Multi-Temporal Monitoring Of Ecological Succession In Tropical Dry Forests Using Angular - Hyperspectral Data (Chris/Proba)

Abstract:

The tropical dry forest is the largest and most threatened ecosystem in Latin America. Remote sensing can effectively contribute to the surveillance of conservation measurements and laws through the monitoring of natural protected areas, at the required temporal and spatial scales. CHRIS/PROBA is the only satellite that presents quasi-simultaneous multi-angular pointing and hyperspectral spectroscopy. These two characteristics permit the study of structural and compositional traces of successional stages within the tropical dry forest. The current study presents the results of mapping the succession of tropical dry forest in the Parque Estadual de la Mata-Seca, in Minas Gerais, Brazil, using a temporal analysis of CHRIS/PROBA images in a time frame of 7 years, between 2008 and 2014. For the purpose the -55° angle of observation has been used, which enhances spectral differences between successional stages. Spectral Angle Mapper has been used for mapping succession of tropical dry forest and afterwards Change Detection Analysis has been performed. Based on our observations, the tropical dry forest in the Parque Estadual de la Mataseca recovers at a fast rate, for the observed period (2008-2014). More than the 50% of the early and intermediate forests has been recovered to a mature forest. Significantly, around a 12% of old pastures have been converted into forest. The spatial analysis also reveals that the areas that recover most rapidly are located in the east of the Park, close to mature forests. The provision of seeds from these forests might be the cause for the fast recovery.