Developing Empirical Models for Estimating Photosynthetically Active Radiation over Akure, South Western, Nigeria

Abstract

In this study, the measured monthly average daily global solar radiation and extraterrestrial solar radiation using the generalized 45% and 40% dataset was utilized to estimate the photosynthetically active radiation (PAR) and extraterrestrial photosynthetically active radiation (PAR₀) for Akure (Latitude 7.17⁰N, Longitude 5.18⁰E, and 375.0 m above sea level) Ondo State located in South Western, Nigeria. The monthly average daily sunshine hours, maximum and minimum temperature data were used to develop nine (9) new PAR sunshine based models and three (3) PAR temperature-based models. The meteorological parameters used in this study covered a period of thirty-one years (1980 – 2010). The newly developed models were tested using statistical indicators of the coefficient of determination (R²), Mean Bias Error (MBE), Root Mean Square Error (RMSE), Mean Percentage Error (MPE), t-test, Nash – Sutcliffe Equation (NSE) and Index of Agreement (IA). The PAR sunshine based models that took a quadratic form and the linear logarithmic PAR temperature-based models were found more suitable for estimating PAR for the location under study. Comparing the PAR sunshine based and temperature-based models indicated that the PAR sunshine based model is more suitable for PAR estimation in Akure. Furthermore, the results showed that PAR is high during the dry season and low during the rainy season. Based on the measured and estimated PAR models; the minimum values were found in July and August while the maximum values in February, March and November. The descriptive statistical analysis shows that the PAR and all the estimated sunshine based PAR data spread out more to the left of their mean value (negatively skewed). Similarly, they have negative kurtosis which indicates a relatively flat distribution and the possibility of platykurtic distribution. The PAR and the PAR logarithmic temperature-based model (equation 17a) data spread out more to the left of their mean value (negatively skewed), while the PAR linear exponent and linear temperature-based models (equation 17b and 17c) data spread out more to the right of their mean value (positively skewed). The PAR and all the estimated PAR temperature-based data have negative kurtosis which indicates a relatively flat distribution and the possibility of platykurtic distribution.