On the Detectability of Conflict: a Remote Sensing Study of the Rohingya Conflict

Abstract: The detection and quantification of conflict through remote sensing modalities represents a challenging but crucial aspect of human rights monitoring. In this work we demonstrate how utilizing multi-modal data sources can help build a comprehensive picture of conflict and human displacement, using the Rohingya conflict in the state of Rakhine, Myanmar as a case study. We show that time series analysis of fire detections from the Moderate Resolution Imaging Spectroradiometer (MODIS) and Visible Infrared Imaging Radiometer Suite (VIIRS) can reveal anomalous spatial and temporal distributions of fires related to conflict. This work also shows that Synthetic Aperture Radar (SAR) backscatter and coherence data can detect the razing and burning of buildings and villages, even in cloudy conditions. These techniques may be further developed in the future to enable the monitoring and detection of signals originating from these types of conflict.