作者： Reinaldo Antonio Petta , Cleyber N. Medeiros , Tomoyuki Ohara
DOI:
关键词:
摘要: This paper provides spatial and georeferenced information related to the susceptibility desertification of several areas Northeast Brazil. We aim test usefulness analysis methodologies capture spatialtemporal heterogeneity from environmental gradients, for assessment process at Remote Sensing data. In this sense, integration geo-environmental variables creation indicators associated with development was performed, based on use modeling procedures applied data semi-arid portion Northeastern Brazil region. A set sixteen-year period Landsat 5-TM 7 ETM+ images were explored vegetation soil study local association variability spectral performed. The data, these indicators, allowed identification five different levels (very high, moderate, low very low), geographic domain each class. Based dynamics cover evaluation field we can establish that main results refer there is a decrease biomass region, either dense caatinga areas, but more important, scrub degraded areas. From an perspective, decreasing level are according negative feedbacks process. Considering last ten-year periods comparisons, variances leave almost different, which means pattern, increasing considerably. fact explicit expansion landscapes, during years.