干旱区生态环境知识资源中心

The mapping and monitoring of Direct Normal Irradiance (DNI) is most relevant for the proper installation of solar power plants that use concentrating solar systems. This paper presents a satellite-based method for estimating DNI and daily direct normal irradiation (Q(n)) that uses a minimal set of regional meteorological information and avoids empirical adjustment with ground-based radiometric data. The focus of the validation and application were on the Northeast Brazil region (NEB), which exhibits high solar radiation levels throughout the year. Two basic parameters for satellite-derived DNI estimates are: (i) DNI under clear-sky conditions DNIclear (computed from the REST model) and (ii) cloud cover C (estimated from GOES visible imagery). The first validations were performed by comparison with measurements from three radiometric stations of the SONDA network in the NEB (Petrofina, Natal and Sao Luis) during 2007-2008. Overall, the results confirmed the good performance of the proposed model. DNI daily cycles in conditions of clear and partially cloudy skies were represented satisfactorily, with daily average errors within +/- 25 W m(-2). Good linearity was found between the measured and satellite estimated Q(n), with coefficients of determination (R-2) ranging from 0.84 to 0.89. The average values of MBE (Mean Bias Error) and RMSE (Root Mean Square Error) over all sites were -1.7% and 18.7%, respectively. In addition, seasonal and annual average maps of direct normal irradiation were generated over NEB for 2008. It was found that the spatial-temporal variability of Q(n) is strongly modulated by meteorological systems acting in the region, such as the ITCZ (Intertropical Convergence Zone). The mapped results reveal that the highest value was approximately 22 MJ m(-2) on an annual average basis, and higher values occurred in areas of semi-arid climate. The proposed satellite-based model may be applied for providing information of DNI (with a spatial resolution of 4 km), particularly over regions without in situ measurements, and suggests good accuracy for climatic and solar resource studies. (C) 2017 Elsevier Ltd. All rights reserved.