UP scientists develop AI model to improve rainfall forecasting

Scientists at the University of the Philippines Diliman College of Science’s Institute of Environmental Science and Meteorology (UPD-CS IESM) have developed an artificial intelligence (AI) model that links past tropical cyclone (TC) tracks to recorded rainfall in order to help improve disaster preparedness.

Dr. Gerry Bagtasa and Cris Gino Mesias cited how the country is often hit by cyclones, whose patterns more often than not repeat.

– Advertisement –

For example, if a typhoon with a certain amount of rainfall passes through Central Luzon, a similar typhoon that passes through the same region again in the future will likely have the same amount and distribution of rainfall.

The said AI model uses the same information about Philippine tropical cyclones, while spotting patterns more quickly and efficiently.

Compared to previous models, the one developed by these UP scientists can run within minutes on a laptop.

“Most predictions of TC 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.

Graphical abstract of the AI-based accumulated tropical cyclone (TC) rainfall model. (Courtesy: Mesias & Bagtasa, 2025)

The scientist explained that the distance of the tropical cyclone and its duration are the parameters that most influenced the AI model’s rainfall predictions, as these mainly determine who will be affected by heavy rains and how much rain the country will experience.

A typhoon near Batanes, for instance, would not be expected to cause heavy rains in Mindanao.

However, slow-moving tropical cyclones that spend more time over land also tend to bring more rainfall overall.

“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,” Bagtasa said.

Fortunately, the model can be updated with fresh data, allowing it to relearn and improve its accuracy.

The AI model developed by Mesias and Bagtasa is different from large language models (LLMs) like ChatGPT and Gemini.

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

Their study “AI-Based Tropical Cyclone Rainfall Forecasting in the Philippines Using Machine Learning” is published in Meteorological Applications. The research was likewise supported by the Department of Science and Technology–Accelerated Science and Technology Human Resource Development Program (DOST-ASTHRDP) and the DOST-Philippine Council for Industry, Energy, and Emerging Technology Research and Development (DOST-PCIEERD).

– Advertisement –

_{Images are for reference only.Images and contents gathered automatic from google or 3rd party sources.All rights on the images and contents are with their legal original owners.}