Estimation of monthly values of atmospheric turbidity using measured values of global irradiation and estimated values from CSR and Yang Hybrid models. Study case: Argentina

For ten sites in Argentina, monthly average Ångström turbidity coefficient β values were estimated using mean monthly global solar irradiation values measured at these sites from the former Argentinean radiometric network REDSOL. The values of the horizontal direct and diffuse components of the global irradiation were estimated using the Yang Hybrid model (YHM) and, to determine the representativeness of those results, they were compared to values from the SWERA database which uses the CSR model, developed by NREL, to make estimates. An inconsistency in the estimated values of the diffuse component was detected, therefore a new expression for that component was proposed, generating a corrected version of the YHM (CYHM). The turbidity coefficient β is considered an independent variable in the equation that estimates the direct transmittance and the values that forces an exact correlation between the measured monthly average global irradiance values with those estimated with the CYHM are considered representative for each site since the RMSE between the monthly global irradiation values of SWERA and CYHM is 11.7%, this value being very close to that found by the designers of the CSR model (10%) when they tested it with meteorological data from USA. The monthly average value estimates of turbidity coefficient β show expected behavior, with values varying according to the season of the year.

► A new method is proposed to estimate monthly average turbidity values from global irradiance data.
► The proposed method includes the Correct Yang Hybrid Model (CYHM).
► A difference of 11.7% was obtained between CYHM estimates and SWERA database values.
► It is assumed that the error in the turbidity coefficient values is also low.