land surface temperature; thermal remote sensing; Amazonian urbanization; green infrastructure.

Mid-sized cities in the Amazon region remain underrepresented in the literature on Surface Urban Heat Islands (SUHI), a context where the combination of accelerated urbanization and an equatorial climate tends to intensify surface warming. This study analyzed the spatiotemporal dynamics of Land Surface Temperature (LST) and its relationship with land use and land cover (LULC) patterns within the urban perimeter of Porto Velho, Rondônia, between 2015 and 2024. Sixty-eight Landsat 8 images (Collection 2, Level 2) were utilized, acquired via Google Earth Engine and processed using R and QGIS environments. LST and the spectral indices NDVI, NDBI, and MNDWI were calculated, and Local Climate Zones (LCZ) were mapped for 2015 and 2024 using the Random Forest classifier applied to Landsat 8 multispectral composites. Statistical analyses included Pearson correlation, the Kruskal-Wallis test with Dunn's post-hoc analysis, and spatial autocorrelation using Global Moran's I and LISA. Ten LCZ classes were identified, achieving an overall accuracy above 0.86 and a Kappa coefficient above 0.82 for both years. Built-up classes expanded their share from 60.9% to 64.9% of the urban area over the studied period. Annual mean LSTs ranged from 39.09 °C to 42.49 °C, with maximum temperatures exceeding 50 °C in all analyzed years. LST exhibited a strong negative correlation with NDVI (r = −0.807), a strong positive correlation with NDBI (r = +0.833), and a moderate negative correlation with MNDWI (r = −0.399), a pattern that remained stable throughout the time series. Compact built-up classes and bare soil surfaces concentrated the highest temperatures, whereas vegetated and water classes recorded the lowest values. These findings confirm the mitigating role of vegetation cover on surface warming and indicate that urban forestry and green infrastructure strategies constitute relevant pathways for the adaptive environmental planning of expanding Amazonian cities.