30.10.2019 - Research on methods: measuring disaster recovery with remote sensing
Remote sensing describes the collection of information from the earth’s surface using sensors attached to airplanes, satellites or drones. A DEval research team has published an article on the potential of using remote sensing data to support the evaluation of post-disaster-recovery, in collaboration with the Faculty of Geo-Information Science and Earth Observation (ITC) of the University of Twente. Specifically, the research project assessed the recovery of seven Philippine municipalities devastated by Typhoon Haiyan in 2013.
The study used high-resolution images to create detailed land use maps of four time points, before and after the typhoon, and analysed these with machine learning techniques. These maps were then linked to detailed questionnaire data collected by DEval as part of a former evaluation of a comprehensive land-planning intervention program.
While the study revealed important benefits of the use of remote sensing data in evaluations, a number of challenges became obvious, including the high costs of commercial remote sensing data and the tendency to overestimate damages. Most importantly, the study highlights the many benefits of integrating remote sensing data with socio-economic survey data to evaluate recovery processes. DEval’s survey data provided valuable contextual insights and helped overcome some of the principal limitations of remote sensing data, which can effectively describe but not explain the reason for differential recovery. For example, recovery rates differed by intervention status, reception of reconstruction support, and perceived corruption within communities.