Case studies on data-rich and data-poor countries

The aim of Work Package 5 is to assess the needs of decision-makers and end-users involved in the process of post-disaster recovery and to provide useful guidance, tools and recommendations for extracting information from the affected area to help with their decisions. This report follows from Deliverables D5.1 “Comparison of outcomes with end-user needs” and D5.2 “Semi-automated data extraction” where the team had set out to explore the needs of decision-makers and suggested protocols for tools to address their information requirements. This report begins with a summary of findings from the scenario planning game and a review of end-user priorities; it will then describe the methods of detecting post-disaster recovery evaluation and monitoring attributes to aid decision making. The proposed methods in the deliverables D2.6 “Supervised/Unsupervised change detection” and D5.2 “Semi-automated data extraction” for use in post-disaster recovery evaluation and monitoring are tested in detail for data-poor and data-rich scenarios. Semi-automated and automated methods of finding the recovery indicators pertaining to early recovery and monitoring are discussed. Step-by-step guidance for an analyst to follow in order to prepare the images and GIS data layers necessary to execute the semi-automated and automated methods are discussed in section 2. The outputs are presented in detail using case studies in section 3. In order to develop and assess the proposed detection methods, images from two case studies, namely Van in Turkey and Muzaffarabad in Pakistan, both recovering from recent earthquakes, have been used to highlight the differences between data-rich and data-poor countries and hence the constraints on outputs on the proposed methods.