To make India more climate-resilient, India Meteorological Department’s upgrade is necessary
The weather forecasting upgrade planned by the India Meteorological Department (IMD) is long overdue. In the past 10 years, the IMD’s ability to make long-range predictions over broad regions has improved appreciably. Yet, there are several days when predictions go wrong, especially during the southwest and northeast monsoons. The challenge today is not to predict average rainfall or temperature over the season, month or week. Extreme weather events that usually occurred once in a few decades now threaten people’s lives and livelihoods throughout the year. Weather forecasters need to find ways to alert farmers, municipal authorities, and office and schoolgoers about copious rainfall at hyper-local levels. Climate-induced vagaries have triggered disasters like the recent landslides in Wayanad and lake bursts in Sikkim and Uttarakhand last year. Administrators have often been caught off guard by the fury of the elements. Last year, when Kalyanapattinum in Tamil Nadu’s Thoothukudi district experienced an entire season’s rainfall in a day, Chief Minister M K Stalin blamed the IMD for not issuing advance alerts. The problem is that predicting such intense rainfall is virtually impossible today. States such as Kerala, Odisha and Maharashtra have sought the help of private agencies to augment IMD’s information. However, the fact remains that large gaps in knowledge will have to be plugged to make the country climate-resilient.
While the previous improvements in IMD focused primarily on augmenting infrastructure, the latest endeavour will be directed at developing computer simulated models tailored to local specifics. This would require data collectors to narrow their focus to district, block, panchayat, village or even ward and street-levels. The scientists will also need to be equipped with a denser network of measuring instruments. Recent research has shown that AI can improve the accuracy of weather forecasts. The IMD has reportedly digitised the country’s weather records going back to more than a century. AI could be used to sift through this corpus to generate knowledge and help forecasters issue timely alerts. The met department would also do well to tap into research conducted in universities and institutions. Scientists at IIT Bombay, for example, used cutting-edge computing technologies this year to predict rainfall in the city with greater accuracy than the IMD. Another ongoing project at IIT Mandi hones into soil characteristics to predict landslides.
Mainstreaming information on erratic weather would also require communicators conversant with local economic and cultural idiosyncrasies. The information will need to be disseminated to the most vulnerable. The IMD’s endeavour will be closely watched.