Ordinarily, I would just ignore this scatter shot mess Penguindreams wrote. However, Jonathan asked me to address Penguindreams’s. It appears Jonathan is worried that if I fail to respond, high school students in Boulder will be deprived of a reasoned rebuttal. I could post this all in your comments, but it’s my policy to typing blog post length responses in comments.

Instead, I’ll post here and assume those students who are interested will find the discussion. I apologize in advance for the meandering nature which is somewhat dictated by the nature of the original criticsm.

Penguindreams seems to dislike my title

Let me begin by illustrating the level of comment included in Penguindreams’s criticism of my recent blog post (which you did not link. This robbed his readers of the opportunity to actually read the post he criticized.)

Here’s the opening:

The title of that article is “IPCC Central Tendency of 2C/century: Still rejected”. Certainly catchy. But let’s look at the substance: Does the IPCC say there’s a central tendency (and if so, what is it that is supposed to have that central tendency) of 2C / century?

I infer from Penguindreams’s words that he thinks my title is not sufficiently substantive. Evidently, the title does not reveal the answer to every question ome might think up? If Penguindreams meant something else, I hope he lets us know. I wouldn’t wish to inaccurately report what he said or meant.

However, if Penguindreams meant what his words appear to say, here’s a news-flash: Titles aren’t full reports. For example, Chicago Tribune once wrote a headline that said “Dewey Defeats Truman!”. It did not say “Dewey Defeats Truman in the hotly contested 1948 election for president of the United States of America, founded in . . . ”

Also, it appears the title of my post lead Penguindreams to ask: “What is it that is supposed to have that central tendency?”

I think the first eight words of the post itself reveal “what”:

“Trends for the global mean surface temperature” are supposed to have that central tendency — according to the WG1 report of the AR4.

Am I supposed to respond to this sort of bizarre substance free nitpicking line by line? Even 6th graders could discern Penguindreams is going on about trivial minutiae.

I’m going to try on items that might be seen as substantive. For the convenience of readers, all chapters and supplementary materials are here.

Now on to meaningful issues:

Did the AR4 project report central tendency for the GMST?

In Penguindreams’s post, he asks his readers:

Does the IPCC say there’s a central tendency (and if so, what is it that is supposed to have that central tendency) of 2C / century?

and answers his own question:

In my reading, which has not been exhaustive, of the two sections, I still found nothing that could be construed as a ‘central tendency’.

“I still found nothing that could be construed as a ‘central tendency’.”?! Wow!

The report by the WG1 for the AR4 is chock full of things that can be construed as central tendencies for a variety of projections. These include discussions of the expected value for the rise in GMST over time, the expected value for precipitation in the future, and the expected rise in sea level. Anyone who knows the definition of “central tendency” should easily find all sorts of things that “could be construed” as a central tendency.

The dark lines on figure 10.4 can certainly be “construed” as central tendencies for the rate of increase in GMST. In fact, they indicate the average of GCM predictions used as the basis for the AR4 projections. Average is the most common use of “central tendencies”, so I should think anyone who knew the definition of “central tendency” could at least construe this as a central tendency for projected trend in GMST.

For that matter, the specific numbers in the table on the bottom of page 763 of chapter 10 or WG1’s report for the AR1 could also “be construed” as central tendencies for the rise in GMST. (Readers can find this figure on page 762 if chapter 10, in the WG1 report of the AR4. )

As it is trivial to show repeated appearances of things that could be construed as “central tendencies” for GMST, it appears, in comments, Penguindreams now tell us “Nor did I say that there’s no central tendency”.

That’s an interesting reply. I guess we could get all Clintonian here and debate what the meaning of “is” is. However, I think many will suggest Penguindreams’s words were chosen to strongly imply that he thought the AR4 does not contain anything that could be construed as a central tendency.

Of course, he may have meant something else.

An alternative explanation might be that he was telling readers he’d skimmed the document quickly, but due to carelessness, haste, or unfamiliarity with the meaning of ‘central tendency’ were unable to find anything that could be construed as a “central tendency”. Then, after skimming carelessly, you fired off a hasty blog post reporting that, as a result of barely looking, you didn’t find what you were seeking.

Is that what he meant? Who knows?

So, on to issues that can be quantified.

What magnitude of central tendency for trend in GMST does the AR4 report?

The central tendency for the AR4 projections for the period one might call “now” is very, very close to 2.C/century. In fact, to one significant figure it is 2C/century.

Here’s how I identified the 2C/century value, using multiple lines of enquiry:

Figure 10.4
The projections for GMST are prominently displayed in Figure 10.4 in chapter 10. This is reproduced regularly at my blog, and my readers are very familiar with it. So, I don’t include the link or reference in every single blog post. However, for your convenience, I recently posted the figure in this post. I posted the page number on the figure– but the figure also reappears on page 79 of the technical summary. So, the IPCC really, really means us to read this figure! 🙂

For the convenience of my readers, I frequently superimpose trend lines on graphic, as I do in the below:

Note that the solid line indicating the average (the most common use of “central tendency” ) is nearly linear during the first two or three decades; 2 C/century goes right through the solid lines for the three SRES.
So, there you have one indication the central tendency is 2C/century.

Table 10.5 in chapter 10 of the WG1 report to the AR4
If your rapid skimming of the report leads you to chapter 10, you will find table 10.5 on page 763 compare the average temperature from 2011-2030 to the average from 1980-1999. As averages these could be “construed” as central tendencies for the rate of increase in GMST.

Because these are averages over two blocks of time, they are not ideally suited for estimating the trend. Slight non-linearities could distort our estimate.

However, the table does indicate the average temperature from 2011-2030 compared to 1980-1999 is expected to rise. For scenario A2, the rise is 0.64 C. If you divide by 3 decades (between the midpoint of the two periods, i.e. 2020-1990), that’s roughly 0.21 C/decade. For scenario A1B, we get 0.69C or 0.23C/decade and for scenario B1, we have 0.66 C which is 0.22 C/century.

So, in fact, we repeatedly find 2C/century. Given various narratives, one must decide if the authors of the IPCC mean 2C/century applies precisely to the first two decades, the first three or some other number of decades. I chose to say “the first few decades”; as the definition of “few” permits the reader to infer some small, indefinite number.

Should Penguindreams wish to quibble over “two”, “three” or few, I will simply point out the precise time frame is not important for the hypothesis test. What matters is that 2C/century appear to apply for the period I am testing, which is 2001-July 2008. As the central tendency shown in figure 10.4 is nearly linear during the first “few” decades, and the test currently focuses on 8 years, there is no practical difference between “one”, “two”, “three” or “few” decades.

But let us looks at the next source for 2C/century!

Page 68 of the technical summary for the WG1
As Penguindreams managed to discover, the authors of the IPCC also mention 2 C/century (the exact number I tested in my blog post) on page 68 of the technical summary. I would think finding the exact number I test at my blog might allay Penguindreams’s suspicions 2C/century was inaccurate.

Evidently not! In this particular paragraph, the authors attributed 2 C/century to the first two decades. Peguindream now jumps on the fact that they included the word “about” and I substituted the word “few” for “two”. In comments he blather about 2C/century possibly being 1.5 C/century or 2.5 C/century,

Sure, in some contexts, “about” might mean a very wide range. However, it happens the authors of the AR4 provide citations to chapter 10 at the end of that paragraph. Chapter 10 contains the figure and table 10.5 mentioned above.

Careful reading of table 10.5 reveals that “about 2 C/century” corresponds to estimates between 2.1C/century to 2.2 C/century for the first three decades.. Inspection of the graph leads us to “about 2 C/century” is about 2C/century for the first few decades. Feel free to add 1.5 C/century or 2.5 C/century lines and see if they line up with the central tendency — meaning the dark line indicating the average.

After drawing your lines, read the narrative near table 10.5 carefully. You will find the authors of the IPCC emphasize that the 0.64C, 0.69C and 0.66C are very similar. . (See table 10.5 in Chapter 10 or the WG1 report for the AR1. It’s on page 763.)

So, in reality, if a reader read the paragraph Penguindreams tells us he read, followed the citations train in the AR4, and read the underlying source, they would discover that when the IPCC authors said “about” 2.0 C/century in the technical summary, they are referring to values like 2.1 and 2.2 C/century for the particular 30 year period, ending in 2030 described above.

They did not mean values that might be as low as 1.5 or as high as 2.5.

Yes, following the practice of the IPCC authors, I wrote 2C/century. Does my rounding down have implications for my hypothesis test? My blog post discusses a hypothesis test: it is also true that if 2 C/century is warming too rapidly to match the recent observations, then both 2.1 C/century and 2.2 C/century are also warming too rapidly. If I did not round, I would find more falsifications!

So, why does Penguindreams “suspect” inaccuracy?
By now, it should be fairly clear to any reader that my blog post actually reports the correct value for the central tendency for the projection over the correct time frame. Yet, Penguindreams seems to suggest the 2C/century might be inaccurate.

Why might it be inaccurate? Evidently, it might be inaccurate because I don’t drop a link to the AR4– a document that can be easily located using Google.

Gosh! Are we being Clintonian again? My reporting might be inaccurate, because I didn’t provide Penguin a link to the AR4 and so it might take him 30 extra seconds to find the AR4 using Google? And then, he might have to read the darn thing to confirm it’s 2C/century?

So, since he hasn’t done all this, 2C/century might be inaccurate?

Is that what he is suggesting? Who knows!

First, I am blogging, not ‘reporting.’ Blogs are ongoing conversations with readers— or at least my blogging is intended that way. For this reason, I can expect readers are aware of previous posts containing links. Moreover, if the readers would like a link, they can ask. You will find this is a widespread practice at blogs. In addition, let me assure you that plenty of people know the IPCC did, indeed, provide a projection with a central tendency of 2C/century!

So, link or no link, my “reporting” the 2C/century for the central tendency of the AR4 projections for the current time is accurate.

Second, there a are many ways to confirm 2C/century the central tendency projected for the GMST in the AR4.

We could calculate the central tendency from the models ourselves!
Should you feel the need to see the average computed over the model runs that have been made available at the climate explorer, I’ve plotted them
here

Note: the trend of about 2C/century applies right now; values of 1.5 C/century or 2.5 c/century would not match the central tendency. Also note, the central tendency is the average over the collection of models. It’s a number. The models also display a standard deviation– but the standard deviation is, by definition a different thing.

We could read the value of the central tendency calculated by Gavin!
Should you feel the need to find Gavin’s interpretation — based on computing the average for model runs. Yes, I read in Penguindreams comments that I should have provided a link when mentioning his post in my comments. Since I did not, readers could have Google the text and found the precise post. You will find Gavin wrote this:

This figure shows the results for the period 2000 to 2007 and for 1995 to 2014 (inclusive) along with a Gaussian fit to the distributions. These two periods were chosen since they correspond with some previous analyses. The mean trend (and mode) in both cases is around 0.2ºC/decade (as has been widely discussed) and there is no significant difference between the trends over the two periods.

So, in other words, Gavin finds a trend of 1.9 C/century averaged over the specific period of 8 years. This is a little lower than the IPCC gets– and Gavin does explain that when you average over the runs you get a slightly different answer than when you average over models. Also, 8 years is different from 20 years (or 30 years, or any specific period one might find mentioned in the IPCC.

So, why don’t I use 1.9 C/century? When testing IPCC projections, I use the information the IPCC authors provide at the level of precision they give. Their document describes a central tendency of 2C/century for the first few decades. They round when reporting this in the Technical Summary.

(If Penguindreams wishes to know if 1.9 C/century falsifies, all he need do is read the tables in the post he criticized and did not link and discover, that, under most statistical tests, it does.)

So, given the wide agreement that the central tendency of the projection is 2C/century, why does Penguindreams blather on about the possible inaccuracy of my saying the IPCC projections have a central tendency of 2C/century? This is truly mystifying

Penguindreams also says vague things about the difference between forecasts and projections.

.
Everyone agrees there can be a difference between projections and forecasts.

However, for the current period of time (2001-2008) there is no practical difference. The emissions trajectories that have occurred are thought to fall well within the range of the SRES. Moreover, in the short term, the projections are is insensitive to a wide range of SRES. So, for the current period, the “projections” are for all practical purposes indistinguishable from forecasts.

If you don’t believe me, here’s what Gavin says in the article I previously mentioned:

The following figure plots the global mean temperature anomaly for 55 individual realizations of the 20th Century and their continuation for the 21st Century following the SRES A1B scenario. For our purposes this scenario is close enough to the actual forcings over recent years for it to be a valid approximation to the simulations up to the present and probable future.

If Pinatubo had erupted in 2002 or if Mars had attacked, then we would be unable to treat the projection as a forecast. Most highschool students will realize that Mars did not attack.

Finally, about the satellite issue
Penguindreams criticizes me for not applying the hypothesis test to global mean surface temperature.

Huh?

It appears inclusion of the satellite data in my comparison has distracted Penguindreams. Careful readers would notice comparison that I also run test with HadCrut, NOAA/NCDC and GISSTemp. Had you provided your readers a link to my post, they likely would have noticed the hypothesis test included these data sets.

Those readers who are familiar with climate science, will be aware that HadCrut, GISS and NOAA are generally considered land based. Their data are thought to provide empirical estimates of GMST (i.e. global mean surface temperature.)

Note also: No matter which statistical method I use, the 2C/century falsifies using the average of the 3 surface based cases, when compared to NOAA/NCDD and when compared to HadCrut. GISSTemp, survives under some of the tests. (It falsifies under others.)

For what it’s worth, in the AR4 the IPCC seems to favor using HadCrut when comparing projections to observations of GMST. Had I adopted their standard, and used only HadCrut, I would conclude the 2C/central tendency falls outside the range consistent with the observations of the earth’s GMST since 2001.

I avoid focusing on HadCrut because it currently shows the most negative trend, leading to the strongest falsification of any tests. In contrast, using UAH — which Penguindreams criticizes — does not falsify under some tests!

Though I am entirely aware the satellites measure over a range of height, I included five separate measurement systems to permit readers to determine whether the diagnosis that the temperature fall outside the bound consistent with data is robust to choice of measurement system. That every measurement system has difficulties is widely admitted here.

Before I leave this section: in addition to complaining I used satellites at all, it appears Penguindreams also complained I use too few satellites. Penguindreams appears to have suggested that I might have chosen to include UAH ad RSS because they give the lowest trends. This guess is wrong: I included them because they are the most widely discussed at various blogs.

If anyone would like me to have a look at the other satellite data sets, feel free to name the data sets and provide links to the data. If the data are freely available, I’d be happy to include them.

About the sun

First: Since, Penguindreams are so insistent on links, do you think maybe he could provide a link to his claim the solar cycle is off its normal 11 year cycle, and that it has been stuck sufficiently long to invalidate the SRES?

One might note that as recently as May, Gavin said:

Note that this is over a period with no volcanoes, and so the variation is predominantly internal (some models have solar cycle variability included which will make a small difference).

(See link above. 🙂 )

So, at least Gavin– who is somewhat familiar with GCM’s– suggest the solar variabilty makes a small difference. He doesn’t suggest any correction for the sun behaving in some way that was not anticipated. But, if you have a theory about that, you are free to propose it.

Second:In my view, the idea that the effect of the sun is predictable, and changes in the solar constant would not affect the projections was an assumption the climate scientists made when creating their projections.

If the projections do not match data because the climate scientists failed to consider the possibility of a slight lengthening in the 11 year solar cycle then their projections would still be falsified. All falsified means is the data do not match the projections.

It appears the projections and the data don’t match.

Presumably, those who notice the data and the model projections are out of whack will seek explanations for the disagreement. If the disconnect is found to arise form incorrect assumptions about solar activity, the climate scientists will presumably improve their models to account for that error.

Oddly, Penguindreams missed the true possible shortcomings

The real possible deficiencies are related to a) whether the statistical model used results in the correct rate of “falsification” and b) whether the spread over the model runs gives a correct rate of “falsification”.

These short comings are discussed in comments here. Unlike the intimation that the IPCC might not have projected 2C/century, both are very real questions.

Conclusion

I apologize to any Boulder high school students that this post is long and scatter shot. Unfortunately, it was difficult to impose structure on such a hodge-podge of separate, unrelated criticisms, many of which simply don’t jibe with the actual contents of my post.

In closing: I think I have dealt with all the issue that are remotely substantive. However, if you have any questions, feel free to ask. This is a blog; comments are open. 🙂

Reposted. Something odd was screwing up comments.

Reposted with old time stamp. Something odd was screwing up comments. It’s still screwing up comments. To comment, go here.