Global Climate Models Underestimate Indian Monsoon Rainfall: Study

A study published in Nature’s Climate and Atmospheric Science journal has highlighted significant shortcomings in global climate models (GCMs) concerning the prediction of the Indian summer monsoon. Conducted by researchers from the University of Science and Technology of China and their collaborators, the study reveals that these models consistently underpredict monsoon rainfall over northern India by 10–20%, raising concerns about the accuracy of flood forecasting and long-term climate projections.

The Monsoon’s Significance

The Indian summer monsoon is a critical weather system, delivering approximately 70–80% of the country’s annual rainfall. This seasonal phenomenon profoundly impacts the food and economic security of over a billion people. Accurate predictions of monsoon patterns are essential for agricultural planning, water resource management, and disaster preparedness.

Identifying the Flaw

The study attributes the underprediction to excessive simulation of light rain over the equatorial Indian Ocean (EIO) by GCMs. This overestimation enhances latent heat release, triggering stronger rainfall in the region. Consequently, it fuels an anomalous Hadley-type circulation—a large-scale tropical pattern where warm air rises at the equator and descends in the subtropics. The descending branch of this circulation suppresses vertical motion over northern India, reducing cloud formation and moisture transport, leading to a persistent “dry bias” in monsoon forecasts.

Methodology and Findings of the study

To investigate this issue, researchers conducted numerical simulations during the active monsoon period in 2020 using two model resolutions: a coarser 60 km model (U60 km) and a high-resolution 3 km model (U3 km). The U60 km model exhibited a pronounced dry bias, underpredicting rainfall over northern India by 15% and overestimating EIO drizzle by 30%. In contrast, the finer U3 km model reduced the dry bias by 8% and cut excessive equatorial rain by 12%, improving accuracy. It also better captured monsoon dynamics, particularly the Somali jet—a key wind current that carries moisture from the Arabian Sea to India. In the coarser model, the jet’s wind speed was 18% weaker than observed, while the high-resolution version closely matched reanalysis data.

Implications for Climate Modeling

The findings highlight the limitations of convection schemes in coarser GCMs, which regulate how clouds and rainfall are simulated. These schemes often overestimate the frequency of light rain, especially over the EIO, where it occurs 25% more often than in real-world observations. This bias shifts the rainfall distribution toward drizzle, intensifying the EIO rainbelt while reducing precipitation over northern India.

The study suggests that improving the spatial resolution of climate models can enhance the accuracy of monsoon predictions. High-resolution models like the U3 km version demonstrate better simulation of monsoon dynamics, offering a more reliable tool for forecasting. As climate change continues to impact weather patterns, refining these models becomes increasingly vital for effective planning and mitigation strategies.

Conclusion

This research highlights a critical area for improvement in climate modeling, particularly concerning the Indian summer monsoon. By addressing the overestimation of light rain over the EIO and enhancing model resolution, scientists can develop more accurate predictions, aiding in the preparation for and response to monsoon-related challenges.

References:

https://timesofindia.indiatimes.com/city/dehradun/global-climate-models-underpredict-indian-monsoon-study/amp_articleshow/120939129.cms

https://www.nature.com/articles/s41612-025-00955-8

https://weather.missouri.edu/gcc/_09-09-13_%20Chapter%201%20Models.pdf

Photo by Parij Photography on Pexels 

https://www.pexels.com/photo/boy-holding-black-umbrella-1483142

Banner Image: 

Photo by Ali Hassan on Pexels https://www.pexels.com/photo/raindrops-falling-on-a-person-s-hand-4029257/