A study published in the journal Meteorological Applications said the AI model developed by climate experts Cris Gino Mesias and Gerry Bagtasa can spot patterns more efficiently while using the same information about the country’s previous typhoons.

“Most predictions of tropical cyclone rainfall rely on dynamic models, which are very difficult to run as they take a lot of computational resources and require high-performance computing,” Bagtasa said.

“When we assessed the AI model, its predictive skill was comparable to a dynamic model that we regularly use. The AI model had better skills for extreme rainfall from tropical cyclones,” he added.

Parameters mostly influencing the AI model’s forecast are cyclones’ distance and duration, which determine who will be affected by heavy rains and how much rain the country will experience, he noted.

Compared to previous models, Bagtasa said the AI model can run within minutes on a laptop.

“This AI model, admittedly, is not perfect. But it can add to the suite of rainfall forecast models available to equip our disaster managers with more information on impending hazards,” he said.

Fresh data can be uploaded to the AI model, allowing it to relearn and improve its accuracy, he added.

ChatGPT and Gemini are large language models that are different from the AI model they developed, Bagtasa noted.

“Some AI models, such as those for weather forecasting, can be useful and more efficient than conventional methods. But there are also some, like large language models, that consume so much energy, leading to environmental impacts that are harmful to the planet,” he said.

Bagtasa and Mesias are from the UP Diliman College of Science’s Institute of Environmental Science and Meteorology.

Their study titled “AI-Based Tropical Cyclone Rainfall Forecasting in the Philippines Using Machine Learning” was supported by the Department of Science and Technology-Accelerated Science and Technology Human Resource Development Program and DOST-Philippine Council for Industry, Energy and Emerging Technology Research and Development.