Fast scheme for determination of direct normal irradiance. Part I: New aerosol parameterization and performance assessment

•A new aerosol direct transmittance scheme is developed and can be used by simple DNI models to consider effect of aerosols.•A fast DNI scheme (SUNFLUX) is improved with the new aerosol scheme for DNI prediction.•The improved DNI scheme is validated against observations and compared with other five...

Full description

Saved in:
Bibliographic Details
Published in:Solar energy 2020-03, Vol.199, p.268-277
Main Authors: Shi, Guoping, Sun, Zhian, Li, Jiangnan, He, Yongjian
Format: Article
Language:English
Subjects:
Aeronet
Aerosol optical depth
Aerosols
Bsrn
Direct normal irradiance
Irradiance
Parameterization
Performance assessment
Radiative transfer
Solar energy
Stations
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:•A new aerosol direct transmittance scheme is developed and can be used by simple DNI models to consider effect of aerosols.•A fast DNI scheme (SUNFLUX) is improved with the new aerosol scheme for DNI prediction.•The improved DNI scheme is validated against observations and compared with other five schemes.•Sensitivity tests are conducted to examine the sensitivity of DNI model to variability of input data. A new aerosol direct transmittance parameterization is developed based on detailed radiative transfer model calculations for improving direct normal irradiance (DNI) for the solar energy forecast in the fast radiation scheme SUNFLUX. The modified SUNFLUX scheme is evaluated against observational datasets collected from the 14 Baseline Surface Radiation Network (BSRN) stations. The aerosol data from AERONET stations collocated with the BSRN network are used in the evaluations. The evaluation results have indicated that the accuracy of the calculated DNI using the new version of SUNFLUX has been improved significantly compared to the old version. The relative mean bias difference (MBD) and relative root mean square difference (RMSD) are −3.03% and 4.95%, respectively. The performance of the modified DNI model is also compared with those from the other five models. The results predicted by the current model are comparable with those from the best reference models. More importantly, the evaluation results show that the new aerosol scheme developed in this study can also be applied to other models to improve their performance.
ISSN:0038-092X
1471-1257
DOI:10.1016/j.solener.2020.02.028