drivers and uncertainties in past and future sea level changes

Drivers and uncertainties in past and future sea level changes

Dewi Le Bars, Physical Oceanographer at KNMI

HSB 1/04/2016

Introduction

1143

Sea Level Change Chapter 13

13

13.1.3 Processes Affecting Sea Level

This chapter focusses on processes within the ocean, atmosphere, land ice, and hydrological cycle that are climate sensitive and are expected to contribute to sea level change at regional to global scales in the coming decades to centuries (Figure 13.1). Figure 13.2 is a navigation aid for the different sections of this chapter and sections of other chap-ters that are relevant to sea level change.

Changes in ocean currents, ocean density and sea level are all tightly coupled such that changes at one location impact local sea level and sea level far from the location of the initial change, including changes in sea level at the coast in response to changes in open-ocean tem-perature (Landerer et al., 2007; Yin et al., 2010). Although both tem-perature and salinity changes can contribute significantly to region-al sea level change (Church et al., 2010), only temperature change produces a significant contribution to global average ocean volume change due to thermal expansion or contraction (Gregory and Lowe, 2000). Regional atmospheric pressure anomalies also cause sea level to vary through atmospheric loading (Wunsch and Stammer, 1997). All of these climate-sensitive processes cause sea level to vary on a broad range of space and time scales from relatively short-lived events, such as waves and storm surges, to sustained changes over several decades or centuries that are associated with atmospheric and ocean modes of climate variability (White et al., 2005; Miller and Douglas, 2007; Zhang and Church, 2012).

Water and ice mass exchange between the land and the oceans leads to a change in GMSL. A signal of added mass to the ocean propagates rapidly around the globe such that all regions experience a sea level change within days of the mass being added (Lorbacher et al., 2012). In addition, an influx of freshwater changes ocean temperature and salinity and thus changes ocean currents and local sea level (Stammer, 2008; Yin et al., 2009), with signals taking decades to propagate around

Figure 13.1 | Climate-sensitive processes and components that can influence global and regional sea level and are considered in this chapter. Changes in any one of the com-ponents or processes shown will result in a sea level change. The term ‘ocean properties’ refers to ocean temperature, salinity and density, which influence and are dependent on ocean circulation. Both relative and geocentric sea level vary with position. Note that the geocenter is not shown.

the global ocean. The coupled atmosphere–ocean system can also adjust to temperature anomalies associated with surface freshwater anomalies through air–sea feedbacks, resulting in dynamical adjust-ments of sea level (Okumura et al., 2009; Stammer et al., 2011). Water mass exchange between land and the ocean also results in patterns of sea level change called ‘sea level fingerprints’ (Clark and Lingle, 1977; Conrad and Hager, 1997; Mitrovica et al., 2001) due to change in the gravity field and vertical movement of the ocean floor associated with visco-elastic Earth deformation (Farrell and Clark, 1976). These changes in mass distribution also affect the Earth’s inertia tensor and therefore rotation, which produces an additional sea level response (Milne and Mitrovica, 1998).

There are other processes that affect sea level but are not associated with contemporary climate change. Some of these result in changes that are large enough to influence the interpretation of observational records and sea level projections at regional and global scales. In par-ticular, surface mass transfer from land ice to oceans during the last deglaciation contributes significantly to present-day sea level change due to the ongoing visco-elastic deformation of the Earth and the cor-responding changes of the ocean floor height and gravity (referred to as glacial isostatic adjustment (GIA)) (Lambeck and Nakiboglu, 1984; Peltier and Tushingham, 1991). Ice sheets also have long response times and so continue to respond to past climate change (Section 13.1.5).

Anthropogenic processes that influence the amount of water stored in the ground or on its surface in lakes and reservoirs, or cause changes in land surface characteristics that influence runoff or evapotranspiration rates, will perturb the hydrological cycle and cause sea level change (Sahagian, 2000; Wada et al., 2010). Such processes include water impoundment (dams, reservoirs), irrigation schemes, and groundwater depletion (Section 13.4.5).

From IPCC AR5

Drivers of sea level change

Global:- Steric effects (warming)- Mass change (land ice melting)

Regional:- Local steric effects (thermosteric and halosteric)- Dynamical(ocean circulation, wind, atm. pressure, modes of climate variability)- Gravitational attraction

Coastal:- Dynamical(wave setup,storm surge, tides)- Earth rebound

KNMI and IPCC scenarios

Outlook

• Global versus local sea level change

• Physical processes: - Thermal expansion (volume change) - Melting of ice sheets (mass change) - Gravitation (displacement)

• Past changes

• Future changes

Global observations: satellite altimetry

AVISO product

1148

Chapter 13 Sea Level Change

13

Frequently Asked Questions FAQ 13.1 | Why Does Local Sea Level Change Differ from the Global Average?

Shifting surface winds, the expansion of warming ocean water, and the addition of melting ice can alter ocean cur-rents which, in turn, lead to changes in sea level that vary from place to place. Past and present variations in the distribution of land ice affect the shape and gravitational field of the Earth, which also cause regional fluctuations in sea level. Additional variations in sea level are caused by the influence of more localized processes such as sedi-ment compaction and tectonics.

Along any coast, vertical motion of either the sea or land surface can cause changes in sea level relative to the land (known as relative sea level). For example, a local change can be caused by an increase in sea surface height, or by a decrease in land height. Over relatively short time spans (hours to years), the influence of tides, storms and climatic variability—such as El Niño—dominates sea level variations. Earthquakes and landslides can also have an effect by causing changes in land height and, sometimes, tsunamis. Over longer time spans (decades to centuries), the influ-ence of climate change—with consequent changes in volume of ocean water and land ice—is the main contributor to sea level change in most regions. Over these longer time scales, various processes may also cause vertical motion of the land surface, which can also result in substantial changes in relative sea level.

Since the late 20th century, satellite measurements of the height of the ocean surface relative to the center of the Earth (known as geocentric sea level) show differing rates of geocentric sea level change around the world (see FAQ 13.1, Figure 1). For example, in the western Pacific Ocean, rates were about three times greater than the global mean value of about 3 mm per year from 1993 to 2012. In contrast, those in the eastern Pacific Ocean are lower than the global mean value, with much of the west coast of the Americas experiencing a fall in sea surface height over the same period. (continued on next page)

Pago PagoManilaAntofagasta

San Francisco Charlottetown Stockholm

Antofagasta

Manila

Pago Pago

−14−12−10−8−6−4−2

02468

101214

Sea

level

chan

ge (m

m yr

-1)

500

250

0

250

500

Sea

level

(mm

)1960 1980 2000

Year

San Francisco

1960 1980 2000Year

Charlottetown

1960 1980 2000Year

Stockholm

-500

-250

0

250

500

Sea

level

(mm

)

1960 1980 2000Year

1960 1980 2000Year

1960 1980 2000Year

FAQ13.1, Figure 1 | Map of rates of change in sea surface height (geocentric sea level) for the period 1993–2012 from satellite altimetry. Also shown are relative sea level changes (grey lines) from selected tide gauge stations for the period 1950–2012. For comparison, an estimate of global mean sea level change is also shown (red lines) with each tide gauge time series. The relatively large, short-term oscillations in local sea level (grey lines) are due to the natural climate variability described in the main text. For example, the large, regular deviations at Pago Pago are associated with the El Niño-Southern Oscillation.

Local: Altimetry and tide gauges, from IPCC AR5

Ice sheet mass balance: MB = SMB - D

Ice discharge D

Snowfall

Meltwater runoff

Surface mass balance (SMB)

Calving of giant iceberg from Ross Ice Shelf, Antarctica

NASA/MODIS

Effect of gravity on sea level change

2200 km6700 km

Effect of gravity on sea level change

1149


13

FAQ 13.1 (continued)

Much of the spatial variation shown in FAQ 13.1, Figure 1 is a result of natural climate variability—such as El Niño and the Pacific Decadal Oscillation—over time scales from about a year to several decades. These climate variations alter surface winds, ocean currents, temperature and salinity, and hence affect sea level. The influence of these processes will continue during the 21st century, and will be superimposed on the spatial pattern of sea level change associated with longer term climate change, which also arises through changes in surface winds, ocean currents, temperature and salinity, as well as ocean volume. However, in contrast to the natural variability, the longer term trends accu-mulate over time and so are expected to dominate over the 21st century. The resulting rates of geocentric sea level change over this longer period may therefore exhibit a very different pattern from that shown in FAQ 13.1, Figure 1.

Tide gauges measure relative sea level, and so they include changes resulting from vertical motion of both the land and the sea surface. Over many coastal regions, vertical land motion is small, and so the long-term rate of sea level change recorded by coastal and island tide gauges is similar to the global mean value (see records at San Francisco and Pago Pago in FAQ 13.1, Figure 1). In some regions, vertical land motion has had an important influence. For example, the steady fall in sea level recorded at Stockholm (FAQ 13.1, Figure 1) is caused by uplift of this region after the melting of a large (>1 km thick) continental ice sheet at the end of the last Ice Age, between ~20,000 and ~9000 years ago. Such ongoing land deformation as a response to the melting of ancient ice sheets is a significant contributor to regional sea level changes in North America and northwest Eurasia, which were covered by large continental ice sheets during the peak of the last Ice Age.

In other regions, this process can also lead to land subsidence, which elevates relative sea levels, as it has at Char-lottetown, where a relatively large increase has been observed, compared to the global mean rate (FAQ 13.1, Figure 1). Vertical land motion due to movement of the Earth’s tectonic plates can also cause departures from the global mean sea level trend in some areas—most significantly, those located near active subduction zones, where one tec-tonic plate slips beneath another. For the case of Antofagasta (FAQ 13.1, Figure 1) this appears to result in steady land uplift and therefore relative sea level fall.

In addition to regional influences of vertical land motion on relative sea level change, some processes lead to land motion that is rapid but highly localized. For example, the greater rate of rise relative to the global mean at Manila (FAQ 13.1, Figure 1) is dominated by land subsid-ence caused by intensive groundwater pumping. Land subsidence due to natural and anthropogenic processes, such as the extraction of groundwater or hydrocarbons, is common in many coastal regions, particularly in large river deltas.

It is commonly assumed that melting ice from glaciers or the Greenland and Antarctic ice sheets would cause globally uniform sea level rise, much like filling a bath tub with water. In fact, such melting results in region-al variations in sea level due to a variety of processes, including changes in ocean currents, winds, the Earth’s gravity field and land height. For example, computer models that simulate these latter two processes predict a regional fall in relative sea level around the melting ice sheets, because the gravitational attraction between ice and ocean water is reduced, and the land tends to rise as the ice melts (FAQ 13.1, Figure 2). However, further away from the ice sheet melting, sea level rise is enhanced, compared to the global average value.

In summary, a variety of processes drive height changes of the ocean surface and ocean floor, resulting in distinct spatial patterns of sea level change at local to regional scales. The combination of these processes produces a complex pattern of total sea level change, which varies through time as the relative contribution of each process changes. The global average change is a useful single value that reflects the contribution of climatic processes (e.g., land-ice melting and ocean warming), and represents a good estimate of sea level change at many coastal loca-tions. At the same time, however, where the various regional processes result in a strong signal, there can be large departures from the global average value.

FAQ13.1, Figure 2 | Model output showing relative sea level change due to melting of the Greenland ice sheet and the West Antarctic ice sheet at rates of 0.5 mm yr–1 each (giving a global mean value for sea level rise of 1 mm yr–1). The modelled sea level changes are less than the global mean value in areas near the melting ice but enhanced further afield. (Adapted from Milne et al., 2009)

−3.0 −2.0 −1.0 0.0 0.2 0.4 0.6 0.8 1.0 1.1 1.2 1.3Sea level change (mm yr-1)

Change of sea level due to melting of Greenland and Antarctic ice sheets at 0.5mm/year each (from IPCC AR5).

Past sea level variations

observations are not unduly weighted in estimating the commonGSL signal gðtÞ.Because a constant-rate trend in gðtÞ could also be interpreted as

a regional linear trend that is present at all reconstruction sites butis not truly global, we condition the model on the assumption thatmean GSL over −100–100 CE is equal to mean GSL over 1600–1800 CE and focus on submillennial variations (Fig. 1A). We chosethe first window to encompass the beginning of the Common Eraand the last window to cover the last 2 centuries before the de-velopment of a tide-gauge network outside of northern Europe.

The priors for each component are characterized by hyper-parameters that comprise amplitudes (for all three components),timescales of variability [for gðtÞ and mðx, tÞ], and spatial scalesof variability [for lðxÞ and mðx, tÞ] (Dataset S1, c). We considerfive priors with different hyperparameters (see Supporting In-formation). The presented rates are taken from prior ML2,1,which is optimized under the assumption that the a prioritimescales of variability in global and regional sea-level changeare the same. Results from the four alternative priors are pre-sented in Supporting Information. Quoted probabilities are con-servatively taken as minima across all five priors. Illustrative fitsat specific sites are shown in Fig. S2.

Results and DiscussionCommon Era Reconstruction. Pre-20th-century Common Era GSLvariability was very likely (probability P= 0.90) between ∼± 7 cmand ±11 cm in amplitude (Fig. 1A and Dataset S1, e). GSL rosefrom 0 CE to 700 CE (P≥ 0.98) at a rate of 0.1 ± 0.1 mm/y (2σ),was nearly stable from 700 CE to 1000 CE, then fell from 1000CE and 1400 CE (P≥ 0.98) at a rate of 0.2 ± 0.2 mm/y (Fig. 1A).GSL likely rose from 1400 CE to 1600 CE (P≥ 0.75) at 0.3 ±0.4 mm/y and fell from 1600 CE to 1800 CE (P≥ 0.86) at0.3 ± 0.3 mm/y.Historic GSL rise began in the 19th century, and it is very

likely (P≥ 0.93) that GSL has risen over every 40-y intervalsince 1860 CE. The average rate of GSL rise was 0.4± 0.5 mm/yfrom 1860 CE to 1900 CE and 1.4± 0.2 mm/y over the 20thcentury. It is extremely likely (P≥ 0.95) that 20th century GSLrise was faster than during any preceding century since at least−800 CE.The spatial coverage of the combined proxy and long-term

tide-gauge dataset is incomplete. The available data are suffi-cient to reduce the posterior variance in the mean 0–1700 CErate by >10% relative to the prior variance along coastlines inmuch of the North Atlantic and the Gulf of Mexico, and parts ofthe Mediterranean, the South Atlantic, the South Pacific, andAustralasia (Fig. 2A). High-resolution proxy records are notablylacking from Asia, most of South America, and most of Africa.Nevertheless, despite the incomplete coverage and regionalvariability, sensitivity analyses of different data subsets indicatethat key features of the GSL curve—a rise over 0–700 CE, a fallover 1000–1400 CE, and a rise beginning in the late 19thcentury—are not dependent on records from any one region(Dataset S1, f). By contrast, the rise over 1400–1600 CE and fallover 1600–1800 CE are not robust to the removal of data fromthe western North Atlantic.On millennial and longer timescales, regional RSL change can

differ significantly from GSL change as a result of GIA, tec-tonics, and sediment compaction (Fig. 2). For example, over 0–1700 CE, RSL rose at 1.5 ± 0.1 mm/y in New Jersey, on thecollapsing forebulge of the former Laurentide Ice Sheet, and fellat 0.1± 0.1 mm/y on Christmas Island, in the far field of all latePleistocene ice cover (Dataset S1, g). Detrended RSL (after re-moval of the average 0–1700 CE rate) reveals notable patterns oftemporal variability, especially in the western North Atlantic,where the highest-resolution reconstructions exist. Rates of RSLchange in New Jersey and North Carolina vary from the long-term mean in opposite directions over 0–700 CE and 1000–1400CE (Fig. 2 and Dataset S1, g). Over 0–700 CE, a period overwhich GSL rose at 0.1± 0.1 mm/y, detrended RSL rose in NewJersey (P≥ 0.91) while it fell in North Carolina (P≥ 0.88). Con-versely, over 1000–1400 CE, while GSL was falling, detrendedRSL fell in New Jersey (P> 0.90) while it rose in North Car-olina (P≥ 0.99). This pattern is consistent with changes in theGulf Stream (16) or in mean nearshore wind stress (19). If drivenby the Gulf Stream, it suggests a weakening or polar migration ofthe Gulf Stream over 0–700 CE, with a strengthening or equatorialmigration occurring over 1000–1400 CE.

Glob

alSe

a Lev

el( cm

)

C

Year (CE)

(change in scale)

E

Marcott et al. (2013)Global sea level (this study) Mann et al. (2009)

D

RCP 8.5RCP 4.5RCP 2.6Maximum range across all RCPs and calibrations

FRCP 2

.6RC

P 4.5 RC

P 8.5

1850 1900 1950 2000 2050 2100-20

0

20

40

60

80

100

120

B1

2

H

A

Glob

al Te

mpe

ratu

re An

omal

y (o C)

Glob

alSe

aLev

el(cm

)Gl

obal

SeaL

evel

(cm)

2000-400 -200 0 200 400 600 800 1000 1200 1400 1600 1800-10

-5

0

5

10

15

20

-400 -200 0 200 400 600 800 1000 1200 1400 1600 1800 2000

-1.0-0.8-0.6-0.4-0.20.00.20.40.6

-400 -200 0 200 400 600 800 1000 1200 1400 1600 1800 2000-10

-5

0

5

10

15

20

Fig. 1. (A) Global sea level (GSL) under prior ML2,1. Note that the model isinsensitive to small linear trends in GSL over the Common Era, so the relativeheights of the 300–1000 CE and 20th century peaks are not comparable. (B)The 90% credible intervals for semiempirical hindcasts of 20th century sea-level change under historical temperatures (H) and counterfactual scenarios1 and 2, using both temperature calibrations. (C) Reconstructions of globalmean temperature anomalies relative to the 1850–2000 CE mean (1, 2). (D)Semiempirical fits to the GSL curve using the two alternative temperaturereconstructions. (E) As in B, including 21st century projections for RCPs 2.6,4.5, and 8.5. Red lines show the fifth percentile of RCP 2.6 and 95th per-centile of RCP 8.5. (F) The 90% credible intervals for 2100 by RCP. In A, B, andD, values are with respect to 1900 CE baseline; in E and F, values are withrespect to 2000 CE baseline. Heavy shading, 67% credible interval; lightshading, 90% credible interval.

2 of 8 | www.pnas.org/cgi/doi/10.1073/pnas.1517056113 Kopp et al.

Sea level reconstruction of the past 2500 years:

“20th century sea level rise is faster than over the previous 27 centuries”(Kopp et al., PNAS 2016)

Past sea level variations

1151


13

heat uptake (Levitus et al., 2001). The models including natural forcing are closer to observations, though with a tendency to underestimate the trend by about 10% (Sections 9.4.2.2 and 10.4.1).

Gregory (2010) and Gregory et al. (2013a) proposed that AOGCMs underestimate ocean heat uptake in their historical simulations because their control experiments usually omit volcanic forcing, so the imposition of historical volcanic forcing on the simulated climate system represents a time mean negative forcing relative to the con-trol climate. The apparent long persistence of the simulated oceanic cooling following the 1883 eruption of Krakatau (Delworth et al., 2005; Gleckler et al., 2006a, 2006b; Gregory et al., 2006) is a consequence of this bias, which also causes a model-dependent underestimate of up to 0.2 mm yr–1 of thermal expansion on average during the 20th century (Gregory et al., 2013a, 2013b). This implies that CMIP5 results may be similarly underestimated, depending on the details of the indi-vidual model control runs. Church et al. (2013) proposed a correction of 0.1 mm yr–1 to the model mean rate, which we apply in the sea level budget in Table 13.1 and Figure 13.7. The corrected CMIP5 model mean rate for 1971–2010 is close to the central observational estimate; the model mean rate for 1993–2010 exceeds the central observational estimate but they are not statistically different given the uncertainties (Table 13.1 and Figure 13.4a). This correction is not made to projec-tions of thermal expansion because it is very small compared with the projected increase in the rate (Section 13.5.1).

In view of the improvement in observational estimates of thermal expansion, the good agreement of historical model results with obser-vational estimates, and their consistency with understanding of the

energy budget and RF of the climate system (Box 13.1), we have high confidence in the projections of thermal expansion using AOGCMs.

13.3.2 Glaciers

13.3.2.1 Observed

‘Glaciers’ are defined here as all land-ice masses, including those peripheral to (but not including) the Greenland and Antarctic ice sheets. The term ‘glaciers and ice caps’ was applied to this category in the AR4. Changes in aggregate glacier volume have conventional-ly been determined by various methods of repeat mapping of surface elevation to detect elevation (and thus volume) change. Mass changes are determined by compilation and upscaling of limited direct observa-tions of surface mass balance (SMB). Since 2003, gravity observations from Gravity Recovery and Climate Experiment (GRACE) satellites have been used to detect mass change of the world’s glaciers.

The combined records indicate that a net decline of global glacier volume began in the 19th century, before significant anthropogenic RF had started, and was probably the result of warming associated with the termination of the Little Ice Age (Crowley, 2000; Gregory et al., 2006, 2013b). Global rates of glacier volume loss did not increase significantly during much of the 20th century (Figure 4.12). In part this may have been because of an enhanced rate of loss due to unforced high-latitude variability early in the century, while anthropogenic warming was still comparatively small (Section 13.3.2.2). It is likely that anthropogenic forcing played a statistically significant role in acceleration of global glacier losses in the latter decades of the 20th

Table 13.1 | Global mean sea level budget (mm yr–1) over different time intervals from observations and from model-based contributions. Uncertainties are 5 to 95%. The Atmo-sphere–Ocean General Circulation Model (AOGCM) historical integrations end in 2005; projections for RCP4.5 are used for 2006–2010. The modelled thermal expansion and glacier contributions are computed from the CMIP5 results, using the model of Marzeion et al. (2012a) for glaciers. The land water contribution is due to anthropogenic intervention only, not including climate-related fluctuations.

Notes:a Data for all glaciers extend to 2009, not 2010.b This contribution is not included in the total because glaciers in Greenland are included in the observational assessment of the Greenland ice sheet.c Observed GMSL rise – modelled thermal expansion – modelled glaciers – observed land water storage.

Source 1901–1990 1971–2010 1993–2010Observed contributions to global mean sea level (GMSL) rise

Thermal expansion – 0.8 [0.5 to 1.1] 1.1 [0.8 to 1.4]

Glaciers except in Greenland and Antarcticaa 0.54 [0.47 to 0.61] 0.62 [0.25 to 0.99] 0.76 [0.39 to 1.13]

Glaciers in Greenlanda 0.15 [0.10 to 0.19] 0.06 [0.03 to 0.09] 0.10 [0.07 to 0.13]b

Greenland ice sheet – – 0.33 [0.25 to 0.41]

Antarctic ice sheet – – 0.27 [0.16 to 0.38]

Land water storage –0.11 [–0.16 to –0.06] 0.12 [0.03 to 0.22] 0.38 [0.26 to 0.49]

Total of contributions – – 2.8 [2.3 to 3.4]

Observed GMSL rise 1.5 [1.3 to 1.7] 2.0 [1.7 to 2.3] 3.2 [2.8 to 3.6]

Modelled contributions to GMSL rise

Thermal expansion 0.37 [0.06 to 0.67] 0.96 [0.51 to 1.41] 1.49 [0.97 to 2.02]

Glaciers except in Greenland and Antarctica 0.63 [0.37 to 0.89] 0.62 [0.41 to 0.84] 0.78 [0.43 to 1.13]

Glaciers in Greenland 0.07 [–0.02 to 0.16] 0.10 [0.05 to 0.15] 0.14 [0.06 to 0.23]

Total including land water storage 1.0 [0.5 to 1.4] 1.8 [1.3 to 2.3] 2.8 [2.1 to 3.5]

Residualc 0.5 [0.1 to 1.0] 0.2 [–0.4 to 0.8] 0.4 [–0.4 to 1.2]

1151


13

heat uptake (Levitus et al., 2001). The models including natural forcing are closer to observations, though with a tendency to underestimate the trend by about 10% (Sections 9.4.2.2 and 10.4.1).

Gregory (2010) and Gregory et al. (2013a) proposed that AOGCMs underestimate ocean heat uptake in their historical simulations because their control experiments usually omit volcanic forcing, so the imposition of historical volcanic forcing on the simulated climate system represents a time mean negative forcing relative to the con-trol climate. The apparent long persistence of the simulated oceanic cooling following the 1883 eruption of Krakatau (Delworth et al., 2005; Gleckler et al., 2006a, 2006b; Gregory et al., 2006) is a consequence of this bias, which also causes a model-dependent underestimate of up to 0.2 mm yr–1 of thermal expansion on average during the 20th century (Gregory et al., 2013a, 2013b). This implies that CMIP5 results may be similarly underestimated, depending on the details of the indi-vidual model control runs. Church et al. (2013) proposed a correction of 0.1 mm yr–1 to the model mean rate, which we apply in the sea level budget in Table 13.1 and Figure 13.7. The corrected CMIP5 model mean rate for 1971–2010 is close to the central observational estimate; the model mean rate for 1993–2010 exceeds the central observational estimate but they are not statistically different given the uncertainties (Table 13.1 and Figure 13.4a). This correction is not made to projec-tions of thermal expansion because it is very small compared with the projected increase in the rate (Section 13.5.1).

In view of the improvement in observational estimates of thermal expansion, the good agreement of historical model results with obser-vational estimates, and their consistency with understanding of the

energy budget and RF of the climate system (Box 13.1), we have high confidence in the projections of thermal expansion using AOGCMs.

13.3.2 Glaciers

13.3.2.1 Observed

‘Glaciers’ are defined here as all land-ice masses, including those peripheral to (but not including) the Greenland and Antarctic ice sheets. The term ‘glaciers and ice caps’ was applied to this category in the AR4. Changes in aggregate glacier volume have conventional-ly been determined by various methods of repeat mapping of surface elevation to detect elevation (and thus volume) change. Mass changes are determined by compilation and upscaling of limited direct observa-tions of surface mass balance (SMB). Since 2003, gravity observations from Gravity Recovery and Climate Experiment (GRACE) satellites have been used to detect mass change of the world’s glaciers.

The combined records indicate that a net decline of global glacier volume began in the 19th century, before significant anthropogenic RF had started, and was probably the result of warming associated with the termination of the Little Ice Age (Crowley, 2000; Gregory et al., 2006, 2013b). Global rates of glacier volume loss did not increase significantly during much of the 20th century (Figure 4.12). In part this may have been because of an enhanced rate of loss due to unforced high-latitude variability early in the century, while anthropogenic warming was still comparatively small (Section 13.3.2.2). It is likely that anthropogenic forcing played a statistically significant role in acceleration of global glacier losses in the latter decades of the 20th

Table 13.1 | Global mean sea level budget (mm yr–1) over different time intervals from observations and from model-based contributions. Uncertainties are 5 to 95%. The Atmo-sphere–Ocean General Circulation Model (AOGCM) historical integrations end in 2005; projections for RCP4.5 are used for 2006–2010. The modelled thermal expansion and glacier contributions are computed from the CMIP5 results, using the model of Marzeion et al. (2012a) for glaciers. The land water contribution is due to anthropogenic intervention only, not including climate-related fluctuations.

Notes:a Data for all glaciers extend to 2009, not 2010.b This contribution is not included in the total because glaciers in Greenland are included in the observational assessment of the Greenland ice sheet.c Observed GMSL rise – modelled thermal expansion – modelled glaciers – observed land water storage.

Source 1901–1990 1971–2010 1993–2010Observed contributions to global mean sea level (GMSL) rise

Thermal expansion – 0.8 [0.5 to 1.1] 1.1 [0.8 to 1.4]

Glaciers except in Greenland and Antarcticaa 0.54 [0.47 to 0.61] 0.62 [0.25 to 0.99] 0.76 [0.39 to 1.13]

Glaciers in Greenlanda 0.15 [0.10 to 0.19] 0.06 [0.03 to 0.09] 0.10 [0.07 to 0.13]b

Greenland ice sheet – – 0.33 [0.25 to 0.41]

Antarctic ice sheet – – 0.27 [0.16 to 0.38]

Land water storage –0.11 [–0.16 to –0.06] 0.12 [0.03 to 0.22] 0.38 [0.26 to 0.49]

Total of contributions – – 2.8 [2.3 to 3.4]

Observed GMSL rise 1.5 [1.3 to 1.7] 2.0 [1.7 to 2.3] 3.2 [2.8 to 3.6]

Modelled contributions to GMSL rise

Thermal expansion 0.37 [0.06 to 0.67] 0.96 [0.51 to 1.41] 1.49 [0.97 to 2.02]

Glaciers except in Greenland and Antarctica 0.63 [0.37 to 0.89] 0.62 [0.41 to 0.84] 0.78 [0.43 to 1.13]

Glaciers in Greenland 0.07 [–0.02 to 0.16] 0.10 [0.05 to 0.15] 0.14 [0.06 to 0.23]

Total including land water storage 1.0 [0.5 to 1.4] 1.8 [1.3 to 2.3] 2.8 [2.1 to 3.5]

Residualc 0.5 [0.1 to 1.0] 0.2 [–0.4 to 0.8] 0.4 [–0.4 to 1.2]

Last two decades: first closure of the budget (IPCC, AR5)

Future sea level rise

1054

Chapter 12 Long-term Climate Change: Projections, Commitments and Irreversibility

12

ACCMIP projected forcing at 2030 (for RCP8.5) and 2100 (all RCPs) is systematically higher than corresponding CMIP5 ERF, although with some overlap between 1-σ ranges. CMIP5 and ACCMIP comprise dif-ferent sets of models and they are related in many but not all cases (Section 8.2.2). Confining analysis to a subset of closely related models also gives higher forcing estimates from ACCMIP compared to CMIP5 so the discrepancy in multi-model ensemble mean forcings appears unrelated to the different model samples associated with the two methods of estimation. The discrepancy is thought to originate mostly from differences in the underlying methodologies used to estimate RF, but is not yet well understood (see also Section 8.5.3).

There is high confidence in projections from ACCMIP models (Shindell et al., 2013b) based on the GISS-E2 CMIP5 simulations (Shindell et al., 2013a) and an earlier study with a version of the HadGEM2-ES model related to that used in CMIP5 (Bellouin et al., 2011), consistent with understanding of the processes controlling nitrate formation (Adams et al., 2001), that nitrate aerosols (which provide a negative forcing) will increase substantially over the 21st century under the RCPs (Sec-tion 8.5.3, Figure 8.20). The magnitude of total aerosol-related forcing (also negative in sign) will therefore tend to be underestimated in the CMIP5 multi-model mean ERF, as nitrate aerosol has been omitted as a forcing from almost all CMIP5 models.

Natural RF variations are, by their nature, difficult to project reliably (see Section 8.4). There is very high confidence that Industrial Era nat-ural forcing has been a small fraction of the (positive) anthropogenic forcing except for brief periods following large volcanic eruptions (Sec-tions 8.5.1 and 8.5.2). Based on that assessment and the assumption that variability in natural forcing remains of a similar magnitude and character to that over the Industrial Era, total anthropogenic forcing relative to pre-industrial, for any of the RCP scenarios through the 21st century, is very likely to be greater in magnitude than changes in natu-ral (solar plus volcanic) forcing on decadal time scales.

In summary, global mean forcing projections derived from climate models exhibit a substantial range for the given RCP scenarios in con-centration-driven experiments, contributing to the projected global mean temperature range (Section 12.4.1). Forcings derived from ACCMIP models for 2100 are systematically higher than those estimat-ed from CMIP5 models for reasons that are not fully understood but are partly due to methodological differences. The multi-model mean estimate of combined anthropogenic plus natural forcing from CMIP5 is consistent with indicative RCP forcing values at 2100 to within 0.2 to 0.4 W m–2.

12.4 Projected Climate Change over the 21st Century

12.4.1 Time-Evolving Global Quantities

12.4.1.1 Projected Changes in Global Mean Temperature and Precipitation

A consistent and robust feature across climate models is a continua-tion of global warming in the 21st century for all the RCP scenarios

(Figure 12.5 showing changes in concentration-driven model simu-lations). Temperature increases are almost the same for all the RCP scenarios during the first two decades after 2005 (see Figure 11.25). At longer time scales, the warming rate begins to depend more on the specified GHG concentration pathway, being highest (>0.3°C per decade) in the highest RCP8.5 and significantly lower in RCP2.6, par-ticularly after about 2050 when global surface temperature response stabilizes (and declines thereafter). The dependence of global temper-ature rise on GHG forcing at longer time scales has been confirmed by several studies (Meehl et al., 2007b). In the CMIP5 ensemble mean, global warming under RCP2.6 stays below 2°C above 1850-1900 levels throughout the 21st century, clearly demonstrating the potential of mitigation policies (note that to translate the anomalies in Figure 12.5 into anomalies with respect to that period, an assumed 0.61°C of observed warming since 1850–1900, as discussed in Section 2.4.3, should be added). This is in agreement with previous studies of aggres-sive mitigation scenarios (Johns et al., 2011; Meehl et al., 2012). Note, however, that some individual ensemble members do show warming exceeding 2°C above 1850-1900 (see Table 12.3). As for the other pathways, global warming exceeds 2°C within the 21st century under RCP4.5, RCP6.0 and RCP8.5, in qualitative agreement with previous studies using the SRES A1B and A2 scenarios (Joshi et al., 2011). Global mean temperature increase exceeds 4°C under RCP8.5 by 2100. The CMIP5 concentration-driven global temperature projections are broad-ly similar to CMIP3 SRES scenarios discussed in AR4 (Meehl et al., 2007b) and Section 12.4.9, although the overall range of the former is larger primarily because of the low-emission mitigation pathway RCP2.6 (Knutti and Sedláček, 2013).

The multi-model global mean temperature changes under different RCPs are summarized in Table 12.2. The relationship between cumu-lative anthropogenic carbon emissions and global temperature is assessed in Section 12.5 and only concentration-driven models are

42 models

39254232

121712

Figure 12.5 | Time series of global annual mean surface air temperature anomalies (relative to 1986–2005) from CMIP5 concentration-driven experiments. Projections are shown for each RCP for the multi-model mean (solid lines) and the 5 to 95% range (±1.64 standard deviation) across the distribution of individual models (shading). Dis-continuities at 2100 are due to different numbers of models performing the exten-sion runs beyond the 21st century and have no physical meaning. Only one ensemble member is used from each model and numbers in the figure indicate the number of different models contributing to the different time periods. No ranges are given for the RCP6.0 projections beyond 2100 as only two models are available.

Scenarios are considered (IPCC, AR5)

Future sea level rise

1180


13

climate change is correlated among all the scenario-dependent contri-butions, while the methodological uncertainties are treated as inde-pendent (see also Supplementary Material).

The sum of the projected contributions gives the likely range for future GMSL rise. The median projections for GMSL in all scenarios lie within a range of 0.05 m until the middle of the century (Figure 13.11), because the divergence of the climate projections has a delayed effect owing to the time-integrating characteristic of sea level. By the late 21st century (over an interval of 95 years, between the 20-year mean of 2081–2100 and the 20-year mean of 1986–2005), they have a spread of about 0.25 m, with RCP2.6 giving the least amount of rise (0.40 [0.26 to 0.55] m) (likely range) and RCP8.5 giving the most (0.63 [0.45 to 0.82] m). RCP4.5 and RCP6.0 are very similar at the end of the century (0.47 [0.32 to 0.63] m and 0.48 [0.33 to 0.63]] m respectively), but RCP4.5 has a greater rate of rise earlier in the century than RCP6.0 (Figure 13.10 and Table 13.5). At 2100, the likely ranges are 0.44 [0.28–0.61] m (RCP2.6), 0.53 [0.36–0.71] m (RCP4.5), 0.55 [0.38–0.73] m (RCP6.0), and 0.74 [0.52–0.98] m (RCP8.5).

In all scenarios, the rate of rise at the start of the RCP projections (2007–2013) is about 3.7 mm yr–1, slightly above the observational range of 3.2 [2.8 to 3.6] mm yr–1 for 1993–2010, because the modelled contributions for recent years, although consistent with observations for 1993–2010 (Section 13.3), are all in the upper part of the observa-

A1B RCP2.6 RCP4.5 RCP6.0 RCP8.5

0.0

0.2

0.4

0.6

0.8

1.0

1.2Gl

obal

mea

n se

a lev

el ris

e (m

)

2081-2100 relative to 1986-2005SumThermal expansionGlaciersGreenland ice sheet (including dynamics)Antarctic ice sheet (including dynamics)Land water storageGreenland ice-sheet rapid dynamicsAntarctic ice-sheet rapid dynamics

Figure 13.10 | Projections from process-based models with likely ranges and median values for global mean sea level rise and its contributions in 2081–2100 relative to 1986–2005 for the four RCP scenarios and scenario SRES A1B used in the AR4. The contributions from ice sheets include the contributions from ice-sheet rapid dynamical change, which are also shown separately. The contributions from ice-sheet rapid dynamical change and anthropogenic land water storage are treated as having uniform probability distributions, and as independent of scenario (except that a higher rate of change is used for Greenland ice-sheet outflow under RCP8.5). This treatment does not imply that the contributions concerned will not depend on the scenario followed, only that the current state of knowledge does not permit a quantitative assessment of the dependence. See discussion in Sec-tions 13.5.1 and 13.5.3 and Supplementary Material for methods. Only the collapse of the marine-based sectors of the Antarctic ice sheet, if initiated, could cause global mean sea level (GMSL) to rise substantially above the likely range during the 21st century. This potential additional contribution cannot be precisely quantified but there is medium confidence that it would not exceed several tenths of a meter of sea level rise.

tional ranges, perhaps related to the simulated rate of climatic warm-ing being greater than has been observed (Box 9.2). In the projections, the rate of rise initially increases. In RCP2.6 it becomes roughly con-stant (central projection 4.5 mm yr–1) before the middle of the century, and subsequently declines slightly. The rate of rise becomes roughly constant in RCP4.5 and RCP6.0 by the end of the century, whereas acceleration continues throughout the century in RCP8.5, reaching 11 [8 to 16] mm yr–1 in 2081–2100.

In all scenarios, thermal expansion is the largest contribution, account-ing for about 30 to 55% of the projections. Glaciers are the next largest, accounting for 15-35% of the projections. By 2100, 15 to 55% of the present volume of glaciers outside Antarctica is projected to be elim-inated under RCP2.6, and 35 to 85% under RCP8.5 (Table 13.SM.2). SMB change on the Greenland ice sheet makes a positive contribu-tion, whereas SMB change in Antarctica gives a negative contribution (Sections 13.4.3.1 and 13.4.4.1). The positive contribution due to rapid dynamical changes that result in increased ice outflow from both ice sheets together has a likely range of 0.03 to 0.20 m in RCP8.5 and 0.03 to 0.19 m in the other RCPs. There is a relatively small positive contri-bution from human intervention in land water storage, predominantly due to increasing extraction of groundwater.

(From IPCC, AR5)

Future sea level rise: KNMI 2014

creating future climate projections on the basis of variousrepresentative concentration pathways (RCPs) (Meinshausenet al 2011) as is done for example in AR5, we use globaltemperature pathways for discriminating between the differ-ent scenarios. There are two main reasons for using dTglobinstead of RCPs. The first is the fact that sea level change, aswell as many atmospheric climate variables, scales well withglobal temperature change (van den Hurk et al 2014b). Sec-ondly, it is thought that dTglob is a variable more easilyunderstood by the general public than the concept of RCPs.

Two distinct temperature pathways are chosen, corre-sponding to a moderate warming scenario (G) and a warmerscenario (W). Table 1 lists the values for three key years,relative to the reference period 1986–2005. For intermediateyears, dTglob is linearly interpolated. The scenarios aretherefore not very extreme in their choice of the dTglob; theG-scenario overlaps with RCP4.5 and W with RCP8.5.

1.2. Processes included in the KNMI’14 scenarios

Many different physical processes contribute to global andregional sea-level change (Church et al 2010, Dangendorfet al 2014a). Primary contributors to present-day and futuresea level change are the expansion of the ocean due towarming and the reduction of the amount of water stored onland, mostly in the form of ice and snow (Church andWhite 2011, Church et al 2011, 2013). All contributingprocesses manifest themselves as a superposition of a slowlyvarying (trend)-signal and faster fluctuating components.These fluctuations imply that the exact future state cannot bepredicted at long lead times. However, also the slowlyvarying (trend) signal is subject to considerable uncertainty,both because different models produce different results(‘ensemble spread’) and because of model inaccuracies,incomplete physics and methodological uncertainties (Hu andDeser 2013). The scenarios developed here aim to describethe slowly varying (climate) component. We now detail theprocesses included in KNMI’14 (Mathematical details insupplementary material appendix B).

1.2.1. Ocean expansion and circulation changes. The oceanmay expand as a result of changes in ocean temperature andsalinity. Changes in the ocean circulation (i.e., dynamics) alsoinfluence the sea level regionally. The ocean has a large heatcapacity, and during warmer periods enormous amounts ofheat are absorbed. These are advected horizontally as well asdownward into the deep ocean, to be gradually releasedduring cooler climatic periods. For this reason the oceansstore a considerable amount of heat of the past climate, and its

response to global warming is generally nonlinear. The oceancirculation changes, such as the response of the meridionaloverturning circulation and the ocean gyres, introduce morecomplexity to the problem, related to density changesresulting from temperature and salinity changes.

1.2.2. Glaciers and small ice caps. This term considerschanges in surface mass balance (SMB) and dynamics of theglaciers and small ice caps. Included here are all glaciersworldwide, including those on Greenland and Antarctica thatare not connected to the main ice sheets. Glaciers respondmore rapidly to climate change compared to the ice sheets, sotheir short-term influence on the global and regional sea-levelchange is expected to be considerable. The total possiblecontribution from all present-day glaciers however, is modest(likely between 31 and 53 cm) compared to the large icesheets (Arendt et al 2012, Huss and Farinotti 2012, Marzeionet al 2012, Grinsted 2013, Radic et al 2014). Importantparameters influencing the mass balance of a local glacier arethe regional climate (temperature and precipitation), as well asits orientation and altitude. Despite such subtleties, the totalcontribution from all glaciers together can be parameterizedapproximately in terms of global mean temperature change(van de Wal and Wild 2001, Slangen and van de Wal 2011).Regionalization is subsequently achieved by a process calleddynamic fingerprinting, explained in section 2.3.

1.2.3. Large ice sheets. This term considers changes of thelargest ice sheets on the planet, those of Greenland (GIS) andAntarctica (AIS). AIS contains by far the most ice. If it wouldmelt completely, it would raise global mean sea level by morethan 60 m. The GIS is much smaller and holds an equivalentof ∼6 m of global sea level. Ice sheets will respond to climatechange in two different ways. First there will be changes intheir SMB, which is the sum of snowfall, summer melt ofsnow and ice resulting in liquid run-off from the ice sheet.Regional climate models forced by atmosphere-ocean globalclimate models (AOGCMs) are used to find suitableparameterizations of SMB in terms of dTglob (Fettweiset al 2013). Secondly, the ice sheets may also show a rapiddynamical response (DYN). The DYN-term describespossible changes in iceberg calving and basal melt oftidewater glaciers by warmer ocean water. Although thelatter process does not contribute to sea-level rise per se, it isassociated with ice flow from grounded glaciers to the floatingice shelves which does contribute to accelerated sea-level rise.Other mechanisms, such as marine ice sheet instability(Joughin and Alley 2011) may play a role. However, todate there is no consensus on the magnitude and time scale ofthe changes (Truffer and Fahnestock 2007, Vaughan andArthern 2007, Little et al 2013), which translates insubstantial uncertainty bands (Horton et al 2014, de Vriesand van de Wal 2014).

1.2.4. Land water change. This term collects changes in theamount of water stored in the form of lakes and rivers,wetlands, as well as the seasonal snowpack at high altitudes

Table 1. Steering values of global-mean temperature change (withrespect to 1986–2005 average) that are used in the KNMI’14 sea-level change scenarios.

Scenario/year 2050 2085 2100

G-scenario +1.0 °C +1.5 °C +1.6 °CW-scenario +2.0 °C +3.5 °C +4.0 °C

2

Environ. Res. Lett. 9 (2014) 115007 H de Vries et al

Unsurprisingly, sea levels along the Dutch coast increasein both scenarios, with the W (warm) scenario giving+45–80 cm in 2085 for dTglob = +3.5 K and the moderatescenario (G, +1.5 K in 2085) +25–60 cm. We state thesenumbers rounded to 5 cm precision, to emphasise that, despiteall efforts in the computation, even the lower and upper limitare intrinsically uncertain. The nonlinear shape of the curvesreveals that the rate of sea-level change increases in bothscenarios. Needless to say is that sea levels will keep risinglong after possible future stabilization of temperature rise.

The largest contribution is from the ocean expansion.Second largest are glaciers and small ice-caps. Especially nearthe end of the 21st century, the likely range strongly widensin both scenarios. This widening is partly caused by therelatively large model-uncertainty in the ocean component,but also by the large possible range of the rapid dynamiccomponent of the Antarctic ice sheet (AIS-dyn), as can beseen from the process contributions in the right margin offigure 3. This component is estimated to be strongly asym-metric around the median (symbol), reflecting the largeuncertainty in the upper-bound. The only negative contribu-tion to global sea-level relates to the SMB changes of Ant-arctica. Precipitation increase and hence increasingaccumulation is the main driver behind this. Changes in GIShave relatively little impact near the Dutch coast.

The grey + symbols show the 5 year running averagedobservations. Clearly there is considerable natural variabilityaround the 30 year running mean. The start of the 21st centuryis characterized by values lower than the average trend, whichis reflected both in the tide-gauge and altimeter record, forreasons to be investigated further. The shaded scenario bandsinclude a rough estimate of the internal variability and are

also shown as vertical dashes. This term is computed heresimply as the variance of the 5 year running average devia-tions from the 30 year running mean, and is seen to fit wellwith the typical fluctuations in the observational record.

Internal variability of sea level occurs on a variety of timescales, from short to very long (Church et al 2010, Dangen-dorf et al 2014b). This makes it more difficult to detectpossible accelerations in regional sea-level rise. Studying5 year running deviations is just one way to discuss qualita-tively the role of internal variability. The shortest time scalesare removed by considering 5 year running means. By sub-tracting these from the 30 year running means, we remove thetrend, as well as most of the variability on time-scales beyond30 year (these should be visible in the black line, but aresmall). The scenario bands have approximately enoughspread to cover variation from natural fluctuations. If, on theother hand, we would have constructed the scenarios byaveraging over a larger basin (for example the North Atlantic,or even the entire globe), the spread would have been reducedconsiderably. Such scenarios would not be suited to describethe natural fluctuations in the North Sea.

3.2. Low-end and high-end scenarios

The right margin in figure 3 shows that all components areaccompanied by considerable uncertainties. Therefore, it isinstructive to compare these for total rises near the lower andupper limit of the estimated ranges. Figure 4 shows such acomparison. It displays the process contributions in 2085 for ahigh W-scenario (values within 2.5 cm of the upper limit ofthe W-scenario) and a low G-scenario (values within 2.5 cmof the lower limit). The low G-scenario has markedly smallercontributions from all components, but most noticeably fromthe steric/dynamic (ocean) component and from AIS-dyn,reflecting the large uncertainty in the latter given it is para-meterised independent of dTglob (supplementary mate-rial B.3).

Figure 3. Scenarios for sea-level rise along the North Sea coast. Thevertical axis denotes 30 year running mean sea-level change in cm,relative to 1986–2005. For years before 2000 and beyond 2085, themean was taken over an increasingly smaller window and is drawnin a different shading. Estimates of natural variability at 5 year time-scale (see text for details) is included in the range and shown asvertical dashes. The black line denotes the 30 year running meanthrough tide-gauge observations along the Dutch coast, grey +symbols the 5 year running means (see text for details). The green xsymbols show 5 year running means from satellite altimetry over theNorth Sea, with respect to 1993–1997 mean. The right margin showsthe ranges from the different processes in 2100.

Figure 4. Contributions (cm) to the high W-scenario (left) and thelow G-scenario (right) in 2085.

5

Environ. Res. Lett. 9 (2014) 115007 H de Vries et al

de Vries et al. 2014

Future sea level rise: Multi-millennial “commitment”

1190


13

Figure 13.14 | (Left column) Multi-millennial sea level commitment per degree Celsius of warming as obtained from physical model simulations of (a) ocean warming, (b) mountain glaciers and (c) the Greenland and (d) the Antarctic ice sheets. (e) The corresponding total sea level commitment, compared to paleo estimates from past warm periods (PI = pre-industrial, LIG = last interglacial period, M11 = Marine Isotope Stage 11, Plio = Mid-Pliocene). Temperatures are relative to pre-industrial. Dashed lines provide linear approximations in (d) and (e) with constant slopes of 1.2, 1.8 and 2.3 m °C–1. Shading as well as the vertical line represents the uncertainty range as detailed in the text. (Right column) 2000-year-sea level commitment. The difference in total sea level commitment (j) compared to the fully equilibrated situation (e) arises from the Greenland ice sheet which equilibrates on tens of thousands of years. After 2000 years one finds a nonlinear dependence on the temperature increase (h) consistent with coupled climate–ice sheet simulations by Huybrechts et al. (2011) (black dot). The total sea level commitment after 2000 years is quasi-linear with a slope of 2.3 m °C–1.

°°

0.42 m °C-1 0.42 m °C-1

1.2 m °C-1

1.8 m °C-1

2.3 m °C-1 2.3 m °C-1

1.2 m °C-1

quantitatively consistent with previous estimates on a millennial time scale (Huybrechts et al., 2011; Goelzer et al., 2012). The sea level contri-bution of the Greenland ice sheet after 2000 years of integration at 560 ppm was plotted against the average Greenland temperature divided by the standard polar amplification of 1.5 between global mean and

Greenland mean temperature increase (Gregory and Huybrechts, 2006, black dot in Figure 13.14h). Taken together, these results imply that a sea level rise of 1 to 3 m °C–1 is expected if the warming is sustained for several millennia (low confidence) (Figure 13.14e, 13.14j).

1190


13


°°

0.42 m °C-1 0.42 m °C-1

1.2 m °C-1

1.8 m °C-1

2.3 m °C-1 2.3 m °C-1

1.2 m °C-1



1190


13


°°

0.42 m °C-1 0.42 m °C-1

1.2 m °C-1

1.8 m °C-1

2.3 m °C-1 2.3 m °C-1

1.2 m °C-1



Thermalexpansion

Mountainglaciers

Greenlandice sheet

Antarcticice sheet

IPCC, AR5

• Sea level and sea level rise are not globally uniform

• Main contributors are expected to be thermal expansion and glaciers

• Sea level rise continues for millennia

• The uncertainty from the Antarctic ice sheet is very big, research is moving fast on this subject

Main points

Questions?

References:Church, J.A., P.U. Clark, A. Cazenave, J.M. Gregory, S. Jevrejeva, A. Levermann, M.A. Merrifield, G.A. Milne, R.S. Nerem, P.D. Nunn, A.J. Payne, W.T. Pfeffer, D. Stammer and A.S. Unnikrishnan, 2013: Sea Level Change. In: Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA

Kopp, R. E., Kemp, A. C., Bittermann, K., Horton, B. P., Donnelly, J. P., Gehrels, W. R., … Rahmstorf, S. (2016). Temperature-driven global sea-level variability in the Common Era, 1–8. http://doi.org/10.1073/pnas.1517056113

Vries, H. De, Katsman, C., & Drijfhout, S. (2014). Constructing scenarios of regional sea level change using global temperature pathways. Environmental Research Letters, 9(11), 115007. http://doi.org/10.1088/1748-9326/9/11/115007

http://doi.org/10.1088/1748-9326/9/11/115007

drivers and uncertainties in past and future sea level changes

Technology