Table 2.
Best-fit stepwise multiple linear regression (MLR) models of monthly average daily solar radiation (MJ m-2 day-1) for Turkey.
| Number of explanatory variables in MLR model | 6 | 7 | 8 |
|---|---|---|---|
| Intercept | -7.156 | -9.853 | -5.206 |
| Ho(S/So) (MJ m-2 day-1) | 0.760 | 0.762 | 0.776 |
| Cloudiness (% month-1) | 0.0834 | 0.0819 | 0.0939 |
| Elevation (m) | 0.00092 | 0.00063 | 0.00066 |
| Month | -0.095 | -0.095 | -0.080 |
| Aspect (compass degree) | 0.00228 | 0.00229 | 0.00248 |
| RHmax (% month-1) | 0.051 | 0.062 | 0.058 |
| Longitude (decimal degree) | 0.054 | 0.048 | |
| Latitude (decimal degree) | -0.126 | ||
| RMSE | 1.59 | 1.58 | 1.58 |
| R2adj (%) | 93.3 | 93.4 | 93.5 |
| Cp | 94.6 | 77.1 | 67.5 |
| V/P ratio (%) | 29 | ||
| R2 for validation (%) | 93.3 | 92.9 | 92.9 |
All the variables are significant at P ≤ 0.001; Ho: monthly average daily extraterrestrial solar radiation on a horizontal surface; S: day length; So: maximum possible sunshine duration; RHmax: maximum relative humidity; RMSE: root mean square error; and V/P: ratio of number of stations of validation dataset to those of parameterization dataset.