Over the years, humanity has deepened its understanding of its actions’ profound impacts on global temperatures. Beyond the rise in global temperatures driven by global warming, local biomes also significantly influence regional temperature variations, often due to changes in land use. Historically, climatologists have attributed local temperature increases to deforestation, urban development, and agricultural activities. Although these factors are well-known, there remains a need for methodologies to objectively quantify the impact of land use changes on temperature increases. This paper proposes an explanation for temperature increase utilizing advanced statistical models such as ARMAX and neural networks with explainable AI components aiming to quantify the effects of these land use changes on local temperature trends over several years. For this purpose, two cities were chosen to apply the methodology. São Félix do Xingu, in the Brazilian Amazon rainforest (a city that has experienced significant deforestation due to agriculture over the last decade), and Cajazeiras, in the Brazilian semi-arid region (a city that has had substantial urban development over the years). The Data for this analysis were sourced from the MapBiomas Brazilian database and temperature records collected by the Brazilian National Institute of Meteorology. Our approach offers insights into human activities’ climatic impact, giving weight to soil changes and providing a methodology for further interdisciplinary studies, potentially serving as a starting point for specialist systems across various domains.