PEMODELAN STATISTIK DAN SIMPLE RADIATIVE MODEL UNTUK MENDUGA RADIASI MATAHARI GLOBAL HARIAN

Abstract

This paper evaluated accuracy and applicability of seven of global solar radiation estimation that use daily ambient temperature, total precipitation and relative humidity at the tropical area of Indoneisan maritime continent. Two models of these evaluated are new proposed model (PM) to estimate solar radiation at the horizantal surface. The first model PM1used tempreature-precipitation as the input (TP-based) while the second PM2 model used temperature-precipitation-relative humidity combination as the input (TPRHbased). All the models are evaluated by error statistics e.g. Fraction of Prediction in one or two factor (FAC2), Mean Bias (MB), Mean Gross Error (MGE), Normalised Mean Bias (NMB), Normalised Mean Gross Error (NMGE), Root Mean Square Error (RMSE), Coefficient Determination (r) and Coefficient of Efficiency (COE). The result shows that TPRHbased model have a good accuracy thanTbased or TP-based. PM2 model showed the best performance among all models while Quej et.al 2016 model has good accuracy but less precision. Tropical climate characteristic where high humidity have significantly affected incoming solar radiation at the surface temporarily or spatially.