behind determining the path of maximum downward gradient is straightforward as described by Jones et al. (1990). Watershed boundaries are delineated with TINS by identifying outlet locations. Once outlets have been selected, flow paths are traced along the path of steepest descent and by combining together triangles whose flow paths pass through a common outlet point.
LUMPED VERSUS DISTRIBUTED MODELS
Creating an accurate hydrologic model of an area is a difficult task. As most hydrologic systems are spatially variable, distributed models may be required to fully describe the system. Distributed models require that calculations be made on a point-to-point basis within the model, and that flow be calculated as a function of time and space throughout the system. Lumped models on the other hand provide a unique representative value for the entire subcatchment. In a lumped model, flow is calculated as a function of time alone (Nelson et al., 1997).
Olivera et al. (1999) discuss how there have been attempts to account for spatially distributed terrain attributes based on lumped models, as the boundary
models and distributed models such as HEC-1, developed by the
Army Corps of Engineers,
the hydrologic system subsystems. HEC-1 watershed outlet.
into subsystems and to apply then routes the responses
lumped models to each of from each subbasin to