Detecção de mudanças ambientais no semiárido brasileiro usando séries temporais e Google Earth Engine

Abstract

Land use and cover change is the leading cause of worldwide land degradation. However, efficient means of detecting and monitoring these environmental changes improve the understanding of this process, especially in developing more effective policies to combat land degradation. The objective of this study was to evaluate the spatio-temporal trends of environmental changes that occurred in the Desertification Nucleus of Cabrobó (DNC), PE, (comprised of the municipalities of Belém do São Francisco, Cabrobó, Carnaubeira da Penha, Floresta, and Itacuruba) between 2001 and 2020. We obtained rainfall data from weather stations belonging to the National Meteorology Institute and the National Water and Sanitation Agency, used to validate the CHIRPS product data on an annual scale via statistical indicators (Pearson correlation coefficient, mean error, Nash-Sutcliffe coefficient of efficiency, percent BIAS and RMSE). Subsequently, we manipulated CHIRPS data, MODIS sensor products related to NDVI (MOD13), surface temperature (MOD11), surface albedo (MCD43), and potential evapotranspiration - PET (MOD16) and MapBiomas data in the Google Earth Engine digital platform (GEE) and QGIS software (version 3.10.9). Using a programming routine, we performed surface albedo and Vegetation Health Index (VHI) calculations, and, after that, we executed a spatio-temporal trend assessment for all biophysical parameters via non-parametric Mann-Kendall (MK) and Sen's Slope Estimator (SSE) tests. In addition, we analyzed land use and land cover changes over the last 20 years using MapBiomas data. The results obtained showed the effectiveness of using these orbital products for long-term environmental analysis and highlighting the municipalities of Belém do São Francisco, Cabrobó, and Floresta. Such counties showed significant trends at the 1% probability level of the biophysical indices (from a decrease in PET, NDVI, and VHI and an increase in albedo and surface temperature) due to greater land use and land cover change, which culminated in a loss of 285.3 km2 (15.6%), 87.04 km2 (5.25%) and 34.82 km2 (0.95%) of their respective original native vegetation cover areas. Such mapping methodology proved to be effective for studies related to the discrimination of spatio-temporal trends in the region of the Desertification Nucleus of Cabrobó, thus being able to be applied in the future in other areas of the Caatinga after the appropriate methodological adjustments.