How can the human dimensions of disaster impacts be more accurately captured and represented in the analysis, modeling and simulation of disasters?
The human dimensions can be more accurately captured and represented through using high-resolution datasets and developing individual-level statistical or simulation (agent-based) models. The impact of a disaster can be broadly categorized into three stages: (1) pre-disaster; (2) during the disaster; and (3) post-disaster. The pre-disaster stage decisions include community members assembling supplies and reinforcements for the disaster. In the event of a disaster;Â the decision processes revolve around protecting life and property. Based on the risk posed by the disaster; agencies might recommend evacuation of residents. Individual households determine if; where; how; and with whom to evacuate (and when to return). Finally; the post-disaster phase involves the recovery of the community and households from the disaster. In the short term; it involves the supply of essentials; restoration of critical infrastructure/lifelines (i.e.; power; road; water) to operation; and rescue efforts. In the medium term; the recovery process includes providing shelter and financial assistance. In the longer term; it includes rebuilding the community while accommodating changes in terms of socio-demographics; built infrastructure; and economic activity. Many aspects of these three stages are associated with human movements and activities. With high-resolution data such as location-based services (LBS) data we can accurately capture human movement/activities that would allow us to predict different behavioral aspects in these three stages such as evacuation; activities during disasters; and recovery trajectory. In addition; if we can develop methodological frameworks (such agent-based simulation models) to incorporate such high-resolution data we can represent and predict human behavior more accurately.
What type of data and supporting research infrastructure would be necessary to enable novel, transdisciplinary approaches to answering these and other human-centered disaster questions?
Human movement data from mobile phones or location-based apps would be the most important data to answer human-centered disaster questions. These datasets should be merged with other datasets such as infrastructure disruptions; point of interest or land use data; and any other data types. Three critical aspects should be considered when building the research infrastructure to incorporate human movement data for disaster research: data availability; privacy; and representativeness. (i) Data availability: human movement data are not widely available or they are very costly to purchase. Only a handful of researchers have access or resources to access such data. This will create biases in the types of questions we will investigate using such data. More research is needed on how such data can be made available while protecting user privacy. Facebook data for good platform has been sharing aggregate movement data; increasing the use of such data; (ii) Privacy protection: regulations and research efforts are needed on how to protect user privacy and share useful data with researchers; (iii) Representativeness: more research is needed to understand the impacts of self-selectivity in location-based apps and how to merge human movement data from location-based apps to ensure population representativeness. Other datasets such as image and social media data may also be useful for understanding disaster damages and socio-economic impacts.
In what ways can US-Japan collaborations advance these questions in new and important ways?
US-Japan collaborations can initiate research that can overcome the above three challenges of data availability; privacy protection and representativeness issues. Collaborations can be initiated to understand the potential and limits of human centered research based on high-resolution movement data. Japanese researchers have already been using cell phone data for disaster research. U.S. researchers can learn from those experiences. U.S. researchers have been conducting research on modeling and simulation of the impacts of disasters to communities across a wide range of disasters from hurricanes to wildfires. Successful collaborations among U.S. and Japanese researchers would allow us to understand the socio-economic impacts of disasters. These collaborations will also allow us to build real-time applications to support disaster response; recovery efforts and infrastructure interdependency issues while a disaster unfolds. There are significant gaps in these directions. In summary; US-Japan collaborations can initiate human-centered research in the following directions: (1) Socio-economic impacts of infrastructure disruptions due to a disaster (2) Interdependency of infrastructure disruptions (3) Building real-time applications for situational awareness; disaster response; and recovery efforts (4) Applications of high-resolution movement data (5) Developing novel approaches to increase the availability; protect user privacy; and ensure representativeness of human movement data.