The India Meteorological Department (IMD) has revolutionized its long-range monsoon forecasting by adopting a
Multi-Model Ensemble (MME) approach, which integrates dynamical global climate models with statistical methods, achieving high accuracy. As of 2025, this system has demonstrated strong, accurate performance for Southwest Monsoon (June-September) rainfall forecasts, with a model error of ± 4%.
Key Aspects of the New Forecasting Strategy
- High Accuracy & Reliability: The new system has consistently provided accurate seasonal forecasts (e.g., 2021-2025). In 2025, the April forecast for 105% of the long period average (LPA) closely matched the actual, demonstrating high skill.
- Six-Month Lead Time: The new framework allows for earlier, more strategic, and updated assessments. The IMD now issues its first-stage LRF in mid-April, allowing for a comprehensive outlook well ahead of the June start, followed by updates in May.
- Dynamical MME Approach: The system uses coupled global climate models (CGCMs) from various global centers, including India’s own Monsoon Mission Climate Forecasting System (MMCFS).
- “Mission Mausam” Integration: The new “Mission Mausam” initiative (launched late 2024/early 2025) boosts this further with advanced AI/ML technologies and increased supercomputing power (using “Arka” and “Arunika” for a combined 28 Peta FLOPS) to improve precision.
- High-Resolution Modeling (BharatFS): The new “Bharat Forecast System” (BFS) can operate at a 6 km resolution—the highest in the world—to provide localized forecasts for smaller areas.Â
What the New Model Changes
- Earlier Actionable Information: The six-month lead time gives planners, policymakers, and agricultural authorities substantial time to prepare for potential deficits or surpluses, strengthening early warning systems.
- Improved Spatial Accuracy: It allows for forecasting regional variations (e.g., Northwest, Northeast, Central India) rather than just a national average, which is critical for agricultural planning.
- Shift from Statistical to Dynamical: Moving from traditional purely statistical methods to MME-based dynamical models reduces dependency on older data correlations and better simulates complex tropical weather patterns.
- Better Agricultural Planning: District-level and block-level forecasts help farmers optimize sowing, irrigation, and crop protection, reducing the risk of crop failure.
- Enhanced Disaster Management: Provides early warnings for heavy rainfall and flooding, allowing for better reservoir management and disaster response.Â
The new system significantly enhances India’s ability to be a “weather-ready and climate-smart” nation by reducing the uncertainty associated with seasonal forecasts, particularly in the context of changing climate patterns.
