Monitoring Deforestation Using Multivariate Bayesian Online Changepoint Detection with Outliers

Abstract

Near real time change detection is important for a variety of Earth monitoring applications and remains a high priority for remote sensing science. Data sparsity, subtle changes, seasonal trends, and the presence of outliers make detecting actual landscape changes challenging. Adams and MacKay (2007) introduced Bayesian Online Changepoint Detection (BOCPD), a computationally efficient, exact Bayesian method for change detection. Incorporation of prior information allows for relaxed dependence on dense data and an extensive stable period, making this method applicable to relatively short time series and multiple changepoint detection. In this paper we conduct BOCPD with a multivariate linear regression framework that supports seasonal trends. We introduce a mechanism to make BOCPD robust against occasional outliers without compromising the computational efficiency of an exact posterior change distribution nor the detection latency. We show via simulations that the method effectively detects change in the presence of outliers. The method is then applied to monitor deforestation in Myanmar where we show superior performance compared to current online changepoint detection methods.