Influence of the rainfall measurement interval on the erosivity determinations in the Mediterranean area

Abstract

Summary The single-storm erosion index, EI, of the USLE and RUSLE models may vary appreciably with the rainfall measurement interval, Δ t . However, the effect of Δ t on EI has not been investigated in the Mediterranean area. Approximately 700 erosive events and 1.5 years of rainfall energies measured by a rainfall impact measurement device were used to evaluate the effect of the rainfall measurement interval (5 min ⩽ Δ t ⩽ 60 min) on the erosivity determinations in the Mediterranean semi-arid area of Sicily. According to both literature and practical considerations, a reference time interval equal to 15 min was used in this investigation. Hourly rainfall data led to an appreciable underestimation of the mean value of EI (i.e., by also a factor of two, depending on the location). In the range 5 min ⩽ Δ t ⩽ 15 min, the effect of the rainfall measurement interval on the predicted erosivity was negligible (i.e., mean values differing by a maximum factor of 1.10) as compared with the uncertainties in the soil loss predictions. Two methods were developed for estimating the reference single-storm erosion index, (EI) 15 , from hourly rainfall data in Sicily. Method 1 converts the erosion index calculated on a 60-min measurement interval basis to (EI) 15 . Method 2 estimates (EI) 15 by using the storm rainfall depth and the maximum rainfall intensity. Testing the two methods against two independent data sets produced a maximum difference between the estimated and the calculated mean values of (EI) 15 equal to 7% for method 1 and 11% for method 2. Both methods may be applied in practice, depending on the available rainfall data. For a given rainfall intensity, the specific power, P , measured at eight time intervals (5 min ⩽ Δ t ⩽ 60 min) was in the range ±10% of the mean of the eight P values.