As climate change makes weather forecasting increasingly challenging, the India Meteorological Department (IMD) on Tuesday launched the country’s first-ever Artificial Intelligence (AI)-enabled monsoon advance forecasting model that would predict the onset of monsoon accurately even at the block level every Wednesday up to four weeks in advance.
Along with this, the Met department also launched a pilot prepared by the National Centre for Medium Range Weather Forecasting (NCMRWF) on high spatial-resolution rainfall forecast for Uttar Pradesh, with a resolution of up to one-kilometre grid, also using advanced AI systems.
The models were launched at a time when the 2026 southwest monsoon is projected to be below normal across most parts of the country, though the IMD categorically said that there is no correlation between the two.
It said both the models aimed at delivering hyper-local, impact-based, and AI-driven weather services across the country have been developed under the guidance of the Ministry of Agriculture and will be passed on to farmers through APIs developed by the ministry and also through the AgriStack platforms.
“The block-level monsoon onset forecast model combines existing numerical models with AI to generate probabilistic forecasts of monsoon progression every Wednesday up to four weeks in advance, but with a model error of four days,” Union Minister for Earth Sciences Dr Jitendra Singh told reporters while launching the two models.
The advance block-level forecasting system will at present cover 3,196 blocks spread across 15 states and one Union Territory, largely lying in the rainfed areas of the country, but will be gradually expanded to cover other regions of the country.
The block-level monsoon advance forecasting system will primarily base its observations on the beginning of a continuous five-day rain spell and absence of a prolonged dry spell for the next 30 days, Dr M Ravichandran, secretary in the Ministry of Earth Sciences, said.
“This innovative system will be of immense help in advance agricultural planning, disaster management, water resources management, etc,” Singh said.
On the pilot for high spatial-resolution rainfall forecast for Uttar Pradesh, Jitendra Singh said that the pilot will generate rainfall forecasts at 1-km spatial resolution up to 10 days in advance.
The system uses advanced AI-driven downscaling techniques and integrates data from Automatic Rain Gauges (ARGs), Automatic Weather Stations (AWSs), Doppler Weather Radars, and satellite-based rainfall datasets. It will also help farmers make informed decisions relating to sowing, irrigation, crop protection, and harvest planning with far greater local precision.
MONSOON ADVANCE: In a related development, the IMD on Tuesday said that conditions are becoming favourable for the onset of southwest monsoon over parts of south Bay of Bengal, Andaman Sea, and Andaman & Nicobar Islands towards the end of this week. The Met department will update on the further progress of the southwest monsoon later this month.
It said in other parts of the country, heatwave conditions are expected at a few places over west Rajasthan during the next 6-7 days and also over east Rajasthan and adjoining central India, Gujarat, and Maharashtra during the next 4-5 days. EoM
AI models to boost block-level monsoon, rainfall forecasts
A) IMD launched AI-enabled block-level monsoon advance model
B) IMD also launched a pilot on high spatial-resolution rainfall forecast for Uttar Pradesh
C) The spatial forecast can accurately predict rainfall in a one-kilometre grid, down from 12.5 km
D) The AI-enabled block-level monsoon onset forecast model now covers 3,196 blocks across 16 states and UTs
E) Most of the blocks are in rainfed areas where monsoon onset is critically important for farming and disaster management
