India Advances AI-Driven Weather Forecasting for Climate Extremes

India’s weather forecasting is evolving rapidly as experts call for integrating artificial intelligence with traditional physics-based models to tackle increasingly unpredictable
climate events. At the India AI Impact Summit 2026, Dr. M. Ravichandran, Secretary of the
Ministry of Earth Sciences presented a clear roadmap during a panel on “Harnessing AI to
Manage Climate Extremes and Build Sustainable Systems.” He emphasized that while numerical models handle large-scale weather patterns effectively, they struggle with
hyper-local, short-term extremes like cloudbursts and flash floods, precisely where AI excels by analyzing time-series data and reducing prediction biases.
Dr. Ravichandran used a vivid analogy: scientists have long tracked the “elephant” of major
weather systems, but climate change demands spotting the “ant” riding it, the subtle,
localized signals that trigger disasters. AI can downscale forecasts to 1 km resolution,
improving accuracy for district-level alerts, agriculture planning, and urban resilience,
especially critical in vulnerable Himalayan regions prone to landslides and glacial lake
outbursts. This hybrid approach refines initial conditions and corrects model errors,
addressing gaps in forecasting convective storms and heavy rainfall events that have
intensified in recent years.
The panel, featuring National Disaster Management Authority Secretary Er. Manish
Bhardwaj and NVIDIA’s Dr. Karthik Kashinath highlighted multidisciplinary collaboration,
breaking silos between physicists, data scientists, biologists, and domain experts, to build
data-driven systems. Organized by the Indian AI Research Organisation, Atria University,
C-DAC, IITM, and LokNeeti, the summit underscores India’s push toward AI-enhanced
meteorology, building on the Ministry of Earth Sciences’ BharatFS models for national resilience amid rising climate threats.