Analog prediction of tropical cyclone-induced precitation in Luzon, Philippines

抄録

Current numerical weather prediction (NWP) models produce relatively large errors in forecasting TC-induced precipitation. Because of that, lots of alternative methods are being developed nowadays. An analog forecasting method has been proposed by Bagtasa (2021) for the Philippines, a country that receives around twenty typhoons yearly. While the methodology produces good results, it still has its limitations due to the techniques and datasets used. This study aims to improve the aforementioned methodology by limiting the area of study to Luzon Island and introducing dynamical adjustments and uncertainties in computing precipitation. Currently, the study is in its early stages. Initial results show promising values that are on par with a previous study. Further implementation of proposed improvements is needed to properly check whether the modified methodology addresses the issues of the original methodology. Ultimately, we believe that this study will further advance the field of forecasting TC-induced precipitation in the region.