Climate models are underestimating future rainfall in India

The most advanced set of climate models used in the Intergovernmental Panel on Climate Change (IPCC) assessment reports are underestimating the impacts of climate change on rainfall in India, a new analysis has found.

The Sixth Coupled Model Intercomparison Project (CMIP6) is a major set of models that “simulate the physics, chemistry, and biology of the atmosphere, land, and oceans” to project the future impacts of climate change under different emissions scenarios. They are among the most advanced set of climate models widely used by researchers and policymakers to understand how climate change will affect weather patterns and socio-economic development pathways around the world.

But while CMIP6 is capable of providing insights at a global scale, it is likely to falter at a more regional scale. A new study that corrects and down-scales CMIP6 estimates has found that at the regional scale in India, CMIP6 is underestimating the magnitude of future long- and short-duration extreme rainfall events. “These findings can provide fundamental insights for formulating national climate change adaptation policies for extreme rainfall events,” the researchers from the Indian Institute of Tropical Meteorology and National Institute of Technology, Rourkela wrote in the paper.

Refining models for more accurate predictions

The study aims to understand the impacts of climate change on the Indian summer monsoon between June and September under different emission scenarios. To do this, the researchers down-scaled and bias-corrected the sea surface temperature and precipitation components of the CMIP6 models. “Scaling essentially involves bringing the coarse resolution from the CMIP6 models to a higher resolution, and bias correction involves bringing the results as close to observations as possible,” explained Jasti S. Chowdhary, a scientist at the Indian Institute of Tropical Meteorology and a co-author of the study.

The baseline period for the models and observations used is India Meteorological Department (IMD) data from 1980 to 2014. While most other downscaled models include only the rainfall component, this study is among the first few studies to also incorporate sea surface temperature, an important determinant of the progress of the Indian summer monsoon, in its projections.

The researchers were able to show that CMIP6 simulations “greatly underestimate the intensity and spatial distribution of rainfall associated with extreme precipitation events” because downscaling and bias correction resulted in a 96 percent improvement in estimates – meaning they were 96 percent closer to the actual observed results.

The researchers found that in the baseline period (1980-2014), eight per cent of the Indian land area experienced heavy rainfall associated with extreme rainfall events during the summer monsoon season, while in the near future (between 2030 and 2060), the high emissions scenario sees a 14 per cent increase compared to the baseline period, and an 18 per cent increase in the distant future (2060 to 2100). Further, in the distant future, the extent of extreme rainfall is projected to be 21 per cent higher than the baseline under the high emissions scenario, while it is 6 per cent in the moderate emissions scenario.

It also found that before 2060, the number of short-term extreme rainfall events is higher than long-duration extreme rainfall events. After 2060, this trend reverses, “leading to a two-fold increase in long-duration extreme rainfall events compared to short-duration extreme rainfall events over the monsoon core region,” the study said.

“Long-duration extreme rainfall events last for several days. Typically, these events result in the total accumulation of rainfall over large geographic areas, often causing persistent water levels in rivers and reservoirs to rise. Short-duration extreme rainfall events last from several hours to a few days. They are characterized by very high rainfall rates, leading to rapid accumulation of water in a short period. They are usually local, affecting small geographic areas,” said Gopi Nadh Konda, lead author of the study.

The CMIP6 models also underestimated the seasonal average and extreme rainfall extent. According to the analysis, according to IMD rainfall, the seasonal average rainfall extremes in India are 48.8 mm per day. The uncorrected CMIP6 model assumed they were 25.8 mm per day, while the corrected rainfall is closer to observations.

Irregular rainfall in India

The IITM study adds to previous research that also found that popular climate models were underestimating the impacts of climate change on rainfall in the subcontinent. A 2020 study by researchers from the Indian Institute of Gandhinagar and Michigan State University found that bias-corrected projections of CMIP6 “project a warmer (3-5°C) and wetter (13-30 per cent) climate over South Asia in the 21st century.”

The research comes at a time when India's monsoon season is becoming increasingly erratic. A tehsil-level analysis of rainfall patterns by the Council on Energy, Environment and Water found that 55 per cent of tehsils saw an increase in southwest monsoon rainfall by more than 10 per cent from 2012 to 2022, and 11 per cent saw a decrease by more than 10 per cent each, compared to the climate baseline (1982-2011).

It also found that 48 per cent of tehsils “received more than 10 per cent rainfall in October, which may be due to the delayed withdrawal of the southwest monsoon from the subcontinent.”

Banner Image: Monsoon in Wayanad, Kerala. Image by Vinoth Chander via Wikimedia Commons (CC BY 2.0).