Director General of the India Meteorological Department Mrutyunjay Mohapatra stated that weather scientists in India have begun utilizing AI and machine learning to improve forecasts.
During an informal conversation with PTI editors, he stated that in the coming years, new technologies will supplement numerical weather forecasting models, which are already widely utilized for weather prediction.
In order to produce mesoscale weather forecasts at the panchayat level or for a region larger than 10 sq km more quickly, he said the meteorological office has been expanding its observing systems. According to Mohapatra, the India Meteorological Department (IMD) has set up a network of 39 doppler weather radars that cover 85% of the nation’s landmass and allow for hourly forecasts for major cities.
“We have started using Artificial Intelligence in a limited way but within the next five years, AI will significantly enhance our models and techniques,” he stated.
According to Mohapatra, artificial intelligence might be used to sort through the vast amount of data that the IMD has digitized—dating back to 1901—to produce understanding about weather patterns.
According to the director general of IMD, artificial intelligence models are data science models that use historical data to provide information that can be applied to improve forecasts rather than delving into the physics of the phenomenon.
He noted that in order to use artificial intelligence, specialist groups have been established at the IMD and the Ministry of Earth Sciences. “To increase forecast accuracy, artificial intelligence and numerical forecasting models will work in tandem. Nobody can take the role of the other, and both will collaborate closely,” Mohapatra remarked.
In response to the demand for hyper-localized forecasts, Mohapatra noted that IMD faced difficulties in providing village-level forecasts for certain dangers.
“We aim to provide forecasts at the Panchayat or village level…tailoring weather information to sector-specific needs in agriculture, health, urban planning, hydrology, and environment,” he stated.
The head of IMD emphasized the significance of data-driven decision-making in the age of abundant information.
“Incorporating AI and machine learning allows us to harness past data to extract valuable insights and improve forecasting accuracy without solely relying on traditional physics-based models,” he stated.
Regarding how climate change affects weather predictability, Mohapatra pointed out that small-scale mesoscale phenomena like convective clouds are emerging and having an impact on nearby towns.
He claimed that in order to combat this, IMD has positioned Doppler weather radars strategically across 85% of the nation.
He said that by enabling the observation and simulation of convective clouds, this sophisticated radar data—which has a resolution of 350 meters per pixel—significantly improves forecast accuracy for extreme occurrences like cyclones and torrential rainfall.
