Driving the seemingly endless climate-treaty negotiations, the most widely feared consequence of Global Warming appears to be a catastrophic rise in sea level (SLR). Environmental advocacy groups are filling the airwaves with lurid images of flooding of Bangladesh and Pacific islands, and raising the specter of hundreds of millions of environmental refugees demanding care and compensation.
Even sober scientists, while not endorsing such obvious scare stories, predict an acceleration of the ongoing global rise, which a system of tidal gauges places at about 18 cm (7 inches) during the 20th century. Economists concerned with trying to estimate a 'social cost' of carbon-dioxide emissions predict huge economic losses from future SLR. Not surprisingly, insurance companies, looking to raise premiums, are cheering them on.
However, more detailed analyses of actual observations suggest an opposite outcome: A climate warming might even slow down SLR -- rather than accelerate it. To understand this counter-intuitive result, one must first get rid of false leads -- just as in a detective story. The misleading argument here is the oft-quoted statement that the climate warmed by 1degF (0.6 C) in the last 100 years and that SL rose by 18 cm. Both parts of the statement may well be true; but the second part does not necessarily follow from the first.
Curiously, Barack Obama predicted a deceleration of SLR when he accepted his party's nomination in 2008: "This was the moment when the rise of the oceans began to slow, and our planet began to heal." Some tidal-gauge data do show deceleration, but starting in 1960. Hey, wasn't that the year during which Obama was conceived?
Sea Level Since Last Glacial Maximum (source)
(Toscano & Macintyre 2003; Fairbanks 1989)
SLR data problems
The principal SLR data have come from tidal gauges, which measure not only tides but storms and everything else. And from these measurements one extracts a steady rise in local sea level. There are about two dozen stations in the world with long-enough records dating back to the early 1900s, which have been used by the international tidal gauge network, located in Liverpool, England.
Global Sea Level 1900 - 1980 (source)
(Trupin & Wahr 1990)
Since 1993, we have also had satellite observations; but these have been plagued with various types of uncertainties -- although in principle, satellites can measure absolute sea level independent of any vertical motion of the coastal land surface. The tidal stations are subject to various corrections as well: they measure relative sea level with respect to the station which is fixed to the land. Since the melting of glacial ice cover from Northern continents several millennia ago, the land surface has rebounded in these places -- a process called 'isostatic adjustment.' But at the same time also, many tidal stations have been sinking -- as the coastal land subsided because of the depletion of groundwater, of oil and gas, and of other processes that led to the compaction of sediments.
It is clear that satellites have an inherent advantage over tidal stations but their figures don't match up. From data gathered by the GRACE satellite system, we can also factor in detailed measurements of local gravity changes but the record is too short to draw firm conclusions. With estimates of past SLR all over the place, how does one proceed?
Leading researcher Bruce Douglas terms SLR a "puzzle" (Physics Today March 2003), while famed Scripps Institution oceanographer Walter Munk calls it an "enigma" (ProcNatlAcadSci 2004). Maybe we should use Churchill's description of Soviet Russia: "A riddle wrapped in a mystery inside an enigma"
The difficulty with projections of sea level rise is nicely illustrated by the IPCC. The estimates of its first assessment report (1990) showed a range of 10-367 cm for sea level in 2100. The second report, published in 1996, narrowed the range to 3-124 cm. Its third report, published in 2001, showed 11-77 cm. The fourth assessment report, published in 2007, showed 14-43 cm in its draft form but changed it to 18-59 cm in the final printed version. As can be seen, the maximum SLR decreased successively as estimates improved. All these IPCC projections are very much smaller than the extreme values of about 600 cm by activist-scientist James Hansen (and by climate multi-millionaire Al Gore) -- which assume excessive and rapid melting of the Greenland icecap.
Seal Level Rise to 2100 (source)
This narrowing of estimates by the IPCC has caused great concern among alarmists who feared that the IPCC was being "too conservative." Probably as a result of this peer-pressure, estimates have now increased -- as will be seen in the fifth assessment report, due in September 2013. As a reviewer of IPCC reports, I have been able to look at the "second order draft," which was recently leaked to the press. It gives values for 2100 of 45-110 cm (16-40 inches) - about double what IPCC estimated just six years ago in their fourth report. (There is no guarantee that these values will survive in the final printed version.) Still, they are very much smaller than some of the extreme estimates that have been written up in newspapers and magazines -- and always blamed on Global Warming (GW) from carbon dioxide, released in the burning of fossil fuels.
There are many problems with the basic SLR data, with no easy resolution. For example, the forthcoming (2013) IPCC report shows zero values before 1880 (presumably based on corals), while other coral data and coastal sediments show positive values. Tidal gauge data show no acceleration during the strong warming of 1920-40, and continue to rise during the slight cooling of 1940-75 and during the "pause" in warming of the past 17 years. However, IPCC-2013 shows increasing values (acceleration) for SLR during the same no-warming period -- and may already have been falsified.
No theory for SLR
There is no overall theory of SLR, encompassing thermal expansion of the oceans, melting of mountain glaciers, and changes, both positive and negative, of Greenland and Antarctic ice sheets. (One may ignore to first approximation the "mining" of fossil groundwater and accumulation of water in reservoirs. Of course, changes in floating sea ice do not affect SLR). A German oceanographer-activist, based in Potsdam, has proposed a "semi-empirical" theory under which SLR is related to sea surface temperature -- and thus to atmospheric CO2 levels (if one accepts the existence of appreciable climate sensitivity). But his theory has no theoretical foundation whatsoever and also disagrees strongly with all observations.
The first clue that there might be something amiss with the logic is hidden in the IPCC report itself. According to their 1996 compilation of data, the contributions to SLR of the past century come mainly from three sources: (i) Thermal expansion of the warming ocean contributed about 4 cm; (ii) the melting of continental glaciers about 3.5 cm. (iii) The Polar Regions, on the other hand, produced a net lowering of SL, mostly from ice accumulation on the Antarctic continent. (The mechanism is intuitively easy to understand but difficult to calculate: A warming ocean evaporates more water, and some of it rains out in the Polar Regions, thus transferring water from the ocean to the polar ice caps.) The surprising result: When one simply adds up all these three contributions (neglecting their large uncertainties), they account for only about 20 percent of the observed rise of 18 cm. The climate warming since 1900 cannot be the cause of the SLR; something is missing here.
But if, as surmised from the absence of observed acceleration during 1920-40, ice accumulation roughly balances ocean thermal expansion and contributions from melting mountain glaciers, why then is SL rising at all? Another riddle requiring a solution.
Why not zero SLR?
The relevant clue comes from corals and from geological observations: It seems that SL has been rising for the past centuries at about the same rate as seen by tidal gauges in the last 100 years -- about 18 cm per century. In other words, SL was rising even during the colder Little Ice age, from about 1400 to 1850 AD. This provides further support for the hypothesis that the observed global SLR since 1900 is reasonably independent of the observed temperature rise.
The explanation for this riddle had been suspected for some time, based on historic data of SLR derived independently from measurements of coral growth and from isotope determinations of ice volume. But the picture was filled in only more recently through estimates of the rate of melting of the West Antarctic Ice Sheet (WAIS), by tracing its shrinkage during past millennia (through the receding position of its "grounding line," i.e., the line of contact of the ice sheet with the underlying continental land mass.) Note that the WAIS is not floating sea ice; like a mountain glacier, its melting contributes water to the global oceans.
We can therefore describe the broad scenario as follows: The strong temperature increase that followed the Last Glacial Maximum (LGM) of about 18,000 years ago has melted enough ice to raise SL by 120 meters (400 feet). The rate of rise was quite rapid at first and controlled by the melting of the huge ice sheets covering North America and the Eurasian land mass. These disappeared about 8000-5000 years ago; but the WAIS continued to melt, albeit at a much lower rate -- and it is still melting at about the same rate today. Other, smaller WAIS-like ice sheets may have existed in the Antarctic, but have already melted away.
The principal conclusion is that this melting will continue for another 7000 years or so, until the WAIS disappears -- unless another ice age takes over before then. Moreover, there is nothing that we can do to stop this future sea level rise! It is as inevitable as the ocean tides -- as long as the Holocene (the present warm interglacial period) survives. Fortunately, coral reefs will continue to grow, as they have in the past, to keep up with SL rise. The rest of us will just have to adapt -- as our ancestors did some 10,000 years ago. At least, we are better equipped to deal with environmental changes.
A final note
What about the effects of putative human-induced global warming on SLR? Will it really increase the rate above its natural value, as predicted by the IPCC? We do have a handle on this question by observing what happened when the climate warmed sharply between 1920 and 1940, before cooling between 1940 and 1975. The answer, first noted in 1997, is quite surprising and could not have been derived from theory or from mathematical models. The data seem to show that SLR slowed down slightly when the climate warmed, and then accelerated when the climate cooled. Evidently, ocean-water thermal expansion and mountain-glacier melting were less important than ice accumulation on the Antarctic continent (which lowers SL). Unfortunately, the SL data are not precise enough to withstand scientific challenge -- and reliable data on ice accumulation over the whole Antarctic continent have not been available.
We can now try to answer our original question: Can a Global Warming really lower sea level rise? It all depends on the time-scale: Yes -- if GW lasts only for some decades or less. No -- if warmer temperatures persist for millennia, the WAIS melting rate would increase -- and so would SLR.
By analogy, a future warming produced, putatively, by an increase in greenhouse gases would give the same result: i.e., reduce the rate of rise of sea level. This is not a recommendation to burn more coal in order to save Venice from drowning. It is a modest appeal to politicians to take note of new scientific developments and recognize that the drastic limits on energy use called for by climate-treaty negotiators will not stop the rising seas.
NB: This essay ignores many less important features of global SLR, such as the "mining" of groundwater and construction of dams. It also ignores important regional and local effects that depend on isostatic adjustments, ocean currents and wind patterns, land subsidence, etc. Efforts are underway to harmonize conflicting data from tidal gauges and from direct measurements of the ocean surface by satellites,