Assessing gully erosion susceptibility in Mayurakshi river basin of eastern India

Abstract

The present study assessed the extent to which the gully erosion susceptibility could be successfully modeled in Mayurakshi river basin as a function of four gully controlling factor clusters (topographical, erosivity, erodibility, and resisting). Gully erosion zones have been identified through two different models (correlation matrix and logistic regression) incorporating 15 conditioning indicators. Both the models have generated quite identical result regarding gully erosion susceptibility. It illustrates that extreme erosion-susceptible zones cover almost 16% of the basin area followed by high susceptibility with almost 28% areal coverage. This susceptibility is dominantly confined within the upper catchment. Older and newer alluvial plains of the lower catchment belong to low susceptibility to relatively safe situation. Models are validated by constructing ROC curve and calculating Kappa statistics. Both the approaches certified the validity of the models. The factor clustering shows that the positively influencing clusters like erosivity factors and erodibility factors are playing a dominant role with, respectively, 76 and 75% correlation with gully erosion susceptibility, whereas the resistant factors achieved only 32% correlation which denotes week significance due to containing sparse vegetation cover having low canopy density as well as poor crop management and conservation practice.