|Timeline:||April 2016 - 17 April 2020|
Post-disaster recovery is vital for communities hit by disasters to survive and return to normal life conditions. This process is multi-dimensional and complex in terms of measuring its duration and quality after a disaster. A robust method to assess post-disaster recovery, both in terms of physical and functional changes, and to understand its drivers is significant for policy-makers, stakeholders, international organizations, and people who donated money and accordingly want to know what the money was spent on. Remote sensing data is a valuable source of information to understand the post-disaster recovery process, and helps to understand the spatial domain from a wide range of areas to small scales, and providing scientists and authorities with objective information for decision making.
In addition to remote sensing data analysis, Agent-Based Modeling (ABM) will be used to simulate the recovery process in a spatial domain. Thus, ABM will be used to explain the post-disaster recovery process and ascertain the effects of physical and socio-economic parameters influencing the process. The ABM will be also utilized to forecast the developments for the recovery processes, and test different scenarios such as linked to changes in policies. In order to generate an ABM, some information about the study area (Tacloban, the Philippines, following the 2013 typhoon Haiyan) is needed such as census data and socio-economic information of the area. Accordingly, fieldwork was done in April 2017 to collect the necessary data and information. The collected data demonstrates the characteristics of the community and affecting factors in the recovery phase, which are needed in creating ABM.
Currently, I am working on my first paper entitled “A review of remote sensing-based proxies for urban disaster risk management”, which will show the result of my studies and surveys on remote sensing-based proxies. In addition, I continue developing new proxies, such as vehicle classification, to monitor post-disaster recovery processes, and completing the conceptual framework for post-disaster recovery assessment. The next step will be to develop automated remote sensing data processing methods based on robust models, such as learning algorithms and graph-cut approaches, to extract necessary information, such as vehicle detection and classification, to monitor the recovery process.