<p>Mediterranean forests play a key role in the functioning of terrestrial ecosystems and in climate regulation, while also providing important ecosystem services to society. However, they remain particularly vulnerable to extreme weather events, the frequency of which may increase due to climate change and anthropogenic pressures. The objective of this study is to assess land use/land cover (LULC) dynamics in the upper central Rif Mountains, located in northern Morocco and subject to a humid bioclimate. It also aims to monitor vegetation responses to global change. Multi-date geospatial data (1990–2023) and support vector machine (SVM) algorithms were employed to analyze the observed changes. A time series of harmonized Landsat images preprocessed with Google Earth Engine was used to generate spectral indices (LST, NDVI, EVI). Non-parametric statistical tests (Mann-Kendall and Theil-Sen slope estimator) were used to monitor and spatialize monotonic trends in spectral indices. Our results reveal a reduction in forest cover in favor of agricultural land. LST exhibits an increase, with clear variations according to LULC categories. Agrosystems and sparse forests show the highest increase, while dense forests show a slight increase. The monotonic trend in vegetation indices (NDVI/EVI) revealed that dense forests are marked by significant increases above 65%, followed by sparse forests with 60%, while agrosystems show the lowest significant trend (55%). A negative correlation was observed between NDVI, EVI, and LST, especially for dense forest, with a coefficient of <i>r</i> = -0.32. These findings highlight the beneficial effect of dense vegetation in mitigating global warming and the vulnerability of open vegetation to these changes, emphasizing the need to implement management and conservation strategies to ensure forest ecosystem sustainability. Furthermore, the combination of remote sensing and machine learning algorithms proves to be an effective tool for monitoring vegetation responses to global change.</p> Graphical Abstract <p>This study investigates the land use and land cover (LULC) changes and vegetation responses to global change. A time series of Landsat images (1990–2023) was collected and preprocessed using the Google Earth Engine platform. Analysis were performed using a support vector machine (SVM) algorithm, spectral indices, and parametric and non-parametric tests. Diachronic analysis reveals a decline in forest cover in favor of agrosystems. The LST index shows an increasing trend, reflecting surface warming. Vegetation as a whole shows significant greening during the growing season. Dense vegetation plays an important role in mitigating temperature increases. Sparse vegetation is most vulnerable to global warming. These findings emphasize the need for strategies that enhance ecosystem resilience, improve biodiversity and support sustainable development goals.</p>