Algorithm improves prediction of sunny days for solar energy applications

Directly benefiting the Philippines’ solar power, agriculture, and other industries, an international team of researchers led by the Ateneo de Manila University and the Manila Observatory has pioneered a way to improve sunny weather forecasts by as much as 94%.

The study, “Application of Kalman filter for post-processing WRF-Solar forecasts over Metro Manila, Philippines,” is published in the journal Solar Energy.

Weather forecasters and scientists around the world rely on computer-generated simulation tools to predict the weather days in advance, with the Weather Research and Forecasting (WRF) Model being one of the most well-known and widely used.

In particular, forecasts of how much sunlight an area receives on a given set of days have all manner of uses—from helping ordinary people decide how to dress up and go about their day, to enabling entire industries to adjust their operations in response to the effects of solar radiation.

The Ateneo-led researchers improved WRF-Solar forecasts by applying a mathematical algorithm called a Kalman Filter (KF). Using data from various Metro Manila weather stations, they found that under some conditions they could minimize the discrepancy between forecasts and actual observations to as little as 6%.

In more technical terms, using KF on WRF-Solar forecasts of global horizontal irradiance for Metro Manila reduced mean bias error (MBE) by up to 94% and root mean square error (RMSE) by 12%, on as short as three days worth of training data. The optimal number of training days varied by season, with 42 days for the dry season (January to March) and 14 for the wet season (June to August).

The KF algorithm also excelled at correcting cloudy-period forecasts, albeit with slight inaccuracies for clear skies due to overcompensation for cloudy periods.

These results suggest that KF is a promising alternative to more computationally expensive forecasting methods for solar energy applications. This pioneering research highlights the potential of combining WRF-Solar and KF to enhance solar energy forecasting, vital for renewable energy planning in the Philippines.

The findings also emphasize the need for further model optimization across diverse Philippine landscapes to ensure reliable solar energy predictions tailored to the country’s unique climatic conditions.

“Results from the study, the first of its kind to assess performance of WRF-Solar and KF over the Philippines, will serve as a basis for a computationally efficient alternative to more intensive higher resolution and multiple ensemble member solar forecasts. Future work intends to focus on applying this method over different topographies in the Philippines, given the availability of irradiance data,” the researchers said.