volcanic cones, with
Kolumbo being the most dominant one in terms of size. In fact, Kolumbo is the only volcanic
centre that has erupted explosively in 1650 A.D., and according to historical references (Fouque
1879) caused significant damage and fatalities on the island of Santorini. In 2006, an important
discovery of R/V Endeavor cruise was an active hydrothermal vent field in the crater of Kolumbo
volcano at a depth of 500 m (Sigurdsson et al. 2006). In 2010, the cruise NA-007 aboard the E/V
Nautilus carried out new oceanographic survey at the area of Anydhros basin using Remotely
Operated Vehicles (ROVs).
DOI: 10.1127/0372-8854/2013/S-00142 0372-8854/13/S-00142 $ 4.75
30 Paraskevi Nomikou et al.
One of the main objectives of this exploration was to determine the morphological structure,
the geological deposits and the hydrothermal venting of Kolumbo volcanic chain and try to relate
them with the tectonic processes that control the orientation of volcanism in this area. For this
purpose, using the ROV Hercules, a large number of samples (lava rocks, pumice deposits, gas
and water) was collected and photographic material, as well.
2 Study area: Geo-tectonic setting
Santorini Island and Kolumbo submarine volcanic chain are located at the centre of the mod-
ern Aegean volcanic arc, developed within the actual Hellenic Orogenic Arc and trench sys-
Fig. 1. Combined bathymetric and topographic Map of Santorini volcanic field based on high resolution
onshore and offshore data.
Morphological analysis and related volcanic features 31
Fig. 2. Simplified map of the present day geodynamic structure of the Hellenic Arc, showing the mod-
ern Aegean volcanic arc developed behind the Hellenic trench, the Peloponnese-Crete island arc and the
Cretan Back-arc basin (Nomikou & Papanikolaou 2010). Note that the African plate to the south subducts
beneath the Eurasian plate to the north along the red lines just to the south of Crete.
tem (Fig. 2). It is formed along the convergent plate boundary of the northwards subducting
African plate underneath the active margin of the European plate (McKenzie 1970, 1972,
LePichon & Angelier 1979). The total migration of the Aegean volcanic arc is approximately
400 km within 40 myears, which corresponds to an average rate of 10 km/myears (Papanikolaou
1993, Royden & Papanikolaou 2011).
The islands of the Santorini group are dominantly composed of layers of volcanic rocks that
have been produced by different eruptions during the last half a million years (Druitt et al.
1989). These several layers of volcanic rocks overlay an alpine basement, mainly late Mesozoic-
early Cenozoic schists and marbles at the south-eastern part of the island (Papastamatiou 1958,
Tataris 1965). Granite also occurs on the island, dating back 9.5 million years (Skarpelis et al.
1992). There have been at least 12 explosive events, many minor eruptions, and at least four cal-
dera collapses (Friedrich 2000) including the great Bronze Age event, 3600 years ago, that prob-
ably affected human populations in the eastern Mediterranean area and led to the demise of the
Minoan civilization. The numerous volcanic eruptions have produced lava flows, pumice and tuff
blankets, volcanic ash and scoria. The most recent activity of Santorini has been located on the
small islands of Palea Kameni and New Kameni in the center of the caldera. Eruptions that have
formed these islands have been relatively weak with a combination of lava flows and some small
explosive events. The most recent one took place in 1950.
Seismic profiles collected in 2006 aboard R/V Aegeao showed that a prominent NE-SW
trending, rift zone, Anydhros basin, extends to the north-east of Santorini, and hosts Kolumbo
and other smaller cones. The transtensional Anydhros basin is bounded by two apparently nor-
mal fault zones: the Ios fault zone towards NW and the Anydhros fault zone. Numerous smaller,
vertical and/or sub-vertical faults, running parallel to the basin’s long axis, crosscut its sedimen-
tary infill and control the spatial distribution of the volcanic cones and the magma rise. This fault
pattern has been interpreted as “flower type” structure developed above a single strike-slip fault,
running NE-SW below the Kolumbo volcanic line (Sakellariou et al. 2010). Localized seismic-
ity below Kolumbo indicates the existence of a magmatic chamber at a depth of about 5 km below
the surface and an overall active volcanotectonic status (Bohnhoff et al. 2006, Dimitriadis
et al. 2010).
Kolumbo is the only centre where an explosive eruption of gas-rich magma has been recorded
in 1650 (Vougioukalakis et al. 1994). The eruption column of ash and pumice raised high
above the volcano and created pyroclastic flows and pyroclastic surges that passed over the sea
surface, spreading several kilometres distance over the ocean in all directions. At least fifty people
died from asphyxiation on Santorini at areas nearer to the volcano and over a thousand livestock
(Fouque 1879). The eruption was strong and was accompanied by earthquakes and the tidal
waves reached up in a great distance from Santorini. Tsunamis generated during the event caused
significant damage along the east coast of Santorini and especially the villages of Perissa and
Kamari brought in ruins (Dominey et al. 2000, Nomikou et al. 2012b).
3 Data collection and instrumentation
All the data which led to the construction of the bathymetric map of the area and the visual data
as well were collected during several oceanographic missions aboard the R/V Aegeo and E/V
Nautilus in 2001, 2006 and 2010.
The mutltibeam data were collected during a survey with the R/V Aegaeo, operated by the
Hellenic Centre for Marine Research (HCMR), using a SEABEAM 2120 swath system working at
20 kHz and until 6000 m water depth. It has an angular coverage sector of 1500 with 149 beams,
covering a swath width from 7.5 to 11.5 times the water depth for depths from 20 m to 5 km. The
maximum swath coverage can reach 9 km at maximum depth and gives satisfactory data qual-
ity at speeds up to 11 knots. GPS navigation (Trimble 4000) provided the average position of the
ship to within +/– 10 m. The multibeam data have been extensively processed by means of data
editing, cleaning of erroneous beams, filtering of noise, processing of navigation data and inter-
polation of missing beams.
The ROV exploration was conducted aboard the E/V Nautilus. The ROVs Hercules and Argus
make up a dual-body deep water system that is capable to reach until the depth of 4000 m. They
receive electrical power from the surface through a fiber-optic cable. A 20-hp electric/hydraulic
pump powers the mechanical functions on Hercules. Hercules is also equipped with to manipula-
tor arms for sampling and high definition cameras for videos and images.
4 Results
4.1 Morphological characteristics of Kolumbo Volcanic Chain
Kolumbo Volcanic Chain (KVC) is located in the basinal area of Anhydros rift northeast of
Santorini, at an average depth of 450 m. It is about 14 km long and its southern edge is Kolumbo
submarine volcano, the largest volcanic cone within this basin. The resulting shaded relief map of
the KVC compiled by multibeam data, shows the morphological structure of the smaller volcanic
domes compared to Kolumbo caldera (Fig.3).
The nineteen volcanic cones are distributed linearly along two different orientations (N 29 °E
and N 42 °E), the common point of which is Kolumbo volcano. Most of the cones have a sym-
metrical or ellipsoidal dome-like shape, whereas the summit of some is a well defined crater.
They are usually circular or concave with very steep slopes, as indicated by dense isobaths, and
are lying in a basinal area. Their highest points range between 130 and 430 meters and their basal
contours between 300 and 450 meters (Nomikou et al. 2012a). As shown on the histogram of
Fig. 4, the height of the cones gradually reduces toward the northeast, far away from Kolumbo.
The larger volcanic cones, VC 11 and VC 12 belong to the northern trend and comprise two sum-
mits separated by a narrow valley as shown in the topographic profile A–B (Fig. 5). Their basal
contour is almost at 410 m depth and their shallowest point is at 210 m below sea level. In the
southern trend, VC 9 and VC 10 are smaller but more uniform in size with their summit up to
360 m depth, except from VC 14, which has a more symmetric conical shape up to 325 m depth
(Profile B–C, Fig. 5).
A further process of the bathymetric data resulted in a slope distribution map (Fig. 6). Accord-
ing to this there are five distinct slope values within this area: (1) areas of mean morphological
slope 0–5 %, (2) areas of 5–15 %, (3) areas of 15–25 %, (4) areas of 15–25 % and (5) areas of 35–50 %.
Areas with low slope magnitudes up to 15 % reflect flat-lying areas of Anhydros basin implying
that the seafloor is smooth with no significant morphological differentiations. On the other hand,
areas of high slope values usually correspond to tectonic or volcanic features. The highest slope
values that are distributed in a linear way correspond to the marginal fault zones of Anhydros
Basin, while the ones distributed in a circular way reflect the abrupt slopes of the volcanic cones.
The Kolumbo volcano is characterized by steep inner crater walls with high slopes > 35 % and
intermediate external crater flanks.
4.1.1 Kolumbo main cone
Kolumbo is located about 7 km NE of the northern coast of Santorini. It is the largest cone in vol-
ume and the point where the two different trends of KVC converge. Its basal diameter is approxi-
mately 3 km and its crater diameter about 1.7 km. The average depth of the crater rim is 150 m,
whilst the shallowest lies at the depth of 18 m towards the south-western side of the volcano. The
smooth crater floor is at a depth of 500 meters (Nomikou 2003). The highest slope values at the
Fig. 3. Shaded relief bathymetric map of the Kolumbo volcanic chain using 10 m isobaths. VC: Volcanic
Cone.
Fig. 4. Histogram showing the distribution of the height of volcanic cones along the volcanic chain from
Kolumbo to the eastern most volcanic cone VC 19 (Nomikou et al. 2012a).
Fig. 5. Topographic profiles A–B and C–D across volcanic domes aligned in the two different trends, oriented
in NE-SW direction.
inner crater walls of Kolumbo most likely represent the remnants of the collapse of a relevant vol-
canic cone. Slope values at the external flanks diminish gradually. On the north-western flanks,
linear to curvilinear areas of abrupt change in slope probably represent an earlier stage of volcan-
ism which created a pre-existing volcanic cone. The volcanic structure of Kolumbo caldera and
the minor small, symmetric volcanic cones at the north-eastern slope of Kolumbo, are clearly
shown in the 3-D stereogram (Fig. 7).
Fig. 6. Slope distribution map of Kolumbo volcanic chain, NE of Santorini volcano (colour scale to the upper
left of map).
Fig. 7. 3-D view of the submarine cone of Kolumbo at depths between 18 m and 505 m.
Volcanic features revealed by ROV exploration

ROV exploration aboard the E/V Nautilus focused on the most important volcanic cones,
and in this paper their morphology and structure are presented as they have been revealed from
captures and video footage.
4.1.2 Kolumbo main cone
ROV exploration at the external flanks of Kolumbo confirmed the existence of the linear to cur-
vilinear features, previously described in paragraph 4.1.1, which were scarps with inward dipping
faces. The seafloor at this area is covered by lapilli sized pumice with occasionally scattered larger
boulders of pumice (Fig. 8).
The inner crater walls show diversity in morphologic features due to the variety of volcanic
lithologies (layers of thick and fine grained pumice, dikes, intrusions, breccias and mass wasting
deposits). At certain points into the pumice sequences, the steep crater walls appeared an undu-
lating character (Fig. 9) carved by erosional processes. In addition, at the lower part of the crater
walls some ridges were observed sloping down to the crater floor. These are the result of massive
lava flows, dikes and intrusions prior to the 1650 A.D. eruption (Fig. 10).
Kolumbo crater floor is generally flat and covered by a layer of bacterial mat, a few centime-
ters in thickness (Fig. 11). At its north-eastern part, a hydrothermal vent field was discovered in
2006 and was explored in detail in 2010 with the ROV Hercules. The ongoing discharge of high
temperature (up to 220 °C) gas and fluids has constructed a number of hydrothermal chimneys
consisting of massive sulfides and sulfates (Fig. 12) (Carey et al. 2011).
Fig. 8. Western external slopes of Kolumbo Volcano comprised with large rhyolitic pumice outcrops.
Fig. 9. Submarine erosional processes have given this part of the crater wall an undulating shape.
Fig. 10. Lava dyke exposed at the lower part of the eastern wall.
Fig. 11. This reddish-orange bacterial mat, a few centimeters in thickness, covers the flat crater floor of
Kolumbo.
Fig. 12. Hydrothermal chimneys from Kolumbo vent field coated with delicate filamentous bacteria.
4.1.3 Secondary volcanic cones (NE of Kolumbo)
Apart from Kolumbo, the rest of the KVC was also explored thoroughly in 2010, during the Nau-
tilus oceanographic cruise. Despite their difference in size and volume in comparison with the
main cone of Kolumbo, a number of noteworthy characteristics were observed.
The slopes of the volcanic cones, in general, become steeper towards their upper parts and
reveal indurated terraced outcrops of biologically encrusted tuffaceous deposits (Figs. 13, 14, 15).
Their lower parts and their flat summits are covered with fine grained sediment with occasional
blocks of pumice. A few linear scarps were also observed at the area of the KVC, which were
interpreted as the surface expression of tectonic features (Fig. 16). Some unique features were
observed in some of the cones that generated further questions for the creation and evolution of
KVC. The slopes of VC 12 are characterized by a step-like morphology (Fig 17).
In VC 7, between the two summits, a fragmented slab like outcrop was observed. Some dark
colored precipitates at the base of a scarp suggest possible hydrothermal activity (Fig 18). VC 14,
which lies at the far end of the easterly trend of KVC, appears very distinctive characteristics.
Small, round shaped depressions at the summit area of cone VC 14 form a special landscape.
They are less than a meter in diameter and a few tens of centimeters in depth. Some linear depres-
sions were also found containing black precipitates indicative of possible hydrothermal activity
(Fig 19). The black sediment was sampled found to consist of frambroidally precipitated manga-
nese oxide indicating that low temperature hydrothermal venting is very active on these cones
(Nomikou et al. 2012a).
Fig. 13. Large slab of scoria traced on the western inner slope of cone VC 3.
Fig. 14. Large lobate outcrop at the eastern slope of VC 10.
Fig. 15. Steep southern slopes at VC 18.

Fig. 16. Linear steps in E-W direction at VC 9.
Fig 17. Very steep slope with uneven topography at VC 12.

Fig 18. The face of an exposed outcrop covered with black precipitation on VC 7.
Fig. 19. Curvilinear depression filled with black manganese precipitates in the summit area of VC 14.
5 Conclusions
Kolumbo submarine volcano exhibits a rather unique site of volcanism, since its type of hydro-
thermal activity is usually met in totally different submarine environments like middle ocean
ridges or Intra-Oceanic island arcs (Hannington et al. 2005). It must have erupted prior to the
1650 A.D. event, judging by the different morphology and lithologies at the different parts of the
crater walls. In addition, the almost vertical character of the crater walls indicates the collapse of
the central part of a pre-existing volcanic dome.
According to integrated evaluation of the swath bathymetry of the Anhydros basin (Sigurdsson
et al. 2006, Nomikou et al. 2012a) and seismic profiling data, the volcanism at the area has evolved
under tectonic control. This NE-SW transtensional zone has formed at least 20 submarine vol-
canic centers that are aligned along two principal trends (Fig. 3). The south-westerly portions of
both trends intersect approximately within the crater of Kolumbo, the largest submarine center
within the NE zone.
Moreover, remotely operated vehicle (ROV) exploration of the cones in the northeast volca-
nic field revealed that the size of the cones located at the more easterly trend is smaller, probably
indicative of their younger age. Judging by the absence of craters NE of Kolumbo, we suggest that
the volcanic activity mostly included dome building rather than explosive events. The outcrops
of cemented pumice and the presence of extensive sediment covers indicate the lack of any recent
volcanic event.
Acknowledgments
This study was supported by Institute for Exploration (IFE / USA) and the collaborative project
“New Frontiers in the Ocean Exploration 2010”. The officers and the crew of the E/V Nautilus
and R/V Aegaeo are gratefully acknowledged for their important and effective contribution to
the field work and sampling. We extend our sincere thanks to the organizing committee of the
15th Joint Geomorphological Meeting for accepting the paper with title “Morphological Analysis
and related volcanic features of the Kolumbo Submarine Volcanic Chain (NE of Santorini Island,
Aegean Volcanic Arc)” and reviewing it for publication.
References
Bohnhoff, M., Rische, M., Meier, T., Becker, D., Stavrakakis, G. & Harjes, H. P. (2006): Microseis-
mic activity in the Hellenic Volcanic Arc, Greece, with emphasis on the seismotectonic setting in the
Santorini-Amorgos zone. – Tectonophysics 423: 17–33.
Carey, S., Bell, K. L. C., Nomikou, P., Vougioukalakis, G., Roman, C. N., Cantner, K., Bejelou, K.,
Bourbouli, M. & Martin, J. F. (2011): Exploration of the Kolumbo Volcanic Rift Zone. – In: Bell,
K. L. C. & Fuller, S. A. (eds.): New Frontiers in Ocean Exploration: The E/V Nautilus 2010 Field
Season. – Oceanography suppl. 24 (1): 24–25.
Dimitriadis, I., Papazachos, C., Panagiotopoulos, D., Hatzidimitriou, P., Bohnhoff, M., Rische, M.
& Meier, T. (2010): P and S velocity structures of the Santorini-Kolumbo volcanic system (Aegean Sea,
Greece) obtained by non-linear inversion of travel times and its tectonic implications. – J. Volcanol. and
Geotherm. Res. 195: 13–30.
Dominey-Howes, D. T. M., Papadopoulos, G. A. & Dawson, A. G. (2000): Geological and historical

investigation of the 1650 Mt. Columbo (Thera Island) eruption and tsunami, Aegean Sea, Greece. – Nat.
Haz. 21 (1): 83–96.
Druitt, T. H., Mellors, R. A, Pyle, D. M. & Sparks, R. S. J. (1989): Explosive volcanism on Santorini,
Greece. – Geol. Mag. 126 (2): 95–126.
Friedrich, W. (2000): Fire in the Sea. The Santorini Volcano: Natural History and the Legend of Atlantis. –
Cambridge Univ. Press, Cambridge.
Fouque, F. (1879): Santorin et ses Ιruptions. – Masson et Cie, Paris.
Hannington, M. D., De Ronde, C. E. J. & Petersen, S. (2005): Supplement to Sea-floor tectonics and
submarine hydrothermal systems. – Econom. Geol., 100th Anniversary Volume: 111–141.
Le Pichon, X. & Angelier, J. (1979): The Hellenic Arc and trench system: a key to the neotectonic evolu-
tion of the Eastern Mediterranean area. – Tectonophysics 60: 1–42.
McKenzie, D. P. (1970): Plate Tectonics of the Mediterranean region. – Nature 226: 239–243.
McKenzie, D. P. (1972): Active tectonics of the Mediterranean region. – Geophys. J. R. Astron. Soc. 30:
109–185.
Nomikou, P. (2003): Santorini and Nisyros: Similarities and differences between the two calderas of the
modern Aegean Volcanic Arc. CIESM Workshop, Santorini, 22–25 October 2003, Greece. – Mono-
graphs 24: 103–108.
Nomikou, P. & Papanikolaou, D. (2010): The morphotectonic structure of Kos-Nisyros-Tilos volcanic area
based on onshore and offshore data. – Proc. XIX Congr. Carpathian-Balkan Geol. Assoc., Thessaloniki,
Greece, 23–26 September 2010, spec. vol. 99: 557–564.
Nomikou, P., Carey, S., Papanikolaou, D., Croff Bell, K., Sakellariou, D., Alexandri, M. & Bejelou, K.
(2012a): Submarine Volcanoes of the Kolumbo volcanic zone NE of Santorini Caldera, Greece. – Glob.
Planet. Change 90–91 (2012): 135–151. doi:10.1016/j.gloplacha.2012.01.001.
Nomikou, P., Papanikolaou, D., Alexandri, M., Sakellariou, D. & Rousakis, G. (2013): Submarine
Volcanoes along the Aegean Volcanic Arc. – Tectonophysics 597–598: 123–146.
Nomikou, P., Carey, S., Croff Bell, K., Papanikolaou, D., Bejelou, K., Cantner, K., Sakellariou, D.
& Perros, I. (2012b): Tsunami Hazard Risk of a future volcanic eruption of Kolumbo submarine vol-
cano, NE of Santorini Caldera, Greece. – Natural Hazards, doi: 10.1007/s11069-012-0405-0.
Papanikolaou, D. (1993): Geotectonic evolution of the Aegean. – Bull. Geol. Soc. Greece XXVII: 33–48.
Papanikolaou, D. & Nomikou, P. (2001): Tectonic structure and volcanic centres at the eastern edge of the
Aegean volcanic arc around Nisyros Island. – Bull. Geol. Soc. Greece 34 (1): 289–296.
Papanikolaou, D. J. & Royden, L. H. (2007): Disruption of the Hellenic arc: Late Miocene extensional
detachment faults and steep Pliocene-Quaternary normal faults or what happened at Corinth? –
Tectonics 26: TC5003, doi:10.1029/2006TC002007.
Papastamatiou, I. (1958): The age of Crystalline Limestones of Thira. – Bull. Geol. Soc. Greece 3 (1):
104–113.
Pavlakis, P., Lykoussis, V., Papanikolaou, D. & Chronis, G. (1990): Discovery of a new submarine vol-
cano in the western Saronic Gulf: The Paphsanias Volcano. – Bull. Geol. Soc. Greece 24: 59–70.
Sakellariou, D., Sigurdsson, H., Alexandri, M., Carey, S., Rousakis, G., Nomikou, P., Georgiou, P.
& Ballas, D. (2010): Active tectonics in the Hellenic Volcanic Arc: the Kolumbo submarine volcanic
zone. – Bull. Geol. Soc. Greece XLIII/2: 1056–1063.
Skarpelis, N., Kyriakopoulos, K. & Villa, I. (1992): Occurance and 40Ar/39Ar dating of a granite in
Thera (Santorini, Greece). – Geol. Rundsch. 81/3: 729–735.
Sigurdsson, H., Carey, S., Alexandri, M., Vougioukalakis, G., Croff Bell, K. L., Roman, C.,
Sakellariou, D., Anagnostou, C., Rousakis, G., Ioakim, C., Gogou, A., Ballas, D., Misaridis, T. &
Nomikou, P. (2006): Marine investigations of Greece’s Santorini volcanic field. – EOS 87 (34): 337, 342.
Tataris, A. (1965): The Metamorphic basement (Hocene) of Thira. – Bull. Geol. Soc. Greece 6: 232–238.
Vougioukalakis, G., Mitropoulos, D., Perissoratis, K., Andrinopoulos, A. & Fytikas, M. (1994):
The submarine volcanic centre of Columbos, Santorini (Greece). – Bull. Geol. Soc. Greece 30 (3):
351–360 (in Greek).
Addresses of the authors:

Paraskevi Nomikou (corresponding author), Dimitros Papanikolaou and Konstantina Bejelou, National &
Kapodistrian University of Athens, Faculty of Geology and Geoenvironment, Panepistimioupoli Zografou,
15784 Athens, Greece; e-mail: evinom@geol.uoa.gr
Matina Alexandri, Institute of Oceanography, Hellenic Centre for Marine Research, POBox 712, 19013
Anavyssos, Greece.
Steven Carey, Katherine L. Croff Bell, Kat Cantner and Julie Fero Martin, Graduate School of Oceanography,
University of Rhode Island, Narragansett, USA.
49
Stuttgart, Comments
Septemberon2013
the neotectonics of the coastal zone of west Messenia
Comments on the neotectonics of the coastal zone of western Messenia

based on luminescence dating and geoarchaeological evidence
Constantin Athanassas, Ioannis Fountoulis †, Ilias Mariolakos, Yiannis Bassiakos, Zafeiris
Karotsieris, Maria Triantaphyllou and Katerina Theodorakopoulou
with 6 figures and 1 table
Abstract. The coastal regions of western Messenia belong to a regime of substancial vertical neotectonic
deformation. This has been an ongoing procedure since at least the Early Quaternary. Along with the crus-
tal displacement, sea level has undergone fluctations over the same period of time. The interaction between
uplifting Messenia and the oscillating sea level has resulted in a flank of wave-cut terraces. So far, the neo-
tectonic configuration of western Messenia was only qualitatively approached. Therefore, a chronological
framework based on numeric dating is necessary to estimate rates of vertical dislocation. Optically Stimu-
lated Luminescence (OSL) dating of sediments has been utilised in this research in order to yield ages of
tectonic and eustatic changes. Nevertheless, saturation of the OSL signal restricted luminescence dating
to the Upper Quaternary only. Wang et al. (2006a, b) proposed that the thermally transferred OSL (TT-
OSL) signals from quartz are suitable for dating older geologic events, thus allowing us to acquire a more
comprehensive picture of the geoenvironmental evolution of western Messenia. Besides, the coastal areas
of Navarino have exhibited evidence of human activity since prehistoric times. Ancient human activity was
controlled by coastal processes and thus its remains may comprise important indicators of coastal displace-
ment. Specifically, Hellenistic and Roman archaeological sites are useful markers of tectonic subsidence and
thus they demonstrate localised diversions from the general pattern of uplift. In this study, geomorphic,
geochronological and geoarchaeological data were merged, in order to understand the neotectonic pattern
that underlines the geoenvironmental evolution of Messenia.
Key words: Messenia, neotectonics, OSL dating, Geoarchaeology
1 Introduction
In western Messenia (Fig. 1), the configuration of the topography and the coastline is control-
led by vertical movements of the upper crust which have been taking place since at least 1.6 Ma
ago (Mariolakos & Fountoulis 1990, Mariolakos et al. 1994a, b). Along with the tectonics,
sea-level underwent oscillations over the same period of time. The interaction between the tec-
tonically uplifting areas of western Messenia and highstanding sea-levels lead to the formation of
a sequence of raised marine terraces. These occupy extensive areas of western Messenia, and sug-
gest continuous uplift of the coastal areas for the entire Pleistocene (Kelletat et. al. 1976). Tec-
tonic and sea level changes have been also recorded in the local biostratigraphy, as revealed from
the study of sediments from the terraces (Marcopoulou-Diacantoni et al. 1990, Fountoulis
& Moraiti 1994, Mariolakos et al. 1994a, b).
In addition, traces of human activity appear on the modern coastline. They capture a period
from the Middle Paleolithic, through the Upper Paleolithic and Neolithic, to the Late Helladic
DOI: 10.1127/0372-8854/2013/S-00143 0372-8854/13/S-00143 $ 3.50
50 Constantin Athanassas et al.
period and historical times (Davis et al. 1997, Zangger et al. 1997, Parkinson et al. 2010).
Nevertheless, posterior archaeological sites, such as Hellenistic and Roman (Korres 1976,
1977) currently appear submerged in respect to the modern sea-level (Kraft et al. 1980).
These conditions constitute important evidence for localised deviations in the general uplifting
trend of western Messenia.
More systematic study of the morphology allows the division of the study area into two dif-
ferent physiographic units: one in the North (Fig. 1 A) and another one in the South (Fig. 1 B).
The northern part includes the aforementioned raised marine terraces (Fig. 2) and is dominated
by a rocky shoreline, lacking coastal plains and sandy beaches. Besides, intense errosion is evi-
dent in area A, in the form of deeply-incised gorges (e.g. Lagouvardos and Filiatrino ravines). To
the contrary, in the southern part, which more or less coincides with the coastal plain bordering
Fig. 1. Three-dimesional representation

of the study area in western Messenia.
Two different physiographic units can
be identified (A and B) on the basis of
coastline morphology and archaeologi-
cal sites. Archaeological evidence is poor
in area A while it is very rich in area B.
Comments on the neotectonics of the coastal zone of west Messenia 51
Bay of Navarino to the north (Fig. 2 Β), the geomorphic conditions are different. There, instead
of well-formed terraces, the area is occupied by alluvial plains and dune fields, and the relief is
much lower in comparison to the northern part. Signficant differentiation in the archaeological
record is also observed between the two parts. In the northern part the archaeological findings
are very rare, whilst in the southern part there is abundance of Paleolithic, Neolithic, Helladic,
Hellenistic, Roman and posterior archaeological sites.
This paper reports on efforts to elucidate the neotectonic processes underlying the differ-
ential morphological evolution of the coastal zone of western Messenia which has controlled
the development of archaeological settlements in the northern and southern parts, respec-
tively. Nevertheless, most previous studies on the neoteconic evolution of the Ionian coast
of the Messenian peninsula were only qualitative (Kelletat et al. 1976, Mariolakos et al.
1994b). Consequently, numeric dating of the raised marine terraces was deemed necessary to
numerically constraint the rate of vertical tectonic movements. Optically Stimulated Lumines-
cence (OSL) dating on sediments from western Messenia was employed in this study. OSL
dates the last exposure of sedimentary mineral grains to sunlight and this event represents the
time of sediment deposition (e.g. Duller 2004, Rhodes 2011). Nevertheless, recent experi-
ments (Athanassas 2011) showed that in conventional OSL dating using quartz from Mes-
senia, saturation of the luminescence signal makes it difficult to estimate ages beyond 100,000
years (Athanassas 2011). It is therefore necessary to develop a new OSL dating methodology
that makes feasible the geochronology of older stages of the Quaternary. Recent progress in
OSL dating methodologies (Wang et al. 2006a, b 2007, Tsukamoto et al. 2008, Pagonis et al.
2008, Athanassas & Zacharias 2010, Porat et al. 2010, Stevens et al. 2010, Kim et al. 2010,
Adamiec et al. 2010) can enhance the age range by engaging as the thermally transferred-OSL
signal (TT-OSL). Indeed, their work showed that TT-OSL is potentially able to date samples
beyond 100 ka, thus allowing us to deepen into the chronology of earlier Quaternary geologic
events of western Messenia.
2 Geomorphic, geologic and archaeological context
Although no agreement has been realized among researchers about the actual number of marine
terraces in western Messinia, (Kelletat et al. 1978, Kourambas & Robertson 2000), our analy-
sis by means of Geographic Information Systems (ArcGIS 9.0) has revealed the existence of at
least four terraces (Ι, ΙΙ, ΙΙΙ, ΙV in Fig. 2), which ascend from W to E and are situated at aver-
age altitudes of 30 m, 90 m, 160 m and > 200 m, respectively (Fig. 2). The terraces exhibit a
characteristic sagittal shape, parallel to the modern coastline. Fieldwork has revealed that these
landforms contain variable marine deposits such as brackish and shallow marine sands/sand-
stones, as well as deeper water sediments (e.g. marls). All of them have been deposited on an
Early Quaternary calci-marly basement. The latter has been verified by paleontological examina-
tion of calcareous nannofossils in this project. Abundance of nannoplanktonic populations such
as Gephyrocapsa spp. and the co-existence of Pseudoemiliana lacunosa, in association with the
absence of Helicosphaea sellii, allow chronostratigraphic association of the claci-malry basement
with the biozone MNN19e (Rio et al. 1990) and constrain the age of the basement of the marine
Fig. 2. Digita terrain model for the northern part of the study area. GIS-analysis of topographic maps
revealed the existence of at least four terraces (Ι to IV), possibly related to Middle-Late Quaternary trans-
gressions.
deposits between 1.25–0.97 million years (Lourens et al. 2004). A typical sequence of the depos-
its contained in the terraces is illustrated in Fig. 3. Stratigraphic alternations, grain size proper-
ties as well as inclusions of fossil fauna reflect the proximity of these deposits to the sea surface,
rendering them appropriate for the study of vertical movements.
Concerning the morpho-neotectonic configuration of the northern part of the study area (A),
central segments are characterised by higher altitudes while elevation tends to progressively
drop in northward and southward directions. This topographic configuration gives the impres-
sion of dome-like structure. Indeed, Mariolakos & Fountoulis (1990) produced subsurface
maps that reveal the morphology of the contact between the Quaternary marine deposits and the
alpine basement. They claimed that subsurface morphology of the contact between the two geo-
logic units seems like an anticline macro-fold, developed across a ENE-WSW virtual axis. They
related this subsurface structure to differential uplift, with the central areas uplifting faster than
the northern and southern rims.
By applying luminescence dating to raised marine terraces, we aim to associate the vertical
movements in western Messenia with certain stages of the Quaternary. Sampling for the purposes
of luminescence dating invoved the collection of sands and marls from natural sections located
on the coastline and hinterland.
As regards to the southern part of the study area, numerous archaeological sites (Gialouris
1966, Korres 1982, 1983, McDonald & Hope Simpson 1964, Pritchett 1965, McDonald &
Rapp 1972, Kraft et al. 1980, Kaltsas 1981, 1982, 1983) appear the periphery of the Gialova
Lagoon (Fig. 4). The majority of the sites, mainly tombs, date to the Hellenistic and Roman peri-
ods, while sites associated with older periods are lesser in number. However, many of the tombs
are currently waterlogged by the lagoon. Specifically, tombs appearing on the edges of the beach
barrier at Gialova have their bases laying about 1 m below modern sea level approximately. Given
that these structures should have been built at some height above the ancient sea-level, the cur-
rent position of the tombs suggests a relative sea-level rise since Hellenistic times. Further to the
Fig. 3. Typical sequence of shallow ma-

rine deposits and sample positions in a
coastal cliff at Filiatra (FTR in Fig. 2).
NW and within area B, a sequence of partially submerged fossil dunes (aeolianites) outcrop at
Petrochori (Fig. 4). Given that coastal dunes are formed higher than the sea level, the current
position of the aeolianites could be attributed to subsidence. According to these observations it
therefore can be hypothesized that not only the lagoon but area B as a whole are under subsid-
ence. It only remains to investigate whether subsidence owes to relative sea level rise, or to neo-
Fig. 4. This map portrays the topography and geographic distribution of archaeological sites in the southern
part of the study area (area B). The light colored band represents a transitional zone across which the kine-
matic behavior of the study area changes. Most probably it coinsides with a fault zone. Signs express the type
of vertical movement (‘+’: uplift, ‘–’: subsidence).
tectonics or to both of these. Currently available archaeological findings did not facilitate the
application of OSL dating to sediments from the archaeological sites at Gialova, allowing only
a qualitative approach to the matter of the neotectonic behaviour of the lagoon. Nevertheless,
samples from the aeolianites at Petrochori were collected and submitted for OSL analysis.
3 Luminescence dating procedures
Optically Stiumlated Luminescence (OSL) consits a natural ‘chronometer’ which can be reset by
direct exposure to sunlight. The last exposure to sunlight prior deposition is called ‘bleaching’
and this is the event that OSL dates. Luminescence signal intensity is dependant on the amount
of exposure to natural radioactivity. The estimation of a luminescence age is a matter of calcu-
lation of two parameters: the amount of absorbed dose which is called the ‘paleodose’ and the
rate at which the dose was absorbed and it is called the ‘dose rate’. Regarding the paleodose esti-
mation, the Single Aliquot Regenerated (SAR) dose protocol for sedimentary quartz (Murray
& Wintle 2000) is the most widespread methodology. The basic advantage of SAR is that cor-
rects any changes in ‘sensitivity’ of quartz which may occur during laboratory measurements.
Calculation of the paleodose using SAR involves measurement of the natural OSL signal and
then its regeneration by artificial doses of known size. The dependence of OSL response to the
increasing dose is described by a saturating exponential curve which is termed the growth curve
(Fig. 5). Projection of the natural’s signal intensity onto the growth curve results in the inquired
paleodose.
With respect to the calculation of the dose-rate, this involves measurement of the sediment’s
content in U, Th and K, plus a cosmic-ray dose rate contribution. Radioelement concentrations
can be converted into radioactivity emitted/ppm of radioelement by using conversion factors
(Adamiec & Aitken 1998). The cosmic ray contribution is calculated by specific equations
which take into account coordinates, sample depth, altitude etc. The ratio of the paleodose over
the overall dose rate represents the luminescence age of the sample.
Nevertheless, the growth of the conventional OSL signal in quartz grains from Messenia
was found (Athanassas 2011) to be nearly saturated at doses higher than 100 Gy (i.e. limit-
ted response to progressively increasing doses). Taking also into account that the average local
dose rate does not exceed 1 Gy/ka on average (see Table 1), saturation restricts the exploration of
paleoenvironmental events in Messenia to the Upper Quaternary only.
On the other hand, the thermally transferred-OSL (TT-OSL) signal from quartz has higher
capacity and can overcome complications risen by saturating conventional OSL signals, allowing
us to extend the dating range (Wang et al. 2006a, b).
TT-OSL generates with additional thermal treatment of the sample after measuring the con-
ventional optical signal. This treatment results in the generation of a new OSL signal but of lower
intensity (Fig. 5a). TT-OSL can linearly grow, in response to a continuously increasing dose, con-
siderably further than ordinary OSL, as shown in Fig. 5b (Wang et al. 2006a, b, 2007, Tsukamoto
et al. 2008, Pagonis et al. 2008, Athanassas & Zacharias 2010, Porat et al. 2010, Stevens
et al. 2010, Kim et al. 2010, Adamiec et al. 2010).
Hereby, TT-OSL dating was applied to sedimentary sand-sized quartz grains from marine
deposits of western Messenia following the SAR methodology introduced by Wang et al. (2007),
but taking also into account Adamiec et al. (2011) considerations. Fig. 5 contrasts the growth
of the TT-OSL signal to that of standard OSL. It can be observed that the response of TT-OSL
increases in a linear manner, while conventional OSL is practically saturated for the same range
of doses. Table 1 summarizes the paleodoses of 10 samples from the terraces of western Messenia,
their dose rates and resulting age estimates (sample sites seen in Fig. 2). The last column in Table 1
arrays the TT-OSL ages in thousands of years (ka). Moreover, the conventional OSL ages are
given for comparisons.
The majority of the OSL-paleodoses appear underestimated in comparisson to these mea-
sured with TT-OSL. This prossibly owes to the overestimation that characterizes paleodoses
resulting from saturated OSL growth curves (Athanassas 2011). Conversely, TT-OSL exhibits
higher luminescense capacity and the TT-OSL paleodoses are the most likely candidates.
Concerning the dose rates, they were calculated by estimating the sediment’s content in U, Th
and K by means of Neutron Activation Analysis (NAA) at the University of Missouri, Research
Reactor Centre. The calculated TT-OSL ages fall in range of 67.76 ± 8.57 – 305 ± 42.5 ka. These
data indicate that the project is dealing with Mid-Late Quaternary deposits.
4 Interpretation
Although the laboratory performance of TT-OSL dating procedure engaged here (i.e. sensitivity
change corrections, recycling ratios, dose recovery) was found to be sufficient (see Athanassas
& Zacharias 2010 for a detailed analysis of the performance, suitability and applicability of
TT-OSL to these samples), unfortunately, no means of independent age control on the TT-OSL
dates was available. Diagnostic nannofosils were present only in the earlier Pleistocene depos-
Fig. 5. a) This figure contrasts a conventional OSL signal to the TT-OSL signal produced after heating at
260 °C a quartz sample from a beachrock in western Messenia. b) Dose response of TT-OSL, in contrast to
conventional OSL for the same quartz sample.
Table 1. Samples, radioelement concentration, dose rates, palodoses and resulting OSL and TT-OSL ages.
Sample U (ppm) Th (ppm) K (wt %) Dose rate OSL P (Gy) OSL Age TT-OSL P (Gy) TT-OSL Age (ka)
(Gy/ka) (ka)
MAT1 0.70±0.09 1.07±0.27 0.56±0.03 0.82±0.03 94.33±18.66 115.04±23.00 73.50±10.07 91.90±12.30

MAT2 0.52±0.11 1.60±0.33 0.54±0.03 0.78±0.03 83.00±18.00 110.66±24.00 57.33±8.20 76.44±10.00
AGS1 0.47±0.50 0.66±0.27 0.82±0.02 1.01±0.03 174.33±21.08 172.61±20.90 145.6±13.30 144.12±13.20
AGS2 0.79±0.11 1.58±0.38 0.47±0.03 0.86±0.04 98.00±27.04 113.95±31.40 90.00±8.90 104.65±10.30
GRG1 1.02±0.11 0.90±0.17 0.07±0.18 0.69±0.02 124.02±12.30 179.74±17.82 100.80±8.64 146.87±12.52
LGV1 0.81±0.15 2.01±0.43 0.25±0.03 0.89±0.02 225.25±7.64 253.01±8.58 231.00±12.20 259.55±13.61
FTR1 0.78±0.13 1.71±0.41 0.42±0.03 0.80±0.03 147.41±19.24 184.26±24.05 84.31±9.07 105.39±11.34
FTR2 0.45±0.49 0.62±0.30 0.78±0.02 0.93±0.03 167.50±41.36 180.11±44.47 105.00±14.11 112.90±15.17
PTC1 0.83±0.06 1.13±0.17 0.17±0.02 0.60±0.01 72.10±10.20 120.00±17.00 ––– –––
KRN6 0.74±0.04 0.61±0.12 0.07±0.01 0.40±0.01 176.50±38.40 438.00±96.00 122.21±17.00 305.52±42.50
Comments on the neotectonics of the coastal zone of west Messenia
57
its whilst the sands and sandy marls submited for TT-OSL analysis were found to be barren.
Because of this, the degree of reliability of the ages genereted here could be assessed only by
contrasting them against the paleoenvironmental context indicated by foraminifera retrieved
from the dated material, as well as by juxtaposing the ages to the sequence of marine terraces
that these samples were associated with. Both approaches indicated a fair degree of agree-
ment between the TT-OSL chronology and the secular sea level record. Reader is referred to
Athanassas & Zacharias (2010) and Athanassas & Fountoulis (2013) for a detailed anal-
ysis of those results.
On the above assumptions, luminescense data are exploited a little further. The OSL ages
of Table 1 are contrasted to the geological and archaeological record. Fig. 6 arrays, in a SE-NW
direction, those samples of Table 1 which are useful in stratigraphic implications. Respective ages,
placed in stratigraphic and geographic order, are plotted against the relative sea-level changes of
the last 350 ka (Lea et al. 2002). Given the relatively large errors appended on the TT-OSL ages,
direct comparisons with individual sea level peaks is not secure. For this reason, ages in Fig. 6 are
illustrated as patterned intervals capturing a time-span.
Firstly, TT-OSL ages associate with warm stages (transgressions) of the Middle-Late Quater-
nary. Indeed, paleontological data support the warm water character of the dated samples (rare
incidence of Globigerinoides ruber), strengthening in this way the reliability of the luminescence
ages.
Fig. 6. SE-NW section of the northern part of the study area displaying sample positions, elevations and
Re-OSL ages and against former sea levels (isotopic data from Lea et al. 2002). Dashed horizontal lines rep-
resent the maximum and minimum oscillation of sea level over the last 350 ka. Arrows indicate roughly the
magnitude of uplift rate for some sites.
Noticeable is, however, the relationship of samples GRG1 and LGV1 with lower peaks (at 147
and 250 ka respectively) at the end of the overall sea-level drop. This observation contravenes the
other sites where ages are related to high sea level events. Nevertheless, taking into account the
strong uplift tendency in the central part of the area, appearance of sediments associated with
lower sea levels above modern sea level is not impossible. In such a case, ages of samples GRG1
and LGV1 might be justified.
In order to estimate uplift rates, Fig. 6 contrasts the TT-OSL ages with their correspond-
ing sea-levels. Comparison indicates higher uplift rates in the middle part of the area than in its
southern and northern edges. These estimations are in agreement with the assumption accord-
ing to which the central parts of the area are uplifting faster than its north and south margins
(Mariolakos & Fountoulis 1990). This striking change in the uplift velocity owes to the activ-
ity of faults which traverse western Messenia, producing a range of tectonic blocks with different
uplifting velocities.
On the other hand, the neotectonic conditions that influence the environment of the south-
ern part (B) of the study area are significantly different (Fig. 4). Luminescence ages of partially
submerged fossil sand dunes (aeolianites) at Petrochori (PTC1 in Fig. 2, Table 1) yielded ages
of 120 ± 17 ka. This result indicates that this part of area B has been under subsidence for at
least during the last 120 ka. Archaeological observations at Gialova are of equal importance.
Immersed tombs are the result of an ongoing subsidence. Nevertheless, a relative sea level rise
in the last 2,000 years should be taken into consideration. Relative sea level curves for the Bay of
Navarino (Kraft et al. 1980) and the nearby Messenian Gulf (Kraft et al. 1975) as well as for
other coastal areas of the Peloponnese and the eastern Mediterranean (Caputo & Pieri 1976,
Lambeck 1996, Lambeck & Purcell 2005) estimate a relative sea level rise of about 1 m during
the last two millennia. Taking also into account that the base of the tombs should be at least 1 m
higher than the Greco-Roman sea-level of -1 m, we can estimate a net tectonic subsidence of at
least 1 m during the last 2,000–2,300 years. This rationale implies a subsidence rate of 0.5 mm/y,
an estimate close to Flemming et al. (1973) calculations for the same site. By combining the
aforesaid observations, it can be concluded that the entire area B is under subsidence. There-
fore, an inversion in the kinematic behavior of western Messenia is taking place. This inversion
potentially occurs on both sides of Selas River and it can be attributed to the activity of a fault
zone (light green area). The inversion of the kinematic behavior is designated by different signs
in Fig. 4.
5 Conclusions
The Ionian coastal zone of Messenia is characterized by a variety of geomorphic features, the
geographic distribution of which allows the division of the area into two parts; one in the north
typified by a rough coastline and another one in the south which is described by a smooth coastal
plain. Novelties in luminescence dating, this is the thermally transferred-OSL, allowed us to
extend the age range in the SW Peloponnese. Application of TT-OSL to raised marine sediments
from the northern study area revealed that this part of western Messenia has been uplifting since
at least the last 300 ka. To the contrary, the southern segments of the area are subsiding, as seen
from sunken archaeological sites. This subsidence pattern can be dated to at least the Graeco-
Roman period, possibly older, as submerged Late Quaternary dunes at Petrochori may imply.
This diversity takes place across Selas River. It is not implausible that the flows of the river coin-
cides with a fault zone which alters the kinematic behaviour of the area across of it.
To sum up, it is the neotectonics which influences the landscape evolution of western Mes-
senia, generating of alluvial lowlands and smooth costal plains in the south but relatively rugged
coastlines to the north. In contrast to the nortern parts, the topography on the bay of Navarino
has favoured the development of human settlements since the Prehistroric times.
Acknowledgments
A major part of this manuscript was written in 2007 as a preliminary report on miscellaneous
observations from a project, and hosted here for the purposes of the Geomorphology meeting
held in Athens in 2011. The reader is referred to Athanassas & Fountoulis 2013; Athanassas
et al. 2012 and Athanassas & Zacharias 2010 for more comprehensive and up-to-date discus-
sions on the matter. Authors highly appreciate comments and suggestions by one anonymous
reviever.
References
Adamiec, G. & Aitken, M. J. (1998): Dose-rate conversion factors: update. – Ancient TL 16: 37–50.
Adamiec, G., Duller, G. A. T., Roberts, H. M. & Wintle, A. G. (2010): Improving the TT-OSL SAR
protocol through source trap characterization. – Radiat. Measurem. 45: 768–777.
Athanassas, C. (2011): Constraints on the precision of SAR in equivalent dose estimations close to satura-
tion in quartz. – Geochronometria 48: 413–423.
Athanassas, C., Bassiakos, Y., Wagner, G. A. & Timpson, M. E. (2012): Exploring paleogeographic con-
ditions at two paleolithic sites in Navarino, Southwest Greece, dated by Optically Stimulated Lumines-
cense. – Geoarchaeology 27: 237–258.
Athanassas, C. & Fountoulis, I. (2013): Quaternary neotectonic configuration of the southwestern Pelo-
ponnese, Greece, based on luminscense ages of marine terraces. – J. Earth Sci. 24: 410-427
Athanassas, C. & Zacharias, N. (2010): Recuperated-OSL dating of quartz from Aegean (South Greece)
raised Pleistocene marine sediments: current results. – Quatern. Geochronol. 5: 65–75.
Caputo, M. & Pieri, L. (1976): Eustatic variation in the last 2000 years in the Mediterranean – J. Geophys.
Res. 81: 5787–5790.
Davis, J. L., Alcock, S. E., Bennet, J., Lolos, Y. G. & Shelmerdine, S. W. (1997): The Pylos regional
archaeological project part I: overview and the archaeological survey. – Hesperia 66: 391–494.
Duller, G. A. T. (2004): Luminescence dating of Quaternary sediments: recent advances. – J. Quatern. Sci.
19: 183–192.
Flemming, N. C., Czartoryska, N. M. G. & Hunter, P. M. (1973): Archaeological evidence for eustatic
and tectonic components of relative sea level change in the south Aegean. – Proc. 23rd Symp. Colston
Res. Soc., Univ. Bristol, 1971.
Fountoulis, I. & Moraiti, E. (1994): Sedimentation, paleogeography, and Neotectonic interpretation of
post-alpine deposits in the Kyparissia-Kalo Nero Basin. – Bull. Geol. Soc. Greece 30: 323–336.
Gialouris, N. (1966): Arcaheologikon Deltion (in Greek), 21: 164–165.
Kaltsas, N. (1981): Archaeologikon Deltion (in Greek), 36: 152.
Kaltsas, N. (1982): Archaeologikon Deltion (in Greek), 37: 137.

Kaltsas, N. (1983): Archaeologikon Deltion (in Greek), 38: 1–77.
Kelletat, D., Kowalczyc, G., Schröder, B. & Winter, K. P. (1976): A synoptic view on the neotectonic
development of the Peloponnesian coastal regions. – Z. dt. Geol. Ges. 127: 447–465.
Kim, J. C., Duller, G. A. T., Roberts, H. M., Wintle, A. G., Lee, Y. I. & Yi, S. B. (2009): Dose dependence
of thermally transferred optically stimulated luminescence signals in quartz. – Radiat. Measurem. 44:
132–143.
Korres, G. (1982): Anaskafi Voidokoilias Pylias. – Praktika Archaeologikis Etairias (in Greek): 191–231.
Korres, G. (1983): Anaskafi Voidokoilias Pylias. – Praktika Archaeologikis Etairias (in Greek): 169–208.
Kourampas, N. & Robertson, A. H. F. (2000): Controls on Plio-Quaternary sedimentation within an active
fore-arc region: Messinia Peninsula (SW Peloponnese), S. Greece. – In: Panayides, I., Xenophontos, C.
& Malpas, J. (eds) (2000): Proceedings of the Third International Conference on the Geology of the
Eastern Mediterranean, 255–285.
Kraft, J. C., Rapp, G. Jr. & Aschenbrenner, S. E. (1975): Late Holocene paleogeography of the coastal
plain of the Gulf of Messenia, Greece, and its relationships to archaeological settings and to coastal
change. – Geol. Soc. Amer. Bull. 86: 1191–1208.
Kraft, J. C., Rapp, G. R. J. & Aschenbrenner, S. E. (1980): Late Holocene palaeogeomorphic reconstruc-
tions in the area of the Bay of Navarino: sandy Pylos. – J. Archaeol. Sci. 7: 187–210.
Lambeck, K. (1996): Sea-level change and shoreline evolution in Aegean Greece since Upper Palaeolithic
Time. – Antiquity 70: 588–611.
Lambeck, K. & Purcell, A. (2005): Sea-level change in the Mediterranean Sea since the LGM: model pre-
dictions for tectonically stable areas. – Quatern. Sci. Rev. 24: 1969–1988.
Lea, D. W., Martin, P. A., Pak, D. K. & Spero, H. J. (2002): Reconstructing a 350 ky history of sea level
using planktonic Mg/Ca and oxygen isotope records froma Cocos Ridge core. – Quatern. Sci. Rev. 21:
283–293.
Lourens, L. J., Hilgen, F. J., Shackleton, N. J., Laskar, J. & Wilson, D. (2004): The Neogene Period. – In:
Gradstein F., Ogg. J. & Smith, A. (eds): A Geologic Time Scale. – Cambridge Univ. Press.
Mariolakos, I. & Fountoulis, I. (1990): Neotectonic macrofolds at the Filiatra area (southwestern Pelo-
ponnesus, Greece). – Bull. Geol. Soc. Greece 15: 19–38.
Mariolakos, I., Fountoulis, I., Marcopoulou-Diacantoni, A. & Mirkou, M. R. (1994a): Some
remarks on the kinematic evolution of Messinia province (SW Peloponnesus, Greece) during the Pleisto-
cene based on Neotectonic Stratigraphic and Paleoecological observations. – Munster Forsch. Geol.
Palaont. 76: 371–380.
Mariolakos, I., Badekas, I., Fountoulis, I. & Theocharis, D. (1994b): Reconstruction of the Early
Pleistocene paleoshore and Paleorelief of SW Peloponnesus area. – Bull. Geol. Soc. Greece XXX/2:
294–304.
Marcopoulou-Diacantoni, A., Mirkou, M. R., Mariolakos, I. & Fountoulis, I. (1991): Stratigraphic
and paleoecological observations on the post alpine sediments at the area of Filiatra (Messinia, Pelo-
ponnesus) and their neotectonic explanation. – Bull. Geol. Soc. Greece 25: 593–688.
McDonald, W. A. & Hope Simpson, R. (1964): Further explorations in the southwestern Peloponnese:
1962–1963. – Amer. J. Archaeol. 68: 229–245.
McDonald, W. A. & Rapp, G. Jr. (1972): The Minnesota Messenia Expedition: Reconstructing a Bronze
Age Regional Environment. – Univ. Minnesota Press, Minneapolis.
Murray, A. S. & Wintle, A. G. (2000): Luminescence dating of quartz using an improved single – aliquot
regenerative – dose protocol. – Radiat. Measurem. 32: 57–73.
Pagonis, V., Wintle., A. G., Chen, R. & Wang, X. L. (2008): A theoretical model for a new dating protocol
for quartz based on thermally transferred OSL (TT-OSL). – Radiat. Measurem. 43: 704–708.
Porat, N., Duller, G. A. T., Roberts, H. M. & Wintle, A. G. (2009): A simplified protocol for TT-OSL. –
Radiat. Measurem. 44: 538–542.
Rhodes, E. J. (2011): Optically Stimulated Luminescence dating of sediments over the last 200,000 years. –
Ann. Rev. Earth Planet. Sci. 39: 461–488.
Rio, D., Raffi, I. & Villa, G. (1990): Pliocene-Pleistocene calcareous nannofossil distribution patterns in
the Western Mediterranean. – In: Kastens, K. A., Mascle, J. et al. (eds): Proc. ODP. – Sci. Res. 107:
513–533.
Pritchett, W. K. (1965): Studies in Ancient Greek Topography, Part I. – Univ. California Press, Berkeley,
Los Angeles.
Stevens, T., Buylaert, J. P. & Murray, A. S. (2009): Towards development of a broadly-applicable SAR
TT-OSL dating protocol for quartz. – Radiat. Measurem. 44: 639–645.
Tsukamoto, S., Duller, G. A. T. & Wintle, A. G. (2008): Characteristics of thermally transferred opti-
cally stimulated luminescence (TT-OSL) in quartz and its potential for dating sediments. – Radiat.
Measurem. 43: 1204–1218.
Wang, X. L., Wintle, A. G. & Lu, Y. C. (2006a): Thermally transferred luminescence in fine-grained quartz
from Chinese loess: basic observations. – Radiat. Measurem. 41: 649–658.
Wang, X. L., Wintle, A. G. & Lu, Y. C. (2006b): Recuperated OSL dating of fine- grained quartz in Chinese
loess. – Quatern. Geochronol. 1: 89–100.
Wang, X. L., Wintle A. G. & Lu, Y. C. (2007): Testing a single-aliquot protocol for recuperated OSL Dating.
– Radiat. Measurem. 42: 380–391.
Zangger, E., Timpson, M. E., Yazvenko, S. B., Kuhnke, F. & Knauss, J. (1997): The Pylos Regional
Archaeological Project: Part II: Landscape Evolution and Site Preservation. – Hesperia 66: 549–564.

Constantin Athanassas and Yiannis Bassiakos, Laboratory of Archaeometry, Institute of Materials Science,
N.C.S.R. ‘Demokritos’, Aghia Paraskevi, 15310, Athens, Greece, e-mail: Hathanassas@ims.demokritos.gr,
bassiakos@ims.demokritos.gr
Ioannis Fountoulis, Ilias Mariolakos and Zafeiris Karotsieris, Department of Dynamic, Tectonic and Applied
Geology, Faculty of Geology and Geoenvironment, University of Athens, Zografou, 15784, Athens, Greece,
e-mail: Fountoulis@geol.uoa.gr, Mariolakos@geol.uoa.gr, karotsieris@geol.uoa.gr
Maria Triantaphyllou, Department of Historical Geology and Paleontology, Faculty of Geology and
Geoenvironment, University of Athens, Zografou, 15784, Athens, Greece, e-mail:mtriant@geol.uoa.gr
Katerina Theodorakopoulou, Department of Geography, Harokopio University, Kallithea, 17671 Athens,
Greece, e-mail: ktheodo@hua.gr
63
Automated
Stuttgart, Identification
September 2013 of Geomorphological and Topographical Features
Automated Identification of Geomorphological and Topographical

Features of the Seabed by Object Based Image Analysis of
Digital Terrain Models
K. Siakavara and D. Argialas
with 13 figures and 1 table
Abstract. Seabed feature identification is of great importance in many engineering and geomorphological
projects. Multibeam echo sounder datasets are often used in those projects for tasks such as producing Digi-
tal Terrain Models (DTM). An effort in automating the interpretation and delineation of seabed geomor-
phological and topographical features from a DTM and its derivatives using an object based image analysis
approach was developed. The approach consists of four basic steps. First objects were created through multi-
resolution segmentation of the DTM. Then the classes of geomorphological features were initially defined
and were successively refined through the multiresolution analysis procedure. The final classes were smooth
ground, channels, scars, mounds, ridges and landslide scars. Subsequently each object was assigned to one of
the classes representing those bedforms by fuzzy classification. Finally in the fourth step a visual qualitative
evaluation of the results was performed.
Key words: seabed, bedforms, object based image analysis, multibeam echo sounder
1 Introduction
Seabed interpretation is considered a rather tedious and time consuming task requiring exper-
tise. Therefore semi-automated or computer assisted processes are crucial in terms of economy.
On the other hand they offer a rather objective formalization of the interpretation process ame-
nable to critical review. Seabed interpretation usually involves the use of backscatter imagery,
sub bottom profiling, sediment samples, in situ photography and groundtruthing video data
as well as bathymetry data and products created through the processing of those data such as
Digital Terrain Models (DTM), Digital Depth Models (DDM), isobath and isopach maps etc.
The acquisition of those datasets is difficult and expensive and thus not easily available for
research. In many case studies there is also a lack of those data such as groundtruthing videos
or core samples.
Chapple (2008) studied the current possibilities of automated detection of objects and espe-
cially mines on the surface of the seabed using imagery from side-scan sonar on Autonomous
Underwater Vehicles (AUV). High-resolution swath bathymetry analysis with a GIS-imple-
mented approach to identify and extract pockmarks on the seabed was successfully applied by
Andrews et al (2010). An approach including as a first step automated segmentation of the raw
echoes of a single-beam acoustic system into clusters (objects) implemented in the software
IMPACT, followed by a labeling of the clusters with descriptive names (classes) by reference to
DOI: 10.1127/0372-8854/2013/S-00144 0372-8854/13/S-00144 $ 3.00
64 K. Siakavara and D. Argialas
other data was tested classifying the study area into different types of bottom and sediment types
(Gleason et al. 2008). Multibeam bathymetry, slope, bathymetric position index and rugosity
were combined in a GIS and assessed with unique algorithms to produce classification maps in
Tutuila, American Samoa by Lundblad et al. (2006). The classification scheme included thir-
teen classes such as narrow depression, local depressions on flat, open slopes etc.. A survey with
airborne laser bathymetry data by Finkl et al. (2005) for the classification of geomorphological
units for a segment of the continental shelf off Delray Beach, Florida included freehand draw-
ing of the geomorphological units (manual segmentation). In general, automated classification
schemes have been applied in many case studies (Szuman et al. 2005) using backscatter informa-
tion, bathymetry imagery as well as DEM derivatives (Bathymetric Position Index, slope, aspect
etc.) but multiresolution segmentation and fuzzy classification, as applied in this paper, has not
been fully investigated.
Our team has worked for years in the interpretation of landforms, physiographic pro-
cesses and topographical forms on earth’s surface (Argialas 1995, Miliaresis & Argialas
1999). Most recently object based image analysis was developed for the interpretation of these
entities using the digital elevation model GTOPO30 of the area of Basin and Range, Nevada.
The study successfully extracted basins, mountains and piedmonts (Argialas & Tzotsos
2002).
Object Oriented or Object Based Image Analysis (OBIA) is a method based on the idea that
important semantic information necessary to interpret an image is mostly not represented in sin-
gle pixels but in meaningful image objects and their mutual relations (BAATZ&SCHAPE 1999)
and has been implemented in the interpretation and classification of satellite imagery and Digital
Elevation Models (DEM) of the land surface (Argialas & Tzotsos 2003). OBIA has also been
used for seabed studies applying several segmentation and classification techniques mainly on
backscatter imagery (Penrose et al. 2005).
The objective of this study was to assess whether a successful OBIA approach is feasible in the
characterization of bedforms using a DTM and certain of its derivatives towards fully automating
seabed classification.
2 Datasets and DTM derivatives
A DTM of the seabed in the area of Rio Antirio in the Gulf of Corinth, Greece, with a resolution
of 2.5 m was used (Fig. 1), courtesy of akti engineering, an engineering, services and consulting
company. The DTM was created through processing multibeam echo sounder data of the area.
The segmentation and classification algorithm created in this study was also applied in a DTM
of the area between the islands Andros and Evia with a resolution of 10 meters in order to test its
applicability in different datasets.
The first step in this approach was the creation of certain derivatives of the DTM, espe-
cially slope gradient and aspect since the extraction of landforms/bedforms depends rather
on these attributes instead of elevation alone. Furthermore, positive openness was considered
for investigation as an additional morphometric index. Slope gradient and aspect images were
Automated Identification of Geomorphological and Topographical Features 65
created according to Zevenbergen & Thorne (1987). Positive openness (Yokoyama et al.
2002) is an angular measure of the relation between surface relief and horizontal distance.
At a DTM grid point it is the average of zenith angles viewed above the surface along each of
the eight azimuths D to a specified distance L. Different L values emphasize fine- or coarse-
scale features. For the purposes of this study the L value was chosen to be rather low, that
was 2.5m, equal to the DTM resolution so that the finer scale features to be visible. Values of
surface openness are displayed in map form (Fig. 2). The DTM derivatives were created in
the open source software SAGA GIS in which positive openness is referred to as Morphometric
Protection Index (MPI).
Fig. 1. Shaded relief images of the DTM Rio-Antirio showing microrelief (a), a big channel (b) and smooth
ground (c).
Fig. 2. Positive Openness-MPI image of the Rio Antirio DTM with L value of 2.5 m.
3 Object Based Image Analysis (OBIA)
OBIA is based on the concept that the interpretation of an image cannot only depend on pixels
and pixel based techniques but on meaningful objects created through segmentation of the image
and their attributes and mutual relationships. The OBIA approach includes two basic procedures:
segmentation and classification and in this study was applied through the eCognition software.
(Baatz & Schape 1999).
Through segmentation, the DTM is segmented into objects. Ecognition provides the algo-
rithm “multiresolution segmentation”, with which the pixels are merged into objects taking into
account user set parameters such as the desirable heterogeneity, the size and the shape of the
objects to be created. Different segmentation parameters can be applied creating different objects
in various resolution levels of hierarchy in the same image. Objects in higher levels are referred to
as super-objects, while in lower levels as sub-objects, creating a hierarchical objects network. The
objects created have shape, spatial and spectral attributes such as size, area, position, relations to
super- or sub-objects etc.. Subsequently, the user defines the classes of interest assigning selected
attributes and their values through fuzzy membership functions to each class. The objects of the
image are then classified through fuzzy logic into the selected classes, each obtaining the proper
membership value to each class.
4 Methodology – Implementation
The presented OBIA approach was designed by a trial and error effort to include six different
levels of resolution. The objects on all levels were designed to represent different scales of seabed
features and to facilitate the extraction. Creating a top to bottom approach, larger objects were
created in higher levels and therefore larger scale geomorphological features were classified, while
in lower levels the finer scale features were identified. For each level a set of classes was designed
with the name of each class containing the number of the level that it was designed for i.e. relief 4
for the level 4 relief class. Through the class related features the classification of different levels
was used to assist in the classification of objects belonging to other levels. Slope gradient, slope
aspect and the Morphometric Protection Index were used in the class descriptions (named in the
software as spectral attributes) as well as shape and spatial attributes. The sequence of creating the
levels was 6-5-3-4-1-2 and is explained at each level description. A description of the segmenta-
tion and classification algorithms of all levels in the order of creation follows.
Objects on the sixth and highest level were created only in order to separate the background
of the image from the actual DTM and no further description of it will be made since it is of no
geomorpological or topographical interest. Level 5 contained objects of a medium scale param-
eter (Fig. 3) and was created for classifying in an early stage, large smooth areas with no con-
cavities or convexities. Two classes for this level were employed, smooth ground_5 and relief_5
(Fig. 3). Given that concavities get high values of MPI and convexities really low, the medium to
low numerical limits were set for the smooth ground_5 class through multiple trials in the feature
view tool. Thus the only characteristic an object should have had in order to be recognized as
smooth ground_5 was a medium to low value of MPI. Every object in this level not fulfilling the
criteria of the class smooth ground_5 was classified as relief_5 and was further on classified as a
bedform in lower levels.
The objects on level 3 were created using a smaller scale parameter of 20 and shape (0.6) and
color (0.4) criteria so as to extract large scale elongated features (Fig. 4) as a first step in identify-
ing channel and ridge like objects.
Fig. 3. Objects of level 5 (left image). The classification result of level 5 (right image).
Fig. 4. The multiresolution segmentation parameters as finally set for level 3. By setting the shape criterion
(0.6), slightly higher than the color criterion (0.4) and the smoothness criterion as 1, objects created have
elongated shapes.
For the classification of level 3 a class hierarchy was created containing four main classes: large
scale convex features like_3 (ridges), large scale concave features like_3 (channels), microrelief_3
and smooth ground_3 (Fig. 5). Objects with:
• high mean MPI value were classified in the class large scale concave features like_3 (chan-
nels),
• low mean MPI value (lower than that set for the class smooth ground_5 in level 5) were
classified as large scale convex features like_3 (ridges),
• a super object in level 5 classified as smooth ground_5 were also classified as smooth
ground_3.
All the other objects that did not fulfill those criteria were classified in the class micro-
relief_3. The objects classified as microrelief_3 were further classified as landslides_3 or as micro-
relief besides landslides_3. For an object to be classified as landslides_3 a set of rules involving
fuzzy membership functions were created, including the conditions to be concave, to be next to
an object classified as large scale concave feature like_3 etc.. Any objects that didn’t fulfill these
landslide rules but were classified as microrelief_3, were classified in the class microrelief besides
landslides_3 and were processed and classified further on in lower levels.
Level 3 was created in order to contribute to the creation of level 4. By using the algorithm
classification based segmentation, new segments were defined for level 4 based on merging of
neighboring objects classified in the same class in level 3. This step was necessary in order to
Fig. 5. The classification result in level 3. By using the classification of higher levels the objects in level 3 were
classified as background_3 and smooth_ground_3.
optimize the shape of the objects to be classified as channels and ridges. Both of those bed-
forms require an elongated shape. By merging neighboring objects classified as large scale con-
cave or convex features_3 and by using shape criteria (Table 1) the new objects of level 4 were
classified as large scale channels_4 and large scale ridges_4, or as wrong large scale channels_4
and wrong large scale ridges_4 to be classified as another type of bedform later on. The class
hierarchy designed for level 4 is shown in Fig. 6. The classification result of level 4 is demon-
strated in Fig. 7.
Next level created was level 1, which was the lowest level containing the smaller objects, simi-
lar to pixel size, in order to achieve minimization of loss of information (Argialas & Tzotsos
2003), thus to ensure that the high resolution information of the DTM is taken full advantage of.
For the creation of this level a really small scale parameter of 3 was chosen. Objects created were
similar to pixel size. In this level:
Fig. 6. The class hierarchy designed to classify

objects in level 4. Through the inheritance fea-
ture, attributes of a parent class (i.e. large scale
channels like_4, large scale ridges like_4) are
automatically inherited to the child classes.
Table 1. A brief description of the fuzzy membership functions developed to describe the attributes of each
class in level 4. For an object to be classified as large scale channel_4 both the length restriction and the exist-
ence of sub object large concave features like_3 must be fulfilled, since large scale channel_4 is a child class of
large scale channels like_4, should be valid.
Cla ss FEATURE Me mbe rship functions Le ft Borde r Right Borde r

large_scale_channels_like_4 Existence of: 0 2
large_concave_features_like_3 Sub
Objects
large_scale_channels_4 Length 17 18
wrong_large_scale_channels_like_4 Invert similarity to: large_scale_channels_4

large_scale_ridges_like_4 Existence of: 0 2
large_scale_convex_features_like_3
Sub objects
wrong_large_scale_ridges_like_4 Invert similarity to: large_scale_ridges_4
large_scale_ridges_4 Length/W idth 10 12
• Every object that had as a super- object one not classified as a bedform (i.e. microrelief besides
landslides_3) was classified in the parent class microrelief_1 with child classes other concave_1
and other convex_1.
• Small parts of the DTM not classified as smooth ground_5 in level 5 because of the large scale
parameter were now classified as new smooth ground_1.
• The rest of the class descriptions used the feature “existence of super objects” and objects
were classified according to the classification of their super objects.
A further refinement of the classification results in level 1 was necessary, in order to correct
mis-classifications as shown in Fig. 8 and Fig. 9, therefore a new level, level 2, was created using
Fig. 7. Four different aspects of the classification of level 4. Not all the objects in this level were classified
(marked with x) since the objective of this level was the recognition of large scale ridges_4 (c) and large scale
channels_4 (b). Some wrong large scale ridges like_4 are also indicated in the figure with (a) and some wrong
large scale channels_4 like with (d).
Fig. 8. Objects classified as channels_1 (either gentle slopes_1 or steep_1) only their size was only one pixel
(circled). Through classification based segmentation those misclassifications were corrected and objects
smaller than a specific area were classified as scars.
the algorithm classification based segmentation, merging neighboring objects classified in specific
classes in level 1.
For the classification of level 2 new classes were inserted in the class hierarchy such as scars
on convex_2, mounds on concave_2, new channels_2, new ridges_2, wrong classified as channels to
become scars_2 etc.. The first two classes were employed in order to describe the complexity of the
seabed. Shape attributes (i.e. area, length to width ratio etc., Fig. 10) as well as DTM derivatives mean
values were used for the descriptions of the classes. The classification in level 2 is shown in Fig. 11.
Fig. 9. On the left image the classified objects of level 1 with scale parameter 3 and on the right image the objects
of level 2 created through classification based segmentation.
Fig. 10. Class description of the attributes of selected classes in level 2.

Fig. 11. Classification of objects in level 2.
Fig. 12. The final classification of the

DTM.
Finally level 1 was re-classified using the classification of super objects in previous levels.
Objects belonging to the same thematic category like new channels_2 and large scale channels_4
were grouped in the final class final channels_1. Fig. 12 and Fig. 13a illustrate the resulting clas-
sification.
5 Discussion
The same segmentation and classification algorithm was applied with little differentiation in the
length and area feature in the description of some of the classes, on the DTM of Andros-Evia with
a resolution of 10 m giving similar results.
Due to lack of other data such as ground truthing imagery, usually used in those kinds of
studies for evaluation, visual qualitative inspection of the results using information provided by
Fig. 13. Zoomed in area of the final clas-

sification (a) where some representa-
tive extracted landslide scars (LS), steep
channels (SC), mounds (M) and general
microrelief are indicated compared with
the shaded relief image of the area (b).
professionals of the field and shaded relief images were employed to conclude that the resulting
classification was satisfactory (Fig. 13).
The basic conclusion of this study was that single use of DTM and some of its derivatives with
multiresolution segmentation and object based image analysis can facilitate the identification of
bedforms in the distant and remote environment of the seabed. As a prospect the combination of
backscatter imagery and DTM as well as its derivatives in an object based image analysis software
is recommended as a step closer to fully automating the recognition of bedforms.
Acknowledgements
The authors would like to thank Mr. Palaiokrassas Aris and the company Akti Engineering (www.
aktieng.gr) for granting the data for this study as well as the organizing committee of the 15th Joint
Geomorphological Meeting for accepting the presentation of this study and reviewing it for pub-
lication. We also do thank Dr. Paraskevi Nomikou, Lecturer in Geological Oceanography of the
Department of National Geography and Climatology, University of Athens for reading and com-
menting on this paper. Finally, we do thank the two anonymous reviewers for their comments.
References
Andrews, B. D., Brothers, L. L. & Barnhardt, W. A. (2010): Automated feature extraction and spatial
organization of seafloor pockmarks, Belfast Bay, Maine, USA. – Geomorphology 124: 55–64.
Argialas, D. (1995): Towards Structured Knowledge Models for Landform Representation. – Z. Geo-
morph. N.F Suppl. 101: 85–108.
Argialas, D. & Tzotsos, A. (2003): Geomorphological Feature Extraction from a Digital Elevation Model
Through Fuzzy Knowledge-based Classification. – Proceedings of SPIE International Conference on
Remote Sensing 4886: 516–527.
Baatz, M. & Schape, A. (1999): Object Oriented and Multi-Scale Image Analysis in Semantic Networks. –
Proceedings of 2nd International Symposium: Operationalization of Remote Sensing, ITC, Nether-
lands.
Chapple, P. (2008): Automated Detection and Classification in High-resolution Sonar Imagery for Auto-
nomous Underwater Vehicle Operations-Maritime Operations Division. – Defense Science and Tech-
nology Organization, Edinburgh, South Australia, 32 pp.
Finkl, C. W., Benedet, L. & Andrews, J. (2005): Seabed Classification Based on Interpretation of Airborne
Laser Bathymetry in Class II Waters off Southeast Florida. –Second International Coastal Symposium
in Iceland at Höfn the Town of Hornafjördur (5–8 June 2005).
Gleason, A. C. R., Reid, R. P. & Kellison, G. T. (2008): Single-beam acoustic remote sensing for coral
reef mapping. – Proceedings of the 11th International Coral Reef Symposium, Ft. Lauderdale, Session
number 17, Florida.
Lanier, A. S. (2006): A Comparison of Seafloor Sonar Classification Methods Through the use of Error
Matrices and 3 Dimensional GIS Visualization: A Multibeam Sonar Investigation of Nehalem Bank,
Oregon. – Oregon State University, Oregon U.S.A., 120 pp.
Lundblad, E., Wright, D. J., Miller, J., Larkin, E. M., Rinehart, R., Naar, D. F., Donahue, B. T.,
Anderson, S. M. & Battista, T. (2006): A Benthic Terrain Classification Scheme for American
Samoa. – Marine Geodesy 29 (2): 89–111.
Miliaresis, G. C. & Argialas, D. P. (1999): Segmentation of Physiographic Features from the Global
Digital Elevation Model / GTOPO30. – Computers & Geosciences 25: 715–728.
Penrose, J. D., Siwabessy, P. J. W., Gavrilov, A., Parnum, I., Hamilton, L. J., Bickers, A., Brooke, B.,
Ryan, D. A. & Kennedy, P. (2005): Acoustic Techniques for Seabed Classification. – Cooperative
Research Centre for Coastal Zone Estuary and Waterway Management. – Technical Report 32: 130 pp..
Prior, D. B. &Hooper, J. R. (1998): Sea floor engineering geomorphology: recent achievements and future
directions. – Geomorphology 31: 411–439, Elsevier.
Szuman, M., Berndt, C., Jacobs, C. & Best, A. (2006): Seabed characterization through a range of high-
resolution acoustic systems – a case study offshore Oman. – Mar. Geophys. Res. 27 (3): 167–180.
Yokoyama, R., Shirasawa, M. & Pike, R. J. (2002): Visualizing topography by Openness: A new application
of image processing to digital elevation models. – Photogramm. Engin. Rem. Sens. 68: 257–265.
Zevenbergen, L. W. & Thorne, C. (1987): Quantitative analysis of land surface topography. – Earth Surf.
Proc. Landf. 12: 47–56.

K. Siakavara, Rural and Surveyor Engineer, 28 Agiou Gerasimou Str., Ilisia 15771 Athens, Hellas (Greece),
e-mail: konstantinasiak@hotmail.com
Prof. D. Argialas, Remote Sensing Laboratory, School of Rural & Surveying Engineering, National Technical
University of Athens, 9 Heroon Polytechniou Str., 15780 Zografou Campus, Athens, Hellas (Greece), e-mail:
argialas@central.ntua.gr
75
Reconstruction
Stuttgart, of modern debris flow activity in the arctic environment
September 2013
Reconstruction of modern debris flow activity in the arctic environment

with the use of dwarf shrubs (south-western Spitsbergen) –
a new dendrochronological approach
Piotr Owczarek, Agnieszka Latocha, Małgorzata Wistuba and Ireneusz Malik
with 13 figures
Abstract. The aim of the paper is to show the potential of tundra dwarf shrubs in dendrochronological
reconstructions of debris flow activity in the High Arctic. The preliminary studies with the use of Salix
polaris and Salix reticulata in dating debris flow activity were conducted on selected study sites in the
south-western Spitsbergen. Wood anatomy characteristics of shrubs, such as: the number of growth rings
in collected specimens, the presence reaction wood and scars were evaluated for their potential in temporal
reconstructions in geomorphology. The age of dwarf shrubs can be determined by counting the number of
growth rings on cross sections of wooden stems. Cross sections were made on the border between branch
systems and roots. The age of shrubs studied indirectly mark the minimal time, which passed since the last
episode of debris flow on the studied surface. Analyses of reaction wood and scars can only support the
dating of debris flows, because these anatomic features can also be developed as results of snow creeping or
animal activities. There is a possibility of dating the periods of intensification in debris flow activity with the
use of dendrochronology and dwarf shrubs, and with the one-year precision. The comparison of the time of
debris flow occurrence with meteorological data, may, in future, allow determining the conditions of debris
flow development in the High Arctic.
Key words: dendrochronology, wood anatomy, dwarf shrubs, Salix polaris, Salix reticulata, debris flows,
talus cones, Spitsbergen, the High Arctic
1 Introduction
Debris cones with large debris flow tracks are one of the most characteristic features of deposi-
tional slopes in the high mountains and polar regions, where scarce or lack of vegetation cover
enhance their morphological expression. Debris cones are often used as valuable sources of
palaeoenvironmental information, based on the interpretation of internal and superficial struc-
ture of the cones (Albjar et al. 1979, Ǻkerman 1984, Hinchcliffe et al. 1998, Curry et al.
2006). Debris flows in the arctic region, together with rock falls, are the most important proc-
esses which shape debris cones and contribute to their build-up (Fig. 1). They are considered to
be triggered by rainfalls during short arctic summers or by fast snow melting (Rapp & Nyberg
1988, Larsson 1982, André 1995).
Field observations indicate that debris flows attain the highest intensity in the first phase of
short arctic summers, but until now there is a lack of methods allowing to precisely dating past
episodes of debris flows occurring in the arctic environment. The previous dating of debris flows
in the High Arctic was based either on their morphometric characteristics, floristic composi-
DOI: 10.1127/0372-8854/2013/S-00145 0372-8854/13/S-00145 $ 5.25
76 Piotr Owczarek et al.
Fig. 1. Location of studied debris flows in the south-western Spitsbergen; A – Gullichsenfjellet debris flows
(G0, G1, G2), B – Jens Erikfjellet debris flow (JE), C – Bratteggdalen debris flow (BD).
Reconstruction of modern debris flow activity in the arctic environment 77
tion of plant communities (relative dating) or lichenometry (absolute dating) using Rhizocarpon
geographicum, which can reach the age of up to 9000 years in the Arctic (Jahn 1976, Rapp &
Nyberg 1981, Larsson 1982, André 1995, Kozłowska & Rączkowska 2002). Activity of this
process in modern and fossil periglacial environment can also be dated by micromorphology. The
microstructures like “tiled” structures, faults, folds and plasmic fabrics are created as a result of
downslope flow movement (Phillips 2006). In different mountain regions of the temperate cli-
matic zone debris flow activity is often investigated by analysing wood anatomy of trees affected
by geomorphic processes, also those growing in the vicinity of the upper tree limit (Baumann &
Kaiser 1999, Gärtner et al. 2003, Stoffel et al. 2005). Age of trees growing on debris flow
tracks gives the information about minimal age of debris flows which occurred on the surfaces
studied in the past (Malik & Owczarek 2009). Debris flows also often kill, wound and tilt trees
growing on debris flow tracks. Dead trees can be dated using traditional dendrochronological
procedure. By this, years of debris flow episodes can be dated indirectly (Schweingruber 1988).
After wounding episodes caused by debris flows, trees gradually heal their trunks/roots. Trees
close wounds by developing soft callus tissue. Obtaining disc samples from wounded trees allows
one to determine the age of scars precisely. Dating debris flows with the use of scarred trees is
done by counting tree rings formed after wounding. Using cores for scar dating is more problem-
atic, so a wedge cut through the overgrowing callus is recommended (Baumann & Kaiser 1999,
Bollschweiler et al. 2007, Arbellay 2010).
Trees sometimes can be tilted by debris flow and as a result they have deformed stems. Exter-
nal deformations of stems affect the anatomical structure of the wood produced. Trees produce
reaction wood and eccentric growth pattern when influenced by mechanical stress and by gravity.
We distinguish two types of reaction wood: compression wood – produced by coniferous trees
and tension wood – formed by deciduous trees. Compression wood is dense and hard, visible as
a dark brown ring composed of thick-walled cells. Tension wood is more difficult to recognize;
its cells are characterized by gelatinous fibres with a very high cellulose and small lignin content
(Schweingruber et al. 2006). They have higher density than normal fibres and are shrunk and
swollen longitudinally. Eccentric growth develops when a tree produces wider rings on one side
of the stem and narrower rings on the other. The pith is displaced as a result of this. The first tree
ring (the oldest one) showing reaction wood or eccentricity provides information on when the
tree had started to react to the impact of debris flow.
In arctic and high-alpine ecosystems, where trees are rare or absent, dendrochronologi-
cal research has been so far limited. The activity of geomorphic processes in the High Arctic,
assessed by means of dendrochronology, have been investigated only in the Svalbard archipelago
(Owczarek 2009), although a large number of dendroecological and dendroclimatological stud-
ies (Warren-Wilson 1964, Kuivinen & Lawson 1982, Shaver 1986, Woodcock & Bardley
1994, Rayback & Henry 2005, Schweingruber & Poschold 2005, Bär et al. 2006, Zaltan
& Gajewski 2006, Au & Tardif 2007) have been carried out on arctic dwarf shrubs – the only
material available for dendrochronological studies in the High Arctic. Considering this, the aim
of presented study is to investigate the possibility to use dwarf shrubs to analyse debris flow activ-
ity in the High Arctic, especially to asses the utility of procedures like: calculation of shrubs age,
dating the reaction wood and scars.
2 Material
2.1 Habits and habitats of dwarf shrubs in the south-western Spitsbergen
Two species of arctic dwarf shrubs, which belong to the Willow family (Salicaceae), were inves-
tigated for their potential in dendrochronological dating of debris flow activity: Salix polaris
and Salix reticulata. We have examined mentioned willow species in populations growing in the
south-western Spitsbergen (the High Arctic) for the presence of wood features such as:
• the number of growth-rings,
• development of reaction wood,
• presence of scars.
Salix polaris (Wahlenb.), commonly known as polar willows (Påhlsson 1985), are decidu-
ous, prostrate, trailing shrubs, usually less than 8 cm tall, which form mats. Only their small,
oval, dark green leaves with fresh branches are visible above the ground level (Fig. 2 B). Wooden
branches and root systems are located underground, in the uppermost part of the permafrost
active layer. Salix polaris shrubs can be met in different locations, both on moraines and dry
screes as well as on wetlands. Salix reticulata (L.) shrubs are commonly called net-leaved wil-
lows (Påhlsson 1985). They are low (8–10 cm) creeping shrubs, which reproduce vegetatively
(Fig. 2 C). In comparison with Salix polaris, Salix reticulata shrubs have larger, oval green leaves.
The branch systems are marginally subterranean. These dwarf shrubs usually do not form mats.
Salix reticulata grow on dry, sunny, debris slopes; they are common on screes.
2.2 Wood anatomy of dwarf shrubs and its application in the mass-movement studies
2.2.1 Growth-rings in Salix polaris and Salix reticulata
Radial growth is a phenomenon widespread among woody plants, but the frequency of species
with clearly visible seasonal rings (developed in consequence of periodical growth) is directly
related to the seasonality of climate (Schweingruber 1996). Clear annual rings are common
among trees and shrubs growing in temperate climates, but growth-ring boundaries have also
been reported in the arctic Salicaceae dwarf shrubs, especially in Salix arctica (Beschel & Webb
1963, Warren-Willson 1964) and Salix alexensis (Zaltan & Gajewski 2006).
Both studied willow species: Salix polaris and Salix reticulata are angiosperms with all wood
features characteristic for deciduous trees. They are semi ring-porous and have well-defined
growth-rings, with boundaries delimited by one or more rows of cells (Owczarek 2009)
(Fig. 3 A, B). Cells are usually smaller in Salix reticulata compared to Salix polaris, thus boundar-
ies of rings are better visible and easier countable in Salix reticulata.
The width of rings distinguished in collected individuals of both species range from relatively
big: 0.8 mm, to extremely small, less than 0.01 mm, usually met in Salix polaris. Discontinuous
growth-rings are very common in analysed specimens (Fig. 3 A). In the arctic environment these
partially absent rings can be linked to the climatic conditions (frost year, physiological drought),
mechanical stress connected with periglacial processes, or partial limitation of growth space of
root and branch system.
The anatomical features of wood within single rings developed by the shrubs studied, together
with the analysis of climate and the course of vegetation period in the south-western Spitsber-
gen, indicate that rings developed by Salix polaris and Salix reticulata are annual rings. Only the
seasonal – annual character of rings developed by the shrubs studied allows using anatomical
features described in the paper in temporal reconstructions of debris flows.
Fig. 2. A – sample collection of Salix polaris from Jens Erikfjellet debris flow track; complete individual of
Salix polaris (B) and Salix reticulata (C) including leaves, branches and root system.
Fig. 3. Wood anatomy features of studied shrubs: A – microtome section of the stem of Salix reticulata
(whole diameter, 46 growth rings) with visible discontinuous growth rings; B – ring boundaries of Salix
polaris composed of two or more cell rows; C – tension wood in Salix polaris (irregular gelatinous fibers
marked by arrows); D – a scar in the Salix reticulata sample developed 9 years ago.
2.2.2 Reaction wood in Salix polaris and Salix reticulata
Reaction wood (tension wood as among all angiosperms) was often visible among collected spec-
imens of Salix polaris and Salix reticulata. Cells in tension wood are characterized by gelatinous
fibers (Schweingruber et al. 2006) and higher density than normal fibers. They are shrunk and
swollen longitudinally (Fig. 3 C). The tension wood is developed on the upper side of tilted, non-
vertical branches and roots, and indicates the change in plant position (Schweingruber 1996).
The presence of reaction wood, if linked with slope surface characteristics, gives the possibility to
reconstruct spatial and temporal patterns of debris flow.
2.2.3 Scars in Salix polaris and Salix reticulata
Scars appear when injuries, of diverse origin, are overgrown by callous tissue produced by the
adjacent remaining cambium around the wound (Fig. 3 D), thus, by counting the number of rings
developed after injury (determining the position of the scar on a cross-section) it is possible to
determine the age of events precisely (Hupp et al. 1987, Stoffel & Bollschweiler 2008). Wounds
in Salix polaris and Salix reticulata were found both in valley bottoms and on steep scree slopes.
The geomorphic factors leading to scar development in the arctic environment are: rock falls, rock
avalanches, and debris flows which collide with exposed roots or branches of dwarf shrubs.
3 Methods
Among debris flow tracks found in the study area we have chosen three, within which plant mate-
rial was most abundant. The collection of samples was carried out during two arctic summers:
2007 and 2008. For the reconstruction of debris flow activity we have applied:
• the age of dwarf shrubs on deposition surfaces,
• event-response dating using ring patterns and wood anatomical changes in the shrubs
affected by debris flows.
A minimum of 10 samples was collected from each debris flow track (Fig. 2 A). Complete
specimens of Salix polaris and Salix reticulata, including their root and branch systems, were col-
lected in the field (Fig. 2 B, C). Each individual was documented by digital photo and sectioned
with GSL 1 sledge microtome. We have taken 15–20 μm cross-sections every 0.5–1.5 cm along
each individual: from 4 to 7 different locations (depending on the shrub length). The cutting
in several sections was necessary for determining the age of shrubs. The oldest part of shrub
is located on the border between the wooden branch system and the root (Schweingruber &
Poschold 2005), which is not visible within the physiognomy of the Salicaceae.
Microtome cross-sections were prepared from the whole stem diameters of selected segments
(Fig. 3). Maximum stem diameters ranged from 0.5–1.1 cm. After staining with 1 % solutions of
safranin and astrablue dyes, digital photographs of the micro-sections were taken for tree-ring
analyses (in search for scars and reaction wood) and measurements (in order to determine the
age of shrubs and their growth pattern, along with ring reductions). Ring widths were measured
along two or three radii using OSM 3.65 and PAST4 software.
The chronologies for Salix polaris and Salix reticulata were constructed for precise dating of
debris flow events. They were built on the basis of samples with minimum numbers of discon-
tinuous rings or scars. The first, the samples were visually cross-dated using skeleton plots. The
next a mean curve for the individual plant was derived on the basis of serial-sectioning curves.
The chronologies were developed using curves from 24 plants. The measurement accuracy and
cross-dating errors were detected through the use of COFECHA (Holmes 1983).
4 Study area
4.1 The Wedel Jarlsberg Land

The study area is located in the Wedel Jarlsberg Land (SW Spitsbergen), on north-western shores
of Hornsund fjord (Fig. 1). This part of the Spitsbergen is built of metamorphosed Precambrain
and Lower Palaeozoician rocks, known under the name Heckla Hoek Succession (Birkenmajer
1990). Longitudinally aligned mountain massifs with elevations of ca. 500–600 m a.s.l., active
coast line and several marine terraces dominate in the landscape. Karczewski et al. (1981) dis-
tinguished 14 raised marine beaches along south-western coast of the Wedel Jarlsberg Land. The
highest terraces (from 100–230 m a.s.l) are connected with the Saalian Glaciation. Three lower-
most (from 4.5–12) are of the Holocene age (Lindner et al. 1991). The slopes of the mountain
ridges are shaped by intensive mass movement processes that affect talus accumulations. Debris
cones with large debris flow tracks are one of the most important element of the relief in the study
area. Three of them were selected for detailed study (Fig. 1 A, B, C).
Southern part of the Spitsbergen has periglacial climate often referred as the “High Arctic
climate” with extremely weak diurnal and strong seasonal patterns (Przybylak 2003). The aver-
age annual air temperature is -4.4 °C, varying from -11,3 °C in January to +4,4 °C in July. The cli-
mate warming was observed in the study area in the last two decades and is typical for the region.
Significant increase of temperature is observed in the beginning of winter season (November –
December), in the spring (April – June) and in August.
The average total precipitation is 433 mm per year (1979–2006). Daily sums of precipitation
can exceed 20 mm, which occur from June to November usually. The extreme daily totals were
observed in August and September (43–58 mm). The long-term pattern of precipitation fluctu-
ates and each season has individual tendency. Growth trend of precipitation include the sum-
mer and autumn seasons. Winter falls indicate short-term tendency. In the Hornsund area both,
records of air temperature and precipitation and the current intensive recession of local glaciers
reflect global warming effects. It may be expected that expanding of the unglaciated areas has
changed the local climate relationships. The impact of warming caused deeper thawing of active
layer of permafrost. The active layer varied in different types of ground reaching maximum depth
1.3 m (Leszkiewicz & Caputa 2004).
Variations of climatic conditions in the Arctic, mainly related to the topography of the area
and the influence of the ocean and glaciers, geology and thickness of snow cover, determine the
development of vegetation. Tundra community vegetation is dominated by low creeping dwarf
shrubs and by various species of mosses, herbs and lichens. The vegetative period is extremely
short. It starts in June and lasts until around the end of August, ranging from 40 to 70 days.
4.2 General characteristics of debris flows in the study area
A total of 81 talus cones were recognized in the study area. They were described in terms of
the following features: morphometric characteristics, debris size and structure, type and extent
of vegetation cover. Further detailed morphometric analyses, involving measurement of depth,
height, length and width of gullies, levées and depositional zone were conducted on talus cones
with the most pronounced debris flow tracks. It provided detailed geomorphologic characteris-
tics of selected sites.
Debris flow features were distinguished on 28 debris cones (ca. 35 % of all cones in the study
area). However, most tracks are old and not active; their furrows are often overgrown or filled
with finer material, probably accumulated during later phases of episodic runoff (Fig. 4 A, B). The
levées and depositional zones are poorly visible in the relief. Landforms which can be connected
with recent debris flow occurrence were found within 13 debris cones. Within most of the debris
cones a few generations of debris flows can be recognized. The exact recognition of the number
of debris flow events, which occurred within a given cone, is however not possible, because older
forms are often destroyed or immersed within the cone material, obscuring their morphological
expression. Moreover, debris flows often use pre-existing gullies of former flows and their effects
are superimposed. Therefore, only a minimal number of debris flows can be estimated.
13 of 81 debris flow tracks have recent activity visible in the relief; they are well preserved in
the landscape and well visible from a distance (Fig. 4 C). The incision of gullies within the debris
cone surface ranges from 0.1 to maximum 1.5 m (Fig. 4 D). In most cases, incision occurs only in
the upper parts of gullies or incised sectors of a gully alternate with stretches where no incision
has taken place. Incision usually does not take place in the lower parts of gullies, even though the
levée benches are well preserved.
5 Results
5.1 Gullichsenfjellet debris flows (G0, G1, G2)
The debris flow tracks are located on the western slope of the Gullichsenfjellet massif (Fig. 1 A,
Fig. 5). The massif is composed of a complex of white and greenish quartzites called Gullichsen-
Fig. 4. Examples of debris flow landforms analysed: A – non-active debris flow tracks (Gullichsenfjellet);
B – non-active debris flow gully filled with finer material and overgrown; C – active, well preserved debris
flow forms (Bratteggdalen); D – fresh debris flow gully with distinct leveés.
fjellet Formation (Birkenmajer 1990). The debris flow tracks cut the talus cone which is 80 m
high and an average slope gradient of ca. 35°. The lower part of the cone is located on a flat raised
marine terrace (12 m a.s.l.). This terrace has varying width and it is the best development on the
Kvartsitt Plain (Fig. 5). The age of this surface was dated at about 10 ka (Lindner et al. 1991).
The debris material is angular and becomes finer at the top and coarser near the margin.
There are three distinct debris flow tracks visible on the cone surface (Fig. 6). The track G0 is
presently shaped by periodic debris flows and used by flowing water. The tracks G1 and G2 are
fossil and are partially vegetated. The length of these two debris flow tracks is about 155 m. Fossil
debris flow tracks have well defined levees, varying from 0.4–1.5 m in height (Fig. 6). Their lower
parts are marked by accumulation forms, fan-like in G1 and finger-like in G2, with particle size
ranging from fine sand to coarse angular blocks of rock.
The samples of Salix polaris and Salix reticulata for dendrochronological analyses were col-
lected from levées, furrows and accumulation forms (Fig. 6). Ring pattern analyses conducted
for individual shrub cross-sections revealed possibility of cross dating of ring series and con-
firmed lack of missing and false rings in studied samples (Fig. 7). The age of dwarf shrubs shows
Fig. 5. The main relief features near the Gullichsenfjellet study site (based on the orthophotomap 1:10 000;
after Kolondra 2002, modified).
the minimum age of disturbance of the cone surface by debris flow event. The sampled dwarf
shrubs started to grow in 1955 (G1) and in 1972 (G2) (Fig. 8). Therefore the debris flow G1 event
must have happened earlier than G2. Additionally wood anatomy features – tension wood and
scars – may show the age of debris flow events of relatively smaller energy without surface-clear-
ing disturbance. These anatomical changes are marked in the plants growing in the bottom of
debris flow tracks. Dating based on tension wood and scars have revealed probable occurrence of
extreme event marked in the 1990s and in the second half of the 1980s (Fig. 8).
Fig. 6. Geomorphic sketch and cross-section of debris flow tracks on a talus cone in the Gullichsenfjellet
study site (G0, G1, G2).
Fig. 7. Results of serial sectioning of Salix polaris – the oldest part of the plant is in cross-section no. 2,
the variability of growth ring within the plant is high. Summary statistics for the plant as derived from the
COFECHA program: *correlation with master chronology, **average mean sensitivity of ring-width series.
Fig. 8. The age of the dwarf shrub samples and the number of probable signals of debris-flow occurrence
obtained from tension wood and scars in the Gullichsenfjellet site (G1, G2).
5.2 Jens Erikfjellet debris flow (JE)
A distinct debris flow track is located on the southern debris slope of the Jens Erikfjellet massif
(576 m a.s.l.) (Fig. 1 B, Fig. 9), which is built of Precambrian black greenstones and carbonate-
chlorite-quartz schists. The average slope angle of the debris cone is c. 30–35°. The marginal part
of the cone contacts the inactive sandur fan (the Elveflya outhwash plane) in the vicinity of the
Werenskiold Glacier terminal moraine (Fig. 9).
The zone of the entire debris flow is around 340 m long and 8–12 m wide (debris flow track),
while the average width of the gully is 4 m, with maximum incision up to 1 m (Fig. 10). The track
has well visible levées, up to 1 m high. The maximum diameter of debris found within the land-
form is 1.2 m; the material is mostly angular in shape. The fan-like accumulation zone is formed
by three superimposed lobes, which can be attributed to different flows episodes. They encroach
onto an old sandur plain, forming a footslope of the mountain range.
The samples of Salix polaris and Salix reticulata were collected from the debris flow track,
levées and lower accumulation forms. All sampled dwarf shrubs have started growing between
1960 and 1985 (Fig. 11 A). It indicates that debris flow event took place before 1960. The debris
flow track may have been disturbed in the 1990s, especially in 1994 and 1996, but without vegeta-
tion clearing. Scars and tension wood appeared during this time (Fig. 11 A).
Fig. 9. The main relief features near the Jens Erikfjellet study site (based on the orthophotomap 1:10 000;
after Kolondra 2002, modified).
Fig. 10. Geomorphic sketch and cross-sections of debris flow track in the Jens Erikfjellet study site (JE; legend:
see Fig. 6).
Fig. 11. The age of dwarf shrub samples and the number of probable signals of debris-flow occurrence
obtained from tension wood and scars in: A – Jens Erikfjellet site (JE), B – Bratteggdalen site (BD).
5.3 Bratteggdalen debris flow (BD)
The debris flow is located on the steep (35–40°), ENE-facing debris slope of the Gullichsenfjellet,
overlooking the Brattegg valley (Fig. 1 C, Fig. 5). The valley is made up of the Middle Proterozoic
amhibolites interbedded with chlorite and biotite schists. The flat middle part of the Brattegg val-
ley is occupied by the Myrktjørna Lake (Fig. 5). This flat marine terrace is raised 80–90 m a.s.l.
and dated at 56±8 ka (Lindner et al. 1991).
The debris flow track is ca. 170 m long and quite narrow, with width varying from 5–10 m in
the upper part of the debris flow zone, reaching the maximum extent (25–30 m in width) in the
accumulation zone, which is elongated and fan-like. The debris flow track has a clear zigzag pat-
tern (Fig. 12). The average gully incision is from 0.5–1 m and becomes shallower in the lowermost
part of the debris flow (ca. 0.5 m), where the gully divides into two separate tracks. The levées are
0.5–1 m high. The maximum size of debris within the track is up to 1.5 m in diameter.
The dwarf shrubs collected from this debris flow track and accumulation forms have started
growing between 1970 and 1986. One individual of Salix polaris is 9 years old. The age of scars
and growth-rings with tension wood allows determining the probable moment, when the debris
flow track was transformed. Similarly to the previous cases, this processes may have took place
in the 1990s (Fig. 11 B).
Fig. 12. Geomorphic sketch and cross-sections of debris flow track in the Bratteggdalen study site (BD; legend:
see Fig. 6).
6 Discussion
The age of dwarf shrubs growing on debris flow tracks enables to estimate the minimal time
passed since the youngest episode of debris flow occurred. The age of dwarf shrubs growing
on individual debris flow tracks is diverse; they have begun to grow between 1956 and 1998
(Fig. 8, 11). However, the number of growth rings in specimens on each single debris flow is
similar. The years in which shrubs started to grow comprise within several-years ranges, e.g. for
Gullichsenfjellet G2: 1973–1984. Even if the period in which dwarf shrubs germinated is long,
as in case of the Bratteggdalen (BD: 1971–1998) most of them began to grow during a few years
(G1: 1956–1961, G2: 1973–1978, JE: 1961–1969, BD: 1971–1975) (Figs. 8, 11). This means that
the age of dwarf shrubs determine the time passed since the youngest episode of debris flow, on
which surface the certain shrubs grow. For a precise dating of debris flow one should add the
time needed for the colonization of the deposition surface to the age of the oldest shrubs. This
time of colonization is dependent on the local site conditions: soil composition, slope inclina-
tion, insolation, snow/water conditions, runoff dynamics, etc. Birks (1980) reports that Salix
starts to colonize ice-cored moraines already after 6–10 years. Ziaja (2006) and Moreau et al.
(2008) delivered similar results about rates of plants succession in the Arctic. The research con-
ducted in the Lovén glacier forefield and Sørkapp Land (Spitsbergen) indicates that vascular
species appeared very quickly after deglaciation, usually earlier than lichens. On the basis of
these results we suppose that analyzed debris flow events took place not later than 10 years
before the appearance of the oldest plant (ca. 60 years ago: G1 and JE; ca. 40–45 years ago: G2
and BD).
Another problem is rejuvenation of debris flow tracks: younger debris flows not necessarily
include the whole track, because of that the age of dwarf shrubs may indicate the occurrence of
several flow episodes within one landform. Similar phenomenon was observed during studies
on debris flows in high and mid mountains of the temperate climatic zone (Gärtner et al. 2003,
Stoffel et al. 2005, Malik & Owczarek 2009).
Results presented in this study, similarly to earlier papers (Schweingruber & Poschold
2005), have shown, that the age of dwarf shrubs (Salix polaris and Salix reticulata) should be
determined by counting the number of growth rings on the border between wooden branch
system and the root. The testing cross sections should also be made 2 cm below and above this
border. In case of dwarf shrubs, as in case of trees, we can use cross-dating and skeleton plots,
which enable to compare data rows from growth rings within several sections of one plant and
to compare data rows between separate shrubs. This procedure should enable to eliminate errors
in determining the age of shrub caused by the presence of false and missing rings. The missing
rings were found in arctic dwarf shrubs growing in Central Spitsbergen (Buchwal 2011). Yet,
our study confirmed that in the south-western Spitsbergen the probability of the appearance of
missing rings is smaller, because of stable temperatures during vegetation period, without severe
decreases of temperature values below the vegetation threshold (Przybylak 2003). Probably
because of that, cross-dating conducted in presented study did not reveal any missing rings in
studied dwarf shrubs.
In studied dwarf shrubs we have identified years in which the reaction wood and scars are
particularly frequent, e.g.: 1994, 1996 (Figs. 8, 11). However, scars and reaction wood may appear
as results of snow sliding within debris flow tracks. Branches and stems of Salix polaris and Salix
reticulata are very flexible and thus prone to tilting of diverse origin. The impact of wind, solifluc-
tion movements and pressure of snow cover may also cause permanent bending of shrubs, which
in consequence results in the development of tension wood by plants. It seems that the trans-
portation of snow can be accompanied by transportation of debris, yet it is difficult to explicitly
link the reaction wood and scars in dwarf shrubs with debris flows. Scars in shrubs may also be
developed as results of treading and chewing by reindeers. A large number of injuries found in
shrubs growing on flat surfaces could have been caused by non-geomorphic factors, like animals,
thus the scar analyses should only include specimens growing in locations with recognized debris
flow morphology.
The geomorphic (connected with debris flows) origin of scars and reaction wood can be
confirmed by the large amounts of precipitation recorded in 1994 and 1996 (Fig. 13). This pre-
cipitation may have intensified the occurrence of debris flows, which may have been recorded in
the anatomy of dwarf shrubs as scars and tension wood. Further studies on shrubs anatomy con-
ducted with regard to the presence of scars and reaction wood will probably allow to determine
the origin of this wood features. In case of trees growing in zones affected by debris flows, it was
possible to distinguish in which part of vegetation period the episode took place and in conse-
quence it was possible to determine what process was responsible for particular scars. Based on
the location of the scar within one annual ring (beginning of the ring formation, earlywood and
latewood layer) we can indicate the snow-melting debris flow in spring or the rainfall-induced
debris flow in summer (Malik & Owczarek 2009). In case of dwarf shrubs short vegetation
period and, in consequence narrow zone of radial growth, make it difficult to date scars and
Fig.13. Precipitation totals in 1980–2007 summer seasons (June–August) in Hornsund compared with
probable signals of debris-flow occurrence obtained from tension wood and scars (marked with arrows).
reaction wood precisely, with accuracy higher than one-year. On the present stage of research
the reaction wood and scars in dwarf shrubs can only support dating of debris flows conducted
mainly by defining the age of shrubs.
Further improvement of the method of shrub dating and further dendrochronological sam-
pling on debris flows in the south-western Spitsbergen should allow distinguishing periods of
intensification in the occurrence of debris flows. Comparison of the time of debris flow episodes
with meteorological data will probably allow describing the conditions of flow appearance. Stud-
ies based on geomorphic and lichenometric methods revealed that debris flows can be corre-
lated with extreme precipitation during summer period (Rapp & Nyberg 1981, Larsson 1982,
Ǻkerman 1984, André 1995) rather than with snow melting in spring time, as was suggested
by previous studies (Jahn 1976). Also the preliminary dendrochronological study indicates that
intense debris flow activity can be correlated with precipitation data for the study area (Fig. 11).
Therefore these are extreme rainfalls that can be assumed as the main triggering factor for debris
flows occurrence. However, it must be remembered that other environmental factors influence
the final effects of precipitation and not all extreme rainfalls are able to induce similar geomor-
phic effects. The occurrence and spatial distribution of debris flows is controlled also by topog-
raphy, mainly slope aspect and inclination, as well as by internal structure of slope covers and
saturation conditions prior to precipitation. It was proved for many areas in Svalbard or northern
Norway, e.g. by Rapp & Nyberg (1981), Ǻkerman (1984), André (1995), using lichenometric
data showed, that intensification of debris flow episodes in Spitsbergen occurred 30–40 years ago.
Research presented in this paper allowed achieving similar results; however the dendrochrono-
logical methods enable to date to occurrence of debris flows in the High Arctic with a one-year
precision.
7 Conclusions
Microscopic analysis of collected dendrochronological samples indicates that dwarf shrub spe-
cies: Salix polaris and Salix reticulata have clearly visible and countable annual rings. The age of
shrubs should be determined by counting the number of growth rings on cross sections on the
border between branch system and roots. For control analyses it is recommended to take testing
samples from cross sections 2 cm below and above this border.
Age of wooden dwarf shrubs (Salix polaris and Salix reticulata) obtained by ring counting
allows to date minimal time passed after debris flow episode. Dwarf shrubs colonize debris flow
tracks and cones several years after debris flow event. Cross-dating conducted for individual
plant specimens should be made in order to eliminate missing and false growth rings.
Scars and reaction wood can be used as indicators of debris flows but they can also occur as
result of snow creeping and animals actions (e.g. reindeers grazing). Therefore scars and reaction
wood can only support and supplement dating of debris flows conducted with the use of shrubs
age.
There is a possibility to indicate the periods of intensification in debris flow activity in the
High Arctic; the dendrochronological dating based on arctic shrubs allows marking the mini-
mal age of landforms with one-year precision. Future comparison of the temporal occurrence of
debris flows with meteorological data may probably allow determining the conditions of debris
flow activation in the High Arctic.
Acknowledgements
The research is supported through the grant no. N N306 601440 “Dendrochronological record
of modern climate change and activity of periglacial processes in the south-western Spitsbergen”
from the Ministry of Science and Higher Education in Poland.
References
Albjar, G., Rehn, J. & Stromquist, L. (1979): Notes on talus formation in different climates. – Geogr. Ann.
61A: 179–185.
André, M. F. (1995): Holocene climate fluctuations and geomorphic impact of extreme events in Svalbard. –
Geogr. Ann. 77A: 241–250.
Arbellay, E., Stoffel, M. & Bollschweiler, M. (2010): Dendrogeomorphic reconstruction of past
debris-flow activity using injured broad-leaved trees. – Earth Surf. Proc. Landf. 35: 399–406.
Au, R. & Tardif, J. (2007): Allometric relationships and dendroecology of dwarf shrub Dryas integrifolia
near Churchil, subarctic Manitoba. – Can. J. Bot. 85: 585–597.
Ǻkerman, J. H. (1984): Notes on talus morphology and processes in Spitsbergen. – Geogr. Ann. 66A:
267–284.
Bauman, F. & Kaiser, K. F. (1999): The Multetta Debris Fan, Eastern Swiss Alps: A 500-year Debris Flow
Chronology. – Arct. Alp. Res. 31: 128–134.
Bär, A., Bräuning, A. & Löffler, J. (2006): Dendroecology of dwarf shrubs in the high mountains of
Norway – A methodological approach. – Dendrochronologia 24: 17–27.
Beschel, R. E. & Webb, D. (1963): Growth ring studies on arctic willows. – In: Muler, F. (ed.): Axel Heiberg
Island, Preliminary Report, McGill University, Montreal: 189–198.
Birkenmajer, K. (1990): Geology of the Hornsund area, Spitsbergen. Explanations to the map 1:75 000
scale. – University of Silesia, Katowice.
Birks, H. J. B. (1980): The present flora and vegetation of the moraines of the Klutlan Glacier, Yukon Terri-
tory, Canada: A study in plant succession. – Quatern. Res. 14 (1): 60–86.
Bollschweiler, M., Stoffel, M., Ehmisch, M. & Monbaron, M. (2007): Reconstructing spatio-
temporal patterns of debris-flow activity using dendrogeomorphological methods. – Geomorphology
87: 37–351.
Buchwal, A. (2011): Arctic shrub chronology development and validation on the example of Salix polaris
Wahlenb. (Central Spitsbergen). – Tree Rings in Archaeology, Climatology and Ecology, May 11–14,
2011, Muséum des Sciences Naturelles, Orléans, France, 6–6.
Curry, A. M., Cleasby, V. & Zukowskyij, P. (2006): Paraglacial response of steep, sediment mantled slopes
to post-Little Ice Age glacier recession in the central Swiss Alps. – J. Quatern. Sci. 21: 211–225.
Gärtner, H., Stoffel, M., Lièvre, I., Conus, D., Grichting, M. & Monbaron, M. (2003): Debris-
flow frequency derived from tree-ring analyses and geomorphic mapping, Valais, Switzerland. – In:
Rickenmann, D. & Chen, C.L. (eds.): Debris Flow Hazards Mitigation: Mechanics, Prediction and
Assessment, Proceedings 3rd International DFHM Conference. – Millpress, Rotterdam (1392 p.):
201–207.
Hinchliffe, S., Ballantyne, C. K. & Walden, J. (1998): The structure and sedimentology of relict talus,
Trotternish, northern Skye, Scotland. – Earth Surf. Proc. Landf. 23 (6): 545–560.
Holmes, R. L. (1983): Computer-assisted quality control in tree-ring dating and measurement. – Tree-Ring
Bull. 43: 69–78.
Hupp, C. R., Osterkamp, W. R. & Thornton, J. L. (1987): Dendrogeomorphic evidence and dating of
debris flows on Mount Sharta, Northern California. – U.S. Geol. Surv. Prof. Pap. 1396-B: 39 pp.
Jahn, A. (1976): Contemporaneous geomorphological processes in Longyeardalen, Vestspitsbergen
(Svalbard). – Biul. Perygl. 26: 253–268.
Karczewski, A., Kostrzewski, A. & Marks, L. (1981): Raised marine terraces in the Hornsund area
(northern part), Spitsbergen. – Pol. Polar Res. 2: 39–50.
Kolondra, L. (2002): Werenskioldbreen, Spitsbergen – orthophotomap 1:10 000. Katowice.
Kozłowska, A. B. & Rączkowska, Z. (2002): Vegetation as a tool in the characterisation of geomorpho-
logical forms and processes: an example from the Abisko Mountains. – Geogr. Ann. 84A: 233–244.
Kuivinen, K. C. & Lawson, M. P. (1982): Dendroclimatic analysis of birch in south Greenland. – Arct. Alp.
Res. 14: 243–250.
Larsson, S. (1982): Geomorphological effects on the slopes of Longyear valley, Spitsbergen, after a heavy
rainstorm in July 1972. – Geogr. Ann. 64A: 105–125.
Leszkiewicz, J. & Caputa, Z. (2004): The thermal condition of the active layer in the permafrost at Hornsund,
Spitsbergen. – Pol. Polar Res. 25 (3–4): 223–239.
Lindner, L., Marks, L., Roszczynko, W. & Semil, J. (1991): Age of raised marine beaches of northern
Hornsund Region, South Spitsbergen. – Pol. Polar Res. 12 (2): 161–182.
Malik, I. & Owczarek, P. (2009): Dendrochronological records of debris flow and avalanche in a mid-
mountain forest zone (Eastern Sudetes – Central Europe). – Geochronometria 34: 57–66.
Moreau, M., Mercier, D., Laffly, D. & Roussel, E. (2008): Impacts of recent paraglacial dynamics on
plant colonization: A case study on Midtre Lovénbreen foreland, Spitsbergen (79 °N). – Geomorphology
95: 48–60.
Owczarek, P. (2009): Dendrogeomorphological potential of Salicaceae from SW Spitsbergen, Svalbard. –
In: Kaczka, R., Malik, I., Owczarek, P., Gärtner, H., Helle, G. & Heinrich, I. (eds.): TRACE –
Tree Rings in Archaeology, Climatology and Ecology, Vol. 7. GFZ Potsdam. – Sci. Tech. Rep. STR
09/03: 181–186.
Påhlsson, L. (1985): List of vegetation types and land forms in the Nordic countries with the plant species
of the vegetation types in Latin, the Nordic languages and English. – Nordic Council of Ministers,
69 pp.
Phillips, E. (2006): Micromorphology of a debris flow deposit: evidence of basal shearing, hydrofracturing,
liquefaction and rotational deformation during emplacement. – Quatern. Sci. Rev. 25: 720–738.
Przybylak, R. (2003): The Climate of the Arctic. – Atmospheric and Oceanographic Sciences Library 26:
288 pp., Kluwer Academic Publishers, Dordrecht, Boston, London.
Rapp, A. & Nyberg, R. (1988): Mass movements, nivation processes and climatic fluctuations in northern
Scandinavian mountains. – Norw. J. Geogr. 42: 245–253.
Rayback, S. A. & Henry, G. H. R. (2005): Dendrochronological potential of the Arctic dwarf-shrub Cassiope
tetragona. – Tree-Ring Res. 61: 43–53.
Schweingruber, F. H. (1988): Tree Rings. Basics and Applications of Dendrochronology. – Kluwer,
Academic press, Dordrecht, Boston, London, 276 pp.
Schweingruber, F. H. (1996): Tree rings and environment dendroecology. – Swiss Federal Institute for
Forest, Snow and Landscape Research, Haupt, 609 pp.
Schweingruber, F. H., Borner, A. & Schulze, A. D. (2006): Atlas of Woody Plant Stems. – Springer-
Verlag, Berlin, Heidelberg, Germany, 229 pp.
Schweingruber, F. H. & Poschold, P. (2005): Growth rings in herbs and schrubs: life span, age determina-
tion and stem anatomy. – Forest Snow and Landscape Res. 79: 415 pp.
Shaver, G. R. (1986): Woody stem production in Alaskan tundra shrubs. – Ecology 56: 40–410.
Stoffel, M. & Bollschweiler, M. (2008): Tree-ring analysis in natural hazards research – an overview.
– Nat. Haz. Earth Syst. Sci. 8: 187–202.
Stoffel, M., Conus, D., Grichting, M., Raetzo, H., Gärtner, H. & Monbaron, M. (2005): 400 Years of
Debris-Flow Activity and Triggering Weather Conditions: Ritigraben, Valais, Switzerland. – Antarctic
and Alpine Res. 37: 387–395.
Warren-Wilson, J. (1964): Annual growth of Salix arctica in the high-Arctic. – Ann. Bot. 28: 71–78.
Woodcock, H. & Bardley, R. S. (1994): Salix arctica (Pall.): its potential for dendroclimatological studies
in the High Arctic. – Dendrochronologia 12: 11–22.
Zaltan, R. & Gajewski, K. (2006): Dendrochronological potential of Salix alaxensis from the Kuujjua River
area, western Canadian Arctic. – Tree-Ring Res. 62: 75–82.
Ziaja, W. (2006): Life expansion in Sørkapp Land, Spitsbergen, under the current climate warming. – Rev.
Environ. Sci. Biotechnol. 5: 187–191.

Piotr Owczarek and Agnieszka Latocha, Institute of Geography and Regional Development, University of
Wroclaw, pl. Uniwersytecki 1, 50-137 Wrocław, Poland; e-mail: piotr.owczarek@uni.wroc.pl, agnieszka.
latocha@uni.wroc.pl.
Małgorzata Wistuba and Ireneusz Malik, Faculty of Earth Sciences, University of Silesia in Katowice,
ul. Będzińska 60, 41-200 Sosnowiec, Poland; e-mail: malgorzatawistuba@gazeta.pl, irekgeo@wp.pl.
97
Shoreline
Stuttgart, displacement
September 2013 and Pineios River diversions in NW Peloponnese
Shoreline displacement and Pineios River diversions in

NW Peloponnese (Greece) as result of the geology,
active tectonics and human activity during the last 100ky
Ioannis Fountoulis †, Spyridon Mavroulis, Emmanuel Vassilakis and
Kyriaki Papadopoulou-Vrynioti
with 7 figures and 1 Table
Abstract. Pineios River is the 3rd longest river in Peloponnese and flows in Kyllini wider area which is located
close to the Hellenic Arc-Trench system. This is one of the most seismically and tectonically active regions
in Greece with a great number of changes in the morphogenetic events taking place during the neotectonic
period, as well as the last 100 ky. Prior to the 18th century A.D., the lower alluvial Pineios River flowed
north of the Kyllini peninsula and into the Ionian Sea southwest of Kotichi Lagoon, but the river now flows
southward into a deltaic swamp and dune region, burying a former lagoon-barrier coastal zone. From this,
it becomes apparent that this river is not monotonous in appearance and therefore is not completely con-
trolled by hydrology and hydraulics. In fact, the lower alluvial Pineios River has reacted to major geological
controls, surface deformation and uplift movements caused by the activity of the recently mapped Pineios
normal fault zone and salt tectonics in Kyllini peninsula resulting in the river flow diversion from north to
south at completely different open sea areas. The effects of the geological, tectonic and neotectonic activity
and the impact of the human presence and influence on the lower Pineios River are presented in this paper
in order to determine the causes of the diversion of the lower (alluvial) Pineios River (NW Peloponnese,
Greece) and shoreline displacements.
Keywords: alluvial rivers, active tectonics, uplift rates, longitudinal river profiles, transverse river valley
profiles, river diversion, sinuosity index, shoreline changes, geoarchaeology
1 Introduction
River responses to active tectonics produce characteristic geomorphological features resulting

from surface deformation in the course of the river. The primary response of rivers to active tec-
tonics is manifested as change in channel slope producing knickpoints, channel depth variations
and convexities or concavities of the longitudinal profile, while secondary changes are reflected
in aggradation/degradation and variations in channel morphology including local widening or
narrowing of channels, variations of levee width or discontinuous levees, local development of
meanders or a braided pattern and any anomalous curve or turn (Holbrook & Schumm 1999,
Burbank & Anderson 2001, Keller & Pinter 2002, Schumm et al. 2002). The aggradation/
degradation process disturbs the existing equilibrium of the discharge-sediment load, which is
further extended in the downstream reaches as tertiary effects (Holbrook & Schumm 1999,
Schumm et al. 2002).
DOI: 10.1127/0372-8854/2013/S-00146 0372-8854/13/S-00146 $ 7.00
98 Ioannis Fountoulis et al.
Deformation can be along faults (shear, normal, or reverse in a downstream sense) result-
ing in the creation of fault blocks that correspond either to grabens or horsts. In general, uplift
regions are characterized by river entrenchment, low width/depth ratio and degradation pro-
cesses, whereas subsidence causes shallow channel, aggradation, frequent flooding and presence
of low-lying area or anomalous ponds, marshy lands and alluvial fills (Ouchi 1985, Holbrook &
Schumm 1999, Marple & Talwani 2000). In addition to local structural features, the entire
valley may be tilted upstream, downstream, or laterally. The actual response to active tectonics,
however, varies from river to river and is a function of the hydrological characteristics of the river
(Gregory & Schumm 1987).
Alluvial rivers flow between banks and on a bed composed of sediments that have been eroded
and deposited by the river (Schumm et al. 2002). Therefore, they are not significantly affected, con-
strained or confined by bedrock or old terraces alluvium. Tectonic activity can significantly con-
trol river patterns and behavior and this is especially true of alluvial rivers. Alluvial rivers response
to vertical deformation in the basin through changes of sediment load, water discharge and varia-
tions of valley floor slope (Schumm 1977, Ouchi 1985) and therefore, in addition to the dramatic
effects when stream channels and terraces are offset along faults (Wallace 1967), deformation of
the valley floor by active tectonics can cause pattern change, aggradation, and degradation.
A great example of alluvial river response to active tectonics is the case of the lower (alluvial)
Pineios River in NW Peloponnese (Greece). Pineios River is the 3rd longest river in the Pelopon-
nese and flows downstream in the Kyllini wider area. The broader Kyllini area, located close to
the Hellenic Arc-Trench system (Fig. 1A), is one of the most active areas in Greece and a great
number of morphogenetic events have taken place during the neotectonic period, as well as the
last 100 ky. It has to be mentioned that in the last 25 years several destructive earthquakes took
place in the broader Kyllini area. The vertical tectonic movements started in the Upper Miocene
after the tangential (thrust) tectonics and continue up to present. This is the initiation of the
drainage development and history of Pineios River (Dufaure 1977) that continues supplying the
basin of NW Peloponnese with sediments up to now.
The Greek historian and geographer Strabo reported in his Geography (1st century A.D. –
Strabo ed. Korais, 1817, VIII, 3, 5, 1) that the delta of Pineios River was situated to the north of
the Kyllini peninsula during the ancient times and the river flowed out into the Ionian Sea south-
west of the Kotichi Lagoon. However, the picture today in the NW Peloponnese is quite different
from the past. Pineios River crosses downstream the study area and flows into the Ionian Sea, south
of the Kyllini peninsula (Fig. 1B). Already published data from drill cores and archaeological studies
(Raphael 1973, 1978, Dufaure & Fouache 1991, Papaconstantinou 1991, Kraft et al. 2005,
Vött 2007, Kontopoulos & Koutsios 2010), topographic maps and literature indicate that the
Pineios River may have shifted north and south of the Kyllini peninsula throughout the Holocene, it
was diverted during the late 18th century A.D. and the current delta has been evolving since then.
The effects of the geological, tectonic and neotectonic structure and the impact of the human
presence and influence on the lower (alluvial) Pineios River are presented in this paper in order
to determine the causes of the diversion of the lower Pineios River and the shoreline displace-
ments of both the former and the present Pineios River delta. The applied methodology consists
of the following: (a) detailed terrain analysis and fieldwork of the northern drainage divide of
Shoreline displacement and Pineios River diversions in NW Peloponnese 99
the western part of the Pineios drainage basin and the Pineios fault zone in 1:5,000 scale, (b)
mapping of the Upper Pleistocene (Tyrrhenian) marine terraces and the Holocene Pineios River
deposits in the wider study area at 1:50,000 scale, (c) study of the neotectonic macrostuctures and
the active fault zones of the NW Peloponnese, (d) estimation of the uplift rates of the identified
neotectonic macrostructures during the last 100 ky, (e) comparison of the current shoreline with
topographic maps, aerial photos and satellite images at different times in the 27-year-period from
1972 to 1999. It is significant to note the important role and effect of the active Pineios normal
fault zone on the lower Pineios River development and history during Holocene, which is char-
acterized as one of the most recently active structures in the NW Peloponnese (Mavroulis 2009,
Mavroulis et al. 2010) and is enlisted in the present study in the analysis and description of the
Pineios River courses and diversions for the first time. All the aforementioned data were evalu-
ated and compared and significant results about shoreline displacements and the lower Pineios
river diversions were outlined and described here in detail.
Fig. 1. (A) The current morpho-neotectonic regime of the Greek territory showing also the distinction of
three sectors with different neotectonic trends: E-W for MNS I, NNW-SSE for MNS II and ENE-WSW for
MNS III (after Mariolakos & Fountoulis 2004). (B) Simplified geological map of Western Peloponnese
showing the study area, the spatial distribution of alpine and the post-alpine formations, the major tectonic
structures (normal and strike-slip faults and fault zones, thrusts and overthrusts) and the main neotectonic
macrostructures of the study area (Pyrgos-Olympia basin and Erymanthos Mts. horst).
2 Geology of the modern Pineios River basin
Pineios drainage basin is located at the Northwestern Peloponnese and covers a total area of
886 km2. The main tributary of the Pineios drainage network is the 7th order lower (alluvial)
Pineios River flowing on the post-alpine formations of the western part of the basin. Then, it
flows into the southern part of Vartholomio, where it forms a bend and flows into the Ionian Sea
(Chelonitis Gulf) south of the Kyllini peninsula.
The geological formations in the wider study area can be divided into two major categories,
alpine and post-alpine. The alpine formations belong to three main geotectonic units, Pindos,
Gavrovo and Ionian and mainly occur in the eastern and the central sector of the study area
(Fig. 1B).
Pindos unit crops out in the eastern sector of the study area and particular in the mountain-
ous chain of Erymanthos that reaches altitudes of 2,200 m. The unit consists of clastic sediments,
pelagic limestones and radiolarites. The western boundary of this unit is well defined by a major
low angle thrust fault, the Pindos thrust (Fig. 1B), representing the most internal structure of
the External Hellenides. The Gavrovo unit is associated with the more western relatively low
relief area. The unit consists of the shallow marine carbonates, which crop out on Skolis Mt., and
the thick flysch, which crops out in the surrounding area (Fig. 1B). Ionian is the least extensive
unit and crops out in the western sector of the study area. It consists of the Triassic evaporates
observed near Kyllini peninsula (Fig. 1B).The post-alpine formations of Pliocene and Quaternary
age occur in the southern and western sector of the study area as a part of the Pyrgos-Olympia
basin (Fig. 1B) and lie unconformably on the alpine basement.
According to geological and neotectonic mapping data from previous works published for
this area (Lekkas et al. 1992, Kamberis et al. 1993), 230Th/238U dating of corals found in the
marine terraces of NW Peloponnese (Stamatopulos et al. 1988) and paleontological findings
and analysis (Athanassiou 2000), the stratigraphical order of the post-alpine formations from
the older to the more recent is the following (Fig. 2): (a) Tyrrhenian marine terraces, (b) the
Holocene alluvium–colluvium including the lower Pineios historic terrace, (c) the Pineios natu-
ral levees and (d) the dunes developed along the coast. These formations are overlying post-
alpine formations of Pliocene-Pleistocene age that form the pre-Tyrrhenian bedrock of the
narrow study area.
3 Neotectonics – active tectonics in the Pineios River basin
3.1 Introduction
NW Peloponnese is one of the most tectonically and seismically active regions of Greece. This is
due to its location on the external part of the present Hellenic orogenic arc and proximity to the
Hellenic Trench that represents the subduction boundary of the African plate under the Euro-
pean one (Fig. 1A). The recorded seismicity levels are possibly the highest in Greece. According
to historical and instrumental records, numerous destructive earthquakes have taken place in the
area since 399 B.C., most, if not all, of which were shallow (h < 20 km) and had high intensities.
Fig. 2. The geological-archeological map of Northwestern Peloponnese showing the distribution of the
alpine and the post-alpine formations, the location of the active Pineios fault zone and the uplift rates of
the identified neotectonic macrostructures during the last 100 ky as well as the shoreline displacements
from Neolithic period to present, the paleo-courses of lower Pineios River during Roman and Ottoman
period, the archeological sites from Helladic to Ottoman period, the location of the lower Pineios River
retaining wall and the present southward course of the lower Pineios River (data from Raphael 1973, 1978,
Kowalczyk & Winter 1979, Mariolakos et al. 1988, Stamatopulos et al. 1988, Dufaure & Fouache
1991, Papaconstantinou 1991, Kraft et al. 2005, Vött 2007, Mavroulis 2009, Mavroulis et al. 2010,
Kontopoulos & Koutsios 2010, Fountoulis et al. 2011 and this study).
The neotectonic structure of NW Peloponnese is characterized by the occurrence of large

neotectonic blocks that correspond either to grabens or horsts bounded by fault zones, which in
some cases are visible or concealed, striking E-W or NNW-SSE (Fig. 1A). These faults create a
complex mosaic with particular neotectonic structural and evolution characteristics in every sin-
gle part of the area (Mariolakos et al. 1985). Major role in the development of the post-alpine
formations was played by the active tectonics during the sedimentation phase, which created
horsts and grabens of smaller order within the main neotectonic macrostructures.
Diapiric movements in the area have caused neotectonic deformation mainly exhibited in the
form of a large, N-S anticline, whose axis coincides with the main outcrop of the evaporites in the
Kyllini peninsula (Underhill 1988, Mariolakos et al. 1991, Lekkas et al. 2000, Maroukian
et al. 2000). This doming is accompanied by significant N-S faulting, has exhumed the alpine
formations that underlie the post-alpine deposits and also led to significant dragging of the beds
of Tyrrhenian marine terraces.
The lower (alluvial) Pineios River development and history takes place in this tectonically
and seismically active area of NW Peloponnese. Especially during Holocene, tectonic activity
was intense and was mainly expressed through brittle deformation (Mariolakos et al. 1991).
The recorded highest seismicity levels in the area as well as the generation of many strong earth-
quakes with Vartholomio (16-10-1988, Mw = 5.8), Pyrgos (26-3-1993, Mw = 5.4), Vartholomio
(2-12-2002, Mw = 5.6) and Andravida (8-6-2008, Mw = 6.4) earthquakes being the most recent
ones, confirm the neotectonic studies, which show that the area is undergoing intense tectonic
deformation.
3.2 Main neotectonic macrostructures
The main neotectonic macrostructures of Pineios River basin are (a) the Erymanthos Mts. horst
and (c) the Pyrgos-Olympia basin (Fig. 1B). These structures result from ongoing vertical move-
ments and are faulted and/or bounded by fault zones.
The alpine tectonic structure includes folds and thrusts normally striking N 10–30° W, which
is the normal strike of fold axes and thrusts in the external Hellenides (Mariolakos et al. 1985,
Mariolakos & Papanikolaou 1987). However, fold axes and thrusts in the western part of the
Erymanthos mountain range trend NE-SW revealing that this block undergoes dextral rotation
due to the existence of normal faults which are characterized by the existence of a sinistral strike-
slip component of movement (Mariolakos et al. 1985, Mariolakos & Papanikolaou 1987).
The Pyrgos-Olympia basin is a large graben structure covering a total area of 1,500 km2. It is
bounded to the north by Skolis Mt., to the northeast by the Erymanthos Mts. horst, to the east by
the Gortynia Mts. horst and to the south by the Lapithas Mt. horst, all of which are built of alpine
formations. The western boundary of this basin is hard to determine due to the prolongation of
the basin to the Ionian Sea. The fault zones that bound these horsts are readily discernible and
form impressive morphological discontinuities. The basin is infilled with post-alpine formations
of Late Miocene-Holocene age, with a maximum thickness of approximately 3,000 m overlaying
uncomformably the well-formed paleo-relief developed on the alpine formations. The neotec-
tonic evolution of the Pyrgos-Olympia basin was not the same in all its extent and differentiated
due to the creation and evolution of smaller size tectonic blocks (Lekkas et al. 1995). These
second order neotectonic structures are considered to be the result of the same stress field despite
the fact that their kinematic evolution has not been uniform. Some of them are strongly related
to the occurrence of evaporates and diapiric phenomena (e.g. Kyllini peninsula). A great number
of faults have been located and studied during neotectonic surveys. Some of these were active in
previous neotectonic periods (Pliocene, Early Pleistocene), whereas others are active structures
of the Holocene (Lekkas et al. 1992, Lekkas et al. 2000, Mavroulis 2009, Mavroulis et al.
2010).
The Pyrgos-Olympia basin is also divided into subbasins filled with Neogene and Quaternary
sediments with thickness varying from less than 500 m northeastwards to more than 4 km south-
eastwards of Kyllini (Kamberis 1987). The minimum thickness of these sediments is observed
in the Amaliada wider area and in the Kyllini peninsula. The areas of maximum thickness are
strictly related to the lower Pineios River sediment loads and transport since the Upper Miocene.
Pineios River paleo-delta was located north of the Kyllini pensinsula with maximum thickness
greater than 2 km and the present delta is located south of the Kyllini pensinsula (Bouka wider
area) with maximum thickness greater than 4 km (Kamberis 1987). Thus, it is concluded that
the study area consists of areas constantly subsiding during the sedimentation phase. Moreover,
the southern area presents higher subsidence rates than the northern one according to the post-
alpine formations thickness and the isobaths presented by Kamberis (1987).
3.3 Active fault zones in the Pineios River basin
The most important fault zone in the lower Pineios River basin is the active Pineios normal
fault zone. It is significant to note that the NE-SW strike-slip seismic fault zone that gave rise
to the Andravida earthquake is also located in the wider study area and more precisely north of
the northern drainage divide of the western part of the Pineios River basin (Mavroulis 2009,
Mavroulis et al. 2010, Vassilakis et al. 2011) (Fig. 1B). Although a clear fault trace was not
visible after Andravida earthquake, many secondary geodynamic phenomena indicating seismic
activity along this fault zone were observed and are described by Mavroulis et al. (2010).
This fault zones combined with other less important faults form several small neotectonic
blocks in the narrow study area including the Gastouni graben (hangingwall of Pineios fault
zone), where the western part of Pineios drainage basin develops, the uplifted area of Varda (foot-
wall of Pineios fault zone) and the Kyllini horst (Fig. 1B).
3.3.1 Pineios fault zone
Pineios fault zone is an active normal fault zone located SSW of Skolis Mt., generally trending
E-W and dipping southwards (Figs. 1B, 2, 3). The footwall consists of a Tyrrhenian marine terrace
overlaying pre-Tyrrhenian bedrock consisting of Pleistocene lacustrine and lagoonal clays with
sand and silt intercalations (Keramidia formation), Pliocene-Pleistocene intercalations of clay,
marl, sand and conglomerate (Vounargo formation) and Pliocene-Pleistocene shallow marine
and fluvial-torrential conglomerates (Peristeri formation) (Lekkas et al. 1992, Kamberis et al.
1993, Lekkas et al. 2000). This post-alpine sequence overlay uncomformably the well-formed
paleo-relief developed on the Mesozoic limestones of the Ionian unit (Figs. 1B, 3) (Kamberis
1987, Kamberis et al. 1993). The hangingwall consists of the aforementioned post-alpine forma-
tions and the recent Pineios River bed deposits (Figs. 1B, 2, 3).
Based on (a) the fact that this fault zone displaces marine terraces whose age ranges from 103
to 209 ky (Stamatopulos et al. 1988), (b) geological criteria including primary fault-generated
landforms such as fault scarps and triangular and trapezoidal facets across active mountain and
hill front, and (c) morphotectonic indices such as (i) the mountain front sinuosity index (Smf )
(Bull & Mcfadden 1977) and (ii) the percent faceting along mountain fronts (Facet %) (WELLS
et al. 1988), it can be deduced that Pineios fault zone is one of the most recently active structures
in NW Peloponnese (Mavroulis 2009, Mavroulis et al. 2010).
Field reconnaissance of the Pineios fault zone wide area has revealed much geomorphic
evidence of active tectonics (Mavroulis 2009, Mavroulis et al. 2010, Fountoulis et al.
2011a, b). The most characteristic diagnostic tectonic landforms associated with the Pineios
fault zone are successive sets of slightly degraded but well-defined and well-preserved triangu-
lar facets along the hillside front of the fault zone (Fig. 3). These tectonic landforms indicate
active faulting and normal displacement along the hill front and reflect rapid recent cumu-
lative hill-front uplift, because only rapid recent uplift can maintain such tectonic signa-
tures in a landscape underlain by porous marine calcareous sandstones of Upper Pleistocene
Fig. 3. (A) Extended part of Pineios fault zone downstream of the modern Pineios artificial dam. In this seg-
ment the footwall consists of a Tyrrhenian marine terrace, while the hangingwall consists of lower Pineios
River recent deposits. The northern drainage divide of the lower Pineios River is clearly defined and formed
by the active Pineios fault zone. (B) Pineios fault zone is characterized by the presence of well-preserved
consecutive triangular facets being developed on Tyrrhenian marine deposits.
(Tyrrhenian) age, marine sands, sandstones and conglomerates of Upper Pleistocene age and
lacustrine and lagoonal clays of Pleistocene age. This rapid recent uplift generally lasts from
103 to 106 years (Cotton 1950, Bull 1987). Taking into account that the Pineios fault zone
displaces Tyrrhenian formations of 103 to 209 ky (age calculated by Stamatopulos et al. 1988),
it is concluded that these periods of high tectonic activity and consequently these triangular
facets are more recent than expected.
Based on the slip-rate-estimation methodology and the relationship between vertical slip
rate and basal facet height developed by DePolo & Anderson (2000), the vertical slip rate
for the Pineios fault zone was calculated as 0.15 mm/yr (Mavroulis, 2009). Based on the
maximum throw of the fault zone (100 m) and the 230Th/238U dating of corals from Tyrrhenian
marine deposits in the footwall of Pineios fault zone [209 ky by Stamatopulos et al. (1988)],
the slip rate for the Pineios fault zone was calculated as 0.48 mm/yr (Mavroulis 2009). The
difference observed among the aforementioned calculated slip rates is due to the different data
used in each calculation. In the first calculation, the data used to define the vertical fault slip
rate is geomorphological and particularly the height of the basal facet, which is the lowermost
facet along a range front. In the second calculation, the data used is geological and particularly
the throw of the fault. Thus, the first value (0.15 mm/yr) refers to a percentage of the total dis-
placement and could be defined as the minimum slip rate, while the second value (0.48 mm/yr)
refers to the total displacement and could be defined as the maximum slip rate of the Pineios
fault zone.
Fault slip rates in the Pyrgos-Olympia basin were also determined as the ratio of the total
geological offset (defined by sediment thickness, height of fault scarp and erosion) to the age
of the fault (biostratigraphical ages) and range from 0.05 to 0.35 mm/yr (Koukouvelas et al.
1996). Particularly, the Vounargo, Prasino, Pyrgos, Varvasaina and Alfioussa fault slip rates
were estimated as 0.35, 0.14, 0.18, 0.05, 0.14 and 0.31 mm/yr respectively. Onshore uplift rates
since Middle Pleistocene have been obtained in Kyparissiakos Gulf (SW Greece) and are up to
0.5 mm/yr (Papanikolaou et al. 2007). Taking into account all the abovementioned data, it is
concluded that the calculated slip rates of Pineios fault zone do not differ substantially from the
slip rates obtained in other neotectonic macrostructures of Peloponnese.
4 New data of the Upper Pleistocene-Holocene active tectonics and the effect on the lower
Pineios River development
Pineios fault zone is the most important factor controlling the lower (alluvial) Pineios River
behavior and morphology since the fault initiation, as will be shown below.
4.1 Pineios fault scarp forming the northern drainage divide of lower Pineios River
First of all, Pineios fault zone is proximal to the lower Pineios River (Figs. 1B, 2, 3, 4, 7) which
flows on the hangingwall and there are no topographic barriers between the river and the fault
zone. The northern drainage divide of the lower Pineios River is clearly defined and formed by
the Pineios fault zone and especially by its impressive fault scarp consisting of primary fault-
generated landforms such as well-preserved consecutive triangular and trapezoidal facets being
developed on Upper Pliocene-Pleistocene formations (Mavroulis 2009, Mavroulis et al. 2010,
Fountoulis et al. 2011a, b) (Fig. 3).
4.2 Asymmetry of the Lower Pineios River network
The activation of Pineios fault zone has caused vertical motion of the resulting blocks accompa-
nied by tilting in a cross-valley direction. Both the footwall and the hangingwall of the Pineios
fault zone generally dip northwestwards. More specifically, the Pleistocene marine formations
of the footwall dip northwestwards at between 4° and 14° and the Upper Pliocene-Pleistocene
marine and lagoonal formations of the hangingwall also dip northwestwards at between 7° and
30° (Mavroulis 2009). The tilting of the fault blocks and the resulted northward migration of
the Lower Pineios River have caused in turn the asymmetry of the drainage network developed
on the hangingwall. Thus, the Lower Pineios River shifts northwards and flows close and parallel
to the northern drainage divide, which is, as previously mentioned, formed and determined by
the Pineios fault zone. This asymmetrical development is clearly reflected in the number, size and
order of the tributaries and subbasins on either side of the river (Fig. 2). North of the 7th order
Lower Pineios River, the 1st and 2nd order drainage basins prevail, while higher order drainage
basins are absent. In contradiction, the 3rd order drainage basins prevail covering an extended
area south of the 7th order Lower Pineios River (Fig. 2).
4.3 Longitudinal profile of Lower Pineios River and transverse Lower Pineios River valley profiles
As it is already known, the paleo-delta of Pineios River was situated to the north of the Kyllini
peninsula until the late 18th century A.D. and the Lower Pineios River flowed out into the Ionian
Sea southwest of the Kotichi Lagoon. Sometime, during the late 18th or early 19th century A.D., the
Lower Pineios River abandoned the northwest course and avulsed to the present-day southwest
course (Kraft et al. 2005). From the observation of the Fig. 4, it is clearly shown that the area
where the lower Pineios River was blocked from flowing to the north is located close to the west-
ern end of Pineios fault zone. In order to assess the impact of the active Pineios fault zone on this
critical area of the Lower Pineios River valley, a longitudinal profile was created.
The longitudinal profile of the Lower Pineios River from the Kentro area (upstream) to the
Bouka area (downstream, in the current Pineios River mouth) (Fig. 4A, B) was obtained from the
available topographic maps of Hellenic Military Geographical Service (HMGS) at scale 1:5,000.
Then, it was divided into three different parts, which are (a) the upstream, (b) the intermediate
and the (c) downstream part respectively (Fig. 4A, B). The upstream (from point K to point M)
part of the profile corresponds to the E-W trending part of the lower Pineios River extended
eastwards from the Kentro area (upstream) and westwards to Markopoulo area (downstream).
The valley floor slope of the upstream part was calculated as 0.36 %. The intermediate (from
point M to point V) part of the profile corresponds to the E-W trending part of the lower Pineios
River extended eastwards from the Markopoulo area (upstream) and westwards to Vartholomio
area (downstream). The valley floor slope of the intermediate part was calculated as 0.10 %. The
Fig. 4. (A) Geological map

of the study area depicting
the location of the trans-
verse profiles (C) and the
riverbed of Pineios River.
(B) Longitudinal profile
of lower Pineios River. (C)
Transverse lower Pineios
River valley profiles.
downstream (from point V to point B) part corresponds to the N-S trending and gently slop-
ing lower course of the lower Pineios River extended northwards from the Vartholomio area
(upstream) and southwards to Bouka area (Pineios River mouth). The valley floor slope of the
downstream part was calculated as 0.056 %.
Significant changes are noted in the valley floor slope from the observation of the longi-
tudinal profile and the valley floor slope values. The valley floor slope of the upstream part of
the longitudinal profile (0.36 %) is higher than the ones calculated in the intermediate and the
downstream part respectively. More precisely, the valley floor slope (0.36 %) of the upstream part
is almost 4-fold higher than the one calculated for the intermediate part (0.10 %) and almost
6-fold higher than the one calculated for the downstream part (0.056 %). Furthermore, the val-
ley floor slope of the intermediate part (0.10 %) is 2-fold higher than the one calculated for the
downstream part.
From the observation of the Fig. 4A it is concluded that the M point, where the first signifi-
cant change of the valley floor slope is situated, is located very close to the western end of the
Pineios fault zone where the lower Pineios River was blocked by the Pineios fault scarp some-
time during the late 18th or early 19th century A.D. resulting in the abandonment of its former
northward courses and channels (Raphael 1973, Kraft et al. 2005). The V point in Fig. 4A,
where the second significant change of the valley floor slope is situated, is located very close to
Vartholomio area, where the characteristic bend of the Lower Pineios River course is located.
As already described, this bend is of tectonic origin, created in the lower western part of the
Gastouni graben and bordered to the west by the uplifted eastern (inland) part of the Kyllini pen-
insula and to the north by the Pineios fault zone.
From the abovementioned data derived from the analysis of the longitudinal profile, it is
concluded that the anomalies observed in the profile correspond to fluvial anomalies due to the
surface deformation of the area.
From the observation of the transverse lower Pineios River valley profiles (Fig. 4C), it is obvi-
ous that the Lower Pineios River flows close and parallel to the northern drainage divide, which
is, as previously mentioned, clearly defined and formed by the Pineios fault zone and especially
by its impressive fault scarp (Figs. 2, 3). It is also shown that there are no topographic barriers
between the river and the Pineios fault zone (Fig. 4C). It is significant to note the warping of the
alluvial deposits of the lower Pineios River valley westwards of the Pineios fault zone.
4.4 Sinuosity index (SI)
Under certain conditions, alluvial rivers tend to evolve as single meandering channels. This
behavior is influenced by tectonic movements, reflected in river channel parameters. Within a
given range of channel gradients, the meander pattern changes as vertical tectonic movements
influence the valley slope. Faults influencing the valley slope will affect the sinuosity of the chan-
nel, since the river tries to keep the channel slope constant in a self- organized manner. A nor-
mal fault down throwing in the upstream direction causes a straightening of the channel whilst
downstream down throwing faults result in increased meandering (Ouchi 1985, Holbrook &
Schumm 1999, Keller & Pinter 2002, Bridge 2005). This process is largely independent of
the river size, once the fluvial system enters the meandering stage. In this way, not only large
rivers are suitable for analysis, but smaller creeks and reaches can also be evaluated, so long as
they are essentially free of human influence. In conclusion, river sinuosity calculations represent
a sensitive tool for recognizing neotectonic activity in low-relief areas, because even the small-
est changes in the topography affect the sinuosity of low gradient rivers (Holbrook & Schumm
1999), providing hints on on-going microtopographic changes.
A quantitative measure of the variation of the meandering pattern is the classic sinuosity
index (SI). The sinuosity index (SI) of modern river channels was defined by Leopold & Wolman
(1957) as the ratio of the thalweg length to the valley length. Brice (1964) proposed a slightly
modified sinuosity index (the ratio between the channel length and the length of the meanderbelt
axis), which has the advantage of allowing for both straight and sinuous meander-belts. In both
papers, an arbitrary value of 1.5 was used to distinguish between low- and high-sinuosity chan-
nels. Although other classification schemes have been proposed (Schumm 1963), this value is
generally accepted. This index may also be applied to smaller segments of the river course, from
which the change in pattern at different reaches of the river can be examined.
The SI values for the lower Pineios River have been calculated here using high resolution dig-
ital elevation models derived from air photos. The calculations were made in the vicinity of the
northern and western tectonic boundary of the Gastouni graben, which is formed by the active
Pineios fault zone and the Kyllini diapiric phenomena respectively and where the lower Pineios
River flows and especially in the three parts of lower Pineios river (KM, MV, VB parts respec-
tively) characterized by different valley floor slope values derived from the longitudinal profile of
lower Pineios presented and described above in detail.
The sinuosity index (SI) values as well as the values of the channel length (S – curvilinear
distance measurement along the centre of the channel) and the valley length (L – horizontal
distance measured in the thalweg of two cross sections in a linear depression between two adja-
cent uplands) are presented in Table 1.
It is reasonable to assume that sinuosity variations along the lower Pineios River are related
to, or strongly coupled with differential tectonic deformation of the study area (differential uplift/
subsidence pattern), as it is concluded from the estimation of the uplift rates of the identified
neotectonic macrostructures during the last 100 ky as well as from the thickness and the iso-
baths of the Neogene and Quaternary formations in the study area (Kamberis 1987). Thus, the
Lower Pineios River responds to on-going faulting that re-activates fault structures by changing
its sinuosity.
Table 1. Sinuosity index (SI) values and valley floor slope values calculated for the lower Pineios River.
Lower Pineios River part S (m) L (m) SI Valley floor slope (%)
KM 13154 11354 1.19 0.36
MV 11206 9175 1.22 0.10
VB 8930 6340 1.40 0.056

4.5 Faulted blocks and uplift rates
As previously mentioned, the footwall of Pineios fault zone consists of Pliocene-Holocene age,
including the Tyrrhenian marine calcareous sandstones, which constitutes an extended marine
terrace. This terrace has a width of almost 7 km with the lower rim at about 30 m a.s.l. and the
inner boundary at the front of Skolis Mt. as high as 140 m a.s.l. and is often covered by a veneer of
reddish sandy and conglomeratic alluvional deposits (Raphael 1973, Stamatopulos et al. 1988,
Dufaure & Fouache 1991, Kraft et al. 2005, Kontopoulos & Koutsios 2010). The age data
of Stamatopulos et al. (1988) indicate Tyrrhenian age for all the sampled layers of three interest-
ing and complete sections in the study area, particularly 103 ky for Psari section at an elevation of
40–45 m, 118 ky for Neapolis section at an elevation of 60–65 m and 209 ky for Aletreika section
at an elevation of 140–145 m located very close to the current northern drainage divide of the
western part of the Pineios basin. Taking into account these datings and the elevation of the sec-
tions, the uplift rates of the footwall of the Pineios fault zone were calculated as 0.39 mm/yr for
Psari area, 0.50 mm/yr for Neapoli area and 0.67 mm/yr for Aletreika area respectively (Figs. 2, 7).
It can be deduced that the maximum uplift rate (0.67 mm/yr) characterizes an area (Aletreika)
located on the footwall side and in close proximity to the fault zone in contrast with the other
areas farther away of the fault zone. The uplift rate of the hangingwall was similarly estimated as
0.19 mm/yr (209 ky at an elevation of 40 m). From these uplift rates, it is clearly showed that the
uplift rate of the hangingwall (0.19 mm/yr) is smaller even than the lowest value of the footwall
uplift rate (0.39 mm/yr) (Figs. 2, 7).
The Kyllini peninsula became an island during the Late Quaternary Tyrrhenian transgres-
sion and this is well documented by the presence of Tyrrhenian marine terraces on the inland
side of the peninsula (Kowalczyk & Winter 1979). Moreover, after the deposition of these
Tyrrhenian terraces, salt tectonics caused strong vertical displacements of these marine beds.
Thus, the uplift rates of Tyrrhenian marine terraces of the inland part of Kyllini peninsula based
on the age [125 ky dated by Mariolakos et al. (1988)] and the elevation (20–60 m) of these
terraces were calculated ranging from 0.16 to 0.48 mm/yr, which are lower than the uplift rate
of Pineios footwall and higher that the uplift rate of Pineios hanging wall (Gastouni graben)
(Figs. 2, 7). It is also significant to note that the northeastern part of the Kyllini peninsula has
higher uplift rate (0.48 mm/yr) than the southeastern part (0.30 mm/yr) (Figs. 2, 7).
From the aforementioned uplift rates of the marine terraces, it is concluded that the study
area consists of faulted blocks characterized by different uplift rates which are outlined in Fig. 7.
These faulted blocks are (Figs. 2, 7): (a) the footwall of Pineios fault zone with uplift rate ranging
from 0.39 to 0.67 mm/yr, (b) the hangingwall of Pineios fault zone with uplift rate estimated as
0.19 mm/yr and (c) the eastern (inland) part of Kyllini peninsula with uplift rate ranging from
0.36 to 0.48 mm/yr. It is very important to note that the minimum uplift rate of the footwall of
the Pineios fault zone is significantly larger than the maximum uplift rate of the eastern (inland)
part of Kyllini peninsula and the Gastouni graben. It is obvious that the hangingwall of the
Pineios fault zone, where the lower Pineios River flows, is uplifted with lower rate in comparison
with the surrounding areas. Hence, the Pineios River naturally gravitates toward the relatively
slower and gentler uplifted area or in other words is forced to flow very close and parallel to
Pineios fault zone and toward the tectonic low of the study area, which is the Gastouni graben
(Figs. 2, 7).
4.6 Deformation of the fault zone area – Throw of the fault
Furthermore, the deformation of the fault zone area shows that the fault zone was initiated from
the east and the age of fault zone initiation progressively decreases from east to west. An increase
from east to west is also observed in the throw of Pineios fault zone. The maximum throw is
observed in the western part and the minimum in the eastern part of the fault zone. The throw of
the western part of the fault zone was gradually increased until a critical period in time (probably
during the late 18th century A.D.). During this period the northward course of the lower Pineios
River was blocked when the uplift rate of the western part of the footwall of the Pineios fault zone
became substantially greater than the rates of the hanging wall uplift and the increase in the river
bed elevation due to the river deposition. This uplift, even if it was very subtle, was sufficient to
alter the northward course of the lower Pineios River and its tributaries. This diversion appeared
as deflections of Pineios northward paleo-currents toward the tectonic low of the area, which is
the Gastouni graben (hanging wall of Pineios fault zone).
Historical records indicate the occurrence of numerous devastating seismic events in West-
ern Peloponnese during the late 18th or the early 19th century A.D., when strong earthquakes took
place in 1783, 1785, 1804, 1806, 1820 in Achaia and Ilia wider area (Papazachos & Papazachou
1989). While an increase of the seismic activity in number of events and magnitude is noted in
Western Peloponnese during the period in which the lower Pineios River was diverted, this diver-
sion cannot be linked to an already known seismic event.
4.7 Vartholomio’s bend
Another important feature in the lower Pineios River morphology is the Vartholomio’s bend. The
lower Pineios River flows in an E-W general direction from Kentro in the east to Vartholomio
area in the west. In Vartholomio area the river presents a characteristic bend and then flows in
an N-S direction until it flows into Ionian Sea (Chelonitis Gulf) south of the Kyllini peninsula.
Dufaure & Fouache (1991) supported that the Vartholomio’s bend could not have been caused
by a river capture, but is an academical case of fluvial antecedence upon the slightly uplifted mar-
gin of a subsiding area (Elis coastal plain) occurred before the historical times.
This bend is the result of the combination of the Pineios fault blocks motion and the uplift
movement of the Kyllini peninsula. As mentioned above, the hanging wall of the Pineios fault
zone, where the Lower Pineios River flows, is uplifted with lower rate in comparison with the
surrounding areas which are the footwall of Pineios fault zone and the eastern part of Kyllini
peninsula (Figs. 2, 7). Hence, the Vartholomio area can be considered as a part of a wider tectonic
low (Gastouni graben), which is bordered to the west by the uplifted eastern (inland) part of the
Kyllini peninsula and to the north by the Pineios fault zone. The lower Pineios River flows to this
tectonic low and has no other option but to flow southwards.
4.8 Conclusions
From the abovementioned data, it is concluded that the sudden stop of the northward course of
the lower Pineios River in Markopoulo area seems to be the manifestation of these fault blocks
movements. Lateral change of course is a preferred way to adjust to changes in slope of river
channels with low stream power (Gregory & Schumm 1987). It seems likely that lower Pineios
River has responded to the tectonic effect by a sudden change in its course.
There is another significant conclusion to be drawn, and that is that all the Pineios River
anomalous behavior discussed in this paper is located on formations of Pleistocene and Holocene
age, which indicates that much of the deformation of the study area may be as young as Holocene
in age.
5 Human activity as a controlling factor in the Pineios drainage network development – The
lower Pineios River retaining wall
Another worth mentioned but less important controlling factor in the lower Pineios River develop-
ment is the human activity on the study area during the ancient times. The lower Pineios River
retaining wall (Fig. 2) is an excellent example of the human activity and effect on the drainage
development, which supported the natural trend and process of the Lower Pineios River south-
ward diversion during 18th century. The lower Pineios River retaining wall reaching a total length
of 3 km was discovered in a distance of 400 m north of the Ancient theatre and the museum of
Elis (Papaconstantinou 1991). This retaining wall was found partially destroyed and inverted.
It was constructed in order to support and protect the northern banks from the destructive lower
Pineios River action and to reduce the width of the river bed up to 60 m. The western end of the
wall cannot be determined due to the fact that is buried under the river deposits (Fig. 2), while the
eastern is located in the area, where the remains of an ancient bridge connecting lower Pineios
River banks were found (Papaconstantinou 1991). The construction of the wall indicates two
chronological phases, the first during the Hellenistic Period (323–146 B.C.) and the second dur-
ing Roman period (146 B.C. to 330 A.D.) based on the construction material often consisting of
architectonic material from the neighboring ruined buildings and temples of the Agora of ancient
Elis as no stone is available around the area up to now.
6 Paleo-courses and paleogeographic evolution of lower Pineios River
Based on the ancient report of Strabo (Strabo ed. Korais 1817), the Late Quaternary changes
in coastal Elis (Raphael 1973, 1978), the Cenozoic history of the Kyllini peninsula outlined by
Kowalczyk & Winter (1979), the effect of the coastal changes on the archaeological settings
in Elis (Kraft et al. 2005), the relative sea level changes and regional tectonic evolution of the
coastal plain of Elis since the mid-Holocene (Vött 2007) and the Late Holocene record of the
environmental changes from Kotichi lagoon (Kontopoulos & Koutsios 2010), an effort was
made to represent the stages of the Pineios River development and history during the last 125 ky
(Fig. 5). The paleo-delta of Pineios River was developed north of the Kyllini peninsula before and
during Neolithic period. The Kotichi lagoon was formed on Pineios paleo-delta during Neolithic
and Helladic periods and more precisely during the period from 6,880 to 3,810 B.P. (4,870 to
1,800 B.C., Kontopoulos & Koutsios 2010). During this period, Pineios River deposited river
material periodically interrupting the predominant lagoonal facies (Fig. 5B).
Fig. 5. Paleogeographic sketches of Northwestern Peloponnese for (A) Tyrrhenian period (125 ky B.P.)
illustrating the paleocourses of lower Pineios River.
During the Hellenistic and Roman periods and especially from 3,810 to 1,400 B.P. (1,800
B.C. to 610 A.D.) facies changed from lagoonal to marginal and then to full terrestrial in the
Kotichi wider area (Kontopoulos & Koutsios 2010) (Fig. 5C). These facies marked a regres-
sive stage, probably as a result of deltaic progradation and intense erosion during Middle and
Upper Bronze Age when Pineios flowed north of Kyllini peninsula as well as during the Classical-
Roman period (146 B.C. to 330 A.D.), when Pineios flowed directly south of the Kotichi lagoon
forming a levee (Fig. 5C), which is now abandoned, eroded and stands as a low sea cliff (Raphael
1973). Undoubtedly it attained an elevation close to the present level of the sea. If the levee scarp
Fig. 5. Paleogeographic sketches of Northwestern Peloponnese for (B) Neolithic–Helladic period (from
8,500 to 3,100 yrs B.P.) illustrating the paleocourses of lower Pineios River.
is projected seaward, it intersects the level of the sea approximately 1.2 km offshore. But this is
a minimum value that does not take into account the subsidence of the levee, the tectonic uplift
of the region, or the possibility that the river debouched into a lagoon. An examination of the
truncated levee south of Kotichi Lagoon has revealed sherds of Roman age (Raphael 1973). The
pottery was found to a depth of 40 cm beneath the crest of the levee, clearly suggesting that this
channel was active during and after Roman period (Raphael 1973).
During the Byzantine period and especially from 1,680 to 1,400 B.P. (330 to 610 A.D.), Pineios
occupied the channel 5 km south of Kotychi lagoon forming another levee (Fig. 5D) standing
Fig. 5. Paleogeographic sketches of Northwestern Peloponnese for (C) Roman period (from 2,150 to 1,700 yrs
B.P.) illustrating the paleocourses of lower Pineios River.
well above the floodplain at the shoreline and also indicates coastal retreat (Raphael 1973). This
channel is now in the process of filling. Since the only occupancy site noted on this levee is of
Ottoman age, the minimum age of this relict levee is about 200 years B.P.. The levee system from
the Ottoman period reveals that the shore was farther seaward during the Middle Ages and has
retreated since then.
From the Lower Pineios River southward diversion during the late 18th or early 19th century
A.D. to present, the facies in the Pineios paleo-delta area change from fully terrestrial to marginal
river flood deposits. This change suggests the beginning of an eastward coastal retreat and delta
Fig. 5. Paleogeographic sketches of Northwestern Peloponnese for (D) Othoman period (180 yrs B.P.) illus-
trating the paleocourses of lower Pineios River.
degradation as a result of the southward diversion of the lower Pineios River, the limited supply
of fluvial sediments and the coastal erosion by the wave action. During the last two centuries, the
dominant geomorphic processes in the modern delta of Pineios River have been progradation
and aggradation with large volumes of sediment entrained in littoral transport and sand blown
inland into large dune fields, resulting in the formation of the coastal swamps of the Pineios River
delta (Kraft et al. 2005).
Currently, the Lower Pineios River flows southward into a deltaic swamp and dune region,
burying a former lagoon-barrier coastal zone and carrying a large volume of sediment into the
Ionian Sea (Kraft et al. 2005). The paleo-channels and few smaller streams can still be seen in
their courses to the northwest, now dry. The Pineios river paleo-delta front is now undergoing
marine transgression and intense coastal erosion, as is to be expected in a former delta now
essentially starved of new sediment (Kraft et al. 2005) (Fig. 6).
7 Conclusions
The lower alluvial Pineios River has mainly reacted to major geological controls and surface
deformation caused by the active faults and the active salt tectonics of the study area resulting in
the river flow diversion from north to south of the Kyllini peninsula at completely different open
sea areas during historic times.
Fig. 6. (A) The remnants of the Ottoman levee can be distinguished as fresh water can be seen while flowing
into the sea. It seems that since Pineios River was diverted, the transported sediment amount decreased and
the erosion impact on the beach is pretty obvious.
Fig. 6. (B) The area of the harbor of Lechena constructed in 2005. The eastbound coastline has changed dra-
matically by the deposition of coastal sediments and algae. This progradation at this part reached 50 meters.
Fig. 6. (C) Characteristic type of beach under severe and recent erosion.
The Pineios normal fault zone is one of the most recently active structures in NW Pelo-
ponnese and the most important factor controlling the lower (alluvial) Pineios River behavior
and morphology and producing distinctive response. This response include (a) the significant
changes of the lower Pineios River valley floor slope in Markopoulo and Vartholomio areas
resulting the sudden change in the course of the river in Markopoulo area and the abandonment
of its former northward courses and channels as well as the river bend in Vartholomio area and
the southward flow respectively and (b) the sinuosity variations and changes of the meandering
pattern. The differential movements along the active Pineios fault zone have also produced tilting
that caused the asymmetrical development of the western part of Pineios River basin, where the
lower Pineios River flows.
All the Lower Pineios River anomalous behavior discussed in this paper is located on forma-
tions of Pleistocene and Holocene age. This indicates that the deformation of the study area may
be as young as Holocene in age.
Taking into account the uplift rates of blocks (Fig. 7) as well as the thickness and the already
published isobaths of the Neogene and Quaternary formations (Kamberis 1987), it is concluded
that the study area is undergoing intense and differential tectonic deformation which has been con-
tinuous since the Pliocene. The areas located northeast (Pineios paleodelta) and southeast (Pineios
modern delta in Bouka wide area) of Kyllini peninsula are constantly subsiding during the sedimen-
tation phase and strictly related to the Pineios delta and river sediment loads and transport (Fig. 7).
Moreover, the southeastern area presents higher subsidence rates than the northeastern one.
The study area is also divided into three subareas that uplift with different rates. These
subareas are (Fig. 7): (a) the footwall of Pineios fault zone with uplift rates ranging from
0.39 up to 0.67 mm/yr, (b) the hangingwall of Pineios fault zone with uplift rate estimated as
0.19 mm/yr, (c) the eastern (inland) part of Kyllini peninsula with uplift rates ranging from 0.36
up to 0.48 mm/yr. From these uplift rates, it is concluded that the western part of Pineios basin is
corresponding to the subarea with the lowest uplift rate which is the hangingwall of the Pineios
fault zone (Gastouni graben). Furthermore, the diversion of Pineios River from north to south
of Kyllini peninsula during the 18th century was the result of the uplift movements of the north-
ern margin of the Gastouni graben due to the gradually increase of the throw along the western
part of the active Pineios normal fault zone during historic times and the western margin of the
Gastouni graben due to the local diapiric phenomena of the Ionian evaporates in the Kyllini
peninsula. This natural process and trend was also supported by the human activity in the study
area during historic times as it is revealed by human constructions in the area and particularly
the retaining wall of Pineios River. It is clear that the lower Pineios River evolution is basically
affected by active tectonics, local gypsum diapirism related to thick basal Triassic evaporites and
strong local halokinetic uplift. Other controlling factors, such as the Holocene sea level changes
and the construction and operation of the modern artificial dam of Pineios River in the late
1960’s, are of minor importance.
Acknowledgements
The author Professor Ioannis Fountoulis passed away (16/2/2013) before the publication of this
article. He was a dear friend and colleague who left too soon. He will always be an inspiration
Fig. 7. Map showing areas with different uplift rates for the last 100 ky. The study area is divided to three
subareas that uplift (+) with different rates. These subareas are: (a) the footwall of Pineios fault zone with
uplift rates ranging from 0.39 up to 0.67 mm/yr, (b) the hangingwall of Pineios fault zone with uplift rate
0.19 mm/yr, (c) the eastern (inland) part of Kyllini peninsula with uplift rates ranging from 0.36 up to
0.48 mm/yr. The areas of maximum Neogene and Quaternary sediments thickness (data from Kamberis,
1987) located northeast and southeast of Kyllini peninsula are constantly subsiding during the sedimenta-
tion phase and strictly related to the Pineios delta and river sediment loads and transport.
to us and Immeasurably missed. The authors would like to thank Dr. E. Fouache and Dr. E.
Draganits for their valuable comments and their constructive suggestions for improving the
manuscript.
References
Athanassiou, A. (2000): Presence of fossil elephants in the area of Penios valley (NW Peloponnesus,
Greece). – Ann. Géol. Pays Helléniques 38 (C): 63–76.
Brice, J. C. (1964): Channel patterns and terraces of the Loup Rivers in Nebraska. – U.S. Geol. Surv. Profess.
Pap. 422–D, Washington.
Bridge, J. S. (2005): Rivers and Floodplains – Forms, Processes, and Sedimentary Record. – Blackwell,
Oxford, 504 pp.
Bull, W. B. (1977): Tectonic geomorphology of the Mojave Desert. – U.S. Geol. Surv. Contact Rep.
14–08–001–G–394: 188 p., Office of Earthquakes, Volcanoes and Engineering, Menlo Park, Calif..
Bull, W. B. & Mcfadden, L. D. (1977): Tectonic geomorphology north and south of the Garlock Fault,
California. – In: Doehring, D. O. (ed.): Geomorphology in Arid Regions: Annual Binghamton Con-
ference, State University of New York at Binghamton, 115–136.
Burbank, D. W. & Anderson, R. S. (2001): Tectonic geomorphology. – Blackwell Science, Oxford, 274 p.
Cotton, C. A. (1950): Tectonic scarps and fault valleys. – Geol. Soc. Amer. Bull. 61: 717–758.
Depolo, C. M. & Anderson, J. G. (2000): Estimating the slip rates of normal faults in the Great Basin, USA. –
Basin Res. 12: 227–240.
Dufaure J.-J. & Fouache, E. (1991): Le Fleuve Pénée et la plaine d’Elide: une erreur de Strabon ? Pour une
solution géomorphologique du problème. – Symposium sur l’Elide et l’Achaie antiques, Athènes, Mai
1989. – Meletimata 13: 293–303.
Dufaure, J.-J. (1977): Néotectonique et morphogenèse dans une péninsule méditerranéenne: Le Pélopon-
nèse. – Rev. Geogr. Phys. Geol. Dynam. 19: 27–58.
Fountoulis, I., Mavroulis, S. & Vassilakis, E. (2011): Shoreline displacement, drainage diversion in
NW Peloponnese (Greece) as result of the geology, active tectonics and human activity during the last
100 ky. – Geomorphology 2011, Proceedings of the International Conference on Fluvial and coastal
systems in tectonic active areas, Athens, Greece, 1–4 June 2011.
Fountoulis, I., Vassilakis, E., Mavroulis, S., Alexopoulos, J. & Erkeki, A. (2011): Quantification
of river valley major diversion impact at Kyllini coastal area (W. Peloponnesus, Greece) with remote
sensing techniques. – In: Grützner, C. et al. (eds.): 2nd INQUA-IGCP-567 International Workshop on
Active Tectonics, Earthquake Geology, Archaeology and Engineering, Corinth, 46–49.
Gregory, D. I. & Schumm, S. A. (1987): The effect of active tectonics on alluvial river morphology. – In:
Richards, K. (ed.): River-environ. Proc., Inst. Brit. Geographers, Spec. Publ. 18: 41–68, Blackwell,
New York.
Holbrook, J. & Schumm, S. A. (1999): Geomorphic and sedimentary response of rivers to tectonic defor-
mation: a brief review and critique of a tool for recognizing subtle epirogenic deformation in modern
and ancient settings. – Tectonophysics 305: 287–306.
Kamberis, E. (1987): Geology and oil–geologic study of the post alpine sediments of NW Peloponnese. –
PhD Thesis, Nat. Tech. Univ. Athens, Greece.
Kamberis, E., Alexiades, X., Philipe, G., Tsaila-Monopoli, S., Ioakim, X. & Tsapralis, B. (1993): Geo-
logical map of Greece 1:50,000, Amalias sheet, IGME, Athens.
Keller, E. A. & Pinter, N. (2002): Active Tectonics – Earthquakes, Uplift, and Landscape (2nd edition). –
Prentice Hall, London, 362 p.
Kontopoulos, N. & Koutsios, A. (2010): A late Holocene record of environmental changes from Kotihi
lagoon, Elis, Northwest Peloponnesus, Greece. – Quatern. Int. 225 (2): 191–198.
Koukouvelas, I., Mpresiakas, T., Sokos, E. & Doutsos, T. (1996): The tectonic setting and earthquake
ground hazards of the 1993 Pyrgos earthquake, Peloponnese, Greece. – J. Geol. Soc. (London) 153 (1):
39–49.
Kowalczyk, G. & Winter, K. (1979): Outline of the Cenozoic history of the Kyllini peninsula, W. Pelopon-
nese. – In: Symeonides, N., Papanikolaou, D. & Dermitzakis, M.: Field Guide to the Neogene of
Megara – Peloponnesus – Zakynthos, No 34, Dept. Geology and Paleontology, S.A.
Kraft, J. C., Rapp, (Rip), G., Gifford, J. A. & Aschenbrenner, S. E. (2005): Coastal change and Archaeo-
logical Setting in Elis. – Hesperia 74: 1–39.
Lekkas, E., Fountoulis, I. & Papanikolaou, D. (2000): Intensity distribution and neotectonic macro-
structure Pyrgos earthquake data (26 March 1993, Greece). – Nat. Haz. 21: 19–33.
Lekkas, E., Papanikolaou, D. & Fountoulis, I. (1992): Neotectonic Map of Greece, Pyrgos – Tropaia
sheets, scale 1:100.000. – Research project of the University of Athens, Department of Geology, Divi-
sion of Dynamic, Tectonic, Applied Geology, Athens.
Lekkas, E., Papanikolaou, D. & Fountoulis, I. (1995): The Pyrgos earthquake – The geological and
geotechnical conditions of the Pyrgos area (W. Peloponnese, Greece). – XV Congress of the Carpatho–
Balcan Geological Association, Seminar on active faults, Geol. Soc. Greece, 42–46, Athens.
Leopold, L. B. & Wolman, M. G. (1957): River Channel Patterns: Braided, Meandering and Straight. –
USGS, Reston, VA: U.S. Geol. Surv. Prof. Pap. 282-B: 39–45.
Mariolakos, I. & Fountoulis, I. (2004): The current geodynamic regime in the Hellenic area. – In:
Mariolakos, I., Zagorchev, I., Fountoulis, I. & Ivanov, M. (eds.): Neotectonic transect Moesia
Apulia, field trip guide book – B26, 32nd Int. Geol. Congr., precongr. field trip B26: 12–16 & B26 Greek
route – appendix: 25 p.
Mariolakos, I., Fountoulis, I., Logos, E. & Lozios, S. (1991): Methods to study the torsional neotectonic
deformation: the case of Kalamata area (SW Peloponnesus, Greece). – In: Qingxuan, Ch. (ed.): Proc.
4th meeting IGCP project No 250, Athens-Kalamata, Sept. 30 – Oct. 8, 1989, Chinese Acad. Geol. Sci.
(CAGS) 3: 15–21, Beijing (China).
Mariolakos, I., Lekkas, E., Danamos, G., Logos, E., Fountoulis, I. & Adamopoulou, E. (1988):
Geological-Tectonic Study of the Earthquake affected areas in Elis Prefecture (Kyllini Peninsula). –
Research project of the University of Athens, Department of Geology, Division of Dynamic, Tectonic,
Applied Geology, 108 p., Athens 1989 (in Greek).
Mariolakos, I., Papanikolaou, D. & Lagios, E. (1985): A Neotectonic Geodynamic Model of Pelopon-
nesus Based on Morphotectonics, Repeated Gravity Measurements and Seismicity. – Geol. Jb. B 50:
3–17.
Mariolakos. I. & Papanikolaou, D. (1987): Deformation type and deformation-seismicity relationship
on the Hellenic Arc. – Bull. Geol. Soc. Greece XIX: 59–76.
Maroukian, H., Gaki-Papanastassiou, K., Papanastassiou, D. & Palyvos, N. (2000): Geomorphologi-
cal observations in the coastal zone of Kyllini peninsula, NW Peloponnesus-Greece, and their relation
to the seismotectonic regime of the area. – J. Coast. Res. 16 (3): 853–863.
Marple, R. T. & Talwani, P. (2000): Evidences for a buried fault system in the Coastal Plain of the Carolinas
and Virginia–Implications for neotectonics in the southeastern United States. – Geol. Soc. Amer. Bull.
112: 200–220.
Mavroulis, S. (2009): Fault activity assessment in NW Peloponnesus – The Andravida Earthquake
(08/06/2008). – MSc Thesis, Inter-University Post-Graduate Studies Programme on “Prevention and
Management of Natural Disasters”, Department of Geology and Geo-Environment of the National and
Kapodistrian University of Athens, and Department of Geoinformatics and Topography of the Techno-
logical Educational Institute of Serres, 622 p.
Mavroulis, S., Fountoulis, I. & Lekkas, E. (2010): Environmental effects caused by the Andravida
(08–06–2008, ML = 6.5, NW Peloponnese, Greece) earthquake. – In: Williams, A., Pinches, G.,
Chin, C., McMorran, T. & Massey, C. (eds.): Geologically Active: 11th IAEG Congress: 451–459,
Taylor & Francis Group, Auckland, New Zealand.
Ouchi, S. (1985): Response of alluvial rivers to slow active tectonic movement. – Geol. Soc. Amer. Bull. 96:
504–515.
Papaconstantinou, E. (1991): Architectonic elements of the ancient Agora of Elis in the retaining wall of
the Pineios River. – First International Symposium on Achaia and Elis in Antiquity, Athens, 329–334.
Papanikolaou, D., Fountoulis, I. & Metaxas, C. (2007): Active faults, deformation rates and Quater-
nary paleogeography at Kyparissiakos Gulf (SW Greece) deduced from onshore and offshore data. –
Quatern. Int. 171–172: 14–30.
Papazachos, V. & Papazachou, K. (1989): The Earthquakes of Greece. – Ziti Publications, Thessaloniki,
Greece, pp. 356 (in Greek).
Raphael, C. N. (1973): Late Quaternary changes in coastal Elis, Greece. – Geograph. Rev. 63: 73–89.
Raphael, C. N. (1978): The Erosional History of the Plain of Elis in the Peloponnesus. – In: Brice, W.C.
(ed.): The environmental History of the Near and Middle East since the last Ice Age, 51–66, Academic
Press, London.
Schumm, S. A. (1963): Sinuosity of alluvial rivers on the Great Plains. – Geol. Soc. Amer. Bull. 74: 1089–
1100.
Schumm, S. A., Dumont, J. F. & Holbrook, J. M. (2002): Active Tectonics and Alluvial Rivers. – Cambridge
University Press.
Stamatopulos, L., Voltaggio, M., Kontopulos, N., Cinque, A. & La Rocca, S. (1988): 230Th/238U dating
of corals from Tyrrhenian marine deposits of Varda area (North-western Peloponnesus), Greece. –
Geogr. Fis. Dinam. Quat. 11: 99–103.
Strabo. ed. Korais, A. (1817): Strabonos Geographicon. Volume II, Book VIII. Paris, 1817.
Underhill, J. (1988): Triassic evaporites and Plio-Quaternary diapirism in western Greece. – J. Geol. Soc.
London 145: 269–282.
Vassilakis, E., Royden, L. & Papanikolaou, D. (2011): Kinematic links between subduction along the
Hellenic trench and extension in the Gulf of Corinth, Greece: A multidisciplinary analysis. – Earth
Planet. Sci. Lett. 303: 108–120.
Vött, A. (2007): Relative sea level changes and regional tectonic evolution of seven coastal areas in NW
Greece since the mid–Holocene. – Quatern. Sci. Rev. 26 (7–8): 894–919.
Wallace, R. E. (1967): Notes on stream channels offset by the San Andreas Fault, southern Coast Ranges,
California. – Stanford Univ. Publ (Geol. Sci.) 11: 6–20.
Wells, S., Bullard, T., Menges, T., Drake, P., Karas, P., Kelson, K., Ritter, J. & Wesling, J. (1988):
Regional variations in tectonic geomorphology along segmented convergent plate boundary, Pacific
Costa Rica. – Geomorphology 1: 239–265.
Address of the authors:

Ioannis Fountoulis, Spyridon Mavroulis, Emmanuel Vassilakis and Kyriaki Papadopoulou-Vrynioti,
National and Kapodistrian University of Athens, Faculty of Geology and Geoenvironment, Depart-
ment of Dynamic Tectonic Applied Geology, Panepistimiopolis Zografou, 157 84 Athens, Greece, e-mail:
evasilak@geol.uoa.gr
125
Exploring the
Stuttgart, September Avyssos-Yali-Strogyli submarine volcanic complex
2013
Exploring the Avyssos-Yali-Strogyli submarine volcanic complex at the

eastern edge of the Aegean Volcanic Arc
Nomikou, P., Bell, K.L.C., Papanikolaou, D., Livanos, I. and Fero Martin, J.
with 12 figures
Abstract. The volcanic centers of Kos, Yali and Nisyros lie at the eastern edge of the Hellenic Volcanic
Arc. Recent swath bathymetric surveys and seismic profiling, conducted by HCMR, led to the discovery
of several submarine volcanic centers and massive underwater volcaniclastic deposits. Further research
aboard the E/V “Nautilus” was conducted at the area in October 2010. Avyssos crater, located northeast
of Strongyli islet, is believed to have been the original location of the massive eruption of Kos ignimbrite
160,000 years ago. Exploration of Avyssos showed that it the seafloor is mostly covered with fine-grained
sediment full with traces of bioturbation. Hydrothermal activity was not evident at any point. Yali and
Strongyli represent Late Pleistocene to Holocene volcanic islands that have developed between the islands
of Kos and Tilos. ROV exploration of the eastern flank of Yali revealed wave-type sediment structures, as
well as linear fractures at various depths. Several smaller craters were also discovered on the northwest
slopes of Strongyli, aligned with ENE-WSW trending fractures with no signs of hydrothermal activity.
Heavy biogenic encrustations cover the volcanic rock outcrops on the flanks of both Yali and Strongyli.
Analysis of recovered samples will provide information about their relationship to the geology of the
nearby islands.
Keywords: Hellenic Volcanic Arc, Nisyros volcano, submarine volcanic morphology, Nautilus cruise, ROV
exploration
1 Introduction
The Aegean Volcanic Arc, including Methana, Milos, Santorini, Kos, Yali and Nisyros has been
formed within a back-arc extensional tectonic regime, in the overall setting of the northeastward
subduction of the Eastern Mediterranean lithosphere beneath the active Hellenic margin of the
European plate (McKenzie 1972, LePichon & Angelier 1979, Papanikolaou 1993, Jackson
1994). Major magmatic activity on the island of Kos has occured since Late Miocene-Pliocene,
whereas one of the largest eruptions in the Eastern Mediterranean, 0.16 Ma ago, is represented
by the Kos ignimbrite (Keller et al. 1990). The products of this eruption covered an area of
more than 3000 km2 (Allen et al. 1999). Nisyros and the surrounding islets of Strongyli, Yali,
Pachia and Pergoussa represent Late Pleistocene-Holocene volcanic islands that have intruded
within an ENE-WSW trending neotectonic graben. The latter has been developed between the
islands of Kos and Tilos, where the Alpine basement rocks form neotectonic horsts (Nomikou &
Papanikolaou 2010b).
Systematic research combining onshore and offshore data has been carried out in the area
around Kos, Nisyros and Tilos islands in the Dodecanese since the late nineties, aiming at a compre-
hensive view of the geodynamic processes in the area (Papanikolaou et al. 1991, Papanikolaou
DOI: 10.1127/0372-8854/2013/S-00147 0372-8854/13/S-00147 $ 3.25
126 Nomikou, P. et al.
et al. 1998, Nomikou & Papanikolaou 2000, Papanikolaou & Nomikou 2001, Nomikou
2004, Pe-Piper et al. 2005, Tibaldi et al. 2008a, Nomikou & Papanikolaou 2011). Recent
multibeam mapping and seismic profiling have revealed several features near Nisyros and
Yali, not previously known, including evidence of an extensive volcanic debris avalanche field
(Tibaldi et al. 2008b) and a number of submarine volcanoes (Nomikou et al. 2004, Nomikou
& Papanikolaou 2010a). The submarine volcanoes are located around Nisyros Volcano. It is
believed that the largest one is a crater NE of Strongyli islet that discharged a large amount of
ignimbrite before 160,000 B.P. (Nomikou 2004). Cruise NA-011 aboard the E/V Nautilus in
October 2010 carried out new exploration of the submarine volcanism in Nisyros Volcanic Field
and investigated the seafloor morphology with high-definition video imaging, using ROV Her-
cules and Argus.
In this paper we report the results of the ROV exploration combined with multibeam data,
providing new information about the main morphological characteristics of the Nisyros subma-
rine volcanic centres.
2 The Study area: Geodynamic setting
The study area lies at the eastern edge of the Aegean Volcanic Arc, developed within the Hel-
lenic arc and trench system (Fig. 1b). The particular morphology and tectonic structures of the
Kos-Nisyros-Tilos volcanic field have been analyzed through a combined onshore/offshore recon-
naissance during the years 1997–2001 (Fig. 1a). The neotectonic structure of the Kos-Nisyros-
Tilos area is a complex tectonic graben between the Kos tectonic horst in the north and the Tilos
tectonic horst in the south. The complexity is produced by the occurrence of the intermediate
Kondellioussa horst, which towards the northeast is substituted by large volumes of recent vol-
canic rocks. It divides the basins on both sides of Nisyros and continues up to the Knidos Penin-
sula in the coastal area of Minor Asia. The volcanic relief is developed both offshore and onshore
and starts from 670 m of depth up to 700 m of elevation. Thus, the overall volcanic relief is approx-
imately 1400 m as shown on the Synthetic Topographic Map given in Fig. 1. The vertical motions
between the neotectonic blocks, as estimated from the throws of the marginal faults, are about
1–2 km. This general subsidence observed at the regional Kos-Tilos graben is equilibrated by the
positive relief produced by the volcanic extrusions that have penetrated till the average depth of
600 m and can be observed over all the marine basins surrounding the Nisyros volcanic field.
The island of Nisyros is composed mostly of Quaternary volcanic rocks represented by
alternating lava flows, pyroclastic layers and viscous lava domes, ranging in age from 160 ka to
25 ka. It forms a truncated cone with a base diameter of 8 km and a central caldera of 4 km in
diameter. The evolution of Nisyros volcano, has been recently subdivided into 5 major stages
(Papanikolaou et al. 1991, Nomikou 2004, Vanderkluysen et al. 2005, Volentik et al.
2005a, b): (1) an underwater volcanic eruption, the products of which were basaltic and andesitic
pillow-lavas that built up the lower volcanic rocks visible on the northern coast near Mandraki.
(2) A 500–700 m high stratovolcano evolved on top of these partially submarine lavas, during a
time interval of more than 100 ka. (3) After several eruptive phases with gas and steam explo-
Exploring the Avyssos-Yali-Strogyli submarine volcanic complex 127
sions, two major rhyodacitic, plinian eruptions emitted huge volumes of pyroclastics covering
the whole island. (4) Subsequently, a major central, vertical collapse of the volcano left a large
caldera (Limburg & Varekamp 1991). (5) The western part of the caldera depression was filled
up, during pre-historical times, by a series of rhyodacitic domes, the highest of which, Profitis
Ilias, reaches 698 m a.s.l.
The first clues for volcanic activity οn Nisyros were derived from Greek mythology. How-
ever, no major eruption has ever been documented in historical times. Only some minor eνents
localised within the caldera during the period 1871–1888 haνe been reported (Gorceix 1873,
1874). The only major destructive earthquake οn the island during the 20th century occurred in
1933, with significant damage, especially to Emborio village, which is situated οn the eastern
Fig. 1. a) Synthetic Topographic Map of Kos-Nisyros-Tilos volcanic field based on onshore and offshore data
(from Nomikou 2004). The study area of the Yali-Strongyli is indicated in the quadrangle, b) the location of
Nisyros Island at the eastern edge of the volcanic arc is shown within the Hellenic arc and trench system.
flanks of the caldera. In 1997, earthquake activity with magnitudes of 4.5 and 5.5 Richter scale
ruined about 30 houses in the capital of Nisyros, Mandraki, and created large landslides within
internal caldera craters (Ioannidis 1998, Nomikou & Papanikolaou 2011).
The Yali volcano is located north of Nisyros Island and exhibits parts of a submerged caldera
which are dissected by a post-caldera N-S fault. The western part of Yali is made of two succes-
sive pumice formations (lower and upper), whereas the eastern part of Yali comprises a rhyolitic
obsidian flow at 24 Ka (Wagner et al. 1976). This is locally covered by pumice deposits, equiva-
lent to the upper pumice of Western Yali. Limestone overlies the pumice succession and forms
an isthmus in the centre of the island between the Yali pumice breccias and the lava (Allen et al.
1999).
Strongyli islet is a steep conical volcano with the base located at 650 m depth and the top at
125 m above the sea level with a diameter of 3.5 km (Nomikou 2004). Conglomerates at 50 m
altitude indicate recent uplift. The entire edifice is made up of gray andesites and minor pyro-
clastics. The age of its formation is unknown, but it must be older than the Yali edifice, since
rhyolitc pumice occurs in several morphological depressions and pockets, most likely deposited
during the Upper Pleistocene (Allen 2001). Northeast of the base of Strongyli volcanic cone
a submarine volcanic structure named Avyssos lies, as indicated by the great depth of its base
around 680 m. The volcanic nature was verified by the analysis of lithoseismic air-gun profiles
(Nomikou & Papanikolaou 2010b), which showed only a few meters of sediments overlying
the volcanic formations. The caldera dimensions are 3 km in the NW-SE direction and 4 km in
the NE-SW direction.
3 Data collection and analysis
The offshore studies, including multibeam bathymetric survey, have been carried out on the
research-vessel R/V AEGAEO of the Hellenic Centre for Marine Research (HCMR), during
three successive cruises conducted within the year 2000 in the area of Kos-Nisyros-Tilos islands.
During the first mission, the area of Nisyros Island and the surrounding small islets have been
mapped using the SEABEAM 1180 (180 kHz) system, which is suitable for depths < 500 m. The
SEABEAM 1180 is a portable, shallow water, compact system integrating transmitter, receiver,
interfaces and power stage within a single unit. The other two missions completed the mapping of
the whole area of Kos-Nisyros-Tilos using the SEABEAM 2120 (20 kHz) system, which is suitable
for depths > 200 m. The SEABEAM 2120 is a relatively new swath system that has been especially
designed to accomplish survey requirements exceeding 6000 m of water-depth, providing a satis-
factory resolution (up to 1° × 1°), without mounting a very large array. By operating the systems
for a total period of 12 days with an average speed of 10 knots, 3500 km2 were covered from very
shallow depths to depths of 2200 m. The resulted bathymetric map was combined with onshore
topographic maps of the islands and the result was a synthetic topographic map of the entire area
both onshore and offshore with a DEM (Fig. 1).
The ROV exploration of submarine volcanism around Nisyros Volcano (Nomikou et al.
2011) took place in October 2010 using the E/V Nautilus, this cruise was a collaborative project
“New Frontiers in Ocean Exploration 2010” between the Graduate School of Oceanography at
the University of Rhode Island (URI-USA), the Faculty of Geology & Geoenvironment of Uni-
versity of Athens (NKUA-GREECE), the Institute for Exploration (IFE-USA), and the Institute
of Geology and Mineral Exploration (IGME-GREECE). The dives of ROVs “Hercules” and
“Argus” were held in order to explore the submarine volcanic centers and active structures previ-
ously detected around Nisyros with the submersible Thetis of HCMR, in 2000. The E/V Nauti-
lus of O.E.T. (Ocean Exploration Trust) is equipped with the ROVs Hercules and Argus, which
are state-of-the-art deep-sea robotic vehicle systems capable of exploring depths up to 4000 m
(Philips et al. 2011). Both vehicles have high-definition video cameras which provide a clear and
high resolution imaging of the seafloor. E/V Nautilus has a high frequency satellite system which
helps in distance learning and research during the journey, through Inner Space Center (ISC)
in URI Graduate School of Oceanography and Exploration Command Consoles (ECCs) located
around the world.
4 Results and Discussion
4.1 Morphological characteristics
The multibeam data have been extensively processed by means of data editing, cleaning of erro-
neous beams, filtering of noise, processing of navigation data and interpolation of missing beams.
The detailed swath bathymetric map of the study area was originally compiled at 1/50,000 scale,
which was greatly reduced for publication, with 12 different colors corresponding to 50 m depth
intervals and with additional isobaths of 10 m (Fig. 2). The map permitted the delineation of
the following three main volcanic features north of Nisyros volcano: (1) Avyssos caldera, NE of
Strongyli volcano, (2) Strongyli Volcanic cone, (3) eastern flanks of Yali Volcano.
Characteristic hyperbolic reflectors were identified on the seismic profiles NE of Strongyli
islet corresponding to volcanic intrusions in the basinal part of Eastern Kos Basin (Fig.1a). This
submarine volcanic structure was named Avyssos because of its great depth (600–700 m) and
the analysis of seismic profiles shows only a few meters of sediments overlying the volcanic for-
mations (Nomikou & Papanikolaou 2000). This locality is believed to correspond to the huge
eruption of the Kos ignimbrite, 160,000 B.P. (Nomikou 2004). The “Avyssos” caldera dimensions
are 3 km in the NW-SE direction and 4 km in the NE-SW direction. A hill of about 1 km length
occurs at the centre of Avyssos with a positive relief of about 60–70 m above the flat sea bottom,
representing subsequent intrusions of volcanic rocks. The stereographic diagramme of Avyssos
was constructed with 2 m isobaths aiming to demonstrate the geometry of the volcanic structure
(Fig. 2b). The peak of the central volcanic dome lies at 610 m depth, rising 66 m above the planar
base of the caldera, which lies at 676 m water depth. The caldera rim lies at about 630 m depth,
forming a submarine circular cliff of 40–50 m of topographic difference.
The Strongyli volcanic cone is located north of Nisyros, emerging from 650 m depth of the
sea-floor with a diameter of 3.5 km and reaching the top of Strongyli islet at 125 m above sea level
(Fig. 2). Its ideal cone-shaped morphology with very steep flanks exhibits a 300m wide crater at
the top. In 2000, submersible dives with THETIS of HCMR at the NW flank of the volcanic cone
revealed an underwater crater with a diameter of several meters and numerous open fractures
across the slope (Papanikolaou & Nomikou 2001). The volcanic cone is separated from Yali
Volcano towards the west by a narrow channel oriented N-S at 360 m depth. The eastern slopes of
Yali volcano reflect an irregular geometry without the presence of the continental shelf and with
longitudinal ridges between water depths of 250 m and 90 m.
The bathymetric map of Yali-Nisyros area was analyzed from the perspective of slope distri-
bution and the results are presented as a slope distribution map (Fig. 3). This map shows the dis-
tribution of slope values within the studied area subdivided into five categories: (1) sub-horizontal
areas of mean morphological slope 0–5 %, (2) low-slope areas of 5–15 %, (3) moderate-slope
areas of 15–25 %, (4) high-slope areas of 15–25 % and (5) very steep slope areas of 35–50 %. The
distribution of slope magnitudes illustrates the homogeneous areas of smooth or intense relief
as well as the zones along which there is an abrupt change of slope which correspond to steep,
abrupt slopes around the volcanic islands.
The slopes of Avyssos caldera are smooth (5–15 %), with a gradual increase of depth from
630 m to the flat basinal part to 670 m of depth. On the contrary, the southwestern slopes of the
caldera towards Strongyli islet are more abrupt with dense isobaths and steep slopes (> 35 %).
Fig. 2. (a) Swath bathymetric map between Strongyli and Yali volcanic islets using 10 m contour intervals.
(Red lines indicate the ROV’s transects, red dashed lines indicate the submarine volcanic features, and,
white dashed lines show the area the inside hills of Avyssos caldera, which have been explored by ROV).
(b) 3-D Map of the submarine caldera of “Avyssos” at depths between 600 m and 700 m, using 2 m isobaths.
Fig. 3. Slope distribution map of Avyssos-Strongyli-Yali area (mind the color scale to the lower right of
map).
The highest slope values (35–50 %) are observed around the external flanks of Strongyli volcano,
shown by dense isobaths, except from the southwestern slopes that reflect a smoother morphol-
ogy. The abrupt slopes of Yali and Strongyli volcano are separated by an elongated channel with
low slope values (5–15 %). The western slopes of Yali gradually reduce from very steep slope areas
of 35–50 % to 25–35 % and kindly down to 15–25 %.
4.2 ROV exploration
The Avyssos crater that had already been discovered from swath mapping and air-gun seismic
profiling was explored in detail during the Nautilus cruise NA011, in 2010 (Fig. 2). The ROV Her-
cules was used to explore the slopes of the crater and the volcanic domes identified on the bathy-
metric map obtained from the multibeam data. The exploration of the Avyssos crater showed that
it is covered with fine-grained sediment full of bioturbation holes and mounds without any evi-
dence of hydrothermal activity (Fig. 4). A few samples were collected from the intra-caldera vol-
canic hills, where the lavas are primarily cemented together with consolidated sediment, forming
a black crust on the surface (Fig. 5). Volcanic rocks were also sampled from the southern part of
the crater, where the very steep lava vertical walls were covered by thick bio-encrustations. Geo-
chronological dating and geochemical analysis of these rocks will indicate their genetic relation
to the “Kos ignimbrite”.
Three transects of Strongyli’s flanks were carried out using the ROV Hercules and Argus dur-
ing Nautilus cruise NA011, in 2010 (Fig. 2). These transects provide important information about
the volcanic stratigraphy and underwater morphology of the volcanic cone. Sea floor observa-
tions indicate that Strongyli volcanic cone is affected by fractures (Fig. 6) oriented at different
directions ranging mainly between NE-SW and E-W. The walls of the fractures consist of heav-
ily cemented bio-encrusted lava-rock. Steep lava walls with evidence of shear, as shown in Fig. 7
where stretching lineation was traced on an outcrop, are observed at the western part of the edi-
fice, partially covered by bio-encrusted sediments. Some other outcrops at the northern slopes
were covered by sediments with numerous small slide scars, but, most of the lava walls are heavily
cemented and encrusted (Fig. 8).
Besides the crater, known from the previous reconnaissance with the submersible Thetis of
HCMR, new craters were discovered at depths around 190–200 m (Fig. 9). These are aligned
in the ENE-WSW direction of the fractures observed along the NW flank of Strongyli volcano,
without any sign of hydrothermal activity. The morphology of the craters region is controlled by
thick pumice layers. Bulk and slurp samples were collected (Fig. 10) together with other samples
from many locations along the ROV transects. The future results from their geochronological
Fig. 4. Fine-grained sediment with

bioturbation holes from megafauna
and mounds without temperature
change in the center of Avyssos cra-
ter at 670 m of water depth.
Fig. 5. Large fracture, 15 m long, at

the northern slope of “Avyssos” cra-
ter with outcrops composed of lava
fragments cemented together by
consolidated sediment.
Fig. 6. Small fracture with en-

crusted lava with white sponges
at 322 m depth at the NW flank
of Strongyli, crossing the flat
sea bottom.
Fig. 7. Very steep vertical lava

wall covered with thick bio-
encrustations at 642 m depth
of the northern flank of Stron-
gyli.
Fig. 8. Steep slope of lava wall

which is heavily cemented and
bio-encrusted at 212 m depth at
the western flank of Strongyli.
The channels on both sides of
the summit indicates transport
sediments.
Fig. 9. Underwater crater ob-

served along the strike of a
fracture zone at 192 m depth
at the northwestern flanks of
Strongyli.
Fig. 10. Suction tube sampling

from the bottom layer of pumice
at 198 m depth in the crater wall
NW of Strongyli.
Fig. 11. Wavy sediment features

at 195 m depth. The dark sedi-
ments occupy the ridges of the
undulations and the light sedi-
ments the troughs.
Fig. 12. View of a wide linear

fracture trending N-S at 159 m
depth covered with heavily en-
crustation.
dating and geochemical analysis will provide insights about the age of the volcanic cone and its
relationship to the nearby Nisyros volcanic complex.
None of the underwater slopes of Yali had been explored prior to the Nautilus cruise NA011
in 2010. The ROV Hercules was used to explore the eastern flank of Yali, identified on the multi-
beam map of the area (Fig. 2). The outcrops of volcanic rocks at the sea bottom are covered by
heavy encrustations or by biogenic and concreted layers. ROV exploration of the eastern flank
revealed spectacular wave type structures in the sediments of the seafloor (ripples) with dark
sediments on the ridges of the undulations and light sediments in the troughs (Fig. 11), indicat-
ing the presence of nearbed current activity. Some linear fractures were also explored at various
depths (Fig. 12). Several samples were collected for dating and geochemical analysis which will
provide us information about the lithology and their relationship to the nearby obsidian flow of
the eastern Yali.
5 Conclusions
The area of the islets around Nisyros is characterized by extended shallow water depths host-
ing a number of volcanic centers. All these volcanoes have been developed within a neotectonic
graben formed by subsidence of the order of 2.5 km between the marginal fault zones of South-
ern Kos and Northern Tilos. The volcanic formations were found at 680 m of depth (Avyssos
in the Eastern Kos Basin and at the eastern base of the Strongyli volcanic cone) up to 700 m of
altitude (Rhyolitic domes of Prophitis Ilias on Nisyros), creating a total volcanic relief of more
than 1400 m which has been produced by geodynamic processes of Upper Pleistocene-Holocene
(Nomikou & Papanikolaou 2010b).
The results of the morphological study and ROV exploration will be useful for the interpreta-
tion of the volcanic stratigraphy and understanding of the underwater morphology of the volca-
nic region and its linkage with the onshore structures. Several geological samples were collected
from many locations along the ROV transects, future analysis of which will provide more infor-
mation about the lithology and their relationship to deposits on the nearby islands.
This recent volcanic morphology demonstrates the intense geodynamic processes of the east-
ern edge of the active Hellenic volcanic arc which can be understood only by the study of the
relief both in the offshore and onshore area.
Acknowledgments
Seabeam mapping was a part of the GEOWARN project (Contract no: IST-1999-12310). The
European Commission is acknowledged for their financial contribution to the project. ROV
exploration was supported by Institute for Exploration (IFE-USA) and the collaborative project
“New Frontiers in the Ocean Exploration”. The officers and the crew of the R/V AEGAEO and
E/V NAUTILUS are gratefully acknowledged for their important and effective contribution to
the field work and sampling. We extend our sincere thanks to the organizing committee of the
15th Joint Geomorphological Meeting for accepting the presentation of this study and reviewing
it for publication.
References
Allen, S. R. (2001): Reconstruction of a major caldera-forming eruption from pyroclastic deposit charac-
teristics: Kos Plateau Tuff, eastern Aegean Sea. – J. Volcanol. Geotherm. Res.105 (1-2): 141–162.
Allen, S. R., Stadelbauer, E. & Keller, J. (1999): Stratigraphy of the Kos Plateau Tuff: Product of a
major Quaternary explosive rhyolitic eruption in the Eastern Aegean, Greece. – Int. J. Earth Sci. 88:
132–156.
Gorceix, M. H. (1873): Sur d’état du volcan de Nisyros au mois de mars 1873. – CR Acad Sci (Paris) 77:
597–601.
Gorceix, M. H. (1874): Phenomenes volcaniques de Nisyros. – C.R. Acad. Sci. (Paris) 78: 444–446.
Jackson, J. (1994): Active tectonics of the Aegean region. – Ann. Rev. Earth. Planet. Sci. 22: 239–271.
Ioannidis, K. (1998): Nisyros Island: observed damages to buildings in Mandraki. – Europ. centre prevent.
and forecast. earthquakes Newslett. 2: 33–35.
Keller, J., Rehren, T. & Stadlbauer, F. (1990): Explosive volcanism in the Hellenic Arc: A summary and
review. – In: Hardy, D. A., Keller, J., Galanopoulos, V. P., Flemming, M. C. & Druitt, T. H. (eds.):
Thera and the Aegean world III, 2: 13–26, Thera Foundation London.
Le Pichon, X. & Angelier, J. (1979): The Hellenic arc and trench system: a key to the evolution of the
Eastern Mediterranean. – Tectonophysics 60: 1–42.
Limburg, E. M. & Varekamp, J. C. (1991): Young pumice deposits on Nisyros, Greece. – Bull. Volcanol. 54:
68–77.
McKenzie, D. P. (1972): Active tectonics of the Mediterranean Region. – Geophys. J. R. Astron. Soc. 30:
109–185.
Nomikou, P. (2004): Geodynamic of Dodecanise islands. Kos and Niyros Volcanic Field. – Phd Thesis,
p. 470, University of Athens, Department of Geology.
Nomikou, P. & Papanikolaou, D. (2000): Active geodynamics at Nisyros, the eastern edge of the aegean
volcanic arc: emphasis on the submarine survey. – Proc. 3rd International Conference on the Geology
of the Eastern Mediterranean, September 1998, 97–103.
Nomikou, P., Papanikolaou, D., Alexandri, S. & Ballas, D. (2004): New insights on the Kos-Nisyros
volcanic field from the morphotectonic analysis of the swath bathymetric map. – Rapp. Comm. Int.
Mer. Medit. 37: 60.
Nomikou, P. & Papanikolaou, D. (2010a): A comparative morphological study of the Kos-Nisyros-Tilos

volcanosedimentary basins. – Bull. Geol. Soc. Greece XLIII/1: 464–474.
Nomikou, P. & Papanikolaou, D. (2010b): The morphotectonic structure of Kos-Nisyros-Tilos volcanic
area basd on onshore and offshore data. – XIX Congress of the Carpathian-Balkan Geological Associa-
tion Thessaloniki, Greece, 23-26 September 2010. – Proceedings Spec. Vol. 99: 557–564.
Nomikou, P. & Papanikolaou, D. (2011): Extension of Active Fault Zones on Nisyros Volcano across the
Yali-Nisyros Channel based on Onshore and Offshore data. – Mar. Geophys. Res., Spec. Iss. on «Sea-
floor Mapping for Geohazard Assessment», DOI 10.1007/s11001-011-9119-z.
Nomikou, P., Bell, K. L. C., Vougioukalakis, G., Livanos, I. & Martin, F. J. (2011): “New Frontiers in
Ocean Exploration” The E/V Nautilus 2010 Field Season. – Oceanography Suppl. 24/1: 27.
Papanikolaou, D. 1993. Geotectonic evolution of the Aegean. – Bull. Geol. Soc. Greece 28/1: 33–48.
Papanikolaou, D., Lekkas, E. & Sakellariou, D. (1991): Geological structure and evolution of the Nisyros
volcano. – Bull. Geol. Soc. Greece 25: 405–419.
Papanikolaou, D. & collaborators (1998): Monitoring of Seismic Activity along the Aegean Volcanic Arc:
with focus on the eastern part of the arc in Kοs and mainly Nisyros Islands. – Newslett. Europ. Centre
Prevent. and Forecast. Earthquakes 2: 10–35.
Papanikolaou, D. & Nomikou, P. (2001): Tectonic structure and volcanic centres at the eastern edge of the
Aegean Volcanic Arc around Nisyros island. – Bull. Geol. Soc. Greece XXXIV/1.
Pe-Piper, G., Piper, D. J. W. & Perissoratis, C. (2005): Neotectonics of the Kos Plateau Tuff eruption of
161 ka, South Aegean Sea. – J. Volcanol. Geotherm. Res. 139: 315–338.
Philips, B., Newman, I. & Gregory, T. (2011): E/V Nautilus Vehicles. – Oceanography Suppl. 24/1: 9,
“New Frontiers in Ocean Exploration”. The E/V Nautilus 2010 Field Season.
Tibaldi, A., Pasquare, F. A., Papanikolaou, D. & Nomikou, P. (2008a): Tectonics of Nisyros Island,
Greece, by field and offshore data, and analogue modeling. – J. Struct. Geol. 30/12: 1489–1506.
Tibaldi, A., Pasquarè, F. A., Papanikolaou, D. & Nomikou, P. (2008b): Discovery of a huge sector col-
lapse at the Nisyros volcano, Greece, by on-land and offshore geological-structural data. – J. Volcanol.
Geotherm. Res. 177: 485–499.
Vanderkluysen, L., Volentik, A. C. M., Principe, C., Hunziker, J. C. & Hernandez, J. (2005): Nisyros
volcanic evolution: the growth of a strato-volcano. – Mem. Geol. (Lausanne) 44: 100–106.
Volentik, A., Principe, C., Vanderkluysen, L. & Hunziker, J. C. (2005a): Explanatory notes on the
“Geological Map of Nisyros Volcano (Greece)”. – In: Hunziker, J. C. & Marini, L. (eds.): The Geology,
Geochemistry and Evolution of Nisyros Volcano (Greece). Implications for the Volcanic Hazards. –
Mém. Géol. (Lausanne) 44: 7–26.
Volentik, A., Vanderkluysen, L., Principe, C. & Hunziker, J. C. (2005b): Stratigraphy of Nisyros vol-
cano (Greece). – In: Hunziker, J. C. & Marini, L. (eds.): The Geology, Geochemistry and Evolution
of Nisyros Volcano (Greece). Implications for the Volcanic Hazards. – Mém. Géol. (Lausanne) 44:
26–67.
Wagner, G. A., Storzer, D. & Keller, J. (1976): Spaltspurendatierungen quartarer Gesteinsglaser aus dem
Mittelmeerraum. – N. Jb. Miner.: 84-94.

Paraskevi Nomikou (corresponding author), Dimitros Papanikolaou and Isidoros Livanos, Kapodistrian
University of Athens, Faculty of Geology and Geoenvironment, Panepistimioupoli Zografou, 15784 Athens,
Greece, e-mail: evinom@geol.uoa.gr
Katherine L. Croff Bell and Julie Fero Martin, Graduate School of Oceanography, University of Rhode
Island, Narragansett, USA.

ZFG Suppl 57 3

Caricato da

Informazioni sul documento

Titolo originale

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

ZFG Suppl 57 3

Caricato da

Copyright:

Formati disponibili

Zeitschrift für Geomorphologie Vol. 57 (2013), Suppl.

Quaternary Geomorphology in Tectonically Active Areas with

© 2013 Gebrüder Borntraeger Verlagsbuchhandlung, Stuttgart, Germany www.borntraeger-cramer.de

Hydrogeomorphological Vulnerability in the Romanian Plain

with 11 figures and 9 tables

Key words: hydrogeomorphological vulnerability, critical values, meanders, channel, geomorphometry,

1 Introduction and objectives

Nr. River L Hizv HGV P*med Ks S Hmed

3 Data and methodology

The analysis is based on several types of data:

4 Analysis and results

4.1.2 The analysis of the river long profiles

max min max min max min max min

4.1.3 The analysis of the drainage system morphometry

Table 3. Synthetic geomorphometric data (Horton-Strahler system).

4.1.4.The morphometric analysis of meanders – the lateral cutting

4.2.1 Multiannual analysis and the identification of threshold values (precipitations)

Fig. 6. Variation of monthly rainfall index Angot – Vedea catchment (1961–2000).

Absolute frequency Cumulative frequency Relative frequency

4.2.2 Values and critical thresholds – heavy precipitation and droughts

4.2.3 The threshold analysis

No. Hidrometric S Hmed Q q (l/ W h

Neajlov/Călugăreni 978 706 586 466 326 225

4.3.1 Extreme hydrological conditions

4.3.2 Critical hydrological thresholds

5 Vulnerability to hydrogeomorphological processes – synthetic map

5.1 Selecting the critical values

From the geomorphometric standpoint, we analyzed the following parameters: hypsometry,

5.2 Diagnosis and map representation

The geomorphometric, climatic and hydrological multi-criteria analysis allowed us to establish

Table 8. The classes of vulnerability to hydrogeomorphological hazards depending on geomorphometric criteria.

Class (score) Vulnerability diagnosis Watershed

6 points stability 1 point/criterion

10.1 – 15 medium Neajlov

15.1 – 20 high Călmăţui

more than 20 very high Vedea

Table 9. Synthetic data for diagnosing the vulnerability to hydrogeomorphological hazards.

Neajlov Mostiştea Călmăţui

Hydrological Number of 40 + 12 48 Lacustrine

Address of the authors:

Morphological analysis and related volcanic features of

Keywords: Kolumbo Volcano, volcanic morphology, ROV dives, swath bathymetry

2 Study area: Geo-tectonic setting

3 Data collection and instrumentation

4.1 Morphological characteristics of Kolumbo Volcanic Chain

4.1.1 Kolumbo main cone

Volcanic features revealed by ROV exploration

4.1.2 Kolumbo main cone

4.1.3 Secondary volcanic cones (NE of Kolumbo)

Fig. 14. Large lobate outcrop at the eastern slope of VC 10.

Fig. 15. Steep southern slopes at VC 18.

Fig. 16. Linear steps in E-W direction at VC 9.

Fig 17. Very steep slope with uneven topography at VC 12.

Dominey-Howes, D. T. M., Papadopoulos, G. A. & Dawson, A. G. (2000): Geological and historical

Addresses of the authors:

Comments on the neotectonics of the coastal zone of western Messenia

with 6 figures and 1 table

Key words: Messenia, neotectonics, OSL dating, Geoarchaeology

Fig. 1. Three-dimesional representation

2 Geomorphic, geologic and archaeological context

Fig. 3. Typical sequence of shallow ma-

3 Luminescence dating procedures

MAT1 0.70±0.09 1.07±0.27 0.56±0.03 0.82±0.03 94.33±18.66 115.04±23.00 73.50±10.07 91.90±12.30