Optimization of numerical weather model parameterizations for solar irradiance prediction in the tropics

Many photovoltaic (PV) systems are connected to electrical power grids, and the grids are at risk of instability due to fluctuation of PV output. Numerical weather prediction (NWP) models are used to forecast solar irradiance and proper grid management. NWP usually has many physical parameterization options, and appropriate schemes of these options should be selected for accurate forecasting. The options should be determined by regional and climatic conditions and other factors. The target country is Thailand, which is in the tropics. In Thailand, cumulus and cumulonimbus clouds frequently appear, and their behavior makes weather forecasting difficult. The optimal combination of schemes in the tropics is determined through a sensitivity analysis of the options. By the optimization the forecasting accuracy increases from 0.773 to 0.814 of the correlation coefficient. It is also found that surface layer and PBL processes make a significant contribution to the improvement of accuracy.