Research on the applicability of global solar radiation estimation model in Henan Province, China

Global solar radiation is very important for designing solar energy systems. However, it is observed at a very limited number of meteorological stations in Henan Province of China, while sunshine duration is abundantly measured in many stations. Two models based on Angström-Prescott equation were tested to estimate daily and monthly global solar radiation from sunshine hours during 1961-2009 in Zhengzhou. The daily temperature range was used to predict model coefficients of a and the altitude, latitude of meteorological station and number of month of year starting from first of January was used for b for the first model; the second model set a and b as empirical constants in Henan Province, respectively, as 0.475 and 0.205. The statistical parameters such as variability explained (R²), the basis of mean absolute error (MAE), mean bias error (MBE), relative mean deviation (RMD) and the root mean square error (RMSE) were used to assess the two models when the daily and monthly global solar radiations were predicted. When the first model was used to predicted the daily global solar radiation, the values of slope for zero-intercept, MAE, MBE, RMD and RMSE were 0.99, 1.65 MJ m^-2, 0.06 MJ m^-2, 0.48% and 16.85%. As for predicting the monthly global solar radiation, the values of slope for zero-intercept, MAE, MBE, RMD and RMSE were 1.00, 25.33 MJ m^-2, 1.93 MJ m^-2, 0.48% and 8.74%. The value of statistical parameters indicated that the highest precision could be expected when the regression coefficients a and b were simulated as dynamic value which also had the capability of simulating the changes of monthly mean global solar radiation.