Monitoramento e projeção futura da vegetação no parque nacional do Itatiaia através de sensoriamento remote

Abstract

Satellite images of earth observation and meteorological sensors have been used for monitoring land use. Recently products obtained from satellite images have been disseminated, among them, several vegetation indices. EUMETSAT, through the Land -SAF, offers, among other products, the Leaf Area Index (LAI). Daily LAI products have beem acquired in raster format corresponding from 01/01/2010 to 30/12/2010. From a pixel located in the central portion of the Itatiaia National Park, a time series was generated, which was analyzed aiming at assessing the dynamics of leaf area index. The tendency observed in this period indicates that LAI decreased during 2010. It was possible to observe that changes in vegetation have close relationship with changes in rainfall and fires that affect the region. The ARIMA (7 1 0) model was able to describe the behavior of the LAI series, producing white noise and indicating correlations among 1, 6 and 7 days among the past observations. The prediction for future values resulted in an average error of 2.74%, indicating the potential of the model to identify changes in vegetation. Models of ARIMA class, in conjunction with orbital products, stand out as promises for use in the analysis of the vegetation of protected areas.