The construction of watershed structures is an important method for flood and sediment control. The cost of constructing these structures at the watershed scale is high, and reducing it requires the use of precise site selection methods. Machine learning models are among the precise site selection methods, and this research uses the Maxent (maximum entropy) model because it does not require absence data. This study was conducted in three watersheds: Aghevlar Talesh, Masouleh Fuman, and Tutkabon Rudbar, in Guilan province, using 21 effective factors in the site selection of watershed structures. The results obtained from the variance inflation factor (VIF) in R software showed that there is collinearity among the variables, and factors with a VIF less than 10 were entered into the modeling process. To evaluate, the data were divided into two parts of 70% and 30% for training and validation of the model, and the ROC-AUC index results showed that the model performance in the training and validation stages in all three watersheds was more than 90%, placing it at an excellent level. Based on the results of the jackknife test, distance to borrow pits, distance to rivers, runoff depth, precipitation, and slope were identified as the most influential variables for siting gabion structures, while distance to rivers, distance to borrow pits, peak flood discharge, runoff depth, and distance to roads were the most important variables for siting masonry walls in all three watersheds. The findings showed that more than 70% of gabion structures and more than 80% of Masonry wall structures in the studied watersheds are located in areas with medium and higher potential for the construction of watershed structures. Out of a total of 923 km of stream networks across the three study watersheds, 63 km and 39 km were identified as the most suitable locations for the construction of gabion and masonry structures, respectively.
Ghaderi Vangah B, Ghazavi R, Dokhani S, Asadi Nalivan O. Selection of the best location for watershed structures in Gilan province using the maximum entropy model. Journal of Rainwater Catchment Systems 2026; 14 (1) : 3 URL: http://jircsa.ir/article-1-609-en.html
تکمیل و ارسال فرم تعارض منافع نویسنده گرامی ، پس از ارسال مقاله ، جهت دریافت فرم، لطفا بر روی کلمه فرم تعارض منافع کلیک نمایید و پس از تکمیل، در فایل های پیوست مقاله قرار دهید.