Relating Rainfall Erosivity and Depth of Rainfall: A statistical Procedure for Data scarce Regions

Rainfall erosivity (R) is the main component in the RUSLE model, used widely for making soil loss estimations in watersheds. The computation of R requires rainfall intensity versus duration curve for every storm event. Few meteorological stations provide this data, while daily rainfall depth is recorded by all the stations. The development of a relation between erosivity and depth of rainfall permit use of all available data and thus could lead to better assessment of R. Four meteorological stations from Krishnagiri reservoir catchment area that record both intensity and depth of rainfall provided data for analysis of 178 rainfall events during 2004 to 2008. Regression analyses between erosivity and depth of rainfall using linear, logarithmic, exponential, power, polynomial and quadratic methods yielded a power function with highest R2 of 0.739. The newly developed power function was used to compute erosivity for the 18 stations having only depth of rainfall data in the Krishnagiri reservoir catchment area. A spatial pattern of erosivity for different seasons were generated using interpolation techniques in Arc GIS 9.0 permitted better delineation of vulnerable zones of the watershed prone to soil erosion. This study proposes a new methodology and a simplified relation between daily rainfall depth and erosivity for a data scarce region.