Sea levels are going to rise

The amount of carbon dioxide in the Earth’s atmosphere directly and indirectly determines the sea level. The more CO2 the higher the sea level. The details matter, the mechanism is complex, and as CO2 levels change, it takes an unknown amount of time for the sea level to catch up.

The present day level of CO2 is just over 400ppm (parts per million). For thousands of years prior to humans having a large effect on this number, the level of atmospheric CO2 was closer to 250. Human release of CO2 into the atmosphere by the burning of fossil fuel, and other human activities, are responsible for this difference. We expect the atmospheric concentration of CO2 to rise considerably by the end of the century. It is remotely possible that by 2100, CO2 will be about where it is now, but only if a significant effort is made to curtail its release. If nothing is done about the release of CO2 by human burning, the number will exceed 1000ppm by 2100. Reasonable estimates assuming the most likely level of effort to change the energy system put CO2 at somewhere around 600 to 700ppm by the end of the century.

So, it is reasonable to ask the question, what is the ultimate sea level likely to be with atmospheric concentrations of CO2 between 500 and 700ppm?

How long will sea level rise take?

Once the CO2 is in the atmosphere, it stays there for a very long time. So, if we curtail emissions and manage to keep the atmospheric CO2 between 500 and 700, that number will not drop for centuries. So, again, we have to expect sea levels to rise to whatever level is “normal” for an atmospheric concentration of CO2 between 500 and 700ppm. That is a conservative and perhaps even hopeful estimate.

A fair amount of research (but not nearly enough) has been produced over the last two or three years with the aim of estimating how fast and to what degree the major glaciers of the earth (in Greenland and Antarctic) will melt with global warming. In the long view, all of this research is irrelevant. The simple fact is that with higher CO2 levels, a lot of that ice will ultimately melt, and sea levels will ultimately go up. But in the short and medium term, that research is some of the most important research being done in climate change today, because it will lead to an understanding of the time frame for this rise in sea level.

None of the research I have seen satisfactorily estimate this rate, but with each new research project, we have a better idea of the process of deterioration of major glaciers. As this research progresses, the glaciers themselves are actually deteriorating, but they are only beginning to do so. As the research advances, we will get a better theoretical model for glacial deterioration. As the glaciers deteriorate we will have increasing opportunity to calibrate and test the models. I expect that in a few years from now (ten or twenty?) there will be active competition in the research world between theoretically based models and empirical observations to provide rate estimates for sea level rise. But at present we have mainly theory (the observational data is important but insufficient) and the theory is too vague.

A new paper came out this week that explores the process of deterioration of a major part of the Antarctic glacial mass. I’ll summarize this research below, but the main point of this post is to put all of this recent research in the context outlined in the first few paragraphs above. How much sea level rise can we ultimately expect, even if we have no good idea of when it will happen?

It may not matter how fast sea level rises

Uncertainty about the time frame for glacial melt is important for all sorts of practical reasons, but an interesting aspect of human culture and economy obviates that uncertainty, and does so rather ironically. In our economic system, we value things in many different ways. There are things that have great value in part because they are fleeting, rare, or ephemeral. People pay a lot for a great meal that is gone in 20 minutes, a random act of erotic pleasure, two minutes of terror on a carnival ride, or a small pile of white powder.

But we also pay good money for things because of their long term value. A classic problem in economics asks why a man (it’s always a man) in, say, Egypt, is willing to plant and tend hundreds of date tree saplings, knowing full well that the first fruit will not be provided until long after his own death. The reason, of course, is that his son will inherit these trees. Of course, his son is not likely to gain much from these trees because they will still be young and small. So the value of this grove of trees to his son is based mostly on the value to the son’s son. And so on.

This is how we place value on real estate. The main reason that a home you might consider buying today is of a certain value is that you can sell it for a similar or greater value in the future. The fact that short term fluctuations may destroy a good part of that value over the next decade does not obviate the longer term value of the property.

Some of the most valuable property in the US and in many other countries is within spitting distance of the ocean. The ocean itself adds this extra value either as commercial or industrial space, or as high-end domestic or tourist space.

If sea level rise sufficient to destroy that property was imminent, so the property would be destroyed this year, then the value of that property would be zero. But considering that the value of the property is always based in large part on the future sale value, then sea level rise sufficient to destroy it, but that won’t happen for a century, is sufficient to destroy that long term component. If you acquire property today that will eventually be flooded by the sea, you might think you own it. But really, you are renting it. When the sea rises up and inundates the lot, the lease is up.

So, in a way, it does not matter how long it will take the sea to rise, say, 10 meters. Any property that would be destroyed with sea level rise is, right now, worth less than a market ignorant of this inevitability would price it.

Why is it hard to estimate the rate of sea level rise?

Most of the ice in Antarctica is sitting on the interior of the continent, well away from the sea. But much of this ice is held in large catchments, or valleys, that have outlets to the sea. Those outlets are plugged with huge masses of ice, and that ice is, in turn, held in place by grounding lines, where part of the ice sits tenuously on the bedrock below sea level.

Behind the groundling lines, upstream, are valleys of various depths, but deeper than the point of grounding itself. It is thought that if the ice sitting on the grounding line falls apart, the plug of ice will deteriorate fairly quickly until a new grounding line, perhaps many kilometers upstream, is established. But that grounding line may be subject to deterioration as well, and eventually, the outlet valley that connects the interior catchment to the sea becomes open water, and the ice in that catchment can also deteriorate, and fall into the sea mainly in the form of ice bergs regularly calved off the glacier, like we see today in Greenland.

The main cause of global warming induced melting of the ice near the groundling line is warm water. The surface of the Antarctic glaciers does not melt very much from warm air, because the air over the southern continent is rarely above freezing. However, with global warming, we expect air temperatures to go above freezing more commonly. This would contribute to thinning of the ice over the grounding lines, and thus, more rapid breakdown of the plug holding most of the ice in place.

It is thought that when a grounding line fails, and the plug of ice begins to move into the ocean, steep cliffs are formed alongside the deteriorating ice. This would cause the ice behind the cliffs to destabilize, causing ice bergs to form at a very high rate. Also, the interior ice, in the large catchments, is generally thought to be unstable, so when Antarctic glaciers reach a certain point of deterioration, those glacial masses may deteriorate fairly quickly.

Each of these steps in glacial deterioration is very difficult to model or predict, as these phenomena have never been directly observed and the process involves so many difficult to measure mechanical and catastrophic events.

Upstream from the grounding line, though the horizontally flowing glacial masses in the plug, and up into the catchments, the sub-ice topography is complex and will likely control, by speeding up or slowing down, glacial deterioration. It is thought that many of the glacial masses that make up Antarctica’s ice have melted and refrozen numerous times, and glacial ice has moved towards the sea again and again, over the last several million years. As glacial ice moves along it carves out valleys or deposits sediments in a complex pattern, which then determines subsequent patterns of ice formation or deterioration. It is reasonable to assume that each time the glaciers melt and reform, the terrain under the ice becomes, on average, more efficient in allowing the movement of ice towards the sea. Thus, any estimate of the rate of glacial movement and deterioration based on past events is probably something of an underestimate of future events.

This has all happened before

We know that the world’s glaciers have melted and reformed numerous times from several sources of evidence, and that this has been the major control of global sea level, as water alternates between being trapped in glacial ice and being in the oceans.

We know that global sea levels have gone up and down numerous times, because we see direct evidence of ancient shorelines above current shorelines, and we have direct evidence that vast areas of the sea have been exposed when glaciers were at maximum size.

We can also track glacial growth and melting by using oxygen isotopes that differ in mass. Glaciers tend to be made of water that contains a relatively high fraction of light oxygen, while the ocean water tends to have relatively more heavy oxygen. This is because water molecules with heavy oxygen are slightly less likely to evaporate, so precipitation tends to be be light. Glaciers are ultimately made of precipitation. There are organisms that live in the sea that incorporate oxygen from sea water into their structures, which are then preserved, and can be recovered from drilling in the deep sea. By measuring the relative amount of heavy vs. light oxygen in these fossils, controlling for depth in the sea cores, and dating the cores at various depths, we can generate a “delta–18” (short for “Difference between Oxygen–18 and Oxygen–16”) curve. This is an indirect but very accurate measure of how much of the world’s free water is stuck in glaciers at any given moment in time.

The Delta–18 curve for Earth for the last ca 800,000 years looks like this:

(from here

That curve is made from deep sea curves, and unfortunately, the deep sea curves don’t go back far enough in time to get a good idea of glacial change, at this level of detail, for enough time to really get a handle on the full history of glaciation. But by piecing together data from many sources, and careful use of dating techniques like Paleomagnetism, it is possible to get a long view of Earth history. The following graphic (from The Panerozoic Record of Global Sea-Level Change by Miller et al, Science 310(1293)) shows the general pattern.

The Earth was warmer many tens of millions of years ago (before the modern ecology, flora, and fauna evolved), fluctuating between warm and very warm over long periods of time. Then, in recent millions of years, things cooled down quite a bit. It is during this cooler period that the modern plants and animals became established, and that humans came on the scene.

Here’s what I want you to get out of this graphic. Look at the purple line. These are global sea levels after about 7 million years ago. Note that sea levels were often tens of meters higher than they are today (relative to the zero line on this graph).

Now have a look at this graph.

This is a very complicated graph and in order to understand all of the details you’ll have to carefully read the original paper. But I can give you a rough idea of what it all means.

Each of the four graphics is a different way that paleoclimatologists can look at the relationship between atmospheric CO2 and sea level, and compile a large number of data points. Think of each graphic as a metastudy of CO2 and sea level using four independent approaches and all of the data available through about 2012.

In each graph, note the dotted lines. The vertical dotted line is CO2 at 280ppm, taken as the pre-industrial value (though we know that is probably an overestimate since lots of CO2 had been released into the environment by human landuse and burning prior to the presumed beginning of the “pre-industrial” period). The horizontal line is the present day sea level.

Each of the little symbols on the graph is a different observation of CO2 and sea level from ancient contexts, using various techniques and with various paleoclimate data sets over many millions of years.

Note that at the point where the pre-industrial CO2 and the modern sea level intersect, there are many points above and below the line. This is partly error or uncertainty, and partly because of the time lag between CO2 reducing over time and subsequent growth of glaciers. It takes many thousands of years for these glaciers to grow (they melt much more quickly).

As we go from pre-industiral levels of CO2 through the values of interest here, getting up to 600 or more parts per million, past sea levels are generally higher than the present. In fact, at 400ppm, where we are now, sea levels are substantially higher than the present for almost all of the data points. This probably means the following, and this is one of the most important sentences in this post, so I’ll give it its own paragraph:

Present day carbon dioxide levels are associated with sea levels many meters above the present sea level; the current Earth’s atmosphere is incompatible with the current Earth’s glaciers, and those glaciers will therefore become much smaller, and the sea level much higher, even if we stop adding CO2 to the atmosphere this afternoon.

The other interesting thing about this graph is that above about 600ppm atmospheric CO2, the sea levels observed in the past are even higher, way way higher. These are time periods when there was virtually nil glacial ice on either pole, or elsewhere. This is what the Earth looks like when virtually all the ice is melted. This is the Earth that we are likely to be creating if we allow CO2 levels to approach 1000ppm, and the track we are on now virtually guarantees that by the end of the century.

One might assume that we would never let that happen, that we would solve this problem of where to get energy without buring fossil fuel before that time. But at this very moment there is about a 50–50 chance that the next president of the United States will be a man who believes that climate change is a hoax. In other words, it is distinctly possible that one of the largest industrial economies in the world, and a globally influential government, will ignore climate change and forestall the transition to clean energy until 2024.

Look at this graph. The upper line, “High Emissions Pathway (RCP 8.5)”

I now officially rename the “Trump Line.”

So, how high will the sea levels rise?

There really is little doubt that we are looking at several meters of sea level rise given our current CO2 levels. How many meters?

From the paper cited above: “…our results imply that acceptance of a long-term 2 degrees C warming [CO2 between 500 and 450 ppm] would mean acceptance of likely (68% confidence) long-term sea level rise by more than 9 m above the present…”

Personally, I think this is a low estimate, and the actual value may be more like 14. Or more. There is no doubt that we are going to add many tens of ppm of CO2 to the atmosphere over the next few decades even if we act quickly on changing our energy system, and the chances are good that we will be close to 600. This is at the threshold, based on the paleodata, between lots of glacial ice melting to produce 9, 15, or so levels of sea level rise, and nearly all of the glacial ice melting, to produce sea level rise of over 30 meters.

And, if we follow the Trump Line we’ll reach that level of CO2 well before the end of the century.

Again, the question emerges, how long will it take for sea level to rise in response to the added CO2? Than answer, again, is we don’t know, but in important ways, as noted above, it matters less than one might think.

We are probably going to be stupid about sea level rise

There is another aspect of this problem that is underscored by the most current research, and several other research projects. For the most part, the glaciers will not melt evenly and steadily. This is not a situation where we can measure how much ice melts off every decade, and extrapolate that into the future. What we now know about the big glaciers is that they will almost certainly deteriorate a little here and a little there, then suddenly and catastrophically break down, losing a huge amount of their ice to the sea, then for a period of time continue to fall apart at a lower but still accelerated rate, which will slow down after a while until some level of stability is reached. Then, that stability will remain threatened for a period of time until the next catastrophic collapse of that particular glacier.

Also, as noted in the research project I’ll report below, some of these catastrophic steps that happen later in the process may in some cases be the largest.

Here’s why this is important. Honestly, do you even believe that we have already added enough CO2 to the atmosphere to flood all of the planet’s major coastal cities, and major areas of cropland, and that this can’t be stopped? Isn’t that a bit extreme, alarmist, even crazy? Of course it seems that way, and even if you accept the science, a significant part of you will have a very hard time accepting this conclusion.

Those in charge of policy, the people who can actually do something about this, are not immune to this sort of cognitive dissonance. So, as long as the glaciers are only adding a foot a century instead of a foot a decade, the massive melting scenario will be on the back burner. Then, of course, one or two or more of these glaciers are going to lose their grounding lines within a few years of each other, and start to add huge amounts of water to the sea, and everyone will freak out and catastrophic coastal flooding will happen, and then the whole thing will slow down and more or less stop for a period of time, and once again, the prospect of sudden and major sea level rise will return to the back burner.

Then it will happen again.

New research on glacial collapse

OK, about that new research. Here is the abstract:

Climate variations cause ice sheets to retreat and advance, raising or lowering sea level by metres to decametres. The basic relationship is unambiguous, but the timing, magnitude and sources of sea-level change remain unclear; in particular, the contribution of the East Antarctic Ice Sheet (EAIS) is ill defined, restricting our appreciation of potential future change. Several lines of evidence suggest possible collapse of the Totten Glacier into interior basins during past warm periods, most notably the Pliocene epoch, … causing several metres of sea-level rise. However, the structure and long-term evolution of the ice sheet in this region have been understood insufficiently to constrain past ice-sheet extents. Here we show that deep ice-sheet erosion—enough to expose basement rocks—has occurred in two regions: the head of the Totten Glacier, within 150?kilometres of today’s grounding line; and deep within the Sabrina Subglacial Basin, 350–550?kilometres from this grounding line. Our results, based on ICECAP aerogeophysical data, demarcate the marginal zones of two distinct quasi-stable EAIS configurations, corresponding to the ‘modern-scale’ ice sheet (with a marginal zone near the present ice-sheet margin) and the retreated ice sheet (with the marginal zone located far inland). The transitional region of 200–250?kilometres in width is less eroded, suggesting shorter-lived exposure to eroding conditions during repeated retreat–advance events, which are probably driven by ocean-forced instabilities. Representative ice-sheet models indicate that the global sea-level increase resulting from retreat in this sector can be up to 0.9?metres in the modern-scale configuration, and exceeds 2?metres in the retreated configuration.

Chris Mooney has written up a detailed description of the research including information gleaned from interviews with the researchers, and you can read that here. In that writeup, Chris notes:

Scientists believe that Totten Glacier has collapsed, and ice has retreated deep into the inland Sabrina and Aurora subglacial basins, numerous times since the original formation of the Antarctic ice sheet over 30 million years ago. In particular, they believe one of these retreats could have happened during the middle Pliocene epoch, some 3 million years ago, when seas are believed to have been 10 or more meters higher (over 30 feet) than they are now.

“This paper presents solid evidence that there has been rapid retreat here in the past, in fact, throughout the history of the ice sheet,” Greenbaum says. “And because of that, we can say it’s likely to happen again in the future, and there will be substantial sea level implications if it happens again.”

And, from the original paper (refer to the graphic below):

The influence of Totten Glacier on past sea level is clearly notable, but for any particular warm period it is also highly uncertain, because the system is subject to progressive instability. Our results suggest that the first discriminant is the development of sufficient retreat to breach the A/B-boundary ridge. This causes an instability-driven transition from the modern-scale configuration to the retreated configuration. Under ongoing ice-sheet loss, the breaching of Highland B causes further retreat into the ASB. Each of these changes in state is associated with a substantial increase in both the absolute and the proportional contribution of this sector to global sea level.

This is a video by Peter Sinclair that dates to a bit earlier than this research was published, but covers the same issue:

Many of the interesting natural areas, like national parks or preserves, have a museum. In the museum there is often a geology exhibit, showing the changes in the landscape over long periods of time. Almost always, there was a period of time when the place you are standing, looking at the exhibit, was part of a “great inland sea.”

Let me introduce you to my little friend … the Great Inland Sea. Because it is coming back.

Human caused greenhouse gas pollution has locked us into a situation where the global sea level will rise, at an unknown rate, high enough to inundate most major coastal cities and vast areas of agricultural land in low lying countries, and wipe out thousands of islands. Entire countries (small, low lying ones, and pacific ocean nations) will either disappear entirely or be made very small. Even as we head towards a likely limit in global food production in relation to increasing demand, large productive agricultural areas will be destroyed. As far as I can tell, there is nothing to stop this from happening, though reducing our greenhouse gas pollution to zero over the next several decades may prevent the global ocean from rising to its absolutely maximum amount.

So sea level rise is important.

The surface of the Earth comes in two forms: Ocean bottom and continent. They are totally different geologically, with the ocean bottom consisting of relatively heavy basaltic rock formed at the margins between spreading plates, and continents of lighter rock, generally formed from below.

The global ocean sits mainly on the oceanic plates, but at its edges (except in a very few special locations), it rests against those continents. Over time the sea rises and falls. When the sea is at its lowest point, with a good amount of its volume reduced because it is trapped in glacial ice, most of the continents are exposed. When the sea is at its highest point, vast areas of the continental margins are inundated. At present, the ocean is pretty high, covering much of the continental margin that it ever covers, but there is room to grow, with large areas of the coastline subject to future inundation.

Rising surface temperatures caused directly or indirectly by human release of greenhouse gas pollution melt glaciers and warm the ocean, both of which are causing the global sea level to rise. This is a long and complicated process. We add greenhouse gas, mainly CO2, to the atmosphere, and this causes warming, enhanced by various positive feedbacks that either cause an increase of additional greenhouse gases such as water vapor, methane, and more CO2, or reduce the ability of certain natural systems to absorb these gasses. The greenhouse gas causes warming, which causes more greenhouse gas, which causes more warming. Meanwhile, most of this extra heat is actually trapped in the ocean where it only contributes a little to melting glaciers, but does contribute to expanding the volume of the ocean. The ultimate amount of heating, and the ultimate amount of sea level rise, takes a long time to be realized, and the rate of this change is only roughly estimated.

What we have already done to the atmosphere will cause sea level rise to continue for a very long time, possibly many centuries, possibly thousands of years. We have increased the amount of CO2 in the atmosphere from the mid 200s parts per million (ppm) to 400ppm, and we expect that increase to continue for decades. Evidence from the past, through the science of paleoclimatology, tells us that when the atmosphere holds between 400ppm and 500ppm of CO2, the global sea level is many meters above the present level.

Understanding sea level change is therefore critically important to understanding the impacts of climate change. We can measure current sea level rise and assume that steady increase over time (even if it is a bit variable) is mostly caused by global warming, heating the ocean and melting glacial ice. But there are problems with these measurements and associated estimates. Recent research has shown that Antarctic, which holds most of the world’s ice, is or could or will contribute a very large amount of water to the sea. But, other recent studies show that some of the expected reduction in glacial size might not be happening at the rate previously estimated. At the moment, sea level is rising at a certain rate, and some research explains a good amount of that increase from melting ice, but other research takes that melting ice out of the equation and leaves that portion of the sea level rise unexplained, for now.

Past sea level change (up or down), prior to the industrial revolution when we started releasing all this greenhouse gas pollution, should give us a baseline against which to assess modern day measurements, and is an essential part of the process of understanding this critically important system. But it is difficult to measure sea level, at present or in the past. We can measure the current position of the sea at a given part of the continental margin by just going there and measuring it. Sea level over recent decades, going back in some places a few centuries, can be estimated using tide gage records. We can sink cores (or trenches) in relatively protected areas (such as behind barrier islands) and find organic material that would have been formed just below the surface of the sea, measure its elevation and date it, to give an estimate of sea level in the past. We can put the tide gage data and the coring data together and get a rough estimate of sea level change.

But that estimate is not just rough, but almost useless, without a lot of careful further study. As the organic material representing older sea levels is buried by later organic material or other sediment, it tends to be compressed and lower in elevation. The study of this process is many decades old, and this can be adjusted for, but it is complicated. The actual sea level at a given point along the coast depends partly on how big the ocean is at the moment (obviously) but also on the position and strength of major currents. At present, and many times in the past, the North Atlantic ocean is bunched up way out at sea because of the movement of currents. This lowers the sea level along the coast in many areas. But if these currents either move or change in their strength, this effect changes, and the coastal sea level goes up or down independently of the global sea level.

Wherever there were large glaciers, the land has been pushed down by the weight of the ice. After the glaciers melt away, the land rebounds. Where this happens along the coast, estimating global sea level from local sea level becomes quite complicated. Meanwhile at the outer edge of the glacial mass, the land is actually pushed up to compensate for the depression caused by the massive glaciers. This is called “forebuldge.” Forebuldge makes the sea level look lower than it should, until the forebuldge reduces and flattens out. Indeed, the rebound effects of enormous glaciers in Canada are still happening, changing the position of the shoreline of Hudson’s Bay fast enough that cabins built on the shore a century ago are now a long walk from the sea.

This is all manageable, and people have been working on collecting these data and figuring out how to use it since the 1960s. But now, this week, what may be the first research project to put most of these data together to provide a pretty good estimate of sea level variation over the last 3,000 years, has been published.

The key result from this paper is the graph at the top of this post.

Robert E. Koppa, Andrew C. Kemp, Klaus Bittermann, Benjamin P. Horton, Jeffrey P. Donnelly, W. Roland Gehrels, Carling C. Hay,b,k, Jerry X. Mitrovica, Eric D. Morrow, and Stefan Rahmstorf’s paper, “Temperature-driven global sea-level variability in the Common Era” (PNAS) does this:

We present the first, to our knowledge, estimate of global sea-level (GSL) change over the last ?3,000 years that is based upon statistical synthesis of a global database of regional sea-level reconstructions. GSL varied by ?±8 cm over the pre-Industrial Common Era, with a notable decline over 1000–1400 CE coinciding with ?0.2 °C of global cooling. The 20th century rise was extremely likely faster than during any of the 27 previous centuries. Semiempirical modeling indicates that, without global warming, GSL in the 20th century very likely would have risen by between ?3 cm and +7 cm, rather than the ?14 cm observed. Semiempirical 21st century projections largely reconcile differences between Intergovernmental Panel on Climate Change projections and semiempirical models.

So now we have a much better idea of the nature of global sea level rise for a couple thousand years prior to human greenhouse gas pollution, and we have a firm demonstration of the effects of this pollution on sea level over the last century or so.

We are fortunate that one of the authors of this paper, Stefan Rahmstorf, is a blogger at Real Climate, where he wrote this post summarizing the original paper (though the original paper, linked to above, is pretty readable!).

Climate Central produced this graphic based on the paper:

Of this, Rahmstorf says, “The fact that the rise in the 20th century is so large is a logical physical consequence of man-made global warming. This is melting continental ice and thus adds extra water to the oceans. In addition, as the sea water warms up it expands.”

How much will sea level rise by the end of the century?

In his post, Rahmstorf brings in a second study on sea level rise, also just published (see the RC post for more details). That research attempts to estimate the amount of sea level rise expectd by 2100. There are four separate studies, each using three different (RCP) assumptions about future human caused climate change, and each combination of study and model provides a range. In centimeters, the lowest numbers are around 25 (close to the amount that has already happened over the last century) and the highest numbers are around 130-150 (so, up to about five feet).

Rahmstorf appears to agree with my thinking on this, which is that these estimates don’t account for catastrophic deterioration of ice sheets and subsequent increase in melting, if such a thing results from what appears to be increasing instability of some of those glacial features. For example, huge parts of the Antarctic ice sheet are in the form of vast glacial rivers pinned in place by a precarious “grounding” of ice on rock near the mouth of those rivers.

If that grounding falls apart, the entire river can start to march to the sea very quickly, establishing a new grounding line upstream. It is possible that such a new grounding line is way upstream. As all that ice falls into the sea, it would likely expose high vertical cliff that would then start producing ice bergs at a very high rate for many years. There may be other features currently deep under the ice that would be exposed, such as pre-melted water near warm spots. In other words, the drainage of meltwater will not be made less efficient by such a collapse, but rather, more efficient, regionally and for a certain period of time. The point is, the impact on the rate of glacial melt of such events is pretty much unknown and very difficult to estimate.

Rahmstorf notes, “The projections on the basis of very different data and models thus yield very similar results, which speaks for their robustness. With one important caveat, however: the possibility of ice sheet instability, which for many years has been hanging like a shadow over all sea-level projections. While we have a pretty good handle on melting at the surface of the ice, the physics of the sliding of ice into the ocean is not fully understood and may still bring surprises. I consider it possible that in this way the two big ice sheets may contribute more sea-level rise by 2100 than suggested by the upper end of the ranges estimated by Mengel et al. for the solid ice discharge, which is 15 cm from Greenland and 19 cm from Antarctica. (The biggest contributions to their 131 cm upper end are 52 cm from Greenland surface melt and 45 cm from thermal expansion of ocean water.)”

He backs this up by reference to other recent studies showing that ice sheets have in the past broken up at surprisingly high rates.

One and a half meters, or five feet, of sea level rise within the lifetime of those born today is possible. Half of that is extremely likely. Double that may even be a possibility. This is expected to continue for centuries, even millennia, or until all the ice melts, whichever comes first.

How many things in your life originate from some thing that happened in the past? The invention of agriculture (that happened many times from about 10,000 to 4,000 years ago), the invention of writing (again, multiple times, thousands of years ago), the modern western system of government and law (depending on where you live, the Magna Carta, the US Constitution) hundreds of years ago. If you are religious, it is likely that your religion’s roots are thousands of years old. The establishment of property rights, water rights, all of that.

If human civilization exists, with some continuity with the present, 1,000 years from now, such a list will include the release of fossil carbon in the form of greenhouse gasses by the people of the 19th, 20th, and 21st centuries. That was the event that caused the sea to rise and engulf so much of the fertile land, causing a major (if possibly slow moving) exodus of most of the settled people of the world. In a thousand years, after we’ve either stopped using fossil fuel, or didn’t but just used it all up, people will still be measuring for the rise of the sea that we are causing right now.

I don’t think they will be thanking us.