Detection of Trend Change-Point in Passive Microwave and Optical Time Series Using Bayesian Inference over the Dry Chaco Forest

The objective of this article was to compare the performance of two vegetation indices (MODIS EVI (optical) and AMSR-E/ and TMI/TRMM LPRM VOD (microwave)) using an offline Bayesian change-point algorithm to monitor vegetation dynamics (retrospective analysis). We tested this model by simulating 8-day EVI and VOD time series with varying amounts of seasonality, noise, length of the time series and by adding abrupt changes with different magnitudes. This model was applied over real time series (optical and microwave) over a dry forest area in Argentina, Dry Chaco Forest (DCF), where deforestation was common. A comparison with common model used over this region was made (visual inspection). The results compared favorably with Redaf dataset, based on Landsat images. These results show the potential to combine optical and passive microwave indices to identify disturb event. Furthermore, the results obtained in this manuscript were relevant for the DCF region, since they provide a fast and alternative model to the traditional visual analysis made by the national forest service and Redaf.