Estimation of urban solar radiation for the City of Recife-PE

Abstract

Urban Global Solar Radiation(RS) is one of the main inputs of urban climatology, and should already be considered from the conception of the construction project of green infrastructures, which mitigate the environmental impacts of urbanization. The objectives of this work were to calibrate and evaluate the performance in estimating the RS, for the capital of Pernambuco, of six existing models in the literature, namely: Hargreaves; Samani (1982), Bristow; Campbell (1984), Jong & Stewart (1993), Chen et al. (2004), Hunt et al. (1998) e Annandale  et al. (2002), referring to a rainy period in the year 2023, with data from an automatic meteorological station installed in one of the campuses of the University of Pernambuco, in the city of Recife-PE. The statistical criteria used for the analysis of the calculated data: the root mean square error (RMSE), the Nash-Sulcliffe coefficient (NSE) and the residual mass coefficient (CMR). The models, in ascending order of greater effectiveness are: Hunt et al. (1998), Bristow; Campbell (1984), Jong & Stewart (1993), Chen et al. (2004), Hargreaves; Samani (1982), and Annandale  et al. (2002). However, it should be noted that for the use of the model proposed by Hunt et al. (1998), it is necessary to have daily precipitation data, which may not be available, especially in historical series. If only temperature values are available, the best model for estimating RS is the one proposed by Bristow; Campbell (1984), which presented a satisfactory performance and very close to that obtained by the model proposed by Hunt et al. (1998).