India Tests Artificial Intelligence To Build Climate Models And Increase Weather Forecast

Parthasarathi Mukhopadhyay, a climate scientist at the Indian Institute of Tropical Meteorology. (photo: dock. Indian Institute of Tropical Meteorologist)

JAKARTA - India is testing artificial intelligence (AI) to build climate models to improve weather forecasts in line with the increasing heavy rains, floods, and drought across the country. This was said by a prominent weather official.

Global warming has sparked more intensive clashes of weather systems in India in recent years, raising extreme weather events, where the independent Center for Science and Environment estimates it has killed nearly 3.000 people this year.

Weather agencies around the world focus on AI, which can lower costs and increase speed, and which the UK says could "revolution" weather forecasts, with Google-funded models recently found to be superior to conventional methods.

Accurate weather forecasts are essential in India, a country with 1.4 billion inhabitants, many of whom are poor, and are the second largest producers in the world for rice, wheat and sugar.

The India Meteorological Department (IMD) provides predictions based on mathematical models using supercomputers. Using AI with expanded observation networks can help generate high-quality forecast data at lower cost.

The department hopes the AI-based climate model and the advice it is developing can help improve forecasts, according to KS Hosalikar, head of climate research and services at IMD told Reuters.

The weather office has used AI to produce public warnings regarding heat waves and diseases such as malaria, Hosalikar said. They plan to upgrade weather observatories, providing data up to the village level, potentially providing data with higher resolutions for forecasting.

The government said on Thursday December 21 that it wanted to produce weather and climate forecasts by combining AI into traditional models, and had set up a center to test the idea through workshops and conferences.

"The AI model does not require high costs involved in running a supercomputer - you can even run it from a good quality desktop," said Saurabh Rathore, an assistant professor at the Indian Institute of Technology-Delhi.

Experts also say that better data is needed to maximize the use of AI.

"Without having high-resolution data in space and time, it is impossible for an AI model to strengthen locations from existing model forecasts," said Parthasarathi Mukhopadhyay, a climate scientist at the Indian Institute of Tropical Meteorology.