The analysis of replacing high-resolution terrain data forforecasting solar radiation in Wuhan, China

Abstract

Abstract This experiment uses the data from European Center for Medium-Term Weather Forcast and the mesoscale numerical model WRF to predict the total solar radiation in Wuhan in January, April, July and October 2018. During the experiment, the Aster global digital elevation model data was used as high-precision topographic data to replace, and two sets of data which are forecasted by WRF were compared with the observation data of Wuhan station. The experiment results show that the four-month forecast has a good correspondence with the observation station data, and which has passed the 0.01 significance test. Meanwhile, the simulation results in January were poor, and the simulation results in July and October were the best. After the improvement of replacing high-precision terrain data, the MAE and RMSE for 4 months have been reduced by 10-20W/m2, and both MAE and RRMSE have also been reduced by 5%-10%.