Answers are grounded in this paper and cite their sources.
Land-use and land-cover (LULC) change is an important concern in rapidly urbanizing semi-arid environments. This study used Sentinel-2 imagery to map annual LULC dynamics in Patos, Brazil, from 2019 to 2024. A supervised Random Forest classifier distinguished three terrestrial classes (Bare soil, Vegetation, and Built-up) while water bodies were delineated separately using the Modified Normalized Difference Water Index (MNDWI). Classification accuracy was consistently high (overall accuracy: 95.6–98.1%; Cohen’s Kappa: 0.930–0.971). The mapped Vegetation class increased from 12.28 km² in 2019 to 18.10 km² in 2020 and subsequently fluctuated between 15.89 and 18.64 km² through 2024, without a clear directional trend. Temporal variation in the Normalized Difference Vegetation Index (NDVI) was evaluated using a linear mixed-effects model. Fixed effects of LULC, year, and their interaction explained 40.7% of NDVI variance (marginal R² = 0.407), while the full model, including the random effect of sampling point, explained 79.7% (conditional R² = 0.797). LULC, year, and the LULC and year interaction were significant ( p < 0.001), supporting the hypothesis that differences in NDVI among land-cover classes varied across years. January-May precipitation was not significantly correlated with NDVI in any LULC class (all p > 0.48), indicating that rainfall alone did not explain interannual NDVI variability over the six-year period. The integrated approach provides a reproducible framework for monitoring urban vegetation dynamics in semi-arid environments.
No takes yet. Share an insight, caveat, or question.
Costa et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: