Extended Abstract
Introduction: Gully erosion is a severe form of soil erosion, but internal gully erosion processes are not well understood, especially at the scale of rainfall event. Nowadays, gully erosion is known as one of the most destructive types of erosion in agricultural lands and natural resources in the world such that it has a significant share of scientific research. Although soil erosion is a natural process, human activities in the past decades have greatly accelerated different types of erosion in nature. Gully erosion is the final and advanced stage of the erosion process, which, if not controlled, can cause huge damage to infrastructure as well as various agricultural parts, natural resources and environment, which either do not compensate for damage or if compensated, takes a long time in nature. In arid and semi-arid regions, due to certain conditions, the creation and development of gully erosion can make tremendous progress. Soil erosion in arid and semi-arid regions is one of the important consequences of climate change or is one of the consequences of environmental and ecological changes. Therefore, the purpose of this study is to rank the effective factors of morphometric erosion in creating gully erosion using statistical methods, as well as preparing gully erosion sensitivity map using maximum entropy model and its sensitivity in arid and semi-arid regions in arid and semi-arid provinces such as Kerman, which in turn provides valuable information on how to create and develop gully erosion in these areas.
Material and Methods: In this study, 79 gullies were identified in Sarab Halil watershed in Kerman province. Then, 15 morphometric information layers were obtained along with gullies distribution map and PCA statistical Analysis was used to determine the most important factors affecting morphometric and finally, the map of gully erosion zoning was obtained using entropy maximum model for morphometric factors. In addition, MaxEnt model is a general model that allows users to evaluate the relationships between a dependent variable and several independent variables in different study contexts. The maximum entropy model based on the principle of entropy specifies the network of connections between dependent and independent variables and are obtained based on the role of each independent variable, its influencing weight, and its response curves. Entropy indicates the degree of uncertainty of the unbalanced distribution of the existing phenomena from the expected information content. Entropy method has been used in various fields such as mathematics, computer and economics in Iran and the world, but it has been used less commonly in geomorphology. In addition, Jackknife test was used to determine the importance of morphometric variables and the area under the curve criterion and acceptor performance specific curve were used to evaluate the accuracy of the model. The graph of the acceptor performance specific curve expresses the presence of the prediction against the accuracy of the absence of the forecast. If the amount of the area under the curve falls between 0.7 and 0.8, the model is considered good, and if the area under the curve ranges from 0.8 to 0.9, the model is considered very good, and if the amount of the area under the curve is more than 0.9, it is considered an excellent model. Meanwhile, the area index under the curve in receiver factor is equal to the probability of correctly distinguishing between the points of presence and absence by a model.
Results and discussion: Gully erosion is one of the most important types of erosion in different climates of the planet, which causes widespread destruction and since it is very scattered in watershed areas, predicting its occurrence with low research costs is very important. The use of morphometric factors in this research, in addition to having low research and field costs, showed that the desired and acceptable results can be achieved without the use of other factors that have higher cost. Map of gully erosion prone areas obtained from entropy maximum model using morphometric factors in the study area showed that gully erosion in northeast, east and south is more likely to gully erosion between 0 and 31%, but in the east of the watershed and southeast, the gully erosion increases slightly and reaches the probability of 92%, but the percentage of the area is very low. However, towards the center, north, northwest, west and southwest of the study area, the probability of gully erosion increases and reaches 92% and sometimes in some parts up to 100%.
Conclusion: The results showed that in the occurrence of gully erosion, the morphometric factors of plan curvature, profile curvature, topographic wetness index, vertical distance to channel network, altitude, and length - slope factor, slope and earth's surface texture are effective in creating the gully erosion.