Assessment of climate change in Upper Jhelum Sub-catchment, India, using nonparametric methods and random forest model

This study examines the present and the future trend in rainfall and temperature in the Upper Jhelum Sub-catchment located in the northwestern Himalayas in India. We used gridded rainfall and temperature data obtained from the India Meteorological Department from 1972 to 2022. Mann–Kendall test and Sen’s slope estimator were utilized to evaluate the trend and quantify changes in the pattern of rainfall and temperature variables. The random forest model was utilized to forecast rainfall and temperature (2023–2047). The accuracy of the model was assessed using performance assessors. The results revealed an annual increasing trend in temperature at the rate of 0.0096 (°C/year) and decreasing trend in rainfall at the rate of − 2.2061 (mm/year) during the pre-monsoon and − 0.8676 (mm/year) during the post-monsoon seasons. A decreasing trend in maximum temperature was recorded during the monsoon and post-monsoon seasons at the rate of − 0.0056 and − 0.0134 (°C/year), respectively. The forecast analysis revealed decreasing trend in the rainfall at the rate of − 0.9256 and − 0.03961 (mm/year) during pre-monsoon and post-monsoon seasons, respectively, while increase in minimum temperature at the rate of 0.0714 , 0.0134 and 0.006 (°C/year) during the pre-monsoon, winter and monsoon seasons, respectively. The random forest model was found effective for forecast analysis of rainfall and temperature variables. The methodological framework utilized in this study may be replicated in other geographical regions for examining climate change.