Assessing the Areal Extent of Drought

George TSAKIRIS, Dialecti PANGALOU, Dimitris TIGKAS, Harris

VANGELIS
Lab. of Reclamation Works and Water Resources Management
School of Rural and Surveying Engineering
National Technical University of Athens
9 Iroon Polytechniou, 15780, Athens – Greece
e-mail: water@survey.ntua.gr

Abstract: Drought is a three dimensional natural phenomenon characterised by its severity,

duration and areal extent. The study deals with the areal extent of drought
proposing two ways of its illustration and assessment: Discretised severity maps
and “or more” cumulative curves of the affected area in relation to severity level.
Using a reference period (e.g. year) instead of duration and a severity index such
as the Standardized Precipitation Index (SPI) or the Reconnaissance Drought
Index (RDI), discretised maps show the areas affected by the corresponding
severity of drought whereas the “or more” cumulative curves produce directly the
percentages of areas belonging to each class of drought severity. Finally an
illustrative application of both ways of drought areal extent estimation is
presented for Eastern Crete, an area facing frequent droughts.

Key words: areal extent, drought indices, “or more” curves, Reconnaissance Drought Index,
Standardized Precipitation Index.

1. INTRODUCTION

Drought is a recurring natural phenomenon characterised by its sevirity, duration

and areal extent. It is therefore a three-dimensional phenomenon, which is difficult to
assess. Significant work has been done over the last decades for devising indices,
which represent the severity of drought (Yevjevich, 1967, Yevjevich et al., 1983,
Rossi et al., 1992). Although the severity of drought could be rationally assessed from
the anticipated consequences, drought indices have been widely used due to their
simplicity, which they offer to both scientists and water management professionals.
Some of the widely used drought indices are the PDSI (Palmer, 1965), the Deciles
(Kinninmonth et al., 2000), the SPI (McKee et al., 1993) and many others.
Comprehensive reviews on these indices may be found in specific papers and
publications (Richard and Heim, 2002, Hayes, 2004, Tsakiris et al, 2005, etc.).
Recently a new index was proposed by Tsakiris (2004), the RDI – Reconnaissance
Drought Index – which uses potential evapotranspiration in conjunction with
precipitation as the variables, which the drought assessment is based upon. In its

© Water Resources Management: New Approaches and Technologies, European Water Resources Association,
Chania, Crete - Greece, 14-16 June 2007.
 2

standardized form, in most of the cases, the RDI responds in a similar way as the SPI
and the various thresholds representing the borders of severity classes are the same
(Tsakiris and Vangelis, 2005, Tsakiris et al., 2006).
By fixing the time reference of analysis, the “duration” dimension may be omitted.
If for instance the hydrological year is selected as the reference period (that is from
October to next September for the Mediterranean countries), drought can be assessed
for each year of the historical record based on the selected drought index and its
spatial distribution.

2. OVERVIEW OF DROUGHT INDICES

For the purposes of this paper two drought indices were used: the SPI and the RDI.
The Standardized Precipitation Index (SPI) was developed for the purpose of
defining and monitoring drought based on a single meteorological determinant, the
precipitation (McKee et al., 1993).
The SPI calculation for any location is based on a series of accumulated
precipitation for a fixed time scale of interest (i.e. 1, 3, 6, 9, 12,… months). Such a
series is fitted to a probability distribution, which is then transformed into a normal
distribution so that the mean SPI for the location and desired period is zero (Edwards
and McKee, 1997). Positive SPI values indicate greater than median precipitation, and
negative values indicate less than median precipitation. Because the SPI is
normalised, wetter and drier climates can be represented in the same way, and wet
periods can also be monitored using the SPI.
The RDI (Reconnaissance Drought Index) is based on the ratio between two
aggregated quantities of precipitation and potential evapotranspiration. It appears in
three forms: the initial value ak , the Normalised (RDIn) and the Standardised (RDIst).
For real world applications if RDI is calculated as a general index of meteorological
drought it is advisable to use periods of 3, 6, 9 and 12 months. In its initial
formulation RDI for a 12 month period could be directly compared with the Aridity
Index produced for the area under study. If a12 for a certain year is lower than Aridity
Index calculated according to UNEP (1992) then the area is suffering from drought
during this year. The Standardised RDI (RDIst), behaves similarly to the SPI and
therefore the interpretation of the results is similar since the same thresholds as SPI
can be used. The equasions for calculating RDI are presented in the Appendix.

3. DROUGHT SEVERITY MAPS AND CUMULATIVE “OR

MORE” CURVES

Drought severity maps have been used in many studies during the last decade (Kim
et al., 2002, Tsakiris and Vangelis, 2004, Loukas and Vasiliades, 2004). Based on a
descretisation procedure, the area under study can be divided into a number of
 3

squares, at which the meteorological variables are transferred. Various methods for
this transfer may be used (e.g. the weight of each station is taken as a reciprocal of
square distance between the centre of the square in question and the location of the
meteorological station in operation). Altitude correction should be also introduced if it
is nessessary. Then, the selected drought index is calculated at each square for the
reference period. Using different colours for each severity class a drought severity
map is produced. Drought maps are very important tools for delineating the particular
parts of the area under study, which are affected by drought.
Although drought maps are useful, they exhibit two major drawbacks:
a) They do not assess drought at the river basin level, at which water
resources management is practised.
b) They do not illustrate the areal extent of drought in relation to the
threshold value of “the critical area percentage”.
In order to correct the first drawback an additional drought map can be produced
by a spatial integration, so that drought is expressed at the river basin scale. For this
purpose it is advisable to transfer the values of the meteorological determinants to the
river basin and particularly at the actual mean elevation of the basin. In case the basin
is very large subbasins could be used as territorial units.
As regards the second drawback, this work proposes to use the cumulative “or
more” curves (ogives), also known as “drought severity – areal extent” curves, which
express directly the percentage of the area under drought, which then can be
compared with the critical area percentage. It is customary to compare the areal extent
of drought with a preset “critical area” percentage. These curves can be produced by
plotting the severity of drought (y-axis) versus the percentage of the affected area (x-
axis). The severity of drought is presented by a drought index and the area refers to
that affected by at least the corresponding severity level. This type of graphs can be
used not only for the characterisation of drought and the determination of its areal
extent, but also for comparisons with the critical area percentage (related to severity)
directly. Clearly, more than one thresholds referring to the percentage of critical area
can be used defining different levels of severity. Since each class of drought severity
has a different threshold, it is obvious that various critical area percentages could be
simultaneously adopted for characterising a drought episode in relation to its areal
extent.

4. APPLICATION

For the application of the proposed methodology, the area of Eastern Crete -
Greece (Prefectures of Heraklion and Lassithi) was selected.
Both annual SPI and RDI were calculated, for the period of 30 hydrological years.
Maps for one wet and one dry year (1964-65 and 1969-70 respectively) appear in
Fig.1. As expected SPI and RDI give comparable results. However discripancies
occur due to the fact that RDI is using an additional meteorological determinant
4

(PET) apart from precipitation.

Further in Fig.2 the most dry year (1989-90) during the examined period from
1962-63 to 1991-92 is presented. In Fig.3, the cumulative “or more” curves for both
SPI and RDI are presented for the wet year 1964-65 and the dry year 1969-70. The
boundaries of the drought severity classes [0,-1] normal to dry, [-1, -1.5] moderate
dry, [-1.5, -2] severe dry and [<-2] extreme dry appear also in Fig 3. Focussing on the
dry year 1969-70 it may be observed that various percentages of “critical area” may
be adopted and may be directly compared with the areal extent of each drought level.
If the RDI is considered it can be seen that about 10% of the total area is under
extreme drought, whereas 25% of the area is under extreme or severe drought and
70% is is under at least moderate drought. In other words, 10% of the area is under
extreme drought, 15% under severe drought and 50% of the area is under moderate
drought.
Finally in Fig.4 the cumulative “or more” curves for the most dry year (1989-90)
during the period examined are presented. From Fig.4 it can be concluded that more
than 80% of the area is under extreme drought.

1964 – 65
SPI RDI

> 2.00 > 2.00

1.50 to 1.99 1.50 to 1.99
1.00 to 1.49 1.00 to 1.49
0 10 20 km -0.99 to 0.99 0 10 20 km -0.99 to 0.99
-1.00 to -1.49 -1.00 to -1.49
-1.50 to -1.99 -1.50 to -1.99
< -2.00 < -2.00

1969 – 70
SPI RDI

> 2.00 > 2.00

1.50 to 1.99 1.50 to 1.99
1.00 to 1.49 1.00 to 1.49
0 10 20 km -0.99 to 0.99 0 10 20 km -0.99 to 0.99
-1.00 to -1.49 -1.00 to -1.49
-1.50 to -1.99 -1.50 to -1.99
< -2.00 < -2.00

Figure 1. Drought Severity Maps for Eastern Crete for one wet (1964-65) and one dry
year (1969-70), based on the severity indices SPI and RDI
 5

1989 – 90
SPI RDI

> 2.00 > 2.00

1.50 to 1.99 1.50 to 1.99
1.00 to 1.49 1.00 to 1.49
0 10 20 km -0.99 to 0.99 0 10 20 km -0.99 to 0.99
-1.00 to -1.49 -1.00 to -1.49
-1.50 to -1.99 -1.50 to -1.99
< -2.00 < -2.00

Figure 2. The most dry year (1989-90) during the examined period 1962-63 – 1991-
92 for Eastern Crete based on the SPI and the RDI.

SPI (1964-1965) RDI (1964-1965)

3.00 3.00

RDI
SPI

2.00 2.00

1.00 1.00

0.00 0.00

-1.00 -1.00

-2.00 -2.00

-3.00 -3.00
100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%
Area Area

SPI (1969-1970) RDI (1969-1970)

3.00
3.00
SPI

RDI

2.00 2.00

1.00 1.00

0.00 0.00

-1.00 -1.00

-2.00 -2.00

-3.00 -3.00
100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%
Area Area

Figure 3. The cumulative “or more” curves for both SPI and RDI for the wet year
1964-65 and the dry year 1969-70.
 6

SPI (1989-1990) RDI (1989-1990)

3.00 3.00

RDI
SPI

2.00 2.00

1.00 1.00

0.00 0.00

-1.00 -1.00

-2.00 -2.00

-3.00 -3.00
100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%
Area Area

Figure 4. The most dry year 1989-90 during the period 1962-63 – 1991-92

5. CONCLUDING REMARKS

The proposed methodology showed that Drought Discretised Maps and

Cumulative “or more” curves can be presented to illustrate the severity of drought and
its areal extent.
In particular, the cumulative “or more” curves describe the areal extent of drought
in direct percentage figures, which may be compared with the adopted critical area
percentage for each class of drought severity. Based on this information, a
comprehensive analysis of past droughts can be assessed and sound preparedness
plans can be devised for facing future droughts.

ACKNOWLEDGEMENTS

The paper is based on the research partially funded by the INTERREG IIIB -
MEDOCC Programme of the European Commission within the framework of the
SEDEMED II Project.

REFERENCES

Edwards D.C., McKee T.B.: 1997, Characteristics of 20th century drought in the
United States at multiple time scales. Climatology Report Number 97–2, Colorado
State University, Fort Collins, Colorado.
Hayes M.: 2004, Drought Indices, National Drought Mitigation Centre.
http://www.drought.unl.edu/whatis/indices.htm.
 7

Kim T., Valdes J. B., Aparicio J.: 2002, Frequency and spatial characteristics of
droughts in the Conchos River Basin, Mexico. Water International; 27(3): 420-
430.
Kinninmonth W.R., Voice M.E., Beard G.S., de Hoedt G.C., Mullen C.E.: 2000,
Australian climate services for drought management. Drought: A Global
Assessment. D.A. Wilhite (Ed.), Routledge; 210-222.
Loukas A., Vasiliades L.: 2004, Probabilistic Analysis of Drought Spatiotemporal
Characteristics in Thessaly Region, Greece. Natural Hazards and Earth Systems
Sciences; 4: 719-731.
McKee T.B., Doeskin N.J., Kleist J.: 1993, The relationship of drought frequency and
duration to time scales. Proceedings of the Eighth Conference on Applied
Climatology, Anaheim, CA, January 17-23, 1993. American Meteorological
Society. Boston, MA; pp. 179-184.
Palmer W.C.: 1965, Meteorological drought. Research Paper No. 45. U.S. Weather
Bureau, Washington, DC.
Richard R., Heim J.: 2002, A Review of Twentieth-Century Drought Indices used in
the United States. American Meteorological Society; 1149-1166.
Rossi G., Benedini M., Tsakiris G., Giakoumakis S.: 1992, On the Regional Drought
Estimation and Analysis. Water Resources Management; 6(4): 249-277.
Tsakiris G.: 2004, Meteorological Drought Assessment. Paper prepared for the needs
of the European Research Program MEDROPLAN (Mediterranean Drought
Preparedness and Mitigation Planning), Zaragoza, Spain.
Tsakiris G., Pangalou D., Vangelis H.: 2006, Regional Drought Assessment Based on
the Reconnaissance Drought Index (RDI), Water Resources Management, OnLine
First, DOI: 10.1007/s11269-006-9105-4, Springer Netherlands.
Tsakiris, G., Rossi, G., Iglesias, A., Tsiourtis, N., Garrote, L., Cancelliere, A. : 2005.
Drought Indicators Report. Report written for the needs of MEDROPLAN project.
Tsakiris G., Vangelis H.: 2004, Towards a Drought Watch System based on spatial
SPI. Water Resources Management; 18(1): 1-12.
Tsakiris G., Vangelis H.: 2005, Establishing a Drought Index incorporating
evapotranspiration, European Water; 9-10: 3-11.
UNEP: 1992, World Atlas of Desertification. Edward Arnold, London.
Yevjevich, V.: 1967, An objective approach to definitions and investigations of
continental hydrologic droughts, Hydrology Paper no. 23, Colorado State
University, Fort Collins, CO.
Yevjevich V., Da Cunha L., Vlachos E.: 1983, Coping with droughts, Water
Resources Publications, Littleton, Colorado.