Information is the most important element in the analysis of fires and its availability is extremely necessary for taking the most appropriate management decisions and actions. Southern Ecuador has a lack of fire information which is urgently needed for land management purposes. In this context, this study concentrates on testing the quality of Rapid Eye imagery to map forest burned areas. Mono-temporal Rapid Eye imagery is analysed in an object based classification environment. The result is obtained after detecting the burned areas, and classifying the land use/land cover. The research proposes a new classification approach to determine the time of burn as well as to discriminate forest burned areas using a single mono-temporal image. The results showed the high suitability of Rapid Eye to map these burned areas (Overall Accuracy 99.1%). Besides, old and recently burned areas can be differentiated by analysing the vegetation cover and charcoal deposition. Finally, Object Based Image Analysis (OBIA) proved to be a powerful tool for detecting forest burned areas in mono-temporal data.
The results of this research were published in the magazine AFZ Der Wald. Source: Corti Meneses N, Kindu M, Schneider T (2017): Brandflächen in Ecuadors Tropenwald per Satellit erfassen. AFZ Der Wald, 19, 25-29. Download