Determining the Spatial Distribution of Gully Erosion Probability Using the MaxEnt Model

Gully erosion is one of the most important types of water erosion, which has a great role in the destruction of land production capacity due to intra-regional and extra-regional effects. Preparing a spatial map of the possibility of gully erosion for better management of land use with the aim of reducing land degradation in areas prone to gully occurrence is very efficient. The distribution of this erosion in Iran and the extent of factors and processes affecting its creation have been a major obstacle in creating a comprehensive model for predicting its occurrence on a large scale. The purpose of this study is to prepare a map of the probability of gully erosion using the machine learning model of maximum entropy in Fars province. In this research, it has been tried to use variables related to terrestrial characteristics, especially soil. According to the results, the area under the ROC curve is above 90%, which shows that the model has been able to evaluate the gully erosion in the study area using the studied data. According to the results of the Jaknaev test, the variables of R horizon probability, soil depth, percentage of coarse grains, pH, and silt particles have the greatest impact on modeling moat erosion in the study area. The spatial distribution map of the occurrence of the gully is a map of land susceptibility to gully erosion. Based on the findings, the highest sensitivity to gully erosion is related to the south of Fars province. The map prepared in this research can be used as a basic map for land management, managers and engineers of urban planning, road construction, natural resources, and agriculture.