Table of Contents  Search  for printer      HOME
The Discovery of Global Warming                      February 2017

Ice Sheets and Rising Seas

A big enough rise of global temperatures would eventually melt the world's glaciers, and indeed a retreat of mountain glaciers since the 19th century was apparent in many regions. That would release enough water to raise the sea level a bit. Worse, beginning in the 1960s, several glacier experts warned that part of the Antarctic ice sheet seemed unstable. If the huge mass slid into the ocean, the sea level rise would wreak great harm, perhaps within the next century or two. While that seemed unlikely (although not impossible), by the 1980s scientists realized that global warming would probably raise sea level enough to affect populous coastal regions. Around 2000, changes in Greenland and Antarctica raised worries that the rise might become dangerously rapid. The problem would be redoubled if stronger storms sent water surging inland.

     Subsections: Suspicions of Instability (1960s-1980s) - Evidence of Instability
- Greenland and Pine Island - Storm Surges and the Future
, seashore flooding, submergence
Glaciologists, the scientists who study how ice behaves in seriously large quantities, have a special interest in floods. They even have their own word, jökulhlaup (from Icelandic), to describe the spectacular outbursts when water builds up behind a glacier and then breaks loose. An example was the 1922 jökulhlaup in Iceland. Some seven cubic kilometers of water, melted by a volcano under a glacier, had rushed out in a few days. Still grander, almost unimaginably grand, were floods that had swept across Washington state toward the end of the last ice age when a vast lake dammed behind a glacier broke loose. In the 1940s, after decades of arguing, geologists admitted that high ridges in the "scablands" were the equivalent of the little ripples one sees in mud on a streambed, magnified ten thousand times. By the 1950s, glaciologists were accustomed to thinking about catastrophic regional floods.        - LINKS -
Also within their purview was flooding on a far grander, but far slower, scale. Since the heroic polar explorations of the late 19th century the world had known that great volumes of water are locked up in ice sheets. If there were substantial melting of the Greenland ice cap, and especially of the titanic volume of ice that buries Antarctica, the water released would raise the oceans in a tide that crept higher and higher for millennia. It had happened before — geologists identified beaches far above the present sea level, cut by waves in warmer periods when the Earth was entirely free of ice. In the last interglacial period, some 125,000 years ago, the planet had reached a temperature about as high as was likely to come from greenhouse warming in the next century or two. Back then, even though most of Antarctica had remained ice-covered, the sea level had been at least seven meters (more than 20 feet) higher than at present. This was about what would be expected if most of Greenland melted. The next time that happened, sea water would swamp coastal regions where a good fraction of the world's population now lived. All this became familiar to anyone who followed scientific discussions of global warming.







=>Public opinion

Up to the 1960s, scientists expected that global warming caused by greenhouse gases, if it happened at all, would steal in gradually over many centuries. So the threat of flooding lay in a comfortably vague and remote future. To be sure, a few scientists had begun to imagine more abrupt change if the melting of the ice itself brought on conditions that accelerated the warming. Transitions between glacial and warm climates — and back again — might come in a matter of only a few centuries if not faster. As one example, in 1947 the New York Times quoted a prominent Swedish geophysicist, Hans Ahlmann, who suggested that a global warming might be underway that could eventually bring a "catastrophic" rise of sea level as glaciers melted. "Peoples living in lowlands along the shores would be inundated," he explained, calling on international agencies to undertake studies as an urgent task. Most scientists, however, expected that within the foreseeable future, the main effect of any global warming on ice would be to shrink the ice pack on the Arctic Ocean. Since that ice was floating, it could melt entirely away without changing sea level at all, just as the level of water in a glass does not change when a floating ice cube melts.(1)


Full discussion in
<=Rapid change




<=Simple models

Suspicions of Instability (1960s-1980s)     TOP OF PAGE  
Glaciers on land could affect sea level, and they were notoriously sensitive to climate. Advances and retreats of glaciers in the Alps in particular had been conspicuous for generations, reacting to small changes not just in temperature but also in the amount of snowfall.(2*) In 1962, John Hollin opened up speculation about how relatively small climate changes might also affect ice in Antarctica. He argued that great volumes of ice there, piled up kilometers high and pushing slowly toward the ocean, were held in place by their fringes. These edge sheets were pinned at the marginal "grounding line" where they rested on the ocean floor. A rise of sea level could float an ice sheet up off the floor, releasing the entire stupendous mass behind it to flow more rapidly into the sea.(3*)  
The idea was picked up by Alex Wilson, who pointed to the spectacle of a "surge." Glaciologists had long been fascinated by the way a mountain glacier might suddenly give up its usual slow creeping, to race forward at a rate of hundreds of meters a day. They figured this happened when the pressure at the bottom melted ice so that water lubricated the flow. As the ice began to move, friction melted more water and the flow accelerated. Could the ice in Antarctica become unstable in this fashion? If so, the consequences sketched by Wilson would be appalling. As the ice surged into the sea, the world's sea-coasts would flood. And that would not be the worst of humanity's problems. Immense sheets of ice would float across the southern oceans, cooling the world by reflecting sunlight. It could bring a new ice age.(4*) Hollin joined in by publishing observations of deposits in England that recorded past sea levels, showing rapid rises of as much as ten meters. It could happen any time, he thought, perhaps in mere decades — or even faster if the sea-level change set off tsunamis. He pointed to unusual features that suggested an abrupt disaster, such as "the curiously intact remains of large mammals" buried whole.(5) Few scientists gave much credence to any of these speculations. The ice that covered most of Antarctica, in places more than four kilometers thick, seemed firmly grounded on the continent's bedrock.





=>Simple models



=>Rapid change

However, in 1968 John Mercer, a bold and eccentric glaciologist at Ohio State University, pointed out a problem: the West Antarctic Ice Sheet (WAIS). This is a smaller — but still enormous — mass of ice, separated by a mountain range from the bulk of the continent. Adventurous traverses of the ice during the International Geophysical Year 1957-58 had shown that much of the base of this mass was below sea level. Mercer argued that it was held back from flowing into the ocean, in a delicate balance, only by the shelves of ice floating at its rim. These shelves might disintegrate under a slight warming. The much larger mass of ice corked up by the shelves would then be released to slide into the ocean and disintegrate into icebergs. Just so, Mercer suggested, a collapse of ice sheets into the Arctic Ocean might have caused the more local, but remarkably sudden, cooling of the North Atlantic around 11,000 years ago that other scientists had identified. A West Antarctic Ice Sheet collapse could be very rapid, Mercer said. The sea level would not rise as far as it would rise if all of Antarctica surged, but it would be bad enough — up to five meters, he estimated (16 feet; calculations decades later pinned down the number at around 11 feet). Much of the world's population lives near the shore. Such a rise would displace more than a billion people and force the abandonment of many great cities. Mercer thought it could happen within the next 40 years.(6*)


The West Antarctic
Ice Sheet

<=Rapid change

= Milestone


Mercer published his worries in an obscure conference report, and although he wrote forcefully, he did not push his views on colleagues in the personal encounters that are crucial in a small community of specialists (he much preferred to be out doing fieldwork, often in the nude). The few specialists who heard of his ideas were not impressed. The problem, one of them complained, "could be argued indefinitely if it is not quantized."(7)  
In fact glaciologists had been working for decades on ways to calculate numbers for the flow of ice masses. In the 1970s they made rapid progress in formulating abstract mathematical models and putting the powerful new computers to work. The calculations, with many approximations, suggested that the West Antarctic Ice Sheet was indeed unstable. Apparently the floating ice shelf that held it back could break up with surprising ease, and the whole mass might begin sliding forward. One scientist who made a landmark calculation, Johannes Weertman, concluded that it was "entirely possible" that the West Antarctic Ice Sheet was already now starting its surge.(8)


<=External input

The scattering of climate specialists and geologists who paid attention to ice sheets viewed the models as highly speculative. It seemed scarcely possible that anything as massive as the the West Antarctic Ice Sheet could disintegrate in less than a few centuries. But if you took a long enough view to be concerned about the next few centuries, a surge that dumped a fifth of a continent of ice into the oceans would be no small thing, and they could not rule it out. The picture fitted with a new feeling that was emerging in the climate community, a feeling that the climate system in general was unstable or even radically chaotic.



<=>Rapid change

Concern sharpened in 1975 when Cesare Emiliani at the University of Miami reported measuring deep-sea cores that showed a shockingly rapid rise of sea level — a rate of meters per decade — around 11,600 years ago. (He remarked that this was exactly the time Plato had given for the fall of Atlantis!) Emiliani thought the cause of the flooding might not have been an Antarctic surge, but water rapidly released from enormous lakes that had been penned up behind the North American ice sheet, a titanic jökulhlaup. In places like Florida where the land sloped gently into the ocean, he wrote, "the sea would have been seen to advance inland 300 feet in... a single summer."(9) Other areas at risk included the Nile Delta and the Netherlands. Science journalists made sure that the more spectacular warnings reached a broad public.(10)

<=Rapid change




=>Public opinion

Meanwhile radar surveys from airplanes showed that the ice of West Antarctica moved toward the sea not as a single sheet but through a set of enormous ice streams. Terence J. Hughes (who started out studying metallurgy but moved on to a different sort of solid material) and other glaciologists developed increasingly elaborate models of ice sheet dynamics.(11) They showed how a slight shift in conditions could prompt an ice shelf to break up into flotillas of icebergs. Looking over the new data and theories, Mercer worried that most climate experts still assumed that ice sheet changes would take many centuries. In 1978, he finally caught their attention with an article in the leading journal Nature, contending that because of global warming from humanity’s use of fossil fuels, "a major disaster... may be imminent or in progress." Mercer admitted that the computer models were loaded with uncertainties, but "there is, at present, no way of knowing whether they err on the optimistic or the pessimistic side."(12)  
Mercer, Hollin and Hughes had a chance to argue their case to a group of experts at a meeting convened in April 1979 in Annapolis, Maryland. One participant noted in his diary that their arguments convinced him that the deglaciation of West Antarctica was "a plausible hypothesis." The majority felt that this was "not a cause for immediate alarm however. We are talking about centuries."(13) In a published review, a trio of experts laid out arguments explaining why the collapse of an ice sheet would probably take several centuries to run its course. Yet they admitted that "little is known about the glaciers," and a 5-meter rise in sea level could possibly happen within a century. "Mercer's warning," they concluded, "cannot be dismissed lightly."(14)  
That continued to be the most common view through the 1980s. Studies found that an ice sheet collapse was likely to take centuries rather than decades, but experts knew too little about the behavior of Antarctica's mammoth ice rivers to agree on any firm conclusion. Field glaciologists, a small but hardy group, measured one or another ice sheet as best they could at a few scattered locations. They found ice streams moving consistently at speeds of hundreds of meters a year, far faster than ordinary mountain glaciers. Meanwhile, their mathematically-minded colleagues back home constructed simplified models for the flow.(15) Some studies foresaw the possibility of a sea-level rise of two or three meters (6-10 feet) by 2100, but most found this unlikely so soon. In particular, for a 1983 National Academy of Sciences report, the dean of oceanographers, Roger Revelle, estimated that within the next hundred years the sea level would probably rise some 70 cm (about two feet). That would be harmful but not catastrophic. He did worry, however, about an Antarctic collapse later on.(16*)  
Some rise of sea level in the coming century seemed not just possible, but nearly certain. The oceans had already risen 10 or 20 centimeters in the 20th century, about ten times as fast as the average sea-level rise in previous millennia. Just where all the water had come from remained uncertain. As one example, it was not until the 1990s that experts realized that significant volumes of water were engaged by human activities like irrigation and building reservoirs, and they could not say whether the net result of such activities was to take water from the oceans or to put more in.(17)  
One contribution to the sea-level rise was entirely clear. Water expands when heated. The consequences may seem obvious, but amid all the talk of melting glaciers, for decades nobody seems to have given a thought to other simple effects. Finally in 1982 two groups separately calculated that the global warming observed since the mid-19th century must have raised the sea level significantly by plain thermal expansion of the upper ocean layers. But a thermal expansion could not account for all of the observed rise. The scientists figured the rest came from melting glaciers (most of the world's small mountain glaciers were in fact shrinking).(18)  
The rising waters might help the West Antarctic Ice Sheet float off its moorings and slowly break up. Even if that never happened, there would still be problems. Scientists warned that tides would probably mount a half meter or even a meter and a half higher by the end of the next century, bringing severe harm to coastal regions. Beaches would erode back hundreds of feet. Salt water would advance into fragile estuaries. Entire populations would flee from storm surges.(19*)  
While the calculations of thermal expansion were straightforward, the actual sea level rise would depend on a much tougher problem — what would happen to the ice sheets of Greenland and Antarctica? So long as they did not surge and disintegrate, global warming would not necessarily make them dwindle. A warmer atmosphere would hold and transport more water vapor, so it would drop more snow. Thus the polar ice sheets might actually grow thicker, withdrawing water from the oceans. The future sea level depended crucially on just what happened to glaciers and ice sheets, one pair of experts concluded, and predicting that would be "a daunting task."(20)  
Evidence of Instability (1980s-2004)      TOP OF PAGE  
To sketch out an answer to the great question of ice-sheet collapse, since the early 1980s increasing numbers of scientists had bundled up in parkas and gone out onto the windswept wastes of Antarctica. It was grueling work, isolated and dangerous. Researchers measuring the ice streams learned to travel with their ski-mobiles roped together like mountaineers (despite precautions, in 2016 glaciologist Gordon Hamilton died when his snowmobile plunged into a hidden crevasse). Their difficult goal was to measure the motions of the immense slow ice currents, using radar pulses, seismic measurements, and boreholes to study how ice moved over the rock beneath. One example was a scientist who had been skeptical of surge models — he recalled that he "felt the whole thing was like a house of cards" — but who changed his mind when he discovered that a kilometer-thick Antarctic ice stream rested not on bedrock but on a layer of slippery mud.Another unsettling discovery was that in recent centuries some of the great ice streams had stopped or started moving, for no clear reason.(21)  
Far more such data would be needed to bring a definitive answer. The dynamics of ice sheets and the streams that fed them turned out to be, like most things geophysical, a complicated snarl of influences. Experts could not even agree on whether the West Antarctic Ice Sheet had disintegrated during previous warm epochs over the past few million years. The past sea level rises might have come from Greenland ice, or from something else entirely. But according to evidence developed in the 1990s, during a dramatic episode at the end of the last ice age, something had once raised the sea level 16 meters within three centuries. The rate of rise might have reached two feet per decade. The West Antarctic Ice Sheet was the most likely source of all that water. The rush of new data fed what one observer called "polite but emotional debate" among experts. And there were indeed WAIS experts now. Since the 1980s a little interdisciplinary, international community had been taking shape in ad hoc workshops at various locations.(22*)  
Meanwhile a powerful new tool, satellite images, revealed that some of the smaller floating ice shelves poking out from the peninsula that projects from the Antarctic continent were rapidly disintegrating.(23) It was not clear whether the changes had anything to say about the possibility of a catastrophic ice-sheet collapse. In these little-known regions, the changes might have been a type of normal, regional event, which just had not been noticed before the age of intensive global monitoring. Yet the public's concern about global warming was reinforced from time to time when satellite images showed tabular icebergs bigger than cities floating off. And scientists began to doubt this was normal. After all, back in 1978 Mercer had called for keeping an eye on just these ice shelves, contending that their breakup would be "one of the warning signs that a dangerous warming trend is under way in Antarctica." He had predicted still more specifically that the collapse of ice shelves would start at the northern end of the Antarctic Peninsula and proceed south, and indeed by 1996 the five most northern ice shelves were shrinking rapidly, but not the more southerly ones.(24)





=>Public opinion

In the 1980s and '90s specialists in glacier flow worked up increasingly complex ice-sheet models, using the rapidly expanding power of computers to incorporate essential features such as heat flow within the ice.(24a) Entirely aside from the question of Antarctic surging, these models might be useful in explaining the ice ages. It seemed increasingly likely that the reason ice sheets came and went in cycles of around 100,000 years had something to do with the length of time needed for a continent of ice to form and flow and melt, while the entire rocky crust beneath it sluggishly sank or rebounded as the weight of ice grew or diminished. Nothing else on Earth seemed to change on the right timescale.




=>Simple models

The models failed to answer the question of how fast a major ice sheet could surge into the ocean. The improved models did show, reassuringly, that there was no plausible way for a large mass of Antarctic ice to collapse altogether during the 21st century. According to these models, if the West Antarctic Ice Sheet diminished at all, it would discharge its contents only slowly over several centuries, not placing too heavy a burden on human society. Yet scientists could not altogether rule out the possibility of a shocking surprise in some future generation. The West Antarctic Ice Sheet remained what one expert had called it a quarter-century earlier — "glaciology's grand unsolved problem."(25)  
Scientists were still less able to answer the question of whether climate change was gradually melting the rest of the world's glaciers and ice caps, or instead was adding snow to them. In "those huge areas where little or no information is available," an expert explained in 1993, "almost anything might be happening." But in 2005 a survey of mountain glaciers around the world found that most of those for which historical records existed had been shrinking since 1900. Some that had survived for many thousands of years were vanishing, a striking sign of unprecedented climate change.Experts could only speculate how far this might affect sea level. Would it be counteracted by the increased snowfall that some models predicted global warming would bring in the remote dry highlands of Antarctica?(26*)

glacier shrinking
Glacier 1875/2004

<=Modern temp's

As scientists turned increasing attention to ice movements, they discovered many kinds of changes, thanks to satellites and airplane overflights as well as increasingly precise measurements by grueling expeditions on the ice itself. "Perhaps the most important finding of the past 20 years," a glaciologist reported in 2002, "has been the rapidity with which substantial changes can occur on polar ice sheets." Warmer ocean waters were melting the underside of ice sheets by tens of meters a year, altering where grounding lines pinned them. Entire floating ice shelves, some of which had been in place for thousands of years, were rapidly thinning, or astonishing experts by breaking up completely. Ice streams that had been held behind the disintegrating shelves accelerated.

=>Rapid change

Ice shelf collapse

Most scientists had figured that even after the air got warm enough to melt the surface of an ice shelf, it would take millennia for the entire great mass to melt away. It turned out, however, that meltwater could seep down into crevasses, refreeze there and wedge them wider, prying apart a thick sheet in months. Meanwhile the gradually warming seawater worked to break up the ice from beneath. None of this had been foreseen by the crude computer models of ice behavior.


Modelers scrambled to incorporate the new concepts. Revised computer simulations and further observations confirmed the idea, originally so speculative, that removing an ice shelf could dramatically speed up the drainage of glaciers "corked up" behind it. In 2004 evidence was published that some of the enormous ice streams leading from the West Antarctic Ice Sheet to the ocean were also speeding up. Scientists were no longer sure how many centuries it might take to drain the entire sheet. "The response time scale of ice dynamics is a lot shorter than we used to think it was," admitted a leader of the research.(27*)



= Milestone

Only a few people were trying to make computer models of any of these processes, and their models remained primitive. A mass of ice is an odd substance, something between a fluid and a solid, no easy thing to simulate. And far too little data had been gathered from the surfaces of these perilous and remote wastes, let alone from their buried base. The most modelers could say was that "The latest theoretical advances have done nothing to allay fears concerning the potential instability of marine ice sheets."(28)  
Greenland and Pine Island       TOP OF PAGE  
Meanwhile, starting around 2000, a few studies raised the additional possibility that the Greenland Ice Sheet, contrary to what most scientists had figured, might not be comfortably stable over the next few centuries. In the warmer summers the snow on the surface would get wet, and become darker. So it would absorb more sunlight and warm still more. Under one speculative scenario, rivers of water would drain through deep holes ("moulins") straight to the bottom of the ice and lubricate it. That might provide, as one team put it, "a mechanism for rapid, large-scale, dynamic responses of ice sheets to climate warming." Another mechanism might be thinning and crevassing at an ice stream's front end, due to warmer ocean water, causing a speedup that propagated upstream. As the flow of its huge ice streams accelerated, the Greenland ice cap would thin around the edges. As the ice surface sank to lower altitudes where the air was warmer, it could melt all the faster. Conceivably, an armada of icebergs would invade the North Atlantic and melt, as had happened around the end of the last ice age. At that time the sea level had risen at a rate that would be catastrophic for coastal areas. The process would presumably take centuries to run its course, or more likely millennia, but glaciologists could only speculate about the probability and timing of such a misfortune.(29*)


Greenland ice stream
Surging ice stream


=>Rapid change

<=The oceans

A 2006 analysis of satellite radar data found that the velocities of large ice streams in southern Greenland had doubled in the past five years — something most experts had thought was impossible. Perhaps the speculations about lubrication of the base of an ice stream were correct? The Greenland ice streams soon slowed down again, however, showing that the lubrication was temporary; a long-range study reported that these particular streams were discharging ice into the sea no faster, on average, than a decade earlier. Glaciologists were not reassured. Considering how ice streams around Antarctica had also been observed to accelerate and slow down suddenly, it seemed that these systems were more sensitive to perturbations than the scientific community at large assumed. Moreover, a new satellite was transmitting disturbing data. It measured gravitational force so sensitively that it could detect changes in the mass of an ice sheet from year to year. Both Greenland and West Antarctica were in fact losing substantial amounts of ice into the oceans. Observers were dismayed to see mass around the margins of Greenland dwindling at a rate that doubled in less than a decade.


Greenland ice loss
Greenland ice loss

=>Modern temp's
=>CO2 greenhouse

What would be the end-point? Some 400,000 years ago the temperature had been only a few degrees higher than at present, but the sea level had been more than six meters higher. An ingenious study of sediments found that southern Greenland had been largely ice-free at that time, and thus the island had been a main contributor to the sea-level rise. By the end of the century, if emissions of greenhouse gases continued, our civilization might well have locked in a grand change of the ice sheet and therefore of sea level.(30a*)  
The losses from Antarctica were still more surprising. Since the 1980s, European and American expeditions had been measuring the ice streams held back behind the Ross Ice Shelf and adjacent regions; they had found no disturbing acceleration. But there was a second region, Pine Island Bay, where narrower ice streams fed into the ocean. This was one of the most inaccessible places on the planet, with terrible weather besides, and it had scarcely been observed even from the air. Yet the terrain was such that already back in 1981, Terry Hughes had suggested that warming might accelerate the ice streams. He had called Pine Island the "weak underbelly" of West Antarctica.(30b)  
The region was lit up as by a flash of lightning with the 1991 launching of the satellite ERS-1, "the single most effective tool ever devised for measuring glacial change." The satellite's radar, peering through clouds and the long Antarctic night, could measure the ice surface with amazing precision. In 1998 Eric Rignot reported that the line where the Pine Island glacier was held back by the seabed had retreated five kilometers within half a decade. He assumed that the glacier's floating tongue was being eroded underneath by the warmer ocean waters. Another group found that the entire glacier basin feeding into the ice stream was losing altitude.(30c)


Eric Rignot Eric Rignot
courtesy NASA

Glaciologists rushed to study the neglected region. By 2009 they found that the ice streams entering Pine Island Bay were rapidly accelerating, and the basin that fed them was dropping 16 meters a year. "We don’t know really know what's going to happen to the ice," remarked a British team leader. Eventually the ice streams might surge much as Hughes had warned. However, the whole process was expected to act at a truly glacial pace. Most experts felt that at worst West Antarctica might contribute ten centimeters or so to the total sea-level rise of the 21st century. Bigger problems would come, gradually but inexorably, in the 22nd century and after. Observational and computer studies published in 2014 confirmed this as a near certainty: there was nothing in the seabed around Pine Island Bay to pin down the Pine Island glacier, the even bigger Thwaites Glacier, and others and keep them from retreating over the next half-dozen centuries. They would spill gigatons of West Antarctic ice into the ocean. The collapse, Rignot announced to widespread press attention, had become "irreversible... it has passed the point of no return." In recent years scientists had been warning in general that the climate system could have "tipping points" (more formally, "critical thresholds"). We had now passed one of these at least.








= Milestone

Some additional light was shed in 2013 by a successful effort, by a consortium of more than a hundred scientists, to extract ice from Greenland as far back as the Eemian period. In that period (about 120,000 years ago) the world had been nearly as warm as it was likely to get in the 22nd century — and the sea level had been roughly 20 feet higher. It turned out that Greenland had not all melted away at that time. Therefore a majority of the sea-level rise must have come from Antarctica. The news added to scientists' worries. Computer models of ice sheets were unreliable and it was a mystery how fast the Antarctic ice might collapse.(30d*)  
Scientists had thought the much larger volume of ice in East Antarctica would be stable due to its extreme cold. But studies in the early 2010s suggested that this sheet too was vulnerable as warming sea water undercut the glaciers that controlled the outflow. Expeditions probing the topography underlying East Antarctica found that, as with West Antarctica, there were regions with massive amounts of ice that were unstable and could collapse into the oceans. It was a problem for future centuries, however, not the present one.(31)  
More immediate worries were raised by news that the Arctic Ocean ice pack was shrinking far more rapidly than any model had predicted. The ice was dwindling to an unprecedented extent not only in area but, still more, in thickness and thus in total volume. By 2007 scientists were predicting that a "Northwest Passage" across northern Canada would be ice-free in summer decades earlier than expected — and already in 2016 a luxury cruise ship did make the trip. Of course, the disappearance of floating ice would not raise the sea level, but it could change weather patterns around the Northern Hemisphere. And the unexpected shrinking confirmed not only that the ice component of the climate system was poorly understood, but that our ignorance was concealing mechanisms that made for rapid changes(31a).

Arctic ice shrinkage
Arctic sea ice loss

=>Public opinion
= Milestone
=>Modern temp's

On the other hand, people who wished to deny that global warming was a problem pointed to the sea ice around Antarctica. In the few decades it had been monitored its area had expanded. To be sure, the increase was irregular and of slight magnitude, only the sum of increases in some regions and decreases in other regions — nothing like the spectacular dwindling of both the extent and the thickness of the Arctic ice pack. A variety of complex factors influenced the formation of Antarctic sea ice. Alongside normal cyclical changes there were global-warming-related changes in wind patterns and ocean currents and even the salinity of the seawater (as the melting glaciers of Antarctica dumped ever-larger volumes of freshwater into the sea). Experts could only agree that the subject was complicated. Indeed, a leading computer modeling group had predicted as far back as 1991 that during the first decades of global warming, the Antarctic Ocean would not warm up in the manner of the Arctic Ocean. The claim that the Antarctic sea ice fluctuations argued against global warming collapsed in 2017 when the area shrank to a record low.(31b)  
One thing was certain. If temperatures climbed a few degrees, as climate scientists now considered likely, the sea level would rise simply because water expands when heated. This is almost the only thing about global change that can be calculated directly from basic physics. The additional effects of glacier and ice sheet melting remained highly uncertain (scientists were still arguing over how much of the 20th century’s sea level rise was due to heat expansion and how much to ice melting). One hint gradually became apparent, and was pinned down after 2000 by satellite measurements: the rate of sea-level rise was accelerating. Since the 1980s the sea level had been rising much faster than at any point since records began in the 1880s. Indeed wide-ranging geological surveys showed that the rise over the past century was beyond anything seen for thousands of years, probably tens of thousands of years. Shooting up on graphs like the blade of a hockey stick, the sea-level rise was increasingly recognized as responsible for a large part of the coastal flooding that had begun to endanger communities on the East Coast of the United States.(31c)



sea level
Sea level rise observations

When the authoritative Intergovernmental Panel on Climate Change (IPCC) issued its 2007 report, the authors found the new ideas about ice sheets altogether uncertain. So for their sea level predictions they stuck with the known rise from warming plus the old, classic models for ice processes. They took no account of the possibilty of surges (the report was based on data published through about 2005, which left out some of the most disturbing reports). Back in 2001 the IPCC had offered a rough guess for the total rise expected by the end of the 21st century — perhaps half a meter, give or take a bit — and the authors of the 2007 report came up with much the same numbers. This refusal to include the possibility of ice-sheet collapse brought sharp criticism from some experts and others who worried about catastrophic sea-level rise.(32)  
Even before the new results came in from Greenland and Antarctica, some scientists had been worrying that the rise might be double the 2007 IPCC report's upper limit. Now they were still less willing to rule out the possibility of a rise of one meter, or even two. Such rapid rates, it turned out, had been experienced in past geological ages similar to the present.(32a) By 2012 many experts were projecting a rise of a meter if not more. Backing this up was a comprehensive international study that found Antarctica and, still more, Greenland were losing mass at rapidly accelerating rates. In August 2012 the entire surface of Greenland was seen to be melting, with pools everywhere (and the pools, darker than the snow surface, would absorb still more sunlight). As one expert ruefully remarked, "The motto for early 21st Century cryospheric science might be, 'that happened faster than I thought it would'."(32b)




Meanwhile some senior glaciologists and others argued that it was a mistake to concentrate on what seemed most probable. What about processes that, although perhaps not likely, would be catastrophic if they did come to pass? Ignoring an unknown did not make it go away.  
In its next major report, issued in 2013, the IPCC responded to the criticism. Now it projected a sea-level rise anywhere from 0.3 meter (if the world promptly launched vigorous emission reductions) to one meter. The latter was a conservative limit for what was "likely," and the panel warned that "there is currently insufficient evidence to evaluate the probability of specific levels above the assessed likely range." But they conceded a possibility of several tenths of a meter more if the WAIS started to collapse, that is, up to a 1.5 meter rise by 2100. (Things would get progressively worse in the following centuries, but few people concerned themselves with that.)




Some experts thought even the IPCC's 1.5 meter rise by 2100 might be an underestimate of what was possible. Computer models of ice-sheet collapse remained speculative; according to one model, the WAIS could collapse in a matter of decades. A prestigious group of scientists warned in 2015 that it was possible for the seas to rise several meters in the space of half a century. That showed, as a journalist noted, "just how far from resolved, scientifically speaking, the question of danger levels remains."(32c)



<=Rapid change


One meter of sea-level rise might not sound like much, but in many areas it would bring the sea inland a hundred meters or more (a few hundred feet), and even farther if storm-driven surges grew stronger. While such a rise would not be a world disaster, in the late decades of this century it would bring significant everyday problems, and occasional storm-surge catastrophes, to populous coastal areas from Bangladesh to New Orleans.


Storm Surges and the Future           TOP OF PAGE  
Scientists had warned for decades that New Orleans was at risk from hurricanes, and some had pointed out that the chance of disaster would mount as global warming raised the sea level and perhaps increased storminess. But after the catastrophe in August 2005 some experts asked whether Hurricane Katrina would have devastated the city, if the heat in the Gulf of Mexico's waters — a main source of the storm's energy — had not been higher than normal? Such a question can never be answered for a single event. The important question is not what global warming did in one case, but what it would mean for the future probability of terrible hurricanes and typhoons.




= Milestone

Scientists had only a sketchy idea of how tropical storms worked. Nevertheless, when the 21st century began, nearly all experts had been confident that tropical storms would not become seriously worse for many decades. Even Kerry Emanuel, who had explained in 1987 how a warmer sea surface would provide energy for greater storms, had not expected a noticeable change anytime soon. But when he analyzed decades of data on tropical storms, he found a disturbing trend. While the number of hurricanes and typhoons had not been increasing, the intensity of the worst storms seemed to have climbed in recent decades. The rapid increase in destructive power, so different from what experts had expected, correlated surprisingly well with the observed rise of sea-surface temperatures. "For the first time in my professional career," Emanuel recalled, "I got alarmed." In mid 2005 he published a warning of gathering danger. It attracted widespread attention three weeks later, when Katrina struck.(33)


Meanwhile a separate group had gotten similar results. But other meteorologists stuck by their earlier conclusions. A fervent, sometimes personal controversy broke out. The experts of the old school insisted that the record of tropical storm intensities was only guesswork for most of the 20th century, especially in the vast, unvisited spaces of the Pacific. If there had indeed been a change in hurricanes, they supposed it was only a phase in a normal North Atlantic cycle. Computer models varied, some projecting little change, others predicting a modest increase in the intensity of great tropical storms by the end of the century. Many observers felt that scientific understanding was so limited that, as one group concluded in 2007, "the question of whether hurricane intensity is globally trending upwards in a warming climate will likely remain a point of debate in the foreseeable future." However, thorough studies published in 2010 and 2013 concluded that Katrina-like events would become increasingly likely. A 2016 review concluded that while the total number of tropical cyclones might actually decrease, especially intense ones (categories 4 and 5) would come more frequently.



=>Public opinion






As for storms outside the tropics, computer models differed on whether they would get stronger, but the models agreed that the storm tracks would shift to different regions. That would probably bring floods more intense than anyone was prepared to withstand. In addition, ambiguous geological evidence raised speculation that the new climate might include "superstorms" far greater than anything seen in historical times. The very uncertainty of the matter was a call to action. If there was a serious risk of increased devastation, it was not something we should leave for the next generation to worry about.(34*)  
The sea-level rise alone makes it likely that low-lying areas where tens of millions of people live will become uninhabitable by the end of this century. Entire island nations are at risk. Then it will get worse. Even if humanity could bring its greenhouse emissions to a full stop, the gases already in the air will capture heat energy that will work its way gradually deeper into the oceans. The tides will continue to creep higher, century after century. Meanwhile, if the planet warms up a few degrees (which is the most likely scenario unless strong restrictions on emissions are promptly introduced), the forces melting polar ice will become irreversible. Eventually, probably after several thousand years, the Greenland Ice Sheet will be gone. In previous geological ages when the CO2 level in the Earth's atmosphere had reached 500 ppm (a level human emissions would bring within decades), much of Antarctica's ice cap too had been gone, raising the sea level tens of meters. Even if nothing happens in Antarctica, the melting of Greenland will put the sea level at least seven meters higher, giving posterity its grandest, but unwelcome, monument of our civilization.(35)

See the summary of expected Impacts of Climate Change.



Rapid Climate Change
The Public and Climate


1. Gladwin Hill, "Warming Arctic climate melting glaciers faster, raising ocean level, scientist says," New York Times, May 30, 1947. Ewing and Donn (1956); Budyko (1962). BACK

2. Glaciers as "sensitive indicators of climate" are stressed in the pioneering theoretical treatment of surges, Nye (1960). BACK

3. "The chief conclusion of this paper is that the greatest glacial fluctuations in Antarctica were produced by changes in sea-level." The paper was motivated by the idea that the timing of Antarctic glacial movements was set by sea-level changes that reflected Northern Hemisphere glaciation. Hollin (1962), p. 174. BACK

4. Wilson (1964); Wilson (1966); Wilson (1969); Wilson's starting-point was the suggestion that the center of Antarctica was at the pressure melting point, see Robin (1962), p. 141, who adds that "one would not expect the ice to surge over a large part of Antarctica at one time"; the role of frictional heat in ice-sheet instability was pointed out back in 1961 (in partial support of Ewing-Donn theory), drawing on earlier work by G. Bodvarsson, by Weertman (1961). BACK

5. Hollin (1965), quote p. 15. BACK

6. Mercer's basic argument was that "fringing ice shelves... will rapidly disintegrate by calving if the average temperature of the warmest month rises above freezing point," Mercer and Emiliani (1970); see Mercer (1968) ; North Atlantic: Mercer (1969); meanwhile a suggestion about a more gradual disappearance of the Greenland ice cap was advanced by Emiliani (1969); earlier, Robin and Adie (1964), said that catastrophic deglaciation of West Antarctica was "unlikely, but not necessarily impossible," p. 117. Later calculations (about 3.2 meters): Bamber et al. (2009). For the history of WAIS research see Thomas (2014). BACK

7. W.J. Campbell in discussion of Wilson (1969), p. 915. BACK

8. Data were analyzed by Hughes (1973); Weertman (1974), "entirely possible," p. 3; the classic theory was Thomas (1973); and Thomas (1973); Flohn (1974) gave a more general model; on ice modeling, see also Hughes (1977). BACK

9. Emiliani et al. (1975); see Science (9/24/76): 1268 for criticism. Quote: Emiliani (1980), p. 87. BACK

10. E.g., Calder (1975); note also the semi-popular article: Emiliani (1980). BACK

11. Hughes (1977); Hughes et al. (1977); Thomas and Bentley (1978). BACK

12. Mercer (1978), quotes pp. 321, 325. On Mercer see Broecker and Kunzig (2008), p. 147. BACK

13. Elliott (1977-89), vol. 1, 4/8/79. BACK

14. Thomas et al. (1979), p. 355. BACK

15. E.g., Herterich (1987). BACK

16. Revelle (1983); similarly Thomas et al. (1979); Bentley (1980) saw a possible ice sheet collapse in the next 500 years; but Bentley (1982) said melting could take thousands of years; this was disputed by Hughes (1982); Hollin (1980) tried to demonstrate an East Antarctic ice sheet surge about 95,000 years ago; for predictions of meter-scale rises, see Jones and Henderson-Sellers (1990), pp. 10-11, 15; a skeptic: Van der Veen (1985); Van der Veen (1988). BACK

17. IPCC (2001a), pp. 657-58. BACK

18. Etkins and Epstein (1982); Gornitz et al. (1982); an influential model-based calculation: Cubasch et al. (1992). BACK

19. "Most workers" project 0.5-1.5m rise in next 50-100 years if warming continues, according to Schneider (1989), p. 777; he cites i.a. Meier et al. (1985); this range was taken as plausible for 2100 in National Research Council (1987); but only a few cm rise by 2025 according to the most cited of these papers, Wigley and Raper (1987). BACK

20. Wigley and Raper (1987), p. 131. BACK

21. Barclay Kamb quoted by Walker (1999); the slippage was predicted by Blankenship et al. (1986). For this and other history see Bindschadler and Bentley (2002). BACK

22. Later work confirmed Antarctic ice as the source: Clark et al. (2002); Weaver et al. (2003); "debate": W. Sullivan, New York Times, May 2, 1995, p. C4. 16m rise: Bard et al. (1990); Hanebuth et al. (2000). Workshops: O'Reilly et al. (2012), p. 714. For all this see Thomas (2014) BACK

23. Doake and Vaughan (1991); Rott et al. (1996). BACK

24. Mercer (1978), p. 325; Vaughan and Doake (1996). BACK

24a. Notably models by Philippe Huybrechts, see O'Reilly et al. (2012), p. 715. BACK

25. Oppenheimer (1998); IPCC (2001a), pp. 678-79; "Unsolved Problem" was the title of Weertman (1976); repeated in Van der Veen and Oerlemans (1987), p. 14. BACK

26. Thomas (1993), p. 398; Oerlemans (1994); Dyurgerov and Meier (2000); Oerlemans (2005) surveyed glacier records around the world and found that "for the period from 1900 to 1980, 142 of the 144 glaciers retreated"; see review by Alley et al. (2005). BACK

27. NOTE that the papers I cite throughout this section are only examples of numerous papers by these and some other authors. A readable summary is in Broecker and Kunzig (2008), pp. 152-56. Wedging: Doake et al. (1998). Ice base melting: Rignot and Thomas (2002), "most important finding," p. 1505. Subsequent work pointing in the same direction included De Angelis and Skvarca (2003), who found that Antarctic grounded ice surged after an ice shelf breakup, and Bindschadler et al.(2003), who reported that a major West Antarctic ice stream started and stopped flowing as the tide went up and down. Breakup of an ice shelf (Larsen) leads to a speedup of glacier movement: Rignot et al. (2004), Scambos et al.(2004) (who also note lubrication by percolating water, see following note). WAIS models: Payne (2004); observed WAIS changes: Thomas (2004), Siegert (2004). Several papers by Rignot and colleagues document other Antarctic changes. "A lot shorter:" Robert Bindschadler in Larry Rohter, "Antarctica, Warming, Looks More Vulnerable," New York Times, 25 Jan. 2005, section D. See Holmes (2004) for discussion. BACK

28. Vaughan and Arthern (2007). BACK

29. Concern about Greenland glacier surging was spurred by Krabill et al. (1999). Lubrication: Zwally et al. (2002). For discussion see, e.g., Schiermeier (2004) and Bindschadler (2006). Darkening (notably by melt pools): Curry et al. (1995). "Mechanism:" Shepherd et al. (2004); for water percolating through the ice see Phillips et al. (2013). Front-end mechanism: Hughes (1986). Hansen raised the question of iceberg armadas, see this website's essay on rapid change and went so far as to call the Greenland ice a "ticking time bomb:" Hansen (2005), p. 275. For the rapid rise at the end of the last ice age see note 22. BACK

30a. Greenland ice stream acceleration: e.g., Rignot and Kanagaratnam (2006), also, small earthquakes in Greenland, caused by sliding glaciers, had become twice as frequent: Ekström et al. (2006). Slowdown: Howat et al. (2007). Mass loss: e.g., Chen et al. (2006); Luthcke et al. (2006); Zwally et al. (2005); another satellite: Pritchard (2009). Accelerating: Rignot et al. (2008); Velicogna (2009). Long-range study: van de Wal et al. (2008). A more recent study showing substantial mass loss: Csatho et al. (2014). Greenland 400,000 years ago: Reyes et al. (2014). BACK

30b. Denton and Hughes (1981), p. 431; "underbelly": Hughes (1982). BACK

30c. Most effective tool: Vaughan (2008), p. 71. Rignot (1998); Wingham et al. (1998). BACK

30d. Andrew Shepherd on BBC News, Aug. 13, 2009, online here; Wingham et al. (2009); NEEM community members (2013). On unreliability of models see O'Reilly et al. (2012), pp. 721-22; "there is still no robust, credible model for the interaction of melting ice sheets with the ocean," Holland and Holland (2015). Irreversible: see Favier et al. (2014); Rignot et al. (2014); Joughin et al. (2014); quote: Justin Gillis and Kenneth Chang, "Scientists Warn of Rising Oceans From Polar Melt," New York Times, May 12, 2014, p. 1. If the collapse had indeed been initiated, another computer study found, all of West Antarctica would "irrevocably" follow, Feldmann and Levermann (2015). Changes in the entire continent's total ice mass were difficult to determine; an increase 1992-2008 reported by Zwally et al. (2015 was widely publicized but disagreed with other studies. BACK

31. Pritchard et al. (2009), Young et al. (2011), Mengel and Levermann (2014), Aitken et al. (2016), Rintoul et al. (2016). BACK

31a. Around 2007 the science-attentive media (Science, Nature, New York Times, etc.) began to cover extensively the sea-ice data compiled by the U.S. National Snow & Ice Data Center. Striking decline in Arctic thickness: Kwok et al. (2009). BACK

31b. Antarctic sea ice: Manabe et al. (1991); Eric Steig, "Clarity on Antarctic sea ice,", Dec. 19, 2014, online here, gives an overview with references. More recent additions to the discussion include Armour et al. (2016), Meehl et al. (2016). BACK

31c. Church and White (2006); Hay et al. (2015); geological: Lambeck et al. (2014); Kopp et al. (2016). On recent acceleration see Fasullo et al. (2016). BACK

32. Despite measurements of total heat absorbed by the oceans by Levitus et al. (2000) and Levitus et al. (2001), "20th-century sea level remains an enigma — we do not know whether warming or melting was dominant, and the budget is far from closed," according to Munk (2003). Classic models: notably Huybrechts (1990). The grids in current models are still too coarse to simulate ice streams. IPCC (2001a), pp. 641-42, projected between 0.1 and 0.9 m rise including ice melting; IPCC (2007b), p. 13 projects 0.2 to 0.6 m explicitly excluding possible ice change surprises. See also Meehl et al. (2005), O'Reilly et al. (2012). BACK

32a. Rapid sea-level changes (10 meters within 1000 years) were found in ancient coral reefs: Thompson and Goldstein (2005); Blanchon et al. (2009) found a "2–3-m jump in sea level" in a century, presumably due to ice sheet instability, during a period warmer than the 20th century. "A rise of over 1 m by 2100 for strong warming scenarios cannot be ruled out," Rahmstorf (2007), extended by Vermeer and Rahmstorf (2009), who project sea-level rise from 1990 to 2100 in the range 75-190cm. The problem will be compounded in many river deltas (Nile, Ganges, Mississippi, etc.) by a half meter or so of subsidence as dams impound sediment and water is withdrawn from aquifers. BACK

32b. Shepherd et al. (2012); Box et al. (2012). Decade-scale collapse: Pollard et al. (2015). In 2016 a new analysis suggested satellite measurements had underestimated recent melting of Greenland ice by some 40%, Khan et al. (2016). On claims of political interference in the IPCC final report see Pearce (2007a) and response by Piers Foster et al., Letter, New Scientist (March 24, 2007), p. 26. I also draw here on my own conversations with glaciologists. See Oppenheimer et al. (2007); also Susan Solomon et al. and Michael Oppenheimer et al., Exchange of Letters, Science 319 (2008): 409-10. Motto: Christina L. Hulbe, "Are you pondering what I'm pondering? Time and change in the West Antarctic Ice Sheet," abstract of lecture, American Geophysical Union meeting, Dec. 13, 2016. BACK

32c.IPCC (2014a), pp. 25-26. Hansen et al. (2016) (published in draft as Hansen et al. (2015)); "far from resolved": Elizabeth Kolbert, "A New Climate-Change Danger Zone?," The New Yorker Daily Comment, July 23, 2015, online here. BACK

33. "For the first time" quoted Kunzig (2006), p.22. Emanuel (1987); Emanuel (2005a), published 4 Aug., found that "longer storm lifetimes and greater storm intensities... correlated with [higher] sea surface temperatures."BACK

34. Webster et al. (2005) (Sept. 16) found in all ocean basins "a large increase... in the number and proportion of hurricanes reaching categories 4 and 5." Also influential was a computer study, Knutson and Tuleya (2004), and a leading expert's insistence, in response to the violent 2004 storm season, that "hurricanes are changing," Trenberth (2005). A more thorough study by Elsner et al. (2008) found that the maximum wind speed of the strongest tropical cyclones had been increasing in the past 30 years by as much as 3 m/sec (6 mph) per decade, which would become significant if the trend continued through the century. See also, among others, Hoyos et al. (2006)Wu et al. (2006). Summary of models: Meehl et al. (2007), pp. 786-88. For the controversy: Pearce (2005b); Kunzig (2006); Witze (2006); Valerie Bauerlein, "Hurricane Debate Shatters Civility of Weather Science," Wall Street Journal, Feb. 2, 2006, p. 1; Emanuel (2005b); Mooney (2007a). "Remain a point of debate:" Kossin et al. (2007)."Future projections based on theory and high-resolution dynamical models consistently suggest that greenhouse warming will cause the globally averaged intensity of tropical cyclones to shift towards stronger storms," Knutson et al. (2010); Grinsted et al. (2013) projected "a twofold to sevenfold increase in the frequency of Katrina magnitude events for a 1°C rise in global temperature." Review: Sobel et al. (2016). An example of a model on extratropical storms predicting a rise in extreme precipitation events: Bengtsson et al. (2009). Superstorms: Hansen et al. (2016). BACK

35. Commitment to sea level change is summarized in Meehl et al. (2007), p. 752. High middle Miocene CO2 and sea levels: Tripati et al. (2009). During the last interglacial before the present the sea level was probably 7-8 or more meters higher: Kopp et al. (2009); see also Dutton and Lambeck (2012), Dutton et al. (2015). Levermann et al. (2013) calculate "a sea-level rise of approximately 2.3 m/ °C within the next 2,000 years." Ice sheet retreat: Levy et al. (2016). BACK

copyright © 2003-2017 Spencer Weart & American Institute of Physics