Performance Analysis of Temperature Models for Environmental Monitoring in Southwest Nigeria
Temperature is a major meteorological parameter driving most of the atmospheric processes vis-à-vis climate change. Therefore, a consistent model is necessary to achieve sustainable development goal 13 (SDG 13) known as climate action. Long-term monthly averages of surface temperature obtained from six southwest states in Nigeria were subjected to five mathematical models, namely the sum of two-Gaussians, the sum of two-Lorentzians, Fourier on four harmonics, Sine wave and Fourth-order polynomial functions. Statistical tools were used to examine the accuracy and fitness of the models. The evaluation showed that the Gaussian and Lorentzian models are good fits of the observed data. Furthermore, the performance indicators such as mean bias error (MBE), root mean square error (RMSE), mean percentage error (MPE) recorded the lowest values for Fourier on the fourth harmonic model. Similarly, its correlation coefficient, R, was the highest ranging from 0.95 to 1. Consequently, the Fourier model presented the best correlation with the observed data and hence recommended for predicting the temperature at the selected locations.