Friday, January 16, 2015

Why I think Climate Sensitivity is Essential for Developing Effective Climate Change Policy

For those of you recent to this blog, my primary readership is not typically experts in climate change science but rather people interested in the policy implications of climate change science. This includes people with interests in renewable energy technologies and governmental decision-making. What this means is that occasionally I need to step back and provide some more detail about the terminology used in the discussion. I normally do this in response to questions and in this case the one question I have received again and again is: why do you keep harping on climate sensitivity?
Personally, I think refining climate sensitivity estimates might be one of the most important things climate scientists can do to help to establish a consensus on the policy and political implications of climate change. Admittedly my opinion is not universally accepted. One of my regular commenters, Dr. Michael Tobis at his blog Planet 3.0, takes an entirely opposite view on the topic. Other blogs present similar views to Dr. Tobis, most of them are based on the argument that sensitivity is reported as a range and as long as a risk of a higher end sensitivity exists then we have to take action now. My response to this will be detailed below but first, for my non-climate scientist readers out there, let's go back to the basics (and in this discussion I mean basics I'm not looking at transient climate sensitivity etc...).

In plain English climate sensitivity represents the anticipated heating expected based on changes in CO2 concentrations in the atmosphere. Climate sensitivity is formally defined as a measure of the anticipated equilibrium temperature change in response to changes of the radiative forcing associated with Tyndall gases. Radiative forcing is the measurement of the capacity of a gas or other forcing agent to affect that energy balance, thereby contributing to climate change. Radiative forcing expresses the change in energy in the atmosphere due to Tyndall gas emissions (reference).

What many of my readers may not know/understand is that due to the chemistry, increases in CO2 concentrations do not have a linear relationship with anticipated heating. Or put another way, 1 unit of COdoes not result in 1 unit of anticipated heating. Rather, each molecule of COadded to the atmosphere has a tiny bit less heating effect than the molecule added before it. Specifically, rather than a linear relationship the relationship is logarithmic. As detailed at the Skeptical Science web site (yes, many of my readers dislike that web site but while it has a deeply political bent it does serve as a useful information source) “this  logarithmic relationship means that each doubling of atmospheric COwill cause the same amount of warming at the Earth's surface. Thus, it takes as long to increase atmospheric CO2 from 560 to 1120 parts per million by volume (ppmv) as it did to rise from 280 to 560 ppmv".

So why do I care about climate sensitivity? Well it becomes a bit of a math game. If climate sensitivity is determined to be 1.5oC then each doubling in atmospheric COconcentrations would result in an increase in 1.5oC. Using the Skeptical Science numbers, an increase in COfrom 280ppmv to 1120 ppmv would represent two doublings, so with a sensitivity estimate of 1.5oC you would expect an increase of 3oC (two times 1.5oC). If your sensitivity estimate is 4.5oC then the same 3oC would occur below 500 ppmv. So you ask why sensitivity is important? Well the math is pretty easy here. If we only have until 500 ppmv to avoid 3oC (which in this thought experiment we will define as a less than scientific "point of disaster") then we need to act immediately. If we have until 1120 ppmv then we can wean our society off fossil fuels more gradually and decrease the pain.

Now that we understand the concept, it becomes clear how fundamental differences with respect to climate sensitivity estimates will affect political and policy implications. In one corner we have extreme Lukewarmers who think that climate sensitivity is around 1oC (note I am not from that group). For them getting to the 3oC (which for our thought  experiment is really bad) would entail three doublings, or atmospheric COconcentrations of 2240 ppmv. Even at our current rate of atmospheric COdeposition that is not happening anytime soon. So trying to convince them that we should divert trillions of dollars to completely re-tool our planetary energy/power systems would be a pretty hard sell. At the other extreme are people who believe climate sensitivity might be closer to 6oC. For them we are not merely approaching the cliff but have already crossed over it and are starting the dive to collapse. For them immediate action is essential to avoid a hard landing at the bottom of the chasm. In both cases we have good, honest people who have a genuine disagreement about the science but their opinion on the science leads to dramatically differing policy positions. One side says that as long as we get this done in the next 50 years or so we are fine and the other thinks that every day's delay will result in more pain and hardship. The first group will deeply resent being asked to make serious sacrifices to address climate change and the second group will think that actions are so necessary that maybe governments should be forcing people to make changes involuntarily. 

To add complexity to the discussion is the point made above by the colleagues of Dr. Tobis. There are lots of things we still do not understand about our climate. Right now we have a pretty good thing going. We live on a planet in an interglacial period with a relatively mild climate with storms and disasters that come at infrequent intervals. Any change could tip that balance towards a world with less good days and more bad days (or maybe even the other direction). Since our civilization has thrived under current conditions isn't it thus a good idea to try and keep it in this current place? Also for the gamblers out there we are reminded that sensitivity represents a range and while there may be 90% (this is a made up number as no one actually has a clue what the real numbers are) chance of a small sensitivity that means there is still a 10% chance of a bigger number and given any non-trivial uncertainty, isn't it better to not role the dice? 

The most important thing to recognize in all this that any uncertainty in the science is going to be magnified by people at both ends of the spectrum who have political and/or financial stakes in either maintaining the status quo or eliminating the status quo. The best way to get these people off message is to reduce the uncertainty. From a policy perspective uncertainty also makes for bad policy as the more contingencies you have to consider the more complex (and thus the less practical) the policy will have to be. Finally, as we work to reduce the uncertainty it would be nice if scientists stopped feeding denialists and catatrophists all that red meat. As a knitter myself, I can say this: stick to your knitting and to mix my metaphors horrendously, stop running around stamping on each other's toes.

Author's note: at this point in this posting at least a couple readers are going to be apoplexic over my use of 3oC in my thought experiment. Please accept that the choice was simply to make the math easier for the discussion. I know that the IPCC has a 2oC target but that number is very much a topic for a future posting.


  1. I think uncertainty over what the real threshold of danger is does need to be folded into the risk management. And I think THAT uncertainty dominates and is what we should be narrowing if we can. I am not sure that we know how to do that. But essentially enough climate science is done that it no longer dominates the uncertainty of what we should be doing.

    In the end, it is possible that the only way we can greatly reduce the uncertainty of what the cataclysm boundary is is to have the cataclysm and then say we shouldn't have gone that far. I think that this is not a good research program.

    But even that isn't really my point.

    My point is that in Rio in 1992, the world was unanimous that the time for policy action had arrived. Bush Sr. signed on to it. This was based on a sensitivity range that notoriously is not much different from the one we have now.

    In the intervening quarter century both accumulated concentrations and emissions have soared. So if it was the right time to act in 1992 on a 3 C sensitivity, arguably it is already past due for half that sensitivity.

    By past due, I mean that the recommended rate of decarbonization is "as fast as is economically feasible". This holds even at the low end of the IPCC sensitivity spectrum, because we are already a quarter century behind the risk management curve.

    Exactly when we catch up to the risk management curve and what to do then is worth discussing, but that is so far away that in the foreseeable decades the sensible policy is not sensitive to the ECS.

    1. In the late 30's several large nations decided to make colossal mistakes which led to tens of millions of dead. In 2003 Mr Bush, aided and abetted by the USA media and both major parties' elites, committed a huge blunder when they invaded Iraq. This tells me that being swayed by past decisions isn't really the intelligent thing to do. You do have to muster a coherent and solid argument to convince me to do what you want. These calls to bow down to presumed authorities really don't go over well when one is old, educated, and knows a little about the subject.

  2. Dr. Tobis

    In my opinion, this is a classic case of two folks talking past each other. You say that the recommended rate of decarbonization should be "as fast as is economically feasible" so the sensitivity does not matter. Our host and I both think that economic feasibility depends on the sensitivity. If the sensitivity is high it is more feasible to spend more today to avert future harm. Unfortunately, there isn’t a technology for widespread replacement of fossil fuels that is close to economic parity. To say we have to do it now means you better be sure that the sensitivity is high or the costs just don’t make sense.

    That is not to say that there aren’t things we can do now. I support continued research for fossil fuel alternatives, more energy conservation and energy efficiency, but I think continued development of fossil fuel resources is necessary and appropriate.

    1. No it's not. I stipulate, and it is obvious, that the recommended rate of decarbonization DOES depend on sensitivity for a range of sensitivities. But our host has stipulated a sensitivity that is high enough that action is overdue.

      Therefore, the best advice has been "decarbonize as quickly and fully as is feasible" for a few years, though perhaps not that very many. (I put Copenhagen as roughly the threshhold of "too late for an near-optimal outcome.)

      Whether or readers believe not my claim that we are past that threshhold of urgency, I would like to know who among them accepts the proposition that such a threshhold might exist as a function of sensitivity.

      So maybe it will be clearer if I split my claim into a few parts:

      1) If we consider that science doesn't really know the sensitivity or the transient response as well as we'd wish, and we contemplate the best strategy as sensitivity (S) increases, there is some value for S for which intervention is too late to avoid very substantial, possibly enormous damage to the biosphere and likely to human well-being on other axes as well.

      2) Once you reach the "too late to avoid damage" the damage climbs very steeply the more a response is delayed.

      3) We may have exceeded that threshhold and are probably are very close to it.

      4) Even if the sensitivity is at or slightly below the bottom of the IPCC range, as Blair stipulates, that is "lukewarmism" in a rather vacuous sense. Even at half the sensitivity, "business as usual" carries unacceptable risks. Given the very slow time constants of human infrastructure, action is either overdue or urgent.

      "As hastily as is consistent with safety on other concerns" is the same, regardless.

      Now if has the sensitivity at 0.5 and the danger threshhold (let's call it D) at 4 C, we have all the time in the world. I understand that.

      But that isn't lukewarmism; that's denial in the literal, psychological sense.

      My argument then isn't that I miss your point. I understand it well. I am just saying that there are values of S and D where it doesn't apply.

  3. If you are being chased by a hungry tiger, there is some distance from the tiger below which the best advice is to sprint for all you're worth. At that point how fast you should run is not strongly dependent on whether your odds of getting away from the tiger are good, tolerable, poor, or terrible, nor on how big the tiger actually is.

    1. The option is not to sprint away but to jump off a cliff. You know the dangers of the jump, but you're not sure about the tiger. Others assure you that you will invent the parachute on the way down. You wonder if it might be wise to stop and think.

  4. If you are being chased by a hungry tiger,

    IF. The key word here is IF.

    If you THINK you are being hunted by a tiger, which you can't actually see but you're told is coming, your best option is probably not to run away in panic.

    Humans don't work best as individual atoms, even if the threat is real, each running for their own life like a caveman might. The best option is almost always not to panic and run away, but to work together using modern techniques.

    The ones advising panic are the ones to avoid. Especially if the threat can't even yet be seen.

    1. Yes of course the key word is "if". But I am personally acquainted with some of the world's leading atmospheric scientists as well as their work. They're pretty convinced there's a tiger. As am I.

      My claim is that once you know the ropes it is obvious.

      But that is all really off topic for the topic at hand. The topic is whether we need to refine our knowledge of S as an input to policy. My answer is, no, not immediately.

      Things are so out of hand that we need to decarbonize as quickly as is possible without seriously damaging economic systems, especially the continued transition of the poorest countries into the modern world.

      Unless we are far more lucky than we deserve, there is no avoiding huge costs any more.

    2. “Things are so out of hand that we need to decarbonize as quickly as is possible without seriously damaging economic systems, especially the continued transition of the poorest countries into the modern world.” And what if all those things are not possible? What if one of the reasons there is so much opposition to the science is because there are no magic solutions and it really, really, really matters how bad things are going to get and how accurate the science is, so that we can target the response rather than flapping about aimlessly?

      If the ambitious targets are to be met then something has to give. Either we will have power or we won’t. Renewables at the moment can’t provide more than a fraction of our needs and then very unreliably. Our current systems need reliability more than we need affordability. There’s no change of that position in sight. So what gives?

      Since we can generate emissions free electricity, we need to be converting everything to run on electricity, which would involve huge capital investment, plus the arrival of efficient, effective alternatives and huge amounts of extra electrical generation. The last of those is currently only possible through the use of nuclear power but for those who profess to be most worried about AGW, nuclear is even more taboo. Without the power, there’s no point in investing in the fancy new electrical society. No point pinning our hopes on electric cars if there’s not enough for essentials, never mind luxuries. To convert before the emissions free generation is sorted would actually increase CO2, rather than decrease it. More importantly it would waste energy, money and public good will. Think of all the people who are investing in new boilers, when they’re told ‘sorry, you’ll have to scrap that expensive gas boiler we insisted you have last week and put this electrical one in instead. Oh, and by the way your bills will be astronomical and your job has gone to China because we agreed it was your turn to suffer.’ It may even be that our existing housing stock and industry can't be made suitably efficient and we need to rebuild almost everything from scratch. What would that take?

      Failing an arrival of copious amounts of cheap, CO2 free electricity and suitably useful equipment to replace what we have now (eg cars), there is only one sure fire way to reduce emissions. Have less, do less, use less. At that point we aren’t seeing the poor climb out of poverty we are sending everyone else sliding back to it. To embrace that, how sure would the human race have to be of climate science?

    3. Thank you. Those are the right questions. (A rare treat in the climosphere.)

      I disagree on a fundamental point. Nuclear is not taboo among those who understand the seriousness of the problem. Very few actual climatologists are anti-nuclear, and those that are are far more anti-fossil-fuel.

      That said, all this delay may have created an opening for new battery technology, which seems to be verging on commercially viable. There is even a case that a largely electric car fleet provides a large fraction of the load balancing we need. Aggressive pricing variation to meet availablity would also help - heavy industry on a low margin would then reduce output at times that the wind and sun weren't cooperating.

      I also agree that we can't really do this faster than the infrastructure turnover rate; that means that you replace your boiler when needed with an electric one, but keep your gas one for now. And similarly for industry and for utilities. And that is exactly why we need to begin the transition long before the consequences of climate change become expensive on the scale of the whole economy.

      These points are excellent, and they fit into the definition of "as fast as possible". Avoiding an economic tailspin is exactly the goal. Having a crash driven by overreaction sooner than a crash driven by procrastination later solves nothing. But the needle we need to thread gets narrower by the year.

      As to your final query "how sure would the human race have to be of climate science" that brings us back to the point that is in dispute.

      Our host has stipulated a sensitivity of half of the IPCC consensus. If you understand the graph I have drawn here, you will see that half the sensitivity does not give us anywhere near as much slack as you might expect.

      The anthropogenic perturbation is historically roughly exponential with a doubling time of decades. And if on the consensus sensitivity we are decades late in addressing this problem, at half the sensitivity, the problem is merely timely.

      Advocating further delay does not mean you are unsure of climate science as you imply. Rationally speaking, it means you are **sure that it is wrong** about S and/or D, and sure that it is wrong in the direction of overstating S and/or understating D. Worst cases weigh heavily in risk management.

    4. Missing link, sorry. "graph I have drawn here" whould point to

    5. See also

    6. Yes of course the key word is "if". But I am personally acquainted with some of the world's leading atmospheric scientists as well as their work. They're pretty convinced there's a tiger. As am I.

      Oh you sweetie you! Well that you like the scientists involved is a red herring of major proportions. Also, some of us have read the work e-mails of some of those leading scientists, or listened to their public pronouncements, and are less than impressed at their professionalism.

      You remember our host's difference between "show us" and "trust us"? Well you obviously still think "trust me" is a valid reason. I've read a fair chunk of the material too, and I come to different conclusions. (To be fair, it is mostly not atmospheric climate science I have issues with, but that was a red herring too.)

    7. Shrug. A PhD in the field qualifies me to have my own opinion.

      If you focus on the worst interpretations of worst things in ten years of emails, you won't, by definition,see people at their best. My personal experience is that these people are every bit as smart as any of the many engineering profs I've met and that the scientific material is rigorous and cohesive and deep. But that's impossible to convey casually.

      You certainly don't get that from rotten-cherry picked emails. I get that.

      Can we move on now?

      The operative question is, how sure are you that climate scientists are fools who are vastly overstating the risk. Keep in mind that you are betting the farm. You better be damn sure that I am lying or crazy.

  5. Blair, I just retired after working in the oil industry for almost four decades. One item I'm trying to connect is the amount of hydrocarbons and coal we can find and produce. This was triggered by the extremely high projections I saw in the IPCC AR5 RCP8.5 (sorry about the jargon).

    I did a quickie estimate using what I felt was a reasonable amount of resources, and I estimated we would get to 630 ppm CO2. This elevates the sensitivity as well as potential geoengineering to the top of my list.

    Right now I'm more worried about overpopulation, erosion of human rights, and the conflict mongering by extremists on all sides. Global warming just doesn't make my top three.

  6. Blair, a much overlooked but very important assumption in the IPCC scenarios is the development of co2 sinks in the future. The full path is: emissions -> concentrations -> temperature. All IPCC vetted models assume that the sinks will saturate in the near future and that the airborne fraction of co2 will explode. However so far the airborne fraction has been remarkable stable.

    1. I don't believe IPCC scenarios make any strong claim about changes in airborne fraction. Chapter and verse, please?

      If I understand correctly, the scenarios are just concentration scenarios. There's no coupled carbon cycle in the CMIP experiments.

      The climate system in GCMs basically only cares about the radiative properties of the atmosphere, and those are what is specified. If we have emissions scenario A with (realistic) 50% AF and scenario B with half the emissions and 100% AF, I believe those are the same scenarios insofar as the GCMs in question are concerned.

      It certainly is the case that biosphere uptake may saturate or even reverse. That would make matters somewhat worse toward 2100 and beyond if we continue to fail to get a grip on this.

      But I don't think it's a front burner issue either way.

  7. It's a relatively trivial point, but since you say you are going to write more on it later:
    The 2C target is not an IPCC target, it's a rather arbitrary thing settled on by some politicians and activists. The IPCC is supposed to be policy neutral.

    1. That is correct; IPCC should not make that recommendation. But since IPCC is essentially the relevant sciences consulting to the policy sector it **is** in their remit to address the interest from the policy sector as to what it would take to hold to that agreed-upon target.

  8. ==> "So trying to convince them that we should divert trillions of dollars to completely re-tool our planetary energy/power systems would be a pretty hard sell."


    How about trying to convince them that investment in renewables will reduce a variety of negative externalities, and might turn out to return a positive value relative to the positive externalities of BAU?

  9. ==> "So trying to convince them that we should divert trillions of dollars to completely re-tool our planetary energy/power systems would be a pretty hard sell."


    How about trying to convince them that investment in renewables will reduce a variety of negative externalities, and might turn out to return a positive value relative to the positive externalities of BAU?