Spatial and temporal variation in rainfall erosivity in a Himalayan watershed

• Published in: Catena (Giessen) 2014-10, Vol.121, p.248-259
• Main Authors: Ma, Xing, He, Yandong, Xu, Jianchu, van Noordwijk, Meine, Lu, Xixi
• Format: Article
• Language: English
• Subjects: Annual rainfall; catchment; china; climate; Daily rainfall erosivity model; Freshwater; High resolution; Kejie watershed; land-cover; Mathematical models; Mountains; r-factor; R-factor maps; R-factor of RUSLE; R-factor variation; Rainfall; river; Stations; Trends; variability; Watersheds
• ISSN: 0341-8162; 1872-6887
• DOI: 10.1016/j.catena.2014.05.017

Global climate change can modify rainfall patterns, leading to more extremes with associated erosion events. Rainfall erosivity, or the R-factor based on the Revised Universal Soil Loss Equation (RUSLE), indicates the potential water erosion risk and it plays an important role in water and soil conservation assessments. However, calculation of the R-factor requires high resolution data series, and thus we present an alternative model that can be used to accurately calculate the R-factor. Our erosivity model uses daily rainfall with advised regression parameters to estimate the R-factor in the watershed, which was selected by comparing the actual R-factor with 10 min high resolution rainfall data and the estimated R-factor with daily rainfall data from 1998 to 2002. The mean annual R-factor map was derived in the study using cokriging. The annual R-factor in the Kejie watershed was classified as medium and medium-strong erosivity, with a mean value of 3264 MJ.mm.ha−1.h−1.yr−1 which represented a range from 2505 to 5538 MJ.mm.ha−1.h−1.yr−1. A simple power relation between annual R-factor and annual rainfall was derived. The long-term change trend analysis showed no significant increasing or decreasing trend observed for the region; however, there was a significant increasing trend observed in two stations in September, one station in March. The annual R-factor with a coefficient of variation of 0.30 indicated inter-annual variation of the R-factor in the watershed was not so apparent. The intra-annual R-factor analysis illustrated the apparent seasonal and monthly distribution, about 65% from the summer season, and the maximum monthly R-factor occurring in July, followed by August and June. Consequently, the adjusted daily model can be applied in this Himalayan mountain area when high-resolution rainfall data is unavailable. The R-factor map and the simple power relation provided a useful tool for land-use planner and agriculture management in the Kejie watershed. •Provide the actual R-factor (EI30) at six sites over five years in a Himalayan watershed;•Recommend a daily model for R-factor estimation for a Himalayan watershed;•Recommend a power equation for annual R-factor estimation for a Himalayan watershed;•Provide a spatial distribution map of mean annual R-factor for a Himalayan watershed.