Document Type : Research Paper
Authors
1
Department of Geography, Faculty of Literature and Humanities, Razi University, Kermanshah, Iran
2
Water Engineering Department. Faculty of Agriculture. Razi University. Kermanshah., Iran
3
Departmt of Geography, Faculty of Literature and Humanities, Razi University, Kermanshah, Iran
4
Department of Water Engineering, Campus of Agriculture and Natural Resources, Razi University, Kermanshah, Iran
Abstract
This study predicts groundwater level (GWL) variations in the Eslamabad plain using the Random Forest (RF) and CMIP6 climate data. GWL data from 20 wells (1997–2014) were collected and, after processing and clustering using the K-Means, modeling was conducted for different climate scenarios in three regions: Jangeh, Barfabad, and Bureg.
The results showed that in the observational period, Barfabad, with variations of 8.7 to 10.2 meters, had the best GWL values, while the Bureg, with 15.5 to 17.3 meters, had the worst. The highest availability was observed in spring and the lowest depth in Fall. Predictions indicate that in the distant future (2076–2100), under the SSP5-8.5 scenario, the greatest GWL increase (3 to 3.5 meters) will occur in Jangeh in Fall. In SSP1-2.6, the greatest decline in Bureg, with a drop of 3.5 to 4 meters, is projected for spring and summer. Under the SSP1-2.6 and SSP2-4.5 scenarios, conditions will be more stable in the distant future.
During the observational period, GWL showed a decreasing trend in all regions, with the highest annual decline (1 meter) recorded in Barfabab. In SSP1-2.6, the GWL decrease in the near future in Bureg (-0.21 m/year) and in the mid-future in Jangeh (-0.14 m/year) will continue. In SSP2-4.5, this decrease in the distant future is significant in all regions, with the highest value (-0.16 m/year) in Bureg. The RF model performed very well (R = 0.97, NSE between 0.89 and 0.98), and the NorESM2-MM model improved prediction accuracy up to 99.5%.
Keywords
Main Subjects
)