xvEPA
United States
Environmental Protection
Agency
Policy, Planning,
And Evaluation
(PM-221)
EPA-230-05-89-052
June 1989
The Potential Effects
Of Global Climate Change
On The United States
Appendix: B
Sea Level Rise
Printed on Recycled Paper
-------
THE POTENTIAL EFFECTS OF GLOBAL CLIMATE CHANGE
ON THE UNITED STATES:
APPENDIX B - SEA LEVEL RISE
Editors: Joel B. Smith and Dennis A. Tirpak
OFFICE OF POLICY, PLANNING AND EVALUATION
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, DC 20460
MAY 1989
-------
TABLE OF CONTENTS
APPENDIX B: SEA LEVEL RISE
PREFACE [[[ iii
THE EFFECTS OF SEA LEVEL RISE ON US. COASTAL WETLANDS ...................... 1-1
Richard A. Park, Manjit S. Trehan, Paul W. Mausel, and Robert C. Howe
NATIONAL ASSESSMENT OF BEACH NOURISHMENT REQUIREMENTS
ASSOCIATED WITH ACCELERATED SEA LEVEL RISE ........................... 2-1
Stephen P. Leatherman
DEFENDING DEVELOPED SHORELINES ALONG SHELTERED
WATERS OF THE UNITED STATES FROM A TWO METER RISE
IN MEAN SEA LEVEL [[[ 3-1
J. Richard Weggel, Scott Brown, Juan Carlos Escajadillo,
Patrick Breen, and Edward L. Doheny
THE COST OF NOT HOLDING BACK THE SEA - PHASE 1
ECONOMIC VULNERABILITY ............................................... 4-1
Gary W. Yohe
-------
PREFACE
The ecological and economic implications of (he greenhouse effect have been the subject of discussion within
the scientific community for the past three decades. In recent years, members of Congress have held hearings
on the greenhouse effect and have begun to examine its implications for public policy. This interest was
accentuated during a series of hearings held in June 1986 by the Subcommittee on Pollution of the Senate
Environment and Public Works Committee. Following the hearings, committee members sent a formal request
to the EPA Administrator, asking the Agency to undertake two studies on climate change due to the greenhouse
effect.
One of the studies we are requesting should examine the potential health and environmental
effects of climate change. This study should include, but not be limited to, the potential impacts
on agriculture, forests, wetlands, human health, rivers, lakes, and estuaries, as well as other
ecosystems and societal impacts. This study should be designed to include original analyses, to
identify and fill in where important research gaps exist, and to solicit the opinions of
knowledgeable people throughout the country through a process of public hearings and
meetings.
To meet this request, EPA produced the report entitled The Potential Effects of Global Climate Change on the
United States. For that report, EPA commissioned fifty-five studies by academic and government scientists on
the potential effects of global climate change. Each study was reviewed by at least two peer reviewers. The
Effects Report summarizes the results of all of those studies. The complete results of each study are contained
in Appendices A through J.
Appendix Subject
A Water Resources
B Sea Level Rise
C Agriculture
D Forests
E Aquatic Resources
F Air Quality
G Health
H Infrastructure
I Variability
J Policy
GOAL
The goal of the Effects Report was to try to give a sense of the possible direction of changes from a global
warming as well as a sense of the magnitude. Specifically, we examined the following issues:
o sensitivities of systems to changes in climate (since we cannot predict regional climate change, we
can only identify sensitivities to changes in climate factors)
o the range of effects under different warming scenarios
o regional differences among effects
o interactions among effects on a regional level
iii
-------
o national effects
o uncertainties
o policy implications
o research needs
The four regions chosen for the studies were California, the Great Lakes, the Southeast, and the Great Plains.
Many studies focused on impacts in a single region, while others examined potential impacts on a national scale.
SCENARIOS USED FOR THE EFFECTS REPORT STUDIES
The Effects Report studies used several scenarios to examine the sensitivities of various systems to changes in
climate. The scenarios used are plausible sets of circumstances although none of them should be considered to
be predictions of regional climate change. The most common scenario used was the doubled CO2 scenario
(2XCO2), which examined the effects of climate under a doubling of atmospheric carbon dioxide concentrations.
This doubling is estimated to raise average global temperatures by 1.5 to 4.5° C by the latter half of the 21st
century. Transient scenarios, which estimate how climate may change over time in response to a steady increase
in greenhouse gases, were also used. In addition, analog scenarios of past warm periods, such as the 1930s, were
used.
The scenarios combined average monthly climate change estimates for regional grid boxes from General
Circulation Models (GCMs) with 1951-80 climate observations from sites in the respective grid boxes. GCMs
are dynamic models that simulate the physical processes of the atmosphere and oceans to estimate global climate
under different conditions, such as increasing concentrations of greenhouse gases (e.g., 2XCO2).
The scenarios and GCMs used in the studies have certain limitations. The scenarios used for the studies assume
that temporal and spatial variability do not change from current conditions. The first of two major limitations
related to the GCMs is their low spatial resolution. GCMs use rather large grid boxes where climate is averaged
for the whole grid box, while in fact climate may be quite variable within a grid box. The second limitation is
the simplified way that GCMs treat physical factors such as clouds, oceans, albedo, and land surface hydrology.
Because of these limitations, GCMs often disagree with each other on estimates of regional climate change (as
well as the magnitude of global changes) and should not be considered to be predictions.
To obtain a range of scenarios, EPA asked the researchers to use output from the following GCMs:
o Goddard Institute for Space Studies (GISS)
o Geophysical Fluid Dynamics Laboratory (GFDL)
o Oregon State University (OSU)
Figure 1 shows the temperature change from current climate to a climate with a doubling of CO2 levels, as
modeled by the three GCMs. The figure includes the GCM estimates for the four regions. Precipitation
changes are shown in Figure 2. Note the disagreement in the GCM estimates concerning the direction of
change of regional and seasonal precipitation and the agreement concerning increasing temperatures.
Two transient scenarios from the GISS model were also used, and the average decadal temperature changes
are shown in Figure 3.
IV
-------
FIGURE 1. TEMPERATURE SCENARIOS
GCM Estimated Change in Temperature from 1xCO2 to 2xCO2
8
o
o
UJ
DC
5
DC
UJ
Q.
UJ
4 '
2
ANNUAL
Great Southeast Great California United
Lakes Plains States'
8
Great Southeast Great California United
Lakes Plains States'
0
SUMMER
n
Great Southeast Great California United
Lakes Plains States'
GISS
GFDL
OSU
* Lower 48 States
-------
FIGURE 2. PRECIPITATION SCENARIOS
GCM Estimated Change in Precipitation from 1xCO2 to 2xCO2
l.U
>- 0.8
Q 0.6
et\
0- °4
UJ
tj °2
5 0.0
=J -0.2
S -0.4
-0.6
ANNUAL
rrn
?%un HI r-i !«•-• ^i Hid 1
BM 1^1 '~~'
1
l.U
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-06 J
WINTER
__.,
m
S
i
P
_,.
i A
^
\
0 ^
1 1
1^-. . — -
1-, M 1
I.U'
0.8-
0.6
0.4-
0.2-
o.o-
-0.2-
-0.4-
-n fi-
No
Change
1
I
i
s
si
®
'//
^
V
I
SUMMER
^
1
§§
™
n
i
&
1
ni
1 r-|
i
Great Southeast Great California United
Lakes Plains States'
Great Southeast Great California United
Lakes Plains States'
Great Southeast Great California United
Lakes Plains States'
* Lower 48 States
-------
4
3.5
£ 3
jg 2.5
H
2
ec
ui
Q.
UJ
1.5
1
0.5
0
O
oc
a
3.5
3
2.5
<
ec
UJ . _
o. 1.5
in
1
0.5
0
3.72
2.99
2.47
1.72
1.36
0.70
0.88
0.30
1980s 1990s 2000s 2010s 2020s 2030s 2040s 2050s
TRANSIENT SCENARIO A
1.26
1.02
0.59
018
0.35
\/ / / /\
7ZZ/
1980s 1990s 2000s 2010s
TRANSIENT SCENARIO B
2020s
FIGURE 3.
GISS TRANSIENTS "A" AND "B" AVERAGE
TEMPERATURE CHANGE FOR LOWER 48 STATES
GRID POINTS.
vu
-------
EPA specified that researchers were to use three doubled CO, scenarios, two transient scenarios, and an analog
scenario in their studies. Many researchers, however, did not nave sufficient time or resources to use all of the
scenarios. EPA asked the researchers to run the scenarios in the following order, going as far through the list
as time and resources allowed:
1. GISS doubled CO2
2. GFDL doubled CO2
3. GISS transient A
4. OSU doubled CO2
5. Analog (1930 to 1939)
6. GISS transient B
ABOUT THESE APPENDICES
The studies contained in these appendices appear in the form that the researchers submitted them to EPA.
These reports do not necessarily reflect the official position of the U.S. Environmental Protection Agency.
Mention of trade names does not constitute an endorsement.
vui
-------
THE EFFECTS OF SEA LEVEL RISE ON VS. COASTAL WETLANDS
by
Richard A. Park
Manjit S. Trehan
Holcomb Research Institute
Butler University
Indianapolis, IN 46208
and
Paul W. Mausel
Robert C. Howe
Remote Sensing Laboratory and
Department of Geography and Geology
Indiana State University
Terre Haute, IN 47809
Cooperative Agreement CR814578-01
-------
CONTENTS
ACKNOWLEDGMENTS iii
FINDINGS 1-1
CHAPTER 1: INTRODUCTION 1-2
CONTEXT 1-2
BACKGROUND 1-2
CHAPTER 2: METHODS 1-3
DATA 1-3
Cover Classes 1-3
Development 1-4
Elevations 1-4
QUALITY ASSURANCE 1-4
DESCRIPTION OF MODEL 1-5
Inundation Model 1-5
Spatial Model 1-8
CHAPTER 3: SIMULATIONS AND SYNTHESIS OF RESULTS 1-11
CASE STUDY: LONG BEACH, NEW JERSEY 1-11
NATIONAL IMPACTS 1-14
REGIONAL TRENDS 1-21
Mid-Atlantic 1-21
Southeast 1-21
West Coast 1-39
Northeast 1-39
CHAPTER 4: SUMMARY 1-51
REFERENCES 1-53
11
-------
ACKNOWLEDGMENTS
This project was made possible by many individuals who often worked overtime to obtain the results in
a timely manner. At Indiana State University, Drs. Richard Hyde and Kamlesh Lulla were involved in the
development of methods used in the analysis of remotely sensed data, with programming support from Nicholas
Vesper and Norman Levine; analysis of Landsat data and digitization of elevational data were performed by
Randy Pearson, Robert Regis, Timothy Gress, Mao Chang, Huen Chen, Mark Jakubauskas, David Kettler,
Steven Kopp, Asilah Mahmud, Chenyge Mao, Mark Karaska, Kai Huang, and Jae Lee. At Butler University,
Michael Magier assisted in the development of software; assistance in editing data, performing simulations, and
preparing results was given by Julie James, Michael Yamanichi, Lisa Guendling, and Joseph Poston; manuscript
editing and review was provided by James Rogers, Thomas Armentano, Paul van der Heijde, and Orie Loucks.
The manuscript benefited from the suggestions of Joy Zedler, Thomas Cavinder, the US. Fish and Wildlife
National Coastal Ecosystem Team, and an anonymous reviewer. The project was funded through Cooperative
Agreement CR814578-01 with the U.S. Environmental Protection Agency; James Titus was the project monitor.
ui
-------
Park
FINDINGS1
During the next century, rising sea level due to global warming will have a profound impact on the
coastal wetlands of the United States and a significant impact on coastal lowlands. With an almost-certain rise
of a half-meter by the year 2100 and with all currently developed areas protected from inundation and erosion,
more than 4,000 mi2 of vegetated wetlands will be lost. With a probable rise of 1 meter by the year 2100,6,441
mi2 or approximately 65% of the coastal marshes and swamps of the contiguous United States could be lost.
With a 2-meter rise, 7,423 mi2 or 77% of the coastal wetlands of the contiguous United States could be lost, and
remaining southeastern marshes could be converted to mangrove swamps. Furthermore, unprotected barrier
islands would be lost through accelerated beach erosion; much of the Florida Everglades and Keys would be
inundated; and low-lying coastal cities such as Charleston, South Carolina, and Long Beach, Mississippi, could
be submerged if not ringed by dikes. In a worst-case scenario, with a 3-meter rise and all dry land protected
from inundation, 10,953 mi2 of marshes and swamps could be lost.
1 Although the information in this report has been funded wholly or partly by the U.S. Environmental
Protection Agency under Cooperative Agreement CR814578-01, it does not necessrily reflect the Agency's views,
and no official endorsement should be inferred from it.
1-1
-------
Park
CHAPTER 1
INTRODUCTION
CONTEXT
The "greenhouse gases" -- carbon dioxide, methane, nitrous oxides, and chlorofluorocarbons that are
transparent to sunlight but absorb reradiated heat energy - are increasing at an alarming rate due to human
activities. CO2 and other greenhouse gases may double by the year 2030 as compared to the amounts present
at the start of the industrial revolution (Villach, 1985), warming the earth's surface between 2 and 4°C (Titus
and Seidel, 1986). If the average temperature increases by 3°C by 2050 and remains constant thereafter, sea level
will probably rise approximately 1 meter (m) by 2100; a global warming of 6°C by 2100 could result in a sea level
rise of 2.3 m (Thomas, 1986).
Such an accelerated rise in sea level would have a serious impact on the distribution of U.S. coastal
wetlands. Salt, brackish, and fresh marshes as well as mangrove and other swamps would be lost due to
inundation and erosion, or would migrate inland as adjacent lowlands not protected by engineering structures
are inundated. The value of these wetlands as habitat for wildlife would be unpaired, and their biodiversity would
decrease. Although many wetlands have kept pace or have increased in area with historic sea level rise due to
sediment entrapment and peat formation (Davis, 1985), vertical accretion of wetlands has not been observed to
occur at rates comparable to those projected for sea level rise in the next century. In fact, the rate of a 1-m rise
by the year 2100 will be equal to that attained during the disintegration of the continental ice sheets at the end
of the Pleistocene (Peltier, 1988), which drowned barrier islands and associated features on the continental
shelves.
Wetlands are vital to the ecology and economy of UJS. coastal areas. Their biological productivity is
equal to that of any other natural or agricultural system (cf. Teal, 1962; Ryszkowski, 1984). Although little of
that productivity may be available to marsh animals and coastal fisheries (Montague et al., 1987), over half the
species of commercially important fishes in the southeastern United States use salt marshes as nursery grounds
(Thurman, 1983). Wetlands also remove pollutants (Pope and Gosselink, 1973) and provide protection from
floods, storms, and high tides (Lugo and Brinson, 1978). Based on these functions, it has been estimated that
marshes provide an annual return to society equivalent to $5,500/acre (Thurman, 1983). For these and other
reasons, the Congress and the U.S. scientific community are seeking the best quantitative estimate of potential
impacts on coastal wetlands under various scenarios of sea level rise and coastal zone management policy.
BACKGROUND
Previous studies by EPA indicated the scope of the problems associated with sea level rise (Earth and
Titus, 1984; Titus 1986, 1988) and showed that a more detailed study was warranted. As a part of one of these
previous studies, Holcomb Research Institute developed a simulation model (SLAMM) that was used to conduct
preliminary regional analyses of the effects of sea level rise on U.S. coastal wetlands. The original model was
based on manually coded data on elevation and cover classes from topographic maps, using a 1-km grid (Park
et al., 1986a,b; Armentano et al., 1988). Simulations suggested that large areas of coastal wetlands would be lost
with sea level rise. However, both data and the model needed refining if the results were to be used for
evaluating policy. The objective of the present study, therefore, is to present and document the refined databases
and calculation procedures, and their limitations, together with improved estimates as to how much of the
nation's coastal wetlands are likely to be lost under various scenarios of sea level rise.
1-2
-------
Park
CHAPTER!
METHODS
DATA
Ninety-three sites were chosen, using an unbiased systematic sampling of U.S. Geological Survey
topographic maps at a scale of 1:24,000. Starting with the easternmost quadrangle in Maine and restricting the
choice to those maps that included some part of the coast, every 15th quadrangle was picked as the center of a
site consisting of one to four quadrangles. (Two sites are represented by 1:64,000 maps and two by 1:25,000
maps.) Of these, a subsample of 46 sites was used in this initial study. Four supplemental sites were also chosen
for verification purposes. For most coastal sites with low slopes and extensive lowland and wetland areas, a cell
size of 500 m by 500 m was used; this is an area of 25 hectares (ha) or 61.75 acres; with this coarse-grained
resolution, a site typically contained four quadrangles. For sites with steep slopes or heterogeneous urban
development, a fine-grained resolution of 250 m by 250 m (6.25 ha or 15.44 acres) was used; normally such a site
was restricted to one quadrangle. A normal coarse-grained site contains as many as 3,400 cells; some sites with
fine-grained resolution and covering four quadrangles contain almost 14,000 cells.
Each cell is represented by information on elevation, percent cover in various classes, development, and
presence of protective engineering structures. Eleven cover classes were distinguished, including upland (above
12 ft or 3.66 m elevation); lowland (above mean high water spring tide level [MHWS] and below 3.66 m
elevation); sandy area (usually backshore and dunes areas, but including both lowland and upland areas
characterized by exposed sand); freshwater marshes, saltmarshes, freshwater swamps, mangrove swamps,
combined beaches and tidal flats, rocky intertidal areas, and water.
Cover Classes
The cover-class data were obtained by analysis of multispectral Landsat data from essentially cloud-free,
geocorrected scenes, augmented by visual interpretation of high-altitude color-infrared photographs (with a scale
of 1:58,000). Numerous studies have shown that these forms of remote sensing can effectively differentiate
wetland and other coastal cover types (Estes and Thorley, 1983).
Well-established methods for quantifying remotely sensed cover-class data were used (Mausel, 1985).
The HINDU algorithm (Dasarthy, 1974) was used to partition the spectral data into classes of repetitive
signatures for a particular site. Usually 20 to 30 clusters were obtained. These were combined by an experienced
analyst to represent the cover classes on the basis of cluster statistics and spectral signatures, confirmed by
information from the maps and photographs. The correspondence between pixels in combined clusters and a
designated cover class, such as low marsh, is not perfect. Preliminary analyses of selected sites suggest that
accuracies for well-defined classes range from 75 to 95+ percent. Features represented by mixed pixels or by
ambiguous spectral responses pose a problem. Residential and commercial developments were not identified
from the spectral data because pavement and roofs could be confused with the "sandy area" class. However,
because of the need to identify all wetlands, the swamp class was used despite an ambiguous signature. Where
possible, high and low marshes were distinguished, based on "dry" and "wet" spectral signatures.
Each Landsat pixel represents an area 57 m by 79 m. The data were resampled and aggregated to form
pixels with an area of 71 m by 63 m, and were printed at a scale of 1:24,000 to facilitate comparison with
topographic maps. Thus with a 500-m by 500-m grid, the unit cell contains 55+ pixels. The percent cover for
each class was stored and reported in increments of 5%.
In the absence of "ground truth" based on detailed field checking, the cover determinations for any
particular site should be considered best-estimates consistent with the regional goals of the study. Topographic
maps and USFWS National Wetland Inventory (NWI) maps were used to confirm the interpretations. Because
the Landsat imagery is often from 1986 and 1987 and supersedes map coverage by a significant number of years,
1-3
-------
Park
the spectral signatures were often accepted in the event of conflict. For example, at a New Jersey site both the
topographic map, photorevised in 1972, and the NWI map, based on 1977 aerial photography, showed an
extensive unbroken marsh; however, the Landsat imagery and the high-altitude photograph provided unmistakable
evidence that a large area had been converted to lowland (probably with dredge fill). Marshes, shadowed
uplands, and uplands with redwoods sometimes exhibited similar spectral signatures; these were reassigned
manually on the basis of topographic maps. Swamps proved to be difficult to distinguish from upland forest; a
dense tree canopy can effectively mask any standing water. For several sites swamps were reassigned manually,
based on topographic maps. Because sandy areas often included beaches and tidal flats, these wetlands are
underestimated in the data, and a transient response is often observed in the simulations as the model
reapportions sandy area to beach.
Development
Residential and commercial developments were identified from topographic maps. If a cell contained
sufficient buildings and other structures worthy of protection (including airports and wharves), the cell was
characterized as developed. Because the maps are older than the Landsat data and the simulations start with
the Landsat coverage, this underestimates the extent of development and subsequent protection. Although it
would be of interest to simulate continued coastal development, that is beyond the scope of this study.
Elevations
Topographic elevations on U.S. Geological Survey maps are relative to the 1929 sea level, known as the
National Geodetic Vertical Datum (NGVD); therefore, elevations of the dry-land classes (upland, lowland, and
sandy area) were corrected for historic sea level rise from 1929 to the date of the topographic map. The
long-term value of 1.2 mm/yr was used for historic sea level rise.
Dry-land elevations were further corrected for changes from the map date to the date of Landsat
coverage, at which time the areas of the classes were observed. This correction is based on both the 1.2 mm/yr
trend and observed subsidence due to regional isostatic adjustments and tectonics, and local compaction due to
withdrawal of water, oil, and gas .
The elevational data were obtained by digitizing the corner elevations of each cell, based on
interpolations of elevations from the topographic maps. Both the elevational and the Landsat data were recorded
using the Universe Transverse Mercator grid and were combined for each cell in the site data file. Elevations
and elevational ranges for classes within a cell were apportioned based on an algorithm that assumes sandy and
rocky areas to have a convex profile, and other lowland and upland areas to have a concave profile. An
elevational mosaic was assumed for saltwater wetlands, with each cover class traversing its full elevational range
within a cell (representing the usual microtopography that occurs with small tidal flats and beaches, tidal creeks,
natural levees, and back-levee areas). Saltwater wetland elevational ranges were computed by assuming constant
relationships of the wetlands to tidal daturas (cf. Lefor et al., 1987), with mean or half tide level (MTL) as 0;
beach and tidal flats extending from mean low water (MLW) to mean high water (MHW), or from MLW to
MTL on coasts with low wave energy and vegetated wetlands; low marsh extending from MTL to MHW; high
marsh extending from MHW to MHWS; and mangrove swamps extending from MTL to MHWS. (Occasionally
saltwater wetlands occur above MHWS, but the area is small and can be ignored.)
QUALITY ASSURANCE
Quality assurance, mandated by the EPA Administrative Procedures Act, received close attention in this
study. All interpretations and implementations received an independent evaluation by another member of the
study team, and records have been kept for all procedures.
Selected sites were visited by the principal investigator and members of the remote sensing team. This
helped provide ground truth for representative sites and promoted inclusion of subtle relationships in the model.
1-4
-------
Park
Discussions of the modeling study were held at several coastal wetland labs, and suggestions for improving the
study were solicited.
DESCRIPTION OF MODEL
Stated simply, the objective of the modeling undertaken in this study has been to consider the dominant
processes involved in vegetative wetland conversions and related shoreline reconfigurations during long-term
sea-level rise. The model, SLAMM2, differs from other wetland models (Day et al., 1973; Wiegert et al., 1975;
Hopkinson and Day, 1977; Browder et al., 1985; Sklar et al., 1985; Costanza et al., 1987; Kana et al., 1988) by
its ability to predict map distributions of wetland cover under conditions of accelerated sea level rise and by its
applicability to the diverse wetlands of the contiguous coastal United States.
Eleven cover-classes were modeled in each of over 3,000 cells for 115 or more years; therefore, the level
of scientific detail represented had to be balanced with efficient computational algorithms. The following
sections present the basic constructs for each of the processes considered in order to establish the basis for the
model results; the constructs are organized in two sections, the basic inundation model and the map-based spatial
model. Simplifying assumptions that may bias the results, and could be investigated more fully in the future, are
emphasized by italics.
Inundation Model
The colonization of newly inundated dry land by wetland vegetation and loss of wetlands due to further
inundation is based on a straightforward geometric relationship, with lag effects for some conversions. Seven
processes are considered as part of the inundation model.
Sea Level Change. Relative sea level (SL) after 1986 was computed for a given year T in the simulation,
using a quadratic equation:
SL = Linear * (T -1986) + Quadratic * (T -1986)2 + Subsidence
where Linear is the historic eustatic trend of 1.2 mm/yr and Quadratic is the second-order parameter with
values depending on the scenario chosen (Figure 1). Subsidence is a site constant for the rate of local
subsidence, and is based on the rate observed at the site or at the nearest place of record. Usually subsidence
reflects regional isostatic and tectonic responses, is not large, and does not vary greatly over the region; however,
subsidence may be large, due to groundwater and petroleum withdrawal and compaction of unconsolidated
sediments, and may be both local and time-varying. Although high rates were not extrapolated to other sites,
they were assumed to apply to the entire site in which they were observed and to be constant over time. Insofar
as pumping may be curtailed, use of high subsidence values may overestimate inundation effects for a few sites.
Conversions between Classes. SLAMM2 is a discrete, algebraic model with a time step of 5 to 25 years
(depending on the rate of sea level rise) that utilizes alternative pathways of change, depending on site and ceU
conditions. Conditions include exposure to open ocean, residential and commercial development, existence of
protective engineering structures, unconsolidated or consolidated substrate (based on geologic maps), prevailing
wave regime, and subtropical climate. Each class occurring in a cell is converted to another class for given
conditions. Generally, the conversion is fractional and is represented by the following equation:
Vulnerability * DeltaT
where LOSSQ^ is the fraction converted in the time step; SLRise is the change in sea level during the time step,
corrected for sedimentation and vertical accretion of wetland where appropriate; Rangeaass is the elevational
range of the class in a given cell; Vulnerability is the susceptibility to change due to factors such as slow death
and colonization (important only for some processes such as inundation of mangrove swamps and only for a
time-step of five years); and DeltaT is the time step. This construct assumes that conversion of area from one
1-5
-------
Park
Ui
u
UI
0)
3
15
2
I.S
1
0.5
0
7
1884 zoco MIS
YCAQ
Figure 1. Sea-level rise scenarios used in the study; the step wise patterns indicate the longer time steps used
•"' for some scenarios.
1-6
-------
Park
class to another is a linear function of the elevational range that is lost due to sea level rise within the cell.
Departures from a linear slope are assumed to be accounted for in the biasing of elevational ranges toward
higher values for convex slopes and lower values for concave. These are simplifying assumptions; therefore,
conversions may occur at a faster or slower rate than simulated, and the simulations should be considered as
approximate responses to be expected for different scenarios and should be used primarily for computing regional
averages.
Tropical Conditions. The conversion of dry land to wetland vegetation is complicated by global climate change.
Mangroves are limited to those areas without appreciable frost. At present, viable populations of small black
mangroves occur as far north as Daytona, Florida, and in the Mississippi River delta and along the barrier
islands of the northern Gulf of Mexico (Sherrod and McMillan, 1985). With only a slight decrease in killing
frosts, these populations could spread to other coastal areas where adjacent water moderates the climate. The
SLAMM2 simulations assume that for all scenarios other than a continuation of the current sea-level trend,
mangrove swamps can become established in the northern Gulf of Mexico and Atlantic coast of Florida beginning
in the year 2000. They are assumed to grow in Georgia and South Carolina at slightly later dates. This is an
approximation and could be sharpened by using projected climatic changes for each area of the coast.
Coastal Engineering Structures. The presence of dikes and levees completely enclosing coastal areas was noted
from the topographic maps. The two protection scenarios represent further construction of dikes around all
developed areas and all dry land, respectively. Areas so protected are not allowed to convert to other classes
in the simulations. Enclosed wetlands are assumed to be maintained in their initial condition.
Death and Colonization. Death and conversion of low marsh to water by inundation and corresponding
conversion of high marsh to low marsh are assumed to occur at a linear rate of 20% per year of the potential
based on inundation alone (Vulnerability = 0.2); with a normal time step of five years or more, this produces
no discernible lag. However, in the model, mangrove swamps are converted to water at the rate of 10% per year
of the potential, producing a 50% lag for a time step of five years. Because they are intolerant to saltwater,
freshwater swamps and marshes are converted instantaneously to water or high marsh, respectively, or to
mangrove swamp if subtropical.
Inundated lowlands and sandy areas are instantaneously converted to either high marsh or mangrove
swamp by the model. This assumes colonization to proceed faster than the normal five-year time step, and it
assumes that paved surfaces are a negligible impediment to colonization. Although relatively rapid colonization
rates have been reported for some areas, these assumptions may lead to an overestimate of vegetated wetlands.
Sedimentation and Accretion. One of the pervasive challenges in developing the model was to incorporate
processes with variable rates known from studies in different areas but not known for the specific study sites.
Perhaps the most important process is vertical accretion of wetlands. In the past in many areas, accretion has
kept pace with local changes in sea level. Although accretion rates vary widely at particular locales and among
areas, including most of the study sites, a pattern emerges that can be used in estimating local rates. In deltaic
areas characterized by extensive marshes, 10 mm/yr seems to be representative, although much higher maximal
values have been observed; in many areas with moderately extensive wetlands, 5 mm/yr seems to be a common
midrange value; in areas with little wetland development, 2 mm/yr seems to be a representative minimal value
(cf. Letzsch and Frey, 1980; Gosselink, 1984; Armentano et al., 1988). Therefore, as a simplifying assumption,
if the percentage of salt wetlands at the start of the simulation is greater than or equal to 30%, an accretion rate
of 10 mm/yr is used for low marsh at the site; if greater than 5% and less than 30%, an accretion rate of 5
mm/yr is used; if less than or equal to 5%, an accretion rate of 2 mm/yr is used. Because marsh areas that are
back from water courses tend to have accretion rates that are approximately half those of streamside marshes
(cf. Gosselink, 1984), the values are halved for high marsh. Areas enclosed by dikes are not affected by these
accretion rates because the marshes are not permitted to change.
Rates of accretion in mangrove swamps seem to depend more on extremely localized conditions and
have not been well studied. In a detailed study of a mangrove swamp in Australia, Bird (1986) reported rates
1-7
-------
Park
varying from 2 to 13 mm/yr. We have assumed 5 mm/yr to be representative and have used that value for all
areas with mangroves.
The assumption of an accretion rate for a particular wetland type that is uniform over a site and constant
through time probably leads to an overestimate of wetland maintenance.
Sedimentation rates in adjacent areas of sheltered water have not been well documented, nor are they
as important as accretion; therefore, the local sedimentation rate for areas of sheltered water is taken to be
one-tenth that of accretion. Admittedly this is an arbitrary value, but the model is not sensitive to it.
Few U.S. beaches are currently prograding or expanding due to sedimentation. The model assumes that
progradation of exposed beaches and tidal flats occurs only under conditions of high sedimentation and a
continuation of historic sea level rise. If sedimentation exceeds sea level rise under any scenario, areas of
sheltered water are converted to tidal flats which in turn are converted to wetlands, using the following
relationships:
LossWMar = (DdteT * Sedimentation - SLRise)/Depth
LossT-|da|flat = (DeltaT * Sedimentation - SLRise)/RangeTldaWat
where LossWatef is the fraction of area of water converted to tidal flat during the time step DeltaT;
Sedimentation is the rate of sedimentation (ra/yr); SLRise is the change in sea level during the time step; Depth
is the average depth of sheltered water; and Range^..^, 's tne elevational range of tidal flats at the site. The
construct treats the fractional losses as uniform for afTcells with sheltered water and tidal flats, and it ignores
subsidence; therefore, it probably overestimates progradation at sites with high sedimentation rates (as indicated
by extensive initial wetlands) for the scenario of continued historic sea level rise.
Salinity. Salinity is not explicitly modeled. This has two consequences in the simulations: low-lying freshwater
marshes may be designated as saltmarshes, and salt pannes and tidal flats in arid regions may be predicted to
convert to saltmarshes, on the basis of elevation.
Spatial Model
In addition to the effects of inundation represented by the simple geometric model described above,
second-order effects occur due to changes in the spatial relationships among the coastal elements. Accordingly,
SLAMM2 incorporates a map-based model component to consider four spatially important processes: coastal
beach erosion, overwash, barrier breaching, and headland erosion..
Erosion. Under equilibrium conditions, erosion and deposition balance and wetlands are not lost. However,
even historic sea level rise coupled with local subsidence has upset coastal equilibrium in many parts of the world
(Bird, 1986; Bruun, 1986). Although the processes of erosion can be expressed by detailed quantitative
relationships, such an approach is beyond the scope of the present study. Rather, qualitative relationships are
defined and used as thresholds for including constant rates of wave erosion in simulating the localized loss of
wetlands. The effects of severe storms are included in these average values. The model can represent several
levels of erosion based on the observed average fetch (the distance across which wind-driven waves can be
formed) of sheltered water. These levels are represented by scalars: "none," "little," "moderate," "heavy," and
"severe"; however, in the present implementation, constant erosion is triggered only when the erosional scalar
is greater than "moderate," which occurs when the average fetch exceeds 9 km. This occurs at only a few sites.
The model also recognizes exposure to open ocean as triggering erosion of wetlands. Cells can become
exposed as protective barrier islands and spits are breached. If a saltmarsh is exposed to the open ocean or
erosion is greater than "moderate," 0.4% of the marsh is lost per year (2 m/yr using a 500-m grid). If a swamp
1-8
-------
Park
is exposed to the open ocean or erosion is greater than "moderate," 0.2% of the swamp is lost per year (1 m/yr
using a 500-m grid). These values are based on the literature (Hall et al., 1986) and on rates observed from
photorevised and sequential maps of areas with eroding wetlands.
Bruun (1962, 1986) has shown that on the average, recession of beaches and backshore areas is one
hundred times the change in sea level. Assuming a beach slope of 12°, an average for many beaches (cf. Dyer
1986), the width is five times the height (tan 12° = 0.2). Therefore,
Recession = 100 * SLRise
WidthBeach = 5 * RangeBeach
LossBeach = Recess«°n/WidthBeach
where Recession is the width of beach lost during a time step (m); SLRise is the change in sea level (m) during
the time step; WidthBeach is the theoretical width of the beach (m); Ran8eBeach IS tne elevational range of the
beach (MHWS-MLW in m); and LossBflach is the fraction of beach lost during a time step.
Assuming that beach width is maintained at the expense of any adjacent backshore and dune area, and
assuming that the beach runs the length of the cell (an assumption that probably results in an overestimation of
beach area), then
AreaBeach = WidthBeach * Scale/10,000
RecessionSandyArea = AreaBeach - AreaBeach
while there is a deficit of beach and there is sandy area to be eroded; if the "protect backshore" scenario is
chosen, recession is not allowed. AreaBeach is the theoretical area of beach (ha), given equilibrium conditions;
Scale is the grid size, usually either 500 or 250 m; 10,000 is a proportionality constant to convert meters to
hectares; Recess'onsandyArea IS the SM1(ty area converted during the time step (ha); and AreaBeach is the area
of the beach before conversion of adjacent sandy area (ha).
Beaches that are developed are assumed to be protected; therefore, developed backshores are not
allowed to erode in the simulations unless a "no protection" scenario is chosen. However, no provision is made
for active formation of natural and artificial dunes on undeveloped beaches, nor is natural nourishment of
beaches due to eroding headlands simulated. Therefore, the model probably overestimates beach erosion on
exposed, undeveloped coasts.
Overwash. As erosion of backshore and dune areas occurs and as other lowlands are drowned, wetlands on the
lee side of coastal barriers are subject to conversion due to overwash, the process by which sediments are carried
over the crest of the barrier and deposited onto adjacent wetlands. This process is simulated only for areas
having a beach and only during the time step in which the lowland is breached. It assumes that 50% of the
adjacent high marsh, 25% of the low marsh, and 5% of mangrove (if present) in the adjacent cell is converted
to beach and tidal fiat; the percentages are educated guesses based on observations of existing overwash areas
(cf. Leatherman and Zaremba, 1986). Adjacent water is converted to beach and tidal flat by an amount based
on the assumption that upper beach sediments (occurring between mean tide level and mean high water spring
tide level) are transported into the water:
(MHWS ' MTL)/Depth
where Loss is the fraction converted.
Exposure to Open Ocean. Preaching of coastal barriers results not only in overwash but also in exposure of
areas to the open ocean. At the beginning of the simulation, the model changes cells from unexposed to exposed
1-9
-------
Park
if they are in line, based on prevailing wave direction, with exposed water and without intervening areas other
than water. Each time overwash occurs, cells on the lee side are changed to exposed, and with each time step
the adjacent ceU on the lee side is changed to exposed if that cell is all water. This effectively limits erosion in
the once-sheltered water areas and mimics the continuation of shoal-water conditions.
Erosion of Headlands. Erosion of sandy areas is simulated to maintain equilibrium with adjacent beaches; but
erosion of lowlands and uplands that do not have a multispectral signature indicating sand is not simulated.
Therefore, recession of sandy areas is probably overestimated by the model, but erosion of other dry lands is
ignored.
In summary, the model probably errs slightly toward maintenance of wetlands and toward accelerated
loss of barrier islands, but overall it provides prudent forecasts useful in guiding policies for coping with sea level
change.
1-10
-------
Park
CHAPTERS
SIMULATIONS AND SYNTHESIS OF RESULTS
As would be expected from the major differences in coastal physiography around the United States, the
potential for loss of wetland resources due to a rise in sea level varies widely from region to region. Thus,
balanced sampling was required for both the regional and national estimates of losses. Each site was chosen
using an unbiased sampling procedure (see DATA) and covers an area from the open ocean to uplands (or diked
lowlands) so that both loss and migration of wetlands can be evaluated. The standard set of simulations for each
site spans a period from the date of the remote imagery to the year 2100 and provides estimates of response to:
- a projection of the historic trend of sea level rise (1.2 mm/yr) (resulting in a eustatic sea level stand
of 0.14 m above 1986 level by the year 2100); and 1/2-, 1-, 2-, and 3-meter sea levels by the year 2100
no protection of dry lands, protection of backshore areas of beaches only, protection of all developed
areas, and protection of all dry land (including enclosed wetlands)
The sea level scenarios are based on projections given by Thomas (1986), assuming different global-warming
scenarios and the possible melting of glacial ice and thermal expansion of the upper layers of the oceans. The
1-meter scenario is considered most probable in the absence of significant efforts to curb global warming.
The protection scenarios are intended to represent different levels of defending coastal areas from
inundation and erosion. With the "no protection" scenario, only those areas already protected by levees and dikes
will be protected in the future. The "protect backshore" scenario proved to be uninteresting and was not used
in the summaries; it was intended to represent the protection of backshore areas (especially on barrier islands)
through sand nourishment and bulkheads, without protection of adjacent lowlands. Protection of developed areas
represents the effects of enclosing all existing developed areas (see DATA) with dikes and levees; it is
conservative in that further development is not considered. The extreme scenario is protection of all dry land;
note that this can lead to protection of freshwater wetlands that are landward of dry land.
CASE STUDY: LONG BEACH, NEW JERSEY
One case study is presented here to illustrate the level of resolution sought in the database, in the
simulation of changes over time, and in the computational and summarization procedure. Long Beach Island,
New Jersey, and the adjacent bay and low-lying mainland comprise a typical barrier island system to the south
of Atlantic City (Figure 2). The data analysis and simulations of this site provide both insights into, and limited
verification of, the SLAMM2 algorithms.
Examination of the map shows undeveloped and developed dry land, saltmarsh, swamp, and water (Figure
3). An enlarged portion of the map (Figure 3B) shows the 500-m by 500-m cells formed by aggregation of
classes from the unsupervised classification of the Landsat multispectral data, represented by the pixel map
(Figure 3C). Interpretation of the Landsat data was facilitated by comparison with the high-altitude infrared
photograph of the same area (Figure 3D). The elevations and locations of developed areas were obtained from
the topographic map (Figure 3E). The National Wetland Inventory map (Figure 3F) was used to confirm the
interpretations of the classes.
Detailed examination of the maps demonstrates two problems. Only the dominant class is plotted in the
computer-generated map (Figure 3A,B); therefore, some of the islands are not shown because the cells happened
to cover more water than marsh. Also, the swamp in the northwestern corner of the site probably includes
forested upland.
1-11
-------
MDELKTON.
NYPATCHO.
MDEASTON.
NJLONGBE
VABLOXOM
VANEWPOR
Dry land
Suanp
Freshnarsh
Saltnarsh
Mangrove
Beach/Flat
Mater
Dike
Deuelooed
Figure 2. Index map of the mid-Atlantic region.
1-12
kmO
l-
miO
300
300
-------
39 45'
NJLONGBE.
BottOM
Park
,ii.;i'jjjlii5Jiij[ '
B
• Dry land
«8 Swanp
Y/. Freshriarsh
ill! Saltnarsh
^ Mangrove
3 Beach/Flat
Matar
I Dike
• Devielooed
..we,
Rgure 3. Maps of the Long Beach, New Jersey, site showing (A) the initial computer-generated map, (B)
an enlargement of part of the map, and corresponding scenes from (C) the Landsat pixel map,
(D) the high-altitude infrared photograph, (E) the topographic map, and (F) the National Wetland
Inventory map.
1-13
-------
Park
The impact of sea level rise perhaps is best visualized by maps representing different scenarios of sea level
rise, with and without protection of developed areas (Figure 4). Marsh persists with a 14-cm rise in sea level
by the year 2100, although a significant area of swamp is lost by inundation (Figure 4B); more likely, some of
the area presently covered by swamp will slowly be converted to saltmarsh, but that transfer is not simulated.
With a 0.5-m rise by the year 2100, most marsh is lost (Figure 4C). There is little change with 1-, 2-, and 3-m
rises when developed areas are protected (Figure 4D-F); however, without protection, saltmarsh would be free
to migrate onto the unpaved areas of developed lowlands such as those west of Ship Bottom (cf. Figure 4G), and
with 2- and 3-m rises Long Beach Island would be breached (Figure 4H,I).
The gradual changes in cover and decline of wetlands with sea level rise are portrayed most effectively by
an area graph (Figure 5). It is apparent that wetlands will be lost rapidly with any sea level rise greater than the
historic trend; furthermore, almost all saltmarsh is lost by the 2080 when an 0.8-m rise is reached, using the most
probable scenario of 1 m by the year 2100.
This projection is quite different from that of Kana et al. (1988) based on detailed transects and application
of a simple geometric model. They concluded that essentially no wetland would be lost by 2075, given the 1-m
scenario. Examination of their composite transect for the area (Figure 6) shows why their results are so
different. They used two feet as the tidal range, whereas we used one foot, as reported on the Long Beach
topographic map; but, more important, the marsh zones are displaced well above normal levels relative to the
tidal range (equivalent to that expected for a 5-foot tidal range). Kana et al. (1985) attribute this to alteration
of the hydrology by mosquito ditches; however, it is possible that the same regional hydraulic gradient that
promotes the adjacent swamps is responsible for maintaining wetter conditions than would be expected from tidal
conditions alone. Compounding the discrepancy in predictions, we interpreted the marsh to be low based on a
"wet" spectral signature, whereas Kana et al. (1985) found that high marsh dominated (Figure 6). We were
unable to duplicate their projection of no marsh loss by imposing a 5-foot tidal range with high marsh instead
of low. They assumed a uniform sedimentation and accretion rate of 5 mm/yr, compared to our assumption of
5 mm/yr for low marsh, 2.5 mm/yr for high marsh, and none for adjacent lowland; however, varying the
accretion rate of low marsh from 2 mm/yr to 10 mm/yr only changed the timing of the loss of saltmarsh
predicted by SLAMM2 by approximately 17 years (Figure 7). It is difficult to reach any conclusions from this
attempted verification, but it is quite possible that our projection of total loss of marsh is more realistic, given
the tendency for flooding.
NATIONAL IMPACTS
Although losses are highly variable around the VS. coastline, the most important estimates are those for
the nation as a whole. Simulations for 98 sites are planned, but for this initial analysis an unbiased subsample
of 46 sites has been used. These sites represent a broad spectrum of temperate and subtropical coastal types,
with varying tidal ranges, subsidence and accretion rates, fetches, and degrees of development (Table 1).
The percent coverage of the coastline with these 46 sites was used to estimate the initial area of wetlands,
as well as the areas projected to be lost under the different sea-level scenarios. For example, the 29 sites in the
mid-Atlantic subsample represent 8.62% of the area of the Mid Atlantic as defined in this study; the reciprocal
of that value yields a transformation factor of 11.6, which was used to scale the subsample results to the entire
Mid Atlantic. These calculations yield a national estimate of 13,145 mi2 of coastal wetlands in 1986 (the time
of Landsat coverage for the respective sites), compared to the estimate of 14,723 for comparable wetlands as
reported by Titus and Green in this volume. (The wetlands for the mid-Atlantic and West coasts are
underestimated in this initial study.)
In Figure 8 the independent variable, time, has been replaced by sea level, making the ordinate more
general but causing it to be expressed on an exponential scale. Mindful of the nonlinearity, saltmarshes are seen
to expand initially and then-decline by approximately 2,000 mi2 with a 0.1-m rise in sea level in all three
protection scenarios; this decline is due to the predicted rapid loss of saltmarshes in Louisiana and the rest of
1-14
-------
Park
Sujrf City
> BottOH
Beach H< iwen
A Present conditions
B Protected, 14 cm
C Protected, 0.5 m
D Protected, 1.0 m
E Protected, 2.0 m
F Protected. 3.0 m
~*
/
G Unprotected, 1.0 m
H Unprotected, 2.0 m
/
t_i KM >_, M i
Unprotected, 3.0 m
Figure 4.
• Dry land 81 Developed & Suanp
'//. Fresfinarsh $! Salt Marsh S> Mangrove
n Beach/Flat D Mater I Dike
Maps of the Long Beach Island, New Jersey, site showing present conditions and predicted
conditions for the year 2100, with and without residential and commercial developments
protected and with sea levels as indicated. Note the loss of marshes with the 0.5-m scenario.
1-15
-------
Park
I I I I I
158* 2010 2030 2C50 2D70
20CO 3r>yn 20*50
2090
21 OD
Figure 5. Changing areas of wetlands at the Long Beach, New Jersey, site with a 1-m sea level rise.
1-16
-------
Park
COMPOSITE TRANSECT OF THE TUCKERTON MARSH
(Tidal Range = 2.0 ft)
WMW
0 1000 2000 3000
TYPIOM. DISTANCE (TO
' Elevations are relative to the 1929 NCVD sea level
toot
Figure 6. Composite transect of the Tuckerton Marsh at the Long Beach, New Jersey, site (Kana et aL, 1989).
1-17
-------
Park
u
6
n
~r\
\\\
\\v
\\\
\\\
^W
2 mm/y
5 mm/y
10 mm/y
ls84lg95:0062D152D252Q352D4520552c652D752DB520ss
Figure 7. Sensitivity of predictions to different accretion rates for low marsh at Long Beach, New Jersey site.
M8
-------
Table 1.
Park
Study site locations, tidal range, subsidence, sedimentation and accretion rates, and fetch for
the sites used in this initial study.
Site
Latitude
Longitude
Oceanic
Tide
in
Inland
Tide
m
Sub-
sidence
mn/yr
Sedimen-
tation
m/yr
Accretion
mm/yr
Fetch
km
NORTHEAST
MEROCKLA.
MAHARBLE.
MAWESTPO.
RIUATCHH.
CNBRIDGE.
Rock I and
Harblehead
Uestport
Watch Hill
Bridgeport
44 07-30
42 37 '30
41 37-30
41 22-30
41 15'
69
70
71
71
73
07-30
52 '30
07 '30
52 '30
15'
2.8956
2.7737
0.9144
0.7924
2.0726
2.8956
0
0.9144
0.7924
2.0726
1
1
1
0
0
.7
.5
.6
.9
0.20
0.20
0.20
0.20
0.20
2.0
2.0
2.0
2.0
2.0
3
0
1
1
1
MID ATLANTIC
NYPATCHO.
NJLONGBE.
DEREHOBA.
MDEASTON.
MDELKTON.
VABLOXOH.
VANEWPOR.
Patchogue
Long Beach
Rehoboth B.
Easton
Elkton
Bloxom
Newport News
40 52-30
39 45'
38 45
38 52-30
39 37-30
37 52-30
37 07-30
73
74
75
76
76
75
76
07 '30
15'
07 '30
07 '30
00'
37-30
30-
0.7924
0.9144
0.3657
o'
0
1.0972
0.762
0.1828
0.3048
0.0609
0.3048
0.7924
1.0972
0.8534
1
2
1
2
2
1
3
.5
.7
.9
.4
.1
.9
.1
0.20
0.50
0.20
0.20
0.20
0.20
0.50
2.0
5.0
2.0
2.0
2.0
2.0
5.0
6
3
4
1
2
1
5
SOUTH ATLANTIC
NCENGELH.
NCLONGBA.
NCUILHIN.
SCCHARLE.
SCHILTON.
GASEAISL.
FLSTAUGU.
FUCAPECA.
Engelhard
Long Bay
Wilmington
Charleston
Hilton Head
Sea Island
St Augustine
Cape Canaveral
35 45-
35 00-
34 15-
80 00-
32 22-30
31 22-30
30 07 '30
28 30'
76
76
78
30
80
81
81
80
07-30
30-
GO-
OD-
52-30
22 '30
30'
45'
0
0.9144
1.1582
1.5849
2.0421
1 .8501
1.3716
1.0668
0
0.6096
1.1887
0.9753
2.286
2.0421
1.3716
0
0
0
0
2
1
1
0
1
.6
.6
.2
.8
.8
.7
0.20
0.50
0.20
0.50
0.50
1.00
0.20
0.20
2.0
5.0
2.0
5.0
5.0
10.0
2.0
2.0
2
4
2
2
2
3
0
5
SOUTH & WEST FLORIDA
FLMIAHI.
FLKEYWES.
FLEVERGL.
FL VENICE.
FLPORTRI.
FLSNIPEI.
Miami
Key West
25 52 '30
24 37 '30
Everglades City 26 00*
Venice
Port Richey
Snipe Is.
27 15'
28 SO-
SO 07-30
80
81
81
82
83
83
15'
52-30
22-30
30'
45-
52-30
0.762
0.3962
0.6096
0.64
0.6096
0
0.6096
0.3962
1.0668
0.64
0.6096
0.6096
1
1
1
1
0
0
.1
.1
.7
.7
0.20
0.20
0.50
0.20
0.50
0.50
2.0
2.0
5.0
2.0
5.0
5.0
3
0
2
0
15
20
NORTH GULF (EXCL LA)
FLFTGADS.
FLAPALAC.
FLSTJOSE.
FLHOLLEY.
MSGULFPO.
HSPASSCH.
TXALLIGA.
TXPALACI.
TXPORTLA.
TXGREENI.
LOUISIANA
LAMAINPA.
LALULING.
LABARATA.
LAGOLDHE.
LAPEL I CA.
LALMISER.
LAGRANOC.
WEST
CAALBION.
CAPTSAL.
CABENICI.
CASANQUE.
ORYAQUIN.
UAANACOR.
WATACOMA.
Fort Gadsden
Apalachicola
St Joseph
Hoi ley
Gulf port
30 00-
29 45'
29 52-30
30 SO-
SO 30'
Pass Christian 30 22-30
Alligator Hole 29 52-30
Palacios
Portland
Green Is.
Main Pass
Luling
Barataria
Golden Meadow
Pelican Pass
Lake Mi sere
Grand Chenier
Albion
Point Sal
Benicia
San Quentin
Yaquina
Anacortes
Tacoma
28 45'
27 52 '30
26 30'
29 22'30
29 52-30
29 45'
29 30'
29 15'
30 07'30
29 52-30
39 15-
35 00'
38 15'
38 00*
44.451
48 45'
47 30'
85
85
85
87
89
89
94
96
97
97
89
90
90
90
90
93
93
123
120
122
122
124
122
122
07-30
07' 30
30'
07 '30
15'
15-
15'30
15'
22'
22'
15'
15'
22 '30
22 '30
22 '30
00'
00'
52 '30
45'
22 '30
30'
07-30
45'
30'
0
0.7010
0.3992
0
0
0.3048
0.3048
0.3048
0
0.3048
0.1524
0
0
0
0.3048
0
0.6096
1.2192
1.2192
1.2192
0
1.8288
1.524
3.3528
0.6400
0.3992
0.3992
0.4572
0.6096
0
0
0.3048
0
0.1524
0.1524
0.0762
0.1524
0.0762
0.3048
0
0.6096
0
1.2192
0.6096
1.2192
1.8268
1.524
2.4384
1
1
0
1
1
1
12
2
2
3
9
8
9
9
13
8
8
0
0
1
0
-1
0
0
.2
.2
.7
.2
.2
.2
.8
.8
.9
.3
.5
.3
.3
.8
.5
.5
.68
.1
.2
.8
0.20
0.20
0.20
0.20
0.20
0.20
0.50
0.20
0.20
0.20
0.50
0.50
0.50
1.00
0.50
0.20
0.50
0.20
0.20
1.00
0.20
0.20
0.20
0.20
2.0
2.0
2.0
2.0
2.0
2.0
5.0
2.0
2.0
2.0
5.0
5.0
5.0
10.0
5.0
2.0
5.0
2.0
2.0
10.0
2.0
2.0
2.0
2.0
2
8
7
6
20
20
0
6
4
7
2
1
4
0
0
6
0
0
0
0
8
2
10
4
1-19
-------
Park
B
Ql
10
SEA LEVEL (m)
9f
00
07
04 ID
SEAl£VQ.(m)
Figure 8. Changing areas of wetlands of the contiguous US. coast with global warming and (A) with all dry
lands protected, (B) with existing developed areas protected, and (C) with no protection.
1-20
-------
Park
the northern Gulf of Mexico. With all dry land protected, saltmarshes will continue to decline until almost all
disappear with a 3-m rise, and mangrove swamps will not change significantly in area; with only developed areas
protected, the decline of saltmarshes will be much less pronounced and mangrove swamps will increase
significantly; with no protection, saltmarshes will decline even less rapidly. With a 1-m rise by the year 2100, and
with all developed areas being protected from inundation, 8,673 mi or approximately 65% of the coastal
wetlands of the contiguous United States will be lost (Table 2).
REGIONAL TRENDS
Mid-Atlantic
The Mid-Atlantic region (Figure 2) has moderate tidal ranges, extensive barrier island and estuarine
systems, and adjacent lowlands that could be colonized by saltmarshes if not protected by engineering structures.
Currently, vegetated wetlands cover approximately 746 mi2, as estimated in this study. Long Beach, New Jersey,
typifies the barrier island systems and has already been examined (Figures 3,4). The regional response for the
total protection scenario (Figure 9) would consist of a gradual decline of wetlands, with tidal flats replacing
saltmarshes.
Southeast
Vegetated wetlands cover 11,735 mi2 in the southeastern states, from North Carolina to Texas; this is 89%
of the vegetated wetlands of the contiguous United States. Therefore, the responses of the southeastern wetlands
to sea level rise (Figure 10) are similar to those presented for the United States. Coastal responses have been
considered separately for several subregions, with examination of sites representing estuarine, deltaic, barrier
island, and subtropical carbonate platform environments.
South Atlantic Coastal Plain. This region (Figures 11, 12) is characterized by relatively high tidal ranges and
by extensive low terraces representing Pleistocene barrier island systems (such as the sea islands of South
Carolina and Georgia). Wetlands are well developed, especially near major rivers where sedimentation and
accretion are highest; at present they cover 3,813 mi . Because of the high tidal ranges and the availability of
lowlands for colonization, these wetlands will be more persistent in the face of rising sea level compared with
other U.S. coastal areas (Figure 13). Mangroves can be expected to spread into the more southerly sites if
temperatures increase.
Charleston. South Carolina. This historic town is at the confluence of several estuaries with large tidal ranges
and extensive marshes. Near the mouth of the harbor Sullivan's Island (celebrated in The Gold Bug by Edgar
Allen Poe) forms a barrier island with a well developed back-barrier marsh. The recent rediversion of the Santee
River has decreased the high sedimentation that helped promote the historic growth of wetlands; however,
assuming no change in historic trends, the model predicts a slight expansion of marshes onto dredge-fill areas
by the year 2100 (Figure 14B). With a half-meter rise, most of the saltmarshes would be inundated and
converted to tidal flats exposed only at low tide (Figure 14C); increasing sea level stands would lead to
widespread inundation of low-lying areas adjacent to the rivers (Figure 14D-F). The 1-meter scenario yields an
estimate of 47% loss of saltmarsh by 2075, which is quite close to the estimate of 51% by 2075 reported by Kana
et al. (1988), based on detailed transects.
Sea Island. Georgia The wetlands of the Georgia Sea Islands have been studied intensively for many years.
Marshes are very well developed in the region, due to a combination of high tidal ranges and high sedimentation
rates. The site covers two topographic quadrangles, including the communities of Sea Island, St. Simons Island,
and part of Brunswick on the mainland (Figure 15). There is no appreciable change in wetlands and adjacent
lowlands until almost a 1-meter rise, due to the existence of a 2-meter tidal range (with marshes occupying the
upper meter) and a low Pleistocene terrace. With a 1-meter rise and a significantly warmer climate, mangroves
would spread onto the low terrace and the marsh would begin to break up. By two meters the community of
1-21
-------
Park
Table 2.
Regional and national estimates of losses of coastal marshes and swamps by the year 2100 (in
square miles) for the baseline case and four sea-level scenarios, considering also three levels of
shoreline protection: total protection of all dry land, standard protection of areas with residential
and commercial development, and no protection. Results are based on simulations using
unbiased samples encompassing variable percentages of total area of respective coasts.
Region Sampled Baseline
Northeast 3.36%
Total
Standard 39
No
Mid Atlantic 11.6%
Total
Standard (39)
No
South Atlantic 10.07%
Total
Standard (59)
No
S&WFIorida 10.71%
Total
Standard (157)
No
Louisiana 13.66%
Total
Standard 2,271
No
Other N. Gulf 13.04%
Total
Standard 270
No
West* 4.87%
Total
Standard (71)
No
United States
Total
Standard 2,254
No
0.5 m
88
55
27
485
201
120
2.295
1,438
1,313
623
92
63
2,450
2,368
2,345
530
396
360
37
(286)
(332)
6,511
4,263
3,904
1 m
93
58
8
520
341
180
2,422
1,669
1,516
829
157
122
3,742
2,732
3,732
1,301
932
918
36
(440)
(518)
8,673
6,441
6,046
2m
100
33
(67)
625
429
361
2,542
1,812
1,606
1,020
165
120
4,758
4.686
4.685
1,121
994
982
39
(651)
(791)
10,206
7,423
6,892
3 m
188
434
335
705
574
465
2,736
2,227
2,044
1,300
665
631
4,801
4,776
4.778
1,170
1,079
1,070
53
(761)
(843)
10,953
8,994
8,480
* subject to correction
1-22
-------
Park
1000
S£A1£VEL
Figure 9. Changing areas <# mid-Atlantic coastal wetlands with global warming and sea level rise, and with
existing developed areas protected.
1-23
-------
Park
10
B
I
QO Q2 07 14 24
Ql Q4 10 1.8
£A LENC. (m)
iO
Figure 10. Changing areas of coastal wetlands of the southeast with global wanning and (A) with all dry lands
protected, (B) with existing developed areas protected, and (C) with no protection.
1-24
-------
• Dry land
& Suanp
% Frcshnarsft
21 Saltnarsh
S? Hanaro we
m Beach/Flat
I Water
I Dike
• Deuelooed
Park
N
kmO
I-
miO
t
100 I
100
NCENGELH.
Figure 11. Index map of study sites in the Carolines.
1-25
-------
Park
GA
miO
• Dry land
$ Suano
% Freshnarsh
3 Saltnarsh
SS Bang rove
g. Beach/Fiat
I Oik*
Figure 12. Index map of study sites in Georpa and Northeastern Florida.
1-26
-------
Park
SEAtEVE.
Figure 13. Changing areas of South Atlantic coastal wetlands with global wanning and sea level rise, and with
existing developed areas protected.
1-27
-------
Park
;ans Is
A Present conditions
B Protected, 14 cm
C Protected, 0.5 m
D Protected, 1.0 m
E Protected, 2.0 m
F Protected, 3.0 m
G Unprotected, 1.0 m
H Unprotected. 2.0 m
,_,Kn
Unprotected, 3.0 m
• Dry land & Developed SS Swanp
'//, FreshMarsh i=ii Saltnarsh $> Mangrove
m Beach/Flat D Mater I Dike
Figure 14. Maps of the Charleston, South Carolina, site showing present conditions and predicted
conditions for the year 2100, with and without residential and commercial developments
protected and with sea levels as indicated. Note the widespread inundation of lowlands.
1-28
-------
Park
ijjjijijjjjiijljijiji:!
s Is,
A Present conditions
i
!:;
B Protected, 14 cm
C Protected, 0.5 m
D Protected, 1.0 m
E Protected, 2.0 m
,_,Kn
F Protected, 3.0 m
• Dry land K Developed SS Suanp
'//. Freshnarsh !=i! Salt Marsh ^ Mangrove
n BeachXFlat D Mater I Dike
Figure 15. Maps of the Sea Island, Georgia, site showing present conditions and predicted conditions for the
year 2100, with protection of residential and commercial developments and sea levels as indicated.
Note the persistence of the marshes and the eventual spread of mangroves.
1-29
-------
Park
Sea Island would either be a diked enclave in the sea or a largely inundated and eroded shoal area. By three
meters the only appreciable dry land left, other than in developed areas, would be the Pleistocene beach ridge
on Little St. Simons Island.
Southern and Western Coasts of Florida. Southern Florida (Figure 16) is a low-lying carbonate platform and
string of coral-reef keys that developed during interglacial epochs when sea level was higher than it is today. A
slight rise in sea level would inundate this area again. The Everglades are maintained as a broad freshwater
marsh by runoff from the interior; the saltwater wetlands are dominated by mangrove swamps which occur over
a narrow elevational range due to the low tidal range. This initial study estimated the area of vegetated wetlands.
Everglades City. Florida. This site is characterized by mangrove swamps and low beach ridges, with Everglades
on the slightly higher ground (Figure 17). As shown by the maps, mangrove swamps would actually increase in
area with sea level rise as the low beach ridges are inundated by a 2-meter rise in sea level. With more than a
2-meter rise, the mangroves would gradually decrease in area.
The west coast north of subtropical southern Florida (Figure 18) has a very gradual slope offshore, marshes
on the open coast, a low tidal range, and a low terrace along the coast. Consequently, it too is quite vulnerable
to sea level rise. Furthermore, with moderating temperatures, the entire region would become subtropical and
mangrove swamps would spread north.
This initial study estimated the area of vegetated wetlands in southern and western Florida to be 1,869 mi2.
Considering the regional response (Figure 19), with a 1-meter sea level rise all saltmarsh would be lost, and
mangroves would expand onto adjacent lowlands at the expense of freshwater swamps. At higher sea level
stands, both the mangrove and freshwater swamps would be gradually lost.
Northern Gulf Coast Excluding Louisiana. The northeastern Gulf Coastal Plain (Figure 20) is characterized by
low tidal ranges, generally narrow barrier islands and spits, and poorly developed low terraces; consequently,
wetlands are not extensive, except in river deltas, and they are vulnerable to sea level rise.
Apalachicola Bay and Fort Gadsden. Florida. This is the last relatively undisturbed bay and estuary in the
eastern Gulf of Mexico. With a 1-meter rise in sea level St. Vincents Island would be inundated, the narrow
barrier island would be breached in several places, and the Apalachicola River estuary would exceed 2 miles in
width (Figure 21D). The swamp would be gradually inundated and mangroves would replace saltmarsh as the
dominant vegetated saltwater wetland, thus converting the area into subtropical estuary. At higher stands of sea
level even more wetlands would be lost.
Gulfport. Mississippi. Gulfport and Long Beach are built on a low beach ridge on Mississippi Sound. The fetch
is large enough that beach erosion can occur, and the elevation is low enough that with a half-meter rise and with
protection of developed areas a lagoon would open up behind Long Beach (Figure 22C); without protection Long
Beach would be lost with a 1-meter rise (Figure 22G). This is one of the most dramatic impacts on an urban
area observed in the study.
With the exception of eastern Texas, which is a continuation of the chenier (old beach) plain of Louisiana,
the coast of Texas (Figure 23) is characterized by higher terraces along the coast. These combined with the low
tidal range will lead to an early loss of wetlands with sea level rise. At the present time high salinities in the
Laguna Madre of South Texas prevent development of marshes and mangroves. This salinity control is not
included in the model, so marshes may be over-represented in the lower sea level rise scenarios. Presumably
with higher sea level there will be breaching of Padre Island and exchange with the open ocean so that salinity
will not be a limiting factor in the maintenance of wetlands.
Considering the entire northern Gulf Coast other than Louisiana, this study estimates the present area of
vegetated wetlands to be 1,218 mi2. A striking pattern of saltmarsh loss would take place, with only a slight rise
1-30
-------
Park
• Dry land
& Swanp
% Fresfmarsh
Si Saltnarsh
3? Mangrove
n Beach/Flat
Mater
I Dike
B Deuelooed
kmQ
miO
100
N
!
100
FLEVERGL
Fort
Lauderdale
Figure 16. Index map of southern Florida sites.
1-31
-------
Park
A Present conditions
B Protected, 14 cm
C Protected, 0.5 m
D Protected, 1.0 m
E Protected, 2.0 m
F Protected, 3.0 m
• Dry land M Developed SJ Suanp
'//, Freshnarsh Si Saltnarsh ^ Mangrove
m Beach/Flat D Water I Dike
Figure 17. Maps of the Everglade City, Florida, site showing present conditions and predicted conditions for the
year 2100 with protection of residential and commercial developments and with sea levels as indicated.
Note the spread of mangroves onto the lowland and, at 3m, inundation of all but the beach ridge and
town.
1-32
-------
FLSNIPEI.
Park
miO
too
Figure 18. Index map of sites on the west coast of Florida
1-33
-------
Park
2500
Ql
Q2 Q7 14
Q* 1.0 ie
SEAIT.S.
Figure 19. Changing areas of south and west Florida coastal wetlands with global wanning and sea level rise,
and with existing developed areas protected.
1-34
-------
Park
MSGULFPO.
Figure 20. Index map of northeastern Gulf Coast coastal sites,
1-35
FLAPALAC.
-------
Park
• Dry land
& Suanp
'//. Freshnarsh
fill Saltiiarsh
^ Mangrove
g Beach/Flat
Water
I Dike
5 Deuelooed
Figure 21. Maps of the Apalachicola and Fort Gadsen, Florida, sites showing present conditions (A), and
predicted conditions for the year 2100 with residential and commercial developments protected and
with sea levels of 14-cm (B), 0.5-m (C), 1.0-m (D), 2.0-m (E), and 3.0-m (F).
1-36
-------
Park
Gulf port
Lonsr Beach
A Present conditions
B Protected, 14 cm
C Protected, 0.5 m
D Protected, 1.0 m
E Protected, 2.0 m
F Protected, 3.0 m
G Unprotected, 1.0 m
H Unprotected, 2.0 m
...KM
I Unprotected, 3.0 m
• Dr« land 8 Developed SS Suanp
'//, Freshnarsh Hi! Saltwarsh ^ Mangrove
m Beach/Flat D Mater I Dike
Rgure 22. Maps of the Gulfport, Mississippi, site showing present conditions and predicted conditions
for the year 2100, with and without protection of residential and commercial developments and
with sea levels as indicated. Note the complete inundation of Long Beach if unprotected.
1-37
-------
Park
Island
Figure 23. Index map of Texas sites.
1-38
-------
Park
in sea level (Figure 24); with a one-meter rise most freshwater marshes would be lost as well. With moderating
climate saltmarshes would be replaced by mangroves, and the northern Gulf would become a subtropical coastal
system.
Louisiana. A surprisingly similar response is predicted for this region (Figure 25). At present vegetated wetlands
cover 4,835 mi . Tidal flats may initially expand as beaches, cheniers (old beaches), and dredge-fill areas are
reclaimed by the sea; but saltmarshes will continue their rapid decline; mangroves would increase; and, prior to
a 1-meter rise, most freshwater marshes would be lost (Figure 26). The pattern predicted for this region is due
to the low tidal range, extensive alluvial and coastal lowlands, high subsidence rates, and high sedimentation rates.
Coastal swamps and marshes cover most of the region. Half the wetlands that would be lost in the southeast,
with a 1-meter rise in sea level and standard protection of developed areas, are in Louisiana (Table 2).
Luling-Pelican. Louisiana. This strip of four sites forms a transect from the western suburbs of New Orleans
(including the airport) to the Gulf of Mexico (Figure 27). The area is now losing wetlands rapidly due to
subsidence and human disturbance. The simulations represent an approximate response to sea level rise, but
do not represent the complex interactions of changing river courses, floodwater diversions, variable sediment
supplies, and salinity controls that affect the coastal ecosystems. These are better treated by region- and
site-specific models such as that by Costanza et al. (1987). The tenuous position of the saltmarshes is shown by
the baseline simulation in which most of the present saltmarshes are lost by the year 2100 (Figure 27B). The
half-meter simulation (Figure 27C) is almost identical to the baseline, which coincides with the flat regional
response for that range of sea levels (Figure 26).
West Coast
California (Figure 28) is characterized by rugged coastlines with low-lying areas. Most wetlands are in San
Francisco Bay, where the high tidal range and high sedimentation rates created extensive wetlands prior to their
destruction by dredge fill and by diking for salt ponds. At present, high subsidence rates and erosion rates are
contributing to the decline of the remaining marshes. Heavy development has left little potential for the spread
of wetlands onto adjacent lowlands. However, the model shows an unexpected spread of wetlands (Table 2).
This appears to be due to errors in digitizing the elevations; also, salt pannes are not modeled in the current
version. Until the data are checked thoroughly, the significant increase in wetlands under all scenarios should
be viewed with suspicion.
Benicia. California. The remaining wetlands, now only a fraction of their original extent, are adjacent to
developed areas and diked lowlands and salt ponds; they are lost with even a half-meter rise in sea level (Figure
29). If lowland areas were not protected, the marshes would migrate inland with sea level rise.
The Pacific Northwest (Figure 30), like California, is characterized by rugged, high-energy coasts and few
low-lying areas. Where the topography is suitable for marsh formation, as in Yaquina Bay, Oregon, high tidal
ranges would help to perpetuate those marshes despite a half-meter sea level rise.
Although the sample size is small and the data should be checked further, this initial study estimates the
area of current vegetated wetland of the West Coast to be 64 mi . As shown in Figure 31 (which illustrates the
total protection scenario and is not subject to elevational errors), aggregated wetlands exhibit a rather flat
response to sea level rise.
Northeast
This diverse region (Figure 32) is similar to the West Coast with its high tidal ranges and rocky coasts, but
includes easily eroded Cape Cod. This initial study estimates the area of vegetated wetlands to be 600 mi2. The
subsample of four sites is too small to define the region well; however, a pattern does emerge. Swamps
developed on poorly drained glacial tills occur at elevations above those that could be inundated by sea level rise
1-39
-------
Park
00
01
IB 2* 10
Figure 24. Changing areas of northern Gulf Coast coastal wetlands (excluding Louisiana) with global warming
and sea level rise, and with existing developed areas protected.
1-40
-------
LALMISER.
Dm land
Suano
Freshnarsh
Saltnarsrt
Mangrove
Beactv'FJat
Hater
I Dike
• Deuelooed
Park
Figure 25. Index map of Mississippi delta and Louisiana-Texas chenier plain sites.
1-41
-------
Park
Q1
Figure 26. Changing areas of Louisiana coastal wetlands with global warming and sea level rise, and with existing
developed areas protected ' - •
1-42
-------
Park
Dry land
Suanp
B
Figure 27. Sites south of New Orleans, Louisiana, showing initial conditions (A) and predicted conditions for the
year 2100 with 0.14-cm (B), 0.5-m (C), and 1.0-m (D) sea level stands.
1-43
-------
Park
Dry land
<$ Suanp
Freshnarsh
ill! Saltwarsh
5 Mangrove
Beaclv^Flat
Uater
Dike
S Deuelooed
Figure 28. Index map of California sites.
1-44
-------
Park
San t
Pablo
Bay
A Present conditions
B Protected, 14 cm
C Protected, 0.5 m
D Protected, 1.0 m
E Protected, 2.0 m
F Protected, 3.0 m
• Dry land & Developed SS Swanp
'//. Freshnarsh iiii Saltnarsh ^ Mangrove
n Beach/Flat D Mater I Dike
Figure 29. Maps of the Benicia, California, site showing present conditions and predicted conditions for the year
2100, with protection, of residential and commercial developments and with sea levels as indicated.
Note the extensive tidal flats at higher sea levels.
1-45
-------
Park
kmo
I-
miO
300
N
t
300
V •
e
• Dry land
*
& Swanp
% Frashnarsh
2 Saltnarsh
$? Mangrove
X Btfach/Flat
Uat«r
I Dika
• Daualooad
Rgure 30. Index map of Pacific Northwest sites.
1-46
-------
Park
SALINE.
Figure 31. Changing areas of West Coast coastal wetlands with global warming and sea level rise, and with all
dry land protected.
1-47
-------
Park km 0
miO
300
300
Dry land
Suanp
Freshmarsh
Saltnarsh
$? Mangrove
Beach/Flat
Mater
Dike
B Deuelooed
Figure 32. Index map of New England sites.
1-48
-------
Park
in the next century and therefore persist. Some coastlines, such as parts of the Maine coast, have no marshes
at present and will not develop marshes due to the rocky substrate. Other areas have marshes protected by
baymouth bars and spits; these marshes will not be affected adversely by lower stands of sea level.
Watch Hill. Rhode Island. This site includes the historic resort town of that name, with a long, narrow spit
maintained by longshore drift to the west. It also includes Misquamicut Beach and adjacent marsh (Figure 33).
The marsh will gradually disappear with higher sea level and eventually Misquamicut Beach will be breached,
according to the simulations.
1-49
-------
Park
MisquANicut
Hatch Hill
A Present conditions
B Protected, 14 cm
C Protected; 0.5 m
,Kn
D Protected. 1.0 m
E Protected, 2.0 m
F Protected, 3.0 m
• Dry land 8 Developed & Suawp
'//. Freshnarsh Hi! Saltnarsh ^> Mangrove
n BeachXFlat D Water | Dike
Figure 33. Maps of the Watch Hill, Rhode Island, site showing initial conditions (A) and. predicted conditions
for the year 2100, with protection of residential and commercial developments and with sea levels as
indicated. Note the breaching of Misquamicut Beach and loss of marshes.
1-50
-------
Park
CHAPTER 4
SUMMARY
Projections of coastal wetland and lowland changes are made using remotely sensed cover-class data,
lowland elevations digitized from topographic maps, and a simulation model based on simple geometric
relationships of inundation and subsidiary influences of spatially varying exposure and erosion relationships.
Approximately 9% of the coast of the contiguous United States was sampled in this phase of the study. Forty-
six sites were chosen using an unbiased, systematic sample of coastal topographic maps; several additional sites
were chosen for purposes of model verification.
Examination of eight representative sites provides insights into local responses of wetlands to accelerated
sea level rise. Some sites, such as Sea Island, Georgia, are well buffered for sea levels up to 1 meter, due to high
tidal ranges and high sedimentation and accretion rates. Other sites, especially in the Gulf of Mexico, are quite
vulnerable to small changes in sea level; Long Beach, Mississippi, could be completely inundated by less than a
1-meter rise. Although the model is perhaps too simple for representing complex deltaic dynamics, it projects
a continuation of current trends for the Louisiana coast south of New Orleans, with the entire undiked area
shown to be at risk without any acceleration in sea level rise.
Composite responses were obtained by summing cover-class areas for each simulation time step for sites
comprising a regional sample, and by transforming the values into regional estimates. The results are presented
in Table 2 and as area graphs showing the patterns of changes in wetland areas with an exponential increase in
sea level up to 3 meters. Regional responses are as follows:
A
• The mid-Atlantic Coast, with 746 mi of diverse wetlands and moderate tidal ranges, would probably exhibit
a gradual decline in vegetated wetlands and an increase in tidal flats; with standard protection of all currently
developed lowlands, 27% of marshes and swamps would be lost with a 0.5-m rise by the year 2100, 46% with a
1-m rise, 58% with a 2-m rise, and 77% with a 3-m rise.
• The South Atlantic Coast, with 3,813 mi2 of vegetated wetlands, is predicted to exhibit a gradual decline
in vegetated wetlands by the year 2100, with mangrove swamps replacing saltmarshes as the dominant saltwater
vegetated wetland at the more southerly sites; with standard protection, 38% of the fresh- and saltwater wetlands
would be lost with a 0.5-m rise, 44% with a 1-m rise, 48% with a 2-m rise, and 58% with a 3-m rise.
^^ f\
• The southern and western coasts of Florida, with 1,869 mi of vegetated wetlands, have large expanses of
low-lying, undeveloped areas suitable for wetlands to migrate onto; therefore, vegetated wetlands would not
decrease greatly in area by 2100 unless extensive dikes were constructed. With standard protection, only 5% of
existing vegetated wetlands would be lost with a 0.5-m rise, 8% with a 1-m rise, and 9% with a 2-m rise; however,
36% would be lost with a 3-m rise. As the entire region becomes subtropical, mangrove swamps would spread
northward and saltmarshes would gradually disappear.
• The northern Gulf Coast, excluding Louisiana, is estimated to have 1,218 mi2 of vegetated wetlands; with
its low tidal ranges and general lack of suitable lowlands for wetlands to migrate onto, saltmarshes would be lost
with low sea-level rises by 2100. With standard protection, 33% of the vegetated wetlands would be lost with
a 0.5-m rise, 77% with a 1-m rise, 82% with a 2-m rise, and 89% with a 3-m rise. Mangroves would take the
place of saltmarshes as the climate moderates.
• Louisiana is estimated to have had 4,835 mi2 of vegetated wetlands at the time the satellite imagery was
obtained; these wetlands are being lost at an alarming rate without accelerated sea-level rise. With standard
protection, this study estimates that by 2100,47% of the vegetated wetlands would be lost with a 0.14-m rise in
eustatic sea level (the historic rise continued), 49% with a 0.5-m rise, 57% with a 1-m rise, 97% with a 2-m rise,
and 99% with a 3-m rise.
1-51
-------
Park
The sample size is too small in this initial study to reach conclusions about the West and Northeast Coasts, but
preliminary results suggest that wetlands would be lost gradually due to the ameliorating effect of high tidal
ranges.
These regional tabulations were summed to obtain estimates and area graphs for the entire coast of the
contiguous United States, which has approximately 14,000 mi2 of coastal wetlands (including 13,145 mi2 of
vegetated wetlands) at the present time. With protection of all existing residential and commercial developments,
by the year 2100 saltmarshes and freshwater swamps would gradually decline in area, mangrove swamps would
gradually increase, and freshwater marshes would gradually decline until a sea level of 0.8 m is reached.
Freshwater marshes would then disappear rapidly (reflecting a pattern seen in both the Gulf and mid-Atlantic
Coasts); 35% of vegetated wetlands would be lost with a 0.5-m rise, 49% with a 1-m rise, 56% with a 2-m rise,
and 68% (almost 9,000 mi2) with a 3-m rise. If all coastal lowlands not already protected by dikes were allowed
to be inundated by a 3-m rise, a more gradual loss of saltmarshes would occur compared to the standard
protection scenarios; 30% of vegetated wetlands would be lost with a 0.5-m rise, 46% with a 1-m rise, 52% with
a 2-m rise, and 65% with a 3-m rise. Conversely, total protection of all dry land would result in an accelerated
decline in saltmarshes and no increase in areas of mangrove swamps; 50% of vegetated wetlands would be lost
with a 0.5-m rise, 66% with a 1-m rise, 78% with a 2-m rise, and 83% with a 3-m rise.
Alternatively, we could consider the sea level rises of a half meter and 2 meters as being reasonable bounds
on a probable 1-meter rise. Therefore, with no protection of dry land other than existing dikes, 3,900 to 6,900
mi2 of vegetated wetlands could be lost by the year 2100. With existing residential and commercial developments
protected, 4,300 to 7,400 mi2 of vegetated wetlands could be lost; with all dry land protected, 6,500 to 10,200 mi2
of vegetated wetlands could be lost.
Wetlands provide important habitat for many fish and wildlife species, including rare and endangered birds
on all three coasts, and over half the commercially important coastal fish species of the Southeast. Wetlands also
remove pollutants and protect inland areas from floods, storms, and high tides. Therefore, policy decisions
should be made to protect these valuable natural resources from the consequences of global warming and sea
level rise.
1-52
-------
Park
REFERENCES
Armentano, T.V., RA. Park, and C.L. Cloonaa 1988. Impacts on wetlands throughout the United States. In
Titus, James G. (ed), Impact of Sea Level Rise on Coastal Wetlands in the United States. Washington: US.
Environmental Protection Agency, pp. 87-149.
Bahr, L.M., Jr., R. Costanza, J.W. Day, Jr., S.E. Bayley, C. Neill, S.G. Leibowitz, and J. Fruci. 1983. Ecological
Characterization of the Mississippi Deltaic Plain Region: A Narrative with Management Recommendations.
FWS/OBS-82/69, Slidell, Louisiana: Fish and Wildlife Service, 189 pp.
Barth, M.C., and J.G. Titus. 1984. Greenhouse Effect and Sea Level Rise. New York: Van Nostrand Reinhold.
Bird, E.C.F. 1986. Mangroves and intertidal morphology in Westernport Bay, Victoria, Australia. Marine
Geology. 69:251-271.
Browder, J A., HA. Hartley, and K.S. Davis. 1985. A probabilistic model of the relationship between marshland-
water interface and marsh disintegration. Ecological Modelling 29:245-260.
Bruun, Per. 1962. Sea Level Rise as a Cause of Shore Erosion. Journal of the Waterways and Harbors Divisioa
Proceedings of the American Society of Civil Engineers, 88(WW1): 117-130.
Bruun, Per. 1986. Worldwide Impacts of Sea Level Rise on Shorelines. In Effects of Changes in Stratospheric
Ozone and Global Climate. Vol. 4: Sea Level Rise, edited by James G. Titus, 99-128. Washington, DC:UJS.
Environmental Protection Agency.
Costanza, R., C. Neill, S.G. Leibowitz, J.R. Fruci, L.M. Bahr, Jr., and J.W. Day, Jr. 1983. Ecological Models
of the Mississippi Deltaic Plain Region: Data Collection and Presentation. ILS. Fish and Wildlife Service,
Division of Biological Services, Washington, D.C. FWS/OBS-82/68, 342 pp.
Costanza, R., F.H. Sklar, M.L. White, J.W. Day, Jr. 1987. A dynamic spatial simulation model of land loss and
marsh succession in coastal Louisiana In: Wetland Modelling, WJ. Mitsch, M. Straskraba, and S.E. Jergensen
(eds.), Elsevier, Amsterdam.
Dasarathy, B.V. 1974. HINDU: Histogram Inspired Neighborhood Discerning Unsupervised Pattern
Recognition System. NASA/Computer Sciences Corp., Huntsville, Alabama, 43 pp.
Davis, RA. 1985. Coastal Sedimentary Environments. New York: Springer-Verlag.
Day, J.W., Jr. W.G. Smith, P.R. Wagner, and W.C. Stowe. 1973. Community Structure and Carbon Budget
of a Salt Marsh and Shallow Bay Estuarine System in Louisiana. Center for Wetland Resources, Louisiana
State University, Baton Rouge. Publ. No. LSU-SG-72-04.
Dyer, K.R. 1986. Coastal and Estuarine Sediment Dynamics. Chichester, U.K.: John Wiley & Sons, 342 pp.
Estes, JA., and GA. Thorley. 1983. Manual of Remote Sensing. Vol. II: Interpretations and Applications.
Second Edition. American Society of Photogrammetry.
Gosselink, J.G. 1984. The Ecology of Delta Marshes of Coastal Louisiana: A Community Profile. FWS/OBS-
84/09, Washington, DC: U.S. Fish and Wildlife Service, 134 pp.
Hall, S.L., W.R. Wilder, and F.M. Fisher. 1986. An analysis of shoreline erosion along the northern coast of
East Galveston Bay, Texas. Journal of Coastal Research 2:173-179.
1-53
-------
Park
Kana, T.W., B J. Baca, and M.L. Williams. 1988a. Charleston case study. In: J. G. Titus (ed.), Greenhouse
Effect, Sea Level Rise and Coastal Wetlands. EPA-230-05-86-013, U.S. Environmental Protection Agency,
Washington, pp. 37-59.
Kana, T.W., W.C. Eiser, BJ. Baca, and M.L. Williams. 1988b. New Jersey case study. In: J. G. Titus (ed.),
Greenhouse Effect, Sea Level Rise and Coastal Wetlands. EPA-230-05-86-013, US. Environmental Protection
Agency, Washington, pp. 61-86.
Leatherman, S.P. and R.E. Zaremba. 1986. Dynamics of a northern barrier beach: Nauset Spit, Cape Cod,
Massachusetts. Geological Society of America Bulletin 97:116-124.
Lefor, M.M., W.C. Kennard, and D.L. Civco. 1987. Relationships of salt-marsh plant distributions to tidal
levels in Connecticut, USA. Environmental Management 11:61-68.
Letzsch, W.S. and R.W. Frey. 1980. Deposition and Erosion in a Holocene Salt Marsh, Sapelo Island, Georgia
Journal of Sedimentary Petrology. 50(2):529-542.
Lugo, A.E., and M.M. Brinson. 1978. Calculations of the value of salt water wetlands. In Wetland Functions
and Values: The State of Our Understanding, pp. 120-130, American Water Resources Association.
Mausel, P.W. 1985. Remote Sensing: Introduction to Computer Assisted Techniques. ISURSL Information
Note 09031985. Indiana State University Remote Sensing Lab, Terre Haute, pp. 72-79.
Montague, C.L., A.V. Zale, and H.F. Percival. 1987. Ecological effects of coastal marsh impoundments: a
review. Environmental Management, vol. 11, pp. 743-756.
Park, RA., T.V. Armentano, and C.L. Cloonan. 1986a. Predicting the impact of sea level rise on coastal
systems. In Supplementary Proceedings for the 1986 Eastern Simulation Conference. Norfolk, Virginia, pp.
149-153.
Park, R A., T.V. Armentano, C.L. Cloonan. 1986b. Predicting the effects of sea level rise on coastal wetlands.
In James G. Titus (ed.), Effects of Changes in Stratospheric Ozone and Global Climate. Vol. 4: Sea Level Rise.
pp. 129-152, Washington: U.S. Environmental Protection Agency.
Peltier, W.R. 1988. Global sea level and earth rotation. Science, vol. 240, pp. 895-901.
Pope, R.M., and J.G. Gosselink. 1973. A tool for use in making land management decisions involving tidal
marshland. Coastal Zone Management Journal. 1:65-74.
Ryszkowski, Lech. 1984. Primary production in agroecosystems. Options Mediterraneennes. Mediterranean
Agronomic Institute of Saragossa, Spain, 84(l):77-94.
Sherrod, C.L., and C. McMillan. 1985. The distributional history and ecology of mangrove vegetation along
the northern Gulf of Mexico. Contributions in Marine Science, 28:129-140.
Sklar, R. Costanza, and J. W. Day, Jr. 1985. Dynamic spatial simulation modeling of coastal wetland habitat
succession. Ecological Modelling 29:261-281.
Swanson, R.L., and C.I. Thurlow. 1973. Recent subsidence rates along the Texas and Louisiana coasts as
determined from tide measurements. Journal Geophysical Research 78:2665-2671.
, ^ '..
Teal, J.M. 1962, Energy flow in the salt marsh ecosystem of Georgia. Ecology 43:614-624.
Teal, J.M., and M. Teal. 1969. Life and Death of the Salt Marsh. New York: Audobon-Ballantine.
1-54
-------
Park
Thomas, R. 1986. Future sea level rise and its early detection by satellite remote sensing. In James G. Titus
(ed.), Effects of Changes in Stratospheric Ozone and Global Climate. Vol. 4: Sea Level Rise, pp. 19-36,
Washington: ILS. Environmental Protection Agency.
Thurman, H.V. 1983. Essentials of Oceanography. Columbus, Ohio: Charles E. Merrill.
Titus, J.G. ed. 1986. Effects of Changes in Stratospheric Ozone and Global Climate, 4 vols. Washington:
U.S. Environmental Protection Agency.
Titus, J.G. ed. 1988. Greenhouse Effect, Sea Level Rise and Coastal Wetlands. EPA-230-05-86-013,
Washington: U.S. Environmental Protection Agency.
Titus, J.G., and S.R. Seidel. 1986. Overview of the effects of changing the atmosphere. In James G. Titus
(ed.), Effects of Changes in Stratospheric Ozone and Global Climate, Vol. 1: Overview, pp. 3-19, Washington:
UJS. Environmental Protection Agency.
Villach. 1985. International assessment of the role of carbon dioxide and of other greenhouse gases in climate
variations and associated impacts. Conference Statement. Geneva: United Nations Environment Program.
Watkins, J.S., M.L. Bottino, and M. Morisawa 1975. Our Geological Environment. Philadelphia: W. B.
Saunders.
Wiegert, R.G., R.R. Christian, J.L. Gallagher, J.R. Hall, R.D.H. Jones, and R.L. Wetzel. 1975. A preliminary
ecosystem model of a coastal Georgia Spartina marsh. In Estuarine Research, Vol. 1, edited by L. E. Cronin,
pp. 583-601. New York: Academic Press.
1-55
-------
NATIONAL ASSESSMENT OF BEACH NOURISHMENT
REQUIREMENTS ASSOCIATED WITH ACCELERATED
SEA LEVEL RISE
by
Stephen P. Leatherman
Laboratory for Coastal Research
University of Maryland
1113 Lefrak Hall
College Park, MD 20742
Contract No. 68-01-72-89
-------
CONTENTS
CHAPTER 1: INTRODUCTION 2-1
OBJECTIVE 2-2
REPORT OUTLINE 2-2
CHAPTER 2: METHODOLOGY 2-3
STUDY SITES 2-3
DATA SOURCES 2-3
COASTAL SEGMENTATION 2-3
CLOSURE DIMENSIONS 2-3
SEA LEVEL RISE SCENARIOS 2-4
SAND VOLUME DETERMINATIONS 2-4
SAND RESOURCE AVAILABILITY 2-4
BEACH NOURISHMENT COSTS 2-6
FLORIDA CASE STUDY 2-7
WAVE CLIMATE AND CLOSURE DEPTH SELECTION 2-7
AREA MEASUREMENTS 2-9
SAND SOURCES AND ASSOCIATED COSTS 2-9
CHAPTER 3: NATIONWIDE RESULTS 2-17
CHAPTER 4: CONCLUSIONS 2-29
REFERENCES 2-30
11
-------
Leatherman
CHAPTER 1
INTRODUCTION1
A significant portion of the United States population lives within the coastal zone, with many buildings
and facilities located at elevations less than 3 meters (10 feet) above sea level. These structures are presently
subject to damage during storms, and this hazard has grown increasingly serious as sea levels have risen during
the twentieth century. Greenhouse-induced warming is expected to raise water levels at historically
unprecedented rates, resulting in increased beach erosion and flooding.
Despite these potential hazards, the coastal population is burgeoning. In fact, development in the coastal
zone is proceeding at rates that more than double inland construction. Hundreds of thousands of beachfront
structures (exquisite single-family houses, high-rise condominiums, and elegant hotels) have been built within a
few hundred feet of an eroding ocean shore: Beachfront property is some of the most valuable real estate in the
country, exceeding $20,000 per linear foot of shoreline along the U.S. mid-Atlantic coast.
The present dilemma and developing disaster have resulted from the tremendous investment in coastal
property at a time when most sandy beaches nationwide are eroding. Best estimates are that 90 percent of the
U.S. sandy beaches are presently experiencing beach erosion (Leatherman, 1986). Accelerated sea-level rise will
increase erosion rates and associated problems.
Public attention is yet to be critically focused on the beach erosion problem. The present (1988) drought
and heat wave have brought about a dramatic awakening and interest of citizens in the greenhouse effect and
climate change. Hopefully, a coastal disaster along an urbanized beach will not be necessary to promote public
awareness of the sea level rise phenomenon and its attendant impacts.
Sea level is a primary control on shore position, which, in human terms, translates to beach erosion when
water levels are rising. While weather is subject to large-scale variations and hence climate change trends are
difficult to measure, rising sea levels are relatively easy to discern and can be thought of as the dipstick of climate
change, reflecting the integration of many earth surface processes.
There are three general responses to accelerated sea level rise: retreat from the shore, armor the coast,
or nourish the beach. Beach nourishment is the focus of this report, wherein sand is artificially placed on the
beach. Other tactics for combatting the sea level rise/coastal erosion problem are discussed elsewhere in this
volume. The proper shore protection response is site-specific on a community or coastal sector basis due to
large differences in environmental and socioeconomic factors. The abandonment alternative is not realistic for
urbanized beaches. For less developed areas along eroding shorelines, planning decisions are less clear cut.
Therefore, the costs and benefits of stabilization vs. retreat must be carefully considered as the cost in either
case is likely to be quite high (National Research Council, 1987).
The principal approach today of protecting coastal property and maintaining recreational beaches is
beach nourishment. Engineering structures, such as groins and seawalls, have often been shown to cause
detrimental effects on adjacent beaches. Also, their construction and maintenance costs are quite high.
Therefore, coastal communities have come to rely upon a "soft" engineering solution - beach nourishment, since
it is environmentally sound, aesthetically pleasing, and up-to-this-time, economically feasible. However, the
projected accelerated sea level rise will cause more rapid rates of beach loss and could make even this alternative
too costly for many resort areas along the United States coastline.
'Although the information in this report has been funded wholly or partly by the U.S. Environmental
Protection Agency under.Contract No. 68-01-72-89, it does not necessarily reflect the Agency's views, and no
official endorsement should be inferred from it.
2-1
-------
Leatherman
OBJECTIVE
The overall objective of this research is to estimate the cost to nourish all the major recreational oceanic
beaches in the UJS., given various sea level rise scenarios. It is clear that developed coastal resort residents
would prefer not to move back and abandon the coast and will attempt to stabilize the shore through beach fill
projects. The approach is to place enough sand on the beach to maintain stable (nonretreating) conditions with
rising sea levels. The quantity of sand required "to hold the line" is evaluated under various sea level rise
scenarios (rise/year combinations) at foot intervals up to a 10-foot rise situation by the year 2100.
REPORT OUTLINE
This Introduction is followed by a general Methodology section. The beach nourishment analysis were
undertaken at the community level, from which state and national totals were determined. Delray Beach,
Florida, was selected as a case study to illustrate the type of analysis conducted for each area Finally, the
national results are presented.. Of the approximately 7,000 miles of sandy shoreline in the U.S., 1,920 miles of
beaches were evaluated in this study. These areas are considered to be the principal recreational beaches in the
country.
2-2
-------
Leatherman
CHAPTER 2
METHODOLOGY
STUDY SITES
This report focuses on the twenty-one coastal states in the United States. Alaska is excluded because
of its undeveloped nature. Some states have only one to a few coastal resort areas (e.g., Ocean City in
Maryland), while others, particularly Florida and New Jersey, are known for their many recreational beaches.
The major recreational beaches in each state are examined; state averages for nourishment needs are then
tabulated from the site-specific calculations. Therefore, cost estimates of beach fills are made based on local
(community or physiographic) conditions to produce statewide and national assessments.
DATA SOURCES
The last national assessment of shore erosion and associated planning implications was undertaken by
the U.S. Army Corps of Engineers in 1971. Their national survey, based on District Corps office reports,
indicated the prevalence of shore erosion. In addition, there are numerous site-specific reports and information
available for various locales. These data were assembled and analyzed to extract information pertinent to the
study. Corps District personnel and State Coastal Zone Management (CZM) officials were also queried for
any up-to-date information and insights. Specifically, the basic information for analysis was largely obtained from
the following sources: U.S.G.S. topographic maps for area! measurements and offshore contours, National
Research Council (1987) report for sea level rise scenarios, supplemented by estimates from Hoffman et al
(1986), baseline relative sea level rise rates for U.S. coast (Lyles et al., 1987), and CERC Inner-Continental Shelf
Studies (ICONS) data sets on offshore sand resources.
COASTAL SEGMENTATION
The coast in each state is divided into three categories: (1) publicly owned, undeveloped; (2) privately
owned, undeveloped, and (3) already developed. Roughly one-third of the U.S. coast falls into each of these
categories. Publicly owned, undeveloped areas (e.g., state parks, national seashores, and NASA installations)
will most likely never be developed, but some areas may be nourished. In general, these areas are not
considered in the nourishment assessment, unless beach fill has already been undertaken and is likely to continue
(e.g., Huntingdon Beach State Park, SC).
Most of the areas contained in the privately owned, undeveloped coastal area category are identified in
the 1983 U.S. Congress COBRA legislation, and usually are excluded from receiving federal assistance in
shoreline stabilization by law. However, these areas still have the potential to be developed, and are therefore
included in the national assessment. Inclusion of these locales represents a worst case scenario in terms of the
total amount of area needing nourishment.
Developed areas have already been urbanized or are already somewhat developed and are likely to be
extensively developed in the future. Beaches that have been nourished in the past or have undergone full-scale
urbanization are the best candidates for further restoration. Areas are delimited along the coast by jurisdic-
tional (e.g., town, city) boundaries or natural demarcations (e.g., inlets) into geomorphic units.
CLOSURE DIMENSIONS
Offshore closure depth is specified for each area on the basis of Hallermeier's (1981) determinations
for the U.S. coast. Hallermeier's (1981) approach relies upon statistical wave data, which is available for the
2-3
-------
Leathennan
entire U.S. coast, and his work represents the state-of-the-art in the field. Some may feel that the derived values
for closure depth are too conservative in certain areas. In this case, a simple ratio of utilized and preferred
values can be used to calculate higher sand volumes and costs.
The US. Geological Survey 7.5-minute quadrangles, supplemented by National Ocean Service nautical
charts, were used to determine the horizontal distance offshore to closure depth for each coastal sector. This
data source was selected since UJS.G.S. quadrangles (1:24,000 scale) are the most commonly utilized maps in
the country, depicting both surficial (e.g., urban development and topography) features and offshore contours.
SEA LEVEL RISE SCENARIOS
A total of six sea level rise (SLR) scenarios are considered in this analysis, evaluated from 1 to 10 feet
at 1-foot intervals but without the time exceeding year 2100. These scenarios are based on previous studies by
the National Research Council (1987) and the US. Environmental Protection Agency (see Titus and Greene,
this volume). The total component of rise can be calculated as follows: T (t) = (0.0012 + M/l,000)t + tb2,
where M is the local (isostatic) factor (in mm/yr) and b describes the concaval upward slope of the quadratic
equation. Estimates of M from Lyles et al. (1987) can be obtained from Table 1 as determined by NRC (1987),
and values of the coefficient b are listed in Table 2 for the six scenarios.
SAND VOLUME DETERMINATIONS
The direct approach of "raising the beach/nearshore profile" is utilized because of its straightforward
application. The beach to offshore closure depth distance (d) represents the active profile dimension. For every
increment (x) of sea level rise, the volume of sand required to "raise the profile" simply corresponds to xd per
unit of shoreline length. This approach overcomes objections to the Bruun Rule formulation regarding
on/offshore sand transport. Also, other methodologies require considerably more data (e.g., Trend Analysis
necessitates knowledge of historical shoreline change and the Sediment Budget Model involves site-specific
information on transport rates; Leatherman, 1985). In this analysis, longshore losses are shown separately in the
tables so that the sand required to mitigate accelerated sea level rise, alone, is clearly stated.
As sea level rises, the land surface becomes relatively lower with respect to mean water level, resulting
in increased frequency and more severe coastal flooding. For barrier islands, the decision will likely be made
at some point to raise the barrier elevations to overcome or lessen the effects of this problem. It is assumed that
after 1 foot of sea level rise, coastal communities will start raising the bayside areas of the island, which are less
than 5 feet above mean sea level. Prior to this point, it can be argued that the cost and nuisance of such actions
would dictate inaction, and people would tolerate the increased flooding. By the time a 4-foot rise in mean sea
level is achieved, the entire barrier surface, including the dunes but excluding wetlands, will have been raised in
concert with water levels to prevent storm overtopping. Therefore, the procedure involves calculation of the
elevational distribution above and below the Moot (MSL) elevation to compute the area and hence volume of
sand required with different scenarios of seaJevel rise. Some barrier- islands and mainland areas had general
elevations above the 15-foot contour line. No mitigating action was deemed necessary for these areas.
SAND RESOURCE AVAILABILITY
Once the quantities of sand required to maintain the recreational beaches for various SLR scenarios have
been established, a determination of available sand resources available to match this projected need must be
undertaken. The preferred borrow site areas are generally located offshore for most states. Backbarrier lagoons
and bays have been utilized in the past for small quantities of material, but environmental objections and
incompatibility of material because of size have precluded further use of such sources. Mining of mainland sand
pits has been employed locally in some areas, but again the resources are limited and this type of activity is not
permitted in most states.
2-4
-------
Leatherman
Table 1. Relative Sea Level for the United States Coast, 1850-1986*
Trend
Location mm/yr
Atlantic Coast
Eastport, ME
Bar Harbor, ME
Portland, ME
Seavey Is., ME
Boston, MA
Woods Hole, MA
Newport, RI
Providence, RI
New London, CT
Bridgeport, CT
Montauk, NY
Port Jefferson, NY
Willets Pt., NY
New Rochelle, NY
New York, NY
Sandy Hook, NJ
Atlantic City, NJ
Philadelphia, PA
Lewes, DE
Baltimore, MD
Annapolis, MD
Solomons Is., MD
Washington, DC
Kiptopeke, VA
Hampton Roads, VA
Portsmouth, VA
Wilmington, NC
Charleston, SC
Ft. Pulaski, GA
Fernandina, FL
Mayport, FL
Miami Beach, FL
Key West, FL
2.7
2.7
2.2
1.8
2.9
2.7
2.7
1.8
2.1
2.1
1.9
2.7
2.4
0.6
2.7
4.1
3.9
2.6
3.1'
3.2
3.6
3.3
3.2
3.1
4.3
3.7
1.8
3.4
3.0
1.9
2.2
2.3
2.2
ft/vr
.009
.009
.007
.006
.010
.009
.009
.006
.007
.007
.006
.009
.008
.002
.009
.014
.013
.008
.010
.010
.012
.011
.011
.010
.014
.012
.006
.011
.010
.006
.007
.008
.007
Location
Gulf Coast
St. Petersburg, FL
Cedar Key, FL
Pensacola, FL
Grand Isle, LA
Eugene Island, LA
Sabine Pass, TX
Galveston, TX
Galveston, TX
Freeport, TX
Rockport, TX
Padre Island, TX
Port Isabel, TX
Pacific Coast
San Diego, CA
La Jolla, CA
Newport, CA
Los Angeles, CA
Santa Monica, CA
Port San Luis, CA
San Francisco, CA
Alameda, CA
Crescent City, CA
Astoria, OR
Neah Bay, WA
Seattle, WA
Friday Harbor, WA
Nawiliwili, HI
Honolulu, HI
Hilo, HI
Trend
mm/vr
2.3
1.9
2.4
10.5
9.7
13.2
6.4
7.5
14.0
4.0
5.1
3.1
2.1
2.0
1.9
0.8
1.8
1.2
1.3
1.0
-0.6
-0.3
-1.1
2.0
1.4
2.0
1.6
3.6
ft/vr
.007
.006
.008
.034
.032
.043
.021
.024
.046
.013
.017
.010
.007
.007
.006
.003
.006
.004
.004
.003
-.002
-.001
-.004
.006
.004
.006
.005
.012
*Source: Lyles et al (1987)
2-5
-------
Leatherman
Table 2. Value of Coefficient "b" for Six Scenarios Considered in This Study
Eustatic Component "b" Coefficient
Scenario* fbv year 2100 in m)** fm/vr2)
I
n
m
rv
V
VI
0.5
1.0
1.5
2.0
2.5
3.0
.00002795
.00006642
.00010490
.00011340
.00018180
.00022030
* Source: Scenarios I, II, and III; NRC (1987)
Scenarios IV, V, and VI; EPA (1986)
**Projections do not exceed year 2100
Inlet sand shoals (termed "ebb tidal deltas") often contain large quantities of beach sand and are located
in close proximity of important resort areas, particularly in the State of Florida. Future plans for dredging of
these inlets and outer shoals for ship navigation should include a provision for sand placement on adjacent or
specified beach areas, rather than dumping sand far offshore in deep water. Unfortunately, some inlet material
is chemically polluted, particularly in the New York-New Jersey metropolitan areas, and there is also the concern
of sand drawndown from adjacent beaches if extensive sand dredging and shoal removal are realized. In fact,
the large ebb tidal delta off Ocean City, Maryland (8 million cubic yards of clean beach sand) probably will not
be used in the forthcoming beach restoration project because of concerns of accelerated, post dredging erosion
of adjacent beach areas. Therefore, only offshore borrow sites are considered in this analysis as ebb-tidal deltas
are generally too small regarding sand volumes or clouded by political concerns (unless specifically recommended
for usage by state authorities).
Offshore sources along the Atlantic Coast were delineated by a series of studies undertaken in the recent
past by the U.S. Army Corps of Engineers Coastal Engineering Research Center (CERC). The Inter-
Continental Shelf (ICONS) study involved all the states from New York to Florida, and these reports are used
to obtain a good estimate of available sand reserves. This information is supplemented for the rest of the coast
and updated by Corps District reports for specific areas. In addition, the U.S.G.S. and M.M.S. (Minerals
Management Services) hi association with various state geological surveys (e.g., Louisiana, Maryland, and Maine)
have expanded and updated these inventories.
BEACH NOURISHMENT COSTS
Sand costs are estimated for the range of alternatives (e.g., various SLR scenarios evaluated at particular
years when a certain sea level has been achieved). Values are based on current rates per cubic yard of material.
A sand cost function was developed from past dredging experience and applied to each coastal sector. It is clear
that as the less expensive, closer-to-shore sand supplies are exhausted, the costs will rise as a step-function
(approximately $1.00 per cubic yard per mile farther offshore as booster pumps are added). The "base rates"
vary regionally so that the actual costs are site-specific. This study gives us the ability to predict for the first time
the nourishment requirements for various SLR scenarios and associated costs for individual resort areas, resulting
in statewide and national estimates.
2-6
-------
Leatherman
FLORIDA CASE STUDY2
The Delray Beach area along the Florida Atlantic coast was chosen as the case study to illustrate the
type of analysis conducted in this report for coastal communities nationwide. The Delray Beach area in Florida
is heavily developed and requires analysis at all levels addressed in this report; profile nourishment, backbarrier
and oceanside elevation raising, sand volume requirements, offshore sediment supplies, and associated dredging
costs. In addition, Delray Beach has been nourished in the past (1973,1978, and 1984), and the state of Florida
proposes to continue such projects in the future (Florida DNR, 1988).
Delray Beach is located in southeast Florida in rapidly growing Palm Beach County. It shares boundaries
to the north and south with Boynton Beach and Boca Raton, respectively. The beach area functions as a barrier
island as it is fronted by the Atlantic Ocean to the east and the Intracoastal Waterway to the west. The barrier
is rather low-lying, with maximum elevations-approximately 15 feet in the dune line. The barrier varies in width,
but averages 1800 feet along its length. The beaches in this locale are composed of medium to coarse shelly
sands and are underlain by similarly composed coquina rock (Florida DNR, 1988).
WAVE CLIMATE AND CLOSURE DEPTH SELECTION
Closure depth represents the seaward limit of significant sediment transport along a beach profile and
is the offshore extent to which beach nourishment should occur. Nourishment of the beach profile to this
distance is imperative for the success and longevity of beach replenishment projects; otherwise, wave action will
rapidly rework the new sediment (which can appear as beach erosion to the casual observer), moving a portion
of it offshore to attain profile equilibrium.
With these considerations in mind, a technique was developed (Hallermeier, 1981) to determine closure
depths at various coastal locations in the U.S. based on local sand characteristics and summary statistics of
annual wave climate (Table 3). The results of Hallermeier's work were used in this study to determine the
appropriate offshore extent of proposed beach nourishment. When Hallermeier's (1981) predictions were not
available for a particular study area, approximate closure depths were extrapolated from the closest given
locations.
Hallermeier defined two offshore limits in his work, d1 and dr The d1 limit was used to estimate closure
depths for this report, as it (d.,) represents the "maximum water depth for sand erosion and seaward transport
by an extreme yearly wave condition, and corresponds to a seaward limit of appreciable seasonal profile change
"(Hallermeier, 1981). The d. value, on the other hand, corresponds to an offshore depth where "expected surface
waves are likely to cause little sand transport" (Hallermeier, 1981). Therefore, to ensure inclusion in the active
profile, sand introduced by beach nourishment should be added out to the offshore depth, d.,. Hallermeier's
(1981) recommended applications for the seaward limit also suggest using dd1 values as the offshore limit for
beach nourishment projects. Although extreme storm events may move sand beyond the d1 location, this report
assumes that beach nourishment projects are based on average wave conditions.
The closure depth used for Delray Beach was 4.2 meters. This depth was determined from Hallermeier's
(1981) work for Boca Raton, Florida (Table 3), the municipality immediately adjacent to southern Delray Beach.
The area is subject to a mild wave climate (average height of 1.59 feet) and closure depths are relatively shallow
and near the shore as a result.
After determining closure depths, United States Geological Survey (USGS) topographic maps (7.5-
minute series) were used to estimate the distance from the shoreline to these depths. For example, the
bathymetric points closest to 4.2 meters were located on the U.S.G.S. map for Delray Beach and measured as
^This section was authored by Ms. Gary Gaunt, Laboratory for Coastal Research, University of Maryland.
2-7
-------
Leathennan
Table 3. Selected Closure Depths for the United States Coastline (Hallermeier, 1981)
Site Closure Depth (m)
Atlantic Coast
Assateague, MD 5.40
Bull Island, SC 4.20
Tybee Lighthouse, GA 7.40
Boca Raton, FL 4.20
Lake Worth, FL 5.04
Atlantic City, NJ 7.04
Virginia Beach, VA 6.25
Nags Head, NC 7.95
Atlantic Beach, NC 5.65
Wrightsville Beach, NC 5.35
Holden Beach, NC 4.72
Gulf of Mexico
Naples, FL 2.98
Destin, FL 4.30
St. Andrews Park, FL 4.50
Crystal Beach, FL 4.80
Gilchrist, TX 4.10
Galveston, TX 3.80
Corpus Christi, TX 5.20
Pacific Coast
Huntington Beach, CA 5.83
San Clemente, CA 7.10
Bolsa Chica, CA 6.10
Pt Mugu, CA 5.70
Pismo Beach, CA 7.50
San Simeon, CA 6.50
Capitola Beach, CA 4.20
Stinson Beach, CA 7.30
Wrights Beach, CA 10.40
Shelter Cove, CA 7.10
Prairie Creek, CA 7.0
Umpqua, OR 7.8
Closure depth is defined based on Hallermeier's d, limit [i.e., "the maximum water depth for nearshore erosion
by extreme (12 hours per year) wave condition"]. The d1 limit represents the recommended seaward limit of
beach nourishment "to ensure its (the sands) inclusion in the annually very active littoral zone (Hallermeier,
1981)." This limit does not consider sand transport due to extreme storm events and thus ignores sand
potentially lost to the outer profile in such scenarios.
2-8
-------
Leatherman
an average of 600 feet offshore. This distance was used in the calculations to determine the area of beach profile
nourishment (i.e., length of beach to be nourished multiplied by offshore distance of closure depth = beach
profile area to be nourished).
AREA MEASUREMENTS
The initial response to rising sea levels is beach profile nourishment. Following this preliminary measure,
it is possible that low-lying areas of the barrier may be raised to a higher elevation by sediment input to prevent
submergence. This report assumes that backbarrier elevations are raised after 1 foot of sea level rise and that
oceanside elevations are raised after 4 feet of sea level rise. Sediment volumes needed to raise these barrier
elevations were approximated in the following manner:
o U.S.G.S. topographic maps (7.5-minute series) were obtained for each study area.
o Backbarrier areas less than S feet above MSL were delineated.
o Oceanside areas greater than 3 feet, but less than 15 feet above MSL, were delineated, unless the
higher elevation represented a dune line. If a dune line was shown, it was included in the oceanside
area measurement, as it is likely that dune elevations would be raised in concert with beach elevations
to maintain the storm buffer.
o Area measurements for each delineated backbarrier and oceanside location were estimated using an
engineer's ruler and the map scale. All area measurements represent the average width and length
for the delineated locations. Only buildable (i.e., not marshy) areas were included. Small, isolated
locations were not included, as the maintenance (e.g., dredging) costs would likely exceed associated
economic benefits.
When tabulating final results, calculations for physiographically similar locations often were lumped
together. Thus, the Delray Beach profile, backbarrier, and oceanside areas were summarized with the results
of Boynton and Highland Beaches and Boca Raton to give the final results:
o Profile Nourishment: 1.693 million cubic yards of sand needed for one foot of sea level rise (SLR);
o Backbarrier Elevations: 0.946 million cubic yards of sand needed for one foot of elevation raising;
o Oceanside Elevations: 3.217 million cubic yards of sand needed for one foot of elevation raising.
Table 4 summarizes the above results for varying SLR scenarios.
Sand volume estimates like those given above were derived for all developed and developable coastal
localities. These individual site results were then cumulated as statewide totals. Table 5 provides an example
summary for the Florida (Atlantic) coast.
SAND SOURCES AND ASSOCIATED COSTS
The VS. Army Corps of Engineers was involved in the late 1960s to mid-1970s in an inventory of the
morphological and sediment characteristics of the Inner Continental Shelf (ICONS Studies) in an effort to locate
sand suitable for beach nourishment endeavors. Using high-resolution seismic reflection surveys and sediment
coring techniques, they performed a preliminary assessment of offshore borrow sites suitable for the restoration
of nearby beaches (Duane and Meisburger, 1969) The ICONS surveys were the primary sources used in this
report to locate sand sources suitable for future beach nourishment projects.
2-9
-------
Leatherman
Table 4.* Sand Volume Requirements for Boynton Beach to Boca Raton, PL (including Defray Beach)
to Raise the Beach/Nearshore Profile and Barrier Elevations with Sea Level Rise**
Sea Level Rise
(feet)
Beach/Nearshore
Profile
(million yd3)
Barrier Elevations
(million yd3)
Backbarrier Oceanside
Total
(million yd )
1
2
3
4
5
6
7
8
9
10
1.693
3386
5.079
6.772
8.465
10.158
11.851
13544
15.237
16.930
0
0.946
1.892
1838
3.784
4.730
5.676
6.622
7568
8514
0
0
0
3.217
6.434
9.651
12.868
16.085
19302
22519
1.693
4332
6.971
12.827
18.683
24539
30395
36.251
42.107
47.963
•Excerpted Table 30 from Florida (Atlantic) Report.
**Shore length considered for nourishment is 14.43 miles.
Table 5.* Summary of Sand Volume Requirements for the Atlantic Coast of Florida to Raise the
Beach/Nearshore Profile and Barrier Elevations with Sea Level Rise
Sea Level Rise
(feet)
Beach/Nearshore
Profile
(million
Barrier Elevations
(million yd3)
Backbarrier
Total
(million yd3)
Oceanside
1
2
3
4
5
6
7
8
9
10
76.981
153.962
230.943
307.924
384.905
461.886
538.867
615.848
692*29
769.810
0
22358
44.716
67.074
89.432
111.790
134.148
156506
178.864
201.222
0
0
0
85.219
170.438
255.657
340.876
426.095
511314
596533
76.981
176320
275.659
460.217
644.775
829333
1,013.891
1,198.449
1383.007
1567565
* Excerpted Table 32 from the Florida (Atlantic) report.
2-10
-------
Leatherman
For example, the Florida Atlantic coast was studied as a part of the ICONS program. Surveys were taken
covering the following areas:
o Miami to Palm Beach (Duane and Meisburger, 1969)
o Palm Beach to Cape Kennedy (Meisburger and Duane, 1971)
o Cape Canaveral (Field and Duane, 1974)
o Cape Canaveral to Georgia (Meisburger and Field, 1975)
Table 6 briefly characterizes sand located by the ICONS studies. These maps and data contained in the ICONS
reports were used to identify the offshore locations of sand suitable for beach replenishment. For example, Field
and Duane (1974) define sand suitable for beach nourishment in the Cape Canaveral area as "medium to coarse,
well-sorted quartzsize-mollusk sand." Seismic profiling and coring suggest that such sand sources lie in large
offshore shoals such as Chester Shoal, which contains an estimated sand quantity of 8.8 x 106 yd3 Table 7
summarizes all reported Florida Atlantic sand reserves in terms of offshore location, quantity, and associated
nourishment cost. The data provided in Table 7 were paramount in calculating final sand costs associated with
beach nourishment.
Initial dredging costs for the Florida Atlantic coast were established at $4.00 per cubic yard, based on
Bruun (1985). Bruun's paper examined some beach nourishment projects in Florida, where he noted that project
costs ranged between $3 and 5.00 per cubic meter ($2.27-3.78 per cubic yard), with costs recently increasing. It
is likely that dredging costs will continue to increase with future beach restoration projects, thus the higher figure
of $5.00 per cubic meter (approx. $4.00 per cubic yard) was used to provide preliminary estimates of future sand
costs for beach nourishment.
The $4.00 per cubic yard value was used only for sand reserves located within one mile of the shore, as
dredging costs increase with greater distance offshore. The Army Corps of Engineer's "rule of thumb" for
dredging cost escalation is $1.00 per cubic yard for each additional mile offshore (Weggel, personal
communication, 1987). This rule only applies for sand reserves within 5 miles of the shore when a floating
pipeline dredge system is used to pump the sand directly to the beach. Beyond 5 miles, the sand must be moved
in two stages; dredging onto a ship for transport to the mainland, followed by truck hauling to location. Costs
for this process are not clearly known, but it is estimated that at least $2.00 per cubic yard would be added to
the highest floating pipeline cost. This rate is used for all cost calculations requiring sand beyond 5 miles. Table
7 summarizes sand costs based on offshore location for the Florida Atlantic coast.
Given data on offshore sand reserves and associated dredging costs, it is possible to project future costs
of beach nourishment projects given various SLR scenarios. This report examines six SLR scenarios and
projected costs at 20-year intervals (2000-2100) given relative sea level rises (RSLR) for each state. Table 8
indicates the RSLR for the Atlantic coast of Florida for each scenario and year studied. These RSLR estimates
were multiplied by the amount of area contained in the beach profile, backbarrier, and oceanside locations to
provide estimates of sand volumes needed for nourishment (Table 9). Recall that beach profiles are nourished
immediately, backbarrier elevations are raised after 1 foot of SLR, and oceanside elevations are raised with 4
feet of rise. An example calculation for the state of Florida (Atlantic) is given as follows (using RSLR Scenario
IV and the year 2100):
o Projected RSLR for Scenario IV by 2100 is 6.94 feet (Table 8)
o Multiply projected RSLR by the appropriate volume of sand needed to nourish 1 foot of each barrier
area -
2-11
-------
Leathennan
Table 6. ICONS Survey Results - Potential Offshore
Sources of Sediment for Beach Nourishment
Study Area
Amount Available
(million yd-*)
Type
Suitability
I.
II.
Miami to
Palm Beach
(Duane and
(Meisburger,
1969)
a)
south of
Boca Raton
b) north of
Boca Raton
Palm Beach
to Cape
Kennedy
(Meisburger
and Duane,
1971)
201; located in
offshore troughs
380; thickly
blanketed over
portions of the
shelf
minimum of 92.2;
located in ridge-
like shoals
Mostly
calcareous
50-50%
quartz and
calcareous
sediments
Medium to
coarse
calcareous
sand
Possibly suitable for
short-term projects,
but are easily
degraded in turbulent
littoral zone and may
become too fine for
long-term projects.
Too fine for success-
ful nourishment of
area's beaches; not
included in
nourishment
assessment.
Sand suitable for
nourishment lies in
three major areas:
- Capron Shoal -
approx. 65.4 x
10B yd3
- Indian River
Shoal - approx.
10.3 x 10ryd3
- Bethel Shoal
approx. 16.5 x
Other shoals in the
area also may have
suitable material,
although evidence is
currently lacking.
2-12
-------
Table 6. Continued.
Leatherman
Study Area
Amount Available Type
(million yd3)
Suitability
III.
Cape
Canaveral
(Field and
Duane, 1974)
approx. 130; large, medium to
south-trending, coarse, well
cape-associated sorted
shoals quartz-
mollusk
sand
IV.
Cape
Canaveral
to Georgia
(Meisburger
and Field,
1975)
minimum of 295;
ten potential
borrow sites
(possibly 21 more
sites) are identi-
fied, with each
having a sand
reserve from 5-
178; total volumes
are unknown - more
study needed;
located in linear
shoals.
Fine to
very quartz
sand in
shoreface;
seaward of
shoreface,
sand is fine
to medium,
well sorted,
predomi-
nantly
quartz sand
Well suited for
nourishment. Sur-
veyed areas show the
following sand
quantities:
- Ohio-Hetzel Shoal
(76.1 x 106 yd3)
- Chester Shoal
(8.8 x 106 yd3)
- The Bull
(31.6 x 106 yd3)
- Southeast Shoal
(15.2 x 106 yd3)
Volumes of suitable
sand in unsurveyed
areas of Chester and
Southeast Shoal are
likely an order of
magnitude larger.
Suitable sand was
identified in the
following locations:
- Jacksonville
(5.0 x 106 yd3)
- Mickler Landing
(178.0 x 106 yd3)
- St. Augustine
(7.4 x 106 yd3)
- Marineland
(39.0 x 106 yd3)
- Qrmond Beach
(66.0 x 106 yd3).
Further study may
show significantly
more sand available.
2-13
-------
Leatherman
Table 7.
Sand Reserves and Approximate Associated Dredging Costs
for the State of Florida (Atlantic Coast)
Distance Offshore
(miles)
< 1 mile
1-2
2-3
3-4
4-5
5-6
6-7
7-8
8-9
9-10
10-11
11-12
12-13
13-14
Sand Amount
(million ydj)
66
87
122.3
48
199.2
-
97.6
-
15.2
76.1
39.0
175
Cost per yd3
($)
4.00
5.00
6.00
7.00
8.00
10.00*
10.00*
10 . 00*
10.00*
10 . 00*
10.00*
10.00*
10 . 00*
10 . 00*
Total Cost
(million $)
264
435
733.8
336
-
1992
-
976
-
152
761
390
-
1750
* At distances > 5 miles offshore, it is highly unlikely that a floating pipeline
dredge would be used. Rather, sand would probably be moved in two stages;
dredging into a ship for transport to mainland, followed by truck hauling to
location. Costs for this process are not clearly known, but it is estimated it
would add at least $2.00 per yd3 to the highest floating pipeline cost.
Year
Table 8.
Amount of Sea Level Rise for Various Year/Scenario
Combinations (in feet)*
II
Scenario
III IV
V
VI
2000
2020
2040
2060
2080
2100
.12
.35
.66
1.04
1.49
2.01
.14
.50
1.03
1.73
2.60
3.65
.17
.64
1.39
2.42
3.72
5.30
.19
.79
1.76
3.11
4.84
6.94
.22
.93
2.13
3.80
5.95
8.57
.24
1.08
2.50
4.49
7.06
10.22
Scenarios I, II, III (NRC, 1987)
Scenarios IV, V, VI (EPA, 1986)
^Relative sea-level rise has averaged 2.2 mm/yr (baseline record from Mayport
and Key West, Florida tide guages). 1986 is base level year for projections.
2-14
-------
Leatherman
-6.94 feet (RSLR) x Table 5 value for Beach/Nearshore Profile (76.981) = 534.248 million yd3
-[6.94 feet (RSLR) - 1 foot] x Table 5 value for backbarrier elevations (22358) = 132.806 million yd3
-[6.94 feet (RSLR) - 3 feet] x Table 5 value for oceanside elevations (85.219) = 335.762 million yd3.
o Add the sand volumes given above (534.248 + 132.806 + 335.762 = 1,002.818 million yd3) to derive
the total sand volumes needed to protect the barriers from encroaching seas (Table 9).
After determining sand volumes needed for each SLR scenario and associated year, nourishment costs
are calculated. Initial dredging costs were used to determine final costs until all sediment supplies within one
mile of the shore were exhausted. In the case of Florida, $4.00 per yd3 was used to project costs for the first
66 million yd3 of sediment (see Table 7). After exhausting these nearby sand reserves, the cost escalation
function was employed for sand each additional mile offshore. For example, Florida's Scenario IV projects that
by the year 2100,1,002.818 million yd3 of sand will be needed for beach nourishment. All of Florida's recorded
offshore sand reserves would be exhausted by such a request; however, costs were calculated assuming the
availability of sand. A sample cost calculation is given as follows:
Scenario IV. year 2100
A
1,002.818 million yd of sand needed for nourishment (see Table 9)...
1,002.818 yd3
- 66.000 vd3 @ $4.00/yd3 (sand w/in 1 mile; see Table 7)
936.818 yd3
- 87.000 vd3 @ $5.00/yd3 (sand 1-2 miles; see Table 7)
849.818 yd3
- 122300 vd3 @ $6.00/yd3 (sand 2-3 miles; see Table 7)
727.518 yd3
48.000 vd3 @ $7.00/yd3 (sand 3-4 miles; see Table 7)
679.518 yd3 @ $10.00/yd3 (sand >5 miles; see Table 7)
Table 9. Sand Volumes Required to Raise the Beach/Nearshore Profile and Barrier
Elevations for the Florida Atlantic Coast (million yd3)
Scenarios
Year
2000
2020
2040
2060
2080
2100
I
9.238
26.943
50.807
80.060
114.702
177314
II
10.777
38.490
79.290
133.177
235.921
340.230
III
13.087
49.268
107.004
218.042
347.183
700.142
IV
14.626
60.815
135.487
286.586
615.246
1002.818
V
16.936
71.592
189.234
355.130
820.105
1,303.647
VI
18.475
83.139
225.989
550.650
1,024.964
1,608.168
LST*
2,800
6.800
10.800
14.800
18.800
22.800
*LST is longshore sediment transport. Average annual rates vary along the Florida Atlantic Coast so a
representative figure of 200,000 yd* is used for illustrative purposes.
2-15
-------
Leatherman
Given the above, total sand costs for Scenario IV, year 2100, for the Florida Atlantic coast are $8,563.980 million
dollars (Table 10).
Tables 9 and 10 also provide sand volume and cost estimates based on average annual longshore transport
rates (200,000 yd3/year) for the Florida Atlantic coast. It is obvious from these figures that sand removed from
the system by longshore transport is insignificant compared to that required to compensate for accelerated sea
level rise.
Sand costs for the Atlantic coast of Florida were derived assuming equal access to offshore sand reserves
for all coastal locations, as it was beyond the scope of this research to evaluate sand supplies and costs on a site-
specific basis. In reality, suitable borrow material is scattered along the coastline; some areas have sand directly
offshore, while other sites are far removed from available supplies. Therefore, the equal access assumption made
in this report may underestimate true sand costs, as sand hauling to distant locations is not adequately examined.
Table 10. Cost of Raising the Beach/Nearshore Profile and Barrier Elevations for the Florida (Atlantic)
Coast ($ millions)*
Year
2000
2020
2040
2060
2080
2100
I
36.952
107.772
203.228
334300
507.510
844.884
II
43.108
153.960
330.450
599.885
1,196.526
1,938.100
III
52348
197.072
469.020
1,089.252
2,007.630
5,537.220
Scenarios
IV
58.504
243.260
611.435
1,511.802
4,688.260
8,563.980
V
67.744
291.960
916.404
2,087.100
6,736.850
11,572.2730
VI
73.900
349.695
1,136.934
4,042300
8,785.440
14,617.480
LST**
11.200
27.200
43.200
59.200
75.200
91.200
*These cost figures assume that the borrow sand is fully compatible in size with the native sand and will remain
on the active beach profile; this is a conservative assumption.
**LST is longshore sediment transport (see footnote for Table 7).
2-16
-------
Leatherman
CHAPTERS
NATIONWIDE RESULTS
The U.S. coastline can be divided on the basis of physiographic regions for discussion purposes. The New
England states typically have small sandy beaches, often consisting of sand spits. Massachusetts has the largest
number of such recreational beaches (Table 11), but those along the Rhode Island coast are perhaps the most
urbanized and have been subject to severe damage during historical hurricanes.
The Mid-Atlantic coast, which extends from New York to Virginia, is in general the most urbanized shore
in the country except for parts of Florida and southern California. The recreational beaches in New York and
northern New Jersey serve as the playgrounds for some 15 million people in the greater New York metropolitan
area Presently, pollution from human waste is adversely impairing their recreational value, but beach erosion
has been a chronic problem, and many nourishment projects have already been completed and others are
planned. Farther south, there are more open stretches of coast (parklands, reserves, etc.) so that the approach
of holding the line by beach fill would be city-specific (e.g., Virginia Beach, VA, Ocean City, MD) rather than
island-wide (e.g., Long Beach Island, NJ).
The U.S. southeastern coast (North Carolina to Florida) is the least urbanized along the Atlantic coast,
but this area has the largest growth potential because of the greatest availability of beachfront property. The
Outer Banks of North Carolina constitute a long chain of barrier islands with development spread out over
long distances (Table 11). While an increasing number of multi-story condominiums are being built, the
traditional type building is the wooden, single-family house that can be readily moved. Therefore, the retreat
alternative becomes more attractive than beach stabilization in many areas. This alternative is plausible to a less
extent in South Carolina and Georgia, but many islands are already too urbanized for this approach (e.g., Hilton
Head, S.C.). Also, the barrier islands in the Georgia bight (southern South Carolina to northern Florida) are
generally higher in elevation, much wider, and more stable than the microtidal barriers found elsewhere along
the Atlantic coast (Leatherman, 1988).
Florida should be considered separately from the others as its immense coastline is the single most
important feature of the state. Almost 300 miles are considered for beach nourishment along the Atlantic coast,
and about 250 miles on the Gulf will require sand fill with accelerated sea level rise. It could be argued that
Florida has the most important beaches in the United States because it serves as a national and even
international resort area. Recreational beaches are the number one source of revenue, and state officials are
considering spending tens of millions of dollars each year for beach nourishment. The Miami Beach project,
completed in 1980 at a cost of $65 million for 10 miles of beach, perhaps represents the scale and magnitude of
future such projects along this rapidly urbanizing coast, which is becoming dominated by the high-density, high-
rise type of development.
The Gulf Coast is the lowest-lying area in the U.S. and consequently is the most sensitive to small changes
in sea level. One of the earliest extensive beach nourishment projects undertaken in the U.S. was in Harrison
County, Alabama, in the 1950s. The beaches have greatly narrowed since this time, and renourishment is now
required. Louisiana has the most complex coastline in the region and also has the distinction of having the most
rapid rate of beach erosion in the nation. While a number of islands are included on the state list for
nourishment (Table 11), much of this proposed work will probably never be undertaken since it is uneconomical
under today's conditions. There are only two recreational beaches in the State of Louisiana-Grand Isle and
Holly Beach. While Grand Isle was recently nourished, it is unlikely that the economics (relative high cost of
sand fill vs. value of property to be protected) will make future projects feasible with accelerated sea level rise.
Texas has the most extensive sandy coastline in the Gulf, but much of the area is little inhabited. Clearly the
City of Galveston will be maintained, but the nearly century-old seawall has been most effective in this regard
largely at the expense of tfie beach. Elsewhere, beach nourishment is probably not the most viable alternative
as much land on the barrier islands is generally available for relocation.
2-17
-------
Leatherman
with Sea-Level
Rise
Sand Volume Needed
State/Site
Maine-Higgins Beach
State Total
New Hampshire beaches (Total)
Massachusetts-Humarock Beach
Siasconset
State Total
Rhode Island beaches (Total)
Connecticut beaches (Total)
New York-Southampton Beach
State Total
New Jersey-Long Beach Island
State Total
Delaware-Rehoboth Beach
Dewey Beach
Bethany/South Bethany
Fenwick Island
North Bethany
State Total
Maryland-Ocean City (Total)
Virginia-Wallops Island
Virginia Beach
Sand Ridge
State Total
North Carolina-Currituck Banks
Currituck Spit
Nags Head area
Buxton
Bogue Banks
Topsail Island/
Beach
lee Island
complex
Miles of
Shoreline
0.62
30.87
8.91
2.75
1.70
100
27.42
63.51
7.05
120
18.03
125
1.55
1.59
2.95
1.12
2.73
9.94
8.94
5.89
5.89
5.21
16.99
2.84
29.47
19.82
0.95
22.67
21.14
3.60
Figure Eight Island 3.64
Profile
0.303
5.229
0.967
0.333
27.390
6.970
43.467
2.756
46.910
5.289
36.668
0.243
0.249
0.462
0.175
0.426
1.555
2.447
5.759
1.728
2.241
9.728
1.667
17.288
11.627
0.556
6.207
4.547
0.774
0.774
Bayside
i. i.
~~~~
3.168
3.778
26.180
.
0.132
0.262
0.394
----
_.. .._
— --
0.519
1.176
0.141
0.096
(million yd3)
Oceanside
-— -
3.139
1.220
___
137.984
3.308
7.663
30.272
2.936
20.362
—-...i
0.189
1.095
0.219
1.503
4.124
0.691
3.995
1.268
5.954
1.944
14.403
7.935
0.611
0.241
1.326
0.281
0.356
2-18
-------
Leatherman
State/Site
Miles of
Shoreline
Area (million yd3)
Profile Bavside Ooeansjde
Wrightsville Beach
Wilmington Beach
area
long Bay area
Bald Beach/
Island
State Total
South Carolina-The Strand
Myrtle Beach
Magnolia Beach
Pawley's Island
Debidue Beach
Capers Island
Dewees Island
Isle of Palms
Sullivan's Island
Folly Beach
Kiawah Island
Seabrook Island
Edisto Beach
Hunting Island
Fripps Island
Hilton Head
State Total
Georgia-Tybee Island
Sea Island
St. Simons Island
Jekyll Island
State Total
Florida-Amelia Island
Seminole/Manhattan
Beaches
Jacksonville area
Ponte Vedra-Vilano Beach
Anastasia Island
Summer Haven-Beverly
Beach
Flagler Beach
Ormond Beach area
Daytona Beach
Wilbur-by-the-sea
New Symrna Beach
South of Symrna Beach
Cocoa Beach
Eau Gallie Beach
4.36
6.44
25.57
2.46
142.96
0.852
1.133
4.500
0.433
50.358
1.932
20.45
9.19
5.19
3.30
3.66
3.03
2.27
1.02
3.69
5.25
7.42
2.14
5.51
5.70
2.71
12.75
93.28
2.65
3.46
2.54
7.44
16.09
10.83
2.78
5.83
23.90
10.13
14.96
5.26
12.42
8.14
5.25
4.15
17.33
8.86
11.10
3.000
2.007
0.879
0.472
0.979
2.370
2.756
6.380
3.900
7.848
9.364
2.427
4.850
13.601
4.634
39.134
104.601
2.738
3.579
2.620
7.685
16.622
3.178
0.762
1.597
6.076
3.170
3.511
1.440
3.887
2.548
1.640
1.379
5.761
3.120
3.038
0.852
1.007
1.
1.
1.
1.
1.
3.
28.956
7.600
0.581
1.411
0.129
.098
.538
0.524
2.839
.499
.106
5.448
4.134
.636
.096
1.766
9.593
43.998
1.400
1.084
0.759
6.129
9.372
1.271
0.436
2.384
4.674
2.378
3.511
3.630
1.133
6.439
.602
.128
.839
.816
6.413
9.767
3.
1.
1.
3,
2-19
-------
Leatherman
State/Site
South of Melbourne Beach
Vero Beach/Rionar
South of Vero Beach
Hutchinson Island
Jupiter Island
South of Jupiter Island
Palm Beach-Lake Worth
Inlet
Boynton & Delray Beaches
Deerfield & Hillsboro
Beaches
Pompano Beach
Ft. Lauderdale
Hollywood/Golden Beach
Miami Beach area
Virginia Key
Key Biscayne
Atlantic Coast Total
Perdido Key
Santa Rosa Island
Moreno Point
Laguna-Biltraore Beach
Mexico Beach-Beacon Hill
St. Joseph's Spit
Money Beach-Indian
Peninsula
St. George Island
Dog Island
St. Teresa Island-
Lighthouse Point
Honeymoon Island
Clearwater Beach
Bellair Shores-John's
Pass
Treasure Island
Long Key
Cabbage Key
Mullet Key
Anna Maria Key
Longboat Key
Lido Key
Siesta Key
Casey Key
Venice Beach
ManasotalGey
Knight/Don Pedro Islands
island
Miles of
Shoreline
L 28.12
3.96
15.26
13.31
10.10
11.78
15.50
J 14.43
5.36
3.84
5.70
9.53
9.03
1.80
4.03
292.69
5.44
7.84
24.36
26.48
L 4.73
4.32
6.08
10.70
6.86
6.91
0.95
2.82
13.67
3,20
4.13
1.14
4.28
7.14
9.40
2.39
6.08
6.36
3.67
11.29
3 6.32,
5.87
Area
Profile
7.700
1.084
4.179
3.645
2.764
1.843
2.427
1.693
0.838
0.752
1.115
1.490
1.325
1.548
3.471
76.981
0.850
1.533
4.760
5.178
1.300
1.267
2.972
1.674
1.073
1.081
0.518
1.545
7.487
2.191
2.180
0
1.674
5.306
2.206
0.560
1.427
1.369
0.862
2.649
2.474
2.526
(million
Bayside
4.467
1.037
1.490
0.956
0.387
0.567
0.660
0.946
..-.. .-
1.558
5.154
0.589
0.917
22.358
0.252
0.479
— •--
2.378
— —
0.292
0.757
0.896
1.119
0.501
0.188
1.828
1.648
0.584
0.867
• — —
1.299
1.326"
Yd3)
Oceanside
8.867
1.998
1.987
1.687
2.307
3.869
3.306
3.217
1.220
1.353
2.085
0.932
1.670
0.352
1.578
85.219
1.280
1.851
4.079
4.332
1.111
0.845
1.025
2.331
0.871
1.036
0.278
0.742
2.520
0.678
0.727
0.200
0.591
0.802
1.857
0.327
2.707
0.622
2.419
1.787
0.761
0.837
2-20
-------
Leatherman
State/Site
Cayo Costa
North Captiva Island
Captiva Island
Sanibel Island
Fort Meyers
Bonita Beach
Naples Park
Naples
Keewatin Island
Marco Island
Key West
Gulf Coast Total
Alabama-Gulf Shores
Morgan Peninsula
Dauphin Island
Dauphin Island-COHRA
State Total
Mississippi-Pascagoula
Belle Fontaine area
Harrison County
Hancock County
State Total
Louisiana-Bastian Bay complex
Grande Terre Islands
Grande Isle
Caminada Pass-Belle
Pass
Tinibalier Islands
Isles Dernieres
Holly Beach
State Total
Texas-Bolivar Peninsula
Galveston Island
Follets Island
Surfside Beach
Bryan Beach/Brazos Spit
Sargent Beach East
Sargent Beach
Matagorda Peninsula
Matagorda Island
St. Joseph Island
Mustang Island
North Padre Island
North Padre Island-COERA
Area (million yd3)
Miles of
Shoreline
7.12
2.84
4.68
12.16
6.53
5.38
4.47
7.86
5.00
2.46
0.57
251.50
3.86
18.00
6.26
7.95
36.07
2.16
5.49
26
9
42.65
18.92
4.85
7.20
13.28
16.04
17.92
7.20
85.41
18.45
28.39
9.81
4.15
9.94
4.45
2.78
50.68
13.13
18.96
15.55
5.28
7.03
Profile
3.064
1.833
3.019
10.224
5.111
3.997
1.923
4.611
2.542
2.311
0.500
95.797
0.756
3.520
1.224
1.556
7.056
0.633
1.611
7.627
2.640
12.511
18.500
5.215
4.222
9.736
15.685
24.526
9.852
87.736
10.822
11.104
4.413
1.622
2.917
2.437
1.333
16.855
5.903
10.010
7.906
2.273
3.435
Bavside
1.115
0.678
0.497
3.632
0.959
0.585
1.848
23.728
— — -.
— — —
2.380
1.858
3.419
2.600
3.137
3.504
1.407
18.305
21.511
32.150
3.907
1.793
8.519
0.870
0.817
35.860
4.600
3.611
Oceanside
1.754
0.167
0.560
2.428
1.022
0.316
0.313
4.071
0.553
1.011
3.250
52.061
1.058
2.013
0.778
3.849
3.852
1.074
5.519
1.193
11.638
— —
-.-- -
..--.-
5.074
15.693
18.739
12.107
5.267
9.834
2-21
-------
Leatherman
South Padre Island-COBRA 29.41 11.504 18.770
South Padre Island 5.38 1.893 2.789
Brazos Island/Boca Chica 6.40 2.504 1.708
State Total 229.79 96.931 113.638 89.981
Calif ornia-Ooean Beach 3.50 1.028
Santa Barbara 2.20 0.644
Carpinteria 1.36 0.400
Buenaventura State 2.84 0.833
Park
Oxnard Beach 2.27 0.667
Malibu/Carbon Beaches 2.46 0.722
Pacific Palasades 2.08 0.611
Santa Monica/Venice 5.40 1.583
Beach
Doctoweller/Marihattan 6.36 1.867
State Park Beach
Herroosa Beach 1.78 0.522
Redondo Beach 2.01 0.589
Long Beach 4.05 1.189
Sea Beach 0.98 0.289
Sunset Beach 1.95 0.572
Huntington Beach 3.60 1.056
Newport Beach 5.17 1.517
Laguna Beach 1.55 0.456
Capistrano Beach 2.88 0.844
Oceanside 2.94 0.861
Carlsbad 2.70 0.792
Solona Beach/Delmar 3.98 1.167
La Jolla Beach 1.14 0.333
Mission Beach 3.50 1.028
Silver Strand 10.04 2.944
liiperial Beach 1.33 0.389
State Total 78.07 22.903
Oregon-Newport 3.56 1.810
State total 27.56 14.012
Washington-Beach Peninsula 20.50 14.032
State total 48.36 33.101
Hawaii-Waikiki Beach 2.97 1.744
State total 64.24 37.722
2-22
-------
Leatherman
The Pacific or West Coast can be divided into two sections - southern California and the rest. Southern
California, which extends roughly from Santa Barbara to San Diego, probably represents the most modified
coastline in the county (although some could argue the same is true of northern New Jersey). This semi-arid
to arid desert-land has been transformed into one of the largest population centers in the U.S., and the explosive
growth is still occurring. Because of extensive and widespread nourishment projects, the beaches are reportedly
wider today than they were a century ago. This is the only area in the country that has successfully reversed
the long-term trend of shore recession through coastal engineering projects (largely beach nourishment).
Considering the value of this real estate, potential growth factor, and history of coastal projects, these public
recreational beaches will undoubtedly be maintained in the future.
Northern California, Oregon, and Washington are characterized by more cliffed and rocky coastlines.
Sandy beaches occur as small pockets between headlands or as sandy spits. Owing to their general inaccessibility
to large numbers of people, beach nourishment will probably be restricted to projects that incorporate inlet
channel dredging as an important benefit in the total project of sand transportation.
The beaches in Hawaii are world renowned. The famous Waikiki Beach has been nourished for quite
some time. Fortunately, most of the beaches are protected from direct attack of oceanic waves by the offshore
coral reefs, which also serve as the source of the white, coralline sand to the adjacent beaches. As long as the
coral reefs are able to maintain pace with accelerated sea level rise, it is believed that there will be plentiful coral
production to be broken up and moved onshore to naturally nourish the sandy beaches. However, state officials
are not relying upon this assumption, and sand sources from other areas (e.g., countries) are being assessed for
its suitability and compatibility with the native beach sand.
The quantity of sand necessary to hold the shoreline in place was assessed on a state-by-state basis for
various sea level rise scenarios (Table 12). This analysis indicates that Florida would have the greatest demand
for sand not only to nourish the beach, but also to raise the low-lying surface elevations. If the higher scenario
values are realized, then Table 12 indicates that Texas would have the second highest requirement for sand,
followed by South Carolina. In practice, the Texas "requirements" will probably not be met as previously
discussed, while the quantities required in South Carolina must be considered more seriously due to the present
value and continued construction along this coastline.
Nationwide estimates of sand quantities required with accelerated sea-level rise are arrayed in Table 13.
It is apparent that tremendous quantities of good quality sand will be necessary to maintain the nation's major
recreational beaches. Almost all of this sand must be derived from offshore, but to date only enough sand has
been identified to accommodate the two lowest scenarios over the long term. Even in these cases, the offshore
sand is not evenly scattered along the U.S. coastline, so that some areas will run out of local (the least expensive)
sand in a few decades. The costs of sand fill presented in Tables 14 and 15 are based on current expense of
offshore dredging and pumping onshore of locally derived material. Obviously, the costs will increase with
inflation, but more importantly the expense could be greatly underestimated if sand must be acquired from
considerable distance from the beach requiring nourishment.
The ranges of costs are arrayed by state for scenarios I to IV (Table 14). The sand cost not only includes
the quantity required, but also the current statewide cost of such nourishment activities. Since the cost per cubic
yard has been traditionally high in Texas, this state is projected to incur the highest expense. While considerable
quantities are also required for California, the costs are by comparison quite low owing to the local availability
of good sandy material at very reasonable rates. Table 15 summarizes the nationwide costs of sand fill required
with accelerated sea level rise. The costs do not seem too unreasonable for the next several decades considering
past expenditures for shoreline stabilization and the U.S. GNP. However, the costs tend to increase in an
exponential fashion due to the increasing rate of sea level rise through tune.
2-23
-------
Leathennan
Table 12. Quantity of Sand by State and Use for Baseline, Half,
One and Two Meters by Year 2100 (in million yd-5)
State
Maine
New Hampshire
Massachusetts
Rhode Island
Connecticut
New York
New Jersey
Delaware
Maryland
Virginia
North Carolina
South Carolina
Georgia
Florida-Atlantic
Florida-Gulf
Alabama
Mississippi
Louisiana
Texas
California
Oregon
Washington
Hawaii
Baseline (12 on)
Profile Bay Ocean
5.790 — —
2.039 — —
10.667 — —
2.718 — —
16.952 — —
18.295 — —
11.706 6.413 1.954
0.606 — —
0.954 — —
3.794 — —
19.640 — —
40.794 — —
6.482 — —
30.023 — —
37.361 — —
2.752 — —
4.879 — —
391.720 — —
37.803 -~ —
8.932 — —
5.465 — —
12.909 — —
14.712 — —
Total
5.790
2.039
10.667
2.718
16.952
18.295
20.073
0.606
0.954
3.794
19.640
40.794
6.482
30.023
37.361
2.752
4.879
391.720
37.803
8.932
5.465
12.909
14.712
2-24
Half Meter
Profile Bay
29.842 ~
9.725 —
62.358 —
15.334 —
86.065 —
98.044 —
79.541 35.378
3.654 0.532
5.750
25.098 —
93.666 —
257.318 —
38.397 —
154.732 22.582
200.216 25.864
14.747 —
26.148 —
449.208 75.417
260.744 192.048
43.516
15.132 —
35.749 —
67.522 —
(Scenario I)
Ocean Total
— 29.842
— 9.725
— 62.358
— 15.334
— 86.065
— 98.044
— 114.919
— 4.196
— 5.750
— 25.098
— 93.666
— 257.318
— 38.397
— 177.314
— 226.080
— 14.747
— 26.148
— 524.625
— 452.792
— 43.516
— 15.132
— 35.749
— 67.522
-------
Leatherman
Table 12. (continued)
One Meter (Scenario II) Two Meters (Scenario IV)
Profile Bay Ocean Total Profile Bay Ocean Total
Maine
New Hampshire
Massachusetts
Rhode Island
Connecticut
New York
New Jersey
Delaware
Maryland
Virginia
North Carolina
South Carolina
Georgia
Florida-Atlantic
Florida-Gulf
Alabama
Mississippi
Louisiana
Texas
California
Oregon
Washington
Hawaii
54.191 — — 54.191
18.353 — — 18.353
107.212 — — 107.212
26.765 — ' — 26.765
157.351 — — 157.351
174.978 — — 174.978
128.766 70.540 21.491 220.797
6.204 1.178 1.503 8.885
9.764 — 4.083 13.847
41.052 — 7.264 48.316
176.757 4.849 — 181.606
428.864 — 48.398 477.262
65.657 — 8.903 74.560
280.981 59.249 — 340.230
357.323 64.777 — 422.100
26.319 — — 26.319
46.666 — — 46.666
593.095 105.437 — 698.532
419.711 378.415 119.675 917.801
81.077 — — 81.077
38.112 — — 38.112
90.034 — — 90.034
129.386 — — 129.386
103.038
35.505
196.920
49.696
299.922
328.846
227.516 141,
11.320 2,
17.814
72.960
341.931 11.
773.001
120.343
534.248 132.
671.537 142.
49.463
87.702
880.869 165.
737.645 751.
156.427
84.072
198.606
253.492
075 76.312
474 6.433
— 17.651
— 26.793
186 109.743
— 193.151
— 39.737
806 335.763
605 208.765
— 15.434
— 46.668
477 —
147 414.812
103.038
35.505
196.920
49.696
299.922
328.846
444.903
20.227
35.465
99.753
462.860
966.152
160.080
1002.818
1022.907
64.897
134.370
1046.346
1903.604
156.427
84.072
198.606
253.492
2-25
-------
Leatherman
Table 13. Nationwide Estimates of Sand Quantities Required with Sea Level Rise (million yd3)
Year
n
Scenarios
ra rv
VI
2000
2020
2040
2060
2080
2100
145.634 166.770 187.645 208.417
404.697 531.097 654.255
777.742
749.914 1067.874 1394.713 1850.035
229.727
900.743
2272343
250.470
1041.429
2658.815
1155.129 1925.232 2667.664 3390.477 4315.144 5428.242
1772^67 2751.612 4314381 6021.119 7469329 9251.228
2424337 4345.477 6767.643 9070.906 11356.659 13655.708
2-26
-------
Leathennan
Table 14. Range of Cost of Sand Fill for Scenarios I to IV
(50 to 200 cm Sea-Level Rise) for Each State
($ millions)
State
Maine
New Hampshire
Massachusetts
Rhode Island
Oonnecticut
New York
New Jersey
Delaware
Maryland
Virginia
North Carolina
South Carolina
Georgia
2000
7.1-11
2.1-3.6
32-49
5.9-9.2
28-50
48-74
47-64
2.0-2.9
2.4-3.4
15-20
35-60
80-118
11-15
Florida-Atlantic 37-59
Florida-Gulf
Alabama
Mississippi
Louisiana
Texas
California
Oregon
Washington
Hawaii
50-77
3.7-5.6
4.5-6.9
219-250
179-251
10-16
0-4.5
0-11
17-32
2020
21-47
6.5-15
92-187
17-36
89-203
136-298
127-231
5.6-11
6.6-13
40-74
109-261
231-433
29-59
108-243
142-312
10-23
13-28
562-755
493-888
29-70
3.9-29
9.3-68
53-136
2040
39-105
12-35
167-406
24-77
167-454
254-663
226-664
10-24
12-28
72-158
208-596
410-927
53-126
203-542
264-690
17-51
24-62
1038-1621
879-3000
55-157
12-74
28-175
104-313
2-27
2060
62-185
20-63
260-704
49-135
263-806
398-1164
342-1240
16-49
18-49
109-271
331-1090
626-1600
82-220
320-1466
421-1416
30-89
37-109
1526-2628
1318-5863
88-279
24-140
57-331
168-560
2080
88-287
29-99
365-1084
69-209
381-1255
571-1804
644-2305
22-102
26-127
152-522
483-2057
876-2900
116-416
474-3981
646-2643
44-168
53-229
2056-3832
2922-11437
128-434
40-228
95-539
245-877
2100
119-412
39-142
490-1546
92-298
516-1800
770-2581
902-3492
34-162
34-213
201-798
656-3240
1158-4348
153-640
787-7746
904-4092
59-260
72-370
2623-5232
4188-17608
174-626
61-336
143-794
338-1267
-------
Leatherman
Table 15. Nationwide Estimates of Cost of Sand Fill Required with Sea Level Rise ($ millions)
Year
2000
2020
2040
2060
2080
2100
I
837
2,333
4,277
6,564
10,524
14,512
n
958
3,032
6,073
11,419
15,874
26,745
m
1,073
3,722
7,896
15,949
26,528
42,765
IV
1,192
4,418
10,956
20,457
37,525
58,002
V
1310
5,112
13,497
26,510
47,672
71,151
VI
1,428
5,911
15,873
33,885
59,502
88379
2-28
-------
Leatherman
CHAPTER 4
CONCLUSIONS
This study represents the first estimation of sand requirements necessary to stabilize the United States
coastline with accelerated sea-level rise. Both the volume of sand and associated costs to nourish the beach
profile and maintain low-lying surface elevations relative to sea level have been considered. A number of
assumptions have been made to make these calculations so that the numbers will be refined as more data
becomes available.
The cost to stabilize the coast through the "soft" engineering approach of sand filling ranges from
approximately $2.3 billion to $5.9 billion for Scenarios I to VI by the year 2020 on a nationwide basis.
Considering the enamorous value of coastal property (e.g., Miami Beach alone is valued at over $1 billion), it
is safe to assume that the densely developed'areas will be nourished and maintained. What is unclear is at what
point moderate-density areas will be forced by economic considerations to choose another approach (e.g., retreat
from the eroding beach). The next step will be to refine these estimates by completing the analysis on a
community-level basis and then comparing these costs with the value of the affected coastal property.
2-29
-------
Leatherman
REFERENCES
Bruun, P. 1985. Cost-Effective Coastal Protection With Reference to Florida and the Carolinas, US A. Journal
of Coastal Research, v.l, 47-55.
Duane, D.B. and E.P. Meisburger. 1969. Geomorphology and Sediments of the Nearshore Continental Shelf,
Miami to Palm Beach, Florida - UJS. Army, Corps of Engineers, TM No. 29.
Field, M.E. and D.B. Duane. 1974. Geomorphology and Sediments of the Inner Continental Shelf, Cape
Canaveral, Florida. U.S. Army, Corps of Engineers, T.M. - No. 42.
Florida Department of Natural Resources. 1988. Florida's Beach Restoration Management Plan.
Hallermeier, RJ. 1981. A Profile Zonation for Seasonal Sand Beaches from Wave Climate. Coastal
Engineering, v. 4, 253-277.
Hoffman, J.S., J.B. Wells, and J.G. Titus, 1986, Future global warming and sea level rise, in Iceland Coastal and
River Symposium, Reykjavik, Iceland, p. 245-266.
Leatherman, S.P. 1985, Geomorphic effects of accelerated sea-level rise on Ocean City, Maryland. In Potential
Impacts of Sea Level Rise on the Beach at Ocean City, Maryland, J.G. Titus, ed. Washington, D.C.: U.S.
Environmental Protection Agency34 pp.
Leatherman, S.P., 1986, Effect of Accelerated Sea-Level Rise on Coastal Ecosystems, U.S. Senate Publication,
S.Hrg. 99-723, Washington, DC, p. 141-153.
Leatherman, S.P. 1988. Barrier Island Handbook, University of Maryland, College Park, MD, 92 pp.
Lyles, S.D., L.E., Hickman, Jr., and HA. Debaugh, Jr., 1987, Sea level variations for the United States, 1855-
1986, NOAA National Ocean Service Report, Rockville, MD, 182 pp.
Meisburger, E.P. and D.B. Duane. 1971. Geomorphology and Sediments of the Inner Continental Shelf, Palm
Beach to Cape Kennedy, Florida. US. Army Corps of Engineers, T.M. - No. 34.
Meisburger, E.P. and M.E. Field. 1975. Geomorphology, Shallow Structure, Sediments of the Florida Inner
Continental Shelf, Cape Canaveral to Georgia U.S. Army Corps of Engineers, T.M. - No. 54.
National Research Council, 1987, Responding to Changes in Sea Level: Engineering Implications, National
Academy of Sciences, Washington, DC, 148 pp.
Weggel, R. 1987. Personal Communication.
U.S. Army Corps of Engineers, 1971, National Shoreline Study, Washington, DC,59 pp.
2-30
-------
THE COST OF DEFENDING DEVELOPED SHORELINES ALONG SHELTERED WATERS
OF THE UNITED STATES FROM A TWO METER RISE IN MEAN SEA LEVEL
by
J. Richard Weggel
Scott Brown
Juan Carlos Escajadillo
Coastal Engineering Program
and
Patrick Breen
Edward L. Doheny
Engineering Geology Program
Department of Civil & Architectural Engineering
Drexel University
32nd & Chestnut Streets
Philadelphia, PA 19104
Contract No. CR-814577-01-0
-------
CONTENTS
CHAPTER 1: INTRODUCTION 3-1
CHAPTER 2: SEA LEVEL RISE SCENARIO 3-4
CHAPTER 3: INDEX SITE CONCEPT 3-5
TOPOGRAPHIC AND SHORELINE LENGTH ANALYSES 3-5
New York Metropolitan Area, New York and New Jersey 3-5
Long Beach Island Area, New Jersey 3-9
Dividing Creek, New Jersey 3-9
Miami and Miami Beach Area, Florida 3-18
Corpus Christi Area, Texas 3-18
San Francisco Bay Area, California 3-18
CHAPTER 4: LONG BEACH ISLAND, NEW JERSEY - IN-DEPTH ANALYSIS 3-28
EVALUATION OF ALTERNATIVES 3-28
SHORELINE AND TOPOGRAPHIC CONDITIONS 3-28
LEVEL OF DEVELOPMENT 3-28
SEA LEVEL RISE SCENARIO 3-32
RESPONSE TO SEA LEVEL RISE 3-37
Reproduce Landward Migration of Barrier Island 3-37
Raise Island in Place 3-42
Dike Around Island and Provide Interior Drainage 3-42
SUMMARY OF COSTS OF ALTERNATIVES AT LONG BEACH ISLAND 3-55
CHAPTER 5: SUMMARY OF ACTIONS AND THE COST OF RESPONDING TO SEA LEVEL
RISE AT INDEX SITES 3-59
COST OF RESPONDING TO SEA LEVEL RISE AT INDEX SITES 3-60
USGS TOPOGRAPHIC DATA ANALYSIS - INDEX SITES 3-60
CHAPTER 6: EXTRAPOLATION OF COSTS TO INCLUDE THE SHELTERED
SHORELINES OF THE US ; 3-71
REFERENCES 3-90
11
-------
Weggel
CHAPTER 1
INTRODUCTION1
Mean sea level is rising in those regions of the world not previously or recently glaciated, not near present
plate boundaries, and not currently subjected to man-caused subsidence. The rise in sea level is eustatic in
nature and worldwide records suggest an overall rise of about 12 cm over the last 100 years (Fairbridge & Krebs,
1962; Gornitz et al., 1982).
Measurements of sea level rise along the coastal margins of the United States show substantial local
variability, but reasonable averages over the last century suggest a 30-cm rise along the East Coast of the U.S.,
an 11-cm rise along the West Coast, and rises ranging from 20 to 100 cm per century along the Gulf of Mexico.
The above estimates of sea level rise are averages based on analysis of available records gathered at
specific locations such as New York, New York, and Atlantic City, New Jersey, over recent decades as well as
readings made at tide gaging stations currently in operation. The estimates filter out the shorter term (2-7 years)
meteorological fluctuations.
Geological data support the observed indications of recent sea level rise and suggest a fluctuating but
persistent rise over the last 1500 years with even more rapid rates of rise extending back over the last 6000 years
(Fairbridge, 1961).
Beyond past and present rates of rise, there is a growing belief, based on theoretical evaluations of the
potential causes of sea level rise, that a sustained or, more likely, an accelerated rate of sea level rise can be
expected in the future. The cause of the projected rise is global wanning due to the "greenhouse" effect (assisted
by ozone depletion), which leads to melting of Alpine glaciers, ice sheets, and ice caps on a global scale,
accompanied by thermal expansion of the oceans.
Projections of future sea levels vary based on the assumptions used. Estimates of the global rise by the
year 2100 generally are between 50 and 200 cm, although Hoffman (1985) concludes that a 350-cm rise is
possible (Titus, 1986; Revelle, 1983; Meier et al., 1985). Because of the quantitative uncertainties and
probabilities involved in such a prediction, the Committee of Engineering Implications of Changes in Relative
Mean Sea Level (National Research Council, 1987) suggests that three plausible variations in eustatic sea level
rise be assumed for design purposes, all of which project an increasing rate of rise relative to the present and
produce rises of 50,100, and 150 cm by the year 2100.
Even recognizing that tide-gage data are subject to influences that tend to "degrade the data" and that
estimates of the future eustatic sea level are based on estimates of factors such as glacial thickness that are not
well known, the projections of accelerated sea level rise cannot be disregarded. In the United States there is an
enormous and growing investment of population, facilities, and real estate in the zone along the Atlantic, Pacific,
and Gulf of Mexico coastal margins. Moreover, the postulated climatic and oceanic models on which increasing
rates of eustatic sea level rise are based are sufficiently well developed to assume that the physical and financial
risks to coastal communities, facilities, and environments can be realistically estimated, and that many areas are
vulnerable to undesirable changes caused by a rising sea level.
The effects of sea level rise on the coastal zone would include effects on physical processes such as:
changes in weather patterns; higher storm surges; increased storm damage; shoreline erosion; increased flood
'Although the information in this report has been funded wholly or partly by the UJS. Environmental
Protection Agency under Contract No. CR-814577-01-0, it does not necessarily reflect the Agency's views, and
no official endorsement should be inferred from it.
3-1
-------
Weggel
frequencies due to backwater effects; changes in river courses and flow rates; increased sedimentation; shoaling
and scouring in inlets; increased need to dredge waterways; loss of agricultural land; increased landslides;
changing offshore current speeds and directions; land subsidence; higher waves; and more frequent barrier island
washover. There could be major influences on systems, facilities, and infrastructure such as: buried utilities,
subway systems, municipal storm and sanitary sewers, transportation systems (roadways, railways, etc.), water
treatment plants, landfills, drainage patterns, water supply systems, ground water supplies, bridges, and coastal
navigation and harbor improvements. Construction-related activities impacted might include: dredging,
tunnelling, drainage, elevated water tables and dewatering, foundation elevations, tidal energy projects, increased
construction costs, and beach nourishment sand sources. Environmental impacts include: changes in wetland
area boundaries, saltwater intrusion, increased energy consumption for pumping, ecological changes, evaporation
ponds, coastal vegetation, inundation, and loss of habitat. Socioeconomic effects include: the public's response
to rising sea level, flood insurance costs, impact on tax base, sources of compensation for losses, legislation,
government versus personal responsibility for costs, the decision making processes, emergency planning and
responses, identification of future problems, water rights, and the financial resources needed to respond to the
rising sea. Finally, it will be necessary to re-map and survey coastal areas and to obtain new bathymetry for the
near offshore.
Responses and response strategies developed when confronted with sea level rise can include some or all
of the following: constructing dikes, bulkheads, sea walls, revetments, groins, beach nourishment, construction
of offshore breakwaters, storm surge barriers, dikes and polders. All of these represent an effort to resist the
landward advance of the sea. In addition, there is always the alternative of abandoning some areas and
structures or relocating structures on higher ground.
In the present study, the effects of sea level rise at six index sites in the United States were evaluated. The
index sites were: the metropolitan New York City area; the Long Beach Island, New Jersey, area; the Corpus
Christi, Texas, area; the Dividing Creek, New Jersey, area (a sparsely developed area); the south San Francisco
Bay, California, area; and the Miami/Miami Beach, Florida, area. Using shoreline length, ground elevation
distributions, and degree of development data, the six sites were evaluated in terms of the type of response
judged to be most appropriate. The cost of responding to sea level rise was determined based on certain
consistent assumptions. The response strategies can be categorized as follows:
1. Abandon those low-lying areas with little development and having a limited economic base.
2. Raise dwellings and other structures where appropriate.
3. Move isolated, structurally sound dwellings and other structures to new locations at higher elevation.
4. Surround low-lying economically developed areas with a dike; install a seawater seepage control
system, an interior drainage system, and storage and pumping facilities to maintain drainage.
Costs for each of the above alternatives were developed by establishing unit costs for each element of a
response strategy and then applying those costs to an assumption as to how each individual index site might
respond to sea level rise. For example, in sparsely developed areas, isolated structures identified on United
States Geological Survey (USGS) quadrangle maps were counted, and the cost of moving some fraction of them
was estimated. (Because much of the data on building locations on the USGS quads is outdated, the number
of buildings obtained from the quads was interpreted as only an indicator of the number of buildings present.)
The cost to construct a dike around isolated communities and to provide an interior drainage system was
estimated for areas with a reasonable level of economic development. The cost of raising and/or replacing
roads that would connect such isolated areas with high land was also estimated. The cost of raising existing
bulkheads and the cost of providing new bulkheads was estimated for those areas where such a response was
believed appropriate. (The total cost of raising bulkheads cannot be allocated against sea level rise, since most
bulkheads have a relatively short lifetime and need to be replaced periodically. Thus an accelerated sea level rise
will result in a shorter lifetime and more frequent replacement. Such costs can be significantly reduced if sea
level rise is anticipated in the initial design.)
3-2
-------
Weggel
The cost of responding to sea level rise for the total U.S. shoreline was determined by extrapolating costs
for the six index sites to the rest of the ILS. coast. Costs at the six index sites were correlated with topographic
and economic development factors, and then digitized topographic data obtained from USGS quad sheets were
used to extrapolate costs to 78 additional sites around the U.S. shoreline. The extrapolation techniques make
use of information on shoreline lengths, the distribution of land elevations, and the level of development.
3-3
-------
Weggel
CHAPTER 2
SEA LEVEL RISE SCENARIO
The sea level rise scenario assumed for the present study is summarized by Titus (this volume). The
assumption is that by the year 2100, sea level will be 2 meters (6.5 feet) above the level it would have attained
had it continued to rise at its local historical rate. Thus, the historical rate is extrapolated to the year 2100, and
an additional 2 meters are added. The historical rate of rise is assumed to continue linearly, while the
superposed 2-meter rise is assumed to increase parabolically. The general equation is given by,
RSIXt) = L(t) + 0.0012 t + 0.0001434 t2 (1)
in which RSL(t) is the relative sea level, L(t) is a local sea level factor that includes the deviation of the local
historical rate from the global rate (a local rate of rise based on historical tide gage records), and it is the
number of years beyond the 1986 base year. For example, the rate of sea level rise in the Long Beach Island,
New Jersey, area has been about 40 cm/100 years or 0.004 m/yr (Lyles et al., 1987); thus, the function L(t) +
0.0012 t = 0.004 t and equation (1) for Long Beach Island becomes,
RSL(t) = 0.004 t + 0.0001434 t2 (2)
Some of the analyses hi the present study are not sensitive enough for the precise rate of sea level rise to
become a factor in determining costs or for determining when certain actions in response to sea level rise will
be triggered; however, for some detailed analyses the rates do enter into the cost calculations. For example,
several alternatives were evaluated for Long Beach Island where the timing of certain actions depends on the
stage of local sea level. On the other hand, several of the analyses simply assume that actions will be taken in
response to sea level rise but the time when those actions are taken is not specified.
3-4
-------
Weggel
CHAPTERS
INDEX SITE CONCEPT
To determine the cost of constructing dikes and other facilities to protect areas from inundation and from
storm surge and wave damage, six index sites were selected for detailed study and used to develop correlations
between data from those sites and data from other coastal sites around the continental U.S. The six sites were
chosen to be a representative sample of the various types of coastal topography and development present in the
U.S. with emphasis on developed areas where it is certain that some action to counter sea level rise will have
to be taken. The sites were selected from among those being analyzed by Park et al. (this volume). The sites
included: a) sections of New York City and its environs; b) Long Beach Island, New Jersey; c) Dividing Creek,
New Jersey; d) Miami and Miami Beach, Florida; e) Corpus Christi, Texas; and 0 portions of California's south
San Francisco Bay area. The boundaries of the specific areas studied were selected to correspond with areas
covered by USGS quad sheets. The sites, the USGS quads analyzed, and their dates of latest revision are given
in Table 1. In addition, a portion of one of these index sites, Long Beach Island, New Jersey, was singled out
for a more in-depth analysis.
Three scenarios of a coastal community's response to sea level rise were evaluated for Long Beach Island.
They included: a) moving the island landward by reclaiming land on the bay side of the barrier while allowing
the ocean side to recede in response to erosion caused by higher sea levels; b) artificially raising the island's
elevation in conjunction with moving the island landward; and c) constructing dikes around the island and
installing an interior drainage system to handle both storm drainage and seawater seepage beneath the dike
system. Under the first two scenarios, houses would be moved to newly reclaimed land or raised in response
to raising the island. In the third scenario, the island would have a protective dike built around it and the houses
would be left in place.
TOPOGRAPHIC AND SHORELINE LENGTH ANALYSES
Topography at the six index sites was analyzed using the USGS quad sheets to obtain basic ground
elevation and shoreline length data. Specifically, the shoreline length was measured using a rolling map-measure
while areas enclosed within various elevation contours were planimetered. As a result, the total shoreline length
was determined for each quad sheet along with that portion of the shoreline that is presently protected by
bulkheads. The area between various contours on the quad sheets was planimetered and the topographic
characteristics determined by plotting the distribution of ground elevations. For example, the distribution of
elevations for the New York City metropolitan area is given in Figure 1. In general, each site has a characteristic
topographic distribution that determines its vulnerability to inundation by a rise in sea level.
New York Metropolitan Area. New York and New Jersey
The New York City metropolitan area is an intensively developed urban area characterized by heavily
populated residential areas as well as large-scale commercial and industrial development. It is perhaps the most
intensely developed metropolitan area of the United States. It also has a long, heavily developed shoreline, most
of which is already protected by structures such as bulkheads or revetments. Major metropolitan subdivisions
include: Manhattan, Brooklyn, Queens, The Bronx, and Staten Island, New York; and Elizabeth, Jersey City,
Union City, and Linden, New Jersey. In general, 26% of the land on the quads lies below the +5 foot contour
while 52% lies below the +10 foot National Geodetic Vertical Datum (NGVD) elevation. (NGVD is sometimes
referred to as the mean sea level datum of 1929.) On the other hand, there are few undisturbed wetlands on
the quads (except for the wetlands along the west bank of the Hackensack River), so that erosion will be more
of a concern than simple inundation. The distribution of land elevations planimetered from the quads is
summarized in Table 2 and shown in Figure 1. (The contour interval of the seven USGS quads that cover the
study area is 10 feet so that the distribution of land elevations below the 10-foot contour cannot be determined
3-5
-------
Weggel
TABLE 1 INDEX SITES AND CORRESPONDING USGS QUAD SHEETS
Nominal
Index Site
USGS Quad Sheets
Original Photo-
Date Revision
New York, NY
Weehawken, NJ
Central Park, NY
Elizabeth, NJ
Jersey City, NJ
Brooklyn, NY
Arthur Kill, NJ
The Narrows, NY
* Long Beach Island, NJ
Barnegat Light, NJ
Long Beach NE, NJ
Ship Bottom, NJ
Tuckerton, NJ
Beach Haven, NJ
Dividing Creek, NJ
Dividing Creek, NJ
Cedarville, NJ
Port Norris, NJ
Fortescue, NJ
Miami, FL
North Miami, FL
Miami, FL
Corpus Christi, TX
Oso Creek NE, TX
Oso Creek NW , TX
Port Ingleside, TX
Crane Islands NW, TX
Portland, TX
Corpus Christi, TX
1967
1966
1967
1967
1967
1966
1966
1953
1951
1952
1952
1951
1956
1956
1956
1956
1962
1962
1968
1968
1968
1968
1968
1968
1981
1979
1981
1981
1979
1981
1981
1972
1972-77
1972
1972
1972-77
1972
1972-82
1972
1972
1969-72
1969
1975
1975
1975
1975
-
1975
San Francisco, CA
Palo Alto, CA 1961 1968-73
Mountain View, CA 1961 1981
Redwood Point, CA 1959 1980
Newark, CA 1959 1980
* Index site selected for more detailed analysis of alternatives
3-6
-------
Figure 1 Distribution of Ground Elevations - New York City
Metropolitan Area.
-------
Weggel
TABLE 2
SUMMARY OF TOPOGRAPHIC CONDITIONS AT NEW YORK, NEW YORK
USGS Quad Area Csq mi) Between Given Elevations (ft) Total Wetlands
0
-------
Weggel
from the quads. The digitized topographic data allowed a better resolution of elevations than the present
analysis.) The USGS quads that cover the present study area include: Weehawken, Arthur Kill, Brooklyn, The
Narrows, Jersey City, Elizabeth, and Central Park. Major bodies of water include the Hudson, Hackensack,
East, Arthur Kill, and Kill van Kull Rivers; Newark Bay, the Verazzano Narrows, Upper Bay, Lower Bay, and
the Raritan Bay. The shoreline lengths are given in Table 3. Table 3a is a partial summary of shoreline lengths
by region. Table 3a does not include the total shoreline shown on all seven quads. Of the 220 miles of shoreline
included in the study area, 155 miles (70%) are bulkheaded.
Long Beach Island Area. New Jersey
Long Beach Island, New Jersey, is a sandy barrier island near the center of New Jersey's Atlantic Ocean
shoreline. The island is located approximately 45 miles south of Sandy Hook and 55 miles north of Cape May
and is bounded by Barnegat Inlet on the north and Beach Haven and Little Egg Harbor Inlets on the south. The
island is about 23 miles long. The mainland behind the barrier island is bordered by extensive wetlands. There
are also numerous wetlands islands in the bays behind the barrier island. The five USGS quads that cover the
study area are: Barnegat Light, Long Beach NE, Ship Bottom, Tuckerton, and Beach Haven. The distribution
of elevations on the five USGS quads is given in Table 4 and in Figure 2. Generally, the barrier island is at
about the -1-3.5-foot NGVD elevation with only a few scattered areas of high dunes where the elevation exceeds
+ 10 feet. (The demarcation between wetlands and fast land on the quads was assumed to be the + 3.5-foot
NGVD contour.) Large areas of the mainland are below the +3.5-foot NGVD elevation, however. The overall
topographic summary indicates that almost 50% of the land area is below +3.5 feet NGVD; about 80% is below
+10 feet, and 93% is below 20 feet.
Shoreline lengths are summarized in Table 5. Long Beach Island's shoreline is about 55 miles long. About
23 miles front the Atlantic Ocean while the remainder is bay shoreline. The shoreline defined by the interface
between the mainland and wetlands is about 33 miles long, while the outer wetlands shoreline is about 55 miles
long.
Long Beach Island was selected for detailed analysis to investigate the cost of several alternative responses
to sea level rise. This analysis is presented in a later section. A more detailed description of Long Beach Island
is also presented there.
Dividing Creek. New Jersey
The Dividing Creek, New Jersey, area is a sparsely developed area bordering the Delaware Bay near where
the Maurice River discharges into the bay. The area considered in the present study is covered by four USGS
Quads and is mostly composed of wetlands having an elevation below +5 feet NGVD; much of it is state-owned
hunting lands. The USGS quads covering the area are: Dividing Creek, Cedarville, Port Norris, and Fortescue.
The region is criss-crossed by small streams and drainage channels. There are several small communities in the
area. They include the bayside towns of Fortescue and Gandys Beach, the town of Dividing Creek on the fast
land behind the wetlands, and the towns of Laurel Lake and Cedarville on the Maurice River and Cedar Creek,
respectively. Laurel Lake and Cedarville are mostly above +10 feet NGVD, while significant portions of the
remaining towns are below +10 feet elevation. The distribution of land elevations obtained by planimetering
the USGS quads is summarized in Table 6 and shown in Figure 3. About half of the land area is below the +5-
foot contour (defined here as wetlands area), while fully 92% of the land area is below the 20-foot contour. The
shoreline lengths on the four quads are summarized in Table 7. The shoreline length is about 97 miles and is
defined here as the interface between the wetlands and Delaware Bay. Dividing Creek is believed to be typical
of much of the undeveloped U.S. coastlines such as areas on the mainland behind barrier islands. For example,
the mainland areas in North Carolina in sounds and bays behind the Outer Banks might be characterized by the
level of development and wetlands of the Dividing Creek area.
3-9
-------
Weggel
TABLE 3 SUMMARY OF SHORELINE LENGTHS - NEW YORK, NEW YORK
USGS Quad
Shoreline Length
Bulk. Unbulk. Total
(mi) (mi) (mi)
Wetlands
Shoreline
(mi)
Weehawken
Central Park
Jersey City
Brooklyn
MANHATTAN ISLAND
0.89 0.15 1.04
13.79
4.32
4.18
5.59
0.15
0
19.39
4.47
4.18
SUB TOTAL
Jersey City
Elizabeth
Arthur Kill
The Narrows
23.19
8.13
4.18
4.03
1.49
5.89
29.08
STATEN ISLAND
0.52 8.65
1.57 5.74
17.89 21.92
6.26 7.75
2.09 *
SUB TOTAL
Elizabeth
17.82
6.93
26.25
44.07
2.09 *
ELIZABETH
7.46 14.39
SUB TOTAL
6.93
7.46
14.39
Weehawken
Jersey City
Elizabeth
JERSEY - UNION CITY
20.43 0 20.43
27.89 *
2.23
10.89
1.49
1.49
3.72
12.38
SUB TOTAL 33.55 2.98 36.53 27.89 *
Wetlands shoreline length not included in total shoreline length,
3-10
-------
Weggel
TABLE 3 (cont.) SUMMARY OF SHORELINE LENGTHS - NEW YORK, NEW YORK
USGS Quad
Weehawken
Brooklyn
Jersey City
The Narrows
SUB TOTAL
TOTAL
TABLE
Region
Jersey City
Brooklyn
Bronx
Queens
Elizabeth
Governors Is.
TOTAL
Shoreline Length
Bulk. Unbulk.
(mi) (mi)
BRONX
16.40 1
10.73
43.11 9
4.10 1
74.34 22
155.83 64
3A SUMMARY
- QUEENS
1.93
0
.09
.27
.29
.87
Wet lands
Total Shoreline
(mi) (mi)
- BROOKLYN
28.33
10.73
52.20
5.37
96.63
220.70 29.98 *
OF SHORELINE LENGTHS BY REGION
Shoreline Length Wetlands
Bulk. Unbulk. Total Shoreline
(mi) (mi) (mi) (mi)
45.56
36.98
6.70
4.76
22.89
2.18
119.05
17.15
1.71
4.29
2.98
3.88
0
30.01
62.71 27.89 *
38.70
11.63
7.75
26.77
2.16
149.06 27.89 *
* Wetlands shoreline length not included in total shoreline length.
3-11
-------
Weggel
TABLE 4
SUMMARY OF TOPOGRAPHIC CONDITIONS ON LONG BEACH ISLAND AND
ADJACENT MAINLAND, NEW JERSEY
USGS Quad
Area (sq mi) Between >}iven Elevations (ft)
0
-------
V
c
% of Area Below Given ElevA
Figure 2 Distribution of Ground Elevations - Long Beach
Island, NJ Area.
-------
Weggel
TABLE 5 SUMMARY OF SHORELINE LENGTHS AT LONG BEACH ISLAND AND ADJACENT
MAINLAND, NEW JERSEY
USGS Quad LBI Ocean LBI Bay Shoreline Mainland Wetlands
Shoreline Bulk. Unbulk. Total Shoreline Shoreline
(mi) (mi) (mi) (mi) (mi) (mi)
Barnegat
Light
Long Beach
NE
Ship
Bottom
Tuckerton
Beach
Haven
TOTAL (mi)
1
2
9
3
6
22
.19
.24
.39
.87
.26
.95
——————
0
0
6
2
5
15
.74
.30
.41
.61
.02
.08
1
1
6
4
2
17
.86
.79
.26
.69
.68
.28
2.
2.
12.
7.
7.
32.
60
09
67
30
70
36
0
0
18.24*
14.60*
0
32.84
4.
0.
26.
18.
4.
54.
32
60
60
17
79
48
* Boundary between wetlands and mainland.
3-14
-------
Weggel
TABLE 6 SUMMARY OF TOPOGRAPHIC CONDITIONS IN DIVIDING CREEK, NEW
JERSEY
USGS Quad Area (sq mi) Between Given Elevations (ft) Total Wetlands
0
-------
0
Figure 3 Distribution of Ground Elevations - Dividing Creek,
NJ Area.
-------
Weggel
TABLE 7 SUMMARY OF SHORELINE LENGTHS AT DIVIDING CREEK, NEW JERSEY
IJSG5 Quad Delaware Bay *
Shoreline
(ni)
Dividing Creek 17.88
Cedarville 51.07
Port Morris 19.22
Fortescue 8.79
TOTAL (mi) 96.96
* Boundary between wetlands and Delaware Bay
3-17
-------
Weggel
Miami and Miami Beach Area. Florida
The Miami/Miami Beach, Florida, area is covered by two USGS quads: Miami and North Miami. The
area is heavily developed, both commercially and residentially. Miami Beach is primarily a resort area while
Miami itself is a major metropolitan area Major municipalities include the City of Miami itself, North Miami,
Miami Shores, and North Miami Beach on the mainland, and Miami Beach on the barrier island. There are
also many smaller political subdivisions. The land is generally low in elevation as shown in Table 8. The
distribution of land elevations is shown in Figure 4 and summarized in Table 8. While only 16% of the land is
below + 5-foot NGVD elevation, 52% is between +5 and +10 feet so that 69% is below +10 feet and 98% is
below +15 feet in elevation. There is virtually no land above the + 20-foot elevation. In addition, most of the
low-lying land is heavily developed and very little of it is undeveloped wetlands. As a result of the low-lying
topography, most of the shoreline has been bulkheaded. Shoreline lengths are summarized in Table 9. Of the
125 miles of shoreline on the USGS quads, more than 100 miles or 80% are bulkheaded. The shorelines in
Table 9 are classified as either ocean or bay shoreline. The major bodies of water are the Atlantic Ocean
fronting the barrier island, and Miami Beach and Biscayne Bay, the bay between the Miami Beach barrier
island and the mainland. Numerous manmade or man-improved islands are located in Biscayne Bay. Many of
them support residential development or they serve to support causeways that connect the barrier island with
the mainland. Most of these small islands' shorelines are completely bulkheaded.
Corpus Christi Area. Texas
Corpus Christi, Texas, is located on the Gulf of Mexico about 150 miles north of the UJS.-Mexican border.
The area is developed around the major city of Corpus Christi and several smaller municipalities such as
Portland and Ingleside. The towns are located on Corpus Christi Bay and are sheltered from the Gulf of Mexico
by Mustang Island and Padre Island, two undeveloped barrier islands. Portions of the undeveloped barrier
islands are included in the present analysis. The area under study is covered by six USGS quads: Corpus Christi,
Crane Island NW, Oso Creek NE, Oso Creek NW, Port Ingleside, and Portland. The distribution of land
elevations in the Corpus Christi area is given in Figure 5 and summarized in Table 10. In general, little land is
below the + 5-foot contour (only 9%), while all of it is below the 30-foot contour. The land elevation distribution
at Corpus Christi is somewhat unique among the index sites, since the function is approximately linear
suggesting a very steep shoreline in an average sense.
The shoreline length and its distribution among the six quads is given in Table 11. Most of the 189-mile-
long shoreline is unbulkheaded.
San Francisco Bay Area California
The portion of the south San Francisco Bay Area considered in the present study is covered by the
following four USGS quads: Redwood Point, Newark, Palo Alto, and Mountain View. The area covered is the
shallow, southernmost portion of San Francisco Bay. The major metropolitan areas in the area are Hayward,
Newark, and Fremont on the east side of the bay, and Palo Alto, Redwood City, Sunnyvale, Mountain View, and
Menlo Park on the west side of the bay. Most of the residential areas associated with these towns are at
sufficiently high elevation to not be significantly affected by sea level rise of the magnitude under consideration
here; however, the low-lying areas surrounding the bay itself are vulnerable. Generally, terrain in the San
Francisco area is quite hilly except for the low-lying areas adjacent to the San Francisco Bay. Most of the bay
shoreline in this area is covered by salt evaporators: portions of the bay separated from the main bay by levees
and used to commercially extract salt from bay water through natural evaporation. The bay is thus subdivided
by the levees. Redwood Creek and Steinberger Slough drain into the bay in this area. The distribution of
elevations is given in Figure 6 and summarized in Table 12 for that portion of the area below the + 30-foot
contour. Because of the evaporators, 59% of the land lower than +30 feet is below +5 feet NGVD elevation.
Shoreline lengths are summarized in Table 13. Two lengths are given in the table: the length along the
outermost levees that separate the evaporators from the bay, and the shoreline length behind the evaporators,
i.e., the shoreline between the evaporators and fast land. This latter shoreline nearly coincides with the +5-
foot contour.
3-18
-------
Weggel
TABLE 8 SUMMARY OF TOPGRAPHIC CONDITIONS AT MIAMI AND MIAMI BEACH,
FLORIDA
USGS Quad
North Miami
Miami
TOTAL
% OF TOTAL
CUM %
North Miami
Miami
TOTAL
% OF TOTAL
CUM %
TOTAL
% OF TOTAL
CUM %
Area i
0 ' Z < 5
1.72
4.38
6.10
60.2
60.2
12.65
3.85
16.50
16.7
16.7
22.60
20.7
20.7
'sq mi) Between Given
5
-------
Figure 4 Distribution of Ground Elevations - Miami/Miami
Beach, FL Area.
-------
Weggel
TABLE 9 SUMMARY OF SHORELINE LENGTHS AT MIAMI BEACH, FLORIDA
USGS Quad
North
Miami
Miami
TOTAL (mi)
Ocean Bay Shoreline * Mainland Shoreline
Shoreline Bulk. Unbulk. Total Bulk. Unbulk. Total
(mi) (mi) (mi) (mi) (mi) (mi) (mi)
8
7
16
.49
.75
.24
15.93 2.98 18.91 6.44 3.84 10.28
88.35 16.08 84.93 10.21 0.82 11.03
34.78 19.06 103.83 16.65 4.66 21.31
* Includes shoreline of islands in bay between Miami and Miami Beach
3-21
-------
Figure 5 Distribution of Ground Elevations - Corpus Christi,
TX Area.
-------
TABL£ iu ooil.lArt/ ur iUrut^Ar'HlvJ OuNDiriONS AT CuixFUS
USGS Quad Area (sq mi) Between Given Elevations (ft) Total
0
-------
Weggel
TABLE 11 SUMMARY OF SHORELINE LENGTHS - CORPUS CHRIST AREA, TEXAS
USGS Quad Shoreline Length
Bulk. Unbulk. Total
(mi) (mi) (mi)
Corpus Christi 9.7
Crane Island NW 1.7
Oso
Oso
Port
Port
Creek NE 7.3
Creek NW
Ingleside 2.6
land 1.7
TOTALS 23.5
41
42
36
6
29
9
165
.3
.4
.2
.0
.8
.8
.5
51
44
44
6
32
11
189
.0
.1
.0
.0 *
.4
.5
.0
Along Oso Creek
3-24
-------
Area &elow 6wen
Figure 6 Distribution of Ground Elevations - South San
Francisco Bay, CA Area.
-------
Weggel
TABLB 12 SUMMARY OF TOPOGRAPHIC CONDITIONS IN SAN FRANCISCO AREA,
CALIFORNIA
USGS Quad Area (sq mi) Between Given Elevations (ft) Total Wetlands
0
-------
Weggel
TABLE 13 SUMMARY OF SHORELINE LENGTHS - SAN FRANCISCO AREA, CALIFORNIA
USGS Quad Inner Shoreline Outer Shoreline
Length * Length **
(mi) (mi)
Redwood Point
Newark
Palo Alto
Mountain View
TOTAL
0.89
14.76
8.35
13.00
37.00
18 19
6.26
0.89
17.00
42.35
* Shoreline between salt evaporators and mainland (approximately
coincides with 5 foot contour.
** Shoreline between evaporators and San Francisco Bay.
3-27
-------
Weggel
CHAPTER 4
LONG BEACH ISLAND, NEW JERSEY - IN-DEPTH ANALYSIS
EVALUATION OF ALTERNATIVES
Long Beach Island, New Jersey, is a barrier island approximately 23 miles long, averaging between 1000
and 3200 feet wide. It comprises an area of about 7.4 square miles. It is bounded on the north by Barnegat
Inlet and on the south by Beach Haven and Little Egg Inlets. The island shelters Little Egg Harbor and
Manahawkin and Barnegat Bays from the Atlantic Ocean. The island is entirely developed with single-family
houses except for the southern 3 miles, which is part of Brigantine National Wildlife Refuge. There are no high-
rise condominiums and only one motel with as many as three stories. It is heavily populated by vacationers
during the summer months but the population during the remainder of the year is relatively small. Access to
the island is by a single bridge from Manahawkin on the mainland to the town of Ship Bottom near the middle
of the island. Long Beach Island's ocean shoreline is protected by dunes along most of its entire length. The
dune crests are at about +10 feet above NGVD with a few rising to +20 feet. The island's bay shoreline is
about 32 miles long; about 15 mUes of the bay shoreline is bulkheaded. See Table 5. The bay shoreline is dotted
with small marinas and other boat launching and docking facilities. There is also a small amount of salt marsh
along the bay shoreline comprising about 0.43 square miles or 5.7% of the island's area.
SHORELINE AND TOPOGRAPHIC CONDITIONS
Shoreline lengths along Long Beach Island and the adjacent mainland are summarized in Table 5, and
topographic conditions are summarized in Table 4. These data are given for each of the USGS quad sheets
analyzed. Long Beach Island is covered by five quads: Barnegat Light, Long Beach NE, Ship Bottom, Tuckerton,
and Beach Haven. The Barnegat Light quad also covers the southern end of Island Beach State Park, the
undeveloped barrier island just north of Long Beach Island, while the Ship Bottom and Tuckerton quads also
cover a significant area of the mainland behind Long Beach Island. The distribution of elevations on Long Beach
Island is given in Figure 7. About 6% of the island is below about 35 feet in elevation (NGVD); about 84% of
the island is between 3.5 feet and 10 feet in elevation. Only 9% is above 10 feet, and less than 1% (a few
scattered high dunes near the northern end of the island) is above 20 feet in elevation. The distribution of
elevations on the mainland is significantly different. (See Figure 7 for the distribution of elevations on both the
mainland and on the barrier island.) Of the mainland area below 40 feet in elevation, about 57% is below + 3.5
feet NGVD. This area is defined as coastal wetlands on the USGS quads.
LEVEL OF DEVELOPMENT
The level of development on Long Beach Island was determined from an analysis of aerial photographs
obtained from the New Jersey Department of Environmental Protection, Bureau of Coastal Engineering. Two
sets of photographs, dated March 23,1982, and March 30,1984, were used to determine the number of buildings
on the island and their location relative to the prevailing high-water shoreline. These data are summarized in
Table 14. The distribution of houses with respect to the shoreline is plotted in Figure 8. The cumulative number
of houses summed along the entire length of the island is plotted as a function of the distance measured
landward across the island from the high water shoreline. Thus if the shoreline were to erode 500 feet from the
ocean side, approximately 4,700 buildings on Long Beach Island would be affected. If the shoreline were to
erode 1,000 feet, 9,800 buildings would be located seaward of the new shoreline. This relationship is nearly
linear, indicating an almost uniform level of development across the island from its seaward side to its bayward
side. It deviates from a linear relationship only for distances exceeding 1,000 feet because of the island's variable
width. Thus the cumulative number of buildings drops off for distances greater than 1,000 feet with that distance
measured from the existing high-water shoreline. The linear portion of the curve can be expressed by the
equation:
3-28
-------
% of Area b£low Swen Blev.
Figure 7 Distribution of Ground Elevation - Long Beach Island
and Mainland and Adjacent to Long Beach Island.
I
-------
Weggel
Figure 8 Proximity of Buildings to High Water Shoreline - Long
Beach Island (Based on Analysis of 1984 Aerial
Photographs).
3-30
-------
Weggel
TABLE 14 DISTRIBUTION OF BUILDINGS ON LONG BEACH ISLAND
Distance from
Ocean Shoreline
(ft)
0
325
650
980
1300
1630
to
to
to
to
to
to
325
650
980
1300
1630
1960
Number of
Buildings
2834
3464-
3110
2468
1756
789
Cumulative
Number
2834
6298
9408
11876
13632
14421
19
43
65
82
94
100
%
.65
.67
.24
.35
.52
.00
3-31
-------
Weggel
N = 10.4 X - 600 (3)
in which N = the number of buildings affected and X = the distance measured landward from the present
high-water shoreline in feet. Equation 3 is valid for values of X between 60 feet and 1000 feet.
Development density varies along the island as shown in Figure 9. In general, building density is lower
in the more northerly communities where lot coverage is lower. The area per building, averaged over the entire
island, is 12,600 square feet per building or a density of 2,200 buildings per square mile. In the northerly
communities of Barnegat Light and Loveladies, about 50% of the homes are of recent construction and are
elevated on pilings one floor above ground level.
SEA LEVEL RISE SCENARIO
By 2100, for purposes of this study, mean sea level at Long Beach Island was assumed to be 2 meters
above 1986 levels plus that portion of the historically observed sea level rise that exceeds 0.0012 meters per year,
i.e., the increase in mean sea level that would occur if the observed historical rate were to prevail (Titus, this
volume). At Long Beach Island, the historical rate of sea level rise has been about 0.004 meters per year. Thus,
mean sea level is given by the equation,
Z = 0.004(YR-1986) + 0.0001424(YR-1986)2 (4)
where Z is the elevation of mean sea level above the 1986 level (meters) and YR is the year. Thus by 2100,
mean sea level at Long Beach Island will be 231 meters above the 1986 level. The first term of Equation 4
gives the historical rate, while the second term gives the rate due to the accelerated "greenhouse effect."
Equation 4 is plotted in Figure 10 along with sea level curves for other historical rates, i.e., for various values
of the coefficient of the first term in equation 4.
A rough estimate of the average rate of shoreline recession for Long Beach Island can be derived from
the rate of sea level rise by using the Bruun rule (Bruun, 1962). For the Atlantic coast of the U.S., the profile
closure depth is approximately 30 feet. At Long Beach Island, the 30-foot contour is approximately 4,000 feet
from shore on average and the dune crests can be assumed to be about 10 feet high. Thus, the profile extending
from dune crest to closure depth is about 40 feet high, and a one foot change in mean sea level will result in
4000/40 = 100 feet of shoreline recession. Combining this rate of erosion with the mean sea level curves results
in the shoreline recession curve in Figure 11. This curve, given by,
X = 328.1{0.004(YR-1986) + 0.0001434(YR-1986)2} (5)
in which X is the shoreline recession in feet. It is simply 100 times the sea level rise curve given by Equation 4.
Coupled with the building density data of Figure 8, the number of houses affected by the rise in sea level
can be determined from the shoreline recession curve. Figure 12 shows the number of buildings on Long Beach
Island affected by sea level rise as a function of time. The cumulative number of buildings is given by,
N = 13.65(YR-1986) + 0.4893(YR-1986)2 - 600 (6)
for the years after 2010.
3-32
-------
Figure 9 Distribution of Building Density a lung Long Beach
Island (Based on Analysis of 1984 Aerial Photographs).
-------
50
-5ft
wo
2000 £020 2040
1080 2100 2120 2140 Z160
Figure 10 Sea Level as a Function of Time - Long Beach Island.
-------
J-r!.im±
n;:;~u
> 11, i .-i
JKU.I;
NW
mo
2000 2020 2040 2060 W
1\W
2f40
Figure 11 Shoreline Recession as a Function of Time - Long Beach
Island.
-------
I i
*«w
2000 2020 2044 2060 £060 2100 21Z0 2140 2160
Figure 12 Number of Houses per Year and Cumulative Number of
Houses Affected by Sea Level Rise as a Function of
Time - Long Beach Island.
-------
Weggel
Because of the setback of existing buildings behind the dunes, the analysis shows that the buildings closest
to the beach are not affected until 2010. Actually, they will feel the effect of sea level rise sometime earlier
because of their vulnerability to storm damage as the dune buffer between them and the sea disappears; however,
based on the given sea level rise scenario, the buildings will remain landward of the shoreline until 2010. The
lower curve of Figure 12, which shows the cumulative number of buildings affected by erosion, should probably
be shifted leftward to account for owners abandoning or moving buildings that experience storm damage more
and more frequently as the shoreline recedes. The annual rate at which buildings are affected by shoreline
recession can be found from the slope of the cumulative number of buildings affected. The rate is given by,
R = 13.65 + 0.9786(YR-1986) (7)
in which R = the number of buildings affected per year. Since the cumulative number of buildings varies
parabolically, the annual rate at which buildings are affected varies linearly with time. Approximately 37
buildings per year will be affected in 2010. By 2100, however, 125 buildings per year will come within reach of
the encroaching shoreline.
RESPONSE TO SEA LEVEL RISE
Reproduce Landward Migration of Barrier Island
One possible response to a rising sea level at Long Beach Island would be to physically introduce a
landward migration of the barrier island. As sea level rises and erodes the island's ocean beaches, additional
land would be created on the bay side of the island by bulkheading and filling. As buildings are affected by the
encroaching ocean shoreline, they would be moved to new land created on the bayward side of Long Beach
Island. Initially, fill to create the new land would be obtained from the bay adjacent to the island, possibly by
deepening, expanding, or moving the Atlantic Intracoastal Waterway. In later stages, however, the cumulative
quantity of fill needed would, like beach nourishment, require the exploitation of offshore sand resources. (See
Leatherman, this volume.)
There are three major elements contributing to the cost of this scenario. They are: a) the cost of creating
new land and subsequently raising the elevation of the island as high tide levels increase; b) the cost of moving
buildings; and c) the cost of replacing infrastructure as it becomes inundated or damaged by the encroaching sea
Some of these costs will be incurred even if sea level does not rise. For example, infrastructure such as roads,
highways, buried utilities, etc., must be replaced as their useful lifetime runs out. Also, some buildings would
be razed and replaced even without sea level rise. In some cases, a rise in sea level may only reduce the
economic lifetime of a structure and hasten its replacement. The true costs attributable to sea level rise are the
additional costs that would not have been otherwise incurred, i.e., the cost of replacing a road that would not
otherwise require replacement or of replacing the road in 5 years rather than in 10 years.
On Long Beach Island, the number of buildings affected by sea level rise in each year following the year
2010 can be computed from Equation 7. The cumulative number of buildings is given by Equation 6. The
average cost of moving a building of the size of those located on Long Beach Island is about $10,000 including
the reconstruction of a new foundation. This is the dost to move the building a distance of less than 1/2 mile.
Thus in 2010, 37 buildings would be either abandoned to the sea or moved at a cost of $10,000 each (1987
dollars).
Determining the cost of creating additional land to which buildings can be moved requires that the volume
of fill needed and its unit cost be determined. The scenario investigated here assumes that fill will be required
to raise the land elevation. The land will initially be raised at the same rate as sea level rises. As sea level rises
further, starting in 2005, new land will be created at an elevation +1.4 feet above spring high tides. As a first
approximation, tidal ranges in the bay behind Long Beach Island were assumed to be unaffected by any increase
3-37
-------
Weggel
in mean sea level. Additional land area will also be created at an elevation to keep it at least 1.4 feet above the
then-prevailing spring high tides. See Figure 13. The depth of the bay dose to shore immediately behind Long
Beach Island averages about 4.1 feet below NGVD datum. Actually it ranges from about 1 foot deep over
relatively large areas to more than 10 feet deep in some small, restricted areas - mostly due to nearshore
navigation channels and deepened small craft mooring areas. The mean tidal range at several locations behind
Long Beach Island is given in Table IS. The average mean tidal range assumed for the present study was 22
feet. The mean spring tidal range was taken to be 3.2 feet. From Figure 7, each foot of sea level rise above
+3.5 feet would inundate about 13% of the island's area. Thus the portion of the island that would have to be
raised is 13% of the island's original 1986 area. New land created after 2005 would be at an elevation exceeding
+3.5 feet NGVD to keep it at least 1.4 feet above spring high tides at the then-prevailing mean sea level.
The scenario adopted for the rate of land creation was: a) starting at present, land would be raised at a
rate to keep its elevation 1.4 feet above spring high tides; and b) starting in 2005, land would be replaced on the
bay side of the island each year at a rate equal to the number of buildings moved each year times an average
building lot area of 12,600 square feet, i.e., 12,600 times the value of R given by Equation 7. (The year 2005
rather than 2010 was used as the start of filling in order to account for the possibility that some owners would
take preventative action before their buildings were damaged.) Figure 14 shows the volume of fill required each
year following 2005 along with the cumulative volume required. The annual amount of fill increases almost
linearly with time so that by 2100, about 680,000 cubic yards of fill will be needed each year. A total of 41
million cubic yards of fill will be required by the year 2100!
For this scenario, 12,600 square foot building lots are re-established on the bay side of the island, and the
annual volume of fill required between the present and the year 2005 is given by,
dV/dt = 12^00 + 877(YR-1986) (8)
where dV/dt is the rate at which sand must be used to create land in cubic yards per year. The cumulative
volume used up to a given year is given by,
V = 12300(YR-1986) + 438(YR-1986)2 (9)
for the years between the present and 2005, with V in cubic yards. For the years following 2005, the annual
volume and cumulative volume are given by,
dV/dt = 73,534 + 5,273(YR-1986) + 0.427(YR-1986)2 (10)
and,
V = -1,957,900 + 73,534(YR-1986) + 2,636(YR-1986)2 + 0.142(YR-1986) (11)
respectively.
These equations are plotted in Figure 14.
3-38
-------
Weggel
fft&
+2.1
^ ground level +&6NGVD
MGVD
-M
U* freeboard
MSL
tide
datum
datum
Figure 13 Daturas and Tide Levels in Bay Behind Long Beach
Island, NJ
3-39
-------
Weggel
TABLE 15 MEAN TIDAL RANGE IN BAYS BEHIND LONG BEACH ISLAND
Location Mean Range
(feet)
Barnegat Light 2.3
Surf City 1.0
Ship Bottom 1.5
Spray Beach 2.2
Beach Haven 2.2
Holgate 2.4
3-40
-------
10
?
HftO MOO 20ZO 20AO 2(760 2060 iJOO 2)20 2140
2160
Figure 14 Volume of Fill and Cumulative Volume of Fill
Required to Raise Long Beach Island, NJ as a
Function of Time.
-------
Weggel
Three major elements contributing to the costs of implementing this alternative are: a) the cost of fill; b)
the cost of replacing infrastructure including roadways and buried utilities (see Table 16); and c) the cost of
raising and moving buildings to the landward side of the barrier island. These costs are shown in Figure 15 as
a function of time.
Raise Island in Place
A second alternative, similar to the preceding one, is to simply raise the elevation of the island but without
moving it landward into shallower water. Buildings would be elevated as necessary and fill placed on the island.
Buildings already raised on pilings would simply have fill placed beneath them. Thus the present trend in coastal
areas of constructing elevated buildings on pilings would simplify implementation of this alternative. Much of
the buried infrastructure might continue to be used. For example, water, and storm and sanitary sewers would
still be serviceable for several years. They would simply end up being buried deeper beneath the ground surface.
Eventually, however, they would have to be replaced as increased seepage into storm and sanitary sewers
becomes a problem because of the relatively higher water table. The sea water environment would also
accelerate deterioration of the pipes. Roadways and sidewalks would have to be replaced at the time the island
is raised. This would be the major infrastructure replacement expense under this alternative.
The amount of fill required for this scenario is slightly less than the preceding scenario; however, the
difference is negligible. The amount of bulkheading is also about the same as for the preceding alternative since
the island's perimeter remains the same. See Figure ISa.
The number of houses involved under this alternative would also be somewhat less than the number
involved in the preceding alternative, particularly if houses, as they are replaced, are replaced with houses
constructed on elevated piling. The cost of simply raising a building in place is also assumed to be less than the
cost of both raising and moving it to another site. For this study, however, the number of houses to be raised
was assumed to be that given by equation 6.
The three primary elements contributing to the cost of this alternative are: a) the cost of fill; b) the cost
of replacing some infrastructure; and c) the cost of raising houses. See Figure ISa.
Dike Around Island and Provide Interior Drainage
A third response to rising sea level along a highly developed barrier island like Long Beach Island, New
Jersey, would be to construct dikes and an interior drainage system. The drainage system would have to handle
both the seepage under the dike resulting from an elevated sea level as well as the interior runoff resulting from
precipitation. The drainage system would have to handle the storm water that might otherwise drain by gravity
into the sea In general, the requirements to handle the runoff from precipitation will initially determine the
size of the storm water storage facilities and pumps needed. Since it would be uneconomical to provide pumps
having the capacity to drain runoff from rare storms with long return periods, storage facilities capable of
holding runoff until it can be pumped into the sea would have to be provided. For the Long Beach Island
scenario, relatively small storage facilities located under the street ends along the back side of the island were
assumed to provide the most economical storage alternative. Since space is generally not available to construct
large storm water storage facilities on Long Beach Island, a number of small tanks capable of holding the runoff
from a cross-section of the island about two blocks wide was assumed. Preliminary dimensions for the tanks
were what might be reasonably constructed beneath the street ends — tanks approximately SO feet wide, 100 to
200 feet long, and 10 to 20 feet deep. The storage capacity of such tanks vary from a minimum of about 50,000
cubic feet to a maximum of about 200,000 cubic feet. Storm water pumping would be episodic with most of it
occurring during and after major rain storms. The average annual precipitation at Long Beach Island is about
45 inches per year. This precipitation is not uniformly distributed over time; rather, it occurs during irregularly
spaced storm periods. Because of Long Beach Island's relatively small area, the precipitation can be assumed
to be uniformly distributed over the island.
3-42
-------
Figure 15 Annual Cost of Fill, Infrastructure Replacement,
and House Moving for Long Beach Island, NJ as a
Function of Time - Raise and Move Island Alternative.
-------
::: :;::i £!IflM*
0
'\%Q mo 2000 2020 2040 2060 2080 2100 2120 2140 2160
Figure 15a Annual Cost of Fill, Infrastructure Replacement,
and House Raising for Long Beach Island, NJ as a
Function of Time - Raise Island Alternative.
-------
Weggel
TABLE 16 HIGHWAYS AND STREETS - LONG BEACH ISLAND, NEW JERSEY
SHORE-PARALLEL ROADS
Primary N-S Road 30.14 mi
Second Major N-S Road 6.85
Third Major N-S Road 5.21
Total Other N-3 Roads 10.83
Total Shore-Parallel 53.03 mi.
SHORE-NORMAL ROADS
Total E-W Roads 70.79 mi.
Total Shore-Normal 70.79 mi.
TOTAL 123.82 mi
3-45
-------
Weggel
A hydrologic analysis of Long Beach Island was performed to determine the amount of storm runoff that
might be expected. This was used to size the storage and pumping facilities needed to handle the runoff.
Rainstorms with return periods ranging from 10 to 50 years were investigated and routed through the proposed
street-end storage tanks to determine how much interior flooding might result during periods of high runoff when
the tank storage capacity is exceeded. Runoff values were calculated using Soil Conservation Service (SCS)
methods outlined in "Urban Hydrology for Small Watersheds," Technical Release 55 (SCS TR-55). This method
relates the runoff per inch of rainfall to the watershed area, slope, land use, and the condition of the land cover.
Rainfall values were obtained from the National Weather Service's Rainfall Frequency Atlas of the United
States, Technical Paper No. 40. Storms having return periods ranging from 10 to 50 years with durations ranging
from 30 minutes to 24 hours were analyzed. Peak discharges were developed for drainage areas of about 24
acres, the contributing area assumed for each of the individual storage tanks. Peak discharges are given in
Table 17 for rainfalls of 24 hours' duration. For a storage tank capacity of 150,000 cubic feet, a pumping capacity
of 30 cubic feet per second is required to minimize flooding the interior of the island during major rainstorms.
If each individual drainage system drains about 24 acres, about 200 such systems are needed to serve Long Beach
Island.
In addition to interior runoff, seawater seepage beneath the dikes will occur, at first only during high tides,
but later, as sea level rises and the difference in elevation between mean sea level and the interior land increases,
seepage will occur during much of the tidal cycle. As time passes and sea level rises, the amount of seepage
will increase. Eventually, it might be necessary to pump almost continuously, albeit at a relatively low rate when
compared to stormwater drainage pumping requirements. To control seepage, an interior drainage system of
buried drain pipes was investigated. These drains would intercept seepage and convey it to the storage tanks
beneath the street ends. During periods of little or no storm runoff, the drainage system would continue to
intercept and hold the seepage until the water reached a given level in the tank, at which time a pump would
turn on and drain the tank.
Dikes to hold back the sea have long been constructed in Holland where much of the land is beneath the
present sea level. Because of its high population density, land reclamation has been economically justified in
Holland. It has been a matter of survival. The question remains whether it is economical to protect areas such
as Long Beach Island, New Jersey, from a rising sea level. At first, it would be economical since it would only
require replacing existing bulkheads with higher, more substantial structures. Bulkheads similar to those now
in existence along much of Long Beach Island's bay shoreline would be adequate to protect the land during
periods of high spring tides. As existing bulkheads deteriorate, higher bulkheads that also penetrate deeper into
the soil would take their place. The cost attributable to sea level rise is only the added cost of building higher,
more substantial bulkheads. However, if the rate of sea level rise is so rapid that the bulkheads must be raised
or replaced before they reach the end of their useful life, the cost of sea level rise is the value of protection for
the remaining lifetime of the bulkhead, which is now no longer adequate to provide protection, plus the added
cost of building a new, more substantial bulkhead to replace it. For the present drainage scenario, a substantial
concrete sheet pile bulkhead backed by an earth embankment was designed. See Figure 16. The estimated cost
per foot of the bulkhead is $500. A sheet pile bulkhead was selected because it can be designed to provide
sufficient soil penetration to limit the rate of seepage beneath it. The earth embankment provides lateral
stability, and the rubble toe protection prevents scour.
The amount of seepage beneath the sheet pile bulkhead was investigated using a computer program that
determines flow patterns beneath a bulkhead and calculates flow rates per unit length of bulkhead into a system
of drains on the interior side of the bulkhead. The number and pattern of drains can be selected. Several
patterns were investigated. In addition, several depths of penetration for the sheet pile bulkhead were
investigated with the computer model. The computed seepage patterns are shown in Figures 17 through 23.
Figures 17 and 18 show seepage patterns under a vertical bulkhead for two different depths of soil penetration.
Penetration depth is 6 feet in Figure 17 and 10 feet in Figure 18. For a soil permeability of 0.0005 feet per
second (typical for sands), the amount of seepage under the wall penetrating 6 feet (Figure 17) is 0.002 cubic
feet per second per foot (cfs/ft) of bulkhead. For 10 feet of penetration, the seepage rate is 0.0017 cfs/ft.
Figures 19 and 20 show the effect of providing a single drain 4 feet below the ground surface and 4 feet behind
the bulkhead, for 6 feet and 10 feet of pile penetration, respectively. The amount of seepage into the drain is
3-46
-------
Weggel
TABLE 17 SUMMARY OF RAINFALL FREQUENCY ANALYSIS - 24 HOUR RAINFALL
Recurrence
Interval
(yrs)
1
5
10
25
50
Ra
n
2
4
5
6
6
infall
epth
in .'
.'8
.5
.5
.0
.Q
Runoff
Depth
( in.)
1
2
2
3
4
.1
.2
.9
.5
.0
Peak *
Discharge
-------
Weggel
LAND
rubble joe
protection
^existing ground
reinforced concrete
sheet
Figure 16 Typical Cross-Section for Concrete Sheet Pile
Bulkhead/Dike.
3-48
-------
Weggel
ber of iterators is :00.
01 = DY 2.00 test.
Perwabihty is 0.00050000 ft/sec.
Discharge unJer wail is 0.00200 cfs/ft.
Figure 17 S-epage Under Sheet Pile Cutoff Wall - 6 foot
Penetration, No Drains.
3-49
-------
Weggel
ai!-5-T3 2 3 3 f 3 5
hiiber of iteratioM is 100.
N = DY = 2.00 fctt.
Pvitability ii 0.00030000 ft/sec.
Discharqi under Mil it 0.00166 cfs/ft.
10
Figure 18
«
Seepage Under Sheet Pile Cutoff Wall - 10 foot
Penetration, No Drains.
3-50
-------
Weggel
Huiber of iterations is
Discharge thru drain is
lischarge under nail is
100.
0.00105 cfs/«,
0.00363 cfs/ft.
Figure 19
e '.'nder Sheet Pile Cutoff Wall - 6 foot
Penetration, One Drain.
3-51
-------
Weggel
Nuibcr of iterations is
Discharge thru drain is
Discharge under nail is
100.
0.00072 cfs/fj;.
0.00297 cfs/ft.
Figure 20
Seepage Under Sheet Pile Cutoff Wall - 10 foot
Penetration, One Drain.
3-52
-------
Kuiter at iterations is 100.
Discharge thru drain 1 is 0.00033 cfs/ft. Weggel
Discharge thru drain 2 is
Discharge thru drain 3 is
Discharge thru drain 4 is
Discharge under vail is
0.00039 cfs/ft.
0.00058 cfs/ft.
0.00094 cfs/ft.
0.00363 cfs/ft.
ac
'0.1
Figure 21
Seepage Under Sheet Pile Cutoff Wall - 6 foot
Penetration, Four Drains 4 feet Below Ground.
,. Huiber of iterations is 100.
Discharge thru drain 1 is 0.00045 cfs/ft.
Discharge thru drain .2-'is 0.00050" cfs/ft.
Discharge thru drain 3 is 0.00079 cfs/ft.
Discharge thru drain 4 is 0.00162 cfs/ft.
Discharge under Mil it 0.00385 cfs/ft.
1.0
Figure 22
Seepage Under Sheet Pile Cutoff Wall - 6 foot
Penetration, Four Drains 8 feet Below Ground.
3-53
-------
Weggel
:f iterations is 100.
Disc'arge thru drain 1 is 0.00044 cfs/ft.
Discharge thru drain 2 is 0.00047 cfs/ft.
Discharge thn drain 3 is O.OOOiS cfs.'ft.
Discharge thru drain 4 is O.G0102 cfs/ft.
Discharge under wall is 0.00307 cfs/rt.
Figure 23 Seepage Under Sheet Pile Cutoff Wall 10 foot
Penetration, Four Drains 8 feet Below Ground.
3-54
-------
Weggel
0.00105 cfs/ft, while the total seepage rate under the wall is 0.00363 cfs/ft. Only about 29% of the seepage is
intercepted by the drain. Figure 23 shows the configuration finally adopted as typical for a drainage system for
Long Beach Island. Drains are located 8 feet below the ground surface, spaced 6 feet apart, with the closest
drain about 4 feet behind the bulkhead. The total discharge under the wall is 0.0031 cfs/ft, while the drains
intercept a total of 84% of the flow beneath the bulkhead. (Note that the total seepage intercepted by the drains
is not 100% because of simplifying assumptions made in the program.) The results of the seepage analysis are
summarized in Table 18.
Elements contributing to the cost of implementing this alternative are: a) the construction of new
bulkheads along the portion of the shoreline not now bulkheaded; b) the cost of constructing raised bulkheads;
c) the cost of the interior drainage system; d) the cost of the storm water detention tanks; e) the cost of the
pumps and pumping stations; and f) the cost of electrical power to operate the pumping system.
SUMMARY OF COSTS OF ALTERNATIVES AT LONG BEACH ISLAND
The cost of each of the three assumed alternative actions at Long Beach Island was evaluated. The unit
costs assumed in the analysis along with the lifetime of each element are given in Table 19. Fill was assumed
to cost about $6.00 per cubic yard. This is an average cost for hydraulically dredged fill obtained from the bay
areas and later from offshore sources as nearshore sources are depleted. The drainage system involves Figure
costs for the storage facilities, pumps, street and sidewalk replacement, and the drainage pipes. Major elements
of the system such as the storage facility were assumed to have a lifetime of 50 years. Other elements, such as
the pumps, were assumed to have shorter lifetimes. The $300,000 cost reflects the shorter lifetime of those
elements such as pumps, etc., and assumes their replacement over the 50-year lifetime of the entire system.
The costs of the three alternatives investigated for Long Beach Island are summarized in Table 20.
Raising the island in place is estimated to cost $136 billion. Most of this cost is associated with replacing
roadways, sidewalks, and other above-ground utilities as the island is raised. The cost of fill is estimated at $247
million, and the cost of raising buildings is estimated at $37 million.
Moving the island landward and raising it in response to sea level rise is estimated to cost $7.67 billion.
The major cost, that of replacing the infrastructure, is $7.35 billion. This includes replacing all of the buried
utilities, a major factor in establishing the higher cost of this alternative when compared with the preceding
alternative. The cost of raising and moving houses under this alternative is $74 million. The cost of raising and
moving a house is assumed to be twice the cost of simply raising a house. Also, with the replacement of houses
by houses elevated on piling over the years, there will probably be fewer houses that will have to be raised at the
time the island is raised.
The third alternative, that of providing a dike around the island and providing an interior drainage system,
appears to be the least expensive alternative with a total overall cost of $542 million. Most of this cost, $285
million, is associated with the construction of a dike system, assumed to occur in the year 2028. Construction
of the interior drainage system and its operation contributes $137 million to the cost. Power was assumed to be
available at $0.12 per kWh and the drainage system was assumed to operate for about 1000 hours per year. Each
of the nearly 200 storage/pumping systems was assumed to have an overall efficiency of 50% and to pump at
the rate of 30 cubic feet per second against a head of 20 feet for the 1000 hours.
Constructing new bulkheads and raising existing bulkheads contributes about $20 million to the cost of this
alternative. These costs are incurred between the present and the year 2028, when the bulkheads would be
abandoned in favor of a major dike system. The cost of raising bulkheads is discussed below. In general, the
only cost included here is the added cost of replacing existing or new bulkheads with higher bulkheads at the end
of their useful lifetime.
3-55
-------
Weggel
TABLE 18
SUMMARY OF SEEPAGE ANALYSIS BENEATH SHEET PILE BULKHEADS
Case
I
(Fig.
II
(Fig.
Ill
(Fig.
IV
(Fig.
V
(Fig.
VI
(Fig.
VII
(Fig.
17)
18)
19)
20)
21)
22)
23)
No. of
Drains
0
0
1
1
4
4
4
Depth of Discharge
Penetration Under Wall
ft) (cfs/ft)
6 0.
10 0.
6 0.
10 0.
6 0.
6 0.
10 ' 0.
0020
0017
0036
0030
0037
0039
0031
Discharge
to Drains
(cfs/ft)
0
0
0.0011
0.0007
0.0022
0.0034
0.0026
Percent
Interception
0
0
29
24
62
87
84
3-56
-------
TABLE 19 ASSUMED UNIT COSTS
Weggel
LIEU
Bulkheads
Fill
Dike System
Raising Houses
Raising & Moving Houses
Drainage System
("storage tanks, pumps
& drainage pipes)
Infrastructure
Roadway 2-lane
Roadway 4-lane
Sanitary sewer
Storm sewer
Water
Gas
LIFETIME
10 yr
50 yr
50 yr
Mil. CQ.S.T
$ 130.00/ft
$ 6.00 cu yd
$ 500.00/ft
$ 5,000 ea
$ 10,000 ea
$ 300,000 ea
TOTAL 2-lane
4-lane
$ 150.00/ft
$ 290.00/ft
$ 180.00/ft
$ 110.00/ft
$ 20.00/ft
$ 20.00/ft
~$480.00/ft
$ 620.00/ft
3-S7
-------
Weggel
TABLE 20 SUMMARY OF COSTS OF THREE LONG BEACH ISLAND
ALTERNATIVES
ALTERNATIVE
(Cost Item)
AVG ANNUAL COST *
CUMULATIVE COST **
Island in. P_lafi£.
Fill
Infrastructure
(roads only)
Raise buildings
TOTALS
$ 2.2 million
9.4 million
0.3 million
$ 11.9 million
$ 247 million
1,072 million
37 million
$ 1,356 million
Raise & Move Island Landward
Fill $ 2.2 million
Infrastructure 64.5 million
(roads & buried utilities)
Raise & move
buildings 0.6 million
TOTALS
$ 67.3 million
$ 247 million
7,352 million
74 million
$ 7,673 million
Dike Island and Provide Interior Drainage Systep
New Bulkheads 0.1 million 12 million
Raise Bulkheads 0.1 million 8 million
(added cost between the years 1986 and 2028)
Dike System 2.5 million 285 million
(constructed in 2028)
Drainage System 0.5 million 57 million
(system operation) 1.6 million 180 million
TOTALS
4.8 million
542 million
* Cumulative total cost divided by 114 years. Note, however, that
all costs may not extend over the entire 114 year period.
** Total costs incurred between the years 1986 and 2100.
3-58
-------
Weggel
CHAPTERS
SUMMARY OF ACTIONS AND THE COST OF RESPONDING TO SEA LEVEL
RISE AT INDEX SITES
The USGS quads for each of the index sites was studied, and a strategy for responding to a rising sea level
was determined for the site. Specifically, the elevation of developed areas was considered, and those areas below
+10 feet NG VD were considered for protection. If a reasonable level of development was present, a dike system
to surround the development or to tie it in with high ground was proposed as the response. The length of the
dike required was determined from the map. The unit cost of constructing a dike was assumed to be about $500
per linear foot of dike. If the diked area was isolated, and connected to surrounding high land by roads at a low
elevation, the cost of raising the elevation of the roadway and/or replacing it was included in the cost of
responding to sea level rise. In general, a two-lane highway was considered as the connecting link and its cost
determined. The unit cost of replacing a two-lane highway was assumed to be $480 per lineal foot. The cost
of replacing a four-lane highway was $620 per foot. These figures include the cost of replacing utilities buried
beneath the roadway, lighting, and drainage.
If the shoreline is already protected by existing bulkheads, the length of the existing bulkhead was
determined and consideration given to raising and/or replacing it as sea level rises. The total cost of replacing
the bulkhead was not attributed to sea level rise, but rather an analysis of the cost of replacing the bulkhead
with a higher bulkhead was made. That portion of the cost of raising bulkheads attributable to an increase in
the mean sea level is made up of two components: the added cost, over and above the bulkhead's replacement
cost, of having to build a higher bulkhead at the end of its useful lifetime, and the cost of replacing the bulkhead
early because it does not provide sufficient protection against a rising sea level. For the present study, only the
former costs are considered, i.e., the bulkhead's design is assumed to adequately consider the projected increase
in sea level so that a rise in sea level does not require its early replacement, but rather, deterioration of the
materials or some other mode of failure is the cause for its replacement.
The cost of raising bulkheads depends on the initial cost per lineal foot of the bulkhead, its initial height,
its useful lifetime (how often it needs to be replaced), and the increase in bulkhead height whenever it needs to
be replaced. The required increase in bulkhead height will vary with time, since the rate at which sea level is
projected to rise varies with time. Bulkhead costs vary approximately with the 1.5 power of the height. Thus,
a cost increase factor was calculated based on the increase in sea level during the bulkhead's lifetime. For
example, for a bulkhead with a lifetime of 20 years, the increase in sea level over 20 years was used to compute
the increase in bulkhead height necessary. Obviously, the increase in height depends on what point in time the
bulkhead is replaced, since the rate at which sea level is rising is projected to increase. The cost increase factor
was defined as the increased bulkhead height divided by its original height raised to the 1.5 power. The cost
increase factor multiplied by the initial bulkhead cost (1986 dollars) gives the new bulkhead cost (1986 dollars).
The length of bulkhead to be replaced each year was taken to be the total bulkhead length divided by the
bulkhead's lifetime.
Assuming an initial bulkhead height of 5 feet, replacement every 10 years, and an initial cost of $130.00
per foot, the added cost per foot of bulkhead per year averaged over the 114-year period between 1986 and 2100
is about $1.60. This will be lower during the early years of the 114-year period but will be higher toward the end
of the next century. The total cost for the time period extending to the year 2100 will be the total bulkhead
length required times the increased cost per foot per year times 114 years. For 1 mile of bulkhead, this amounts
to expending more than $1.2 million just to replace existing bulkheads at the end of their lifetime with the
required higher bulkheads.
For areas that are presently unbulkheaded, consideration was given to the need for new, additional
bulkheading to protect vulnerable, low-lying areas. New bulkheading was assumed to cost $130 per lineal foot.
This represents the present (1988) cost of aluminum sheet piling, about 13 feet long with a concrete cap and
anchors.
3-59
-------
Weggel
For buildings located in sparsely developed areas where construction of a dike system would not be
economical, consideration was given to moving individual structures to higher ground. Again, not all structures
would be moved under such circumstances. The decision to move a structure would depend on its value, its
condition, its movability, and the availability of high land within a reasonable distance to which the structure
could be moved. It was assumed that about one-half of the buildings identified on the USGS quads would be
moved. That is, half of the isolated buildings were considered to be candidates for moving; however, the
number of buildings shown on the undeveloped (unshaded) areas of the USGS quads may not be representative
of the actual number of buildings present because of development that may have occurred since the time the
quads were last corrected. Therefore, the number of buildings on the quads was used to estimate the cost of
moving structures. The unit cost of moving a structure was assumed to be about $10,000 based on recent costs
to move a 1000 square foot house a distance of about 1/2 mile. This figure includes the cost of preparing a
new, simple foundation system (simple footings or concrete slab). Obviously, larger structures and moves of
more than 1/2 mile would increase this cost.
The summary of actions that might be taken at each of the index sites is given in Table 21.
COST OF RESPONDING TO SEA LEVEL RISE AT INDEX SITES
The costs of responding to sea level rise at the six index sites are given in Table 22. The costs in the table
are the total costs that would be incurred during the 114-year period between 1986 and 2100, which are
attributable just to the increase in sea level. Structure replacement costs that would be incurred if sea level
remained at its present level have been subtracted. Where bulkheads are already present or are needed to
respond to higher sea levels, the cost of replacing them with higher bulkheads at the end of their useful lifetime
overshadows most of the other costs considered. The cost for the New York index site is $275 million, with
$205 million of that attributable to raising the existing and proposed bulkheads. Where bulkheading is not
considered to be practical because of low levels of economic development, the costs are significantly less. For
example, at Dividing Creek, New Jersey, only 6.1 miles of new bulkheading appear justified and the cost is only
$5.8 million, most of it associated with raising 7.2 miles of highway. The Long Beach Island area cost figures
on Table 22 are given for both the mainland behind the barrier island and for the island and mainland combined.
Note that the cost of raising the roads on Long Beach Island are not included.
USGS TOPOGRAPHIC DATA ANALYSIS - INDEX SITES
In order to extrapolate the results from the six index sites to a sub-set of the sites investigated by Park et
at. (this volume), the topographic and economic development conditions at the six index sites were compared
with the conditions at Park's sites. These comparisons were subsequently used to determine the cost of
responding to sea level rise at Park's sites. Topography digitized from USGS quads (Park et al., this volume)
was used to determine the distribution of land elevations and shoreline lengths at the six index sites. For most
of the sites, a spatial matrix of ground elevations averaged over a 500- by 500-meter pixel was available. For the
New York and Corpus Christi areas, elevations were averaged over a 250- by 250-meter pixel. A histogram of
the distribution of elevations was determined by summing the number of pixels having an elevation within a given
elevation interval. For the land elevation histograms, 10 intervals of 1 foot each between elevations 0 and 10 feet
NGVD were selected for the analysis. In addition, the number of pixels with elevations above the 10-foot
elevation and the total number of pixels with elevation above 0 feet NGVD were determined. In addition, a
simple algorithm was developed to estimate the shoreline lengths for each of the index sites. The length of the
shoreline was determined using an algorithm that sweeps the elevation matrix two columns at a time and two
rows at a time and assigns a shoreline length to the resulting four-pixel pattern, depending on the number of
pixels with elevations above zero. If all four pixels are above 0 or below 0, no shoreline length is assigned. If
one or three pixels are at 0 or below, the shoreline length assigned is 1.414 times the length of the side of a pixel.
If two pixels are at 0 elevation or below, the length of shoreline assigned is either 1.0 times the length of a side
or 2.828 times the length of side, depending on the pattern of land and water pixels. If the land pixels are
3-60
-------
rABL£ - J. ^j-ii"rt~v'
Index Site & USGS Quad
Weehawken
Arthur Kill
Brooklyn
The Narrows
Jersey City
Elizabeth
Central Park
SUB TOTAL
Ship Bottom *
Tuckerton *
Long Beach NE *
Beach Haven *
SUB TOTAL
* Excludes Long Beach Is
Port Norris
Fortescue
Cedarville
Dividing Creek
SUB TOTAL
North Miami
Miami
SUB TOTAL
Oso Creek NE
Portland
Crane Island NW
Port Ingleside
SUB TOTAL
** Includes trailer park
Redwood Point
Newark
Palo Alto
Mountain View
SUB TOTAL
TOTALS
"S3 REQUIrt^D A- "aA " •*"/
New Bulk. Raise Exist.
Bulkhead
(mi) (mi)
HEJ. YQfiiL. HI AREA
29.50 21.32
5.93 4.03
2.28 14.91
6.69 3.50
11.25 58.89
9.92
17.52 30.19
33.09 132.84
LQH£ BEACH. ISLAND. Hi AREA
0.67
3.73
-
— —
4.40 0
land itself.
DIVIDING CREELo. ILL AREA
2.38
2.68
-
1.04
6.10
MlAMl^ EL, AREA
12.83 22.37
2.83 79.06
15.66 101.43
CQRPJLS. CHRISTI . TX AREA
15.86 9.10
-
2.39
2.98
15.86 14.47
structures
SAH. FRANCISCO. CA AREA
0.89
0.60 6.56
0.89 6.41
1.49 3.43
3.87 16.40
113.32 265.14
3-61
**i. r\ . ~j •* "} e\
XJ XJ k V 1. ij t^ M* f* »
Buildings
to Move
<*)
21
27
-
-
-
-
-
48
139
135
-
—
274
152
20
190
116
478
-
27
27
82 **
8
83
111
284
83
28
8
84
203
1287
N " ""' A -J L *. — J
Highway
to Raise
(mi)
2 .89
-
-
-
-
-
—
2.89
1.49
-
-
—
1.49
0.37
0.45
3.28
3.06
7.16
2.54
—
2.54
5.97
3.73
2.24
4.18
16.12
1.49
2.53
1.34
0.75
6.11
36.31
-------
Weggel
Table 22. COSTS ASSOCIATED WITH SEA LEVEL RISE AT INDEX SITES
New York Area. NY & NJ
New Bulkheads
Raise Bulkheads
Move Buildings
Raise Highways
Lone Beach Island Area. NJ (Mainland
New Bulkheads
Raise Bulkheads
Move Buildings
Raise Highways
Lone Beach Island Area. NJ (Mainland
New Bulkheads
Raise Bulkheads
Move Buildings
Raise Highways
Dividing Creek Area. NJ
New Bulkheads
Raise Bulkheads
Move Buildings
Raise Highways
Miami & Miami Beach Area. FL
New Bulkheads
Raise Bulkheads
Move Buildings
Raise Highways
Corpus Christi Area. TX
New Bulkheads
Raise Bulkheads
Move Buildings
Raise Highways
San Francisco Bay Area. CA
New Bulkheads
Raise Bulkheads
Move Buildings
Raise Highways
83.1 mi
215.9 mi
48
2.9 mi
onlv)
4.4 mi
4.4 mi
270
1.5 mi
and bulkheadine on back
17.2 mi
36.7 mi
270
1.5 mi
6.1 mi
6.1 mi
478
7.2 mi
15.7 mi
117.1 mi
27
26 mi
15.9 mi
303 mi
284
16.1 mi
3.9 mi
203 mi
203
6.1 mi
$ 57.0 million
205.3 million
0.5 million
9.5 million
Total $272.3 million
$ 3.0 million
4.2 million
2.7 million
3.8 million
Total $«13.7 million
side of island)
$ 11.9 million
35.0 million
2.7 million
3.8 million
Total $ 53.4 million
$ 4.2 million
5.8 million
4.8 million
18.2 million
Total $ 33.0 million
$ 10.8 million
1113 million
03 million
8.3 million
Total $130.7 million
$ 10.9 million
28.8 million
2.8 million
40.9 million
Total $ 83.4 million
$ 2.7 million
193 million
2.0 million
20.0 million
Total $ 44.0 million
3-62
-------
Weggel
diagonally opposed, the shoreline length factor is 2.828, while if the land pixels are adjacent to one another, the
shoreline length factor is 1.0. Because of the coarse size of the pixels (500 meters x 500 meters), in most cases,
the algorithm underestimates the length of the actual shoreline, since small variations in the actual shoreline are
replaced by straight line segments. For engineering purposes, the estimate is probably sufficient since
erosion/flood control structures such as bulkheads are constructed in straight-line segments to minimize their
length rather than along the lines of a tortuous shoreline. A correlation between shoreline lengths determined
from the digitized data and shoreline lengths found from planimetering the USGS quads is shown in Figure 24.
The index site corresponding to each point is indicated on the figure. The data are scattered about the 45 degree
line of equality.
Table 23 provides topographic information on the six index sites where the overall distribution of ground
level elevations is given. These data were determined by planimetering the USGS quads. Table 24 provides
topographic, shoreline length, and the slope of the land below given elevations as determined from the quads,
while Table 25 provides similar data as determined from the digitized topographic data. Shoreline lengths
obtained from the digitized topographic data analysis are given in Table 26 for the six index sites as well as for
78 additional coastal locations around the US.
3-63
-------
Weggel
TABLE 23 SUMMARY OF INDEX SITE TOPOGRAPHIC CONDITIONS
(Based on total land area)
Elevation (ft) Z<5
USGS Area below
Elevation 22.5
USGS % below
Elevation 49.1
Digitized USGS
% below Elev 51.0
USGS Area below
Elevation 48.2
USGS % below
Elevation 55.2
Digitized USGS
% below Elev 35.6
USGS Area below
Elevation 22.6
USGS % below
Elevation 20.7
Digitized USGS
% below Elev 29.7
USGS Area below
Elevation 14.1
USGS % below
Elevation 20.5
Digitized USGS
% below Elev 33.1
USGS Area below
Elevation
USGS % below
Elevation
Digitized USGS
% below Elev 23.0
USGS Area below
Elevation 57.7
USGS % below
Elevation 61.1
Digitized USGS
% below Elev 20.7
<10 Z<15 Z<20 Z<25 Z<30
LOSS. BEACH ISLAND. NJ. ABEA
Z<35 Z<40 Total
36.5
79.4
61.6 76.
DIVIDING
72..6
83.2
49.3 98.
MIAM
42
92
8 77
CBEEE
87
92
7 99
I, EL
.6
.8
.1 78.4
^ M ABEA
.3
.8
.0 99.4
ABEA
45.9
100.0
78.7
94.1
100.0
99.6
78.5 107.2 109.1 109.1 109.1
71.9 98.3 100.0 100.0 100.0
67.9 99.7 100.0 100.0 100.0
CQBPJIS. CHRISTI. IX AREA
25.5 42.8 68.9 90.7 109.5
37.0 62.1 100.0 100.0 100.0
44.0 89.7 94.6 97.0 97.1
HE1 XQBL. HI ABEA
78.1 - 122.7 - 148.8
52.5 - 82.5 - 100.0
31.4 87.8 89.1 90.1 90.2
SAH FRANCISCO. CA AfiEA
68.1 - 83.6 - 94.4
72.1 - 88.5 - 100.0
25.1 85.7 86.3 86.6 86.8
45.9 45.9
100.0 100.0
78.7 78.7
94 .1 94 .1
100.0 100.0
99.7 99.3
109.1 109.1
100.0 100.0
100.0 100.0
157.2 157.2
100.0 100.0
97.4 97.4
148.8 148.8
100.0 100.0
90.5 90.5
100.0 100.0
87.0 87.1
45.9
100.0
100.0
94. 1
100.0
100.0
109.1
100.0
100.0
157.2
100.0
100.0
178.5
100.0
100.0
94.4
100.0
100.0
-------
Weggel
TABLE 24 AVERAGE GROUND SLOPE NEAR SHORELINE
(Based on data obtained from USGS quads)
Index Site Area Below . Shoreline
Given Elevation Length
(sq mi) (mi)
<5 ft <10 ft
New York - 78. 14 220.7
Long Beach
Island 22.55 36.46 109.8
Dividing
Creek 48.19 72.58 97.0
Miami 22.60 78.48 141.4
Corpus
Christi 14.14 25.46 189.0
San
Francisco 57.67 70.54 42.4
Slope of Land Below
Given Elevation
<5 ft <10 ft
0.0054
0.0046 0.0057
0.0019 0.0025
0.0059 0.0034
0.0127 0.0141
0.0007 0.0011
3-65
-------
Weggel
TABLE 25 AVERAGE GROUND SLOPE NEAR SHORELINE
(Based on digitized US6S topo data)
Index Site Area
Given E
(sq
<5 ft
New York
Long Beach
Island
Dividing
Creek
Miami
Corpus
Christi
San
Francisco
43
18
46
32
26
17
.04
.73
.80
.91
.18
.32
Below
levat
mi)
<10
58
22
64
75
34
20
ion
ft
.70
.59
.81
.24
.83
.98
Shoreline
Length
(mi)
333
77
122
119
130
175
.4
.4
.3
.4
.1
.1
Slope of Land Below
Given Elevation
<5 ft <10 ft
0
0
0
0
0
0
.0073
.0039
.0025
.0034
.0047
.0096
0
0
0
0
0
0
.0108
.0065
.0036
.0030
.0071
.0158
3-66
-------
Weggel
soo
^>
200
-o
ItQ
ifiGTjZU
"7T"
I ; ! i I
:§£:
-H-
100
in
•toftffi
/injft^l
-U.
-v^r
:--tifttt:m^^^T=
^\- i i i I i I I i I I i I i i i I I i I i I 1-4-
Shorelirie Leng+h from
Topo
\00
100
WQ
Figure 24
Correlation Between Shoreline Lengths Measured
from USGS Quads and Shoreline Lengths Obtained
from Digitized USGS Topographic Data.
3-67
-------
Weggel
TABLE 26 SUMMARY OF COASTAL SITE SHORELINE LENGTHS AND AREAS
Site
Abbrev .
MEFREEPO
MEROCKLA
MEJONESP
MAMARBLE
MAWESTPO
MAOBLEAN
RIWATCHH
CNBRIDGE
* NYBROOKL
NYNARROW
NYPATCHO
NYSOUTHA
* NJDIVIDC
* NJLONGBE
DEREHOBA
MDEASTON
MDCOVEPT
MDELKTON
MDMIDDLE
VACOLBEA
VABLOXOM
VANEWPOR
VAWILTON
NCENGELH
NCWILMIN
NCLONGBA
NCCAMPLE
SCHILTON
SCCHARLE
SCBROOKG
GASEAISL
Shoreline
Length (mi)
141.9
75.2
183.2
57.2
121.2
144.4
106.9
66.8
333.4
252.0
85.3
184.6
122.3
77.4
99.1
73.2
9.6
53.2
108.1
252.0
81.5
143.0
170.4
85.1
150.1
159.1
115.7
233.8
164.5
87.2
145.8
Area
;5 ft
22.1
10.5
38.1
4.1
15.8
37.9
12.2
12.7
43.0
15.0
12.6
11.0
46.9
18.7
11.0
11.0
0.2
4.5
4.9
15.2
28.9
27.7
24.8
169.4
32.0
66.1
23.2
83.4
74.3
63.6
49.2
(sq mi)
<10 ft
24.1
12.2
39.6
22.3
22.3
42.8
17.8
23.4
58.7
41.5
17.0
20.5
64.9
22.6
18.9
20.8
0.2
6.3
10.2
41.8
33.4
49.2
50.2
190.1
44.0
95.8
30.0
103.6
109.2
72.4
55.9
Devel .
1.0
0.6
0.4
1.5
1.5
24.8
1.6
12.4
52.0
15.0
5.1
2.0
3.0
2.3
1.2
0.5
0.0
0.1
0.2
0.0
9.3
1.6
1.1
27.3
2.2
1.4
9.9
12.2
1.6
1.1
3-68
-------
Weggel
TABLE 26 (cont. ) SUMMARY OF COASTAL SITE SHORELINE LENGTHS AND AREAS
Site
Abbrev.
FLLOSTMA
FLCARDSO
FLFTGADS
FLAPALAC
* FLMIAMI
FLVENICE
FLSTAUGU
FLEVERGL
FLCAPECA
FLSNIPEI
FLKEYWES
FLHOLLEY
FLFORTMY
FLPORTRI
FLSTJOSE
ALGRANDI
MSPASSCH
MSGULFPO
LALULING
LABARATA
LAGOLDME
LABELLEC
LACAMERO
LAPONCHA
LASULPHU
LALMISER
LAGRANDC
LAPELICA
LAMAINPA
TXALLIGA
TXGREENI
* TXPORTLA
TXPALACI
TXRIVIER
TXSMITHP
TXT I VOL I
Shoreline
Length (mi)
182.5
93.9
130.4
96.7
119.4
50.2
91.9
215.9
159.1
51.5
** 33.6
114.6
63.1
0.1
89.6
60.3
62.0
29.6
186.7
236.7
279.6
354.3
262.8
70.9
99.4
120.7
202.5
202.4
229.0
186.2
114.4
175.1
141.8
112.0
109.6
137.4
Area
<5 ft
187.1
84.8
99.5
21.9
32.9
4.2
26.0
204.2
36.8
37.5
3.4
19.4
31.6
0.0
12.3
24.3
10.2
6.6
160.4
141.0
201.1
174.6
116.1
86.1
92.6
-
170.1
66.5
50.8
148.3
80.1
26.2
38.9
33.5
46.0
42.7
(sq mi)
<10 ft
184.1
85.5
141.5
34.0
75.2
10.1
41.9
204.2
70.2
72.3
5.0
28.1
75.1
0.0
27.6
35.0
13.4
11.4
176.4
141.1
201.1
177.3
122.3
92.4
139.4
212.6
170.6
66.5
50.8
194.1
89.4
34.8
66.2
77.2
55.6
77.5
Devel.
0.0
0.1
0.7
1.9
46.3
1.9
5.9
1.0
54.4
0.0
0.0
0.6
12.3
0.0
0.0
0.6
1.0
1.6
20.0
1.5
2.2
7.9
2.1
2.9
5.3
2.1
1.2
0.1
0.0
0.2
0.0
6.0
0.5
0.0
1.3
1.4
3-69
-------
Weggel
TABLE 26 (cont.) SUMMARY OF COASTAL SITE SHORELINE LENGTHS AND AREAS
Site
Abbrev.
CABENICI
CAANONUE
CASANQUE
CAOCEANS
CAPTSAL
* CAPALOAL
CAALBION
CAFERNDA
CATIJUAN
CAPTMUGU
ORPORTOR
ORYAQUIN
W A AN AGO R
WAGARDNI
WANEMAH
WAPORTGA
WATACOMA
Shoreline
Length (mi)
** 358.7
57.3
** 130.1
** 28.9
** 27.1
133.6
34.2
85.9
72.6
64.0
51.0
109.6
105.9
90.1
158.8
83.8
115.7
Area
<5 ft
52.2
3.8
17.3
0.8
0.6
84.6
0.0
12.5
150.1
39.8
1.5
0.3
6.5
2.4
51.4
6.6
2.5
(sq mi)
<10 ft
56.8
4.0
21.0
1.6
0.6
98.7
0.0
40.0
157.9
60.3
1.93
0.6
7.0
3.0
54.5
7.6
4.0
Devel .
6.3
0.0
9.0
0.1
0.0
17.8
0.0
0.8
68.9
0.0
0.3
0.1
0.4
0.3
4.1
1.7
0.7
* Denotes Index Site
** Denotes Fine Grid Data (250 m x 250 m)
3-70
-------
Weggel
CHAPTER 6
EXTRAPOLATION OF COSTS TO INCLUDE THE SHELTERED
SHORELINES OF THE UJS.
A regression analysis was made using the costs of responding to sea level rise at the six index sites in order
to determine the costs at 78 other coastal sites for which digitized topographic data were available (Park et al.,
this volume). These 84 sites comprise about 13.6% of the US. shoreline. The regression analysis developed
equations for the amount of new bulkheading required, the amount of bulkheading that will be needed to
respond to sea level rise in the years between the present and the year 2100, the number of buildings to be
moved, and the number of miles of highway that would have to be raised to provide access to nearshore areas.
These variables were related to topographic and development variables such as the percentage of the land below
the + 5-foot NGVD contour that is economically developed, and the average slope of the land below the + 5-
foot contour.
The percentage of the shoreline's length that is bulkheaded correlated with the percentage of the land
below the + 5-foot contour that is economically developed and with the average land slope below the + 5-foot
contour (see Figure 25). For steep land slopes (SL), the slope of the line on the figure (SI) is lower. For
flatter nearshore slopes (lower SL), SI is greater. (The slope of the land is defined here as the + 5-foot
'elevation divided by the average distance between the + 5-foot contour and the shoreline. The average distance
of the + 5-foot contour from the shoreline is equal to the land area below +5 feet divided by the shoreline
length.) Similarly, the length of bulkheading to be raised in response to sea level rise was also related to the
percentage of land below +5 feet that is economically developed and the average land slope (see Figure 26).
Again, the slopes of the lines on the figure (S2) were found to be a function of the land slope below the +5-
foot contour (SL). The relationship between the slopes of the lines on Figures 25 and 26 and the nearshore
land slope is shown in Figure 27. The slopes of the lines on the figures vary almost linearly with SL. The
equation is,
B% = SI (%AD5) (12)
in which B% is the percentage of the shoreline that is presently bulkheaded, SI is the slope of the line in Figure
25 and %AD5 is the percentage of land below the +5 foot contour that is economically developed. Similarly,
R% = S2 (%AD5) (13)
in which R% is the percentage of the shoreline length that will be bulkheaded by the year 2100 due to sea level
rise, i.e., the amount of bulkheading that will have to be raised in order to provide continued protection because
of rising sea level. S2 is the slope of the lines in Figure 26. The slopes SI and S2 are related to the land slope
near shore by the relationships,
SI = -146.8 SL + 1.85 (14)
and S2 = -167.6 SL + 2.20 (15)
See Figure 27. The length of new bulkheading that will be needed is simply the amount that will eventually be
needed (the amount that would have to be raised), R%, minus the amount that is presently there, B%.
3-71
-------
Weggel
1D:-
0
Figure 25 Correlation Between X of Shoreline Bulkheaded
and % of Land Below +5 feet that is Developed
and Slope of Land Near Shore.
3-72
-------
Weggd
Figure 26
Correlation Between Anount of Bulkheading to be
Raised as a % of Shoreline Length and % of Land
Below +5 feet that is Developed and Slope of Land
Near Shore.
3-73
-------
Weggel
.001 .004 .006 .006 ,010 ML .014
Figure 27
Relationships for Slopes of Regression Lines for
Bulkheading Variables and Slope of Land Hear Shore,
3-74
-------
Weggel
The number of houses that are candidates for moving was related to the percent of the shoreline that is
presently bulkheaded by the relationship shown in Figure 28. The number of houses is given by,
N = 0.04762 B%2 - 9.762 B% + 500 (16)
N can be determined by first determining B% from equation 12 and then using equation 16.
No satisfactory expression could be established to determine the number of miles of highways that would
have to be raised or relocated. To estimate the costs of highway relocation, the number of miles was simply
expressed as a fraction of the developed land area below +5 feet in elevation. Thus,
LH = 0.45 ADS (17)
in which LH is the length, in miles, of highways that have to be replaced and AD5 is the area, in square miles,
of developed land below +5 feet in elevation.
Costs were estimated for constructing new bulkheads, for periodically raising the existing and new
bulkheads, for moving houses, and for raising/relocating highways in low-lying areas. Costs were estimated in
the same way as they were estimated for the index sites. The results of the analysis are given in Table 27.
The numbers in parentheses in Table 27 are the numbers of miles of new and total bulkheading, the
number of buildings to be moved, and the number of miles of highways to be raised or relocated, respectively.
For the 84 sites for which shoreline lengths and developed areas were determined, the total cost of responding
to sea level rise is about $3.36 billion. This is the low estimate with the cost of bulkheading/diking at $130.00
per foot. If this cost is increased to $500.00 per foot, the cost of responding to sea level rise increases to $10.8
billion (see Table 28). Since these 84 sites represent 13.6 of the U.S. sheltered shorelines, the low estimate for
the entire U.S. shoreline is $24.6 billion, while the high estimate is $80.2 billion. These figures are obviously
biased because of the characteristics of the 84 coastal sites on which the extrapolation is based. The sites favor
the rural, little-developed southeastern U.S. coastal regions while ignoring the heavily developed Northeastern
States and the Pacific Northwest. While several heavily developed areas such as the New York metropolitan
area, the Miami/Miami Beach area, and the San Francisco Bay area have been included in the present analysis,
a more representative sample of coastal sites would probably provide significantly higher cost estimates. The
costs are summarized by coastal region in Table 29.
3-75
-------
Weggel
Figure 28
Relationship for Nunber of Houses to be Moved as a
Function of the % of Shoreline that is Bulkheaded.
3-76
-------
Weggel
TABLE 27 THE COST OF SEA LEVEL RISE AT 84 SITES ALONG THE
U.S. COASTLINE - LOW ESTIMATE
Site Cost of New
Bulkheads
MEFREEPO $
MEROCKLA $
MEJONESP $
MAMARBLE $
MAWESTPO $
MAORLEAN $
RIWATCHH $
CNBRIDGE $
NYBROOKL $
NYNARROW $
NYPATCHO $
NYSOUTHA $
NJDIVIDC $
NJLONGBE $
4.09
5.96 mi)
2.44
3.55 mi)
1.60
2.33 mi)
0.00
0.00 mi)
5.99
8.72 mi)
85.61
124.72 mi)
6.22
9.06 mi)
45.88
66.84 mi)
156.07
227.38 mi)
0.00
0.00 mi)
21.54
31.38 mi)
0.00
0.00 mi)
4.39
6.40 mi)
- 8.04
$
$
$
$
$
$
$
$
$
$
$
$
$
$
Cost to
Raise
Bulkheads
6.99
( 7 . 36
4.. 20
( 4.42
2.69
( 2.83
0.00
( 0.00
10.40
( 10.93
137.26
(144.35
11.04
( 11.61
63.56
( 66.84
271.45
(285.47
0.00
( 0.00
36.95
( 38.86
0.00
( 0.00
7.31
( 7.69
13.48
$
mi )
$
mi )
$
mi )
$
mi )
$
mi )
$
mi )
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
Cost to
Move
Bu i Idings
4.60
( 460)
4.55
( 455)
4.88
( 488)
5.00
( 500)
4.32
( 432)
0.12
( 12)
4.21
( 421)
0.00
( 0)
0.56
( 56)
5.00
( 500)
2.05
( 205)
5.00
( 500)
4.50
( 450)
3.63
$
( 0
$
( 0
$
( o
$
( o
$
( o
$
( 11
$
( 0
$
( 5
$
( 17
$
( 6
$
( 2
$
( o
$
( o
$
Cost to
Replace
Highways
1
.44
0
.26
0
.18
2
.87
1
.65
28
.16
1
.74
14
.56
43
.09
17
.74
5
.29
2
.91
1
.74
2
. 11
mi )
.86
mi )
.47
mi)
.20
mi)
.65
mi)
.30
mi)
.87
mi)
.10
mi)
.32
mi)
.08
mi)
.82
mi)
.32
mi)
.88
mi)
.53
$
$
$
$
$
$
$
$
$
$
$
$
$
$
Total
Cost
16
11
9
7
22
251
23
123
471
22
66
7
18
27
.79
.85
.64
.20
.36
.28
.33
.53
.39
.08
.36
.32
.08
.68
11.71 mi) ( 14.18 mi) ( 363) ( 1.00 mi)
-------
Weggel
TABLE 27 (cont.) THE COST OF SEA LEVEL RISE AT 84 SITES ALONG THE
U.S. COASTLINE - LOW ESTIMATE
Site Cost of New
Bulkheads
DEREHOBA $
<
MDEASTON $
<
MDCOVEPT $
<
MDELKTON $
<
MDMIDDLE $
<
VACOLBEA $
<
VABLOXOM $
<
VANEWPOR $
<
VAWILTON $
<
NCENGELH $
<
NCWILMIN $
<
NCLONGBA $
<
NCCAMPLE $
(
SC HILTON $
(
4.29
6.25 mi)
2.03
2.95 mi)
0.00
0.00 mi)
0.18
0.26 mi)
0.00
0.00 mi)
0.00
0.00 mi)
0.00
0.00 mi)
37.15
54.13 mi)
6.54
9.53 mi)
0.42
0.61 mi)
75.52
110.02 mi)
3.15
4.60 mi)
3.05
4.44 mi)
24.70
35.98 mi)
$
$
$
$
$
$
$
$
$
$
$
$
$
$
:ost -. j
Raise
Bulkheads
C
(
(
(
(
(
(
(
(
(
(
(
(
(
7.68
8.C5
3.47
3.65
0.00
0.00
0.38
0.40
0.00
0.00
0.00
0.00
0.00
0.00
62.75
66.00
11.24
11.82
0.69
0.72
127.11
133.68
5.24
5.51
5.14
5.41
41.12
43.25
$
.-n i )
$
mi )
$
mi )
$
mi)
$
mi)
$
mi)
$
mi )
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
Cost to
Move
Buildings
4.40
( 440)
4.61
( 461)
5.00
( 500)
4.95
( 495)
5.00
( 500)
5.00
( 500)
5.00
( 500)
1.99
( 199)
4.47
( 447)
4.93
( 493)
0.40
( 40)
4.72
( 472)
4.63
( 463)
3.61
( 361)
$
( o
$
( o
$
( o
$
( o
$
( o
$
( o
$
( o
$
( 4
$
( o
$
( o
$
( 8
$
( 0
$
( 0
$
( 3
Cost to
Replace
Highways
1
.52
0
.22
0
.00
0
.05
0
.09
0
.31
0
.00
10
.17
1
.70
0
.31
22
.81
1
.57
0
.35
10
.96
.32 $
mi )
.55 $
mi)
.00 $
mi)
.11 $
mi)
.22 $
mi)
.78 $
mi)
.00 $
mi)
.56 $
mi)
77
-------
Weggel
TABLE 27 (cont.) THE COST OF SEA LEVEL RISE AT 84 SITES ALONG THE
U.S. COASTLINE - LOW ESTIMATE
Site Cost of New
Bulkheads
SCC AH RLE $
<
SCBROOKG $
<
GASEAISL $
(
FLCARDSO $
(
FLLOSTMA $
(
FLFTGADS $
(
FLAPALAC $
(
FLMIAMI $
(
FLVENICE $
FLSTAUGU $
FLEVERGL $
FLCAPECA $
FLSNIPEI $
FLKEYWES $
(
21.96
31.99 mi)
2.56
3.73 mi)
2.84
4.13 mi)
0.13
0.19 mi)
0.00
0.00 mi)
0.74
1.07 mi)
2.88
4.20 mi)
65.02
94.73 mi)
1.21
1.76 mi)
11.14
16.23 mi)
1.20
1.75 mi)
93.40
136.08 mi)
0.00
0.00 mi)
0.00
0.00 mi)
$
$
$
$
$
$
$
$
$
$
$
$
$
$
Cost to
Raise
Bulkheads
36.47
( 38.35
4.24
( 4.46
4.73
( 4.97
0.21
( 0.22
0.00
( 0.00
1.22
( 1.28
4.85
( 5.10
108.72
(114.34
2.69
( 2.82
18.62
( 19.58
1.98
( 2.08
151.28
(159.10
0.00
( 0.00
0.00
( 0.00
$
mi)
$
mi)
$
mi)
$
mi)
$
mi )
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
Cost to
Move
Buildings
3.28
( 328)
4.59
( 459)
4.73
( 473)
4.98
( 498)
5.00
( 500)
4.92
( 492)
4.58
( 458)
0.25
( 25)
4.66
( 466)
3.42
( 342)
4.92
( 492)
0.13
( 13)
5.00
( 500)
5.00
( 500)
$
<
$
<
$
<
$
<
$
<
$
<
$
<
$
<
$
<
$
(
$
(
$
<
$
(
$
(
Cost to
Replace
Highways
10
4.22
1
0.74
1
0.44
0
0.05
0
0.00
0
0.22
0
0.35
22
8.73
0
0.35
3
1.52
1
0.44
28
11.34
0
0.00
0
0.00
.69
mi)
.87
mi )
.11
mi)
.11
mi)
.00
mi)
.56
mi)
.88
mi)
.13
mi)
.89
mi)
.85
mi)
.11
mi)
.74
mi)
.00
mi)
.00
mi)
$
$
$
$
$
$
$
$
$
$
$
$
$
$
Total
Cost
72
13
13
5
5
7
13
196
9
37
9
273
5
5
.39
.26
.40
.44
.00
.43
.19
.12
.45
.04
.21
.56
.00
.00
3-79
-------
Weggel
TABLE 27 (cont.) THE COST OF SEA LEVEL RISE AT 84 SITES ALONG THE
U.S. COASTLINE - LOW ESTIMATE
Site Cost of New
Bulkheads
FLHOLLEY $
FLFORTMY $
FLPORTRI $
FLSTJOSE $
ALGRAND $
MSPASSCH $
MSGULFPO $
LALULING $
LABARATA $
LAGOLDME $
LAGRANDC $
LAPELICA $
LAMAINPA $
LABELLEC $
2.42
3.52 mi)
26.44
38.52 mi)
0.00
0.00 mi)
0.00
0.00 mi)
1.51
2.20 mi)
2.01
2.93 mi)
2.95
4.29 mi)
19.53
28.45 mi)
2.87
4.18 mi)
3.51
5. 11 mi)
1.60
2.33 mi)
0.30
0.43 mi)
0.00
0.00 mi)
17.29
25.20 mi)
Cost to Cost to
Raise Move
Bulkheads Buildings
(
$
(
$
(
$
(
$
(
$
(
$
(
$
(
$
<
$
<
$
(
$
<
$
<
(
4.1!
; 4.33
43.86
; 46.13
0.00
; o.oo
0.00
: o.oo
2.51
; 2.64
3.43
; 3.60
4.95
; 5.21
32.30
; 33.97
4.75
; 5.00
5.81
; 6.11
2.64
: 2,78
0.50
[ 0.52
0.00
: o.oo
28.70
: 30.18
mi ) c
$
mi) (
$
mi ) (
$
mi) (
$
mi) (
$
mi) (
$
mi) (
$
mi) (
$
mi) (
$
mi) (
$
mi) <
$
mi) <
$
mi) <
$
mi) . <
i.70
470)
0.81
81)
5.00
500)
5.00
500)
4.65
465)
4.55
455)
3.69
369)
3.62
362)
4.83
483)
4.82
482)
4.89
489)
4.98
498)
5.00
: 500)
4.33
: 433)
$
( 0
$
( 5
$
( o
$
( o
$
( o
$
( o
$
( 0
$
( 6
$
( o
$
( 1
$
( o
$
( o
$
( o
$
( 3
Cost to
Replace
Highways
0
.26
13
.52
0
.00
0
.00
0
.27
0
.22
0
.35
16
.52
1
.69
2
.00
1
.52
0
.05
0
.00
9
.56
.66
mi)
.98
mi)
.00
mi)
.00
mi)
.67
mi)
.55
mi)
.89
mi)
.51
mi)
.76
mi)
.53
mi)
.32
mi)
.11
mi)
.00
mi)
.03
$
$
$
$
$
$
$
$
$
$
$
$
1
$
$
Total
Cost
11
85
5
5
9
10
12
71
14
16
10
5
5
-:.aa
.90
.10
.00
.00
.35
.54
.47
.96
.21
.67
.45
.89
.00
.35
3-80
-------
Weggel
TABLE 27 (cont.) THE COST OF SEA LEVEL RISE AT 84 SITES ALONG THE
U.S. COASTLINE - LOW ESTIMATE
Site Cost of New Cost to
Bulkheads Raise
Bulkheads
LACAMERO $ 5.51 $
( 8.03 mi) (
LAPONCHA $ 2.85 $
( 4.15 mi) i
LASULPHU $ 6.67 $
( 9.71 mi) (
LALMISER $ 1.80 $
( 2.62 mi) <
TXALLIGA $ 0.14 $
( 0.21 mi) <
TXGREENI $ 0.00 $
( 0.00 mi) (
TXPORTLA $ 18.33 $
( 26.71 mi) (
TXPALACI $ 1.65 $
( 2.40 mi) <
TXRIVIER $ 0.00 $
( 0.00 mi) <
TXSMITHP $ 3.13 $
( 4.56 mi) (
TXTIVOLI $ 4.22 $
( 6.14 mi) <
CABENICI $ 22.70 $
( 33.08 mi) (
CAANONUE $ 0.00 $
( 0.00 mi) (
CASANQUE $ 31.73 $
( 46.23 mi) <
.,
: 9
4
: 4
11
: 11
2
: 3
0
( o
0
: o
31
: 33
2
C 2
0
: o
5
: s
7
: 7
39
; 41
0
: o
54
: 57
1 J
.62
.70
.94
.02
.59
.97
.12
.24
.25
.00
.00
.42
.05
.75
.90
.00
.00
.20
.47
.03
.40
.00
.01
.00
.00
.98
.83
5
mi )
$
mi )
$
mi )
$
mi)
$
mi )
$
mi)
$
mi )
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
Cost to
Move
Buildings
4.71
( 471)
4.45
( 445)
4.09
( 409)
4.79
( 479)
4.99
( 499)
5.00
( 500)
3.62
( 362)
4.84
( 484)
5.00
( 500)
4.60
( 460)
4.57
( 457)
4.14
( 414)
5.00
( 500)
2.13
( 213)
$
<•
$
<
$
<
$
<
$
<
$
<
$
<
$
<
$
<
$
<
$
<
$
<
$
<
$(
(
I.
1.
2.
0.
0.
0.
1.
0.
0.
0.
0.
2.
0.
3..
Cost to
Replace
Highways
2
04
3
31
6
39
2
96
0
05
0
00
4
95
0
22
0
00
1
57
1
61
6
42
0
00
8
44
.63
mi )
.31
mi )
.07
mi )
.43
mi )
.11
mi)
.00
mi)
.95
mi)
.56
mi)
.00
mi)
.44
mi )
.55
mi)
.14
mi)
.00
mi)
.71
mi)
$
$
$
$
$
$
$
$
$
$
$
$
$
$
Total
Cost
22
15
27
11
5
5
58
9
5
14
17
71
5
97
.00
30
.85
.98
.49
.00
.33
.80
.00
.37
.37
.98
.00
.56
3-81
-------
Weggel
TABLE 27 (cont.) THE COST OF SEA LEVEL RISE AT 84 SITES ALONG THE
U.S. COASTLINE - LOW ESTIMATE
Site Cost of New Cost to
Bulkheads Raise
Bulkheads
CAOCEANS $
CAPTSAL $
CAPALOAL $
CAALBION $
CAFERNDA $
CATIJUAN $
CAPTMAGU $
ORPORTOR $
ORYAQUIN $
WAANACOR $
WAGARDNI $
WANEMAH $
WAPORTGA $
HATACOMA $
0.00 $
0.00 mi) (
0.00 $
0.00 mi.) (
26.46 $
38.55 mi) (
0.00 $
0.00 mi) (
3.25 $
4.73 mi) <
40.81 $
59.46 mi) <
0.00 $
0.00 mi) <
0.00 $
0.00 mi) <
0.00 $
0.00 mi) <
0.00 $
0.00 mi) (
0.00 $
0.00 mi) <
12.19 $
17.77 mi) i
1.12 $
1.63 mi)
0.00 $
0.00 mi)
0.00 $
0.00 mi)
0.00 $
0.00 mi)
43.82 $
; 46.08 mi)
0.00 $
; 0.00 mi)
5.58 $
; 5.87 mi)
67.34 $
; 70.82 mi)
0.00 $
: 0.00 mi)
0.00 $
: 0.00 mi)
0.00 $
: 0.00 mi)
0.00 $
C 0.00 mi)
0.00 $
( 0.00 mi)
20.33 $
( 21.38 mi)
3.63 $
( 3.82 mi)
0.00 $
( 0.00 mi)
Cost to Cost to
Move Replace
Buildings Highways
5.00
( 500)
5.00
( 500)
2.58
( 258)
5.00
( 500)
4.48
( 448)
0.20
( 20)
5.00
( 500)
5.00
( 500)
5.00
( 500)
5.00
( 500)
5.00
( 500)
3.97
( 397)
4.81
( 481)
5.00
( 500)
$ 0.08 $
( 0.03 mi)
$ 0.00 $
( 0.00 mi)
$ 17.07 $
( 6.74 mi)
$ 0.00 $
( 0.00 mi)
$ 0.88 $
( 0.35 mi)
$ 78.60 $
( 31.01 mi)
$ 0.00 $
( 0.00 mi)
$ 0.33 $
( 0.13 mi)
$ 0.11 $
( 0.05 mi)
$ 0.44 $
( 0.18 mi)
$ 0.33 $
( 0.13 mi)
$ 4.62 $
( 1.82 mi)
$ 1.98 $
( 0.78 mi)
$ 0.78 $
( 0.31 mi)
Total
Cost
5.08
5.00
89.93
5.00
14.19
186.95
5.00
5.33
5.11
5.44
5.33
41.11
11.54
5.78
3-§2
-------
Weggel
TABLE 28 THE COST OF SEA LEVEL RISE AT 84 SITES ALONG THE
U.S. COASTLINE - HIGH ESTIMATE
Site Cost of New
Bulkheads
MEFREEPO $
(
MEROCKLA $
(
MEJONESP $
(
MAMARBLE $
(
MAWESTPO $
(
MAORLEAN $
(
RIWATCHH $
(
CNBRIDGE $
(
NYBROOKL $
(
NYNARROW $
(
NYPATCHO $
(
NYSOUTHA $
(
NJDIVIDC $
(
NJLONGBE $
(
15.74
5.96 mi)
9.37
3.55 mi.)
6.15
2.33 mi)
0.00
0.00 mi)
23.03
8.72 mi)
329.27
124.72 mi)
23.91
9.06 mi)
176.46
66.84 mi)
600.28
227.38 mi)
0.00
0.00 mi)
82.85
31.38 mi)
0.00
0.00 mi)
16.90
6.40 mi)
30.92
11.71 mi)
Cost to Cost to
Raise Move
Bulkheads Buildings
$
$
$
$
$
$
$
$
$
$
$
$
$
$
2:3.30
C 7.36
16-. 15
( 4.42
10.36
( 2.83
0.00
( 0.00
39.99
( 10.93
527.91
(144.35
42.44
( 11.61
244.44
( 66.84
1044.02
(285.47
0.00
( 0.00
142.11
( 38.86
0.00
( 0.00
28.11
( 7.69
51.85
( 14.18
$
mi )
$
mi )
$
mi)
$
mi )
$
mi )
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
4.60
( 460)
4.55
( 455)
4.88
( 488)
5.00
( 500)
4.32
( 432)
0.12
( 12)
4.21
( 421)
0.00
( 0)
0.56
( 56)
5.00
( 500)
2.05
( 205)
5.00
( 500)
4.50
( 450)
3.63
( 363)
$
$
$
$
$
$
$
$
$
$
$
$
$
$
Cost to
Replace
Highways
1
0.44
0
0.26
0
0.18
2
0.87
2
0.65
37
11.16
2
0.74
18
5.56
57
17.09
22
6.74
7
2.29
3
0.91
2
0.74
3
1.00
.48
mi)
.88
mi )
.62
mi)
.93
mi)
.20
mi)
.73
mi)
.49
mi)
.80
mi)
.75
mi)
.78
mi)
.76
mi)
.09
mi)
.51
mi)
.38
mi)
$
$
$
$
$
$
$
$
$
$
$
$
$
$
Total
Cost
48
30
22
7
69
895
73
439
1702
27
234
8
52
89
.71
.96
.01
.93
.54
.02
.05
.70
.61
.78
.77
.09
.01
.78
3-83
-------
Weggel
TABLE 28 (oont.) THE COST OF SEA LEVEL RISE AT 84 SITES ALONG THE
U.S. COASTLINE - HIGH ESTIMATE
Site
DEREHOBA
MDEASTON
MDCOVEPT
MDELKTON
MDMIDDLE
VACOLBEA
VABLOXOM
VANEWPOR
VAWILTON
NCENGELH
NCWILMIN
NCLONGBA
NCCAMPLE
SCHILTON
Cost of New
Bulkheads
$ 16.49
( 6.25 mi)
$ 7.80
( 2.95 mi)
$ 0.00
( 0.00 mi)
$ 0.69
( 0.26 mi)
$ 0.00
( 0.00 mi)
$ 0.00
( 0.00 mi)
$ 0.00
( 0.00 mi)
$ 142.90
( 54.13 mi)
$ 25.16
( 9.53 mi)
$ 1.61
( 0.61 mi)
$ 290.46
(110.02 mi)
$ 12.13
( 4.60 mi)
$ 11.73
( 4.44 mi)
$ 95.00
( 35.98 mi)
Cost to Cost to
Raise Move
Bulkheads Buildings
$
$
$
$
$
$
$
$
$
$
$
$
$
$
29.44
( 8.05
13.36
( 3.65
0.00
( 0.00
1.46
( 0.40
0.00
( 0.00
0.00
( 0.00
0.00
( 0.00
241.36
( 66.00
43.21
( 11.82
2.65
( 0.72
488.89
(133.68
20.16
( 5.51
19.78
( 5.41
158.16
( 43.25
$
mi) (
$
mi) (
$
mi) (
mi) (
$
mi) <
$
mi) (
$
mi) <
$
mi) (
$
mi) <
$
mi) <
$
mi) <
$
mi) (
$
mi) <
$'
mi) <
4.40
440)
4.61
461)
5.00
500)
4.95
: 495)
5.00
: 500)
5.00
: 500)
5.00
; 500)
1.99
: 199)
4.47
: 447)
4.93
: 493)
0.40
: 40)
4.72
C 472)
4.63
C 463)
3.61
( 361)
$
( 0
$
( o
$
( 0
( 0
$
( o
$
( 0
$
( o
$
( 4
$
( o
$
( 0
$
( 8
$
( 0
$
( 0
$
( 3
Cost to
Replace
Highways
1
.52
0
.22
0
.00
0
.05
0
.09
1
.31
0
.00
14
.17
2
.70
1
.31
29
.81
1
.57
f
1
.35
13
.96
.76
mi)
.73
mi)
.00
mi )
.15
mi )
.29
mi)
.03
mi)
.00
mi)
.08
mi)
.36
mi)
.03
mi)
.77
mi)
.92
mi)
.17
mi)
.37
mi)
$
$
$
$
$
$
$
$
$
$
$
$
$
$
Total
Cost
52.
26.
5.
7.
5.
6.
5.
400.
75.
10.
809.
38.
37.
270.
11
50
00
25
29
03
00
33
20
22
52
93
31
14
3-84
-------
Weggel
TABLE 28 (cont.) THE COST OF SEA LEVEL RISE AT 84 SITES ALONG THE
U.S. COASTLINE - HIGH ESTIMATE
Site Cost of New
Bulkheads
SCC AH RLE $
SCBROOKG $
GASEAISL $
FLCARDSO $
FLLOSTMA $
FLFTGADS $
FLAPALAC $
FLMIAMI $
FLVENICE $
FLSTAUGU $
FLEVERGL $
FLCAPECA $
FLSNIPEI $
FLKEYWES $
84.46
31.99 mi)
9.85
3.73 mi)
10.91
4.13 mi)
0.50
0. 19 mi)
0.00
0.00 mi)
2.83
1.07 mi)
11.10
4.20 mi)
250.09
94.73 mi)
4.64
1.76 mi)
42.85
16.23 mi)
4.61
1.75 mi)
359.24
136.08 mi)
0.00
0.00 mi)
0.00
0.00 mi)
$
$
$
$
$
$
$
$
$
$
$
$
$
$
Cost to
Raise
Bulkheads
140
( 38
16
( 4
18
( 4
0
( 0
0
( 0
4
( 1
18
( 5
418
(114
10
( 2
71
( 19
7
( 2
581
(159
0
( 0
0
( 0
.25
.35
.30
.46
.18
.97
.82
.22
.00
.00
.68
.28
.64
.10
.16
.34
.33
.82
.61
.58
.62
.08
.85
.10
.00
.00
.00
.00
$
mi )
$
mi )
$
mi )
$
mi )
$
mi )
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
Cost to
Move
Buildings
3.28
( 328)
4.59
( 459)
4.73
( 473)
4.98
( 498)
5.00
( 500)
4.92
( 492)
4.58
( 458)
0.25
( 25)
4.66
( 466)
3.42
( 342)
4.92
( 492)
0.13
( 13)
5.00
( 500)
5.00
( 500)
$
$
$
$
$
$
$
$
$
$
$
$
$
$
4
0
0
0
0
0
0
8
0
1
0
11
0
0
Cost to
Replace
Highways
14
.22
2
.74
1
.44
0
.05
0
.00
0
.22
1
.35
29
.73
1
.35
5
.52
1
.44
38
.34
0
.00
0
.00
.25
mi)
.49
mi )
.48
mi )
.15
mi)
.00
mi )
.75
mi)
.17
mi)
.50
mi)
.19
mi)
.14
mi)
.48
mi)
.32
mi)
.00
mi)
.00
mi)
$
$
$
$
$
$
$
$
$
$
$
$
$
$
Total
Cost
242.
33.
35.
6.
5.
13.
35.
698.
20.
123.
18.
979.
5.
5.
24
23
30
45
00
17
49
00
82
03
63
54
00
00
3-85
-------
Weggel
TABLE 28 (cont.) THE COST OF SEA LEVEL RISE AT 84 SITES ALONG THE
U.S. COASTLINE - HIGH ESTIMATE
Site Cost of New
Bulkheads
FLHOLLEY $
(
FLFORTMY $
(
FLPORTRI $
(
FLSTJOSE $
(
ALGRAND $
(
MSPASSCH $
(
MSGULFPO $
(
LALULING $
(
LABARATA $
<
LAGOLDHE $
(
LAGRANDC $
(
LAPELICA $
(
LAMAINPA $
(
LABELLEC $
(
9.31
3.52 mi)
101.69
38.52 mi)
0.00
0.00 mi)
0.00
0.00 mi)
5.81
2.20 mi)
7.74
2.93 mi)
11.33
4.29 mi)
75.12
28.45 mi)
11.03
4.18 mi)
13.50
5.11 mi)
6.14
2.33 mi)
1.15
0.43 mi)
0.00
0.00 mi)
66.52
25.20 mi)
$
$
$
$
$
$
$
$
$
$
$
$
$
$
Cost to Cost to
Raise Move
Bulkheads Buildings
15.32
( 4.33
168.71
( 46.13
0.00
( 0.00
0.00
( 0.00
9.66
( 2.64
13.18
( 3.60
19.04
( 5.21
124.22
( 33.97
18.28
( 5.00
22.34
( 6.11
10.16
( 2.78
1.91
( 0.52
0.00
( 0.00
110.37
( 30.18
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
4.70
( 470)
0.81
( 81)
5.00
( 500)
5.00
( 500)
4.65
( 465)
4.55
( 455)
3.69
( 369)
3.62
( 362)
4.83
( 483)
4.82
( 482)
4.89
( 489)
4.98
( 498)
5.00
( 500)
4.33
( 433)
$
(
$
(
$
(
$
(
$
(
$
(
$
(
$
(
$
(
$
(
$
(
$
(
$.
(
$
(
Cost to
Replace
Highways
0
0.26
18
5.52
0
0.00
0
0.00
0
0.27
0
0.22
1
0.35
22
6.52
2
0.69
3
1.00
1
0.52
0
0.05
0
0.00
12
3.56
.88
mi )
.64
mi)
.00
mi)
.00
mi)
.90
mi)
.73
mi)
.19
mi)
.02
mi)
.34
mi)
.38
mi)
.76
mi)
.15
mi)
.00
mi)
.04
mi)
$
$
$
$
$
$
$
$
$
$
$
$
$
$
Total
Cost
30.
289.
5.
5.
21.
26.
35.
224.
36.
44.
22.
8.
5.
193.
71
85
00
00
03
20
24
98
49
04
95
19
00
26
3-86
-------
Weggel
TABLE 28 (cont.) THE COST OF SEA LEVEL RISE AT 84 SITES ALONG THE
U.S. COASTLINE - HIGH ESTIMATE
Site Cost of New
Bulkheads
LACAMERO $
(
LAPONCHA $
(
LASULPHU $
(
LALMISER $
(
TXALLIGA $
(
TXGREENI $
(
TXPORTLA $
(
TXPALACI $
(
TXRIVIER $
(
TXSMITHP $
(
TXTIVOLI $
(
CABENICI $
(
CAANONUE $
(
CASANQUE $
(
21.19
8.03 mi)
10.94
4.15 mi)
25.64
9.71 mi)
6.91
2.62 mi)
0.56
0.21 mi)
0.00
0.00 mi)
70.52
26.71 mi)
6.33
2.40 mi)
0.00
0.00 mi)
12.03
4.56 mi)
16.21
6.14 mi)
87.33
33.08 mi)
0.00
0.00 mi)
122 . 04'
46.23 mi)
$
$
$
$
$
$
$
$
$
$
$
$
$
$
Cost to Cost to
Raise Move
Bulkheads Buildings
35.19
( 9.62
18.08
( 4.94
42.39
( 11.59
11.41
( 3.12
0.92
( 0.25
0.00
( 0.00
120.85
( 33.05
10.59
( 2.90
0.00
( 0.00
19.99
( 5.47
27.05
( 7.40
149.99
( 41.01
0.00
( 0.00
.211.47
( 57.83
$
mi)
$
mi )
$
mi)
$
mi)
$
mi)
$
mi )
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
4.71
( 471)
4.45
( 445)
4.09
( 409)
4.79
( 479)
4.99
( 499)
5.00
( 500)
3.62
( 362)
4.84
( 484)
5.00
( 500)
4.60
( 460)
4.57
( 457)
4.14
< 414)
5.00
( 500)
2.13
( 213)
$
( 1
$
( 1
$
( 2
$
( 0
$
( o
$
( o
$
< 1
$
( o
$
( o
$
( o
$
( o
$
( 2
$
( o
$
( 3
Cost to
Replace
Highways
3
.04
4
.31
8
.39
3
.96
0
.05
0
.00
6
.95
0
.22
0
.00
1
.57
2
.61
8
.42
0
.00
11
.44
.51
mi)
.41
mi)
.09
mi)
.24
mi)
.15
mi)
.00
mi)
.60
mi)
.75
mi)
.00
mi)
.92
mi)
.07
mi)
.18
mi)
.00
mi)
.62
mi)
$
$
$
$
$
$
$
$
$
$
$
$
$
$
Total
Cost
64
37
80
26
6
5
201
22
5
38
49
249
5
347
.60
.88
.21
.34
.62
.00
.59
.51
.00
.54
.91
.64
.00
.26
-------
Weggel
TABLE 28 (cont.) THE COST OF SEA LEVEL RISE AT 84 SITES ALONG THE
U.S. COASTLINE - HIGH ESTIMATE
Site Cost of New
Bulkheads
CAOCEANS $
CAPTSAL $
CAPALOAL $
CAALBION $
CAFERNDA $
CATIJUAN $
CAPTMAGU $
ORPORTOR $
ORYAQUIN $
WAANACOR $
WAGARDNI $
WANEMAH $
WAPORTGA $
WATACOMA $
0.00 $
0.00 mi)
0.00 $
0.00 mi)
101.76 $
38.55 mi)
0.00 $
0 .00 mi )
12.49 $
4.73 mi)
156.97 $
59.46 mi)
0.00 $
0.00 mi)
0.00 $
0.00 mi)
0.00 $
0.00 mi)
0.00 $
0.00 mi)
0.00 $
0.00 mi)
46.90 $
17.77 mi)
4.30 $
1.63 mi)
.• o.oo . $
0.00 mi)
Cost -.0
Raise
Bulkheads
0.
( 0 .
0.
( 0.
168.
( 46.
0.
( 0.
21.
( 5.
258.
( 70.
0.
( o.
0.
( 0.
0.
( o.
0>.
( 0.
0.
( o.
78.
( 21.
13.
( 3.
, 0.
( o.
00
00
00
00
54
08
00
00
47
87
98
82
00
00
00
00
00
00
00
00
00
00
19
38
95
82
00
00
&
$
mi )
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
$
mi)
mi)
$
mi)
$
•mi)
$
mi)
Cost to
Move
Buildings
5 . 00
( 500)
5.00
( 500)
2.58
( 258)
5.00
( 500)
4.48
( 448)
0.20
( 20)
5.00
( 500)
5.00
( 500)
5.00
( 500)
5.00
( 500)
5.00
( 500)
3.97
( 397)
4.81
( 481)
5.00
( 500)
$
(
$
(
$
(
$
(
$
(
$
(
$
(
$
(
$
(
$
(
$
(
$
(
$
(
$
(
Cost to
Replace
Highways
0
0.03
0
0.00
22
6.74
0
0.00
1
0.35
104
31.01
0
0.00
0
0. 13
0
0.05
0
0.18
0
0.13
6
1.82
2
0.78
* 1
0.31
.11
mi)
.00
mi)
.76
mi)
.00
mi)
.17
mi)
.80
mi)
.00
mi)
.44
mi)
.15
mi)
.59
mi)
.44
mi)
.16
mi)
.65
mi)
.03-
mi)
$
$
$
$
$
$
$
$
$
$
$
$
$
$
Total
Cost
5
5
295
5
39
520
5
5
5
5
5
- 135
25
6
.11 '
.00
.64
.00
.61
.96
.00
.44
.15
.59
.44
.22
.72l
.03
3-88
-------
Weggel
TABLE 29 NATIONWIDE ESTIMATE ($ millions)
Low High
Northeast 6,932 23,607
Mid Atlantic 4,354 14,603
Southeast 9,249 29,883
West 4,097 12,802
USA 24,633 80,176
3-89
-------
Weggel
REFERENCES
Bruun, P. (1962). "Sea Level Rise as a Cause of Erosion," Journal of the Waterways and Harbors Division,
ASCE, Vol. 88, No. WW1, February 1962, pp. 117-130.
Fairbridge, R.W. (1961). "Eustatic changes in sea level. Physics and Chemistry of the Earth, Vol. 4, pp. 99-185
Fairbridge, R.W. & OA. Krebs (1962). Sea Level and the Southern Oscillation," Geophysical Journal, Royal
Astron. Proceedings, Vol. 6, No. 4, pp. 532-545.
Gornitz, V., S. Lebedeff, & J. Hansen (1982). "Global sea level trends in the past century," Science, Vol. 215,
pp. 1611-1614.
Hoffman, (1985). "Estimates of Future Sea Level Rise" in Greenhouse Effect and Sea Level Rise, M.C. Earth
& J.G. Titus, editors, Van Nostrand Reinhold Co., Inc., New York, NY.
Lyie et al. (1987). "Sea Level Variations in the U.S.," National Ocean Survey, Rockville, MD, 1987.
Meier, M.F. et al. (1985). Glaciers, Ice Sheets and Sea Level, National Academy Press, 1985.
National Research Council (1987). Responding to Changes in Sea Level, Engineering Implications, Committee
on Engineering Implications of Changes in Relative Mean Sea Level, Marine Board, National Academy Press
Washington, DC.
Revelle, R. (1983). "Probable Future Changes in Sea Level Resulting from Increased Atmospheric Carbon
Dioxide," in Changing Climate, Carbon Dioxide Assessment Committee, Washington, DC, National Academy
Press, pp. 433-447.
Titus, J. (1986). "Greenhouse Effect, Sea Level Rise and Coastal Zone Management," Coastal Zone
Management Journal, Vol. 14, No. 3.
Soil Conservation Service (1986). "Urban Hydrology for Small Watersheds," Technical Release No. 55, U.S.
Department of Agriculture, Soil Conservation Service, June 1986.
3-90
-------
THE COST OF NOT HOLDING BACK THE SEA PHASE 1
ECONOMIC VULNERABILITY
by
Gary W. Yohe
Department of Economics
Wesleyan University
Middletown, CT 06457
Cooperative Agreement No. CR-814927-01-0
-------
CONTENTS
FINDINGS 4-1
CHAPTER 1: THE THEORY BEHIND MEASURING VULNERABILITY 4-2
THE VALUE OF THREATENED STRUCTURE 4-2
THE VALUE OF THREATENED PROPERTY 4-3
THE SOCIAL VALUE OF THREATENED COASTLINE 4-4
CHAPTER 2: VULNERABILITY FOR LONG BEACH ISLAND, NEW JERSEY 4-7
CHAPTER 3: DISCUSSION 4-12
CHAPTER 4: EXTENSIONS TO A NATIONAL STUDY 4-19
CHAPTER 5: EXTENSIONS FROM VULNERABILITY TO COST 4-20
REFERENCES 4-22
11
-------
Yohe
FINDINGS1
The first step in estimating the cost to the United States of allowing the oceans to rise in response to
greenhouse warming against unprotected coastlines is to develop a methodology by which researchers can catalog
and measure the current value of real sources of economic wealth that might be threatened. Such measures
represent initial, if naive, estimates of the social cost that would be incurred at each site if a decision to forego
any protection from rising seas were made. If the sites chosen for application of the methodology are also part
of a national sample, the localized estimates that they support can eventually be used to judge the potential cost
of a universally applied decision of no protection. They can, in other words, be used to produce a first cut at
a measure of economic vulnerability across the United States to greenhouse-induced sea level rise.
This paper reports on the first steps of a process which will lead to this national estimate. The first
three chapters are designed to outline the methodology by which site-specific cost estimates were made for Long
Beach Island, New Jersey, and to record the results of its application. In the first, the underlying theory of the
measurement is described. There are three areas of focus: the value of threatened structure, the value of
threatened property, and, where appropriate, the social value of threatened coastline. The results of applying
the theory to Long Beach Island are recorded in Chapter 2, while discussion found in the third chapter tries to
put these local results into some perspective.
Broader perspective is drawn in the last two chapters. Chapter 4 begins the extension by describing
more fully the sampling methodology by which local estimates of vulnerability can lead to a national estimate.
The site selection process and more generally applicable estimation procedures are of particular interest there.
A final chapter concludes with an outline of the issues that will have to be confronted if measures of economic
vulnerability are to be translated into measures of economic cost. Discounting, uncertainty, growth, depreciation,
and frictional adjustment costs will all have to be considered; identifying their precise role in the translation
process certainly will be the focus of subsequent research.
'Although the information in this report has been funded wholly or partly by the UJS. Environmental
Protection Agency under Cooperative Agreement No. CR-814927-01-0, it does not necessarily reflect the Agency's
views, and no official endorsement should be inferred from it. The author expresses appreciation for comments
offered and direction given by two anonymous contributors to the peer review process. They contributed to the
quality of the exercise, but -not to any errors that might remain.
4-1
-------
Yohe
CHAPTER 1
THE THEORY BEHIND MEASURING VULNERABILITY
The cost of not holding back the sea should flow from at least four separate sources: (1) the value of
lost structure, (2) the value of lost property, (3) the value of lost social "services" delivered from the existing
coastline, and (4) adjustment costs associated with redeploying productive resources once applied to the lost land.
The present effort considers only the first three of these sources, postponing any thorough consideration of the
frictional costs of redeployment until later.2 They relate more to immediate measures of vulnerability; the last
relate more to adjustments that will be required to translate vulnerability to cost.
Land and structures are, for example, stores of economic wealth; even threatening their loss would likely
produce macroeconomic reductions in aggregate demand, the effects of which would extend well beyond the
shoreline. The extent of their potential contractionary influence on economic activity and long-term growth is
thus another concern which should be investigated when the dimension of likely shoreline loss is more fully
understood.3 Social services provided by coastlines are, similarly, major components of economic and social well-
being for which people have demonstrated a significant and immediate willingness to pay. Each of the first three
sources of potential cost is, therefore, significant and deserving of individual attention.
THE VALUE OF THREATENED STRUCTURE
The precise notion employed to compute the economic value of threatened structure is that people will
abandon a structure when the land upon which it sits is covered by water during mean spring high tide. In fact,
the inundation scenarios upon which the vulnerability calculations are based are not sufficiently detailed to apply
that notion exactly. The shoreline retreat scenarios provided by Park et al. (this volume) indicate, for each site
in the national sample, only the percentages of developed cells (usually 500 meters square) that are flooded when
the seas rise 50 cm, 100 cm, and 200 cm through 2100. In practice, therefore, the percentage of structure
currently located in each cell and deemed abandoned with each increment of sea level rise must be taken to be
the percentage of that cell that is flooded.
More precisely, the current value of structure located within any specific cell can be estimated from tax
records or housing and business census data on the basis of a sample of structures presently located within its
boundaries. To be sure, neither tax records nor census data necessarily reflect current market value. A
reasonable translation from recorded value to current market value can, however, be accomplished by noting (1)
the percentage of market value reported by the assessor's office, and (2) some degree of inflation since the last
assessment. The accuracy of the translation can, in addition, be validated by comparing the assessed values of
structures now on the market with their quoted prices. Moving to an estimate of the value of threatened
structure within that cell can then be accomplished using the percentages indicated by the inundation scenarios.
See Chapter 5 for a brief discussion of the anticipated role of adjustment cost in the process which translates
economic vulnerability to economic cost. Adequate treatment of frictional adjustment costs will involve more
sophisticated intertemporal modeling.
Real estate markets are assumed to be efficient, so the economic value of public goods and services which
are also threatened by inundation is capitalized in the values of land and structure. No separate accounting of
public goods and services is therefore necessary. No notion of critical mass is employed, as a result, so some
early vulnerability estimates for regions which will essentially disappear may be too low; they will capture the
total loss of the value of public activity only when the last piece of property is lost even though, in fact, public
activity probably stopped years earlier.
4-2
-------
Yohe
If, for example, a 50-cm sea level rise is expected to put x% of the region under water by the year 2075, then
it can be assumed that x% of the estimated value of the structure located in that region is lost by 2075. Adding
across all threatened cells can Finally produce a site-specific cost estimate of potential structure loss.
One sampling procedure upon which the estimation process can rest looks at strips of land running
inland from the shoreline past an inland point at which (1) property and structure are no longer threatened by
sea level rise, and (2) property values no longer reflect surplus location rent derived from proximity with the
shore. Series of real estate valuations along these strips should be sufficient to support aggregate potential cost
estimates subject, of course, to some sampling error. Sampling error could be avoided completely if the
inundation scenarios were more detailed and if tax records were digitized, but neither of these conditions is met
in reality. Resulting estimates must rely, instead, on the efficient operation of real estate markets to keep the
sampling errors low; a small number of strips in each sample should, in fact, be sufficient to keep the t-statistics
around sample means of (e.g.) structure values, in excess of 10.
The technicalities of sampling aside, a procedure which uses current value as a measure of potential
future cost can be criticized for several reasons. For one thing, the sites being studied will surely enjoy economic
growth over the next half century or so. Current value misses that growth entirely. For another, structure prices
tend to inflate more quickly than the general Consumer Price Index. Estimates based on current value might,
therefore, be conservative to the degree that they ignore either or both of these phenomena
On the other hand, using current value sidesteps both the vagaries of social discounting and the potential
that threatened structures will be allowed to fall into disrepair when it becomes known that they may be under
water in the foreseeable future. Inasmuch as the cost of not holding back the sea will be compared with the
cost of protection on a year-to-year (or decade-to-decade) basis as various future scenarios unfold, however, the
problems created by not discounting are not necessarily as severe as they might at first appear. They may involve
discounting over a decade's time, for example, and not over a half-century. Moreover, it may turn out that the
growth and relative inflation trends just noted proceed over the long term at a rate which roughly offsets the
effect of discounting on the real value of threatened structure. Current value and present value would then
match over the long term if not over decades.
The issue of not maintaining structure is also one of timing. For example, if the owner of a $200,000
structure that will be inundated in the year 2050 were to ignore its physical upkeep over the 25-year period from
the year 2025 to 2050, then the owner would suffer a smaller loss in 2050 than he would otherwise. How much
smaller? The present value, in 2050, of the money that he did not spend maintaining the property since the year
2025 net of the reduced rent that he received as the property deteriorated. If, however, it were known that the
structure were going to be abandoned in 2050, then the market value of that structure would begin to decline
well before 2050. An accurate accounting of the economic loss might therefore also start recording this decline
in value years ahead of the 2025 collapse, thereby moving the loss forward and increasing its current present
value. Which effect would dominate is, at this point, anybody's guess; but it is certainly an issue which warrants
further consideration.
All of these intertemporal issues will be considered, when vulnerability measures are adjusted to reflect
cost, with an eye toward keeping track of precisely "Who knows what and when?" Discounting must, for
example, be considered to the extent that decisions to protect ponder investment at some time certain in
anticipation of avoiding loss sometime in the future. Its implications will be clear, however, only in the context
of modeling, which also allows for economic growth, depreciation, market expectations, and uncertainty. For the
moment, it must be emphasized that only current value estimates are provided here.
THE VALUE OF THREATENED PROPERTY
The same sampling procedure outlined above can also be used to produce estimates of the current value
of lost property which are subject to virtually the same set of concerns. To the degree that current values miss
4-3
-------
Yohe
the effects of higher relative inflation, they likely to be too low. To the degree that they are not discounted, they
are likely to be too high. Market value erosion might also be expected; it would be based on the same rational
response to anticipated inundation, and it would happen automatically through the operation of the marketplace.
In fact, the only caveat that no longer applies is the analog to an owner's ability to run down a structure. The
value of the land upon which something might be built cannot be significantly diminished by neglect. It may
become unsightly, but the marketplace will continue to acknowledge its intrinsic value derived from location and
other relatively unalterable characteristics.
There is, however, one additional wrinkle that must be considered — exactly what piece of property is
lost when the sea rises? For structures, the answer to this question is simple; the structure that is abandoned
is the one that is lost. For property, though, loss of a shoreline lot means that the next lot is now a shoreline
lot. Economic loss should, therefore, be measured at some interior point.
To see this more precisely, consult Figure 1; a hypothetical property value gradient for one-eighth acre
lots is displayed there. Note that values start at $100,000 on the shoreline and eventually stabilize at $50,000
some 500 feet from the shoreline. Were the sea to rise so that the first lot were lost, then the second lot would
become a shoreline lot and assume the $100,000 value originally attributed to the first. The value of the third
lot would climb to $90,000, and so on. The community would, in effect, lose the economic value of an interior
lot located initially more than 500 feet from the shoreline. The true economic loss would be the equivalent of
a $50,000 lot instead of the shoreline $100,000 lot; there would be a distributional effect, to be sure, but the net
social loss would be $50,000. Where appropriate and accessible, this sort of accounting procedure can be applied
in the property value loss calculations. The strip sampling method is, in fact, specifically designed to provide
enough information to support its application. Note, as well, that the interior valuation process works from all
directions for an island. The value of an interior plot of land can, as a result, rise, at least for a while. Proper
sampling design for an island therefore involves looking at strips that run its entire length or width.
THE SOCIAL VALUE OF THREATENED COASTLINE
The final source of potential economic loss from sea level rise can be traced to the social value of the
coastline that may be lost. Beaches are recreational areas, for example, which are generally available for use at
the price of a beach badge; estimation of even their recreational value is therefore extremely difficult. The
literature, building on work by Clawson (1966), suggests using transportation cost to construct at least a partial
measure of value. More specifically, if using the beach is essentially free except for the cost of getting there and
getting home, then the prices that families (e.g.) pay to use the beach are simply equal to the expenses that they
incur simply getting to the beach and getting back home. Use surveys can then be employed to construct a
demand curve for beach services by matching these prices with quantities demanded (people living various
distances from the beach pay different prices to enjoy its services). The contribution of the beach to general
social welfare can then be taken to be the usual consumer surplus area under this demand curve.
There are, of course, an array of other benefits generated by our coastlines which are not captured by
this travel cost measure, and the problem of estimating the cost of losing a coastline region is one of measuring
the value of all of these benefits. One approach that showed some promise in moving toward a more general
measure was developed by Knetsch (1964) and David (1968). They both noted that property values increase with
proximity to a recreation area like a beach. Since these increases reflect, quite simply, a willingness to pay for
the general amenities provided by a beach, e.g., Knetsch and David argued that the sum of these increases could
be employed as a measure of the value of that beach. As the beach disappears, then, the economic cost might
be estimated by keeping track of the losses in proximity-generated surplus economic rents.
There are, however, several difficulties in applying the Knetsch-David notion directly. Some of the
amenity, and thus some of the slope in a property value gradient, comes from views of the ocean that please
residents with or without a beach. Attributing the entire slope to the beach proximity would therefore produce
an overestimate of beach value. On the other hand, there are many people who do not live near the beach but
4-4
-------
Yohe
A
8
8
8
X
(i
H
3
4»
&
(I
0
i
100
90
80
70 ^
60
50
40
YH
•*<
Ualue
Distance to Shove
Figure 1. Distance to shore.
4-5
-------
Yohe
who nonetheless use the beach. Using a property value gradient exclusively would miss the value of the beach
services that they enjoy, and would thus produce an under-estimate of beach value. Finally, there is considerable
storm protection value provided to inland property by a beach and its associated dune structure which is captured
by neither transportation cost surveys nor property value gradients. Still, a rough Knetsch-David style estimate
can provide context ~ an order of magnitude guess against which to judge more careful estimates derived in
other ways.
The alternative procedure employed here attempts to account for all of the sources of value to the
degree actually recognized by shoreline communities by judging beach value from community behavior when
beaches are threatened. As a matter of law, in some places like Texas (Texas Open Beach Act), and of practice,
in other places like New Jersey and North Carolina, a structure located along a beachfront must be abandoned
and/or torn down when the land upon which it sits is inundated during the mean spring high tide. This allows
the beach and presumably its dune to migrate inland, albeit at the expense of the property owner whose structure
was in the way, but to the good of the inland community. By revealed preference, therefore, the social value of
a beach must be at least as high as the value of beachfront structures which would be abandoned if the beach
were to erode. It is, in other words, reasonable to assume that a beachfront structure is sacrificed to preserve
the social value of coastline whenever a sea level rise scenario brings the water within a certain minimum
distance of its foundation. Titus and Greene (this volume) submit that that minimum width is 40 feet.
Refer again to Figure 1 to see how this procedure might work operationally. Suppose, for the sake of
argument, that $200,000 structures were located on each lot and that there were a 40-foot beach on the ocean
side of the first lot. Recall that the lots are all 100 feet long moving away from the water. Now let the ocean
rise, eroding 100 feet of beach and dune. What has been the cost? Any structure on the first lot is now within
40 feet of the ocean. To maintain the minimum beach width, therefore, that structure must be abandoned and
perhaps torn down; the loss, attributable to the social value of the beach, is thus at least $200,000 derived from
the lost structure. What about the property? An additional $25,000, representing half of the property value of
an interior lot, has been lost, as well, because half of the first lot is gone. Should this loss be added to the
property loss accounting outlined in the previous subsection, or should it be attributed to the beach value
accounting just noted? Ultimately, the answer to this question does not matter as long as it is not added in both
places. Total vulnerability is, after all, the sum of the losses attributed to structure, coastline, and property. To
emphasize the importance of preserving the social services provided by coastline, though, the accounting
procedure adopted here attributes all property and structure loss associated with maintaining a coastline to the
value of preserving that coastline.
Presumably the value of the next lot has increased according to the earlier story.
4-6
-------
Yohe
CHAPTER!
VULNERABILITY FOR LONG BEACH ISLAND, NEW JERSEY
Estimates of economic vulnerability for Long Beach Island were prepared from a systematic sampling
of assessed property and structure values along 25 separate strips of land. Two of the strips were designed to
sample from atypical developments on the bay side of the northern part of the island. The remaining 23 were
each approximately 200 feet wide, evenly distributed along the 18-mile length of the island and extending from
the ocean to the bay; they were designed to sample from the more traditional development pattern of the
majority of the island. Table 1 identifies the sample sites.
The general cross-sectional topography of the island, and thus of 23 of the 25 strips, is portrayed in
Figure 2. There was some variation in development pattern. The north shows big houses on large lots and
located well away from wide beaches; the south shows smaller houses on smaller lots packed up against narrower
beaches. Nonetheless, their remarkable consistency made it possible to extrapolate inundation scenarios for each
strip into integrated inundation scenarios for the entire island.
Beginning on the bay side, significant inundation will usually begin after a 1-foot rise; there are places
where the bulkhead is a bit higher, but rarely could it restrain more than a 3-foot rise. Once begun, inundation
will proceed quickly over the virtually flat area located between the bay and Long Beach Boulevard. On the
ocean side of the Boulevard, the rate of inundation will slow as elevations rise more quickly, but it will by no
means stop until the island is completely underwater. Ten feet above mean high tide is the usual maximum
altitude of developed property at the base of the ocean-side protecting dunes.
Turning now to the ocean side, 100 feet of beach is lost on Long Beach Island for every 1 foot of sea
level rise (Weggel et. al., this volume). Since the beach is less than 50 feet wide in some spots, particularly on
the south end of the island with houses build up the inland sides to the tops of the dunes, maintaining the beach
for social value will involve some economic loss even with a 6-inch rise. The cost accelerates until, at about 4
feet of sea level rise, nearly 75% of the $2 billion value of the island is lost.
With inundation boundaries defined along each strip of the sample (and, by interpolation, along the
entire length of the island) for 6-inch, 1-foot, 18-inch, 2-foot, 3-foot, 4-foot, and 6-foot sea level rise scenarios,
it remained only to estimate the property, structure, and beach values threatened by each step of the process
according to the procedure outlined in Chapter I. Estimates for both property and structure, normalized per
eighth-acre lots, were produced directly from recent tax maps and a complete grand list for each level of
inundation within each sampling strip. A comparison between asking price and assessed value for properties
currently listed in the real estate market revealed a close match; no disparities of more than 10% were
discovered, and no consistent bias in either direction was noted. Moving from these sampling estimates to
property, structure, and beach value estimates for the entire island was finally accomplished by extrapolation,
taking note of both the area inundated by each increment of sea level rise within the sample sites and the likely
area inundated by each increment between sample strips.
Table 2 records the results of this entire process; it shows cumulative vulnerability estimates for the
entire island for increment of sea level rise. Sampling errors (1 standard deviation) for the sample means are
registered in the parentheses; the market works so well that thorough incorporation of the values recorded
within the sample of 25 strips was sufficient to support t-statistics consistently well in excess of 20, in most cases,
and never less than 10.
Notice that the total value attributed to the beach over the entire range of sea rise is $353 million.
Comparing property values on Long Beach Island with the average of a small sample taken in Manahawkin (just
across the bay), produced a rough Knetsch/David estimate of $346 million in total property value differential
between the island and the mainland. There is, in addition, an estimated $89 million location premium for island
4-7
-------
Yohe
Table 1. Sample Sites -- Long Beach Island, New Jersey
Number
1
2
3
A
*+
5
6
7
8
9
10
11
12
13
14
15
16
17
18
XU
19
X 7
90
£V/
99
^£.
23
^*J
24
25
Tax ID
A-6
A-33
A-52
AOA
- O\J
D-27
E-22
F-38
H-ll
J-22
K-10
L-13
M-24
0-11
0-32
0-62
0-98
0-128
R-90
£m \J
R-fi9
\J £.
R-100
AW
T-7/8
A / / W
T-&0
X *T V
T-144
X X't'T
T-176
W-5/6
Southern Street
Cleveland Avenue
Carolina Avenue
Joshua Avenue
17th Street
25th Street
33rd Street
Marine Lane
Mississippi Avenue
Kansas Avenue
Cape Cod Lane
Rhode Island Avenue
Burwell Avenue
Dupont Avenue
Beards ley Avenue
46th Street
37th Street
D f\ v -i A A*\7AT11 1 0
R.\JA.±.& nvciiuc
R7^Vi ^1"Vftot"
Of \*Ll (JL.A6CU
North- south through
Amherst Road
Northern Street
McKinley Avenue
Inlet Avenue
Magnolia Avenue
Mflfdinl 1 A^roTiiiA
1*1 CLJ. OllaJUX f\ VCil vlC
18th Street
26th Street
34th Street
Ryerson Lane
Idaho Avenue
Lillie Avenue
Ocean View Drive
Massachusetts Avenue
Dayton Avenue
Goldsborough Avenue
Kirkland Avenue
45th Street
36th Street
TJ"? Tl/lttfSS Y"H U f\H H
W J.lLvLWclL U- JcxOclU
T !3 ffOOTl l?O£t/^
l^ogOUIl IxUaCL
T rtTTO 1 fl /1 1 O G T iS n A
JjU VcxaU.J.cS i^aile
Tlaaoon Dvf ir
-------
Yohe
Elevation
30'
20'
10'
.beach
^densely packed houses
Long Beach Boulevard
bulkhead
J
1 ocean
bay
South end of the island with a long stretch of land west of Long Beach Boulevard vulnerable
to inundation from the bay and development packed up to land on top of the dune of a narrow
beach.
Elevation
30'
20'
10'
0«
beach
w construction
'large houses on larger lots
Long Beach Boulevard
ocean
bulkhead
bay
North end of the island with less property to the west of Long Beach Boulevard and larger
houses on larger lots placed further from the dune and a wider beach. Some new construction
is going in on the west side of the dunes.
Figured
4-9
-------
Yohe
Table 2. Economic Vulnerability*
Sealevel
Rise
0-6 inches
6-12 inches
12-18 inches
18-24 inches
2-3 feet
3-4 feet
4-6 feet
Property
$0
(0)
$0
(0)
$80
(4)
$70
(4)
$129
(9)
$315
(8)
$175
(4)
Structure
$0
(0)
$0
(0)
$83
(4)
$72
(4)
$137
(8)
$345
(7)
$184
(5)
Beach
$15
(1)
$40
(2)
$62
(2)
$50
(2)
$115
(5)
$45
(2)
$26
(1)
INCREMENT
$15
(1)
$40
(2)
$225
(6)
$192
(6)
$381
(13)
$705
(11)
$385
(7)
TOTAL
$15
(1)
$55
(2)
$270
(6)
$462
(9)
$843
(16)
$1548
(19)
$1932
(20)
Measured in millions of dollars. The numbers in parentheses represent
standard errors of estimation around the sample means of total or
incremental dollar vulnerability. The total value of the island stands
at approximately $2 billion.
4-10
-------
Yohe
property in direct proximity with the ocean and bay shorelines.5 A total property value increment of $435 million
can therefore be supported by a crude application of the Knetsch/David technique, suggesting that the
structure/property based estimate of the social value of the beach reported in the tables is conservative.
5This additional premium is computed by looking at the property value gradients along both the bay side and
the ocean side of the island along its entire 18-mile length.
4-11
-------
Yohe
CHAPTERS
DISCUSSION
Relating the vulnerability estimates of Table 2 to temporal, greenhouse induced sea level rise scenarios
requires incorporating the natural 3.9 mm increase per year trend of the ocean off New Jersey. Table 3 tracks,
in 10-year increments, sea level scenarios that attribute 50-cm, 100-cm, and 200-cm increases to greenhouse
wanning, respectively. Each includes nearly 1.5 feet in historical trend sea level rise between now and the year
2100. Table 4 translates the cumulative cost estimates of Table 2 into time-dependent estimates for each of the
three scenarios; Figure 3 portrays each trajectory graphically. Annual loses are reflected in Figure 4 and Table
5. Both highlight the losses which can be expected on an annual basis for the decade following the indicated
year. The figures show that marginal costs do not always climb; for the 2-meter scenario, e.g., marginal cost
at 2100 is zero because the island was completely lost by the year 2090.
When real estate markets work well, market values reflect the discounted value of a stream of housing
service income, implicit in the case of owner-occupied housing or explicit in the case of rental property. It is
therefore interesting to consider the trajectory of lost economic rent that would have supported property values
that were lost. Figure 5 illustrates lost economic rent embodied in cumulative economic cost for an assumed
10% return on investment. Higher returns would, of course, produce higher loss profiles; lower returns, lower
profiles.
4-12
-------
Yohe
Table 3. Amount of Sea Level Rise for Various Scenarios
Year
2000
2010
2020
2030
2040
2050
2060
2070
2080
2090
2100
*
Scenario
50 cm. 100 cm.
.14
,31
.51
.73
.98
1.25
1.56
1.89
2.25
2.63
3.05
.15
.36
.63
.94
1.32
1.74
2.22
2.76
3.34
3.99
4.68
200 cm.
.18
.47
.87
1.38
1.99
2.71
3.55
4.49
5.53
6.69
7.95
Measured in feet, including the natural trend of 3.9mm per year.
The scenario identification indicates the amount of sea level rise
attributed to greenhouse warming above and beyond this natural trend.
4-13
-------
Yohe
Table 4. Cumulative Economic Vulnerability*
Year
2000
2010
2020
2030
2040
2050
2060
2070
2080
2090
2100
50 CM.
$3
$9
$15
$34
$56
$155
$280
$315
$405
$518
$873
100 CM.
$4
$11
$23
$49
$175
$355
$527
$720
$1041
$1540
$1651
200 CM.
$6
$14
$39
$215
$457
$671
$1168
$1633
$1831
-H-
-H-
Measured in millions of dollars. The scenarios are identified in
Table 3; the source of the cost estimates is Table 2.
-H- The entire island is lost at this point.
4-14
-------
Yohe
A
c
0
•H
H
N
£ 2000
v 1580
W 1000
0
0
500
•H
£
0
a
Fifty CX,
•
OneH,
x
IvoK,
Vears
Rgure3.,
4-15
-------
Yohe
a
Fifty A
•
OneH.
x
IwH,
Figure 4.
4-16
-------
Yohe
Table 5. Annual Increase in Economic Vulnerability*
Year
2000
2010
2020
2030
2040
2050
2060
2070
2080
2090
2100
50 CM.
$0.0
$0.4
$0.6
$1.3
$2.0
$6.0
$11.2
$3.0
$9.0
$10.2
$18.4
100 CM.
$0.0
$0.5
$1.0
$1.8
$7.8
$14.3
$16.6
$18.3
$25.7
$41.0
$30.5
200 CM.
$0.0
$0.7
$1.2
$9.8
$20.9
$22.8
$35.5
$48.3
$33.7
$17.0
++
Measured in millions of dollars. The scenarios are identified in
Table 3; the source of the cost estimates is Table 2.
++ The entire island was lost in 2090.
4-17
-------
Yohe
c
o
200
a 150
3 100
w
o
Ol
PC
*
3
D
Fifty Ql,
•
toed,
x
toH,
Vears
Figure 5.
4-18
-------
Yohe
CHAPTER 4
EXTENSIONS TO A NATIONAL STUDY
Straightforward application of the basic methodology recorded in Chapter 1 to a coastal sampling
conducted by Park et al. (this volume) can be used to produce national and regional estimates of economic
vulnerability. Park's study looked at the effects of 0.5-, 1-, and 2-meter sea level rise scenarios on 46 sites
selected at regular intervals around the country. Together, these sites accounted for 10% of the U.S. coastal
zone. Taking every other site to generate a first cut at an estimate of total vulnerability would therefore cover
5% of the coastal zone, and guarantee that particular regions would be included in the estimate roughly in
proportion to their area. Basing a national estimate on this subsample would not support a precise result, but
it would be sufficient to support an order of magnitude estimate of vulnerability. Going further may, in fact, give
the spurious impression of increased precision given the uncertainties with which we view the distant future.
More importantly, using the 5% subsample should certainly identify regions for which initial translations of
vulnerability to cost would be most productive.
The precise details of applying the theory of Chapter 1 to the Park sample results need not be covered
here, but at least one limitation should be mentioned. The Park surveys for each site usually record the effects
of sea level rise for quadrants measuring 500 meters by 500 meters. Applying the notions of property value
gradients outlined above to Park grids whose patterns frequently include quadrants extending 1000 feet inland
is therefore troublesome, at best. It should, as a result, be expected that estimating vulnerability on the basis
of average property values for each quadrant, taken from tax maps or housing and business census data, is the
greatest precision which the inundation scenarios will support. How much accuracy is thereby lost? Initial
comparisons of estimates derived from Park scenarios for Long Beach Island and the estimates reported in
Chapters 2 and 3 above suggest that the answer to this question is "Not much." The law of large numbers seems
to apply quite nicely, but any work toward a national estimate based on the Park surveys will include, as a quality
check, a careful comparison with the more detailed work on Long Beach Island reported here.
4-19
-------
Yohe
CHAPTERS
EXTENSIONS FROM VULNERABILITY TO COST
Frictional adjustment costs were first mentioned in Chapter 1, but they were dismissed there as being
more closely related to costs than vulnerability. It is immediate, therefore, that modeling needs to be done to
reflect their potential as the focus moves away from measuring the economic vulnerability to sea level rise and
toward measuring the economic cost of sea level rise. Their very nature is, however, extremely suggestive. If
the rate of greenhouse-induced sea level rise were known with certainty and there were enough time to respond,
it is possible that the economic cost of sea level rise would be confined to adjustment costs and the value of the
inundated land. Structure is mobile and would presumably be moved; coastal services can be provided by the
new coastline. The question becomes, then, a matter of determining what happens when time is short and our
foresight is imperfect and uncertain.
An initial line of analysis should look at the simpler component of this question. Some long-term growth
modeling along a certain sea level rise trajectory should unravel the dependence of both relocation costs and the
lost value of structure abandoned because time was too short on rates of economic growth, rates of economic
depreciation, and rates of dislocation. It should include a thorough analytical structure which reflects how people
and markets might reasonably respond to the effects of the trajectory, so it can reflect the time dependence of
intertemporal costs. Only then can a decision whether or not to protect a particular piece of coastline be cast
in a context that considers both the timing and the degree of protection.
A second line of analysis should then build on the first to incorporate uncertainty and risk. Critical here
should be not only how people and markets respond to uncertainty over the long term, but also how people and
markets learn what is going on. Figure 6 illustrates the current state of our knowledge about greenhouse-
induced sea level rise, and the decisions we make now are dependent upon the relative likelihoods that we place
on each possibility shown there. Our subjective distribution of possible futures will be different in the year 2000,
and 2010, and so on; so we should expect that our decisions might change. Protection decisions, contingent upon
certain events actually occurring, should, in fact, be considered explicitly -- perhaps as exogenous changes in
economic environment made at certain times, but perhaps as endogenous variables in the modeling itself. In
either case, it becomes important to explore the value of the information upon which decisionmakers weigh the
costs of protecting a region against the costs of not protecting that region.
This second phase of the theoretical work will be difficult, so it should be conducted as one part of a
two-part extension of the certainty modeling. With the results of the first cost analysis well established, it should
also prove fruitful to apply the insights that it provides to specific regions taken from the vulnerability subsample.
Looking at likely sources of costs for specific regions along the three scenarios will reveal which of the theoretical
issues are more important than others and whether or not the ranking of their relative importance depends upon
the region selected. Extension of the region-specific analysis to the more complex uncertainty modeling will then
be able to focus on the most productive issues without wasting time on concerns that turn out, at least for one
region, to be less significant.
4-20
-------
Yohe
w
o>
• Hoffman (1983) High
UJ
>
UJ
-------
Yohe
REFERENCES
David, E. L., "Lakeshore Property Values: A Guide to Public Investment in Recreation," Water Resources
Research, (vol. 4, pp. 697-707), 1968.
Clawson, M. and J.L. Knetsch, Economics of Outdoor Recreation. Johns Hopkins, Baltimore, 1966.
Gibson, J., "Recreational Land Use", in The Valuation of Social Cost, (ed. David Pearce), George Allen &
Unwin, Boston, 1978.
Hoffman, J.S., D. Keyes and J.G. Titus, "Projecting Future Sea Level Rise", U.S. EPA, Washington, 1983.
Hoffman, J.S., J. Wells and J.G. Titus, "Future Global Warming and Sea Level Rise" in Sigbjarnarson, G. (ed.)
Iceland Coastal and River Symposium. 1986.
Knetsch, J. L., "The Influence of Reservoir Projects on Land Values," Journal of Farm Economics, (vol. 46, pp.
231-43), 1964.
Meier, M.F. et. al., Glaciers. Ice Sheets, and Sea Level. National Academy Press, Washington, 1985.
Revelle, R. "Probable Future Changes in Sea Level Resulting from Increased Atmospheric Carbon Dioxide," in
Changing Climate. National Academy Press, Washington, 1983.
4-22
-------
AN OVERVIEW OF THE NATIONWIDE IMPACTS
OF SEA LEVEL RISE
by
James G. Titus and Michael S. Greene
Office of Policy Analysis
Office of Policy, Planning & Evaluation
U.S. Environmental Protection Agency
401 M Street, SW
Washington, DC 20460
-------
CONTENTS
Page
FINDINGS 5-1
CHAPTER 1: INTRODUCTION 5-2
CHAPTER 2: OBJECTIVES AND STRATEGY 5-3
CHAPTER 3: SCENARIOS 5-7
CHAPTER 4: CASE STUDY OF LONG BEACH ISLAND 5-9
METHODS USED IN THE LBI CASE STUDY 5-9
Loss of Coastal Wetlands 5-9
Sand Costs to Raise LBI and Maintain the Shoreline 5-9
Alternatives to Raising Barrier Islands 5-10
Value of Threatened Structures 5-10
RESULTS AND IMPLICATIONS FOR THE NATIONAL STUDY 5-11
Wetlands Losses 5-11
Is Raising Barrier Islands In Place a Reasonable Scenario? 5-13
CHAPTER 5: THE NATIONAL STUDIES: ADDITIONAL METHODOLOGICAL
CONSIDERATIONS 5-15
SITE SELECTION 5-16
LOSS OF COASTAL WETLANDS AND DRYLANDS 5-15
DEFENDING SHELTERED SHORES 5-18
RAISING THE PROFILE: A SIMPLE PROCEDURE FOR ESTIMATING SAND
REQUIREMENTS DUE TO SEA LEVEL RISE 5-19
COST OF PROTECTING THE OPEN COAST 5-21
CHAPTER 6: RESULTS 5-22
RESULTS: NATIONWIDE LOSS OF WETLANDS AND DRYLANDS 5-22
Estimates of Park et al 5-22
Analysis: Sampling Error 5-22
RESULTS: THE NATIONWIDE COST OF PROTECTING SHELTERED SHORES 5-29
Estimates of Weggel et al 5-29
Analysis: Interpolating Results of Weggel et al. for 0.5- and 1-Meter Sea Level Rise
Scenarios 5-29
RESULTS: NATIONWIDE COST OF PROTECTING THE OPEN COAST 5-34
Leatherman Estimate of Sand Costs 5-34
Analysis: An Increasing-Cost Scenario for Sand 5-34
Analysis: The Cost of Elevating Buildings and Roadways 5-42
Census Data 5-42
Extrapolation Procedure 5-42
Results of Extrapolation 5-46
CHAPTER 7: SUMMARY AND CONCLUSIONS 5-48
FINAL CAUTION 5-48
SUMMARY RESULTS 5-48
NEXT STEPS 5-51
Toward Better National Assessments 5-51
Toward Protecting Coastal Wetlands 5-51
REFERENCES 5-53
APPENDIX 1 5-55
u
-------
Titus
FINDINGS1
Global warming could cause sea level to rise 0.5 to 2.0 meters by 2100. Such a rise would inundate
wetlands, erode beaches, exacerbate coastal flooding, and increase the salinity of estuaries and aquifers.
A 1-meter rise by the year 2100 could drown approximately 25 to 80% of the U.S. coastal wetlands. Their
ability to survive would depend largely on whether they could migrate inland or whether levees and
bulkheads block their path of migration.
A 1-meter rise could inundate 5,000-10,000 square miles of dryland if shores are not protected, and 4,000-
9,000 square miles of dryland if only developed areas are protected.
Most coastal barrier island communities would probably respond to sea level rise by raising land with sand
pumped from offshore. Wide and heavily urbanized islands may use levees, while communities on lightly
developed islands may adjust to a gradual landward migration of the islands.
The long-term survival of coastal wetland ecosystems can be ensured if society takes measures to explicitly
declare that developed low lands will be vacated as sea level rises. If implemented today, the purchase of
future development rights required to follow such a strategy will be relatively inexpensive; if delayed, those
same purchases will become too expensive, and forcing landowners to vacate their coastal property without
just compensation would be considered an unconstitutional taking.
Although the information in this report has been funded wholly or partly by the U.S. Environmental
Protection Agency, it does not necessarily reflect the Agency's views, and no official endorsement should be
inferred from it.
5-1
-------
Titus
CHAPTER 1
INTRODUCTION
In the last six years, coastal scientists, engineers, and policy makers have gradually begun to consider the
prospect of a rise in sea level of one to ten feet over the next century. Because the interest of coastal zone
managers in the practical consequences of sea level rise predated a widespread interest at the national level, most
case studies on the effects of sea level rise have examined the implications for the specific decisions people make
today, rather than estimating nationwide impacts. This paper summarizes the first nationwide assessment of the
implications of future sea level rise.
5-2
-------
Titus
CHAPTER 2
OBJECTIVES AND STRATEGY
Ideally, the goal of our research would be to know the economic and environmental impacts of all of the
various scenarios of sea level rise for all possible policy response options and for every coastal community in the
nation. Every community would then have sufficient information to rationally consider how they should respond.
Moreover, we could estimate the nationwide impact by picking the best policy response option for each
community and adding the costs. Because such a comprehensive analysis is not yet possible, this analysis had
a more limited objective: a nationwide assessment that included as many factors as possible.
Our first step was to choose which of the impacts to study; we chose shoreline retreat for several reasons.
First, we excluded saltwater intrusion because only one case study had been conducted; the processes are too
complicated to meaningfully represent without detailed models; and the unavoidable economic and environmental
impact of increased salinity appeared to be an order of magnitude less than shoreline retreat—and much more
sensitive to drought than to sea level rise (Hull and Titus, 1986). We would have liked to have included flooding,
which is closely related to shoreline retreat, but the cost of applying flood models to a large number of sites was
prohibitive, and models of the resulting property damage are inaccurate without detailed surveys of the elevation
and types of structures.
By contrast, examining the costs of (1) natural shoreline retreat and (2) holding back the sea seemed
feasible. Estimating inundation of dryland simply requires that one know its elevation; wetland loss requires
the elevation and an assumption regarding how rapidly the wetlands might accrete; a first-order estimate of
beach erosion can be developed using topographic maps and a simple mathematical formula; and the value of
lost property can be estimated using tax maps. The costs of holding back the sea are also fairly straightforward.
The additional wetland loss is the area of developed property that would be protected (if these areas were not
protected, wetlands would be created); the cost of nourishing beaches can be derived using data collected by the
Corps of Engineers, and the cost of elevating land, houses, and infrastructure, and of erecting shore-protection
structures, can be calculated by engineers based on experience.
Moreover, the procedures for assessing shoreline retreat tend to implicitly account for flooding caused by
storm surges (at least after the first foot of sea level rise). Where development is protected from sea level rise,
levees and pumping systems used for preventing inundation would also limit flooding; and raising barrier islands
and the structures on them by the amount of sea level rise would leave flood risks constant.2 Where
development is unprotected, the estimates of lost land and structures probably account for the costs of increased
flooding; although flood plains would move inland, the value of structures standing in the new flood plain would
approximately be balanced by the inundated structures that are lost. Nevertheless, for the first foot of sea level
rise, examining shoreline retreat probably does not account for flooding: if development is protected, major
coastal engineering measures probably would not be taken to counteract the first foot of rise, so the frequency
of flooding would increase. If development is not protected, the first foot of rise would increase all flood surges
but would not threaten many structures.4
assumes that climate change has no impact on storm frequency of magnitude.
*This implicitly assumes that the development density of the coastal plain is uniform. Development in coastal
areas tends to have the highest density in the first few rows of housing closest to the water (to obtain a
waterfront view). The density of development in the rows just out of sight of the sea is slightly less.
One foot is somewhat arbitrary; some areas might require levees with a smaller rise. However, we doubt
that many areas will build levees or elevate land and structures until the sea rises at least one foot, but all coastal
areas will experience incremental increases in flooding for any rise in sea level.
5-3
-------
Titus
At the outset, it was clear that it would not be possible to estimate both the cost of shoreline retreat and
the cost of holding back the sea in time to meet the report's congressional deadline. From previous studies
and conversations with hundreds of state and local officials, it was clear that if we had to choose between the
two, the cost of holding back the sea--at least in developed areas-was a more reasonable representation of the
nationwide impact of sea level rise. We would learn little, for example, from estimating the value of buildings
on Manhattan Island that would be lost if the sea was not held back; because of its value, the area would have
to be protected. Furthermore, coastal scientists and engineers had been studying the physical implications of sea
level rise, but few economists had investigated the economic implications.
By contrast, the wetlands study could examine the impacts of not holding back the sea as well. Estimating
the net loss of wetlands requires one to consider (1) the conversion of existing wetlands to open water and (2)
the conversion of dryland to wetlands. The former, which generally does not depend on coastal protection
policies, is difficult because it depends on wave erosion and the ability of wetlands to accrete vertically; by
contrast, the latter, which does depend on coastal protection, is fairly easy to estimate because over the course
of a century virtually any sheltered area with an elevation between mean and spring hightide would convert to
wetlands if not protected by levees and other structures.
Given the funding constraints and disciplinary boundaries for assessing sea level rise, we defined four
studies:
(1) Park et al. would compile elevation data to estimate the inundation of dryland with and without
the protection of developed areas and the loss of coastal wetlands for various shore-protection
options.
(2) Leatherman would estimate the cost of dredging sand to nourish recreational beaches and,
where necessary, to raise barrier islands in place, and develop data on the areas of developed
barrier islands.
(3) Weggel et al. would estimate the cost of protecting developed areas along sheltered coasts, and
would develop rough estimates of the cost of elevating houses and rebuilding infrastructure to
accommodate a raising of Long Beach Island, New Jersey.
(4) Yohe would estimate the costs of losing land and structures, starting with Long Beach Island,
New Jersey.
Besides providing an overview of the four papers, this paper undertakes a number of supplementary
analyses. We examine statistical uncertainty due to sampling for the Park et al. and Weggel et al. studies. For
the cost of protecting sheltered shorelines, we also combine the Park et al. estimates of inundated lowlands with
the Weggel et al. cost assumptions for bulkheads and levees to develop cost estimates (1) for sea level rise
scenarios that Weggel et al. did not consider and (2) that explicitly calculate the mix of levees and bulkheads
necessary to protect each site. For the cost of protecting the open coast, we combine the cost factors of Weggel
et al. and Leatherman's estimates of the area of UJS. barrier islands with Census data on housing densities to
estimate the (non-sand) cost of raising barrier islands in place in response to a rise in sea level. We also use
Leatherman's results to estimate national and state sand costs if unit sand costs increase as nearshore deposits
are exhausted.
Figure 1 illustrates the relationships between the studies. The assessment of the nationwide costs of holding
back the sea began with a case study of Long Beach Island, New Jersey. In the study of Long Beach Island,
Leatherman and Park et al. followed the same procedures they would subsequently apply nationwide. Weggel
et al. examined additional shore protection options as a check to ensure that the option used by Leatherman was
reasonable.
5-4
-------
Titus
Park
/ \
Loss of
Wet/Dry
Land
Elevations
Cost of
Levees
Island
Retreat
CASE STUDY
NATIONAL ANALYSIS
Cost to
Rebuild
Infra-
structure
Cost
of
Sand
Erosion
Value
of
Threatened
Property
Decision to Use
Island Raising
Scenario
Loss of
Wet/Dry
Land
Cost of
Protecting
Sheltered Coasts
2-m Scenario
Cost of Nourishing
Beaches and Raising
Barrier Elevations
(assuming fixed costs
of sand)
(^ Titus J) (^
Non-Sand Cost
of Raising
Coastal Barriers
Titus
Confidence
Intervals
for Wetland
Loss
i
r
Cost of Protecting
Sheltered Shores
50 and 100cm
Scenarios
Titus
Increasing Sand
Cost Scenario
Figure 1. Overview of sea level rise studies — authors and impacts.
5-5
-------
Titus
The choice of shore protection options used in this study results from the need to develop a nationwide estimate
of the costs associated with sea level rise; they are not necessarily presumed to be the most appropriate responses.
The assumption of a uniform nationwide approach to shore protection was a computational necessity and not a
reflection of how we expect society to respond. The justifications we provide show why these are reasonable options,
but they should not be construed as an endorsement
5-6
-------
Titus
CHAPTERS
SCENARIOS
We chose three scenarios of sea level rise for this study: 50, 100, and 200 centimeters (cm) by the year
2100. Following the convention of a recent National Research Council report (Dean et al., 1987), the rise was
interpolated throughout the 21st century using a parabola, as shown in Figure 2. For each site, local subsidence
was added to determine relative sea level rise. In addition to the three accelerated sea level rise scenarios, we
also included a baseline scenario, which assumes that sea level will continue to rise at the historical rate of 12
cm per century or 14 cm by the year 2100.
We also devised three alternative scenarios of shoreline protection: no protection, standard protection,
and total protection. In the no protection case, we assumed that no shoreline would be defended from a rising
sea. For standard protection, we assumed that densely developed, sheltered coasts would be defended by either
seawalls or levees (the cost of which was calculated without drainage systems for all but the Long Beach Island
case study).5 For the total protection case, every mile of sheltered coastal lowlands would be protected with
either bulkheads or levees.
*The exclusion of drainage system costs for the national assessment of protecting developed sheltered
shorelines gives us a conservative (e.g., low) estimate of the costs that the country may face.
5-7
-------
Titus
2100
Figure 2. Sea level rise scenarios for Miami Beach (including local subsidence).
5-8
-------
Titus
CHAPTER 4
CASE STUDY OF LONG BEACH ISLAND
We picked Long Beach Island (LBI), NJ., for the case study because we had experience there and because
it provided most of the features that had to be considered: a narrow well-developed barrier island and a low-
lying, partially developed mainland with extensive marshes.
Although the Park et al. and Leatherman studies of Long Beach Island were to be similar to their respective
approaches in the nationwide assessment, we needed a more detailed engineering assessment from Weggel et
al.: Leatherman was going to estimate only the cost of pumping sand onto beaches and coastal barriers; he was
not going to estimate the cost of elevating buildings and roads. By examining Long Beach Island, Weggel et al.
would provide us with engineering cost factors that we could apply to other communities. Moreover, we wanted
to confirm the reasonableness of the hypothesis on which Leatherman's study would be based: that raising
barrier islands was a reasonable assumption for estimating the cost of defending the open coast. Therefore,
Weggel et al. also examined two alternative options: (1) a landward migration of the barrier island and (2)
building a levee and drainage system. Finally, it would not have been feasible for Weggel et al. to visit every site
the study would assess. Therefore, the investigators used Long Beach Island and five other sites to develop
engineering rules-of-thumb that could be applied to a broader selection of sites.
METHODS USED IN THE LBI CASE STUDY
Loss of Coastal Wetlands
Park et al. sought to (1) compare the results of their model of wetland loss around Long Beach Island
with the survey-based estimates from Kana et al. (1988), and (2) determine the impact of shore-protection efforts
on wetland loss. First, the elevations of both wetlands and dryland had to be characterized. For wetlands,
satellite imagery was used to determine plant species for 60- by 80-meter parcels. Using estimates from the
literature on the frequency of flooding that various wetland plants can tolerate, it was then possible to estimate
the percentage of time a particular parcel is underwater. Park et al. then used estimates of local tidal ranges
to calculate the corresponding wetland elevation. For dryland, spot elevation measurements were used to
interpolate between contours on USGS topographic maps and the elevations defined by the upper boundary of
tidal wetland vegetation. To keep computations manageable, Park et al. aggregated the results into 500-meter
cells; however, they also kept track of the percentages of the ceU that corresponded to various elevations and
wetland types.
Park et al. estimated the loss of wet and dryland for no protection and protection of developed areas. For
the no-protection scenario, estimating the loss of dryland is straightforward. However, for calculating wetland
loss, Park et al. had to consider the wetlands' vertical growth. For the baseline scenario, published rates of
vertical accretion were used. For accelerated sea level rise, allowance was made for some acceleration of vertical
accretion in areas with ample supplies of sediment, such as the tidal deltas.
Sand Costs to Raise LBI and Maintain the Shoreline
Leatherman sought to estimate the cost of pumping enough sand to maintain the shoreline and gradually
raise Long Beach Island. This required estimating the area of (1) the beach system, (2) the low bayside, and
(3) the slightly elevated oceanside of the island. Leatherman used the "raising the profile" method, which we
some areas, vertical accretion is limited by sea level rise, not available sediment. If sea level rise
accelerates some sediment flow that would otherwise wash onto beaches, sandbars, or into deep water, the
sediment will instead wash into the wetlands.
5-9
-------
Titus
explain in a following section ("National Studies: Additional Methodological Considerations"). The area of the
beach system was found by multiplying the length of the island times the length of the beach profile, which
Leatherman calculated based on a 1-year storm, which in this case implied the 23-foot contour.7 Topographic
maps were used to estimate the area of land above and below 5 feet NGVD (1929),8 which is about 4.5 feet
above current sea level.
Given the beach areas, the volume of sand was estimated by assuming that the beach system would be
raised by the amount of sea level rise, and that the bay and ocean sides of the island would be raised after a rise
in sea level of one and three feet, respectively. Leatherman assumed that sand would cost $7.85 per cubic yard,
based on published sand inventories conducted by the Corps of Engineers.
Alternatives to Raising Barrier Islands
Examining the practicality of raising barrier islands required an assessment of two alternatives that had
received more attention in previous studies: (1) protecting the island with seawalls and levees or (2) allowing
it to migrate landward (imitating the natural barrier island overwash process by allowing the oceanside of the
island to erode and pumping sand into the bayside to build land). After visiting Long Beach Island and the
adjacent mainland, Weggel et al. designed and estimated the cost of an encirclement scheme consisting of dikes,
levees, and a drainage system involving pumping and underground retention of stormwater. For island migration,
the Bruun Rule (see Leatherman, this volume, for a description of this rule) was used to estimate oceanside
erosion, and navigation charts were used to calculate the sand necessary to fill the bay landward by an amount
equivalent to the oceanside erosion. For island raising and island migration, Weggel et al. used the literature
to estimate the cost of elevating and moving houses and of rebuilding roads and utilities.
In preparation for the national study, Weggel et al. estimated the cost of protecting the mainland shore
in the vicinity of Long Beach Island and compared the detailed shoreline information with the rougher data
provided by the 500-meter cells of Park et al.
Value of Threatened Structures
Yohe's case study is the only part of his report that was contracted to be available by publication of this
report; the national estimate of the value of threatened structures and property is expected in the latter half of
1989. His objective was to provide economic information necessary to place the estimates of shore protection
into their proper context (by estimating the value of land and structures that would be lost if the sea were not
held back) after the other studies had been completed.
On the bay side of the island, the approach was relatively straightforward: any structures or land flooded
at high tide would be considered lost. However, on the more elevated oceanside, Yohe had to specify the timing
7Large storms have an impact on sediment transport farther out to sea than do small storms. With larger
storms come larger waves and excessively high and low tides. Thus, the larger the storm considered, the farther
out to sea the beach profile extends.
8Because sea level has been rising, a contour that was five feet above sea level fifty years ago may only be
4.5 feet above sea level today. To avoid potential confusion, most maps today express elevations with respect
to the "National Geodetic Vertical Datum" (NGVD) reference plane, which is a fixed reference unaffected by
changes in sea level.
9Only the direct costs associated with raising the barrier island are included in the analysis. Indirect and less
tangible costs that may be felt (i.e., the inconvenience suffered by the community when roads and utilities are
dug up and raised, or when houses must be raised or moved as sand is pumped onto the island) are not included.
These costs may be substantial and may change the outcome of our analysis. We are, however, unequipped to
estimate them.
5-10
-------
Titus
of the removal of the structures. Would they be removed only after the beach was lost and the structures were
flooded at high tide or when the beach narrowed to a critical width? The latter approach was chosen for three
reasons. First, the former approach would require Yohe to estimate the demand for beach area, something
that was beyond the study's resources. Second, houses in front of the dunes would be vulnerable to storm
damage and probably would not be able to stand in front of the dunes for more than a few years. Finally, in
most cases, the value of the beach is greater than the value of the oceanfront structure, since the beach is one
of the main reasons people buy property on or travel to barrier islands. Therefore, it seemed reasonable to
assume that if a community decided to allow its shores to retreat, it would also require that structures be
removed before they disrupted use of the beach. (The Texas Open Beaches Act already requires houses to be
removed if they are within the dune vegetation area, and many other areas administratively follow this policy
where possible.)
Nevertheless, some narrowing of the beach would probably be tolerated. Particularly along the northern
part of the island, houses are generally set back over one hundred feet from spring high tide. To be
conservative, Yohe settled on a minimum distance of forty feet from a structure to spring high tide, which is
about equal to the distance from the crest of the dunes to the wet part of the beach. Because Yohe did not
estimate the demand for beach area, he could not estimate the recreational benefits that would be lost if the
beach narrowed. Nor did he consider the diminished flood protection value.
Given these behavioral assumptions, and estimates of erosion and inundation from Leatherman, Yohe
determined which property would be lost from sea level rise for a sample of 25 strips spanning the island from
ocean to bay. He then consulted tax maps to estimate the value of the land and structures that would be lost.
Because a 50- by 100-foot oceanfront lot is typically worth about $100,000 more than an interior lot of similar
size, Yohe had to consider the fact that this oceanfront premium would be transferred to another owner, not lost
to the community. Thus, the value of land lost is the value of an interior lot.10
RESULTS AND IMPLICATIONS FOR THE NATIONAL STUDY
Table 1 summarizes the case study results for Long Beach Island for two policy options: (1) raising the
island and bulkheading mainland sheltered shorelines and (2) natural shoreline retreat.
Wetlands Losses
The estimates by Park et al. largely confirmed previous estimates by Kana et al. (1988) that a 2-meter rise
in sea level would drown 80% of the wetlands around Long Beach Island if shores were not protected.
However, the results were not consistent with the hypothesis that wetlands loss would be substantially greater
if the shores were protected; bulkheading the shores decreased the area of surviving wetlands by less than 10%
in all scenarios. Two possible explanations are:
(1) A large portion of the wetlands are on marsh islands that would not be bulkheaded under any
circumstances.
(2) At the coarse (500-meter) scale Park et al. used, the assumption of only protecting developed areas
amounted to not protecting a number of mainland areas where the shoreline is developed but areas
behind the shoreline are not.
As a result of the latter reason, it seemed reasonable to investigate the implications for coastal wetlands
of protecting all coastal lowlands (total protection).
1 Adjustments were made to these data to ensure that the information was up-to-date. See Yohe's paper
in this volume.
5-11
-------
Titus
Table I. Results For Long Beach Island Case Study
*
Wetland Losses
(Percent of Original Area)
Response 50 cm 100 cm 200 cm
Raise Island/Bulkhead Mainland 45% 70% 78%
Natural Shoreline Retreat 50% 73% 80%
**
Response Costs
(Millions of 1986 Dollars)
Response 50 cm 100 cm 200 cm
Raise Island/Bulkhead Mainland:
Sand Cost
Elevating Houses/Roads
Total
Natural Shoreline Retreat:
Loss of Rents
158
457
615
2,663
303
856
1,156
6,096
611
1.109
1,720
9,696
Sources:
Wetland Loss -- Park et al.
Sand Costs - - Leatherman
Other Engineering Costs for 2-meter scenario -- Weggel et al.
Lost Rents -- Yohe
NOTE: All researchers added approximately 20 cm of local subsidence to the
global sea level rise scenarios. Therefore, our derivations of other
engineering costs for 50- and 100-cm scenarios are based on Weggel et al.
estimate of the cumulative cost in the 2-meter scenario when local sea level had
risen 70 and 120 cm, respectively.
Wetland loss estimates reflect Park's original run. His current paper reports
on a subsequent run and results are substantially different.
Dollar figures are cumulative, not discounted.
5-12
-------
Titus
Is Raising Barrier Islands In Place a Reasonable Scenario?
The results by Weggel et al. clearly indicate that it would be much less expensive to raise Long Beach
Island than to allow it to migrate landward, as shown in Table 2. A casual glance at the table also suggests that
the option of building a levee around the island would be even less expensive. However, other considerations
suggest that island raising would be more reasonable. First, a levee would eliminate the bayfront view. Second,
because most of the levee costs would have to be borne all at once, financing it would be difficult. Third,
Weggel et al. concluded that the levee would have to be built in the 2020's, and island raising could take place
gradually between 2020 and 2100. This implies that the (discounted) present value of the levee cost would be
greater.'1 For example, by the middle of the 2020's, the present value of the cost for island raising through the
end of the century would be $400 million at a 3% discount rate and $200 million at a 10% discount rate, which
would be far less than the $800 million for the levee system.12 Fourth, a levee would alter the ecology of the
undeveloped tracts of land. Finally, some people would feel unsafe residing on a barrier island below sea level.
Therefore, we concluded that it would be reasonable for Leatherman to assume that entire developed
coastal barriers like Long Beach Island are raised in place, given that he could examine only one option for his
national study. However, we caution that this assumption would probably not be reasonable for islands whose
characteristics are vastly different. A very lightly developed island might find migration cheaper. For example,
the analysis of Weggel et al. shows that landward migration is more expensive than island raising primarily
because of the increased costs of rebuilding water and sewer lines and other utilities. But considerably less sand
is required. (See Titus 1987b for a discussion of the institutional challenges this option would face.)
On the other hand, levees might be more practical for wide barrier islands where most people do not have
a waterfront view. The most noteworthy example in the United States is Galveston, Texas, which is already
protected by a seawall. Recently, people there have discussed totally encircling Galveston Island with a levee.
The cost of raising an island with a given development density depends on the area of the island, while the cost
of a levee depends on the island's perimeter. Thus, levees are least practical for narrow barrier islands. But for
an island as long as Long Beach Island but five times as wide, the cost of a levee would be about the same, and
the cost of raising the island would be five times as great.
A final question concerning the reasonableness of raising islands is whether the costs would be so great
that it would be better to simply abandon the island. This is not likely to be a serious option for Long Beach
Island in the next century, even for the 2-meter sea level rise scenario. Figure 3 compares the annual cost for
elevating the island under the 2-meter scenario with Yohe's estimate of the value of the economic returns (rental
income) lost in a particular year due to the cumulative loss of land and structures. The annual cost for elevating
the island would gradually rise to $22 million/year by 2100. By contrast, the annual loss of rents (and property
as well) would reach this level by the 2030's; by 2100, the annual loss of rents would be about $200 million, ten
times the cost of shore protection. Right from the start, shore protection at Long Beach Island would be
cost-effective and it would continue to be so. (The fact that shore protection is cost-effective means that the
island has the resources to protect itself, but it does not address whether the residents, taxpayers, or contributors
of greenhouse gases should bear the costs.)
11 In principle, some of the costs of a levee and drainage system could be deferred by raising the levee in
stages, but the initial cost would be more than half of the total cost due to the need for land purchases, pumping
systems, and design.
12The term "discounting" refers to a procedure by which economists equate dollars in one year with dollars
in another year, generally by using a rate of return (interest rate). The present (discounted) value of one dollar
Y years hence at a discount rate of R is:
5-13
-------
Titus
Table 2. Results For Long Beach Island Case Study
Protection
Costs for 2 -Meter
Sea Level Rise Scenario
(Millions of 1986 Dollars)
*
Sand Costs:
Beach
Land Creation/Maint.
Moving/Elevating Houses
Roads/Utilities
**
Levee and Drainage
Total
Encir-
clement
290
0
0
0
542
832
Island Island
Raising Migration
290 0
270 321
37 74
1,072 7,352
0 0
1,669 7,747
Sources: Leatherman, Weggel et al.
NOTE: Weggel et al. estimated sand cost of $560 mil. ("Island Raising" above)
differs from Leatherman's estimate of $611 mil. (Table 1 for 200 cm scenario)
because each investigator made different assumptions regarding the closure depth
of the beach profile (i.e. they assume different widths for the beach profile).
Sand costs include the incremental periodic beach nourishment costs to raise
entire beach profile.
Designed to withstand 100-year storm.
5-14
-------
Titus
CHAPTERS
THE NATIONAL STUDIES: ADDITIONAL METHODOLOGICAL CONSIDERATIONS
SITE SELECTION
Ideally, we would have studied the entire coastal zone of the United States. Unfortunately, the cost of
satellite data collection and interpretation made it impossible for Park et al. to encode more than 20% of the
U.S. coast; because Weggel et al. used the same sites, they were similarly limited.
Site selection was motivated by two concerns: First, we wanted the sample to be unbiased and to yield
statistically efficient estimates. Second, state and local coastal zone agencies expressed a need for information
to be as site-specific as possible, and certainly aggregated at no more than the state level. Leatherman sought
to satisfy both needs by examining every recreational beach in the country. To date, he has examined all of the
beaches in all of the coastal states except Hawaii, Washington, Oregon, and Maine through New Jersey. In
these states, which account for 20% of the nation's recreational open-coast beaches, he examined one beach per
state.
For the other studies, however, we knew that sampling would be necessary. Thus we had to choose between
a random sample and sampling at regular intervals. We decided to adopt the latter approach because it
guarantees that particular regions will be represented in proportion to their total area in the coastal zone, while
a random sample would have left open the possibility that Louisiana or another atypical region would have a
disproportionate fraction of the sites. Such a condition would be more likely to significantly bias the nationwide
estimates and might have left an important region uncovered. Accordingly, 92 sites were picked at regular
intervals along the coast, accounting for 20% of the U.S. coastal zone. (This paper discusses only the results
from a subsample of 46 sites for the Park et al. study, while the study of Weggel et al. includes the entire
sample.) These studies did not consider Hawaii or Alaska
In presenting results from the studies of Park et al. and Weggel et al., we group the sites into seven coastal
regions, four of which are in the Southeast: New England, Mid-Atlantic, South Atlantic, South Florida/Gulf
Coast Peninsula, Louisiana, Other Gulf (Texas, Mississippi, Alabama, Florida Panhandle), and Pacific Coast.
Figure 4 illustrates these regions.
LOSS OF COASTAL WETLANDS AND DRYLANDS
Park et al. sought to test a number of hypotheses that previous publications had put forth:
(1) A rise in sea level greater than the rate of vertical wetland accretion would result in a net loss of
coastal wetlands.
(2) The loss of wetlands would be greatest if all developed areas are protected (total protection), less if
only densely developed areas are protected (standard protection), and least if shorelines retreat
naturally (no protection).
(3) The loss of coastal wetlands would be greatest in the southeast, particularly Louisiana.
Park et al. applied the same procedures to the nationwide study that had been used in the case study.
The only major difference was that for sites in the Southeast, where they considered the gradual replacement
of salt marshes by mangrove swamps as areas became warmer.
5-15
-------
Titus
Lost Rent From Not
Raising the Island
1MO
tiae
1140
Figure 3. Annual cost of elevating Long Beach Island versus lost economic rent from not raising island.
5-16
-------
Titus
WEST COAST
\
OTHER \ LOUISIANA1,
GULF ^ ,
NORTHEAST
SOUTHERN
WEST FLORIDA
Figure 4. Coastal regions used in this study.
5-17
-------
Titus
The greatest uncertainty in the analysis of this study comes from the poor understanding of vertical
accretion rates. This uncertainty could substantially affect the results for the baseline and 0.5-meter sea level
rise scenarios. However, for a rise of one meter by 2100, there is no evidence that wetland accretion could
keep pace with the 1- to 2-cm/yr accelerated rise that the scenario implies for the second half of the 21st
century. For a 2-meter rise, the uncertainty regarding accretion rates has little, if any, practical significance; no
natural amount of accretion would be able to keep up with sea level.
Another limitation is that Park et al. do not consider the potential implications of alternative ways of
managing river flow. This is a particularly serious limitation for application to Louisiana, where a wide variety
of measures have been proposed for increasing the amount of water and sediment delivered to the wetlands.
Finally, the study makes no attempt to predict which undeveloped areas might be developed in the next century.
If currently undeveloped areas are developed and protected, wetland losses will be higher.
DEFENDING SHELTERED SHORES
Weggel et al. sought to estimate the cost of protecting developed areas along sheltered waters. Their
approach was to examine a number of index sites in depth to develop generalized cost estimates for protecting
different types of shorelines; to use the topographic information collected by Park et al. for the sample sites to
determine the area and shoreline length that had to be protected; to apply the cost estimation factors to each
site; and to extrapolate the sample to the entire coast.
After assessing Long Beach Island, less detailed studies of protecting developed areas from a 2-meter rise
were conducted at five other index sites: metropolitan New York; Dividing Creek, NJ.; Miami and Miami
Beach; the area around Corpus Christi, Texas; and parts of San Francisco Bay. Weggel et al. then developed
high- and low-cost estimates for the entire sample of sites, based primarily on shoreline lengths for the other
sites provided by Park et al. The high estimates assumed that levees would have to be built; the low estimates
assumed that only bulkheads would be necessary. The estimates netted out costs that would normally occur
without sea level rise, such as rebuilding existing bulkheads as they reached the end of their useful lives.
The most serious limitation of this study is that cruder methods are used for the extrapolation than for
the index sites. Even for the index sites, the cost estimates are based on the literature, not site-specific designs
that take into consideration wave data for bulkheads or the potential savings from tolerating substandard roads.
The exclusion of drainage systems in the nationwide costs understates the cost of protection since drainage
systems will be necessary for areas protected by levees. The engineering assessment of Weggel et al. does not
assess the environmental impacts of artificial drainage on water quality. Finally, the investigators were able to
examine only one scenario: a 2-meter rise by 2100. Although the 1-meter rise is more likely, we chose to
interpolate the impacts of a 1-meter rise from the 2-meter estimates of Weggel et al.; we felt this would yield
more accurate estimates than if we chose to extrapolate from a 1-meter rise to a 2-meter rise.
Compared with the Leatherman and Park et al. studies, the methods of Weggel et al. are crude. This
relative inaccuracy results more from the relative difficulty of achieving the Weggel et al. objective than from
failure on the investigators' part to employ better methods. While literature, maps, and remote sensing provided
Leatherman and Park et al. with sufficient data for all sites, a similarly valid sample for Weggel et al. would have
required a few dozen prohibitively costly engineering assessments.
Nevertheless, the approach of Weggel et al. seems sufficient to provide a useful, conservative (that is, likely
to understate the cost) first approximation. It is useful because it considers the length of shorelines that would
have to be protected and uses typical cost-estimation factors for bulkheads and levees that should be accurate
within a factor of two for a large sample. It is conservative because it does not include the cost of the extensive
drainage systems that would accompany levees. In the Long Beach Island case, Weggel et al. estimate that the
drainage system would be almost as expensive as the levees. Barrier islands have a large amount of shoreline
for a small area; because drainage costs primarily depend on the area being drained rather than upon shoreline
5-18
-------
Titus
length, the costs for mainland areas could be several times the cost of building levees. Thus it is possible that
the nationwide cost estimates of Weggel et al. are severalfold too low.
RAISING THE PROFILE: A SIMPLE PROCEDURE FOR ESTIMATING SAND REQUIREMENTS DUE
TO SEA LEVEL RISE
Several studies on the impacts of sea level rise have estimated shoreline retreat, but only two studies of
Ocean City, Maryland (Everts, 1985; Kriebel and Dean, 1985) have estimated the quantity of sand necessary to
maintain the current shoreline, and the methods employed by those researchers require substantial amounts of
data However, a rough estimate can be developed simply by assuming that the entire beach profile is raised
as much as the sea rises. For the this analysis, we define the variables as follows:
D = duration.
E = erosion.
H = vertical height of the beach profile from dune crest to closure depth
(outer point of sediment transport).
L = horizontal length of the beach profile from dunes to point of closure.
P = average slope of the beach profile.
S = sand volume required to raise beach profile by amount of sea level rise.
SLR = sea level rise for a particular scenario by the year 2100 in feet.
The assumption of raising the beach profile by the amount of sea level rise is a corollary of the Bruun
rule (1962),
(1) E = SLR * L / H.
Erosion can be counteracted if S = E*H. Substituting E=S/H into equation 1 and multiplying by H, we
have,
(2) E = SLR * L,
which is the same as raising the entire beach profile.
In the Ocean City report, Titus (1985) called this approach "Bruun" because the study was designed to
compare estimates of shoreline retreat and sand requirements for different methods, and there was no point in
changing names. However, in this report, we use the term "raising the profile method" because it more
accurately conveys the procedure. Moreover, the fact that it is a corollary of the Bruun rule does not mean that
all of the requirements for the Bruun rule must be satisfied for it to apply. Critics of the Bruun rule such as
Devoy (1987) are concerned with the two-dimensional formulation's inability to predict the alongshore transport
(e.g., from headland to embayment) that might be induced by sea level rise, as well as the fact that it ignores
present day alongshore sand transport. However, when the objective is to estimate the sand requirements for
an entire coast, the net alongshore transport is negligible.
5-19
-------
Titus
Both methods share the requirement of defining the profile. Strictly speaking, we should not view the
profile as f(x), but as f(x, D, epsilon), with D equal to the period of time the profile has had to adjust to sea
level rise, with epsilon equal to the rate of sand transport viewed as sufficient for a location to be considered as
being within the profile, and with x and f(x) equal to the horizontal and vertical dimensions, respectively. The
domain of this function would extend farther inland and out to sea as D increases and as epsilon decreases (and
hence the profile would be longer).
If epsilon is equal to the amount of transport necessary to bury one's feet in 60 seconds and D equals 10
minutes, the profile would be confined to the relatively small breakers and swash zone. If D is 12 hours with
the same epsilon, the profile may extend over an extra 100 feet or so as tide goes in and out.
For a smaller epsilon, with D equal to a year, the profile might extend out to around the -20* contour
along much of the Atlantic coast, and with a 50-year D, out to the -30* contour. Hallermeier (1981) outlines
procedures for estimating the profile length for a particular D; topographic maps and navigation charts enable
one to estimate height given the length. Thus, a time-dependent application of the Bruun rule would be,
0
(3) E(t) = / SLR'(t-D) * F6D,
-oo
where SLR' equals the rate of sea level rise and I" is the derivative of the ratio L(D)/H(D). Similarly, a
time-dependent means of raising the profile would be,
0
(4) S(t) = / SLR'(t-D) * L'(D)«D.
-oo
As a practical matter, the profile does not get much longer as D increases beyond SO years, and most
people are content to pick a single value for D and use Hallermeier's estimates; Leatherman's study also follows
this convention. However, we hope that future studies will use more general formulations such as:
T
(5) E(t) = £ [SLR'(t-D) * (L(D)/H(D) - L(D-1)/H(D-1)}]
D=0
where T ranges from 50 to 100 years, and
T
(6) S(t) = £ [SLR'(t-D) * (L(D) - L(D-1)}].
D=0
The function L(D) could be approximated by fitting a polynomial through published estimates for specific values
of D; H(D) could be approximated with whatever functional specification one uses to describe the shape of the
beach profile.
5-20
-------
Titus
We note that our formulation assumes that over a period of D years the profile adjusts to a D-year storm.
This represents a maximum likelihood estimate, but not necessarily an unbiased or median estimate. For
example, while we assume that the worst storm in a 100-year period will be the 100-year storm, the probability
that the worst storm will be milder is .99100= 37%, while there is a 63% chance that it will be worse.
Given that Leatherman did not have time to employ the more general formula, he had to pick a value for
D. Because the general approach was to underestimate the cost of sea level rise, Leatherman picked a 1-year
storm.
COST OF PROTECTING THE OPEN COAST
Leatherman applied the "raising the profile" methods for all recreational beaches from Delaware Bay to the
mouth of the Rio Grande plus California, which account for 80% of the nation's beaches. He also examined one
representative site in each of the remaining states.
Leatherman's analysis provides a state-of-the-art assessment of the beach nourishment costs for the nation,
with two caveats: (1) Although the sample of sites in the Northeast and Northwest are representative, complete
coverage would have been more accurate; and (2) Leatherman used very conservative assumptions in estimating
unit costs of sand. Generally, a fraction of the sand that is placed on a beach washes away because of insufficient
grain size; and as the least expensive sand is used and dredges have to go farther offshore, sand costs will
increase. For Florida, Leatherman used published estimates of the percentage of fine-grain sand, and assumed
that the dredging cost would rise $l/cubic yard for every additional mile offshore the dredge had to go. For the
other states, however, he assumed that the deposits mined would have no fine-grain sand and that dredging costs
would not increase. Leatherman is also underestimating sand costs by assuming that the beach profile extends
out only to the point where the annual severe storm would deposit sand.
A final limitation of the Leatherman study is that it represents the cost of applying a single technology
throughout the ocean coasts of the United States. Undoubtedly, there are areas where communities would
choose to erect levees and seawalls-particularly Galveston and other wide barrier islands in Texas-or to accept
a natural shoreline retreat.
5-21
-------
Titus
CHAPTER 6
RESULTS
RESULTS: NATIONWIDE LOSS OF WETLANDS AND DRYLANDS
Estimates of Park et al.
Table 3 compares the current area of wetlands estimated by Park et al. with the estimates from a recent
NOAA inventory of coastal wetlands (Alexander et al., 1986). For the nation and for five of the seven regions,
the differences between the estimates are within the sampling margin of error. However, the sample of Park
et al. substantially underestimates the coastal wetlands of the middle Atlantic and the Pacific Coast, which
together account for about 15% of U.S. coastal wetlands.
The Park et al. results generally supported the hypotheses suggested by previous studies (see Table 4).
For a 1-meter rise in sea level, 66% of all coastal wetlands would be lost if all shorelines were protected, 49%
would be lost if only currently developed areas were protected, and 46% would be lost if shorelines retreated
naturally.13
As expected, the greatest losses would be in the Southeast. Figure 5 illustrates the loss of coastal wetlands
for this region for three scenarios of shore protection. Even under the baseline scenario, a substantial fraction
of Louisiana's wetlands would be lost. (See Chapter 6 of the main report to Congress.) Most other areas would
experience slight gains in wetland areas. Of the 6046-8673 square miles of U.S. wetlands that would be lost
from a 1-meter rise, 90-95% would be in the Southeast.
Analysis: Sampling Error
Table 4 illustrates wetland loss (in square miles and percentage terms) by region for each of the scenarios,
along with sampling error (i.e., the standard deviation times the square root of the sample size). For the total
protection scenario, the estimated loss of wetlands was greater than the sampling error for all regions, and hence
could be viewed as statistically significant at the 60-70% level of confidence. However, to be significant at the
95% level, Student's-t distribution would require the losses to be 2.1-3.1 times as great as error (depending on
sample size).14 At this level of confidence, only the South Atlantic, Louisiana, and (barely) Mid-Atlantic show
statistically significant losses for the total protection scenario; for other shore protection scenarios, only the
South Atlantic and Louisiana are significant.
Given the small regional samples, the lack of statistical significance at the 95% level for area of wetlands
lost could have been expected. However, we note that the uncertainty results largely from the fact that different
sites had varying amounts of initial wetland coverage. We had hoped that this problem could be circumvented
by considering percentage losses of coastal wetlands. Unfortunately, we found large standard deviations for
percent losses as well, largely because most of the regions had at least one outlier (e.g., most of the sites in a
region show 40-60% losses but one site shows a 1000% gain). We did not have time to undertake more
13When all shorelines are protected, as sea level rises, the protective structures limit wetlands forming upland
everywhere. On the other hand, with protection limited to developed shorelines, wetlands can form upland in
undeveloped areas. Thus, the net area of wetlands is less after sea level rise under the total shore protection
scenario than under the standard shoreline protection scenario.
14In the following analysis of wetlands and dryland losses, we have presented our results using 95%
confidence intervals. We chose this level because we believe that even the losses at the low end of the interval
are high enough to induce decision makers to plan ahead for sea level rise.
5-22
-------
Titus
Table 3. Comparison of Park Baseline Data For Vegetated
Wetlands (1985) and NOAA Wetlands Inventory
(Square Miles)
Region
(Sample Sites)
Percent
of Coast
In Sample
Park
Estimate
Sampling
Error
NOAA
Estimate
Northeast (4) 3.4
Mid-Atlantic (7) 8.6
South Atlantic (8) 10.1
Louisiana (7) 13.7
Gulf Except LA (14) 12.2
W/SW Florida (6) 10.7
Other Gulf (8) 13.1
West (6) 4.9
USA 9.7
600
746
3813
4835
3087
1869
1218
64
13145
389
245
848
876
1169
957
673
45
2105*
382
2080
3967*
4491
3608
na
na
195
14723
na - not available.
Alexander et al. (1986) also estimate the area of tidal flats for several
states; we present only the sum of their estimates for vegetated wetlands.
We have modified data from Alexander et al. to account for differences in the
definition of coastal wetlands for North Carolina. Alexander et al. include
all wetlands in coastal counties regardless of elevation, while Park et al.
excluded wetlands above 12 ft NGVD. Because of extensive swamps above 12 ft
in North Carolina's coastal counties Alexander et al. found the area of
coastal swamps to be 8.4 times the area of marsh, while the boundaries of the
sample of Park et al. found only 1.6 times as much.
Standard deviation of the estimate of the sum (i.e., sample standard
deviation times the square root of the sample size).
5-23
-------
Titus
-------
Table 4. Regional and National Coastal Wetland Losses
Region
Northeast
Best estimate (sq mi)
Best estimate (X)
Sampling error (sq mi)
Mid-Atlantic
Best estimate (sq mi)
Best estimate (X)
Sampling error (sq mi)
South Atlantic
Best estimate (sq mi)
Best estimate (X)
Sampling error (sq mi)
Standard
Protection
Baseline
39
7
20
-39
-5
87
59
-2
201
South/Gulf Coast of Florida Peninsula
Best estimate (sq mi)
Best estimate (X)
Sampling error (sq mi)
Louisiana
Best estimate (sq mi)
Best estimate (X)
Sampling error (sq mi)
-157
-8
110
2271
52
421
50 cm
88
15
56
485
65
197
2295
60
526
623
33
274
2450
56
338
Florida Panhandle. Alabama. Mississippi. Texas
Best estimate (sq mi)
Best estimate (X)
Sampling error (sq mi)
270
22
209
530
44
279
Total
Protection
100 cm
93
16
60
520
70
200
2422
64
506
829
44
380
3742
85
735
1031
85
761
200 cm
100
17
64
625
84
208
2542
67
510
1020
55
477
4758
99
882
1121
92
812
50 cm
55
9
29
201
27
273
1438
38
516
92
5
230
2368
54
385
396
33
306
Standard
Protection
100 cm
58
10
34
341
46
238
1669
44
558
157
8
313
3732
85
735
932
77
779
200 cm
33
6
25
429
58
247
1812
48
621
165
9
474
4686
99
902
994
82
821
50 cm
27
5
43
120
16
270
1313
34
517
63
3
235
2354
54
385
360
30
319
No
Protection
100 cm 200 cm
8
1
69
280
38
232
1516
40
573
122
7
320
3732
85
735
918
75
779
-67
-11
112
361
48
241
1606
42
656
120
6
481
4685
99
902
982
81
823
H
*•*
£
-------
Table 4. Regional and National Coastal Wetland Losses (continued)
West Coast
Best estimate (sq mi)
Best estimate (X)
Sampling error (sq mi)
Southeast (original 11.735)
Wetlands Lost:
Best estimate (sq mi)
Best estimate (X)
95X low (sq mi)
95X high (sq mi)
Wetlands Left: b
Best estimate (sq mi)
Best estimate (X)
95X low (sq mi)
95X high (sq mi)
l
United States (original 13.145)
Wetlands Lost:
Best estimate (sq mi)
Best estimate (X)
- 95X low (sq mi)
95X high (sq mi)
-*»
Wetlands Left:6
Best estimate (sq mi)
Best estimate (X)
95X low (sq mi)
95X high (sq mi)
-71
-111
44
2,325
+20
a
a
NC
NC
NC
NC
2255
17
1168(9)
3341(25)
NC
NC
NC
NC
37
58
22
5899
50
4408(37)
7390(63)
NC
NC
NC
NC
6511
50
4944(38)
8077(61)
NC
NC
NC
NC
36
56
21
8024
68
5535(47)
10513(90)
2843
24
779
4907
8673
66
a
a
4472
34
2302(18)
6642(51)
39
61
23
9443
80
a
a
2294
20
848
3740
10206
78
a
a
2897
22
1302(10)
4492(34)
-286
-447
202
4296
37
2783(24)
5809(50)
NC
NC
NC
NC
4263
32
2591(20)
5934(45)
NC
NC
NC
NC
-440
-688
282
6491
55
3976(34)
9006(77)
NC
NC
NC
NC
6441
49
3813(29)
9068(69)
NC
NC
NC
NC
-651
-1017
336
7647
65
4722(40)
10572(90)
NC
NC
NC
NC
7423
56
4350(33)
10495(80)
NC
NC
NC
NC
-332
-519
209
4090
35
2563(22)
5617(48)
NC
NC
NC
NC
3904
30
2216(17)
5592(43)
NC
NC
NC
NC
-518
-809
280
6289
54
3572(30)
8826(75)
NC
NC
NC
NC
6046
50
3388(26)
8703(66)
NC
NC
NC
NC
-791
-1236
371
7342
63
4379(37)
10305(88)
NC
NC
NC
NC
6892
52
3758(29)
10025(76)
NC
NC
NC
NC
NC = Not calculated.
a = Confidence intervals not calculated for cases where sampling error exceeds best estimate.
= Wetlands left only calculated for cases when sampling error exceeded best estimate for wetlands lost.
-------
Titus
Table 5. Loss of Dryland
Northeast
(Square Miles)
Baseline 50 cm 100 cm 200 cm
No Protection nc 139 235 472
Standard Protection 32 71 126 262
Mid Atlantic
No Protection nc 904 1205 1771
Standard Protection 448 705 928 1385
South Atlantic
No Protection nc 1094 1600 2561
Standard Protection 493 886 1272 2023
South Florida and Gulf
Coast of Florida Peninsula
No Protection nc 768 1278 2035
Standard Protection 272 717 1196 1907
Louisiana
No Protection nc 1364 1417 1638
Standard Protection 1178 1249 1295 1449
Florida Panhandle. Alabama.
Mississippi, and Texas
No Protection nc 905 1091 1548
Standard Protection 563 809 976 1405
Pacific Coast
No Protection nc 511 903 1771
Standard Protection 92 444 867 1537
United States
No Protection
Best Estimate nc 5313 7727 11793
Error nc 989 1289 1783
95% High nc 7311 10330 15394
95% Low nc 3315 5123 8191
Standard Protection
Best Estimate 3078 4164 6661 9967
Error 804 982 1250 1747
95% High 4686 6147 9186 13496
95% Low 1470 2180 4136 6438
5-27
-------
Titus
Table 5. Loss of Dryland (continued)
(Square Miles)
Baseline 50 cm 100 cm 200 cm
Southeast
No Protection
Best Estimate nc 4131 5386 7782
Error nc 890 1084 1478
95% High nc 5929 3196 4796
95% Low nc 2333 7576 10767
Standard Protection
Best Estimate nc 3661 4739 7116
Error nc 888 1075 1460
95% High nc 5455 6910 10065
95% Low nc 1867 2567 4166
nc - not calculated.
5-28
-------
Titus
sophisticated approaches such as discarding outliers; hence we simply accepted the lack of significance at the
regional level.
The one instance where we were able to reduce estimated sampling error concerns the "total protection"
scenario. In a number of cases, we found that the standard deviation of remaining wetlands was much less than
that for wetlands lost, as the table illustrates.
Despite the lack of significance for most regions, at the nationwide and southeast-wide levels of aggregation,
the results are highly significant. The 95% confidence intervals for the nationwide percentage wetland loss are
38-61, 50-66, and 66-90% for the 50-, 100-, and 200-cm sea level rise scenarios and the total protection case and
17-43, 26-66, and 29-76% for the no-protection scenario. Statistical significance for the loss of dryland followed
the same pattern. The best estimates indicate that if shorelines retreat naturally (no protection), a 1-meter rise
would inundate 7727 square miles of dryland, an area the size of Massachusetts; with a 2-meter rise, 11,793
square miles could be lost. Again, most of the land loss would occur in the Southeast, particularly Florida,
Louisiana, and North Carolina. The corresponding 95% confidence intervals are 3,000-8,000, 5,000-10,000, and
8,000-15,000 square miles lost for the 50-, 100-, and 200-cm sea level rise scenarios, respectively. Of course, with
total protection of coastal lowlands, there would be no losses for any of the sea level rise scenarios.
RESULTS: THE NATIONWIDE COST OF PROTECTING SHELTERED SHORES
Estimates of Weggel et al.
Table 6 shows estimates from Weggel et al. for the index sites and the nationwide estimate. The index sites
represent two distinct patterns. Because urban areas like New York would be entirely protected by levees, the
cost of moving buildings and rebuilding roads and utilities would be relatively small. On the other hand,
Weggel et al. concluded that in more rural areas like Dividing Creek, NJ., only the pockets of development
would be protected. The roads that connected them would have to be elevated or replaced with bridges, and
the small number of isolated buildings would have to be moved.
Weggel et al. estimate that the nationwide cost of protecting developed shorelines from a 2-meter rise in
sea level would be $25 billion if only bulkheads are necessary and $80 billion if levees are required. Unlike
wetland loss, the cost of protecting developed areas from the sea would be concentrated more in the Northeast
than the Southeast, because a much greater portion of the coast is developed in the Northeast. (The Southeast
still accounts for a large percentage of total costs owing to its majority share of the U.S. sheltered shorelines.)
Analysis: Interpolating Results of Weggel et al. for 0.5- and 1-Meter Sea Level Rise Scenarios
Our objectives were to (1) interpolate the 2-meter sea level rise cost estimates developed by Weggel et al.
to the 50 and 100-cm scenarios, (2) develop statistical confidence intervals of the costs of protection, and (3)
explicitly consider whether particular sites would be protected with levees or bulkheads.
Weggel et al. assumed that even in the baseline scenario, bulkheads must be rebuilt every ten years. Their
estimate for the cost of sea level rise is the cost of the additional height required by sea level rise. He assumes
that in the baseline scenario, a five-foot bulkhead is necessary, at $130 or $500 per foot, and that costs rise with
height to the 1.5 power. Thus, if SLR(t) represents the sea level rise in feet by the year t, the cost of bulkheads
for the $130/foot estimate is simply,
(7) Bulkhead Cost - ((5 + SLR)/5)15 * 130,
15One reviewer noted that the cost of protecting Miami, Florida, may be too low. The city is located on a
porous limestone base, a factor that may cause severe seepage and drainage problems.
5-29
-------
Titus
Table 6. Cost of Protecting Sheltered Waters
Against a 2-Meter Sea Level Rise
(Millions of
New
Index Sites Bulkhead
New York 57
Long Beach Island 3
Dividing Creek 4
Miami Area 11
Corpus Christi 11
San Franciso Bay 3
Nationwide Estimate
Northeast
Mid Atlantic
Southeast
West
USA
Source: Weggel et al.
This is the cost for the
**
Raise Old
Bulkhead
205
4
6
111
29
19
Low
6,932
4,354
9,249
4,097
24,633
1986 Dollars)
Raise
Move Roads &
Building Utils.
0.5 9.5
2.7 3.8
4.8 18.2
0.3 8.3
2.8 40 . 9
2.0 20.0
High
23,607
14,603
29,883
12,802
80,176
Total
272.3
13.7
33.0
130.7
83.4
44.0
low estimate only.
Assumes no extraordinary seepage problems.
5-30
-------
Titus
and the incremental cost due to sea level rise is,
(8) Cost = ((5 + SLR)/5)1 5 * 130 130
We now present our procedure for the 1-meter scenario; the 50-cm scenario is analogous. The ratio of
costs due to sea level rise incurred for a particular year between the 2-meter and 1-meter scenarios is,
(9) RATIO(t) = COST_lm(t)/COST_2m(t) =
(5+SLR_lm(t)/5)15
(5+SLR_2m(t)/5)1'5 -
(Although the elasticity of total cost with respect to sea level rise is 1.5, the elasticity of costs due to sea level
rise is only 1.08 over the 50- to 200-cm range.)
We are interested in estimating the cumulative cost for the 50- and 100-cm scenarios, which requires
considering,
2100 2100
(10) £ COST_lm(t)/£COST_2m(t)
t=1986 t=1986
Because Weggel et al. reported the denominator (i.e., the sum) and not the cost for specific years, we could
not calculate this ratio precisely. Instead, we use a conservative approximation, COST_lm(2100)/
COST_2m(2100). (See Appendix 1 for proof that this ratio provides a conservative approximation.)
We wanted our analysis to explicitly consider the suggestion of Weggel et al. that $130/foot applies to areas
above sea level that simply need bulkheads, while $500/foot applies to areas that would be inundated and hence
need levees and pumping systems. Unfortunately, Weggel et al. were not able to determine the portion of
developed shores that would require levees. However, Park et al. provide estimates of lost lowland. We
assumed that the percentage of the developed shoreline requiring a levee would be equal to the percentage of
coastal lowlands (below 12 feet NGVD) flooded by spring high tide under the no-protection scenario.16
Thus, we define the cost for protecting developed sheltered shorelines as,
(11) Cost = [500/130 * (Lowland. Lost/Lowlands. 1986) + (l-Lowland_ Lost /Lowlands, 1986)]
* Weggel_Cost_130 * Ratio(SLR)
where:
Weggel_Cost_130 = Weggel's estimate of the cost of protecting a site in
the 2-meter scenario assuming $130/foot.
Lowlands. Lost = Area of lowlands lost at time t.
T"his assumption is conservative in that it underestimates the area of land that would need levees. Even
today, cities like New Orleans that are completely protected by levees often have substantial areas above sea
level, for two reasons: (1) even if only a small portion of the total land is low enough to need a levee, the entire
shoreline can consist of lowland that needs protection, and (2) levees may be needed to protect areas from
flooding during storms. The fraction of the shore requiring a levee would be less than the fraction of area below
sea level only in unusual cases, such as a site with a straight lowland shore accompanied by uplands that jut into
the sea like fingers.
S-31
-------
Titus
Lowlands. 1986 = Area of lowlands in 1986.
Ratio(SLR) = ((5 + SLR)/(5+6.56))15
SLR = Sea level rise for a particular scenario by year 2100 in feet.
6.56 = The rise in sea level by 2100 in feet for the 2-meter scenario.
Using this equation, we calculated the cost of protecting developed shores for each of the 46 sites in the Park
et al. subsample.
As a final step, we sought to incorporate the additional information from the other 49 sites Weggel et al.
examined. As Table 7 shows, the estimates from the full sample were within the sampling error for the
Northeast, mid-Atlantic, and Southeast. However, by chance, the Park et al. subsample of the west coast had
excluded all of the sites that would require significant amounts of bulkheads and levees; the full sample results
in an estimate over six times as great. We followed the simple procedure of adjusting the estimates from
equation 11 by the ratio of the Weggel et al. cost estimates for each of the four major areas of the nation; we
adjusted statistical error by the ratio of errors for the two samples.
5-32
-------
Titus
Table 7. Comparison of Weggel Low Cost Estimates For Park's
Subsample and Full Sample
(Billions of 1986 Dollars)
Sub sample Full Sample
estimate error estimate error
Northeast 4.91 3.28 6.93 3.65
Mid-Atlantic 2.85 1.23 4.35 2.45
Southeast 11.22 3.43 9.25 2.12
West 0.65 0.02 4.10 0.37
Ratio*
estimate error
1.41 1.11
1.53 1.99
0.82 0.62
6.31 23.40
Ratio of Full Sample to Subsample.
5-33
-------
Titus
Table 8 illustrates estimated costs of protecting sheltered shores for both the sample and subsample for
each of the four major regions, and confidence intervals for the nation. For the three sea level rise scenarios,
our estimated confidence intervals are 5-13,11-33, and 29-100 billion dollars. Thus, for the nation at large, the
elasticity of total cost with respect to sea level rise is 1.4; that is, a quadrupling of sea level rise from 50 to 200
cm increases costs sevenfold. The elasticity would have been greater if the levee estimates had included the
cost of drainage systems.
RESULTS: NATIONWIDE COST OF PROTECTING THE OPEN COAST
Leatherman Estimate of Sand Costs
Table 9 illustrates Leatherman's estimates. A total of 1920 miles of shorelines would be nourished. An
area of 931 square miles would be raised, 235 of this after a 1-foot rise in sea level. As the table shows,
two-thirds of the nationwide costs would be borne by four southeastern states: Texas, Louisiana, Florida, and
South Carolina. Figure 6 illustrates the cumulative nationwide costs over time. For the 50 and 200 cm
scenarios, the cumulative cost would be $2.3-4.4 billion through 2020, $11-20 billion through 2060, and $14-58
billion through 2100.
Analysis: An Increasing-Cost Scenario for Sand
In the past, we have shown that if unit sand costs increase substantially over time, a community that
chooses to hold back the sea at first may eventually decide to migrate landward (Titus, 1987). However, that
analysis was based on hypothetical increases in dredging costs. We wanted this study to provide at least a
first-order estimate of how costs might escalate. In this section, we use the sand cost function Leatherman
developed for Florida to develop an increasing-cost scenario. We emphasize that unlike our other estimates, no
statistical interpretation can be attached to this estimate. We hope that this crude estimate encourages other
researchers to consider cost-escalation in the future.
Leatherman's cost function for total available sand off Florida's Atlantic coast was based on the following:
Distance offshore Available Sand Unit Cost
(miles) (millions cu yd) (dollars)
0-1 66 4
1-2 87 5
2-3 122 6
3-4 48 7
4-5 0 8
5+ Plenty 10
17Elasticities are used to measure the effect a change in one variable has upon another. In this case, the
elasticity is calculated with the equation In^/C^/ln^LR^SLRg), where "C" is cost and "SLR" is sea level
rise.
5-34
-------
Titus
Table 8. Cost of Protecting Developed Sheltered Shorelines Through 2100
(Billions of 1986 Dollars)
Park Subsample Weggel Full Sample
Base-
Region** line 50 cm 100 cm 200 cm 50 cm 100 cm 200 cm
Northeast (4/8)
total 0.41
error 0.29
Mid Atl (7/15)
total 0.31
error 0.11
S. Atl (8)
total 0.58
error 0.23
S/W Fl. (6)
total 0.15
error 0.13
Louisiana (7)
total 0.11
error 0.06
Other Gulf (8)
total 0.07
error 0.03
Southeast (29/54)
total 0.91
error 0.27
95% low nc
95% high nc
Pacific (6/17)
total 0 . 04
error 0.00
United States
total 2 . 00
error 0.41
95% low 2.80
95% high 1.20
I
1
1
1
0
2
1
0
0
0
0
0
0
4
1
0
0
7
1
.89
.34
.33
.54
.86
.09
.65
.56
.37
.18
.28
.11
.16
.24
nc
nc
.14
.00
.52
.90
nc
nc
4
3
3
1
7
3
1
1
0
0
0
0
10
3
0
0
19
5
.41
.84
.35'
.45
.75
.13
.64
.39
.65
.42
.81
.24
.82
.46
nc
nc
.29
.01
.86
.37
nc
nc
16
12
9
4
21
8
4
8
2
1
1
0
29
12
0
0
55
17
.06
.00
.13
.31
.64
.91
.44
.87
.12
.21
.64
.65
.84
.65
nc
nc
.65
.02
.68
.94
nc
nc
2.66
1.49
2.
1.
3.
0.
1.
4.
0.
0.
9.
1.
4.
13.
.03
,07
nc
nc
nc
nc
nc
nc
nc
nc
43
77
87
99
88
08
00
99
98
02
6.22
4.26
5.12
2.88
nc
nc
nc
nc
nc
nc
nc
nc
8.91
2.14
4.59
13.23
1.82
0.16
22.07
5.57
10.82
33.32
22
13
13
8
24.
7.
8.
40.
4.
0.
65.
17.
29.
100.
.64
.32
.97
.58
nc
nc
nc
nc
nc
nc
nc
nc
59
82
79
39
10
37
30
67
60
99
Full sample estimates are based on the ratios calculated in Table 7.
Baseline was not calculated for the full sample.
Numbers in parenthesis after each region are the number of sites in the
subsample and full sample, respectively.
5-35
-------
Titus
Table 9. Cumulative Cost of Placing Sand on U.S. Recreational
Beaches, Coastal Barrier Islands, and Spits
(Millions of 1986 Dollars)
Baseline 50 cm 100 cm
200 cm
Maine**
**
NH ^
Mass
RI**
**
CT
**
NY^
NJ**
Del
MD
VA
NC
SC
GA
FL (AT)*
FL (G)
AL
MS
LA
TX
CA**
OR**
**
WA
HA**
Nation
SE
23
8
168
16
102
144
158
5
6
30
137
184
26
120
149
11
13
1956
350
36
22
52
74
3,790
2,946
119
39
490
92
516
770
902
34
35
201
656
1158
154
787
904
59
72
2623
4188
174
61
143
338
14,515
10,601
217
73
842
161
944
1374
1733
71
83
387
1271
2148
263
1938
1688
105
128
3493
8490
324
153
360
647
26,893
19,524
412
142
1546
298
1800
2581
3493
162
213
798
3240
4348
640
8565
4092
260
370
5232
17608
626
336
794
1268
58,824
44,355
Incremental cost due to relative sea level rise only.
Indicates states where estimate was based on extrapolating a representative
site to the entire state. All other states have 100% coverage.
* Florida Atlantic estimates account for the percentage of fine grain
sediment, which generally washes away, and for cost escalation as least
expensive sand deposits are exhausted. All other estimates conservatively
ignore this issue.
Source: Baseline Costs derived from Leatherman.
5-36
-------
60
SO
40
30
20
Titus
200 CM
100 CM
60 CM
1980 2000 2020 2040 2060 2080 2100
Figure 6. Nationwide cost of sand for protecting ocean cost.
5-37
-------
Titus
The sand quantities are based on surveys by the Corps of Engineers. Unit cost estimates up to five miles
assume that dredges pump the sand onto the shore, and use the generally accepted rule-of-thumb that each
additional mile offshore adds $l/cubic yard for the booster pumps that would be necessary. Leatherman
assumed that for distances greater than five miles, pipelines would be infeasible, and barges and dump trucks
would be employed at a cost of $10/cubic yard, regardless of how far offshore one travels. Leatherman did not
consider the possibility that improved technologies would reduce costs, nor the possibility that higher energy
prices would increase costs.
Let Cp^SAND) represent this cost function for Florida, and C§(SAND) represent the function for a state
O*
Ideally, we would like dC/dSAND to recognize the differences in sand availability, state-specific economic
and environmental factors that influence the cost of dredging, and the fact that, all else equal, the amount of
sand a particular distance from the shore is proportional to the amount of coastline. By scaling the Florida
equation using state-specific data provided by Leatherman, we can accurately account for the latter factor and
crudely attempt to account for the first two.
First, we scale the cost function Cp, by 293 miles, Leatherman's estimate of the length of the recreational
beaches of Florida's Atlantic coast. This new function, which we call C , refers to the cost of nourishing one
mile of beach. The unit cost of nourishing one mile of beach is C '. Then,
(12) C*(SAND) = CFL(SAND/293),
and
(13) C*'(SAND) = CFL'(SAND/293).
If we define Cg(SAND) as total sand cost and C'g(SAND) as the unit cost for a particular state and a given
amount of sand (e.g., C'(0) is the current unit cost), we can scale for differences in current sand costs and
shoreline lengths:
(14) CS'(SAND) = C*'(SAND / SHORELINE_LENGTH(STATE)) + C$'(0) - $4.00.
This equation simply says that the unit cost of sand for a state increases by the same pattern as the cost in
Florida, but (1) the base cost is whatever Leatherman found it to be for that particular state, and if a state's
shoreline is L times that of Florida, it can dredge L times as much sand as Florida can before it must go
another mile out to sea. Our additive incorporation of the current cost into this equation is probably
conservative for states where the current cost is greater than in Florida, and too liberal in the minor case of
Mississippi where the cost is less. Greater costs for sand often imply that one must already go farther out to
sea than Florida, which may indicate that there is less sand at any distance from shore for such a state than
there is for Florida Although this situation would suggest a multiplicative relation, we decided to keep the
additive formulation because cost diffences due to other factors such as wage rates, average deposit size, and
using barges would not increase with the distance from shore.
Table 10 illustrates the shoreline lengths, base costs, and sand required for the three sea level rise
scenarios for each state. Table 11 shows the implied sand requirements and cost per mile assuming constant
costs. Table 12 shows the cumulative costs by state for the increasing cost scenario. Excluding Louisiana, the
cost elasticity is 1.25 (i.e., costs rise with the 1.25 power of sea level rise).
18Louisiana was excluded from this calculation because even without sea level rise, the Louisiana coast will
require large amounts of nourishment.
5-38
-------
Titus
Table 10. Increasing Marginal Sand Cost Scenario For Dredging Sand
State
Maine
NH
MASS
RI
CT
NY
NJ
DE
MD
VA
NC
SC
GA
FL(Atl)
FL(Gulf)
AL
MS**
LA**
TX
CA
OR
WA
HA
Source :
* _
Developed
Sandy Ocean
Shoreline
(miles)
31
9
100
27
64
120
125
10
9
17
143
93
16
293
251
36
43
85
230
78
28
48
64
Leatherman
Unit Cost
of sand
(S/vd3)
4.00
4.00
7.87
6.00
6.00
7.85
7.85
8.00
6.00
8.00
7.00
4.50
4.00
4.00
4.00
4.00
2.75
5.00
9.25
4.00
4.00
4.00
5.00
San<
Sea Le'
(millior
50 cm
30
10
62
15
86
98
115
4
6
25
94
257
38
177
226
15
26
525
453
44
15
36
68
i Required for
rel Rise Scenario
is of cubic yards)
100 cm 200 cm
54
18
107
27
157
175
221
9
14
48
182
477
74
340
422
26
47
698
917
81
38
90
129
103
35
197
50
300
329
445
20
35
100
463
966
160
1003
1023
65
134
1046
1903
156
84
199
253
The calculations in this table are based on the assumption that National
Parks and Wildlife Refuges would not be protected. Areas included under the
Coastal Barrier Resources Act (COBRA) are not included unless connected to
mainland by a bridge.
None of the barrier islands in Mississippi and only one barrier island in
Louisiana are developed. These calculations assume that all Louisiana
barriers are raised for storm protection, and that the beaches and low resort
communities behind Mississippi's barriers are raised.
5-39
-------
Titus
Table 11.
Average Quantity and Cost of Sand Per Mile of Shoreline
State
Sand per mile
(millions of cubic yards)
50 cm 100 cm 200 cm
Cost
(millions of 1986 dollars)
50 cm 100 cm 200 cm
Maine
NH
MASS
RI
CT
NY
NJ
DE
MD
VA
NC
SC
GA
FL(Atl)
FL(Gulf)
AL
MS
LA
TX
CA
OR
WA
HA
0.97
1.11
0.62
0.56
1.34
0.82
0.92
0.40
0.67
1.47
0.66
2.76
2.38
0.60
0.90
0.42
0.60
6.18
1.97
0.56
0.54
0.75
1.06
1.74
2.00
1.07
1.00
2.45
1.46
1.77
0.90
1.56
2.82
1.27
5.13
4.63
1.16
1.68
0.72
1.09
8.21
3.99
1.04
1.36
1.88
2.02
3.32
3.89
1.97
1.85
4.69
2.74
3.56
2.00
3.89
5.88
3.24
10.39
10.00
3.42
4.08
1.81
3.12
12.31
8.27
2.00
3.00
4.15
3.95
5.0
6.1
5.4
3.7
11.1
7.3
8.3
3.4
4.6
16.0
5.2
24.0
18.8
2.9
4.7
1.9
2.1
63.0
25.1
2.6
2.5
3.8
6.8
12.4
15.0
9.9
7.3
24.4
15.2
19.5
8.3
13.7
34.5
11.5
48.9
41.3
6.6
11.8
3.6
4.6
85.3
55.9
5.6
8.6
13.8
17.2
28.2
33.9
22.3
17.2
51.3
33.0
44.3
23.0
41.7
77.3
37.1
104.1
95.0
29.2
35.8
13.1
22.3
130.4
121.1
15.0
25.0
36.5
38.5
Scenario assumes that distribution of sand off Florida's Atlantic Coast is
typical of sand distribution off all states' coasts.
Note: Cost escalation is based on the equation,
C*'(SAND) - A +-$4.00 for SAND = 0 to 225,000 cubic yards/mile A + $5.00
for SAND - 225,000 to 522,000 cu yd/mi
A + $6.00 for SAND - 522,000 to 938,000 cu yd/mi
A + $7.00 for SAND - 938,000 to 1,100,000 cu yd/mi
A + $10.00 for SAND > 1,100,000 cu yd/mi
<• .' ,*•".•'
for 0-1 miles off shore, 1-2, 2-3, and 5+ miles of shore, respectively
(Leatherman's cost function shows no sand in the 4-5 mile range off Florida's
coast, and our calculations for the other states follow this assumption, also).
C*' is the unit cost of sand and A is the difference between the initial unit
cost for a state and that of Florida ($4.00).
5-40
-------
Titus
Table 12. Cost by State^of Protecting Open Coast Under Increasing Sand
Cost Scenario
(Millions of 1986 Dollars)
50 cm 100 cm 200 cm
Maine
NH
MASS
RI
CT
NY
NJ
DE
MD
VA
NC
SC
GA
FL(Atl)
FL(Gulf)
AL
MS
LA
TX
CA
OR
WA
HA
USA (Increasing Cost)
USA (Fixed Cost)"
Southeast (1C)
USA (1C - Excluding LA)
155
55
540
100
710
876
1,038
34
41
272
744
2,232
301
850
1,180
68
90
5,355
5,773
203
70
182
435
21,304
14,515
16,593
15,949
384
135
990
197
1,562
1,824
2,438
83
123
587
1,645
4,548
661
1,934
2,962
130
198
7,251
12,857
437
241
662
1 , 101
42,950
26,893
32,186
35,699
874
305
2,230
464
3,283
3,960
5,538
230
375
1,314
5,305
9,681
1,520
8,556
8,986
472
959
11,084
27,853
1,170
700
1,752
2,464
99,075
58,824
74,416
87,991
The calculations in this table are based on the assumption that National
Parks and Wildlife Refuges would not be protected. Areas included under
the Coastal Barrier Resources Act (COBRA) are not included unless
connected to the mainland be a bridge.
Leatherman national estimate minus the difference between Leatherman
estimate for Florida (increasing cost) and the estimate implied by a
constant cost of $4.00/cu yd.
5-41
-------
Titus
Analysis: The Cost of Elevating Buildines and Roadways
If all the nation's developed barrier islands were developed like Long Beach Island, we could simply
multiply the unit cost estimates described in the case study by the area of barrier islands that had to be raised.
However, most islands are developed less densely. However, for barrier communities which are more densely
developed, like Ocean City, Maryland, the infrastructure cost would not be proportionately greater. For the
most part, a greater density reflects the presence of high-rises instead of single-family homes; the density of
roads and utilities is not necessarily much greater. Thus, we use the following procedure:
(1) Collect census data for a random sample of coastal barrier communities on the number of buildings
and divide by area to get density.
(2) Calculate the mean values and confidence intervals for densities in three coastal regions: the
Gulf Coast, the Southeast Atlantic (Florida to North Carolina), and the mid and Northeast
Atlantic States.
(3) Develop equations relating cost to shoreline length, area, and density, and apply the equation
to the confidence interval for densities, for each of the three scenarios.
(4) Adjust the estimates (usually downward) for the 50- and 100-cm scenarios to account for the fact
that the relative elevated ocean sides may not have to be raised for these scenarios.
Census Data
Table 13 illustrates census data on densities for the sites in our random sample. We note that there may
be an upward bias, in that the Bureau of Census does not provide data if there are not at least 1000 year-round
residents. On the other hand, there is a downward bias in that some of the barriers are lumped in with a
township that extends to the mainland, which is generally less dense than the barrier. If census data were not
available for a particular site, data from a nearby locale were used. In some instances, data from a nearby
coastal (instead of a barrier) town had to be used. Another problem with census data is that it provides the
number of housing units and the number of single-family homes, but not the total number of buildings, which
we need to estimate road density. Thus, we were forced to use the number of single family homes as a proxy for
building density. This last assumption effectively treats multi-unit structures as vacant lots; it is still more
accurate than treating a condominium with 100 units as 100 houses.
Extrapolation Procedure
With means and confidence intervals for the densities calculated, we can now use the estimates of Weggel
et al. to calculate the cost of elevating infrastructure for the nation's developed barrier islands. Because these
costs are mostly related to roads, we begin with an equation relating Long Beach Island's development density
(and, hence, infrastructure) to other barrier islands' level of development and size:
5-42
-------
Titus
Table 13. Building Densities For a Sample of Coastal Barriers
Site
- (units/mile2) -
Area Housing Single Housing Single
(mile2) Units Units Density Density
Calveston, TX
Freeport, TX
Grand I., LA
Gulf Shores, AL
Panama, FL
Belleair, FL
Siesta Key, FL
Manasota, FL
Ft. Myers, FL
Naples, FL
Marco I., FL
South Atlantic
35.5
7.2
4.8
8.8
15.8
4.5
27.0
4.5
2.6
8.6
7.9
27,850 17,908
4,978
1,719
1,567
2,525
1,023
6,817
1,264
5,685
12,204
5,901
3,629
1,294
1,327
1,136
904
2,502
748
2,376
6,432
4,166
785
691
358
178
160
227
252
281
2,187
1,419
747
504
504
270
151
72
201
93
166
914
748
527
Mean Single Density: 377
Standard Deviation: 282
Standard Deviation of the Mean:
85
Key Biscayne, FL 1.3
Lauderdale-
By-Sea, FL.
Palm Beach, FL
Vero Beach,FL
Cocoa Beach, FL
Daytona Beach, FL
Fernandina, FL
St. Simona, GA
Folly Beach, SC
Hilton Head, SC
Myrtle Beach, NC
Nags Head Area, NC 9
Wrightsville, NC 5
Long Bay, NC 5
4,635
1,928
3,433
0.4 2,254 699 5,123
3.2 8,664 3,249 2,708
6.6 8,983 5,408 1,361
4.2 6,246 2,942 1,847
3.0 1,267 212 422
9.9 3,356 2,544 339
7.4 3,400 2,591 459
1.9 1,128 774 594
43.2 9,768 7,922 226
16.8 10,107 5,508 602
0 4,632 4,025 515
8 2,251 1,015 388
3 2,967 2,314 560
Mean Single Density: 586
Standard Deviation: 469
Standard Deviation of the Mean: 130
1,428
1,589
1,015
819
700
71
257
350
407
183
328
447
175
437
5-43
-------
Titus
Table 13. Building Densities for a Sample of Coastal Barriers (continued)
Site
Area
(mile2)
Housing
Units
Single
Units
- (units/mile2) -
Housing Single
Density Density
Mid-Atlantic/Northeast
Va. Beach, VA
Ocean City, MD
Rehobeth, DE
Beach Haven, NJ
N. Beach Haven,
Ship Bottom, NJ
Surf City, NJ
Sea Isle, NJ
Wildwood, NJ
Seaside Heights,
Ocean Beach, NJ
Long Beach, NY
Atlantic Beach,
144.5
4.5
1.9
1.0
NJ 1.7
0.6
0.7
2.3
4.3
NJ 0.4
0.8
2.0
NY 0.4
92,032
18,221
3,111
2,379
5,326
1,781
2,530
4,595
16,664
2,728
4,022
15,203
975
74,362
3,116
1,593
1,734
3,920
1,322
1,801
2,762
8,267
1,004
3,877
5,123
760
637
4,049
1,637
2,379
3,133
2,968
3,614
1,998
3,875
6,820
5,028
7,602
2,438
515
692
838
1,734
2,306
2,203
2,573
1,201
1,922
2,510
4,846
2,562
1,900
Narragansett , RI
Pier
Town of
W. Yarmouth, MA
2.6
3.8
12.4
1,576
6,587
784
953
5,395
417
606
1,733
63
367
1,420
34
Mean Single Density: 1726
Standard Deviation: 1177
Standard Deviation of the Mean:
294
5-44
-------
Titus
(15) Road_ Mileage = Length + LBI_ Secondary, Road_ Density * Density / LBI_ Density * Area
where:
(16) LBI_Secondary_Road_Density = (LBI_Roads - LBI_Length) / LBI_Area,
and Area and Length refer to the barrier island under analysis. Equation (15) says that if density is zero, the
road mileage is equal to the length of the island in question. If the island is twice as long as LBI and has the
same width and density, the road mileage is twice that of LBI. If the island has the same length but twice the
area, the road length is not quite double that of LBI, since secondary roads are double but the primary road is
the same length.
The secondary road density estimated by Weggel et al. is 14.3 miles per square mile. Weggel et al. also
estimate the building density at 1949 per square mile; however, because only 73% are single-family houses, we
adjust this downward to 1420, to be consistent with our approach of using census data for single family houses.
Because the area of the island is 7.4 square miles, equation (15) becomes:
(17a) Road_Mileage = Island, Length + 0.01007042 * Density * Area,
or,
(17b) Road_ Mileage = Island. Length + 0.01007042 * Single_ Houses.
In the discussion of the case study, we noted that the cost for elevating houses and infrastructure worked
out to $457, $856, and $1358 million for the 50-, 100-, and 200-cm scenarios, respectively. Because most of these
costs apply to rebuilding infrastructure along roadways, we assume that the costs are proportional to road
mileage. Given the island's 124 miles of roads, we multiply equation (17b) by the cost per mile of road for each
of the scenarios, and we get:
(18a) Cost(50 cm) = 3,685,000 * Length + 37,109 * Buildings
(18b) Cost(100 cm) = 6,903,000 * Length + 69,518 * Buildings
(18c) Cost(200 cm) = 10,952,000 * Length + 110,287 * Buildings.
The intercept term, ranging from $3.7 to $11 million per mile, appears reasonable, when one considers
that the roads are being replaced more than once in the high scenarios. However, the cost of $37-110 thousand
per single house seems somewhat high at first glance. The cost results in part because Weggel et al. assume
that communities would rebuild roads to normal engineering standards; however, this assumption is offset by the
assumption that no other infrastructure would be necessary.
The cost does not seem quite so high if one remembers that the costs are incurred continuously over the
course of a century; even in the high scenario, it is less than one thousand dollars per year per building.
Moreover, the co-efficient also includes costs attributable to multi-unit housing and would be 25% less if we
included all buildings.
5-45
-------
Titus
Because we have sampled for density, we rewrite equation (18) as follows:
(19a) Cost(50 cm) = 3,685,000 * Length + 37,109 * Area * Density
(19b) Cost(100 cm) = 6,903,000 * Length + 69,518 * Area * Density
(19c) Cost(200 cm) = 10,952,000 * Length + 110,287 * Area * Density.
Although the 200-cm scenario involves raising the entire island, the 50- and 100-cm scenarios might only
require that the low bay sides be raised. Therefore, we need to scale the equation to account for the part of
the island that would be raised. We do this by multiplying the equations by the ratio bayside_ area/area and
dividing by the value of that ratio for Long Beach Island, 0.56. (Thus, for Long Beach Island, the equation is
unchanged.)
(20a) Cost(50 cm) = 6,580,4000 * Length * Bayside_Area / Area + 65,947 * Bayside_Area * Density
(20b) Cost(100 cm) = 12,328,000 * Length * Bayside.Area / Area + 123,542 * Bayside_Area * Density
Because ocean as well as baysides would have to be raised in the 2-meter scenario, we do not bother to scale
this equation, and simply use equation (19c).
We also use the unsealed equation (19b) as an alternative to equation (20b) for the 1-meter scenario, to
account for the possibility that ocean sides of barrier islands might have to be raised even with a 1-meter rise.
Leathennan assumed this to be the case, largely because for many islands, much of the land above 5 ft NGVD
is within 2 feet of this contour. However, for Long Beach Island and other coastal barriers, most of the ocean
side is above 8 ft NGVD. (Even land at the 10-ft contour might have to be raised with a 1-meter rise. Most
of the islands with little land below 5 ft NGVD are along the Atlantic Coast. With a typical spring tidal range
of 7 feet (and the fact that sea level is 6 inches above the NGVD reference elevation), land at the 10-foot
contour is only 6 feet above spring ocean tide; with a 4-foot relative rise, it would only be 2 feet above the
ocean's spring high tide. If the dunes were eroded by a prolonged northeaster, such low elevations of ocean side
lots would greatly increase the risk of an inlet breach. If bay sides of barrier islands were already being raised,
local officials recognizing that the sea would continue to rise after the year 2100 might conclude that raising
ocean sides would be worthwhile as well.)
Results of Extrapolation
Table 14 illustrates our estimates of the non-sand costs of elevating barrier islands in place. For a 50 cm rise
in sea level, Gulf coast barrier islands account for over 50% of the $11 billion cost, largely due to their lower
elevations. By contrast, for a 2-meter rise, the Mid-Atlantic and Northeast would account for over 50% of the
$96 billion cost because they are on average the most densely developed. Our estimates imply a cost elasticity
of 1.6.
5-46
-------
Titus
Table 14. Cost of Elevating Roads and Structures Assuming
That Costs are Proportional to Building Density
Shoreline Miles"
Bayside Area (mi2)
Oceanside Area (mi2)
Single Unit
Building Density:
Mean
Standard Dev
Total Cost (Billions
50 cm
100 cm
100 cm (Alt Cost )
200 cm
Sampling Error
50 cm
100 cm
100 cm (Alt Cost
200 cm
Gulf
565
181
325
377
282
of 1986
5.8
10.9
17.2
27.2
1.0
1.9
3.0
5.6
South
Atlantic
522
24
167
586
469
Dollars)
1.4
2.6
11.4
18.1
0.19
0.37
1.63
2.64
Mid Atlantic
& Northeast
511
30
204
1726
1177
3.6
7.2
28.1
50.1
0.57
1.09
4.75
7.60
USA*
1,598
235
696
nc
nc
10.8
20.7
56.6
95.4
1.17
2.22
5.85
9.86
Cost(50 cm) - 6,580,400 * Length * Bayside_Area/Area
+ 65,947 * Bayside_Area * Density
Cost(100 cm) - 12,327,000 * Length * Bayside_Area/Area
+ 123,542 * Bayside_Area * Density
Alt Cost(100 cm) = 6,903,000 * Length + 69,518 * Area * Density
Cost(200 cm) - 10,952,000 * Length + 110,287 * Area * Density
Results are for the Atlantic and Gulf Coasts only; the Pacific Coast has
no barrier islands. Sampling error is based solely on the variation of
building density.
Shoreline lengths are from Leatherman and refer to developed as well as
developable sandy ocean shorelines.
5-47
-------
Titus
CHAPTER?
SUMMARY AND CONCLUSIONS
FINAL CAUTION
This paper has discussed the methods and results of four studies and our own analysis of the nationwide
impact of a rise in sea level of 50-200 centimeters by the year 2100. The analysis was structured to enable
consideration of three broad policy options: protect all shores, protect no shores, and protect only the areas that
were developed by the middle of the 1980s.
It is the nature of first-cut national assessments to underestimate the cost of any undertaking, and this
study is no exception. We have not identified every important cost; we have not estimated the magnitude of
every cost we have identified; and we use assumptions that tend to understate the impact at each step of the
calculations.
The estimates for barrier island raising are conservative because they assume (1) today's level of
development and (2) do not consider the sand losses caused by major storms. The estimates for protecting
sheltered shorelines are conservative because the complete cost of defending land from inundation would
generally be greater than $500 per foot of shoreline; they only evaluate the cost of defending areas that are
densely developed today; and our assumptions understate the portion of shorelines that would require levees.
We did not attempt to estimate the cost of protecting water supplies from saltwater intrusion or the cost
of protecting lowlands from flooding. Nor did we examine the cost of saving Louisiana's coastal wetlands or
of rebuilding municipal drainage systems.
Although the study provides nationwide estimates of wetland loss for three alternative policy options, it only
provides a partial cost estimate for protecting currently developed areas. Future studies will have to assess the
value of the (currently) undeveloped land that would be lost, as well as the economic losses that would occur
from the loss of coastal wetlands.
Consideration of all three policy options should be conducted for specific areas; but nationwide cost
estimates for the no-protection and total protection options would not be particularly meaningful. The
no-protection costs would assume an eventual abandonment of the nation's beach resorts, as well as major
portions of coastal cities such as Miami, Charleston, New York, and Boston-such an abandonment does not
seem plausible given the relatively low cost of shore protection. The total-protection cost would assume a
complete armoring of all UJS. tidal shorelines, which again does not seem reasonable.
SUMMARY RESULTS
Tables 15 and 16 summarize the nationwide and southeast-wide results of the papers comprising this
volume. Fourteen thousand square miles of land could be lost to the sea from a 1-meter rise if shores are not
protected, with dry and wet land each accounting for about half the loss. For approximately $100 billion, a
thousand square miles of currently developed areas (accounting for about 7% of the threatened land) could be
protected from inundation, but the loss of coastal wetlands would be greater.
Our estimates suggest that the cumulative cost of shore protection would be approximately $140,000 per
acre. Thus, at the national level, protecting developed coastal areas appears to be cost-effective. Even if one
merely compares this figure with the value of land and structures on barrier islands and coastal cities, these areas
appear worth protecting. But this cumulative estimate implies that even at the end of the century, the annual
cost of protection would be about $3,000 per acre—hardly a welcome prospect for coastal property owners but
nevertheless, one well worth bearing in order to maintain the properly.
5-48
-------
Titus
Table 15. Summary of Nationwide Results
50 cm
If No Shores Are Protected
Land Lost
Dryland Lost (sq mi) 3315-7311
Wetlands Lost (%) 17-43
Value of Lost Property ($bil) Y
Cost of Coastal Defense ($bil) •
Open Coast 0
Sand 0
Elevate Structures 0
Sheltered Shores 0
If Developed Areas Are Protected
Land Lost
Dryland Lost (sq mi) 2200-6100
Wetlands Lost (%) 20-45
Value of Lost Property Y
Cost of Coastal Defense 32-43
Open Coast
Sand 15-20f
Elevate Structures 9-13
Sheltered Shores 5-13
If All Shores Are Protected
Land Lost
Dryland Lost (sq mi) 0
Wetlands Lost (%) 38-61
Value of Lost Property 0
Cost of Coastal Defense ?
Open Coast
Sand 15-20f
Elevate Structures 9-13
Sheltered Shores ?
100 cm
5123-10330
26-66
Y
0
0
0
0
4100-9200
29-69
Y
73-111
27-41f
21-57f
11-33
0
50-82
0
?
27-41f
21-57f
?
200 cm
8191-15394
29-76
Y
0
0
0
0
6400-13500
33-80
Y
169-309
58-100f
75-115
30-101
0
66-90
0
?
58-100f
75-115
?
Note: All dollar figures are in billions.
Symbols: Y signifies value that Yohe
will calculate
in future
report.
? signifies value not currently being assessed.
f Interval represents estimates based on alternative formulae. All
other intervals represent statistical uncertainty, except for
totals, which contain both.
5-49
-------
Titus
Table 16. Summary of Southeastern Results
50 cm
100 cm
200 cm
2300-5900
22-48
Y
If No Shores Are Protected
Land Lost
Dryland Lost (sq mi)
Wetlands Lost (%)
Value of Lost Property
Cost of Coastal Defense
Open Coast
Sand
Elevate Structures
Sheltered Shores
If Developed Areas Are Protected
Land Lost
3200-7600
30-75
Y
4800-10800
37-88
Y
0
0
0
0
0
0
0
0
0
0
0
0
Dryland Lost (sq mi) 1900-5500
Wetlands Lost (%) 24-50
Value of Lost Property
Cost of Coastal Defense
Open Coast
Sand
Elevate Structures
Sheltered Shores
If All Shores Are Protected
Land Lost
Dryland Lost (sq mi)
Wetlands Lost (%)
Value of Lost Property
Cost of Coastal Defense
Open Coast
Sand
Elevate Structures
Sheltered Shores
Y
19-28
10-15f
5-9
2-5
0
38-63
0
?
10-15f
5-9
?
2600-6900
34-77
Y
42-75
19-30f
10-40f
5-13
0
47-90
0
?
19-30f
10-40f
?
4200-10100
40-90
Y
127-174
44-74f
60-75
9-41
0
68-93
0
?
44-74f
60-75
?
Note:
All dollar figures are in billions.
Symbols: Y signifies value that Yohe will calculate in future
report.
? signifies value not currently being assessed.
f Interval represents estimates based on alternative formulae. All
other intervals, represent statistical uncertainty, except for
totals, which contain both. >
5-50
-------
Titus
But cost-effectiveness is not the sole criterion society should use to determine whether our shores should
be protected from the effects of a rising sea; the value of resources like wetlands that might be lost if we protect
our more tangible economic assets should be considered as well. Fortunately, it might be possible to protect
both. The effort to protect coastal wetlands would be most successful if focused upon areas that are not yet
densely developed. Abandoning developed areas would only increase the area of surviving wetlands by 5 to 10%
-but at great cost. By contrast, limiting coastal protection to areas that are already developed (and allowing
currently undeveloped areas to flood) would increase the area of surviving wetlands by 40 to 100%.
NEXT STEPS
Toward Better National Assessments
Although our nationwide estimates are based on samples of coastal sites, we have analyzed the implications
of these results only at the national level, with the exception of the Long Beach Island case study. Although this
is probably adequate for providing national policy makers with a sense of the magnitude of the threat from global
warming and the need to develop anticipatory policies, it does little to suggest what those policies might be.
Once Yohe's assessment of coastal property values is complete, we will have a nationwide data base that
will be adequate for conducting preliminary economic analyses of coastal policy options. For each sea level rise
scenario, it will be possible to estimate whether and for how long particular coastal sites would be worth
protecting (ignoring the value of wetlands, which market forces generally do). Such assessments will make it
possible for national analyses to use a more realistic management scenario which assumes that all areas will be
protected that are worth protecting; such a scenario will almost certainly fall between our total and standard
protection scenarios.
Additional refinements in shore-protection cost estimates should focus on alternatives to raising barrier
islands on the open coast and explicitly incorporating drainage costs in areas that would require levees. The
former is necessary because island migration and levees may be more viable for some communities on the open
coast; the latter is necessary because it is potentially as significant as construction of levees.
Estimates of the value of lost property will also have to be improved. Our proposed market-based
shore-protection scenario will substantially understate the dryland protected and the wetlands lost as long as we
assume today's level of development. Particularly in the Southeast, there is considerable low-lying forest and
farm land that may be developed in the next thirty to one hundred years.
Toward Protecting Coastal Wetlands
Because a substantial acceleration in sea level rise is still decades in the future, we have argued that it is
still too soon for society to implement most of the responses that sea level rise will eventually necessitate, but
that one important exception is the protection of coastal wetlands (Titus, 1984; Titus, 1986; and Titus, 1988).
Our hypothesis has been that coastal states outside Louisiana could maintain most wetland shorelines and
minimize the loss of wetland acreage most efficiently by enacting a policy of "presumed mobility" that required
that areas that are developed in the future (and perhaps a limited number of areas that are currently lightly
developed) revert to nature 75-100 years hence if the sea rises enough to inundate them. There are many ways
of implementing such a policy (e.g., restrictions on bulkhead reconstruction, state regulations, long-term leases,
conditional land ownership), but the success of this option requires that it be implemented soon.
In our view, presumed mobility is legally, economically, and administratively more feasible than the
alternative of prohibiting coastal development. First, prohibitions of development are often ruled as contrary to
the "due process" clause of the Constitution. By contrast, requirements to yield property to the state as shores
erode have been part of the riparian laws of many states since colonial times; bulkheading restrictions are
commonplace; and the courts have found that a coastal management action is not a "taking" if the impact is a
negligible fraction of a property's value. (Note: the present value of losing a $100,000 property fifty years hence
is less than $1,000, and one-hundred years hence, less than $10.)
5-51
-------
Titus
The latter aspect also explains why the presumed mobility policy would be more economically feasible than
prohibiting development. To purchase a few million acres of land that might be wetlands with a one or two
meter rise today, would cost tens of billions of dollars; and it would be a poor investment if the greenhouse effect
were curtailed and the sea did not rise as projected. By contrast, some ways of implementing the presumed
mobility approach (e.g., bulkhead regulations) would require little if any public expenditures; even eminent
domain purchases of the option to take over property as sea level rises would be at most a few percent of the
property values. If sea level does not rise as projected, the investment to protect coastal wetlands would be lost,
just as farmers hedging against decreases in crop prices lose their investments in commodity options if prices rise
and homeowners lose the value of their fire insurance premiums if their houses do not burn. In either event,
economic theory generally finds hedging and insurance to be rational investments.
Finally, the presumed mobility approach is more administratively feasible because governmental decisions
are confined to setting an environmental constraint—the long-term protection of wetlands-rather than dictating
the methods by which private landholders meet the constraint. Unlike prohibitions of development, one must
concede the eventuality of sea level rise before opposing a policy of presumed mobility. Moreover, preventing
development of coastal lowlands would require drawing a line on a map beyond which the land would not be
developed; drawing the line would require a decision regarding what level of sea level rise to anticipate, which
would in turn require policy makers to (1) rely on a projection of sea level rise, and (2) pick a year after which
the policy would be ineffective. By contrast, presumed mobility allows (in fact, requires) real estate markets to
incorporate the assessments of buyers and sellers regarding how much the sea will rise and the present value of
losing land at some future date.
So far, our argument in favor of the presumed mobility approach has been nothing more than a logical
hypothesis; it has not been possible to estimate the practical significance of our arguments. The State of Maine
has extended its coastal protection policies to include long-term mechanisms to ensure the survival of coastal
wetlands as sea level rises; but no one else has yet followed suit. Whether this is because coastal policy makers
disagree with our hypothesis or they simply feel that the matter is not sufficiently urgent to require action, we
believe that they need an assessment that estimates costs, remaining wetland acreage, and percentage of the coast
with a band of at least (for example) 100 meters of wetlands, for various policy options and implementation
dates.
We are now close to having sufficient geomorphic and engineering data to conduct such an assessment for
a sample of 46 sites throughout the nation; but we still need projections of economic development for the areas.
Moreover, economic models have to be developed to project both annual economic losses and current impacts
on property values for a given time profile of the probability distribution of future sea level rise. These models
will have to address strategies for physical and economic depreciation of properties. Property owners certain to
lose their property in five to ten years would allow their properties to physically depreciate by avoiding major
repairs. Owner-occupied property likely to be inundated 20-30 years later could be sold to rental investors who
viewed the future more than twenty years hence as irrelevant. People concerned about permanent family
ownership of property would tend to buy property in areas that were likely to be protected, but fairly modest
price differences would induce investors to purchase property even if inundation were only 20-30 years away.
Nevertheless, these strategies would not avoid all of the losses experienced by long-term property owners, who
might have attachments to a community or might have invested in additions that command little premium on the
rental market.
Incorporating these issues appears manageable. We have an opportunity to evaluate a policy whose benefits
may be an order of magnitude greater than the costs. We hope that such an assessment can be undertaken soon.
Because the situation is very different than for the rest of the nation's coast, this report has not focused on
Louisiana. Unlike wetland loss elsewhere, the implications of sea level rise for this coastal state appears almost
certain to require federal action, because the federal government manages the flow of the Mississippi River. A
recent EPA/Louisiana Geological Survey report outlined the analysis necessary to evaluate options to protect
Louisiana's coastal wetlands. With 40% of the nation's coastal wetlands at risk and the federal government
preventing freshwater and sediment from reaching the marshes and swamps, wetland loss in Louisiana cannot
realistically be viewed as the parochial concern of a single state.
5-52
-------
Titus
REFERENCES
Alexander, C.E., MA. Broutman, and D.W. Field. 1986. An Inventory of Coastal Wetlands of the USA.
Rockville, MD: National Oceanic and Atmospheric Administration.
Earth, M.C. and J.G. Titus. 1984. Greenhouse Effect and Sea Level Rise: A Challenge for This Generation.
New York: Van Nostrand Reinhold.
Bruun, P. 1962. "Sea Level Rise as a Cause of Shore Erosion. Journal of the (ASCE) Waterways and Harbors
Division 88 (WW1):117-130.
Dean, R.G. et al. 1987. Responding to Changes in Sea Level. Washington, D.C.: National Academy Press.
Devoy. R J. 1987 Sea Surface Studies. New York: Croom Helm.
Everts, C. 1985. In Potential Impacts of Sea Level Rise on the Beach at Ocean City. Maryland. Washington,
D.C.: Environmental Protection Agency.
Hallermeier, RJ. 1981. "A Profile Zonation for Seasonal Sand Beaches from Wave Climate." Coastal
Engineering 4:253-277
Hansen et al. 1986. In Titus, J.G. (ed). Effects of Changes in Stratospheric Ozone and Global Climate.
Washington D.C.: Environmental Protection Agency and United Nations Environmental Program.
Hull, C.HJ., L. Thatcher, and R. Tortoriello. 1986. "Impact on Delaware River Salinity." In: Hull and Titus,
op. cit.
Hull, C.HJ. and J.G. Titus. 1986. Greenhouse Effect. Sea Level Rise, and Salinity in the Delaware Estuary.
Washington, D.C.: Environmental Protection Agency.
Kana, T.W., B. Baca, and M.L. Williams, 1984. Potential Impact of Sea Level Rise on Wetlands Around
Charleston. South Carolina Washington, D.C.: Environmental Protection Agency.
Kana, T.W., B.J. Baca, and M.L. Williams. 1988. "Charleston Case Study." In: Titus (ed), 1988.
Kriebel and Dean 1985. Potential Impacts of Sea Level Rise on the Beach at Ocean City. Maryland.
Washington, D.C.: Environmental Protection Agency.
Louisiana Wetland Protection Panel. 1984. Saving Louisiana's Coastal Wetlands: The Need for a Long-Term
Plan of Action
Sorensen, R.M., R.N. Weisman, and G.P. Lennon. 1984. "Control of Erosion, Inundation, and Salinity Intrusion
Caused By Sea Level Rise." In: Earth and Titus, op. cit.
Titus, J.G. 1986. "Greenhouse Effect, Sea Level Rise, and Coastal Wetlands.' Coastal Zone Management
Journal. 14:3.
Titus, J.G. 1985. "Sea Level Rise and the Maryland Coast." In Potential Impacts of Sea Level Rise on the
Beach at Ocean City. Maryland. Washington, D.C.: Environmental Protection Agency.
Titus, J.G. (ed) 1988. Greenhouse Effect. Sea Level Rise, and Coastal Wetlands. Washington, D.C.:
Environmental Protection Agency.
5-53
-------
Titus
Titus, J.G., T. R. Henderson, and J.M. Teal. 1984. "Sea Level Rise and Wetlands Loss in the United States."
National Wetlands Newsletter Environmental Law Institute: Washington, D.C. 6:5.
Titus, J.G., T. R. Henderson, and J.M. Teal. 1984. "Sea Level Rise and Wetlands Loss in the United States."
National Wetlands Newsletter Environmental Law Institute: Washington, D.C. 6:5.
Titus, J.G., C. Kuo, MJ. Gibbs, T. Laroe, K. Webb, and J. Waddel, 1987. Greenhouse Effect. Sea Level Rise.
and Coastal Drainage Systems. ASCE Journal of Water Resources Management.
5-54
-------
Titus
APPENDIX 1.
Proof that the Ratio of Costs in the Year 2100 for different sea level rise scenarios is a conservative
approximation of the ratio of total costs, assuming the Weggel formula.
First, we simplify notation by letting x(t) and y(t) represent the costs of the 1 and 2 meter scenarios, and
A represents the ratio x(2100)/y(2100). It is commonly known that if x(t) >. A * y(t) for t over a given range
(in our case 1990 to 2100), then,
2100 2100
(Al) /x(t)dt//y(t)dt >: A
t=1986 t=1986
Therefore, x(2100)/y(2100) is a conservative estimate provided that x(t)/v(t) is in fact greater than
x(2100Vvf2100) for t<2100. which we now show.
First, we note that sea level rise accelerates over time, which means that,
(A2) SLR_lm(t)/SLR_2m(t) > SLR_lm(2100)/SLR_2m(2100),
throughout the period 1986 to 2100. We define the latter ratio (for 2100) as B, which is less than one.
Therefore, recalling that,
((5+SLR_lm(t))/5)1 5-l
(A3) x(t)/y(t) = ,
((5+SLR_2m(t))/5)1 5-l
We substitute B*SLR_2m(t)
((5+SLR_2m(t))/5)1 5 -
For clarity, we redefine SLR_2m(t) as z(t),
((5+B*z(0)/5)15-l
(A5) x(t)/y(t) >
Since z(t) is monotonically increasing, and B is positive but less than 1, it is clear that x(t)/y(t) is
monotonically decreasing. Therefore, x(t)/y(t) is in fact greater than x(2100)/y(2100) for all years before 2100,
and the assertion is proven.
5-55
------- |