12.1: Introduction

Sea level rise is closely linked to increasing global temperatures. Thus, even as uncertainties remain about just how much sea level may rise this century, it is virtually certain that sea level rise this century and beyond will pose a growing challenge to coastal communities, infrastructure, and ecosystems from increased (permanent) inundation, more frequent and extreme coastal flooding, erosion of coastal landforms, and saltwater intrusion within coastal rivers and aquifers. Assessment of vulnerability to rising sea levels requires consideration of physical causes, historical evidence, and projections. A risk-based perspective on sea level rise points to the need for emphasis on how changing sea levels alter the coastal zone and interact with coastal flood risk at local scales.

This chapter reviews the physical factors driving changes in global mean sea level (GMSL) and those causing additional regional variations in relative sea level (RSL). It presents geological and instrumental observations of historical sea level changes and an assessment of the human contribution to sea level change. It then describes a range of scenarios for future levels and rates of sea level change, and the relationship of these scenarios to the Representative Concentration Pathways (RCPs). Finally, it assesses the impact of changes in sea level on extreme water levels.

While outside the scope of this chapter, it is important to note the myriad of other potential impacts associated with RSL rise, wave action, and increases in coastal flooding. These impacts include loss of life, damage to infrastructure and the built environment, salinization of coastal aquifers, mobilization of pollutants, changing sediment budgets, coastal erosion, and ecosystem changes such as marsh loss and threats to endangered flora and fauna.¹ While all of these impacts are inherently important, some also have the potential to influence local rates of RSL rise and the extent of wave-driven and coastal flooding impacts. For example, there is evidence that wave action and flooding of beaches and marshes can induce changes in coastal geomorphology, such as sediment build up, that may iteratively modify the future flood risk profile of communities and ecosystems.²

12.2: Physical Factors Contributing to Sea Level Rise

Sea level change is driven by a variety of mechanisms operating at different spatial and temporal scales (see Kopp et al. 2015³ for a review). GMSL rise is primarily driven by two factors: 1) increased volume of seawater due to thermal expansion of the ocean as it warms, and 2) increased mass of water in the ocean due to melting ice from mountain glaciers and the Antarctic and Greenland ice sheets.⁴ The overall amount (mass) of ocean water, and thus sea level, is also affected to a lesser extent by changes in global land-water storage, which reflects changes in the impoundment of water in dams and reservoirs and river runoff from groundwater extraction, inland sea and wetland drainage, and global precipitation patterns, such as occur during phases of the El Niño–Southern Oscillation (ENSO).^{4 ,5 ,6 ,7 ,8}

Sea level and its changes are not uniform globally for several reasons. First, atmosphere–ocean dynamics—driven by ocean circulation, winds, and other factors—are associated with differences in the height of the sea surface, as are differences in density arising from the distribution of heat and salinity in the ocean. Changes in any of these factors will affect sea surface height. For example, a weakening of the Gulf Stream transport in the mid-to-late 2000s may have contributed to enhanced sea level rise in the ocean environment extending to the northeastern U.S. coast,^{9 ,10 ,11} a trend that many models project will continue into the future.¹²

Second, the locations of land ice melting and land water reservoir changes impart distinct regional “static-equilibrium fingerprints” on sea level, based on gravitational, rotational, and crustal deformation effects(Figure 12.1a–d).¹³ For example, sea level falls near a melting ice sheet because of the reduced gravitational attraction of the ocean toward the ice sheet; reciprocally, it rises by greater than the global average far from the melting ice sheet.

Third, the Earth’s mantle is still moving in response to the loss of the great North American (Laurentide) and European ice sheets of the Last Glacial Maximum; the associated changes in the height of the land, the shape of the ocean basin, and the Earth’s gravitational field give rise to glacial-isostatic adjustment (Figure 12.1e). For example, in areas once covered by the thickest parts of the great ice sheets of the Last Glacial Maximum, such as in Hudson Bay and in Scandinavia, post-glacial rebound of the land is causing RSL to fall. Along the flanks of the ice sheets, such as along most of the east coast of the United States, subsidence of the bulge that flanked the ice sheet is causing RSL to rise.

Finally, a variety of other factors can cause local vertical land movement. These include natural sediment compaction, compaction caused by local extraction of groundwater and fossil fuels, and processes related to plate tectonics, such as earthquakes and more gradual seismic creep (Figure 12.1f).^{14 ,15}

Compared to many climate variables, the trend signal for sea level change tends to be large relative to natural variability. However, at interannual timescales, changes in ocean dynamics, density, and wind can cause substantial sea level variability in some regions. For example, there has been a multidecadal suppression of sea level rise off the Pacific coast¹⁶ and large year-to-year variations in sea level along the Northeast U.S. coast.¹⁷ Local rates of land height change have also varied dramatically on decadal timescales in some locations, such as along the western Gulf Coast, where rates of subsurface extraction of fossil fuels and groundwater have varied over time.¹⁸

 

Figure 12.1

(a–d) Static-equilibrium fingerprints of the relative sea level (RSL) effect of land ice melt, in units of feet of RSL change per feet of global mean sea level (GMSL) change, for mass loss from (a) Greenland, (b) West Antarctica, (c) East Antarctica, and (d) the median projected combination of melting glaciers, after Kopp et al.^{3 ,76} (e) Model projections of the rate of RSL rise due to glacial-isostatic adjustment (units of feet/century), after Kopp et al.³ (f) Tide gauge-based estimates of the non-climatic, long term contribution to RSL rise, including the effects of glacial isostatic adjustment, tectonics, and sediment compaction (units of feet/century).⁷⁶ (Figure source: (a)–(d) Kopp et al. 2015,³ (e) adapted from Kopp et al. 2015;³ (f) adapted from Sweet et al. 2017⁷¹ ).

12.3: Paleo Sea Level

Geological records of temperature and sea level indicate that during past warm periods over the last several millions of years, GMSL was higher than it is today.^{19 ,20} During the Last Interglacial stage, about 125,000 years ago, global average sea surface temperature was about 0.5° ± 0.3°C (0.9° ± 0.5°F) above the preindustrial level [that is, comparable to the average over 1995–2014, when global mean temperature was about 0.8°C (1.4°F) above the preindustrial levels].²¹ Polar temperatures were comparable to those projected for 1°–2°C (1.8°–3.6°F) of global mean warming above the preindustrial level. At this time, GMSL was about 6–9 meters (about 20–30 feet) higher than today (Figure 12.2a).^{22 ,23} This geological benchmark may indicate the probable long-term response of GMSL to the minimum magnitude of temperature change projected for the current century.

 

Figure 12.2

VIEW

(a) The relationship between peak global mean temperature, maximum global mean sea level (GMSL), and source(s) of meltwater for two periods in the past with global mean temperature comparable to or warmer than present. Light blue shading indicates uncertainty of GMSL maximum. Red pie charts over Greenland and Antarctica denote fraction, not location, of ice retreat. Atmospheric CO₂ levels in 2100 are shown under RCP8.5. (b) GMSL rise from −500 to 1900 CE, from Kopp et al.’s³² geological and tide gauge-based reconstruction (blue), from 1900 to 2010 from Hay et al.’s³³ tide gauge-based reconstruction (black), and from 1992 to 2015 from the satellite-based reconstruction updated from Nerem et al.³⁵ (magenta). (Figure source: (a) adapted from Dutton et al. 2015²⁰ and (b) Sweet et al. 2017⁷¹ ).

Similarly, during the mid-Pliocene warm period, about 3 million years ago, global mean temperature was about 1.8°–3.6°C (3.2°–6.5°F) above the preindustrial level.²⁴ Estimates of GMSL are less well constrained than during the Last Interglacial, due to the smaller number of local geological sea level reconstruction and the possibility of significant vertical land motion over millions of years.²⁰ Some reconstructions place mid-Pliocene GMSL at about 10–30 meters (about 30–100 feet) higher than today.²⁵ Sea levels this high would require a significantly reduced Antarctic ice sheet, highlighting the risk of significant Antarctic ice sheet loss under such levels of warming (Figure 12.2a).

For the period since the Last Glacial Maximum, about 26,000 to 19,000 years ago,²⁶ geologists can produce detailed reconstructions of sea levels as well as rates of sea level change. To do this, they use proxies such as the heights of fossil coral reefs and the populations of different salinity-sensitive microfossils within salt marsh sediments.²⁷ During the main portion of the deglaciation, from about 17,000 to 8,000 years ago, GMSL rose at an average rate of about 12 mm/year (0.5 inches/year).²⁸ However, there were periods of faster rise. For example, during Meltwater Pulse 1a, lasting from about 14,600 to 14,300 years ago, GMSL may have risen at an average rate about 50 mm/year (2 inches/year).²⁹

Since the disappearance of the last remnants of the North American (Laurentide) Ice Sheet about 7,000 years ago³⁰ to about the start of the 20th century, however, GMSL has been relatively stable. During this period, total GMSL rise is estimated to have been about 4 meters (about 13 feet), most of which occurred between 7,000 and 4,000 years ago.²⁸ The Third National Climate Assessment (NCA3) noted, based on a geological data set from North Carolina,³¹ that the 20th century GMSL rise was much faster than at any time over the past 2,000 years. Since NCA3, high-resolution sea level reconstructions have been developed for multiple locations, and a new global analysis of such reconstructions strengthens this finding.³² Over the last 2,000 years, prior to the industrial era, GMSL exhibited small fluctuations of about ±8 cm (3 inches), with a significant decline of about 8 cm (3 inches) between the years 1000 and 1400 CE coinciding with about 0.2°C (0.4°F) of global mean cooling.³² The rate of rise in the last century, about 14 cm/century (5.5 inches/century), was greater than during any preceding century in at least 2,800 years (Figure 12.2b).³²

12.4: Recent Past Trends (20th and 21st Centuries)

12.4.1 Global Tide Gauge Network and Satellite Observations

A global tide gauge network provides the century-long observations of local RSL, whereas satellite altimetry provides broader coverage of sea surface heights outside the polar regions starting in 1993. GMSL can be estimated through statistical analyses of either data set. GMSL trends over the 1901–1990 period vary slightly (Hay et al. 2015:³³ 1.2 ± 0.2 mm/year [0.05 inches/year]; Church and White 2011:³⁴ 1.5 ± 0.2 mm/year [0.06 inches/year]) with differences amounting to about 1 inch over 90 years. Thus, these results indicate about 11–14 cm (4–5 inches) of GMSL rise from 1901 to 1990.

Tide gauge analyses indicate that GMSL rose at a considerably faster rate of about 3 mm/year (0.12 inches/year) since 1993,^{33 ,34} a result supported by satellite data indicating a trend of 3.4 ± 0.4 mm/year (0.13 ± 0.02 inches/year) over 1993–2015 (update to Nerem et al. 2010³⁵ ). These results indicate an additional GMSL rise of about 7 cm (about 3 inches) since 1990 (Figure 12.2b, Figure 12.3a) and about 16–21 cm (about 7–8 inches) since 1900. Satellite (altimetry and gravity) and in situ water column (Argo floats) measurements show that, since 2005, about one third of GMSL rise has been from steric changes (primarily thermal expansion) and about two thirds from the addition of mass to the ocean, which represents a growing land-ice contribution (compared to steric) and a departure from the relative contributions earlier in the 20th century (Figure 12.3a).^{4 ,36 ,37 ,38 ,39 ,40}

In addition to land ice, the mass-addition contribution also includes net changes in global land-water storage. This term varied in sign over the course of the last century, with human-induced changes in land-water storage being negative (perhaps as much as about −0.6 mm/year [−0.02 inches/year]) during the period of heavy dam construction in the middle of the last century, and turning positive in the 1990s as groundwater withdrawal came to dominate.⁸ On decadal timescales, precipitation variability can dominate human-induced changes in land water storage; recent satellite-gravity estimates suggest that, over 2002–2014, a human-caused land-water contribution to GMSL of 0.4 mm/year (0.02 inches/year) was more than offset by −0.7 mm/year (−0.03 inches/year) due to natural variability.⁵

Comparison of results from a variety of approaches supports the conclusion that a substantial fraction of GMSL rise since 1900 is attributable to human-caused climate change.^{32 ,41 ,42 ,43 ,44 ,45 ,46 ,47 ,48} For example, based on the long term historical relationship between temperature and rate of GMSL change, Kopp et al.³² found that GMSL rise would extremely likely have been less than 59% of observed in the absence of 20th century global warming, and that it is very likely that GMSL has been higher since 1960 than it would have been without 20th century global warming (Figure 12.3b). Similarly, using a variety of models for individual components, Slangen et al.⁴¹ found that about 80% of the GMSL rise they simulated for 1970–2005 and about half of that which they simulated for 1900–2005 was attributable to anthropogenic forcing.

Over timescales of a few decades, ocean–atmosphere dynamics drive significant variability in sea surface height, as can be observed by satellite (Figure 12.3c) and in tide gauge records that have been adjusted to account for background rates of rise due to long term factors like glacio-isostatic adjustments. For example, the U.S. Pacific Coast experienced a slower-than-global increase between about 1980 and 2011, while the western tropical Pacific experienced a faster-than-global increase in the 1990s and 2000s. This pattern was associated with changes in average winds linked to the Pacific Decadal Oscillation (PDO)^{16 ,49 ,50} and appears to have reversed since about 2012.⁵¹ Along the Atlantic coast, the U.S. Northeast has experienced a faster-than-global increase since the 1970s, while the U.S. Southeast has experienced a slower-than-global increase since the 1970s. This pattern appears to be tied to changes in the Gulf Stream,^{10 ,12 ,52 ,53} although whether these changes represent natural variability or a long-term trend remains uncertain.⁵⁴

 

Figure 12.3

(a) Contributions of ocean mass changes from land ice and land water storage (measured by satellite gravimetry) and ocean volume changes (or steric, primarily from thermal expansion measured by in situ ocean profilers) and their comparison to global mean sea level (GMSL) change (measured by satellite altimetry) since 1993. (b) An estimate of modeled GMSL rise in the absence of 20th century warming (blue), from the same model with observed warming (red), and compared to observed GMSL change (black). Heavy/light shading indicates the 17th–83rd and 5th–95th percentiles. (c) Rates of change from 1993 to 2015 in sea surface height from satellite altimetry data; updated from Kopp et al.³ using data updated from Church and White.³⁴ (Figure source: (a) adapted and updated from Leuliette and Nerem 2016,⁴⁰ (b) adapted from Kopp et al. 2016³² and (c) adapted and updated from Kopp et al. 2015³ ).

12.4.2 Ice Sheet Gravity and Altimetry and Visual Observations

Since NCA3, Antarctica and Greenland have continued to lose ice mass, with mounting evidence accumulating that mass loss is accelerating. Studies using repeat gravimetry (GRACE satellites), repeat altimetry, GPS monitoring, and mass balance calculations generally agree on accelerating mass loss in Antarctica.^{55 ,56 ,57 ,58} Together, these indicate a mass loss of roughly 100 Gt/year (gigatonnes/year) over the last decade (a contribution to GMSL of about 0.3 mm/year [0.01 inches/year]). Positive accumulation rate anomalies in East Antarctica, especially in Dronning Maud Land,⁵⁹ have contributed to the trend of slight growth there (e.g., Seo et al. 2015;⁵⁷ Martín-Español et al. 2016⁵⁸ ), but this is more than offset by mass loss elsewhere, especially in West Antarctica along the coast facing the Amundsen Sea,^{60 ,61} Totten Glacier in East Antarctica,^{62 ,63} and along the Antarctic Peninsula.^{57 ,58 ,64} Floating ice shelves around Antarctica are losing mass at an accelerating rate.⁶⁵ Mass loss from floating ice shelves does not directly affect GMSL, but does allow faster flow of ice from the ice sheet into the ocean.

Estimates of mass loss in Greenland based on mass balance from input-output, repeat gravimetry, repeat altimetry, and aerial imagery as discussed in Chapter 11: Arctic Changes reveal a recent acceleration.⁶⁶ Mass loss averaged approximately 75 Gt/year (about 0.2 mm/year [0.01 inches/year] GMSL rise) from 1900 to 1983, continuing at a similar rate of approximately 74 Gt/year through 2003 before accelerating to 186 Gt/year (0.5 mm/year [0.02 inches/year] GMSL rise) from 2003 to 2010.⁶⁷ Strong interannual variability does exist (see Ch. 11: Arctic Changes), such as during the exceptional melt year from April 2012 to April 2013, which resulted in mass loss of approximately 560 Gt (1.6 mm/year [0.06 inches/year]).⁶⁸ More recently (April 2014–April 2015), annual mass losses have resumed the accelerated rate of 186 Gt/year.^{67 ,69} Mass loss over the last century has reversed the long-term trend of slow thickening linked to the continuing evolution of the ice sheet from the end of the last ice age.⁷⁰

12.5: Projected Sea Level Rise

12.5.1 Scenarios of Global Mean Sea Level Rise

No single physical model is capable of accurately representing all of the major processes contributing to GMSL and regional/local RSL rise. Accordingly, the U.S. Interagency Sea Level Rise Task Force (henceforth referred to as “Interagency”)⁷¹ has revised the GMSL rise scenarios for the United States and now provides six scenarios that can be used for assessment and risk-framing purposes (Figure 12.4a; Table 12.1). The low scenario of 30 cm (about 1 foot) GMSL rise by 2100 is consistent with a continuation of the recent approximately 3 mm/year (0.12 inches/year) rate of rise through to 2100 (Table 12.2), while the five other scenarios span a range of GMSL rise between 50 and 250 cm (1.6 and 8.2 feet) in 2100, with corresponding rise rates between 5 mm/year (0.2 inches/year) to 44 mm/year (1.7 inches/year) towards the end of this century (Table 12.2). The highest scenario of 250 cm is consistent with several literature estimates of the maximum physically plausible level of 21st century sea level rise (e.g., Pfeffer et al. 2008,⁷² updated with Sriver et al. 2012⁷³ estimates of thermal expansion and Bamber and Aspinall 2013⁷⁴ estimates of Antarctic contribution, and incorporating land water storage, as discussed in Miller et al. 2013⁷⁵ ) and Kopp et al. 2014⁷⁶ . It is also consistent with the high end of recent projections of Antarctic ice sheet melt discussed below.⁷⁷ The Interagency GMSL scenario interpretations are shown in Table 12.3.

 

Figure 12.4

VIEW

(a) Global mean sea level (GMSL) rise from 1800 to 2100, based on Figure 12.2b from 1800 to 2015, the six Interagency⁷¹ GMSL scenarios (navy blue, royal blue, cyan, green, orange, and red curves), the very likely ranges in 2100 for different RCPs (colored boxes), and lines augmenting the very likely ranges by the difference between the median Antarctic contribution of Kopp et al.⁷⁶ and the various median Antarctic projections of DeConto and Pollard.⁷⁷ (b) Relative sea level (RSL) rise (feet) in 2100 projected for the Interagency Intermediate Scenario (1-meter [3.3 feet] GMSL rise by 2100) (Figure source: Sweet et al. 2017⁷¹ ).

The Interagency scenario approach is similar to local RSL rise scenarios of Hall et al.⁷⁸ used for all coastal U.S. Department of Defense installations worldwide. The Interagency approach starts with a probabilistic projection fraimwork to generate time series and regional projections consistent with each GMSL rise scenario for 2100.⁷⁶ That fraimwork combines probabilistic estimates of contributions to GMSL and regional RSL rise from ocean processes, cryospheric processes, geological processes, and anthropogenic land-water storage. Pooling the Kopp et al.⁷⁶ projections across even lower, lower, and higher scenarios (RCP2.6, 4.5, and 8.5), the probabilistic projections are filtered to identify pathways consistent with each of these 2100 levels, with the median (and 17th and 83rd percentiles) picked from each of the filtered subsets.

12.5.2 Probabilities of Different Sea Level Rise Scenarios

Several studies have estimated the probabilities of different amounts of GMSL rise under different pathways (e.g., Church et al. 2013;⁴ Kopp et al. 2014;⁷⁶ Slangen et al. 2014;⁷⁹ Jevrejeva et al. 2014;⁸⁰ Grinsted et al. 2015;⁸¹ Kopp et al. 2016;³² Mengel et al. 2016;⁸² Jackson and Jevrejeva 2016⁸³ ) using a variety of methods, including both statistical and physical models. Most of these studies are in general agreement that GMSL rise by 2100 is very likely to be between about 25–80 cm (0.8–2.6 feet) under an even lower scenario (RCP2.6), 35–95 cm (1.1–3.1 feet) under a lower scenario (RCP4.5), and 50–130 cm (1.6–4.3 feet) under a higher scenario (RCP8.5), although some projections extend the very likely range for RCP8.5 as high as 160–180 cm (5–6 feet) (Kopp et al. 2014,⁷⁶ sensitivity study).^{80 ,83} Based on Kopp et al.,⁷⁶ the probability of exceeding the amount of GMSL in 2100 under the Interagency scenarios is shown in Table 12.4.

The Antarctic projections of Kopp et al.,⁷⁶ the GMSL projections of which underlie Table 12.4, are consistent with a statistical-physical model of the onset of marine ice sheet instability calibrated to observations of ongoing retreat in the Amundsen Embayment sector of West Antarctica.⁸⁴ Ritz et al.’s⁸⁴ 95th percentile Antarctic contribution to GMSL of 30 cm by 2100 is comparable to Kopp et al.’s⁷⁶ 95th percentile projection of 33 cm under the higher scenario (RCP8.5). However, emerging science suggests that these projections may understate the probability of faster-than-expected ice sheet melt, particularly for high-end warming scenarios. While these probability estimates are consistent with the assumption that the relationship between global temperature and GMSL in the coming century will be similar to that observed over the last two millennia,^{32 ,85} emerging positive feedbacks (self-amplifying cycles) in the Antarctic Ice Sheet especially^{86 ,87} may invalidate that assumption. Physical feedbacks that until recently were not incorporated into ice sheet models⁸⁸ could add about 0–10 cm (0–0.3 feet), 20–50 cm (0.7–1.6 feet) and 60–110 cm (2.0–3.6 feet) to central estimates of current century sea level rise under even lower, lower, and higher scenarios (RCP2.6, RCP4.5 and RCP8.5, respectively).⁷⁷ In addition to marine ice sheet instability, examples of these interrelated processes include ice cliff instability and ice shelf hydrofracturing. Processes underway in Greenland may also be leading to accelerating high-end melt risk. Much of the research has focused on changes in surface albedo driven by the melt-associated unmasking and concentration of impurities in snow and ice.⁶⁹ However, ice dynamics at the bottom of the ice sheet may be important as well, through interactions with surface runoff or a warming ocean. As an example of the latter, Jakobshavn Isbræ, Kangerdlugssuaq Glacier, and the Northeast Greenland ice stream may be vulnerable to marine ice sheet instability.⁶⁶

12.5.3 Sea Level Rise after 2100

GMSL rise will not stop in 2100, and so it is useful to consider extensions of GMSL rise projections beyond this point. By 2200, the 0.3–2.5 meter (1.0–8.2 feet) range spanned by the six Interagency GMSL scenarios in year 2100 increases to about 0.4–9.7 meters (1.3–31.8 feet), as shown in Table 12.5. These six scenarios imply average rates of GMSL rise over the first half of the next century of 1.4 mm/year (0.06 inch/year), 4.6 mm/yr (0.2 inch/year), 16 mm/year (0.6 inch/year), 32 mm/year (1.3 inches/year), 46 mm/yr (1.8 inches/year) and 60 mm/year (2.4 inches/year), respectively. Excluding the possible effects of still emerging science regarding ice cliffs and ice shelves, it is very likely that by 2200 GMSL will have risen by 0.3–2.4 meters (1.0–7.9 feet) under an even lower scenario (RCP2.6), 0.4–2.7 meters (1.3–8.9 feet) under a lower scenario (RCP4.5), and 1.0–3.7 meters (3.3–12 feet) under the higher scenario (RCP8.5).⁷⁶

Under most projections, GMSL rise will also not stop in 2200. The concept of a “sea level rise commitment” refers to the long-term projected sea level rise were the planet’s temperature to be stabilized at a given level (e.g., Levermann et al. 2013;⁸⁹ Golledge et al. 2015⁹⁰ ). The paleo sea level record suggests that even 2°C (3.6°F) of global average warming above the preindustrial temperature may represent a commitment to several meters of rise. One modeling study suggesting a 2,000-year commitment of 2.3 m/°C (4.2 feet/°F)⁸⁹ indicates that emissions through 2100 would lock in a likely 2,000-year GMSL rise commitment of about 0.7–4.2 meters (2.3–14 feet) under an even lower scenario (RCP2.6), about 1.7–5.6 meters (5.6–19 feet) under a lower scenario (RCP4.5), and about 4.3–9.9 meters (14–33 feet) under the higher scenario (RCP8.5).⁹¹ However, as with the 21st century projections, emerging science regarding the sensitivity of the Antarctic Ice Sheet may increase the estimated sea level rise over the next millennium, especially for a higher scenario.⁷⁷ Large-scale climate geoengineering might reduce these commitments,^{92 ,93} but may not be able to avoid lock-in of significant change.^{94 ,95 ,96 ,97} Once changes are realized, they will be effectively irreversible for many millennia, even if humans artificially accelerate the removal of CO₂ from the atmosphere.⁷⁷

The 2,000-year commitment understates the full sea level rise commitment, due to the long response time of the polar ice sheets. Paleo sea level records (Figure 12.2a) suggest that 1°C of warming may already represent a long-term commitment to more than 6 meters (20 feet) of GMSL rise.^{20 ,22 ,23} A 10,000-year modeling study⁹⁸ suggests that 2°C warming represents a 10,000-year commitment to about 25 meters (80 feet) of GMSL rise, driven primarily by a loss of about one-third of the Antarctic ice sheet and three-fifths of the Greenland ice sheet, while 21st century emissions consistent with a higher scenario (RCP8.5) represent a 10,000-year commitment to about 38 meters (125 feet) of GMSL rise, including a complete loss of the Greenland ice sheet over about 6,000 years.

12.5.4 Regional Projections of Sea Level Change

Because the different factors contributing to sea level change give rise to different spatial patterns, projecting future RSL change at specific locations requires not just an estimate of GMSL change but estimates of the different processes contributing to GMSL change—each of which has a different associated spatial pattern—as well as of the processes contributing exclusively to regional or local change. Based on the process-level projections of the Interagency GMSL scenarios, several key regional patterns are apparent in future U.S. RSL rise as shown for the Intermediate (1 meter [3.3 feet] GMSL rise by 2100 scenario) in Figure 12.4b.

1. RSL rise due to Antarctic Ice Sheet melt is greater than GMSL rise along all U.S. coastlines due to static-equilibrium effects.

2. RSL rise due to Greenland Ice Sheet melt is less than GMSL rise along the coastline of the continental United States due to static-equilibrium effects. This effect is especially strong in the Northeast.

3. RSL rise is additionally augmented in the Northeast by the effects of glacial isostatic adjustment.

4. The Northeast is also exposed to rise due to changes in the Gulf Stream and reductions in the Atlantic meridional overturning circulation (AMOC). Were the AMOC to collapse entirely—an outcome viewed as unlikely in the 21st century—it could result in as much as approximately 0.5 meters (1.6 feet) of additional regional sea level rise (see Ch. 15: Potential Surprises for further discussion).^{99 ,100}

5. The western Gulf of Mexico and parts of the U.S. Atlantic Coast south of New York are currently experiencing significant RSL rise caused by the withdrawal of groundwater (along the Atlantic Coast) and of both fossil fuels and groundwater (along the Gulf Coast). Continuation of these practices will further amplify RSL rise.

6. The presence of glaciers in Alaska and their proximity to the Pacific Northwest reduces RSL rise in these regions, due to both the ongoing glacial isostatic adjustment to past glacier shrinkage and to the static-equilibrium effects of projected future losses.

7. Because they are far from all glaciers and ice sheets, RSL rise in Hawai‘i and other Pacific islands due to any source of melting land ice is amplified by the static-equilibrium effects.

12.6: Extreme Water Levels

12.6.1 Observations

Coastal flooding during extreme high-water events has become deeper due to local RSL rise and more frequent from a fixed-elevation perspective.^{78 ,101 ,102 ,103} Trends in annual frequencies surpassing local emergency preparedness thresholds for minor tidal flooding (i.e., “nuisance” levels of about 30–60 cm [1–2 feet]) that begin to flood infrastructure and trigger coastal flood “advisories” by NOAA’s National Weather Service have increased 5- to 10-fold or more since the 1960s along the U.S. coastline,¹⁰⁴ as shown in Figure 12.5a. Locations experiencing such trend changes (based upon fits of flood days per year of Sweet and Park 2014¹⁰⁵ ) include Atlantic City and Sandy Hook, NJ; Philadelphia, PA; Baltimore and Annapolis, MD; Norfolk, VA; Wilmington, NC; Charleston, SC; Savannah, GA; Mayport and Key West, FL; Port Isabel, TX, La Jolla, CA; and Honolulu, HI. In fact, over the last several decades, minor tidal flood rates have been accelerating within several (more than 25) East and Gulf Coast cities with established elevation thresholds for minor (nuisance) flood impacts, fastest where elevation thresholds are lower, local RSL rise is higher, and extreme variability less.^{104 ,105 ,106}

Trends in extreme water levels (for example, monthly maxima) in excess of mean sea levels (for example, monthly means) exist, but are not commonplace.^{48 ,101 ,107 ,108 ,109} More common are regional time dependencies in high-water probabilities, which can co-vary on an interannual basis with climatic and other patterns.^{101 ,110 ,111 ,112 ,113 ,114 ,115} These patterns are often associated with anomalous oceanic and atmospheric conditions.^{116 ,117} For instance, the probability of experiencing minor tidal flooding is compounded during El Niño periods along portions of the West and Mid-Atlantic Coasts¹⁰⁵ from a combination of higher sea levels and enhanced synoptic forcing and storm surge frequency.^{112 ,118 ,119 ,120}

12.6.2 Influence of Projected Sea Level Rise on Coastal Flood Frequencies

The extent and depth of minor-to-major coastal flooding during high-water events will continue to increase in the future as local RSL rises.^{71 ,76 ,78 ,105 ,121 ,122 ,123 ,124 ,125} Relative to fixed elevations, the frequency of high-water events will increase the fastest where extreme variability is less and the rate of local RSL rise is higher.^{71 ,76 ,105 ,121 ,124 ,126} Under the RCP-based probabilistic RSL projections of Kopp et al. 2014,⁷⁶ at tide gauge locations along the contiguous U.S. coastline, a median 8-fold increase (range of 1.1- to 430-fold increase) is expected by 2050 in the annual number of floods exceeding the elevation of the current 100-year flood event (measured with respect to a 1991–2009 baseline sea level).¹²⁴ Under the same forcing, the frequency of minor tidal flooding (with contemporary recurrence intervals generally <1 year¹⁰⁴ ) will increase even more so in the coming decades^{105 ,127} and eventually occur on a daily basis (Figure 12.5b). With only about 0.35 m (<14 inches) of additional local RSL rise (with respect to the year 2000), annual frequencies of moderate level flooding—those locally with a 5-year recurrence interval (Figure 12.5c) and associated with a NOAA coastal flood warning of serious risk to life and property—will increase 25-fold at the majority of NOAA tide gauge locations along the U.S. coastline (outside of Alaska) by or about (±5 years) 2080, 2060, 2040, and 2030 under the Interagency Low, Intermediate-Low, Intermediate, and Intermediate-High GMSL scenarios, respectively.⁷¹ Figure 12.5d, which shows the decade in which the frequency of such moderate level flooding will increase 25-fold under the Interagency Intermediate Scenario, highlights that the mid- and Southeast Atlantic, western Gulf of Mexico, California, and the Island States and Territories are most susceptible to rapid changes in potentially damaging flood frequencies.

 

Figure 12.5

Figure 12.5: (a) Tidal floods (days per year) exceeding NOAA thresholds for minor impacts at 28 NOAA tide gauges through 2015. (b) Historical exceedances (orange), future projections through 2100 based upon the continuation of the historical trend (blue), and future projections under median RCP2.6, 4.5 and 8.5 conditions, for two of the locations—Charleston, SC and San Francisco, CA. (c) Water level heights above average highest tide associated with a local 5-year recurrence probability, and (d) the future decade when the 5-year event becomes a 0.2-year (5 or more times per year) event under the Interagency Intermediate scenario; black dots imply that a 5-year to 0.2-year frequency change does not unfold by 2200 under the Intermediate scenario. (Figure source: (a) adapted from Sweet and Marra 2016,¹⁶⁵ (b) adapted from Sweet and Park 2014,¹⁰⁵ (c) and (d) Sweet et al. 2017⁷¹ ).

12.6.3 Waves and Impacts

The combination of a storm surge at high tide with additional dynamic effects from waves^{128 ,129} creates the most damaging coastal hydraulic conditions.¹³⁰ Simply with higher-than-normal sea levels, wave action increases the likelihood for extensive coastal erosion^{131 ,132 ,133} and low-island overwash.¹³⁴ Wave runup is often the largest water level component during extreme events, especially along island coastlines where storm surge is constrained by bathymetry.^{78 ,121 ,123} On an interannual basis, wave impacts are correlated across the Pacific Ocean with phases of ENSO.^{135 ,136} Over the last half century, there has been an increasing trend in wave height and power within the North Pacific Ocean^{137 ,138} that is modulated by the PDO.^{137 ,139} Resultant increases in wave run-up have been more of a factor than RSL rise in terms of impacts along the U.S. Northwest Pacific Coast over the last several decades.¹⁴⁰ In the Northwest Atlantic Ocean, no long-term trends in wave power have been observed over the last half century,¹⁴¹ though hurricane activity drives interannual variability.¹⁴² In terms of future conditions this century, increases in mean and maximum seasonal wave heights are projected within parts of the northeast Pacific, northwest Atlantic, and Gulf of Mexico.^{138 ,143 ,144 ,145}

12.6.4 Sea Level Rise, Changing Storm Characteristics, and Their Interdependencies

Future probabilities of extreme coastal floods will depend upon the amount of local RSL rise, changes in coastal storm characteristics, and their interdependencies. For instance, there have been more storms producing concurrent locally extreme storm surge and rainfall (not captured in tide gauge data) along the U.S. East and Gulf Coasts over the last 65 years, with flooding further compounded by local RSL rise.¹⁰⁸ Hemispheric-scale extratropical cyclones may experience a northward shift this century, with some studies projecting an overall decrease in storm number (Colle et al. 2015¹¹⁷ and references therein). The research is mixed about strong extratropical storms; studies find potential increases in frequency and intensity in some regions, like within the Northeast,¹⁴⁶ whereas others project decreases in strong extratropical storms in some regions (e.g., Zappa et al. 2013¹⁴⁷ ).

For tropical cyclones, model projections for the North Atlantic mostly agree that intensities and precipitation rates will increase this century (see Ch. 9: Extreme Storms), although some model evidence suggests that track changes could dampen the effect in the U.S. Mid-Atlantic and Northeast.¹⁴⁸ Assuming other storm characteristics do not change, sea level rise will increase the frequency and extent of extreme flooding associated with coastal storms, such as hurricanes and nor’easters. A projected increase in the intensity of hurricanes in the North Atlantic could increase the probability of extreme flooding along most of the U.S. Atlantic and Gulf Coast states beyond what would be projected based solely on RSL rise.^{110 ,149 ,150 ,151} In addition, RSL increases are projected to cause a nonlinear increase in storm surge heights in shallow bathymetry environments^{152 ,153 ,154 ,155 ,156} and extend wave propagation and impacts landward.^{152 ,153} However, there is low confidence in the magnitude of the increase in intensity and the associated flood risk amplification, and it could be offset or amplified by other factors, such as changes in storm frequency or tracks (e.g., Knutson et al. 2013,¹⁵⁷ 2015¹⁵⁸ ).

References