Working Group I: The Scientific Basis

Other reports in this collection

11.4 Can 20th Century Sea Level Changes be Explained?

In order to have confidence in our ability to predict future changes in sea level, we need to confirm that the relevant processes (Section 11.2) have been correctly identified and evaluated. We attempt this by seeing how well we can account for the current rate of change (Section 11.3). We note that:

  • some processes affecting sea level have long (centuries and longer) time-scales, so that current sea level change is also related to past climate change,
  • some relevant processes are not determined solely by climate,
  • fairly long records (at least 50 years according to Douglas, 1992) are needed to detect a significant trend in local sea level, because of the influence of natural variability in the climate system, and
  • the network of tide gauges with records of this length gives only a limited coverage of the world’s continental coastline and almost no coverage of the mid-ocean.
Table 11.14: Sea level rise 1990 to 2100 due to climate change derived from AOGCM experiments following the IS92a scenario, including the direct effect of sulphate aerosols. See Tables 8.1 and 9.1 for further details of models and experiments. Results were extrapolated to 2100 for experiments ending at earlier dates. The uncertainties shown in the land ice terms are those discussed in this section. For comparison the projection of Warrick et al. (1996) (in the SAR) is also included. Note that the minimum of the sum of the components is not identical with the sum of the minima because the smallest values of the components do not all come from the same AOGCM, and because for each model the land ice uncertainties have been combined in quadrature; similarly for the maxima, which also include non-zero contributions from smaller terms.
Experiment Sea level rise (m) 1990 to 2100
  Expansion Glaciers Greenland Antarctica a Sum b
min max min max min max min max min max
CGCM1 GS 0.43 0.03 0.23 0.00 0.07 –0.07 0.02 0.45 0.77
CSIRO Mk2 GS 0.33 0.02 0.22 –0.01 0.08 –0.12 –0.04 0.29 0.60
ECHAM4/OPYC3 GS 0.30 0.02 0.18 –0.02 0.03 –0.17 –0.06 0.19 0.48
GFDL_R15_a GS 0.38 0.02 0.19 –0.01 0.09 –0.09 –0.01 0.37 0.67
HadCM2 GS 0.23 0.02 0.17 –0.01 0.05 –0.09 0.00 0.21 0.48
HadCM3 GSIO 0.24 0.02 0.18 0.00 0.05 –0.13 –0.03 0.18 0.46
MRI2 GS 0.11 0.01 0.11 0.00 0.03 –0.04 0.00 0.11 0.31
DOE PCM GS 0.19 0.01 0.13 –0.01 0.06 –0.13 –0.04 0.12 0.37
Range 0.11 0.43 0.01 0.23 –0.02 0.09 –0.17 0.02 0.11 0.77
Central value 0.27 0.12 +0.04 -0.08 0.44
SAR Best estimate 0.28 0.16 +0.06 -0.01 0.49 Range   0.20 0.86
a Note that this range does not allow for uncertainty relating to ice-dynamical changes in the West Antarctic ice sheet. See Section for a full discussion.
b Including contributions from permafrost, sedimentation, and adjustment of ice sheets to past climate change.

Figure 11.11:
Global average sea level rise 1990 to 2100 for the IS92a scenario, including the direct effect of sulphate aerosols. Thermal expansion and land ice changes were calculated from AOGCM experiments, 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. For the models that project the largest (CGCM1) and the smallest (MRI2) sea level change, the shaded region shows the bounds of uncertainty associated with land ice changes, permafrost changes and sediment deposition. Uncertainties are not shown for the other models, but can be found in Table 11.14. The outermost limits of the shaded regions indicate our range of uncertainty in projecting sea level change for the IS92a scenario.

The estimated contributions from the various components of sea level rise during the 20th century (Table 11.10, Figure 11.9) were constructed using the results from Section 11.2. The sum of these contributions for the 20th century ranges from –0.8 mm/yr to 2.2 mm/yr, with a central value of 0.7 mm/yr. The upper bound is close to the observational upper bound (2.0 mm/yr), but the central value is less than the observational lower bound (1.0 mm/yr), and the lower bound is negative i.e. the sum of components is biased low compared to the observational estimates. Nonetheless, the range is narrower than the range given by Warrick et al. (1996), as a result of greater constraints on all the contributions, with the exception of the terrestrial storage terms. In particular, the long-term contribution from the ice sheets has been narrowed substantially from those given in Warrick et al. (1996) by the use of additional constraints (geological data and models of the ice sheets) (Section 11.3.1).

The reason for the remaining discrepancy is not clear. However, the largest uncertainty (by a factor of more than two) is in the terrestrial storage terms. Several of the components of the terrestrial storage term are poorly determined and the quoted limits require several of the contributions simultaneously to lie at the extremes of their ranges. This coincidence is improbable unless the systematic errors affecting the estimates are correlated. Furthermore, while coupled models have improved considerably in recent years, and there is general agreement between the observed and modelled thermal expansion contribution, the models’ ability to quantitatively simulate decadal changes in ocean temperatures and thus thermal expansion has not been adequately tested. Given the poor global coverage of high quality tide gauge records and the uncertainty in the corrections for land motions, the observationally based rate of sea level rise this century should also be questioned.

In the models, at least a third of 20th century anthropogenic eustatic sea level rise is caused by thermal expansion, which has a geographically non-uniform signal in sea level change. AOGCMs do not agree in detail about the patterns of geographical variation (see Section 11.5.2). They all give a geographical spread of 20th century trends at individual grid points which is characterised by a standard deviation of 0.2 to 0.5 mm/yr (Gregory et al., 2001). This spread is a result of a combination of spatial non-uniformity of trends and the uncertainty in local trend estimates arising from temporal variability. As yet no published study has revealed a stable pattern of observed non-uniform sea level change. Such a pattern would provide a critical test of models. If there is significant non-uniformity, a trend from a single location would be an inaccurate estimate of the global average. For example, Douglas (1997) averaged nine regions and found a standard deviation of about 0.3 mm/yr (quoted by Douglas as a standard error), similar to the range expected from AOGCMs.

A common perception is that the rate of sea level rise should have accelerated during the latter half of the 20th century. The tide gauge data for the 20th century show no significant acceleration (e.g., Douglas, 1992). We have obtained estimates based on AOGCMs for the terms directly related to anthropogenic climate change in the 20th century, i.e., thermal expansion (Section, ice sheets (Section, glaciers and ice caps (Section (Figure 11.10a). The estimated rate of sea level rise from anthropogenic climate change ranges from 0.3 to 0.8 mm/yr (Figure 11.10b). These terms do show an acceleration through the 20th century (Figure 11.10a,b). If the terrestrial storage terms have a negative sum (Section 11.2.5), they may offset some of the acceleration in recent decades. The total computed rise (Figure 11.10c) indicates an acceleration of only 0.2 mm/yr/century, with a range from -1.1 to +0.7 mm/yr/century, consistent with observational finding of no acceleration in sea level rise during the 20th century (Section The sum of terms not related to recent climate change is -1.1 to +0.9 mm/yr (i.e., excluding thermal expansion, glaciers and ice caps, and changes in the ice sheets due to 20th century climate change). This range is less than the observational lower bound of sea level rise. Hence it is very likely that these terms alone are an insufficient explanation, implying that 20th century climate change has made a contribution to 20th century sea level rise.

Recent studies (see Sections 2.3.3, 2.3.4) suggest that the 19th century was unusually cold on the global average, and that an increase in solar output may have had a moderate influence on warming in the early 20th century (Section This warming might have produced some thermal expansion and could have been responsible for the onset of glacier recession in the early 20th century (e.g., Dowdeswell et al., 1997), thus providing a possible explanation of an acceleration in sea level rise commencing before major industrialisation.

Other reports in this collection

IPCC Homepage