Sea level projections to AD2500 with a new generation of climate change scenarios

https://doi.org/10.1016/j.gloplacha.2011.09.006Get rights and content

Abstract

Sea level rise over the coming centuries is perhaps the most damaging side of rising temperature (Anthoff et al., 2009). The economic costs and social consequences of coastal flooding and forced migration will probably be one of the dominant impacts of global warming (Sugiyama et al., 2008). To date, however, few studies (Nicholls et al., 2008; Anthoff et al., 2009) on infrastructure and socio-economic planning include provision for multi-century and multi-metre rises in mean sea level. Here we use a physically plausible sea level model constrained by observations, and forced with four new Representative Concentration Pathways (RCP) radiative forcing scenarios (Moss et al., 2010) to project median sea level rises of 0.57 for the lowest forcing and 1.10 m for the highest forcing by 2100 which rise to 1.84 and 5.49 m respectively by 2500. Sea level will continue to rise for several centuries even after stabilisation of radiative forcing with most of the rise after 2100 due to the long response time of sea level. The rate of sea level rise would be positive for centuries, requiring 200–400 years to drop to the 1.8 mm/yr 20th century average, except for the RCP3PD which would rely on geoengineering.

Highlights

► We estimate sea level rise of 0.57  1.10 m by 2100 with four new RCP scenarios. ► Sea level will continue to rise for several centuries reaching 1.84  5.49 m by 2500. ► Due to long response time most rise is expected after stabilization of forcing. ► 200–400 years will require dropping the rate to the 1.8 mm/yr- 20th century average.

Introduction

The conventional approach to estimate the sea level rise has been to model the major components: ocean thermal expansion, melting from ice sheets and glaciers and terrestrial storage (Meehl et al., 2007, Pardaens et al., 2011). However, measurements of all these components are fraught with difficulty; hence models of their behaviour rely on significant extrapolation from a small observational dataset (Meehl et al., 2007). Conceptually the best way to estimate future rises in sea level would be physical models of all the water storage reservoirs on the planet and how they behave under a changing climate. This task is complex and the subject to intense research efforts, and at present the behaviour of the large ice sheets is limited by physical understanding of dynamics and to a lesser degree by lack of computing power and geophysical observations (Durand et al., 2009, Goldberg et al., 2009). Physically based climate models simulate the thermal expansion component and surface mass balance of Greenland and Antarctic ice sheets while the numerous smaller glaciers budget is parameterized (Meehl et al., 2007, Pardaens et al., 2011). At present, there are very few estimates of dynamical ice sheet loss which are not simply statistical extrapolations (Katsman et al., 2011) or expert opinion (Pfeffer et al., 2008) and all models lack a proper representation of key processes such as calving (Graversen et al., 2010, Price et al., 2011). The best estimates from these modelled components amount to only 1/3 of observed 20th century sea level rise (Gregory et al., 2006), or about 2/3 of that for the past 50 years (Hegerl et al., 2007).

Another approach is to simulate observed sea level using physically plausible models (von Storch et al., 2008) of reduced complexity that respond to histories of global temperature (Rahmstorf, 2007a, Grinsted et al., 2010) or radiative forcing (Jevrejeva et al., 2009, Jevrejeva et al., 2010). Sea level rise in these models is caused by changes in global ice volume and global ocean heat content as a response to changes in global temperature or radiative forcing with a characteristic response time. This characteristic response time is assumed to be infinite (Rahmstorf, 2007a) or estimated by the model as a probability density function spanning a wide range of time scales (Jevrejeva et al., 2009, Grinsted et al., 2010). All semi-empirical models, by construction, simulate recent past and present sea level rise very well. In addition, the latest semi-empirical models (Grinsted et al., 2010, Jevrejeva et al., 2010) reproduce climate system modelled sea level behaviour at scales from centennial to multi-annual, e.g. the impact of volcanic eruptions on sea level simulated by semi-empirical models is in excellent agreement with that given by a coupled climate model (Moore et al., 2010). Semi-empirical simulation of 1993–2006 sea level rate is 3–4 mm/yr (Rahmstorf et al., 2007, Grinsted et al., 2010), which is very similar to the rate of 3.3 mm/yr calculated from satellite altimetry observations; in contrast process based models estimate of the rate is 1.9 mm/yr (Church et al., 2001). Vermeer and Rahmstorf (2009) have concluded that there is a good agreement between thermal expansion simulated by their semi-empirical method and two coupled climate models for the past 1000 years. Conversely, the analysis by von Storch et al. (2008) of their ECHO-G millennial run to simulate one of the component of sea level, thermal expansion of the ocean (also used by Vermeer and Rahmstorf, 2009), suggests that it is difficult to make an estimate of thermal expansion forced by global temperature on centennial timescales. However, they find that sea level forced by radiative forcing (as used in this study) is significantly better on all timescales than forcing with temperatures. Projections by semi-empirical models are based on the assumption that sea level in the future will respond as a linear system, so that future response is analogous to the past. This may not hold in the future if potentially non-linear physical processes come into play (e.g. ice-sheet dynamic feedbacks). Another limitation of semi-empirical models is the lack of spatial variability, hence regional sea level rise prediction is beyond the scope of this paper. There has also been some discussion of the statistical procedures used in some semi-empirical studies (Holgate et al., 2007, Rahmstorf, 2007a, Rahmstorf, 2007b, Schmith et al., 2007, Vermeer and Rahmstorf, 2009, Taboada and Anadón, 2010, Vermeer and Rahmstorf, 2010), however the models used here (Jevrejeva et al., 2009, Grinsted et al., 2010) have not attracted statistical criticism.

In this study, a semi-empirical model (Jevrejeva et al., 2009, Grinsted et al., 2010) is constrained by the 300 years of global sea level records from tide gauges (Jevrejeva et al., 2008) and driven by various radiative forcing time series (solar, volcanic, greenhouse gases and aerosols) over the past 1000 years (Crowley et al., 2003, Goosse et al., 2005, Tett et al., 2007), shown in Table 1. We assume that global sea level is an integrated response of the entire climate system to the changes in radiative forcing that reflects alteration in the dynamics and thermodynamics of the atmosphere, ocean and cryosphere. The use of radiative forcing removes the substantial uncertainties in the relationship between forcing and temperature response and subsequent sea level response and implicitly includes the effects of feedback mechanisms.

In this study we do not include any changes in sea level associated with non-climate related components such as contribution from groundwater mining, urbanization and water storage in reservoirs. This is in contrast to the approach by Vermeer and Rahmstorf (2009) where the contribution from reservoir construction of − 0.55 mm/yr (Chao et al., 2008) was taken into account, but not the potentially cancelling effects of groundwater mining (0.55–0.64 mm/yr; Huntington, 2008) and urbanization (Cazenave and Nerem, 2004). Hence we follow the suggestion of Lettenmaier and Milly (2009) that land, overall, contributes essentially nothing to sea-level rise today. This is consistent with closure of the sea level budget using only climate related components since 1955 (Moore et al., 2011).

Here we use suites of observationally tuned models driven by the four new RCP radiative forcing scenarios (Fig. 1) to project future sea level by AD2100 and to explore the range of uncertainties in sea level rise by AD2500 associated with changes in radiative forcings.

Section snippets

Description of the new Representative Concentration Pathways scenarios

Interest in modelling climate system components, such as global sea level, the oceans and the ice sheets, has created the demand for emission scenarios to extend well beyond the end of 21st century. The new Representative Concentration Pathways (RCPs) scenarios (Moss et al., 2010) of future radiative forcings have been developed since the Fourth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC), providing a framework for modelling in climate change research up to 2500.

Sea level projections by 2100

The sea level responses to radiative forcing from four new RCP scenarios by the end of the 21st century are presented in Fig. 3. Sea level is insensitive to RCP forcing until 2050 with a range of about 0.32–0.38 m above the 1980–2000 reference level. However, by the end of the 21st century there are clear consequences depending on which scenario is followed, with sea level rise ranging from 0.57 to 1.10 m by 2100 (with lower and upper 5–95% confidence limits of 0.36 m to 1.65 m, Table 3), largely

Discussion

It is unclear how the climate system will respond to the changes in radiative forcing envisaged by the new scenarios, since long-term feedbacks will affect climate sensitivity, greatly increasing uncertainty in projections of long-term climate change. The main uncertainty for the sea level projections is the response of the ice sheets in Greenland and Antarctica to hundreds of years of warmer temperatures, which is the focus of several ice sheet dynamical modelling initiatives (e.g. Timmermann

Conclusion

The sea level rise due to ocean thermal expansion and melting of glaciers and ice sheets has a characteristic timescale of 100–200 years (Grinsted et al., 2010, Jevrejeva et al., 2010). This is comparable to the residence time of CO2 in the atmosphere and hence the radiative forcing timescale. Thus sea level and anthropogenic climate forcing are linked by two multi-centennial time scales. Sea level rise of 0.57–1.10 m by 2100 has been estimated as medians from 2,000,000 runs by our model.

Acknowledgements

We are grateful to the anonymous reviewers for constructive criticisms which helped to improve an earlier draft of this manuscript. Financial support from: NSFC No. 41076125 and China's National Key Science Program for Global Change Research (No. 2010C8950504) and NERC consortium “Using Inter-glacials to assess future sea level scenarios” (NE/1008365/1).

References (57)

  • H. Goosse et al.

    Internal and forced climate variability during the last millennium: a model-data comparison using ensemble simulations

    Quaternary Science Reviews

    (2005)
  • K. Mosegaard et al.

    Probabilistic approach to inverse problems international

  • R.B. Alley et al.

    Ice-sheet and sea level changes

    Science

    (2005)
  • D. Anthoff et al.

    The Economic Impact of Substantial Sea-Level Rise

  • D.B. Bahr et al.

    Sea‐level rise from glaciers and ice caps: a lower bound

    Geophysical Research Letters

    (2009)
  • S.M. Barbosa

    Quantile trends in Baltic sea-level

    Geophysical Research Letters

    (2008)
  • P. Blanchon et al.

    Rapid sea-level rise and reef back-stepping at the close of the last interglacial highstand

    Nature

    (2009)
  • A.E. Carlson et al.

    Rapid early Holocene deglaciation of the Laurentide Ice Sheet

    Nature Geoscience

    (2008)
  • A. Cazenave et al.

    Present-day sea level change: observations and causes

    Reviews of Geophysics

    (2004)
  • B.F. Chao et al.

    Impact of artificial reservoir water impoundment on global sea level

    Science

    (2008)
  • J.A. Church et al.

    Changes in sea level

  • T.J. Crowley et al.

    Modeling ocean heat content changes during the last millennium

    Geophysical Research Letters

    (2003)
  • G. Durand et al.

    Full Stokes modeling of marine ice sheets: influence of the grid size

    Annals of Glaciology

    (2009)
  • M. Goes et al.

    The economics (or lack thereof) of aerosol geoengineering

    Climatic Change

    (2011)
  • D. Goldberg et al.

    Grounding line movement and ice shelf buttressing in marine ice sheets

    Journal of Geophysical Research Earth Surface

    (2009)
  • R.G. Graversen et al.

    Greenland's contribution to global sea-level rise by the end of the 21st century

    Climate Dynamics

    (2010)
  • J.M. Gregory

    Vertical heat transports in the ocean and their effect on time-dependent climate change

    Climate Dynamics

    (2000)
  • J.M. Gregory et al.

    Simulated global-mean sea-level changes over the last half-millennium

    Journal of Climate

    (2006)
  • A. Grinsted et al.

    Reconstructing sea level from paleo and projected temperatures 200 to 2100AD

    Climate Dynamics

    (2010)
  • G.C. Hegerl et al.

    Understanding and attributing climate change, Climate Change 2007: the physical science basis

  • S. Holgate et al.

    Comment on “A semi-empirical approach to projecting future sea-level rise”

    Science

    (2007)
  • T.G. Huntington

    Can we dismiss the effect of changes in land-based water storage on sea-level rise?

    Hydrological Processes

    (2008)
  • S. Jevrejeva et al.

    Recent global sea level acceleration started over 200 years ago?

    Geophysical Research Letters

    (2008)
  • S. Jevrejeva et al.

    Anthropogenic forcing dominates sea level rise since 1850

    Geophysical Research Letters

    (2009)
  • S. Jevrejeva et al.

    How will sea level respond to changes in natural and anthropogenic forcings by 2100?

    Geophysical Research Letters

    (2010)
  • M.M. Johansson et al.

    Scenarios for sea level on the Finnish coast

    Boreal Environment Research

    (2004)
  • C.A. Katsman et al.

    Exploring High-end Scenarios for Local Sea Level Rise to Develop Flood Protection Strategies for a Low-lying Delta — The Netherlands as an Example, Climatic Change

    (2011)
  • R.E. Kopp et al.

    Probabilistic assessment of sea level during the last interglacial stage

    Nature

    (2009)
  • Cited by (197)

    • Pathways for adapting tourism to climate change in an urban destination – Evidences based on thermal conditions for the Porto Metropolitan Area (Portugal)

      2022, Journal of Environmental Management
      Citation Excerpt :

      According to several regional, national and international reports, an average sea level rise of between 0.17m and 0.38m is expected by 2050, and between 0.26 and 0.82 by the end of the 21st century (global projections) (Câmara Municipal do Porto, 2016; IPCC, 2014, 2021). Other authors indicate that this increase could reach 1.10 m in 2100 (global projections – Jevrejeva et al., 2012). In any case, the rise in mean sea level may cause more serious impacts, when combined with the sea level rise associated with storms (storm surge – Johnston et al., 2021; Karim and Mimura, 2008).

    View all citing articles on Scopus
    View full text