The Sea Also Rises

by Christina Hulbe

It is estimated that 1.2 billion people live within 100 kilometers of a coastline and 100 meters of mean sea level in communities whose average densities are nearly 3 times larger than the global average. Of those, more than 100 million live within 1 meter of sea level. One of the more dramatic graphics in Al Gore's movie, An Inconvenient Truth, involved the flooding of Manhattan due to sea level rise in the event of collapse of the Greenland or West Antarctic Ice Sheet. You've probably seen similar graphics of flooded Florida and flooded Bangladesh. So how likely are these Waterworld-lite scenarios? Good question.

Global mean sea level is currently rising at a rate of 3 +/- 0.5 millimeters per year, a value that is larger than the 20th century mean, 1.7 +/- 0.3 millimeters per year. That is, the rate of sea level rise appears to be increasing. The recent rates (since 1992) are calculated using satellite observations of sea surface height while calculations for earlier times rely on a global network of tide gauges. The total observed increase in sea level over the 20th century is 0.15 to 0.2 meters.

Causes of sea level change
In order to put those numbers into context, we need to start with some framing concepts. First, we are interested in the mean sea surface height measured with respect to a fixed reference. Second, changes in sea level can have both local and global causes. Locally, the elevation of the land surface may be changing. For example, subsidence yields an apparent local rise in sea level. Globally, sea level changes are related to changes in volume, of either the ocean basins or the water contained within them. Changes in sea level due to changes in volume are referred to as eustatic.

Ocean basin volume is determined primarily by plate tectonics (more specifically, by seafloor spreading) and secondarily by sedimentation in ocean basins. For example, a large-scale increase in the rate of mid-ocean ridge growth would displace ocean water onto the land surface, causing sea level to rise. Such changes transpire over millions of years and are not of interest here.

In the context of the ongoing global warming and its cultural impacts, there are two important sources of eustatic sea level rise: thermal expansion of ocean water and mass loss from glaciers and ice sheets ( see the USGS glacier glossary for definitions and great pictures). Changes in the volume of water stored in aquifers and surface reservoirs can play a modest role as well.

Thermal expansion
All materials experience volume changes as they are heated or cooled. If you have a tank-style water heater at home, it was built to accommodate thermal expansion: 50 gallons of water at 40 degrees Fahrenheit (4.4 degrees Celsius) will expand to about 51 gallons when heated to 140 degrees Fahrenheit (60 degrees Celsius). The relationship between water density and water temperature is not completely simple, water reaches its maximum density at 4 degrees Celsius, but that's another story. The upper few 100 meters of the global ocean are indeed warming and the resulting thermal expansion is estimated to account for between 1/4 and 1/3 of the total observed sea level rise.

Climate models used to simulate 20th century climate change and accompanying sea level rise can account for 3 to 5 of the observed 15 to 20 centimeter total. These models consider only thermal expansion of ocean and not contributions from melting glacier ice. The difference between modeled and observed sea level rise must be the result of model under-estimation of ocean heat uptake and associated thermal expansion, melting of glacier ice unaccounted for in the models, or a combination of the two.

The cryosphere
Glaciers and ice sheets hold 75% of Earth's fresh water, the equivalent of about 75 meters of sea level. Most of that water, about 90%, is stored in the Antarctic ice sheets. Most of the rest, equivalent to about 7 meters of global sea level, is stored in the Greenland Ice Sheet. While modest in their total mass, equivalent to about 0.65 meters of sea level, mountain glaciers are an important part of the story because, globally, they are retreating at a rapid pace.

Estimating the change in mass of the world's glaciers is difficult because only a small fraction of them have been measured in detail. Happily, advances in satellite remote sensing (for example, NASA's ICESat mission) mean that estimates of glacier change are improving all the time. The best current estimate of sea level rise due to glacier retreat of which I am aware is 0.51 millimeters per year over the period from 1961 (when sufficient data become available) to 2003. As with many other global-warming related processes, the rate appears to be increasing over time (pdf of a technical paper).

If the mean 1961 to 2003 glacier wastage rate applies throughout the 20th century, mountain glaciers are the source for something like 0.05 meters of the observed 0.15 to .2 meters of sea level rise. Not too shabby but if we add this to the upper bound computed for thermal expansion, we still only account for 1/2 to 2/3 of the observed rise. Some of that missing water must be from the Greenland Ice Sheet but significant mass loss there is a very recent phenomenon (more in a future post). There are a couple of hot spots for change in Antarctica as well but probably not enough to make up the difference (more in a future post). How do we resolve this? The models may be underestimating ocean heat uptake and thermal expansion. The glacier guys may be underestimating the mass wastage from the world's mountain glaciers. Were I a betting woman, I'd bet on both.

So where do we go from here?
We are committed to additional warming and sea level rise even if we stopped burning fossil carbon tomorrow. This is because the climate system is still adjusting to our 20th century fossil carbon emissions and will continue to do so for hundreds of years (pdf of a technical paper).

Climate modelers at the National Center for Atmospheric Research conducted a set of model experiments recently to estimate the range of future warming and sea level rise to which we have committed ourselves via 20th century fossil fuel use. Supposing CO2 emissions had stopped entirely in the year 2000, we were already committed to at least an additional 0.08 to 0.1 meters of sea level rise just due to thermal expansion (the models do not account for changes in glaciers and ice sheets). The study also considered several future emissions scenarios. Supposing moderate anthropogenic contributions to atmospheric CO2 over the 21st century and an end to emissions at that point, the sea level rise commitment in these experiments is 0.2 to 0.26 meters by the year 2100 and 0.36 to .48 meters by 2200. For "business as usual" types of CO2 scenarios, the numbers would be larger. Again, these models consider only thermal expansion.

What about the contribution from glaciers and ice sheets? This is a difficult variable to estimate because these systems don't simply grow and shrink according to how much snow they receive in the winter and how much ice they lose to melting in the summer. They are dynamic fluid flows that respond to climate (and other) perturbations on time scales that vary according to their size, temperature, and geographic setting. In general, the larger the ice mass, the longer the time scale for response to climate change (centuries to millennia) but we are learning that even the great ice sheets can respond quickly (decades) to the right perturbations, at least around their margins (for example, the speedup of Greenland outlet glaciers). But there's a lot left to understand about the physical processes and time scales over which rapid events transpire (it makes a big difference if the speedup in Greenland last for a decade or for several). With perhaps a few exceptions, mountain glaciers are going to continue to retreat and contribute to sea level rise in the decades ahead. What happens to larger ice masses is much less clear but current trends in Greenland are certainly cause for concern.

For now, let's make a simple guess at future sea level rise by assuming that the 20th century can be used as an analog for the 21st. If the present-day relative contributions from thermal expansion and glacier melt hold in the future, then we need to multiply the climate model projections reported above by at least 3 to estimate a total sea level rise through the 21st century. This will give us the thermal expansion commitment we make by continuing to burn fossil fuels and a perhaps conservative contribution from the cryosphere. For the moderate CO2 emissions scenario reported above, the total rise would be 0.6 to 0.78 meters. That's within splashing distance of 100 million people.