Climate Change 2001:
Working Group I: The Scientific Basis
Other reports in this collection

11.2.6 Tectonic Land Movements

We define tectonic land movement as that part of the vertical displacement of the crust that is of non-glacio-hydro-isostatic origin. It includes rapid displacements associated with earthquake events and also slow movements within (e.g., mantle convection) and on (e.g., sediment transport) the Earth. Large parts of the earth are subject to active tectonics which continue to shape the planet's surface. Where the tectonics occur in coastal areas, one of its consequences is the changing relationship between the land and sea surfaces as shorelines retreat or advance in response to the vertical land movements. Examples include the Huon Peninsula of Papua New Guinea (Chappell et al., 1996b), parts of the Mediterranean (e.g. Pirazzoli et al., 1994; Antonioli and Oliverio, 1996), Japan (Ota et al., 1992) and New Zealand (Ota et al., 1995). The Huon Peninsula provides a particularly good example (Figure 11.6) with 125,000 year old coral terraces at up to 400 m above present sea level. The intermediate terraces illustrated in Figure 11.6 formed at times when the tectonic uplift rates and sea level rise were about equal. Detailed analyses of these reef sequences have indicated that long-term average uplift rates vary between about 2 and 4 mm/yr, but that large episodic (and unpredictable) displacements of 1 m or more occur at repeat times of about 1,000 years (Chappell et al., 1996b). Comparable average rates and episodic displacements have been inferred from Greek shoreline evidence (Stiros et al., 1994). With major tectonic activity occurring at the plate boundaries, which in many instances are also continental or island margins, many of the world's tide gauge records are likely to contain both tectonic and eustatic signals. One value of the geological data is that it permits evaluations to be made of tectonic stability of the tide gauge locality.

Table 11.9: Recent estimates of sea level rise from tide gauges. The standard error for these estimates is also given along with the method used to correct for vertical land movement (VLM).
  Region VLMa Rate ± s.e.b
Gornitz and Lebedeff (1987) Global Geological 1.2 ± 0.3
Peltier and Tushingham (1989, 1991) Global ICE-3G/M1 2.4 ± 0.9c
Trupin and Wahr (1990) Global ICE-3G/M1 1.7 ± 0.13
Nakiboglu and Lambeck (1991) Global Spatial decomposition 1.2 ± 0.4
Douglas (1991) Global ICE-3G/M1 1.8 ± 0.1
Shennan and Woodworth (1992) NW Europe Geological 1.0 ± 0.15
Gornitz (1995)d N America E Coast Geological 1.5 ± 0.7c
Mitrovica and Davis (1995),
Davis and Mitrovica (1996)
Global far field (far from former ice sheets) PGR Model 1.4 ± 0.4c
Davis and Mitrovica (1996) N America E Coast PGR Model 1.5 ± 0.3c
Peltier (1996) N America E Coast ICE-4G/M2 1.9 ± 0.6c
Peltier and Jiang (1997) N America E Coast Geological 2.0 ± 0.6c
Peltier and Jiang (1997) Global ICE-4G/M2 1.8 ± 0.6c
Douglas (1997)d Global ICE-3G/M1 1.8 ± 0.1
Lambeck et al. (1998) Fennoscandia PGR Model 1.1 ± 0.2
Woodworth et al. (1999) British Isles Geological 1.0
a This column shows the method used to correct for vertical land motion. ICE-3G/M1 is the Post Glacial Rebound (PGR) model of Tushingham and Peltier (1991). ICE-4G/M2 is a more recent PGR model based on the deglaciation history of Peltier (1994) and the mantle viscosity model of Peltier and Jiang (1996). Nakiboglu and Lambeck (1991) performed a spherical harmonic decomposition of the tide-gauge trends and took the the zero-degree term as the global-average rate. They indicated that a PGR signal would make little contribution to this term. The use of geological data is discussed in the text.
b The uncertainty is the standard error of the estimate of the global average rate.
c This uncertainty is the standard deviation of the rates at individual sites.
d See references in these papers for estimates of sea level rise for various other regions.

Over very long time-scales (greater than 106 years), mantle dynamic processes lead to changes in the shape and volume of the ocean basins, while deposition of sediment reduces basin volume. These affect sea level but at very low rates (less than 0.01 mm/yr and 0.05 mm/yr, respectively; e.g., Open University, 1989; Harrison, 1990).

Coastal subsidence in river delta regions can be an important contributing factor to sea level change, with a typical magnitude of 10 mm/yr, although the phenomenon will usually be of a local character. Regions of documented subsidence include the Mediterranean deltas (Stanley, 1997), the Mississippi delta (Day et al., 1993) and the Asian deltas. In the South China Sea, for example, the LGM shoreline is reported to occur at a depth of about 165 m below present level (Wang et al., 1990), suggesting that some 40 m of subsidence may have occurred in 20,000 years at an average rate of about 2 mm/yr. Changes in relative sea level also arise through accretion and erosion along the coast; again, such effects may be locally significant.

11.2.7 Atmospheric Pressure

Through the inverse barometer effect, a local increase in surface air pressure over the ocean produces a depression in the sea surface of 1 cm per hPa (1 hPa = 1 mbar). Since water is practically incompressible, this cannot lead to a global-average sea level rise, but a long-term trend in surface air pressure patterns could influence observed local sea level trends. This has been investigated using two data sets: (i) monthly mean values of surface air pressure on a 10°x5° grid for the period 1873 to 1995 for the Northern Hemisphere north of 15°N obtained from the University of East Anglia Climatic Research Unit, and (ii) monthly mean values on a global 5°x5° grid for the period 1871 to 1994 obtained from the UK Met Office (see Basnett and Parker, 1997, for a discussion of the various data sets). The two data sets present similar spatial pattens of trends for their geographical and temporal overlaps. Both yield small trends of the order 0.02 hPa/yr; values of -0.03 hPa/yr occur in limited regions of the high Arctic and equatorial Pacific. As found by Woodworth (1987), trends are only of the order of 0.01 hPa/yr in northern Europe, where most of the longest historical tide gauges are located. We conclude that long-term sea level trends could have been modified to the extent of ± 0.2 mm/yr, considerably less than the average eustatic rate of rise. Over a shorter period larger trends can be found. For example, Schönwiese et al. (1994) and Schönwiese and Rapp (1997) report changes in surface pressure for the period 1960 to 1990 that could have modified sea level trends in the Mediterranean and around Scandinavia by -0.05 and +0.04 mm/yr respectively.

Other reports in this collection