«DIRECT TESTIMONY OF JAMES E. HANSEN Q. Please state your name and business address. A. My name is James E. Hansen. My business address is 2880 ...»
A. We compare different climate states in the Earth’s history, thus obtaining a measure of how much climate responded to climate forcings in the past. In doing this, we must define climate forcings and climate feedbacks clearly. Alternative choices for forcings and feedbacks are appropriate, depending on the time scale of interest.
8 A famous definition of climate sensitivity is from the ‘Charney’ problem, in which it is assumed that the distributions of ice sheets and vegetation on the Earth’s surface are fixed and the question is asked: how much will global temperature increase if the amount of CO2 in the air is doubled? The Charney (1979) climate sensitivity is most relevant to climate change on the decadal time scale, because ice sheets and forest cover would not be expected to change much in a few decades or less. However, the Charney climate sensitivity must be recognized as a theoretical construct. Because of the large thermal inertia of the ocean, it would require several centuries for the Earth to achieve its equilibrium response to doubled CO2, and during that time changes of ice sheets and vegetation could occur as ‘feedbacks’, i.e., as responses of the climate system that engender further climate change. Feedbacks can either magnify or diminish climate changes, these effects being defined as positive and negative feedbacks, respectively.
Climate feedbacks include changes of atmospheric gases and aerosols (fine particles in the air). Gases that change in response to climate change include water vapor, but also the longlived greenhouse gases, CO2, CH4 and N2O.
Q. Is water vapor not a stronger greenhouse gas than these others?
A. Yes, and that is sometimes a source of confusion. Water vapor readily evaporates into and condenses out of the atmosphere. The amount of H2O in the air is a function of the climate, primarily a function of temperature. The air holds more water vapor in the summer than in winter, for example. Water vapor is a prime example of what we call ‘fast’ feedbacks, those feedbacks that respond promptly to changes of climate. Because H2O causes a strong greenhouse effect, and tropospheric H2O increases with temperature, it provides a positive feedback.
The Charney climate sensitivity includes the effects of fast feedbacks such as changes of water vapor and clouds, but it excludes slow feedbacks such as ice sheets. We obtain an empirical measure of the equilibrium Charney climate sensitivity by comparing conditions on Earth during the last ice age, about 20,000 years ago with the conditions in the present interglacial period prior to major human-made effects. Averaged over a period of say 1000 years, the planet in each of these two states, glacial and interglacial, had to be in energy balance with space within a small fraction of 1 W/m2. Because the amount of incoming sunlight was practically the same in both periods, the 5°C difference in global temperature between the ice age and the interglacial period had to be maintained by changes of atmospheric composition and changes of surface conditions. Both of these are well known.
Figure 5 shows that there was a lesser amount of long-lived greenhouse gases in the air during the last ice age. These gases affect the amount of thermal radiation to space, and they have a small impact on the amount of absorbed solar energy. We can compute the climate forcing due to the glacial-interglacial change of CO2, CH4, and N2O with high accuracy. The effective climate forcing (Hansen et al. 2005a), including the indirect effect of CH4 on other gases, is 3 ± 0.5 W/m2.
Changes on the Earth’s surface also alter the energy balance with space. The greatest change is due to the large ice sheets during the last ice age, whose high albedo (‘whiteness’ or reflectivity) caused the planet to absorb less solar radiation. Smaller effects were caused by the altered vegetation distribution and altered shorelines due to lower sea level during the ice age.
The climate forcing due to all these surface changes is 3.5 ± 1 W/m2 (Hansen et al. 1984).
Thus the glacial-interglacial climate change of 5°C was maintained by a forcing of about
6.5 W/m2, implying a climate sensitivity of about ¾°C per W/m2. This empirical climate 9 sensitivity includes all fast feedbacks that exist in the real world, including changes of water vapor, clouds, aerosols, and sea ice. Doubled CO2 is a forcing of 4 W/m2, so the Charney climate sensitivity is 3 ± 1°C for doubled CO2. Climate models yield a similar value for climate sensitivity, but the empirical result is more precise and it surely includes all real world processes with ‘correct’ physics.
Q. This climate sensitivity was derived from two specific points in time. How general is the conclusion?
A. We can check climate sensitivity for the entire past 425,000 years. Ice cores (Figure 5) provide a detailed record of long-lived greenhouse gases. A measure of surface conditions is provided by sediment cores from the Red Sea (Siddall et al. 2003) and other places, which yield a record of sea level change (Figure 6a). Sea level tells us how large the ice sheets were, because water that was not in the ocean was locked in the ice sheets. Greenhouse gas and sea level records allow us to compute the climate forcings due to both atmospheric and surface changes for the entire 425,000 years (Hansen et al. 2007a).
When the sum of greenhouse gas and surface albedo forcings (Figure 6b) is multiplied by the presumed climate sensitivity of ¾°C per W/m2 the result is in remarkably good agreement with ‘observed’ global temperature change (Figure 6c) implied by Antarctic temperature change.
Therefore this climate sensitivity has general validity for this long period. This is the Charney climate sensitivity, which includes fast feedback processes but specifies changes of greenhouse gases and surface conditions.
It is important to note that these changing boundary conditions (the long-lived greenhouse gases and surface albedo) are themselves feedbacks on long time scales. The cyclical climate changes from glacial to interglacial times are driven by very small forcings, primarily by minor perturbations of the Earth’s orbit about the sun and by the tilt of the Earth’s spin axis relative to the plane of the orbit.
Q. Can you clarify cause and effect for these natural climate changes?
A. Figure 7 is useful for that purpose. It compares temperature change in Antarctica with the greenhouse gas forcing. Temperature and greenhouse gas amounts are obtained from the same ice core, which reduces uncertainty in their sequencing despite substantial uncertainty in absolute dating. There is still error in dating temperature change relative to greenhouse gas change, because of the time needed for ice core bubble closure. However, that error is small enough that we can infer, as shown in Figure 7b, that the temperature change tends to slightly precede (by several hundred years) the greenhouse gas changes. Similarly, although the relative dating of sea level and temperature changes are less accurate, it is clear that warming usually precedes ice melt and sea level rise.
These sequencings are not surprising. They show that greenhouse gas changes and ice sheet area changes act as feedbacks that amplify the very weak forcings due to Earth orbital changes. The climate changes are practically coincident with the induced changes of the feedbacks (Figure 7). The important point is that the mechanisms for the climate changes, the mechanisms substantially affecting the planet’s radiation balance and thus the temperature, are the atmospheric greenhouse gases and the surface albedo. Earth orbital changes induce these mechanisms to change, for example, as the tilt of the spin axis increases both poles are exposed to increased sunlight. Changed insolation affects the melting of ice and, directly and indirectly, the uptake and release of greenhouse gases.
10 Q. What is the implication for the present era and the role of humans in climate?
A. The chief implication is that humans have taken control of global climate. This follows from Figure 8, which extends records of the principal greenhouse gases to the present. CO2, CH4 and N2O (not shown) are far outside their range of the past 800,000 years for which ice core records of atmospheric composition are available.
Q. Yet the global warming also shown in Figure 8 does not seem to be commensurate with the greenhouse gas increases, if we were to use the paleoclimate as a guide. Can you explain that?
A. Yes. Observed warming is in excellent agreement with climate model calculations for observed greenhouse gas changes. Two factors must be recognized.
First, the climate system has not had enough time to fully respond to the human-made climate forcings. The time scale after 1850 is greatly expanded in Figure 8. The paleoclimate portion of the graph shows the near-equilibrium (~1000 year) response to slowly changing forcings. In the modern era, most of the net human-made forcing was added in the past 30 years, so the ocean has not had time to fully respond and the ice sheets are just beginning to respond to the present forcing.
Second, the climate system responds to the net forcing, which is only about half as large as the greenhouse gas forcing. The net forcing is reduced by negative forcings, especially human-made aerosols (fine particles).
Q. But is not the natural system driving the Earth toward colder climates?
A. If there were no humans on the planet, the long term trend would be toward colder climate.
However, the two principal mechanisms for attaining colder climate would be reduced greenhouse gas amounts and increased ice cover. The feeble natural processes that would push these mechanisms in that direction (toward less greenhouse gases and larger ice cover) are totally overwhelmed by human forcings. Greenhouse gas amounts are skyrocketing out of the normal range and ice is melting all over the planet. Humans now control global climate, for better or worse.
Another ice age cannot occur unless humans go extinct, or unless humans decide that they want an ice age. However, ‘achieving’ an ice age would be a huge task. In contrast, prevention of an ice age is a trivial task for humans, requiring only a ‘thimbleful’ of CFCs (chlorofluorocarbons), for example. The problem is rather the opposite, humans have already added enough greenhouse gases to the atmosphere to drive global temperature well above any level in the Holocene.
Q. How much warmer will the Earth become for the present level of greenhouse gases?
A. That depends on how long we wait. The Charney climate sensitivity (3°C global warming for doubled CO2) does not include slow feedbacks, principally disintegration of ice sheets and poleward movement of vegetation as the planet warms. When the long-lived greenhouse gases are changed arbitrarily, as humans are now doing, this change becomes the predominant forcing, and ice sheet and vegetation changes must be included as part of the response in determining long-term climate sensitivity.
It follows from Figure 7 that equilibrium climate sensitivity is 6°C for doubled CO2 (forcing of 4 W/m2) when greenhouse gases are the forcing, not 3°C. (Note: the Antarctic 11 temperature change, shown in Figure 7, is about twice the global mean change.) To achieve this full response we must wait until ice sheets have had time to melt and forests have had time to migrate. This may require hundreds of years, perhaps thousands of years. However, elsewhere (Hansen et al. 2007a) we have discussed evidence that forests are already moving and ice sheet albedos are already responding to global warming, so climate sensitivity is already partially affected by these processes.
Thus the relevant equilibrium climate sensitivity on the century time scale falls somewhere between 3°C and 6°C for doubled CO2. The expected temperature change in the 21st century cannot be obtained by simply multiplying the forcing by the sensitivity, as we could in the paleoclimate case, because a century is not long enough to achieve the equilibrium response.
Instead we must make computations with a model that includes the ocean thermal inertia, as is done in climate model simulations (IPCC 2007; Hansen et al. 2007b). However, these models do not include realistically all of the slow feedbacks, such as ice sheet and forest dynamics.
Q. The huge climate changes over the past few hundred thousand years show the dramatic effects accompanying global temperature change of only a few degrees. And you infer climate sensitivity from the documented climate variations. Yet the climate changes and mechanisms are intricate, and it is difficult for the lay person to grasp the details of these analyses. Is there other evidence supporting the conclusion that burning of the fossil fuels will have dramatic effects upon life on Earth?
A. Yes. Climate fluctuations in the Pleistocene (past 1.8 million years) are intricate, as small forcings are amplified by feedbacks, including ‘carbon cycle’ feedbacks. Atmospheric CO2 varies a lot because carbon is exchanged among its surface reservoirs: the atmosphere, ocean, soil, and biosphere. For example, the solubility of CO2 in the ocean decreases as the ocean warms, a positive feedback causing much of the atmospheric CO2 increase with global warming.
That feedback is simple, but the full story of how weak forcings create large climate change is indeed complex.
A useful complement to Pleistocene climate fluctuations is provided by longer time scales with larger CO2 changes than those caused by orbital oscillations. Larger CO2 changes occur on long time scales because of transfer of carbon between the solid earth and the surface reservoirs. The large CO2 changes on these long time scales allow the Earth orbital climate oscillations to be viewed as ‘noise’. Thus long time scales help provide a broader overview of the effect of changing atmospheric composition on climate.
A difficulty with long time scales is that knowledge of atmospheric composition changes is not as good. Samples of ancient air preserved in ice cores exist for only about one million years. But there are indirect ways of measuring ancient CO2 levels to better than a factor of two beyond one million years ago. Atmospheric composition and other climate forcings are known well enough for the combination of Pleistocene climate variations and longer-term climate change to provide an informative overview of climate sensitivity and a powerful way to assess the role of humans in altering global climate.
Q. What determines the amount of CO2 in the air on long time scales?