Less than a week before the southwest monsoon is set to reach the Andaman Islands, the Union Ministry of Earth Sciences (MoES) and the India Meteorological Department (IMD) have launched two advanced, artificial intelligence-driven non-conventional weather forecasting systems designed to deliver hyper-local rainfall and other weather information for farmers, disaster management officials, and administrators as well as citizens.
The launch, held in New Delhi on May 12, was led by MoES minister Jitendra Singh in the presence of department secretary M. Ravichandran, director general of IMD Mrutyunjay Mohapatra, and director of IITM Pune Suryachandra Rao, as well as senior scientists and officials.
Monsoon prediction
Two AI-enabled systems—”Forecast of Monsoon Advance over Different Parts of the Country” and a “High Spatial Resolution Rainfall Forecast for Uttar Pradesh”—have been developed jointly by the India Meteorological Department (IMD), Indian Institute of Tropical Meteorology (IITM), Pune, and National Centre for Medium Range Weather Forecasting.
Minister Singh claimed that the new AI-enabled forecast mechanisms would provide probabilistic weather forecasts every Wednesday up to four weeks in advance.
“The products combine AI-based forecasting models, extended range prediction systems, and statistical techniques to provide operationally useful forecasts for agricultural planning and preparedness of farmers across 16 states and more than 3,000 sub-districts through the dissemination framework of the Ministry of Agriculture and Farmers’ Welfare,” shared a communiqué of the union government available with The Plurals.
Better than conventional predictions
According to information shared, IMD has launched a pilot project in Uttar Pradesh providing 1-km spatial resolution rainfall forecasts valid up to 10 days in advance. To achieve this, AI systems integrate data from Automatic Rain Gauges (ARGs), Automatic Weather Stations (AWS), Doppler Weather Radars (DWRs), and satellite-based datasets.
“…The number of Doppler weather radars, which was 16 to 17 a decade ago, has now reached 50, and there are plans to have another 50 under Mission Mausam,” said the minister, claiming that the modernization of India’s weather infrastructure has significantly improved the forecasting capabilities and early warning system across the country.
Doppler stations are ground-based radar facilities to measure the velocity and direction of moving objects, such as precipitation or wind, and provide real-time, high-resolution data crucial for weather forecasting.
The ministry claims that “AI-driven models have shown to significantly reduce mean absolute errors (up to 11.1% in some studies) compared to conventional interpolation methods.” According to him, India has witnessed nearly 40 percent improvement in forecast accuracy for severe weather events.
Service to be expanded
Singh pointed out that early warnings are now being disseminated through multiple channels, including mobile applications, SMS alerts, WhatsApp, Kisan portals, television, and other digital platforms to ensure wider public outreach and last-mile connectivity.
“…Systems have been developed in response to increasing demand from agriculture,” said Ravichandran, adding that “similar services would gradually be expanded to other parts of the country as observational infrastructure continues to grow.”
The ministry claimed that the newly launched initiative would be particularly useful for managing water resources, renewable energy, urban planning, and disaster management apart from the agricultural sector.
“While advanced technology is always welcome in weather forecasting, we have to keep in mind that implementing advanced forecasting tools requires significant investment in specialized software, technology, and hiring highly skilled data personnel; getting all of which in tandem may turn out to be key challenges,” opined a weather expert.