India Meteorological Department (IMD) chief Mrutyunjay Mohapatra said that weather scientists in India are now using artificial intelligence (AI) and machine learning to improve weather predictions.
Mohapatra, speaking to news agency PTI, emphasised that these emerging technologies will significantly enhance current numerical weather forecasting models in the next five years.
“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.
To spearhead this development, a dedicated team of expert from within IMD and the Ministry of Earth Sciences (MoES) has been formed. Collaborations with institutes such as Indian Institutes of Technology (IITs) and Indian Institutes of Information Technology (IIITs) have also been established to leverage their expertise in AI and machine learning, Mohapatra noted.
He mentioned that the weather office is expanding observational systems to provide mesoscale weather forecasts at the panchayat level or covering areas of over 10 square kilometres more rapidly.
Mohapatra noted that the IMD has deployed a network of 39 doppler weather radars, covering 85% of the country’s landmass, enabling hourly forecasts for major cities.
The IMD has digitised weather records dating back to 1901, and Mohapatra suggested that AI could be employed to analyse this extensive data to better understand weather patterns.
Explaining artificial intelligence models as data science tools, Mohapatra highlighted their ability to utilise historical data to improve forecasts, without delving into the physics of the phenomena.
“Both artificial intelligence and numerical forecasting models will complement each other to improve forecast accuracy. Both will work hand in hand and nobody can replace the other,” Mohapatra said.
AI reduce costs and boost efficiency
Weather agencies across the globe are increasingly turning to AI to enhance forecasting capabilities, aiming to reduce costs and boost efficiency. The predictive power of AI is that, it not only aids in understanding weather patterns but also plays a crucial role in combating climate change.
Various AI tools are being developed for accurate weather predictions globally. Projects like watsonx.ai, a collaboration between NASA and IBM, monitor environmental shifts and make future forecasts based on gathered data.
Similarly, Google’s DeepMind has developed GraphCast, an AI-driven weather forecasting model capable of delivering 10-day predictions within a minute.
According to a report by WEF, scientists at Google DeepMind said that GraphCast has outperformed conventional weather prediction methods, boasting a 90% verification rate, marking a significant advancement in weather forecasting and decision-making.
Why AI Outperforms Conventional Methods in Weather Prediction?
Unlike traditional forecasting methods reliant on costly computing power, AI models like GraphCast leverage historical weather data to predict weather patterns more economically and accurately. By identifying intricate patterns within datasets, AI enhances forecast precision in ways conventional equations cannot match.
Moreover, GraphCast demonstrates energy efficiency, being approximately 1,000 times more cost-effective than conventional forecasting techniques, the Financial Times reported.
(With inputs from agencies)
Comments
0 comment