Probabilistic evaluation of three generations of climate models for simulating precipitation over the Western Himalayas

The current study aimed to assess the ability of Global Climate Models (GCMs) from the third, fifth, and sixth phases of the Coupled Model Intercomparison Project (CMIP) to reproduce observed rainfall over the Western Himalaya Region (WHR) between 1901 and 2000. A probabilistic approach was adopted to evaluate the models’ skill in capturing the observed precipitation from 22 India Meteorological Department stations by comparing their probability distribution function (PDF) with that of the observed rainfall. The results indicated that all three generations of climate models could adequately reproduce the PDF of observed precipitation, surpassing the observed PDFs in both annual and monsoon seasons. Subsequently, each model was assigned a rank based on its skill score. The Geophysical Fluid Dynamics Laboratory model from CMIP3, the community Earth system model from CMIP5, and the CMIP6 model from the Met Office Hadley Centre achieved the highest ranks, accurately capturing the distribution of observed precipitation over WHR. Remarkably, the Japanese model from the Model for Interdisciplinary Research and the UK model from Hadley Centre exhibited consistently high performance across all three CMIP generations. Overall, the top-ranked models from each generation can be deemed reliable for use by researchers and decision-makers concerned with the impact of climate change on the Western Himalaya Region.