Understanding 20th Century Sea-level Rise and Projections for the Future
Dr John Church
Polar Waters Subprogram Leader, Antarctic CRC and CSIRO Marine Research
INTRODUCTION
Sea-level rise is one of the most well known and more emotive impacts of anthropogenic climate change. Today, storm surges can, and do, have a serious impact on coastal communities. The most severe storm surges have resulted in the loss of hundreds of thousands of people in a single event. Questions often asked include what will be the impact of sea-level rise, what will it mean for low lying countries, and how will the intensity and frequency of extreme events change?
The most recent report of the Intergovernmental Panel of Climate Change (IPCC, 2001) includes a chapter examining historical and future sea-level rise. The chapter (Church et al, 2001) is an assessment of our current knowledge of the science associated with sea-level change since the last glacial maximum, over the last few hundred years, an assessment of likely change by the end of the 21st century and potential changes on the longer term. An international group of authors was responsible for completing the assessment and steering it through the review process. Two independent Review Editors were charged with the responsibility of ensuring that the chapter was a balanced and thorough assessment of the state of science and that the authors responded appropriately to each and every one of the hundreds of comments received from individual scientists and from Governments around the word. This paper largely draws on the sea level chapter of the IPCC Third Assessment Report (Church et al, 2001).
There have been significant advances is our understanding of sea-level change since the second IPCC assessment (Warrick et al, 1996). These advances are a result of improvements in general circulation models (GCMs) of the climate system, the use of a range of GCMs for addressing sea-level change, improvements in models of the response of glaciers and ice sheets, improved data sets and the use of a broader set of constraints to understand historical change. However, significant uncertainties remain.
Observed Changes in Sea Level
Changes since the Last Glacial Maximum
Over the last several hundred thousand years, sea-level has fluctuated by over 100m as the mass of ice stored on land has changed. At the last glacial maximum, sea level was over 120m below present day levels. The majority of the ice required to lower sea level at this time was stored in the northern hemisphere ice sheets with lesser amounts stored in Antarctica and elsewhere. The most rapid rise in sea level occurred between 15,000 and 6,000 years ago at an average rate of 10 mm/yr, but with peak rates as high as 40 mm/yr. Over the past 6,000 years sea level rise has been much slower. Geological data indicate an average rate of rise over this period of 0.5 mm/yr, and perhaps 0.1 to 0.2 mm/yr over the last 3,000 years.
These large transfers of mass from the ice sheets to the ocean result in vertical land movements (post-glacial rebound) that are ongoing today. In regions distant from the former ice sheets and near the center of former ice sheets the land is rising whereas in some regions surrounding the former ice sheets land is currently falling. Allowance for these vertical land motions are essential if we are to accurately infer sea-level changes from measurements of ‘relative’ sea-level change as measured by coastal tide gauges.
Instrumental Observations of Sea-level Change
Relative sea level (ie the height of the ocean surface relative to the land) is measured by coastal and island tide gauges. Unfortunately for estimating global sea-level rise, the historical distribution of tide gauges is far from optimum. Most long records are from gauges sited along northern hemisphere (particularly European) coastlines with relatively few records at mid-ocean island sites or in the southern hemisphere. The available long records, some dating back to the 16th century, indicate an increase in the rate of sea-level rise of 0.4 to 0.9 mm/yr/century.
To estimate the absolute rate of sea-level rise, it is necessary to correct the tide gauge data sets for vertical land motion resulting from post-glacial rebound and other tectonic motions. These corrections are based on either geological observations or geophysical models. Even during the 20th century, the distributions of tide gauges necessary for measuring global average sea-level rise is far from ideal and estimates of 20th century sea-level rise depend on the selection of tide gauge sites and the method used for correcting for land motions. The IPCC assessment (Church et al, 2001) was that the rate of 20th century sea-level rise was between one and two mm/yr and that no increase in the rate of sea-level rise has been detected from the 20th century data set. In this assessment, we recognised that the spatial distribution of the gauges was potentially an issue biasing various estimates. Model results (Gregory et al, 2001) indicate that the bias introduced by the spatial sub-sampling was small while a recent analysis of estimates of ocean thermal expansion (Cabanes et al, 2001) indicated a larger bias. However, the data set used for the estimates of ocean thermal expansion is itself incomplete and the magnitude of the spatial bias inferred by Cabanes et al (2001) has been questioned (Church, 2001). The model data also indicate that it would be difficult to detect any 20th century increase in the rate of sea-level rise from the available tide gauges in the presence of spatial and temporal variability.
Satellite observations are revolutionising our ability to measure sea-level change. Satellite altimeter missions like TOPEX/POSEIDON give near global observations of sea level. Since its launch in late 1992, TOPEX/POSEIDON data indicates a rate of global-mean sea-level rise of greater than 2 mm/yr. However, the record is as yet to short to infer a recent increase in the rate of sea-level rise. Satellites have relatively short life times but a successful launch of Jason-1 on 7 December, 2001 (and subsequent missions) will ensure an ongoing record of sea-level change. It is necessary to carefully control biases within and between satellite missions using in situ tide gauges. Here again modern technology is assisting with more accurate and stable tide gauges and the ability to measure vertical land motions using the satellite based Global Positioning System.
Understanding 20th Century Changes
The sea-level rise equivalent stored in the various water reservoirs on earth is 3,750m for the ocean, 0.5m for glaciers and ice caps, 7.2m for the Greenland ice sheet, 61m for the Antarctic ice sheet and 12m for ground water (less than 750m depth). Changes in all of these reservoirs need to be considered to understand global sea-level change.
The Ocean
As the ocean warms, the density of sea water decreases and thus even at constant mass the volume of the ocean increases. This ocean thermal expansion is a major contributor to 20th century sea-level rise and is likely to be the largest single contributor to 21st century sea-level rise. Estimates of ocean thermal expansion can be made from observations of ocean temperature changes and knowledge of the equation of state of sea water. (Salinity also affects the ocean density and thus local sea level.)
Unfortunately there are insufficient observations to estimate the rate of warming throughout the ocean over the 20th century. The longest time series of ocean observations are from near Bermuda in the North Atlantic. This data indicates that over the period 1922-95 ocean thermal expansion was about 0.9 mm/yr. Comparisons of observations of ocean temperatures from trans-ocean top-to bottom measurements with historical data from the 1950s and 1960s indicate rates of thermal expansion of about 1 mm/yr. The only global analysis indicates smaller rates of thermal expansion over the period 1955-95 of 0.55 mm/yr. However, data gaps may bias this estimate low (Church, 2001).
Coupled atmosphere ocean general circulation models can also be used to estimate ocean thermal expansion during the 20th century and to project rates of thermal expansion into the future. Over the period 1910-90, these models indicate a rate of sea-level rise of between 0.3 and 0.8 mm/yr. For the period 1960-90, the model estimates of ocean thermal expansion are in the range of 0.6 to 1.1 mm/yr. This is in general agreement with the observational estimates but few rigorous comparisons have been completed to date.
Glaciers and Ice Caps
There are over 100,000 glaciers and ice caps. Excluding the ice sheets of Greenland and Antarctica, glaciers and ice caps contain the equivalent of about 0.5m of global averaged sea level. They gain mass by the accumulation of snow at higher elevations and lose mass by melting at the surface or base and possibly by the discharge of ice into floating ice shelves.
There has been widespread retreat of glaciers during the 20th century but only a small fraction of the glaciers and ice caps have been monitored to allow the direct estimation of their contribution to sea-level change. Instead, changes in glacier mass are calculating by estimating their sensitivity to changes in climate (both temperature and precipitation). Observational and model studies indicate a contribution to sea-level change of between 0.2 and 0.4 mm/yr averaged over the 20th century. In general, climate models project that increases in temperature (resulting in increased melting) will have a much larger impact on glacier mass than anticipated changes in precipitation. In making projections, it is necessary to consider both the regional and seasonal changes in temperature and the evolution of glacier area.
Ice Sheets
The annual average snowfall on the ice sheets of Greenland and Antarctica is equivalent to 6.5 mm of sea level. For the Greenland ice sheet, snowfall is balanced by melting and the discharge of glaciers. In contrast, Antarctic surface temperatures are so low that there is insignificant surface melting and the primary balance is between snowfall and the discharge of glaciers. Present day estimates of precipitation and melting and discharge are too inaccurate to provide useful constraints on the changes in mass of the ice sheets. Also, ice sheets have long response times and as a result they are still responding to changes since the last glacial maximum. To estimate this ongoing response we have used constraints from changes in sea-level over the last several millennia and from the use of numerical models. These two constraints indicate an ongoing contribution of between zero and 0.5 mm/yr of sea level for the combined contribution of the two ice sheets.
In addition to this ongoing response, anthropogenic climate change will also cause changes in ice sheet mass. For Greenland, projections are that increased melting from higher temperature will exceed increases in precipitation. For Antarctica, temperatures are too cold for there to be significant melting and thus increased precipitation will be stored as increased mass in Antarctica. This will partially offset positive sea-level rise contributions from other sources.
In the future, changes in the mass of the Greenland and Antarctic ice sheets will be estimated from changes of the ice sheet height and changes in gravity fields as measured with satellites. Changes in earth rotation parameters will form another useful constraint on estimates of changes in ice sheet mass.
Terrestrial Storage
The area of greatest uncertainty at present is the storage of water on land. The two major papers in this area, Gornitz et al (1997) and Sahagian (2000), arrive at somewhat different conclusions. Terrestrial contributions come from ground water mining, increased storage in lakes and reservoirs, infiltration into aquifers from reservoirs and irrigation and changes in runoff from urbanisation. Attention is required to improve estimates of the terrestrial storage term so that the 20th century global water budget can be more satisfactorily closed.
Synthesis of Understanding of 20th Century Changes
In summary, ocean thermal expansion, glacier melting, recent increased melting of the Greenland ice sheet and the ongoing response of the Antarctic ice sheet to changes since the last glacial maximum are all contributing to sea-level rise. In addition, there are small contributions to sea-level rise from changes in permafrost and sediment deposition in the ocean. Recent increases in precipitation over Antarctica are likely to partially offset these positive contributions to sea-level rise. For the 20th century, our estimate of the sum of terms is a global averaged rate of sea-level rise of between 0.3 and 1.8 mm/yr (with a central value of 1 mm/yr), compared with the observed rate of between 1 and 2 mm/yr. The highly uncertain terrestrial storage terms widen the sum of contributions to -0.8 to + 2.2 mm/yr, with a central value of 0.7 mm/yr. With or without the terrestrial storage terms, the model calculations are biased low compared with the observed sea-level rise. The reason for the discrepancies between observations and models is unclear. As discussed above, the accuracy of the observational estimate of sea-level rise has been questioned. It is also possible that we are underestimating one or more of the contributions to sea-level rise.
Sea-Level Rise Projections
1990 to 2100
Projections for sea-level rise for the period 1990 (to allow comparisons with earlier studies) to 2100 are made with a number of different greenhouse gas scenarios. These are the IS92a scenario used in the IPCC Second Assessment Report and the emission scenarios documented is the Special Report on Emission Scenarios (SRES; IPCC, 2000). For the IS92a scenario, projections are made using coupled atmosphere-ocean general circulation models, glacier models and ice sheet models. To make projections for the full range of 35 SRES scenarios, simple atmosphere-ocean models, tuned to reproduce the results of each of the available general circulation models, were used.
For the period 1990-2100, the largest contribution comes from ocean thermal expansion, followed by the glacier and ice cap contributions and the Greenland contribution. As expected, increases in precipitation in Antarctica result in some offsetting of other contributions. For the IS92a scenario, the sea-level rise projections are slightly less than in the SAR.
For the six illustrative scenarios, the available model projections were averaged. The projected sea-level rises (the central lines in Figure 1) barely diverge from each other prior to 2050; ie prior to 2050 the projected sea-level rise is not strongly dependent on which greenhouse gas scenario is used. This is a result of the long time scales associated with the ocean and the fact that sea level is still responding to past increases in greenhouse gas concentrations. After 2050, the projections begin to diverge and by 2100 the projections range from 30-50cm above 1990 levels. The projected sea-level rise for the full set of SRES scenarios and the full set of models (shown by the light shaded region in Figure 1) ranges from about 20-70cm. Including a generous allowance for land-ice uncertainties results in the full range of projections of sea level rise from 1990-2100 of between 9cm and 88cm.
Figure 1: Global-average sea-level rise 1990-2100 for the SRES scenarios. Thermal expansion and land ice changes were calculated using a simple climate model calibrated separately for each of seven AOGCMs, and contributions from changes in permafrost, the effect of sediment deposition and the long-term adjustment of the ice-sheets to past climate change were added. Each of the six central lines appearing in the key is the average of AOGCMs for one of the six illustrative scenarios. The region in dark shading shows the range of the average of AOGCMs for all 35 SRES scenarios. The region in light shading shows the range of all AOGCMs for all 35 scenarios. The region delimited by the outermost lines shows the range of all AOGCMs and scenarios including uncertainty in land-ice changes, permafrost changes and sediment deposition. (From Church et al, 2001.)
The use of GCMs for the projections allows estimates of the regional distribution of sea-level rise (Gregory et al, 2001). However, our confidence in these regional projections remains poor as there are significant differences in the patterns produced by the nine available models. Two common features of the models are a minimum in sea-level rise in the Southern Ocean and a maximum in the Arctic Ocean.
Sea-level rise will not only be felt through impacts of changes in the mean level (eg inundation) but also through changes in the intensity and frequency of extreme events (storm surges). There are two contributions to changes in extreme events - changes in the mean level and changes in the meteorological forcing of storm surges. To date there have been only a few studies of the combined impact of these two contributions. The example (Figure 2) given in the IPCC Assessment is for Immingham on the east coast of England. Under present day conditions a surge height of 1.5m is a one in forty year event. However, under increased meteorological forcing this 1.5m event becomes a one in 20 year event. For the combination of increased meteorological forcing and a change in the mean level in the mid-range of the projections for 2100, the 1.5m event becomes a one in three year event.
Figure 2: Frequency of extreme water level, expressed as return period, from a storm surge model for present-day conditions (control) and the projected climate around 2100 for Immingham on the east coast of England, showing changes resulting from mean sea-level rise and changes in meteorological forcing. The water level is relative to the sum of present-day mean sea-level and the tide at the time of the surge. (From Lowe et al, 2000.)
Longer-term Projections
Ocean thermal expansion and rising sea levels will continue long after greenhouse gas concentrations are stabilised. After 500 years, sea-level rise from ocean thermal expansion may be only half its eventual level. Models indicate that for a doubling of CO2 levels, sea-level rise from ocean thermal expansion might eventually be 0.5-2m (for a quadrupling, 1-4m).
Glaciers will continue to retreat and a substantial loss in glacial mass is likely.
For a sustained warming above 3ºC, melting of the Greenland ice sheet is projected to exceed precipitation. As a result, even if ice discharge to the ocean is reduced to zero, the Greenland ice sheet is projected to melt, leading eventually to a sea-level rise of up to 7m over millenia. For a warming over Greenland of 5.5°C, the projection is for a rise of sea level of over 1m by 2500 and 3m by 3000. (Projected global averaged temperature rises by 2100 range from 1.4°C to 5.8°C. For a range of models, projected temperature increases over Greenland are a factor of 1.2 to 3.1 greater than the global average temperature rise.)
Current models of the West Antarctic Ice Sheet (which is grounded below sea level) indicate that it is unlikely to make a significant contribution to sea level in the 21st century but its dynamics are inadequately understood, especially for projections on a longer time scale. Current ice dynamic models suggest that the West Antarctic Ice Sheet could contribute up to 3m over 1,000 years, but these results are strongly dependent on model assumptions.
Summary
Since the Second Assessment Report of the IPCC, significant progress in understanding the processes controlling sea-level rise has been made. However, significant uncertainties remain. Model results for the 20th century rate of sea-level rise encompass the observed range but uncertainties of all of the components (particularly the terrestrial storage terms) means that the modeled error bounds are broader than the observational range. Also, the central estimates from the models are biased low compared to the observations. The use of general circulation models has allowed estimates of the regional distribution of sea- level rise. However, confidence in these distributions remains low. Improving our ability to understand 20th century sea-level rise and to improve projections for the future requires continued satellite and in situ observation programs, continued model development and most importantly the rigorous testing of models.
Not only will sea-level rise change the mean level but it will also change the intensity and frequency of extreme events. If coastal development continues unabated at its present rate and for a mid-range sea-level rise scenario, Nicholls et al (1999) estimate the number of people who will have to respond to storm surges by 2100 will be tens to hundreds of millions. Sea-level rise is expected to continue well after 2100 and for centuries after greenhouse gas concentrations have stabilised. The eventual rise could be measured in metres. The issues associated with anthropogenic climate change and its impacts, including sea-level rise, are profound and require urgent attention by all levels of society.
References
- Cabanes, Cecile, Anny Cazenave and Christian Le Provost, 2001. Sea level rise during the 1990s and past 40 years: New insight from satellite and in situ observations. Science, 294, 840.
- Church, John A, 2001. How fast are sea levels rising? Science, 294, 802-803.
- Church JA, Gregory JM, Huybrechts P, Kuhn M, Lambeck K, Nhuan MT, Qin D, and Woodworth PL, 2001. Changes in Sea Level. In Climate Change (2001). The Scientific Basis. Contribution of Working Group 1 to the Third Assessment Report of the Intergovernmental Panel on Climate Change. Houghton, JT, Ding Y, Griggs DJ, Noguer M, van der Linden P, Dai X, Maskell K and Johnson CI, eds, Cambridge University Press, 639-694.
- Gornitz V, Rosenzweig C and Hillel D, (1997). Effects of anthropogenic intervention in the land hydrological cycle on global sea level rise. Global and Planetary Change, 14, 147-161.
- Gregory, J.M., J.A. Church, G.J. Boer, K.W. Dixon, G.M. Flato, D.R. Jackett, J.A. Lowe, S.P. O’Farrell, E. Roeckner, G.L. Russell, R.J. Stouffer and M. Winton, 2001. Comparison of results from several AOGCMs for global and regional sea-level change 1900-2100. Climate Dynamics, in press.
- IPCC (2001). Climate Change 2001: The Scientific Basis. Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change [Houghton, JT, Ding Y, Griggs DJ, Noguer M, van der Linden P, Dai X, Maskell K and Johnson CI, eds.]. Cambridge University Press, 881pp.
- IPCC (2000). Special Report on Emission Scenarios. A Special Report of Working Group III of the Intergovernmental Panel on Climate Change. Nakicenovic, N, Alcomo J, Davis G, de Vries B, Fenhann J, Gaffin S, Gregory K, Grübler A, Jung TY, Kram T, La Rovere EL, Michaelis L, Mori S, Morita T, Pepper W, Pitcher H, Price L, Riahi K, Roehrl A, Rogner H-H, Sankovski A, Schlesinger M, Shukla P, Smith S, Swart R, van Rooijen S, Victor N and Dadi Z, eds, Cambridge University Press, Cambridge, UK, 599pp.
- Lowe, J.A., J.M. Gregory and R.A.Flather, 2001: Changes in the occurrence of storm surges around the United Kingdom under a future climate scenario using a dynamic storm surge model driven by the Hadley Centre climate model. J Climate., in press.
- Nicholls RJ, Frank M.J. Hoozemans, Marcel Marchand. 1999. Increasing flood risk and wetland losses due to global sea-level rise: regional and global analyses. Global Environmental Change 9, S69-S87.
- Sahagian, D, (2000). Global physical effects of anthropogenic hydrological alterations: sea level and water redistribution. Global and Planetary Change, 25, 39-48.
- Warrick, R.A., C. Le Provost, M.F. Meier, J. Oerlemans, P.L. Woodworth, 1996: Changes in Sea Level. In: Climate Change 1995, The Science of Climate Change, Houghton, J.T., L.G. Meira Filho, B.A. Callander, N. Harris, A. Klattenberg, K. Maskell (Ed.). Cambridge University Press, 359-405.
|
