Prediction of Soil Loss in the Northern Part of Somali Region of Ethiopia Using Empirical Soil Erosion Models

Abstract

Two empirical soil loss models, namely Universal Soil Loss Equation (USLE) and Soil Loss Estimation Model for Southern Africa (SLEMSA) were applied to assess extent of erosion hazard at 12 selected sites in the northern part of Somali region of Ethiopia. The amount of estimated soil loss for 10 out of 12 sites by using the USLE is by two to three and half times greater than that estimated by SLEMSA. The difference between the values of soil loss estimated by the two models can be attributed to the difference in the sensitivity of the models to their input factors. In general, since SLEMSA under estimate the amount of soil loss from a given site, it is recommended to use USLE to estimate soil loss in the northern part of the Somali region of Ethiopia. The sensitivity of both models to their 20% increase or decrease in one of their input variable at a time while keeping other variables constant was analysed. The analysis indicated that USLE was highly sensitive to slope gradient factor (S), soil conservation practice factor (P) and rainfall erosivity factor (R) but less sensitive to slope length (L) and vegetal cover factor (C). Whereas, SLEMSA was highly sensitive to change in rainfall kinetic energy (E) and soil erodibility (F) and was less sensitive to slope gradient (S), slope length (L) and percent cover factor (C).