New empirical relationships for determining global PAR from measurements of global solar radiation, infrared radiation or sunshine duration

Abstract

Data on global solar radiation (H), global photosynthetically-active radiation (PAR), downward infrared sky radiation (IR) and sunshine duration (n) for a 2-year duration are analysed on monthly (monthly mean of daily values) basis at Ilorin (8°32′N, 4°34′E), Nigeria. The data were taken using a precision spectral pyranometer, quantum sensor, pyrgeometer and Campbell–Stokes recorder, respectively. Two thermistor circuits were built into the pyrgeometer for monitoring the dome and body temperatures so as to take care of the dome heating effect. From monthly values of IR, PAR and H, a correlation model relating the three parameters was established and is of the form: (1) where fp is the ratio of PAR to H. Further analysis showed that the use of maximum likelihood quadratic fit to relate the three parameters does not improve the coefficient of determination. Also from the monthly mean values of PAR and n, equations of the Angstrom–Prescott one-parameter model of linear and quadratic forms are developed for the estimation of global PAR at this location and locations with similar climatic conditions. The advantage of these models over the traditional method of estimating PAR, by assuming it to be of constant ratio to insolation, is that PAR can be estimated even at locations where insolation values are not available since they make use of only solar-related astronomical parameters (extraterrestrial radiation and maximum sunshine duration) and duration of sunshine. The results show that in practice, the simple linear regression model is sufficient and hence the use of a quadratic model is only slightly justifiable. On developing the models based on the two seasons (rainy and dry) encountered here, the rainy season model showed a high correlation coefficient, R, of 0.9793 against the poor R of 0.5075 for the dry season. The poor correlation during the dry season period is probably due to large differences in the characteristics of the sky during this period. Copyright © 2000 Royal Meteorological Society