Intercomparison of the parameterized Linke turbidity factor in deriving global horizontal irradiance

This study presents a comparative assessment of nine parameterized methods to estimate Linke turbidity (TL) in the five Köppen–Geiger climate zones. The evaluation is intensified by simulating all TL values into a global horizontal irradiance (GHI) model, called the Hammer model. The results found that atmospheric turbidity on the Korean Peninsula varies per season. The atmosphere tends to be more transparent in fall and winter, with lower turbidity. Meanwhile, the turbidity consistently rises in spring and peaks and summer, indicating a significant decline in atmospheric transparency. All nine TL methods generated varied degrees of accuracy according to their Köppen–Geiger climate zones. In terms of relative root mean square error (rRMSE) and relative mean bias error (rMBE), TL_Valko1, TL_Valko2, and TL_Capdrou consistently perform poorly. Further, these three methods can only classify the atmospheric turbidity into two types: least and fully turbid. This study recommends the following methods for evaluating atmospheric turbidity in various climate types: TL_Remund, TL_Dogniaux1, TL_Dogniaux2, TL_Ineichen, TL_Molineaux, and TL_Greiner. This study concludes that none of the six methods can be considered “the most superior” due to minor differences in rRMSE and rMBE values.