Kerala’s rice granary faces steep losses as the climate warms, a modelling study has found. It projects that Palakkad paddy yields in two farm blocks of the Bharathapuzha basin could fall by more than half by the end of the century. In the worst case, the fall reaches about 74%.
The paper, “Assessing climate change effects on streamflow and paddy production in the Bharathapuzha Basin Kerala”, was published in Scientific Reports. The first author is Gulja S. Nair of the Department of Agricultural Engineering at the Nehru Institute of Technology, Coimbatore. The corresponding authors are D. Karunanidhi of the Department of Civil Engineering at Mohan Babu University, Tirupati, and Berihu Abadi Berhe of the School of Earth Sciences at Mekelle University, Ethiopia.
What the researchers did
The Bharathapuzha rises at 1,965 m in the Anamalai Hills and runs about 209 km to the Arabian Sea at Ponnani. Its basin covers about 6,186 km² of Palakkad, Thrissur and Malappuram districts, with small parts in Tamil Nadu. The study looked closely at two blocks in Palakkad district: Palakkad and Alathur. The paper calls the Palakkad block the “granary of Kerala”.
The team used a computer model called SWAT. It tracks how rain turns into river flow and soil moisture, and how much water a crop gets. They first tested it against real records from 1998 to 2017. These included rainfall and temperature from the India Meteorological Department, river flow at Kuttipuram from the Central Water Commission and paddy yields from Kerala’s farm statistics.
They used 1998 to 2008 to tune the model and 2009 to 2017 to test it. Rainfall in the baseline years swung widely, from about 700 mm in 2003 to more than 2,100 mm in 2007. The model matched the past closely. On a common test where 1 means a perfect match, it scored 0.74 for monthly river flow during calibration. For paddy yields it scored 0.80 in Palakkad and 0.68 in Alathur.
The researchers then fed in future weather from three global climate models, known as EC-Earth3-Veg, INM-CM5-0 and MPI-ESM1-2-HR. The data had been corrected and scaled down to about 25 km squares by IIT Gandhinagar. They used two futures: a middle path, known as SSP2-4.5, and a high-emissions path, SSP5-8.5. They checked yields in 2035, 2060, 2080 and 2100 against the 1998 to 2017 average.
What they found
Losses start early. By 2035, the model shows Palakkad block yields down by 24% to 33% across the models and scenarios. Alathur shows a wider spread, from 16% to 42%.
The losses deepen with time. By 2060, Palakkad yields are down 38% to 47% and Alathur yields 41% to 54%. By 2080, Palakkad is down 49% to 58% and Alathur 48% to 69%. By 2100, the model puts Palakkad paddy yields 54% to 68% below today’s and Alathur’s 63% to 74% below. The single worst figure is a 73.86% fall in Alathur under high emissions.
The authors point mainly to water as the cause. Warmer weather dries the soil faster and lowers river flow in the dry months. The crop then runs short of water when it flowers and fills its grain. At the other extreme, heavier monsoon rain brings bigger floods. Floods waterlog fields, damage seedlings and delay transplanting. Citing earlier studies, the paper notes that tropical rice loses 3% to 10% of its yield for every 1 °C of warming.
Alathur is hit harder in the near term, the authors found, while Palakkad’s losses grow more later in the century. Their rainfall projections show more ups and downs from year to year. River flood peaks sharpen, especially under high emissions after 2060.
Limits listed by the authors
The authors are open about the model’s limits. It used standard rice settings that were not tuned to local varieties, because detailed field data were not available. The exact size of the losses is therefore uncertain, though the authors consider the direction of change reliable.
The model also kept land use, soils, irrigation and farming practices fixed. In effect, it assumes farmers change nothing for 80 years. River flow was checked at one gauge only, and the projections rest on three climate models and two scenarios. The paper’s summary figures also differ slightly from its own tables; this report uses the table values.
The study arrives as Kerala’s rice area has already shrunk. Citing earlier work, the authors note that the state’s paddy area fell from more than 880,000 ha in the 1970s to under 200,000 ha in recent years.
The authors list practical steps to soften the blow. These include short-duration and flood-tolerant varieties and moving sowing and transplanting dates away from flood peaks and dry spells. They also suggest alternate wetting and drying, in which fields are allowed to dry briefly between irrigations to save water, repair of canals and tanks, and limits on groundwater pumping. Where losses pass about 70%, they suggest moving to crops that need less water. Because losses of a quarter or more appear in the Palakkad block as early as 2035, the window for its farmers and planners to act is short.
Sources: Nair, G. S., Karunanidhi, D. and Berhe, B. A. (2026). Assessing climate change effects on streamflow and paddy production in the Bharathapuzha Basin Kerala. Scientific Reports, 16, 21091. https://doi.org/10.1038/s41598-026-52356-3. Licence: CC BY-NC-ND 4.0. Earlier work cited: Thomas (2011); Kumar and Kunhamu (2021); Prasad and Kuruvila (2024), as cited in the paper. all findings are from the paper.
