Investigation of natural and human factors affecting land use changes in Shadegan wetland

Shadegan wetland is located at the end of Jarahi watershed. In order to monitor the changes in water cover, vegetation and uncovered or dried lands, first the area of Tala Shirin in its most waterlogged conditions was determined. Landsat TM, ETM + and OLI satellite data from 1988 to 2018 in June were used. Supervised vector classification (SVM) was used and the images were classified into three classes: water, vegetation and soil. Image classification accuracy was calculated with two indicators, overall accuracy and kappa. The area of each class was zoned in GIS and calculated as a percentage. Due to the fact that determining the share of each of the components affecting different levels of Shadegan wetland in the period (1988-2018) is the main purpose of this study, the radius and multilayer perceptron functions of neural networks were used. And after comparing the efficiency of the models, the best model was selected and based on it, the effect of each parameter on the different levels of classification in the study period was determined. The results showed that the share of human and natural factors in drought was 55.4% and 44.6% in the wetland, 70.8% and 29.2% in the water level of the wetland and 62% and 38% in the vegetation, respectively. Therefore, Shadegan wetland, more than anything else, requires human management at the watershed level and the provision of the required water supply in order to be able to play a role and continue its life as a wetland ecosystem.