Feature

Are the Waters Rising?

Are we seeing ocean levels rise as a result of global warming or geological processes?

Tim Stern

Sea level change is a phenomenon that has far-reaching consequences for many societies, particularly those that live in coastal regions. It is a subject that has provoked much discussion in both scientific and lay circles, yet it is a complex process that is not always well understood.

It has become of particular interest in recent years because of its association with the greenhouse effect, where the discharge of man-made gases is causing gradual warming of the Earth's atmosphere and oceans due. Sea level rises associated with the greenhouse effect are due to two related phenomena: as the sea warms it expands and, secondly, as the air temperature rises, more ice melts from the Antarctic and Greenland ice cap and the volume of the water in the oceans increases.

A quick glance through the talk summaries of the Greenhouse 94 meeting reveals that both government and private research groups are interested in exploring the link between climate and sea level change. However it is still not widely appreciated that there is more driving sea level than merely climatic effects. These "natural" effects need to be fully understood before observed sea level changes can be ascribed to anthropologic factors.

On the Rebound

Foremost of the natural processes contributing to sea level change is the effect of an Earth that is actively "rebounding " from two or three giant ice caps that were present on the Earth just a few thousand years ago. There is clear geological evidence that ice caps one to three kilometres thick and 1,000-2,000 km in diameter existed over Canada (centred on Hudson Bay) and Scandinavia just 20,000 years ago -- this is very short time on a geological time-scale. A large amount of extra ice is also thought to have been on Antarctica, but this is harder to establish from geological evidence. By 5,000 years ago, most of the ice from these ice caps had melted. It is the manner and rate at which the Earth and oceans respond to this ice melting, and the effect that this has on changes in present day sea level, that is of interest here.

Rocks at the surface of the earth appear solid enough -- hit a rock with a hammer and it will smash or chip into small bits. However, if rocks are subjected to a large pressure for very long periods (i.e. hundreds to millions of years) they will flow like treacle.

Similarly, my 4-year-old daughter came home from kindy the other day with a ball of silly putty. She excitedly showed me how one could bounce a ball of this material on the kitchen floor, yet when in her hand she could squeeze the putty into any shape she liked. This is analogous to how the Earth operates: bouncing the ball on the floor is a short term "load" and the silly putty responds elastically and bounces back, but slowly squeeze the putty (a long term "load") and it responds by flowing in a plastic manner.

Thus as ice caps build up and weigh down the Earth's surface, parts of the Earth at depths of about 100 km, called the asthenosphere (from the Greek word athenes meaning no strength), respond by flowing out from underneath the ice to create a saucer shaped dent in the Earth's surface into which the ice cap sits.

This does not occur instantaneously, however. The asthenosphere is viscous -- that is, it is a fluid with internal friction -- so that there is a characteristic time taken for the asthenosphere to flow when subjected to a force. Just as the asthenosphere takes a specific time to flow away from under a newly developed ice cap, the reverse also applies -- when the ice melts, a dent is observed in the Earth's surface and it takes some time for the asthenosphere to flow back so the dent will pop out again. As the land beneath former ice sheets slowly rises back to sea level, a series of beaches are cut around the periphery of the depression.

An analysis of raised beaches around Hudson Bay suggests there has been a maximum of 120 metres of uplift in the centre of Bay, while the region of the peripheral bulge, which extends from the Canada-US border to the Florida Keys, has undergone up to 10 m of subsidence.

Dating of such raised beaches suggests that it takes about 10-20,000 years for the asthenosphere to completely flow back once a load has been removed from the Earth's surface. This time-delay is then the critical point when trying to understand why present day sea level changes are still partially linked to rebound from the melting of large ice caps that ended their melting about 5,000 years ago.

From Ice to Water

There are three processes from ice cap melting that contribute to changing sea level as measured today:

(1) The melting ice adds extra water to the oceans and hence this raises mean sea level -- the Antarctic ice cap is volumetrically equivalent to 70 m of sea level change over the whole globe.

(2) As ice disappears from the continent, asthenosphere material flows back beneath the dent left by the ice and generally raises the surface of the solid Earth. This has the effect of making sea level appear to drop in the area of rebound. But as the asthenosphere flows in to fill the dent, the areas peripheral to the ice cap will suffer a corresponding subsidence. So for these areas that are coastal, it appears as though the sea level is rising.

(3) As melt water from the ice caps enters the oceans, the extra water pushes down on the ocean floor. This in turn drives the viscous asthenosphere outwards from under the oceans towards the continents, thus pushing the continents back up.

Clearly the final pattern of uplift and subsidence right around the whole globe is complicated both in space and time -- there is no stable reference point anywhere on the globe to measure true sea level change. Instead, solving the problem requires making many observations of "apparent" sea level changes around the globe and comparing these observations with a complex computer model that simulates the changing loads of ice and water, and the induced motion of the viscous asthenosphere, over the globe. Critical features of the model that must be known, or at least estimated, are:

• the distribution and dimensions of past ice sheets
• when they started and finished melting
• the depth and distribution of the oceans
• the viscosity ( or "stiffness") of the Earth's asthenosphere

Two leading earth science institutions are prominent in trying to solve this problem. One team is under the leadership of Professor Peltier at the University of Toronto and the other is Professor Lambeck's group at the Australian National University in Canberra. Although there is a difference in detail between the conclusions of the two groups, the one aspect they do agree on is the necessity to have melting of southern hemisphere ice within the past 10,000 years in their models.

Evidence from the Earth

Antarctica is the obvious place for such ice to have originated. And although there is some evidence suggesting that the Antarctic ice sheet was indeed larger than at present, the case is far from unequivocal. Peltier and his colleagues recently published a map of the Earth showing, for every shore line around the globe, the predicted rate of sea level rise or fall for the present day based on his best estimate model for post-glacial rebound.

Large apparent sea level falls are predicted in the regions peripheral to the main ice sheets, such as the eastern United States sea board. This prediction is consistent with observations of swampy areas along the eastern seaboard of the United States from Chesapeake Bay in the Baltimore area to the Florida Keys -- a 2,000km-long zone that is peripheral to the old ice sheet that used to be centred on Hudson Bay in Canada.

Another area of serious subsidence and flooding that could be linked to post-glacial processes is Holland, which is peripheral to the Scandinavian ice sheet. Here, models predict that Holland and parts of northern Germany and the eastern UK should have suffered 10m of subsidence in the last 6,000 years. In distant regions, many thousands of kilometres away (such as the South Pacific), sea level drops of the order of 1-3 mm/y are predicted.

There are two important points in the data and model produced. Firstly, once corrected for the predictions of sea level change, the observed global pattern of sea level change becomes geographically more consistent. This suggests that the model is, at least to the first order, reasonable. The second, and probably the most crucial, aspect of the model results is that once the global observations of sea level rise are corrected for the effects of post-glacial rebound, there appears to be a global rise of 2.0-11.0 mm/y. This rate is higher than estimates based on data that are uncorrected for post-glacial rebound.

Or, in other words, the present day response of post-glacial rebound is disguising the true global rise in sea level. This conclusion is, however, a tentative one and it needs to be emphasised that the model results are sensitive to modest changes in the analysis procedure.

In conclusion, the contribution of post glacial rebound to sea level change is a "first-order" environmental problem, and needs to be understood before we can meaningfully interpret reported sea level changes. From our present state, post-glacial rebound effects may be disguising a global sea rise that is twice what is apparent from direct measurements. This finding needs to be checked by refining our model and searching for better constraints on post-glacial movements of the Earth.

Perhaps the most important region to continue this research is on our own back door -- Antarctica. In particular the geological evidence from Antarctica of an enlarged ice sheet in the last 20,000 years should be a principal research goal.

Tim Stern is Senior Lecturer in Geophysics at the Research School of Earth Sciences, Victoria University of Wellington.