Abstracts and Responses to Prompts

Flooding during tropical storms and hurricanes can be devastating to healthcare facilities, coming at the time when medical assistance is most needed. For example, Tropical Storm Allison in 2001 devastated Texas Medical Center (TMC) and caused over 1,000 patients to be evacuated, closed hospitals, and caused $2 billion in infrastructure damage. Two major flood mitigation projects were undertaken following Tropical Storm Allison to mitigate future flood damage. Hurricane Harvey in 2017 was the first major test of these mitigation efforts. Although floodwaters were 150 percent higher, there was minimal flood damage to TMC, and facilities remained open. We calculated the return on investment of two major mitigation efforts: Improvements to the Brays Bayou resulted in a 21.9-fold return on investment, and installation of submarine-style waterproof doors at TMC resulted in a 400-fold return on investment. Importantly, these savings only considered damage prevented to physical infrastructure, and were calculated for a single event, even though both efforts will continue to function for any future flooding events. We further identified 91 healthcare facilities in Harris County floodplains that may benefit from similar mitigation investments.

This research investigates the unfolding experience of ‘swarming,’ or collective, emergent action, in the context of Hurricane Harvey. The work is founded on an argument that disaster preparedness and response, as presently constructed, represents a problematic misunderstanding of publics to negotiate disaster in autonomous ways (otherwise known as ‘resilience’). We balance experiences as both citizen responders during the hurricane, and as members of homeland security institutions, with that of our involvement in scholarly research. The paper represents in-depth, ongoing work consisting of interview, observational, and other data analyzed both individually and collaboratively with a range of participants, as we engage a co-autoethnography and participatory action research. Our intellectual contribution, as well as practical impact, is to trace tools and technologies used in the moment and over time to facilitate contextualized, situated understandings of disaster. We also detail socio-technical constructions of disasters past, present, and future, and the translation of these understandings to human action. Findings identify how swarming unfolds, but also becomes neglected within institutional homeland security regimes to offer new insights into improved management of disasters.

After every natural disaster, it is difficult to answer elementary questions on how to provide high quality water supplies and health services. There is no existing digital infrastructure to scientifically determine the hurricane impact on drinking water quality, the severity of a hazard to human health, or baseline data on the sophistication, connectivity, and operations of the distributed physical and related digital infrastructure systems. We test data publication mechanisms after Hurricane Maria in Puerto Rico to understand risks to human health by assessing the spatial and temporal presence of waterborne pathogens in multiple types of systems, demonstrating usability of the Consortium of Universities for the Advancement of Hydrologic Science, Inc. (CUAHSI) HydroShare system as a clearinghouse for data related to Hurricanes Maria, Harvey, and Irma, and developing a prototype cyberinfrastructure to assess environmental and public health impacts. Our resulting archive and research software engineering practices provide a prototype cyberinfrastructure system for researchers to study natural disasters.

The Minority Scholars from Under-Represented Groups in Engineering and the Social Sciences (SURGE) Capacity in Disasters pilot project focuses on the inclusion of minority scholars from science, technology, engineering, and math (STEM) and social science fields to solve complex problems faced in the aftermath of disasters. SURGE, which is a National Science Foundation INCLUDES project, takes an interdisciplinary service-learning approach and is designed to increase the number of underrepresented minorities in STEM disciplines who are interested in hazards mitigation and disaster research. Increasing the involvement of qualified minorities will help reduce broader vulnerability concerns in marginalized communities and help advance scholarship through increasing diversity of thought, perspective, and problem-solving capabilities. SURGE creates a new and replicable model that addresses the diversity concerns in both STEM and disaster fields. Through the utilization of workshops, post disaster reconnaissance experiences, and a multifaceted mentorship program, SURGE will help contribute to resilience efforts in communities across the nation. This project will be of interest to policymakers, educators, and the general public.

Field observations are particularly important in geotechnical engineering because it is difficult to replicate in the laboratory the response of soil deposits built by nature over thousands of years. Detailed mapping and surveying of damaged and undamaged areas provides the data for the well-documented case histories that drive the development of many of our design procedures. Thus, documenting key insights from extreme events advances research and practice. This is a primary goal of the National Science Foundation-sponsored Geotechnical Extreme Events Reconnaissance (GEER) Association.

This study examined college students’ perspectives on how the August 2016 flood in Louisiana affected their educational experiences at four-year universities. The data are derived from qualitative interviews completed six to eight months following the flood with 32 college students who were either affected personally or who had family members who experienced flooding. The students described numerous ways that the flood affected their educational experience that have yet to be described in the academic literature, including attendance, course schedule changes, grade fluctuations, and motivation for schooling. Results show that these impacts varied, in accordance with socioeconomic status, as well as family connectivity. However, all students expressed changes in motivation, which became a consistent theme affecting their educational experiences. This research indicates a need for further studies on college students following a disaster because students are an underrepresented vulnerable population and occupy a socially constructed space between childhood and traditional adulthood.

In our collaborative National Science Foundation Rapid Response Research project titled “How Decision-Making About Risk and Interdependencies Impact Well-Being: A Baseline Study of Communities Affected by Hurricane Harvey,” researchers employ decision-making methods to interview residents impacted by Hurricane Harvey across six counties in southeastern Texas. How do people make decisions when faced with rebuilding their lives and livelihoods following a major disaster? What sorts of interdependencies shape those choices? Do perceptions of risk shift over time and over the course of decisions? This presentation considers these kinds of decision-making processes related to risk and interdependencies situated within stories of recovery and overall well-being. Browne will present preliminary data from interviews with 30 Hurricane Harvey survivors in southeastern Texas across urban, rural, and peri-rural communities. She will discuss methodology and early findings.

The risk of U.S. electrical grid failure is growing. We interviewed 42 households in Highlands and Orange Counties, Florida, as an empirical case study of household resilience to large-scale power failure following Hurricane Irma in fall 2017. Using content analysis and ordinal regression, we determined that residents relied on technical, social, and intellectual resources to prepare for, mitigate, and recover from the power failure event. Respondents with fewer resources demonstrated less resilience overall, suffering more during the event, and experiencing more barriers during recovery. In addition, respondents of lower socioeconomic status self-reported higher levels of stress and lower levels of agency and pathways thinking after the storm, possibly due to their constrained adaptive capacities. Understanding the patterns and specific mechanisms that lead to differential outcomes in hazardscapes that include power failure may help address gaps in preparedness and response efforts that often leave those in most need at highest risk.

We hypothesized that runoff from Hurricane Harvey, the most extreme rain event in our country’s history, mobilized contaminants, including metals and microorganisms. Forty grab samples were collected from different locations, mainly from southwestern Houston, representing a variety of (sub)urban land use patterns, including residential, industrial, commercial, educational, and religious activities. All samples were first filtered through sterile 0.22 micrometer membrane filters. Fifty representative, transition, and rare earth elements and dissolved organic carbon were analyzed in the filtrate. Iron (Fe), strontium (Sr), and barium (Ba) were moderately elevated, regardless of land usage. Isolated strong peaks in potassium (K), lithium (Li), and aresenic (As) may be the result of landscaping and construction materials. Additionally, DNA was extracted from 12 filters using a bead beating protocol. The microbial community composition was determined using 16S ribosomal RNA next-generation sequencing. Results showed the variation of relative abundances of bacterial community composition across the different sampling locations. 

Post-disaster blood demand and supply surges often lead to undesirable situations, such as delivery delays (e.g., the events of September 11, 2001) and a waste of donated blood (e.g., 2008 Wenchuan earthquake). However, research on blood donation and management has focused on military conflicts and the supply side (e.g., psycho-social aspects of blood donation). We contribute to this domain of research by examining the end-to-end chain, starting from upstream donor management to downstream hospital usage. Our study focuses on two scenarios, namely routine emergencies and disasters, and uses two lenses, namely, operations and organizational-institutional. With Asia rapidly urbanizing and becoming the most disaster-prone region globally, we study three urban settings located in the region. Our study uses a mixed method approach, with qualitative (e.g., archives, interviews, field observation) and quantitative (e.g., modeling and simulations) approaches. We will present our preliminary findings from our ongoing fieldwork on a Southeast Asian city and highlight the potential contribution from our comparative and multidisciplinary approach.

Historical markers, monuments to disaster victims, and flood levels painted on structures are seen in many communities. The erection of such markers and other commemorative events are often part of the recovery process following a major disaster. Social scientists have noted how such activities promote both personal and community recovery—physically, emotionally, socially, and economically. This paper explores the effectiveness of historical markers, monuments, and flood level indicators in communicating hazard vulnerability to residents and visitors long after the event. A survey of Wisconsin’s 72 county emergency management directors regarding their awareness of the markers and their use, or potential use, in communicating hazard vulnerability indicates divergent knowledge and opinions, ranging from approval to skepticism. Managers in counties with such markers were much more likely to indicate that they were “very effective” in enhancing community awareness of hazards and appropriate responses.

This study will provide an understanding of the current strategies employed to support disaster recovery in schools and their effectiveness as judged by school personnel. The ultimate goal of this study is to provide research-based guidance to educators and policymakers on which responses are associated with the best outcomes for students. This project studies areas affected by Hurricane Harvey (August 2017) and Hurricane Matthew (October 2016) by addressing three research questions relevant to school recovery: What impacts did the storms have?; How did schools respond to support students?; and How successful were those responses? The research team conducted interviews, focus groups, and surveys with personnel from schools, districts, and state agencies in a sample of 20 districts in Texas and North Carolina that were impacted by the hurricanes. Early results suggest that schools in North Carolina had little evidence to inform their recovery efforts after Hurricane Matthew.

Hurricanes Harvey and Irma, and the Santa Rosa wildfires, provided ample opportunities to learn about the ways in which people make decisions surrounding pets and evacuation. Through mixed methods data collection, we targeted two key groups: evacuees and program coordinators. Evacuees for this project are people who either left or considered leaving because of the disasters. We conducted face-to-face and web-based data collection for both of these target groups. Preliminary results indicate that evacuees place a strong emphasis on emotions such as regret, grief, and stress associated with being separated from their animals. Furthermore, stress is linked to logistics and supplies (e.g., having space, having a kennel, and moving the animal safely from place to place). For program coordinators, pet overpopulation is seen as linked to the overall animal care ecosystem and directly impacts the ways in which the evacuation will unfold for animals and their humans. 

This project examines how organizations and businesses contribute to community resilience through the work they do in order to minimize interruptions caused by disaster. Field research has been ongoing since late September 2017 in Houston, following the flooding caused by Hurricane Harvey. For this research, we have recruited businesses and organizations to facilitate site visit observations, documented collection and analysis, and conducted interviews with owners, leaders, and human resources managers. We aim to triangulate informant interviews with internal documents to assess resilience as an organizing and communicating process.

We develop and validate the concept of mobile energy storage (ES) units that can provide backup flexibility to power grids. Using the real-life predictions and ex post data from Hurricane Harvey, we designed a data-driven approach to enable the optimal routing, reconfiguration, and control of mobile ES units to enhance power grid resiliency. We also developed a new electrochemical ES mechanism that is capable of generating two transportable energy carriers: redox-active species (such as those used on redox flow batteries) and hydrogen (e.g., available in refineries along the U.S. Gulf Coast). Our results demonstrate that such technologies are effective to overcome the effects of natural disasters.

With the frequency of recent hazard events and the ease with which an increasing number of images can be collected, data collected to document the consequences of various types of disasters is growing. To keep up with this data deluge, artificial intelligence and machine learning must be exploited to automate the processing of these unstructured image data. In response to this challenge, we have developed these capabilities and integrated them into the Automated Reconnaissance Image Organizer (ARIO). ARIO provides the data collector, from the field, the power to automatically organize all of the images collected to help make informed decisions. The tool has several built-in classifiers and executes quickly, requiring less than 0.5 seconds per image. The focus of this presentation will be on the development of ARIO, the lessons learned by using real image data from past missions, and how to access and use the tool.

Hazardous wind events frequently occur in the United States. Hurricane Irma, which hit the southeastern coast in 2017, left a majority of damage concentrated on low-rise buildings. Roof-to-wall connections play an important role in the behavior of wood-frame buildings when exposed to wind-induced loadings. Wind actions on roof-to-wall connections have not been adequately estimated yet. An extensive, large-scale aerodynamic testing study is being conducted at the National Science Foundation Natural Hazard Engineering Research Infrastructure Wall of Wind Experimental Facility to investigate wind actions resulting from simulated hurricane winds on a gable wood frame building model with roof-to-wall connections. The study adopted a length scale of 1:4 and a velocity scale of 1:2. Load cells were mounted at the roof-to-wall connection level to measure the effective net wind forces. Also, internal pressures were measured using pressure taps. The model was tested under different wind speeds and directions varying from 0° to 315°. Results were compared to pressure coefficients recommended by the American Society of Civil Engineers 7 code for cases of Component and Cladding and Main Wind-Force Resisting System. Experimental results showed a better match with the ASCE loadings based on the Component and Cladding case.

The Earthquake Engineering Research Institute (EERI) is a multidisciplinary, international, nonprofit, technical society with a long history of deploying multidisciplinary teams after an earthquake through their Learning from Earthquakes (LFE) program. EERI members understand the need for streamlined information sharing after a damaging earthquake to assist with reconnaissance and research. This presentation covers EERI’s newly-founded Virtual Earthquake Reconnaissance Team (VERT) and the organization’s new earthquake clearinghouse web presence, and how the two work together to develop an efficient approach that provides targeted information to the earthquake science and engineering community on specific topics of interest. This partnership also informs decisions about how EERI should respond to an earthquake. VERT harnesses enthusiasm of motivated younger members to perform virtual post-earthquake reconnaissance. VERT members prepare teams for the field, provide crucial information to EERI before a decision is made to respond, and help teams post-process data upon returning from a reconnaissance trip. 

Hurricane Matthew in 2016 demonstrated how families responded to adverse impacts and, in particular, how they evacuated and what factors influenced their decisions to evacuate. There are a great deal of empirical studies investigating how households make hurricane evacuation decisions. However, those studies considered each household as a decision-making unit and, thus, largely failed to capture how differences within families could affect decision-making. By conducting a massive mail survey with households in the metropolitan area of Jacksonville, Florida, we explored how previously disconnected bodies of literature—from disciplines including emergency management, family science, transportation logistics, and others—can be synthesized to address this gap in the research. We proposed an updated Protective Action Decision Model, where previously neglected intra-family factors were added. Preliminary findings from the household survey implied significantly different perceptions between two household members on adequate financial resources, power distribution, family relationship prioritization, equal access to and control over financial resources, and specialization in major decision-making.

Big spatial data acquisition and processing technologies are increasingly used for post-disaster assessment, generating large spatiotemporal datasets. Acquisition and analysis of these datasets offers tremendous opportunities for advancing natural hazards research. In spite of these great values, the vast size and complex processing requirements of these datasets make it challenging to effectively use them in hazard research. In this paper, we describe the progresses, challenges, and an envisioned data infrastructure for accelerated knowledge discovery from large disaster datasets. We use a National Science Foundation Geotechnical Extreme Events Reconnaissance-funded Hurricane Harvey study as a use case to characterize data integration and sharing needs. The preliminary findings are that federation and fusion of multi-sourced data can greatly facilitate damage assessment, content-based retrieval is key to efficient exploration of large disaster datasets, and cloud-based infrastructures provide promising capabilities in data archiving and sharing, but require more research on dedicated data analytics.

In October 2016, Hurricane Matthew caused flooding throughout much of North Carolina. The Center of Excellence for Risk-Based Community Resilience Planning, with collaborators from the National Institute of Standards and Technology, conducted a field study in November 2016 in Lumberton, North Carolina, a racially and socioeconomically diverse community. The goals of the study were to collect perishable data about housing damage and dislocation and to allow for longitudinal recovery modeling through documentation of a baseline. The interdisciplinary team conducted damage assessments and household surveys for a random sample of housing units. Damage assessments were used for developing damage fragilities. Logistic regression models were developed, linking damage with household data to understand how different factors shape dislocation. Findings indicate that dislocation has impacted a substantial proportion of households in Lumberton, and for many this has been a protracted process. Renter and minority households in structures with higher damage were most likely to dislocate.

The U.S. Virgin Islands were devastated by back-to-back Category 5 hurricanes in the summer of 2017. We carried out community surveys and water quality analysis on St. Thomas island three months after Hurricane Irma. The survey results showed that most of the residents (59 percent) were satisfied with governmental responses to managing water in the post-disaster period. However, less than 17 percent of the residents in the lower income group considered the water “safe” to drink. Microbial water quality analyses results showed the cistern samples contained diverse microbial populations. Microbial compositions in three of the cisterns had a relatively higher portion of Proteobacteria and higher salinity, implying cross contamination by coastal water. We are in the process of identifying members of these microbial populations that have pathogenic potential. These results will shed light on potential health risks associated with water supplies in the post-hurricane U.S. Virgin Islands.

Resilience is a socially constructed concept and, thus, it's important to understand the public's perception of community resilience, especially among people living in high-risk areas. In this paper, we explore the correlations between perceived community resilience and livelihood assets and household preparedness activities. A 2018 questionnaire that collected survey data from 1,000 individuals in three villages in Weinan, Shanxi Province, China, was used for analysis. The perceived community resilience is measured by the Communities Advancing Resilience Toolkit (CART), and varied forms of livelihood capital (e.g., social, financial, human, natural, and physical) and household preparedness activities are covered. Linear regression analyses demonstrate that people with higher social capital and awareness preparedness would have significantly higher perceived degrees of community resilience, while the physical capital and material preparedness associations with perceived community resilience are not significant.  

The 2013 Uttarakhand flash flood was such a surprise that the predominant sources of information for those at risk were environmental cues and, secondarily, peer warnings rather than official warnings. Of those who received warnings, few received information other than the identity of the threat. A survey of 316 survivors found that most people’s first response was to immediately evacuate, but some stayed for additional information, or to engage in evacuation preparations. Unfortunately, engaging in these behaviors delayed their final evacuations. Regression analyses revealed that immediate evacuation and evacuation delay were both predicted best by information search and positive affect, but correlation analyses indicated that a number of other models were also plausible. Final evacuation was best predicted by immediate evacuation and, to a significantly lesser extent, household togetherness. For this dependent variable also, correlation analyses indicated that a number of other variables were almost as predictive as household togetherness. The data from this event revealed some similarities but also some differences from other studies of rapid onset disasters, indicating that further research is needed to determine the relative importance of situational and cultural characteristics in producing the observed differences in household response to flash floods.

As climate-induced disasters are increasingly causing displacement of large populations, understanding resettlement implications is a growing need. Because relevant studies require long-term observation in each unique setting, scholarly knowledge is still limited. This research targets coastal residents who resided in a zone that was designated no-build after Typhoon Haiyan, and explores differences in resettlement processes, livelihood changes, and future prospects of residents currently living in original neighborhoods and relocation sites. Since 2014, we conducted six sets of 40 semi-structured interviews in both the coastal and relocation sites over the three-year period, in addition to interviews with city officers to understand relocation progress. Observation explains that, first, timing of involvement in the relocation process made a difference in people’s current location; second, living in relocation sites is more complicated than original sites; and third, relocation became less favored among the target population, although they were initially attracted for safety and housing.

In the aftermath of Hurricane Harvey, a significant volume of waste and debris was generated in Houston and Beaumont, Texas. The management of post-disaster debris is important because it contributes an estimated 25 percent to disaster management costs, can generate five to fifteen times the annual waste generation rates of the affected community, and specifically leads to the most federal funding being lost due to mismanagement. As a result, the Federal Emergency Management Agency will reimburse consulting companies for the costs of professional debris monitoring services. These companies deploy a state-of-the-art automated debris management system to electronically collect, store, and manage debris project data and information. However, estimates of waste volume are still qualitative. A National Science Foundation Rapid Response Research grant facilitated field visits to Beaumont to obtain records of waste debris weight, consultant monitoring tickets, and to explore the application of smartphones and unmanned aerial vehicles to quantify waste debris piles using photogrammetry techniques.  

Hurricane Harvey led to unprecedented flooding in the Houston metropolitan area. The ability to construct flood hydrographs in urban areas in real time during flash floods, hurricanes, and other extreme weather events is difficult because of the low spatial density of water level measurements and the complex interactions of built infrastructure, ground topography, and natural landscapes with flowing water. To achieve better representation of flood hydrographs, we leveraged time-lapse images and videos and image processing algorithms. In particular, this research explores the application of image segmentation to detect water in photos and subsequently estimate the flood level using a reference object (e.g., a bridge pier, light pole, or car). A case study of a bridge overpass in Houston is provided to demonstrate the workflow for flood hydrograph reconstruction.

Hurricane Harvey caused unprecedented devastation to huge parts of Southeast Texas, particularly damaging the wastewater infrastructure resulting in release of sewage contamination into environmental waters. The main objective of this study was to conduct a preliminary assessment of fecal indicator bacteria (i.e., E. coli and Enterococci) and human-associated fecal genetic markers (i.e., human-associated Bacteroidales), measured using quantitative polymerase chain reaction (qPCR) assays, across a Texas river impacted by Hurricane Harvey. Water samples were collected along the Guadalupe River from September through December 2017. The most heavily flooded sites showed the highest abundance of fecal indicator bacteria and human-associated Bacteroidales markers, indicating that a large number of sewage overflows and stormwater runoff occurred during Harvey flooding. These findings suggest that high levels of human fecal contamination were introduced into waterways draining into the Gulf of Mexico, and impaired surface water quality. In general, results of this initial microbiological contaminant assessment will serve as baseline information for follow-on studies to monitor existing and emerging public health risks to residents of Texas and potential long-term environmental impacts upon the water resources in the impacted regions.

This research seeks to better bring together the parallel processes yielding international agreements and action on disasters, climate change, and sustainable development. In examining basic definitions and modes of action, significant separation among the topics tends to be found, especially with climate change seeking to distance itself from other topics. No reason exists for climate change to be separated from wider disaster risk reduction and sustainable development processes. Instead, the evidence provides a resolution through placing climate change within wider disaster risk reduction and, in turn, placing disaster risk reduction within sustainable development. Based on this research, a conceptual approach for policy and practice is provided, although entrenched territorialism and inertia remain barriers.

Using Hurricane Harvey as a case study, this study examines the extent to which communities are aiming to build resilience during the disaster recovery process. We conducted key informant interviews with local health departments and one office of emergency management to determine if and how their communities are incorporating public health considerations into the visioning, planning, implementation, and assessment phases of disaster recovery. Utilizing a combined inductive and deductive approach, we coded and thematically analyzed interview notes and/or transcripts. We found that communities are continuing to miss opportunities to “build back better” and to build resilience through disaster recovery. Local health departments would benefit from additional resources, support, and technical assistance designed to facilitate working across sectors and resilience-building during recovery.

Hurricanes Harvey, Irma, and Maria created an unprecedented path of destruction before making landfall in U.S. territories. In response, the presenter coordinated a community-wide reconnaissance effort for all three storms, deploying teams to Texas, Florida, Puerto Rico, and the U.S. Virgin Islands between September and November 2017 under a unified set of policies, practices, and data standards. The efforts resulted in thousands of door-to-door damage assessments using a Fulcrum mobile smartphone app, enabling a unified assessment framework feeding into a rigorous quality control process. At select locations, unmanned aerial surveys, Light Detection and Ranging (LiDAR) technologies, and coastal surveys generated additional valuable information. This paper will introduce the lessons learned in this extended coordinated reconnaissance effort across diverse geographies and the advantages of pre-establishing regional nodes at universities across the impacted areas.

On August 8, 2009, Typhoon Morakot invaded Taiwan and caused the highest loss in the history of natural hazards in Taiwan. This study used the first-year, post-disaster recovery survey to verify the effectiveness of various rehabilitation assistance from family, friends and relatives (FR), government, and nongovernmental organizations (NGOs) on subjective recovery time. The findings are that victims who once lived in shelters have longer recovery times, indicating that the environment of shelter may not be good for recovery; receiving goods and manpower from FR can significantly reduce the recovery time; receiving financial aid and accommodations from FR delays the recovery that may result from dependency; the psychological counseling provided by government and NGOs delays the recovery that may be caused by post-traumatic stress disorder; religion can accelerate recovery; and the manpower provided by government and NGOs can only accelerate recovery in some areas.

With the advent of the big data era, real-time human response data extracted from social media could provide new opportunities for studying disparities in disaster resilience. This paper will describe an ongoing interdisciplinary project designed to examine the issue of social and geographical disparities of Twitter use and how it might affect resilience. The project compares Twitter data from two events, Hurricane Isaac and Hurricane Sandy, both of which occurred in 2012. The project includes four components: development of data-mining algorithms for the evaluation and visualization of Twitter data, analysis of social media and resilience disparities at the community (zip-code) level across the emergency management cycle, an online survey of Twitter and non-Twitter users to understand effects of social media use on individual resilience, social networks, social support, social disparities, and analysis of bidirectional communication among residents and governmental and nongovernmental organizations throughout each event cycle. Knowledge gained from this project should help develop strategies to reduce disparities, create effective social media campaigns and emergency management outreach, and promote resilience to disasters.  

This National Science Foundation RAPID project focuses on the changing roles and effects of social media use in disaster resilience through the case of Hurricane Harvey. A new phenomenon emerged during this event: Many resorted to social media to call for rescue when 911 was overloaded. This changing use of social media marks Harvey as one of the very first disastrous events in which social media played an important role in fast-responding rescue missions. The overarching research question is: How effective is social media in enhancing resilience such as its new role of response and rescue, and do we see an increase or decrease in the geographical and social disparities of social media use that may have affected the outcome and the resilience of individuals and communities? This interdisciplinary project collected time-sensitive Twitter data and online surveys of individuals and organizations in the flood-affected communities in the Houston region to examine the patterns, effects, and new roles of social media in disaster resilience. We will present preliminary results from analyses of the Twitter and individual survey data. 

To understand the environmental impact of Hurricane Harvey, we monitored water quality and fecal indicator bacteria levels at three stations within Clear Lake, an estuary between Houston and Galveston, and at three stations in bayous that feed into the system. Before Harvey, salinity within Clear Lake ranged from 9-11 practical salinity units (PSU). Immediately after the storm, salinity dropped to less than 1 PSU and then gradually increased to pre-storm levels over two months. Dissolved inorganic nutrient levels were also relatively low immediately after Harvey. Fecal indicator bacteria levels were elevated immediately after the storm; however, after one week, the E. coli levels had decreased to less than 100 most probable number of bacteria per 100 milliliters effluent (<100 MPN/100 mL) and remained relatively low in samples collected in the bayou stations for the next two months. These results suggest that the hurricane had a dramatic, short-term impact on nutrients and fecal contamination of surface waters, but conditions returned to background levels within a week.

Food insecurity is a chronic problem in the United States that affects over 40 million people annually under normal conditions. This reality dramatically worsens after a disaster, as retailers and food banks struggle to meet local needs. Food supply disruptions exacerbate the situation precisely when large numbers of people are thrust into disaster-related food insecurity. Unfortunately, very little systematic research exists on the consequences of disaster events for food bank operations. This research specifically examines the food aid landscape after Hurricane Harvey in 2017, and seeks to understand how it affected the local food aid environment both spatially and quantitatively. This National Science Foundation RAPID research project utilizes survey methods, census data, and food bank distribution data from a coastal food bank that has endured numerous hurricanes. Geographic information system (GIS) mapping coupled with statistical tests are used to identify significant changes in distribution patterns of food assistance.

The Federal Emergency Management Agency (FEMA) is beginning to operationalize crowdsourcing to enhance situational awareness and decision-making at the National Response Coordination Center (NRCC). Crowdsourcing is an open call for voluntary assistance from a large group of individuals, either in person or online. FEMA harnessed crowdsourcing through the collective power of digital volunteers collecting, analyzing, and validating social media, and crowdsourced data to enhance situational awareness and decision-making during the response and recovery efforts during the 2017 hurricane season. For the first time, FEMA appointed two crowdsourcing coordinators at the NRCC in response to Hurricane Maria. The coordinators engaged with six digital volunteer networks with over 5,700 total digital volunteers collecting, validating, and analyzing imagery, social media, and other open data. This resulted in 10 real-time, crowdsourced maps, which enabled rapid damage assessments, enhanced situational awareness, and supported operational decision-making for emergency management.

Spontaneous volunteers are individuals or groups who self-deploy on impulse shortly after large-scale disaster events. This study identifies unique aspects of spontaneous volunteer management from an operations research perspective compared to other labor assignment problems in the for-profit sector, as well as volunteer scheduling in non-crisis situations. Our findings are based on field work with nonprofit organizations who utilized spontaneous volunteers in relief distribution efforts following Hurricane Harvey in Houston.

The massive tsunami after the magnitude 9.0 Great East Japan Earthquake (GEJE) of 2011 devastated communities along 500 km (311 miles) of northeast Japan's Tohoku coastline. This paper analyses housing recovery support programs post-GEJE in the context of recent housing reconstruction policy development history. Although Japan has well-established post-disaster recovery precedents and reconstruction policies, new policies were introduced after the vast and complex GEJE. Characteristics of the ongoing recovery include: selection of projects by local governments relying on national government subsidies, strong land use control, and large-scale residential relocation. Both U.S. and Japanese housing recovery policy is characterized as primarily market-driven with limited government involvement in private housing reconstruction. However, Japanese programs focus on public housing provision, with varied support for private housing. Through consideration of Japanese housing recovery policy after the GEJE, this paper clarifies key differences from the U.S. system, and resulting implications for residents' life recovery.

Despite legislation requiring remediation of earthquake-prone buildings (EPB), there are significant obstacles to retrofitting these buildings. This research examines mitigation actions following the 2010 and 2011 Canterbury earthquakes. The study obtained data on commercial and public buildings removed from the Wellington City Council EPB List from 2012 through 2016 due to mitigation actions (e.g., strengthening). It also examined voluntary home assessments. Results show removal of significant numbers of EPBs from the EPB Register, with strengthening being the most frequent reason. This suggests that incentives are enhancing earthquake preparation, before the legislative deadline and above the minimum standard, suggesting an emerging norm. In contrast, the home assessments data show a brief spike after the 2013 Cook Strait earthquakes. This spike suggests that in the absence of legislation or insurance incentives, citizens’ actions are only briefly influenced by experiencing earthquakes. These contrasting findings suggest the value of legislation and norms to drive mitigation actions.

Researchers have often sought to understand group behavior in disasters, including behavior in emergent, expanding, extending, and established organizations. While there has been significant research on these four types of organizations, there has been less research on volunteer emergent groups that quickly and effectively transitioned into expanding organizations that self-deployed to another state to provide lifesaving capabilities, as the Cajun Navy did. This group conducted boat rescues in 2016 Louisiana flooding and Hurricanes Harvey, Irma, and Maria. This research uses qualitative interviews with members of the Cajun Navy to understand the motivations, processes, and resources they leveraged, as well as interviews with emergency management about the integration of spontaneous volunteers into formal relief efforts. The work contributes to theories on social organization, volunteer behavior and motivations, social capital, and technology and social media adoption. We will present preliminary results from the National Science Foundation RAPID project. 

Long-term recovery remains one of the least studied but most complicated phases of disaster. Supporting unmet needs is one crucial aspect of this phase. Local communities often form long-term recovery groups as either organizations or committees to support residents with unmet recovery needs. This process requires coordination and collaboration between government, private, and nonprofit organizations. This presentation provides results from ongoing research about how communities structure long-term recovery coordination. We will present results from a community affected by Hurricane Harvey, along with results from six other communities we have been studying for the past four years. This research involved interviews with more than 120 people from public, private, and nonprofit organizations, and field observations of recovery meetings. Our results highlight challenges and opportunities in recovery management, a need for nonprofit training on disaster recovery specifically, and often missed opportunities to promote disaster resilience. 

The RAPID Facility—part of the National Science Foundation’s Natural Hazards Engineering Research Infrastructure—is developing custom-made fieldwork software, RApp, to afford users the ability to identify, capture, aggregate, organize, store, and manage social science, engineering, and geoscience reconnaissance data and, to a lesser extent, disseminate, analyze, and visualize those data. A requirement of RApp is to promote communication, coordination, and collaboration among researchers across disciplines to foster the next generation of post-event reconnaissance. To meet this objective, RApp is being designed with a human-centered approach that includes user research, ideation, prototyping, and iteration. Ultimately, RApp will help researchers specify relationships between collected data in ways not currently possible with off-the-shelf data collection and management tools.

The National Institutes of Health Disaster Research Response Program (DR2) was created by the National Institute for Environmental Health Sciences and the National Library of Medicine to facilitate vital data collection and research into national response and recovery efforts. The program reduces the deployment time of health research responses after disasters through the promotion of survey instruments, processes, and relationships. Accomplishments include: public web-based access to over 350 data collection tools; a novel Institutional Review Board (IRB) review protocol that can be modified for future disaster situations; national workshops; large-scale tabletop exercises; and new networks linking academia, public health officials, and impacted communities to test initiatives and address ethical and IRB considerations. Together with multiple partners, including Canada and Japan, progress is being made to advance our collective disaster research capabilities to improve preparedness, response, and recovery efforts.

Extreme events test physical infrastructure, cascading to socioeconomic disruptions in ways and on a scale that cannot be replicated in a laboratory. The National Institute of Standards and Technology (NIST) leads a multidisciplinary Disaster and Failure Studies Program within their Engineering Laboratory intended to standardize disaster field deployment, assessment, and reporting protocols. The Center of Excellence for Risk-Based Community Resilience Planning (CoE) has identified field studies as one of its major research areas to help validate predictive models. NIST and the CoE are collaborating on a longitudinal study to assess the impact that riverine flooding had on the small city of Lumberton, North Carolina and its subsequent recovery. The collection, handling, and long-term storage of spatially collected data can be difficult for large field teams. The field team used geolocation approaches to aggregate collected data onto a map of the affected area using geographic information system (GIS) software, and a GIS-enabled web viewer to preserve the data, creating a baseline of the damage and socioeconomic data.

Large-scale damage to the power infrastructure from hurricanes and high-wind events can have devastating ripple effects on infrastructure lifelines and communities. Using Hurricane Irma’s impact on Florida as a case study, we examined the differences in electric power outages and restoration rates between urban and rural counties, the effect of hurricane force winds on power outages, and associations with socioeconomic vulnerability. Using spatial autoregressive lag models, we found that rural counties, predominantly served by rural cooperatives and municipally-owned utilities, experienced longer power outages and much slower and more uneven recovery. The results also show that wind swath helps predict the duration of power outages. Finally, there is positive spatial dependence between power outages and several social vulnerability indicators. The findings from our study highlight the need for additional research to deepen our understanding of how power restoration after hurricanes contributes to and is impacted by the socioeconomic vulnerabilities of communities.

Most heat-related studies and health campaigns fail to target pregnant women and fetuses despite a growing body of literature that documents the vulnerabilities of these subgroups; heat exposure exacerbates the risk of shorter gestation, lower birth weight, and fetal death. Researchers are calling for health care providers to communicate these dangers to pregnant patients and encourage them to reduce heat exposure. But what perception and knowledge do providers have on the topic? We report data from a pilot study in El Paso, Texas. First, we report on our analysis of 12 interviews with El Paso maternal health providers. The interview participants were mostly unaware of the health risks associated with heat exposure—in fact, many of them normalized such exposure because of the region they serve. Second, we report on a “Heat Exposure and Maternal Health” workshop offered in El Paso, where we saw important gains from participants' pre- to post-surveys. 

The objective of this study was to examine the relationship among the structure of stakeholders' social networks, the level of integration among the network of hazard mitigation and infrastructure plans, and the extent of community vulnerability to flooding risks. First, a countywide stakeholder survey was designed and deployed to several state, county, and city officials; infrastructure agencies; private companies; and non-governmental organizations in Harris County, Texas, and in the aftermath of Hurricane Harvey. A total of 198 responses from 138 organizations were collected. Using the survey results, the social network of stakeholders involved in hazard mitigation and infrastructure development was mapped and analyzed. Second, we analyzed the level of integration of the Harris County/Houston network of plans that govern the location and design of different infrastructures and land uses and ecosystem protection in flood hazard areas. The results uncovered new insights regarding the relationships among the stakeholder networks, the level of plan integration, and the community’s physical vulnerability.

Parallel governance structures, or entities vested with special authority and powers that allow them to bypass the established norms and procedures of governance, have increasingly become the normative model for governments to coordinate and execute recovery from catastrophic disasters in international contexts. Yet, the academic literature on these structures is extremely limited. Based primarily on interviews and review of secondary sources, this paper compares parallel governance structures established after the 2001 Gujarat earthquake in India (Gujarat State Disaster Management Authority), the 2011 Great East Japan Earthquake and tsunami in Japan (Reconstruction Agency), and the 2015 Gorkha earthquake in Nepal (National Reconstruction Authority). It suggests that, although parallel governance structures are often framed as a panacea in post-disaster contexts, they fail to ensure a speedy, efficient, and transparent recovery, and they undermine local institutions. The paper concludes with future directions for research on disaster governance and policy recommendations for national governments.

This research develops a conceptual framework that represents decision makers and the information and uncertainty linkages and flows in hurricane evacuation decision-making. We investigate various sources of uncertainties, based on the extant literature and our preliminary research findings, which affect the overall decision-making process. The framework conceptually models how the information and uncertainty propagates through multiple stakeholder groups prior to and during evacuation. The representation starts with meteorological data flowing to weather forecast offices, the National Hurricane Center, and private sector meteorological broadcast entities. Predictions then flow to emergency management agencies and mass and social media. Both predictions and evacuation messages flow to individual households, who also obtain information from other sources. In this presentation, we highlight how the results of a survey of operational meteorologists and initial findings from a household survey in the context of Hurricane Matthew provide details for various components of the conceptual framework.

The purpose of this community resilience field study was to provide a comprehensive set of initial conditions data about flooding impacts to households, housing, and critical infrastructure (e.g., electric power network, water, gas). It was critical that the data collection activities generated representative data that could be utilized to improve flood hazard fragilities for residential housing, develop population dislocation models, and establish baseline data for housing recovery modeling in a longitudinal study. An important addition to these general goals was to develop a field research strategy for undertaking systematic interdisciplinary engineering and social science data collection activities that can be replicated and improved upon over time. This paper reviews the sampling and field research strategy developed and provides an illustration of how engineering and social science data collected using this interdisciplinary approach are combined to model household dislocation. 

This National Science Foundation EArly-concept Grant for Exploratory Research (EAGER) project will establish two scientific platforms and coordinating networks for Social Science Extreme Events Reconnaissance (SSEER) and Interdisciplinary Science and Engineering Extreme Events Reconnaissance (ISEEER). SSEER and ISEEER are designed to build the capacity of social science, engineering, and interdisciplinary hazards and disaster research communities. The ultimate vision is to prepare individual researchers and teams to carry out extreme events rapid reconnaissance research that is coordinated, comprehensive, coherent, ethical, and scientifically rigorous. Drawing on the science of team science literature and convergence theory, this project is rooted in a commitment to bringing together diverse disciplinary perspectives to study, understand, and solve complex problems.

After Hurricane Harvey in August 2017, Virginia Tech and Texas A&M University’s AgriLife Extension provided free well water sampling kits and well user surveys to residents in flood-impacted counties in Texas to characterize water quality in private wells after flooding and to examine short- and long-term testing, resource, and recovery needs. During our first sampling campaign (September through November 2017), our team analyzed 630 well water samples from 21 counties in Texas. The primary concern for flood-impacted wells is contamination associated with flood water and septic system failures, and we observed that 12 percent of wells tested positive for E. coli bacteria, an incident rate three times higher than observed during non-flooding conditions. Not surprisingly, residents were no longer confident in the safety of their water after flooding, and there were barriers to access recovery information and resource needs. Results from this effort provide insight into water quality and resource challenges encountered by unregulated well users during hurricane recovery.

Our team mobilized sampling campaigns for three natural disaster events: Louisiana’s summer flood of 2016, Texas’ Hurricane Harvey, and Florida’s Hurricane Irma. Our goal was to evaluate drinking water quality and recovery needs in flood-impacted communities served by unregulated private drinking water wells. The impact of our response varied considerably in these areas; 113 wells were sampled in Louisiana, 1,190 wells in Texas, and 178 wells in Florida. During this presentation, we describe our strategies for mobilizing emergency sampling campaigns for everything from advertising and coordinating testing opportunities to developing and refining survey tools. We aim to provide an overview of the various barriers, hardships, lessons learned, and opportunities encountered by our team in both the field and lab in order to help other teams interested in mobilizing a successful emergency response well sampling campaign.

In the United States and abroad, prisons remain institutions characterized by social and geographic isolation. Therefore, the role of prison labor in disasters remains drastically under-researched. Studies examining the practice have focused on prisoners fighting wildfires in states such as California and Wyoming. This study extends this research by examining prisoner assistance in Georgia in response to Hurricane Irma. Georgia uniquely relies on prisoners, not only as a source of general disaster labor, but also as traditional firefighters for emergency response. Through qualitative interviews with various corrections, fire, emergency, and government officials, this study examines the role of the Georgia Department of Corrections Fire and Life Safety inmate firefighter program and general prisoner assistance throughout the life cycle of Hurricane Irma. Results reveal the institutional structure of deploying prisoner labor in disasters, the types of assistance prisoners provide, and how officials manage the perceptions of prisoners as a hazardous population and a source of labor. 

The Survey of Trauma, Resilience, and Opportunity in Neighborhoods in the Gulf (STRONG) was designed to assess the long-term mental and behavioral effects of the Deepwater Horizon Oil Spill on residents in the Gulf Coast. It is a population-representative sample of 2,520 coastal residents surveyed April through August 2016, in Texas, Louisiana, Alabama, Mississippi, and Florida. Exposure to the spill, represented by resource loss attributed to the spill, was associated with positive screens for depression and anxiety. In the aftermath of the 2017 Gulf Coast hurricane season, with National Science Foundation RAPID funding, Texas residents were re-interviewed; this helped us procure private support for follow-up interviews with respondents in the other states. These re-contacts provide insights into how exposure to the hurricanes contributed to mental and behavioral health outcomes. Such data can help policymakers and other stakeholders develop targeted approaches to foster resilience, particularly among vulnerable populations, and thereby mitigate the effects of future disasters.

Harvey was a catastrophic tropical cyclone that made landfall as a Category 4 hurricane near Rockport, Texas. In days after the landfall, Harvey meandered around Southeast Texas while deluging the region, the heart of the U.S. energy and chemical industry, with rainfall in excess of 40 inches. The resulting flooding was unprecedented, leading to significant damage of critical infrastructure, including drinking water and wastewater systems, and the potential release of chemicals into surface water and stream sediment. The current presentation will discuss the results of our ongoing study to assess the environmental impacts of Harvey along the Neches River around Beaumont, Texas. The goal of this study is to identify high-risk chemicals; understand their distribution, fate, and transport in the floodwaters, flood plain, and bank soils; and evaluate their likely impacts on the source water for city water supply.

DesignSafe is the cyberinfrastructure (CI) platform supporting research in natural hazards engineering. Part of the National Science Foundation-funded Natural Hazards Engineering Research Infrastructure network, DesignSafe allows researchers to more effectively share, publish, and find data; perform numerical simulations using high performance computing; and integrate diverse datasets. DesignSafe embraces a cloud strategy, with all data, simulation, and analysis taking place on the server-side resources of the CI, accessible from the desktop. Three important DesignSafe components are the Data Depot, Discovery Workspace, and Reconnaissance Portal. The Data Depot is the central shared data repository supporting the full research lifecycle, from data creation to analysis to curation and publication. The Discovery Workspace is the place for researchers to perform simulations using the most sophisticated computational tools (including OpenSees), as well as to analyze, visualize, and transform their data using cloud-based tools such as MATLAB or Jupyter Notebook. The Reconnaissance Portal represents a unique interface to find data within the Data Depot (and other locations) that are associated with reconnaissance activities.  

Social media provides quick methods for disseminating information to the public and responding to questions from residents during natural disasters. Bidirectional communication between residents, public agencies, and nongovernmental organizations enhances the adaptive capacity of communities to prepare, respond, and support recovery, enhancing overall community resilience. We conducted a survey of over 300 organizations involved in emergency-assistance functions during Hurricanes Isaac in Louisiana and Sandy in the Northeast United States, both in 2012, and Hurricane Harvey in Texas in 2017, to determine the extent of social media use, key obstacles encountered, and adjustments made to improve social media use. We found evidence of a clear regional disparity in 2012, with greater use of social media in urban communities of the Northeast; generalized concern about the need to “fact check” information before disseminating to the public; and persistent technical staff shortages that inhibit full realization of social media’s utility in the context of disasters.

Sampling in Houston bayous and the Galveston Bay estuary proceeded immediately following Hurricane Harvey in an effort to understand the longitudinal response of the stream network to the more than 50 inches of rainfall. Different watersheds with varying degrees of urbanization and differing hydrologic and hydrodynamic characteristics were sampled multiple times over the course of several weeks post-Harvey. Results showed impacts from the hurricane on dissolved oxygen levels, pH, total suspended solids, total dissolved solids, and trace metals. Significant amounts of sediment were mobilized, transported, and deposited within the studied waterways; their relative impact on water and sediment quality post Harvey remains largely uncharacterized. Persistent organic pollutant concentrations post-Harvey indicated elevated concentrations of dioxins and polychlorinated biphenyls. While indicator bacteria were within historical norms, bacteria in sediment exhibited shifts in diversity and response to pollutants. Less urban watersheds recovered to historical norms within a few weeks after Harvey.

Hurricane Maria made landfall in Puerto Rico on Sept. 20, 2017. Its severity significantly impacted the infrastructure of the island, leaving residents without electricity and potable water service for weeks. Due to this, there were several concerns regarding water quality conditions. These conditions potentially exposed populations to higher levels of pathogens, disinfection byproducts, and other contaminants, which have the potential for widespread public health impacts. In response to Hurricane Maria, a National Science Foundation RAPID project was launched to perform a comprehensive assessment of the water quality in Puerto Rico. For this project, tap water samples from the Puerto Rico Aqueducts and Sewer Authority (PRASA) system, centralized water systems, small community systems (non-PRASA), and unmonitored surface streams were collected. In this presentation, we will share the results of the sampling, including the presence of pathogens, metals, and disinfection byproducts, as well as highlight the relationship between chemical and microbiological water quality.

Hurricane Harvey struck the Texas coastline on Aug. 25, 2017, as a Category 4 hurricane. Wind gusts over 130 mph and storm surge as high as 12.5 feet caused widespread damage to buildings and critical infrastructure in coastal communities, including Rockport and Port Aransas. Over a 12-day survey period, the authors assessed the performance of more than 1,000 individual, geo-located residential buildings. The goals of the study are to evaluate to what extent the wind speed to damage relationship in hurricanes is affected by additional hazards, such as storm surge, or unique phenomenon such as mesovortices within the hurricane eyewall, and whether the wind to damage relationship differs from that of other extreme wind hazards, specifically tornadoes. The influence of ground roughness, building age, structural characteristics, and presence of multiple hazards (i.e., wind and surge) are evaluated by means of the fragility functions.

Mobile homes are the single largest source of unsubsidized affordable housing in the United States.  Many mobile homes are located in land-lease communities called mobile home parks. Mobile home parks are frequently affected by disasters. Yet, there is little published research on their contribution to disaster risk. The landfall of Hurricane Harvey in the Houston metropolitan area shines a spotlight on this understudied issue. In this presentation, we report on a unique, large-scale study of mobile home park exposure during major disaster. First, we describe the role that mobile home parks play in the Houston housing system. Next, we detail the exposure of mobile home parks in the Houston area to floodwaters caused by Hurricane Harvey. Using a unique mixed methods research design, we find that over 400 mobile home parks were at least partially exposed. Third, we contextualize these findings by describing the population who lived in exposed parks, based on U.S. census data. Finally, we describe the future trajectory of our project, which will document the recovery pathways for exposed parks.

Despite federal policy directives to strengthen the resilience of critical infrastructure systems to extreme weather and other adverse events, several knowledge and governance barriers currently frustrate progress toward policy goals, namely: a lack of awareness of what constitutes resilience in diverse infrastructure applications, a lack of judgement about how to create resilience, a lack of incentives that motivate resilience creation, and obstacles that prevent action or reform—even where incentives exist—within existing governance systems. We describe each of these barriers in greater detail, provide a catalog of theories for overcoming them, and describe examples of catastrophe management (e.g., the near collapse of the Oroville Dam) through this lens of these theories. Additionally, we discuss the difficulty of education and training for resilient infrastructure systems and propose simulation gaming as an integrative research and education approach for capturing lessons learned from historical catastrophes, play-testing scenarios, sharing knowledge, and training a workforce prepared for the challenges of the post-industrial infrastructure age. 

Key informant interviews were conducted with hospital administrators in Southeast Texas about evacuation in Hurricane Harvey. To date, we have conducted eight interviews with 15 individuals representing 22 hospitals. Early results suggest the following: First, while many individuals did not have direct experience with hurricanes or with evacuating their hospital, the overall health system had significant experience that has been shared between hospitals. Second, most of the hospitals had just finished their annual hurricane exercise, which was very fresh in their minds as they made decisions about Hurricane Harvey. All of the interviews suggested that the hospital administrators were very familiar with their hospitals and the level of impact it could sustain without interruption of services. Finally, all of the hospitals stood up their operations centers early (seven days before landfall) and received regular updates from their jurisdiction and the Southeast Texas Regional Advisory Council. The question is: Will any of this translate to a jurisdiction that hasn't had the repeated exposure that Southeast Texas has had in the past 20 years?  

This presentation will offer preliminary results of a study, funded by a National Science Foundation RAPID award, that surveyed a representative probability sample of adults across the state of Florida (N=1,637) during the 60 hours before Hurricane Irma made landfall (Wave 1; 58% participation rate) and again four to six weeks later (Wave 2; N=1,478, ~90% retention rate). We examined significant predictors of pre-hurricane risk perceptions, sought to answer questions about who evacuates as a hurricane is approaching—and why—and sought to understand the extent to which media coverage of hurricanes plays a role in risk perceptions and emotional response to the disaster. The strongest predictor of evacuation was pre-landfall perceptions of risk; about a quarter of the sample reported evacuating their homes, and a quarter reported living in an evacuation zone but not evacuating. Data will illustrate how pre-hurricane factors predict post-hurricane responses and will demonstrate how vulnerable populations of individuals who are at risk for exposure to future hurricanes are coping with these recurrent stressors.

In the aftermath of major natural disasters, governments, aid agencies, and affected populations engage in practices of sense-making to gauge the extent and severity of the crisis, direct response activities, and coordinate recovery planning. To understand the conduct and implications of these practices, we examined the official damage assessment implemented by the Government of Nepal following the April 2015 earthquake. In addition, we undertook participatory mapping to examine the consequences of this assessment in the Langtang Valley, a severely-affected area of the country. We argue that the informatics of post-disaster damage assessment in Nepal played a primary role in narrating the events of the 2015 earthquake, legitimating particular paths toward recovery in the aftermath, and limiting opportunities for alternative configurations of social life that emerge during disasters. Our research demonstrates the ways that forms of sense-making afforded by information technologies play central roles in enacting repair work following crisis and breakdown.

Several aspects of the school context make children distinctively vulnerable during disaster. Schooling takes place in unique settings with respect to location, time of day, organization of activities, building construction, adult-to-child ratio, and need for student supervision. The unique physical, psychological, and cognitive characteristics of children during the developmental period contribute to the need for emergency planning. Children with disabilities, for example, may need more time to evacuate or to take shelter. Cases of school-related disasters highlighting how the school context intersected with natural hazards and subsequently affected schoolchildren were investigated. Purposeful maximal sampling was used to select school-related disaster cases for their geographical diversity, different instructional contexts, and types of initiating hazards. Among these, seven cases with broad considerations pertaining to safe school environments were chosen. A within-case analysis was conducted of each case, followed by a cross-case thematic analysis of preexisting causes of disproportionate risk.

In January 2018, the Center of Excellence on Risk-Based Community Resilience Planning, with collaborators from the National Institute of Standards and Technology, conducted an interdisciplinary, recovery-based field study in Lumberton, North Carolina. This field study was the second wave in what is to be a series of longitudinal field studies aimed at documenting the initial damage and recovery process of a diverse community impacted by a major flood event caused by Hurricane Matthew in October 2016. The residential and business sectors were the primary focus of the second field study. Structured surveys were carried out across a stratified sample of housing units and businesses, within and outside the inundation area, to document damage, repair progress, availability, and timing of recovery resources, and recovery progress using a series of sector-specific indicators. Preliminary findings indicate that one year after the flooding, the community is only in the early stages of recovery.

In 2015, the Center for Risk-Based Community Resilience Planning, a National Institute of Standards and Technology funded Center of Excellence (NIST-CoE) at Colorado State University brought together an interdisciplinary team of researchers, programmers, NIST collaborators, and graduate students to begin development of a robust computational environment for community resilience planning and post-disaster recovery strategies that include physical, social, and economic systems. The NIST-CoE team is made up of more than 100 individuals across 15 universities, spanning the disciplines of engineering, social sciences, and economics. In October 2016, the NIST-CoE conducted its first rapid reconnaissance field study to investigate the damage caused by Hurricane Matthew in Lumberton, North Carolina. In 2018, the team returned for a follow-up study to document recovery over the long term. This study explores the interconnectivity between structural damage, school closures, housing dislocation, and business recovery. We will discuss challenges and opportunities that arose during the planning and execution of this study.

Existing social vulnerability metrics rely on aggregate, place-based information, often masking how individuals (i.e., people, households) are themselves vulnerable to a hazard. This paper explores allocating census microdata to small areas to simultaneously account for individual and place-level vulnerabilities. We examine a case study of New York City's Rockaway Peninsula during Hurricane Sandy in 2012. Household-level responses from the American Community Survey 2012 Public Use Microdata Sample are allocated to block groups, and then used to identify clusters of places whose households are similar based on combined demographic, socioeconomic, and housing stock characteristics. We compare place clusters to Federal Emergency Management Agency Individual Assistance inspection rates post-Sandy, controlling for physical exposure and household application rates. Preliminary results show decreased inspection rates in areas featuring largely renter occupants of mid/high-rise housing with reduced socioeconomic status. These findings illustrate how individual factors aid clearer identification of place-specific vulnerability drivers.

Although scholars have argued that businesses benefit from communities adopting mitigation measures, insufficient attention has been paid towards examining the extent to which community-level mitigation activities impact business continuity and disaster recovery efforts. The present study addresses this gap in research by examining business continuity and disaster recovery efforts following Hurricane Irma in relation to whether the business is located in a county that participates in the Federal Emergency Management Agency's Community Rating System (CRS) program. The CRS program is a voluntary program designed to incentivize communities to implement floodplain management activities that surpass the minimum required under the National Flood Insurance Program. Using data gathered from 25 semi-structured interviews with business owners (n=19) and local government staff (n=6), results indicate that businesses located in higher CRS-participating communities sustained less impact and recovered faster than businesses located in lower CRS-participating communities. Findings also indicate that business continuity and recovery from Hurricane Irma were hampered by dependence on other businesses and customers, stress and emotional reactions, financial considerations, personnel issues, contractor responsiveness, challenges with insurance companies, and supply chain issues.

The Natural Hazards Engineering Research Infrastructure Natural Hazard Reconnaissance Facility (or “RAPID”), headquartered at the University of Washington (UW), is a collaboration between UW, Oregon State University, Virginia Tech, and the University of Florida. The RAPID facility provides investigators with the hardware, software, and support services needed to collect, process, and assess perishable interdisciplinary data following extreme natural hazard events. The facility commenced operations in September 2016 and will begin field deployments in September 2018. This presentation will provide an overview of the vision for the RAPID facility, the equipment that will be available for use, the facility’s operations, and opportunities for user training and facility use.

Our research examined business recovery in Lumberton, North Carolina, fifteen months after Hurricane Matthew. Done as part of the larger interdisciplinary effort through the Center of Risk-Based Community Resilience Planning with partner researchers from the National Institute of Standards and Technology, the data collection was a collaboration between social scientists, economists, and engineers. The study looked at several indicators of business recovery and their relationship to the built environment, including reported physical damage to the business (e.g., the structure and its contents), damage to households (e.g., worker and customer loss), and effects of utility disruption. We sampled both inside and outside of the inundation area to understand the full range of business impacts. Preliminary findings indicate that just under a third of businesses in our observed sample were still closed at the time of the study, with mixed indicators across recovery categories for those that have reopened.

Rural and agriculturally-dominated regions of the United States are distinctly different from more urban and suburban areas in terms of population density, land cover, and infrastructure systems. While all of these characteristics contribute to the impact and recovery of a rural community due to a natural hazard, this research focuses on the response of the unique structural and infrastructure systems found in agriculturally-dominated rural areas through a case study in Fremont, Nebraska, following a windstorm in June 2017. Within one week of the event, a team was deployed to assess structural and nonstructural damage in the Fremont area using both handheld digital cameras and an unmanned aerial system with an onboard camera for documentation. Over a 12-square mile area, 18 center-pivot irrigation systems were observed to be at least partially collapsed, with multiple damage patterns and collapse directions, indicating a complex wind field. In contrast, only one instance of structural damage was observed over the same 12-square mile area, highlighting the unique vulnerabilities of rural areas in the face of windstorms.

Forensic studies of structures damaged by high severe windstorms give valuable insights into wind speeds and effective construction techniques and help engineers design structures to better protect life and property. Structural damage serves as a proxy method for estimating otherwise unknown tornado speeds whenever the resistance of affected structures can be estimated. If wind speeds are reasonably well known (as with most U.S. landfalling hurricanes), correlation of damage levels and winds speeds facilitates the creation or validation of insurance risk models. Forensic analysis requires significant time and access to critical damage evidence, neither of which may be available to field researchers. Remote-sensing technologies (e.g., laser scanning, unmanned aircraft systems, and structure-from-motion) enable the preservation of three-dimensional damage scenes for subsequent detailed forensic analysis. This presentation highlights the use of such technologies for recent National Science Foundation RAPID investigations of structural damage from the 2015 Pampa, Texas, tornado and Hurricane Harvey in 2017 along the Texas coast.

Hurricanes are among the most severe natural disasters along the Gulf Coast. They have become more frequent in recent years, and their destructiveness has gradually increased. Taking Hurricane Harvey as an example, multiple incidents caused by the hurricane could potentially impact environmental quality, such as river backwash, Superfund site flooding, and chemical explosions and spills. In turn, hazardous chemicals and toxic contaminants were accidentally released into the environment, impacting the environmental quality through ozone pollution, surface water pollution, and other types of pollution, which elevated the risks for public health. The scale of Harvey far exceeded the capacity of common environmental quality management practices, especially for water treatment facilities. This paper will present the environmental impacts associated with hurricanes in a region heavily concentrated with oil refining and chemical processing industries, and discuss strategies to enhance regional resilience from an environmental quality management perspective.