DESCRIBING LONG-TERM TRENDS IN TEMPERATURE OF JOS REGION OF NIGERIA, USING GENERALIZED ADDITIVE MODELS (GAMS).

Descriptions of how temperature patterns change through time can be relevant in light of the present worry about climate change. These trends can be deduced from monthly temperature data collected over the past few decades. To model temperature patterns, generalized linear models are frequently utilized. These models can represent temperature variations over the course of a year, but they have limitations when it comes to predicting long-term trends, especially when they are non-linear. By fitting smooth functions to the data, generalized additive models (GAMs) provide a framework for modeling non-linear connections. Using data from the Department of Meteorology and Climatology University of Jos from January 1986 to December 2023, this research illustrates how GAMs can enhance the flexibility of models to reflect seasonal patterns and longterm trends in temperature. Smoothed model estimations provide useful graphical depictions of a rise in temperature patterns at this area during the previous 38 years. GAMs are very useful for looking for nonlinear correlations in data. Smooth functions must be chosen with care to ensure that they are appropriate for the data and modeling goals. [ABSTRACT FROM AUTHOR]