«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 ...»
The 41 Ky climate variability is apparent in Figure 10b and is present in almost all climate records. However, glacial-interglacial climate variations became more complex in the most recent 1.2 My, with large variations at ~100 Ky periodicity, as well as ~41 Ky and ~23 Ky periods. As the planet became steadily colder over the past several million years, the amplitude of glacial-interglacial climate swings increased (Figure 10b) as ice sheet area increased. Ice sheets on Northern Hemisphere continents, especially North America, extended as far south as 45N latitude. Similar ice sheets were not possible in the Southern Hemisphere, which lacked land at relevant latitudes.
Hemispheric asymmetry in ice sheet area allows two additional Earth orbital parameters, which work in concert, to come into play. Gravitational tugs of the planets cause the eccentricity of the Earth’s orbit about the sun to vary from near zero (circular) to an eccentricity of about
0.06. When the orbit is significantly non-circular, this allows another orbital parameter, axial precession, to become important. Precession, which determines the date in the year at which the Earth in its elliptical orbit is closest to the sun, varies with a periodicity of ca. 23 Ky. When the 16 Earth is closest to the sun in Northern Hemisphere winter, thus furthest from the sun in summer, ice sheet growth in the Northern Hemisphere is encouraged by increased winter snowfall and cool summers. The effect of eccentricity + precession on ice sheet growth is opposite in the two hemispheres, so the effect is important only when the area of high albedo ice and snow is much different in the two hemispheres, as it has been in the past million years. Climate variations then include all three periodicities, ~23 Ky precession, ~41 Ky tilt, and ~100 Ky eccentricity, as has been demonstrated for the recent ice age cycles (Hays et al 1976).
Q. What are the current Earth orbital parameters?
A. Precession has the Earth closest to the sun in January, furthest in July, which would favor growth of Northern Hemisphere ice. But eccentricity is small, about 0.016, so the precession effect is not large. Tilt is about midway between its extremes headed toward smaller tilt, the next minimum tilt occurring in ~10 Ky. Smaller tilt favors ice sheet growth, so, if it were not for humans, we might expect a trend toward the next ice age. But the trend may have been weak, because, by the time tilt reaches its minimum, the sun will be closest to the sun in Northern Hemisphere summer. Thus in this particular cycle the two mechanisms, tilt and eccentricity + precession, will be working against each other, rather than reinforcing each other. In any event, this natural tendency has become practically irrelevant in the age of fossil-fuel-burning humans.
Q. Why is the natural glacial-interglacial cycle irrelevant?
A. Earth orbital changes were only pacemakers for glacial-interglacial climate change, inducing changes of ice area and greenhouse gases. Changes of surface albedo and greenhouse gases were the mechanisms for climate change, providing the immediate causes of the climate changes. We showed in Figure 6 that these two mechanisms account for the glacial-interglacial climate variations.
Now humans are responsible for changes of these climate mechanisms. Greenhouse gases are increasing far outside the range of natural glacial-interglacial variations (Figure 8) and ice is melting all over the planet. The weak effect of slow orbital changes is overwhelmed by the far larger and faster human-made changes.
Humans are now entirely responsible for long-term climate change (Figure 13).
However, it would be misleading to say that humans are “in control”. Indeed, there is great danger that humans could set in motion future changes that are impossible to control, because of climate system inertia, positive feedback, and tipping points.
Q. Can we finally finish with this paleoclimate discussion?
A. Please allow one final comment. For the record, since I could only estimate broad ranges for CO2 in the Cenozoic era, I should show at least one estimate from the proxy CO2 data. Figure 14A shows estimated CO2 for the entire Phanerozoic eon, the past 540 million years. I show this longer time interval, because it includes CO2 changes so large as to make the errors in the proxies less in a relative sense.
Geologic evidence for ice ages and cool periods on this long time frame (Figure 14B) shows a strong correlation of climate with CO2. Climate variations were huge, ranging from ice ages with ice sheets as far equatorward as 30 degrees latitude to a much warmer planet without ice. Although other factors were also involved in these climate changes, greenhouse gases were a major factor.
17 Q. Are climate models consistent with paleoclimate estimates of high climate sensitivity and with observed global warming in the past century?
A. Yes. Climate models yield equilibrium sensitivity (the response after several centuries) of typically about 3°C for doubled CO2. Figure 15B shows the resulting global warming when such a climate model ( one with ~3°C sensitivity for doubled CO2) is driven by climate forcings measured or estimated for the period 1880-2003 (Figure 15A). The calculated and observed warmings are similar. Good agreement might also be obtained using a model with higher sensitivity and a smaller forcing or using a model with lower sensitivity and a larger forcing. But the sensitivity of this model (Hansen et al. 2007b) agrees well with the empirical sensitivity defined by paleoclimate data.
Q. I am confused. Did you not say earlier that climate sensitivity is about 6°C for doubled CO2?
A. Yes. That is an important point that needs to be recognized. We showed that the real world climate sensitivity is 6°C for doubled CO2, when both fast and slow feedback processes are included, based on data that covered climate states ranging from interglacial periods 1°C warmer than today to ice ages 5°C cooler than today. That 6°C sensitivity is also the appropriate estimate for the range of warmer climates up to the point at which all ice sheets are melted and high latitudes are fully vegetated.
This higher climate sensitivity, 6°C for doubled CO2, is the appropriate sensitivity for long time scales, when greenhouse gases are the specified forcing mechanism and all other slow feedbacks are allowed to fully respond to the climate change. The substantial relevant slow feedbacks are changes of ice sheets and surface vegetation.
Q. Yet you employed a climate model with 3°C sensitivity, a model excluding these slow feedbacks. Does this cause a significant error?
No, not in simulations of the 20th century climate change as in Figure 15. Feedbacks come into A.
play not in response to climate forcing but in response to climate change. Ocean thermal inertia introduces a lag, shown by the climate response function in Figure 15c. The response function is the fraction of the equilibrium surface response that is achieved at a given time subsequent to introduction of the forcing. About half of the equilibrium response occurs within a quarter century, but further response at the Earth’s surface is slowed by mixing of water between the ocean surface layer and the deeper ocean. Nearly full response requires several centuries.
Furthermore, the response time to a climate forcing increases in proportion to the square of climate sensitivity (Hansen et al. 1985), so the response time for 6°C climate sensitivity is about four times greater than that shown in Figure 15c. The explanation for this strong dependence of response time on climate sensitivity is simple: the rate of heating is fixed, so to warm the ocean mixed layer would take twice as long for 6°C sensitivity as for 3°C sensitivity.
But this additional time allows more mixing of heat into the deeper ocean. For diffusive mixing it follows analytically, as shown in the referenced paper, that the response time goes as the square of climate sensitivity.
In addition, some climate feedback processes can increase response time above that associated with ocean thermal inertia alone. A fast feedback such as atmospheric water vapor amount occurs almost instantly with temperature change. However, ice sheets require time to disintegrate or grow, and vegetation migration in response to shifting climate zones also may require substantial time.
18 Ice sheet and vegetation responses were not important factors affecting the magnitude of 20 century global warming, so simulations of 20th century global temperature change were not th compromised by exclusion of those feedbacks. However, with a substantial and almost monotonic global warming now in place (Figure 1A), the ice sheet and vegetation feedbacks should begin to contribute significantly to climate change in the 21st century. Ice sheet and vegetation changes will continue to alter the planetary energy balance over century time scales and must be accounted for in projecting future climate change.
Q. Can we move on from this technical discussion of feedbacks and response time?
A. Please allow one final comment. The 6°C sensitivity (for doubled CO2) is valid for a specified change of greenhouse gases as the climate forcing. That is relevant for human-made change of atmospheric composition, and this sensitivity yields the correct answer for long-term climate change if actual greenhouse gas changes are used as the forcing mechanism. However, climate model scenarios for the future usually incorporate human-made emissions of greenhouse gases.
Atmospheric greenhouse gas amounts may be affected by feedbacks, which thus alter expected climate change.
Greenhouse gas feedbacks are not idle speculation. Paleoclimate records reveal times in the Earth’s history when global warming resulted in release of large amounts of methane to the atmosphere. Potential sources of methane include methane hydrates ‘frozen’ in ocean sediments and tundra, which release methane in thawing. Recent Arctic warming is causing release of methane from permafrost (Christensen et al. 2004; Walter et al. 2006), but not to a degree that has prevented near stabilization of atmospheric methane amount over the past several years.
Hansen and Sato (2004) have shown from paleoclimate records that the positive feedbacks that occur for all major long-lived greenhouse gases (carbon dioxide, methane, and nitrous oxide) are moderate for global warming less than 1°C. However, no such constraints exist for still larger global warming, because there are no recent interglacial periods with global warming greater than about 1°C. Based on other metrics (avoiding large sea level rise, extermination of species, and large regional climate disruption) we argue that we must aim to keep additional global warming, above the level in 2000, less than 1°C. Such a limit should also avert massive release of frozen methane.
Q. Observed (and modeled) global warming of 0.8°C in the past century seems small in view of the large changes of greenhouse gases shown in Figure 8. Why is that?
A. There are two reasons.
First, there is the large thermal inertia of the ocean. It takes a few decades to achieve just half of the global warming with climate sensitivity of 3°C for doubled CO2, as shown in Figure 15C. And the slow feedbacks that contribute half of the paleoclimate change are now just beginning to come into play.
Second, the greenhouse gases are not the only climate forcing. Human-made tropospheric aerosols, Figure 15A, are estimated to cause a negative forcng about half as large as the greenhouse forcing, but opposite in sign.
Q. There must be some uncertainty in the climate forcings, especially the aerosol forcing. Can you verify that the estimated forcings are realistic?
A. Yes. The aerosol forcing is difficult to verify directly, but there is an exceedingly valuable diagnostic that relates to the net climate forcing. Given that the greenhouse gas forcing is known 19 accurately, the constraint on net forcing has implications for the aerosol forcing, because other forcings are either small or well-measured (Figure 15A). The diagnostic that I refer to is the planetary energy imbalance (Hansen et al. 2005b).
The Earth’s energy imbalance, averaged over several years, is a critical metric for several reasons. First and foremost, it is a direct measure of the reduction of climate forcings required to stabilize climate. The planetary energy imbalance measures the climate forcing that has not yet been responded to, i.e., multiplication of the energy imbalance by climate sensitivity defines global warming still “in the pipeline”.
A good period to evaluate the Earth’s energy imbalance is the eleven-year period 1995because this covers one solar cycle from solar minimum to solar minimum. A climate model with sensitivity ~3°C for doubled CO2, driven by the climate forcings in Figure 15A, yields an imbalance of 0.75 ± 0.15 W/m2 for 1995-2005. Observations of heat gain in measured portions of the upper 700 m of the ocean yield a global heat gain of ~0.5 W/m2. Measured or estimated heat used in sea ice and land ice melt, warming of ground and air, and ocean warming in polar regions and at depths below 700 m yield a total estimated heat gain of 0.75 ± 0.25 W/m2 (Hansen 2007b).
The observed planetary energy imbalance thus supports the estimated climate forcings used in the climate simulations of Figure 15. This check is not an absolute verification, because the results also depend upon climate sensitivity, but the model’s sensitivity is consistent with paleoclimate data. Indeed, the existence of a substantial planetary energy imbalance provides confirmation that climate sensitivity is high. Climate response time varies as the square of climate sensitivity, so if climate sensitivity were much smaller, say half as large as indicated by paleoclimate data, it would not be possible for realistic climate forcings to yield such a large planetary energy imbalance.
Comment: The planetary energy imbalance is the single most critical metric for the state of the Earth’s climate. Ocean heat storage is the largest term in this imbalance; it needs to be measured more accurately, present problems being incomplete coverage of data in depth and latitude, and poor inter-calibration among different instruments. The other essential measurement for tracking the energy imbalance is continued precise monitoring of the ice sheets via gravity satellite measurements.
Q. How much is global warming expected to increase in the present century, and how does this depend upon assumptions about fossil fuel use?