-
Climate Change in Coastal Areas in Florida:Sea Level Rise
Estimation and
Economic Analysis to Year 2080
Dr. Julie Harrington, DirectorCenter for Economic Forecasting
and Analysis
Florida State UniversityInnovation Park
2035 E. Paul Dirac Dr.Suite 129 Morgan BldgTallahassee Fl
32310
Dr. Todd L. Walton, Jr., DirectorBeaches and Shores Resource
Center
Florida State UniversityInnovation Park
2035 E. Paul Dirac Dr.Suite 203 Morgan BldgTallahassee Fl
32310
February 2007
This research was supported by a grant
from the National Commission on Energy Policy.
-
Summary Results
The science and economics of sea level change has been
researchedin the past, with primary focus based on erosion of the
shoreline,and human adaptation. This study serves to add to sea
level riseresearch from two perspectives: one, through the
estimation of sealevel rise based on historical local tidal gage
data, and the other per-spective, to examine storm event and
corresponding cost damages(based on intensity and storm surge
return period) and associatedchanges in property values in a six
coastal county case site area ofFlorida (Dade, Duval, Monroe,
Escambia, Dixie and Wakulla).
The following project had a two-pronged approach. The data
col-lection and estimation of sea level change, based on historical
tidalstation gage data, was performed by the FSU Beaches and
ShoresResource Center. The approaches to SLR estimation examined
theeffects of accelerated SLR, including subsidence. In addition,
thisproject utilized projected changes in eustatic SLR for years
2030and 2080, based on the Intergovernmental Panel on Climate
Change(IPCC) 2001 estimates. Based on the FSU Beaches and Shores
Re-source Center findings and IPCC data, the FSU Center for
EconomicForecasting and Analysis (CEFA) estimated storm event
return pe-riod and cost damages, storm surge, and property values
affected asa result of sea level rise, for the six county case
sites in Florida.
• Summary Result
– Sea Level Rise Estimate
1. The FSU Beaches and Shores Resource Center foundthat although
there was a wide distribution of differentgage sites over the
Florida Peninsula, the projected sealevel rise in year 2080 does
not vary substantially, thelargest value being 0.35 meters in St.
Petersburg, FL,while the smallest value is 0.25 meters in
Fernandina,FL.
2. This study is the first known work to explore sea levelrise
forecasting methods beyond the traditional poly-nomial linear
estimation forecasting methods utilizinggage data. The second order
approach used in this
2
-
study is in accordance with climate modeling scenariosthat
project an exponential sea level rise due to green-house gas
effects.
– Economic Analysis of Sea Level Rise Estimate
1. There will be an increasing return period frequencywith an
associated sea level rise. That is, more se-vere storm events will
become more frequent than inthe past. Cost damages associated with
storm eventscan also be expected to increase with respect to
sealevel rise. Based on the IPCC sea level rise estimates,regarding
a number of coastal counties, people livingalong the coast are
projected to experience propertylosses at twice the rate of
normal.
2. The storm model findings convey a considerable lossof
property, in terms of property values and land area,due to varying
sea level rise for years 2030 and 2080.
This study was able to establish a reasonable range of low to
highsea level rise estimates to year 2080, with FSU providing the
lowerbound for sea level rise based on historical gage data, and
the IPCC2001 sea level estimates providing the upper bound, based
on cli-mate modeling scenarios. The results of this project
underscore theimportance of including sea level rise as a critical
component in thehazard preparedness and mitigation planning for
coastal communi-ties. In addition, the SLR impact on property
values was analyzedwith the GIS Storm Model Regressions. Finally,
the damage costestimates of storm surge for each 10 cm SLR were
compared to theproperty values at risk with similar SLR.
Key Words:Sea Level Rise (SLR), Property Value, Damage Cost,
Storm Surge,Return Period
3
-
Contents
1 Introduction 9
2 Projected Sea Level Rise (to 2080) in Florida, Basedon Tide
Station Records 142.1 Introduction . . . . . . . . . . . . . . . .
. . . . . . . 142.2 Data Sources . . . . . . . . . . . . . . . . .
. . . . . 162.3 Methodology For Sea Level Forecasts . . . . . . . .
. 182.4 Summary . . . . . . . . . . . . . . . . . . . . . . . .
23
3 Economic Analysis of Sea Level Rise 263.1 Regression Model(s)
For Hurricane Return Years and
Damage Cost . . . . . . . . . . . . . . . . . . . . . . 273.2
Hurricane Return Year(s) . . . . . . . . . . . . . . . 293.3 Damage
Cost Assessment . . . . . . . . . . . . . . . 403.4 Economic
Analysis of Property Loss for the Six Florida
Coastal Counties . . . . . . . . . . . . . . . . . . . . 453.5
”Storm Model” Results . . . . . . . . . . . . . . . . . 463.6 Storm
Surge Results . . . . . . . . . . . . . . . . . . 473.7 Alternate
SLR Impact Analysis Using the GIS Storm
Model Regression Models . . . . . . . . . . . . . . . . 483.8
Summary and Conclusions . . . . . . . . . . . . . . . 56
4 Appendix A 65
5 Appendix B 66
6 Appendix C 72
7 Appendix D 86
4
-
List of Figures
1 Map of Six Florida Counties . . . . . . . . . . . . . . 132
Fernandina Gage Station Forecast Filtered Sea Level
Rise . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
Key West Gage Station Forecast Filtered Sea Level
Rise . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
St. Petersburg Gage Station Forecast Filtered Sea
Level Rise . . . . . . . . . . . . . . . . . . . . . . . . 245
Cedar Key Gage Station Forecast Filtered Sea Level
Rise . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
Pensacola Gage Station Forecast Filtered Sea Level
Rise . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
Reduction of Hurricane Return in Wakulla County
Year(s) by Elevation Based on IPCC and FSU SeaLevel Rise
Estimates . . . . . . . . . . . . . . . . . . 31
8 Reduction of Hurricane Return Year(s) in Dade Countyby
Elevation Based on IPCC and FSU Sea Level RiseEstimates . . . . . .
. . . . . . . . . . . . . . . . . . 33
9 Reduction of Hurricane Return Year(s) in Dixie Countyby
Elevation Based on IPCC and FSU Sea Level RiseEstimates . . . . . .
. . . . . . . . . . . . . . . . . . 34
10 Reduction of Hurricane Return Year(s) in Duval Countyby
Elevation Based on IPCC and FSU Sea Level RiseEstimates . . . . . .
. . . . . . . . . . . . . . . . . . 36
11 Reduction of Hurricane Return Year(s) in EscambiaCounty by
Elevation Based on IPCC and FSU SeaLevel Rise Estimates . . . . . .
. . . . . . . . . . . . 38
12 Reduction of Hurricane Return Year(s) in MonroeCounty by
Elevation Based on IPCC and FSU SeaLevel Rise Estimates . . . . . .
. . . . . . . . . . . . 39
13 Damage Cost and Storm Surge Estimates in WakullaCounty . . .
. . . . . . . . . . . . . . . . . . . . . . . 42
14 Damage Cost and Storm Surge Estimates in DadeCounty . . . . .
. . . . . . . . . . . . . . . . . . . . . 42
15 Damage Cost and Storm Surge Estimates in DixieCounty . . . .
. . . . . . . . . . . . . . . . . . . . . . 43
16 Damage Cost and Storm Surge Estimates in DuvalCounty . . . .
. . . . . . . . . . . . . . . . . . . . . . 43
5
-
17 Damage Cost and Storm Surge Estimates in Escam-bia County . .
. . . . . . . . . . . . . . . . . . . . . . 44
18 Damage Cost and Storm Surge Estimates in MonroeCounty . . . .
. . . . . . . . . . . . . . . . . . . . . . 44
19 Property Values at Risk with Sea Level Rise of DadeCounty . .
. . . . . . . . . . . . . . . . . . . . . . . . 49
20 Storm Surge Damage Cost with Sea Level Rise ofDade County . .
. . . . . . . . . . . . . . . . . . . . 49
21 Property Values at Risk with Sea Level Rise of DixieCounty .
. . . . . . . . . . . . . . . . . . . . . . . . . 50
22 Storm Surge Damage Cost with Sea Level Rise ofDixie County .
. . . . . . . . . . . . . . . . . . . . . 50
23 Property Values at Risk with Sea Level Rise of DuvalCounty .
. . . . . . . . . . . . . . . . . . . . . . . . . 51
24 Storm Surge Damage Cost with Sea Level Rise ofDuval County .
. . . . . . . . . . . . . . . . . . . . . 51
25 Property Values at Risk with Sea Level Rise of Es-cambia
County . . . . . . . . . . . . . . . . . . . . . 52
26 Storm Surge Damage Cost with Sea Level Rise ofEscambia County
. . . . . . . . . . . . . . . . . . . . 52
27 Property Values at Risk with Sea Level Rise of Mon-roe County
. . . . . . . . . . . . . . . . . . . . . . . 53
28 Storm Surge Damage Cost with Sea Level Rise ofMonroe County .
. . . . . . . . . . . . . . . . . . . . 53
29 Property Values at Risk with Sea Level Rise of WakullaCounty
. . . . . . . . . . . . . . . . . . . . . . . . . . 54
30 Storm Surge Damage Cost with Sea Level Rise ofWakulla County
. . . . . . . . . . . . . . . . . . . . . 54
31 Hurricane Return Year(s) and Cost Damage in WakullaCounty . .
. . . . . . . . . . . . . . . . . . . . . . . . 66
32 Hurricane Return Year(s)and Cost Damages in DadeCounty . . .
. . . . . . . . . . . . . . . . . . . . . . . 66
33 Hurricane Return Year(s) and Cost Damages in DixieCounty . .
. . . . . . . . . . . . . . . . . . . . . . . . 67
34 Hurricane Return Year(s) and Cost Damages in Du-val County .
. . . . . . . . . . . . . . . . . . . . . . . 67
35 Hurricane Return Year(s) and Cost Damages in Es-cambia County
. . . . . . . . . . . . . . . . . . . . . 68
6
-
36 Hurricane Return Year(s) and Cost Damages in Mon-roe County .
. . . . . . . . . . . . . . . . . . . . . . 68
37 Increase Damage Cost in Wakulla County . . . . . . 6938
Damage Cost and Return Period for Dade County . . 6939 Damage Cost
and Return Period for Dixie County . . 7040 Damage Cost and Return
Peroid for Duval County . 7041 Damage Cost and Return Period for
Escambia County 7142 Damage Cost and Return Period for Monroe
County 7143 Dade County . . . . . . . . . . . . . . . . . . . . . .
7244 Dixie County . . . . . . . . . . . . . . . . . . . . . . 7445
Duval County . . . . . . . . . . . . . . . . . . . . . . 7646
Monroe County . . . . . . . . . . . . . . . . . . . . . 8047
Wakulla County . . . . . . . . . . . . . . . . . . . . . 8248
Escambia County . . . . . . . . . . . . . . . . . . . . 84
7
-
List of Tables
1 The Potential Cost of Sea-Level Rise Along the De-veloped
Coastline of the United States (Billions of1990 Dollars) . . . . .
. . . . . . . . . . . . . . . . . 10
2 Eustatic Sea Level Rise Scenarios (in meters) . . . . . 143
Florida Tide Stations Used with at Least 50 Years of
Historical Data . . . . . . . . . . . . . . . . . . . . . 174
Table 4. Forecast Relative Sea Level Rise From 2006
to 2080 by Florida County . . . . . . . . . . . . . . . 215
Forecast Relative Sea Level Rise from 2006 to 2080 . 226 Regression
Results for Elevation Return Years by County 287 Hurricane Return
Year(s) for Recent Hurricane Events
by County . . . . . . . . . . . . . . . . . . . . . . . . 298
Storm Events Totals (in 2006 $)by County of Loss
Occurrence, 2004 . . . . . . . . . . . . . . . . . . . . 409
Storm Event Totals (in 2006 $) by County of Loss
Occurrence, 2005 . . . . . . . . . . . . . . . . . . . . 4010
Damage Cost and Storm Surge Regression Equations
for Recent Hurricane Events by County . . . . . . . . 4111 Sea
Level Rise, Cost Damage and Storm Surge Esti-
mates by County . . . . . . . . . . . . . . . . . . . . 4112
Land Value and Land Area at Risk from Four Sea
Level Rise Scenarios For Six Florida Counties . . . . 4713 Storm
Surge Projections Based on FSU and IPCC
SLR Estimates . . . . . . . . . . . . . . . . . . . . . 4814
Regression Results of Property Values at Risk . . . . 5515
Comparison of Marginal Storm Surge Damage Cost
and Marginal Property Values at Risk Assessment . . 56
8
-
1 Introduction
The economic impact of sea level change has been researched
inthe past, with primary focus based on erosion of the shoreline,
andhuman adaptation. This study will provide an estimate (to
year2080) of sea level and hurricane prediction changes (based on
inten-sity and storm surge return period) and projected economic
analysisof affected property values in a six coastal county case
site area ofFlorida (Dade, Duval, Monroe, Escambia, Dixie and
Wakulla).
The US has approximately 12,400 miles of coastline and morethan
19,900 miles of coastal wetlands. Frequently cited studies ofthe
’quantitative’ economic impacts of global climate change
includeNordhaus (1991), Cline (1992), Fankhauser (1995) and Tol
(2002)1.These studies estimated the costs of climate change in the
US in theareas of farming, forestry, fisheries, energy, water
supply, ecosystemloss, human amenity, life /morbidity, migration,
air pollution, waterpollution and natural hazards. The summary of
the results showthat annual costs range from $48.6 to $121.3
billion with associatedtemperature rises from 2.5 to 4◦C, and with
corresponding sea levelrises of one meter (to year 2100). Those
annual costs are compa-rable with a loss of between 1-2.5% of US
GDP. Based on a studyin 2000, by Neuman, et al., Table 1
illustrates the potential cost ofsea level rise along the developed
coastline of the U.S., in 1990 dol-lars2. Another recent study
(Greenpeace, 2006) estimated the costsof adapting to a one meter
sea level rise in the US would amount toUS $156 Billion (3 percent
of GDP).3
1Tol, R. S. J. 2002. Estimates of the Damage Costs of Climate
Change. Part 1: BenchmarkEstimates . Environmental and Resource
Economics 21: 41-73.
2Neumann, J.E., G. Yohe, R. Nicholls, M. Manion, 2000. Sea-Level
Rise and GlobalClimate Change: A Review of Impacts to U.S.
Coasts.
3http://www.greenpeace.org/international/campaigns/climate-change/impacts/sea
level rise
9
-
Table 1 — The Potential Cost of Sea-Level Rise Along
theDeveloped Coastline of the United States (Billions of1990
Dollars)
Global Sea-Level Rise(source)
Measurement AnnualizedEstimate
CumulativeEstimate
AnnualEstimate in2065
100 cm (Yohe, 1989) Property at riskof inundation
n/a 321 1.37
100 cm (EPA,1989) Protection n/a 73-111 n/a100 cm (Nordhaus,
1991) Protection 4.9 n/a n/a100 cm (Fankhauser, 1994) Protection
1.0 62.6 n/a100 cm (Yohe, et al., 1996) Protection and
Abandonment0.16 36.1 0.33
50 cm (Yohe, 1989) Property at riskof inundation
n/a 138 n/a
50 cm (Fankhauser, 1994) Protection 0.57 35.6 n/a50 cm (Yohe et
al., 1996) Protection and
Abandonment0.06 20.4 0.07
50 cm (Yohe andSchlesinger, 1998)
Expected Protec-tion and Aban-donment
0.11 n/a 0.12
100 cm (Yohe andSchlesinger, 1998)
Expected Protec-tion and Aban-donment
0.38 n/a 0.40
41 cm mean (Yohe andSchlesinger, 1998)
Protection andAbandonment
0.09 n/a 0.10
10 cm, 10th percentile (Yoheand Schlesinger, 1998)
Protection andAbandonment
0.01 n/a 0.01
81 cm, 90th percentile (Yoheand Schlesinger, 1998)
Protection andAbandonment
0.23 n/a 0.31
The costs of sea level rise can be expressed as the capital cost
ofprotection plus the economic value of land and structures at loss
orat risk. Agricultural impacts can be expressed as costs or
benefitsto producers and consumers. Non-market impacts, such as the
im-pacts on ecosystems or human health is more difficult to
quantify,although there is a broad array of literature on valuation
theoryand application, primarily in non-climate change journals
(Tol etal., 2000).
Currently, economic impact projections based on sea level
changehave not been reported for the state of Florida. However,
there area number of climate change studies that have been
conducted inFlorida. Yohe, et al., conducted sea level change
research for cer-tain sites in Florida (Miami, Key West, Port
Richey, Apalachicola,and St. Joseph) on ”no foresight” and ”perfect
foresight” scenarios,
10
-
regarding gradual erosion loss and adaptation of the market
withrespect to the sea level predictions, using a benefit-cost
decisionmaking framework for estimating the human response to sea
levelrise. The Yohe estimates examined 33, 67, and 100 cm SLR
scenar-ios, and relied on relatively low-resolution elevation data
by today’sstandards. The first option assumes sufficient advance
warning ofSLR and fairly rapid market response to the perceived
threat. Thesecond option reacts to the imminent loss of property at
the timeof inundation, while the last option accepts protection as
given andsimply seeks to minimize its costs. In general, costs for
the advancedforesight option are lower than for the wait-and see
option, especiallyfor the two higher SLR scenarios, but this
advantage requires moreprecise knowledge of the course of SLR and
and effective market-based retreat policy. Costs are highest for
permanent protection.For Florida, Roger Pielke, Jr. (1998) found
that hurricane damagefor a 1926 Miami hurricane, in normalized 1992
dollars4, was $39billion.
Florida is particularly vulnerable to sea level change. Florida
isthe fourth most populated state (17.5 million people in 2005)
andprojected to increase 47% by the year 2025, according to the
U.SBureau of Census. Approximately 4,500 square miles (of the
to-tal 66,000 square miles) in Florida are within 4.5 feet of sea
level.According to a recent EPA study, a one foot rise would erode
mostFlorida beaches at least 100-200 feet unless mitigation
measures wereused. The Florida South Water Management District, in
a recentstudy of the impact of sea level rise on the water
resources of theregion, found that a 15 cm sea level rise would
result in flooding insoutheastern coastal Florida and a greater
need for water use cut-backs. They also found that certain areas
throughout the Districtwould need additional freshwater deliveries
to offset the increasinglyhigher saline water intrusion. Another
recent study conducted bythe National Wildlife Federation and
Florida Wildlife Federation,of sea level rise (using a mid-range
scenario of 15 inches) for ninecoastal areas in Florida found that
about 50 percent (23,000 acres)of saltmarsh and 84 percent (167,000
acres) of tidal flats at thesesites would be lost by year 2100. The
area of dry land is projected to
4Assumption was that losses are proportional to three factors:
inflation, wealth, and pop-ulation.
11
-
decrease by 14 percent (175,000 acres) and about 30 percent
(1,000acres) of ocean beaches and 2/3 (5,880 acres) of estuarine
beacheswould disappear.
The tourism industry is acutely at risk to sea level rise in
Florida.Tourism is the number one industry in the state, primarily
due to itsnatural resources, a favorable climate, an immense
shoreline, themeparks, professional and major university sports,
major airports andcruise industry ports, cultural events and
retirement communities.The number of Florida tourists reached a
record 84.6 million in2006, and is projected to grow to 89 million
by 2010 (Visit Florida,2006). Currently, 1.3 million Florida jobs
are directly or indirectlyrelated to tourism, and projected to grow
to between 1.5 and 1.8million by 2010. The 2005 Florida Visitor
Study reported that thestate collected $3.7 billion in
tourism/recreation sales taxes in 2005.That is, $62 billion was
infused into the state’s economy duringthe year through tourist
expenditures. (Visit Florida, 2005). Inaddition to tourism, the
loss of coral reefs, coastal estuaries and as-sociated fisheries
due to sea level rise would also negatively impactFlorida’s
economy. There are options for adaptation to sea levelrise in
Florida including: elevating existing areas, building sea wallsand
flood control structures, and encouraging relocation. In a re-cent
study by the Natural Resources Defense Council, they foundthat
beach re-nourishment alone would cost between $50-$60
billion(current dollars) over the next 100 years.
There was a two-pronged approach regarding this project. Thedata
collection and estimation of sea level rise, based on
historicaltidal station gage data, was performed by the FSU Beaches
andShores Resource Center. The approaches to SLR estimation
in-cluded the effects of accelerated SLR and subsidence. This
studyalso included eustatic SLR for years 2030 and 2080,
correspondingto the Intergovernmental Panel on Climate Change
(IPCC) 2001 es-timates (Table 2). Based on the FSU Beaches and
Shores ResourceCenter findings, the FSU Center for Economic
Forecasting and Anal-ysis (CEFA) used the FSU SLR and IPCC
estimates to examinevarious SLR scenarios with respect to
associated hurricane/stormevent cost damages and property values
for the six county case sitesin Florida (dependent on the level of
forecasted inundation).
12
-
In addition, this project utilized projected changes in
eustaticSLR for years 2030 and 2080, based on the Intergovernmental
Panelon Climate Change (IPCC) 2001 estimates (Table 2). In a
recentstudy conducted by Rahmstorf, et al., (2007) of carbon
dioxide con-centration, global-mean air temperature and sea level
changes, theyfound that the IPCC may have underestimated, in
particular, thesea level rise projections. They concluded that the
rate of sea levelrise for the last 20 years is 25% higher than any
other 20 year periodin the preceding 115 years. Although the time
interval is relativelyshort and could be attributed to internal
decadal climate variability,the authors do stress that the largest
contribution to the rapid sealevel rise come from ocean thermal
expansion and the melting fromnon-polar glaciers, and include
increasing evidence that the ice sheetcontribution is also rapidly
increasing.
Based on the FSU Beaches and Shores Resource Center findingsand
IPCC data, the FSU Center for Economic Forecasting and Anal-ysis
(CEFA) estimated storm event return period and cost damages,storm
surge, and property values affected as a result of sea level
rise,for the six county case sites in Florida.
Figure 1 — Map of Six Florida Counties
13
-
Table 2 — Eustatic Sea Level Rise Scenarios (in meters)Year Low
Middle High2030 0.05 0.10 0.152080 0.10 0.30 0.65
*IPCC Estimates, 2001
2 Projected Sea Level Rise (to 2080) in Florida,Based on Tide
Station Records
2.1 Introduction
Rising sea level has important economic consequences for
Floridawhich has a relatively low lying coastal zone. Rising sea
level willinundate low coastal areas of Florida and cause salt
water intrusioninto coastal aquifers and coastal estuaries.
Additionally, beach anddune recession will occur as a result of the
rising sea level by creatinga sediment budget deficit in the
offshore area. This shore recessionis a function of the sea level
rise rate, the active beach profile width,and the depth of closure
as first postulated by Bruun (i.e. see Deanand Dalrymple (2002)).
Just what sea level in Florida5 will looklike in 2080 is an
unanswered question but deserving of scientificinvestigation.
Looking at the broader picture, a global rise in sea level is
pre-dicted by various climatic modelers in the following
references: Hoff-man et al.(1983), National Academy of Sciences
(1983,1985) , EPA(1989), Barth and Titus (1984), National Research
Council (1987),IPCC (1990), Houghton et al (1990), National
Research Council(1990), Church et al (1991), Wigley and Raper
(1992), Titus andNarayanan (1995), and U.S. National Report to IUGG
(1995). Inmore recent findings noted above, the Intergovernmental
Panel onClimate Change (IPCC) report (Houghton et al.(1990))
suggests ascenario of global warming and consequent global sea
level changeof 0.18 meters by 2030 and 0.44 meters by 2070. An
independentlyresearched climatic scenario of Church et al. (1991)
has calculateda sea level rise of 0.35 meters by 2050. In 2001, the
IPCC projectedthat sea level rise would increase by 0.09 to 0.88
meters by 2100 over1990. The range is mainly the result of
uncertainties about green-
5Todd L. Walton Jr.Director, Beaches and Shores Resource Center,
Institute of Scienceand Public Affairs, Florida State
University
14
-
house gas emissions scenarios, temperature sensitivity of the
climatesystem, and glacial melt. Recent studies show that net
melting fromGreenland and the West Antarctic Ice sheet may be
happening (e.g.,Thomas et al., 2004; Cook et al., 2005; Chen et
al., 2006; Luthcke etal., 2006; Velicogna and Wahr, 2006; Shepherd
and Wingham, 2007.In 2007, the IPCC slightly lowered its estimate
of sea level rise to0.18 to 0.59 meters. The lowering was mainly
the result of changesin estimates of the contribution of thermal
expansion of the oceansto sea level. The IPCC admitted that it is
challenging to factorin the future contribution from melting ice
sheets. Using an em-pirical approach (comparing observed sea level
and temperatures)Rahmstorf (2006) projected that sea level will
rise 0.5 to 1.4 metersby 2100 over 1900. Empirical approaches can
be very sensitive torecent trends. These climatic modelers suggest
that recent globalwarming due to greenhouse gas accumulation in the
environmentfrom industrial consequences has brought about sea level
rise ac-celeration. The U.S. National Report to IUGG (1995) notes
that ”Recent analyses indicate that global sea level has risen at
somethingclose to 2 mm per year for at least the last century or
so” based onthe work of Peltier and Tushingham( 1989), Trupin and
Wahr(1990)and Douglas(1991), and also states that ”.... for the
next centuryvarious authors plausibly argue that global sea level
will rise at amuch faster rate than at present because of global
warming”. Thusmuch climatic modeling work is predicated on the
working basisthat global sea level is not only rising but also
accelerating due toincreasing levels of greenhouse gases in our
environment.
Just how these global sea level projections translate to
Floridaof vital importance to economic projections for both Florida
coastaldevelopment decisions as well as population growth
decisions.
The overall acceptance of an acceleration in sea level rise may
notbe an agreeable conclusion to all scientists based on past water
leveldata record analysis. For example, The IPCC (2007) in
reviewinginfomration on observed sea level rise concluded that:*
Since 1961, 80% of the climate system’s warming has been ab-sorbed
by the oceans and warming has been observed up to 3000meters in
depth.* Mountain glaciers (not included Antarctica and Greenland)
and
15
-
snow cover have decreased, contributing to sea level.* Since
1993, losses of glacial mass from Greenland and Antarcticahave
contributed to sea level rise.* Global sea level rose at a faster
rate in the 20th century than inthe 19th century. The rate of sea
level appears to have increasedsince the early 1990s, although
natural variability cannot be ruledout as a cause of the
increase.
In fact, it is felt by the author that future prediction should
bepragmatic in providing for worst case economic scenarios that
in-clude the possibility of sea level rise acceleration and let the
dataspeak for itself when it comes to future projections of sea
level. Pre-vious water level data related estimates of ”relative”
sea level risetrends have assumed a linear rise trend in sea level
which does notallow for acceleration/deceleration in relative sea
level with time.Such findings have primarily been based on
statistical findings thatan acceleration component of sea level
rise is not statistically dif-ferent from zero at a given
confidence level (see for example Zervas(2001)). Although it can
always be shown at some level of ”sta-tistical” significance that
higher terms in a non-linear approach tosea level rise may be
insignificant, it is rationalized here that from apragmatic
standpoint due to the physics behind climatic modelingand due to
the existing climatic studies suggesting acceleration ofsea level,
a higher order trend should be considered in sea level mod-eling to
assess forecasts based on the data while allowing for
suchacceleration/deceleration. Additionally, in a relatively
tectonicallystable area such as Florida, global sea level
acceleration as providedby climatic modelers would translate to a
relative sea level accelera-tion trend. It is on this basis that
the following forecasts of sea levelrise for the year 2080 are
estimated.
2.2 Data Sources
The data utilized in the present projected sea level rise
scenar-ios is from the National Oceanic and Atmospheric
Administration(NOAA) primary tide gage station network in Florida.
Althoughnumerous coastal tide stations exist in Florida, most have
opera-tional data for only short record periods and are not
suitable for
16
-
the analysis provided herein. Necessary length of data record is
aconsideration in picking which stations to utilize in sea level
riseanalysis as too short a record will not reflect a proper trend
whiletoo long a record will allow for non-stationarity in the data
seriesto hide important shorter term fluctuations that may govern
theforecast period. Pugh (1987) demonstrated that 10 year trends at
asite can have different signs, depending on the time interval
chosen.In a similar approach Douglas (1991) used the San Francisco
tidegage data ( the longest continuous record (140 years) in the
U.S.)and found that 30 year trends computed anywhere in the entire
se-ries varied from -2 to +5 mm per year using linear trend
analysis.His findings are suggestive that a 30 year record would be
too shortfor analysis (and consequent forecasting/extrapolation).
In anotherfinding, Emery and Aubrey (1991) noted strong coherence
of resultsfor sea level records longer than 40-50 years which might
be sug-gestive that such a period is reasonable for forecasting
future sealevels. Roemmich investigated sea level records at
Bermuda andCharleston, S.C., and found that coastal and nearby
mid-ocean sealevel trends differ markedly over several decades. His
conclusionssuggest that 50 year records of sea level are necessary
to understandthe fluctuations at a given coastal location. In
concurrence with thefindings above, the tide stations utilized all
have 50 years or longerof available historical data record.
Stations and station numbersutilized in the analysis are listed
below in Table 3.
Table 3 — Florida Tide Stations Used with at Least 50 Years
ofHistorical Data
Station Name Station NumberFernandina, FL 8720030Key West, FL
8724580St.Petersburg, FL 8726520Cedar Key, FL 8727520Pensacola, FL
8729840
All of the tide station gages utilized are in somewhat
protectedwaters, which is the reason for the more complete analysis
recordsavailable. Although open coast tidal stations might be
expected tohave a higher water level due to the effects of wave
setup, since theanalysis herein is aimed at projecting differences
in water level frompresent to the year 2080, this difference in
water level is not im-
17
-
portant for the purposes herein. The fact that the data records
areless contaminated by wave setup effects is in fact a benefit for
thepresent analysis which aims at projecting the low frequency
waterlevel rise over an approximately 75 year period.
The monthly mean sea level series was utilized from each of
theabove gages for the analysis herein. An additional gage with a
longhistorical period of record is that of Mayport ,FL (Station
Number8720220) but data was not available for the period 1999
through2005, hence this gage was not analyzed. Additionally, the
proximityof the Mayport and Fernandina gages was investigated and
it wasfound that the two gages had a strong linear correlation
between thedata sets showing that either of the gages could be a
proxy for theother. The Fernandina gage also had an extended period
of datamissing (1960-1969) but as the data was missing from the
middleportion of the historical series rather than at the end of
the seriesit was felt that the Fernandina record would provide a
more mean-ingful analysis period than the Mayport historical
series.
2.3 Methodology For Sea Level Forecasts
To keep the sea level rise scenario projections on the same
timeperiod footing, a starting date of January 1941 was used to
providethe historical parameter fitting (except for the St.
Petersburg se-ries for which available data started in 1947). This
was the earliestmonthly mean data available for the Key West tide
gage stationand hence the other series records were accordingly
shortened (ex-cept for St. Petersburg) for the analysis provided
herein. The fithistorical series period of record was from January
1941 (January1947 for St. Petersburg series) through December of
2005, while theforecast period of record was from 2006 to 2080 with
the projectedestimates at year end 2080 provided. Since the
historical period ofrecord utilized in fitting spanned
approximately 69 years (63 yearsfor the St. Petersburg series) it
is believed that projection (extrap-olation) to a forecast period
of approximately 75 years is reasonable(i.e. roughly equal time
spans of historical fit and future forecast).
18
-
Although the investigated station series are complete for the
mostpart, there are missing values in station records for some
months(as shown in later graphics) which do not allow for analysis
tech-niques such as linear or non-linear filtering which typically
requirecomplete data series. Rather than attempt to provide
estimates ofunmeasured data to fill in the incomplete series (i.e.
see Walton(1996)), techniques utilized in the analysis were limited
to both lin-ear and non-linear least squares analysis along with
seasonal meanestimation which can be applied to incomplete series
data.
As noted previously the model fitting is predicated on a
workingassumption that global sea level (and consequently for
Florida rela-tive sea level) is not only rising but also
accelerating due to climaticinfluence of greenhouse gases. This
scenario will later be comparedusing the same approach and data
only with standard linear leastsquares estimation for the
non-acceleration assumption.
As climatic modelers have provided the global sea level rise to
bean exponential rise in form, the nature of the model for relative
sealevel rise for the Florida sea level stations was chosen of a
similarform, i.e.
y(t) = p1 + p2 ∗ exp(p3 ∗ t)
(1)
which can be expanded in series form to
y(t) = p1 + p2(1 + p3 ∗ t + (p3)2 ∗ t2/2 + ....higher order
terms)
(2)
where t represents the time component (i.e. the monthly meansea
level index) and where The modeled y is a seasonally filteredwater
level developed by removing the seasonal (monthly) meansof the
monthly mean sea level series. It should be noted that the
19
-
y(t) series being fit is not the raw data but rather the
deseasonal-ized data residual where the actual fit or forecast data
would be themodeled or forecast y(t) with the seasonal (month)
average added.The use of the seasonal averaging filter was to
reduce the noiseof the fit and hence provide more parameter
stability, along withthe fact that missing data in the raw data
series did not allow fortypical linear or non-linear filtering
approaches without making ”es-timates” about the missing data. The
approach utilized makes noapriori assumptions regarding missing
(unavailable) data. It shouldbe noted here that although the model
to be fit is assumed to be ofa non-stationary exponential form in
this expansion approach, a se-ries expansion of a stationary
harmonic model approach can also beshown to lead to a higher order
polynomial model with dependentcoefficients.
Equation 2 can be simply reformulated as a linear higher
orderpolynomial which in the case provided has been terminated
withthe second order. The approaches for model fitting involved
botha linear least squares approach to model parameter estimation
aswell as a non-linear least squares approach to model parameter
es-timation. As non-linear estimation routines require information
re-garding starting parameter values, a linear approach was
utilizedto formulate estimated starting values for model fitting in
the non-linear least squares approach. It should be noted that
non-linearestimation techniques are not guaranteed to provide
stable fit pa-rameter values but as will be seen, in many of the
water level gageseries fits to the data, the non-linear forecast
sea level rise was foundto be very close to the linear second order
forecast sea level rise thusconfirming the validity of the utilized
approaches. Due to the factthat most of the gages fit provided
comparable values by the twotechniques, the linear second order
forecast sea level rise was chosenfor projecting final sea level
rise scenarios in the year 2080. The lin-ear first order sea level
rise is also provided for comparison purposes.
The fit y(t) series are shown as the ordinate values in Figures1
through 5 where the abscissa is the monthly time index (i.e.
1=January 1941, [ except for St. Petersburg where 1= January
1947]).The de-seasonalized monthly mean sea level data during the
his-torical period are shown as points on the graphs and the solid
line
20
-
represents the de-seasonalized historical sea level data fit
during thespan of the historical data and the de-seasonalized sea
level forecastcurve during the forecast period. The estimate of sea
level rise fromyears 2006 to 2080 is the difference in the solid
line between thefinal forecast time (2080) and the final historical
time (2005) andis summarized in Table 4. For informational
purposes, a second setof forecast sea level rise values is also
provided in Table 4 for theshorter time forecast horizon from 2006
to year 2030.
Table 4 — Table 4. Forecast Relative Sea Level Rise From 2006
to2080 by Florida County
County 2006-2030 2006-2080(in meters) (in meters)
Monroe 0.0845 0.310Escambia 0.0887 0.343Dade 0.0845 0.310Dixie
0.0714 0.275Duval 0.073 0.254Wakulla 0.0827 0.319
As a check on the utilized approach to filtering the data series
byremoving the monthly means, a second approach was used on theKey
West data series in which the data series was not
de-seasonalized(i.e. monthly means were not removed), but the
entire data set wasfit utilizing two additional parameters to
represent the monthly se-ries as a harmonic. This approach
considered a monthly cycle of theform with T=12 [months] for the
yearly cycle and the two unknownsbeing Amp= Amplitude (m) , and
Phase = Phase of cycle (radians).For the Key West series, a
forecast to the year 2080 produced theexact same sea level rise
forecast result (to 2 decimal places) as theprevious
de-seasonalized approach and additionally produced simi-lar
Gaussian residual magnitudes.
y(t) = Amp ∗ (cos(2 ∗ pi ∗ t/T − phase)) + p1m + p2m ∗ t + p3m ∗
t2
(3)
An interesting result from the analysis is that for the
differentgage sites which are widely spaced over the Florida
Peninsula, the
21
-
projected sea level rise in year 2080 does not vary by much
withthe largest value being 0.35 meters in St. Petersburg, FL,
while thesmallest value is 0.25 meters in Fernandina, FL.
Additionally in allbut the Fernandina gage data, the second order
non-linear term hasa parameter that is statistically different from
zero (and positive)at a 95 percent significance level. It is
believed that the Fernand-ina series provided a less than
significant second order term due tothe large gap in the data and
the higher tidal range that may beresponsible for magnification of
error in the residual modeled.
Table 5 is provided to compare the sea level rise by the
threemodeling approaches utilized (i.e. the linear first order, the
linearsecond order, and the non-linear exponential)
Table 5 — Forecast Relative Sea Level Rise from 2006 to 2080
Station Relative SeaLevel Rise(inmeters) 1stOrder
Relative SeaLevel Rise(inmeters) 2ndOrder
Relative SeaLevel Rise(inmeters)Exponential
Fernandina, FL 0.16 0.25 0.27Key West, FL 0.15 0.31
0.28St.Petersburg, FL 0.18 0.35 0.36Cedar Key, FL 0.11 0.27
0.16*Pensacola, FL 0.13 0.34 0.21*
* parameter estimation suspect due to convergence problems
Table 4 suggests that for gages where non-linear estimation
con-vergence was obtained, both the second order linear model and
theexponential model were comparable as previously noted. This
tablealso shows that the linear first order sea level rise
estimates were onthe order of one half of the linear second order
sea level rise esti-mates. Similar linear first order estimates can
be projected from sealevel rise rates provided in Zervas (2001).
The fact that the secondorder forecasts provided greater sea level
rise than that providedby the linear ”standard” approach suggests
that the scenario of anacceleration in sea level rise rather than a
deceleration in sea levelrise is a more likely scenario on the
basis of the actual data avail-
22
-
able. Residuals from the data fitting procedure for the Florida
gagesare provided in Figures 1 through 5, and show that the data
resid-uals for the methodology utilized provide reasonable Gaussian
bellshaped curves suggestive that the higher order fitting is
satisfactory.
2.4 Summary
Relative sea level rise has been forecast from the present
(year2006) to the future year 2080 for long term water level gages
aroundthe Florida Peninsula by three different methods. The second
orderlinear approach is recommended in the final analysis for
projectingeconomic scenarios of future costs due to sea level rise
due to itsinclusion of a higher order term that allows for
acceleration in sealevel rise in accord with climate modeling
scenarios that project anexponential sea level rise due to
greenhouse gas effects. Althoughthe present work is not definitive
in regard to an accelerating sealevel rise, it is clear at least
from the data available that trends areconsistent that there is not
a deceleration in sea level rise and apragmatic approach to future
economic planning should be in tunewith climatic model scenarios
that suggest the strong possibility ofan accelerating sea level
rise in Florida and future values of sea levelrise on the order of
the magnitude herein.
Figure 2 — Fernandina Gage Station Forecast Filtered Sea
LevelRise
23
-
Figure 3 — Key West Gage Station Forecast Filtered Sea Level
Rise
Figure 4 — St. Petersburg Gage Station Forecast Filtered Sea
LevelRise
24
-
Figure 5 — Cedar Key Gage Station Forecast Filtered Sea
LevelRise
Figure 6 — Pensacola Gage Station Forecast Filtered Sea Level
Rise
25
-
3 Economic Analysis of Sea Level Rise
Property owners and visitors along hurricane-prone coastal
areasshould be aware of hurricane information including path,
intensityand potential corresponding damage in order to prepare for
propertyprotection and survival. While precise hurricane forecast
informa-tion from the National Hurricane Center is easily obtained,
infor-mation is less readily available regarding potential damage
costsassociated with hurricanes. Insurance and re-insurance
companiesoften perform these types of analyses, however, their
results are notfrequently released to the public. The public still
does not have con-clusive and consensus evidence of increased
intensity and frequencyof storms as a result of climate change. The
estimation of dam-age cost is also variable, based on different
methodologies. Windspeed along the hurricane path is generally used
to predict damagecosts. In order to gain more precise damage
estimates associatedwith sea level rise, the objective of this
section was to link damagecost and storm return years (based on
storm surge) as a means ofproviding beach protection options or
mitigation strategies such asflood proofing, elevated structure,
and building codes, among oth-ers. Since storm surge is positively
correlated with wind speed, (CN-MOC, 2005) one can predict a
positive relationship between flooddamage and wind damage. For this
analysis, storm surge data basedon return period was linked to
historical damage cost data to yielda potential future total damage
cost for various size hurricanes6 foreach of the six counties. The
economic analysis involved two var-ied methodologies to measuring
cost damage associated with sealevel rise. One is termed ”Hurricane
Return Years” and the otheris ”Damage Cost”7. The statistical
methodology for ”HurricaneReturn Years” estimation and attendant
confidence and predictionlimits has been detailed by Johnson and
Watson (1998). The ”Dam-age Cost” analysis is based on insurance
claims data provided by theFlorida Office of Insurance Regulation
(FOIR). The FOIR empha-sizes that although they do analyze the
insurance claimant data forcompleteness and reasonability, they do
not formally audit nor verify
6Hurricane strength in terms of Category 1-5.7In order to fit
both graphs in one figure, CEFA selected hurricanes that would fit
most
appropriately on both the reduction of hurricane return year and
hurricane cost damageassessment figures. Hence, a few counties
figures will extend slightly beyond the slope on oneor two
figures.
26
-
the data. It should be noted that the following results are
location-specific for the six Florida counties and on an individual
hurricanebasis, and are not representative in application to all
storm eventsin all Florida counties.
3.1 Regression Model(s) For Hurricane Return Years andDamage
Cost
From the late 1980’s to current years, the Federal Emergency
Man-agement Agency (FEMA) estimated stillwater elevations and
stormreturn year(s) for Florida counties. The storm surge
stillwater eleva-tions are a function of tidal and wind setup
effects, and contributionsfrom wave action. The storm surge
elevations for the storm returnyear (10, 50, 100 and 500-year)
floods were estimated by FEMA foreach of the six counties. FSU CEFA
expressed this linear relation-ship as equation (4) for the study
area(s).
The purpose of the FEMA Flood Insurance Studies was to
in-vestigate the existence and severity of flood hazards in the
areas ofFlorida Counties, and to aid in the administration of the
NationalFlood Insurance Act of 1968 and the Flood Disaster
Protection Actof 1973. FSU CEFA applied the Florida Office of
Insurance Regu-lation data to generate storm surge and historical
damage cost ofeight hurricanes in 2004 and 2005 (using
interpolation and regres-sion) presented as equation (5). The data
sources for Equation 5were obtained from the Hurricane Summary
Data, Florida Office ofInsurance Regulation, 2006 (in 2006 Dollars)
and the Tropical Cy-clone Report for 2005.
Hurricane Return Years8
Y = a + bX
(4)
whereY = Surge(ft)9, X = Y ear
8Return periods capture the essence of uncertainty in extreme
meteorological phenomena(storm surge, wave, and wind) associated
with hurricanes.
9Storm surge is simply water that is pushed toward the shore by
the force of the windsswirling around the storm.
27
-
Cost Damage Assessment10
Y = c + dM
(5)
whereY = Surge(ft), M = Damage Cost
For example, in Wakulla County, the regression equation forstorm
surge and storm return period is presented as:
Y = 0.271412 + 5.90899logX
and the regression for storm surge and historical damage cost
is
Y = 4.763688 + 0.922944M
(M in million dollars).
The regression summary results for regression equation 4
(elevationreturn years) are presented in Table 6.
Table 6 — Regression Results for Elevation Return Years by
County
County T Stat P-Value R SquareDade Intercept 17.26 Intercept
0.003* 0.99
X variable 13.69 X variable 0.005*Dixie Intercept 4.89 Intercept
0.039* 0.99
X variable 14.89 X variable 0.004*Duval Intercept 2.30 Intercept
0.26 0.99
X variable 19.30 X variable 0.03*Escambia Intercept -1.39
Intercept 0.30 0.99
X variable 68.39 X variable 0.0002*Monroe Intercept 1.02
Intercept 0.42* 0.90
X variable 4.33 X variable 0.05*Wakulla Intercept 0.56 Intercept
0.63 0.99
X variable 23.69 X veriable 0.002*
Note:*95% Level of Significance
10Regression with 95% confidence
28
-
3.2 Hurricane Return Year(s)
Table 7 illustrates the results of the storm surge elevation
regres-sion(s) for current conditions, for the FSU SLR estimates to
Year2080, and the corresponding IPCC SLR estimates to Year 2080,for
the study area counties. FSU CEFA selected recent hurricaneevents
which were representative of the counties, and with
readilyavailable data for cost damage and storm surge elevations.
This sec-tion highlights the reduction of hurricane return years
findings foreach county. As can be expected, with both the FSU and
IPCC es-timates for SLR, there is a consistent reduction in
hurricane returnperiod. That is, one would expect that with SLR, a
hurricane eventwith similar storm surge would occur more
frequently. For example,a hurricane with surge of 7 feet would be
expected under normalconditions to occur once every 76 years.
Including FSU and IPCCSLR estimates, the same hurricane storm surge
can be expected ev-ery 21 and 5 years, respectively.
Table 7 — Hurricane Return Year(s) for Recent Hurricane Eventsby
County
County Hurricane Surge(ft)
ReturnPeriod(yrs)
SLR Estimates (ft) ReturnPeriod(yrs)
Dade Wilma 7.00 75.79 FSU 2030 0.28 51.252080 1.02 20.60
IPCC 2030 0.49 39.522080 2.13 5.18
Dixie Dennis 9.00 13.59 FSU 2030 0.06 13.282080 0.90 9.59
IPCC 2030 0.49 11.262080 2.13 6.00
Duval Frances 5.9 100.00 FSU 2030 0.24 80.172080 0.83 47.00
IPCC 2030 0.49 63.682080 2.13 14.03
Escambia Dennis 12 845.99 FSU 2030 0.29 731.802080 1.12
470.03
IPCC 2030 0.49 657.352080 2.13 272.41
Monroe Wilma 2.76 7.35 FSU 2030 0.28 6.042080 1.02 3.61
IPCC 2030 0.49 5.222080 2.13 1.65
Wakulla Dennis 9 29.96 FSU 2030 0.27 27.142080 1.05 20.40
IPCC 2030 0.49 25.052080 2.13 13.70
29
-
Wakulla County -
The ground elevations are typically verylow throughout the
county, ranging fromsea level to ten feet National Geodetic
Ver-tical Datum (NGVD) near the coast togreater than 25 feet near
the northern partof the county. As mentioned in the FEMAstudy11,
the main flood hazard in terms ofdamage to Wakulla County is the
inunda-tion of low-lying coastal areas during thepassage of a
severe hurricane or tropicalstorm. The coastal area is very prone
toextreme storm tides. The storm surge elevations are higher is
cer-tain areas (west and south of Apalachee Bay) for two reasons;
shal-low water depths extend a great distance offshore, thereby
increas-ing the effect of bottom and wind friction which results in
higherstorm surge elevations and second, storm generated winds out
ofthe south-southeast create a flow of water in a north-west
directionalong Florida’s west coast into Apalachee Bay.
A hurricane with an elevation of 12.5 feet, would have a return
pe-riod12 of 100 years (i.e., that hurricane with associated storm
surge12.5 feet in elevation would most likely occur once every 100
years).Hurricane Dennis resulted in a 9 feet high surge in Wakulla
County.Based on FEMA’s study, it was classified as a 30 year event
hur-ricane (using aforementioned Equation 4). For a sea level rise
of0.27-feet and 1.05-feet scenarios (FSU estimates to year 2030
and2080), the same hurricane storm surge as Dennis will be
reducedfrom 30 years to 27.14 years and 20.40 years respectively;
for 0.49-feet and 2.13-feet scenarios (IPCC estimates to year 2030
and 2080),the same hurricane storm surge will be reduced to 25.05
years and13.70 years. Figure 7 indicates the scenario of reduction
of hurricanereturn year(s) in Wakulla County, according to the IPCC
and FSUSLR estimates to year 2080 . We can expect an increasing
return pe-
11Federal Emergency Management Agency, Flood Insurance Study,
Wakulla County Unin-corporated Areas, 1986. Community Number -
120315.
12Hurricane Return Years capture the essence of uncertainty in
extreme meteorologicalphenomena (storm surge, wave, and wind)
associated with hurricanes.
30
-
riod frequency with a higher sea level rise scenario. This
translatesinto a significant impact for coastal area residents.
Since frequencyof the same extent storm return is increasing, the
importance of en-suring adequate housing and beach protection seems
critical in thethreatened areas.
Figure 7 — Reduction of Hurricane Return in Wakulla
CountyYear(s) by Elevation Based on IPCC and FSU SeaLevel Rise
Estimates
31
-
Dade County -
The Dade County Flood InsuranceStudy was used for developing
actuarialflood insurance rates, to update existingfloodplain
regulations and by local and re-gional planners to further promote
soundland use and floodplain development13. Thestudy encompassed
all of Dade County,with the exception of the Everglades Na-tional
Park (about 1/3 of the county). DadeCounty is flat and low with
elevations gen-erally below ten feet NGVD. The westernand southern
areas are marshy with a mean elevation of around 5feet mean sea
level (MSL).
Hurricane Wilma resulted in a seven feet high surge in
DadeCounty. Based on FEMA’s study, it was classified approximately
asa 75.80 year event hurricane. For a sea level rise of 1-feet
scenario(historical tidal gauge records FSU SLR estimate to year
2080 of 31cm), the same density hurricane as Wilma would be reduced
froma 75.80 year to a 20 year event; for the IPCC SLR estimate to
year2008, 2.132 ft (65 cm), it will reduced to be 5.20 years
(Figure 8).Additionally, for a sea level rise of 0.28-feet scenario
(FSU estimatesto year 2030), the same hurricane storm surge as
Wilma will be re-duced from 75.80 years to 51.25 years; for
0.49-feet scenario (IPCCestimates to year 2030), the same hurricane
storm surge will be re-duced to 39.52 years.
13Federal Emergency Management Agency, Flood Insurance Study,
Dade County and In-corporated Areas, 1994. Community Number -
120635 - 120661.
32
-
Figure 8 — Reduction of Hurricane Return Year(s) in Dade
Countyby Elevation Based on IPCC and FSU Sea Level
RiseEstimates
Dixie County -
The purpose of the Dixie County FloodInsurance Study was to
investigate the exis-tence and severity of flood hazards in
DixieCounty, and to aid in the administration ofthe National Flood
Insurance Act of 1968and the Flood Disaster Protection Act
of197314. The study was also used to con-vert Dixie County to the
regular program offlood insurance by the FEMA. In addition,local
and regional planners use the studyto promote sound flood plain
management.The study covered all of the unincorporated areas of
Dixie County.Dixie County is on the North Florida Gulf Coast, about
90 milessouth of Tallahassee, and about 100 miles north of Tampa.
DixieCounty is low in elevation, with gently sloping and poorly
drained
14Federal Emergency Management Agency, Flood Insurance Study,
Dixie County and Un-incorporated Areas, 1983. Community Number -
120336.
33
-
marshy areas. The elevations range from 10 feet NGVD to
higherareas in the northern portion of the county, extending to 60
feetNGVD. Hurricane Dennis resulted in a 9 feet high surge in
DixieCounty. Based on FEMA’s study, it was classified approximately
asa 15 year event hurricane. For a sea level rise of 1-feet
scenario (his-torical tidal gauge records FSU SLR estimate to year
2080 of 31 cm),the same hurricane storm surge as Dennis would be
reduced from a15 year to less than a 10 year event; for the IPCC
SLR estimate toyear 2008, 2.132 ft (65 cm), it will reduced to be
9.60 years (Figure9). For a sea level rise of 0.06-feet scenario
(FSU estimates to year2030), the same hurricane storm surge as
Dennis will be reducedfrom 13.60 years to 13.28 years; for
0.49-feet and 2.13-feet scenarios(IPCC estimates to year 2030 and
2080), the same hurricane stormsurge will be reduced to 11.26 years
and 6 years.
Figure 9 — Reduction of Hurricane Return Year(s) in Dixie
Countyby Elevation Based on IPCC and FSU Sea Level
RiseEstimates
34
-
Duval County -
The flood insurance study for DuvalCounty was used by the
community to up-date existing floodplain regulations as partof the
regular phase of the National FloodInsurance Program15. The
information wasalso used by local and regional planner topromote
sound land use and floodplain de-velopment. Due to flat terrain,
many ar-eas inland experience shallow flooding af-ter a heavy
rainfall. The City of NeptuneBeach is partially protected from the
At-lantic Ocean by a seawall.
Hurricane Frances resulted in a 5.9 feet high surge in
DuvalCounty. Based on FEMA’s study, it was classified as a 100
yearevent hurricane. For a sea level rise of 0.24-feet and
0.83-feet sce-narios (FSU estimates to year 2030 and 2080), the
same hurricanestorm surge as Frances will be reduced from 100 years
to 80.17years and 47 years respectively; for 0.49-feet and
2.13-feet scenar-ios (IPCC estimates to year 2030 and 2080), the
same hurricanestorm surge will be reduced to 63.68 years and 14.03
years. Figure10 indicates the scenario of reduction of hurricane
return year(s) inDuval County, according to the IPCC and FSU SLR
estimates toyear 2080 .
15Federal Emergency Management Agency, Flood Insurance Study,
City of Neptune BeachDuval County 1989. Community Number -
120079.
35
-
Figure 10 — Reduction of Hurricane Return Year(s) in DuvalCounty
by Elevation Based on IPCC and FSU SeaLevel Rise Estimates
Escambia County -
The goal of the Escambia County FloodInsurance Study was used
for developingactuarial flood insurance rates, to updateexisting
floodplain regulations and by lo-cal and regional planners to
further pro-mote sound land use and floodplain devel-opment, and to
aid in the administration ofthe National Flood Insurance Act of
1968and the Flood Disaster Protection Act of197316. The study
covered the entire geo-graphic area of Escambia County. The
ter-
16Federal Emergency Management Agency, Flood Insurance Study,
Escambia County andIncorporated Areas, 2000. Community Number -
120080,120082,120084,125138.
36
-
rain in Escambia County is highly variable. Level to
moderatelysloping terrain are indicative of the southwest portion
of the county(west of Pensacola). The soils here are somewhat
impermeable,and poorly drained. In addition, in the southwestern
area are flat,low and marshy areas. In the central and northern
portions of thecounty, there are rolling forested hills and
moderately steep slopes.Elevations in these areas may reach up to
300 feet. Flooding inEscambia County is normally a result of tidal
surge and overflow ofstreams and swamps associated with rainfall
runoff.
Hurricane Dennis resulted in a 12 feet high surge in
EscambiaCounty. Based on FEMA’s study, it was classified as
approximatelya 900 year event hurricane. For a sea level rise of
0.29-feet and1.12-feet scenarios (FSU estimates to year 2030 and
2080), the samehurricane storm surge as Dennis will be reduced from
855 years to731.80 years and 470.03 years respectively; for
0.49-feet and 2.13-feet scenarios (IPCC estimates to year 2030 and
2080), the samehurricane storm surge will be reduced to 657.35
years and 272.41years. Figure 11 indicates the scenario of
reduction of hurricanereturn year(s) in Escambia County, according
to the IPCC and FSUSLR estimates to year 2080 .
37
-
Figure 11 — Reduction of Hurricane Return Year(s) in
EscambiaCounty by Elevation Based on IPCC and FSU SeaLevel Rise
Estimates
Monroe County -
The flood insurance study (FIS) forMonroe County was used to aid
in the ad-ministration of the National Flood Insur-ance Act of 1968
and the Flood Disasterprotection Act of 1973. The FIS was alsoused
to develop flood risk data for variousareas of the county that were
used to estab-lish actuarial flood insurance rates and as-sist the
county to promote sound floodplainmanagement17. The FIS covered the
entiregeographic area of Monroe County. Resi-dential, commercial
and industrial development in Monroe Countyoccurs mainly along the
Florida Keys. The mainland remains largely
17Federal Emergency Management Agency, Flood Insurance Study,
Monroe County andIncorporated Areas, 2005. Community Number -
12087CV000A.
38
-
undeveloped and includes the Big Cypress National Preserve
andEverglades National Park. Coastal areas bordering the
AtlanticOcean and Gulf of Mexico are subject to storm surge
flooding dueto hurricanes and tropical storms. Flood protection
measures arenot known to exist in Monroe County.
Hurricane Wilma resulted in a 2.76 feet high surge in
MonroeCounty. Based on FEMA’s study, a 100 year event hurricane
wouldbe associated with a 6.75 feet high storm surge. For a sea
level riseof 0.28-feet and 1.02-feet scenarios (FSU estimates to
year 2030 and2080), the same hurricane storm surge as Wilma will be
reducedfrom 7.30 years to 6.04 years and 3.61 years respectively;
for 0.49-feet and 2.13-feet scenarios (IPCC estimates to year 2030
and 2080),the same hurricane storm surge will be reduced to 5.22
years and1.65 years. Figure 12 indicates the scenario of reduction
of hurricanereturn year(s) in Monroe County, according to the IPCC
and FSUSLR estimates to year 2080 .
Figure 12 — Reduction of Hurricane Return Year(s) in
MonroeCounty by Elevation Based on IPCC and FSU SeaLevel Rise
Estimates
39
-
3.3 Damage Cost Assessment
There are significant cost damages associated with hurricane
events.FSU CEFA compiled data from the latest two years (2004 and
2005)from the Hurricane Summary Data Reports, presented in Tables
8and 9. Although the storm events listed below were of the
strongercategory rating, the level of corresponding cost damages
was highlyvariable. The cost damages were a function of storm
intensity andproximity of the study area county to the associated
hurricane.
Table 8 — Storm Events Totals (in 2006 $)by County of
LossOccurrence, 2004
County Charley Frances Ivan JeanDade $3,008,721 $70,468,075
$2,865,950 $16,170,268Dixie $36,408 $4,945,128 $63,237
$971,682Duval $5,906,950 $72,322,498 $1,649,646 $22,404,237Escambia
$1,001,182 $12,980,961 $2,010,001,983 $19,105,056Monroe $663,804
$4,945,128 $363,295 $133,665Wakulla $14,047 $1,854,422 $214,588
$193,451
Table 9 — Storm Event Totals (in 2006 $) by County of
LossOccurrence, 2005
County Dennis Katrina Rita WilmaDade $5,976,177 $585,157,998
$4,396,620 $2,152,438Dixie $59,559 $1,742 $661 $33,104Duval
$361,426 $831,764 $151,072 $1,055,752Escambia $70,706,486
$11,341,048 $150,867 $283,996Monroe $4,400,998 $27,907,960
$11,329,370 $215,335,831Wakulla $4,418,483 $588,457 $1,274
$28,279
*Hurricane Summary Data, Florida Office of Insurance Regulation,
2006
As mentioned previously in the model description and
regressionequation(s), the regression for storm surge and
historical damagecost is Y = 4.76+0.92M (in $2006 dollars).
Overall, the damagecost estimates for recent hurricanes range from
thousands to billionsfor the six counties. By using storm surge
data and damage costestimates data; the FSU SLR estimates (which
varied by county)and the IPCC standard SLR estimates of 2.13 feet
(or 65 cm) wereapplied to the initial storm surge level to express
a new regressionfor the damage cost estimates. The damage cost
estimate equationsare presented in Table 10.
40
-
Table 10 — Damage Cost and Storm Surge Regression Equationsfor
Recent Hurricane Events by County
County Regression EquationWakulla Y = 4.76+0.92MMonroe Y =
0.86+0.02MDade Y = 1.40+2.67MDuval Y = 0.87+0.07MDixie Y =
2.82+0.1MEscambia Y = 4.93+0.10M*Y=Storm Surge(feet), M=$2006
Dollar
Table 11 — Sea Level Rise, Cost Damage and Storm SurgeEstimates
by County
County Hurricane Surge(ft)
DamageCost
SLR FSUEsti-mates(ft)18
DamageCost
SLRIPCCEsti-mate(ft)
DamageCost
Dade Wilma 7.00 $2.10B 1.02 $2.48B 2.132 $2.90BDixie Dennis 9.00
$61
Thous0.91 $70.90
Thous2.132 $83.12
ThousDuval Frances 5.90 $69
Thous0.82 $80.15
Thous2.132 $98.00
ThousEscambia Dennis 12.00 $73
Thous1.12 $84.51
Thous2.132 $95.00
ThousMonroe Wilma 2.76 $233M 1.02 $298.40M 2.132 $370.13MWakulla
Dennis 9.00 $4.59M 1.05 $5.73M 2.132 $6.90M
A storm similar to Hurricane Dennis in Wakulla County couldbring
50% more damage from $4.59 million to almost $7 million,as depicted
in Figure 12. Dade County, from a Hurricane Wilmadamage cost
perspective, would bring about a 40% increase in costdamages to
around $3 billion (Figure 13). Dixie County experi-enced relatively
low cost damages (based on filed reports) so wouldexperience a 37%
increase in cost damages based on damages associ-ated with
Hurricane Dennis (Figure 14). Duval County (HurricaneFrances
strength hurricane) would result in a 42% increase (Figure15).
Escambia County (Hurricane Dennis size) would result in a30%
increase, to $95 Million (Figure 16). Monroe County (Hurri-cane
Wilma, Figure 17) would experience a 60% increase in costdamages
associated with a Wilma-sized hurricane in the year 2080(according
to IPCC estimates of 65 cm).
41
-
Figure 13 — Damage Cost and Storm Surge Estimates in
WakullaCounty
Figure 14 — Damage Cost and Storm Surge Estimates in
DadeCounty
42
-
Figure 15 — Damage Cost and Storm Surge Estimates in
DixieCounty
Figure 16 — Damage Cost and Storm Surge Estimates in
DuvalCounty
43
-
Figure 17 — Damage Cost and Storm Surge Estimates in
EscambiaCounty
Figure 18 — Damage Cost and Storm Surge Estimates in
MonroeCounty
44
-
3.4 Economic Analysis of Property Loss for the Six
FloridaCoastal Counties
The objective of this phase of the project was to determine
thevalue of land that would be affected by SLR over time for the
sixFlorida coastal counties19, based on the 2030 and 2080 sea level
risescenarios. CEFA used both the FSU Beaches and Shores
historicaltide gage data (for years 2030 and 2080), and the IPCC
highest es-timates (for years 2030 and 2080). The four sea level
rise scenariosprovide a range of property values affected by sea
level rise over time.
FSU CEFA and Industrial Economics Inc., under the guidanceof
James Neumann, created a GIS ArcView Model Builder tool toprocess
Florida parcel data20 and combine it with Digital ElevationModel
(DEM) elevation data (assigned to a tax parcel centroid).The model
was termed the ”’storm model”’. DEM is a digital rep-resentation of
ground surface topography or terrain.
The DEM data was downloaded from http://seamless.usgs.gov/for
each of the six counties. For some of the counties, 10 meterDEM was
available for only certain sections of the county. FSUCEFA decided
that if 10M was not available for the entire county,then 30 meter
resolution DEM would be feasible21. Using the DEMdata, FSU CEFA
next generated a set of contour maps for eachcounty. CEFA used a
surface analysis tool (3D Analyst) to createthe contours. The first
contour was baselined at zero with a Z factor(increment) of one,
then the next contour was drawn for one meter,and so forth. The
parcel data map was then placed (overlaid) overthe contour map. A
centroid points file (with unique identifier) wasnext created using
X Tools Pro. One centroid was attributed to eachparcel.22 was then
used to generate the parcel elevation file and to
19Dade, Duval, Dixie, Escambia, Monroe and Wakulla
Counties20Florida Department of Revenue DR-590 (12D.8) parcel data.
CEFA did not ver-
ify/validate the FDOR data, thus, the results of this study are
based on the assumptionof the quality or correct topology of the
FDOR parcel data. CEFA used the variable ”JustValue” to best
reflect property values. The Department of Revenue uses just value
for prop-erty tax assessment and recommended it’s use (personal
communication, Dr. Ke-tsai Wu,FDOR Statistician)
21Dade, Escambia and Wakulla County were available for 10 meter
DEM. Dixie and MonroeCounty were available for 30 meter DEM. Duval
County only had a small portion of the countyavailable in 10 meter
DEM, primarily inland.
22The storm model derives centroid elevation values from a
spatial analyst function usingDEM data (parcels that are not within
the shoreline range of the DEM are deleted). The
45
-
process the output based on user defined sea level conditions.
Themodel works for any SLR or storm surge value, within the limits
ofthe DEM data, and appears to be returning both reasonable val-ues
and maps of inundated areas. The results presented in Table
6,provide an estimate of land value and land area23at risk, for the
sixcounty-wide data based on sea level rise scenarios.
3.5 ”Storm Model” Results
The results presented in Table 12, provide an estimate of
landvalue and land area at risk for the six county-wide data based
onsea level rise scenarios. The storm model findings convey a
consider-able loss of property due to varying sea level rise for
years 2030 and2080. For example, Monroe County will incur
significant damages inall four scenarios, ranging from
approximately $1 billion (Year 2030,FSU Beaches and Shores SLR
estimate) to $5.7 billion (Year 2080,IPCC SLR estimate), in 2007
dollars. Similarly, according to theIPCC 2080 SLR scenario, Duval
County will experience $644 mil-lion in damages. In general,
concerning the more rural areas, suchas Wakulla and Dixie Counties,
one would expect lower propertyvalues compared to the more
urbanized counties. However, with sealevel rise, all the counties
experience substantial property losses overtime.
The total land acres were much more difficult to quantify.
Forexample, Duval County, during that same time period
(2030-2050),can expect to lose approximately 31,000 land acres,
based on thedata provided to FSU CEFA by the Department of Revenue
landparcel database that was used for this study. This translates
to ap-proximately 6.3% of total land area of Duval County.
model joins centroid (with DEM values) to the parcel data based
upon parcel ID. It creates aselect statement that is user defined,
such that parcels below a given elevation are lost (andthis value
is calculated). Lastly, the model produces a shapefile of lost
values for both justvalues and land values.
23Based on the parcel database and the associated errors with
land values, these resultsare strictly based on the quality of the
parcel database. Dade, Dixie, Escambia and MonroeCounties are using
year 2005 parcel database; Duval County is using year 2004 parcel
database;Wakulla County is using year 2007 parcel database.
46
-
Table 12 — Land Value and Land Area at Risk from Four Sea
LevelRise Scenarios For Six Florida Counties
County SLR Estimates PropertyValuesat Risk*($2007)
TotalLandatRisk(Acres)
%Cov-eredbyDEM
% ofLandAreaatRisk
CountyLandArea(Acres)
Dixie FSU 2030 0.0174m $1,953,152 93 100 0.02 450,5602080 0.275m
$2,753,065 140 0.03
IPCC 2030 0.15m $2,656,830 122 0.032080 0.65m $4,275,514 190
0.04
Escambia FSU 2030 0.0887m $141,101,223 1,107 39.4 0.26
423,9362080 0.343m $279,520,579 3,196 0.75
IPCC 2030 0.15m $151,447,789 1,741 0.412080 0.65m $977,130,041
10,580 2.50
Monroe FSU 2030 0.0845m $922,092,573 4450 100 0.70 638,0162080
0.310m $3,264,205,082 14,671 2.30
IPCC 2030 0.15m $1,528,820,629 6,319 0.992080 0.65m
$5,706,227,005 22,756 3.57
Wakulla FSU 2030 0.0827m $269,960 47 68.5 0.01 388,2882080
0.319m $1,055,766 83 0.02
IPCC 2030 0.15m $409,620 57 0.012080 0.65m $2,230,401 181
0.05
Dade FSU 2030 0.0845m $14,835,863 148 57.8 0.01 1,245,5042080
0.310m $56,980,539 195 0.02
IPCC 2030 0.15m $29,852,956 159 0.012080 0.65m $128,065,286 243
0.02
Duval** FSU 2030 0.073m $35,010,480 4,354 $100 0.88 495,1682080
0.254m $263,831,552 11,284 2.28
IPCC 2030 0.15m $171,618,164 6,307 1.272080 0.65m $643,562,482
30,945 6.25
* Based on Florida Department of Revenue DR-590 (12D.8) Parcel
Data Just Values for2005.** Duval County had only a small portion
of 10 meter DEM data available, mostly inland.
3.6 Storm Surge Results
Table 13 presents storm surge estimates based on the flood
in-surance studies and for the insurance regulation data (for the
sixcounties) for sea level rise estimates to 2080.
47
-
Table 13 — Storm Surge Projections Based on FSU and IPCC
SLREstimates
County Hurricane Surge(ft)
SLRFSUEsti-mates(ft)
ProjectionsBasedon FSUSLR Es-timates(ft)
SLRIPCCEsti-mate(ft)
ProjectionsBased onIPCCSLR Esti-mates
Dade Wilma 7.00 1.02 8.02 2.132 9.13Dixie Dennis 9.00 0.91 9.91
2.132 11.13Duval Frances 5.90 0.82 6.72 2.132 8.03Escambia Dennis
12.00 1.12 13.12 2.132 14.13Monroe Wilma 2.76 1.02 3.78 2.132
4.89Wakulla Dennis 9.00 1.05 10.05 2.132 11.13
Appendix A describes the project data sources for the GIS
map-ping portion of this study, in greater detail. Appendix B
(Figures31 through 42) depict the joint representation of hurricane
returnperiod and associated cost damages based on SLR estimates.
Ap-pendix C (Figures 43 through 48) illustrate sample coastal area
sec-tions of the six county storm model results for years 2030 and
2080,based on FSU Beaches and Shores and IPCC sea level rise
estimates.As can be expected with higher and variable SLR
estimates, greaternumbers of land parcels become inundated over
time.
3.7 Alternate SLR Impact Analysis Using the GIS StormModel
Regression Models
To better illustrate the marginal increase rate of property
valuesat risk with associated SLR, we implemented the
aforementioned’Storm Model’. Staff used 10 cm as the interval to
measure howmuch value of property became inundated with each 10 cm
increaseof sea level rise; at 10 cm, 20 cm, 30 cm, 40 cm, 50 cm and
60cm increments of sea level rise. Appendix D describes the
detailedmethodology. Scenarios greater than 60 cm are of course
possible,but would not lead to a greater qualitative insight, as
the highestIPCC estimate of SLR for year 2080, is 65 cm.
Although the potential economic developments in the next
fewdecades are difficult to predict, an approximation of basic
imple-mentation costs is possible. The following calculations are
roughapproximations, however, they provide a useful first estimate
and
48
-
guideline for future data collection efforts. Figures 19 through
30show the marginal or added cost of property values at risk of
each 10cm sea level rise increase of each county. In addition, a
comparisonwas made regarding property values at risk and the
previous sectiondamage assessment analysis on the cost of storm
surge by (relativeto each 10 cm of sea level rise). Both property
values at risk anddamage cost are measured in 2007 Dollars.
Figure 19 — Property Values at Risk with Sea Level Rise of
DadeCounty
Figure 20 — Storm Surge Damage Cost with Sea Level Rise ofDade
County
49
-
Figure 21 — Property Values at Risk with Sea Level Rise of
DixieCounty
Figure 22 — Storm Surge Damage Cost with Sea Level Rise ofDixie
County
50
-
Figure 23 — Property Values at Risk with Sea Level Rise of
DuvalCounty
Figure 24 — Storm Surge Damage Cost with Sea Level Rise ofDuval
County
51
-
Figure 25 — Property Values at Risk with Sea Level Rise
ofEscambia County
Figure 26 — Storm Surge Damage Cost with Sea Level Rise
ofEscambia County
52
-
Figure 27 — Property Values at Risk with Sea Level Rise ofMonroe
County
Figure 28 — Storm Surge Damage Cost with Sea Level Rise ofMonroe
County
53
-
Figure 29 — Property Values at Risk with Sea Level Rise
ofWakulla County
Figure 30 — Storm Surge Damage Cost with Sea Level Rise
ofWakulla County
54
-
Among the six counties, Monroe County has the highest
marginalproperty value at risk. For each 10 cm increase in sea
level, the prop-erty values at risk increase by $ 797.27 Million.
The lowest marginalvalue at risk is in Dixie County. An equivalent
rise in sea level wouldincrease the property values at risk by
$168.92 Thousand. Based onthe results from the GIS ’Storm Model,’
the property values at riskwith sea level rise of each county are
presented in Table 14.
Table 14 — Regression Results of Property Values at RiskCounty
Regression Equation R-SquareDade Property Values at Risk = -1.69 +
205.91 Sea Level Rise 0.98Dixie Property Values at Risk = 2.01 +
1.40 Sea Level Rise + (2.9
E-06) (Sea Level Rise2̂)0.94
Duval Property Values at Risk = -36.39 + 1060.67 Sea Level Rise
0.96Escambia Property Values at Risk = 201.22 - 974.36 Sea Level
Rise +
0.003 (Sea Level Rise2̂)0.98
Monroe Property Values at Risk = 340.50 + 7972.70 Sea Level Rise
0.99Wakulla Property Values at Risk = -0.04 + 3.40 Sea Level Rise
0.99
*Y: Sea Level Rise in Meters; M: Property Values at Risk in
million 2007 $.
The damage cost from a storm surge would also increase after
arise in sea level, as compared to an equivalent storm surge in
theabsence of a sea level rise. Comparing with the increase in
prop-erty values at risk, the previous damage cost analysis with
similarSLR, for Dade County, the percentage is1%. For Dixie County,
thepotential increase in damage costs associated with a storm
surgeassessment, mentioned in the previous section, ranges from
0.2% to0.5% of increase in property values at risk. For Duval
County, thepotential increase in previous storm surge damage costs
estimatesranges from 0.003% to 0.009% of total increase in property
values atrisk. For Escambia County, the percentage of potential
increase inprevious storm surge damage costs estimates comparing to
increasein property values at risk ranges from 0.002% to 0.005%.
For Mon-roe County, the potential increase in previous storm surge
damagecosts estimates ranges from 1%-2% of the total increase in
propertyvalues at risk. For Wakulla County, the potential increase
in previ-ous storm surge damage costs estimates of an equivalent
storm surgein the absence of a sea level rise is about 0.1% of
increase in prop-erty values at risk with same SLR. Table 15
details the percentageof comparing the previous storm surge damage
costs analysis withproperty values at risk associated with a 10 cm
sea level rise.
55
-
Table 15 — Comparison of Marginal Storm Surge Damage Cost
andMarginal Property Values at Risk Assessment
Sea LevelRise
Dade Dixie Duval Escambia Monroe Wakulla
0-10 cm 1% 0.2% 0.009% 0.003% 1% 0.1%11-20 cm 1% 0.2% 0.005%
0.004% 2% 0.1%21-30 cm 1% 0.4% 0.003% 0.005% 2% 0.1%31-40 cm 1%
0.5% 0.004% 0.005% 2% 0.1%41-50 cm 1% 0.5% 0.005% 0.003% 2%
0.1%51-60 cm 1% 0.5% 0.004% 0.002% 2% 0.1%
3.8 Summary and Conclusions
Sea Level Rise Estimation - The FSU Beaches and Shores Re-source
Center performed the sea level rise estimation based on
site-specific tidal gage records data (historical 60 years) for the
followingsix counties in Florida: Dade, Duval, Monroe, Escambia,
Dixie andWakulla; for years 2030 and 2080. An interesting result
from theanalysis is that although there was a wide distribution of
differentgage sites over the Florida Peninsula, the projected sea
level rise inyear 2080 does not vary substantially, the largest
value being 0.35meters in St. Petersburg, FL, while the smallest
value is 0.25 metersin Fernandina, FL.
This study is the first known work to explore sea level rise
fore-casting methods beyond the traditional polynomial linear
estimationforecasting methods utilizing gage data. The second order
linear ap-proach is recommended in the final analysis for
projecting economicscenarios of future costs due to sea level rise.
The second orderlinear approach includes a higher order term that
accounts for ac-celeration effects in sea level rise. The second
order approach is inaccordance with climate modeling scenarios that
project an expo-nential sea level rise due to greenhouse gas
effects. Although thepresent work is not definitive in regard to an
accelerating sea levelrise, it is clear at least from the data
available that trends are con-sistent and that there is no
deceleration in sea level rise over time. Apragmatic approach to
future economic planning should be in tunewith climatic model
scenarios that suggest the strong possibility ofan accelerating sea
level rise in Florida.
56
-
Economic Analysis of Sea Level Rise - Given a 65 cm(Year 2080,
IPCC estimate) sea level rise scenario, a similar stormevent will
appear again sooner rather than later. This translates toa
situation in a number of coastal counties, for storm events suchas
hurricanes, where people living along the coastal areas will
ex-perience property losses at twice the current rate. For
example,Hurricane Dennis resulted in a 9 feet high surge in Wakulla
Countyand according to FEMA’s study, was classified as a 30 year
eventhurricane. For a sea level rise of 2.13-feet scenario (IPCC
estimateto year 2080 of 65 cm), the same hurricane storm surge as
Denniswill be reduced from a 30 year to a 13.7 year event. There
will bean increasing return period frequency with an associated sea
levelrise.
Cost damages associated with storm events can also be expectedto
increase with respect to sea level rise. For example, for Dade
andMonroe Counties, a Wilma-like hurricane with corresponding SLRof
65 cm in Year 2080, can expect to generate a 40% increase and a60%
increase, respectively, in cost damages.
The storm model findings convey a considerable loss of
propertydue to varying sea level rise for years 2030 and 2080. For
example,Monroe County will incur significant damages in all four
scenar-ios, ranging from approximately $1 billion (Year 2030, FSU
Beachesand Shores SLR estimate) to $5.7 billion (Year 2080, IPCC
SLRestimate), in 2007 dollars. Duval County will lose
approximately31,000 land acres (or 6.2% of the total county area)
due to sea levelrise by Year 2080. In general, concerning the more
rural areas, suchas Wakulla and Dixie Counties, one would expect
lower propertyvalues compared to the more urbanized counties.
However, with sealevel rise, all the counties experience
substantial property losses overtime.
The GIS Storm Model Regressions analyzed the SLR impact
onproperty values. Each 10 cm of SLR will result in a
substantialnumber of properties at risk of inundation. In addition,
the damagecost estimates of storm surge for each 10 cm SLR were
compared tothe property values at risk with similar SLR. For
example, each 10cm of SLR will result in $ 797.27 million of
property values at risk
57
-
for Monroe County. Damage Cost associated with a same
scenariostorm surge can be expected to increase by $13.75
million.
According to a recent report by the Southeast Regional
Assess-ment Team on Preparing for a Changing Climate, changes in
dis-turbance patterns (e.g., hurricanes, floods, etc.) are possibly
moresignificant in terms of potential economic losses than longer
termchanges in precipitation and temperature. Facing gradual sea
wa-ter level rise, coastal residents should certainly employ
mitigationstrategies regarding protection of their coastal
properties. Thisstudy was able to establish a reasonable range of
low to high sea levelrise estimates to year 2080, with FSU
providing the lower bound forsea level rise based on historical
gage data, and the IPCC 2001 sealevel estimates providing the upper
bound, based on climate model-ing scenarios. The results of this
project underscore the importanceof including sea level rise as a
critical component in the hazard pre-paredness and mitigation
planning for coastal communities.
58
-
References
[1] Beth, M. C. and Titus,J.G. (Editors). 1984. Green-house
Effect and Sea Level Rise: A Challenge for thisGeneration Van
Nostrand Reinhold Company Inc.
[2] Bin, Okmyung. et al. 2006. Measuring the Impactsof Climate
Change on North Carolina Coastal Re-sources.
[3] Bosello, F., M. Lazarin, R. Roson, and R. S. Tol.
2004:Economy-Wide Estimates of The Implications of Cli-mate Change:
Sea Level Rise. JEL Classification
[4] Chen, J. L., C. R. Wilson, and B. D. Tapley. 2006.”Satellite
Gravity Measurements Confirm Acceleratedmelting of Greenland Ice
Sheet.” Science 313: 1958-1960.
[5] Church, J.A., Godfrey,J.S., Jacket,D.R. and MacDou-gal,T.J.,
1991. A Model of Sea Level Rise Caused byOcean Thermal Expansion,
J. Climate, 4, no. 4, 438-456.
[6] Cook, A. J., A. J. Fox, D. G. Vaughan, and J. G. Fer-rigno.
2005. ”Retreating Glacier Fronts n the Antarc-tic Penninsula over
the Past Half-Century.” Science308: 541-544.
[7] Dean, R.G., and Dalrymple, R.A., 2002. CoastalProcesses with
Engineering Applications, CambridgeUniversity Press, New York,
N.Y.
[8] Douglas, B. C., 1991. Global Sea Level Rise, J. Geo-phys.
Res., 96(C4), 6981-6992.
[9] Emery, K. O., and D. G. Aubrey, 1991. Sea Levels,Land
Levels, and Tide Gauges, Springer-Verlag, NewYork, N.Y.
[10] Federal Emergency Management Agency, Flood In-surance
Study. Dade, Duval, Dixie, Escambia, Mon-roe and Wakulla Counties
(various years, see footnotetext).
59
-
[11] Florida Office of Insurance Regulation, Florida Insur-ance
Study. Hurricane data summary. 2006.
[12] Gornitz, V. and A. Solow, 1991. Observations of LongTerm
Tide Gauge Records for Indicators of Acceler-ated Sea Level Rise,
in Greenhouse Gas-Induced Cli-matic Change: A Critical Appraisal of
Simulationsand Observations, edited by M. E. Schlesinger,
pp.347-367, Elsevier, Amsterdam.
[13] Hansen, J., L. Nazarenko, R. Ruedy, M. Sato,J.Willis, A. D.
Genio, D. Koch, A. Lacis, K. Lo, S.Menon, T. Novakov, J. Perlwitz,
G. Russell G. A.Schmidt, N. and Tausnev. 2005. Earth’s Energy
Im-balance: Confirmation and Implications. Science 303,1434.
[14] Hebert, P.J., J.D. Jarrell, and B.M. Mayfield. 1997.The
Deadliest, Costliest and Most Intense UnitedStates Hurricanes of
this Century (and Other Fre-quently Requested Hurricane Facts),
NOAA, Tech-nical Memorandum NWS-TPC-1, 30 pp.
[15] Hoffman,J.S., Keyes,D. and Titus,J.G., 1983. Project-ing
Future Sea Level Rise, U.S. Environmental Pro-tection Agency,
Washington, DC.
[16] Houghton, J.T., Y. Ding, D.J. Griggs, M. Noguer,P.J. Van
der Linden, D. Xiaoxu, and K. Maskell (ed.).2001. Climate Change
2001: The Scientific Basis. NewYork: Cambridge University
Press.
[17] Houghton, J. T., Jenkins,G. J. and Ephraums,J.
J.,(Editors), 1990. Climatic Change; the IPCC ScientificAssessment,
Cambridge U. Press, Cambridge, U.K.
[18] Intergovernmental Panel on Climate Change. 2001.Climate
Change: The IPCC Scientific Assessment,Cambridge University Press,
New York, N.Y.
[19] Knogge, T., M. Shirmer, and B. Schuchardt.
2004.Landscape-Scale Socio-Economics of Sea-Level Rise.Ibis, 146
(Suppl.1), pp. 11-17
60
-
[20] Luthcke, S. B., J. J. Zwally, W. Abdalati, D. d. Row-lands,
R. D. Ray, R. S. Nerem, F. G. Lemoine, J.J. McCarthy, and D. S.
Chinn. 2006. ”Recent Green-land Ice Mass Loss by Drainage System
from SatelliteGravity Observations.” Science 314: 1286-1289.
[21] National Academy of Sciences. 1983. Changing Cli-mate,
National Academy Press, Washington, DC.
[22] National Academy of Sciences. 1985. Glaciers, IceSheets,
and Sea Level,Mark Meier, Chairman, Na-tional Academy Press,
Washington, DC.
[23] National Research Council. 1987. Responding toChanges in
Sea Level,National Academy Press, R.G.Dean, Chairman, Washington,
DC.
[24] National Research Council (NRC) report on Sea LevelChange,
1990.National Academy Press, Washington,DC.
[25] Neumann, J. E., G. W. Yohe, R. Nicholls, M. andManion.
2000. Sea-Level Rise and Global ClimateChange: A Review of Impacts
to U.S. Coasts.
[26] Palmer, J.I., Jr. 2002. Supplemental State-ment: Governance
of Coastal and Ocean
Assets.http://www.epa.gov/gmpo/governance.html
[27] Parmesan, C., and G. W. Yohe. 2003