Sea level is obviously directly related to extremes of climate change. As a gross example, at the beginning of the last deglaciation 18,000 years ago sea level was everywhere about 100 meters lower than it is now. The more subtle question is, can we observe future small (fractional mm) changes of sea level quickly enough, and also interpret them, to provide a useful index of smaller but important changes of climate?
Using estimates of eustatic sea level rise as an indicator of climate change faces the difficulty that sea level rise is an output combining many individual effects. Some of these effects can offset others, so that the exact response of global sea level to climate change remains somewhat uncertain. It is necessary to work out a hydrological and geophysical budget for the various contributors to local and global sea level change. For example, global warming will cause both expansion of the ocean and changes of circulation [ Church et al., 1991]. In addition, the melting of small glaciers, while difficult to quantify, is also significant [ Meier, 1984]. These two effects together can account for about one mm per year of sea level rise over the last century, assuming that a global warming of about 0.5 degree C has occurred.
Contributions of the Greenland and Antarctic ice sheets can play a different role. In the case of Greenland, Zwally et al. [1989] claimed that satellite altimeter data showed a thickening of ice there from 1978-87 equivalent to a fall of sea level of a few tenths of a mm per year. If this scenario held true over an extended time, then a possible increase of eustatic sea level rise from global warming, for example from thermal expansion, could be offset by increased storage of water in the form of ice on Greenland because of increased precipitation there. Douglas et al. [1990] vigorously disputed the adequacy of the satellite altimeter data analysis technique used by Zwally et al. [op. cit.], and further calculated that the alleged increase of ice thickness (about 20 cm per year) would have changed the angular momentum of the earth in a manner that was not in fact observed. Van der Veen [1993] has also disputed the Zwally et al. [1989] result on purely glaciological grounds. But the basic issue remains. If global warming causes increased precipitation at very high latitudes with concomitant storage of water in the form of ice, sea level rise due to thermal expansion of the ocean or melting of small glaciers could be offset to a greater or lesser extent.
One must also consider the matter of water storage in artificial reservoirs that would otherwise have flowed into the oceans. Chao [1991] calculated that the increasing storage of water in both large and small above-ground reservoirs was equal to a fall of global sea level of 0.7 mm per year over the last 40 years. Since this figure must be added to the current eustatic rate to accurately reflect the real situation, an unexpectedly large source of water must be found to account for it. Sahagian, et al., [1994] found a smaller effect, but ignored the very large contribution of small reservoirs.
From these considerations it is clear that simply obtaining a value for global sea level rise in the past, or detecting an increase in the future, is not enough for sea level rise to serve as an unambiguous indicator of global climate change. Global sea level, whether observed to increase, stay the same, or decrease, must be analyzed and understood in terms of all of the factors that affect it for meaning to be attached to it.