The Philippines ranked first as the most vulnerable country to climate-related risks, according to the World Risk Index 2024 report. With extreme weather events becoming more frequent and intense, there is an urgent need to enhance disaster preparedness efforts in the country.  

As climate patterns grow increasingly unpredictable, traditional forecasting methods alone may no longer be sufficient to mitigate risks effectively. Recognizing this challenge, the Philippine government is leveraging artificial intelligence (AI) technology to strengthen the country’s resilience against climate threats.  

AI approach to weather forecasting 

In 2023, the Department of Science and Technology (DOST) signed an agreement with U.S.-based AI meteorology company Atmo Inc., leading to the establishment of the DOST initiative called the AI-Powered Weather Forecasting for a Resilient Philippines (AI-4RP). 

The AI-4RP team, alongside the Atmo Inc. team, poses for a photo with DOST Secretary Dr. Renato U. Solidum Jr. And DOST Undersecretary Maridon O. Sahagun. Photo courtesy of Atmo Inc. 

Spearheaded by the DOST-Philippine Atmospheric, Geophysical and Astronomical Services Administration (DOST-PAGASA) in collaboration with the DOST-Advanced Science and Technology Institute (DOST-ASTI), the AI-4RP Project aims to develop and deploy an AI-powered system that will augment the country’s current weather forecasting capabilities. 

In 2024, Atmo successfully deployed the country’s first online AI-based weather forecasting system. This system delivers high-resolution, ultra-precise weather forecasts, enabling weather patterns to be tracked and predicted at a more localized, neighborhood level. Calibrated specifically for the Philippines’ unique weather patterns, the system is now in its second iteration and is currently in the process of training multiple future versions. 

Enhancing predictions through data sharing 

For the AI-based forecasting system to achieve greater accuracy, it requires large volumes of high-quality data. To address this, an agreement on sharing of meteorological data was signed between DOST-PAGASA and DOST-ASTI in collaboration with Atmo through a ceremonial signing event on 18 February 2025 at the Novotel Manila Araneta City, Cubao, Quezon City. 

Left to right: Ms. Maria Cristina C. Uson (DOST-PAGASA, AI-4RP Project Leader), Dr. Marcelino Q. Villafuerte II (DOST-PAGASA, Deputy Administrator for R&D), Dr. Franz A. De Leon (DOST-ASTI, Director), Ms. Maridon O. Sahagun (DOST, Undersecretary for Scientific and Technical Services) 

As part of this partnership, a data-sharing agreement was signed, providing Atmo Inc. with access to historical meteorological data through the country’s national research and education network. This agreement strengthens the partnership among DOST-ASTI, DOST-PAGASA, and Atmo Inc., ensuring that the latter’s AI models can leverage the richest possible dataset to deliver even more precise, accurate, and timely forecasts tailored to the Philippines’ unique climate challenges. 

Moreover, DOST-ASTI is contributing additional datasets through its Meteorological Data Acquisition Stations for Information Dissemination (MASID) and Forecasting the Weather Using Lightning Observations as Safeguard for Hazards (FLASH) Projects. Data collected from the PhilSensors weather stations and POTEKA devices will also be integrated into Atmo’s system, further improving its forecasting capabilities.  

Strengthening local expertise in AI Meteorology 

Following the signing of agreements, Atmo Inc. team designed a workshop to equip DOST personnel with the knowledge and skills to maximize the potential of the AI-powered forecasting system. 

The workshop began with a data assimilation session focused on integrating local meteorological data into Atmo’s AI system. The training covered key topics such as data curation for Philippine weather data, weather model development within the Atmo AI system, and AI model performance analysis.

Atmo Project Scientist Giorgo Sarro leads a training session on analyzing AI model performance. 

As part of the initiative, the project team from DOST-PAGASA evaluates the AI models using specialized verification procedures and compares their performance against the numerical weather prediction models currently in use. 

DOST-ASTI Director Dr. Franz A. De Leon highlighted the transformative impact of this collaboration on the country’s approach to weather forecasting and disaster mitigation. 

“This collaboration is a call to move beyond the possibilities in the fields of weather science and technology. Together, with the help of artificial intelligence, we will ensure that the Philippines is not only prepared for the challenges of today but also empowered to face those of the future,” he emphasized. 

Atmo, Inc. President and CEO Alexander Levy emphasized the importance of secure data-sharing for the project.  

“The data-sharing agreement we signed today established the necessary framework for secure and responsible data use, reinforcing our shared commitment to scientific innovation and national resilience. Together, we are setting a new standard in meteorology—harnessing cutting-edge AI meteorology technology to protect lives, safeguard infrastructure, and build a more weather-resilient future,” he noted. 

By integrating cutting-edge AI technology into weather forecasting and fostering knowledge-sharing among international partners and local experts, the Philippines is making significant strides toward strengthening its resilience against extreme weather events. As climate change continues to pose challenges, these efforts represent a crucial step in ensuring that the country remains proactive, prepared, and empowered in the face of increasingly unpredictable weather. 

 

About the AI-4RP Project 

The AI-4RP Project aims to enhance weather forecasting in the Philippines using advanced AI technologies, complementing the current, traditional weather forecasting methods used in the country. This DOST-GIA-funded project is a collaborative project between DOST-PAGASA and DOST-ASTI, in partnership with Atmo Inc., a U.S.-based AI meteorology company. The project’s goal is to deliver faster forecasts with higher accuracy, for longer periods of time.