Artificial intelligence (AI) is transforming weather prediction, but it will remain a tool — not a replacement — for human meteorologists, according to the Philippine Atmospheric, Geophysical and Astronomical Services Administration (PAGASA).

At the HANDA Pilipinas 2025 Visayas Leg, PAGASA Administrator Dr. Nathaniel Servando said AI is helping improve the accuracy, timeliness, and impact-based nature of weather forecasts in a country hit by around 20 tropical cyclones each year.

“Artificial intelligence offers a sharper edge by improving forecast reliability through data integration from satellites, radar, ground stations, and environmental sensors,” Servando said.

He acknowledged that traditional forecasting still faces challenges, including uncertainty, short lead times for extreme events like flash floods, and difficulty predicting localized weather systems.

Despite the advancements, Servando emphasized that AI is meant to assist, not replace, human expertise.

“AI does not replace meteorology or the expertise of our meteorologists,” he said. “It supports them by processing vast and complex data sets, generating timely and understandable guidance for decision-makers. Meaning, AI will not replace our jobs in the future.”

Servando said AI-driven systems can improve decision-making in areas such as evacuation planning, agriculture, and infrastructure protection, but their success depends on proper implementation.

“We need robust infrastructure to provide adequate computing and storage capacity, ongoing capacity-building for our data users, and effective strategies to introduce AI technologies to the public,” he said.

PAGASA showcased its AI4RP (Artificial Intelligence for a Resilient Philippines), also known as Project GABAY, a pioneering initiative that uses AI to enhance weather forecasting accuracy and make climate information more accessible to communities.

“This enhancement represents our commitment not only to improve accuracy but also to ensure that information reaches those who need it most — communities that must understand and act on it,” Servando said.

Launched in 2024, Project GABAY aims to develop a high-resolution AI-driven weather forecasting model tailored to the Philippines. The system can analyze historical weather data, recognize patterns, and generate faster, more precise forecasts.

Under the project, PAGASA is testing AI models that increase resolution from three kilometers to two, extend forecast lead times from two days to 14, and cut computation time from three hours to just 15 minutes.

“The AI technique produces outputs faster and requires less computation,” Servando said. “This means more timely forecasts for local decision-makers.”

Servando concluded that while technology enhances forecasting, saving lives still requires collaboration among government, communities and other stakeholders.

“Technology alone is not enough,” he said. “The success of our warning systems depends on partnerships, trust, and collective action.”