Spatial-temporal analysis of vegetation cover and soil degradation from Landsat time-series – A case study in the Caatinga, Brazil

Abstract

Monitoring environmental changes in semi-arid regions is essential to identify its impacts on their ecosystems. By using remote sensing techniques, it is possible to detect these changes efficiently. The Caatinga biome in Brazil is a hotspot semi-arid region, challenged by increasing human activities. This study aimed to analyze the spatiotemporal dynamics of vegetation cover and soil degradation in the Caatinga Biome for the last decades. Land use and occupation maps (LULC), Normalized Difference Vegetation Index (NDVI), Leaf Area Index (LAI), Normalized Difference Water Index (NDWI) and Normalized Difference Moisture Index (NDMI), were determined based on the annual average of the years 1985, 1998, 2010 and 2022. NDVI, SAVI, LAI and NDMI indices showed a decreasing trend over the years, decreasing the reduction of vegetation, while NDWI showed variations in the presence of water. In 2022, natural Caatinga ecosystems (mainly shrubby vegetation) is still the dominant class, covering 50.7 % of the area. However, it was decreased by 5.4 % from 1985, coupled with an expansion in pasture (9.6 %) and agriculture (2.2 %) areas. Land cover and vegetation indices revealed a decline in native Caatinga vegetation due to agricultural expansion, especially pastures. These changes affect microclimates and rainfall patterns, indicating ecosystem degradation. Vegetation indices are effective tools for identifying critical areas and supporting sustainable land management and public policy planning in the biome.