Storms of My Grandchildren (35 page)

Life seemed to be influenced profoundly by the final snowball event, which occurred about 600 million years ago. Prior to that time, the most complex organisms on Earth were unicellular protozoa and filamentous algae—in other words, the only life on the planet was green scum. The final snowball was followed promptly by the Cambrian explosion of life. Eukaryotes, cells with a membrane-bound nucleus, expanded rapidly into eleven different body plans. These eleven animal phyla still encompass all animals that have ever inhabited Earth.

At the end of the last snowball Earth, the sun’s brightness was within 6 percent of its present value. There will never be another snowball Earth, because the sun continues to get brighter. In fact, with humans on the planet, there will never be another ice age. Sorry to distract you with an aside, but I need to clarify a point, one that is relevant to the present discussion.

A few geologists continue to speak as if they expect Earth to proceed into the next glacial cycle, just as it would have if humans were not around. That glacial period would begin with an ice sheet developing and growing in northern Canada. But why would we allow such an ice sheet to grow, and flow, and eventually crush major cities, when we could prevent it with the greenhouse gases from a single chlorofluorocarbon factory? Humans are now “in charge” of future climate. It is a trivial task to avoid the negative net climate forcing that would push the planet into an ice age (moving conditions toward the left in figure 30). But it is not an easy task to find a way to stop the growth of atmospheric greenhouse gases, most notably carbon dioxide (which moves conditions toward the right in figure 30), as we have been discussing.

How will the sun’s continuing evolution alter Earth’s climate on long time scales? Our sun is a very ordinary medium-size star. It is about 4.6 billion years old, still “burning” hydrogen, producing helium by nuclear fusion in the sun’s core, releasing energy in the process. It is slowly getting brighter. As the hydrogen fuel is exhausted, leaving inert helium in the core, the sun will expand enormously to its Red Giant phase as it burns hydrogen in its outer shell. The expanding sun will toast and eventually swallow Earth about 5 billion years from now. Nothing for you or your grandchildren to worry about. By that time, if humanity still exists, the people 200 million generations from now may have the technology to escape to another solar system.

Humanity will need to figure out climate control long before our sun approaches the Red Giant phase. In one billion years the sun will be about 10 percent brighter than it is today. That climate forcing, about 25 watts per square meter, is surely enough to push Earth into the runaway greenhouse effect, evaporate the oceans, and exterminate all life on the planet. But you and your grandchildren do not need to worry about the long-term change of the sun’s brightness either, because that trend is negligible in comparison with what humans are doing by adding greenhouse gases to the air. Besides, long before the sun becomes 10 percent brighter, humans will have realized that they need to shade the sun a bit, if they want life of the sort that we know to continue. The required “geo-engineering” would be of a simple, direct kind, reflecting a fraction of incident sunlight back to space, a task that will surely be easy for a civilization that exists in future millennia, assuming that it exists.

This geo-engineering comment requires one more digression, to answer the inevitable question: Why not use such a geo-engineering trick to solve our present global warming problem, thus avoiding the need to draw down carbon dioxide to less than 350 parts per million? There are several reasons. First, carbon dioxide must be less than 350 ppm to avoid ocean acidification problems. Second, sun shielding at present is far more expensive and difficult to implement than rational alternatives such as energy efficiency, renewable energy, and nuclear power. Third, it is generally a bad idea to try to cover up one pollution effect by introducing another; such an approach is likely to have many unintended effects. It is hard to match nature. Better to keep atmospheric composition and solar irradiance at the levels to which humanity and nature are adapted. The purpose of sun shielding in the very distant future would be to keep solar irradiance at the level to which life is adapted.

Allow me to elaborate just a bit on the second of these reasons, why implementation of such geo-engineering does not make sense now. Geo-engineering costs money. In contrast, some of the more attractive alternatives would more than pay for themselves. Pay-for-itself is true, for example, for energy efficiency and nuclear power, at least in the mode that nuclear power would be used in places such as India and China, countries that would be expected to choose modular designs and limit the ability of antinukes and bureaucratic lethargy to delay construction and drive up costs. The earliest third- and fourth-generation nuclear power plants will be expensive relative to coal without carbon capture, but nations that choose to limit construction delays should be able to produce nearly carbon-free nuclear energy that is cost-effective. Some renewable energies are expensive relative to fossil fuels, but there are instances where renewable energy is already cost-effective, and these instances should increase with future economies of scale. Although first priority should be given to energy efficiency, renewable energies, and nuclear power, it does make sense to carry out geo-engineering research to define options in the event that continued business-as-usual energy policies create a planetary emergency that demands rapid changes.

Now we are ready for the important part—trying to figure out how close we are to the climate forcing that will cause a runaway greenhouse effect. Until recently I did not worry much about that. Why? Because I knew that at some times in the past there was much more carbon dioxide in the air than today, probably a few thousand parts per million. Even burning all of the fossil fuels will not exceed that amount, so we should be safe, right?

Wrong, unfortunately. It turns out that there are three factors or circumstances that alter the picture, and each of them works in the bad direction.

Circumstance 1 is not the biggest factor, but I start with it because it is substantial and we understand it accurately. Circumstance 1 is the irradiance of the sun. At earlier times, when atmospheric carbon dioxide was more abundant, the sun was dimmer. For example, 250 million years ago the sun was about 2 percent dimmer than it is now. A 2 percent change of solar irradiance is equivalent to doubling the amount of carbon dioxide in the atmosphere. So if the estimated amount of carbon dioxide 250 million years ago was 2,000 ppm, it would take only about 1,000 ppm of carbon dioxide today to create a climate equally as warm, assuming other factors are equal. (As explained earlier, a 2 percent change of solar irradiance and a doubling of carbon dioxide are equivalent forcings, each being about 4 watts per square meter.) In other words, the fact that some scientists have estimated that CO2 was much larger earlier in Earth’s history, perhaps even by a few thousand parts per million, does not mean that we could tolerate that much carbon dioxide now without hitting runaway conditions, because the sun is brighter now.

Circumstance 2 is the “measurement,” or estimation, of past carbon dioxide amounts. Actually, we have direct measurements of past carbon dioxide only for recent glacial to interglacial climate changes, the period with ice core data—essentially the past million years. Maximum carbon dioxide amount in that period was about 300 ppm, until humans started burning fossil fuels. Estimates for more ancient times are based on indirect (“proxy”) measures, but such indirect inferences have great uncertainties. Results from many different methods are compared in our 2008 “Target Atmospheric CO2” paper’s supplementary material. Some methods yield carbon dioxide amounts in the early Cenozoic era, between 65 and 50 million years ago, as great as 2,000 ppm, but other methods suggest that the maximum amount was less than 1,000 ppm.

Certain methods of estimating ancient carbon dioxide levels explicitly depend on assumptions about how much carbon dioxide would have been needed to cause the recorded climate change. In other words, these methods depend on assumed climate sensitivity. For example, with standard assumptions about climate sensitivity, it was estimated that unfreezing a hard “iceball” Earth—an Earth with oceans frozen solid to depths of one kilometer or more—would require a huge amount of carbon dioxide. However, we now recognize that a hard iceball is an unrealistic picture of what actually would have been snowball Earth conditions. Also, the transient phase of unfreezing snowball Earth is not directly relevant to figuring out the climate forcing needed for the runaway greenhouse effect. In other words, although the carbon dioxide amount in the air may have been large just before and during the planet’s thawing, atmospheric carbon dioxide would decrease dramatically during the thawing process, long before the planet could reach runaway greenhouse conditions.

To understand the significance of the new PETM analysis and its relevance to the runaway greenhouse effect, we need to go back to figure 13 (page 106), which shows the deep ocean temperature over the past 65 million years. In the “Target Atmospheric CO2” paper, we used that temperature curve to estimate the carbon dioxide amount over the 65-million-year period, given a very simple assumption about climate sensitivity. Specifically, we assumed that the “fast feedback” climate sensitivity was 3 degrees Celsius (for doubled carbon dioxide) for the entire 65 million years. Total climate sensitivity was greater during the most recent 34 million years because of slow feedbacks, specifically changes of ice sheet area. In justifying the assumption of a 3-degree sensitivity for the ice-free planet, we pointed to the fact that today’s climate seems to be in the middle of a rather flat portion of the (fast feedback) climate sensitivity curve shown in figure 30. We also felt that it was best to employ the simplest assumption until better information became available. Under the assumption of a 3-degree Celsius climate sensitivity, we inferred that the maximum carbon dioxide amount was probably in the range of 1,000 to 1,400 ppm.

Figure 30 suggests that warmer climates may have larger climate sensitivity, indeed, that today’s climate is not terribly far from the runaway situation. However, it is a model result—other models may differ. Empirical results are more meaningful. The Zeebe-Zachos-Dickens paper provides such an empirical result. In it, they show, based on the depths to which the ocean acidified and dissolved carbonate sediments, that the carbon increase that caused the PETM warming was at most 3,000 gigatons of carbon. They infer that atmospheric carbon dioxide would have increased about 700 ppm, from a baseline of approximately 1,000 ppm to about 1,700 ppm. Such a carbon dioxide increase, less than a doubling, would increase global temperature about 2 degrees Celsius, if doubled carbon dioxide sensitivity is 3 degrees Celsius. Zeebe, Zachos, and Dickens conclude, “Our results imply a fundamental gap in our understanding of the amplitude of global warming associated with large and abrupt climate perturbations.”

Their conclusion is the appropriate one from a scientific perspective. I believe we can take it one step further, suggesting that their analysis is evidence that climate sensitivity in the warmer early Cenozoic was greater than 3 degrees Celsius for doubled carbon dioxide. It also favors a smaller value for the carbon dioxide abundance in the early Cenozoic. Both inferences (that the carbon dioxide level in the early Cenozoic may have been less than generally assumed, and that climate sensitivity in the early Cenozoic was greater than today) are reason for increased concern about the long-term effects of burning all fossil fuels. The PETM results would be easier to understand if the baseline carbon dioxide, prior to the PETM warming, was closer to 500 ppm. But even so, the magnitude of the PETM warming implies a climate sensitivity greater than 3 degrees Celsius for doubled carbon dioxide.

My conclusion regarding Circumstance 2 is that recent data suggest that past carbon dioxide amounts were not as great as once believed. These empirical paleoclimate data also suggest that climate sensitivity was greater when the planet was warmer, consistent with the world having been closer to the runaway greenhouse conditions when the carbon dioxide amount was greater.

Circumstance 3 concerns the time scale of climate forcings and response. Carbon dioxide that caused climate change during Earth’s history was introduced much more slowly than the human-made perturbation. Slower introduction allows negative (diminishing) feedbacks in the carbon cycle to come into play. Even the solid Earth reservoir takes up carbon on millennial time scales. The negative feedbacks are the reason that, after the rapid injection of methane (which is quickly oxidized to form carbon dioxide) during the PETM, the carbon dioxide amount and global temperature recovered on fairly rapid geologic time scales (figure 18 on page 153).

The human injection of fossil fuel carbon into the atmosphere, if we choose to burn all fossil fuels, will occur so fast, on the time scale of a century or two, that carbon cycle diminishing feedbacks will not have time to come into play. If we burn all fossil fuels, the forcing will be at least comparable to that of the PETM, but it will have been introduced at least ten times faster. The time required for the ocean to respond to this forcing is only centuries. Thus, carbon cycle diminishing feedbacks will not significantly reduce the ocean warming. The warming ocean can be expected to affect methane hydrate stability at a rate that could exceed that in the PETM, where the rate of change was driven by the speed of the methane hydrate climate feedback, not by the nearly instantaneous introduction of all fossil fuel carbon.