DESCRIBING RATE VARIABILITY OF STORM EVENTS FOR INFILTRATION PREDICTION

This article deals with the influence of rainfall rate variability on the infiltration predicted with the Chu model. At first, we determined the dependence of infiltration on both peak intensity and time to pea k intensity of idealized, unimodal storm events. Secondly we examined the influence of the temporal resolution of natural storms measurement on infiltration by using rainfall data aggregated in 5- to 60-min tim e intervals. Finally, we tried to derive a simplified procedure to des cribe rate variability of observed storm events for infiltration predi ction. The study showed that the infiltration predictions were signifi cantly affected by the rainfall peak intensity of a unimodal storm eve nt for relatively low values of the ratio between mean rainfall intens ity and saturated hydraulic conductivity, and that time to peak intens ity did not appreciably influence model results. For the selected set of natural storm events, characterized by either high rainfall depth o r high rainfall intensity, the effect of the temporal resolution of ra infall measurement on the infiltration estimates strongly depended on the soil type. In soils with either high or low permeability, this eff ect was practically negligible. In soils with intermediate permeabilit y characteristics, an increase of the temporal scale of rainfall data aggregation determined an increase of the infiltration values and a de crease of their variability. For these soils, a rough relationship bet ween simple descriptors of both infiltration estimates and rainfall in tensities variability was recognized. A simplified description of stor m intensities, which assumes an exponential rainfall intensity distrib ution and requires only the knowledge of mean and maximum storm intens ities, allowed a satisfactory reproduction of mean infiltration estima tes of a series of storm events.