TEMPERATE climates, attractive scenery, ease of navigation, and access to ocean food supplies have put coastlines at the forefront of human development throughout history and around the world. The United States is no exception. Today, counties touching the coast account for 39 percent of total U.S. population and 28 percent of national property by value. Coastal living carries risk, particularly on the East Coast and along the Gulf of Mexico, where hurricanes and other coastal storms inflict billions in property and infrastructure damage each year. Climate change elevates these risks. Rising sea levels will, over time, inundate low-lying property and increase the amount of flooding that occurs during coastal storms. Moreover, as discussed in chapter 4, warmer sea-surface temperatures may change the frequency and intensity of those storms.
BACKGROUND
A growing body of research assesses the potential impacts of sea-level rise (SLR) on coastal communities. Early studies focused on developing a methodology for site-specific estimates of damage from SLR that could be used as a model for nationwide assessments (Yohe 1990). Several compared the cost to coastal property of damage from mean sea-level rise with the cost of protecting that property with sea walls, structural enhancements, and other adaptive measures (Yohe et al. 1996; Yohe and Schlesinger 1998).
Subsequent work expanded to regional assessments. One of the first was conducted by the U.S. Environmental Protection Agency (Titus & Richman 2001), which identified areas vulnerable to inundation from higher sea levels along the Atlantic and Gulf coasts. A subsequent U.S. interagency assessment of the Mid-Atlantic region simulated a 1-meter SLR running from New York through Virginia and estimated the associated impacts on residential property and coastal residents (CCSP 2009). The first robust national estimate of potential inundation damage from SLR, as well as the cost of protective measures, was published in 2011 (Neumann et al. 2011) using the National Coastal Property Model (NCPM) developed by Industrial Economics, Inc., for the U.S. Environmental Protection Agency.
Permanent inundation from mean sea-level rise is only one of the risks climate change presents to coastal property and infrastructure. Higher average sea levels lead to higher storm surges and elevated flooding risks (Frumhoff et al. 2007), even if the intensity or frequency of storms remains unchanged (Frazier et al. 2010). Kemp and Horton (2013) found that while the record 13.9-foot storm tide in New York Harbor during Superstorm Sandy was primarily due to the coincidence of the strongest winds with high tide, SLR driven by historical climate change added more than 1 foot to that 13.9-foot total.
A number of recent studies have assessed coastal communities’ vulnerability to future SLR-driven increases in storm surge. At a local scale, after Superstorm Sandy, the New York City Panel on Climate Change analyzed the risk to the city’s property and infrastructure from future climate-driven changes in sea levels and storm activity (NPCC 2013). California conducted an assessment of the impact of SLR on the Bay Area’s 100-year floodplains for coastal storms (San Francisco Bay Conservation and Development Commission 2011; Heberger et al. 2012), and Harrington and Walton (2008) estimated the impacts on coastal property for six coastal counties in Florida. Neumann and colleagues have incorporated projected increases in storm surge as a result of both mean SLR and potential changes in hurricane intensity and frequency into the NCPM for select cities (Neumann et al. 2014).
OUR APPROACH
Alongside the academic and policy-oriented work described earlier, private companies have developed sophisticated models to estimate potential losses from coastal storms. These models are used by the insurance industry in underwriting flood and wind insurance products, by the finance industry in pricing catastrophe bonds, and by local officials in coastal communities in preparing for and responding to hurricanes and other coastal storms. While not traditionally used in this way, they are also incredibly powerful tools for understanding how climate change will likely shape both industry and coastal community risk exposure in the years ahead.
Risk Management Solutions (RMS) is a leading provider of such tools, along with models for quantifying and managing other catastrophic risks, from earthquakes to terrorist attacks to infectious disease; RMS is a partner in this assessment. To assess the value of property at risk from future SLR, we mapped the probabilistic local SLR projections described in chapter 4 against RMS’s detailed exposure data set, which covers buildings, their contents, and automobiles for all coastal counties in the United States. To analyze the impact of local SLR on storm surge and flood damage during hurricanes and nor’easters, we used RMS’s North Atlantic Hurricane Model. This model combines state-of-the art wind and storm surge modeling and a stochastic event set that represents more than 100,000 years of hurricane activity and spans the range of all possible storms that could occur in the coming years (see appendix C).
The result of this analysis is the first comprehensive, nationwide assessment of the risk to coastal communities from mean SLR and SLR-driven increases in storm surge from hurricanes and nor’easters under a full range of climate futures, and at a very high level of geographic resolution. Taking this work one step further, we explore the impact of changes in hurricane frequency and intensity projected by Knutson et al. (2013) for RCP 4.5 and Emanuel (2013) for RCP 8.5 on both future storm surge and wind damage (see chapter 4). (While we capture projected change in frequency and intensity from the cyclogenesis models used by Knutson et al. and Emanuel, we do not capture projected change in landfall location. This could have a meaningful impact on the geographic distribution of hurricane-related losses and is worthy of considerable additional research.)
There is considerable uncertainty surrounding future coastal development patterns, which makes accurate cost projections challenging. Over the past few decades, population and property values in coastal counties have grown faster than the national average, putting more people and assets at risk. It is unclear the extent to which this trend will continue going forward, given constraints to further development and expansion in many coastal areas. Rather than attempt to predict how the built environment will evolve in the decades ahead, we assess the impact of future changes in sea level and storm activity relative to the American coastline as it exists today. Damage is reported in current dollars against current property prices.
Inundation from Mean Sea-Level Rise
While all coastal states are at risk from rising sea levels, some are much more vulnerable than others. Under RCP 8.5, for example, between 4.1 percent and 5.5 percent of total insurable residential and commercial property in the state of Louisiana will likely be below mean sea level (MSL) by 2050 (excluding that property already below MSL), growing to 15 to 20 percent by 2100 (figure 11.1). Florida is the second most vulnerable state in percentage terms, with 0.4 to 0.6 percent of current statewide property likely below MSL by 2050, growing to 1 to 5 percent by 2100. In dollar terms, between $33 billion and $45 billion worth of current Louisiana property will likely be below MSL by 2050, growing to $122 billion to $164 billion by 2100. The total value of current Florida property at risk is similar, with between $15 billion and $23 billion likely below MSL by 2050, growing to $53 billion to $208 billion by 2100 (figure 11.2).
FIGURE 11.1. Share of Current Property Below MSL in 2100 Under RCP 8.5
FIGURE 11.2. Value of Current Property Below MSL by 2100
RCP 8.5, billion 2011 U.S. dollars
Nationwide, we find that between $66 billion and $106 billion worth of current coastal property will likely be below MSL by 2050 under RCP 8.5 unless protective measures are taken (table 11.1), growing to $238 billion to $507 billion by 2100 (table 11.2). The value of current property likely under MSL falls to $62 billion to $85 billion by 2050 in both RCP 4.5 and 2.6. By 2100, nationwide property likely below MSL is $175 billion to $339 billion in RCP 4.5 and $150 billion to $276 billion in RCP 2.6.
TABLE 11.1 Additional current property below MSL and MHHW by 2050 (in billion 2011 U.S. dollars, at current property prices)
| Probability |
RCP 8.5 |
RCP 4.5 |
RCP 2.6 |
| MSL |
MHHW |
MSL |
MHHW |
MSL |
MHHW |
| 1-in-100 chance above |
156 |
523 |
143 |
472 |
129 |
456 |
| 1-in-20 chance above |
126 |
465 |
107 |
400 |
106 |
397 |
| Likely range |
66 to 106 |
323 to 389 |
62 to 85 |
294 to 366 |
62 to 85 |
287 to 360 |
| 1-in-20 chance below |
61 |
256 |
60 |
240 |
60 |
226 |
| 1-in-100 chance below |
52 |
186 |
51 |
181 |
50 |
172 |
MSL, mean sea level; MHHW, mean higher high water.
TABLE 11.2 Additional current property below MSL and MHHW by 2100 (in billion 2011 U.S. dollars, at current property prices)
| Probability |
RCP 8.5 |
RCP 4.5 |
RCP 2.6 |
| MSL |
MHHW |
MSL |
MHHW |
MSL |
MHHW |
| 1-in-100 chance above |
1,114 |
1,636 |
719 |
1,433 |
613 |
1,332 |
| 1-in-20 chance above |
701 |
1,432 |
495 |
1,135 |
431 |
990 |
| Likely range |
238 to 507 |
724 to 1,144 |
175 to 339 |
759 to 926 |
150 to 276 |
430 to 830 |
| 1-in-20 chance below |
166 |
509 |
134 |
400 |
116 |
362 |
| 1-in-100 chance below |
131 |
383 |
105 |
313 |
102 |
246 |
MSL, mean sea level; MHHW, mean higher high water.
Two factors explain this relatively small difference in inundation between the RCPs. First, the expanding ocean and melting ice sheets respond to both the amount of warming and the length of exposure to elevated temperatures. Temperatures begin to diverge significantly between RCPs only in the second half of the century; sea level, which integrates temperature, diverges later. Second, the largest sources of uncertainty in sea level are potential positive feedbacks in the behavior of ice sheets, particularly the West Antarctic Ice Sheet (WAIS). For example, for parts of the sea floor that are appropriately sloped, it is possible that, as a warming ocean eats away at the base of the WAIS (which unlike most of the Greenland and East Antarctic Ice Sheets largely sits below sea level), it will expose more of the ice sheet to the ocean, which will accelerate melt, exposing still more ice, and so forth. Such feedbacks are poorly understood at present; the uncertainties arising from this low level of understanding are independent of emissions and therefore cause the projected ranges of sea-level change for all the RCPs to overlap considerably.
At the tails of the SLR probability distribution, inundation damage is considerably worse than the likely range. For example, there is a 1-in-20 chance more than $346 billion of current Florida property (8.7 percent) could be below MSL by the end of the century under RCP 8.5 (see figure 11.2) and a 1-in-100 chance that more than $681 billion of current Florida property (17 percent) could be lost by 2100 unless defensive measures are taken. Nationwide, there is a 1-in-20 chance that more than $701 billion of current property will be below MSL by 2100 and a 1-in-100 chance it will be more than $1.1 trillion (see table 11.2).
While roughly two thirds of all current property likely below MSL by 2050 is in Louisiana and Florida, which may become three quarters of all property by the end of the century, Maryland, Texas, Massachusetts, North Carolina, New York, New Jersey, and California also face meaningful inundation risk. In Maryland, for example, between $13 billion and $23 billion (0.7 and 1 percent) of current statewide property will likely be below MSL by 2050, with losses concentrated in Queen Anne’s and Talbot counties located on the east side of the Chesapeake Bay. In Texas, up to $44 billion of current property will likely be below MSL by the end of the century, including important industrial and energy infrastructure.
Inundation risk from SLR extends beyond those properties underwater at average tide levels. There is currently $1.6 trillion in coastal property that is above MSL, but at or below peak high-tide levels, often referred to as mean higher high water (MHHW) levels. Most of this property is protected by shoreline defense built up over the course of decades or even centuries. As MSLs rise, the high-tide mark will rise as well, putting additional property in the line of fire. Without defensive investments (see chapter 22 for a discussion), these properties risk significant damage.
Figures 11.3 and 11.4 show the share and value of additional current property below MHHW by 2100 due to mean SLR. The value is two to four times larger than that for property below MSL, depending on time frame and SLR scenario (see tables 11.1 and 11.2).
FIGURE 11.3. Share of Current Property Below MHHW Caused by SLR by 2100 Under RCP 8.5
FIGURE 11.4. Value of Additional Current Property Below MHHW Caused by SLR by 2100
RCP 8.5, billion 2011 U.S. dollars
To illustrate the risks presented by local SLR to coastal communities, we map inundation levels in 2100 at MSLs in the median, 1-in-100, and 1-in-200 projections for RCP 8.5 for Miami, Norfolk, Houston, and Wilmington, North Carolina, in figures 11.5 through 11.8. The inundation threat to Miami is particularly grave at a citywide level, but will also challenge the viability of several neighborhoods in New York City, Wilmington, and elsewhere. In Houston, while the center of the city is reasonably safe, critical energy infrastructure is at risk. In Norfolk, major naval installations are threatened by SLR. This choice of examples is illustrative only. Many other cities in the country face significant SLR risk. These maps also do not show property at or below MHHW but above MSLs.
FIGURE 11.5. Wilmington MSL Projections in 2100 Under RCP 8.5
FIGURE 11.6. Miami MSL Projections in 2100 Under RCP 8.5
FIGURE 11.7. Norfolk MSL Projections in 2100 Under RCP 8.5
FIGURE 11.8. Houston MSL Projections in 2100 Under RCP 8.5
Storm Surge
As mentioned earlier, higher sea levels also mean greater flooding during hurricanes and other coastal storms. These storms currently result in roughly $27 billion in average annual commercial and residential property damage and business interruption costs along the East Coast and Gulf of Mexico, with roughly half of that occurring in Florida. The impact of climate change on flooding during coastal storms is larger and more immediate than the impact of gradual SLR-driven inundation discussed earlier. Assuming current hurricane activity continues, SLR under RCP 8.5 will likely increase average annual losses by $2 billion to $3.5 billion per year as early as 2030, a 7 to 13 percent increase over current levels. This increase in storm damage, like the storms themselves, will not be evenly spread across time. These numbers reflect the expected average annual loss of all storms across different scenarios for SLR.
As with inundation from SLR, this climate-driven increase in expected storm damage hits some states harder than others (figures 11.9 through 11.14). The largest relative likely increases occur in Delaware (16 to 39 percent by 2030), New Jersey (14 to 36 percent), New York (11 to 27 percent), and Virginia (13 to 28 percent). In absolute terms, Florida faces a far greater increase in expected storm damage due to higher sea levels than any other state. By 2030, average annual losses likely grow by $738 million to $1.3 billion.
FIGURE 11.9. Relative Increase in Average Annual Coastal Storm Damage Caused by Higher Sea Levels in 2030
Percent change from 2010 expected average annual losses
FIGURE 11.10. Relative Increase in Average Annual Coastal Storm Damage Caused by Higher Sea Levels in 2050
Percent change from 2010 expected average annual losses
FIGURE 11.11. Relative Increase in Average Annual Coastal Storm Damage Caused by Higher Sea Levels in 2100
Percent change from 2010 expected average annual losses
FIGURE 11.12. Absolute Increase in Average Annual Coastal Storm Damage Caused by Higher Sea Levels in 2030
Million 2011 U.S. dollars relative to 2010 expected average annual losses
FIGURE 11.13. Absolute Increase in Average Annual Coastal Storm Damage Caused by Higher Sea Levels in 2050
Million 2011 U.S. dollars relative to 2010 expected average annual losses
FIGURE 11.14. Absolute Increase in Average Annual Coastal Storm Damage Caused by Higher Sea Levels in 2100
Million 2011 U.S. dollars relative to 2010 expected average annual losses
By 2050, average annual losses from hurricanes and nor’easters will likely grow to $5.8 billion to $13 billion nationwide under RCP 8.5, a 21 to 48 percent increase from current levels, due just to mean SLR. Average annual losses in New Jersey will likely increase by between 64 and 174 percent, by 53 to 155 percent in Delaware, and by 45 to 110 percent in Virginia. In absolute terms, Florida will likely see an additional $1.9 billion to $4 billion a year in storm damage by 2050 unless protective measures are taken, while New York will likely see an additional $658 million to $3 billion in coastal storm-related costs each year.
By 2100, SLR-driven increases in average annual hurricane and nor’easter damage will likely grow by $19 billion to $33 billion under RCP 8.5, a 71 to 122 percent increase from current levels. There is a 1-in-20 chance that damage could grow by more than $42 billion by 2100 and a 1-in-100 chance that damage could grow by more than $50 billion. Conversely, there is a 1-in-20 chance that average annual losses will only grow by $15 billion or less and a 1-in-100 chance of a less than $8.6 billion increase. Florida will likely see a $7 billion to $14 billion, or 60 to 104 percent, increase above current levels. New York will likely see a $2.6 billion to $5.2 billion increase, or 159 to 313 percent, and New Jersey will likely see a $1.4 billion to $3.7 billion increase, or 208 to 414 percent.
Averaged over the two-decade intervals used for other impact categories, the likely SLR-driven increase in average annual coastal storm damage is $2 billion to $3.6 billion on average by 2020–2039, $5.7 billion to $12 billion on average by 2040–2059, and $18 billion to $28 billion on average by 2080–2099 (figure 11.15).
FIGURE 11.15. Increase in Expected Annual Property Losses as a Result of SLR
Assuming no change in hurricane activity (billion 2011 U.S. dollars)
The relatively small difference in SLR in this century between RCPs translates into a relatively small difference in SLR-driven surge damage between RCPs as well. In RCP 4.5, the likely increase in average annual coastal storm damage due to mean SLR is between $5 billion and $11 billion by 2040–2059 and between $15 billion and $22 billion by 2080–2099. In RCP 2.6, the likely range falls to $4.6 billion to $10 billion on average by 2040–2059 and $12 billion to $19 billion on average by 2080–2099.
Another way to think about the risk to coastal property from SLR-driven increases in storm surge is to map the change in extent of flooding during a 1-in-100 year flood; or, put another way, areas with a 1 percent chance of being flooded in any given year. Buildings within the 100-year floodplain are generally required to purchase flood insurance by the federal government. Projected local SLR will materially change the 1-in-100 year floodplain in many communities, and as soon as the next 10 to 20 years. Figures 11.16 and 11.17 show the change in the 100-year floodplains of New York City and Norfolk, respectively, as a result of projected SLR in our median RCP 8.5 scenario.
FIGURE 11.16. The New York City 100-year Floodplain Under Median RCP 8.5 SLR
Assumes historical hurricane activity
FIGURE 11.17. The Norfolk 100-year Floodplain Under median RCP 8.5 SLR
Assumes historical hurricane activity
CHANGES IN HURRICANE FREQUENCY AND INTENSITY
There is considerable uncertainty about how climate change will influence the frequency and intensity of hurricanes going forward, but the impact of potential hurricane activity change is significant. For example, using ensemble projections from Emanuel (2013) for changes in hurricane frequency and intensity under RCP 8.5, average annual damage from East Coast and Gulf of Mexico hurricanes and nor’easters will likely grow by $3.0 billion to $7.3 billion by 2030, an 11 to 22 percent increase from current levels (figure 11.18). By 2050, the combined impact of higher sea levels and modeled changes in hurricane activity likely raise annual losses by $11 billion to $23 billion, roughly twice as large of an increase as from changes in local sea levels alone. By the end of the century, the combined likely impact of SLR and modeled changes in hurricane activity raise average annual losses by $62 billion to $91 billion, three times as much as higher sea levels alone.
FIGURE 11.18. Increase in Average Annual Losses with Historical and Projected Hurricane Activity
Billion 2011 U.S. dollars: RCP 8.5 ensemble tropical cyclone activity projections from Emanuel (2013). Blue is historical, green is projected.
Under RCP 4.5, using changes in hurricane activity projected by Knutson et al. (2013), the increase in average annual commercial and residential property damage as a result of climate change is likely $2.7 billion to $7.0 billion by 2030, $11 billion to $22 billion by 2050, and $56 billion to $80 billion by 2100. Averaged over the two-decade intervals used for other impact categories, the increases are $3.6 billion to $5.7 billion by 2020–2039, $11 billion to $22 billion by midcentury, and $47 billion to $65 billion by late century (figure 11.19). The increase in damage resulting from either Emanuel’s or Knutson and colleagues’ projections for future changes in hurricane activity are due to both greater storm surge (even without climate-driven SLR) and greater wind damage.
FIGURE 4.19. Increase in Average Annual Losses with Projected Hurricane Activity
Billion 2011 U.S. dollars.
Source: RCP 8.5 tropical cyclone activity projections from Emanuel (2013); RCP 4.5 projections from Knutson et al. (2013)
While examining different RCPs, both Emanuel and Knutson and colleagues find significant changes in hurricane activity as a result of warmer sea-surface temperatures. Should this finding turn out to be correct, changes in storm activity could be a more important determinant of climate-driven changes in hurricane damage than SLR alone in the years ahead.
KEEPING OUT THE SEA
There are a number of steps individual building owners, community organizations, and policy makers at the local, state, and national levels can take to guard against some of these coastal effects. These include strengthening buildings, constructing sea walls, and nourishing beaches. In part 5 of this book, we analyze the extent to which these adaptive measures can reduce the risk that coastal communities face.
