Sei sulla pagina 1di 24

HYDROLOGICAL PROCESSES

Hydrol. Process. 15, 2151– 2174 (2001)


DOI: 10.1002/hyp.281

Effects of pipeflow on hydrological process and its


relation to landslide: a review of pipeflow studies in
forested headwater catchments
Taro Uchida,* Ken’ichirou Kosugi and Takahisa Mizuyama
Laboratory of Erosion Control, Department of Forestry, Faculty of Agriculture, Kyoto University, Kyoto 6068502, Japan

Abstract:
Since the 1980s, several field studies of pipeflow hydrology have been conducted in forested, steep headwater
catchments. However, adequate information is lacking with regard to questions as to how representative these previous
studies are and how widespread the phenomena might be. Thus, the aim of this study is to review some studies of
pipeflow hydrology on forested steep hillslopes. Several points were clarified. (1) The maximum discharge of pipeflow
was mainly determined by the soil pipe diameter. When the condition of both the soil pipe diameter and the slope
gradient in forest soil were similar to those in peaty podzol, the maximum discharge of pipeflow on forest slopes was
slighter than that in peaty podzol. (2) Pipeflow was delivered from a variety of sources, and the contributing area of
pipeflow extended as the soil layer became wetter. Therefore, the dominant contributor (new water and old water)
was varied and the contribution of pipeflow to streamflow increased with the increase of rainfall magnitude. (3) The
roles of pipeflow in the storm runoff generation processes demonstrated two roles: the concentration of water and
the rapid drainage to downslopes. Thus, soil pipes can increase the peak discharge from the hillslope and decrease
the peak lag time of the storm hydrograph, when compared to the unpiped hillslope. (4) The roles of pipeflow on
hydrological process suggested that the soil pipes contribute to the slope stability by increasing the rate of soil drainage
and limiting the development of perched groundwater conditions. However, if the rate of water concentration to the
soil pipe network is in excess of the pipeflow transmission capacity, the cavity of the soil pipe could readily be filled
with water during a rain event, increasing pore water pressure in the surrounding matrix. In this case, the soil pipe
induced slope instability. (5) Moreover, pipe erosion has significant impact on the runoff characteristics of pipeflow,
since pipe erosion contributed to a change in the limited drainage capacity of soil pipe. Thus, pipe erosion plays an
important role in shallow landslide initiation. Copyright  2001 John Wiley & Sons, Ltd.
KEY WORDS pipeflow; forested steep slope; storm runoff generation; landslide initiation; hillslope hydrology; pipe
erosion; preferential flow

INTRODUCTION
A chain of connected macropores, developed nearly parallel to the soil surface, is commonly found in slopes,
ranging from subarctic wetland (Woo and diCenzo,1988) and boreal forest (Roberge and Plamondon, 1987) to
tropical rain forest (Elsenbeer and Lack, 1996) and semiarid land (Newman et al., 1998). These macropores
are commonly referred to as ‘soil pipes’ and concentrated subsurface flow of water in these natural soil pipes is
called ‘pipeflow’ (e.g. Jones, 1971, 1981; Beven and Germann, 1982; Pierson, 1983). However, as indicated
by Faeh (1997), no clear distinction is evident between macropores and soil pipes which are commonly
described in the literature (e.g. Bouma, 1981; Beven and Germann, 1982; Hornberger et al., 1990). According
to Beven and Germann (1982), soil pipes are macropores with diameters greater than 4 cm. However, in
forested hillslope, macropores with diameter less than 4 cm are sometimes called ‘soil pipes’ (e.g. Kitahara,
1996; Uchida et al., 1996). Recently, Faeh (1997) proposed that macropores and pipes were differentiated not

* Correspondence to: T. Uchida, Department of Forestry, Faculty of Agriculture, Kyoto University, Kitashiraka-Oiwakecho, Kyoto 6068502,
Japan. E-mail: uchy@kais.kyoto-u.ac.jp

Received 31 March 2000


Copyright  2001 John Wiley & Sons, Ltd. Accepted 25 July 2000
2152 T. UCHIDA, K. KOSUGI AND T. MIZUYAMA

on the basis of their cross-sectional size, as their names might imply, but rather with regard to their lengths
and orientation, which differ according to the factors resulting in their formation. Macropores are preferential
flow channels being predominantly vertically orientated (e.g. Beven and Germann, 1982; Germann, 1990).
Their effect is most pronounced on infiltration. On the other hand, soil pipes are parallel channels which are
sufficient in length to influence the flow processes at the hillslope scale (Faeh, 1997). Their effect is most
pronounced on subsurface hillslope drainage (e.g. Tani, 1997; Uchida et al., 1999a). We accept Faeh’s (1997)
classification and definition of soil pipes.
The mechanisms of subsurface storm runoff in forested headwater catchments in which overland flow
is not the source of storm runoff have been debated since 1930 (Kirkby, 1988; Anderson and Burt, 1990;
Buttle, 1994; Bonell, 1998; Leibundgut et al., 1999). The rapid hydrograph response in forested headwater
catchments can originate from a variety of hydrological processes, including groundwater ridging, translatory
flow and kinematic waves (Buttle, 1994; Bonell, 1998; McDonnell and Buttle, 1998). It has been considered
that pipeflow is an important hydrological process which may account for the rapid transfer of pre-event
water to stream channels in forested hillslopes (e.g. McDonnell, 1990a; Mulholland, 1993; Peters et al., 1995;
Elsenbeer et al., 1995; Keppeler and Brown, 1998; Weiler et al., 1998). Moreover, recent studies emphasized
that pipeflow has a large impact on stream chemistry (e.g. Muscutt et al., 1990; Chapman et al., 1993;
Elsenbeer et al., 1995; Hagedorn, 1999; Hill et al., 1999; Katsuyama et al., 2000).
However, Sklash et al. (1996) indicated that detailed field studies of pipeflow hydrology outside the UK
are less common. This indication suggested that the information about pipeflow hydrology in steep, forested
slopes was inadequate, since the hillslopes in the UK were relatively gentle (the relief ratio was less than 0Ð3)
and the vegetation is commonly included in humid moorland (McCaig, 1983; Jones and Crane, 1984; Wilson
and Smart, 1984; Jones, 1987; Muscutt et al., 1990; Chapman et al., 1993; Sklash et al., 1996; Jones et al.,
1997). Since the 1980s, several field studies of pipeflow hydrology have been conducted in Japanese forested,
steep headwater catchments (e.g. Tanaka et al., 1988; Tsukamoto et al., 1988; Tanaka, 1992, 1996; Kitahara,
1996; Tani et al., 1998). However, it has been considered that adequate information is lacking with regard to
questions as to how representative these previous studies are and how widespread the phenomena might be
(e.g. Selby, 1993; Bryan and Jones, 1997; Jones, 1997a). Thus, the first aim of this study is to review some
studies of pipeflow hydrology on forested steep hillslopes.
Since soil pipes have significant impact on the effective hydraulic conductivity (e.g. Montgomery and
Dietrich, 1995) and the storm runoff generation processes (e.g. Kitahara, 1994), it has been considered that
pipeflow may contribute to landslide initiation (e.g. Wieczorek, 1993; Sidle, 1994; Onda et al., 1996; Iverson
et al., 1997). Numerical simulation showed that relatively slight hydraulic conductivity contrasts of less than
one order of magnitude markedly affect the slope failure potential (Reid and Iverson, 1992). Indeed, soil pipe
outlets are often found in scars of shallow landslides (e.g. Brand et al., 1986; Jenkins et al., 1988; Selby,
1993). Since the 1980s, a variety of processes have been suggested to explain the effects of pipeflow on the
slope stability (e.g. Pierson, 1983; Sidle et al., 1985; Shindo, 1993; Uchida et al., 1996; Onda et al., 1996).
However, the effects of pipeflow on shallow landslide initiation have not been fully understood, because
of insufficient information about the runoff generation processes of pipeflow. Recent studies of pipeflow,
especially in Japan, have improved the knowledge of the runoff generation processes of pipeflow. Thus,
the second aim of this study is to present the various pipeflow effects on hydrological processes related to
landslide initiation.

SOIL PIPE MORPHOLOGY


The outlets of soil pipes varied in size and the diameters ranged from 0Ð1 to 300 cm, as shown in Table I. The
soil pipes in peaty podzols ranged from 3 to 50 cm in diameter, whereas the diameters of soil pipe outlets
in forest soils varied even more widely (0Ð1–50 cm). In some cases, the pipe diameter in forest soil was less
than 1 cm. The densities of soil pipes at excavated trenches on forested slopes ranged from 0Ð28 to 15Ð5

Copyright  2001 John Wiley & Sons, Ltd. Hydrol. Process. 15, 2151– 2174 (2001)
EFFECTS OF PIPEFLOW ON HYDROLOGICAL PROCESS AND LANDSLIDES 2153

Table I. Summary of soil pipe outlet observationa

Site Diameter Densityb Pipe location Flow References


(cm) (pipes/m2 ) type
Soil profile Topography Material

Forest
Yakeyamagawa Japan 0Ð1–9Ð1 6Ð3 ts hillslope brown forest pe Kitahara (1994)
soil
Jozankai Japan 0Ð8–2Ð1 9Ð7 ls hillslope brown forest ep Kitahara and
soil Nakai (1992)
Hitachi Ohta Japan 0Ð2–4 3Ð8 ts/ls valley head brown forest ep Tsuboyama
soil et al. (1994)
Hayano Japan 8, 10 — ls valley head clay loam ep Yasuhara et al.
(1984)
Hakyuchi 1 Japan 5 — ls valley head clay loam ep Shindo (1983)
Hakyuchi 2 Japan 1–50 11Ð6 ts/ls valley head clay loam ep Tsukamoto and
Ohta (1988)
Hakyuchi 3 Japan 1Ð5–30 2Ð3 ls valley head clay loam pe Tanaka et al.
(1984)
Hakyuchi 4 Japan 10 15Ð5 ts hillslope clay loam — Tsukamoto
et al. (1988)
Hakyuchi 5 Japan 10 3Ð4 ts/ls valley head clay loam — Tsukamoto
et al. (1988)
Nozasa Japan 20 — — valley head decomposed ep Ogawa (1997)
granite
Obara Japan 3 — ls hillslope decomposed — Onda (1994)
granite
Akatsu Japan 25–60 — ls valley head decomposed ep Terajima and
granite Sakura
(1993)
Ryuo Japan 5 — ls hillslope brown forest pe Uchida
soil (unpublished)
Ashiu Japan 5 — ls valley head brown forest ep Uchida et al.
soil (1995a)
Hiruzen Japan 7–50 0Ð28 br hillslope Andosol ep Koyama (1994)
Casper Creek USA 2–45 — ls valley head brown forest ep Ziemer and
soil Abright
(1987)
Alum Creek USA 0Ð8–2Ð8 — ts hillslope silt/loam soil ep Navar et al.
(1995)
Panola USA 1–6 — ts? hillslope red–yellow ep Burns et al.
podzolic (1998)
soil
Pointe Baptiste Dominica 10 — ls stream bank volcanic pe Walsh and
clay soil Howells
(1988)
La Cuenca Peru 8 — ts hillslope red–yellow ep Elsenbeer et al.
podzolic (1995)
soil
Maimai NZ 0Ð3–10 — ts/ls hillslope yellow–grey ep McDonnell
podzolic (1990a)
soil
Moorland, grassland
Nant Gerig UK 50 — — valley bottom peaty podzol pe Sklash et al.
(1996)

(continued overleaf )

Copyright  2001 John Wiley & Sons, Ltd. Hydrol. Process. 15, 2151– 2174 (2001)
2154 T. UCHIDA, K. KOSUGI AND T. MIZUYAMA

Table I. (Continued )

Site Diameter Densityb Pipe location Flow References


(cm) (pipes/m2 ) type
Soil profile Topography Material

Nant Gerig UK 10 — — hillslope peaty podzol ep Sklash et al.


(1996)
Cerrig-yr-Wyn UK 4Ð6 — — hillslope peaty podzol — Gilman and
Newson
(1980)c
Slithero Clough UK 3 — — hillslope peaty podzol ep McCaig (1983)
Maesnant UK 5 — — hillslope peaty podzol ep Jones (1987)
Maesnant UK 8–10 — — river terrace peaty podzol pe Jones (1987)
Cwm Llwch UK 6 — ts steep side slope brown earth pe Wilson and
soil Smart (1984)
Afon Cyff UK 5–15 — ts hillslope peaty podzol ep Muscutt et al.
(1990)
upper Wye UK 5–10 — ls hillslope peat gleys — Bell (1972)c
and
podzols
James Bay Canada 6–7 — ls stream bank silt soil ep Woo and
diCenzo
(1988)
Drakensberg South 85 — ls valley head gley soil ep Garland and
Africa Humphrey
(1992)
East Twin Brook UK 2Ð5–5 — — valley bottom peat soil — Weyman (1971,
1975)c
Badland, semiarid land
Yangdaogou China 20–100 — — hillslope loess ep Zhu (1997)
Umsweswe Zim- 50 — ls valley head sand — Stocking (1981)
babwe
Negev — 10 — — hillslope shale — Yair et al.
(1980)c
— — 91 — — flat upland loess — Fuller (1922)c
San Pedro valley USA 6–46 — — river terrace alluvial soil — Jones (1968)c
Arizona USA 5–610 — ls alluvial valley alluvial soil — Brown (1961)c
North Dakota USA 300 — — alluvial terrace clay loam — Bell (1968)c

a Abbreviations: ts, topsoil; ls, lower soil; br, bedrock; pe, perennial; ep, ephemeral.
b The density of soil pipes at excavated trenches.
c Compiled from Jones (1981); for these references see Jones (1981).

pipes/m2 (Ohta et al., 1983; Tsukamoto et al., 1988; Kitahara and Nakai, 1992; Koyama, 1994; Tsuboyama
et al., 1994).
Based on the position of soil pipes in the soil profile, the soil pipes in forest soils were classified into
two groups (Tsukamoto et al., 1988; Kitahara, 1994; Uchida et al., 1996). The first group of soil pipes is
referred to as ‘topsoil pipe’. Soil pipe formation processes in topsoils (A horizon) are different from those
in lower soils (C horizon). Tsukamoto et al. (1988) demonstrated the three fundamental conditions for soil
pipe formation in soils of naturally developed hillslopes: (i) initiation of soil pipes, (ii) development of soil
pipes and (iii) sustenance of soil pipes. In topsoils, high biotic activities provided many soil pipes (Tsukamoto
et al., 1988; Tsukamoto, 1997). Ziemer and Albright (1987) and Onda and Itakura (1997) indicated that biotic
activities were important pipe initiation factors in forested hillslopes. However, these soil pipes are soon
replenished with organic matter and soils because of less frequent water passages (Tsukamoto et al., 1988).

Copyright  2001 John Wiley & Sons, Ltd. Hydrol. Process. 15, 2151– 2174 (2001)
EFFECTS OF PIPEFLOW ON HYDROLOGICAL PROCESS AND LANDSLIDES 2155

Previous studies suggested that the erosion by lateral water movement is one of the most important factors
of pipe development in humid regions (Wilson and Smart, 1984; Tsukamoto et al., 1988; Selby, 1993; Bryan
and Jones, 1997). Therefore, it can be summarized that the pipe initiation agents in topsoils are rich, but both
the forces of pipe development and their sustenance are relatively small (Tsukamoto et al., 1988).
The last type is called a ‘lower soil pipe’. In lower soil layers, biotic activities were small compared with
topsoils. However, especially above a soil–bedrock interface, the pipe developing force is rich, since the lateral
water movement occurs frequently. Thus, soil pipe networks commonly developed above bedrock or imperme-
able horizons, parallel to them (Pierson, 1983; Yasuhara et al., 1984; Tsukamoto and Ohta, 1988; Walsh and
Howells, 1988; McDonnell, 1990a; Kitahara and Nakai, 1992; Tsuboyama et al., 1994; Uchida et al., 1995a).
Jones et al. (1997) investigated the distribution of soil pipes in the UK and reported that 70% of soil
pipes are found in moorland catchment covered by peats or podzols. They associated the soil pipe formation
primarily with soils of low permeability in upland regions. In Japanese forested slopes, the soil pipes are
found in a variety of soils, including forest brown soil, clay loam, decomposed granite and Andosol (see
Table I). Tsukamoto (1997) conducted a survey of two piped slopes (Hakyuchi (clay loam) and Nishimikawa
(decomposed granite); refer to Tables I and VI) to investigate the effects of soil properties on the conditions
for soil pipe formation. He suggested that the pipe developing rate by pipe erosion in the clay loam was
lower than that in the sandy soil (decomposed granite), since the cohesion of the clay loam is greater than
that of the sandy soil. Dunne (1990) and Onda et al. (1996) also demonstrated that the cohesion of a soil had
impacts on the resistance of the soil to erosion by subsurface flow. Moreover, the suggestion by Tsukamoto
(1997) corroborated the previous observation, which showed that the sediment yields from the soil pipes were
commonly greater in forested hillslopes covered by decomposed granites (Terajima and Sakura, 1993; Sonoda,
1993; Onda, 1994; Ogawa, 1997). Further, Tsukamoto (1997) suggested that once soil pipes are formed, soil
pipes in the clay loam show relatively rich sustenance agents determined by soil hardness compared with the
sandy soil. Thus, he suggested that periods from the initiation of the specific soil pipe to the disappearance
in the clay loam may be longer than those in the sandy soil (Tsukamoto, 1997). These investigations (Jones
et al., 1997; Tsukamoto, 1997) indicated that soil properties, especially permeability, cohesion and hardness,
had a large impact on the conditions for soil pipe formation.
Numerous investigations about soil pipe morphologies have been conducted in British moorlands, which are
commonly covered by peats and/or podzols (see previous reviews by Jones (1981) and Jones et al. (1997)).
Soil pipe networks in badlands have also been observed in detail (Stocking, 1981; Jones, 1981; Zhu, 1997).
However, as indicated by Noguchi et al. (1997, 1999), adequate information is lacking with regard to the
upslope connectivity of soil pipe network systems in forest soil. It is difficult to clarify the morphological
characteristics of soil pipe in forest soil, since soil pipes are commonly small (Table I), short and dynamic.
Kitahara and Nakai (1992) and Uchida et al. (1995a) reported that the responses of one pipeflow to storms
were similar to those of the adjacent pipeflow. Onda (1994) showed that a decrease in the discharge rate
from one pipe caused an increase in that of the adjacent pipe. These results suggested that the soil pipes
developed as a network. Indeed, Koyama (1994) reported that two soil pipes at the same excavated trench
were connected with each other. Uchida et al. (1995a) indicated that the shapes of the outlets were almost
circular, with a diameter of about 5 cm, and the round shape of the outlet extended for about 30 cm. However,
beyond this length, the soil pipe was joined by several secondary pipes (the openings in the gravel) (Uchida
et al., 1995a). Moreover, a recent dye test demonstrated that even though lengths of individual macropores
were rather short, the coupling of these flow paths with the soil matrix, living and decayed roots and bedrock
fractures produced complex networks (Noguchi et al., 1999).

RUNOFF GENERATION PROCESS OF PIPEFLOW


Runoff characteristics of pipeflow
Pipeflow velocity ranged between 0Ð54 and 80 cm/s (Table II). These values were comparable to the velocity
of overland flow (0Ð1–21 cm/s) summarized by McCaig (1983). The peak discharge rate of pipeflow varied

Copyright  2001 John Wiley & Sons, Ltd. Hydrol. Process. 15, 2151– 2174 (2001)
2156 T. UCHIDA, K. KOSUGI AND T. MIZUYAMA

Table II. Subsurface flow velocities through soil pipe

Site Velocity Slope Diameter References


(cm/s) gradient (cm) (cm)

Forest soil
Hakyuchi 3 18Ð3 0Ð11 30 Tanaka (1992)
Jozankei 0Ð73–0Ð90 0Ð88 0Ð83–2Ð1 Kitahara (1994)
Yakeyamagawa 0Ð54–3Ð1 0Ð19–0Ð25 0Ð1–2Ð2 Kitahara (1994)
Peaty podzol
East Twins Brook 10–30 0Ð04 2Ð5–5 Weyman (1971, 1975)a
East Twins Brook 30 0Ð04 2Ð5–5 Stagg (1974)a
Nant Gerig 10 0Ð36 5 Newson and Harrison (1978)
Maesnant 15 0Ð07 9 Jones (1975, 1978)a
Cwn Llwch 20–80 0Ð24 6 Wilson (1977)a
Slithero Clough 10 0Ð15 3? McCaig (1979)a

a Compiled from Jones (1981); for these references see Jones (1981).

Forest soil (0.6 < I) Qpmax = 0.00038*d 3.3 (R2 = 0.794)


Forest soil (0.15 < I < 0.6) Qpmax = 0.00081*d 2.2 (R2 = 0.617)
100

10
Maximum pipeflow (l/s)

1 Forest soil (0.6 < I)


Forest soil (0.15 < I < 0.6)
0.1
Peaty podzol (I < 0.15)
0.01 Peaty podzol (0.15 < I < 0.6)

0.001

0.0001

1E − 05
1 10 100
Pipe diameter (cm)
Figure 1. Relationship between maximum discharge rate of pipeflow and diameter of soil pipe outlet

widely, and ranged from 0Ð0002 to 59Ð3 l/s (Table III). Total pipeflow during one storm flow period ranged
from 300 to 135 000 l (Table III). Based on data from Table III, the relationship between the maximum
discharge rates of pipeflow and the diameters of soil pipe outlets is shown in Figure 1. Data points in this
figure are classified into four categories for slope gradient at soil pipe outlets (I) and soil materials. Maximum

Copyright  2001 John Wiley & Sons, Ltd. Hydrol. Process. 15, 2151– 2174 (2001)
Table III. Summary of pipeflow observationa

Site Pipe characteristics Pipe location Peak pipeflow Lag time Total pipeflowb References

Type D (cm) n Topography Angle Soil Max. Conditions Start Peak n Mean (I) n
profile value (h) (h)
Rt (mm) Ip (mm/h)

Forest soil
Yakeya- pe 9Ð1 1 hs 0Ð19 ls 0Ð03 — — — — — — — Kitahara (1994)
magawa
Jozankei ep 0Ð9–2 1 hs 0Ð88 br 0Ð006 39 18 2 1 3 — — Kitahara and

Copyright  2001 John Wiley & Sons, Ltd.


Nakai (1992)
Jozankei ep 0Ð9–2 1 hs 0Ð88 ts 0Ð002 39 18 1 4 1 — — Kitahara and
Nakai (1992)
Jozankei ep 0Ð9–2 1 hs 0Ð88 ts 7E  04 39 18 1 4 1 — — Kitahara and
Nakai (1992)
Jozankei ep 0Ð9–2 1 hs 0Ð88 ts 4E  04 24Ð5 8 0–1 0–1 2 — — Kithara et al.
(1994)
Jozankei ep 0Ð9–2 1 hs 0Ð88 ts 2E  04 24Ð5 8 0–1 0–1 2 — — Kithara et al.
(1994)
Jozankei ep 0Ð9–2 1 hs 0Ð88 ts 0Ð001 24Ð5 8 0–1 0–1 2 — — Kithara et al.
(1994)
Jozankei ep 0Ð9–2 1 hs 0Ð88 ts 0Ð001 24Ð5 8 0–1 0–1 2 — — Kithara et al.
(1994)
Jozankei ep 0Ð9–2 1 hs 0Ð88 ts 0Ð001 24Ð5 8 0–1 0–1 2 — — Kithara et al.
(1994)
Jozankei ep 0Ð9–2 1 hs 0Ð88 ts 6E  04 24Ð5 8 0–1 0–1 2 — — Kithara et al.
(1994)
Hiruzen ep 50 1 hs 0Ð47 ls 1Ð85 167 34 34 28 3 — — Koyama (1994)
Hiruzen ep 10 1 hs 0Ð47 ls 0Ð005 167 34 — 31 2 — — Koyama (1994)
Hiruzen ep 7–10 2 hs 0Ð47 ls 0Ð092 167 34 — 25 2 — — Koyama (1994)
Hakyuchi 1 ep 5 1 vh 0Ð44 — 0Ð012 — 10 0 1 1 — — Ohta
et al.(1983)
Hakyuchi 2 pe 5 1 vh 0Ð52 ls 0Ð5 75 22 9 5 1 19400 14 Ohta
EFFECTS OF PIPEFLOW ON HYDROLOGICAL PROCESS AND LANDSLIDES

et al.(1983)
Hakyuchi 2 ep 11 1 vh 0Ð52 ts 0Ð21 75 22 18 6 1 9900 14 Ohta
et al.(1983)
Ryuo pe 5 1 hs 0Ð76 — 0Ð12 158 54Ð5 58 1 1 — — Uchida
(unpublished)

(continued overleaf )

Hydrol. Process. 15, 2151– 2174 (2001)


2157
2158

Table III. (Continued )

Site Pipe characteristics Pipe location Peak pipeflow Lag time Total pipeflowb References

Type D (cm) n Topography Angle Soil Max. Conditions Start Peak n Mean (I) n

Copyright  2001 John Wiley & Sons, Ltd.


profile value (h) (h)
Rt (mm) Ip (mm/h)

Hitachi Ohta ep 0Ð3–0Ð7 3 hs 0Ð81 ts 4E  04 61Ð4 18Ð6 10 6 3 — — Sidle et al.


(1995)
Hitachi Ohta ep 0Ð7–0Ð8 2 hs 0Ð81 ls 0Ð005 61Ð4 18Ð6 8 4 3 — — Sidle et al.
(1995)
Hitachi Ohta ep 0Ð3–1 4 hs 0Ð81 ts 1E  04 61Ð4 18Ð6 8 2 3 — — Sidle et al.
(1995)
Toino ep 5 1 vh 0Ð71 ls 0Ð18 194 25Ð5 12Ð1 3Ð7 26 1300 75 Uchida et al.
(1995,1996)d (1999a)
Toino ep 5 1 vh 0Ð71 ls 0Ð074 75Ð5 9Ð5 11 1Ð6 18 3700 27 Uchida et al.
(1993)d (1999a)
Toino ep 5 1 vh 0Ð71 ls 0Ð17 70 17 12Ð6 1Ð3 9 900 25 Uchida et al.
(1999a)
La Cuenca ep 8 1 hs 0Ð51 — 0Ð22 38 34 0 0 2 316 6 Elsenbeer and
Lack (1996)
T. UCHIDA, K. KOSUGI AND T. MIZUYAMA

Casper Creek pe 60–80 1 vh 0Ð31 ls 8Ð3 — — — 6Ð3 7 — — Keppelar and


Brown
(1998)
Casper Creek pe 50 1 vh 0Ð45 ls 28 — — 10? 8? 1 Keppelar and
Brown
(1998)
Peaty podzol
Afon Cyff ep 5–10 1 hs 0Ð25 ts 1Ð5 63 2Ð5 5 6 1 35000 1 Muscutt
et al.(1990)
Cwn Llwch ep 6 1 hs 0Ð24 — 1Ð5 — — — — — — — Wilson (1977)c
Cerrig- — 5 hs 0Ð21 — 1 — — — — — — — Gilman and
yr-Wyn Newson
(1980)c
East Twins — 2Ð5–5 1 vf 0Ð04 — 0Ð75 — — — — — — — Stagg (1974)c
Brook

Hydrol. Process. 15, 2151– 2174 (2001)


Maesnant ep 5 1 hs 0Ð27 — 2Ð3 — — 11–13 1–3 190 38200 190 Jones (1987)
Maesnant ep 5 1 hs 0Ð27 — 7Ð1 — — 11–13 1–3 190 73400 190 Jones (1987)
Maesnant ep 5 1 hs 0Ð27 — 6Ð3 — — 11–13 1–3 190 135000 190 Jones (1987)
Maesnant ep 5 1 hs 0Ð27 — 3Ð7 — — 11–13 1–3 190 40000 190 Jones (1987)
Maesnant ep 5 1 hs 0Ð27 — 1Ð6 — — 11–13 1–3 190 33500 190 Jones (1987)
Maesnant pe 8–10 1 rt 0Ð07 — 6Ð6 — — 13Ð2 7Ð8 190 9600 190 Jones (1987)
Maesnant pe 8–10 1 rt 0Ð07 — 8Ð3 — — 13Ð2 11Ð8 190 7400 190 Jones (1987)
Maesnant pe 8–10 1 rt 0Ð07 — 59Ð3 — — 8Ð6 7Ð7 190 10600 190 Jones (1987)
Maesnant pe 8–10 1 rt 0Ð07 — 2Ð9 — — 8–11 4–5 190 5300 190 Jones (1987)
Maesnant pe 8–10 1 rt 0Ð07 — 9Ð4 — — 8–11 4–5 190 3200 190 Jones (1987)
Others
James Bay ep 6–7 1 rt 5E  04 ls 0Ð7 22 8 0 1 1 — — Woo and
diCenzo
(1988)

Copyright  2001 John Wiley & Sons, Ltd.


Drakens- ep 85 1 vh 0Ð15 ls 3Ð47 — — — 1–7Ð5 26 — — Garland and
berg Humphrey
(1992)

a Abbreviations: D, diameter; n, number of observed soil pipe outlets. Several studies observed total discharge rate from several pipes; Angle, ground surface angle at outlet of
soil pipe; Rt , total rainfall amount; Ip , peak one hour rainfall intensity; pe, perennial; ep, ephemeral; hs, hillslope; vh, valley head; vf, valley floor; rt, river terrace; ls, lower soil;
br, bedrock; ts, topsoil.
b Total pipeflow during one storm flow period.
c Compiled from Jones (1981); for these references see Jones (1981).
d In spite of the same pipe, runoff characteristics in 1993 were different from those in 1995– 96 (Uchida et al., 1999a).
EFFECTS OF PIPEFLOW ON HYDROLOGICAL PROCESS AND LANDSLIDES

Hydrol. Process. 15, 2151– 2174 (2001)


2159
2160 T. UCHIDA, K. KOSUGI AND T. MIZUYAMA

discharge rate of pipeflow (Qp max ) in forest soil can be quantitatively related to the diameter of soil pipes (d)
as follows:
Qp max D ˛dˇ

where ˛ and ˇ are coefficients. The value of ˇ is 2Ð2 for 0Ð15 < I < 0Ð6 and 3Ð3 for 0Ð6 < I. The calculated
Reynolds number of pipeflow in forest soil (Kitahara, 1994) ranges from 3Ð9 to 205, indicating that generally
pipeflow in forest soil is nonturbulent, similar to the pipeflow in peaty podzols (Jones, 1981). This indication
suggests the possibility that the discharge rate of pipeflow in forest soil can be described by the theory of
Hagen–Poiseuille flow. However, the calculated ˇ value (2Ð2–3Ð3) is smaller than the theoretical value (4Ð0)
of Hagen–Poiseuille flow.
When the condition that both the diameter of soil pipe (d) and the slope gradient at the soil pipe outlet (I)
were almost the same, the maximum rate of pipeflow in forested soils was lower than that in peaty podzol
(Figure 1). Generally, the pipeflow velocities in forested hillslopes are also lower than those in peaty podzols
(Table III). These differences may imply a difference between the roughness and/or the tortuosity of soil pipes
in forest soil and those in peaty podzol.
Moreover, it has been considered that the limited drainage capacity of soil pipes was determined by the
morphological characteristics of soil pipe networks (Tsukamoto et al., 1988; McDonnell, 1990b; Loukas and
Quick, 1993; Kitahara, 1994). The correlation between the soil pipe diameter and the maximum rate of
pipeflow (Figure 1) supports these suggestions. Uchida et al. (1999a) reported that the sediment discharge
from a soil pipe had an impact on the maximum value of pipeflow discharge. This observation suggested that
the soil pipe morphological change, which was caused by the sediment yield from the soil pipe, contributed to
the increase in the limited drainage capacity of the soil pipe. However, it is difficult to quantify the relationship
between the morphological characteristics of soil pipe and runoff characteristics of pipeflow in forest soil,
since adequate information is lacking with regard to the upslope connectivity of soil pipe network systems in
forest soil (refer to the previous section).
Although the pipeflow velocities are comparable to that of overland flow, the start lag times (lags in start
times of storm runoff from soil pipe in comparison with the start times of rainfall burst) are often greater
than 10 h (Ohta et al., 1983; Ziemer and Albright, 1987; Koyama, 1994; Uchida et al., 1999a) (Table III).
The delayed start of pipeflow was also reported in moorland catchments in the UK (Jones and Crane, 1984;
Jones, 1987; Sklash et al., 1996). Wilson et al. (1990) and Sidle et al. (1995a) showed that pipeflows were
more likely to occur under wet antecedent moisture conditions. Additionally, Ziemer and Albright (1987)
and Luxmoore et al. (1990) reported that ephemeral pipes required some initial rainfall before they began to
flow. Uchida et al. (1999a) indicated that this trigger amount of rainfall varied depending on antecedent soil
moisture conditions. These results are similar to the results in the forested hillslope, which generate pipeflow
(McDonnell, 1990a; Burns et al., 1998). They indicated that a rainfall threshold was necessary for significant
hillslope contributions to channel storm flow. Moreover, the previous studies in forested slopes suggested that
a minimum groundwater stage was required before a quick response can occur in soil pipes (Roberge and
Plamondon, 1987; Terajima et al., 1996a), as similar to the results of moorland catchments in the UK (Jones,
1981; McCaig, 1983; Sklash et al., 1996).
The lags in peak ephemeral pipeflow time in comparison with the peak rainfall intensity (peak lag time) were
relatively short (1–7Ð5 h), with the exception of Hiruzen (Koyama, 1994) (Table III). Koyama (1994) reported
that the storm pipeflow hydrograph had two peaks. The first peak was almost in conjunction with the peak of
the rainfall intensity, similar to other sites. The second peak was large and lagged the peak of rainfall intensity
by 28 h. Moreover, he indicated that the runoff water temperature at the second great peak was significantly
lower than that at the first small peak. This result suggested that the source of pipeflow at the second peak
shifted to a deeper zone, compared with the first peak, since his observation was conducted in the summer
season (the soil temperature decreased with an increase in soil depth). Uchida et al. (1999b) also reported a
similar response of the pipeflow temperature; the pipeflow temperature in late summer immediately decreased
after the end of a rain burst. Recent works showed that the deep groundwater (e.g. bedrock groundwater)

Copyright  2001 John Wiley & Sons, Ltd. Hydrol. Process. 15, 2151– 2174 (2001)
EFFECTS OF PIPEFLOW ON HYDROLOGICAL PROCESS AND LANDSLIDES 2161

caused the delayed large peak of the streamflow in headwater catchments (Hirose et al., 1994; Lakey and
Krothe, 1997; Onda et al., 1999). These results suggested that the deep groundwater may contribute to the
storm runoff of pipeflow in Hiruzen. The result in Koyama (1994) implies that the deep groundwater sometimes
causes the long lags in times of peak pipeflow in comparison with the peak times of rainfall intensity.

New water vs. old water


High cation and silica concentrations of pipeflow suggested that ‘old’ (pre-event) water was an important
contributor to pipeflow (Walsh and Howells, 1988; Koyama, 1994). A two-component model using geochemi-
cal and isotopic tracers also indicated that the runoff water in a lateral preferential flowpath was dominated by
old water (Sklash et al., 1986; Pearce et al., 1986; Burns et al., 1998). These results concur with the pipeflow
chemistry and the hydrograph separation of pipeflow in moorland catchments in the UK (Chapman et al.,
1993; Sklash et al., 1996).
However, not all pipeflow showed old water dominance. Several investigations suggested that the cation
concentration in pipeflows was lower than that of soil matrix flow (Burns et al., 1998; Katsuyama et al.,
2000). In some fields, new water played an important role in the runoff generation of pipeflow (Elsenbeer
et al., 1995; Turton et al., 1995).

Conceptual model of pipeflow pathway


These differences (the hydrological response of pipeflow to rainfall and the ratio of new water to
old water) suggested that the pipeflow might have a variety of sources. Noguchi et al. (1999) evaluated
morphological characteristics of macropores and the distribution of preferential flow pathways in forest soil
using a combination of staining agents. Noguchi et al. (1999) indicated that although the length of individual
macropores was less than 62 cm, they were connected into longer preferential flow paths by a variety of flow
paths.
From the previous investigations, a conceptual model describing the runoff generation processes of pipeflow
is shown in Figure 2. Recent work on hillslope hydrology emphasized the importance of water flow through
bedrock fractures (see pathway 1, Figure 2) at a variety of headwater catchments (e.g. Wilson and Dietrich,
1987; Anderson et al., 1997; Onda et al., 1999). Moreover, Noguchi et al. (1999) reported some macropores
connecting to the bedrock fracture. Adequate information is lacking with regard to the contribution of bedrock
groundwater on pipeflow, but Koyama (1994) and Uchida et al. (1999b) suggested that water movement in
the bedrock zone may have important roles in the storm runoff processes of pipeflow (refer to the previous
section).
Noguchi et al. (1999) also reported that in some cases soil water flows only through macropores without
interacting with the soil matrix (pathway 2, Figure 2). Moreover, Uchida et al. (1999b) found that streamflow
delivered groundwater from an upslope area to a stream via soil pipes, shortcutting the mixing process with
the soil matrix water (pathway 3, Figure 2), similar to the UK’s moorland covered by peats (Sklash et al.,
1996).
The linkage of these flowpaths (pathways 2 and 3) has been supported by Wilson et al. (1991) and Elsenbeer
et al. (1995), who indicated that the variation of new water in pipeflows was sensitive to the variation in the
rainfall intensity. Moreover, Hill et al. (1999) and Hagedorn (1999) also showed the linkage of these two
flowpaths, based on the measurements of nitrogen leaching. This linkage of flowpaths may have contributed
to the results of Turton et al. (1995), Burns et al. (1998) and Katsuyama et al. (2000), which indicated that
new water played an important role in the runoff generation of pipeflows.
However, recent tracer and staining tests showed the interaction of pipeflow with the adjacent soil matrix
(pathway 4, Figure 2). The staining experiment by Noguchi et al. (1999) revealed that certain macropores
interacted with the surrounding mesopores to enlarge these preferential flow paths during wet conditions
(the lowest figure in Figure 3). This pathway was also indicated by the previous tracer tests (Luxmoore and
Ferrand, 1993; Tsuboyama et al., 1994; Navar et al., 1995). Tsuboyama et al. (1994) conducted sprinkling tests

Copyright  2001 John Wiley & Sons, Ltd. Hydrol. Process. 15, 2151– 2174 (2001)
2162 T. UCHIDA, K. KOSUGI AND T. MIZUYAMA

(ii) Preferential flow through decayed and living


(iii) Shortcutting normal
roots, and vertical connected macropores
mixing process with soil
→ new water dominant
matrix water
→ quick response
→ new water dominant
(Wilson et al., 1991; Elsenbeer et al., 1995)
→ quick response
(Wilson et al., 1991;
Uchida et al. 1997)

Top soil pipe Soil


Bedrock fracture
Bedrock

Lower soil pipe (i) Exfiltration from bedrock fracture


→ old water dominant ?
→ delayed peak
(Koyama , 1994; Uchida et al., 1999b)
(v) Interaction of macropore
with surrounding soil matrix
(e.g., Tsuboyama et al., 1994; (iv) Water movement from transient saturated groundwater above bedrock
Navar et al., 1995) or impeding layer into soil pipes
→ old water dominant
→ delayed start and quick peak
(e.g., McDonnell, 1990; Peters et al., 1995; Sklash et al., 1996; Tani, 1997)

Figure 2. A schematic-perceptual model describing runoff generation processes of pipeflow in forested slope segment. Black arrows indicate
water movement. Based on Noguchi et al. (1999) (Reproduced by permission of soil society of America, Inc.) and Sidle et al. (2000).
(Storm flow generation in steep forested headwater a linked hydrogeomorphic paradigm, Hydrological Processes, Sidle RC, Tsuboyama Y,
Noguchi S, Hosoda I, Fujieda M, Simizu T. 2000.  John Wiley & Sons Limited. Reproduced with permission.)

with applied conservative tracers and measured subsurface flow and tracer breakthrough from five portions
(the organic-rich soil layer including macropores, the mineral soil matrix, and three groups of macropores in
the mineral soil layer) at an excavated soil trench. They analyzed subsurface flow and tracer breakthrough
curves using a convective–dispersive model and calculated effective pore volumes. This calculation showed
that the pore volume for the entire profile increased only slightly with the increasing rainfall application rate,
but the relative proportions of macropore increased as antecedent hydrologic conditions became wet. These
results indicated that the ‘hydrologically active’ preferential flow pathways extended with the expansion of
interactive zones, as shown in a schematic illustration (see the expansion process 2, Figure 3).
Tani (1987), McDonnell (1990a), Peters et al. (1995) and Buttle and Turcotte (1999) proposed similar
conceptual models which indicated that pipeflow on hillslopes can result in large old water contributions to
storm flow. These models showed that new water infiltrates rapidly into permeable forest soils via vertical
preferential flowpaths or soil matrices to a soil–bedrock interface, where a perched groundwater zone forms,
producing a mixing of event water with large volumes of stored old soil matrix water. This saturated water

Copyright  2001 John Wiley & Sons, Ltd. Hydrol. Process. 15, 2151– 2174 (2001)
EFFECTS OF PIPEFLOW ON HYDROLOGICAL PROCESS AND LANDSLIDES 2163

water movement

Dry condition Soil pipe

hydrologically active area

(i) Expansion of
hydrologically active area

(iii) Extension of
macropore network
upslope

(ii) Expansion of
interactive zone

interactive zone
Wet condition

Figure 3. Conceptual model of an expansion of hydrologically active area at soil matrix–pipe interface, the expansion of surrounding soil
that interacts with water in macropores and the extension of macropore network upslope. Based on Tsuboyama et al. (1994) (Tsuboyama Y,
Sidle RC, Noguchi S, Hosoda I, Water Resources Research, 30: 879– 890, 1994. Copyright  1994 by the American Geophysical Union.)
and Uchida et al. (1997)

is displaced rapidly downslope via soil pipes along the soil–bedrock interface (pathways 3 and 5, Figure 2).
These conceptual models were supported by the previous measurements, which reported that a minimum
groundwater level was required before a quick response can occur in soil pipes (e.g. Roberge and Plamondon,
1987; Terajima et al., 1996a). However, Uchida et al. (1997) reported that the presence of pipeflow depends
not only on the peak of soil pore water pressure but also on the duration of the saturated zone formation.
This is probably because the area of interface between the soil pipes and the soil matrix where water moves
smoothly (‘hydrologically active area’) increases as the duration of the saturated zone formation increases
(Uchida et al., 1997) (see the expansion process 1, Figure 3).
Moreover, Tani (1997), Buttle and Turcotte (1999) and Sidle et al. (2000) indicated that the contributing
area of pipeflow expanded with the extension of the saturated groundwater zone (see the expansion process
3, Figure 3). Isotopic and/or chemical compositions of pipeflow supported the concept of expansion of the
contributing area, and indicated that much of storm runoff water travelled considerable distances via subsurface
soil pipes and was not necessarily contributed to by surface runoff processes (Peters et al., 1995). The
expansion of the contributing area and the ‘hydrologically active’ soil pipes may bring about a change of
the dominant contributor of pipeflow. Turton et al. (1995) reported that new water becomes important during
infrequent large storms, while old water is dominant during frequent small storms.

Copyright  2001 John Wiley & Sons, Ltd. Hydrol. Process. 15, 2151– 2174 (2001)
2164 T. UCHIDA, K. KOSUGI AND T. MIZUYAMA

EFFECTS OF PIPEFLOW ON STREAMFLOW


When the rainfall magnitude was small, the pipeflow contribution to streamflow was commonly small (Walsh
and Howells, 1988; Kitahara et al., 1994; Sidle et al., 1995a, 2000; Uchida et al., 1999a). Terajima et al.
(1996a) and Uchida et al. (1999b) also showed that the Darcian groundwater flow plays a significant role in
the discharge phenomena during the low flow period, based on the relationship between the soil pore water
pressure distribution and the discharge rate of streamflow.
As the rainfall magnitude increased, the role of the slope discharge shifted towards the pipeflow (Terajima
et al., 1996a; Uchida et al., 1999b) and the ratio of pipeflow to streamflow increased (Kitahara et al., 1994;
Sidle et al., 1995a, 2000). The increase of pipeflow contribution with the increase of rainfall magnitude may
be caused by the extension of both the contributing area of pipeflow and the ‘hydrologically active’ soil pipes
(Figure 3; refer to the previous section). Under the heaviest rainfall, the contribution of pipeflow to the peak
subsurface flow from excavated trenches on forest hillslopes was sometimes greater than 90% (Tsukamoto
and Ohta, 1988; Kitahara et al., 1994), although the maximum discharge rate from the individual pipe was
small (Table III) (Kitahara et al., 1994).
Numerical simulations also examined the contribution of pipeflow to the storm hydrograph (Barcelo and
Nieber, 1981, 1982; Watanabe and Imai, 1984; Nieber and Warner, 1991; Tani and Abe, 1996; Himeno, 1997;
Kato, 1998). Methods of numerical simulation are listed in Table IV. These researches demonstrated that
the soil pipe exerted a significant influence on the peak discharge, the peak lag time and the soil moisture
distribution in the hillslope, but the impact of pipeflow on the baseflow hydrograph was relatively small
(Tani and Abe, 1996). Moreover, Barcelo and Nieber (1981) reported that the effects of pipeflow on storm
hydrograph increased with an increase in rainfall magnitude, and Nieber and Warner (1991) showed that
pipeflow can contribute to the majority of total subsurface storm flow.
However, as mentioned previously, soil pipes have limited drainage capacity. Thus, when the discharge rate
of pipeflow came up to the limited drainage capacity, the contribution of the individual pipeflow to streamflow
decreased with an increase in rainfall magnitude (Uchida et al., 1999a). This concurs with the result in the
UK’s piped catchment, which demonstrated that the peak in the percentage that contributed to the streamflow
from the pipe network is related to the limited capacity of the pipe network to carry the water (Jones, 1987).
Moreover, Uchida et al. (1999a) reported that once the pipeflow, which was located at 6 m upstream from

Table IV. Previous numerical simulations


References Model Matrix flow Water movement between Water movement in soil pipe
soil pipe and soil matrix

Barcelo and Nieber longitudinal 2D saturated– tile drainage theory storage of soil pipe
(1981) unsaturated flow assumed to be zerob
Barcelo and Nieber pseudo 3D saturated– tile drainage theory storage of soil pipe
(1982) unsaturated flow assumed to be zerob
Watanabe and Imai 3D saturated– Darcy–Backingham eq.a large hydraulic
(1984) unsaturated flow conductivitya
Nieber and Warner 3D saturated flow constant head storage of soil pipe
(1991) assumed to be zerob
Tani and Abe longitudinal 2D saturated– seepage face storage of soil pipe
(1996) unsaturated flow assumed to be zerob
Himeno (1997) longitudinal 2D saturated– tile drainage theory storage of soil pipe
unsaturated flow assumed to be zerob
Kato (1998) longitudinal 2D saturated– tile drainage theory storage of soil pipe
unsaturated flow assumed to be zerob

a Pipeflow is assumed to match the Darcy–Backingham equation with large saturated hydraulic conductivity.
b The discharge rate of pipeflow equals the rate of water movement from soil matrix to soil pipe.

Copyright  2001 John Wiley & Sons, Ltd. Hydrol. Process. 15, 2151– 2174 (2001)
EFFECTS OF PIPEFLOW ON HYDROLOGICAL PROCESS AND LANDSLIDES 2165

the watershed exit, came up to the limited drainage capacity, additional pipeflows occurred from other soil
pipes located 10 m upstream from the watershed exit.
These observed results fitted well with the pipeflow model describing the timing of pipeflow occurrence and
the contribution of pipeflow to storm runoff, which was applied for the moorland catchment in the UK (Wilson
and Smart, 1984). This indication suggested that the process of pipeflow initiation in moorland catchments is
similar to that in forested catchments, although both the amounts and the velocities of pipeflow in moorlands
are different from those in forested catchments (Tables II and III). Wilson and Smart (1984) observed that
once the total discharge rate of subsurface flow from the upslope exceeded the subsurface seepage capacity,
then water was delivered to the pipeflow system. This threshold response is central to this model. This model
indicated that once pipeflow occurred, the contribution of pipeflow increased with the increasing rainfall
magnitude until the total discharge rate of subsurface flow from the upslope exceeded the capacity of both
subsurface seepage and pipeflow.
Dunne (1978), Kirkby (1988) and Anderson and Burt (1990) had suggested that differences in the peak
runoff rate and the lag between peak rainfall and peak runoff (peak streamflow lag time) among various
watersheds could be attributable to the mechanism of storm flow generation as well as the watershed size
(Figure 4). Moreover, based on the observation results in Maesnant (mid-Wales) (refer to Table I), Jones
(1997a, b) plotted pipeflow in the same context and indicated that pipeflow responses were intermediate
between subsurface flow and saturation overland flow (Figure 4). The peak responses at Japanese forested
catchments, which generate pipeflow, are plotted in the same context (Figure 4). Some characteristics of these
catchments are listed in Table V. The drainage basin areas are determined by surface topography.
Figure 4 indicates that the peak responses in Japanese pipeflow catchments were similar to that in Maesnant
(Figure 4). Most of the peak streamflow lag times in forested pipeflow catchments were intermediate between
subsurface flow catchments and saturation overland flow catchments. The largest peak runoff in forested

Table V. Summary of Japanese piped watershed. Results of peak streamflow responses are shown in Figure 4a

Site Pipe location Weir location Observed storm References


Topography Soil Rt (mm) Ip (mm/h)
profile

Jozankei vh ls end of 0 order basin snow melt event Terajima et al. (1996a)
Hitachi Ohta hs ts/ls end of 0 order basin 48 21 Sidle et al. (1995a, 2000)
Hakyuchi 1 vh ls end of 0 order basinb 67 5 Shindo (1993)
Hakyuchi 2 vh ts/ls end of 0 order basinb 72 23 Tsukamoto and Ohta (1988)
Hakyuchi 3 vh ls end of 0 order basin 178 25 Tanaka et al. (1984)
Obara vf ls end of 0 order basin 144 14 Onda (1994)
Inabu hs — sideslope of first order 244 — Tanaka et al. (1990)
stream
Kanedaira vh ls end of second order basin 67 11 Himeno (1997)
Toino vh ls end of 0 order basin 194 26 Uchida et al. (1999a)
106 31 Uchida et al. (1999a)
Tatsunokuchi SL hs ls sideslope of first order 60 12 Tani (1997)
streamb
Tatsunokuchi MN hs ls second order stream 60 12 Tani (1997)
Tatsunokuchi KT hs ls second order stream 60 12 Tani (1997)
Hiruzen hs ls sideslope of first order 167 33 Koyama (1994)
streamb
Amatani hs ls second order stream 167 33 Koyama (1994)

a Abbreviations: vh, valley head; hs, hillslope; vf, valley floor; ls, lower soil; ts, topsoil; Rt , total rainfall amount; Ip , peak one hour rainfall
intensity.
b Runoff measured at excavated trench.

Copyright  2001 John Wiley & Sons, Ltd. Hydrol. Process. 15, 2151– 2174 (2001)
2166 T. UCHIDA, K. KOSUGI AND T. MIZUYAMA

Figure 4. Peak lag times and peak runoff rates for hillslope process. Based on Dunne (1978) (Field studies of hillslope flow process, Dunne T.
Editor Kirkby MJ. 1978.  John Wiley & Sons Limited. Reproduced with permission.), Anderson and Burt (1990) (Subsurface runoff. In
Process Studies in Hillslope Hydrology, Anderson MG, Burt TP. Editors Anderson MG, Burt TP. 1990.  John Wiley & Sons Limited.
Reproduced with permission.), Jones (1997a) (Pipeflow contribution areas and runoff response, Hydrological Processes, Jones JAA. 1997. 
John Wiley & Sons Limited. Reproduced with permission.) and Uchida et al. (1999a) (Reprinted from Journal of Hydrology, 222, Uchida T,
Kosugi K, Mizuyama T. Runoff characteristics of pipeflow and effects of pipeflow and effects of pipeflow on rainfall-runoff phenomena in
a mountainous watershed, 18–36. Copyright 1999, with permission Elsevier Science.) with additional information from the authors. Some
characteristics of the observed watersheds are listed in Table V

Copyright  2001 John Wiley & Sons, Ltd. Hydrol. Process. 15, 2151– 2174 (2001)
EFFECTS OF PIPEFLOW ON HYDROLOGICAL PROCESS AND LANDSLIDES 2167

pipeflow catchments was almost equal to the smallest value in catchments, which generate saturation overland
flow. The smallest peak runoff was similar to the largest value in catchments, which generate subsurface storm
flow (Figure 4). Numerical simulations concurred with Figure 4, indicating that soil pipes can increase the
peak discharge from the hillslope and decrease the peak lag time of the storm hydrograph (Barcelo and Nieber,
1981, 1982; Himeno, 1997; Kato, 1998), when compared to the unpiped hillslope.

EFFECTS OF PIPEFLOW ON SLOPE STABILITY


Soil pipes in landslide scars
Pipeflow may contribute to landslide initiation and/or slope stability, since the spatial variation in hydrologic
response is attributed to the influence of pipeflow (e.g. Pierson, 1980; Brand et al., 1986; Jenkins et al., 1988;
Sidle, 1994; Fannin et al., 2000). Ohta et al. (1981) and Pierson (1983) reported that 50–90% of landslide
scars had soil pipes in headscarps. Ohta et al. (1981) also reported that the mean number of soil pipe outlets
in a scar of landslide was 3Ð57.
As shown in Table VI, soil pipe outlets were often found in scars of shallow landslides which occurred
during heavy rainfalls in Japan (Shindo, 1983; Tanaka, 1989; Takaya, 1995; Miyagi et al., 1998; Kaibori
et al., 1999; Hiramatsu et al., 1999a, b). This table indicated that soil pipes could be seen in scars of various
magnitude landslide (25–3000 m3 ), geological condition (crystalline schist, sedimentary rock, tuff, basalt and
granite) and slope gradient (0Ð2–0Ð8). The diameter of soil pipes in landslide scars (1–30 cm) was similar
to that commonly found in forested hillslopes (Table I). Both ephemeral and perennial pipeflows occurred in
the landslide scars (Table VI).

Table VI. Observation of soil pipes at scars of landslidea

Site Date Pipeflow Magnitude of landslide Geology Angle References

Flow D Volume Width Length Depth


type (cm) (m3 ) (m) (m) (m)

Kitatakami, 24 Sep. 1999 pe — — 15 20 2 crystalline 0Ð84 Hiramatsu et al.


Kochi schist (1999a)
Nishiiya, 29 Jun. 1999 ep 5–10 3000 20 50 3 crystalline 0Ð73 Hiramatsu et al.
Tokushima schist (1999b)
Okedera, 29 Jun. 1999 ep — — — — 1 granite 0Ð43 Kaibori et al.
Hiroshima (1999)
Nishigo, 27 Aug. 1998 ep 1–10 2500 — — — tuff 0Ð31 Miyagi et al.
Fukushima (1998)
Nishigo, 27 Aug. 1998 ep 1–10 1000 — — — tuff 0Ð21 Miyagi et al.
Fukushima (1998)
Nishigo, 27 Aug. 1998 ep 1–10 40 — — — tuff 0Ð20 Miyagi et al.
Fukushima (1998)
Nishigo, 27 Aug. 1998 ep — — 15 30 2–3 tuff 0Ð27 Miyagi et al.
Fukushima (1998)
Ryugamizu, 6 Aug. 1993 ep 30? — 30 100 0Ð5–5 basalt 0Ð73 Takaya (1995)
Kagoshima
Inabu, Aichi 30 Jun. 1985 pe 5 25 5–5Ð3 4Ð8–5Ð3 — granite 0Ð73 Tanaka (1989)
Naganuma, 12 Sep. 1982 ep — — 3–20 — — mudstone 0Ð62 Shindo (1983)
Tokyo
Nishimikawa, Jul. 1972 — 17Ð5 — — — — granite — Ohta et al.
Aichi (1981)

a Abbreviations: D, diameter; pe, perennial; ep, ephemeral.

Copyright  2001 John Wiley & Sons, Ltd. Hydrol. Process. 15, 2151– 2174 (2001)
2168 T. UCHIDA, K. KOSUGI AND T. MIZUYAMA

Effects of pipeflow on development of perched groundwater conditions


A variety of processes have been suggested to explain the contributions of pipeflow on landslide initiation.
From the previous investigations, a schematic illustration describing the processes, which explained the effects
of pipeflow on the landslide initiation, is shown in Figure 5.
The effects of pipeflow on runoff generation, which control the occurrence of landslides or otherwise, can
be classified into (1) the concentration of water into the soil pipe network (the first effect) (Pierson, 1983;
Selby, 1993; Onda, 1994) and (2) the rapid drainage of water to downslope (the second effect) (Mosely,
1982; Kitahara, 1989; Tanaka, 1992) (Figure 5). These effects suggested that pipeflows may contribute to
slope stability by increasing the rate of soil drainage and limiting the development of perched groundwater
conditions, just as field drainage pipes facilitate drainage on agricultural land (e.g. Jenkins et al., 1988; Johnson
and Sitar, 1990; McDonnell, 1990b; Van Asch et al., 1999).
Under normal rainfall conditions, the rate of water concentration to the soil pipe network may be smaller
than the limited drainage capacity of the soil pipe, which is determined by the soil pipe morphological
characteristics. In this case, the perched water table may be quickly dissipated by pipeflow (McDonnell,
1990b). Bench scale experiments with artificial soil pipes support this suggestion, indicting that the soil pipes
may decrease the slope failure susceptibility by enhancing the effective slope drainage and the pore water
dissipation (Tani et al., 1987; Sidle, 1994; Sidle et al., 1995b; Uchida et al., 1995b) (Figure 5). However,
when the soil pipe was blocked or was a deadend passageway (‘closed ended pipe’), the limited drainage
capacity of the soil pipe (the second hydrologic effect of soil pipe) may be very small. Thus, it can be
thought that the cavity of the soil pipe could readily be filled with water during a rain event, increasing
pore water pressure in the surrounding matrix, especially the lower end of the closed ended pipe. This
suggestion was evidenced by both the numerical simulation (Watanabe and Imai, 1984) and the bench scale
experiments (Pierson, 1983; Uchida et al., 1995b). Moreover, Sidle (1984) and Brand et al. (1986) observed
the asymmetry response of soil pore water pressure (rapid rise and gentle recession) in hillslopes. They

Soil pipe network development Biotic activity

Water concentration into soil pipe


Rapid drainage
-expansion of source area

Pipe erosion

Rate of water concentration Rate of water concentration


larger than the limited smaller than the limited
drainage capacity drainage capacity
Block of soil pipe
- decreased drainage
capacity

Factors of landslide initiation Factors of slope stability


- increased soil pore water - dissipated soil pore water
pressure in the surrounding pressure
matrix
Figure 5. Schematic diagram describing the processes which explained effects of pipeflow on the landslide initiation

Copyright  2001 John Wiley & Sons, Ltd. Hydrol. Process. 15, 2151– 2174 (2001)
EFFECTS OF PIPEFLOW ON HYDROLOGICAL PROCESS AND LANDSLIDES 2169

suggested that the rapid increase of soil pore water pressure, which can trigger landslides, is due to the closed
ended pipes.
Under the heaviest rainfall, the rate of water concentration to the soil pipe network (the first effect) is in
excess of the pipeflow transmission capacity (the second effect). In this case, the soil pipe induced the slope
instability by the rapid increase in the soil pore water pressure surrounding the soil pipe network, although
the lower ends of the soil pipes were not blocked (McDonnell, 1990b; Miyagi et al., 1998). As a result of a
build-up of pore water pressure, pipeflow may produce upthrust pressures (Roger and Selby, 1980).
In low permeable soils, soil pipes may not effectively contribute to the rapid dissipation of elevated pore
pressure (McDonnell and Taratoot, 1995). They suggested that the soil pipe was ineffective in reducing pore
water pressure build-up, since the opportunity to feed water to the soil pipe system (the first hydrological
effect of soil pipe) was low. That is, soil pipes contribute to the effective soil drainage system only when two
hydrological effects (concentration and rapid drainage) of soil pipe are well combined.

Effects of pipe erosion on slope stability


Once a pipe outlet is formed, convergence of groundwater flow into the soil pipe occurs (Pierson,1983;
Selby, 1993; Onda, 1994). When the water flow becomes a non-Darcian flow as it enters the soil pipes, it
accelerates (Kitahara, 1989). As a consequence, the collapse of the pipe wall and roof and sediment discharge
from the soil pipe may occur (Selby, 1993; Terajima et al., 1993). Numerical simulation showed that contrasts
in hydraulic conductivity have the most pronounced and diverse effects on groundwater seepage force (Reid
and Iverson, 1992). This simulation result suggested that the soil pipe may have a large impact on groundwater
erosion. Indeed, sediment yields from the soil pipes are commonly found in forested catchments (Ziemer,
1992; Sonoda, 1993; Terajima and Sakura, 1993; Onda, 1994; Ogawa, 1997; Uchida et al., 1999a).
Although very few studies have demonstrated the relationship between the sediment discharge from the
soil pipe and the change of the pipeflow response to storms, it has been predicted that soil pipes modify their
hydraulic properties by groundwater erosion to produce an efficient drainage network (Wilson and Smart,
1984; Tsukamoto et al., 1988; Terajima et al., 1996b). The sediment discharge from the soil pipe causes
an increase in the cavity volume of the soil pipe. Both the numerical simulations (Barcelo and Nieber,
1981, 1982; Nieber and Warner, 1991) and the bench scale experiment with artificial soil pipes (Sidle et al.,
1995b) showed that the peak runoff increased as the cavity volume of soil pipe increased, supporting the
previous predictions (Wilson and Smart, 1984; Tsukamoto et al., 1988). As mentioned previously, recent field
observations indicated that the sediment discharge from soil pipes contributed to an increase in the limited
drainage capacity (Uchida et al., 1999a). According to these indications, the soil erosion by pipeflow may
contribute to the slope stability by enhancing the effective slope drainage.
In contrast, some studies have shown that an increase in the cavity volume in soil layers by pipe erosion
may foster a reduction in the slope stability (e.g. Onodera, 1990; Shindo, 1993; Onda, 1992, 1994). Onda
(1992) reported that pipe erosion causes small slope failures, which were much smaller in size than landslides
which occurred during a heavy rainfall. He concluded that the repetition of pipe erosion and resulting small
slope failure played important roles in a geomorphic evaluation of convex slope.
Harp et al. (1990) performed artificial irrigation experiments on natural slopes to investigate variations
in soil pore water pressure during induced slope failure. Piezometer records showed a consistent trend of
increasing pressure during the early stages of infiltration and abrupt decreases in pressure from 5 to 50 min
before failure. They suggested that the dramatic drop of pore water pressure prior to landslide movement
may be due to pipe erosion. Harp’s suggestion accords with Uchida et al. (1999a), indicating pipe erosion
enhanced the effective slope drainage and the pore water dissipation. Although soil pipes may facilitate rapid
drainage, the landslide still occurred. That is, it can be thought that in Harp’s experiment, the former effect
of pipe erosion, reported by Wilson and Smart (1984) and Uchida et al. (1999a), is less than the latter effects,
reported by Onda (1992) and Shindo (1993). Thus, the soil pipes may play a contributing factor in landslides.
Onda (1994) and Zhu (1997) demonstrated that the collapse of the pipe wall causes clogging of a pipe
outlet, so that a new pipe outlet is formed nearby. However, if the new soil pipe outlet is not formed, the

Copyright  2001 John Wiley & Sons, Ltd. Hydrol. Process. 15, 2151– 2174 (2001)
2170 T. UCHIDA, K. KOSUGI AND T. MIZUYAMA

limited drainage capacity of the soil pipe declines (Terajima et al., 1996b, 1997). Consequently, the soil pipe
may become a contributing factor to landslides (Ohta et al., 1981; Uchida et al., 1996), like a ‘closed ended
pipe’ (Figure 5).

CONCLUSIONS
The aim of this study was to review some studies of pipeflow hydrology in forested steep hillslopes, where
adequate information is lacking with regard to the effects of pipeflow on storm runoff generation process.
Thus, several points were clarified.

(1) The maximum discharge rate of pipeflow was mainly determined by the soil pipe diameter. When the
conditions of both the soil pipe diameter and the slope gradient in forest soil were similar to those in
peaty podzol, the maximum rate of pipeflow in forest slopes was slighter than that in peaty podzol.
(2) Soil pipes in hillslopes can increase the peak discharge and decrease the peak lag time of the storm
hydrograph. However, at baseflow, the effects of pipeflow on streamflow are commonly small. The
contribution of pipeflow to streamflow increases with the increase of rainfall magnitude, since the
contributing area of pipeflow increases as the soil layer becomes wet.
(3) Pipeflow is delivered from several sources. In some cases, runoff water from soil pipes flows only through
macropores without interacting with the soil matrix. In other cases, pipeflow may be delivered from a
bedrock fracture. Therefore, the dominant contributor of pipeflow varied.
(4) The roles of pipeflow in storm runoff generation processes were classified into two processes:
(i) concentration of water into the soil pipe network and (ii) rapid drainage of water to downslope. The
soil pipes contribute to the slope stability by increasing the rate of soil drainage and limiting the devel-
opment of perched groundwater condition, only when these two hydrological effects of soil pipe are well
combined. However, if the rate of water concentration to the soil pipe network is in excess of the pipeflow
transmission capacity, the cavity of the soil pipe could readily be filled with water during a rain event,
increasing pore water pressure in the surrounding matrix. In this case, the soil pipe induced the slope
instability.

ACKNOWLEDGEMENTS

We thank Drs Makoto Tani, Masaharu Fujita and Nobuhito Ohte (all of Kyoto University) for their valuable
suggestions. We also thank Dr. Koichiro Kuraji (Tokyo Institute of Technology) for his valuable comments.

REFERENCES
Anderson MG, Burt TP. 1990. Subsurface runoff. In Process Studies in Hillslope Hydrology, Anderson MG, Burt TP (eds). Wiley: Chichester;
365–400.
Anderson SP, Dietrich WE, Montgomery DR, Torres R, Conrad ME, Loague K. 1997. Subsurface flow paths in a steep unchanneled
catchment. Water Resources Research 33: 2637– 2653.
Barcelo MD, Nieber JL. 1981. Simulation of the hydrology of natural pipes in a soil profile. American Society of Agricultural Engineering,
Paper No. 81-2026.
Barcelo MD, Nieber JL. 1982. Influence of a soil pipe network on catchment hydrology. American Society of Agricultural Engineering, Paper
No. 82-2027.
Beven KJ, Germann PJ. 1982. Macropores and water flow in soils. Water Resources Research 18: 1311– 1325.
Bonell M. 1998. Selected challenges in runoff generation research in forests from the hill-slope to headwater drainage basin scale. Journal
of American Water Resources Association 34: 765– 785.
Bouma J. 1981. Soil morphology and preferential flow along macropore. Agricultural Water Management 3: 235–250.
Brand EW, Dale MJ, Nash JM. 1986. Soil pipes and slope stability in Hong Kong. Quarterly Journal of Engineering Geology 19: 301–303.
Bryan RB, Jones JAA. 1997. The significance of soil piping process: inventory and prospect. Geomorphology 20: 209–218.

Copyright  2001 John Wiley & Sons, Ltd. Hydrol. Process. 15, 2151– 2174 (2001)
EFFECTS OF PIPEFLOW ON HYDROLOGICAL PROCESS AND LANDSLIDES 2171

Burns DA, Hooper RP, McDonnell JJ, Freer JM, Kendall C, Beven K. 1998. Base cation concentrations in subsurface flow from a forested
hillslope: the role of flushing frequency. Water Resources Research 34: 3535– 3544.
Buttle JM. 1994. Isotope hydrograph separations and rapid delivery of pre-event water from drainage basins. Progress in Physical Geography
18: 16–41.
Buttle JM, Turcotte DS. 1999. Runoff processes on a forested slope on the Canadian Shield. Nordic Hydrology 30: 1–20.
Chapman PJ, Reynolds B, Wheater HS. 1993. Hydrochemical changes along stormflow pathways in a small moorland headwater catchment
in Mid-Wales, UK. Journal of Hydrology 151: 241– 265.
Dunne T. 1978. Field studies of hillslope flow process. In Hillslope Hydrology, Kirkby MJ (ed.). Wiley: Chichester; 227– 293.
Dunne T. 1990. Hydrology, mechanics, and geomorphic implications of erosion by subsurface flow. In Groundwater Geomorphology,
Higgins CC, Coate DR (eds). Geological Society of America: Boulder, CO; 1–28.
Elsenbeer H, Lack A. 1996. Hydrometric and hydrochemical evidence for fast flowpaths at La Cuenca, western Amazonia. Journal of
Hydrology 180: 237–250.
Elsenbeer H, Lack A, Cassel K. 1995. Chemical fingerprints of hydrological compartments and flow paths at La Cuenca, western Amazonia.
Water Resources Research 31: 3051– 3058.
Faeh AO. 1997. Understanding the processes of discharge formation under extreme precipitation, A study based on the numerical simulation
of hillslope experiments. Mitteilungen 150, Versuchsanstalt für Wasserbau, Hydrologie und Glaziologie der Eidgenössischen Technischen
Hochschule: Zürich; 197.
Fannin RJ, Jaakkola J, Wilkinson JMT, Hetherington ED. 2000. Hydrologic response of soils to precipitation at Carnation Creek, British
Columbia, Canada. Water Resources Research 36: 1481– 1494.
Garland G, Humphrey B. 1992. Field measurements of discharge and sediment yield from a soil pipe in the Natal Drakensberg, South
Africa. Zeitschrift für Geomorphologie N.F. 36: 15–23.
Germann PF. 1990. Macropores and hydrologic hillslope processes. In Process Studies in Hillslope Hydrology, Anderson MG, Burt TP
(eds). Wiley: Chichester; 327–364.
Hagedorn F. 1999. Transformation and transport of nitrogen in subalpine catchment: effects of elevated nitrogen deposition, Dissertation
ETH No. 13167, Zürich; 145.
Harp EL, Wells WG, Sarmiento JG. 1990. Pore pressure response during failure in soils. Geological Society of America Bulletin 102:
428–438.
Hill AR, Kemp WA, Buttle JM, Goodyear D. 1999. Nitrogen chemistry of subsurface storm runoff on Canadian Shield hillslope. Water
Resources Research 35: 811– 821.
Himeno J. 1997. Analysis of runoff in a hillslope—the effects of pipe flow. Master’s thesis, Nagoya University; 23 (in Japanese with English
summary).
Hiramatsu S, Ishikawa Y, Miyoshi I, Chiba A. 1999a. Actual state of sediment-related disaster on September 24–25, 1998 in Kochi
Prefecture, Japan (Prompt report). Journal of the Japan Society of Erosion Control Engineering 51(5): 43– 47 (in Japanese with English
summary).
Hiramatsu S, Ishikawa Y, Osanai N, Miyoshi I. 1999b. Report of sediment-related disaster on June 29, 1999 in Nishiyayama Village
Tokushima Prefecture, Japan (Prompt report). Journal of the Japan Society of Erosion Control Engineering 52(3): 44–49 (in Japanese).
Hirose T, Onda Y, Matsukura Y. 1994. Runoff and solute characteristics in four small catchments with different bedrocks in Abukuma
mountains, Japan. Transactions, Japanese Geomorphological Union 15A: 31–48.
Hornberger GM, Beven K, Germann P. 1990. Influence about solute transport in macropores forest soils from time series models. Geoderma
46: 249–262.
Iverson RM, Reid ME, LaHusen RG. 1997. Debris-flow mobilization from landslide. Annual Reviews of Earth and Planetary Science 25:
85–138.
Jenkins A, Ashworth PJ, Ferguson RI, Grieve IC, Rowling P, Stott TA. 1988. Slope failures in the Ochil hills, Scotland, November 1984.
Earth Surface Processes and Landforms 13: 69–76.
Johnson KA, Sitar N. 1990. Hydrologic conditions leading to debris-flow initiation. Canadian Geotechnical Journal 27: 789–801.
Jones JAA. 1971. Soil piping and stream channel initiation. Water Resources Research 7: 602–610.
Jones JAA. 1981. In The Nature of Soil Piping—A Review of Research. GeoBooks: Norwich; 301.
Jones JAA. 1987. The effects of soil piping on contributing area and erosion pattern. Earth Surface Processes and Landforms 12: 229–248.
Jones JAA. 1997a. Pipeflow contribution areas and runoff response. Hydrological Processes 11: 35–41.
Jones JAA. 1997b. The role of natural pipeflow in hillslope drainage and erosion: extrapolating from the Maesnant data. Physics and
Chemistry of the Earth 22: 303–308.
Jones JAA, Crane FG. 1984. Pipeflow and pipe erosion in the Maesnant experimental catchment. In Catchment Experiments in Fluvial
Geomorphology, Burt TP, Walling DE (eds). GeoBooks: Norwich; 55– 72.
Jones JAA, Richadson JM, Jacob HJ. 1997. Factors controlling the distribution of piping in Britain: a reconnaissance. Geomorphology 20:
289–306.
Kaibori M, Ishikawa Y, Ushiyama M, Kubota T, Hiramatsu S, Fujita M, Miyoshi I, Yamashita Y. 1999. Debris flow and slope failure
disasters in Hiroshima Prefecture caused heavy rainfall in June 1999 (Prompt report). Journal of the Japan Society of Erosion Control
Engineering 52(3): 34–43 (in Japanese with English summary).
Kato K. 1998. A study of the hillslope percolation process of rain water concerning the effects of small scale terrain. Master’s thesis, Nagoya
University: 32 (in Japanese with English summary).
Katsuyama M, Ohte N, Uchida T, Asano Y, Kimoto A. 2000. Effects of the difference of hydrological processes on the streamwater
chemistry. Journal of Japan Society of Hydrology and Water Resources 14: 227– 239 (in Japanese with English summary).
Keppeler E, Brown D. 1998. Subsurface drainage processes and management impacts. In Proceedings of the Conference on Coastal
Watersheds: The Caspar Creek Story, Ziemer RR. (ed.). Pacific Southwest Research Station, Forest Service, US Department of Agriculture:
25–34.

Copyright  2001 John Wiley & Sons, Ltd. Hydrol. Process. 15, 2151– 2174 (2001)
2172 T. UCHIDA, K. KOSUGI AND T. MIZUYAMA

Kirkby MJ. 1988. Hillslope runoff process and models. Journal of Hydrology 100: 315– 339.
Kitahara H. 1989. Characteristics of pipe flow in a subsurface soil layer on a gentle slope (II). Hydraulic properties of pipes. Journal of
Japanese Forestry Society 70: 317– 322 (in Japanese with English summary).
Kitahara H. 1994. A study on the characteristics of soil pipes influencing water movement in forested slopes. Bulletin of the Forestry and
Forest Products Research Institute 367: 63–115 (in Japanese with English summary).
Kitahara H. 1996. The research history of pipe flow and macropore. Water Science 227: 81–114 (in Japanese).
Kitahara H, Nakai Y. 1992. Relationship of pipe flow in to streamflow on a first order watershed. Journal of Japanese Forestry Society 74:
49–52 (in Japanese).
Kitahara H, Terajima T, Nakai Y. 1994. Ratio of pipe flow to through flow. Journal of Japanese Forestry Society 76: 10–17 (in Japanese
with English summary).
Koyama K. 1994. Observation of pipeflow at a volcanic headwater catchment. Master’s thesis, Tottori University; 79 (in Japanese).
Lakey B, Krothe NC. 1997. Stable isotopic variation of storm discharge from a perennial karst spring, Indiana. Water Resources Research
32: 721–731.
Leibundgut C, McDonnell JJ, Schultz G. 1999. In Integrated Methods of Catchment Hydrology—Tracer, Remote Sensing and New Hydro-
metric Techniques. International Association of Hydrological Sciences Publication 258, Wallingford; 284.
Loukas A, Quick MC. 1993. Hydrologic behaviour of a mountainous watershed. Canadian Journal of Civil Engineering 20: 1–8.
Luxmoore RJ, Ferrand LA. 1993. Towards a pore-scale analysis of preferential flow and chemical transport. In Water Flow and Solute
Transport in Soils, Russo D, Dagan G (eds). Springer-Verlag: Berlin; 45– 60.
Luxmoore RJ, Jardine PM, Wilson GV, Jones JR, Zelazny LW. 1990. Physical and chemical controls of preferred path flow through a
forested hillslope. Geoderma 46: 139– 154.
McCaig M. 1983. Contributions to storm quickflow in a small head-water catchment. The role of natural pipes and soil macropores. Earth
Surface Processes and Landforms 8: 239–252.
McDonnell JJ. 1990a. A rationale for old water discharge through macropores in a steep, humid catchment. Water Resources Research 26:
2821– 2832.
McDonnell JJ. 1990b. The influence of macropores on debris flow initiation. Quarterly Journal of Engineering Geology 23: 325– 331.
McDonnell JJ, Buttle JB. 1998. Comments on ‘A deterministic-empirical model of the effect of capillary-fringe on near-stream area runoff.
I. Description of the model’ by Jayatilaka CJ, Gillham RW. Journal of Hydrology 207: 280– 285.
McDonnell JJ, Taratoot M. 1995. Soil pipe effects of pore pressure redistribution in low permeability soils. Geotechnical Engineering 26:
53–61.
Miyagi T, Furuya T, Umemura J, Chiba N, Marui H, Chigira M. 1998. Prompt report of sediment-related disaster in south Fukushima
Prefecture caused by heavy rainfall on August, 1998. Journal of Japan Landslide Society 35(2): 91–98 (in Japanese).
Montgomery DR, Dietrich WE. 1995. Hydrologic processes in a low-gradient source area. Water Resources Research 31: 1–10.
Mosely MP. 1982. Subsurface flow velocities through selected forest soils, South Island, New Zealand. Journal of Hydrology 55: 65–92.
Mulholland PJ. 1993. Hydrometric and stream chemistry evidence of three storm flowpaths in Walker Branch watershed. Journal of Hydrology
151: 291– 316.
Muscutt AD, Wheater HS, Reynolds B. 1990. Stormflow hydrochemistry of a small Welsh upland catchment. Journal of Hydrology 116:
239–249.
Navar J, Turton D, Miller EL. 1995. Estimating macropore and matrix flow using the hydrograph separation procedure in an experimental
forest plot. Hydrological Processes 9: 743–753.
Newman BD, Campbell AR, Wilcox BP. 1998. Lateral subsurface flow pathways in a semiarid ponderosa pine hillslope. Water Resources
Research 34: 3485– 3496.
Nieber JL, Warner GS. 1991. Soil pipe contribution to steady subsurface stormflow. Hydrological Processes 5: 329– 344.
Noguchi S, Tsuboyama Y, Sidle RC, Hosoda I. 1997. Spatial distributed morphological characteristics of macropores in forested soil of
Hitachi Ohta experimental watershed, Japan. Journal of Forest Research 2: 207– 215.
Noguchi S, Tsuboyama Y, Sidle RC, Hosoda I. 1999. Morphological characteristics of macropores and distribution of preferential flow
pathways in a forested slope segment. Soil Science Society of America Journal 63: 1413– 1423.
Ogawa K. 1997. Study of profile patterns and moisture variations in surface soil layers on slopes of a mountain region. Research Bulletin
of Hokkaido University Forest 54: 87–141 (in Japanese with English summary).
Ohta T, Tsukamoto Y, Noguchi H. 1981. An analysis of pipeflow and landslide. In Proceedings of Annual Meeting of the Japan Society of
Erosion Control Engineering; 92– 93 (in Japanese).
Ohta T, Noguchi H, Tsukamoto Y. 1983. Study on the behavior of storm water in a small forested watershed (IV)—Role of pipeflow on
runoff generation process of storm water. Transaction of the 93rd Meeting of the Japanese Forestry Society; 459–461 (in Japanese).
Onda Y. 1992. Influence of water storage capacity in the regolith on hydrological characteristics, slope processes, and slope form. Zeitschrift
für Geomorphologie N.F. 36: 165– 178.
Onda Y. 1994. Seepage erosion and its implication to the formation of amphitheater valley head: a case study at Obara, Japan. Earth Surface
Processes and Landforms 19: 627– 640.
Onda Y, Itakura N. 1997. An experimental study on the burrowing activity of river crabs on subsurface water movement and piping.
Geomorphology 20: 279– 288.
Onda Y, Okunishi K, Iida T, Tsujimura M. 1996. In Hydrogeomorphology: The Interaction of Hydrological and Geologic Processes.
Kokonshoin: Tokyo; 267 (in Japanese).
Onda Y, Komatsu Y, Tsujimura M, Fujiwara J. 1999. Possibility for predicting landslide occurrence by analyzing the runoff peak response
time. Journal of the Japan Society of Erosion Control Engineering 51(5): 48–52 (in Japanese).
Onodera S. 1990. Discharge process of the subsurface water and its hydrogeomorphological characteristics in a hillslope. Transactions,
Japanese Geomorphological Union 11: 369– 385 (in Japanese with English summary).

Copyright  2001 John Wiley & Sons, Ltd. Hydrol. Process. 15, 2151– 2174 (2001)
EFFECTS OF PIPEFLOW ON HYDROLOGICAL PROCESS AND LANDSLIDES 2173

Pearce AJ, Stewart MK, Sklash MG. 1986. Storm runoff generation in humid headwater catchments 1. Where does the water come from?
Water Resources Research 22: 1263– 1272.
Peters DL, Buttle JB, Taylor CH, LaZerte BD. 1995. Runoff production in a forested, shallow soil, Canadian Shield basin. Water Resources
Research 31: 1291– 1304.
Pierson TC. 1980. Piezometric response to rainstorms in forested hillslope drainage depressions. Journal of Hydrology (NZ) 19: 1–10.
Pierson TC. 1983. Soil pipes and slope stability. Quarterly Journal of Engineering Geology 16: 1–11.
Reid ME, Iverson RM. 1992. Gravity-driven groundwater flow and slope failure potential. 2. Effects of slope morphology, material properties,
and hydraulic heterogeneity. Water Resources Research 22: 939– 950.
Roberge J, Plamondon AP. 1987. Snowmelt runoff pathway in a boreal forest hillslope, the role of pipe throughflow. Journal of Hydrology
95: 39–54.
Roger NW, Selby MJ. 1980. Mechanisms of shallow translational landslide during summer rainstorm: North Island, New Zealand.
Geografiska Annaler 62A: 11–21.
Selby MJ. 1993. In Hillslope Materials and Processes. Oxford University Press: Oxford; 451.
Shindo S. 1983. Landslide and groundwater dynamics at valley heads. Report for the Fund of Monbusho for Scientific Research (57025009);
94 (in Japanese).
Shindo S. 1993. Convergent flow of subsurface water concerned with slope failure. The Quaternary Research 32: 315– 322 (in Japanese
with English summary).
Sidle RC. 1984. Shallow groundwater fluctuations in unstable hillslopes coastal Alaska. Zeitschrift für Gletscherkunde und Glazialgeologie
20: 79–95.
Sidle RC. 1994. Hydrological process related to landslide initiation. In Proceeding of the SABO Workshop on Experimental Research on
Landslide Mechanism; 19–27.
Sidle RC, Pearce AJ, O’Loughlin CL. 1985. In Hillslope Stability and Land Use. American Geophysical Union: Washington, DC; 140.
Sidle RC, Tsuboyama Y, Noguchi S, Hosoda I, Fujieda M, Simizu T. 1995a. Seasonal hydrologic response at various spatial scales in a
small forested catchment, Hitachi Ohta. Journal of Hydrology 168: 227–250.
Sidle RC, Kitahara H, Terajima T, Nakai Y. 1995b. Experimental studies on the effects of pipeflow on throughflow partitioning. Journal of
Hydrology 165: 207–219.
Sidle RC, Tsuboyama Y, Noguchi S, Hosoda I, Fujieda M, Simizu T. 2000. Stormflow generation in steep forested headwater a linked
hydrogeomorphic paradigm. Hydrological Processes 14: 369– 385.
Sklash MG, Stewart MK, Pearce AJ. 1986. Storm runoff generation in humid headwater catchments 2. A case study of hillslope and
low-order stream response. Water Resources Research 22: 1273– 1282.
Sklash MG, Beven KJ, Gilman KK, Darling WG. 1996. Isotope studies of pipeflow at Plynlimon, Wales, UK. Hydrological Processes 10:
921–944.
Sonoda M. 1993. Seasonal change of sediment discharge in a forested small basin in Japan. Transactions, Japanese Geomorphological Union
14: 53–70 (in Japanese with English summary).
Stocking M. 1981. A model of piping in soils. Transactions, Japanese Geomorphological Union 2: 263– 278.
Takaya S. 1995. On debris flow disasters on steep slopes facing the Kagoshima Bay, caused by the torrential rain of August 6, 1993, Kyushu.
Journal of the Japan Society of Erosion Control Engineering 47(6): 24–27 (in Japanese).
Tanaka T. 1989. A study of pipe flow measured on a surface failed hillslope. Journal of the Japan Society of Erosion Control Engineering
41(5): 14–22 (in Japanese with English summary).
Tanaka T. 1992. Storm runoff processes in a small forested drainage basin. Environmental. Geology and Water Sciences 19: 179– 191.
Tanaka T. 1996. Recent progress in Japanese studies on storm runoff processes. Geographical Reviews of Japan 69B: 144– 159.
Tanaka T, Yasuhara M, Sakai H, Marui A. 1988. The Hachioji experimental basin study—storm runoff processes and the mechanism of its
generation. Journal of Hydrology 102: 139– 164.
Tani M. 1987. On researching effects of forest soil on runoff properties. Water Science 177: 35– 61 (in Japanese).
Tani M. 1997. Runoff generation processes estimated from hydrological observation on a steep forested hillslope with a thin soil layer.
Journal of Hydrology 200: 84–109.
Tani M, Abe T. 1996. Recession characteristics of runoff produced from a soil layer a hillslope with widely-developed preferential pathways.
Journal of Japan Society of Hydrology and Water Resources 9: 425–437 (in Japanese with English summary).
Tani M, Abe T, Kobayashi C, Hattori S. 1987. Effects of underdrains on runoff processes of a hillslope. Transaction of the 99th Meeting of
the Japanese Forestry Society; 751–752 (in Japanese).
Tani M, Kosugi K, Tsuboyama Y, Kubota J. 1998. Toward understanding effects of forest soil properties on characteristics of runoff from
catchment with various kinds of heterogeneities. Journal of Japanese Forestry Society 80: 44–57 (in Japanese).
Terajima T, Sakura Y. 1993. Effect of subsurface flow on a topographic change at a valley head of granitic mountains. Transactions,
Japanese Geomorphological Union 14: 365– 384 (in Japanese with English summary).
Terajima T, Kitahara H, Nakai Y. 1993. Experimental consideration related to the effects of subsurface flow on subsurface hydraulic erosion
(I). Relationship between the variety of soil pipe length and the hydraulic action of subsurface flow. Transaction of the 104th Meeting of
the Japanese Forestry Society; 701–704 (in Japanese).
Terajima T, Kitahara H, Sakamoto T, Nakai Y, Kitamura K. 1996a. Pipe flow significance on subsurface discharge from the valley head of
a small watershed. Journal of Japanese Forestry Society 78: 20–28 (in Japanese with English summary).
Terajima T, Sakamoto T, Nakai Y, Kitamura K. 1996b. Subsurface discharge and suspended sediment yield interactions in valley head of a
small forested watershed. Journal of Forest Research 1: 131– 137.
Terajima T, Sakamoto T, Nakai Y, Kitamura K. 1997. Suspended sediment discharge in subsurface flow from the head hollow of a small
forested watershed, northern Japan. Earth Surface Processes and Landforms 22: 987–1000.
Tsuboyama Y, Sidle RC, Noguchi S, Hosoda I. 1994. Flow and solute transport through the soil matrix and macropores of a hillslope
segment. Water Resources Research 30: 879– 890.

Copyright  2001 John Wiley & Sons, Ltd. Hydrol. Process. 15, 2151– 2174 (2001)
2174 T. UCHIDA, K. KOSUGI AND T. MIZUYAMA

Tsukamoto Y. 1997. In Conservation of Forest, Water and Soil . Asakurashoten: Tokyo; 134 (in Japanese).
Tsukamoto Y, Ohta T. 1988. Runoff process on a steep forested slope. Journal of Hydrology 102: 165– 178.
Tsukamoto Y, Minematsu H, Tange I. 1988. Pipe development in hillslope soils in humid climate. Rolling Land Research 6: 268–280 (in
Japanese with English summary).
Turton DJ, Barnes Jr DR, Navar JJ. 1995. Old and new water in subsurface flow from a forest soil block. Journal of Environmental Quality
24: 139–146.
Uchida T, Kosugi K, Kobashi S, Mizuyama T. 1995a. Observation and survey of underground pipes at Toinotani Basin in Ashu Experimental
Forest. Bulletin of Kyoto University Forest 67: 58–67 (in Japanese with English summary).
Uchida T, Kosugi K, Ohte N, Mizuyama T. 1995b. An experimental study on influence of soil pipe for groundwater level. Transaction of
the 106th Meeting of the Japanese Forestry Society; 505–508 (in Japanese).
Uchida T, Kosugi K, Ohte N, Mizuyama T. 1996. The influence of pipe flow on slope stability. Journal of Japan Society of Hydrology and
Water Resources 9: 330– 339 (in Japanese with English summary).
Uchida T, Kosugi K, Mizuyama T. 1997. Analysis of the relationship between groundwater level and discharge rate of pipe flow at a valley
head. Journal of Japanese Forestry Society 79: 202–210 (in Japanese with English summary).
Uchida T, Kosugi K, Mizuyama T. 1999a. Runoff characteristics of pipeflow and effects of pipeflow on rainfall-runoff phenomena in a
mountainous watershed. Journal of Hydrology 222: 18– 36.
Uchida T, Kosugi K, Mizuyama T, Katsuyama M. 1999b. Effects of pipeflow on runoff generation process at a headwater catchment in
Toinotani, Kyoto University forest in Ashiu. Forest Research, Kyoto 71: 55–64 (in Japanese with English summary).
Van Asch ThWJ, Buma J, Van Beek LPH. 1999. A view of some hydrological triggering systems in landslides. Geomorphology 30: 25– 32.
Walsh RPD, Howells KA. 1988. Soil pipes and their role in runoff generation and chemical denudation in humid tropical catchment in
Dominica. Earth Surface Processes and Landforms 13: 9–17.
Watanabe K, Imai H. 1984. Transient behavior of groundwater flow in idealized slope having a soil pipe in it. Journal of the Japan Society
of Engineering Geology 25: 1–8 (in Japanese with English summary).
Weiler M, Naef F, Leibundgut C. 1998. Study of runoff generation on hillslope using tracer experiments and a physically-based numerical
hillslope model. International Association of Hydrological Sciences Publication 248: 353– 360.
Wieczorek GF. 1993. Assessment and prediction of debris-flow hazards. Proceedings of Hydraulic Engineering’93 2: 1272– 1283.
Wilson CJ, Dietrich WE. 1987. The contribution of bedrock groundwater flow to storm runoff and high pore water pressure development
in hollows. International Association of Hydrological Sciences Publication 165: 49–59.
Wilson CM, Smart PL. 1984. Pipes and pipe flow process in an upland catchment, Wales. Catena 11: 145– 158.
Wilson GV, Jardine PM, Luxmoore RJ, Jones JR. 1990. Hydrology of forested watershed during storm events. Geoderma 46: 119–138.
Wilson GV, Jardine PM, Luxmoore RJ, Zelanzny LW, Lietzke DA, Todd DE. 1991. Hydrogeochemical processes controlling subsurface
transport from an upper subcatchment of Walker Branch watershed during storm events 1, Hydrologic transport process. Journal of
Hydrology 123: 297–316.
Woo M, diCenzo P. 1988. Pipe flow in James Bay coastal wetlands. Canadian Journal of Earth Science 25: 625–629.
Yasuhara M, Tanaka T, Marui A. 1984. Effect of hydrological factors on soil pipe development in two headwater basins in the Tama Hills,
Japan. Transactions, Japanese Geomorphological Union 5: 125–137 (in Japanese with English summary).
Zhu TX. 1997. Deep-seated, complex tunnel systems—a hydrological study in a semiarid catchment, Loess Plateau, China. Geomorphology
20: 255–267.
Ziemer, RR. 1992. Effects of logging in subsurface pipeflow and erosion: coastal northern California, USA. International Association of
Hydrological Sciences Publication 209: 187– 197.
Ziemer RR, Albright JS. 1987. Subsurface pipeflow dynamics of north-coastal California swale systems. International Association of
Hydrological Sciences Publication 165: 71–80.

Copyright  2001 John Wiley & Sons, Ltd. Hydrol. Process. 15, 2151– 2174 (2001)

Potrebbero piacerti anche