Overview
With increasing extreme weather events and a changing climate, there is an urgent need to assess, manage, and monitor flooding related risks and communicate such risks to impacted communities in an efficient and timely manner. Yet, achieving these objectives has been complicated by several socio-technical challenges, including 1) the unavailability of appropriate platforms that enable near real-time, two-way communication of flood related hazards for the impacted communities; 2) the lack of high-fidelity models of flooding risk and risk trajectories at a fine spatial and temporal scale; and 3) even when such models are developed, there is limited knowledge about the level of uncertainty embedded in these data and models and how that informs decision-making. To address these challenges, this project aims to design and develop a scalable Community Resilience Information System (CRIS-HAZARD) that leverages crowdsourced, volunteered geographic information (VGI), and social media data for training high-fidelity flooding-risk models to inform citizens about potential flood-related hazards. An important component of CRIS-HAZARD will include robust uncertainty assessments of the predicted risks and their implications on decision-making. This system will support critical community-based planning and policy decisions by providing short and long-term projections of flooding related vulnerabilities and their potential error margins to enhance community resilience.
The proposed CRIS-HAZARD will extend the capabilities of an existing primary platform, the Community Resilience Information System (CRIS) developed by researchers at the University of South Florida (USF) in partnership with AT&T and Argonne National Laboratory (ANL) for Pinellas County, Florida. This project will combine the capabilities of CRIS developers at USF, local stakeholders, and community engagement networks, with the expertise of spatial planning analytics at Georgia Tech, to develop this early warning and knowledge transfer system.
Intellectual Merit
This research will advance the understanding of risk and resilience in coastal communities experiencing persistent flooding events due to extreme weather and a changing climate. It breaks new ground in integrating user-supplied data (crowdsourced) together with near real-time flood prediction models and novel forms of uncertainty analysis to inform decisions about mitigating risks and improving resiliency in coastal communities. It will generate new knowledge in understanding the margins of error when predictions are made using data science tools and citizen-supplied data at high temporal and spatial resolutions. The entire workflow will be incorporated within one platform called CRIS-HAZARD. This platform will be a citizen science-driven initiative using citizens as sensors and citizen engagement as a framework for increasing resiliency. By connecting diverse communities through the process of information and knowledge exchange and by examining how socioeconomically diverse communities are engaging, we will ultimately establish a template for developing smart and resilient communities and realize the goal of a holistic smart city.
Broader Impacts
Successfully deploying CRIS-HAZARD will harness community knowledge through direct and indirect engagement efforts to inform decision-making. We will groom citizen scientists to assist us as partners in our model calibration, information processing, and dissemination efforts. Our proposed project, which leverages USF's ongoing project, the Initiative on Coastal Adaptation and Resilience (iCAR), will connect communities to decision-makers and serve as a roadmap to provide "information on risks and vulnerability to individuals and communities [that] is transparent and easily accessible to all" as outlined by the 2030 goals of National Research Council for resilient communities [NRC 2012]. Over time, we will capture significant training data that will become publicly available for research/development of similar platforms in other U.S. regions. We will also gain insights into the differential impacts of this project on diverse communities in the context of the digital divide and marginalization. We will offer a hybrid graduate course in 'coastal resilience' that will teach students i) integrated modeling with CRIS-HAZARD and uncertainty analysis and ii) community engagement efforts to aid resilience.
Reference
National Research Council. Disaster Resilience: A National Imperative. Tech. rep., The National Academies Press, Washington, D.C., 2012.