PREDICTING A MONOTONIC TREND OF THE SURFACE AIR TEMPERATURE USING GENERALIZED ADDITIVE MODEL (GAM) OVER NORTH-EASTERN REGION OF NIGERIA

Abstract

Considering the current concern over climate change, descriptions of how temperature patterns vary over time may be pertinent. A few decades' worth of monthly temperature data can be used to infer these tendencies. Generically adjusted linear models are widely used to model temperature patterns. The models have limits in predicting long-term trends, particularly when they are non-linear, but they are capable of representing temperature fluctuations over a year. Generalized Additive Models (GAMs) offer a framework for describing non-linear connections by fitting smooth functions to the data. This study demonstrates how GAMs can improve models' flexibility to capture seasonal variations and long-term temperature trends using data from the Nigerian Metrological Agency (NiMet) during a 41-year period (1981–2022). The area's temperature patterns have been raised over the past 41 years, as shown graphically using smoothed model estimations. GAMs are a great tool for spotting non-linear connections in data. Smooth functions need to be carefully selected to make sure they fit the data and modeling objectives.