Effect of spatial variability of effective rainfall on direct runoff by a geomorphologic approach

The importance of spatial variability of effective rainfall input to a direct runoff model was analyzed by the geomorphologic approach based on the use of different paths determined according to the Horton-Strahler ordering scheme. Actual effective rainfall distributions, derived from observed rainfall-runoff events on four large basins located in Central Italy, together with synthetic storms, preserving effective rainfall depths, were employed. Storms of different time distribution were used, and for each of them sensitivity of basin response to the spatial variability of effective rainfall was computed. The sensitivity was only minor. A comparison of the geomorphologic approach with the Clark method was also carried out for spatially variable synthetic storms of 1 h duration. The Clark method was more sensitive to spatial variability of effective rainfall. It is argued that it is unreasonable to use the geomorphologic approach in computing basin response for spatially variable effective rainfall as model input. An analysis of the distribution throughout the basin of overland regions composing each path revealed that the averaging scheme of the channel network, following the Horton-Strahler ordering scheme, was not appropriate.