In a significant step towards technology-driven agriculture, the Government of India has successfully conducted a pilot project using Artificial Intelligence (AI) to generate localized monsoon onset forecasts. The initiative aims to help farmers make timely and informed sowing decisions during the Kharif season. Implemented by the Department of Agriculture and Farmers’ Welfare, the project leverages advanced technology and weather data to deliver more precise and actionable insights to farmers.
The pilot was carried out in selected regions across 13 states for the Kharif 2025 season in collaboration with Development Innovation Lab-India. It used an open-source blended model combining Neural GCM, the AI Forecasting System (AIFS) developed by the European Centre for Medium-Range Weather Forecasts, and 125 years of historical rainfall data from the India Meteorological Department.
The forecasts specifically focused on predicting the local onset of the monsoon, a critical factor in determining the ideal sowing window. These probabilistic forecasts were shared with nearly 38.8 million farmers across 13 states through SMS alerts via the M-Kisan portal. To ensure accessibility, the information was delivered in five regional languages—Hindi, Odia, Marathi, Bengali, and Punjabi—enabling wider reach and usability.
Following the dissemination of forecasts, telephonic feedback surveys were conducted through Kisan Call Centers in Madhya Pradesh and Bihar. The findings revealed that 31% to 52% of farmers adjusted their agricultural decisions based on the information received. These adjustments included changes in land preparation, sowing dates, crop selection, and input usage, highlighting the practical value of accurate weather intelligence in farming operations.
Encouraged by the pilot’s success, the government is now preparing to integrate this capability into the national agricultural system. This will be achieved through an in-house technological framework developed in collaboration with leading institutions such as the Indian Institute of Tropical Meteorology, India Meteorological Department, and Indian Space Research Organization. Notably, dynamical models developed by IITM have demonstrated strong performance in predicting local monsoon onset and are set to be integrated into the AI framework by 2026.
FAQs:
What is the Khetivyapar AI monsoon forecast initiative?
It is a government-backed AI pilot project providing localized monsoon onset forecasts to assist farmers in timely sowing decisions.
Which states are covered under the AI pilot project?
The pilot was rolled out in selected regions across 13 states during the Kharif 2025 season.
How do farmers receive the forecasts?
Forecasts are sent via SMS on the M-Kisan portal in five regional languages: Hindi, Odia, Marathi, Bengali, and Punjabi.
How accurate are the local monsoon forecasts?
The AI model combines Neural GCM, AIFS, and 125 years of historical rainfall data, ensuring highly precise predictions.
How can Khetivyapar help farmers use these forecasts?
Khetivyapar provides guidance and actionable insights to help farmers adjust land preparation, sowing, and crop selection based on AI forecasts.
