Evaluating the rainfall erosivity (R-factor) from daily rainfall data: an application for assessing climate change impact on soil loss in Westrapti River basin, Nepal

Rainfall is one of the active drivers of soil loss that detaches and displaces soil particles. It is difficult to assess precisely the distribution, the size, and the terminal velocity of raindrops in relation to soil detachment and displacement. Therefore, many empirical methods are used based on rainfall for various time steps (daily, monthly, and yearly). Evaluation of rainfall erosivity (R-factor) precisely needs rainfall data of higher temporal resolution (< 15-min interval) for several years, which is not commonly available, especially in developing countries like Nepal. In this study, we developed an R-factor estimation model using daily rainfall which was firstly compared with R-factor derived using sub-hourly rainfall. Westrapti River basin (WRB) of Nepal is the study domain where 5-min interval rainfall data are available 2011 onwards. The estimation model was calibrated for 3 years (2011–2013) and validated for 2 years (2014–2015). The performance of the model was evaluated using six statistical indicators. We then used the model for 30 years (1986–2015) as a baseline and three future time periods: (1) near-future, 2025–2049, (2) mid-future, 2050–2074, and (3) far-future, 2075–2099. We used five climate models under two warming scenarios (RCPs 4.5 and 8.5) for projected analysis during future time periods. The estimated mean annual R-factor of the study area was 3514.6 MJ mm ha−1 h−1 year−1 during the baseline and is expected to increase by 10% under both warming scenarios during the far-future time period. July was the highest erosive month followed by August. Our result showed that the change in rainfall patterns increased the erosivity density in the WRB in recent time. We found that the mean soil loss for the WRB for the baseline time period was estimated at 8.1 t ha−1 year−1 and is expected to increase by about 10% under both warming scenarios during the far-future time period. The result of this study is useful for sediment and watershed management.