Comparative analysis of CMIP5 and CMIP6 in conjunction with the hydrological processes of reservoir catchment, Chhattisgarh, India

Abstract

Study regionMahanadi reservoir project complex (MRP), Chhattisgarh, India Study focusThis study assesses and compares the performance of the Coupled Model Intercomparison Project Phase 5 (CMIP5) and CMIP6 models in modeling the hydrological regime. The study compared the performance of 13 CMIP6 GCMs in two Shared Socioeconomic Pathways (SSEPs) with that of 16 CMIP5 GCMs in two Representative Concentration Pathways (RCPs). Before being used to drive the Soil and Water Assessment Tool (SWAT) for streamflow prediction, the raw CMIP outputs were adjusted and downscaled using Bias Correction and Spatial Disaggregation methods (BCSD). In addition, the water deficits were also assessed between the mean monthly released by the optimized model and the mean monthly water demand. Results of this research projected the domestic water demand for future scenarios (2023–2099). New hydrological insights for the study regionThe results of this study revealed that the genetic algorithm (GA) outperforms the various optimization techniques for CMIP5, with deficits observed by the GA algorithm. In addition, CCCmaCanESM2 was found to be the most efficient among CMIP5 GCMs, whereas MPI-ESM1–2-HR was used for CMIP6 GCMs. Overall, the CMIP6 multi-model mean ensemble (MMME) outperformed the CMIP5 MMME in simulating streamflow over the study area at annual and seasonal timescales. CMIP6 MMME also reduced maximum and minimum temperature biases over the study region significantly. In overall conclusion, the CMIP6 ensemble offers a lower margin of uncertainty for future climate projections and better credibility for hydrological effect analysis. The results further indicate that satisfactory steam flows can be obtained from the SWAT model while slight underestimations can be seen for higher discharges. In addition, the projected domestic water demand based on future population and irrigation demand clearly showcases significant annual increases.