DOST to use AI in weather forecasting for more frequent rain predictions  

MANILA (PIA) – The Department of Science and Technology (DOST) is set to revolutionize weather forecasting by incorporating artificial intelligence (AI).

This initiative, announced by DOST Secretary Renato Solidum Jr. on Wednesday, September 11, aims to provide rain forecasts every 15 minutes, significantly increasing the frequency of predictions.

During budget deliberations for the DOST and its attached agencies, including the Philippine Atmospheric, Geophysical and Astronomical Services Administration (PAGASA), Solidum explained that the AI integration will replace the current three-hour computation cycle, enabling more timely and accurate weather information.

“Soon we will operate through our research program to use AI in weather forecasting. So instead of every three hours, our target is every 15 minutes,” he said.

He added that the goal is to enhance PAGASA’s forecasting capabilities to provide rain predictions for up to 14 days, compared to the current five-day forecast.

Senator Juan Miguel “Migz” Zubiri, who presided over the Senate finance subcommittee hearing, raised concerns about the number of operational Doppler radars in the country. 

According to PAGASA Administrator Nathaniel Servando, only 11 out of 19 radars are currently operational, with some damaged and others needing replacement. 

Solidum indicated that the country requires at least 21 radars to ensure redundancy in case some units fail.

“When it comes to Doppler radars, the reason we need 19 or 21 is because of redundancy. If one or several are not operational, the existing ones can cover,” Solidum explained. 

He noted that aging equipment necessitates replacements and that acquiring new radars involves a lengthy procurement process.

Zubiri expressed support for funding additional Doppler radars, highlighting that other countries can predict the exact timing of rain, a capability currently lacking in Philippine forecasts. 

Solidum acknowledged the challenges of forecasting in an archipelagic nation, saying, “There is a big difference between continental meteorology and meteorology in an archipelago like the Philippines.”

He also mentioned that PAGASA is moving toward impact-based forecasting, which assesses the effects of rain rather than simply predicting its occurrence. 

“PAGASA produces several rainfall warnings and 24-hour rainfall forecasts, which disaster managers should read for their areas,” Solidum said. 

However, he noted that there is a need for standardized operating procedures at the local level to effectively respond to these forecasts. (JVD/PIA-NCR)