The VORTEX-SE Community Forum is intended for anyone involved in the mitigation of tornado risk in the Southeast U.S. We use a Slack workspace where we can openly discuss any issues related to this risk, VORTEX-SE news and announcements, weather events, etc. This web site is used as a place to store information in more detail than can be readily communicated on Slack. It should be a good place to go just to browse and understand the variety of research that is being conducted by VORTEX-SE.
What sorts of things will you find here? Under the Grants tab, organized by year, you will find listings of all the grants made by VORTEX-SE, with their titles, investigators, organizations, objectives, publications, and a recent Executive Summary. As we build the content of this site, we will try to add informal publications and posters with links from these individual grant listings.
Using the site’s search feature, you should be able to find information about topics, events, and researchers.
In 2020, we will be creating a “Case Encyclopedia” under the Data tab. This will be a place where you can browse all of the events studied by VORTEX-SE, and find out where the data can be accessed. We are hoping that some of these cases will be sufficiently interesting that you will dig into them a little! Also under the Data tab, you can access the relevant NOAA Data Sharing policies.
VORTEX-SE grant results are summarized by topic under the Findings tab of this website. As always, users of the Community Forum should be careful to note which findings are final, typically being conveyed in formal publications, as opposed to preliminary results. Users should be respectful of the work being done by other researchers, being careful to give credit and form collaborations whenever appropriate.
Hello out there! My name is Tracie Sempier and I am new to the VORTEX-SE Community Forum, but I am not new to the impact of tornados. In fact, I can remember in 2000 when the F4 tornado hit my hometown of Tuscaloosa and left many homes damaged and 11 people dead. Since then, I have moved to coastal Mississippi where the majority of my work has focused on coastal storms. Then again in 2011, I can remember driving up to Tuscaloosa to help pass out supplies in the parking lot of my childhood church that was mostly destroyed. We had a community Thanksgiving meal out in the parking lot under a large tent since the building could not be occupied. Fast forward to 2020 where I now have the opportunity to take what I have learned about communicating risk and apply it to coordinating a new outreach program. The VORTEX-SE Community Outreach Program will help vulnerable communities in the Southeast prepare for, respond to, and recover from the impacts of tornados. As part of this effort, we will be creating a model for a regional extension program that will synthesize findings from VORTEX-SE researchers and develop ways to inform application of the research at the local level.
If you are reading this blog post, I am sure you are aware of the amazing work being conducted by the VORTEX-SE team of researchers (physical and social scientists as well as broadcast meteorologists). In my new role, I will be looking for practical ways a community can increase their resilience to severe weather that is based on sound science and can be implemented at the local level. I have many years of experience as an extension specialist, trying to assess the needs of those I serve, which will come in handy as I embark on this new assignment. The goal of this outreach program will be to better understand the vulnerabilities of those living in tornado prone areas, use this information to inform VORTEX priorities and research questions, and work together to co-produce solutions to keep people safe. Connecting people in their neighborhoods to resources, providing trainings on tornado safety, and partnering with organizations that already have a strong presence in the community are a few of the ways we hope to build capacity for communities to respond.
As you can imagine, communicating risk involves a great deal of trust and credibility between our team and those we serve. We want to involve members of the public who may have never been involved before, but who can be key drivers of change in their neighborhoods and help build social networks that will last long after any formal program ends. Therefore, we will also focus on informal networks and recruit thought leaders within communities that can help us understand the gaps that exist, but also share their local knowledge about how to reach vulnerable populations in their area. The goal is to increase tornado and severe weather knowledge but also help residents develop and practice skills that can be used in the event of an inclement weather event.
I am excited about the start of this program and what it will mean for the health and safety of all those living in the Southeast and how they approach preparedness in the future. Well-equipped and knowledgeable community members are the best resource a neighborhood can hope for when severe weather hits. Fostering a sense of purpose and ability to take action is a key component of increasing community resilience. Let’s get started! If you or your organization is interested in learning more or would like to be part of this grassroots effort, please contact me at email@example.com.
(Ed. note: this posting originally appeared in the NOAA VLab version of the Community Forum on 12 Juy 2019)
Laura Myers, PhD
The Center for Advanced Public Safety
The University of Alabama Tuscaloosa
In my Vortex-SE study, “Collaborative Research: Understanding How Uncertainty in Severe Weather Information Affects Decisions,” I studied residents of Alabama to determine their local knowledge and awareness of weather vulnerabilities, their connection to the weather enterprise network, and their perceived strengths and gaps in weather warning communication. I used a longitudinal cohort approach to understand the changes in public perceptions over 18 months with the ability to classify households by critical variables, such as household type and household characteristics.
A major finding from the public perception analyses was the public’s interpretation of probabilities and uncertainty. In the days and hours leading up to severe weather events, the weather enterprise partners convey the probabilities and the uncertainties associated with the event. Understanding at what point in this process the public began to have confidence in the information and thereby began to take the situation seriously was enlightening. The five-day outlook provided by the Storm Prediction Center (SPC) is not taken seriously until about the two-day mark. The public becomes more aware and tuned into their trusted sources from that point forward.
At that two-day point in the process, there are certain elements of the messaging that make a difference to the public and that draw attention to the forecast information. The public notes how they detect a change in the seriousness of the message at that point and they listen for that seriousness. They also desire an explanation of uncertainty to assist with their decision making. They do not want to interpret that uncertainty themselves. They have an element of trust in the messenger which increases their confidence in the message even when the message does not verify. Analyses that took place after the April 2011 tornadoes in Dixie Alley, revealed significant frustration over false alarms in Dixie Alley, but as weather enterprise partners began being more explanatory and dealt with uncertainty head on, that frustration has diminished.
The weather enterprise partners in Dixie Alley began changing their messaging after 2011 to include more of the elements that made a difference to the public and caused the public to become more aware and to take action. The weather enterprise partners, including broadcast meteorologists, NWS personnel, and emergency managers, used more visual images of previous events and real-time current images to capture attention. They improved their criteria for warnings and reduced false alarms. Over time, my research revealed that the public was becoming more knowledgeable about the warning process. A dialogue was happening between the weather enterprise and the public, in which the public was educated about the need for multiple warning modalities and having emergency plans to respond to the warnings.
As more warning modalities were suggested to the public, analyses of the usage and functionality of the modalities revealed the messaging content that was most useful to the public. Warning modalities function differently by context. WEA alerts, alert notifications, and mobile apps are the modalities used mostly by the public. Sirens and NOAA Weather Radio are also important to the public, depending on their needs for weather awareness. Television, visual imagery and their trusted sources at the NWS and local television stations are very important to the public. Radar imagery is used by the public no matter if they understand it or not. There is a need to educate them on the proper interpretation of radar.
The public indicates that particular content from their warning modalities is most useful in making their protective action decisions. The public wants information about location and timing. Location is critical because people do not want to change their behavior unless required. Timing is also a critical issue for the public because they want to know when they should prepare to take action. This research has indicated that location and timing are probably two of the most critical elements in the messaging process. The public also wants to know about potential impacts and how those impacts will compare to past events. The public tends to use previous events as a reference point to understand the severity and seriousness of impacts. These elements work best for the public when there is certainty, good timing information, and targeted location information. Many events, however, are not so certain, which makes it harder to provide useful information to the public. For example, the complexity of severe thunderstorms and severe thunderstorm warnings needs to be explained to the public. They need to know that severe thunderstorms can create straight-line winds that can do as much damage as a tornado.
Calls to action are also important in warning messaging. Direct calls to action provide guidance when the time is right. The public does not want to have to figure out what to do when the messaging emerges. They put a significant amount of trust in their weather authorities. Broadcast meteorologists and other weather professionals provide preparedness information in advance of severe weather, indicating what the public should do when the time comes. This preparedness information is often provided through television broadcasts, website content, and especially social media content. Relevant calls to action are included in that content and when the actual messaging is disseminated those calls to action are found in the message. The public knows what they should do and why they should do it. They do not have to think about it at the time of the crisis. Just act! Follow your plan!
A significant element missing in messaging appears to be the “all clear” indicator. The public perceives there is minimal information provided regarding when the danger has passed. They may come out of their shelters too soon or they may stay too long in their shelters and become agitated because they do not know when they will be safe. They are likely not to take shelter in the next event.
However, people mostly feel they have been warned properly. They appear to understand probabilities and uncertainties with the weather warning process. When asked about false alarms, they are not unhappy with that outcome. They understand that meteorology and forecasting is not a perfect science and they know that their weather enterprise has done everything they can to reduce the number of false alarms. There is a small percentage of the public who do complain when they think weather professionals “got it wrong.” But for the most part, the public feels they have been prepared for weather events and that overall, the messaging process helps them decide what they should do.
A significant issue with the warning process is reaching vulnerable populations. Every community has people who are somewhat disconnected from communications and who also believe they have no way to make themselves safe from weather events. It is essential that connected and disconnected populations be identified and taken into account in every community. The weather enterprise partners work very hard to enhance communication for both populations in their communities. It is much easier to enhance communication for the connected populations through various technologies and communication techniques than it is to reach disconnected populations. Cell phones and mobile applications have become very common warning modalities, but some people do not have nor use these technologies effectively. And even when they do possess these technologies, they do not know how to use the information or in some cases they do not know what the information means to them. For example, people living in mobile homes have reported that they received the WEA alert on their phone indicating they should take shelter. These people report that they did not understand that the WEA alert was meant for them at their location, so they did not take action. They also indicate that they did not have a plan to shelter which would involve having to leave their mobile home for a stronger shelter.
These results have been shared with local emergency managers and National Weather Service personnel who want to find better ways to reach these disconnected populations. This research suggests that the weather enterprise should create profiles of the disconnected to develop plans for reaching them and assisting them with responding to warnings. People with mobility issues, language differences, and communication difficulties, such as deaf and hard of hearing populations, have to be reached in unique ways.
One solution that has emerged in some communities is the use of nodal networks to enhance weather information communication. Understanding how to better reach disconnected populations involves determining how these populations currently communicate within their neighborhoods. Where do they get their information? Do they get it from friends and family, from church members, radio, word-of-mouth? Are there language challenges in these communication networks?
Once there is an understanding of current communication networks, the weather enterprise partners can then determine how best to disseminate information to the major nodes in these networks for further dissemination within their networks. This takes a lot of effort to accomplish, but many of the weather enterprise partners in Dixie Alley are developing efforts to do so.
Enhancing communication with disconnected populations is critical. However, it is essential that all populations are educated about their sheltering options. When interviewed, many people indicate that they had no emergency plan and no sheltering options. It is important to share with them that they do have sheltering options and that they can use an emergency plan that works for them. They often indicate that they do not have a personal home shelter and they cannot transport themselves to a community shelter. When they are told that there are options within their home where they can survive, they are surprised. It is important to help them understand what their best options might be and help them identify a plan that works with their particular situation. If they cannot transport themselves to a better shelter, then they need a plan in place to have someone assist them with their transportation needs.
The methods of communication used to reach all populations have been enhanced in recent years. Television and radio have traditionally been the primary methods of communication. Now, social media, alert notification systems, cell phones, and other technologies are being used to convey warning information. So, how useful are these new methods and how useful is television today? Broadcast meteorologists have enhanced their efforts with social media, sophisticated graphics and analytics, as well as community outreach. Many of the television stations in Dixie Alley go to wall-to-wall coverage as an event is approaching.
The public indicates that it is typically the other modalities that drive them back to the television. They receive information through an alert notification, siren or a mobile app which causes them to turn to their television to see how serious the event is and to see whether it will affect them. A critical issue with all of these modalities, as well as the television, is to provide location, timing, and impact information. Words, images, graphics, numbers, and other forms of communication are used to quickly convey location, timing, and impacts. The colors used in the graphics can be inconsistent across modalities and even across television stations which can create confusion for the public. The colors chosen to represent seriousness, severity, location, and even timing can be confusing because people interpret the colors to have meaning. While numerous colors are used, the public indicates that just a few colors mean something to them. The color “red” conveys the most seriousness, followed by orange and yellow. Other colors are much less meaningful and generate different interpretations, so color choice is extremely critical and requires more research attention to help inform this process.
Conveying location is often accomplished with mapping graphics. Showing a map, typically a county map, briefly with color graphics or other overlaid images is intended to help people identify the impacts approaching their location. This research reveals that the public has a difficult time determining their location on these maps. It is not that they do not know where they live, but rather they are trying to figure out where their location is on a brief graphic. If the public is educated pre-event about the mapping graphics that will be used and they are asked to find their location pre-event, then the public finds these maps more useful.
After the tornadoes in 2011 in Dixie Alley, many of these new modalities were adopted and efforts were made to enhance all forms of communication. The weather enterprise partners educated the public about using these modalities and also taught them how to understand the information being provided. This has created a two-way communication process between the public and the partners, both pre-event and during the event, in which the public will ask the partners for individualized answers to their questions. Such questions include “Will this by like April 27th, 2011?” or “What about my location?” The partners attempt to answer these individual questions, as well as answer those questions in such a way as to answer all the questions that the public may have. This has led to a significant pre-event burden for the partners. In the days and hours ahead of an event, in a time period where the partners are gathering data and communicating with each other to prepare, they are now responding to queries to the point of exhaustion. They find themselves worn out and exhausted by the time of the actual event, which may span a significant amount of time. The partners worry that they are not at their fullest potential when they need to be, as the event unfolds. The partners are using these research results to make their messaging more consistent and to provide the information most needed by their publics with the goal of reducing the need for so many individualized responses. These research results have been provided to these partners to help them improve their pre-event communication in order to reduce this burden. Knowing that location, timing, and impacts are important, as well as providing the explanations of uncertainty, has helped reduce this burden significantly.
The results of this research add to the social science weather research discipline through the examination of forecasters, broadcasters, emergency managers, and publics before, during, and after severe weather events. This research focused on integrating these partner communication networks to reveal the often unexpected and underutilized special “local knowledge” that these partner networks already have at hand – special knowledge about local weather and seasonal climate patterns; knowledge about particular populations, especially vulnerable populations; knowledge about local understandings of and vulnerabilities to severe weather – that can be used to improve the flow of information within these networks and the efficacy of decisions made to protect life and property.
The results of this research include social science teaching and training of weather enterprise personnel on these results for use in their messaging, forecasting, and dissemination activities and are being used to transform the functionality of products and services designed to warn the public about weather. The types of information needed by the public to take appropriate action in severe weather events are being integrated into physical, institutional, and information resources.
This Science Assessment involved hundreds of hours of work undertaken by the VORTEX-SE Steering Committee co-chairs, ex-officio members, and other committee members and contributors (huge THANK YOU!). It is a summary of all of the findings of VORTEX-SE to date, as well as a description of a large body of work in-progress as of summer 2020. Further, the Science Assessment discusses the future research needs of the project.
The Science Assessment is the best one-stop source of information about VORTEX-SE, and is essential reading for those interested in developing new collaborations and proposals for funding. Anyone interested in VORTEX-SE directions or findings should join in the conversation on the VORTEX-SE Slack workspace.
On this web site, the science assessments can be found under the “Findings” menu, or simply follow the link at the top of this posting.
Tony Lyza has added information about the recently-begun grants awarded through the early-2020 NOFO funding competition. This new page can be found under the Grants menu. As progress is reported for these grants, the listings will be updated to reflect the progress. Investigators are encouraged to let Erik Rasmussen, Tony Lyza (physical sciences), or Justin Sharpe (social and behavioral sciences) know if they would like their listings modified.
This blog describes research carried out recently in OU-CIMMS’ Behavioural Insights Unit (BIU) in an attempt to draw together the wide-ranging literature related to tornado epidemiology. The literature included Centers for Disease Control (CDC) reports, grant reports and NOAA Service Assessments alongside academic journal articles for the first time.
It was recognised that there are gaps in our collective knowledge, due to gaps in records or a series of methodologies that don’t necessarily talk to each other. Undertaking a meta-analysis to better understand the specificities of tornado fatalities, can provide a baseline for understanding the direct causes as much as possible, as well as examining previously overlooked literature was proposed as a method for identifying such gaps.
Meta-analyses often look at large data sets, examining literature from the medical sciences field, calling for a quantitative analytical approach. Such an approach makes an examination of study selection, summarizes measures such as risk ratio, difference in means, among others. However, while undertaking the initial literature searches it became clear that this approach was not suitable here and a new approach might need to be taken. For instance, some studies (e.g. Brooks and Doswell, 2002 and Hammer and Schmidlin 2002), take historical approaches, whereas CDC reports take purely epidemiological approaches, with some contextual information offered in places. Consequently, this variability of included literature lends itself to a more open method of synthesis, using meta-ethnography as an approach.
Meta-ethnography as a method allowed for key themes to rise to the surface. Some of these themes are well known, such as vulnerability of mobile home residents, but the layers of complexity underpinning their vulnerability is also explored by linking this to other so-called ‘vectors of vulnerability’ (defined and discussed below), as a way of showing the levels of complexity that exist when attempting to reduce tornado impacts on communities.
A key finding is that complexity exhibited by these vectors of vulnerability is key to understanding drivers of risk, and how these impact on survivability. There is not one simple solution to reducing risk of fatality because there are multiple vulnerabilities which when coupled with the magnitude of the hazard (provided by the EF scale) all underpin varying degrees of risk that likely impact on survivability of tornadoes.
The key themes that were discovered and analysed are discussed below. The full report is available here or by clicking on the image below (Note this will open in a new browser window).
Theme One: Contextual Epidemiology
The theme of contextual epidemiology brought insights into where victims were, what protective action they took, how effective it was/was not and if they received a tornado warning in enough time to react appropriately. Contextual information across the literature included in this study, included details on how much of the housing structure was destroyed and whether this was seen to be a vector of vulnerability in terms of driving higher fatality numbers.
Other contextual information related to warning lead times, as well as whether victims were able to take sheltering action, what this was and its effectiveness. The two emerging themes from this data, ‘housing structure’ and ‘warnings’ are both complex vectors of vulnerability.
Of note, CDC reports or NOAA Service Assessments are only carried out for large scale or high fatality events. This means that potentially there is data not being recorded adequately or with enough context to make accurate analysis of risk, vulnerability and specific causes of fatalities. Epidemiology studies and investigations that were undertaken, are able to be included in analyses such as this, allowing contextual epidemiological information that contributes to a broader pattern of understanding of the multiple risks that are likely to derived from vectors of vulnerability, described below.
Theme Two: Vector of Vulnerability
Analysis of literature highlighted the multiple vulnerabilities that drive increased risk of loss of life or injury. These do not exist in isolation but conspire together to weaken survivability outcomes. In order to understand these and frame them in a manner that acknowledges their complexity, a new terminology was needed. Consequently, vector of vulnerability is defined as a vulnerability resulting from a specific vector that improves or worsens the impact for individuals in the path of a tornado. In the same way that a mosquito is a vector for diseases such as malaria, dengue fever and others, by dint of the way it carries and spreads these diseases, a tornadic vector of vulnerability is one that has the capacity to carry and spread increased vulnerability, with each of vector having their own weight in terms of strengthening or weakening survivability of those in the path of a tornado.
This weighting is likely to be reflected by the way each vector increases fatality rates or enhances survivability. There is also complexity within each vector of vulnerability as they are also impacted by access to finances, social networks and education and learning opportunities. Although many different vectors of vulnerability were identified, there were certain vectors that were seen as having the greatest impact on driving mortality:
Location in a house or building (floor, near a window etc.,);
Warnings given that may have been adequate, inadequate or not received;
Access to a basement in permanent homes;
Access to a shelter.
Figure 1, illustrates how the threads currently connect as vectors of vulnerability. If severed, tornado fatalities and their impacts might be reduced by tackling vectors with the highest impact. Returning to our mosquito and malaria example, this would include having mosquito nets (with and without DEET, therefore showing further layers of complexity, cost and risk assessment by an individual), as well as simpler and finance free measures such as removing standing water where mosquito larvae are laid. The questions we need to ask and provide reasonable solutions for are:
What are the most effective ways of reducing vectors of vulnerability?
Which vectors have the highest weight, cost and practical applications which might encourage or hinder actions to counteract them?
Figure 1:A MaxQDA generated code relations map showing the interaction of codes for Vectors of Vulnerability
High frequency vectors of vulnerability such as mobile homes, are also made up of a variety of sub-vectors that include, mobile homes’ construction, lack of stability when not tied to foundations, age, lack of maintenance and over-confidence in their ability to withstand tornadic forces.
There is also evidence that residents of mobile homes are also aware of the appropriate action to take when a tornado occurs but don’t have access to shelter. Nevertheless, even following the correct advice can have tragic consequence when tornadoes directly hit and have high wind speeds.
The literature indicated that people are acutely aware of the risks, how these impacts on their vulnerabilities and prepare appropriately within the bounds of their resources (such as access to shelter, transport, alternative measures, hazard awareness and safety actions). This tends to refute arguments made regarding ‘complacency’ among residents towards tornado preparation and actions (e.g. Biddle, 1994 and Ashley, 2007).
This leaves the thorny issue of how to tackle some of the gaps that still exist and make mobile home residents so vulnerable to tornadoes to the extent that fatality rates are up to 30 times that of a resident of a permanent home. It is likely that the answer is not to suppose, theorise or expostulate, but to directly connect with communities with large mobile home populations. The phrasing of this last sentence is chosen carefully. There aren’t really ‘communities’ of mobile home residents, especially in the SE United States, but there are communities and networks bound by common experience, religion and culture. Perhaps the answer is to have research focus led by asking these communities that happen to reside in mobile homes what they need.
Theme Three: Risk
Risk as a part of the overall story of tornado epidemiology was focused around risk perception, with a specific focus on tornado warnings. This is because the issue is inextricably linked currently to temporality. This temporality was deemed to be at a number of times scales:
Minutes in terms of tornado warnings given;
Hours in terms of tornado watches;
Diurnally in terms of day or night tornadoes, with many nocturnal tornadoes going ‘unwarned’ from the perspective of citizens who may be asleep when warnings are given;
Seasonally in terms of expectations of citizens and the idea of a tornado season.
There may be improvements in communication of risk (in terms of watches, warnings and how citizens react to them) that can be achieved, but there are also issues of how information is received, whether it is trusted and most importantly, the options that citizens have to act upon warned tornadoes. There are known vulnerabilities, and these are inextricably linked to the nature of the hazard (in terms of its magnitude) as part of the conceptual risk equation Risk= Hazard X Vulnerability (Burton et al, 1978 and Wisner et al, 2004).
The literature also points out that many citizens are aware of risk and take appropriate action such as going to an interior bathroom or closet but that the structural integrity of the mobile home coupled with poor anchorage and maintenance of tie-downs struts in some cases are drivers of risk and are separate vectors of vulnerability that enhance that risk.
Large loss of life in mobile homes and indeed in permanent homes is also linked to lack of access to shelters. It is likely that this gap might be closed by allowing communities to map their own risk (better understanding it) as well as evaluating their own capacities in terms of sheltering options that currently exist locally.
Theme Four: Capacity
Something that is missing in the literature is capacity of individuals. This concept needn’t be nebulous and intangible, but real and actionable behaviours. These might come from initiatives in which communities of practice (citizens, emergency managers, the weather enterprise and researchers) learn to work together to identify what capacities currently exist and building upon them to enhanced future capacities.
A key part of identifying capacity is knowledge about the survivability of tornadoes. One of the principal guiding aims of the research was to provide a baseline for survivability as well as acknowledging and understanding what drives fatalities. But baselines require a knowledge of exposure rates in tornado events so that there is an understanding of both fatality rates and survivability rates. This is an area of future research that is likely to reap significant benefits and perhaps needs to be a research priority.
Key to reducing individual, and by extension a community’s vulnerability is by enabling them to become more resilient to the shocks that a potential disaster may force upon them. But this requires learning rather than education or knowledge transfer that are often grasped for improving resilience. Consequently, socially constructed and transformative learning with the potential to build self-efficacy has an important bridging role to play in allowing communities to prepare for and adapt to such events, irrespective of wealth, status and level of formal education.
Transformative Learning (TL) approaches facilitate learners to question their assumptions, which then change as a result experience. Mezirow and Taylor suggested that it is teaching for change (Mezirow and Taylor, 2009), while Armitage et al. noted that Mezirow (1995) proposed that, “an outcome of transformative learning is the development of liberated, autonomous and socially responsible individuals with the capacity to move from critical examination of their experiences to action” (Armitage et al, 2007, p.88).
In particular, transformative learning, although challenging, has the potential to empower individuals to adapt to shocks and surprises that disasters may cause. As a result, individuals, families and groups may evolve into communities of practice with better coping mechanisms allowing them to view natural hazards as inevitable but disasters as not.
Currently capacity is not a widely expressed as underpinning survivability. In order to understand and enhance capacity there may be a need to augment our collective approach. This means moving away from from top-down and expert led approaches to bottom up and embedded ones, allowing for experts to contribute from across communities. By widening communities of practice, it is more likely that we will be able to identify and close value-action gaps between intentions and behaviour, building capacity to survive tornadoes as an outcome.
Justin Sharpe, PhD, Research ScientistatOU-CIMMand the social science co-ordinator for the VORTEX-SE project.
Armitage, D., Berkes, F., Doubleday, N. (Eds.), 2007. Adaptive Co- Management: Collaboration, Learning and Multi-Level Governance. University of British Columbia Press, Vancouver.