Reassessment of a semi-analytical field-scale infiltration model through experiments under natural rainfall events

Operative distributed/semi-distributed rainfall-runoff formulations need a mathematically tractable component for estimating the expected areal-average infiltration rate, I, while point infiltration models are usually adopted for this purpose. A model for I, suitable for hydrological applications, was proposed by Morbidelli et al. (2006, Hydrological Processes, 20: 1465–1481). The main model tenet is the use of cumulative infiltration as an independent variable and replacement of time by its expected value. While this model - as with many other models for field-scale infiltration - was tested through Monte Carlo methods, it remains untested with experimental observations. In this study, a reassessment of the model's applicability is conducted through experimental evidence from several rainfall-runoff events observed under natural conditions at the plot scale over a slight slope characterized by a silty loam soil. During each experiment rainfall rate, surface runoff at the plot outlet and soil water content vertical profiles were monitored at short time intervals. An overall analysis of our results suggests that the model reliably simulates I and surface runoff as functions of time and is also suitable for applications under time-dependent rainfall events. The process of redistribution under rainfall hiatus is not addressed by this model.