Statistical Analysis of Some Meteorological Variables Data for Sokoto and its Vicinity

Abstract

This study aims at studying the characteristics of some meteorological variables in Sokoto and its vicinity using probability distribution models. The thirty one (31) years data (i.e 1980-2010) were collected at Nigerian Meteorological Agency (Nimet) Oshodi and the data were subjected to various probability distribution analyses in order to resolute the best fit probability functions for each meteorological variables. The variables measured consist of Relative humidity, Rainfall, Temperature, Sunshine hours, Solar radiation, wind speed and Evaporation pitche. Whereas the probability distribution models adopt were Normal, Gumbel, Pearson type III and Log-Pearson type III distribution functions. Numerical equation were recognized and used to forecast the variables. Goodness of fit tests such as chi-square, correlation coefficient and coefficient of determination were carried out to obtain the reliability of the forecasted values. The model that satisfies the statistical tests conditions mostly was selected as the best fit model. The study revealed that Rainfall, wind speed, evaporation pitche are best fitted by Log-pearson type III probability distribution model, whereas the Relative humidity, solar radiation, sunshine hours and temperature the best model is Gumbel probability distribution.