A hurricane evacuation management decision support system (EMDSS)

This article describes the challenges confronting local authorities who must decide if and when to initiate evacuations from tropical cyclones. This problem can be decomposed into the behavior of the hurricane that is relevant to evacuation and the behavior of evacuees that is relevant to the hurricane. The uncertain behavior of these two systems can be modeled in an evacuation management decision support system (EMDSS). The hurricane EMDSS described here displays information about the minimum, most, and maximum probable evacuation time estimates (ETEs) in comparison to the earliest, most, and latest probable estimated times of arrival (ETAs) for storm conditions. In addition, EMDSS calculates the cost of false positive (the economic cost of an evacuation) and false negative (lives lost in a late evacuation) decision errors. EMDSS is being used in experiments to assess different information displays, team compositions, community characteristics, and hurricane scenarios. In addition, it will be used in training and actual hurricane operations. Finally, definition of the program’s requirements has identified further research needed to build a better empirical base for its input data.     

Agent-Based Modeling and Analysis of Hurricane Evacuation Procedures for the Florida Keys

The unique geography of the Florida Keys presents both high risk of hurricane landfall and exceptional vulnerability to the effects of a hurricane strike. Inadequate hurricane shelters in the Keys make evacuation the only option for most residents, but the sole access road can become impassable well in advance of a major storm. These extraordinary conditions create challenges for emergency managers who must ensure that appropriate emergency plans are in place and to ensure that an orderly exodus can occur without stranding large numbers of people along an evacuation route with inadequate shelter capacity. This study attempts to answer two questions: (1) What is the minimum clearance time needed to evacuate all residents participating in an evacuation of the Florida Keys in advance of a major hurricane for 92,596 people – a population size calculated based on the 2000 US Census population data, census undercounts, and the number of tourists estimated to be in the area? (2) If a hurricane makes landfall in the Keys while the evacuation is in progress, how many residents will need to be accommodated if the evacuation route becomes impassable? The authors conducted agent-based microsimulations to answer the questions. Simulation results suggest that it takes 20 h and 11 min to 20 h and 14 min to evacuate the 92,596 people. This clearance time is less than the Florida state mandated 24-h clearance time limit. If one assumes that people evacuate in a 48-h period and the traffic flow from the Keys would follow that observed in the evacuation from Hurricane Georges, then a total of 460 people may be stranded if the evacuation route becomes impassable 48 h after an evacuation order is issued. If the evacuation route becomes impassable 40 h after an evacuation order is issued, then 14,000 people may be stranded.     

The logistics of household hurricane evacuation

Although there is a substantial amount of research on households’ hurricane evacuation decision making, there is much less research on the logistical issues involved in implementing those evacuations. The limited research on household evacuation logistics has consistently shown that most evacuees stay in the homes of friends and relatives or in commercial facilities rather than in public shelters. However, evacuation logistics—which can be defined as the activities and associated resources needed to reach a safe location and remain there until it is safe to return—encompasses a much broader range of behaviors than this. The present study extends previous research by reporting data on other aspects of evacuation logistics such as departure timing, vehicle use, evacuation routes, travel distance, shelter type, evacuation duration, and evacuation cost. Hurricane Lili evacuation data at the county level are generally consistent with the data from previous hurricanes, but there is notable variation across counties studied here. There were only modest correlations of demographic and geographic variables with the evacuation logistics variables, a result that indicates further research is needed to better understand what happens between the time an evacuation decision is made and the time re-entry is begun. Moreover, research is needed to understand the logistics of evacuation by special populations such as transients and households with disabled members.     

Time to leave: an analysis of travel times during the approach and landfall of Hurricane Irma

Hurricane Irma caused widespread evacuation activity across Florida and some of its neighboring states in September of 2017. The researchers gathered estimated travel times from the Google Distance Matrix API over about a month to identify and analyze evacuation periods on roads in Florida, Georgia, and South Carolina during this time. Travel time data were mathematically adjusted to show more realistic estimations. Both sets of travel times were then graphed, with the assumption that elevated travel times prior to and during hurricane landfall were indicative of evacuation activity. The study generally corroborated the well-established daytime evacuation preference. However, not all evacuation periods followed the daytime travel preference, and at least one nighttime evacuation may have been caused by flooding. In another case, later elevated travel coincided with significant power loss. Finally, the Florida data suggest that most of the evacuation traffic departed before local jurisdictions’ recommended evacuation start times.

     

Risk perception and evacuation decisions of Florida tourists under hurricane threats: a stated preference analysis

Though most hurricane evacuation studies have focused on residents, tourists are also a vulnerable population. To assess their perceptions of risk and evacuation likelihood under different hurricane conditions, we surveyed 448 tourists visiting central Florida. Respondents viewed four maps emulating track forecast cones produced by the National Hurricane Center and text information featuring variations of storm intensity, coast of landfall, centerline position relative to the survey site, time until landfall, and event duration. We performed chi-square tests to determine which hurricane conditions, and aspects of tourists such as their demographics and previous hurricane experience, most likely influenced their ratings of risk and evacuation likelihood for respondents located on Pinellas County beaches or inland near Orlando, FL. Highly rated scenarios featured a Category 4 hurricane making landfall along the Gulf Coast with the centerline passing over the sampling site. Overall, tourists that indicated the highest risk and evacuation ratings were not previously affected by a hurricane, had a trip duration of less than 6 days, and had checked for the possibility of a hurricane strike before departure. However, results for other tourist attributes differed between tourists in coastal and inland locations. We found that although somewhat knowledgeable about hurricanes, tourists misinterpreted the track forecast cone and hurricane conditions, which led to a lower perception of risk and subsequent likelihood to evacuate. Tourists, particularly those from outside of Florida, need to be better educated about the risks they face from hurricanes that make landfall.     

The perceived landfall location of evacuees from Hurricane Gustav

Hurricane evacuations in the United States are costly, chaotic, and sometimes unnecessary. Many coastal residents consider evacuation after viewing a forecasted graphic of where the storm is anticipated to make landfall. During the evacuation process, hurricane tracks commonly deviate from the forecasted landfall track and many evacuees may not pay attention to these track deviations after evacuating. Frequently, a disconnect may occur between the actual landfall track, the official forecasted track, and the perceived track of each individual as they made their evacuation decision. Specifically for evacuees, a shift in track may decrease the hazards associated with a landfalling hurricane since evacuees perceive their threat level to be high at the time of evacuation. Using survey data gathered during the evacuation from Hurricane Gustav (2008) in coastal Louisiana (USA), we calculated a type of Z-score to measure the distance error between each evacuee’s perceived landfall location and the actual landfall location from each evacuee’s home zip code. Results indicate a personal landfall bias in the direction of home zip code for evacuees of three metropolitan regions. Evacuees from the greater New Orleans area displayed the highest error, followed by evacuees from greater Lafayette. Furthermore, we validate the authenticity of the previous results by employing two additional methods of error assessment. A large regional error score might possibly be a predictor of evacuation complacency for a future hurricane of similar magnitude, although there are many other variables that must be considered.     

The effects of transportation network failure on people’s accessibility to hurricane disaster relief goods: a modeling approach and application to a Florida case study

Following the catastrophic and devastating Atlantic Hurricane seasons in 2004 and 2005, there has been increased interest in formulating planning directives and policy aimed at minimizing the societal impacts of future storms. Not all populations will evacuate an area forecast to be affected by a hurricane, so emergency managers must plan for these people who remain behind. Such planning includes making food, water, ice, and other provisions available at strategic locations throughout an affected area. Recent research has tackled problems related to humanitarian and relief goods distribution with respect to hurricanes. Experience shows that the torrential rains and heavy winds associated with hurricanes can severely damage transportation network infrastructure rendering it unusable. Scanning the literature on hurricane disaster relief provision, there are no studies that expressly consider the potential damage that may be caused to a transportation network by strong storms. This paper examines the impacts of simulated network failures on hurricane disaster relief planning strategies, using a smaller Florida City as an example. A relief distribution protocol is assumed where goods distribution points are set up in pre-determined locations following the passage of a storm. Simulation results reveal that modest disruptions to the transportation network produce marked changes in the number and spatial configuration of relief facilities. At the same time, the transportation network appears to be robust and is able to support relief service provision even at elevated levels of hypothesized disruption.     

Highway congestion during evacuation: examining the household’s choice of number of vehicles to evacuate

Hurricane evacuations in coastal counties have been reviewed and analyzed for the role of household preparedness and decisions before and during a disaster. However, one of the several emerging problems in the hurricane evacuation is transportation. Transportation issues have become more important in coastal evacuations as traffic problems impinge on people’s ability to get out of harm’s way and ultimately influence their decisions to evacuate. To add to the complexity, when families evacuate in multiple vehicles, it leads to additional vehicular traffic on roads and increased pressure on the transportation systems. However, little has been investigated on the characteristics that influence a household’s decision to evacuate in one or multiple vehicles. The outcome from such an analysis can help both the emergency managers and the transportation planners to targets groups that report taking more vehicles to develop policies that result in efficient evacuation. This study investigates the responses of evacuees surveyed after Hurricane Rita in the counties of Galveston, Brazoria and Harris County. The ordinary least square regression analysis revealed that access to transportation characteristics of a household such as number of registered vehicles in a household and number of eligible drivers was positively and significantly related to evacuating in more vehicles. Meanwhile, the risk of reaching destination safely was negatively related to taking more vehicles for evacuation even though both the risk index and deterrence index were positively significant. The time of decision and evacuation did not report any statistical significance.     

Modeling the integrated roles of insurance and retrofit in managing natural disaster risk: a multi-stakeholder perspective

This paper introduces a new modeling framework to understand and improve regional natural disaster risk management in the USA, including the interactions among key stakeholders and between the two important risk management mechanisms of insurance and retrofit. The framework includes a stochastic programming optimization to represent insurer decisions, which interacts with a utility-based model of individual homeowners’ decisions to insure and/or retrofit. Reinsurer and government roles are represented as inputs, and the decision models are integrated with a detailed regional catastrophe loss estimation model. This modeling framework is applied to a full-scale, realistic case study for hurricane risk to residential buildings in Eastern North Carolina. Several alternative system configurations are considered that affect the incentives for adoption of alternative risk management methods. They include providing a government subsidy for insured homeowners to encourage retrofit, providing both a government subsidy and insurance rebate to reduce retrofit costs, and mandating insurance purchase with a cap on insurance premiums. For each configuration, outcomes are presented from the perspectives of all key stakeholders—primary insurer, homeowners (insured and uninsured, in high- and low-risk areas), reinsurers, and the government. Results suggest that it is possible to design policies in which all stakeholders can be better off simultaneously. Retrofit incentives for insured homeowners can be effective in linking and strengthening the benefits of retrofit and insurance. Mandatory insurance coupled with capped profit loading factors and possibly retrofit rebates from the insurer to the homeowner can also reduce overall system risk.     

On the use of NOAA's storm surge model,SLOSH, in managing coastal hazards — the experience in Puerto Rico

A numerical-dynamic, tropical storm surge model, SLOSH (Sea, Land, and Overland Surges from Hurricanes), was originally developed for real-time forecasting of hurricane storm surges on continental shelves, across inland water bodies and along coastlines and for inland routing of water -either from the sea or from inland water bodies. The model is two-dimensional, covering water bodies and inundated terrain. In the present version available at the University of Puerto Rico a curvilinear, polar coordinate grid scheme is used. The grid cells are approximately 3.2 × 3.2 km in size.The model has been used in a revision of all coastal Flood Insurance Rate Maps (FIRM) for Puerto Rico and the U.S. Virgin Islands, and in hurricane evacuation studies. The FIRM's, since they are based on the 100 year stillwater elevation, are also used by the state Planning Board for regulatory purposes. The hurricane evacuation studies are used by emergency planners and personnel to assign shelters, escape routes, and delimit coastal zones that need to be evacuated during a hurricane threat.Recently, the acquisition of data from hurricane Hugo has allowed the first comparison of model results and observations for Puerto Rico and the other islands.     

A GIS-based analysis of constraints on pedestrian tsunami evacuation routes: Cascais case study (Portugal)

Tsunami hazard in coastal areas susceptible to flooding, although reduced (in terms of probability of occurrence), may pose a high risk. Therefore, in these areas, a detailed evacuation planning of the affected population is required as a risk mitigation measure. The knowledge and enforcement of evacuation routes may reduce the population vulnerability, making it more resilient and reducing risk. This paper presents a GIS approach for modelling evacuation routes based on the optimal path search problem, of the graph theory, which is implemented on ArcCasper tool. The methodology proposed considers the elements involved in the evacuation process, the worst credible tsunami inundation scenario (hazard extent and travel time), the number of people that needs to be evacuated in different time scenarios, the safe areas or destination points of the evacuation routes, the roads network characteristics and finally the time available to evacuate. The knowledge of those elements allows predicting some possible outcomes of the evacuation, such as the arrival time of the evacuees to a shelter and the identification of congestion hot spots resulting from the application of a flocking model which simulates the path to be used by evacuees avoiding obstacles. The municipality of Cascais was used to test the methodology proposed in this study. Cascais is one of the largest urban centres located about 25 km west of Lisbon, Portugal, with a high density of infrastructure along the coastline whereby most of the population and economic activities are exposed to a tsunami. The results, presented in the form of maps, allow identifying the optimal evacuation routes as well as the unfeasible routes. This crucial information could be used to the evacuation optimization regarding the location of meeting points and vertical shelters as well as to improve the accessibility of the areas to be evacuated.     

Virtual tools for volcanic crisis management, and evacuation decision support: applications to El Chichón volcano (Chiapas, México)

Decision making regarding massive evacuation of a population threatened by a probable volcanic eruption is a major problem in crisis management. Such a decision is general on the number of people to be evacuated, available resources and infrastructure, quantity and quality of the escape routes and shelters, and the economic, social and political costs involved in the operation, coupled with the updated information provided by scientists about the forecast of future activity and probable eruption scenarios. Knowing time-lapse between the evacuation decision-making time and the time in which the evacuation is completed is another critical issue that must be carefully considered in densely populated areas. In such areas, it is really important to estimate in advance this time-lapse, as the forecast must be released with enough time to complete all the evacuation process before the destructive manifestations of the eruption begin. In this context, evacuation planning is a crucial component of emergency management. It is common for Emergency Plans to include pre-established strategies. However, an evacuation procedure should be flexible, depending on the above-mentioned timing, and on the decisions, evacuation schemes, environmental characteristics and other factors. In this work, several hazard models such as a lava flow model based on a Monte Carlo algorithm, a pyroclastic density current based on energy cone model, a semi-empirical inversion model to estimate the thickness of ash deposits, and all available information about the El Chión volcano have been used to obtain the area that should be evacuated in case of an eruption. Then, multiple evacuation strategies at El Chichón volcano have been designed, considering not only the characteristics of the eruption forecast, but also environmental factors (e.g., weather conditions) and social factors (e.g., tourism and farming seasons). The variable scale evacuation model has been used to estimate the evacuation time. In the paper, those virtual tools are briefly described as well as the information obtained from the drill of 2009. In addition to the optimization of evacuation under variable conditions and situations, one of the main objectives of this work is to provide a reliable estimation of the mitigation action time, for an Emergency Plan.     

Generation of a landslide risk index map for Cuba using spatial multi-criteria evaluation

This paper explains the procedure for the generation of a landslide risk index map at national level in Cuba, using a semi-quantitative model with ten indicator maps and a cell size of 90 × 90 m. The model was designed and implemented using spatial multi-criteria evaluation techniques in a GIS system. Each indicator was processed, analysed and standardised according to its contribution to hazard and vulnerability. The indicators were weighted using direct, pairwise comparison and rank-ordering weighting methods, and weights were combined to obtain the final landslide risk index map. The results were analysed per physiographic region and administrative units at provincial and municipal levels. The Sierra Maestra mountain system was found to have the largest concentration of high landslide risk index values while the Nipe–Cristal–Baracoa system has the highest absolute values, although they are more dispersed. The results obtained allow designing an appropriated landslide risk mitigation plan at national level and to link the information to the national hurricane early warning system, allowing also warning and evacuation for landslide-prone areas.     

The sun-hurricane connection: Diagnosing the solar impacts on hurricane frequency over the North Atlantic basin using a space–time model

The authors define a spatio-statistical response of hurricane frequency to the solar cycle. Previous research indicates reduced (increased) hurricane intensities and frequency in the western (eastern) tropical Atlantic. However, no formal quantitative relationship has been spatially established between hurricane frequency and solar activity. The authors use a Bayesian hierarchical space–time model, an increasingly popular approach due to its advantage in facilitating regression modeling of space–time phenomena in the context of large data sets. Regional hurricane frequency over the period 1866–2010 is examined in response to September sunspot number (SSN) while controlling for other relevant climate factors. The response features a 13 % reduction in probability of annual hurricane occurrence for southeastern Cuba, the southern Bahama islands, Haiti, and Jamaica when the SSN is 80 sunspots. In contrast, hurricane risk in regions of the southeastern Atlantic is predicted to increase by 73 % when the SSN is 160 sunspots. The model can be ported to explore other relationships over contiguous space.     

Location-allocation modeling for emergency evacuation planning with GIS and remote sensing: A case study of Northeast Bangladesh

This work developed models to identify optimal spatial distribution of emergency evacuation centers(EECs) such as schools, colleges, hospitals, and fire stations to improve flood emergency planning in the Sylhet region of northeastern Bangladesh.The use of location-allocation models(LAMs) for evacuation in regard to flood victims is essential to minimize disaster risk.In the first step, flood susceptibility maps were developed using machine learning models(MLMs), including: Levenberg–Marquardt back propagation(LM-BP) neural network and decision trees(DT) and multi-criteria decision making(MCDM) method.Performance of the MLMs and MCDM techniques were assessed considering the area under the receiver operating characteristic(AUROC) curve.Mathematical approaches in a geographic information system(GIS) for four well-known LAM problems affecting emergency rescue time are proposed: maximal covering location problem(MCLP), the maximize attendance(MA), p-median problem(PMP), and the location set covering problem(LSCP).The results showed that existing EECs were not optimally distributed, and that some areas were not adequately served by EECs(i.e., not all demand points could be reached within a 60-min travel time).We concluded that the proposed models can be used to improve planning of the distribution of EECs, and that application of the models could contribute to reducing human casualties, property losses, and improve emergency operation.     

Optimistic bias and the consistency of hurricane track forecasts

Forecast graphics depicting a hurricane track and uncertainty cone have become pervasive in the communication of hurricane risk. This study examined whether the effect of hurricane tracks on risk perception is influenced by the consistency and optimistic bias. Specifically, it focused on the differences between forecasts that remain consistent compared to those that veer away for a forecast period. To answer the research question, this study conducted online surveys in which respondents from two coastal universities were asked risk perception questions based on a series of forecast graphics. Other variables measured included dispositional optimism, general hurricane risk perception, and hurricane information use. Optimistic bias was calculated from two of the risk perception questions. Results did not indicate strong support for an influence of optimistic bias or changing forecast track on risk perception. There was limited evidence that a veering track scenario may lead to differences in risk judgments about another location, but most measures of personal risk estimation were not influenced by the track. Dispositional optimism was not related to optimistic bias or many of the risk perception variables tested, including general hurricane risk perception. There did appear to be an interaction between track scenario and optimistic bias with more relationships being significant among those who received the consistent track scenario.     

浅谈数字国土工程建设

数字国土工程属于国家的基础建设, 涉及内容广泛, 它的目标是通过国土资源的信息化实现资源的合理利用, 进而促进国民经济的可持续性发展; 数字国土工程涉及海量数据的生产、传输、更新等, 需要宽带网络和“3S”集成技术及相关领域理论成果的支持; 数字国土工程的数据多样性决定了必须制定统一的数据库建设标准, 才能实现系统的资源共享; 工程的模型是多维的, 支持国土规划、土地管理、矿产开发决策等, 能够切实保护国土资源; 工程建设必须遵循长期规划与阶段目标统一的原则.      

Hurricane wind risk in Louisiana

A statistical procedure for estimating the risk of strong winds from hurricanes, known as the Hurricane Risk Calculator, is demonstrated and applied to several major cities in Louisiana. The procedure provides an estimate of wind risk over different length periods and can be applied to any location experiencing this hazard. Results show that an area 100 km around the city of New Orleans can expect to see hurricane winds blowing at 49 ms^?1 (44.3–53.7) [90 % confidence interval (CI)] or stronger, on average, once every 20 years. In comparison, for the same time period, the capital city of Baton Rouge and the surrounding area can expect to see hurricane winds of 43 ms^?1 (38.2–47.8) (90 % CI) or stronger. Hurricane track direction is also analyzed at the cities of interest. For Morgan City, Lafayette, Lake Charles, and Alexandria, tropical cyclones with winds at least 18 ms^?1 travel from the southeast to northwest. New Orleans and Baton Rouge tropical cyclones have a greater tendency to turn toward the east while within 100 km of the city, historically giving them a southwesterly approach. Tropical cyclones within 350 km off the south-central Louisiana coast occur most often in September, and the most extreme of these events are becoming stronger through time as shown with quantile regression.     

A multiple scale modeling system for coastal hurricane wind damage mitigation

Hurricane wind damage constitutes the largest percentage of catastrophic insured losses in the US. Yet the complicated wind structures and their changes are not fully understood and, thus, have not been considered in current risk catastrophic models. To obtain realistic landfall hurricane surface winds, a large eddy simulation (LES) framework in a weather forecasting mode has been developed from a multiple nested Weather Research & Forecasting (WRF) model to explicitly simulate a spectrum of scales from large-scale background flow, hurricane vortex, mesoscale organizations, down to fine-scale turbulent eddies in a unified system. The unique WRF-LES enables the high resolution data to be generated in a realistic environment as a hurricane evolves. In this paper, a simulation of the landfalling Hurricane Katrina is presented to demonstrate various features of the WRF-LES. It shows that the localized damaging winds are caused by the large eddy circulations generated in the hurricane boundary layer. With a sufficient computational power, WRF-LES has the potential to be developed into the next generation operational public wind-field model for hurricane wind damage mitigation.     

A novel agent-based model for tsunami evacuation simulation and risk assessment

Tsunami evacuation is an effective way to save lives from the near-field tsunami. Realistic evacuation simulation can provide valuable information for accurate evacuation risk assessment and effective evacuation planning. Agent-based modeling is ideal for tsunami evacuation simulation due to its capability of capturing the emergent phenomena and modeling the individual-level interactions among agents and the agents’ interactions with the environment. However, existing models usually neglect or simplify some important factors and/or mechanisms in tsunami evacuation. For example, uncertainties in seismic damages to the transportation network are not probabilistically considered (e.g., by simply removing the damaged links (roads/bridges) from the network). Typically a relatively small population (i.e., evacuees) is considered (due to computational challenges) while neglecting population mobility. These simplifications may lead to inaccurate estimation of evacuation risk. Usually, only single traffic mode (e.g., on foot or by car) is considered, while pedestrian speed adjustment and multi-modal evacuation (e.g., on foot and by car) are not considered concurrently. Also, pedestrian–vehicle interaction is usually neglected in the multi-modal evacuation. To address the above limitations, this study proposes a novel and more realistic agent-based tsunami evacuation model for tsunami evacuation simulation and risk assessment. Uncertainties in seismic damages to all links in the transportation network as well as uncertainties in other evacuation parameters are explicitly modeled and considered. A novel and more realistic multi-modal evacuation model is proposed that explicitly considers the pedestrian–vehicle interaction, walking speed variability, and speed adjustment for both the pedestrian and car according to traffic density. In addition, several different population sizes are used to model population mobility and its impact on tsunami evacuation risk. The proposed model is applied within a simulation-based framework to assess the tsunami evacuation risk assessment for Seaside, Oregon.