Artificial intelligence is significantly enhancing the accuracy and speed of weather forecasting in India, according to officials from the India Meteorological Department. Scientists are now using AI tools to analyze vast amounts of weather data more efficiently, enabling more reliable predictions of changing weather patterns, including short-term variations that were previously harder to detect.
One major improvement comes from advanced forecasting systems like the Bharat Forecast System, which operates on a higher-resolution grid (6 km compared to the earlier 12 km). This upgrade has led to a 30% increase in accuracy for predicting extreme rainfall events, making forecasts more precise at local levels.
The use of powerful supercomputers such as ‘Arka’ and ‘Arunika’ has also transformed forecasting efficiency. These systems have reduced the time required to generate forecasts from around 12 hours to just 3–6 hours, allowing meteorologists to deliver faster and more timely updates on events like monsoons, fog, and western disturbances.
Additionally, AI-driven models and techniques—such as ensemble forecasting, impact-based forecasting, and machine learning systems—have improved severe weather prediction accuracy by 30–40% over the past decade. Looking ahead, the IMD plans to adopt advanced AI models like GraphCast and Pangu, further strengthening India’s ability to predict extreme weather and improve disaster preparedness.
