It is time to stop quivering in our boots in pointless fear of the future and just roll up our sleeves and build it.
- Ray Pierrehumbert

Friday, October 30, 2015

Some Miscast Pearls: On Climate Sensitivity

I wrote this on Tom Fuller's. I hope you appreciate it more than his crowd did:

Others may be accused of shifting goalposts, but I am innocent of the charge.
The ECS is probably between 2.5 and 3, where it has always been. The TCR is ill-defined, as it has always been. Global warming is not the key socially relevant metric of climate disruption, as I said before anyone was talking about the hiatus and as I still say now that the hiatus is apparently ending.
Changing the radiative properties of the atmosphere to a larger extent than the Milankovic forcing will cause substantial climate change.
The changed climate will almost certainly be substantially warmer, whence all the obsession on global mean surface temperature, which is after all relatively easy to measure. But the warmed global climate is not the cause of climate change – it is a consequence of climate change.
The main issue that people will have to deal with is not this global measure, but that this measure is the result of climate changing locally everywhere in ways that are hard to predict and hard to manage. The extent of change can already be expected to be comparable in magnitude to the Milakovic cycle changes but much faster, leaving local adaptations, both artificial and natural, stressed and at risk of decline. The longer we delay in reaching carbon neutrality, the larger the disruption we have committed to.
Nothing in the recent observational data calls any of this into question in the least.
The process is slow on political time scales. Talk about what will happen to GMST in 2017 is a fine example of missing the point. People may be silly enough that this will affect politics. But the climate system won’t care a bit.
The response of the climate system is dominated by how much fossil carbon gets emitted from the beginning of the industrial revolution until the day we reach carbon neutrality. The rest is secondary.

8 comments:

Steve Bloom said...

"But the warmed global climate is not the cause of climate change – it is a consequence of climate change. The main issue that people will have to deal with is not this global measure, but that this measure is the result of climate changing locally everywhere in ways that are hard to predict and hard to manage."

The GMST increase is a result of climate changing locally everywhere? I think I see what you mean, but it's a subtle point that doesn't seem necessary or even helpful to your argument.

Tom said...

https://thelukewarmersway.wordpress.com/2015/10/29/bygones-forgetting-the-climate-past-reinventing-the-climate-future/

Michael Tobis said...

Steve, I think it's important. A

ll physics is local. Global temperature emerges from local phenomenology. So often people speak as if it were the other way around.

sylas said...

"But the warmed global climate is not the cause of climate change – it is a consequence of climate change."

I have trouble with this one. As I see it...

The major disrupting factor for climate is a change in the composition of the atmosphere so that it interacts more strongly with IR radiation.

The immediate consequence of this is warming. The surface has to be warmer to get rid of the heat absorbed from the Sun.

The consequence of this warming is climate change; but the changes to climate are strongly local. The atmospheric changes are -- to a good first approximation -- global; because the atmosphere is fairly well mixed. The reduced efficiency of radiative cooling is, to a first approximation -- global. The climate consequences depend strongly on location; they are local.

So I would have put it the other way around. Climate is changing everywhere, in a whole range of different ways. The major cause of changing patterns of climate in the present epoch is a global reduction in the efficiency of radiative cooling. Averaged out, climate is warming; and that is because the CAUSE of the changing patterns of climate is a global change in the capacity of the planet to cool (to shed heat absorbed from the Sun) or more simply; present changes in the climate are caused by a global warming.

To say that global warming is caused by climate change seems to me to be just wrong. Sure; you can only measure global warming by combining lots of local changes. But it is no accident that the result is warming... because physically, the causation goes from global warming to local climate change.

andthentheresphysics said...

Changing the radiative properties of the atmosphere to a larger extent than the Milankovic forcing will cause substantial climate change.

I think this is a key point. We're going to make a change comparable to that - in magnitude - of the change between a glacial and an inter-glacial. We're also going to do this faster than its been done at almost any time in the history of the planet. The climate is going to change. Some of the changes are likely to be substantial and damaging. The only way to avoid this is to get net emissions to zero as soon as we can.

Michael Tobis said...

"The major disrupting factor for climate is a change in the composition of the atmosphere so that it interacts more strongly with IR radiation."

yes, certainly.

"The immediate consequence of this is warming. The surface has to be warmer to get rid of the heat absorbed from the Sun."

This is something of an abstraction. We have to think in terms of an entire atmospheric column for this to occur, and each atmospheric column reacts independently. There are first order terms in the adjustment that have to do with the column's relationship to its neighbor columns. And those in turn are affected to first order by large scale emergent processes.

Conceivably, some regions could have enhanced low cloudiness which could keep the global balance by increasing albedo. That indeed was Lindzen's hypothesis. It appears to be false, on paleo evidence alone. But it is plausible enough that it means the argument from local column radiative imbalance to global surface warming is not a simple one.

Another point, of course, is that ice sheets could be destabilized, leading to warming in Greenland and a very long period of cooling elsewhere. This idea appears also to be unlikely on current evidence, but it's not just a simple global forcing argument.

My point is that the imbalance affects the physics locally. The adjustment process is global. The global average column model warms, but in practice there can be local phenomena that could affect the key terms of the of the average in ways that aren't explicit in the global average column model.

Yes, global warming is a near-certain result in the very long term. But large scale processes are emergent processes which emerge from the statistics of instantaneous local responses. It is not the other way around. All physics is local.

(At least, except for some very weird quantum effects which do not concern us much outside physics labs. Screws up my aphorism. Grumble. But really not relevant.)

sylas said...

"All physics is local."

I prefer: "All phenomena emerge from local interactions."

Physics deals with both the local, and the global. IMO most of physics is about combined emergent phenomena. Temperature itself only exists as a emergent property of lots of local interactions; thermodynamics is all about physics of combined local interations.

In the post you say "But the warmed global climate is not the cause of climate change – it is a consequence of climate change."

In a comment, you explained this as "Global temperature emerges from local phenomenology."

I'm okay with the latter phrasing; but not with the original phrasing in terms of "causation".

Local phenomenology is a result of long chains of local interactions coming from all over the globe. A single direct measurement is local in time as well as location, but its "causes" are global.

Changes in the composition of the atmosphere distribute over the globe, and have physically predictable global consequences; it makes it harder for the Earth to shed heat. The local phenomena we measure are caused by arbitrarily complex chains of local interactions and are not predictable in the same way. But physics -- especially thermodynamics -- lets us draw some definite conclusions about combined effect of all those local interaction without needing to identify and analyze them all individually.

Climate is currently changing more rapidly than usual and (in this instance) the rapid pace of change is a physical consequence of a rapid change in the atmosphere. The physical mechanism by which the change in the atmosphere is presently driving climate change is through increasing temperatures. As you say, locally it's possible to get reduced temperature -- most certainly if we speak of local in time as well as place. Physics still means there is a global increase, because the changes in the atmosphere make it harder to shed heat.

Hence I prefer to say: climate change -- the changes in climate we measure and experience locally -- are caused by long chains of interactions from all over the globe. Overall warming of the globe is causing the particular (hard to predict) patterns of local change now underway. The measurement of global temperature is not a single direct measurement, but a calculation from many local measurements. The increase in global temperature we calculate from distributed data sets is being caused by changes in the atmosphere which is in turn causing a whole range of diverse local changes around the globe.

PS. Sorry if this comes as a second comment. I lost the page while writing the first time and I don't know if it got submitted. If you see a poorly edited comment from me a bit prior to this, feel free to delete it. And thanks for the thought provoking post!

Michael Tobis said...

Sylas, thanks, but I'm sticking to my guns.

Consider that, as RPSr is always emphasizing, significant climate change is possible in the absence of any change in GMST.

With greenhouse forcing (leaving aside all the other anthropogenic forcings which may well be important) it's true the forcing is global, and it's true that a global mean column model gives us a meaningful instantaneous sensitivity, but the complexities missing from that column model are a consequence of the fact that the model is averaging out an immense variety of local radiative balances, each of which responds to local changes in condition. That's why we need GCMs after all.

Also it's worth noting that the Milankvoc forcing has a global direct forcing of zero. So plugging that into a global mean column model makes it disappear, taking the glacial cycle along with it. That doesn't mean it wasn't a mite colder 20Ka ago.

Yes, of course there are emergent properties from aggregating all those local physics. The existence of the word "climate" in the first place demonstrates that. But they emerge as a result of local processes.

It's true that there is global warming in response to greenhouse gas accumulation. But even if there were little or no global warming owing to something like Lindzen's iris effect or the Day After Tomorrow scenario, there would be climate change. Global warming is very likely to continue as greenhouse gases accumulate, but accelerating climate change is absolutely certain.

Stephen Schneider once said something very like "You can't add four watts per square meter and have nothing happen". A near-exact quote, I think. Yes. That.