IDENTIFICATION OF FLUVIAL SENSITIVITY INDEXES TO OIL SPILLS IN ENVIRONMENTS OF THE LOWER AMAZON (CASE STUDY: SANTARÉM-PARÁ)

Abstract

This work proposes to use RapidEye image processing combined with field checking steps to generate a River Sensitivity Index map for oil spills to the border of the municipality of Santarém - Pará. The unsupervised classification which resulted in the generation of the map of identified classes, composed of the classes water, urban area, river beach, floodplain area. The Kappa Geral index was ~ 0.86%, which is considered an excellent result in terms of the accuracy of the classification. The class-map together with the data checked in the field provided sufficient data for the elaboration of the river sensitivity index (RSI) map to the oil spill in the studied area, consisting of 5 indexes (RSI 1- artificial structures, RSI 4- escarpment/ravine, RSI 6- Beach or sheltered sand/pebble, RSI 10A- macrophyte bank and RSI 10B- flooded vegetation). RSI's 1 and 4, which are considered less sensitive to damage caused by contact with oil products, represent 5.1% of the total area of application of the indices. RSI's 6, 10A and 10B represent 94.9% of the total area of application of the index, these are more sensitive indexes.