Hurricane Floyd Symposium
Permanent URI for this collection
Browse
Recent Submissions
Item Open Access Trauma Written in Plywood and Flesh: Hurricane Graffiti, Post-Katrina Tattoos, and the Value of Narratives to Hazards Research(2009-09-18) Alderman, Derek H.; Ward, Heather; Gentry, GlennDr. Alderman explained that narratives are an important way to understand how people were impacted by a hurricane. Graffiti and tattoos are visually evocative narratives written on plywood and flesh as opposed to paper. Hurricane graffiti can serve as a practical tool or as an indicator of tensions and needs. Post-hurricane tattoos serve as memoirs, vehicles for retelling hurricane stories, and as a way to express trauma. A visual content analysis was used to identify specific themes evident in hurricane graffiti inscriptions. Some of the major themes included history, defiance, desperation, territoriality, humor, politics, and prayer. An example of graffiti as history is a piece of graffiti outside of a store that said, "1) Charles, 2) Frances, 3) Ivan, 4) For sale." An example of graffiti as defiance is, "Take a hike Ike." An example of graffiti as desperation is, "We need power!" An example of graffiti as territoriality is, "Looters will be killed." An example of graffiti as humor is, "Santa’s naughty list: Charles, Frances, Ivan." In the tattoo study, tattoo artists were interviewed. One artist, Tom, decided to get a tattoo of an X on his leg after Hurricane Katrina, which references the rescue symbol. This tattoo, like many tattoos, elicits stories from others, allows him to retell his story, and serves to provide a collective memory. Narratives are significant because they are personal and social and provide tremendous insight. For this reason, these types of narratives are important to natural hazards research.Item Open Access Preserving Assets in Low-Income Communities Affected by Disaster(2009-09-18) Rausch, Christina; Windley, TikiMDC helps organizations and communities close gaps that separate people from opportunities and helps people to try to find a path out of poverty through education, work, and asset-building. The group has been active in disaster preparedness for low-income communities, which are typically less prepared and suffer greater loss. Preparation is a way that people can build and preserve their assets in the face of disaster. Many low-income people own businesses and cannot recover after a disaster. When a disaster hits a small business, people become unemployed, which leads to less consumer spending, making it tough on the community because the tax base decreases. The barriers to resilience in low-income communities include lack of access to capital and affordable insurance. MDC took a community development approach to creating disaster preparedness in low-income communities. They identify barriers for socially vulnerable communities and came up with solutions . They recommended improvement to land-use planning in hazardous areas and sufficient buy-outs when necessary. There is a need to reduce the costs of mitigation and provide insurance-sponsored mitigation. The researchers suggested developing relationships with trusted community agents, more effective disaster plans, and expanding access to financial products for recovery. Public–private partnerships and microloans could be critical in recovery for low-income communities.Item Open Access Fatal Tradeoff? Toward a Better Understanding of the Costs of Not Evacuating from a Hurricane in Affected Landfall Counties(2009-09-18) Czajkowski, Jeffery; Kennedy, EmilyThe researchers examined an economic behavior model of the evacuation decision. Hurricane fatalities decreased each decade until the 2000s, when deaths increased due to Hurricane Katrina. As more people move into coastal region, we need a new look at the cost of evacuation compared to the value of waiting. For mitigation and minimization of fatalities, the researchers examined the creation of an Ike dike that costs $40 million and questioned its value. The researchers modeled fatalities as related to the direct and indirect factors of storm characteristics, socioeconomic factors, forecasting technology, evacuation, and time. Most research is focused on coastal counties, but deaths often occur inland. For Floyd, of the 70 fatalities, only three were direct coastal fatalities. The researchers compared deaths from different storms, 93% with zero fatalities. They found a lowering of fatalities over time and high levels of evacuations. As evacuation levels rise, there were less fatalities. If there was no evacuation for Hurricane Andrew, the model predicted ten additional deaths. Early storms cause more fatalities, and fewer fatalities occur among people with higher median incomes. As forecast error is worsens, expected fatalities decrease because more people are warned and are better prepared. Although this research focused on the coastal region, an important extension would be to look at inland communities and compare with rainfall data.Item Open Access A Methodology to Inject Sea-Level-Rise-Enhanced Storm Surge Modeling into the Long-Range Comprehensive Plans of Coastal Communities(2009-09-18) Fraziers, Tim G.; Yarnal, Brent; Wood, NathanThis project implemented a comprehensive vulnerability assessment framework that used mapping and stakeholder input to create long-range land-use plans that took into account sea-level rise. They compared regular storm surge maps with sea-level-rise-enhanced storm surge maps in Sarasota, Florida, which is in the county with the second oldest population in the nation. The researchers used the SLOSH model with sea-level rise to predict flood inundation. They found that the storm surge inundation area increased and doubled the number of vulnerable elderly people. When looking at mitigating and planning for the future, they found that a category five hurricane places the entire population of the county in the storm surge zone. Using tax parcel information, the researchers found that several communities would not be very resilient since most of the buildings are in the storm surge zone. The researchers separated focus groups into planners, engineers, environmental advocates and business managers, and assigned them the task of looking at the hazard and finding solutions. When the groups saw the storm surge maps, they concluded that location of development was not appropriate. Overall the groups felt that the current urban plan needed to be revised and that urban boundaries should be relaxed to encourage development in less hazardous zones. The community also needed to relocate infrastructure and develop alternative evacuation routes in response to the impacts expected from sea-level rise.Item Open Access Severe Weather-Related Risk and Emergency Communication in Coastal Communities(2009-09-18) Kain, Donna; Smith, CatherineThe researchers want to learn how people in eastern North Carolina receive and use information regarding hurricanes. The researchers conducted face-to-face interviews with residents, businesses, and local government officials in six eastern counties; phone surveys were used in the 20 CAMA counties. They found that residents in coastal counties were more likely to have evacuated in the past than residents of other CAMA counties. All residents seek information to know how to respond to emergencies. Almost half of people never have evacuated during a hurricane, but they do seek information on how to prepare and make that decision. Current research shows that more people are likely have a disaster plan and know the location of a nearby emergency shelter than in past studies. Coastal residents are more likely than others to know if their homes are covered by evacuation orders, and 61% of residents said they were more likely to evacuate if an evacuation order was issued. Residents get most of their information from television, but social networks and the Internet are becoming more important. Coastal residents rate the quality of information high for television and low for newspapers and social networks. In interviews and focus groups, the researchers heard skepticism about television in terms of sensationalizing a storm. Future work will identify opportunities for using technologies already in residents’ hands, such as cell phones and the Internet. They also are looking at watch and warning text messages, but they need to confront the problem that people don’t understand the difference between a watch and a warning and they don’t understand the cone of uncertainty in hurricane forecasting.Item Open Access Hurricanes and Homeowner Decision-Making(2009-09-18) Burrus, Robert T.; Dumas, Christopher F.; Graham Jr., J. EdwardThe researchers conducted surveys between 2001 and 2002 that examined homeowner decisions concerning wind damage to homes from hurricanes. They collected information about current mitigation practices, expectation of damage, and willingness to pay for future mitigation improvements. Their key findings were that income, size of the deductible, and expected damages from a category three hurricane—but not category four or five hurricanes—were the main determining factors for people that do mitigation. Homeowners generally dismissed low probability events. They found that homeowner willingness to pay for mitigation devices is negative and requires subsidy in excess of $14,000. Willingness to pay is greater if insurance is not available. They concluded that most homeowners are not willing to purchase excessive mitigation devices.Item Open Access Home-buyer Sentiment and Hurricane Landfalls(2009-09-18) Burrus, Robert T.; Graham Jr., J. Edward; Hall, WIlliam W.; Schuhmann, Peter W.The researchers looked at how hurricanes impact real estate markets and home-buyer sentiment. Sentiment is related to the perception of risk by investors in the securities markets, but is not quantifiable, so the researchers looked at developing proxies. They used three proxies to determine the most meaningful one, which included the spread between listing and selling prices, the average days of a house on the market, and the number of single-family houses sold per month. They looked at homeowner sentiment from 1995 to 2002 in the Cape Fear region and the impact of Hurricanes Fran, Bonnie, and Floyd on the market. When they looked at the prices and days on the market, they found that after Bonnie there was not a difference in sentiment. Then after Fran there was some difference. Then after Floyd, more difference. The proxy impacted most was the days a home was on the market. The researchers concluded that the market suffers after successive hurricane landfalls, but that sentiment recovers a year or more after the hurricane.Item Open Access Property Values and Flood Risk: What Happens to Premiums over Time?(2009-09-18) Bin, Okmyung; Landry, Craig E.After Hurricane Floyd, property values in Pitt County were reduced when the properties were determined to be located in the flood plain. Bin and Landry used GIS data to look at the difference between homes sold in the flood plain between 1996–2002 and 2003–2008. It was difficult to distinguish the effect of Hurricane Floyd from other changes at the time, so they made a comparison between properties within flood zones and properties outside flood zones. They found that homes in a flood zone had a lower value in the 1998–2002 period. The home values reflected the awareness of the risk of flooding. There was no significant effect of location in the flood plain in the 2002–2008 period, which indicated that flood risk premiums associated with lower flood risk diminish over time.Item Open Access Flood Insurance Coverage in Dare County: Before and After Floyd(2009-09-18) Landry, Craig E.Dr. Landry started with an overview of how the flood insurance program works. Flooding is a catastrophe risk since flooding events cause multiple losses that are correlated across space; and given the rarity of flooding events, reliable information may not be available to predict likelihood of loss. Government provision for disaster relief can cause a “charity hazard” in that people may opt not to insure. As such, private companies have traditionally exhibited little interest in providing insurance against flooding loss, and the government has stepped in with a public option: the National Flood Insurance Program (NFIP). The program started by creating flood inundation maps and offering subsidized insurance, but most people did not buy it. Then the government made flood insurance mandatory if you had a federally backed mortgage. This has increased participation in NFIP. This study compared 1998 and 2008 flood insurance in Dare County, NC, a vulnerable area. In 1998 many properties did not have flood insurance, but in 2008 more did. There are also more mortgaged properties than ten years ago, and these have higher assessed values and higher amounts covered. Demand for insurance is not responsive to price, so raising or lowering flood insurance prices is not likely to have a large effect on coverage. Since people are required to purchase insurance if they have a mortgage, they don’t have the option not to buy. Demand for flood insurance also increases with income and education level. Insurance coverage is greater for higher value buildings and riskier areas.Item Open Access Charting the Course(2009-09-18) Furgione, Laura K.Furgione reviewed NOAA’s activities at the time of Hurricane Floyd in comparison to today’s technology and integration. She reviewed the tracks and impacts of the 1999 series of storms—Hurricanes Dennis, Floyd, and Irene—and how important forecasting was to reducing the death toll. Since that time, the National Hurricane Center has gone from six to ten forecasters and has expanded its webpage to allow more users. In 2008, there was less storm track error and the forecast has been extended from three to five days. Wind speed probability and storm surge probability products are also new. Tornado forecasting lead time was ten minutes in 2001, and now it is 15 minutes. The hydrologic prediction service now has over 2,237 hydrographs, and the system is 56% complete. Future projects include advanced flood inundation mapping for North Carolina. Because of Floyd, North Carolina is ahead of other states in flood mapping. There is now a greater understanding of the ecological impacts of big storms and their run-off. NOAA is also working on information to help communities become more resilient and suffer less financial loss in the future. They have used the NOAA satellites to assess damage after Hurricane Isabel, and the Storm Ready program has also been developed. Finally, Furgione showed that climate change has not had a discernable effect on hurricane activity over the last 100 years. Global sea-level measurements, however, show that sea-level rise is occurring. Increased sea level is expected to increase storm surge impacts in the future.Item Open Access Prototyping a Hurricane-Flood-Landslide-Continuum Prediction System: A CI-FLOW Contribution to North Carolina and Broader Coastal Regions(2009-09-18) Hong, Yang; Gourley, J.J.; VergaraArrieta, Humberto J.; Liao, Zonghu; Van Cooten, Suzanne; Kelleher, KevinA partnership between NASA and the US Agency for International Development (USAID) is examining prototyping a hurricane-flood-landslide continuum as part of CI-FLOW (Coastal and Inland Flooding Observation and Warning project). Using high resolution satellites, hydrological data can be used to monitor global flooding. The Terra and Aqua satellites use a key instrument called MODIS (Moderate Resolution Imaging Spectroradiometer) to map flood inundation. Aster flood maps are used to calibrate hydrologic modeling. Additionally, inland river flow and storm surge modeling is being done as a part of CI-FLOW. Part of the problem with the models is determining how to calibrate them effectively. Researchers examined a full spectrum of data calibration techniques. The guided DREAM technique performed the best in terms of minimizing bias. They also examined Hurricane Floyd, and, using a discharge simulation, the guided DREAM preformed well. Landslides can accompany floods, and LIDAR (Light Detection and Ranging) data can be useful in examining slope stability and rainfall. In order to effectively manage hazards, an integrated approach is necessary, using hydrological data and flood prediction systems. Recently, the prediction system has been implemented successfully.Item Open Access CI-FLOW: Evaluating and Testing New Technologies for Accurate and Timely Identification of Inland and Coastal Floods in the Tar-Pamlico and Neuse River Basins of Coastal North Carolina(2009-09-18) Van Cooten, Suzanne; Thigpen, Jack; Bacon, RobertCI-FLOW is a new technology being utilized to identify flood hazards. CI-FLOW stands for the Coastal and Inland Flooding Observation and Warning project. CI-FLOW was implemented ten years ago by the directors of Sea Grant and NOAA, along with other North Carolina partners and state agencies. The National Weather Service has one forecast point in Louisburg, NC, on the Tar River. The addition of additional USGS gauges is important for effective identification of flood hazards. The accurate and timely identification of flood hazards is important given the growing and seasonally fluctuating population of Dare County. Approximately one-half of the housing in Dare County is seasonal, and emergency mangers must account for this. For instance, it is estimated that in 2030, if a category three hurricane hit during a time of peak tourism, emergency managers would need to clear 30,000 vehicles in 31 hours. Given the importance flood identification, an integrated approach is needed for accurate and timely identification. The utilization of storm surge models and observations of weather and river levels, is key in flood identification. Three-dimensional images linked with real-time data can help assess the dangers of a flood and aid in emergency management decisions. CI-FLOW is being used to time crests and discharges along with a coupled model which links water and storm surge discharge. In conclusion, the CI-FLOW project is a multi-agency evaluation of new technologies to better identify floods in the Tar-Pamlico and Neuse river basins.Item Open Access African Easterly Waves and Rainfall Variability in Niger during the 2006 AMMA Field Campaign(2009-09-18) Ferreira, Rosana Nieto; Rickenbach, ThomasDr. Ferreira showed that when Africa gets more rain, North Carolina gets more hurricanes. More than half of hurricanes, including Hurricanes Dennis and Floyd, form as African Easterly Waves (AEWs). The African Easterly Jet forms due to an interaction of warm air over the Sahara with cold air over the Gulf of Guinea, and the inter-tropical convergence zone causes convection. In this unique region, heating gets stronger and makes AEWs. Warmer air is north of the equator, not at the equator, so wind blows west. During the AMMA field campaign, NASA and MIT put radar off Africa to find out how these storms develop and if they can be used to predict Atlantic hurricanes. A total of 28 squall lines in 2006 produced 82% of all rainfall that year. AEWs propagate at two different latitudes depending on the jet. There is significant difference in the structure of systems further south and further north. More AEWs means more potential hurricanes, but factors like El Niño and wind shear will affect the production of hurricanes. Understanding the interactions between squall lines and AEWs may lead to better hurricane prediction.Item Open Access Material Transport in Coastal North Carolina following Hurricanes: A Remote-Sensing Perspective of Hurricane Floyd's Impact(2009-09-18) Miller, Richard L.; Riggs, Stanley R.; Buonassissi, Christopher J.A hydrograph of the Tar River depicts an unprecedented amount of rainfall during Hurricane Floyd. This excess rainfall transported carbon in the form of dissolved organic carbon or Colored Dissolved Organic Matter (CDOM). It is important to understand the transport of CDOM because it can photo-oxidize, yielding carbon dioxide and carbon monoxide, which has implications for global warming. Using remote sensing, the color photographs from satellites of the Tar and Neuse rivers were examined before and after Hurricane Floyd. Pre-Floyd, sediments that were stirred up by Denis had begun to settle and the Tar and Neuse rivers were brightly colored. Post-Floyd, on September 17, 1999, the Tar and Neuse rivers were dark with CDOM. Ultimately, researchers want to be able to quantify the amount of carbon released during a storm to better understand the potential climate change implications.Item Open Access The 1999 Flood of the Century: Extraordinary Hydrometeorological Event or Human-Induced Catastrophe?(2009-09-18) Lecce, Scott A.; Kotecki, Erica S.In 1999 the effects of Floyd, Dennis, and Irene caused unprecedented flooding, but was this a natural event or a human disaster? The researchers examined photographs of the effects of Hurricane Floyd and some other floods to compare flood stage. They also looked at drainage basins and census data to examine land use changes and how they have affected the drainage basins. Rainfall data at 21 stations were collected during Hurricane Floyd. Hurricane Dennis, which dumped 10–20 inches of rainfall in the Greenville area, saturated the soils, but did not cause major flooding. The three-week period including Hurricanes Dennis and Floyd was as much as 85% of annual rainfall. At these stations, there were 200- and 500-year recurrence intervals, which means this was an extreme event. Using parametric and non-parametric statistics they find a significant trend in annual mean discharge and flow, but not an association with human activities. There was no trend through time. In comparison to Mississippi River flooding, they found that floods are self-similar events and recurrence intervals are subject to a large amount of uncertainty. The largest events drown out human impacts since there is limited storage capacity of wetlands. During a period of rapid urbanization, there was little change in the stream flow at the scale of watersheds. Extreme precipitation plus the sequencing of storms means that there was not a difference due to human land use issues.Item Open Access Changes in Flood Characteristics after a Major Event: Re-evaluating the Effect of Hurricane Floyd on Future Flood Response(2009-09-18) Arrigo, JenniferHurricane Floyd had immediate impacts on the relatively slow-changing environment and undeveloped, unregulated watershed. The total flood water was 95% of the volume of Pamlico Sound, and many locations exceeded the record for rainfall in a 24-hour period. North Carolina flood mapping and hydrology projects were started after Floyd. The high water mark signs in the community reflect the social and policy legacy of the storm. When looking at long-term effects, the researchers did not see many changes in flow or channel characteristics based on stage–discharge relations and USGS rate curves. Data from three gauges were examined based on the NWS flood modeling program. At Rocky Mount, upstream of the Tar reservoir, flow is regulated. The inundation model is highly correlated with a well established relationship between stage and discharge. The timeline of Floyd follows the model very well; however, a moderate flood causes a different relationship between stage and discharge. The Tar River at Tarboro, using the same rating curve, showed change after Floyd. In Greenville, a partial duration series showed that, during extreme floods, the relationships are about the same, but there were changes at minor to moderate flood stage. Minor to moderate flood inundation maps might be different than the typical inundation maps currently used.Item Open Access Back to the Future: Satellite Precipitation as a Tool to Reanalyze Hurricane Floyd and Forecast Probabilities of Extreme Rainfall in Eastern North Carolina(2009-09-18) Curtis, ScottIn order to answer questions about how much rain fell during Hurricane Floyd and its input into rivers, researchers looked at satellite, river gauge, and radar data. They took a radar map and put it into a GIS with watershed maps and the other data. The location of rainfall was tracked through gauges, and much of it was interpolated. The extent of the storm totals were compared using these different ways measuring rainfall. For the Tar River, rainfall input measures were close to the gauge measures. Researchers used a calculation in places where there weren’t gauges or radar to estimate discharge. Greenville received 22.9 inches of rain during Floyd. Satellites measured rainfall over the ocean to get the full picture of rainfall during a storm. Extreme rainfall events are increasing due to climate change. The 1999 season was not the highest, but part of an upward trend. Prior to Dennis and Floyd, there was an extended pool of warm ocean water, and this led to increased rainfall. Rainfall rankings are a measure of relative climate change and show that Floyd was extreme in the global sense. Similar global rankings that include storms from the past may be helpful in predicting the future.Item Open Access Facing Disaster: Forecasting and Assessing Floyd and its Impacts on North Carolina(2009-09-18) Orrock, JeffJeff Orrock reviewed the year of Hurricane Floyd, 1999. It was not a big hurricane year overall in the Atlantic, but very active for North Carolina. Hurricane Dennis was a welcome storm in that it brought 5–8 inches of rainfall to break the drought. As Dennis moved out, Floyd was named. Just before Floyd’s landfall, there was precursor rainfall. Then Floyd’s outer bands moved onshore, and a cold front interacted with the hurricane. At that time, forecasters had just started to use hydrographs to measure river flooding. The river crests, which lasted for days, were above the record. The force of the water moved mobile homes and put them against trees. Precipitation forecasting was close to the actual rainfall. Forecasts are critical to predicting river crests since the first whole day of rainfall will help determine the amount of flooding for following days. The tools are much improved since 1999 when forecasters said flooding was going to be worse than Fran, but they didn’t know by how much. Radar has improved for rainfall estimates. Now rainfall is digitized and can be mapped in GIS to see exactly where it is going. Hurricane forecasting has also been improved so that forecast are 3–4 days before landfall, enough time to evacuate people and prepare shelters. Forecasters need to further advance understanding so that we can predict and communicate better.Item Open Access Welcoming Remarks (Day 2)(2009-09-18) Curtis, Scott; White, Dean Alan; Mageean, Deirdre; Landry, Craig E.Hurricane Floyd SymposiumItem Open Access Summary Report: Hurricane Floyd Symposium(2009-09-18) Covi, Michelle P.; Huffer, Hillary; Anderson, Benjamin; Siepert, S. DanielOn the morning of Thursday, September 16, 1999, Hurricane Floyd made landfall at the mouth of the Cape Fear River. Rains associated with Hurricanes Dennis, Floyd, and Irene resulted in extensive flooding over a two-month period in eastern North Carolina, with most river basins exceeding the 500-year flood level. Flooding was worst along the Tar River, leaving the towns of Rocky Mount, Tarboro, Princeville, Greenville, and Washington devastated. In North Carolina, Floyd was directly responsible for 35 fatalities and several billions of dollars in property damages. This publication provides a summary of the presentations in the public forum and the research conference. The value of an endeavor like the symposium comes from how the information shapes the way we move forward. Therefore, this publication also serves as a benchmark so that we can measure our progress on hurricane risk mitigation.