Evaluation of gully erosion Sensitivity using Maximum Entropy model in Shoor river watershed (Mohr Township)

gully erosion is one of the most important land degradation processes that leads to reduced land use. The southern and southeastern parts of the Shoor -Mohr basin have been affected by gully erosion. The consequences of this type of erosion include disturbing the ecological balance of the region, endangering biological resources, which has necessitated the study of gully erosion in the region. The purpose of this study is to prepare a zoning map for gully erosion sensitivity using the maximum entropy model and GIS. To achieve this goal, 15 variables including height, slope, slope direction, distance from waterway, drainage density, distance from road, soil texture, geology, land use, rainfall, distance from fault, vegetation, cross section curvature, flow strength index (SPI), Topographic Moisture Index (TWI) was selected and linearity test was performed using (Variance Inflation Factor) and (Tolernace) indices. Out of 100 selected gullys, 70% were randomly selected for educational data and 30% for classification validation and Bootstrap sampling method was used to run the model. Jackknife test was used to determine the most important variables and ROC curve was used to determine the predictive power of the model. Test results showed alignment between vegetation, fault distance, cross-sectional curvature, SPI and TWI were thus omitted. According to Jack Knife test, the variables of height, average annual rainfall, soil texture, drainage density, geology, distance from the road, distance from the river had the greatest effect on gully erosion, respectively. The ROC curve represents 95% accuracy in the validation stage of the model. According to this model, more than 15% of the basin (8445.83 hectares) is very sensitive to gully erosion.