Jitendra Singh, Minister of State (Independent Charge) for Earth Sciences, told the Lok Sabha in written replies that Artificial Intelligence (AI), Machine Learning (ML) and Big Data are being integrated into weather forecasting and early warning under Mission Mausam. Most of these AI/ML systems are still experimental, he said, so no separate quantitative assessment has been made of their contribution to reducing loss of lives, property and livelihoods.
The National Centre for Medium Range Weather Forecasting (NCMRWF), under the Ministry of Earth Sciences, is combining AI/ML-based guidance with data assimilation, coupled Earth System modelling, ensemble prediction systems, high performance computing and conventional Numerical Weather Prediction (NWP) models. The India Meteorological Department (IMD) uses this guidance to improve forecasting skill. AI/ML-derived products have also been integrated into the Multi-Hazard Early Warning Decision Support System (MHEW-DSS), a GIS-based system developed by IMD.
The Minister said AI complements conventional NWP by speeding up forecast generation, reducing systematic biases, producing probabilistic guidance for extreme events and giving high-resolution downscaled forecasts. The listed initiatives include:
On the user side, MausamVani is a Generative AI application, being developed as a Retrieval-Augmented Generation platform, for weather-based decision support in regional languages. MausamGPT, a large language model application, is under development to produce short multilingual summaries of forecasts and warnings. An AI/ML-dynamical hybrid model produced localised monsoon onset advisories, and during the 2026 southwest monsoon season SMS advisories reached about 5.28 crore farmers across 15 States and one Union Territory.
IMD sends forecasts through mobile applications, SMS, television, radio, websites and social media, with AI-enabled multilingual tools including Bhashini for regional languages. Impact-Based Warnings also travel through Central and State agencies, disaster management authorities and media.
IMD's network has 50 Doppler Weather Radars, 1,008 Automatic Weather Stations and 6,885 rain gauge stations. The Minister said severe weather forecast accuracy has improved by 40-50% over the past decade, with lower cyclone track, landfall and intensity errors. For floods, 25 Flood Meteorological Offices issue daily 7-day Quantitative Precipitation Forecasts for 220 sub-basins to assist the Central Water Commission (CWC).
The National Seismological Network has grown from 86 observatories in 2014 to 174, with 89 in Seismic Zones IV and V. INCOIS is the designated Tsunami Service Provider for 28 Indian Ocean Rim nations, and 26 coastal villages in Odisha have been recognised under the UNESCO-IOC Tsunami Ready programme. The proposed 2026-27 allocation for NCPOR under the Polar Science and Cryosphere Research (PACER) scheme is ₹270 crore.
IMD said it issued district-level rainfall forecasts and colour-coded impact-based warnings for all Upper Assam districts, including Jorhat, Golaghat, Sivasagar and Charaideo, for heavy to very heavy rainfall on 18-20 July 2026. The events were captured with a 7-day lead time. A monsoon trough, a low-pressure area over Gangetic West Bengal and upper air cyclonic circulations over Assam, Bihar and Jharkhand sustained the rainfall.
IMD's verification for Day 1 to Day 5 shows lower skill at longer leads. In West Assam, Probability of Detection (POD) falls from 0.6345 to 0.1209 while the False Alarm Ratio (FAR) rises from 0.738 to 0.9613. Central Assam's POD falls from 0.6929 to 0.0757. In Northeast Assam the Percentage of Correct forecasts (PC) goes from 58.47% to 51.61%, lower than the other regions, though its FAR of 0.3913 to 0.5538 is the lowest.
IMD's seasonal outlooks use dynamical and statistical models with ENSO and the Indian Ocean Dipole (IOD) among their inputs. Moderate El Niño conditions currently prevail with a neutral IOD. El Niño is generally linked to below-normal monsoon rainfall over most of India, but the ministry said its relationship with Northeast rainfall is weak and varies from event to event, because rainfall there depends on Bay of Bengal moisture, monsoon circulation and orographic effects. Flood forecasts are made by the CWC from observed and forecast rainfall, river conditions and hydrological modelling.
IMD is working on an integrated National AI-based Early Warning Platform to produce Impact-Based Forecasts and Risk-Based Warnings. For the Northeast, the 6 km BharatFS, run on the Arunika and Arka computing systems, has been operationalised, and the region's observation network is being upgraded.
With reference to the use of Artificial Intelligence in weather forecasting in India, consider the following statements:
Which of the statements given above is/are correct?
Consider the following pairs:
Initiative / Application : Description given in the article
How many of the pairs given above are correctly matched?
Consider the following statements:
Statement-I: The ministry says El Niño has only a weak and event-to-event varying relationship with rainfall in Northeast India, although El Niño is generally linked to below-normal monsoon rainfall over most of India.
Statement-II: Rainfall in Northeast India depends on Bay of Bengal moisture, monsoon circulation and orographic effects.
Which one of the following is correct in respect of the above statements?
