Advancements in extreme precipitation projections for South Asia: A comparative evaluation of CMIP5 and CMIP6

Authors: KAMRUZZAMAN, MOHAMMAD - Bangladesh Rice Research Institute; WAHID, SHAHRIAR - Commonwealth Scientific And Industrial Research Organisation (CSIRO); MAINUDDIN, MOHAMMED - Commonwealth Scientific And Industrial Research Organisation (CSIRO); CHIEW, FRANCIS - Commonwealth Scientific And Industrial Research Organisation (CSIRO); ISLAM, ABU REZA MD - Begum Rokeya University; Thorp, Kelly; SHAHID, SHAMSUDDIN - Universiti Teknologi Mara (UITM)

Submitted to: Journal of Hydrology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 5/15/2025
Publication Date: 5/26/2025
Citation: Kamruzzaman, M., Wahid, S., Mainuddin, M., Chiew, F., Islam, A.T., Thorp, K.R., Shahid, S. 2025. Advancements in extreme precipitation projections for South Asia: A comparative evaluation of CMIP5 and CMIP6. Journal of Hydrology. https://doi.org/10.1002/joc.8915.
DOI: https://doi.org/10.1002/joc.8915

Interpretive Summary: Global climate models are continuously updated to provide more accurate predictions of future climate conditions across the globe. However, the reliability of global climate models to estimate extreme precipitation events requires further evaluation. This study compared the performance of global climate models at locations throughout south Asia. The results demonstrated substantial variability in performance among individual models, but improvements are being made as compared to previous model versions. Results of the study will be useful for intergovernmental agencies focusing on mitigation of climate change threats to food security.

Technical Abstract: Global Climate Models (GCMs) are crucial in understanding and projecting future climate changes. However, the ability of GCMs to reliably simulate extreme precipitation remains a significant concern, particularly in South Asia. The objective of this study was to evaluate the performance of Coupled Model Intercomparison Project phase 5 (CMIP5) and CMIP6 GCMs in replicating 11 extreme precipitation indices estimated by ERA5 for 1975–2005. The results indicated substantial variability in the performance of individual models, with certain models excelling or underperforming for specific indices. Generally, CMIP6 models demonstrated superior performance compared to their CMIP5 counterparts, although inconsistencies persisted. Both CMIP5 and CMIP6 multi-model ensemble means (MME) exhibited higher relative root mean square error (RMSE) than the best-performing individual models, underscoring the necessity for model enhancement. CMIP6 models generally surpassed their CMIP5 counterparts, with an average Kling-Gupta efficiency (KGE) of 0.42, exceeding the KGE of 0.38 achieved by CMIP5. This improvement was further corroborated by Taylor diagrams, where CMIP6 models exhibited superior agreement with observed data, evident in their higher average r2 value of 0.65, compared to 0.59 with CMIP5. CMIP6 also demonstrated lower uncertainty in inter-annual monthly precipitation variability projections than CMIP5. EC-Earth3 (CMIP6) and EC-Earth (CMIP5) consistently demonstrated strong correlation with various indices, while MIROC-ESM also performed well in both models. Both MMEs captured inter-annual variability patterns; however, the CMIP6 MME performed better, with a greater KGE (0.48) and r2 (0.71) as compared to the CMIP5 MME (0.44 and 0.67, respectively). Future projections of extremes for the 2040s and 2080s suggest substantial alterations in precipitation extremes for various emission scenarios. The study concludes that CMIP6 exhibits advancements over CMIP5, though further improvements are required to simulate precipitation extremes accurately. This study highlights the progress made by CMIP6 models over CMIP5 in simulating extreme precipitation events and emphasizes the continued need for model refinement.