GEOPHYSICAL RESEARCH LETTERS, VOL. 33, L01602, doi:10.1029/2005GL024826, 2006

A 20th century acceleration in global sea-level rise

John A. Church1,2 and Neil J. White1,2

Received 6 October 2005; revised 22 November 2005; accepted 1 December 2005; published 6 January 2006.

[1] Multi-century sea-level records and climate models

2. Reconstructing Monthly Sea Levels

indicate an acceleration of sea-level rise, but no

20th

[4] We use the same techniques as in our earlier study

century acceleration has previously been detected. A

[Church et al., 2004] of using tide-gauge data to determine

reconstruction of global sea level using tide-gauge data

the changes in amplitude between consecutive months of a

from 1950 to 2000 indicates a larger rate of rise after 1993

selected number of these EOFs. These techniques were

and other periods of rapid sea-level rise but no significant

developed to estimate historical values of surface atmo-

acceleration over this period. Here, we extend the

spheric pressure and sea surface temperatures [Kaplan et al.,

reconstruction of global mean sea level back to 1870 and

2000; Rayner et al., 2003]. First differences of the tide-

find a sea-level rise from January 1870 to December 2004

gauge data are used because it is not possible to relate all of

of 195 mm, a 20th century rate of sea-level rise of 1.7 Â±

the separate records to a single vertical datum. The first

0.3 mm yr 1 and a significant acceleration of sea-level rise

differences of the EOF amplitudes are integrated backward

of 0.013 Â± 0.006 mm yr 2. This acceleration is an important

in time to estimate sea-level fields and hence GMSL each

confirmation of climate change simulations which show an

month. A revised scaling of the EOFs results in realistic

acceleration not previously observed. If this acceleration

formal error estimates.

remained constant then the 1990 to 2100 rise would range

[5] We calculate the EOFs from 12 years (compared with

from 280 to 340 mm, consistent with projections in the

9 years in our earlier study) of satellite altimeter data (T/P

IPCC TAR. Citation: Church, J. A., and N. J. White (2006), A

and Jason-1) from January 1993 to December 2004. All

20th century acceleration in global sea-level rise, Geophys. Res.

standard corrections except the inverse barometer correction

Lett., 33, L01602, doi:10.1029/2005GL024826.

are applied, including corrections for the drift of the water-

vapour measurements [Keihm et al., 2000; MacMillan et al.,

1. Introduction

2004] for both T/P and Jason-1 and for the drift of the T/P

sea-level measurements

[Mitchum, 2000]. We map the

[2] Most estimates of 20th century sea-level rise have

altimeter data to a 1

1 month grid. We remove

depended on averaging the rates of rise from the few, long,

the seasonal signal and a linear trend in GMSL, as we will

high-quality tide-gauge records that are available [Douglas,

use the EOFs to model variability about the time-varying

1991, 2001; Peltier, 2001]. However, these records contain

GMSL. An additional spatially uniform field is included in

significant decadal variability, obscuring any acceleration

the reconstruction to represent changes in GMSL. Omitting

[Woodworth, 1990; Douglas, 1992]. Even when a global-

this field results in a much smaller rate of GMSL rise,

mean sea level (GMSL) record is known to contain an

inconsistent with tide-gauge data

(in the mean and at

acceleration (as in numerical models), an acceleration is

individual sites) and earlier studies [e.g., Douglas, 1991],

difficult to detect in an average of a small number of records

and results in unrealistically large spatial variability in

[Gregory et al., 2001].

regional trends as a finite number of EOFs cannot

[3] The TOPEX/Poseidon (T/P) and Jason-1 satellite

adequately represent a substantial change in mean sea level.

altimeters have produced high quality measurements of near

Trends from our reconstructed time series agree well with

global (66 S to 66 N) sea level from 1993. The spatial

trends from long tide-gauge records. The EOFs provide

correlations from this data set, expressed as Empirical

information on global correlations of sea-level variability.

Orthogonal (eigen)Functions

(EOFs), together with the

There is no assumption that the spatial pattern of sea-level

longer but sparse tide-gauge data set, have been used to

rise for 1993 to 2004 is maintained over a longer period. We

produce estimates of reconstructed global sea-level variabil-

have not tried to detect the regional pattern of sea-level rise

ity [Chambers et al., 2002] and rise [Church et al., 2004]

resulting from the elastic response of the earth to present

for 1950 to 2000. As these estimates explicitly account for

day contributions from glaciers and ice sheets [Mitrovica et

the spatial redistribution of sea level, the temporal variabil-

al., 2001].

ity is at least an order of magnitude lower than that present

[6] We use monthly sea-level data downloaded from the

in individual records. The estimates also allow for the

Permanent Service for Mean Sea Level

(PSMSL

possibility of spatial variability in the rate of sea-level rise

[Woodworth and Player, 2003]) web site (http://www.pol.

[Nakiboglu and Lambeck, 1991].

ac.uk/psmsl/) in February 2003. Careful selection and edit-

ing criteria as given by Church et al. [2004] are employed.

1CSIRO Marine and Atmospheric Research, Hobart, Tasmania,

Where there are multiple records near a single satellite grid

Australia.

point, the changes in height at each time step were averaged.

2Antarctic Climate and Ecosystems Cooperative Research Centre,

The error estimates of first differences of

50 mm (the

Hobart, Tasmania, Australia.

solution is not sensitive to the value used) was computed

from the rms of the differences between the few sets of

0094-8276/06/2005GL024826$05.00

nearby (within about 100 km) sea-level records. The impact

L01602

1 of 4

L01602

CHURCH AND WHITE: AN ACCELERATION IN GLOBAL SEA-LEVEL RISE

L01602

corrections (0.09 mm yr

1 from the rms difference between

GMSL trends calculated using three different GIA models

[Church et al.,

2004]) and uncertainties in the EOFs

(0.1 mm yr 1, see below). For 1950 to 2000, the linear

least squares trend is 1.75 mm yr

1, consistent with earlier

estimates [Church et al., 2004; Douglas et al., 1991, 2001;

Peltier,

2001]. The yearly-averaged reconstructed GMSL

agrees with the T/P/Jason-1 satellite altimeter data from

1993 to within the error estimates (Figure 2). It also agrees

well with an average estimated directly from the tide

gauges, but has much smaller error bars.

[9] Fitting a quadratic to the GMSL time series gives an

acceleration

(twice the quadratic coefficient) of

0.013 Â±

0.006 mm yr 2 (95%) for 1870 to 2001. The differences

between the quadratic and the GMSL time series have an

rms value of only 7.5 mm (Figure 2b), less than the error

estimates for most of the record. For the 20th century alone,

Figure 1. The number and distribution of sea-level records

the acceleration is smaller at 0.008 Â± 0.008 mm yr

(95%).

available for the reconstruction. (a) The number of locations

Another approach, given the clear change of slope at

with sea-level data.

(b),

(d) the distribution of

1930, is to do linear regressions on the two halves

gauges in the 1980s, 1950s and 1900s.

(1870 - 1935 and 1936 - 2001) of the record. The slopes

are 0.71 Â± 0.40 and 1.84 Â± 0.19 mm yr

1 respectively,

implying an acceleration of 0.017 Â± 0.007 mm yr

(95%).

on measured sea level of the ongoing response of the earth

to changes in surface loading following the last glacial

maximum was removed using the same estimate of glacial

isostatic adjustment

(GIA) as in our earlier study, as

calculated by Mitrovica and colleagues

[Davis and

Mitrovica, 1996; Milne et al., 2001].

[7] As the analysis technique allows us to ingest data

from a time-varying array of tide gauges in a consistent way,

we can use many more gauges to estimate GMSL than the

traditional approach of estimating sea-level rise [Douglas,

1991, 1992, 2001; Peltier, 2001]. This leads to better spatial

coverage and reduced errors in the estimate of GMSL as the

few very long records available are clustered in small

regions (mostly NW Europe, North America and a few in

Australia and New Zealand). The number of sea-level

records available for each year is a maximum in the

1980s (Figure 1) but decreases rapidly in the 1990s as a

result of late submission of data sets. Running back in time,

the number of gauges drops to about 140 in 1950, to just

under 50 in 1900, with the majority of gauges in the

northern hemisphere. By the 1870s, there are only 10 gauges

(none in the southern hemisphere) and by 1860 there are

only five gauges available, too few to extend the recon-

struction back past 1870. The number of records limits our

ability to reconstruct GMSL to January1870 to December

2001. To ensure that we always solve an over-determined

problem we use 5 EOFs prior to 1900 and 10 after 1900, but

the number of EOFs has virtually no impact on the

Figure 2. Global mean sea level from the reconstruction

dominant EOF amplitudes or on GMSL.

for January 1870 to December 2001. (a) The monthly global

average, the yearly average with the quadratic fit to the

yearly values and the yearly averages with the satellite

3. Global-Mean Sea Level From 1870 to 2004

altimeter data superimposed are offset by 150 mm. The one

[8] From the start of the reconstruction in January 1870

(dark shading) and two (light shading) standard deviation

to the end of the altimeter data in December

2004

error estimates are shown. (b) Departures of the GMSL

(135 years), the total GMSL rise is 195 mm (Figure 2),

from the quadratic fit to the data. (c) Linear trends in sea

an average of 1.44 mm yr

1. For the 20th century, the rise is

level from the reconstruction for overlapping

year

about 160 mm and the linear least-squares trend is 1.7 Â±

periods. The trend for each period is plotted at the centre

0.3 mm yr 1 (95% confidence limits). This error includes

time of the period. The error estimates of GMSL are a

allowance for the serial correlation of the time series, (four

minimum of about 5 mm in the 1980s rising to about 22 mm

years of data per degree of freedom), uncertainties in GIA

in 1870.

2 of 4

L01602

CHURCH AND WHITE: AN ACCELERATION IN GLOBAL SEA-LEVEL RISE

L01602

The rms residual to the linear fits is lower at

5.8

short-term (years to a decade or so) reductions in GMSL

(cf

7.5

mm), consistent with much of the acceleration

following major volcanic eruptions. The post-1960 major

occurring in the first half of the 20th century rather than a

volcanic eruptions of Mt. Agung (1963), El Chichon (1982)

smooth acceleration over the whole period. Recent esti-

and Mt Pinatubo (1991) offset about 0.005 mm yr

2 of the

mates of regional- and global-MSL constructed using very

acceleration that is otherwise present, perhaps explaining

different techniques are qualitatively similar to ours, includ-

why little acceleration has been detected over the second

ing a significant acceleration in the first half of the 20th

half of the

20th century. The 1930s acceleration occurs

century (S. A. Jevrejeva et al., Nonlinear trends and multi-

during a period of little volcanic activity.

year cycles in sea level records, submitted to Journal of

[13] The quadratic implies that the rate of rise was zero in

Geophysical Research, 2005).

about 1820 when GMSL was about 200 mm below present

[10] While the GIA corrections are essentially constant

day values. This level is consistent with estimates from

from 1870 to 2000, it is possible that a temporally varying

bench marks carved in rock in Tasmania in 1840 [Hunter et

tide-gauge array may combine with errors in the GIA

al.,

2003] and the height of ancient Roman fish tanks

corrections to give an error in the computed acceleration.

[Lambeck et al., 2004], which implies virtually no long-

Tests, including assuming a 100% error in the GIA, indicate

term average change in GMSL from the first century AD to

negligible impact on the computed acceleration. Using

1800 AD.

EOFs obtained from the recent altimeter data is another

[14] The 19th century commencement of the acceleration

potential source of error. Tests using EOFs determined from

is consistent with geological data [Donelly et al., 2004] and

different subsets of the 12 year altimeter record lead to 1-s

long tide-gauge records [Woodworth, 1990, 1999; Maul and

uncertainty in GMSL trends of about 0.06 mm yr

1 and

Martin, 1993] but this is the first time a post 1870 (and 20th

negligible effects on the estimates of the acceleration. El

century) acceleration of GMSL has been detected. This

Nino-Southern Oscillation variability (ENSO, the dominant

acceleration is an important confirmation of climate simu-

signal in the low order EOFs) has been present for millen-

lations [Gregory et al., 2001] which show an acceleration

nia. However, over time, changes in ENSO patterns could

not previously detected in observations. Sea-level rise from

occur and we therefore almost double our uncertainty

20th century climate simulations [Church et al., 2001] is

estimate, assigning a total

1-s error from uncertainty in

somewhat less than that inferred from observations, perhaps

EOFs of 0.1 mm yr 1. Note also that any non-stationarity

because the acceleration of sea-level rise commenced during

of patterns will also be reflected in the error estimates of

the 19th century and by 1870, at the start of our recon-

GMSL. This conclusion of the relatively minor impact of

struction, was already rising at a rate of about 0.6 mm yr

the assumption of stationarity of EOFs on GMSL trends is

Both the rate of rise and the observed acceleration should be

consistent with previous work

[Chambers et al.,

2002;

valuable constraints to test the next round of climate

Church et al., 2004] and the application of these techniques

simulations.

to the estimate of sea surface temperature variations [Rayner

et al., 2003].

[15] Acknowledgments. This paper is a contribution to the CSIRO

Climate Change Research Program and the CSIRO Wealth from Oceans

Flagship and was supported by the Australian Governmentâs Cooperative

4. Implications and Conclusions

Research Centres Programme through the Antarctic Climate and Ecosys-

tems Cooperative Research Centre. T/P and Jason-1 data were obtained

[11] If this acceleration was maintained through the 21st

from the NASA Physical Oceanography DAAC at the Jet Propulsion

century, sea level in 2100 would be 310 Â± 30 mm higher

Laboratory/California Institute of Technology. Sea-level data is from the

Permanent Service for Mean Sea Level.

than in 1990, overlapping with the central range of projec-

tions in the Intergovernmental Panel on Climate Change

References

Third Assessment Report (IPCC TAR) [Church et al.,

Chambers, D. P., C. A. Melhaff, T. J. Urban, D. Fuji, and R. S. Nerem

2001]. For 1910 to 1990, the acceleration in ocean thermal

(2002), Low-frequency variations in global mean sea level: 1950 - 2000,

expansion (only) in these climate models range from 0.005 Â±

J. Geophys. Res., 107(C4), 3026, doi:10.1029/2001JC001089.

0.003 mm yr 2 to 0.014 Â± 0.004 mm yr 2 (Table 11.2 of the

Church, J. A., et al. (2001), Changes in sea level, in Climate Change 2001:

IPCC TAR), consistent with the present estimates of

The Scientific Basis, edited by J. T. Houghton et al., pp. 639 - 694, Cam-

bridge Univ. Press, New York.

0.013 mm yr 2 and 0.017 mm yr 2 for the 132 year period

Church, J. A., N. J. White, R. Coleman, K. Lambeck, and J. X. Mitrovica

and the 0.008 mm yr 2 for the 20th century.

(2004), Estimates of the regional distribution of sea-level rise over the

[12] Between 1930 and 1960, GMSL rises faster than the

1950 to 2000 period, J. Clim., 17, 2609 - 2625.

Church, J. A., N. J. White, and J. M. Arblaster (2005), Significant decadal-

quadratic curve at a rate of about 2.5 mm yr

(Figure 2c),

scale impact of volcanic eruptions on sea level and ocean heat content,

following (with about a 20 year lag) the 1910 to 1940

Nature, 438, 74 - 77.

period of more rapid global temperature rise [Folland et al.,

Davis, J. L., and J. X. Mitrovica (1996), Glacial isostatic adjustment and the

anomalous tide gauge record of eastern North America, Nature, 379,

2001]. Variability in GMSL trends prior to 1930 are not

331 - 333.

significant. After 1960, there are minima in the rates of rise

Donelly, J. P., P. Cleary, P. Newby, and R. Ettinger (2004), Coupling in-

in the 1960s and 1980s, each followed by more rapid rates

strumental and geological records of sea-level change: Evidence from

of rise (peaking at over 3 mm yr

1), consistent with Holgate

southern New England of an increase in the rate of sea-level rise in the

late

19th century, Geophys. Res. Lett., 31, L05203, doi:10.1029/

and Woodworth [2004]. From 1993, the rates of rise

2003GL018933.

estimated from tide gauge and altimeter data (after correc-

Douglas, B. C. (1991), Global sea level rise, J. Geophys. Res., 96, 6981 -

tion for GIA effects [Douglas and Peltier, 2002]) are about

6992.

Douglas, B. C. (1992), Global sea level acceleration, J. Geophys. Res., 97,

3 mm yr 1 [Leuliette et al., 2004; Church et al., 2004],

12,699 - 12,706.

faster than the quadratic (about 2.3 mm yr

1) at this time.

Douglas, B. C. (2001), Sea level change in the era of the recording tide

Model simulations and data [Church et al., 2005] show

gauge, in Sea Level Rise, Int. Geophys. Ser., vol. 75, edited by B. C.

3 of 4

L01602

CHURCH AND WHITE: AN ACCELERATION IN GLOBAL SEA-LEVEL RISE

L01602

Douglas, M. S. Kearney, and S. P. Leatherman, pp. 37 - 64, Elsevier, New

Milne, G. A., J. L. Davis, J. X. Mitrovica, H.-G. Scherneck, J. M.

York.

Johansson, M. Vermeer, and H. Koivula (2001), Space-geodetic con-

Douglas, B. C., and W. R. Peltier (2002), The puzzle of global sea-level

straints on glacial isostatic adjustment in Fennoscandia, Science, 291,

rise, Phys. Today, 55, 35 - 40.

2381 - 2385.

Folland, C. K., et al. (2001), Observed climate variability and change, in

Mitchum, G. T. (2000), An improved calibration of satellite altimetric

Climate Change 2001: The Scientific Basis, edited by J. T. Houghton et

height using tide gauge sea levels with adjustment for land motion,

al., pp. 99 - 181, Cambridge Univ. Press, New York.

Mar. Geod., 23, 145 - 166.

Gregory, J. M., et al. (2001), Comparison of results from several AOGCMs

Mitrovica, J. X., M. Tamisiea, J. L. Davis, and G. A. Milne (2001), Recent

for global and regional sea-level change 1900 - 2100, Clim. Dyn., 18,

mass balance of polar ice sheets inferred from patterns of global sea-level

225 - 240.

change, Nature, 409, 1026 - 1029.

Holgate, S. J., and P. L. Woodworth (2004), Evidence for enhanced coastal

Nakiboglu, S. M., and K. Lambeck (1991), Secular sea level change, in

sea level rise during the 1990s, Geophys. Res. Lett., 31, L07305,

Glacial Isostasy, Sea Level, and Mantle Rheology, edited by R. Sabatini,

doi:10.1029/2004GL019626.

K. Lambeck, and E. Boschi, pp. 237 - 258, Springer, New York.

Hunter, J., R. Coleman, and D. Pugh (2003), The sea level at Port Arthur,

Peltier, W. R. (2001), Global glacial isostatic adjustment and modern in-

Tasmania, from 1841 to the present, Geophys. Res. Lett., 30(7), 1401,

strumental records of relative sea level history, in Sea Level Rise, Int.

doi:10.1029/2002GL016813.

Geophys. Ser., vol. 75, edited by B. C. Douglas, M. S. Kearney, and S. P.

Kaplan, A., Y. Kushnir, and M. A. Cane (2000), Reduced space optimal

Leatherman, pp. 65 - 95, Elsevier, New York.

interpolation of historical marine sea level pressure, J. Clim., 13, 2987 -

Rayner, N., et al. (2003), Global analyses of sea surface temperature, sea

3002.

ice, and night marine air temperature since the late nineteenth century,

Keihm, S., V. Zlotnicki, and C. Ruf (2000), TOPEX microwave radiometer

J. Geophys. Res., 108(D14), 4407, doi:10.1029/2002JD002670.

performance evaluation, IEEE Trans. Geosci. Remote Sens., 38, 1379 -

Woodworth, P. L. (1990), A search for accelerations in records of European

1386.

mean sea level, Int. J. Climatol., 10, 129 - 143.

Lambeck, K., M. Anzidei, F. Antonioli, A. Benini, and A. Espositol (2004),

Woodworth, P. L. (1999), High waters at Liverpool since 1768: The UKâs

Sea level in Roman time in the central Mediterranean and implications for

longest sea level record, Geophys. Res. Lett., 26, 1589 - 1592.

recent change, Earth Planet. Sci. Lett., 224, 563 - 575.

Woodworth, P. L., and R. Player (2003), The Permanent Service for

Leuliette, E. W., R. S. Nerem, and G. Mitchum (2004), Calibration of

Mean Sea Level: An update to the 21st century, J. Coastal Res.,

TOPEX/Poseidon and Jason altimeter data to construct a continuous

19, 287 - 295.

record of mean sea level change, Mar. Geod., 27, 79 - 94.

MacMillan, D. S., B. D. Beckley, and P. Fang (2004), Monitoring the

TOPEX and Jason-1 microwave radiometers with GPS and VLBI wet

zenith path delays, Mar. Geod., 27, 703 - 716.

Maul, G. A., and D. M. Martin (1993), Sea level rise at Key West, Florida,

J. A. Church and N. J. White, CSIRO Marine and Atmospheric Research,

1846 - 1992: Americaâs longest instrument record, Geophys. Res. Lett.,

GPO Box 1538, Hobart, Tas 7001, Australia.

(john.church@csiro.au;

20, 1955 - 1958.

neil.white@csiro.au)

4 of 4