PARLIAMENT QUESTION: LONG RANGE FORECAST SYSTEM

Since 2021 the India Meteorological Department (IMD) under the Ministry of Earth Sciences has employed an advanced Multi‑Model Ensemble (MME) approach for long‑range forecasting. The MME system, which combines multiple coupled dynamical climate models, has reduced the mean absolute forecast error to 2.2 % of the Long Period Average (LPA) for the 2021‑2025 period, compared with 7.8 % for 2016‑2020. Table data show that, for all‑India monsoon rainfall, the first‑stage forecasts (issued in April) and second‑stage forecasts (issued in May) have stayed within the ±5 % and ±4 % error bands respectively, with only one exceedance (14 % error) in 2019. The government is further enhancing forecasting capability through upgrades of numerical weather prediction models, expansion of observation networks (Doppler radars, AWS, ARG, satellite platforms) and the use of AI/ML tools, with forecasts disseminated via mobile apps, APIs, SMS, TV, radio and social media.