India Meteorological Department (IMD) has given great news amid the 150th anniversary of its establishment. The department has now focused on improving weather forecasting models using AI and fast supercomputers.
IMD Director General Mrityunjay Mohapatra said that the weather office is trying to better understand the process of formation of clouds that cause thunderstorms and heavy monsoon rains.
For this, experiment centers are being established in Odisha and Madhya Pradesh.
The IMD will launch ‘Panchayat Mausam Seva’ on Monday, marking its 150th anniversary, which aims to take weather forecast to every farmer in every village.
Mahapatra said the IMD has come a long way, from issuing cyclone warnings at Kolkata port in the late 19th century to weather forecasts reaching mobile phones.
He said it now plans to use the latest advances in AI and machine learning to improve its forecasting capabilities.
Help is also being taken from NIT and IIT
Mohapatra said, ‘Being an old organisation, IMD has weather data since 1901 which has been digitalised.
“AI is a data science where we can use all this historical data to develop tools and methods for prediction.”
Mahapatra said IMD has constituted an expert group on AI-machine learning and is collaborating with NITs (National Institutes of Technology), IITs (Indian Institutes of Technology), IIITs (Indian Institutes of Information Technology), academic and research institutes for joint development of tools. Also collaborating with development (R&D) institutions. He said it could be used to improve forecast accuracy and in field applications.
Fast supercomputer will also be used
The IMD chief said the weather office is further developing high-powered computing systems to improve its numerical modeling capability.
He said the faster supercomputers will help the IMD gather more area-specific and location-specific information. The weather office was criticized for wrongly forecasting heavy rains in southern Tamil Nadu in December.
When asked about this, Mahapatra admitted that the forecasting system is not capable of predicting weather phenomena like cloud bursts and small-scale severe weather events. Mahapatra said, ‘I encourage criticism.
There should be criticism so that we learn lessons and improve our forecasting system. But, still there are some challenges like cloud burst incident. Rainfall of 60–90 cm occurs in very exceptional cases. This system is not currently capable of predicting these types of weather systems.
Post Views: 5