Determination of erosion rainfall criteria based on natural rainfall measurement and its impact on spatial distribution of rainfall erosivity in the Czech Republic

Rainfall erosivity is the main factor of the USLE or RUSLE equations. Its accuracy depends on recording precision and its temporal resolution, number of stations and their spatial distribution, length of recorded period, recorded period, erosion rainfall criteria, time step of rainfall intensity and interpolation method. This research focuses on erosion rainfall criteria. A network of 32 ombrographic stations, 1-min temporal resolution rainfall data, 35.6-year period and experimental runoff plots were used. We analysed 8951 rainfalls from ombrographic stations, 100 rainfalls and caused soil losses and runoffs from experimental runoff plots. Main parameter which influenced the number of erosion rainfalls was the precondition AND/OR which determines if conditions of rainfall total (H) have to be fulfilled simultaneously with rainfall intensity (I-15 or I-30) or not. We proved that if parameters I-15 > 6.25 mm/15 min AND H > 12.5 mm were fulfilled, then 84.2% of rainfalls caused soil loss > 0.5 t/ha and 73.7% >= 1 t/ha. In the case of precondition OR only 44.6% of rainfalls caused soil loss > 0.5 t/ha and 33.9% >= 1 t/ha. If the precondition AND was fulfilled, there were on average 75.5 rainfalls, average R factor for each rainfall was 21 MJ/ha.cm/h (without units below in the text, according international unit: 210 MJ/ha.mm/h) and average annual R factor was 45.4. In the case of precondition OR there were on average 279 rainfalls but average R factor for each rainfall was only 9.1 and average annual R factor was 67.4. Therefore if the precondition OR is used, R factor values are overestimated due to a high number of rainfalls with no or very low erosive potential. The resulting overestimated soil losses calculated using USLE/RUSLE subsequently cause an overestimation of financial expenses for erosion-control measures.