ASSESSMENT OF PROBA-V DATA FOR DISCRIMINATING BURNED AREAS IN MINAS GERAIS STATE, BRAZIL

摘要

High spatio-temporal resolution optical remote sensingrndata provides opportunities to monitor and discriminaternburned area in a accurate way. This study has thernpurpose to assess the discriminatory performance ofrnmulti-spectral reflectance values of PROBA-V sensorrnand on normalized difference spectral indices (NDSIs),rnsuch as the Normalized Difference Vegetation Indexrn(NDVI) in burned land discrimination a in differentrnland-cover types across Minas Gerais state, Brazil. ThernM separability index was calculated in each land-coverrntype including Cerrado (Tropical Savanna), AtlanticrnForest, agricultural crops, and pastures, to determine thernmost powerful band(s) combinations among thernPROBA-V reflective bands for discriminationrnbetween burnt and unburnt areas The results showedrnthat the BLUE channel is potentially effective for burntarearndiscrimination in the majority of all land coverrntypes ,. Moreover results showed that spectral indexesrnused for discriminating burned areas are vegetation typerndependant.