IMD’s Multi-Model Forecasting Cuts Monsoon Prediction Error to 2.2%: Centre
New Delhi, July 29 (TNT): The India Meteorological Department (IMD) has substantially improved the accuracy of its long-range monsoon forecasts since adopting an advanced Multi-Model Ensemble (MME) forecasting system in 2021, reducing the average forecast error to 2.2 per cent of the Long Period Average (LPA) during 2021-25, the Centre informed the Lok Sabha on Wednesday.
In a written reply, Minister of State (Independent Charge) for Earth Sciences Dr. Jitendra Singh said all long-range forecasts issued since 2021 have remained within the prescribed error limits.
He said the average absolute error in the first-stage monsoon forecast during 2021-25 was 3.1 per cent of the LPA, while the second-stage forecast recorded an average error of 2.2 per cent, compared with 7.8 per cent during 2016-20.
According to the minister, the forecast error exceeded 10 per cent of the LPA only once in the last decade, in 2019, when the absolute error was 14 per cent.
Dr. Singh said the MME system combines forecasts from multiple coupled dynamical climate models, reducing uncertainties associated with individual models and improving the representation of large-scale climate drivers such as the El Niño-Southern Oscillation (ENSO) and the Indian Ocean Dipole (IOD), resulting in more reliable seasonal monsoon predictions.
The minister said the government has taken several measures to further improve long-range weather forecasting and prediction of extreme weather events, including upgrading numerical weather prediction models, expanding meteorological observation networks under Mission Mausam, strengthening satellite-based observations, deploying high-performance computing and artificial intelligence-based forecasting tools, and enhancing dissemination of weather forecasts and warnings through digital and conventional communication platforms.
He said these measures are being implemented uniformly across the country and no state-wise allocation has been made.
TNT KS
