## Is the science settled?

#### Posted on 24 March 2010 by John Cook

A common skeptic refrain is that "the science isn't settled", meaning there are still uncertainties in climate science and therefore action to cut CO2 emissions is premature. This line of argument betrays a fundamental misunderstanding of the nature of science. Firstly, it presumes science exists in a binary state - that science isn't settled until it crosses some imaginary line after which it's finally settled. On the contrary, science by its very nature is never 100% settled. Secondly, it presumes that poor understanding in one area invalidates good understanding in other areas. This is not the case. To properly answer the question, "is the science settled?", an understanding of how science works is first required.

Science is not about absolute proofs. It never reaches 100% certainty. This is the domain of mathematics and logic. Science is about improving our understanding by narrowing uncertainty. Different areas of science are understood with varying degrees of confidence. For example, while some areas of climate science are understood with high confidence, there are some areas understood with lower confidence, such as the effect on climate from atmospheric aerosols (liquid or solid particles suspended in the air). Aerosols cool climate by blocking sunlight. But they also serve as nuclei for condensation which leads to cloud formation. The question of the net effect of aerosols is one of the greater sources of uncertainty in climate science.

What do we know with high confidence? We have a high degree of confidence that humans are raising carbon dioxide levels in the atmosphere. The amount of CO2 emissions can be accurately calculated using international energy statistics* *(CDIAC). This is double checked using measurements of carbon isotopes in the atmosphere (Ghosh 2003). We can also triple check these results using observations of falling oxygen levels due to the burning of fossil fuels (Manning 2006). Multiple lines of empirical evidence increase our confidence that humans are responsible for rising CO2 levels.

We also have a high degree of confidence in the amount of heat trapped by increased carbon dioxide and other greenhouse gases. This is otherwise known as radiative forcing, a disturbance in the planet's energy balance. We can calculate with relatively high accuracy how much heat is trapped by greenhouse gases using line-by-line models which determine infrared radiation absorption at each wavelength of the infrared spectrum. The model results can then be directly compared to satellite observations which measure the change in infrared radiation escaping to space. What we find in Figure 1 is the observed increased greenhouse effect (black line) is consistent with theoretical expectations (red line) (Chen 2007). These results can also be double checked by surface measurements that observe more infrared radiation returning to Earth at greenhouse gas wavelengths (Evans 2006). Again, independent observations raise our confidence in the increased greenhouse effect.

*Figure 1: Increased greenhouse effect from 1970 to 2006. Black line is satellite observations. Red line is modelled results (**Chen 2007**).*

So we have a lower understanding of aerosol forcing and a higher understanding of greenhouse gas forcing. This contrast is reflected in Figure 2 which displays the probability of the radiative forcing from greenhouse gases (dashed red line) and aerosol forcing (dashed blue line). Greenhouse gas forcing has a much higher probability constrained to a narrow uncertainty range. Conversely, the aerosol forcing has a lower probability and is spread over a broader uncertainty range.

*Figure 2: Probability distribution functions (PDFs) from man-made forcings. Greenhouse gases are the dashed red curve. Aerosol forcings (direct and indirect cloud albedo) are the blue dashed curve. The total man-made forcing is the solid red curve **(IPCC AR4 Figure 2.20b)*

The important point to make here is that a lower understanding of aerosols doesn't invalidate our higher understanding of the warming effect of increased greenhouse gases. Poorly understood aspects of climate change do not change the fact that a great deal of climate science is well understood. To argue that the 5% that is poorly understood disproves the 95% that is well understood betrays an incorrect understanding of the nature of science.

Nedat 00:30 AM on 28 March, 2010WV feedback(whatever its magnitude or sign for all but the very recent *atmospheric* temperature changes has *already happened*. There may be some other temperature changes coming along "in the pipeline", but WV adjusts to a temperature change in a period of months or so.That's right; if we eliminated all other forcings then the water vapor feedback would settle at its new equilibrium level very rapidly. However, it seems likely that we'll keep emitting CO2 for some time now. The water vapor feedback will continue to increase as long as CO2 continues to increase, and it will persist as long as CO2-induced warming persists (i.e., thousands of years).Tom Daytonat 02:33 AM on 28 March, 2010chrisat 03:16 AM on 28 March, 2010at equilibrium(for a [CO2] increase from 286-386 ppm), and for a 3 oC sensitivity, ~1.25 oC of surface warming. Since we’vealreadyhad 0.85 oC of warming without taking account the aerosol effect and the climate response time), it’s very unlikely that the climate sensitivity can be lower than 2 oC of warming per doubling of atmospheric CO2. A similar conclusion was recently obtained from an obviously (!) much more detailed analysis of the Earth’s energy balance since 1950 (Murphy et al. (2010) -------------------------------------------- [***] delta T = (ln([CO2]final/[CO2]start))*s/ln(2) where deltaT is the surface temperature change expected from a change in [CO2] from [CO2]start to [CO2]final in ppm, and s is the climate sensitivity in oC.chrisat 04:16 AM on 28 March, 2010R. Knutti and G. C. Hegerl (2008) The equilibrium sensitivity of the Earth's temperature to radiation changesNature Geoscience 1, 735-743Knutti and Hegerl (2008) and:Murphy DM et al. (2009) An observationally based energy balance for the Earth since 1950J. Geophys. Res.114 art. #D17107Murphy et al. (2009)chrisat 05:13 AM on 28 March, 2010There is no high confidence that the warming in the last 30 years is exceptional, nor that anything in the pattern of warming in the last 200 years is closely related to the concentration of CO2.Since paleoanalysis indicates that the warming of the last 30 years is exceptional at least in the context of the last millennium and likely last two millennia, I don’t think your statement is supported by the science (see e.g. overlay of paleotemp reconstructions, and more recent paleoanalysis). Likewise, the pattern of warming is entirely consistent with the expected effects of enhanced [CO2], both in degree (see chris) and its nature (polar amplification; tropospheric warming associated with stratospheric cooling etc.).The relationship of the global temperature with patterns of ocean circulation is much better than the relationship with the CO2 concentration.That’s certainly incorrect. Analysis of ocean current contributions to 20th century warming indicates that these have made close to zero net contribution to warming over the last 100 years, and in fact reinforce the dominant role of enhanced greenhouse gas contributions ( Swanson et al (2009)[*]There is a lot of evidence, that the climate is much more complicated than could follow from the high understanding of heat trapping by CO2 alone.I don’t think anyone would say otherwise. That doesn’t negate the fact that we have rather high scientific certainty that raised greenhouse gas levels have dominated the Earth’s surface temperature rise during the last 100-150 years. [*]K.L. Swanson et al. (2009) Long term natural variability and 20th century climate changeProc. Natl. Acad. Sci. USA 106, 16120-16130shawnhetat 06:42 AM on 28 March, 2010Riccardoat 07:15 AM on 28 March, 2010chrisat 07:21 AM on 28 March, 2010My point was directed to folks who might think that WV feedback on the forcing we've already added was yet to happen.That's not quite right. As you indicate we've had the water vapour feedback on thewarmingthat's occurred so far. However since we haven't yet had all the warming that will eventually result from theforcingwe've already added, there is still some water vapour feedback to accrue as a reult of the current forcing.shawnhetat 17:33 PM on 28 March, 2010Riccardoat 19:45 PM on 28 March, 2010shawnhetat 03:00 AM on 29 March, 2010Philippe Chantreauat 14:04 PM on 30 March, 2010shawnhetat 08:17 AM on 31 March, 2010Philippe Chantreauat 09:21 AM on 31 March, 2010Berényi Péterat 13:30 PM on 31 March, 2010sis a monotonic functionof average IR optical depthHyof atmosphere:s =whereH(y)for allH(y_{1}) >H(y_{2})y. 2 For a given optical depth_{1}> y_{2}ythere is an equilibrium temperature_{0}sso that_{0}s. 3 This equilibrium is stable against small transient perturbations. 3.1 If_{0}=H(y_{0})is considered to be a functional acting on optical depth historiesHysuch that_{0}+ y(t)y(t)is bounded (y) and zero outside_{0}>> |y(t)|tfor some_{1}> t > 0t, then_{1}tends toH(y_{0}+ y(t))sin the long run. 4_{0}isHsmootharoundy, that is if the integral of_{0}y(t)squared is sufficiently small, there is some linear transformHsuch that5H(y_{0}+ y(t)) =H(y_{0}) + Hyis time shift invariant. That is ifHh(t) =, thenH(y(t))h(t+tfor all_{1}) =H(y(t+t_{1}))t. In this case the linear transform_{1}Hdefined above is a filter and is fully specified by its impulse response function or the Fourier transform of it, the transfer function. 6 LetHbe a first order lowpass filter. It's easier to visualize its response to a step functiony(t)which isyfor_{1}t > 0and zero otherwise. If this forced increase in optical depth (relative to the equilibrium value ofy) induces a long term increase of_{0}sin SST, the response function defined by_{1}Hish(t) = swhere_{1}/y_{1}(1-e^{-t/t0})tis_{0}relaxation time. Now. Average water contents of the atmosphere is somewhere around 4000 ppmv, highly variable. It is more than ten times the current CO_{2}level. Also, H_{2}O has much more absorption lines in thermal infrared, so even tiny changes in humidity imply changes in overall IR optical depth. Also, as the story goes, vapor pressure of H_{2}O over open water surfaces increases with temperature, so overall optical depth is also expected to increase. As average annual precipitation on Earth is close to 1000 mm and atmospheric moisture is low (only 0.24% by weight), turnover time has to be short (approx. 9 days). Therefore atmospheric IR optical depth change should be an almost instantaneous response to a change in SST. We have already postulated a rise ofsin SST in response to an increase in optical depth of_{1}y. Now it is done the other way around. If SST is increased by_{1}s, it causes an immediate increase of optical depth by_{1}f*y(with some coefficient_{1}f). This is the water vapor feedback. From now on attention is restricted to the supposed linear regime around the equilibrium state defined above, so only anomalies are dealt with. Letxbe the IR optical depth anomaly due to GHGs other than H_{2}O. We have two equations:s = Hy(1)y = f*y(2) From these we have_{1}/s_{1}*s + xs = (H(3) Let's switch to the frequency domain. The Fourier transform of^{-1}-f*y_{1}/s_{1})^{-1}x = GxHisandHwis angular frequency. In this case(4) If it is put back to (3)Hw = s_{1}/y_{1}/(1+j*t_{0}*w)(5) whereGw = 1/(1-f)*s_{1}/y_{1}/(1+j*t_{1}*w)tis obtained. From this (by inverse Fourier transform) the response to a step function of magnitude_{1}= t_{0}/(1-f)xin GHG induced increase of IR optical depth is_{1}g(t) = 1/(1-f)*s(6) Indeed, an amplification factor of_{1}/x_{1}*(1-e^{-t/t1})1/(1-f)is seen which is larger than one if1 > f > 0. There is no runaway warming in this case. However, we also have this relaxation time thingy. Could anyone give an order-of-magnitude guess about how large it is supposed to be? Also, the assumptions going into the WV amplification theory are made explicit, so they can be scrutinized.adelagardeat 05:36 AM on 5 April, 2010philcat 08:26 AM on 6 April, 2010fydijkstraat 18:44 PM on 8 April, 2010chrisat 22:19 PM on 9 April, 2010science. If you are interested in paleoproxyanalysis of temperature without recourse to tree ring studies, it makes much more sense to look at the properly peer-reviewed science. This supports the conclusion that if one analyses paleoproxydata, eliminating tree-ring data sets, that the late 20th century and contemporary warming is anomalous in the context of the last millennium and more. And why go to some stuff that someone posted on the web to address the question of warming since the LIA and ocean circulation effects when this can be addressed by looking at the properly peer-reviewed science. This shows that the contribution of “natural oscillations” to the warming of the last 30 years was likely negligible. One can create theimpressionof uncertainty by basing one’s information on stuff from dodgy sources that are designed to confuse the issue (and that we know are incorrect). But if we’re really interested in these issues one really should address thescience.