Geoscience Reference
In-Depth Information
conditions, such as initial moisture conditions in the catchment and the magnitude of the
precipitation.
In some cases the coexistence of different mechanisms can give rise to some unusual
phenomena. For instance, under low initial conditions in a watershed in central Cote
d'Ivoire, Masiyandima
et al
. (2003) observed double-peaked hydrographs resulting from
the same rainfall burst; the first peak, which occurred while itrained, was produced by
the rainfall on the saturated valley bottom; the second peak, which came minutes to hours
after the first, resulted from the rain that had fallen on the area surrounding the valley
bottom and that had traveled to the stream channel by subsurface flow.
All this underscores again the extreme complexity of the stream generation process.
These observations suggest that a single unifying runoff model may not be possible nor
even desirable, and they have profound implications for the development of modeling
strategies for predictive purposes in applied hydrology.
Identification of major mechanisms
In order to keep the formulation sufficiently simple and parsimonious, it may be necessary
to identify and include only the dominant mechanisms for any given set of conditions, and
to accept some inevitable uncertainty resulting from the omission of the remaining minor
mechanisms. On the basis of a knowledge of these local conditions, the analyst must
then decide which mechanisms are the major ones that must be considered to represent
a particular catchment. The insight gained by the recent field observations can also give
some guidance inthis. For instance, different kinds of subsurface flow can be assumed
to dominate the runoff process in humid areas with an active vegetation. Well-developed
mineral soils undoubtedly favor the development of preferential flow paths, whereas thin
porous soils with organiclitter probably lead to shallow lateral flow of the perched water
above the less permeable soil or bedrock. Wetland areas near the stream may allow rapid
mobilization or ridging of the water table, and the development of partial and variable
source areas, on which saturation excess overland flow can take place. Infiltration excess
overland flow will be prevalent during large precipitation events on unvegetated surfaces
inaridregions and in areas subject to intense human activities.
Objectives of the analysis
Thiswide variety in possible mechanisms also means that, in the development of mod-
eling strategies for engineering and other applied purposes, for a given catchment it may
be advisable to adopt different formulations depending on the objectives of the analysis.
For example, the prediction of disastrous flash flooding, under extreme precipitation
conditions, may require an entirely different approach from those needed to describe
solute transport and water quality in the environment, to analyze possible climate change
scenarios under more normal flow regimes, or to assess the potential for erosion or land-
slides. For flood prediction, mainly the flows at a certainpoint along the river may be of
interest; for climate change scenarios, surface-atmosphere interactions are of paramount
importance; and for water quality purposes it may be crucial to know the pathways, in
order to determine the fate and transport of admixtures and water pollutants; finally,
