Can anyone explain Roy Spencer's climate model?

Spencer's work is a variation on a theme. That graph from the link above looks familiar. Golly, see here:

http://www.realclimate.org/index.php/archives/2008...

Spencer, for all his bluster about god and so forth, should be glad there isn't a commandment to always adhere to the truth.

edit: Keep in mind that Figure 4 is an "optimized" plot so that the PDO forcing and mixed-layer depth are very likely not the same for the two curves. That explains why the PDO Only is different than the PDO+CO2 curve, the conditions are different. If Spencer had fixed PDO forcing and mixed-layer depth, the PDO+CO2 curve would lie a little above the PDO Only curve. It's all wrong anyway, but I like where Spencer complains his paper submitted to GRL was rejected. It's going to be harder and harder for Spencer to get his stuff through peer review since he has burned bridges at a lot of journals by publishing stuff later found to contain misdirections and data cooking. Soon, he will be relegated to Energy and Environment where he belongs.

Dana,

I must admit that I haven’t read anything about this yet (so many papers, so little time), but if I understand you correctly, what you’re saying is…

Either CO2 had no effect at all in the early part of the 20th century – in which case the lines should be identical – or CO2 was having an effect, but if it was, then the PDO + CO2 line should be *higher* than the PDO only line. Since, on the graph the PDO+CO2 line is lower than the PDO only line, he must be suggesting that CO2 has a cooling effect.

Is that what you’re saying?

The short answer is, I’m not sure, you’d have to ask Roy, but a quick look at the next paragraph reveals the following quote…

“If I include the CO2 and other forcings during the 20th Century complied by James Hansen with the PDO-forced cloud changes (solid line labeled PDO+CO2), then the fit to observed temperatures is even closer.”

I think the relevant part of that sentence is “and other forcings”. I can only guess (and this is nothing but a guess – I don’t *know*) does James Hansen’s “CO2 and other forcings” data show a net cooling in the early part of the 20th century?

Just a thought.

"Does this mean that Spencer's model somehow attributes a cooling effect to the increasing CO2 in the early 20th century? What the heck is going on here?"

No. If you had read a little further, you would have seen this quote by Spencer:

"If I include the CO2 and other forcings during the 20th Century complied by James Hansen with the PDO-forced cloud changes (solid line labeled PDO+CO2), then the fit to observed temperatures is even closer."

So the PDO+CO2 is not just PDO and CO2, but other forcings as well.

Benjamin wrote

"CO2 will be negated because of negative feedback from changing cloud cover. It looks as if he has just proven his own hypothesis wrong."

Clouds are tricky--generally, low cloud cover (LCC) is a positive feedback, while HCC is a negative one. But then again, LCC can also provide a negative, and HCC a positive. The trick is averaging their effects over time over the entire globe.

But here is the difference with Spencer's model--he does not prescribe cloud cover as a feedback (well, technically he does, at least partially. It is novel in its approach), but rather an actual forcing.

I don't have a lot of time right now, but i'll take a better look at it in a few hours. So far, it doesn't look particularly promising.

How can anyone claim that they are following sound principles of science when they put more wait on theoretical evidence than empirical evidence. Over the past century temperature trends have followed the PDO.

http://icecap.us/images/uploads/PDO_Easterbrook.JP...

(Easterbrook, 2001, 2006, 2007) predicted that late 20th century warming would last until this time as the PDO shifts phases

http://icecap.us/index.php/go/new-and-cool/shiftin...

Temperatures for the past two years have dropped severely at the exact same time the PDO has shifted phases.

In the first link I provided, notice when the PDO was in its cool phase, how many and how severe the La Ninas were. Notice also how few El Ninos there were. Now look at the warm phase, notice how the opposite occurred.

It does not matter how much cooling we have. Even if temperatures drop by 0.1 degrees as it during the mid century (it actually dropped more before the Hansen changed the data to support his theory), input those variables in a computer model and tell me what is the future temperature prediction.

The little outtakes you've provided sound more like a grade school article than someone with a PHD in Meteorology. He even admits that 50% of emissions by man are not being absorbed yet is looking completely past what other scientists, and most of the world for that matter, see as fact in that CO2 is a greenhouse gas.

I also found his figure 4 interesting. His model of PDO and PDO+CO2 are basically the same, and correlate nicely with the thermometer data up until about the 1970's. FYI, solar output corrolates well with global average temperatures up until the 1960's-1970's as well. After the 1970's, the PDO+CO2 model is a much better fit with observed global temperatures than the model for only PDO.

This tells us that CO2, and not the PDO, is the main forcing behind our recent global warming, right? Spenser even states that, "If I include the CO2 and other forcings during the 20th Century complied by James Hansen with the PDO-forced cloud changes ... then the fit to observed temperatures is even closer."

Now, as far as his modeling of the early 20th century, the PDO-only model is a hair warmer than the PDO+CO2 model. Please note that the PDO model is really a "PDO-induced cloud cover" model. Perhaps what Spencer is showing is that changing cloud cover is a positive feedback that will enhance global warming.

Cloud cover being a positive feedback that enhances global warming is kind of the opposite of what Spencer has been saying over the last few years. Spencer had been claiming that global warming due to rising concentrations of CO2 will be negated because of negative feedback from changing cloud cover. It looks as if he has just proven his own hypothesis wrong. Roy Spencer should probably stick to having his findings peer-reviewed for clarity instead of posting them on web blogs.

I agree with JS, there's still a lot to learn about climate change. If we discovered that the climate is less sensitive to CO2 than we currently believe, it'd be a big relief. I wonder if all of the skeptics would suddenly trust the scientists?

He is modeling PDO and cloud cover (and CO2). Even when he has CO2 + PDO showing slight temperature increase versus PDO alone, it is not an indication that CO2 is driving the direction of temperatuer increase. I am sure you like models that have many more assumptions and are prone to many more errors. He is not suggesting that CO2 is causing cooling in the early 20th Century. Models never perfectly fit nature unless you had all variables and knew everything and accounted for every molecule. I don't think he would argue that his model is a perfect predictor either. He is demonstrating that CO2 is not as sensitive as IPCC suggests and some of IPCC assumptions are probably wrong.

He sums up his conclusions nicely:

The evidence continues to mount that the IPCC models are too sensitive, and therefore produce too much global warming. If climate sensitivity is indeed considerably less than the IPCC claims it to be, then increasing CO2 alone can not explain recent global warming. The evidence presented here suggests that most of that warming might well have been caused by cloud changes associated with a natural mode of climate variability: the Pacific Decadal Oscillation.

The IPCC has simply assumed that mechanisms of climate change like that addressed here do not exist. But that assumption is quite arbitrary and, as shown here, very likely wrong. My use of only PDO-forced variations in the Earth’s radiative energy budget to explain three-quarters of the global warming trend is no less ‘biased’ than the IPCC’s use of carbon dioxide to explain global warming without accounting for natural climate variability. If any IPCC scientists would like to dispute that claim, please e-mail me at roy.spencer (at) nsstc.uah.edu. (two months later, as of late December, 2008, I’ve still not received a response.)

It should be noted that the entire modern satellite era started in 1979, just 2 years after the PDO switched to its positive phase during the ‘Great Climate Shift’ of 1977. Thus, our satellite data records are necessarily biased toward conditions existing during the positive phase of the PDO, and might not correspond to ‘normal’ climate conditions. Indeed there might not be any such thing as ‘normal’ climate conditions.

About figure 4, Spencer freely admits that for those resutls he only cherry-picked the results that seemed to fit the temperature model. In other words, the PDO and PDO + CO2 lines placed over the temperature data may bear no direct relationship to each other out of the thousands of runs he made using his unique model.

It's interesting that he freely admits that CO2 is required to obtain the best fit:

"If I include the CO2 and other forcings during the 20th Century complied by James Hansen with the PDO-forced cloud changes (solid line labeled PDO+CO2), then the fit to observed temperatures is even closer."

The heat transfer he's talking about is +/- 2 per square meter, or a peak-to-trough change of 4 watts per square meter!

"Now, the average PDO forcing that was required by the model for the two curves in Fig. 4 ranged from 1.7 to 2.0 Watts per square meter per PDO index value. In other words, for each unit of the PDO index, 1.7 to 2.0 Watts per square meter of extra heating was required during the positive phase of the PDO, and that much cooling during the negative phase of the PDO."

It seems like a very promising line of inquiry. I hope he's able to get more peer reviewers to consider his paper for publication, which might enable the topics of PDO influence and cloud cover to get more attention. The PDO and cloud cover may or may not pan out to be a mitigating factor, but more knowledge is better, correct?

Wouldn't you agree that the possibility of discovering that climate is not as sensitive to CO2 as is currently feared is the best possible outcome that we could hope for?

Given the noise present in both recent (125 year) temperatures and historical temperature estimates, it seems likely that there are sources of heat transfer that we don't fully understand yet, and the reservoir of heat transported in the oceans is a logical place to look for just such an influence.

As Mr. Spencer points out, the PDO model does not disprove an influence from CO2. We simply might have a little more time to work on fossil fuel replacement if the sensitivity of climate to CO2 is lower.

We'll know more as the PDO switches polarity and we observe how global temperatures fare over the next 10 to 30 years.

Global warming is clearly not real