ArticlePDF Available

Tornado shelter-seeking behavior and tornado shelter options among mobile home residents in the United States

Authors:

Abstract

Residents of 401 mobile homes in Georgia, Mississippi, Illinois, and Oklahoma were surveyed after they heard a tornado warning. Most residents (69%) did not seek shelter during the warning. Half of those who sought shelter went to the frame house of a friend, neighbor, or relative, and 25% of those sought shelter in a basement or underground shelter. Some of the places where residents sought shelter were of dubious quality, such as their own mobile home, another mobile home, or in an out-building. Twenty-one percent of mobile home residents believed that they had a basement or underground shelter available as shelter during a tornado warning, and about half of those said they would drive to the shelter. Residents said they would drive if the shelter was more than 200m away. Fifteen percent actually had a basement or underground shelter suitable as shelter within 200m of their mobile home, but only 43% of the residents would use those shelters. The most common reason cited for not using the shelters was that they did not know the people who lived there. Likewise, a frame house or other sturdy building was within 200m of 58% of the mobile homes, but only 35% of the residents stated they would use those houses for shelter. Thirty-one percent of mobile home residents had a ditch that was at least 0.5m deep within 200m of the mobile home. However, 44% of these ditches had utility lines overhead, 23% had water in them, and 20% had trees overhead. The limited tornado shelter options among mobile home residents in the United States needs to be incorporated into safety instructions so that residents without nearby shelter are allowed to drive to safer shelter.
ORIGINAL PAPER
Tornado shelter-seeking behavior and tornado shelter
options among mobile home residents in the United
States
Thomas W. Schmidlin Æ Barbara O. Hammer Æ Yuichi Ono Æ Paul S. King
Received: 17 December 2007 / Accepted: 19 May 2008 / Published online: 11 June 2008
Springer Science+Business Media B.V. 2008
Abstract Residents of 401 mobile homes in Georgia, Mississippi, Illinois, and Oklahoma
were surveyed after they heard a tornado warning. Most residents (69%) did not seek shelter
during the warning. Half of those who sought shelter went to the frame house of a friend,
neighbor, or relative, and 25% of those sought shelter in a basement or underground shelter.
Some of the places where residents sought shelter were of dubious quality, such as their own
mobile home, another mobile home, or in an out-building. Twenty-one percent of mobile
home residents believed that they had a basement or underground shelter available as shelter
during a tornado warning, and about half of those said they would drive to the shelter.
Residents said they would drive if the shelter was more than 200 m away. Fifteen percent
actually had a basement or underground shelter suitable as shelter within 200 m of their
mobile home, but only 43% of the residents would use those shelters. The most common
reason cited for not using the shelters was that they did not know the people who lived there.
Likewise, a frame house or other sturdy building was within 200 m of 58% of the mobile
homes, but only 35% of the residents stated they would use those houses for shelter. Thirty-
one percent of mobile home residents had a ditch that was at least 0.5 m deep within 200 m
of the mobile home. However, 44% of these ditches had utility lines overhead, 23% had
water in them, and 20% had trees overhead. The limited tornado shelter options among
mobile home residents in the United States needs to be incorporated into safety instructions
so that residents without nearby shelter are allowed to drive to safer shelter.
Keywords Tornado Mobile home Warning Shelter
T. W. Schmidlin (&) B. O. Hammer Y. Ono
Department of Geography, Kent State University, Kent, OH 44242, USA
e-mail: tschmidl@kent.edu
Present Address:
Y. Ono
United Nations International Strategy for Disaster Reduction (UN/ISDR), Geneva, Switzerland
P. S. King
779 Ringwood, Ithaca, NY, USA
123
Nat Hazards (2009) 48:191–201
DOI 10.1007/s11069-008-9257-z
1 Introduction
Persons occupying mobile homes are highly susceptible to injury or death if a tornado
strikes the home (AMS 1997, 2000). About one-half of tornado deaths occur to people in
mobile homes (AMS 2000; Ashley 2007). Daley et al. (2005) reported that people in
mobile homes were 35 times more likely to die and 12 times more likely to suffer serious
injury than people in frame houses during a violent tornado. Only people outdoors suffered
greater risks of death or injury.
The National Weather Service (NWS) tornado preparedness guidelines state that
‘Mobile homes, even if tied down, offer little protection from tornadoes and should be
abandoned’ when a tornado warning is issued (NOAA 1995). If no building is available
for shelter, then mobile home residents who abandon the mobile home and find themselves
outdoors are advised to ‘lie flat in a nearby ditch or depression’ for the duration of the
warning. No advice is given for locations where a ditch is not available. An updated on-line
source (NOAA 2007) does not include the long-held recommendation to lie down in a
ditch, but mobile home residents are given no advice on where to go if a nearby shelter is
not available during a tornado warning. It is not currently known whether suitable tornado
shelters are widely available to mobile home residents in the United States.
Schmidlin et al. (2001) found that only 33% of mobile home parks in the United States had
a designated ‘tornado shelter’ and most of these ‘shelters’ were above-ground rooms with
large windows, commonly used as an office, storage room, party/community room, or laundry
room. Even in Oklahoma, where the tornado frequency is greatest, about 40% of mobile home
parks do not have a tornado shelter, and it is not clear whether mobile home residents are
willing to pay additional lot rent to cover the cost of shelters (Simmons and Sutter 2007).
Furthermore, most (63%) mobile homes were not located in mobile home parks (O’Hare and
O’Hare 1993) but on rural property, and the availability of tornado shelters for the residents of
these rural mobile homes is not known. In addition, it is unclear whether mobile home
residents follow current NWS advice during tornado warnings or take a different course of
action. Our earlier field work showed that residents of mobile homes often do not have
underground shelter nearby and they shun the recommendation to lie down in a ditch during a
tornado warning (Schmidlin et al. 1998). Previous research (reviewed by Sorensen 2000)
suggested that an individual’s perceptions of both the hazard and the possible courses of
action directly influence the individual’s choice of behavior when faced with that hazard.
What mobile home residents perceive as potential shelter, or what they believe to be a
feasible course of action during a tornado warning, is unknown. Since about one-half of
tornado-induced deaths in the United States occur among residents of mobile homes (AMS
2000), it is important to understand the tornado shelter-seeking behavior and the tornado
shelter options of mobile home residents. This understanding could be used to improve
tornado safety recommendations and reduce risk of injury or death in this vulnerable
population.
The goals of this research were to assess the shelter-seeking behavior of mobile home
residents during a tornado warning and to determine both the actual and the perceived
availability of shelter for those residents.
2 Methods
Surveys were completed in-person for one adult resident in each of 401 mobile homes in
Georgia, Mississippi, Oklahoma, and Illinois. Counties were chosen that had a recent
192 Nat Hazards (2009) 48:191–201
123
tornado warning issued by the NWS during times of the day when residents were most
likely to be awake and aware of the warning (0700–2300 local time). The warning was
issued for the entire county in all cases. Two to four researchers traveled to the counties
for the fieldwork. The survey (available from Schmidlin) asked how the resident heard
about the warning, whether they felt threatened, whether they sought shelter during the
warning, and, if so, where they sought shelter, the distance to the shelter, and whether
they walked or drove to the shelter. All surveyed residents, regardless of whether they
sought shelter during the warning, were asked where they believed was the nearest
available basement or underground shelter, where they believed was the nearest available
sturdy building, how long it would take them to get to those shelters, and whether they
would walk or drive to the shelter. The researchers also assessed whether there was a
basement or underground shelter within 200 m of the mobile home and whether there
was a frame house or other sturdy building within 200 m of the mobile home, how far
these potential shelters were from the mobile home, and any limitation on the resident’s
use of the building as a tornado shelter. Demographic characteristics were collected
about each respondent.
Field-work was completed in Rankin and Neshoba counties, Mississippi, during 22–
29 March 2000 following tornado warnings on 15 March; in Tulsa County, Oklahoma,
during 26 April–1 May 2000 following tornado warnings on 26 March; in Champaign,
Ford, Livingston, Logan, and Macon counties, Illinois, during 19–23 May 2000 fol-
lowing tornado warnings on 12 and 18 May; and in Troup County, Georgia, during 19–
23 April 2001 following tornado warnings on 3 April. All of the warnings were
Doppler radar-based warnings and did not state that a tornado had been observed. No
tornadoes were reported in the warned counties in Mississippi or Georgia. A brief
touchdown of an F0 tornado with no damage occurred in Tulsa County, Oklahoma,
during the warning. Eleven tornadoes were reported among the five Illinois counties on
12 and 18 May 2000. All were F0 tornadoes, 10–50 yards wide, with no injuries and
no property damage or crop damage reported. Thus, it is unlikely that the appearance
of a tornado affected the shelter-seeking response during the tornado warnings in these
samples.
In Mississippi and Oklahoma, a list of addresses of all mobile homes in the selected
counties was obtained from the county administration office. Addresses for 120 residents
to be surveyed were randomly selected. Teams of two researchers traveled to each
randomly selected address. If an adult was at home, and if that person was at home and
aware of the warning when it was issued, then the brief (5–10 min) survey was com-
pleted. If there was no one home, or if they were not at home when the warning was
issued, or were not aware of the warning, then the researchers went to the nearest mobile
home and continued until a survey was completed in that area. Nearly all of the mobile
homes in the surveyed Illinois counties were located in mobile home parks. These parks
were most often found near the edges of cities and villages. Rural roads were driven in
search of rural mobile homes, but most surveys were completed within the mobile home
parks. Similarly, convenience samples were completed at rural mobile home sites and in
mobile home parks in Georgia.
Teams completed 12–20 surveys per day in rural areas and 15–30 surveys per day in
mobile home parks. Surveying was done during the day and in the evening, up until
darkness, to capture the broadest span of demographics in the sample. A total of 401
surveys were completed, 111 in Mississippi, 103 in Oklahoma, 110 in Illinois, and 77 in
Georgia.
Nat Hazards (2009) 48:191–201 193
123
3 Results
3.1 Demographics and source of warning
Selected demographic characteristics of the 401 mobile home residents surveyed are shown
in Table 1. Demographic characteristics of the overall sample are similar to national
statistics for residents of mobile homes, as taken from 1999 U.S. Census Bureau housing
data (U.S. Census Bureau 1999). Demographic characteristics of samples were similar
among the states. Exceptions included more Blacks in Mississippi (21%) and Georgia
(12%) than Oklahoma (1%) and Illinois (2%). About half of the surveyed mobile homes in
Georgia and Oklahoma were in mobile home parks, while that number was 21% in Mis-
sissippi and 90% in Illinois. More (*87%) of the surveyed residents in Oklahoma and
Illinois reported tornado warning sirens nearby than in Mississippi and Georgia (*52%).
Regional analyses of results are not presented here, although local NWS offices and
emergency managers may find it useful to study local characteristics for their areas.
Answers to survey questions showed that 51% of the mobile home residents had seen a
tornado in the past, 12% had previous home damage from a tornado, and 1% had been
injured by a previous tornado. Research results are mixed on whether previous experience
with a hazard or previous losses from a hazard affect actions taken when confronted with
the hazard again (Drobot et al. 2007). Seventy-one percent of residents said they had
tornado warning sirens nearby but only 57% reported hearing the sirens during the
warning. Only 13% owned a NOAA weather radio. Nearly all of the residents (96%) had a
Table 1 Demographic characteristics of the sample of mobile home residents
Census 1999 Overall sample
Sample size 401
Mean age 45 46
% [65 years 28 21
% Female 62
% White, non-Hispanic 84 89
% Black, non-Hispanic 9 9
% Hispanic 6 2
% In mobile home park 53
% Married 57
% With high school diploma 69 68
% With college degree 10 12
% One-person households 27 18
Mean number in household 2.5 2.9
% With children under 18 years 38 48
Mean number of vehicles when warned 1.7
% With at least one vehicle 93 96
% With family member with mobility limitations 14
% Who own a NOAA weather radio 13
% Who have warning sirens nearby 71
The column headed ‘‘Census 1999’’ is data taken for residents of mobile homes from the American Housing
Survey for the United States: 1999
194 Nat Hazards (2009) 48:191–201
123
motor vehicle at the home, 45% had two vehicles, and 13% had three or more. Fifty-seven
percent of respondents were married, 19% divorced or separated, 13% widowed, and 12%
single. Most residents (82%) reported having a pet and, of those who sought shelter, 73%
took a pet with them.
Because we only surveyed residents who were at home and aware of the tornado
warning when it was issued, everyone in our sample was aware of a tornado warning. The
source of the tornado warning was overwhelmingly from the television (85%), followed by
siren (53%), local radio (11%), a personal phone call (9%), and NOAA Weather Radio
(2%). Many people reported hearing the warning from more than one source. About 2/3 of
the sample reported first hearing about the warning from television, followed by siren
(24%), local radio (4%), and a personal phone call (4%). The importance of television in
conveying hazard warnings is well-known (Hammer and Schmidlin 2002; Hayden et al.
2007). Hayden et al. (2007) showed that choice of media for flood warnings varies among
demographic groups, and not surprisingly, younger people prefer the internet or cell phones
for weather information. Hayden et al. (2007) recommended that warnings be targeted to
distinctive segments of the public through diverse media.
3.2 Shelter-seeking behavior
Overall, 31% (126/401) of the mobile home residents said that they sought shelter during
the tornado warning. Of those who sought shelter, 52% (66) went to the frame house of a
relative, friend, or neighbor. In addition to those who sought shelter in frame houses,
another 15% (19) sought shelter in what were assumed to be reasonably sturdy buildings,
such as a storm shelter, church, fire station, apartment building, or underground parking
garage. Thus, of those who sought shelter, 67% (85) were in sturdy buildings. This was
21% of the total sample. Of the 126 who sought shelter, 25% (31) were underground (22 in
basements and 9 in storm shelters). This was about 8% of the total sample of mobile home
residents.
Among those who said that they sought shelter, 39% went less than 200 m to reach the
shelter. The median distance that the residents traveled to shelter was 805 m (1/2 mile) and
22% traveled 4.8 km (3 miles) or more. Most (73%) of those who sought shelter got in a
vehicle to drive to the place of shelter. Among those who drove to a shelter, 73% said they
wore a seat belt. The time that the person remained at their reported place of shelter was
most often given as 1 h (43%), one-half hour (13%), or 2 h (13%).
Among the 31% who said that they sought shelter during the warning, some chose a
shelter location that provided little or no reduction of risk for injury compared to staying in
the mobile home. This may have been a result of a poor decision on their part or to their
perception that there was no better shelter available nearby. Among examples of ‘shelter’
choices that provided little reduction in risk, 9% of those who sought shelter chose their
own mobile home or the mobile home of a friend or relative so that they would have
company during the storm. This shelter-seeking behavior did not actually reduce their risk
of injury. Some said they went to the hallway, bathroom, or closet of the mobile home,
apparently trying to minimize risk by adopting the instructions given to residents of frame
houses. One person drove to park under an overpass and another drove to a car wash and
parked inside during the storm. Neither resident thought that location would provide
protection from the tornado, but rather they sought to protect their vehicles from hail
damage. More than three-quarters (76%) of mobile homes nationally do not have a garage
or carport (U.S. Census Bureau 1999). Other locations sought for shelter may have been
Nat Hazards (2009) 48:191–201 195
123
safer than the mobile home but were of dubious safety in high winds. These included a car
dealership, retail stores, movie theater, and out-buildings. Several people drove away from
the mobile home without a sturdy building or shelter as a destination. Some of these sought
to escape the known hazards of being in a mobile home during high winds, while others
who lived in a mobile home among trees sought a better vantage from which to watch the
sky for severe weather.
In summary, 31% of mobile home residents said they sought shelter during the tornado
warning, and 21% went to a sturdy shelter, including the 8% who went to a basement or
underground shelter. Some of those who said they sought shelter made poor choices for the
shelter.
3.3 Factors associated with seeking shelter
To assess personal, situational, and environmental factors associated with the action of
seeking shelter during a tornado warning, Pearson correlation was calculated in SPSS
between the outcome of whether the mobile home resident sought shelter and 26 other
variables. Variables that were significantly (p \ 0.05) correlated with seeking shelter were
the belief that they were in the path of the tornado, the belief that they were in danger,
knowledge of a sturdy building available for shelter, the mobile home located on a private
lot, male gender, presence of children in the house, having a high school diploma, and no
reference during the interview to ‘God’s will.’ Researchers noticed early in the surveying
that residents often volunteered a response called the ‘God’s will’ comment. This most
often occurred when they were asked the questions, ‘‘At any time did you take shelter?’’ or
‘At any time did you feel in danger?’ during the recent tornado warning. After answering
‘No’ (typically), they would mention statements such as, ‘If it’s my time the Lord will
take me,’ ‘If I go, its God’s will,’ ‘If it’s the good Lord’s will for me to go,’ ‘The
good Lord will take care of me,’ or ‘If it’s my time, God will get me wherever I am.’
Although not part of the survey, researchers tallied these responses. Overall, 9% of resi-
dents gave a ‘‘God’s will’ response. Those who mentioned ‘‘God’s will’’ tended not to take
shelter, to have had previous tornado damage to their home, to live on a private lot rather
than in a mobile home park, not be a graduate of high school, and to have a person in the
house with limited mobility. Variables in the overall sample that were not correlated with
seeking shelter included previous experience with a tornado, hearing warning sirens,
knowledge or ownership of a NOAA weather radio, age, race/ethnicity, number of people
in the family, college degree, and presence of a person in the home with physical
disabilities.
3.4 Shelter availability assessed by the residents
Residents were asked, ‘Where do you believe is the nearest basement or underground
shelter that you could use for shelter?’ and ‘Where do you believe is the nearest sturdy
building that you could use for shelter?’ If they indicated either type of shelter was
available, researchers asked how far away that shelter was, how long it would take to reach
the shelter, and whether they would walk or drive to the shelter.
Overall, 21% of mobile home residents stated that they had a basement or underground
shelter available as a tornado shelter. The median distance to the shelter was 135 m, the
median time to reach the shelter was 2.5 min, and 46% of the residents with a basement or
underground shelter available to them would drive to the shelter. Of those who said a
196 Nat Hazards (2009) 48:191–201
123
basement or underground shelter was available within 200 m of their home, 93% would
walk (7% would drive) to the shelter. In contrast, of those for whom the nearest available
basement or underground shelter was more than 200 m away from their home, only 13%
would walk (87% would drive) to the shelter. Thus, by the residents’ own admission, they
were unlikely to walk more than 200 m to a basement or underground shelter as a tornado
shelter. Beyond that distance, most would drive to the shelter. Of the residents who said
they had a basement or underground shelter available, 5% said it was more than 10 km
away (and as far as 24 km) with a median driving time of 15 min to those shelters more
than 10 km away. A basement or shelter at a neighbor’s or friend’s house was the most
frequently cited location for the shelter (10% of the total sample). Another 5% said a
relative’s house had the nearest available basement or underground shelter and 2% had a
shelter on their own property. Among other locations for the nearest available basement or
underground shelter were mobile home park shelters, churches, commercial buildings,
school, underground parking garage, fire station, hospital, and an old school bus buried in
the yard.
Overall, 59% of mobile home residents said that they had a sturdy building available as
a tornado shelter. Of those who said they had a sturdy building available for shelter, the
median distance to the building was 350 m, the median time to reach the building was
3.0 min, and 61% would drive to the building. About a quarter (26%) of the residents said
they had a sturdy building available within 200 m of their mobile home. Of those who said
a sturdy building was available for shelter within 200 m, 87% would walk (13% would
drive) to the shelter. In contrast, of those for whom the nearest available sturdy building
was more than 200 m away from their home, only 2% would walk (98% would drive) to
the shelter. Again, by the residents’ own admission, they are unlikely to walk more than
200 m for a sturdy building as tornado shelter. Beyond that distance, most would drive to
the shelter. A relative’s house was the most frequently cited (23% of total sample) sturdy
building available as a tornado shelter. Another 15% stated a friend’s or neighbor’s house
was their nearest sturdy building, although one in seven of those respondents said they did
not know the neighbor and would not go there for shelter. Commercial buildings were cited
as the nearest sturdy building by 7% of the mobile home residents. These included
WalMart, offices, shopping centers, grocery stores, VFW hall, beauty salon, and a bar.
Various out-buildings, such as a garage, workshop, barn, and club house were cited by 5%
of the residents as their nearest ‘sturdy buildings’ but were of dubious strength. Other
structures cited as the sturdy building were a church, fire station, hospital, school, motel,
and an underground airport parking garage.
3.5 Shelter availability assessed by the researchers
The researchers assessed the availability of tornado shelter within 200 m of the mobile
home, independent of the residents’ perceptions of shelter. Two hundred meters was about
the greatest distance that mobile home residents would walk or run to a shelter. At greater
distances, most would drive to the shelter. By visual inspection and inquiries to the resi-
dents, researchers determined whether there was a basement or underground shelter, a
frame house, or other sturdy building within 200 m, the actual distance to the structure, and
any limitation to its use by the resident during a tornado warning.
Fifteen percent of mobile home residents had a basement or underground shelter within
200 m of their home. The median distance to the basement or underground shelter was
55 m but only 43% of the residents who had a basement or underground shelter within
200 m would use that location as a tornado shelter. The most common reasons given for
Nat Hazards (2009) 48:191–201 197
123
non-use were that they did not know the owners of the basement or underground shelter, it
could not hold everyone, it was fenced off, and it was full of stored items.
Fifty-eight percent of mobile home residents had a frame house (46%) or other sturdy
building (12%) within 200 m of their home (some of these also had a basement or
underground shelter within 200 m). The median distance to the frame house or sturdy
building was 70 m but only 35% of the residents who had a frame house or sturdy building
within 200 m would use that location as a tornado shelter. The most common reasons given
for non-use were that they did not know the owners of the nearest frame house, it was
fenced off, it was not a well-built home, the building was locked, and the resident was
disabled so could not get into the house. By the researchers’ assessment, 31% of the mobile
home residents had no basement, underground shelter, or sturdy building within 200 m of
their mobile home.
Mobile home residents who do not have sturdy shelter nearby have been advised to go
outdoors and lie flat in a ditch or depression for the duration of the tornado warning
(NOAA 1995). No research has been found that shows ditches offer protection from
injuries during a tornado. Researchers found that 31% of mobile home residents had a ditch
at least 0.5 m deep within 200 m of their home. The median depth of the ditches was 1 m
and only 2% of the ditches were more than 2 m deep. The median distance to the ditch was
40 m. There were many limitations to using the ditch as a tornado shelter, including utility
lines overhead that may fall into the ditch (44%), the ditch was full of water (23%), trees
along the ditch may fall during high winds (20%), and the ditch was fenced off (6%). Some
residents noted that there were snakes or fire ants in the ditch. No one in our sample moved
to a ditch for shelter during the tornado warning.
4 Discussion
This in-person survey of 401 mobile home residents in four states who were at home and
aware of a previous tornado warning showed that 69% (275) did not seek any sort of
shelter during the warning. Of those who did seek shelter, about 33% (41) sought types of
‘shelters’ that offered little if any reduction of risk over the mobile home. Thus, 79%
(316) of mobile home residents at home and aware of a tornado warning did not evacuate
to a reasonably sturdy shelter. This number would be even higher for the general mobile
home population, some of whom would not be aware of the warning. On the other hand,
this number may be lower (more people seeking shelter) for situations in which a tornado
actually approached the mobile home. People often wait to take action until the threat from
the tornado is confirmed or seems imminent, but by then there may not be enough time to
reach shelter (Schmidlin and King 1995, 1997; Hammer and Schmidlin 2002).
Most residents who sought shelter during the warning drove away from the mobile
home to the shelter. This action is contrary to NWS tornado safety recommendations.
NOAA (2007) directs vehicle occupants to ‘Get out of automobiles’ when a warning is
issued and if you are in a car, ‘‘leave it immediately.’’ However, recent research shows that
being in a vehicle may be less risky than being in a mobile home or outdoors during a
tornado warning (Duclos and Ing 1989; Carter et al. 1989; Brown et al. 2002; Schmidlin
et al. 2002; Daley et al. 2005). Therefore, driving to a shelter when a warning is issued
may be a reasonable choice for mobile home residents whose shelter is more than 200 m
away.
Only 21% of mobile home residents said they had a basement or underground shelter
available for tornado shelter and 46% said they would drive to that location. A little over
198 Nat Hazards (2009) 48:191–201
123
half (59%) of mobile home residents said they had a frame house or other sturdy building
available as a tornado shelter, and 61% said that they would drive to that building. Gen-
erally, mobile home residents said they would walk/run to shelters within 200 m of their
home but would drive if the shelter was more than 200 m away.
By the researchers’ own assessment, 15% of the surveyed mobile home residents had a
basement or underground shelter within 200 m of their home and 58% had a frame house
or other sturdy building within 200 m of their home. However, fewer than half of the
mobile home residents said they would actually go to those closest potential shelters,
mostly because they did not know the owners. Thirty-one percent did not have any
underground shelter or sturdy building within 200 m of their home. Thirty-one percent of
mobile home residents had a ditch [0.5 m deep within 200 m of their home, but most
ditches had limitations to their use as a ‘shelter’ and the residents surveyed did not lie down
in ditches during the warning.
5 Recommendations
Stronger efforts and better preparedness methods are necessary to convince residents of
mobile homes to seek shelter immediately when a tornado warning is issued and not wait
until the storm is upon them. School, church, and library programs for children of all ages
and television public service announcements aimed at children may lead to improved
household response and lead to a generation with improved tornado response (regardless of
housing type).
The limited tornado shelter options among mobile home residents in the United States
needs to be incorporated into safety instructions so that residents without nearby shelter
are ‘allowed’ to drive to the closest safer shelter. The NWS and American Red Cross tell
residents of mobile homes that they should not use their vehicles to drive to shelter during a
tornado warning. This effectively limits the use of shelters to those structures within 200 m
of the mobile home. Tornado preparedness literature and programs directed to mobile
home residents should acknowledge that a large portion of this targeted, vulnerable
audience does not have a sturdy building within 200 m and most do not have a ditch within
200 m.
Residents of frame houses should be encouraged to extend invitations to their neighbors
in mobile homes to come to their frame house when a tornado warning is issued or when
the mobile home residents feel threatened by severe weather. Sturdy frame homes within
200 m of mobile homes are under-used as tornado shelters because the residents of mobile
homes do not feel that they know their neighbors well enough to seek shelter there. This is
tragic. Instead of going to the neighbor’s frame house for shelter, these mobile home
residents drive to shelter or stay in their mobile home, in either case leading to greater risk
than going to a frame house. Perhaps the reluctance to seek shelter in a neighbor’s frame
house is due to class differences, a wish to not bother someone, a sense of privacy, or other
reasons. This neighborly gesture could be facilitated by county emergency managers,
police, fire departments, churches, and service organizations.
Residents in mobile home parks need a place of shelter during tornado warnings and
other wind storms. This is a nearly universal sentiment, yet little progress is made.
Objections to providing shelters commonly include cost, liability, security, and access.
These should be overcome so that a wind shelter is provided for these residents.
America needs an inexpensive, in-ground, portable, accessible family tornado shelter to
be invented and marketed to rural mobile home residents and to residents of frame houses
Nat Hazards (2009) 48:191–201 199
123
without basements. A substantial monetary award to the best invention would make this
happen.
Future research should examine the effects of the wording of the tornado warning on
residents’ responses (following the work of Smith and Piltz 1999). Does the report of a
‘tornado on the ground’ or stating this is a ‘dangerous situation’ or ‘this tornado has a
history of damage’ or a ‘tornado emergency’ affect the response of mobile home resi-
dents? Do residents living in or near communities mentioned in the warning have a
different response than those farther from the stated path? Previous experience with false
alarms in tornado warnings may affect response and should be investigated. Proximity to a
major television market and the statements of local television weathercasters may affect
response and could be investigated to improve the message. The ‘God’s will’’ response to
the warning and the reluctance to seek shelter should be investigated so that this response
can be understood and shelter provided. The mechanisms of how residents of mobile
homes perceive a threat when under a tornado warning should be studied so that realistic
threat levels can be communicated more effectively. Further study of why some residents
of mobile homes do not seek shelter in a neighbor’s sturdy house may reveal methods to
overcome that reluctance.
Acknowledgments This research was funded by the National Science Foundation, Geography and
Regional Science grant #9904402. We appreciate assistance in our field work from Megan Olsen. Jan
Winchell assisted with data analysis, and Mark Bradac, Jeremy Hunter, and Margaret Dixon assisted with
data entry.
References
AMS (1997) Policy statement: mobile homes and severe windstorms. Bull Am Meteorol Soc 78:850–851
AMS (2000) Policy statement: Tornado preparedness and safety. Bull Am Meteorol Soc 81:1061–1065.
doi:10.1175/1520-0477(2000)081\1061:PSTPAS[2.3.CO;2
Ashley WS (2007) Spatial and temporal analysis of tornado fatalities in the United States: 1880–2005.
Weather Forecast 22:1214–1228. doi:10.1175/2007WAF2007004.1
Brown S, Archer P, Kruger E, Mallonee S (2002) Tornado-related deaths and injuries in Oklahoma due to
the 3 May 1999 tornadoes. Weather Forecast 17:343–353. doi:10.1175/1520-0434(2002)017
\0343:TRDAII[2.0.CO;2
Carter AO, Millson ME, Allen DE (1989) Epidemiologic study of deaths and injuries due to tornadoes. Am J
Epidemiol 130:1209–1218
Daley WR, Brown S, Archer P et al (2005) Risk of tornado-related death and injury in Oklahoma, May 3,
1999. Am J Epidemiol 161:1144–1150. doi:10.1093/aje/kwi142
Drobot SD, Benight C, Gruntfest EC (2007) Risk factors for driving into flooded roads. Environ Hazards
7:227–234. doi:10.1016/j.envhaz.2007.07.003
Duclos PJ, Ing RT (1989) Injuries and risk factors for injuries from the 29 May 1982 tornado, Marion,
Illinois. Int J Epidemiol 18:213–219. doi:10.1093/ije/18.1.213
Hammer B, Schmidlin TW (2002) Response to warnings during the 3 May 1999 Oklahoma City tornado:
reasons and relative injury rates. Weather Forecast 17:577–581. doi:10.1175/1520-0434(2002)017
\0577:RTWDTM[2.0.CO;2
Hayden MH, Drobot S, Radil S, Benight C, Gruntfest EC, Barnes LR (2007) Information sources for flash
flood warnings in Denver, CO and Austin, TX. Environ Hazards 7:211–219. doi:10.1016/j.
envhaz.2007.07.001
NOAA (1995) Tornadoes Nature’s most violent storms: a preparedness guide including safety infor-
mation for schools. U.S. Department of Commerce, National Weather Service
NOAA (2007) Tornadoes Nature’s most violent storms. http://www.nssl.noaa.gov/edu/safety/tornado
guide.html. Cited 12 Dec 2007
O’Hare W, O’Hare B (1993) Upward mobility. Am Demogr 15:26–34
Schmidlin TW, King PS (1995) Risk factors for death in the 27 March 1994 Georgia and Alabama
tornadoes. Disasters 19:170–177. doi:10.1111/j.1467-7717.1995.tb00367.x
200 Nat Hazards (2009) 48:191–201
123
Schmidlin TW, King PS (1997) Risk factors for death in the 1 March 1997 Arkansas tornadoes. Quick
Response Report 98, Natural Hazards Center, University of Colorado, Boulder. http://www.colorado.
edu/hazards. Cited 12 Dec 2007
Schmidlin TW, King P, Hammer B, Ono Y (1998) Risk factors for death in the 22–23 February 1998 Florida
tornadoes. Quick Response Report 106, Natural Hazards Center, University of Colorado, Boulder.
http://www.colorado.edu/hazards. Cited 12 Dec 2007
Schmidlin TW, Hammer B, Knabe J (2001) Tornado shelters in mobile home parks in the United States.
J Am Soc Prof Emerg Plan 8:1–15
Schmidlin TW, Hammer B, King P, Ono Y, Miller S, Thumann G, Golden J, McCarthy D, Goldhammer R
(2002) Unsafe at any (wind) speed: testing the stability of motor vehicles in severe winds. Bull Am
Meteorol Soc 83:1821–1830. doi:10.1175/BAMS-83-12-1821
Simmons KM, Sutter D (2007) Tornado shelters and the manufactured home parks market. Nat Hazards
43:365–378. doi:10.1007/s11069-007-9123-4
Smith RD, Piltz SF (1999) Situation specific tornado warnings at the National Weather Service Forecast
Office, Tulsa, Oklahoma. Nat Weather Dig 23:41–44
Sorensen JH (2000) Hazard warning systems: review of 20 years of progress. Nat Hazards Rev 1:119–125.
doi:10.1061/(ASCE)1527-6988(2000)1:2(119)
U.S. Census Bureau (1999) American Housing Survey of the United States: 1999, Table 2–7. http://www.
census.gov/hhes/www/housing/ahs/ahs99/tab27.html. Cited 12 Dec 2007
Nat Hazards (2009) 48:191–201 201
123
... Manufactured housing and MHPs have a greater likelihood than the general housing stock of being located in unincorporated (Pierce et al. 2018) and rural or semirural areas (Pierce and Jimenez 2015). Previous studies identified outsized impacts of climate-related disasters on mobile home residents, including those from hurricanes (Kusenback et al. 2010), tornadoes (Schmidlin et al. 2009;Sutter and Simmons 2010;Liu et al. 2019), and floods (Baker et al. 2012;Rumbach et al. 2020). While the connection between households living in manufactured housing units and climate change risk is understudied, some literature does depict a clear relationship between social vulnerability and the impact of wildfires and extreme heat (Wigtil et al. 2016;Cova et al. 2013). ...
... However, from the perspective of social inequality, different individuals or population groups may have varied mobility responses even under the same level of environmental stress because of effect modification of socio-economic factors (Fig. 5). The disparate exposure to risks across different social groups is widely documented (e.g., Lewis, 2019;Sun et al., 2017) and explained from multiple perspectives, such as spatial segregation, home and car ownerships, work environment, access to amenities, and urban lifestyles (e. g., Huang et al., 2022;Laumbach et al., 2015;Schmidlin et al., 2009;Sun et al., 2017;White and Borrell, 2011). For example, rich people work indoors and have a high probability of car ownership, which protect them from the outdoor hazards; by contrast, the poor are more likely to commute by public transit or bicycle and work in outdoor occupations (Sun et al., 2017). ...
Article
Full-text available
In view of the increasing occurrence of urban environmental hazards which can threaten our quality of life and health outcomes, how citizens react to adverse environmental impacts is of great concern. Mobility, as an important part of urban everyday life, is a key concept in understanding individuals' experience in cities and the society ; however, this notion has long been understudied with regard to avoiding the harmful effects of environmental hazards. By highlighting the research perspective of averting behavior, we conceptualize mobility as a response (MaaR) to environmental hazards to inform a new interdisciplinary dialogue for transportation, urban planning, and environmental health research. We review a variety of literature in environmental hazards, (im) mobility, and inequality and reveal the challenges posed by MaaR. MaaR would provide empirical evidence on whether, how, and to what extent the effects and social costs of environmental hazards are underestimated and existing environmental inequality are exacerbated. This article puts forward that MaaR is a promising analytical framework through which we can enrich our knowledge on the person−environment relationships.
... The potential for loss is a common construct for many social vulnerability definitions, for example, "social vulnerability to natural hazards is the potential for loss and is complex interaction among risk, mitigation, and the social fabric of a place." [12] To validate the index as a classifier, a realized disaster loss (DL) event must be used corresponding to an actual disaster loss for each one of the n locations (e.g., number of deaths for a census tract or county). Consequently, performance assessment of the SVI classifier is easily accomplished by comparing the SVI (i.e., potential for loss) and the DL severity (i.e., actual losses or harm) rankings of each location. ...
Article
Considering the potential for widespread adoption of social vulnerability indices (SVI) to prioritize COVID-19 vaccinations, there is a need to carefully assess them, particularly for correspondence with outcomes (such as loss of life) in the context of the COVID-19 pandemic. The University of Illinois at Chicago School of Public Health Public Health GIS team developed a methodology for assessing and deriving vulnerability indices based on the premise that these indices are, in the final analysis, classifiers. Application of this methodology to several Midwestern states with a commonly used SVI indicates that by using only the SVI rankings there is risk of assigning a high priority to locations with the lowest mortality rates and low priority to locations with the highest mortality rates. Based on the findings, we propose using a two-dimensional approach to rationalize the distribution of vaccinations. This approach has the potential to account for areas with high vulnerability characteristics as well as to incorporate the areas that were hard hit by the pandemic.
Article
Full-text available
Analysis of tornado strength winds interacting with a highway overpass structure is presented with emphasis on air flow patterns above and under the bridge. Experiments were performed in a wind tunnel with the scaled geometry of an overpass. Velocity and dynamic pressure measurements were obtained independently at four locations as the overpass was rotated about its vertical axis between air flow angles of approach between 0° and 90°, at 10° increments. Lift and drag forces on the overpass geometry were also measured. To compare various highway overpass locations with the surroundings, the measured dynamic pressure and velocity, drag and lift forces, and drag coefficients at each of the locations and approach angles were examined. It was found that at all locations, the measured velocities never exceeded the freestream velocity of 190.2 ft/s (58 m/s; 130 mph), with the maximum Re occurring above the overpass and between the I-beams. A theoretical maximum pressure drop for the tornado center was calculated to be 0.5 psi for an Enhanced Fujita 2 scale tornado and compared with the highest pressure drop of 0.278 psi, determined from the experiments. Calculated pressure coefficients Cp were mostly <0 and some close to one dynamic head less than ambient. The drag coefficients Cd remain primarily in the laminar region with later transition to turbulence. Using experimental data from the literature, drag forces on an average size man in crouching and laying positions between the overpass I-beams section were determined to be a maximum of 31 lbf.
Article
On March 3, 2019, an EF4 tornado devastated the rural Alabama communities of Beauregard and Smith Station, killing 23 people and causing direct injuries to another 97. This storm was unusually devastating, with twice the predicted casualty rate based on the tornado’s power, the impacted population, and impacted housing stock. In this paper, we apply qualitative methods from anthropology, geography, and planning to better understand the social context of this unusually devastating tornado. Recognizing that there are multiple formulations of the problem of disasters, we aim to highlight how interdisciplinary qualitative research can deepen our understanding of tornado disasters. Combining policy analysis, political economic critique, and ethnographic description, we seek to showcase how qualitative research enables us to interrogate and reimagine the problem of disasters. Rather than simply juxtaposing qualitative and quantitative methods, we emphasize how the heterogeneity of qualitative research methods can strengthen interdisciplinary research projects by generating dialogue about the multiple contexts relevant to understanding a social problem. While problem definition remains a central challenge to establishing a dialogue between anthropology and social work, here, we intend to extend this discussion to larger interdisciplinary collaborations. Situating the issue of problem formation within a broader ecology of qualitative inquiry, we highlight how dialogue about problem definition can, itself, produce meaningful insights into how we understand disasters.
Article
The coronavirus disease 2019 (COVID-19) pandemic resulted in unprecedented challenges that dramatically affected the way of life in the United States and globally in 2020. The pandemic also made the process of protecting individuals from tornadoes more challenging, especially when their personal residence lacks suitable shelter, particularly for residents of mobile homes. The necessity of having to shelter with other families – either in a public shelter or at another residence – in order to protect themselves from a tornado threat conflicted with the advice of public health officials who recommended avoiding public places and limiting contact with the public to minimize the spread of COVID-19. There was also a perception that protecting against one threat could amplify the other threat. A survey was undertaken with the public to determine the general viewpoint to see if that was indeed the case. The results found that it was possible to attenuate both threats provided that careful planning and actions were undertaken. Understanding how emergency managers should react and plan for such dual threats is important to minimize the spread of COVID-19 while also maintaining the safety of the public. As there was no precedence for tornado protection scenarios amidst a pandemic, both short-term and long-term recommendations were suggested which may also be useful in future pandemic situations.
Article
Nocturnal tornadoes are challenging to forecast and even more challenging to communicate. Numerous studies have evaluated the forecasting challenges, but fewer have investigated when and where these events pose the greatest communication challenges. This study seeks to evaluate variation in confidence among US residents in receiving and responding to tornado warnings by hour-of-day. Survey experiment data comes from the Severe Weather and Society Survey, an annual survey of US adults. Results indicate that respondents are less confident about receiving warnings overnight, specifically in the early morning hours (12 AM to 4 AM local time). We then use the survey results to inform an analysis of hourly tornado climatology data. We evaluate where nocturnal tornadoes are most likely to occur during the time frame when residents are least confident in their ability to receive tornado warnings. Results show that the Southeast experiences the highest number of nocturnal tornadoes during the time period of lowest confidence, as well as the largest proportion of tornadoes in that time frame. Finally, we estimate and assess two multiple linear regression models to identify individual characteristics that may influence a respondent’s confidence in receiving a tornado between 12 AM and 4 AM. These results indicate that age, race, weather awareness, weather sources, and the proportion of nocturnal tornadoes in the local area relate to warning reception confidence. The results of this study should help inform policymakers and practitioners about the populations at greatest risk for challenges associated with nocturnal tornadoes. Discussion focuses on developing more effective communication strategies, particularly for diverse and vulnerable populations.
Article
Residents of the Oklahoma City metropolitan area are frequently threatened by tornadoes. Previous research indicates that perceptions of tornado threat affect behavioral choices when severe weather threatens, and as such are important to study. In this paper, we examine the potential influence of tornado climatology on risk perception. Residents across central Oklahoma were surveyed about their perceptions of tornado proneness for their home location, and this was compared to the local tornado climatology. Mapping and programming tools were then used to identify relationships between respondents’ perceptions and actual tornado events. Research found that some dimensions of the climatology, such as tornado frequency, nearness, and intensity have complex effects on risk perception. In particular, tornadoes that were intense, close, and recent had the strongest positive influence on risk perception, but weaker tornadoes appeared to produce an “inoculating” effect. Additional factors were influential, including sharp spatial discontinuities between neighboring places that were not tied to any obvious physical feature or the tornado climatology. Respondents holding lower perceptions of risk also reported lower rates of intention to prepare during tornado watches. By studying place-based perceptions, this research aims to provide a scientific basis for improved communication efforts before and during tornado events, and for identifying vulnerable populations.
Article
Previously published claims of large regional (northern vs southern states) differences in risks of fatality associated with tornadoes in the United States are reexamined. This new study extends earlier claims to include 1) data from a much longer time frame, 2) injuries as well as fatalities, and 3) more precise estimates of meteorological features of tornado events (specifically, a precise calculation of daytime vs nighttime and pathlength). The current study also includes formal mediation analyses involving variables that might explain regional differences. Results indicate that significant increases in the risk of fatality and injury do occur in southern states as compared with northern states. Mediation models show that these regional differences remain significant when meteorological factors of nocturnal occurrence and pathlength are included. Thus, these meteorological factors cannot explain regional differences in risk of fatality and injury, a failure that is unlikely to reflect a lack of data or a lack of precision in the measurement of potential mediators.
Article
Full-text available
During the evening hours of 3 May 1999, 58 tornadoes occurred in Oklahoma. One tornado reached F5 intensity and left a widespread path of death, injury, and destruction in and around the Oklahoma City metropolitan area. Other communities across the state were also affected. Data on persons who died or were injured were collected from medical examiner reports, hospital medical records, and follow-up questionnaires. In addition, community surveys were conducted in the damaged areas, and a random telephone survey of people in the Oklahoma City area was conducted. A total of 45 persons died, and 645 survivors were treated at a hospital for injuries/health conditions related to the tornadoes. Detailed analyses were conducted on the 40 deaths and 512 nonfatal injuries that resulted directly from the tornadoes. The average age of persons who died was significantly higher than that of injured survivors. Almost three-fourths (74%) of survivors were treated in an emergency department and released; 26% were hospitalized, with an average stay of 7 days. The most common kinds of injuries were soft-tissue injuries (cuts, bruises, and scrapes), fractures/dislocations, and brain injuries. The most common causes of injuries for survivors were being hit by flying/falling debris and being picked up or blown by the tornado. Probable causes of death included multiple injuries (50%), head injuries (23%), chest trauma (18%), and traumatic asphyxia (10%). The proportion of deaths among injured persons was higher in mobile homes, apartment buildings, and outdoors than in houses. The vast majority of persons received television warnings; few persons received warnings by weatherband radio. Survey data indicated that since the tornadoes there had been only a slight increase in the proportion of persons with a storm shelter on their home premises and in the proportion of persons who know the location of the nearest shelter. Effective preparedness plans should be developed in advance of situations with tornadic potential and implemented in a timely manner when warnings are issued. The accessibility and knowledge of the location of shelters also needs to be increased. Warning systems should be integrated and duplicative, and they should include traditional methods and new technology.
Article
Full-text available
Residents of homes that sustained F4 or F5 damage in the deadliest of the 3 May 1999 tornadoes were surveyed to determine their responses to the tornado warning reasons for their responses, and relative injury rates. There were 190 people in 65 surveyed houses at the time that warnings were issued. Television was the most commonly cited source of the warning (89%), followed by a telephone call (37%), sirens (37%), and AM/FM radio (25%), and 55% received the warning from more than one source. Nearly one-half (47%) of the residents fled their homes before the tornado struck. Of those who fled, 65% went to a tornado shelter, of whom 70% ran to the shelter (median distance 30 m) and 30% drove to the shelter (median distance 4.8 km). About one-half (53%) of those who fled their homes left in a vehicle. None of those who fled their homes, by foot or by vehicle, were injured. Of those who stayed in the home, 39% sought shelter in a bathroom, 38% in a closet, 9% in a hallway and 15% in other rooms. Reasons for not leaving included believing the storm would not strike their house. believing it was too late or too dangerous to leave, having no transportation available, or having no alternative shelter available. Thirty percent of those who remained in their homes were injured and 1% killed. The rate of serious injury was not significantly different for those in a closet (14%), hallway (20%), or bathroom (23%). Tornado preparedness and warning programs should recognize that long tornado warning lead times and street-level television coverage allow residents to make reasoned decisions to minimize risk and that those decisions may include driving out of the path of the tornado.
Article
Full-text available
A dataset of killer tornadoes is compiled and analyzed spatially in order to assess region-specific vulnerabilities in the United States from 1880 to 2005. Results reveal that most tornado fatalities occur in the lower-Arkansas, Tennessee, and lower-Mississippi River valleys of the southeastern United States-a region outside of traditional "tornado alley." Analysis of variables including tornado frequency, land cover, mobile home density, population density, and nocturnal tornado probabilities demonstrates that the relative maximum of fatalities in the Deep South and minimum in the Great Plains may be due to the unique juxtaposition of both physical and social vulnerabilities. The spatial distribution of these killer tornadoes suggests that the above the national average mobile home density in the Southeast may be a key reason for the fatality maximum found in this area. A demographic analysis of fatalities during the latter part of the database record illustrates that the middle aged and elderly are at a much greater risk than are younger people during these events. Data issues discovered during this investigation reveal the need for a concerted effort to obtain critical information about how and where all casualties occur during future tornado and hazardous weather events. These new, enhanced data, combined with results of spatially explicit studies exploring the human sociology and psychology of these hazardous events, could be utilized to improve future warning dissemination and mitigation techniques.
Article
Full-text available
The United States has no comprehensive national warning strategy that covers all hazards in all places. Instead, public warning practices are decentralized across different governments and the private sector. Uneven preparedness to issue warnings exists across local communities; hence, people are unevenly protected from the surprise onset of natural disasters. Without changes in this situation, inequalities will grow larger, and the gains made in saving lives over the past decades may well be reversed. Since the first assessment of research on natural hazards was completed in 1975, there have been significant improvements in forecasts and warnings for some hazards but only marginal improvements for others. Forecasts for floods, hurricanes, and volcanic eruptions have improved most significantly, and public dissemination of warnings has improved the most for hurricanes. However, a 100% reliable warning system does not exist for any hazard.
Article
Full-text available
The goal of this research was to examine the relative safety and stability of stationary motor vehicles exposed to severe winds. The focus was on private passenger vehicles. 1) The behavior of two instrumented storm-chase vehicles that were exposed to severe winds, 2) the behavior of 291 vehicles exposed to a tornado, and 3) the wind speed required to upset a sedan and a minivan exposed to winds in a wind tunnel were studied. A wind as strong as 47 m s1 (105 mph) has been measured by a storm-chase pickup truck and 44 m s1 (98 mph) by a storm chase sedan. The vehicles were not adversely affected by the wind. Also studied were 291 vehicles parked outdoors at homes struck by tornadoes, and the behavior of the vehicles was compared to the F-scale damage to the house. At sites with F1 or F2 damage, 72% of the vehicles were not moved by the wind and 96% were not tipped over. At sites with F3 or F4 damage, 50% were not moved by the wind and 82% were not tipped over. Wind tunnel tests on a sedan and minivan showed they were most vulnerable to upset (lifting of one tire from the ground) with wind directions near 45° and 135°, as measured from the front. When modeled with 5° of suspension tilt to the side, the sedan was found to be upset at wind speeds of 51-67 m s1 (115-150 mph), and the minivan was upset at wind speeds of 58-80 m s1 (130-180 mph). Although an underground shelter or sturdy building offer the best protection from severe winds, it is found that a vehicle may be a relatively stable place and may be safer than a mobile home or the outdoors. These findings may warrant changes to public recommendations made during tornado warnings and other severe storm situations.
Article
Full-text available
Manufactured or mobile homes represent a fast growing portion of the housing market but are particularly vulnerable to tornadoes. In the US over 40% of tornado fatalities occur in mobile homes even though they comprise about 8% of US housing units. We examine the market for tornado shelters in manufactured home parks in Oklahoma. Almost 60% of parks in the state have shelters, with 90% of the shelters underground. Parks with shelters are not concentrated in urban areas but spread across the state, with parks with shelters in 32 counties. We find that rents for lots in parks with shelters are 5–8% higher, which generates sufficient revenue to approximately pay for shelters, but the point estimate is statistically significant in only one specification.
Article
The goal of this research was to examine the relative safety and stability of stationary motor vehicles exposed to severe winds. The focus was on private passenger vehicles. 1) The behavior of two instrumented storm-chase vehicles that were exposed to severe winds, 2) the behavior of 291 vehicles exposed to a tornado, and 3) the wind speed required to upset a sedan and a minivan exposed to winds in a wind tunnel were studied. A wind as strong as 47 m s-1 (105 mph) has been measured by a storm-chase pickup truck and 44 m s-1 (98 mph) by a storm chase sedan. The vehicles were not adversely affected by the wind. Also studied were 291 vehicles parked outdoors at homes struck by tornadoes, and the behavior of the vehicles was compared to the F-scale damage to the house. At sites with F1 or F2 damage, 72% of the vehicles were not moved by the wind and 96% were not tipped over. At sites with F3 or F4 damage, 50% were not moved by the wind and 82% were not tipped over. Wind tunnel tests on a sedan and minivan showed they were most vulnerable to upset (lifting of one tire from the ground) with wind directions near 45° and 135°, as measured from the front. When modeled with 5° of suspension tilt to the side, the sedan was found to be upset at wind speeds of 51-67 m s-1 (115-150 mph), and the minivan was upset at wind speeds of 58-80 m s-1 (130-180 mph). Although an underground shelter or sturdy building offer the best protection from severe winds, it is found that a vehicle may be a relatively stable place and may be safer than a mobile home or the outdoors. These Findings may warrant changes to public recommendations made during tornado warnings and other severe storm situations.
Article
This research examines sources of information for flash floods in two large metropolitan areas, Denver, CO, and Austin, TX. Previous research has noted that information delivery systems for weather forecasts are geared toward the cultural majority and suggests that inadequate warnings are a primary contributor to deaths and injuries from hazards. This investigation used chi-square analysis to determine the prime warning source preferences and preferred time of day for receiving different media. Results indicate that successful warning messages need to be targeted toward specific sub-populations if the warning is to be received, understood, and responded to properly.