Michael Beeson's Home Page

Changes in sea level since the last glaciation

22,000 years ago, the Earth was in a glacial period, and since glaciers are made of water, sea level was more than 130 meters lower than today. Since that time, the glaciers have been melting. It takes thousands of years for all the ice to melt, and we still have a lot of ice in Greenland and Antarctica, which could conceivably melt and add water to the oceans.

It is worth noting that floating ice does not affect sea level when it melts--floating objects displace water equal to their weight, so when they melt, that water is no longer displaced, but the same weight of water is added to the reservoir. (You can verify this in your kitchen.) Technically, since floating ice does not incorporate salt, it displaces salt water equal to its weight, so there is a small error in this argument, but it is not significant.

The melting was slow at first, but over the last 14,000 years, sea level has risen 130 meters, as shown in the following graph:

A more detailed explanation of the graph can be found here. Although it is interesting and not trivial how these numbers are reached, it doesn't seem to be at all controversial. Everyone agrees that this is what happened to sea level since the last glacial maximum. Note that North America was populated, apparently over a land bridge from Asia, ten to fifteen thousand years ago, when the lower sea level allowed for the existence of that land bridge.

The most recent 9000 years is shown in the following graph, in which the vertical scale allows for more detail:

A more detailed explanation of this graph can be found here.

It looks like the graph is leveling off, doesn't it? So what's going to happen next? In the past there have been many cycles of glaciation, separated by "interglacials" such as the present one. So it seems likely that there will be another glacial period, and sea levels will decrease. I can't find any evidence of disagreement about the future thousands of years ahead: another glacial period is coming. Big glaciers will cover northern lands as they did before. Humans will need technology and ingenuity if civilization is to survive. But that will take thousands of years, and for us it is a point of interest what the maximum will be before the sea level decreases again.

An obvious question is, what happened in previous interglacials? The Wikipedia article on Sea Level Rise states, without any references,

"During the previous interglacial about 120,000 years ago, sea level was for a short time about 6 m higher than today, as evidenced by wave-cut notches along cliffs in the Bahamas. There are also Pleistocene coral reefs left stranded about 3 metres above today's sea level along the southwestern coastline of West Caicos Island in the West Indies. These once-submerged reefs and nearby paleo-beach deposits are silent testimony that sea level spent enough time at that higher level to allow the reefs to grow (exactly where this extra sea water came from—Antarctica or Greenland—has not yet been determined). Similar evidence of geologically recent sea level positions is abundant around the world."

That conclusion is supported in the Report of the Working Group on Paleoclimatology of the IPCC, page 458, section 6.4.3.3, where we find "Direct sea level measurements based upon coastal sedimentary deposits and tropical coral sequences (e.g., in tectonically stable settings) have clearly established that eustatic [globally averaged] sea level was higher than present during this last interglacial by approximately 4 to 6 m." That section goes on to discuss ice core data, which seems to lead to the conclusion that the ice in Canada and sourthern Greenland melted completely, but some remained in northern Greenland, and that since this isn't enough to account for the sea level rise, some Antarctic ice must have melted too.

Hence: either

This is the longer-term context in which we should consider the much-debated question of what sea rise to expect by 2100.

If the ice on Greenland, or the ice on Antarctica, were to melt overnight, there would be a several-meters rise in sea level, flooding coastal areas and causing quite a bit of trouble for humans. These dramatic scenarios have stimulated both public and scientific interest in the question of how likely such a melt-off might be. Facts relevant to that question will be discussed on other pages (devoted to Arctic ice and Antarctic ice). Here we note only that at the present rate of melting, it would take many thousands of years; hence dramatic scenarios involve unforeseen accelerations of the meltoff due to unknown processes. That raises the question, whether in past interglacials, there was anything dramatic about the rate of change of sea level as the maximum sea level was approached.

Ian Plimer's book, Heaven and Earth, addresses this question on p. 312. Here is what he says:

"Ice core drilling in both Greenland and Antarctica has given a continuous record of some 800,000 years of past glacials and interglacials. In past interglacials when the temperature was at least 5 degrees C warmer than now for about 10,000 years, the polar ice caps did not completely melt. Sea level has been rising and falling by about 130 meters over the past 800,000 years of alternating glacials and interglacials. The ice sheets could not have melted, otherwise there would be no old ice to drill. Well before Greenland was covered with ice, Antarctica had a thick ice sheet which remained during a very warm period of 4 million years."

This paragraph from Plimer contains errors and misleading statements. First, a casual reading might seem to imply that both the Arctic and Antarctic ice caps did not melt during the previous interglacial period, while it seems that the Greenland ice core data only goes back to the beginning of the last glacial period, about 116,000 years ago. At present I don't know for certain whether that is because that is the bottom of the ice cap, which would imply that the ice cap melted in the previous interglacial, or because that was just as deep as could be successfully drilled. Second, the Antarctic ice data goes back only 420,000 years, according to Muller et al. (who are writing a book about the Ice Ages) rather than the 800,000 mentioned by Plimer. Third, if there were interglacial periods when the temperature was 5 degrees warmer than now for 10,000 years, they don't show up in the Antarctic ice record (see the discussion on this page. However, the data does show three periods when the temperature was about two degrees warmer than 1950 (about the same temperature as today) for several thousand years. This claim by Plimer is important, because it would imply that we don't need to worry about the Antarctica ice cap melting from a few hundred years at a few degrees warmer than now. But although Plimer's book has many footnotes, there are no footnotes to support these apparently erroneous claims. Technically, Plimer's paragraph doesn't say that the periods when the temperature was 5 degrees warmer than now occurred in the last 420,000 years, but Muller et al. also contains a temperature graph (derived from oxygen isotope ratios in sea floor sediments) going back 3 million years, and only near the beginning of that period were temperatures as much as five degrees warmer than 1950.