The White Planet: The Evolution and Future of Our Frozen World
Page 23
At least they couldn’t argue about that point using recent data, but skeptics always appeal to the records from the last millennium. It is true that the warming of the last decades is on the same order of magnitude as the variations in temperature associated with the Little Ice Age or with the medieval optimum, both of which were of natural origin. However, the recent warming is global while those two events were more regional. To attribute the current warming to natural causes, in particular to variations in solar activity, as is often advanced by skeptics, does not hold up to scrutiny. During the last centuries, the variations in radiative forcing due to those of solar activity are close to ten times less than those due to human activity over the same period. It would thus be necessary for the effect of solar activity to be considerably amplified in relation to that due to our activities for the Sun to lead the way today. Let’s imagine, however, that this has been the case; with the amplification working in the two directions, the year 2007 should have been relatively colder than five or six years earlier, since the eleven-year cycle that characterizes solar activity was at its minimum. This wasn’t true. The year 2007 was the fifth warmest year, and this in spite of the cooling of part of the Pacific associated with La Niña. Another difficulty for the skeptics: a variation in solar activity should produce a rather homogeneous modification in temperature over the entire vertical thickness of the atmosphere, whereas with an increase in the greenhouse effect we expect a warming of the troposphere and a cooling of the stratosphere. And this is indeed what was observed; the cooling of the stratosphere is in a certain sense the signature of the role of human activity.
We hope we have been convincing: attributing the recent warming to human activity, described by the IPCC as very probable, is built on solid reasoning in which it is really difficult to find any holes. And when we turn to our white planet, the evidence that warming is occurring is abundant. Indeed, this was the key conclusion that was reached as an outcome of the debate organized by the French Académie des Sciences in September 2010 following the above-mentioned open letter signed by more than six hundred scientists involved in climate research. French academicians have concluded that “the greenhouse gas increase, a large part of which is due to human activities, plays an essential role in the current warming.”
The White Planet on the Front Lines of Global Warming
When we examine the geographic distribution of the recent warming it appears that it has been less rapid at the surface of the oceans than on the continents, in fact twice as slow if we consider just the last two decades. The ocean has great inertia: it takes time for the additional heat to be transferred from the atmosphere to its surface layers. Thus it has trouble following the rhythm of warming, which explains a larger warming over the continents. This amplification is even greater in the regions of the Arctic, which during the twentieth century warmed on average twice as fast as the rest of the world, although in an uneven way.
One of the reasons for this polar amplification resides in the shrinking of the snow-covered surfaces of the Northern Hemisphere. At its highest level at the beginning of the spring, the snow cover, which reached 38 million km2 in the 1920s, was henceforth closer to 35, with a significant dip in the 1980s. During the last twenty-five years these snow-covered surfaces have decreased by around 2 million km2, about five times the surface of California; that amount of reflective surface was replaced by ground and vegetation, which absorb solar radiation. This change in albedo is far from being the only cause of polar amplification, as changes in the atmospheric circulation might also have contributed in a more important way.14 The snow cover has also been modified in mountain regions, as is seen by a thinner layer of snow, by the melting of the earlier and less abundant snow, or by a rise in the line of equilibrium corresponding to the isotherm 0°C.
The ice cover of the Arctic Ocean has also decreased, as we mentioned at the beginning of this work, to a remarkable degree when we take into account the data from 2007. Let’s recall that this ice is formed by the freezing of the seawater to a thickness of several meters at the most, generally less for ice formed during the year. The ice surface is smallest at the end of the summer, and that minimum level has decreased considerably. Between the beginning of the 1980s and 2005 it has gone from around 7.5 to 5.2 million km2. And to everyone’s surprise it suddenly decreased to 4.2 million km2 in September 2007, a much more rapid decrease than any of the sea ice models had indicated (Figure 12.6). This unexpected acceleration of the disappearance of sea ice is probably the result not only of warming but also of changes in atmospheric circulation. It is accompanied by a very clear decrease in the thickness of that ice, of which a smaller proportion corresponds to ice formed over many years. Because young ice is more fragile, one might think that 2007 was a hard year for this ice cover, which would have trouble coming back in the coming years. However, this minimum level in the covering of ice increased in 2008 and then again in 2009. It then decreased, reaching in 2011 a quite similar small sea ice extent as that observed in 2007. And 2012 has been again a record year with a minimum extent of ~ 3.4 million km2, considerably less than in 2007—a difference corresponding to about twice the size of California.
This change in the characteristics of the sea ice in the Arctic Ocean is also seen in the increased speed of the drift of the ice, as has been witnessed by the expedition of the sailboat Tara, which voluntarily became a prisoner of that ice in September 2006. Nine months later it had already traveled 1,000 kilometers, twice as rapidly as predicted, and came back to the open ocean at the beginning of 2008; its return had not been expected until the following summer. This acceleration in the drifting of the ice can moreover be held partly responsible for the decrease in the surface covered by the sea ice at the end of the summer of 2007. We noted earlier that this acceleration of the consequences of warming, particularly great in the Arctic, also seems to affect the coastal regions of Greenland. There, too, the recent data are extremely impressive because of the possible consequences in terms of sea level or other climatic events. The ice is flowing from the inside of the ice sheet toward the ocean through huge glaciers, true rivers of ice, at a flow rate that is already quite high but which, for some glaciers, has more than doubled during the last five years, reaching speeds of forty meters per day at the front of the glacier. That acceleration seems to be linked to two phenomena. The first is that the base of the glacier is lubricated and slides better as a result of the infiltration of liquid water linked to the melting that is accelerating in the coastal regions of Greenland. The second is the disappearance, here too due to warming, of certain ice shelves, which until recently were present at the mouth of these glaciers and slowed their flow into the ocean. Recent estimates, thanks to the interferometry radar work led by a team headed by Éric Rignot from the University of California,15 have shown that in 2005 Greenland lost more than 200 billion tons of ice, around twice as much as it did in 1996. As we indicated in chapter 2, this loss of mass has continued and has even increased.
Figure 12.6. Arctic September sea ice extent: observations and model runs. This illustrates the minimum extension of the sea ice, which decreased very sharply in 2007, and the predictions of models. Source: National Snow and Ice Data Center, after J. Stroeve, M. M. Holland, W. Meier, T. Scambos, and M. Serreze, “Arctic Sea Ice Decline: Faster than Forecast,” Geophysical Research Letters 34 (2007), L09501, doi:10.1029/2007GL029703.
The acceleration of these glaciers is recorded by networks of seismometers placed to detect and record earthquakes. When these enormous flows accelerate then slow down, they generate low-frequency seismic waves that can be distinguished from those caused by a regular earthquake and enable us to follow at a distance the way in which the flow of Greenland is evolving. The results corroborate those obtained from the satellite observation of the glaciers. A team of American researchers, led by Göran Ekström,16 has shown an increase in these mini glacial earthquakes since 2002. In 2005 we counted twice as many as in any other year prior to 2002.
Additional proof of this warming in high latitudes of the Northern Hemisphere is in the lakes and rivers that on average freeze later and thaw earlier than they did in the middle of the nineteenth century—by around six days in both directions. The permafrost is also affected, with a warming on the surface of more than 3°C compared to the temperature in 1980 and whose base thaws by a few centimeters every year. The same is true of the ground that freezes and thaws each year, whose surface has decreased by 7% since the middle of the twentieth century and whose maximum thickness has decreased by 30 centimeters.
And what about Antarctica? Unlike the Arctic, warming there is not as noticeable compared to the global average, probably because it is surrounded by a huge ocean and not by continents. There were no sufficiently viable data for the sea ice covering around the Antarctic continent before the satellite data were available at the beginning of the 1970s, and since then the minimum values of the sea ice covering, which are quite variable from one year to the next, do not show a significant change. In fact, the most visible events around Antarctica, whose coastal regions are colder than those of Greenland and thus a priori less sensitive to a warming of 1 or 2°C, concern the very thick ice shelves that carry the ice cap far into the ocean. Not the huge Ross and Filchner ice shelves, which are as large as France and not much in the news, but those that border the Antarctic Peninsula.
Among them is the Larsen B Ice Shelf, which was in the headlines in February and March 2002. Despite a thickness greater than 200 meters in some places, more than 3,000 km2, close to the size of Rhode Island, has broken off and scattered in the form of enormous icebergs, which have then melted in the warmer regions of the ocean. Over the preceding five years that ice shelf had already lost almost the equivalent surface, and during the last few decades around 12,500 km2 of ice shelves have disappeared. And this in a region where they were present since the beginning of the current Holocene period, around 10,000 years ago.17 It is probable that this disappearance is linked to the particularly pronounced warming of the Antarctic Peninsula (by more than 2°C during the last fifty years), but it is tricky to make a connection between that occurrence and global warming. It is important to note that the gradual warming of the ocean waters circulating under such ice shelves also likely plays a role. Finally, the disappearance of the Larsen B Ice Shelf resulted in an acceleration of the glaciers in this part of the Antarctic Peninsula, whose flow toward the ocean is henceforth much easier. At the beginning of 2008 part of the Wilkins Ice Shelf, located at the southwest of the Antarctic Peninsula, broke apart.
Thanks to the work of Éric Rignot and his coauthors,18 we are beginning to better assess the mass of the different basins of Antarctica over the recent period. Rignot’s study, which covers the period 1992–2006, confirms that East Antarctica is in quasi-balance but shows very notable losses over West Antarctica and the Antarctic Peninsula, which have reached close to 200 billion tons in 2006. These figures are comparable to those that we have cited for Greenland. The phenomenon has amplified in the past few years such that Antarctica on the whole henceforth has a negative mass.
Wherever we look, with the exception of the sea ice surrounding the Antarctic continent and East Antarctica, all the components of the cryosphere—snow-covered surfaces, frozen ground, frozen rivers and lakes, sea ice, ice shelves, glaciers, and ice sheets—thus appear to be affected by climate warming, with a clear acceleration over the last decade. The consequence of this is a significant contribution of glaciers, ice caps, and ice sheets to a rise in the sea level; thus over the period 1993–2003 those factors are at the origin of nearly half of the three-centimeter observed increase in the water, the largest part of which is due to glaciers and ice caps.
The 400 billion tons that sum up the summer losses of the ice sheets, in 2005 for Greenland and in 2006 for Antarctica, have contributed to a rise in the sea level of more than one millimeter, and this over the course of just one year. Since then, the annual losses to the ice sheets have, on average, been greater.
CHAPTER 13
What Will the Climate Be in the Future?
Attributing climate change to human activity has been the subject of debates ever since the creation of the IPCC. It will take another few years, probably a decade, perhaps more, for this to become an uncontested assertion. Continuing to acquire quality data, better understand the role of aerosols, better identify the natural causes, know more precisely the sensitivity of the climate—in other words its reaction vis-à-vis a modification in the radiative forcing, as well as its natural variability and causes—are the directions in which we must collectively progress. However, there is a near certainty that we must face when we turn to the future: the climate will continue to warm. Why is this? Quite simply because even if, by waving a magic wand, we were able to instantly stabilize the greenhouse effect, the climatic system has an inertia such that it would still warm by an amount almost equivalent to what has occurred over the twentieth century. In fact, limiting warming to 2°C, even compared to the climate at the end of the twentieth century, is a true challenge, whose different facets we will examine in a later chapter.
Let’s first look at a climatic future in which the need to stabilize the greenhouse effect would not be taken into account. This is in fact the process widely undertaken by the IPCC experts in the 2007 report of the scientific group, which we will look at broadly here for that which involves the global climate and in more detail for that which concerns our white planet: glaciers, sea ice, frozen ground. But we will also look at the sea level, whose rise is largely associated with the behavior of polar ice caps and inland ice sheets, and at the Gulf Stream, whose evolution is strongly influenced by that of the climate in high latitudes.
A True Upheaval if We Aren’t Careful
The only way to predict the climate of the future is to use climatic models. Granted, there were periods in the past that were warmer than the current climate, for example, during the last interglacial period 125,000 years ago. But the origin of that warming related to a different amount of sun radiation has nothing to do with an increase in the greenhouse effect. That climate cannot therefore be used as an analogue of the future warmer climate, even if it is full of information, for example, about the evolution of the sea level.
These global climate models have been considerably improved since the work of the first climate modelers, who took only the atmosphere into consideration. The aim of such a model is to simulate the movements of the atmosphere but also to predict precipitation, evaporation, and more generally everything that has to do with the cycle of water. The basic equations for dynamics of the air, of physics for that which bears on the water cycle, are similar to those used to predict the weather. But whereas the meteorologist follows the evolution of perturbations from their origin until he loses any trace of them at the end of a few days, the climatologist looks at the average readings of temperatures, winds, and precipitation.
In the 1980s interactions between land surfaces and the atmosphere were integrated, whereas oceanic models were developed separately. At the beginning of the 1990s coupled models included the atmosphere, the continents, the oceans, and the ice, then they began to take into account the role of aerosols and the carbon cycle, and more recently the chemistry of the atmosphere and the dynamics of vegetation. There are currently more than twenty of these models throughout the world, including two in France, one at Météo France, the other at the Institut Pierre-Simon-Laplace; in the United States, General Circulation Models were first developed at Princeton (Geophysical Fluid Dynamics Laboratory [GFDL]), New York (NASA/GISS), and Boulder (National Center for Atmospheric Research [NCAR]). All these models differ; some emphasize the carbon cycle or the role of aerosols, while others focus on vegetation or on the interactions between the atmosphere and the ocean. This diversity is a rich source of information.
The confrontation with reality illustrates the quality of these climatic models developed during the last decades. Their ability to simulate broad
characteristics of atmospheric circulation was quickly demonstrated. They take into account the rise of humid air in equatorial regions, the fall of dried air in subtropical regions where the great deserts are located, and the more horizontal circulation in our latitudes around depressions and anticyclones. We can verify that these models reproduce the climate remarkably well, not only during a year month after month, season after season, but that they are also able to describe the climate over longer periods. This is seen in the comparison of simulations of the climate of the twentieth century with variations observed over that period (Figure 12.4) or climates reigning over other planets: Mars and Venus, with such different conditions, or over our planet Earth during the extreme conditions of the Last Glacial Maximum. Or, closer in time, of those that prevailed six thousand years ago with a more pronounced cycle of monsoons in response to a sharp difference between the amount of sun in the summer and the winter.
Despite all of this progress, these climatic simulations, which require many computing hours, still convey some uncertainty. Thus our knowledge of climate sensitivity has scarcely progressed in thirty years. And yet there is a simple exercise, identical for each model, which consists of simulating the climate for a concentration of CO2 multiplied by two. Between 1.5 and 4.5 in 1979, the range of the results has tightened only slightly; it goes from 2 to 4.5°C in 2007 with the most probable number of 3°C. Somewhat anecdotally, we were correct in indicating, as seen in the article published in 1990 conjointly with Jim Hansen and Hervé Le Treut, that the paleoclimatic data were consistent with a climatic sensitivity of 3 to 4°C. But more important is the fact that our scientific community henceforth has confidence in the lower value of that sensitivity, probably equal to 2°C and not lower than 1.5°C. There are thus indeed processes of amplification, connected first to water vapor, that are at work in the climatic system and not mechanisms of compensation such as those suggested by the American climatologist Richard Lindzen.