Process realism: flow paths and storage - Runoff Prediction in Ungauged Basins: Synthesis across Processes, Places and Scales - page 56

Figure 4.2. (a) Wadi Al-Khoudh in the coastal plains of Oman. Photo G. Al-Rawas. (b) Wadi Tiwi in the coastal mountains of Oman.
Photo R. Kirnbauer.

topography exert time invariant controls on gradients that
drive lateral flows (because each represents an inclined
material interface across which flow resistance changes
significantly). At the event scale, topography controls both
the directions of surface and subsurface flows and the
strength of the forces that drive these flows. This controls
flow redistribution at seasonal time scales, which in
turn may control soil moisture patterns (Western et al.,
1998a , b , 2002 ), which affect flow resistances in the soil
matrix, activation of apparent preferential flow paths
and separation of surface and subsurface flows. At longer
time scales of soil formation and landscape evolution,
topography indirectly controls water flow through co-
evolutionary soil (pedological) and vegetation (ecological)
processes. The connectivity of surface flow paths is essen-
tial for the actual stream runoff, as not all of the runoff
generated locally may reach a stream channel, but may re-
infiltrate along the way (Kirkby, 2005 ; Western et al.,
2001b ). Similarly, the connectivity of subsurface flow
paths is often highly important.
The soil characteristics play a key role in the partitioning
of rain water into surface and subsurface flow paths. At the
local scale, the partitioning can be seen as essentially a soil
physical problem. Soil physical characteristics may be
inferred from widely available data such as soil texture
(e.g., Wösten et al., 1999 ; Nyberg, 1995 ; Hernandez
et al., 2000 ). However, macropores and other preferential
flow paths often dominate the infiltration behaviour even
more than the characteristics of the soil matrix (Bouma
et al., 2011 ). Also, as one moves up in scale to hillslopes
and the catchment, the layering of the soil and its spatial
arrangement, in particular along hillslopes, become
increasingly relevant. Because of the co-evolution of soils
with vegetation and water movement within the catchment,

soils tends to be spatially organised, a notion represented in
the catena concept (Milne, 1935 ; Jenny, 1941 ).
In pristine areas the evolution of vegetation and soils is
strongly interlinked (Markart et al., 2004 ). Typically, soil
profiles evolve along the hillslopes driving mass, nutrient
and heat flows with feedbacks on the potential gradients and
vertical flow resistances within the profiles, as both are
controlled by soil texture and the pore size distribution.
These may lead to long-term feedbacks involving the
regimes of water, heat and nutrient flows and hence the
habitat for soil organisms and vegetation, which in turn alter
the macropore and root networks. Land management in
urban, forest and agricultural areas will impact on this
organised architecture and will likely transform the runoff
pathways (see Chapter 2 ). Surface sealing and soil compac-
tion tend to increase the importance of surface flow paths
relative to subsurface flow, contributing to a flashier runoff
response (Moglen, 2009 ). Agricultural activities may lead
to compaction and soil disturbance, and consequently dis-
turbance of soil organisms with feedbacks on preferential
pathways, depending on the nature of tillage practices (e.g.,
Ndiaye et al., 2005 ). Similarly, forestry practices, including
forest roads, may affect flow paths significantly (e.g., Luce,
2002 ; Buttle, 2011 ). These may exhibit scale effects where
land use change effects tend to taper off with increasing
catchment size (Blöschl et al., 2007 ). In general, the above-
mentioned land use changes, from urbanisation to agricul-
ture and to forest harvesting, tend to increase peak runoff
and reduce water storage in individual catchments, but the
relationships tend to be more complex and not easily
decipherable, as illustrated by numerous paired catchment
experiments and modelling studies on floods (Bronstert
et al., 2002 ; Robinson et al., 2003 ) and water yield
(Andréassian, 2004 ; Brown et al., 2005 ).