In the United States,
hurricanes are an annual threat to the eastern and gulf coastal states. During
the period from 1900 to 2002, more than 166 direct hits by hurricanes
(including 65 major hurricanes) occurred on the mainland coastline, as recorded
in Table 1–1. The table enumerates the total number of hurricane
strikes experienced by the eastern and gulf coastal states.
Table
1–1. U.S. Mainland
Hurricane Strikes by State (1900 – 2002)
|
Area
|
Major Hurricanesa
|
All Hurricanes
|
|
Texas
|
16
|
37
|
|
Louisiana
|
12
|
27
|
|
Mississippi
|
6
|
9
|
|
Alabama
|
5
|
10
|
|
Florida
|
24
|
60
|
|
Georgia
|
0
|
5
|
|
South
Carolina
|
4
|
14
|
|
North
Carolina
|
11
|
27
|
|
Virginia
|
1
|
5
|
|
Maryland
|
0
|
2
|
|
Delaware
|
0
|
0
|
|
New
Jersey
|
0
|
1
|
|
New
York
|
5
|
9
|
|
Connecticut
|
3
|
8
|
|
Rhode
Island
|
3
|
5
|
|
Massachusetts
|
2
|
6
|
|
New
Hampshire
|
0
|
2
|
|
Maine
|
0
|
5
|
|
Totalb
|
65
|
166
|
Of these, the 30
most costly and deadliest hurricanes are listed in Table 1–2, which identifies the year; hurricane by name, state,
or location; category; costs incurred; and number of deaths. The Year 2000 Cost
column lists the costs in inflation-adjusted Year 2000 dollars; the Normalized
Cost column lists the expected cost if a similar hurricane hit the same
location today. The 30 costliest hurricanes are estimated
to have caused a cumulative $132 billion of damage, and the average expected
cost if one of these hurricanes occurred today is about $14 billion. The
cumulative number of deaths for the 30 deadliest hurricanes is almost 15,000.
Table
1–2. The Costliest and Deadliest U.S. Hurricanes
(1900 – 2000)
|
Year
|
Hurricane
|
Category
|
Cost
|
Deathsa
|
|
Year 2000a
|
Normalizeda
|
|
1900
|
(N TX)
|
4
|
928
|
32,090
|
8000
|
|
1906
|
(SE FL)
|
2
|
|
|
164
|
|
1906
|
(MS, AL, NW FL)
|
3
|
|
|
134
|
|
1909
|
(LA)
|
4
|
|
|
350
|
|
1909
|
(TX)
|
3
|
|
|
41
|
|
1910
|
(SW FL)
|
3
|
|
|
30
|
|
1915
|
(N TX)
|
4
|
1,544
|
27,190
|
275
|
|
1915
|
(LA)
|
4
|
|
|
275
|
|
1918
|
(SW LA)
|
3
|
|
|
34
|
|
1919
|
(S TX)
|
4
|
|
6,448
|
600
|
|
1926
|
(SE FL, AL)
|
4
|
1,738
|
87,167
|
243
|
|
1928
|
(SE FL)
|
4
|
|
16,631
|
1836
|
|
1932
|
(TX)
|
4
|
|
|
40
|
|
1933
|
(S TX)
|
3
|
|
|
40
|
|
1935
|
(FL Keys)
|
5
|
|
|
408
|
|
1938
|
(New England)
|
3
|
4,749
|
20,046
|
600
|
|
1940
|
(GA, SC, NC)
|
2
|
|
|
50
|
|
1944
|
(NE US)
|
3
|
1,221
|
7,790
|
390
|
|
1944
|
(SW FL)
|
3
|
|
20,331
|
|
|
1945
|
(SE FL)
|
3
|
|
7,611
|
|
|
1947
|
(SE FL, LA, AL)
|
4
|
930
|
10,015
|
51
|
|
1949
|
(SE FL)
|
3
|
|
7,038
|
|
|
1954
|
Carol (NE US)
|
3
|
3,134
|
10,929
|
60
|
|
1954
|
Hazel (SC, NC)
|
4
|
1,911
|
8,486
|
95
|
|
1955
|
Diane (NE US)
|
1
|
5,541
|
12,335
|
184
|
|
1957
|
Audrey (SW LA, NW TX)
|
4
|
|
|
390
|
|
1960
|
Donna (FL, Eastern US)
|
4
|
2,408
|
16,631
|
50
|
|
1961
|
Carla (N TX, CE TX)
|
4
|
2,551
|
8,522
|
46
|
|
1964
|
Dora (NE FL)
|
2
|
1,541
|
3,747
|
|
|
1964
|
Hilda (CE LA)
|
3
|
|
|
38
|
|
1964
|
Cleo (SE FL)
|
2
|
|
2,936
|
|
|
1965
|
Betsy (SE FL, SE LA)
|
3
|
8,517
|
14,990
|
75
|
|
1967
|
Beulah (S TX)
|
3
|
1,113
|
|
|
|
1969
|
Camille (MS, SE LA,
VA)
|
5
|
6,992
|
13,219
|
256
|
|
1970
|
Celia (S TX)
|
3
|
2,016
|
4,024
|
|
|
1972
|
Agnes (NW FL, NE US)
|
1
|
8,603
|
12,904
|
122
|
|
1975
|
Eloise (NW FL)
|
3
|
1,489
|
|
|
|
1979
|
Frederic (AL, MS)
|
3
|
4,965
|
7,587
|
|
|
1983
|
Alicia (N TX)
|
3
|
3,422
|
4,890
|
|
|
1985
|
Juan (LA)
|
1
|
2,419
|
2,892
|
|
|
1985
|
Elena (MS, AL, MW FL)
|
3
|
2,016
|
|
|
|
1985
|
Gloria (Eastern US)
|
3
|
1,451
|
|
|
|
1989
|
Hugo (SC)
|
4
|
9,740
|
11,307
|
|
|
1991
|
Bob (NC, NE US)
|
2
|
2,005
|
|
|
|
1992
|
Andrew (SE FL, SE LA)
|
4
|
34,955
|
39,896
|
|
|
1994
|
Alberto (NW FL, GA, AL)
|
TS
|
|
|
30
|
|
1995
|
Opal (NW FL, AL)
|
3
|
3,521
|
3,617
|
|
|
1996
|
Fran (NC)
|
3
|
3,670
|
3,735
|
|
|
1998
|
Georges (FL Keys, MS, AL)
|
2
|
2,495
|
|
|
|
1999
|
Floyd (NC)
|
2
|
4,667
|
4,680
|
56
|
Clearly, the
potential still exists for a hurricane to cause tremendous damage to coastal
regions today.
One method used
to reduce the number of deaths, and to a lesser degree, costs caused by
hurricanes, is to evacuate those areas that might be impacted. The importance
of this approach has grown with recent advances in the ability of forecasters
to more accurately predict the track of a hurricane, thus reducing the number
of unnecessary evacuations. However, hurricane evacuations remain
difficult transportation activities to manage.
Hurricane Floyd
was a large hurricane that peaked in intensity as a Category 4 hurricane in the
Bahamas.
Although it dropped in intensity, weakening to Category 2 by the time it
reached landfall in North Carolina, it’s large
size resulted in evacuations of roughly 3 million people from a 4-state area
consisting of parts of Florida, Georgia, North Carolina,
and South Carolina.
This large-scale evacuation resulted in traffic jams across the affected
regions as motorists flooded the highways. For example, travel time between Charleston and Columbia,
South Carolina, normally only
2-1/2 hours, increased to as much as 18 hours during the period of peak
congestion.
This breakdown in
the effectiveness of the transportation system during the evacuation spurred a
renewed interest in evacuation planning both within individual states and at
the Federal level. For example, many states recognized the advantages of using
lane reversals (or contraflow) to facilitate evacuations. State traffic and
emergency management officials have since modified their evacuation plans and
the highways as well to better support lane reversals in the future. State
emergency management officials also took steps to improve coordination of
evacuation and evacuation planning activities between states; monitor and
control the transportation infrastructure during evacuations; and disseminate
information to the public.
In May 2002, the
Federal Highway Administration (FHWA) funded grants to nine southeastern states
(Texas, Louisiana,
Mississippi, Alabama,
Florida, Georgia,
South Carolina, North
Carolina, and Virginia)
to improve transportation operations as part of their emergency management
program for hurricanes evacuations. Following are brief descriptions of the
activities each state proposed to fund with these grants.
·
Texas.
Before receiving this grant, Texas had
recently completed hurricane evacuation plans that included contraflow on parts
of I-37 from Corpus Christi to San Antonio to increase traffic flow. With
this grant, nine additional methods of increasing evacuation traffic flow from Corpus Christi were
evaluated.
·
Louisiana.
Louisiana was
in the process of deploying traffic count stations to facilitate real-time
monitoring of evacuation route traffic during hurricane evacuations. These
stations, which include both traffic and hydrology instruments, were being
deployed as a collaborative effort of Louisiana Department of Transportation
and Development (LA DOTD) and U.S. Geological Survey (USGS) that use an
existing USGS hardware platform and satellite communication system to transmit data.
Louisiana
used its grant to extend this program to include additional detector stations.
·
Mississippi.
Mississippi recently confirmed plans to
implement contraflow on I-59 to facilitate hurricane evacuations, particularly
evacuation from New Orleans.
With the grant, Mississippi
produced and distributed a brochure in the fall of 2003 that explains hurricane
evacuation procedures, and in particular, the use of contraflow on I-59 during
hurricanes.
·
Alabama.
Alabama
developed a reverse lane plan for portions of I-65 in 2000, and used the
Federal grant to develop and implement a public information program concerning
the lane-reversal plans during a hurricane evacuation.
·
Florida.
Florida used
the Federal grant to help develop a geographic information system (GIS)-based
hurricane evacuation software system known as HEADS UP (Hurricane Evacuation
Analysis and Decision Support Utility Program). This program extends the
capabilities of ETIS (Emergency Transportation Information System) by including
additional data. In the future, ETIS will include a model that will compute
dynamic clearance times.
·
Georgia.
Georgia
first developed a contraflow plan in 1995, and has recently updated the plan.
With the Federal grant it received, Georgia
produced evacuation route maps and distributed those maps to Welcome Centers, Georgia
Department of Transportation (GDOT) district offices, and Georgia State Patrol
posts in the southern half of the state.
·
South
Carolina. South Carolina has recently revised its
evacuation plans to consider broader effects of a hurricane evacuation and to
extend the evacuation area further inland. With its Federal grant, South Carolina printed
and distributed hurricane evacuation route maps and hurricane guides that
included information on these updates.
·
North
Carolina. North Carolina identified a need for
real-time traffic information to better monitor hurricane evacuation
activities. North Carolina
plans to use the Federal grant to deploy traffic-monitoring detectors at key
locations along hurricane evacuation routes.
·
Virginia.
Virginia
applied its Federal grant to develop an abbreviated clearance time model that
would be easier for a non-traffic planner to run and applied this model to the
Hampton Roads area.
The purpose of
this evaluation report is to draw some lessons learned from the activities
pursued using the Federal grants that were received by the nine states. Two
approaches were pursued for doing so. The first approach expands on the brief
descriptions of the state activities listed in Section 1.2, and also notes
other activities called out by those states in recent presentations at the
Transportation Evacuation Planning and Operations Workshop held in New Orleans, Louisiana
on April 14 – 15, 2003. The primary purpose of this review is to gain insight
into the areas that the states deem important for supporting hurricane
evacuations by documenting what their traffic and emergency management
officials are pursuing. This information was synthesized to identify lessons
learned that might have broad application among the states and is documented in
Section 2.0.
The second
approach focuses on the Louisiana partnership
with the USGS to deploy Hydrowatch stations
(identified by Louisiana
as “Information Stations”) that also monitor traffic. It is believed that this
type of partnership might be a cost-effective approach for deploying traffic
count stations at remote locations in many states. This portion of the
evaluation is documented in Section 3.0 of this report, and provides a case
study of the Information Station deployment and how similar deployments could
occur in other states.
Although each of
the nine states received equal funding, individual state activities were quite
different. The activities depended on the perceived needs in those states, and
how well the states could use the FHWA funding to complement existing hurricane
preparedness activities. In general, the type of activities pursued can be
divided into two groups:
·
The states of Mississippi,
Alabama, Georgia,
and South Carolina
used the funds to support public information activities. These activities
concentrated on providing information to the public on evacuation routes and
contraflow that might be used on some of those routes.
·
The other five states, Texas,
Louisiana, Florida,
North Carolina, and Virginia, used the funds to support
technical activities related to hurricane evacuations.
This section
provides information on the activities pursued by each of these states, with
the activities of the states pursuing public information activities described
in section 2.1 and those pursuing technical activities described in
section 2.2.
Each of the
following four states used the FHWA funds to support public information
activities.
Unlike many of
the other states that received hurricane evacuation grant money, Mississippi has a
relatively small, less populated coastal region, and evacuation routes from
these regions are expected to be sufficient. However, there is significant
potential for large cross-border evacuations from New
Orleans, and to a lesser extent, from Alabama. In fact, the potential for large
cross-border evacuations from New Orleans on
I-59 creates the potential need for Mississippi
to implement contraflow on I-59 to better support evacuation of Louisiana residents.
Lesson Learned 1 – Coordinate
plans that cross state lines.
|
However, these
plans were put on hold in October 2002 as Mississippi
wrestled with the issue of how to ensure appropriate services to residents of Mississippi while supporting the evacuation of Louisiana residents into Mississippi. By June 2003, Mississippi
and Louisiana had reached a revised agreement
for Mississippi to use contraflow on I-59 in Mississippi to support Louisiana’s
evacuation when contraflow was implemented on I-59 in Louisiana.
To educate the
public on hurricane evacuation procedures, including contraflow on I-59, Mississippi released a
new hurricane evacuation brochure in September of 2003. Mississippi elected to devote its grant
money towards a publicity campaign to increase awareness of and confidence in
these brochures. This publicity campaign consisted of billboard and radio
advertisements, coverage by local media, and insertion of the brochures in
local newspapers. The benefit of this grant usage was a high degree of public
awareness that resulted in the distribution of the entire initial run of
100,000 brochures was. A reprint of 50,000 additional brochures is being
produced to meet the excess demand.
Lesson Learned 2 – Share
information.
|
In addition, Mississippi used a small
amount of the FHWA grant to help sponsor a symposium that gathered participants
from multiple states to discuss emergency management practices. The symposium
participants all agreed that future conferences of the same nature would be
extremely important.
In 2000, Alabama developed a contraflow plan for I-65 from North
of Mobile to just south of Montgomery.
While implementation and operation exercises have helped confirm that the plan
can quickly and safely set up this section for contraflow, Alabama felt that there was a need to
educate the public on contraflow operations so that they could participate more
easily – and quickly – with a contraflow-assisted evacuation. Alabama used the FHWA grant money to
evaluate various methods of disseminating information about the I-65 contraflow
plan. This evaluation led to the use of reversed direction signing, variable
message signs, Alabama Emergency Radio, and annual implementation exercises as
the primary means of educating the public at this time. After this evaluation, Alabama used the
remaining FHWA funds (along with other funds) to purchase two Highway Advisory
Radio (HAR) units with permanent quick disconnect antennas and advanced
notification signage. The signs and HAR units will be used on both ends of the
contraflow route on I-65 that runs from Mobile
to Montgomery.
While these signs
have not been used during a hurricane evacuation, Alabama believes that the use of HAR at the
contraflow entry and exit points will significantly improve traffic flow at
these evacuation decision points.
Georgia, like
the other coastal, hurricane-prone states, continues to improve its
preparations for hurricane evacuations. Some of the key features of these
preparations are:
·
Signage improvements. Defining and
improving hurricane evacuation routes with signage on those routes, including
contraflow plans for I-16 with drop gate barriers on the east bound entrance
ramps to prevent access in the east bound direction.
·
Improved traffic flow. Planning
activities to improve traffic flow on evacuation routes, including
pre-evacuation clearing of Interstate highway shoulders, manned push button
traffic control at important signal-controlled intersections along evacuation
routes, and coordination with railroads to help ensure that trains do not block
evacuation routes.
·
Improved evacuation routes. Implementing
improvements to evacuation routes to increase capacity (e.g., by adding lanes)
and decrease the likelihood that the road is blocked (e.g., by raising the road
elevation to avoid flooding).
·
Expanded traveler information. Expanding
traveler information during evacuations through portable HAR, variable message
signs, and cooperative agreements with Georgia Public Radio stations.
·
Expanded traveler assistance. Expanding
traveler assistance by the Highway Emergency Response Operator (HERO) incident
response vehicles.
Lesson Learned 3 – Educate the public about contraflow.
|
Despite all of
these preparations, Georgia
noted that the effectiveness of these preparations is limited if evacuees are
not aware of the available resources. Consequentially, Georgia used
the grant money to create information sheets and posters that contained information
on evacuation routes (including maps), what to do, who to call, and other
information and things to remember when evacuating. One example was a
suggestion to pack supplies and to fill the car up with gas before evacuating.
The information sheets where double-sided, letter-sized sheets that were
distributed to Georgia
State patrol, local GDOT
offices, and local Georgia Emergency Management Agency personnel for
redistribution to the public. The posters, which were larger versions of the
information sheets, were positioned at state patrol offices and Georgia rest
areas and welcome centers.
Since the state
of Georgia
did not sustain a hurricane this past season, it was difficult to identify
specific benefits that were achieved. However, citizens did mention that they
appreciated the fact that all the pertinent hurricane evacuation information
was on a single sheet. Georgia
believes that many citizens kept this sheet handy to guide them in the case
that a hurricane occurred.
2.1.4.
South Carolina
During Hurricane
Floyd, South Carolina
noted several problems with the hurricane evacuation routes, including the
following:
·
Conflicting needs. The evacuation routes
were developed from individual scenarios for each population area, so there was
potential for conflicting needs during more wide scale evacuations.
·
Traffic impediments. Evacuation routes
from different areas sometimes crossed, which impeded evacuating traffic at
those points.
·
Insufficient evacuation length. The
evacuation routes only reached 50 miles inland.
Recently, the
evacuation routes were revised to address these problems, which created a need
to prepare better evacuation maps. South
Carolina used the grant funds to print and distribute
hurricane evacuation route maps and hurricane guides and to make similar maps
and guides available from its Website.
Each of the
following five states used the funds to support technical activities related to
hurricane evacuations.
Texas has a large coast with 22 coastal
counties that are subject to a significant risk from hurricanes. For most of
these counties, a combination of relatively low populations and a good road
network make evacuations relatively quick – typically less than 10 hours. The
exceptions, listed in Table 3, are several of the more densely populated
counties and counties that include particularly remote locations. Table 3
identifies some of the Texas
counties and areas affected by hurricane Categories 1 – 5, and the estimated
evacuation time in hours residents would take to reach safer locations.
Table
2–1.
Estimated Evacuation Times (in Hours) for Some Texas Coastal Counties
|
Area
|
Hurricane Category
|
|
1
|
2
|
3
|
4
|
5
|
|
Cameron County
(Brownsville)
|
15
|
21
|
28
|
32
|
33
|
|
Nueces County
(Corpus Christi)
|
11
|
20
|
28
|
31
|
32
|
|
San Patricio
|
8
|
11
|
15
|
17
|
18
|
|
Brazoria
|
7
|
9
|
13
|
15
|
17
|
|
Galveston County
(Galveston)
|
14
|
20
|
28
|
32
|
33
|
|
Harris County
(Houston)
|
14
|
20
|
28
|
32
|
33
|
|
Chambers
|
10
|
13
|
17
|
19
|
19
|
|
Orange County
(Orange)
|
14
|
20
|
29
|
33
|
34
|
Since exact
hurricane landfall location is often determined less than 24 hours from
landfall, it is important to take steps to decrease evacuation clearance times.
Consequentially, a contraflow evacuation plan was developed in 2000 for parts
of I-37 to support evacuation from Nueces
County. The development
of this plan brought to light several difficulties (e.g., reluctance of
the Department of Public Safety to use contraflow, manpower demands of
implementing contraflow, and dangers of unofficial entry to contraflow lanes
from rural frontage roads). Chief among these difficulties was the fact that
much of the contraflow capacity would remain unused during an evacuation due to
bottlenecks in the system leading up to the contraflow area.
To address these
difficulties, Texas chose to conduct a study
of the following six alternatives for improving evacuations from Nueces County
during a hurricane:
Lesson Learned 4 – Locate your bottlenecks.
|
·
Land addition. Analyze adding a lane on
the I-37 evacuation route from Corpus
Christi. Analysis indicates that by reducing the
inside shoulder to 4 feet and the lane widths to 11 feet, the current
3-lane northbound I-37 could be restriped to 4 lanes without significantly
impacting vehicle speeds. The added lane would help remove the potential
bottleneck, thereby increasing access to the contraflow portion of I-37.
·
Shoulder access. Assess using the
existing I-37 shoulder as a hurricane evacuation lane. The primary concerns for
this approach includes the lack of a shoulder on which to move disabled
vehicles; the potential for confusion at exit ramps; and impediments often
found in the shoulder (e.g., rumble strips, raised pavement markings at
exits).
·
Other potential entry points. Identify
other potential contraflow entry points. The current contraflow entry point is
just north of the I-37 and Route 77 interchange. The Nueces
River Bridge
just south of this point is the only nearby crossing that could allow emergency
vehicles to travel south to Corpus
Christi.
Extending the contraflow region south across that bridge was not deemed
feasible.
·
Signage needs. Identify signage needs and
other impediments to smooth traffic flow on other evacuation routes from Corpus Christi. An
alternate route on Route 43 was identified as a feasible evacuation route and a
signage plan for this route was developed.
·
Additional technologies. Consider
technologies to provide real-time traffic data on I-37 during an evacuation.
·
Improve hurricane evacuation map. Improve
the hurricane evacuation map for the Corpus
Christi area.
Lesson Learned 5 – Leverage
the USGS streamgaging programming.
|
In Louisiana, the need
existed for better traffic monitoring during evacuations. Because of the high
potential for flooding to block routes, there needed to be better monitoring of
water levels near roads. Since USGS also has an interest in monitoring water
levels and the Louisiana Office of Emergency Planning (LOEP, now called the
Louisiana Office of Homeland Security Preparedness), has an interest in flood
detection, the Louisiana Department of Transportation and Development (LA DOTD)
developed a collaboration with these agencies to deploy combined traffic and
hydrology monitoring stations, which they called Information Stations. This
approach pooled funding from three agencies to provide a cost effective
approach for LA DOTD to obtain real-time traffic and road flooding information.
The primary
benefits of the project are that LA DOTD now has access to near real time
traffic information from seven Information Stations located along key
evacuation routes near Lake Pontchartrain.
Because of the success of this initial deployment, LA DOTD now plans to deploy
an additional fifteen stations in the same area. The collaboration with USGS
has also resulted in plans for USGS to provide LA DOTD with flood alarms for
flood-prone roads on which hydrowatch stations are located.
For more
information on the Louisiana Information Station project, see section 3.0.
Florida has in place a number of measures to
help support hurricane evacuations. Five contraflow plans have been developed and
information has been publicized about these routes, including a significant
amount of information available online from the Florida Division of Emergency
Management. Evacuation routes in Florida
are signed with flip up signs used for hurricane-specific traveler information
(e.g., shelter locations). A series of radio stations provide coverage for
disseminating traveler information during hurricanes. Florida has also equipped these contraflow
routes with traffic counters and will be upgrading many of these traffic count
stations to include traffic video cameras.
With all of these
elements in place already, Florida
chose to apply the grant funding, combined with other funding, to help
integrate some of these resources by developing a hurricane evacuation decision
support software tool called HEADS UP. Planned features for HEADS UP include:
·
Data links. Currently, HEADS UP links to
the following elements: shelter status (e.g., location, capacity, current
population); road closure status; traffic counts (through a link to a
Florida DOT site); road construction and real-time traffic (via http://www.myflorida.com/); traffic
incidents (via http://www.fhp.state.fl.us/);
and weather information (via http://fawn.ifas.ufl.edu/).
For data that is needed for HEADS UP calculations, there is an active effort to
consolidate that data into the HEADS UP database rather than just link to it.
For example, one of the Phase 2 objectives is to eliminate the need to enter
data in both HEADS UP and ETIS by creating software tools that will keep the
two systems synchronized.
·
Time stepped calculations. The HEADS UP
traffic model is a time-stepped model that uses a mixture of measured and
estimated parameters to predict current and future traffic conditions.
·
Calculate shut down times for areas. If
traffic counters indicate traffic is backed up on a link and a queue has
formed, HEADS UP will estimate the amount of time it will take to clear that
link. This action will help to devise alternate route plans.
·
Sheltering calculations. HEADS UP will
calculate information necessary to help estimate sheltering requirements, such
as the number of evacuees expected to select a site as a final destination and
the number of pass-through evacuees.
·
Compare to actuals. HEADS UP will compare
predicted/estimated values (e.g., traffic counts) to measured values so
that estimates can be improved. For example, if HEADS UP anticipated an
evacuation rate of 20,000 vehicles per hour from a county, but traffic counts
indicate that a much lower rate of evacuations is occurring, then this
information will feed back into the model.
·
Integrate mesoscale weather (proposed).
HEADS UP might integrate mesoscale weather predictions into the model. For
example, if weather predictions indicate rainfall of 1.5 inches/hour on an
important evacuation route, then HEADS UP might decrease the capacity on that
route because of the poor weather conditions.
In particular, Florida used the FHWA
grant to support HEADS UP, Phase 1, which included only some of these features.
The success of the Phase 1 implementation of HEADS UP led to the Phase 2
version of that software, which includes most of the features listed above. For
example, HEADS UP now includes an Abbreviated Transportation Model (see section
2.2.5) that helps calculate dynamic clearance times based
on specific storm information.
Florida views this program as the next
generation of the ETIS software, and is interested in working with other states
that might be interested in using the software.
2.2.4.
North Carolina
In reviewing
needs that could be addressed by the FHWA grant, North Carolina identified the need for
real-time traffic information that could be monitored during hurricane
evacuations. Real-time traffic information on evacuation routes, especially
along the contraflow portion of I‑40, was of particular interest. North Carolina
advertised its interest in receiving bids for a best-value deployment of
monitoring stations that would deliver speed, volume, occupancy, and video data
to the state Traffic Operations Center (TOC) and the Website www.ncsmartklink.org.
However, there were no responses. Apparently, the cost of providing the field
hardware to take and transmit traffic measurements and the software costs to
fuse that data with existing systems and provide a user interface to access the
data made the low price unappealing to vendors.
Lesson Learned 7 – Develop
simple-to-use decision-support tools.
|
North Carolina then began looking for other
funding sources that could be combined with the FHWA grant to deploy a larger
traffic monitoring project that would attract vendor participation. This search
identified a Federal ITS earmark that had been granted to support traveler
information for Johnston
County, The location of
the planned exit point of the I-40 contraflow. This earmark would be spent on
deploying four cameras, four detectors, a communication network to transmit the
data to the TOC, and a GUI interface to manage it all. The hurricane evacuation
grant was added to these earmark dollars to deploy an additional four detectors
that are at locations important for monitoring I-40 contraflow. The project is
currently in the design phase, should be advertised in the spring of 2004, and
installed by next hurricane season. Eventually, the additional evacuation data
will feed into the traveler information system.
Because the
project is still in the planning stages, no direct benefits of the project have
been observed. However, North Carolina expects that the availability of
real-time traffic data near the exit point of the contraflow portion of I-40
will help them better manage evacuations in the future.
Lesson Learned 7 – Develop
simple-to-use decision-support tools.
|
Virginia used the FHWA grant to update
estimated hurricane evacuation clearance times for the at-risk population in
the Hampton Roads area by developing an interim abbreviated transportation
model (ATM). In the past, the application of clearance time models has been
limited in two ways. First, since the models were difficult to update as
population and behavioral parameters change, such updates were infrequent. For
example, the clearance times for the Hampton Roads area had not been updated
since the early 1990’s, despite significant changes in the population and population
distribution in that area. Second, the models were difficult for
non-transportation planners to understand and use. This has meant that the
models were seldom used to estimate the impact of new developments on clearance
times.
In response to
these limitations, the Hurricane Evacuation Study process has recently been
modified to produce an ATM. The ATM is designed to be relatively easy to update
to estimate the impact of expected or actual population changes on evacuation
clearance times. Virginia
is already in the process of updating the HES for the Hampton Roads area, and
this update will produce an ATM. However, Virginia was anxious to have improved
evacuation clearance time estimates before the completion of the HES. To this
end, Virginia
used the FHWA grant to develop an interim ATM that is based on current census
data and evacuation phasing strategies developed in the 1980’s and updated in
2001.
The first benefit
of developing the interim ATM was that it indicated that, despite the
population increases, the updated evacuation strategies resulted in clearance
times that had not increased significantly since the completion of the last
HES. This allowed evacuation planners to focus their attention on other issues
will waiting the completion of the HES. For example, the interim ATM was used
to evaluate what-if scenarios of what would happen if a major evacuation route
were blocked in order to better plan mitigation activities.
The second
benefit occurred because, once completed, the interim ATM was distributed to
each hurricane risk jurisdiction in the Hampton Road area, and planners have been
using the interim ATM to assess the impact of new developments on hurricane
evacuations. For example, when a developer comes in and looks at siting a major
housing development, the planners have used the interim ATM to estimate the
impact of changing demographic and population figures on the evacuation
clearance times given the existing roadways.
A third benefit
is that working on the interim ATM opened communication pathways between
hurricane response organizations in Virginia
and North Carolina.
While these communications were originally necessary to develop the interim
ATM, they have continued (e.g., via bi-state meetings) because they have
helped everyone improve their hurricane evacuation planning and response
activities.
3.0
The Louisiana
Information Station Deployment
This section
describes the Louisiana Information Station Deployment for monitoring traffic
and water level conditions and transmitting the data in near real time to
appropriate traffic and emergency operations.
Like many U.S. coastal states, Louisiana is at risk for hurricane damage.
There are several elements related to the geography of Louisiana that put this state at particular
risk. First, there is a large coastal area southeast of New
Orleans (the parishes of Jefferson, Plaquemines, and St. Bernard)
with long evacuation routes that pass through New Orleans. Since New Orleans, too, will be
evacuated if a significant hurricane is expected to strike the southeast coast
of Louisiana, the evacuation times for these counties is lengthened by the
already long evacuation times expected for New Orleans.
Since most of New Orleans is below sea level and a large hurricane has
the potential to put most of the city under water, evacuees must escape the New Orleans basin before
they can expect to be safe. Another result of the low elevations of much of Southeast Louisiana is that important hurricane
evacuation routes are subject to flood. One of the greatest challenges facing Louisiana is that it
must manage long evacuations across terrain that is subject to significant
flooding.
One of the
priorities that Louisiana
needed to address in order to better manage such evacuations was that of gathering
real-time data on traffic and water level conditions for evacuation routes.
This led Louisiana
to work with the United States Geological Survey (USGS) to design and deploy
Information Stations that simultaneously monitor traffic and water level conditions
and transmit that data in near real time. Deploying the Information Stations
led to several benefits that could benefit other states that might use a
similar approach to gathering real-time traffic and hydrographic data:
·
Cost sharing. Because the Information
Stations provide data of value to both the LA DOTD and USGS, the costs of
installing, operating, and maintaining those stations are shared between those
organizations and other cost-sharing partners. This reduces the effective cost
for each organization.
·
Reduced infrastructure requirements.
Because the Information Stations rely on satellite communications and solar
power, they can be installed where access to communication and power utilities
would be expensive. Also, the NOAA communications and data processing
infrastructure that already supports data collection from USGS hydrographic
stations can be used to facilitate DOT access to the traffic data collected
from Information Stations.
·
Ease of deployment in remote locations.
Because the Information Stations do not rely on access to power and
communication utilities, they can be easily installed at remote locations that
are not part of the power and communication grids used to support other
transportation monitoring field devices.
This combination
of benefits may make this collaborative approach to deploying traffic
monitoring devices attractive to states other than Louisiana.
Because of its
low-lying landscape, Southeastern Louisiana
and its roadways have long been subject to flooding. While a flooded roadway
can be dangerous in any circumstance, it can be disastrous if it occurs during
a hurricane evacuation. The most recent such occurrence in Louisiana was when a
low-lying section of I-10 that passes beneath the Southern Railroad underpass
flooded 12 hours before Tropical Storm Isidore hit land in 2002. Since this
section of I-10 is part of an important evacuation route from New Orleans, losing this section of road
during a hurricane could strand many evacuees. A recent notice from the NOAA reinforced the significance of this problem
by pointing out that many roads in South Louisiana
are sinking, which increases their risk of being flooded.
Source: http://water.usgs.gov
Figure 3–1.
USGS Hydrowatch Stations
|
In response to
this problem, Louisiana
began working with USGS to develop a Flooded Road Alert System. The basis of
this system is 165 USGS Hydrowatch stations (see Figure 3–1) that monitor in near real-time stream stage,
streamflow, and other hydrographic information in Louisiana. (Other Hydrowatch stations exist
in Louisiana
that do not monitor stream stage; these are not used by the Flooded Road Alert
System.) Because 95 percent of these stations are located on LA DOTD
bridges, they are an ideal source of information on the water level beneath the
bridges. By combining this stream stage information with data on the height of
the roadbed above the stream, alerts can be generated when the measured stream
stage approaches the height of the roadbed.
In working with
USGS on the Flooded Road Alert System, LA DOTD became aware of the potential
for a collaborative effort between LA DOTD and USGS to deploy additional
stations that combine the hydrographic measurement instrumentation common on
Hydrowatch stations with traffic count instrumentation desired by LA DOTD. The
combined station, dubbed an Information Station, could leverage the data
collection and communication infrastructure already supported by USGS for its
network of Hydrowatch stations. This combined system could provide traffic
count information at a cost below that for using separate instrumentation for
each. LA DOTD approached USGS with the Information Station concept, and USGS agreed
to deploy prototype Information Stations if LA DOTD could provide traffic count
instrumentation that was plug-compatible with the existing Hydrowatch station
configuration.
At the same time,
the Louisiana Office of Emergency Planning (LOEP), now called the Louisiana
Office of Homeland Security Preparedness, was encouraging LA DOTD to build on a
failed 1997 effort to deploy real-time traffic count stations to support
hurricane evacuations. LA DOTD and LOEP agreed to support development, testing,
and prototype deployment of the Information Stations. LA DOTD then began
working with a vendor, PEEK, to develop such a plug-compatible loop detector
instrument. Development and testing of the loop detector instrument for the
Information Station required about 1.5 years.
During that
period, LA DOTD also worked with LOEP to identify 22 flood-prone locations
around Lake Pontchartrain for which real-time
traffic and flooding information would be critical for supporting hurricane
evacuations. Funding for seven of these sites was committed by three agencies:
LA DOTD; LOEP; and USGS. Deployment of prototype stations began in the Spring
of 2003. At the time of this report, seven Information Stations are deployed
and operating. Early indications showed that these stations operated well,
reliably providing LA DOTD with near real-time traffic information and water
level in all weather conditions. Additional deployments are planned, but LA
DOTD has delayed those deployments while they determine the type of traffic
detector to use in future stations.
An Information
Station (see Figure 3–2) is a USGS Hydrowatch station that is fitted with a
plug-compatible traffic detector. A typical Hydrowatch station is a set of
measurement devices – typically, water level, wind speed, wind direction, and
rainfall – powered by a battery with a solar array for recharging and equipped
with a satellite-based communication system. Typically, the data logger has
unused ports that are free to be used to register other information, such as
traffic information.
Without the solar
array or satellite communication, the station is designed to run unattended for
several weeks. Periodically, such a station must be visited to replace or
recharge the battery and retrieve the logged data. With the solar array and
satellite communication, less frequent and less costly equipment maintenance
visits are required. USGS personnel still visit each site, on average, once
every 6weeks in order to check on the integrity of the site and to take
streamflow measurements to confirm the calibration between stream gage and
streamflow.
Each measurement
device is connected to a data logger, which records the measured results from
the devices. For new Hydrowatch stations, the data from the station is
transmitted once per hour via the National Environmental Satellite, Data, and
Information System (NESDIS). Each station is allocated a 10-second time slot
once per hour to transmit data over a 300-baud channel. (Older stations only
transmit data once every 4 hours using a 1-minute slot on a 100-baud channel.)
When needed
(e.g., during emergency operations), stations can also access additional
communication channels to transmit data once every 15 minutes. For example, a
Hydrowatch station can be programmed to use these extra channels to transmit an
alert notice when the measured water level approaches a pre-selected flood
level. LA DOTD is currently working with USGS to provide three different alerts
as the water level approaches flood stage at a Hydrowatch station: a first
alert when the water level is within 2 feet of the road level, a second alert
at 1 foot, and a final alert when the water level equals the road level.
Many stations are also equipped with a phone line or cell phone connection, in
which case the station can be polled as needed.
The transmitted
data passes via satellite to a receiving station in Virginia. It is then transmitted to the
regional USGS office, archived in a database, and forwarded to an Internet
server. The data can be accessed via the regional Hydrowatch home page and can
be pushed to another server using an Internet-based connection. The typical
latency is between 2 and 3 minutes from the time the data is transmitted from
the field device and when it is available at the regional Web page.
An Information
Station is a Hydrowatch station in which some of the unused ports are connected
to traffic detectors. Loop detectors were used
in the initial deployments, but LA DOTD is considering the use of other types
of traffic detector. The data logger merges the traffic count and hydrographic
data, and transmits the combined data as part of the hourly satellite transmission.
The transmitted data can then be retrieved from the regional Hydrowatch home
page via the Internet-based connection mentioned previously. The traffic data
from an Information Station is not available on the regional Hydrowatch home
page. Thus, an Information Station allows LA DOTD to collect traffic count data
in near real time in a way that leverages existing USGS capabilities,
eliminating the need to connect the station to either a power grid or a LA DOTD
communication network.
At the time of
this report, Louisiana
had deployed seven information stations and had plans for deploying 15 others.
The expected cost breakdown for one of these stations is given in Table 3–1. The table also specifies the type of equipment,
purchase or installation cost, and the anticipated maintenance costs associated
with the various types of equipment for an average total deployment and
maintenance cost of nearly $49,000 per unit.
Table 3–1.
Expected Costs an Information Station at Highway 90 Near Pearlington, MS
|
Type of Equipment
|
Purchase / Install Cost
|
Maintenance Cost
|
|
Hydrowatch Equipment
|
|
Data Collection
Platform
|
$12,400
|
$ 4,500
|
|
Rain Gage
|
$ 1,000
|
$ 1,000
|
|
Wind Speed and
Direction
|
$ 3,000
|
$ 1,000
|
|
O&M
|
---
|
$ 6,000
|
|
Traffic Equipment
|
|
Traffic Counter*
|
$ 7,930
|
---**
|
|
Permanent Loop (6 x 6)
|
$ 1,500
|
---**
|
|
TOTALS
|
$25,830
|
$12,500
|
While the costs
in the Table 3–1 are typical, considerable variation can exist in the
costs of the Hydrowatch equipment. At LA DOTD, the Information Station project
began with $60,000 in funding from FHWA and $140,000 in funding from LOEP. This
funding, along with some Hydrowatch deployment and maintenance funds available
to USGS, was sufficient to test, deploy, and maintain seven Information
Stations. Plans approximate the costs at about $426,000 to deploy and maintain
an additional 15 stations.
The Mississippi
USGS also provided information on the costs of deploying and operating
streamgages. The typical deployment cost for streamgages
in Mississippi
is about $12K, with operational costs varying depending on the type of station.
The operating costs for a station that continuously monitors streamflow is
about $12K per year, with costs of about $8K per year if a station monitors
streamflow only during floods and costs of about $6K per year if a station is
only used to monitor streamstage. These costs are comparable to the costs
listed in Table 3–1 after removing the costs of the meteorological and
traffic instruments.
For comparison,
the Transportation Statistics Office of the Florida DOT provided cost estimates
for deploying and maintaining their collection of telemetered traffic
monitoring sites (TTMSs). Florida’s
TTMS program includes over 300 data collection stations, most of which collect
traffic count, speed, and vehicle classification information. The stations use
batteries for power with solar panels for recharging the batteries and use
modems for transmitting data back to the Transportation Statistics Office. Most
stations collect data continuously and transmit the data each night. The typical cost for deploying a TTMS
station for collecting traffic counts only (i.e., not speed or vehicle
classification) ranges from $10K to $15K per station. Operating costs for the
stations range between $20 and $30 per month for phone charges, and maintenance
costs average about $100K per month for the all of the stations, or about $4K
per station per year. In addition, the Transportation Statistics Office
maintains a bank of about eight receiving modems that are used to poll the TTMS
stations each night. (For real-time access via modems, a larger bank of modems
would be required.) The costs of deploying and operating the modem bank for
receiving this data is not included in the preceding listed operating and
maintenance costs.
Because an
Information Station is based on a Hydrowatch station, it is important to
understand the USGS program that is responsible for deploying Hydrowatch
stations – the USGS Streamgaging Program. The USGS Streamgaging program
collects streamflow information from a network of about 7,000 streamgages
nationwide, with about 5,000 of these gages equipped with satellite
communication equipment so that collected data can be relayed to USGS in near
real-time. The collected data is available at http://waterdata.usgs.gov/.
Funding for these
gages comes from a variety of sources – more than 700 Federal, State, and local
agencies cooperate with USGS to fund, in whole or in part, about
93 percent of USGS-operated stations. The majority of these stations
(about 4,000) are funded under the Federal-State Cooperative Program, under
which USGS provides up to 50 percent of the funds required to operate the station.
A large number of stations (about 2,000) are funded by other Federal agencies,
with about 500 stations funded directly from USGS Congressional appropriations.
While this
diversity of funding sources has enabled USGS to develop a streamgage network
that is substantially larger than one that could be created using USGS
resources alone, this network recently was reduced in size because funding
partners discontinued support for some streamgages. This change induced USGS to
begin developing the USGS National Streamflow Information Program (NSIP). Under this program, USGS intends to
directly support streamflow measurement activities for a core set of critical
streamgages and the network used to collect, validate, and disseminate the
collected data. USGS would continue to support non-critial streamflow gages
through the Cooperative Program. Because the network expenses for each gage
would be covered by NSIP rather than by the partners funding each gage, it is
expected that costs for streamgages under the Cooperative Program would
decrease.
Because NSIP is
not fully funded at this time and the Cooperative Program funding in most
states is fully utilized, USGS funding for new sites may be unavailable in some
states. For transportation agencies, this will decrease the cost-effectiveness
of Information Stations in states where USGS Cooperative Program matching funds
are not available. More details on the factors that should be considered before
deciding to deploy Information Stations are listed in section 3.6, and section 3.7 lists some hypothetical examples that demonstrate
some of the factors that impact the cost-effectiveness of Information Stations.
Deploying an Information
Station for measuring near real-time traffic counts rather than a traditional
real-time traffic count station seems like a win-win proposition for both a DOT
and USGS. However, several elements must be considered before the benefits of
this approach can be achieved, which are described as follows:
·
Operating costs of streamgages can be higher
than traffic count stations. USGS uses water-level data to estimate
streamflow by applying a site-specific rating curve that relates water-level to
streamflow. Because the streambed can change significantly over time, USGS
technicians visit streamgages that measure streamflow about once every 6 weeks to measure the flow directly. This results
in significantly higher operating costs than might be expected by state DOTs
considering the operating costs of remote traffic count stations, which require
less frequent service. Because USGS will typically want to share the operating
costs for new streamgage installations, the high operating costs of streamgages
could make an Information Station uneconomical for most state DOTs. Note that
some USGS streamgages only measure water level, not streamflow, in which case
lower operating costs may be possible. Other strategies for making these costs
more economical are considered in some of the paragraphs below.
·
It is possible to retrofit existing
streamgages with traffic count instrumentation. Because an Information
Station employs unused data ports on the data logger to connect to the USGS
streamgage equipment, it may be possible to retrofit existing streamgages with
traffic count instrumentation. One approach for ensuring that an Information
Station program benefits both USGS and a state DOT could be to use a mix of new
streamgaging stations partially supported by the DOT – the costs to the DOT for
these new stations may be higher than other approaches for gathering traffic
count information, and retrofits of existing stations that are supported from
other sources – the costs to the DOT for
these new stations will be below that of other approaches. By using both
low-cost add-ons to existing stations and higher cost deployment of new
stations, the overall cost to the DOT may be below that of other alternatives.
See section 3.7 for some hypothetical examples that demonstrate the
impact of this factor on the cost of Information Stations.
·
Identify funding sources other than the DOT
and USGS. USGS actively collaborates with numerous partners in funding the
streamgage network. In Louisiana,
LOEP is already participating in a program to fund new streamgage stations in
areas likely to flood. LOEP also has a natural interest in helping with
hurricane evacuation routes, so the combined benefits of an Information Station
that could monitor for flooding during normal operations and measure
evacuation traffic and monitor for flooding during hurricanes was very
appealing to LOEP. Emergency planning organizations in other states may find
that combination of benefits appealing, as well. Local municipalities may have an
interest in improving stream monitoring in flood-prone regions. By working with
USGS to identify other potential funding partners, a DOT can reduce its portion
of the costs for Information Stations.
·
Consider the full range of benefits of
deploying Information Stations. An Information Station does not just
provide traffic count data to a state DOT, but can also provide warnings if the
stream height approaches flood levels. A secondary benefit is that accurate
streamflow data can help improve downstream flood stage predictions so that
appropriate precautions can be taken to safeguard both the affected roads and
the community. An information station that includes wind speed measurements can
be used to identify when dangerous driving conditions exist due to excessive
winds, and could also be used to identify when to close bridges during storms.
Also, the USGS streamgaging network has proven itself to provide reliable data
during all sorts of adverse conditions when other systems may fail. For
example, damage to phone lines during a hurricane or spikes in cell phone
communications can eliminate phone or cell phone communications from remote
detectors. This type of failure does not affect the satellite communication
network used by USGS for collecting streamgage data.
·
The reliability of the USGS streamgaging
network. USGS has placed considerable emphasis on the reliability of the
streamgaging network. The stations are capable of withstanding strong weather
events. The battery/solar power supply and satellite communications used are
less vulnerable to disruptions in utility services than many other approaches
(e.g., storms downing phone and power lines, cell phone users overloading
cells during emergency situations). A redundant receiver station for the
satellite downloads is being considered. Many state USGS offices have arranged
partnerships with other state USGS offices to provide redundancy for the
Websites that make the collected data available. These actions lead to a high
level of reliability that may be difficult for individual state DOTs to match
at a comparable cost.
·
Consider the full range of cost savings of
deploying Information Stations. Deploying a typical permanent traffic count
station requires trenching to connect to a local power and phone service. In many
locations, these trenching costs can be a significant fraction of the overall
deployment costs. Also, a data collection system (e.g., modem bank with
software for polling stations and storing retrieved data) must be developed and
maintained, and monthly power and communication charges will be assessed. The
Information Station does not require any power or phone connections (though
phone or cell phone connections are often built-in for backup communication),
and the USGS data collection system is already in place. These factors can
significantly reduce the overall operating cost of an Information Station as
compared to a typical station.
·
The Information Station approach can
complement other DOT data collection activities. The Information Station
approach to gathering near real-time traffic data is at its best in remote
locations where access to power and a statewide communications network is more
costly or does not exist. Even if a DOT can use an existing statewide
communication network for many locations, the Information Station approach may
still make sense for more remote locations.
·
Compromises in the location of the
Information Stations may be required. Information Stations must be deployed
at locations where streamgage measuremen
