Near real-time change monitoring in tropical forest using MODIS satellite image time series

Monitoring forest cover changes in tropical regions is critical for addressing how deforestation and degradation is impacting carbon storage, biodiversity, and other socio-ecological processes. Satellite remote sensing enables cost-effective and accurate monitoring of forest change at frequent time steps over large areas. However, there is a need for methods that enable fast and accurate analysis of satellite image time series to detect forest change in near real time. More and more change detection techniques become available that are able to process satellite image time series data to detect changes using historical satellite image time series. However, methods to detect changes near-real time are lacking. We are proposing an approach to monitor and detect change in near-real time by comparing it with a seasonal-trend model fitted onto the historical time series. As such, identification of normal and abnormal change in near-real time becomes possible when new image data is captured. The method is based on the “Break For Additive Seasonal Trend” (BFAST) concept (http://bfast.r-forge.r-project.org/). Validation is done (1) simulating 16-day MODIS NDVI time series (2000-2010) with different amount of noise, seasonality and containing abrupt change at the end of the time series (2) by application on real MODIS satellite image time series (MOD13Q1) to detect “forest change” in near real-time. Preliminary results illustrate that abrupt changes at the end of time series are successfully detected while being robust for strong seasonality and atmospheric noise. Cloud masking however remains important as the clouds can be detected as an abnormal change. Once further testing is finalised the method will be made publicly available within the BFAST package for R. The proposed method is a automatic and robust change detection approach that can be applied on different types of data (e.g. Landsat data and future sensors like the Sentinel constellation that provide higher spatial resolution at regular time steps).