Strong Negative Feedback Found in Radiation Budget

[uke2se]

Run along. The adults are going to talk now.

What are you implying, mhaze?

Why didn't you give a cite? Is this another instance of us giving sources while you don't, but you see that as you giving sources while we don't?

 

[mhaze]

How about showing a chart with an R2 value greater than 0.54.

That's low enought to say the two variables are not related at all.

You are repeating what Spencer said in his opening slide:

http://www.drroyspencer.com/wp-cont...rcing-Feedback-AGU-09-San-Francisco-final.pdf

But note the issue is that the straight line fit cannot be applied to the data and useful results gathered. That, instead of the low correlation, is the point.

http://en.wikipedia.org/wiki/Autocorrelation

So then....what does one do?

 

[At In To]

Regarding the need or utility for a longer time series and more data points, consider the following.

Suppose you had a full 30 year cycle (eg 30 years for ocean AMO and PDO effects). Is there some sense in which say the warming side of the PDO cycle would reverse the feedbacks found by Spencer? Spencer looked at short term stronger effects, irregardless of sign. I can't see that. Maybe someone else can produce an argument in that direction.

I'm not even sure that Spencer has found evidence of strong negative feed backs, that's the problem. And he has limited data, with a new method, and he hasn't applied his method to other data sets. He thinks his simple model can explain it, but I didn't see anything in there about attribution of those loops.

Tsonis showed most variation in global temperature to be the product of aligning or separation of the effects of four major ocean cyclic systems plus a background "secular trend". That background trend is quite small (I'll return to this shortly).

Because major variations could be mathematically modeled in this method (and predicted) Tsonis suggested that there was no reason for the "post 1950s temperature rise" to be considered as due to greenhouse gases, and there to be no reason to consider the 1950-1970 cool period to be due to some kind of offsetting effect by aerosol cooling.

Major variations? Spencer is arguing for reduced climate sensitivity because of his hypothesized strong negative feedbacks. Tsonis is saying that the climate shift in the 70's could be a result of convergent climate modes shifting the climate state, and superimposed on anthropogenic warming. If Spencer is right, then we shouldn't even see major variations in climate.

You're trapped in a corner if you simultaneously think that the climate is insensitive to radiative forcing due to increased atmospheric opacity, but that the observed warming can be modeled by intrinsic natural variability.

 

[CapelDodger]

Let's look at that. Spencer indicates that short time intervals have to be used to separate out forcing and feedbacks in estimates of feedbacks. Now that's not been done before, right? If it has been I'm sure you can point us to the citations?

More accurately, Spencer has to choose a short time interval to get the result he wants. There's no telling how long he'll be picking over the 2000's for confirmation that what's happened over the last thirty years (and will contiinue to happen) never could. The 2000's were the best decade he'll ever get, the Golden Decade that denialists will wish they could keep living in.

Otherwise a new and improved method is to be frowned upon because it's a "challenge to the accepted science" which said science did not have such a method? That's rather strange. So you are in favor of broader uncertainty in calculations, instead of lowering those uncertainties?

New and improved? How does that work? It's either one or the other. I think you've paid too much attention to advertising copywriters.

Short periods in climate are necessarily more uncertain than long periods, since there's a lot of noise in the system. It's pretty certain that Spencer's brave and determined efforts will fail to echo down the centuries. A footnote in history perhaps, but nothing at all in science.

Odd, that anyone would consider a clarification to "one of the major uncertainties that still remain within current models" as a "challenge to the existing science".

But Hey!

This is "climate science"!

Only in your mind and Spencer's (and in certain like minds) is this a clarification of anything. Many of us have a pretty clear picture of such minds already.

 

[CapelDodger]

Major variations? Spencer is arguing for reduced climate sensitivity because of his hypothesized strong negative feedbacks. Tsonis is saying that the climate shift in the 70's could be a result of convergent climate modes shifting the climate state, and superimposed on anthropogenic warming. If Spencer is right, then we shouldn't even see major variations in climate.

Exactly.

(mhaze has been taking the name of Tsonis in vain for years now, despite every effort to educate him. He's hardly alone in that, of course. Tsonis must find it very frustrating that his work is abused in this way.)

 

[mhaze]

More accurately, Spencer has to choose a short time interval to get the result he wants....

You really can be amusing, Capel. I hand it to you for that. But what "he wants" if he's a scientist, is decent results on the Durbin-Watson statistical measure over offset time periods. I assume that's in line with what he's got.

That's what he says he's got (although in the articles so far he doesn't mention that particular test).

I'm not even sure that Spencer has found evidence of strong negative feed backs, that's the problem. And he has limited data, with a new method, and he hasn't applied his method to other data sets. He thinks his simple model can explain it, but I didn't see anything in there about attribution of those loops. ....

Valid points, although attribution was pretty much handled in the 2008 paper I linked to, in GRL I think it was. I'd like to see a rigorous proof of the validity of short term climate data for seven odd years being of merit, or if not, then the confidence levels of the results with that time period.

.... Tsonis is saying that the climate shift in the 70's could be a result of convergent climate modes shifting the climate state, and superimposed on anthropogenic warming. If Spencer is right, then we shouldn't even see major variations in climate.

You're trapped in a corner if you simultaneously think that the climate is insensitive to radiative forcing due to increased atmospheric opacity, but that the observed warming can be modeled by intrinsic natural variability.

Here I may differ from you, because these seem completely congruous to me. Which is why I brought Tsonis up. In fact, Tsonis suggested lower sensitivity as a solution to the riddle.

If for example we have

Global(T) = y1*f(A,B,C,D)+y2*f(R)

where A,B,C, and D are independently varying ocean cyclic effects multidecadal and R is radiative forcing

Then this is opposed to constants used in a simplied "Warmer view" relevant to the discussion of

Global(T) = y3*f(CO2) + y4*f(aerosols)

Where the first term is considered positive, the second negative, and the net effect covers the cooling and warming periods of the last 60 years.

For convenience I've added the constants y1..y4 so that one can see how the two equations may be equalized by changing the ratios. It just adds up to higher or lower natural variability and higher or lower greenhouse effects.

If there's a trap in the corner I'm not seeing it. Of course I've blundered into a few corners, too.

 

[At In To]

Tsonis is not saying that climate sensitivity is low. Here is what he and Swanson are saying with regards to sensitivity:

Finally, it is vital to note that there is no comfort to be gained by having a climate with a significant degree of internal variability, even if it results in a near-term cessation of global warming. It is straightforward to argue that a climate with significant internal variability is a climate that is very sensitive to applied anthropogenic radiative anomalies. If the role of internal variability in the climate system is as large as this analysis would seem to suggest, warming over the 21st century may well be larger than that predicted by the current generation of models, given the propensity of those models to underestimate climate internal variability.

It's a trap because global temperature responds to forcings. If the climate sensitivity is low, then it doesn't matter what A, B, C, and D are. They have radiative forcings on the climate, and thus the climate should also be insensitive to those forcings.

Climate sensitivity and climate variability are related. If climate sensitivity is low, then climate variability is low. If climate sensitivity is high, then the variability should be high. You can't have high climate variability with low climate sensitivity.

 

[mhaze]

Tsonis is not saying that climate sensitivity is low. Here is what he and Swanson are saying with regards to sensitivity:....

Yes, this is a quote from him (or Swanson, I can't recall which). Doesn't matter, because the previous comment stands also. They are not mutually excluded from the set of possible realities.

...It's a trap because global temperature responds to forcings. If the climate sensitivity is low, then it doesn't matter what A, B, C, and D are. They have radiative forcings on the climate, and thus the climate should also be insensitive to those forcings.

Climate sensitivity and climate variability are related. If climate sensitivity is low, then climate variability is low. If climate sensitivity is high, then the variability should be high. You can't have high climate variability with low climate sensitivity.

I believe I've covered these cases with the previous suggestion that constants y1..y4 be considered with the two equations:

• Global(T) = y1*f(A,B,C,D)+y2*f(R)
• Global(T) = y3*f(CO2) + y4*f(aerosols)

Obviously, there could be several choices of the "y" variables that equalized these two formulas over the 20th century time series of temperature. You've described one. But focusing on that one solely would be erroneous, without some solid reasoning. That's at the core of all of this.

Further there is something that you may have missed. Tsonis noted a background "secular trend". Assuming that is 100% CO2 (and that's a really big stretch, but for grins let's assume that) then we can back compute CO2 sensitivity from the 20th century secular trend. No other accommodation for CO2 and it's effects is needed.

Doing this, you get a "low sensitivity". That is what I mentioned was quite in line with the results which Spencer found.

And yes we are simplifying all this about fourteen ways from zero.

 

[mhaze]

A further comment. You said:

It's a trap because global temperature responds to forcings. If the climate sensitivity is low, then it doesn't matter what A, B, C, and D are. They have radiative forcings on the climate, and thus the climate should also be insensitive to those forcings.

A, B, C, and D are simply oceanic cycles such as PDO. They don't add or subject energy from the climate system, over time they average out to zero. Or we think they do. You could say, of course, that their immediate effect was to add IR or subtract IR energy to the atmosphere. The incremental temperature numbers used for the A,B,C,D are total effects, net of all feedbacks and forcings.

If the past 1950 global temperature rise was due in large part from several heat releasing oceanic cycles coming together, then it was not due to CO2. Simple as that. No complicated forcings or feedbacks need to be brought in.

Now what you are saying is that if we have IR in the air, it will have certain effects based on the climate sensitivity - and it does not matter whether the IR is from CO2 or from oceans releasing heat. But the secular curve subtracted out from Tsonis's data was the total effect of feedbacks and forcings from everything other than the four oceanic cycles he later examined.

Assume that at some moment in time 1C is contributed by oceanic cycles and 0.5C by CO2, for a total of 1.5C "IR in the air". CO2 increases, and reaches an effect of 1.0C. We then have 2.0C "IR in the air". You appear to want to add to that some positive feedback.

There are two problems with this.

1. the 0.5C is measured and is net of all feedbacks and forcings. There is no other effect to be added. So going to double the CO2 from that amount which gave us the 0.5C change, would be going to 1.0C in effect EXCEPT FOR:
2. Spencer indicates net negative feedback exists.

I think that makes sense.

 

[At In To]

If the past 1950 global temperature rise was due in large part from several heat releasing oceanic cycles coming together, then it was not due to CO2. Simple as that. No complicated forcings or feedbacks need to be brought in.

Except that there is nothing complicated about an increasing greenhouse effect and ocean storage of that heat. The heat has to come from somewhere. The cycles go to zero with time, but that doesn't mean they don't have large impacts in both directions...

You're misreading Tsonis. They state quite implicitly that in this study they find large variability. It doesn't matter that these forcings go to zero with time, the impact they are having is large with the full amplitude of the pseudo-cycles, and that means a sensitive climate. You simply cannot have it both ways.

 

[BenBurch]

Except that there is nothing complicated about an increasing greenhouse effect and ocean storage of that heat. The heat has to come from somewhere. The cycles go to zero with time, but that doesn't mean they don't have large impacts in both directions...

You're misreading Tsonis. They state quite implicitly that in this study they find large variability. It doesn't matter that these forcings go to zero with time, the impact they are having is large with the full amplitude of the pseudo-cycles, and that means a sensitive climate. You simply cannot have it both ways.

You'd have to assume that those who quote Tsonis have intellectual honesty as a goal before you can assume that it matters to them that they are misrepresenting him.

 

[mhaze]

Except that there is nothing complicated about an increasing greenhouse effect and ocean storage of that heat. The heat has to come from somewhere. The cycles go to zero with time, but that doesn't mean they don't have large impacts in both directions...

You're misreading Tsonis. They state quite implicitly that in this study they find large variability. It doesn't matter that these forcings go to zero with time, the impact they are having is large with the full amplitude of the pseudo-cycles, and that means a sensitive climate. You simply cannot have it both ways.

Oh, I understand what you are saying quite well. I just don't see it is supported by the published articles.

"it doesn't matter that these forcings go to zero with time"

Of course it does matter. That means they can be separated out. That leaves the background curve as already discussed - the so called "secular curve". You can't have it both ways. Either you accept that natural cycles can and did generate the post 1950s "warming" - not CO2, or you don't accept that.

Are you trying to assert that because of positive feedbacks predominating, those four natural cycles which go to zero will not go to zero or will be much larger but still go to zero in the future? If the former, Tsonis already subtracted those effects out. If the latter, you have to reject Spencer's assertion of net negative feedback.

Which doesn't bother me, of course.

 

[BenBurch]

Again a clean miss in comprehension.

 

[mhaze]

Again a clean miss in comprehension.

Ben!

I understand now! It's all about going higher, stepping on the clouds, and finding Warmer and warmer Salvation up there! The preachers of Warmerism profitzing done did well. I have been so sorely mistaken in not following the true and Righteous path of Warmers to find Redemption.

http://www.youtube.com/watch?v=z5bygrq32Vk&feature=related

I'm going higher...higher...higher...stepping on those clouds. Hockey sticks - Hallelujah! I sees them now!

 

[At In To]

"it doesn't matter that these forcings go to zero with time"

Of course it does matter. That means they can be separated out. That leaves the background curve as already discussed - the so called "secular curve". You can't have it both ways. Either you accept that natural cycles can and did generate the post 1950s "warming" - not CO2, or you don't accept that.

No, it really doesn't matter. We are discussing variability and sensitivity here. Maybe an example will be illustrative. If you have a climate where a particular cycle produces one degree of warming at the high end, followed by one degree of cooling at the low end, it still goes to zero with time. If you have a different climate where the same cycle produces 0.01 degrees of warming at the high end and 0.01 degrees of cooling at the low end, it still goes to zero with time.

One climate is insensitive to the variations in that cycles forcing. The other is not. The average being zero doesn't mean that the climate is insensitive, it means that it is a cycle, which averages to zero. Cycles should wash out to zero, unless they aren't really cycles at all. The variability is what matters, and Tsonis' work with climate states clearly implies that variability is large, thus the sensitivity is also large.

Think of a real world example. Is the ocean insensitive to the gravitational influence of the moon? No, it definitively is sensitive to the influence of the moon, and sun.

Are you trying to assert that because of positive feedbacks predominating, those four natural cycles which go to zero will not go to zero or will be much larger but still go to zero in the future?

...NO. Positive feedback has nothing to do with what I am saying...I honestly have no idea how you derived that from what I am telling you.

 

[mhaze]

So are you saying that if co2 sensitivity is low, but climate variability is high, then that is a problem later when there is more co2, because it is more or less a multiplier on the variability?

And this has nothing to do with feedbacks?

Yep, I'm not seeing that. Putting it in plain English: On the premise that CO2 has a low sensitivity (and therefore Copenhagen, cap and trade etc are all economically and scientifically wrong) somehow high natural variability is a bad thing?

Well, maybe all those Copenhagen people could go on to natural variability?

 

[At In To]

So are you saying that if co2 sensitivity is low, but climate variability is high, then that is a problem later when there is more co2, because it is more or less a multiplier on the variability?

No, I'm not saying anything about CO2 sensitivity. I'm talking about climate sensitivity, whether you choose that as a doubling of carbon dioxide, or some other perturbation with an equal radiative forcing.

What I'm saying is:

Climate sensitivity is the magnitude of the response to a forcing. So if the sensitivity is low, then it means you need very large changes to get even small amounts of climate change.

And this has nothing to do with feedbacks?

Again, not understanding where anything I said implied this. We got off of the Spencer feedback when you brought in an incorrect statement of what Tsonis and colleagues have found.

Yep, I'm not seeing that. Putting it in plain English: On the premise that CO2 has a low sensitivity (and therefore Copenhagen, cap and trade etc are all economically and scientifically wrong) somehow high natural variability is a bad thing?

On the premise that the climate sensitivity is low, then high natural variability is not possible. You need to invent entirely new physics if you wish to show these to be compatible (low sensitivity and high variability).

 

[mhaze]

Okay, you are talking about climate sensitivity in the generic. That makes a bit more sense.

Now what is left to make sense out of your stand is some specific meaning, range or clearer definition of what is "high" and "low".

As an example, the IPCC midpoint of a doubling of CO2 causing a 3C temperature change I consider "high". Spencer's calculation of 0.3-0.6C for the doubling of CO2 I consider "low".

The differences are in the feedback parameters which is why I asked about that.

 

[bit_pattern]

If CO2 sensitivity really were that low then we'd have a terrible time trying to explain how the planet recovers during the interglacials in the way that it does.

 

[CapelDodger]

You really can be amusing, Capel. I hand it to you for that. But what "he wants" if he's a scientist, is decent results on the Durbin-Watson statistical measure over offset time periods. I assume that's in line with what he's got.

What Spencer wants is confirmation of his belief, just as he does when finding ways to reject evolution by natural selection (which also contradicts his beliefs). As a scientist he must be a very conflicted human being, but his beliefs come first.

Surely you can empathise with that? Even as a non-scientist.

What Spencer has arrived at, rather desperately, fails the test of what has actually happened in the last few decades, and will continue to fail the test. He also falls into the trap (as Lindzen did before him) of proving that the current inter-glacial never could have happened. His data-mined negative feedback from a slice of the 2000's would have prevented it, just as Lindzen's Iris would have had it ever manifested. Just as negative feedback from clouds would have. Just as every other desperate expedient would have.

Yet here we are, and here we go. I don't doubt that you'll still be here in five years time grasping at some other straw, but I'm less sanguine about Spencer. Being as conflicted as he is, and fighting so many losing but never (in his own mind) lost battles cannot be good for his health. And he's no spring chicken, as the saying goes.