A Philippine government research project is testing whether artificial intelligence can help produce more detailed weather forecasts farther into the future and process forecasting data more quickly.
The AI-Powered Weather Forecasting for a Resilient Philippines, or AI-4RP, is being developed by the Department of Science and Technology-Advanced Science and Technology Institute (DOST-ASTI) with the Philippine Atmospheric, Geophysical and Astronomical Services Administration (PAGASA).
The project uses deep-learning neural networks to develop a medium-range forecasting model specifically for the Philippines.
DOST-ASTI lists three major technical targets: forecasts extending up to 14 days, a horizontal resolution of two kilometers by two kilometers, and forecasts at 15-minute intervals.
DOST Secretary Renato Solidum Jr. has also identified computing time as one area where artificial intelligence is being tested.
PAGASA processes large volumes of meteorological information in producing forecasts. Solidum said processing currently takes about three hours, while researchers are experimenting with AI in an effort to reduce that period.
“The computing time to process the data would be every three hours. We are experimenting with the use of AI to shorten it to 15 minutes so that the forecasts are more up-to-date,” Solidum said.
The 15-minute target remains experimental and is not being presented by DOST as an existing operational processing time.
AI-4RP is one part of a broader set of science and technology projects being applied to disaster preparedness.
Another addresses a different problem: bringing hazard information from multiple sources into a system that users can query.
The HANDA Integrated Platform for Preparedness and Response consolidates disaster and hazard information to make data more accessible for preparedness and planning.
Philippine Institute of Volcanology and Seismology (PHIVOLCS) Director Teresito Bacolcol has compared the way people can interact with the system to using a conversational AI service.
“It’s like the ChatGPT of disaster preparedness,” Bacolcol said.
A user, for example, can ask about active faults affecting a particular location and retrieve relevant hazard information.
HANDA is connected to another DOST initiative known as GATES, or Geospatial Analytics for Transformative Earth Systems.
The program is designed to bring together geospatial, hazard, climate and other datasets from government science agencies, including PAGASA and PHIVOLCS, in an AI-ready national data platform.
DOST is working with the Japan International Cooperation Agency on the data infrastructure supporting the initiative.
A separate system focuses on observations rather than predictive modeling.
PAGASA’s National Hydro-Met Observing Network, or PANaHON, provides access to near-real-time information gathered from weather monitoring stations across the country.
The department is also using technology for the operational side of disaster response.
Its Mobile Command and Control Vehicle, or MoCCoV, is designed to provide a deployable command facility in locations where a permanent operations center may not be immediately available.
A MoCCoV unit turned over to the Tuguegarao City government can be set up in approximately 20 minutes.
The vehicle contains facilities for command operations and conferences and can also support rescue or medical triage activities. It is equipped with solar panels and a backup generator that can operate for up to 32 hours.
The projects perform different functions within disaster management: AI-4RP focuses on weather prediction, PANaHON provides observational weather information, HANDA makes hazard information accessible, GATES consolidates government datasets, and MoCCoV provides mobile infrastructure during emergency operations.
Solidum said DOST’s objective is for scientific information to be incorporated into decisions affecting communities.
“Science must enter the decisions that shape people’s lives,” he said.
“Science and the information it provides must guide policies and influence investment.”
The projects are being developed and deployed as the government expands its use of artificial intelligence, geospatial information, weather monitoring and communications technology in disaster preparedness and response.




