Modelling Extremal Temperatures using Extreme Value Theory: A case for Namibia

Temperature extremes (commonly referred to as cold spell and heat waves) are regarded as the most significant climate events and have been extensively studied over the last several decades. The heat waves and drought during the summer of 2017 in Namibia sent shock waves to the energy industry, farming and water resources infrastructures and health system in terms of system vulnerability and management. This paper focuses on modeling of extreme temperatures using Extreme Value Theory (EVT). The aim is to explore the frequency of occurrences of extremely low and extremely high temperatures. The dataset consists of 27900 daily maximum temperatures from 1990 to 2019 collected from the Namibia Meteorological services. Block maxima (BM) approach is used to fit the Generalized Extreme Value Distribution (GEVD) while the Peak Over Threshold approach is employed to fit the Generalized Pareto Distribution (GPD) model which analyzes the upper and lower tails of the distribution of the data. The Maximum Likelihood Estimation (MLE) technique is used to estimate the parameters in two distributions. The model’s goodness of fit is assessed graphically means, such as probability plots, quantile-quantile plots and mean excess plots as well using some empirical goodness-of-fit tests. Results indicate that, the models under consideration provide overall good fits for the data Subject Classifications: 68Txx, 68Pxx, 68Nxx.