Changing sea levels and natural hazards

Sea levels have changed throughout geological time (e.g. Haq et al., 1987) in response to a range of different and sometimes interacting isostatic, eustatic and tectonic processes (Dawson, 1992; Box 6.1). Natural hazards related to such sea-level change are surprisingly many and varied, and the relationship between the two is often far from clear. Broadly speaking, rising sea levels can be expected to increase the threat to coastal zones, primarily owing to the inundation or flooding of low-lying terrain (see Chapter 3) but also through increased erosion, déstabilisation and collapse of elevated coastlines. Higher sea levels will also exacerbate the impact and destructive potential of storm surges and tsunami, partly because of the elevated level of the sea surface but also through increasing the exposure of many coastlines as a result of inundation of wetlands and other protective environments. The hazard implications of falling sea levels are less obvious, although it has been suggested that rapid drops in sea level may trigger submarine landslides (e.g. Maslin et al., 1998; Rothwell et al., 1998). On a much broader scale, a number of authors have proposed that large sea-level changes - either up or down - may trigger

Box 6.1 Causes of sea-level change in the geological record

Eustatic sea-level changes Eustatic changes in sea level represent changes in the form and level of the surface of the Earth's oceans that exist in equilibrium with the planet's gravitational field. Eustatic sea-level changes occur as a result of three geophysical mechanisms. Glacio-eustatic changes reflect vertical changes to the global ocean surface as a result of variations in water volume due to the growth and melting of ice caps and glaciers. During the Quaternary, major fluctuations in sea level were largely glacio-eustatic, occurring in response to changes in global ice volume during glaciation-déglaciation cycles. Tectono-eustatic changes arise from modifications to the shapes of the ocean basins as a result of plate tectonic processes such as the widening of the ocean basins due to sea-floor spreading or their shrinkage through the consumption of ocean floor at subduction zones. For example, Bloom (1971) proposed that owing to sea-floor spreading during the late Quaternary, the ocean basins are capable of holding 6 per cent more water by volume now than during the last interglacial. Geoidal-eustatic changes involve modifications to the shape of the ocean surface. Like the land, the surface of the ocean possesses a 'topography' with swells and depressions that can nowadays be contoured and monitored using precise satellite altimetry. This watery 'topography' is defined as the geoid, which is the equipotential surface of the Earth's gravity field. Gravitational attraction is affected by, for example, density variations within the crust and mantle and the sizes of ice masses at high latitudes. Variations in geoidal sea-surface altitudes are today measured using satellite instruments (such as TOPEX/POSEIDON and Jason-1), which have revealed vertical differences in the sea surface of the order of 200 m. In contrast to glacio- and tectono-eustatic sea-level changes, those due to variations in the geoid are not global in their effects. The importance of the nature of geoidal sea-level variations lies in the fact that many published curves of Quaternary sea-level change may be accurate on only a regional rather than a global scale.

Isostatic sea-level changes Isostatically related sea-level changes are a consequence of accommodations in the equilibrium that exists between the lithosphère and the underlying asthenosphere on which it 'floats'. During the Quaternary, most isostatic changes in sea level reflected loading and unloading effects associated with the redistribution of ice and water during glaciation-déglaciation cycles. Glacio-isostatic changes relate to the depression and rebound of glaciated continental and adjacent areas, while hydro-isostatic changes are related to the subsidence (and, arguably, the uplift) of continental margins that are respectively loaded and unloaded by changing ocean volumes. Owing to the elastic behaviour of the lithosphere, depression and succeeding uplift in response to ice sheet growth and decay can exceed 1 km. Over the past 7000 years alone, Chappell (1974) has proposed that sea-level rise has resulted in an overall depression of the ocean floor by 8 m, accompanied by an average 16 m rise of the continents, although these numbers have been questioned.

Tectonic sea-level changes In addition to large-scale tectono-eustatic changes related to modifications to the ocean basins, local sea-level changes may also result from vertical crustal movements, particularly in tectonically active settings such as island arcs and collisional plate margins. Such behaviour has drastically modified the global or regional Quaternary sea-level change records for a number of tectonically active areas including the Mediterranean region, Japan and the Aleutian Islands.

increased volcanism (e.g. Wallmann et al, 1988; Nakada and Yokose, 1992; McGuire et al., 1997) and seismicity (Nakada and Yokose, 1992) along continental margins. Proposed mechanisms range from global spin (Matthews, 1969) or hydro-isostatic crustal (Anderson, 1974; Chappell, 1975; Rampino et al., 1979; Nakada and Yokose, 1992) readjustments related to the redistribution of planetary water, to responses by individual coastal and island volcanoes to large sea-level changes (Wallmann et al, 1988; McGuire et al, 1997).

6.1.1 Sea-level changes in the Quaternary

Most of our knowledge of sea-level change and its impact derives from the rapid and dramatic sea-level changes that occurred during the Quaternary - the past 1.65 million years - when successive glaciation-déglaciation cycles led to changes in global sea level of over 130 m (Shackleton, 1987). These large Quaternary sea-level fluctuations were primarily glacio-eustatic and can be broadly attributed to changes in global ice volume. At a regional scale, however, isostatic processes were also probably important, with the wholesale redistribution of planetary water resulting in loading and unloading of the lithosphere, particularly along continental margins and on glaciated continental terrain, sufficient to cause significant spatial variations in relative sea level. The rate of global eustatic Quaternary sea-level change, particularly at glacial terminations, was very rapid, and at times during o

the Holocene (the past ~10,000 years) may have reached several metres per millennium. Global eustatic sea-level changes over the past 140,000 years are relatively well constrained from the oxygen isotope record determined primarily from planktonic foraminifera (see Dawson, 1992, for more detailed discussion of oxygen isotopes and their application to the late Quaternary). At the same time, analysis of foraminifera in deep-sea sediment cores provides information on the history of global continental ice volume, thereby permitting the glacio-eustatic component of sea-level change to be determined (e.g. Shackleton and Opdyke, 1973). These high-resolution records of ocean volume changes represent, however, only a first approximation to variations in global sea level as they are not consistent with sea-level records derived from dated marine terraces in areas undergoing uniform crustal uplift (e.g. Bloom, et al, 1974; Dodge et al, 1983). Shackleton (1987; Fig. 6.1) attempted to resolve this situation by producing a general eustatic sea-level curve that presents a more detailed record of ocean volume changes than that from raised terraces. Discrepancies remain between the two records, but both do show the same general trends in sea level (Chappell and Shackleton, 1986). It is worth reiterating here, however, that although a global eustatic sea-level curve is a valid concept, owing to changes in the shape of the geoid (see Box 6.1), sea-level changes over the entire planet, during glaciation-déglaciation cycles, were neither of the same magnitude nor even the same sense. This was a consequence of the large-scale redistribution of planetary

Figure 6.1 The oxygen isotope global sea-level history compared with sea-level data estimated from marine terraces in New Guinea. (After Shackleton, 1987)

0 20 40 60 80 100 120 140

Thousands of years before present

Figure 6.1 The oxygen isotope global sea-level history compared with sea-level data estimated from marine terraces in New Guinea. (After Shackleton, 1987)

0 20 40 60 80 100 120 140

Thousands of years before present water, which had a profound effect on the topography of the geoid through the massive transfers of load from the land to ocean basin (and vice versa), loading and unloading effects on the upper mantle, and the growth and decay of ice masses (Tooley and Turner, 1995). The large variations in sea-level change that occurred during the Holocene have important implications for future global warming-related sea-level rise, which may not affect all parts of the globe equally.

Sea-level changes during the past 140,000 years (approximately corresponding to the late Quaternary, which started 130,000 years ago) are important because they provide a backdrop against which we can view the current period of rising sea levels. Over this time, sea-level behaviour can be subdivided into three broad categories (Box 6.2), each linked to the prevailing global climate and therefore also to ice-cap growth and decay. As might be expected, reconstructions of the timing and magnitude of sea-level changes are better constrained for the late Pleistocene and Holocene, a period covering the past 30,000 years. The early part of this time-span, from 30 ka to 18ka bp, is characterised by a large fall in sea level from between 45 and 80 m to a base level of around -120 m. The former figure is based upon oxygen isotope data (Shackleton, 1987) and the latter on the ages of marine terraces (Bloom et al., 1974). Depending on which starting level is accepted, a time-averaged rate in sea-level fall of between 3.75 and 6.55 mm/y is obtained.

Subsequent rises in sea level after 18ka bp coincide with progressive warming at the end of the last glaciation, and are dated by incremental drowning of coral reefs in the Caribbean-Atlantic region (Fig. 6.2). If we look more closely at time-averaged sea-level curves (e.g. Fairbanks, 1989), we find that they obscure the real nature of sea-level rise, which is actually episodic and step-like. Some of these rises are in fact revealed to be near-instantaneous, and Blanchon and Shaw (1995) report three metre-scale catastrophic rise events (CREs) in the postglacial record: a 13.5m rise dated at 14.2 ka bp, a 7.5-m rise at 11.5ka bp, and a 6.5-m rise as recently as 7.6 ka bp. Although sea level rose, on average, by around 10 m per millennium during the period of rapid post-glacial ice sheet melting, Blanchon and Shaw's reef-drowning studies from the Caribbean and the Atlantic suggest that the CREs may have involved an increase in sea level by as much as 13.5 m in less than 290 ± 50 years. These rates - of up to 46 mm a year - are overprinted on an annual general post-glacial rise of between 4 and 13 mm, and reflect major and rapid changes in global climate and geography. The 7.5-m rise at 11.5 ka bp, for example, coincided with the end of the Younger Dryas cold phase, when changing dust concentrations and snow accumulation rates suggest that global climate change might have

Box 6.2 Sea-level changes over the past 140,000 years

Dates are in thousands of years bp; sea-level changes are compared to current sea level.

Rapid rises in sea level at the end of major glacial stages (glacial terminations): The principal rises occurred from 140 to 128 (-130 m to 0 m) and from 18 to 7 (-125 m to +5m)

Stable sea levels during interglacial periods (warm phases between distinct glacial episodes): from

128 to 115, during interstadials (short periods of climate amelioration within otherwise cold phases) from 105 to 100,95.5 to 94,82 to 80,60 to 58 and 40 to 36, and during one stadial (glacial period) from 70 to 60.

Step-like falls in sea level during the last (Weichselian/Devensian) glacial period: with major fall between 115 and 110 (-8 m to -50 m), 80 and 74 (-27 m and -60 m), 72 and 70 (-50 m and -80 m) and 18 and 29 (-75 m to -120 m).

0 1 5 10 15 Age (thousands of years)

Figure 6.2 Sea-level curve for the past 18,000 years based upon the dating of submerged corals from the Barbados region of the Caribbean (after Fairbanks, 1989). Such time-averaged curves obscure the real nature of post-glacial sea-level rise, which is actually episodic and step-like.

occurred within 20-50 years (Mayewski et al, 1993; Alley et al, 1993). Others may reflect the breaching of continental meltwater lakes that then catastrophically emptied into the ocean. Blanchon and Shaw (1992) do, in fact, explain their CREs in terms of the spectacular release of meltwater megafloods from glacial and proglacial reservoirs, which led to unprecedented sea-level rises over very short periods of time. Discharge from Lake Agassiz in North America around 8000 years ago is estimated (Hillaire-Marcel et al, 1981), for example, to have caused a eustatic sea-level rise of between 20 and 42 cm. Such dramatic rises were succeeded in the late Holocene by a more steadily consistent rise in sea level, the rate of which also began to slow, first to 5.45 mm per year between 7000 and 4800 years ago, and since then to a time-averaged annual rate of 1.25 mm.

6.1.2 Recent changes in sea level

How then do recent changes in global sea level compare with the major variations observed in the post-glacial record? First, it must be remembered that sea-level change measured at a point on a coastline can reflect either vertical displacement of the land or the changing volume of ocean, or a combination of the two. Consequently, great care must be taken when interpreting the raw data in terms of rising sea levels on a global scale, and effects that are not a result of environmental change must be subtracted before a true record of absolute sea-level rise is achieved. These are many and varied, and occur on both a local and a regional scale. Examples of the former include the compaction of coastal sediments and the abstraction of water from coastal aquifers, while the latter include tectonic uplift and continuing post-glacial rebound of the crust. In order to obtain a meaningful global average for absolute sea-level change, the effects of such local or regional vertical movements of the crust must be removed from the observed record, and many such records from numerous geographical locations across the planet combined. The situation is further complicated if one wishes to extract that part of the sea-level rise signal that reflects anthropogenic global warming. This is because sea levels have been slowly rising since long before human activities could have had any effect on global temperatures.

Despite these difficulties, a number of different approaches have been utilised in order to constrain better estimates of sea-level rise over the past 150 years. These include (a) averaging all tide gauge observations from stable coastal locations, (b) determining the long-term late-Holocene rise from dated shorelines and subtracting this from the local tide gauge record, and (c) generating a geophysical model that simulates local tectonic and isostatic effects and subtracting this from the local tide gauge data. A number of estimates for recent sea-level rise have been obtained using such approaches, ranging from 1.8 ± O.lmm/y (Douglas, 1997) for the period from 1880 to 1980 to 2.4 ± 0.9mm/y (Peltier and Tushingham, 1989) for the twentieth century up to that date. These rates are significantly higher than those prior to the middle of the nineteenth century, for which some estimates (e.g. Douglas, 1995)

Figure 6.2 Sea-level curve for the past 18,000 years based upon the dating of submerged corals from the Barbados region of the Caribbean (after Fairbanks, 1989). Such time-averaged curves obscure the real nature of post-glacial sea-level rise, which is actually episodic and step-like.

suggest that the rise was either negligible or even indistinguishable from zero. Although it seems clear, then, that twentieth-century sea level rose probably more rapidly than at any time during the past 1000 years or more, so complex are the interdecadal and other variations in the sea-level record that a statistically significant acceleration remains to be identified (Douglas, 1992). This is a major part of the problem that has resulted in such disparate estimates of just how much sea level will have risen by the end of this century.

Continue reading here: Perspectives on future sealevel change

Was this article helpful?

0 0