logistics of hurricane evacuation during hurricane ike

Chapter 6

LOGISTICS OF HURRICANE EVACUATION DURING HURRICANE IKE Hao-Che Wu, Michael K. Lindell, Carla S. Prater, and Shih-Kai Huang Hazard Reduction and Recovery Center, Texas A and M Univ., College Station, TX, US

ABSTRACT Although there is a fairly substantial amount of research on households’ hurricane evacuation decision making, there is much less research on the logistical issues involved in implementing evacuations. This research examines household hurricane evacuation logistics—the activities and associated resources needed to reach a safe location and remain there until it is safe to return—during Hurricane Ike in past 2008 compared with studies of hurricane evacuation logistic data from Hurricanes Lili (2002), Katrina (2005), and Rita (2005). Evacuation logistics variables include evacuation route information sources, evacuation departure dates, vehicles taken, evacuation routes and destinations, travel distances and times, shelter accommodations, and costs of transportation, food, and lodging. These studies found that, during hurricane evacuations, evacuees are most likely to choose the homes of 

Corresponding author: Hao-Che Wu. Research Assistant, Hazard Reduction and Recovery Center, Texas A and M Univ., College Station, TX 77843-3137. E-mail: wu2046@neo. tamu.edu.

128

Hao-Che Wu, Michael K. Lindell, Carla S. Prater, and Shih-Kai Huang friends/relatives as their shelter accommodations and most of them evacuated either on the day the hurricane watch or hurricane warning was issued. Additionally, this study produced data on evacuation distances and costs—as well as associations with demographic and situation..al variables. Finally, the results suggested that evacuees’ home county did make a difference on evacuation distance and this result had hnot been found in previous research.

INTRODUCTION Hurricanes are the most severe tropical storm in the meteorological hazard category. Hurricanes are tropical storms with winds that exceed 74 mph. According to the National Hurricane Center (NHC), the Atlantic hurricane season begins June 1st and ends on Nov. 30th; the Eastern Pacific hurricane season begins May 15th and also ends on Nov. 30th. The infamous Hurricane Katrina was a Category Three hurricane. About 1,500 fatalities were directly caused by the force of Katrina spread across Louisiana, Mississippi, Florida and Georgia (NHC, 2005). A few weeks later, Hurricane Rita struck the Texas area as another Category Three hurricane. During Hurricane Rita, more than two million people evacuated in Texas and fifty-five indirect fatalities were reported (NHC, 2006). A Category One or Two hurricane can result in serious damage to the public as well. Hurricane Ike, the costliest hurricane in Texas history, was a Category Two hurricane when it made land fall on Galveston Island in 2008 (NHC, 2008). In an average year, there are six hurricanes per year and only two of those actually strike the US coastal area (Lindell et al., 2007). In past twenty years, the most costly hurricane in the US was Hurricane Katrina. Hurricane Katrina cost $81.2 billion in 2005, followed by Hurricane Andrew’s $40.7 billion during the 1992 hurricane season (NHC, 2009). Hurricanes Ike was one of the ten costliest hurricanes in US history. Hurricane Ike (US$ 29.5 billion) made landfall on Galveston Island around 2:30 am CDT on Saturday, September 13, 2008, as a Category Two hurricane. Despite the evacuation of over 1.3 million people Ike caused 112 fatalities in the US and over 300 victims went missing (Huang et al., 2012). The NHC issued hurricane warnings and local governments in the risk areas issued evacuation orders before landfall for Hurricane Ike. The NHC issued a hurricane watch at 4:00 pm CDT on Wednesday, September 10 and a warning at 10:00 am on Thursday, September 11. Although local emergency managers had developed evacuation plans before the hurricanes struck, there were still problems during the evacuations. Problems such as these call attention to the

Logistics of Hurricane Evacuation in Hurricane Ike

129

need for emergency managers to understand household evacuation logistics. Based on definitions of logistics by Ballou (1987), Johnson and Wood (1996), and Rushton, Oxley, and Croucher (2000), Lindell, Kang, and Prater (2011), Wu, Lindell and Prater (2012) defined household evacuation logistics as comprising ‘‘the activities and associated resources needed to reach a safe location and remain there until it is safe to return.’’ Lindell (2008) believes that it is important for emergency managers to see a complete picture of household evacuation logistics and that will require an assessment of people’s evacuation route information sources, departure times, vehicle usage, routes and destinations, distances and times, shelter accommodations, durations, and costs (see also Lindell and Prater, 2007). In previous work, researchers who studied household evacuation logistics often focused on the first phase of this processes, which is the interval between the time an evacuation decision was made and the time households leave their homes (Dixit, Pande, Radwan and Abdel-Aty, 2008; Fu and Wilmot, 2004; Fu and Wilmot, 2006; Fu, Wilmot, Zhang and Baker, 2007), and analyzed the activities during this first phase (Kang, Lindell and Prater, 2007; Lindell, Lu, and Prater, 2005). The followings are some articles which focused on the evacuation decision behavior issues and discussed some evacuation logistic issues. Evacuation information: One of the most important psychological factors of evacuation behavior was how evacuees obtain information (Baker, 1991). There are many different information sources, such as environmental cues, others’ behavior, news media, authorities and peers, which could affect evacuees’ decisions (Mileti and Sorenesen, 1987; Tierney, Lindell, and Perry, 2001; Lindell and Perry, 2004). Lindell et al. (2011) found that the evacuation information such as written materials received before the event; local officials’ recommendations during the event; news media recommendations during the event; and expectation about the relative time were positively correlated with each in Hurricane Lili evacuation study. Only evacuees’ past experience was not correlated with other information source variables. In this study, evacuees who chose their evacuation route based on previous experience were less likely to rely on other sources of route information. In contrast, Baker (1991) found that previous personal experience is sometimes negatively related to evacuation. Zhang, Prater and Lindell (2004) indicated that information provided by officials and media is an important method of controlling evacuation traffic and previous evacuation experience has limited value. Wu et al. (2012) found past evacuation experience was significantly correlated with written materials and media recommendations during hurricanes Katrina and Rita, although the

130

Hao-Che Wu, Michael K. Lindell, Carla S. Prater, and Shih-Kai Huang

correlation was low. Household evacuation plans: Perry, Lindell and Greene (1981) indicated that evacuation plans that include a destination, routes, and transportation mode can encourage evacuation. However, they found that 42% of the evacuees lacked evacuation plans before leaving flooded areas. Similarly, Burnside (2007) reported that those who had an evacuation plan had a 22.4% higher chance of evacuating during Hurricane Katrina. Evacuation vehicle usage: Perry, Lindell and Greene (1981) found that 74% of evacuees used their own vehicles during flood evacuations and 13% of them either rode with relatives or friends or took public transportation. However, recent research findings indicated that more people try to evacuate in their own vehicles. Baker (2000) found that only 5% of evacuees rode with peers during Hurricane Floyd, whereas Siebeneck and Cova (2008) found that 90.5% of evacuees used their own vehicles during the Hurricane Rita. On the other hand, evacuees try to travel light, they generally take 1.5 to 1.7 vehicles per household during hurricane evacuation (Dash and Morrow, 2001; Siebeneck and Cova, 2008). Dow and Cutter (2002) reported that only 25% of the evacuees took two or more cars during Hurricane Floyd in South Carolina. Lindell, Kang and Prater (2011) reported that 90% of evacuees took their own vehicles and less than 1% of them used public transit, and number of the average vehicles taken per household was 1.7 during Hurricane Lili. Shelter accommodation: Researchers have found that relatively few evacuees use the public shelters or congregate care facilities. Mileti, Sorensen and O’Brien (1992) reported that only 15% of evacuees go to public shelters, which are primarily occupied by ethnic minorities, the poor and later departures. Baker (2000) found 15% of the evacuees used public shelters during Hurricane Floyd. In 2003, Whitehead found only 5.5 % of the evacuees stayed in public shelters, 15.7% of them stayed in hotels/motels, and 70% of them stayed with friends or family during Hurricane Bonnie. 54% of evacuees stayed with peers, 29% of them stayed in commercial facilities, and only 3% of them went to public shelters during Hurricane Lili (Lindell, Kang and Prater, 2011). During Hurricanes Katrina and Rita, 61% of the evacuees went to friends and relatives’ homes and 18% of the evacuees evacuated to hotels and motels (Wu et al., 2012). Evacuation departure time: Baker (2000) reported that less than 15% of evacuees left before local officials issued a warning during Hurricane Floyd. Dow and Cutter (2002) found that 5% of the evacuees evacuated before the evacuation order, 61% of the evacuees left on the day that the local officials issued evacuation orders, and 31% of them evacuated the day after the order was issued during Hurricane Floyd in South Carolina. During Hurricane Lili, 31% of the evacuees left two days before the landfall, 62% left one day before the hurricane hit, and 7% left on the day the hurricane hit


131

(Lindell, Kang and Prater, 2011). During Hurricanes Katrina and Rita, Wu et al. (2012) found that 43% of the Katrina respondents and 53% of the Rita respondents evacuated 2 days before hurricane landfall; therefore, the evacuation peak during Hurricanes Katrina and Rita was two days before hurricane landfall. Evacuation distance: Dow and Cutter (2002) found 9% the evacuees evacuated to locations within their own county, 32% of them evacuated within state, and 56 % of them evacuated out of state. Lindell, Kang and Prater (2011) reported that the average evacuation distance was 192.50 miles and the evacuees stayed away from home for 2.33 days on average during Hurricane Lili. During Hurricane Rita, evacuees traveled an average of 198 miles (Siebeneck and Cova, 2008). Wu et al. (2012) found 38% of the evacuees from New Orleans area evacuated to other states during Hurricane Katrina, but 42% of the evacuees from Texas coastal areas evacuated to east Texas during Hurricane Rita. Evacuation cost: On average, evacuees spent $68.35 dollars on transportation, $46.90 on food, and $28.47 on other expenses during Hurricane Lili, and the daily average cost was $111.84 (Lindell, Kang and Prater, 2011). However, during Hurricanes Katrina and Rita, evacuees actually spent a great deal more than Lindell et al. (2011) found during Hurricane Lili. Wu, Lindell and Prater (2012) found evacuees, on average, spent $340.99 on transportation, $333.48 on food, and $405.16 on lodging during Hurricanes Katrina and Rita. In summary, many studies have addressed sheltering and evacuation issues. However, only few articles focused on evacuation logistics. In fact, by recognizing the evacuation logistics information, the emergency managers could be able to get a clearer picture about how residents organize their evacuation decisions. The logistics information contend not only sheltering and vehicles, but also evacuation timing, distance, duration, and cost. Therefore, this study will try to analyze the above evacuation logistics variables during the Hurricane Ike and summarize the Ike evacuation logistic findings with other recent hurricane evacuation studies. Five research hypotheses will be tested in this study: H1: Demographic variables will be significantly related to evacuation vehicle access, evacuation vehicle use, and shelter accommodation type. H2: The different locations of counties will be significantly related to vehicle use and shelter accommodation. H3: The location of counties will have significant difference impact on evacuation cost, vehicle use, evacuation distance, evacuation travel time.

132

Hao-Che Wu, Michael K. Lindell, Carla S. Prater, and Shih-Kai Huang H4: Shelter accommodation type will be related to total evacuation cost. H5: Evacuees’ departure day will have significant impact on sheltering type, evacuation distance, evacuation cost. H6: Evacuation distance will be related to evacuation travel time, additional travel time, and evacuation cost.

METHOD This research is based on the survey of Hurricane Ike evacuees conducted by the Texas AandM University Hazard Reduction and Recovery Center during 2008. These questionnaires included demographic items, hurricane evacuation experience, and evacuation logistics such as evacuation travel time, evacuation vehicles, evacuation distance and sheltering types.

Sample The Hurricane Ike surveys were mailed to households in Galveston, Brazoria, Harris and Jefferson counties five months after Hurricane Ike struck Galveston, TX (Figure 1). This study used a disproportionate stratified sampling procedure to identify 200 households from each of these areas. The questionnaire was sent out following Dillman’s (1978) procedure. Each household in the sample was sent as many as three copies of the questionnaire. For Hurricane Ike, 1,426 questionnaires were mailed out and 808 households responded with a response rate of 56.6%. This response rate is higher than the response rates obtained in other HRRC studies of coastal residents (25.8% by Prater, et al. 2000; 22.4% by Lindell, et al. 2001; 50.5 % by Lindell, et al. 2011; 30.9% by Wu et al.).

Measures The questionnaire included both open-ended and closed-format questions. Respondents provided their demographic characteristics in the final section of the questionnaire.


133

Figure 1. Surveyed Areas.

These characteristics included age, gender (Female =1, Male=0), ethnicity (African American, Asian/Pacific Islander, Caucasian, Hispanic, Native American, Mixed, Other), marital status (married, single, widowed, divorced), household size, the highest level of education ( some high school, high school graduate/GED, some college/vocational school, college graduate, graduate school), yearly household income (less than $15,000; $15,000-24,999; $25,00034,999; $35,000-49,999; more than $50,000), and home ownership (renter or owner).

134


Also, the respondents were asked to provide their evacuation logistics data. The evacuation date was coded by 1 to 4 (1 is three days before the hurricane hit, 2 is two days before the hurricane hit, 3 is one day before the hurricane hit, 4 is the day hurricane hit). Evacuation transportation was measured by four variables. (1) The number of registered vehicles, (2) the number of vehicles taken, (3) the number of trailers taken, (4) the evacuation transportation mode if the household did not take their own vehicles (e.g. rode with someone else, used public transit, or other). The respondents were asked to provide their evacuation shelter type, which was coded as home of friend/relative, hotel/motel, public shelter, or other. The evacuation travel time is the time period that evacuees took to reach their final destinations measured in minutes. The respondent’s evacuation destination city was open-ended item. This item was used to calculate the evacuation distance. Evacuation cost was the sum of the respondent’s transportation and lodging cost during evacuation.

RESULTS Table 1 shows the number of responses (N), mean (M), and standard deviation (SD) for each Ike variable. Similar to Hurricanes Katrina and Rita’s mail surveys, the Hurricane Ike survey over-represented white, married, homeowning households. The mean household size is 2.46 people. The education level and income level were not substantially skewed. Most of the responders have at least some college /vocational school education level and yearly household income is between US $25,000 and US $34,999. During the Ike evacuation, there were 52.4 percent of the respondents that evacuated on the two days before the hurricane hit. 32.5 percent of them evacuated one day before and 13.4 percent of them evacuated three days before. The respondents owned 2.03 registered vehicles average. During the hurricane Ike evacuation, evacuees took an average of 1.25 vehicles and .12 trailers. 87 percent of the evacuees took their own car. For those responders who did not take their own car, 55 percent of them rode with someone else and 20 percent of them used another form of transportation. There were only 4.8 percent of the evacuees who took public transit during Ike evacuation. Most of the evacuees stayed with their friends or relatives (63%), but some stayed in hotels or motels (26%), and only a few stayed in public shelter (less than 1%). The average evacuation costs were US $780.49 dollars per household, and Ike evacuees traveled 156.92 miles on average during the evacuation.


135

Table 1. Means, standard deviations, and sample sizes Variable 1. Age 2. Female 3. White

M 56.77 0.49 0.72

SD 15.24 0.50 0.45

N 551 553 547

4. Married

0.67

0.47

549

5. HHSize

2.46

1.33

541

6. Children

0.48

0.91

543

7. Educ 8. Income (US $1000) 9. HmOwn 10. EvacDay

3.39

1.12

546

40.73

12.62

508

0.90 2.23

0.30 0.69

551 292

11. RegVeh

2.03

1.00

351

12. VehNum 13. TrlrNum 14. Pooled 15. PubTrsp

1.25 0.12 0.69 0.06

0.72 0.63 0.47 0.24

346 334 51 51

16. TravTotal (min)

294.77

301.09

334

17. TravAdd (min)

80.42

162.26

323

18. FrRelHom

0.63

0.48

338

19. HotelMotel 20. PubShltr 21. CostTotal (US$)

0.30 0.02 780.49

0.46 0.13 924.65

339 338 271

22. EvacDist (mile)

156.92

144.90

336

Description Respondent’s age Female gender Respondent’s ethnicity Respondent’s marital status Total number of persons in household Number of children under 18 Respondent’s education Annual household income in thousands of US dollars Homeowner Date of evacuation Number of registered vehicles Number of vehicles taken Number of trailers taken Rode with someone else Used public transportation Total evacuation travel time Difference between evacuation and normal travel time Stayed with friends or relatives Stayed in a hotel or motel Stayed in a public shelter Total evacuation cost Direct distance from home to evacuation destination

Table 2 shows the intercorrelations among all the variables, allowing us to draw the following conclusion. First, female gender was positively correlated with car pool use (r=.30); in contrast, marital status (r= -.33) and household size (r= -.32) were negatively correlated with car pool use. Second, evacuees took

136


more vehicles during Ike evacuation if they are younger (r= -.20), married (r= .15), having bigger house hold size (r= .28), with higher income (r= .31), or having more registered vehicles (r=.45). Third, the evacuation costs were significantly positively correlated with household size (r= .14), total travel time (r= .24), and staying in hotels/motels (r= .22). Research hypotheses were tested using ANOVA, cross-tabulation analysis, and correlation coefficient. H1: The results for hypothesis 1 are showed in the Table 2. Partially consist with the hypothesis, age was negatively correlated with vehicle access (r= -.16) and evacuation vehicle use (r= -.20). Female genter was also negatively correlated with vehicle access (r= -.16). The white population tended to use more vehicles for evacuation (r= .15) and not to go to public shelter (r= -.16). Married households had more vehicle accessibility (r= .34) and actually used more vehicle during the Ike evacuation (r= .30). Household size was positively correlated with vehicle access (r= .34), vehicle usage(r= .28), staying in hotels/motels (r= .13), and staying in public shelters (r= .12). Education level was only positively correlated with people who stayed with friends/ relatives during Ike evacuation (r= .13). Income status was positively correlated with evacuation vehicle access (r= .34), vehicle usage (r= .31) and staying with friend/relatives during evacuation (r= .18); it is negatively correlated with staying in the public shelter during evacuation (r= -.24). Homeownership was positively correlated with evacuation vehicle access (r=.23) and negatively correlated with staying with friend/relatives during evacuation (r= -.11) and staying in the public shelter during evacuation (r= -.16). Notice that trailer usage and public transportation usage are not significantly correlated with any of the demographic variables. H2: There were no significant differences among counties on evacuating vehicle usage. On the other hand, there were significant differences among counties on the sheltering variables. Table 3 shows that most of the evacuees stayed with friends or relatives, especially the evacuees from Galveston County (74.7%) and Harris County (74.4%). The second most popular shelter accommodation was hotels/motels; only a few evacuees stayed in public shelter. This result is similar to results in the Katrina and Rita evacuation study (Wu et al., 2012). The analyses of H3 showed that evacuees’ total travel time, additional travel time and evacuation distance differed significantly across home county location (Table 4).

Table 2. Intercorrelations among variables 1 1. Age 2. Female 3. White 4. Married 5. HHSize 6. Children 7. Educ 8. Income 9. HmOwn 10. EvacDay 11. RegVeh 12. VehNum 13. TrlrNum 14. Pooled 15. PubTrsp 16. TravTot 17. TravAdd 18. FrRelHom 19.HotelMotel 20. PubShltr 21. CostTotal 22. EvacDist

2 -.06 .08 -.11 -.37 -.42 -.08 -.22 .29 -.04 -.16 -.20 -.04 .13 .02 .10 -.04 -.07 .03 .02 -.02 .90

3 -.03 -.29 -.05 .03 -.03 -.18 -.08 .04 -.16 -.04 -.05 .30 -.01 -.01 .05 .03 .00 -.01 .10 -.10

4

.09 -.09 -.07 .23 .26 .12 -.01 .08 .15 .00 -.03 .13 -.13 -.18 .09 -.09 -.16 -.05 .02

5

6

7

8

9

10

11

12

13

14

15

16 17

18

19 20 21 22

.40 .17 .75 .08 .02 .00 .43 .20 .09 .40 .23 .00 -.07 .02 .31 -.08 -.03 -.05 -.03 .05 .01 .47 .34 .10 .08 .34 .23 .02 .30 .28 .07 .07 .31 .08 .06 .45 .08 .01 .01 -.03 .07 .07 .00 .05 .09 -.33 -.32 -.18 .02 .00 -.20 -.23 -.33 -.29 -.14 .20 .11 -.08 .09 -.06 .08 .04 .31 -.03 -.08 -.37 -.08 .11 .11 -.11 -.25 .01 -.21 .02 -.11 .02 .12 .02 -.11 .11 .07 -.06 -.25 -.08 -.13 .05 -.01 .02 .03 -.02 .61 .09 .03 .07 .13 .18 .08 .01 .08 .08 -.04 .07 .25 -.05 -.09 .00 .13 .07 -.08 -.11 -.11 .04 -.04 -.10 -.02 .05 -.21 .13 .18 -.65 -.08 .12 .08 .06 -.24 -.16 .03 -.14 -.18 -.03 -.01 -.06 .04 .13 -.08 .01 .07 .14 .03 -.09 -.03 .01 .03 .22 .02 .03 .12 -.18 .24 .07 -.15 .22 -.05 .03 .05 .04 -.04 -.04 .00 -.19 .08 -.12 .05 .03 .13 .44 .22 -.12 .16 -.02 .11

Shaded correlations are statistically significant at the 0.05 level (2-tailed).

-

138 Hao-Che Wu, Michael K. Lindell, Carla S. Prater, and Shih-Kai Huang Table 3. Choice of Shelter Type by County Sheltering County

FrRel

Brazoria County Galveston County Harris County Jefferson County Total (Count) 2 𝜒9 = 36.59, P < .001.

58.3% 74.7% 74.4% 45.7% 213

HotMot 33.3% 15.6% 14.0% 41.1% 87

PubShltr .0% 1.9% .0% .0% 3

Other 8.3% 7.8% 11.6% 13.2% 35

Total (Count) 12 154 43 129 338

Table 4. Difference in Variables by County Evacuation Distance (mile) County n m S.D. n Brazoria 173.6 172.7 251.79 183.07 14 44.13 70.62 13 14 County 9 4 Galvesto 182.3 14 138.2 129.9 15 242.53 183.92 151 59.92 n County 4 9 4 3 2 Harris 102.7 185.98 166.95 46 48.67 87.89 45 92.00 43 County 4 Jefferson 156.6 11 163.4 163.4 12 404.51 414.48 123 123.14 County 0 6 5 5 7 162.2 32 156.9 144.9 33 Total 294.78 301.08 334 80.42 6 3 2 0 6 Statistics F3,330 = 9.78, P < .001 F3,319 = 4.40, P < .01 F3,332=6.23, P< .001 min = minutes; m = mean, S.D. = standard deviation, n = frequency. Variable

Total travel time(mins) m S.D.

Additional travel time(mins) m S.D. n

People from Jefferson County had the longest total travel time (404.51 minutes) and additional travel time (123.41 minutes). People from Brazoria County spent the second highest total travel time (251.79 minutes) and people from Galveston County spent the second highest additional travel time (59.92 minutes). Evacuees from Brazoria County had the longest evacuation distance (173.60 miles) and evacuees from Jefferson County had the second longest evacuation distance (163.45 miles). On the other hand, unlike hurricane Katrina and Rita data (Wu et al., 2012), county location had no statistical significant impact on evacuation cost.


139

The analysis for H4 is presented in Table 2. Shelter accommodation types were partially correlated with evacuation cost. Staying with friend/relatives were negatively correlated with evacuation cost (r= -.15) and staying in hotels/motels during the evacuation were positively correlated with evacuation cost (r = .22). On the other hand, staying in shelters was not significantly correlated with the evacuation cost variable. The analysis of Research Hypothesis 5 showed that evacuees’ total travel time, evacuation distance and total evacuation cost differed significantly over the departure days (Table 5). Those who evacuated the day of the hurricane watch had the longest total travel time (338.84 mins) and evacuation distance (193.15 mile). Those who evacuated on the day of hurricane landfall spent substantially more on total cost (US $2,101.25). However, sheltering type did not have significant differences over departure days. H6 was partially supported by our correlation analysis. Evacuation distance was significantly correlated with total travel time (r = .44) and additional travel time (r=.22), but it was not significantly correlated with travel cost (Table 2). Table 5. Difference in Variables by County Variable

Total travel time(mins) m S.D. n

Evacuation distance (mile) m S.D. n

Total evacuation cost (US $) m S.D. n

Departure Day 3 days before 338.84 193.43 39 193.15 108.94 39 670.24 633.41 29 (Watch) 2 days before 306.46 215.47 150 175.64 181.70 152 823.43 853.77 123 (Warning) 1 day before 225.34 186.83 94 114.92 89.61 94 643.84 665.73 78 Landfall 183.00 95.10 5 119.28 104.05 3 2101.25 2183.88 4 Total 282.23 206.08 288 157.61 150.25 288 766.43 822.33 234 Statistics F3,284 = 4.62, P < .01 F3,284 = 4.18, P < .01 F3,230=4.63, P< .01 min = minutes; m = mean, S.D. = standard deviation, n = frequency.

140 Hao-Che Wu, Michael K. Lindell, Carla S. Prater, and Shih-Kai Huang

DISCUSSION This research focused on household hurricane evacuation logistics in hurricane Ike. It analyzed people’s actions from evacuation preparedness until they arrived at their evacuation destination. The results indicate that there is a difference among the counties in evacuation travel time and evacuation distance. However, unlike during hurricanes Katrina and Rita (Wu et al., 2012), Harris County did not have the longest evacuation travel time and evacuation distance. During the Hurricane Ike evacuation, Jefferson County had the longest evacuation travel time and Brazoria County had the longest distance. Compared to hurricanes Bonnie, Lili, Katrina, and Rita studies (Whitehead, 2003; Lindell et al, 2011; Wu et al, 2012), the average evacuation distance during hurricane Ike evacuation was relatively shorter. On average, evacuees took 1.25 vehicles to evacuate during hurricane Ike, which is consisted with previous study results (Dash and Morrow, 2001; Siebeneck and Cova, 2008; Lindell et al, 2011; Wu et al., 2012). In addition evacuees also took an average of .12 registered trailers. Similar to the Hurricanes Lili and Katrina/Rita studies (Lindell, et al., 2011; Wu et al, 2012), this study also found that the most common way to evacuate was for households to take their own vehicles no matter where they lived or when they left. There were only 4.8 percent of the evacuees who took public transit during Ike evacuation. The public transit usage is slightly higher than in the hurricane Katrina and Rita study (2%) and the Lili study (1%) but it is still very low (Lindell, et al., 2011; Wu et al., 2012). This study and the recent hurricane evacuation studies found that there was a very low percentage of evacuees who stayed in public shelters (Whitehead, 2003; Lindell et al, 2011; Wu et al., 2012). The results of this study and the previous evacuation logistic studies also indicate that sheltering choices were not related to evacuation departure day (Lindell, et al. 2011; Wu et al., 2012); on the other hand, the evacuees’ home county made a difference in the choice of shelter type in the Lili, Katrian/Rita and Ike evacuation logistic studies (Lindell, et al. 2011; Wu et al., 2012). Fifty two percent of the evacuees left on the day of the hurricane warning and 32 percent left on the day after the warning. This conclusion is consisted with Hurricane Floyd (Dow and Cutter, 2002), Lili (Lindell et al. 2011) and Katrina/Rita (Wu et al., 2012) studies. Evacuation departure day did not make a difference in sheltering type; but it did make a difference in total travel time, evacuation distance and total cost during the evacuation. Ike evacuees spent the longest total travel time on the day that hurricane watch was issued (338.84


141

minutes); however, this was not the day most of the evacuees left. Nevertheless, evacuees still spent quite a long travel time on the day most of the evacuees left (306.46 minutes). This research also analyzed the relations among evacuation departure days with some evacuation logistic variables. Unlike the Hurricane Katrina and Rita study, which found that the departure day only made a difference in additional travel time and other costs (Wu et al., 2012), we found that evacuation departure day also made difference in evacuation distance and total evacuation cost but not in sheltering choice. These inconsistent research findings indicate that more evacuation logistic research is need to form a clear picture of these issues. The average evacuation costs were US $780.49 dollars per household during the Ike evacuation. This number is much lower compared to the Hurricane Katrina and Rita study’s US $1,137 dollars per household in 2005 (Wu et al., 2012), but higher than the Lili evacuation study’s US $111.84 dollars in 2002 (Lindell et al, 2011). The evacuation cost for Hurricanes Lili, Katrina, Rita studies changes if we count for the Consumer Price Index (CPI) Inflation (Bureau of Labor Statistics, 2012). For the evacuation cost during Hurricanes Katrina and Rita, US $1,137 in 2005 has the same buying power as US$ 1,253.45 in 2008. For the evacuation cost during Hurricane Lili, US $ 111.84 dollars in 2002 has the same buying power as 113.85 in 2008. The results stay the same. The total cost of the Hurricanes Katrina and Rita study are the highest among other studies. In summary, it is important for emergency managers to realize how evacuees prepare for their evacuation, how they reach their evacuation destinations and, finally, how they return to their homes after it is safe to do so. This research and the Lili study (Lindell et al. 2011), the Katrina/Rita study (Wu et al., 2012) found that the evacuees who come from different counties could have different preferences on information sources, travel times, food cost, lodging cost, sheltering place, evacuation destinations, evacuation routes and evacuation distance. On the other hand, this study has its limitations. The response rate was 56.6% and the questionnaire was returned disproportionately by elderly and white households. Statistically this moderate response rate does not appear to bias central tendency estimates (Curtin, Presser and Singer, 2000; Keeter, Miller, Groves and Presser, 2000, Lindell and Perry, 2000) and is not likely to affect correlations (Lindell and Perry, 2000). Another limitation of this study is that shelter type was only a single item in the questionnaire; but information in the open response comment section indicated that the evacuees might had multiple shelter locations during the evacuation, as well as multiple destinations. The above problems are some important issues that can be addressed in the future studies. Still, this research has produced important

142 Hao-Che Wu, Michael K. Lindell, Carla S. Prater, and Shih-Kai Huang information about evacuation logistics that emergency managers will find useful.

REFERENCES Baker, E. J. (1991). Hurricane evacuation behavior. International Journal of Mass Emergencies and Disasters, 9, 287-310. Baker, E. J. (2000). Hurricane evacuation in the United States. In: R. Pielke, Jr. and R. Pielke, Sr. (Eds.) Storms (1st ed., pp. 308-319). London: Routledge. Ballou, R. H. (1987). Basic business logistics: Transportation, materials management, physical distribution (2nd ed.). Englewood Cliffs, NJ: Prentice-Hall. Burnside, R., Miller, D. S. and Rivera, J. D. (2007). The impact of information and risk perception on the hurricane evacuation decision-making of Greater New Orleans Residents. Sociological Spectrum, 27,727-740. Buresu of Labor Statistics. (2012), Consumer Price Index Inflation Calculator, Retrieved Dec 5, 2012, from < http://www.bls.gov/data/inflation_ calculator.htm>. Curtin, R., Presser, S. and Singer, E. (2000). The effects of response rate changes on the index of consumer sentiment. Public Opinion Quarterly, 64, 413-428. Dash, N. and Morrow, B. H. (2001). Return delays and evacuation order compliance: The case of Hurricane Georges and the Florida Keys. Environmental Hazards, 2, 119-128. Dixit, V. V., Pande, A., Radwan, E., and Abdel-Aty, M. (2008). Understanding the impact of a recent hurricane on mobilization time during a subsequent hurricane. Transportation Research Record, 2022, 94-102. Dillman, D. A., (1978). Mail and telephone surveys: The total design method. New York: Wiley. Dow, K. and Cutter, S. L. (2002). Emerging hurricane evacuation issues: Hurricane Floyd and South Carolina. Natural Hazards Review, 3, 12-18. Fu, H. and Wilmot, C. G. (2004). Sequential logit dynamic travel demand model for hurricane evacuation. Transportation Research Record, 1882, 19-26. Fu, H. and Wilmot, C. G. (2006). Survival analysis-based dynamic travel demand models for hurricane evacuation. Transportation Research Record, 1964, 211-218. Fu, H. and Wilmot, C. G., Zhang, H., and Baker, E. J. (2007). Modeling the hurricane response curve. Transportation Research Record, 2022, 94-102.


143

Huang, S. K., Lindell, M. K., Prater, C. S., Wu, H. C., and Siebeneck, L. K. (2012) Household Evacuation Decision Making in Reponse to Hurricane Ike. Natural Hazards Review, 12, 283-296 Johnson, J. C. and Wood, D. F. (1996). Contemporary logistics (6th ed.). Upper Saddle River, NJ: Prentice-Hall. Kang, J. E., Lindell, M. K. and Prater, C. S. (2007). Hurricane evacuation expectations and actual behavior in Hurricane Lili. Journal of Applied Social Psychology, 37, 881-897. Keeter, S., Miller, C., Kohut, A., Groves, R. M., and Presser, S. (2000). Consequences of reducing nonresponse in a national telephone survey. Public Opinion Quarterly, 64, 125-148. Lindell, M. K. and Perry, R. W. (2000). “Household adjustment to earthquake hazard.” Environment and Behavior, 32, 590-630. Lindell, M. K. and Perry, R. W., (2004). Communicating environmental risk in multiethnic communities. Thousand Oaks, CA: Sage. Lindell, M. K., Lu, J. C. and Prater, C. S. (2005), Household decision making and evacuation in response to Hurricane Lili. Natural Hazards Review, 6, 171-179. Lindell, M. K. and Prater, C. S. (2007). Critical behavioral assumptions in evacuation analysis for private vehicles: Examples from hurricane research and planning. Journal of Urban Planning and Development, 133, 18-19. Lindell, M. K. (2008). EMBLEM2: An empirically based large-scale evacuation time estimate model. Transportation Research Part A, 42, 140-154. Lindell, M. K., Kang J. E. and Prater, C. S. (2011). The Logistics of Household Hurricane Evacuation. Natural Hazards, 58, 1093-1109. Mileti, D. S. and Sorensen, J. H., (1987). Why people take precautions against natural disaster. NY: Cambridge University Press, Mileti, D. S., Sorensen, J. H. and O’Brien, P. W. (1992). Toward an explanation of mass care shelter use in evacuations. International Journal of Mass Emergencies and Disasters, 10, 25-42. National Hurricane Center. (2005), Tropical Cyclone Report Hurricane Katrina, Retrieved Dec 5, 2012, from < http://www.nhc.noaa.gov/pdf/ TCRAL122005_Katrina.pdf >. National Hurricane Center. (2006), Tropical Cyclone Report Hurricane Rita, Retrieved Nov 5, 2012, from < http://www.nhc.noaa.gov/pdf/TCRAL182005_Rita.pdf >. National Hurricane Center. (2008), Tropical Cyclone Report Hurricane Ike, Retrieved Nov 5, 2012, from < http://www.nhc.noaa.gov/pdf/TCRAL092008_Ike_3May10.pdf >.

144 Hao-Che Wu, Michael K. Lindell, Carla S. Prater, and Shih-Kai Huang National Hurricane Center. (2009), Hurricane History, Retrieved Dec 5, 2012, from < http://www.nhc.noaa.gov/HAW2/english/history.shtml#rita>. Perry, R. W., Lindell, M. K. and Greene, M. R. (1981). Evacuation planning in emergency management. Lexington, MA: Heath Lexington Books. Rushton, A., Oxley, J. and Croucher, P. (2000). The handbook of logistics and distribution management. London: Kogan Page. Siebeneck, L. K. and Cova, T. J. (2008). An assessment of the return-entry process for Hurricane Rita 2005. International Journal of Mass Emergencies and Disasters, 26, 91-111. Tierney, K., Lindell, M. K. and Perry, R. W., (2001). Facing the unexpected: Disaster preparedness and response in the United States, Washington, DC: Joseph Henry Press. Whitehead, J. C. (2003). One million dollars per mile? The opportunity coasts of Hurricane evacuation. Ocean and Coastal Management, 46, 1069-1083 Wu, H. C., Lindell, M. K. and Prater, C. S. (2012). Logistics of hurricane evacuation in hurricanes Katrina and Rita. Transportation Research Part F, 15, 445-461. Zhang, Y., Prater, C. S. and Lindell, M. K. (2004). Risk area accuracy and evacuation from Hurricane Bret. Natural Hazards Review, 5, 115-120.

N.G.