No abstract

No abstract

The popularity of cutting‐edge machine learning ensemble approaches has solved many climate change research and prediction issues. The six top‐performing GCMs obtained from Technique for Order Preference by Similarity to an Ideal Solution were ensembled using seven machine learning ensemble methods such as Random Forest Regressor (RFR), Support Vector Regressor (SVR), Linear Regression (LR), Adaptive Boosting Regressor (AdaBoost), Extreme Gradient Boosting Regressor (XGBR), Extra Tree Regressor (ETR), Multi‐Layer Perceptron neural network (MLP) and simple Arithmetic Mean (AM) over the diverse geo‐climatic basins. Precipitation is best simulated by EC‐Earth3 and BCC‐CSM2‐MR. Maximum temperature by MPI‐ESM1‐2‐HR, EC‐Earth3‐Veg, INM‐CM5‐0 and MPI‐ESM1‐2‐LR. Minimum temperature by INM‐CM5‐0 and MPI‐ESM1‐2‐LR model. The MME of XGBR and RFR stand out for their superior performance across all six basins, with exceptional performance over the per‐humid basins, while AdaBoost, SVR and the AM underperform. Examining the interseasonal variability of the simulated MMEs over the basins highlights the reliability of these MME models. The anticipated change in maximum and minimum temperature in the SSP245 and SSP585 in the future horizon corroborates the undeniable rise in temperature by all the MMEs with a dramatic change in future temperature in AM and AdaBoost in precipitation with a factor of two rises in the far future over the recent past. Though climate change is expected to increase precipitation, atmospheric stabilization over the Ghats will affect the spatiotemporal distribution of precipitation. We recommend a comprehensive testing and validation approach to generate ensembles in regional investigations involving complicated and diverse precipitation mechanisms.

No abstract

No abstract