Greenland GISP Ice Core O

Figure 1.14: Oxygen isotope data from the GISP-2 Greenland ice core. Larger (less negative) values correspond to warmer temperatures.

of a glacial lake into the ocean, but generally speaking the mechanisms of both the Younger Dryas and of millennial variability remain as Big Questions that are yet to be resolved. This is an especially important question because the flat, quiescent period at the end of the Greenland record - the Holocene period we have lived in for all of the history of civilization - represents an abrupt cessation of high amplitude millennial variability. What accounts for the uncommonly stable climate of the Holocene, and what would it take to break this situation? This is a Big Question with considerable import indeed.

1.11 Holocene climate variation

The climate variations of the Holocene have been (at least so far) more subtle than the massive variations we have discussed previously. In part, this is simply because the Holocene is a short period of time, and there hasn't been enough time for something really dramatic to happen, given the relatively slow pace of many of the geological processes that modify climate. But the limited span of the Holocene is not the whole story. The Holocene has not witnessed extreme millennial scale variability of the sort seen in the preceding glacial time. Indeed, one of the Big Questions concerning the Holocene is the physical basis for the relatively stable Holocene climate. In particular, given the massive assault on climate by industrial society (see Section 1.12) it becomes all the more pressing to understand what it would take to break the equable and steady Holocene climate enjoyed during the rise of civilization.

Still, the Holocene has not been without its points of interest so far as climate variations go, the more so because there were human civilizations around to witness and be affected by these variations. A key driver of Holocene climate change is the precessional cycle, to be discussed in Chapter 7. It is the same precessional cycle that plays a role in the rhythm of the Pleistocene ice ages. The Earth's spin axis precesses like a top, so that the way the "tilt seasons" line up with the "distance seasons" associated with the varying distance of the Earth from the Sun goes through a cycle lasting about 22,000 years. A quarter cycle is only 5500 years, which brings these cycles within the span of recorded history. At present, the Earth is farthest from the sun when the Northern Hemisphere points towards from the Sun (i.e. during Northern summer), and is closest to the Sun when the Northern Hemisphere points away (i.e. during Northern winter). This gives us relatively warm winters and relatively cool summers. 11000 years ago the situation was reversed, and Northern Hemisphere summers received considerably more sunlight than they do today, particularly at high latitudes. This should have made polar regions considerably warmer than today, but for reasons that are only partly understood, the time of warmest summers was delayed several thousand years, to a time called rather tendentiously the Climatic Optimum about 7000 years ago. It is not entirely clear what the climate was "optimal" for, and this is in any event a Northern-centric view as Southern hemisphere continents if anything experienced a weak cooling at this time (as one would expect from the nature of the precessional cycle). The alternate term Altithermal is preferred, as being less value-laden. In any event, at this time the Northern polar regions were getting up to 10% more solar radiation in summer than they are today, leading to warmer summers. The warming is expected to be greatest over land, since oceans average out the warm summers and cold winters. Tree ring records indicate that high-latitude land masses were between 2C and 4C warmer in the summer at the time of the Altithermal. These estimates are corroborated by an expansion of the tree line to higher altitude land in the European Arctic. Cold-tolerant trees are primarily sensitive to growing-season temperatures, and are little effected by a moderate decrease in winter temperatures; in fact, more severe winters can in some cases be favorable to tree growth, since cold winters interrupt the life cycle of various insect pests. While the Altithermal is certainly connected with the precessional cycle, the magnitude of the warming, the cause of the delay in warming (partly associated with leftover glaciers from the last ice age), and the role of ocean circulations and vegetation changes in the altithermal, are all active subjects of research.

The precessional cycle has also had a profound effect on the distribution of precipitation in low latitudes. The Sahara is a desert today, but the dry river features known as wadis were flowing with water six thousand years ago, at which time the desert was a savannah grassland. This wet period commenced about 14,500 years ago and the Sahara abruptly reverted to desert about 4700 years ago. There are also intriguing indications that the time of initiation of present tropical mountain glaciers follows a precessional cycle. For example, the Peruvian Andean glaciers of the Southern Hemisphere date back to the last ice age, while the Kilimanjaro ice fields were laid down during the African Humid Period about 10,000 years ago and Himalayan glaciers of the northern subtropics tend to be even younger. The connection between the seasonal cycle of solar radiation and the tropical precipitation distribution involves atmospheric circulations - monsoons and the Hadley circulation - that cannot be treated without a full understanding of atmospheric fluid dynamics. We will therefore have only limited opportunities to pursue the precessional precipitation cycles in the course of this book, though the treatment of the precessional cycle in solar forcing will provide the student with the necessary background for further study.

The Little Ice Age is another Holocene climate fluctuation of considerable interest. This term refers to a period of generally cool Northern Hemisphere extratropical land temperatures extending from approximately 1500 to 1800. Tree ring estimates suggest the Northern Hemisphere mean temperature dropped by something over 0.5C between the year 1400 and 1600. The cooling is corroborated by records of advances of mountain glaciers, sailors' observations of sea ice, and agricultural records. The Little Ice Age is too short and too recent to have anything to do with the precessional cycle, and while it is possible that fluctuations in the ocean circulation could have produced the cooling, the prime candidate for an explanation of the Little Ice age is a temporary slight dimming of the Sun. Sunspot observations do indicate a cessation in the normal solar sunspot cycle - called the Maunder Minimum- at about the time of the Little Ice Age. However, most estimates of the associate solar output change are far too small to yield a significant cooling.

Various mechanisms are under investigation which could amplify the response to the small solar fluctuation, but in the grand scheme of things the Little Ice Age is a rather subtle event, and accordingly hard to understand, particularly in terms of simple models.

To put the Holocene in perspective, it is salutory to note that the "abrupt" PETM event discussed in Section 1.9.1 lasted nearly 200,000 years, and set in over a period of around 10,000 years - as long as the full length of the Holocene. There really hasn't been much time for things to happen in the Holocene, and the time span of Figure 1.9 no doubt contains many 10,000 year periods as quiescent as the Holocene has been up until recently. Short as the Holocene is, we will see next that human activities have been able to cause some dramatic changes in the composition of the atmosphere and consequently in the Earth's climate. One wonders what it would take to trigger a hyperthermal event such as the PETM, which is over in the wink of an eye by the standards of Figure 1.9, but has a duration twenty times as long as the span of human civilization to date.

1.12 Back to home: Global Warming

We have seen that CO2 has been a major factor in determining climate througout Earth's history, and that life, in turn, has greatly shaped the carbon cycle. Life is in the midst of disrupting the carbon cycle once more, but this time it's technological life that has provided the necessary innovations. Over billions of years, a great deal of organic carbon has been sequestered in the Earth's crust without oxidizing. Most of this carbon is in very dilute forms which cannot easily be tapped to provide economically useful amounts of energy. However, a very small portion winds up in nearly pure forms that are moreover chemically altered in a fashion that makes them especially convenient as fuels. These are the fossil fuels - coal formed from land plants and oil formed in marine environments. Natural gas can be produced by thermal alteration of either coal or oil. Fossil fuels represent concentrated solar energy stored in the form of organic carbon, which has been accumulating over hundreds of millions of years. This pool of readily oxidizable carbon exists precisely because it is in geological formations that have kept it apart from oxygen over the ages. It is only the evolution of technical civilization that is making it possible to dig up and oxidize hundreds of millions of years worth of stored fossil carbon within a few centuries.

In the year 2005, over 8 gigatonnes (8 • 1012kg) of carbon were released by fossil fuel burning, and annual emissions continue to grow rapidly. There are several ways to see that this is a very big number - a major upset to the natural carbon cycle. First, the pre-industrial atmosphere contained about 600 gigatonnes of carbon, so the 2005 annual emission is fully 1.3% of the undisturbed atmospheric content. If the same amount were released into the atmosphere each year, it would take only 75 years to double the atmospheric CO2 content, provided all the released CO2 stayed in the atmosphere. Alternately, one could compare the fossil fuel emissions to the volcanic outgassing which in the long term balances silicate weathering and sustains the carbon cycle. Precise estimates of volcanic outgassing are hard to come by, but generally are on the order of 0.1 gigatonnes of carbon per year or less. Thus, fossil fuel carbon emissions are eighty timers larger than background volcanic outgassing. In fact, the very largest carbon flux number involved in the whole carbon cycle is the net CO2 carbon fixed into organic carbon each year by worldwide photosynthesis, and fossil fuel emissions even look impressive when compared to this number. Based on satellite chlorophyll observations, it has been estimated that photosynthesis fixes 100 gigatonnes of carbon each year, about half on land and half in the oceans. The year 2005 fossil fuel emissions were fully 8% of this number. In other words, worldwide photosynthetic productivity would have to increase by 8% to take up the fossil fuel CO2 and 100% of that carbon would have to be buried as organic matter without being recycled by respiration. That, of course, would be a completely absurd situation, as virtually all of the photosynthetically fixed carbon is quickly respired back into the atmosphere, largely by bacteria who have had several billion years to become proficient at making use of organic carbon wherever they find it. As an example, land photosynthesis fixes about 50 gigatonnes of carbon each year, but the flux of organic carbon to the oceans in all the world's rivers is a mere 0.4 gigatonnes per year (one twentieth of fossil fuel carbon emissions). And there is no evidence that much of the remainder of the photosynthetically fixed carbon is remaining on land as soil organic carbon. To say that humans have become a force of geological proportions vastly understates the case, for by this measure human influences on the carbon cycle overwhelmingly dominate the natural sources.

The result of all our busy digging and burning has been a steady increase in atmospheric CO2. Figure 1.15 shows the time series of atmospheric CO2 concentration since 1750. CO2 has a very long atmospheric lifetime, so it is well-mixed. In consequence, one finds nearly the same CO2 concentration whereever one measures it, so long as the measurement is not in the immediate vicinity of major sources or sinks. The part of the record since 1950 comes from direct analyses of air samples at the Mauna Loa observatory, whereas the earlier part of the record comes from air trapped in bubbles in the ice of the Siple Dome site, Antarctica, but the two records match up well where they meet. At the dawn of the industrial era CO2 concentrations are near 280 molecules per million ( ppmv for short), right where they were left at the end of the most recent ice age. After 1750 the concentrations begin to rise, and by 2007 the concentrations have exceeded 380ppmv - fully 35% above the pre-industrial value. Most of the increase has happened since the mid-twentieth century, and the rate of increase seems to be accelerating along with population and economic growth.

Not all of the carbon released by fossil fuel burning has remained in the atmosphere. Estimates based on careful historical inventories suggest that only about half of the total carbon released to date remains in the atmosphere as carbon dioxide. Most of the remainder has slowly infiltrated the ocean, with a lesser amount having been taken up by the terrestrial ecosystem (net of deforestation). In fact, it has been shown that the rate at which the ocean can take up the excess CO2 is limited by the mixing between the upper ocean and the deep ocean. This is a slow process, and if all fossil fuel burning were to suddenly cease, it would take in excess of 600 years for 80% of the excess CO2 to be taken out of the atmosphere. The remainder would stay in the atmosphere for millennia longer, owing to certain chemical processes (discussed briefly in Chapter 8) which limit the ability of the ocean to take up CO2. The slow net removal rate of CO2 allows fossil fuel emissions to accumulate in the atmosphere. Another consequence of the long lifetime of CO2 in the atmosphere is that the climatic effects of elevated CO2 will persist for centuries to millennia, even after any (much to be hoped-for) dramatic restriction of fossil-fuel burning. Allowing for uptake by the ocean, there are enough fossil fuel reserves- primarily in the form of coal -to ultimately increase the atmospheric CO2 concentration to at least six times the pre-industrial value. The number could go much higher if the ocean sink were to become less efficient, or if land ecosystems were to turn around and become a CO2 source rather than a sink.

This all leads us to a series of very Big Questions: if the rise in CO2 is allowed to continue to a doubling of the pre-industrial value, how much will the Earth warm? How will the warming be distributed? How much will sea level rise as a result of melting land ice and thermal expansion of ocean water? What will happen to precipitation patterns? How will all of this affect human societies and natural ecosystems? The basic physics needed to treat these questions is identical to what is used to account for the influence of CO2 and other long lived greenhouse gases on past climates. The problem in this instance has more immediacy as many generations of our descendents will be living with the consequences of our fossil fuel emissions in the next several

400 380 360 340 320 300 280

-Mauna Loa ■Siple Dome

1750 1800

1850 1900 Year

1950 2000

Figure 1.15: Annual mean CO2 concentration from 1750 to the time of writing. The earlier part of the record is from air trapped in bubbles in the Siple Dome Antarctic ice core. The more recent part of the record is from instrumental measurements at the Mauna Loa observatory. The units are molecules of CO2 per million molecules of air.

decades. In order to understand what kind of planet we are leaving these descendents, there is a demand for greater detail in the understanding of the climate changes to be wrought by these rapid increases in atmospheric CO2.

Interest in the effect of CO2 changes on climate long predates the kind of data shown in Figure 1.15 which showed that CO2 was on the increase, and in fact predates the realization that human activities really could cause CO2 to increase appreciably. Likewise, global warming was a concern long before it was confirmed that the Earth really was warming in response to increases of CO2. These things were all anticipated theoretically a century or more before global warming burst onto the scene as an issue of political consequence, and the driving force was basic curiosity about the physics governing planetary temperature. It's a line of inquiry that extends right back to Fourier's pathbreaking inquiry into how an atmosphere affects the energy budget of a planet, and hence its temperature. The discovery of global warming is a great triumph of two centuries of developments in fundamental physics and chemistry. It is not a matter of people having noticed that both CO2 and temperature were going up, and concluding that the first must be somehow causeing the second. Both the rise of CO2 as a consequence of fossil fuel burning, and the consequent rise in temperature as a response to the Earth's perturbed energy balance, were anticipated long before either was observed.

After Fourier, the tale resumes with Tyndall, whose work on the infrared absorption of CO2 and water vapor was mentioned near the beginning of this chapter. Tyndall was interested in these gases because of the questions raised by Fourier regarding the factors governing planetary temperature. He was also interested in the recently-discovered phenomenon of the ice-ages, and with several contemporaries thought perhaps ice ages could arise from a reduction in CO2. In that, he was partly right; the Pleistocene ice ages are cold partly because of the glacial interglacial CO2 cycle, even though the ultimate pacemaker of the ice ages is the rhythm of Earth's orbital parameters. Tyndall died, however, before he ever had the chance to translate his measurements into a computation of the Earth's temperature. That task was left to the Swedish physical chemist Svante

Arrhenius, who in 1896 performed the first self-consistent calculation of the Earth's temperature incorporating the greenhouse effect of water vapor and CO2. Interestingly, Tyndall's measurements were not sufficient to provide the information about weak absorption over long path lengths, so for the absorption data he needed he turned to Langley's observations of infrared emitted by the Moon. It was a felicitous re-use of data intended originally for determination of the Moon's temperature, and indeed was a more correct use of the data than Langley was able to accomplish. This shows the benefit of curiosity-driven science: measurements taken to satisfy curiosity about lunar temperature wound up being instrumental in permitting an evaluation of the effect of the Earth's atmosphere on the Earth's temperature. Astronomers initiated the study of infrared as an observational technique, but the radiative transfer work stimulated by their needs soon provided the crucial tool needed to understand planetary climate. Arrhenius not only estimated the Earth's then-current temperature, but also estimated how much it would warm if the amount of CO2 in the atmosphere were to double. Using clever scaling analyses from Langley's data, he was able to do this without a firm knowledge of just what the atmosphere's CO2 content actually was. Not long afterwards, he realized that industrial burning of coal was dumping CO2 into the atmosphere, and could eventually bring about a doubling; he described this process as "evaporating our coal mines into the atmosphere." At then-current rates of consumption, it appeared that a doubling would take up to a millennium, and Arrhenius would no doubt have been surprised to know that his own great-grandchildren could well live to witness the doubling. This takes our story to about 1900. What happened then?

A long hiatus. In part, there was little sense of urgency, because a failure to anticipate the explosive growth of fossil fuel use looming in the coming decades led to a belief that any problem was off in the far-distant future. Besides that, two unfortunate turns of events held back the study of global warming for decades. The first was a highly touted experimental study published in 1900 by the prominent physicist Knut Angstrom, which purported to show that the radiative effects of CO2 are "saturated," i.e. that the gas already absorbs as much as it can at the atmosphere's then-present concentration, so increases would have no effect. A concomitant and closely associated (and equally wrong) idea was that the strong absorption of water vapor would completely swamp any effect CO2 might have. The experiment turned out to be wrong, but such was Angstrom's reputation and such was the resistance to the idea that humans could change climate that it was decades before anybody definitively checked the result. Moreover, it turns out that even if Angstrom had been right, it would not have negated the greenhouse effect; this misunderstanding hinged on the poorly developed understanding of radiative transfer in a temperature stratified atmosphere. The "grey gases" we will study in the first half of Chapter 4 are "saturated" in the sense of Angstrom, but nonetheless allow for an increase in the greenhouse effect as more greenhouse gas is added to the atmosphere. The second barrier to progress was the belief that the huge carbon content of the ocean would buffer the atmosphere, overwhelming anything human industry could have thrown at it. The carbonate chemistry needed to defeat this idea was largely worked out by the 1930's;indeed, it requires nothing more than is taught routinely in high-school chemistry courses today. However, it had not been assimilated into a coherent and widely appreciated picture of the uptake rate of CO2 by the oceans, in part because of lack of knowledge of the rate of mixing between the upper ocean and the deep ocean. A paper published by Revelle and Suess in 1957 is widely credited with having broken the logjam, but in fact mentions the essential carbonate buffering mechanism (fully worked out by earlier researchers, and cited as such) almost as an afterthought. The attempt of Revelle and Suess no doubt helped to revive interest in the question of oceanic CO2 uptake rates, but in fact the paper came to exactly the wrong conclusion - that fossil fuel emissions were unlikely to lead to any significant increase in atmospheric CO2 concentration. The true implications of the carbonate buffer for CO2 increase due to fossil fuel burning was finally brought out clearly in a paper two years later, by Bert Bolin and Erik Eriksson.

Despite Revelle and Suess's conclusion, the idea gained hold that somebody should actually systematically check and see what atmospheric CO2 was doing. This program was initiated by Charles Keeling while at Caltech, and was subsequently encouraged by Revelle. Keeling's work culminated in the Mauna Loa data shown in Fig. 1.15. The techniques for recovering past CO2 from air bubbles trapped in ice were not to be developed until the 1970's, so Keeling had to wait a decade or so before it was clear that CO2 was really rising, and a bit more time after that before there was a clear idea of just how high CO2 already was relative to the pre-industrial value. This work, together with developments in infrared radiative transfer stimulated by astronomical observation and military interest in infrared target detection lead to new breakthroughs in the formulation of radiative transfer. The work culminated in 1967 with the calculation by Manabe and Wetherald of the Earth's temperature using modern radiative physics. They were also able to calculate the warming due to a doubling of CO2, allowing for expected changes in water vapor content as the planet warmed. This was not the end of the story, which indeed continues today, since there was much to be done in terms of embedding the radiative transfer in a fully consistent computation incorporating the fluid dynamics of the atmosphere - a general circulation model. It was, however, the beginning of the modern chapter of the study of global warming. With the publication of the Charney report by the US National Academy of Sciences in 1979, global warming began to be perceived as a real threat. The powers that be were slow to awaken to the magnitude of the problem, and several more years were to pass before the creation of the Intergovernmental Panel on Climate Change in 1988, which initiated regular, comprehensive surveys of the state of the science surrounding global warming. At the time of writing, the world still awaits substantive action to curb fossil fuel emissions.

All aspects of the essential chemistry, radiative physics and thermodynamics underlying the prediction of human-caused global warming have been verified in numerous laboratory experiments or observations of the Earth and other planets. Other aspects of the effect of increasing greenhouse gases rely on complex collective behavior of the interacting parts of the climate system; this includes behavior of clouds and water vapor, sea ice and snow, and redistribution of heat by atmospheric winds and ocean currents. Such things are impossible to test in laboratory experiments. To some extent, aspects of our theories of the collective behavior have been tested against the seasonal cycle of Earth, interannual variability, and past climates, as well as attempts to simulate other planetary climates. The ultimate test of the theory, though, is to verify it against the uncontrolled and inadvertent experiment we are conducting on Earth's own climate. Can we see the predicted warming in data? This is not an easy task. For one thing, the atmospheric CO2 increase is only a small part of the way towards doubling, and the climate has not even fully adjusted to the effect of this amount of extra radiative forcing: oceans take time to warm up, and delay the effect for many years (for reasons to be discussed in Chapter 7. Thus, so far the signal of the human imprint on climate is fairly small. Set against that is a fair amount of noise complicating the detection of the signal. Climate, even unperturbed by human influence, is not steady from year to year, but is subject to a certain amount of natural variability. This can be due to volcanic eruptions and subtle variations in the brightness of the Sun. There are also various natural cycles in the ocean-atmosphere system that cause the planet to be a bit warmer or colder from one year to the next. Chief among these is the El Niño phenomenon of the tropical Pacific. During El Niño years, the coupled dynamics of the tropical ocean and atmosphere causes warm water to spread throughout the Pacific, leading to a warming of mean surface temperatures both in the tropics and further afield. La Niña years represent a bunching up of the warm water, and an accentuated upwelling of cold water, leading to cold years. The two phases alternate erratically, with a typical time scale of three to five years.

The fact that the signal is hard to detect does not mean that global warming is of little consequence. The difficulty arises precisely because we are trying to detect the signal before it becomes so overwhelmingly large as to be obvious. Given the long lifetime of CO2 in the atmosphere, it would be highly desirable to keep the signal from ever getting that large, as if it ever does it will take many centuries to subside. Let's now take a look at some of the data, and see if there are any signs that the theoretically anticipated warming is really taking place.

Figure 1.16 shows a times series of estimated global mean surface temperature, based on recorded temperatures measured with thermometers. There is a lot of arduous statistics and data archaeology behind this simple little curve. Particularly for the data going into the early part of the curve, there has been a need to standardize measurements to allow for the various different ways of taking a temperature reading. Most of the oceanic measurements, for example, were taken by commercial or military ships of one sort or another, and some of the entries in ships' logs record things like "bait tank temperature" or "engine inlet temperature." There has also been a need to screen out stations that have been strongly affected by local land use changes such as urbanization, and to avoid spurious trends due to changes in the spatial distribution of temperature measurements (e.g. fewer Antarctic readings once Antarctic whaling essentially ceased). To help correct for biases in individual classes of temperature measurements, the long-term trends are presented as anomalies relative to the average of a station's long-term average standardized to a fixed base period (e.g. 1951-1980 for the data in Fig. 1.16).

There is little temperature trend between 1880 and 1920, but between 1920 and 2005 the temperature has risen by nearly 1C. The rise hasn't been steady and uninterrupted, however. It takes the form of an early rise between 1920 and 1940, followed by a 30 year period when temperatures remained fairly flat, whereafter the temperature rise resumes and has continued to the present. Given that CO2 has been rising at an ever-increasing rate over the industrial period, why was the warming interrupted between 1940 and 1970? The answer lies largely in another effect of human activities on climate. Burning of fossil fuels, and especially coal, releases sulfur compounds into the atmosphere which form tiny highly reflective droplets known as sulfate aerosols. By 1995,the effect was finally quantified with sufficient accuracy to permit reasonable estimates of the effect, and it began to appear that most of the evolution of climate of the twentieth and twenty-first centuries could be accounted for by a combination of rising greenhouse gases (mainly CO2 ) due to human activity, with an offsetting cooling effect of sulfate aerosols. The reason small particles are so good at scattering light back to space is discussed in Chapter 5, where the optical properties of sulfate aerosols will be discussed in detail. By the year 2000, the greenhouse warming signal had unquestionably risen above both the noise of natural variability and the offsetting effect of aerosol cooling. Sulfur is an active element in many actual and hypothetical planetary atmospheres, and so the study of sulfate aerosols on Earth informs other planetary problems, including the clouds of Venus and Venus-like extrasolar planets.

The Earth's emissions in the microwave spectrum have been monitored continuously by satellite-borne instruments since 1979, and these observations make it possible in principle to obtain reconstructions of atmospheric temperature trends which are independent of the somewhat inhomogeneous surface station network. Processing the microwave data acccurately enough to obtain reliable temperature trends proved very difficult, and there were many false steps along the way. Nonetheless, the main problems were resolved by early in the twenty-first century. The microwave temperature retrievals give the temperature of the atmosphere averaged over fairly deep layers, in constrast to the surface stations which measure near-surface air temperature. The left panel of Figure 1.17 shows the satellite retrieval of temperature in layer of the atmosphere known as the lower troposphere - extending from sea level to roughly 5 km in altitude. The satellite record tracks the GISS surface station record very closely, with the exception of the very strong 1997 El

Ï GISS Surface Station Data


1880 1900 1920 1940 1960 1980 2000


1880 1900 1920 1940 1960 1980 2000


Figure 1.16: Global average annual mean surface temperature since 1870, estimated from surface temperature observations. Data source is the NASA GISS surface station analysis. The temperature is given as an anomaly relative to the mean temperature for the years 1951 to 1980. To turn these into actual global mean temperatures in degrees Celsius, add 14C to the anomaly.

Niño, during which the satellite indicates that the lower tropospheric layer warmed considerably more than the near-surface air. Both satellite and GISS records reproduce the cooling caused by the El Chicon eruption (which overwhelmed the 1982 El Niño) and the Pinatubo eruption (which accentuated the La Niña cooling following the 1991 El Niño, leading to a very cold year). The substantial agreement between the satellite and surface station record proves beyond doubt that the warming observed in recent times is not an artifact of any supposed inadequacies of the surface station record.

The situation looks quite different higher up in the atmosphere. The right panel of Fig. 1.17 shows the temperature trend in a portion of the atmosphere called the lower stratosphere, extending from about 15 to 25 km in altitude. Here, volcanic eruptions produce a pronounced warming, as opposed to the cooling seen at lower layers. This suggests that volcanic aerosols heat the upper atmosphere by absorbing sunlight. The pattern of upper level cooling and lower level warming produced by high altitude solar-absorbing layers will be discussed in Chapter 4, and the reflective effect of aerosols will be brought into the picture in Chapter 5. Leaving out the warm spikes associated with volcanic eruptions, the lower stratosphere appears to have undergone a pronounced cooling over the span of the satellite record. Is stratospheric cooling compatible with CO2-induced warming in the lower troposphere? This is a Big Question that is resolved in Chapter 4. Ozone destruction also cools the stratosphere, since ozone absorbs sunlight. That portion of the cooling should go away as ozone recovers as a consequence of the Montreal Protocol banning ozone-destroying chlorofluorocarbons.

The Big Question of how much the Earth will warm upon a doubling or quadrupling of CO2, and how fast it will do so, engages a number of associated questions. Insofar as water vapor is itself a powerful greenhouse gas, any tendency for water vapor content to increase with temperature will amplify the warming caused by CO2. This is known as water vapor feedback. This feedback is now considered to be on quite secure ground, but the study of the behavior of water vapor in the atmosphere offers many challenges, and is a problem of considerable subtlety. In subsequent chapters, we'll provide the underpinnings needed for a study of this host of questions. Clouds present an entirely greater order of difficulty, as they warm the planet through their effect on

Figure 1.17: Atmospheric temperature time series derived from analysis of microwave satellite data. Left panel: Mean temperature anomaly for the layer of the atmosphere below about 5km, compared with the GISS instrumental record. Temperature is given as an anomaly relative to the same base value as used in the GISS instrumental record. Right panel: Temperature anomaly for the lower stratosphere (layer from about 15 km to 25 km). In both figures, the El Chichon and Pinatubo volcanic eruptions are marked. In the left panel, major El Nino events are indicated by upward open-shafted arrows.

Figure 1.17: Atmospheric temperature time series derived from analysis of microwave satellite data. Left panel: Mean temperature anomaly for the layer of the atmosphere below about 5km, compared with the GISS instrumental record. Temperature is given as an anomaly relative to the same base value as used in the GISS instrumental record. Right panel: Temperature anomaly for the lower stratosphere (layer from about 15 km to 25 km). In both figures, the El Chichon and Pinatubo volcanic eruptions are marked. In the left panel, major El Nino events are indicated by upward open-shafted arrows.

outgoing infrared radiation, but cool the planet through their reflection of solar radiation. The net effect depends on the complex processes determining cloud height, cloud distribution, cloud particle size and cloud water or ice content. The infrared effects of clouds will be discussed in Chapter 4 and the reflective effects of clouds on sunlight will be discussed in Chapter 5. Uncertainties about the behavior of clouds are the main reason we do not know precisely how much warmer the planet will ultimately get if we double the CO2 concentration. Typical predictions of equilibrium global average warming for a doubling of CO2 range from a low of around 2C to a high of around 6C, with some potential for even greater warming with a low (but presently unquantifiable) probability. Because of other uncertainties in the system (particularly the magnitude of the aerosol effect and especially the indirect aerosol effect on cloud brightness) simulations with a range of different cloud behaviors can all match the historical climate record so far, but nonetheless yield widely different forecasts for the future. There is no analysis at present that excludes the possibility of the higher end of the forecast range, for which the effects would likely be catastrophic. There are other feedbacks in the climate system that complicate the forecast. These include feedbacks from melting snow and ice, and from the dynamics of glaciers on land. They also include changes in vegetation, and changes in the ocean circulation which can affect the delay due to burial of heat in the deep ocean.

Global warming - perhaps more aptly called "global climate disruption" - is an event of geological proportions, but one which is caused by human activities. The natural range of CO2 for the past 800,000 years, and almost certainly for the entire two million years of the Pleistocene, has been 180 to 280 molecules per million. Owing to human activities, the CO2 concentration is already far above the top of the natural range that has prevailed for the entire lifetime of the human species, and without action will become much higher still. The human species and the natural ecosystems we share the Earth with have adapted over the Pliocene and Pleistocene to glacial-interglacial cycles, but a world with doubled CO2 will subject them in the course of two centuries or less to a temperature jump to levels far warmer than the top of the range to which societies and organisms have adapted. Even if climate sensitivity is at the low end of the predicted range and if human societies hold the line at a doubling of CO2 , the resulting 2C warming represents a substantial climate change; it takes a great deal to change the mean temperature of the entire globe, and a 2C global mean increase is a summary statistic that masks much higher regional changes and potentially quite massive effects on sea ice, glaciers and ecosystems. If climate sensitivity turns out to be at the high end, the warming could be 4C or more, and if that is compounded by an increase to four times pre-industrial CO2 the global mean increase could reach 8C. That is twice the degree of warming in the PETM, and though the PETM looks abrupt, it is very likely to have set in on a longer time scale than it would take human industrial society to burn the remaining reserves of fossil fuels. If this is allowed to happen, it will take thousands of years for the climate to recover to a normal state. Could global warming disrupt the natural glacial-interglacial cycle? What would the consequences of that be? Those are indeed Big Questions.

As seen by paleoclimatologists ten million years in the future, whatever species they may be, the present era of catastrophic release of fossil fuel carbon will appear as an enigmatic event which will have a name of its own, much as paleoclimatologists and paleobiologists refer today to the PETM or the K-T boundary event. The fossil carbon release event will show up in 13C proxies of the carbon cycle, in dissolution of ocean carbonates through acidification of the ocean, through mass extinctions arising from rapid warming, and through the moraine record left by retreating mountain glaciers and land-based ice sheets. As an event, it is unlikely to permanently destroy the habitability of our planet, any more than did the K-T event or the PETM. Still, a hundred generations or more of our descendents will be condemned to live in a planetary climate far different from that which nurtured humanity, and in the company of a greatly impoverished biodiversity. Biodiversity does recover over the course of millions of years, but that is a very long time to wait, if indeed there are any of our species left around at the time to do the waiting. Extinction may not be precisely forever, but it is close enough.

1.13 The fate of the Earth,the lifetime of biospheres

Even if a planet enters a habitable phase at some stage in its life, it will not remain habitable forever; various kinds of crises can bring its habitability to an abrupt or gradual end. This brings us to the Big Question of lifetime of biospheres; the answer has implications for how likely it is that complex or intelligent life will have had time to evolve elsewhere in the Universe.

Certainly, the Earth's habitability will end when the Sun leaves the main sequence and expands into a Red Giant. Perhaps some of the outer planets or their satellites will enter a brief habitable phase at that time, but it will not be long lasting. That particular crisis is about four billion years in Earth's future, but other habitability crises are likely to set in long before then. In particular, as the Sun continues to brighten, at some point the brightness will outstrip the ability of the silicate weathering process to compensate by drawing down CO2. At that point the Earth would succumb to a runaway greenhouse, become lethally hot, and eventually lose its water to space. When will that happen? That is a Big Question, and some current estimates put the remaining natual lifetime of Earth's biosphere at as little as a half billion years. Given that it took four billion years of Earth History before intelligent life emerged, that makes our existence look like quite a close call. Even before the runaway stage, silicate weathering will draw down CO2 to the point where most forms of photosynthesis will no longer be able to operate. Can more efficient forms of photosynthesis fill in the gap? That's a Big Question as well, but one of a primarily biological nature that we will not attempt to answer.

As the Sun's luminosity increases, Earth may become uninhabitable, but other planets in the Solar System may become more hospitable; in any event they will go through interesting transformations. Mars will warm up, but given that it has little or no active tectonics to generate a new atmosphere, it is unlikely to become Earthlike unless some artificial means is found to give it a more substantial atmosphere. Could Europa melt and become a waterworld? What will happen to Titan as the Sun gets brighter?

Alternately the end could come by ice rather than fire. Earth's life and climate are ultimately maintained by a brew consisting of solar energy and the CO2 outgassing from the interior. The CO2 has a warming effect of its own, which can be modified by organisms that intercept it and transform it into oxygen, methane, or other compounds. If the tectonic release of CO2 ceases, as it will once the Earth exhausts its interior heat sources, all that will come to an end. There will be nothing to offset silicate weathering, and CO2 will draw down until the Earth turns into a snowball - unless the runaway greenhouse from a brightening Sun gets us first.

This class of questions naturally generalizes to the question of how the time scales that limit the biosphere's lifetime would be different for planetary systems around other stars. We have already mentioned that hotter stars have a shorter life on the main sequence, while cooler stars last longer; the former will have planets with short-lived biospheres compared to the latter. The question of how long the silicate weathering thermostat can cope with changing stellar luminosity, and how long outgassing can sustain the climate, is far subtler, and will have interesting dependences on the planet's size, composition and orbit. There could well be other chemical cycles other than silicate weathering and CO2 outgassing, which could provide climate regulation; the search for such possibilities is still in its infancy. There could also be novel habitability crises, associated with long term evolution of planetary systems with highly eccentric orbits or systems perturbed by binary star companions.

All this climate catastrophe presupposes no intervention by the inhabitants. In fact, there are quite realistic possibilities for technologically adept inhabitants to stave off the catastrophe at least until their star leaves the Main Sequence. A runaway greenhouse could be prevented by simply reducing the effective stellar brightness, through orbital sunshades or injection of reflecting aerosols into the upper atmosphere. Indeed, such geoengineering fixes have been proposed to offset the global warming effect of anthropogenic CO2 increases. They are a rather desperate and alarming prospect as a solution to global warming, since they offset a climate forcing lasting a thousand years or more with a fix requiring more or less annual maintainence if catastrophe is not to strike; far better to keep CO2 from getting dangerously high in the first place. However, if the alternative a half billion years out is a runaway greenhouse, the risk of maintaining sunshades will no doubt seem quite acceptable. The loss of CO2 outgassing as Earth's tectonic cycle ceases also has a relatively easy technical fix. Inhabitants could use a small portion of the energy received from the star to cook CO2 back out of carbonates, in a process nearly identical to that by which cement is manufactured. Given the slow rate of silicate weathering, only modest quantities of carbonate would have to be processed. All this can be done,but it would appear to require long term planning and intelligent intervention. A good understanding of the principles of planetary climate will be needed by any beings contemplating such interventions. This book, we hope, will be a good place to start.

1.14 For Further Reading

Many of the problems and explorations in the Workbook sections of this book require the use of some basic numerical analysis. The necesary algorithms are enumerated and exercised in the Workbook section of this chapter. The essential reference for the derivation and implementation of the algorithms is the Numerical Recipes series. The current edition is:

• Press WH, Teukolsky SA, Vetterling WT and Flannery BP 2007: Numerical Recipes 3rd Edition: The Art of Scientific Computing. Cambridge University Press

The algorithm description is independent of programming language, but the implementations given in this particular edition are based on c++. Earlier editions are available for other programming languages. The c language implementations in the 1992 edition Numerical Recipes in c provide a clean basis for re-implementation in other programming languages, particularly convenient for the reader who wishes to avoid some of the intricacies of c++. The reader will need only a very small part of the material covered the Numerical Recipes opus; the relevant sections are pointed out in the Workbook for this chapter.

Conditions during the earliest period of Earth's history, the time required for the crust to form and cool, and the time evolution of heat flux from the interior of the Earth to the surface are discussed in:

• Sleep NH, Zahnle K and Neuhoff PS 2001: Initiation of clement surface conditions on the earliest Earth, Proc Nat Acad Sci, 98, 3666-3672.

• Turcotte DL 1980: On the thermal evolution of the Earth, Earth Planet. Sci. Lett., 48, 53-58.

The long-term evolution of the brightness of the Sun and similar stars is discussed in:

• Gough, DO 1981: Solar interior structure and luminosity variations, Solar Physics, 74, 21-34.

• Sackmann I-J, Boothroyd AI, and Kraemer KE 1993: Our Sun. III. Present and Future, Astrophysical Journal, 418, 457-468.

For a very engaging introduction to what we know about life on the Early Earth, the reader is directed to the book

• Knoll A 2004: Life on a Young Planet. Princeton University Press

The original paper on the Faint Young Sun problem is

• Sagan C and Mullen G 1972: Earth and Mars: Evolution of atmospheres and surface temperatures , Science, 177, 52-56.

A good review of the history of oxygen on Earth and the proxy methods used to infer this history can be found in

• Canfield DE 2005: The Early History of Atmospheric Oxygen: Homage to Robert M. Garrels Annu Rev Earth Planet Sci, 33, 1-36. doi:10.1146/

For a general introduction to the Snowball Earth problem, see

• Hoffman, PF and Schrag DP 2002: The snowball Earth hypothesis: testing the limits of global change. Terra Nova 14, 129-155.

The discussion of late Cretaceous and Cenozoic paleoclimate drew largely on the following papers:

• Sluijs A et al 2006: Suptropical Arctic Ocean temperatures during the Palaeocene/Eocene thermal maximum. Nature 441 doi:10.1038/nature04668 .

• Moran K et al 2006: The Cenozoic palaeoenvironment of the Arctic Ocean. Nature 441 doi:10.1038/nature04800

• Pearson PN et al 2007: Stable warm tropical climate through the Eocene Epoch. Geology 35 doi: 10.1130/G23175A.

• Forster A et al 2007: Geology 35 doi:10.1130/G23874A.

• Mix, A.C., et al. 1995. Benthic foraminifera stable isotope record form Site 849, 0-5 Ma: Local and global climate changes. Pages 371-412 in N.G. Pisias et al. editors, Proceedings of the Ocean Drilling Program, Scientific Results 138, College Station, Texas, USA.

Evolution of Phanerozoic climate, occurrence of glaciations, and evolution of CO2 content of the atmosphere are discussed in

• Crowley TJ and Berner RA 2001: CO2 and climate change, Science, 292, 870-872. DOI: 10.1126/science.1061664

• Zachos J et al. 2001: Trends, Rhythms and Aberrations in Global Climate 65 Ma to Present, Science, 292 , 686-693. DOI: 10.1126/science.1059412

Veizer's long term fossil 18O tropical temperature record, discussed in Crowley and Berrner (2001), is not generally considered reliable.

The GISP, Vostok and EPICA ice core records are described in

• Grootes PM, and Stuiver M. 1997: Oxygen 18/16 variability in Greenland snow and ice with 103 to 105-year time resolution. J. Geophys. Res. 102 26455-26470.

• Petit, JR et al. 2001: Vostok Ice Core Data for 420,000 Years, IGBP PAGES/World Data Center for Paleoclimatology Data Contribution Series No. 2001-076. NOAA/NGDC Paleo-climatology Program, Boulder CO, USA.

• Petit, JR et al.: 1999, Climate and Atmospheric History of the Past 420,000 years from the Vostok Ice Core, Antarctica, Nature 399 429-436.

• Siegenthaler TF et al. 2005: Stable Carbon Cycle-Climate Relationship During the Late Pleistocene. Science 310 1313-1317.

The intellectual history surrounding anthropogenic climate change ("global warming") is surveyed in the following book and acccompanying web site:

• Weart S 2008: The Discovery of Global Warming. Harvard University press

Additional information can be found in The Warming Papers, a set of critical readings with essays by D. Archer and the author, forthcoming from Wiley/Blackwell circa 2009.

Was this article helpful?

0 0

Post a comment