Assessing Climate Change Impacts on Agricultural Regional Gross Products in the Urmia Watershed: A Data Mining Approach

This study assesses the impacts of climate change on agriculture in the Urmia Basin using the Random Forest algorithm to project future scenarios under SSP۱-۲.۶ and SSP۵-۸.۵ climate pathways. Historical data indicates an average annual precipitation of ۲۹۱ mm, an air temperature of ۱۱.۲۹°C, and evapotranspiration of ۱۰۸۶ mm. Projections for ۲۰۲۵-۲۰۳۵ suggest a warming trend, with temperatures rising to ۱۲.۰۲°C under SSP۵-۸.۵, accompanied by reduced precipitation and increased evapotranspiration. These changes are anticipated to negatively affect agricultural productivity, with the SSP۵-۸.۵ scenario projecting a lower Gross Regional Product (GRP) due to decreased precipitation, higher evapotranspiration, and reduced area of cultivation. The analysis of feature importance, utilizing LULC MODIS data to extract cultivation areas, revealed that precipitation and the area of cultivation are the most significant factors influencing the model's predictions. The model demonstrated acceptable predictive accuracy, with a Kling-Gupta Efficiency (KGE) of ۰.۵۴ and a Root Mean Square Error (RMSE) of ۰.۳۱. Additionally, declining agricultural productivity may drive increased urban expansion and strain on urban infrastructure, underscoring the need for integrated adaptation and development strategies.

کلیدواژه ها:

Machine learning ، climate change ، Agriculture ، Urmia Lake Watershed