AI-powered applied sciences enhancing accuracy of climate forecasts: IMD official
Abhishek Anand, Deputy Director on the IMD centre, mentioned AI-powered instruments are enabling extra exact information evaluation, enhancing reliability in predicting climate sample modifications.
“AI tools and other advanced technologies are being used by IMD scientists for data analysis, enabling more reliable predictions about changes in weather patterns with less computational power. With the use of these new technologies, even short-term changes in weather patterns are being easily detected,” he mentioned whereas addressing a programme right here.
Anand mentioned the Ranchi centre adopted IMD’s Bharat Forecast System (Bharat FS) in May final yr, that includes a 6-km decision grid, an enchancment over the sooner 12 km mannequin, leading to a 30 per cent rise in excessive rainfall prediction accuracy.
He mentioned supercomputing upgrades similar to ‘Arka’ and ‘Arunika’ have lowered forecast runtime from 12 hours to 3 to 6 hours, enabling sooner and extra correct short-range forecasts for monsoons, western disturbances and fog.
Mission-specific instruments, together with ensemble forecasting, Impact-Based Forecasting (IBF), and AI/ML fashions similar to ‘Mithuna-FS’, have improved extreme climate prediction accuracy by 30-40 per cent over the previous decade, Anand mentioned.
The IMD can also be planning to induct AI-based climate fashions similar to ‘GraphCast’, ‘Pangu’ and ‘ForecastInternet’, whereas Doppler radars are proposed to be put in in Ranchi and Deoghar to strengthen real-time monitoring of nor’westers and heavy rainfall, he mentioned.
Anand additionally emphasised the necessity for collective duty in direction of environmental safety to assist extra correct climate forecasting.
