In India this summer 38 million farmers received AI-assisted monsoon forecasts produced by a hybrid model called NeuralGCM. The forecasts arrived about four weeks before the usual monsoon onset. Researchers said the model correctly signalled a three-week pause in the monsoon after it made landfall in early June and started moving north.
NeuralGCM combines physics-based forecasting with machine learning and was developed by Google. The University of Chicago team benchmarked it against conventional physics models and other AI models and reported strong performance and greater computational efficiency. Chicago researchers, supported by the Gates Foundation, are benchmarking models for East and West Africa, focusing on rainy seasons and heatwaves.
Officials and project partners say the forecasts were used to advise farmers about planting and risk management. The software can run on a laptop, making forecasts more widely available than models that need expensive supercomputers. The initiative currently works with five countries and plans more expansion in 2026 and 2027.
Difficult words
- predict — to say what will happen in the futurepredicting, prediction
- forecast — a statement about what will happen in the futureforecasting, forecasts
- agriculture — the practice of farming and growing crops
- technology — the use of scientific knowledge for practical purposes
- development — the process of growth or improvement
Tip: hover, focus or tap highlighted words in the article to see quick definitions while you read or listen.
Discussion questions
- How do you think AI will change farming in the future?
- What other technologies could help farmers?
- Why is accurate weather information important for agriculture?
Related articles
Study finds flaws in cloud password managers
Researchers at ETH Zurich tested three cloud-based password managers and found multiple attacks that could expose or change users' passwords. They followed responsible disclosure, gave companies time to fix the issues, and recommended stronger encryption and audits.