UPDATE: I forgot to include the closing when I published this. I’ve added it.
This post is a continuation of my post Tamino Once Again Misleads His Followers, which was cross posted at WattsUpWithThat here. There Tamino’s disciples and his other followers, one a post author at SkepticalScience, have generally been repeating their same tired arguments.
The debate is about my short-term, ARGO-era graph of NODC Ocean Heat Content (OHC) data versus the GISS climate model projection. This discussion is nothing new. It began in with Tamino’s unjustified May 9, 2011 post here about my simple graph. My May 13, 2011 reply to Tamino is here, and it was cross posted at WUWT on the same day here. Lucia Liljegren of The Blackboard added to the discussion here.
A graph that’s similar to the one Tamino and his disciples think is fake is shown in Figure 1. It’s similar but different, sort of a short-term OHC model-data comparison Modoki. We’ll get back to it.
First, let’s discuss…
…THE BLATANTLY OBVIOUS ERROR IN TAMINO’S RECENT FAILED CRITIQUE
Tamino’s recent failed critique is titled Fake Predictions for Fake Skeptics. Some people have noted that a fake climate skeptic would be an anthropogenic global warming proponent, but we’ll move on from the implications of that.
Tamino linked to the NODC Ocean Heat Content webpage as the source of his data. There, the NODC provides a spreadsheet of their global OHC data (here). Tamino presents a few preliminary graphs and observes:
Note that there’s a decided hot fluctuation in 2003. So we’ll “predict” the time span 2003 to the present, based on data from 1993 to 2003.
His next graph (my Figure 2) is identified only with the heading of “Ocean heat content 0-700m”. It includes a linear trend that Tamino claims is “based on data from 1993 to 2003”. The length of the trend line is assumedly based on the same period as data. But the period of his trend line does not include the “decided hot fluctuation in 2003.”
Figure 2 (Tamino’s 3rdgraph)
Tamino’s next graph, my Figure 3, includes an extension of his linear trend. In addition to the heading, the graph identifies the red trend line as “fit 1993-2003”, but his “fit 1993-2003” still does not include the “decided hot fluctuation in 2003.”
Figure 3 (Tamino’s 4rdgraph)
In the paragraph after that graph, Tamino states:
We can see that observations don’t follow the prediction exactly — of course! The main difference is that during 2003, the observations were hotter than the prediction. For that time span at least, the oceans had more heat than predicted.
He reinforces that the 2003 data is warmer, yet he and his disciples fail to observe that the 2003 data is not included in his “fit 1993-2003”.
The NODC clearly marks their quarterly data in the spreadsheet they provide here. The Global OHC value at 2002-12 is 6.368768 [*10^22 Joules], and the value at 2003-12 is clearly marked 11.6546 [*10^22 Joules]. But the data included in Tamino’s trend does not include the 4thquarter 2003 data at 11.6546 [*10^22 Joules].
If you’re having trouble seeing that, Figure 4 is similar to Tamino’s graph with the hash marks included. It shows the trend based on the period of 1993 to 2002.
And Figure 5 gives you an idea what Tamino’s graph would look like if he had actually included 2003 data in his trend.
Animation 1 compares the two. (You may need to click on it to get the animation to work.)
How could Tamino and his disciples have missed such an obvious mistake? Some of you might even think it wasn’t a mistake on Tamino’s part, and that his disciples purposely overlooked that blatant error. I’m sure you’ll have a few comments about that.
HANSEN ET AL (2005) OFFSETS THE OHC DATA
A recent comment noted that the observations-based dataset in Hansen et al (2005) was not NODC OHC data, that it was the OHC data based on Willis et al (2004). I never stated that I was providing Willis et al (2004) data. My OHC update posts have always been about the NODC dataset.
My Figure 6 is Figure 2 from Hansen et al (2005). Note that the data starts at about zero watt-years/m^2 in 1993. And my Figure 7 is Figure 3 from Willis et al (2004). Note that the Willis et al data starts at -1 * 10^8 Joules/m^2 at 1993. Hansen et al converted the data, which is not in question since I’ve done the same thing, and Hansen et al have offset the data, which I have done also.
Mysteriously, Hansen et al can shift the data without comment from Tamino, but when I do it, it’s interpreted by Tamino and his disciples, and by those from SkepticalScience, as a fake graph.
THE BEST WAY TO COMPARE THE MODELS TO THE OBSERVATIONS-BASED OHC DATA
Obviously, the best way to present the GISS Model-ER projection for Ocean Heat Content would be to use the actual GISS Model-ER data. The RealClimate annual model-data updates here and herepresent the Model-ER data. But the Model-ER OHC simulations are not available in an easy-to-use format like at the KNMI Climate Explorer. If it was available, all of this nonsense about my shifting data, my misrepresenting data, etc., would disappear. Why?
I have stated in comments at WUWT that I would use the ensemble mean of the Model-ER data and the NODC OHC observations for my future model-data comparisons. I’ve also stated I would use the base years of 1955-2010 to avoid the possibility of being accused of cherry-picking the base years.
Why? I presented this in a June 14, 2011post. And that post has been linked to all OHC updates since then.
Figure 8 is a graph from a 2008 presentation by Gavin Schmidt of GISS. It includes the OHC simulations of the Model-ER for the period of 1955 to 2010, which is the model data shown in the RealClimate model-data posts. It also includes the older version of the global NODC OHC data.
1. replicate the ensemble mean data of the GISS Model-ER,
2. replace the older NODC OHC data with the current version, and
3. use the base years of 1955-2010 so that no one can complain about cherry-picked base years,
Figure 9 would be a reasonable facsimile of the long-term comparison from 1955 to 2010. Notice where the ensemble mean of the GISS Model-ER intersects with the data near the ARGO era. Sure looks like 2003 to me. Figure 1 at the top of this post confirms how closely the GISS Model-ER would intersect with the NODC OHC data at 2003.
That graph in Figure 1 looks familiar, doesn’t it? It sure does look like the ARGO-era graphthat Tamino and his disciples dislike so much.
In my January 28, 2012 at 6:18 pmcomment at the WUWT cross post I wrote the following:
I offered in a comment above to use the base years of 1955-2011 for my short-term ARGO-era model-data comparison. That way there can be no claims that I’ve cherry picked the base years or shifted the data inappropriately. I do not have the capability to process the GISS Model-ER OHC hindcast and projection data from the CMIP3 archive. So I cannot create the ensemble member mean of the global data, on a monthly basis, for the period of 1955 to present. But some of you do have that capability. You could end the debate.
If you choose to do so, please make available online for all who wish to use it the Global GISS Model-ER hindcast/projection ensemble member data on a monthly basis from 1955 to present, or as far into the future as you decide.
I will revise my recent OHC update and reuse that model data for future OHC updates. That way we don’t have to go through this every time I use that ARGO-era comparison graph as the initial graph in my OHC updates.
I know the trend of the OHC data is not the model mean, but for those who are wondering what Tamino’s NODC OHC graph might have looked like if he had actually included the 1993 through 2003 data in his trend AND then compared it to the period of 2003 to 2011, refer to Figure 10.
And if he had lopped off the data before 2003, because it isn’t presented in the graph that he complains about so much, the result would look like Figure 11.
Pingback: Tamino Once Again Misleads His Followers | Bob Tisdale – Climate Observations
Not bloody likely !!!
Great work Bob !!
Pingback: The Comment of the Week | Bob Tisdale – Climate Observations
A quick note regarding Tamino plotting a trend “from 1993 to 2003,” I plotted in that exact period, verbatim, at woodfortrees (with SST data). Read the link below.
Trend and data are actually plotted at that site from Jan 1993 to December 2002. It’s a convention used frequently – “to” 2003, means up to but not including. It appears that’s what Tamino meant, and that’s what he did.
I’m guessing that Tamino avoided incorporating any data from your plot (ie, Jan – Dec 2003) when making a trend prediction. So he ran a trend right up to the very last data point before yours begins, and started in 1993 because that was the beginning of the original reference period (Hansen after Willis).
The modelled slope for OHC at 750 meters in Hansen was 0.6 W
m2/yr. Obs based on Wilis 04 gave a trend of 0.6 for the ocean total.
Tamino’s slope (NODC) is 5.2 Wm2/yr. 1993 through 2003 slope is 0.64 Wm2, But these figures are for the upper 700 meters – NODC isn’t exactly the same metric as Willis 2004.
barry: Regardless of a WoodForTrees trend showing a convention among statisticians, there are no climate change-related papers that I’m aware of where the end year did not include all months of data during that year. Further to that, Tamino set the precedent for himself with his Foster (Tamino) and Rahmstorf (2011) paper. The title of the paper is “Global temperature evolution 1979–2010”. The last sentence of their abstract reads, “The adjusted data show warming at very similar rates to the unadjusted data, with smaller probable errors, and the warming rate is steady over the whole time interval. In all adjusted series, the two hottest years are 2009 and 2010.” They had to have included 2010 temperature data in their analysis; otherwise they could make no claims about 2010 temperatures.
Here’s a link to the paper:
Click to access 1748-9326_6_4_044022.pdf
And here’s a link to Tamino’s post about the paper. It contains a link to the data he used in the paper, and it runs through December 2010 (2010.958 on the spreadsheet).
If Tamino intentionally excluded 2003, the question is why, since Hansen et al used the period of 1993 to 2003, not 1993 to 2002. You can attempt to justify it anyway you like, but the outcome was obviously misleading.
I’ll ask Tamino about it at his blog, but it seems pretty straightforward to me that none of the data used for the prediction should coincide with the observational data. You started in Jan 2003, so the trend prediction should use no data past Dec 2002.
If consistency is important then there are all sorts of problems with how you and Tamino have handled the data. You are using a different observational data set to Hansen for starters, with different ocean depth – do you know if the extra 50 meters makes a difference? You are using a statistical model, as is Tamino, where Hansen used a physical one.
No one is comparing apples to apples here.
barry says: “I’ll ask Tamino about it at his blog, but it seems pretty straightforward to me that none of the data used for the prediction should coincide with the observational data. You started in Jan 2003, so the trend prediction should use no data past Dec 2002.”
Do you think I or any of the readers here would believe any answer Tamino might provide? He’d need a multitude of links to references specifically stating that there can be no overlap between the two.
barry says: “You are using a different observational data set to Hansen for starters, with different ocean depth – do you know if the extra 50 meters makes a difference?”
The IPCC has no qualms about comparing 0-700m with 0-750m OHC datasets:
Notice the difference in the 1993-2003 trends between Ishii et al and Willis et al.
Then let me try one more time to get you to answer my criticism. I provide a visual aid.
I’ve computed a linear trend from 1990 to present, and offset the trend line so that it is zeroed to 2000. I want to see how the obs compare to trend after 2000.
The warming since 2000 has well exceeded the overall trend. I conclude that in the last ten years the warming has been well above the estimated trend.
Can you explain how the method for this spurious result is any different from yours.
Barry says: “Then let me try one more time to get you to answer my criticism. I provide a visual aid.
“I’ve computed a linear trend from 1990 to present, and offset the trend line so that it is zeroed to 2000. I want to see how the obs compare to trend after 2000.
Thanks, Barry. You’ve now reached troll stage. And that means you’re likely done here.
You’ve obviously gone to the Tamino School on how to misrepresent the graph from Bob’s post. I don’t provide the offset you want to insert in the graph. And I did not use TLT data. Here’s my graph with a Barry-type offset added:
And here’s the difference between my graph and the way Tamino would likely portray it if he had actually used the Hansen projection and added his usual offset:
I’ve been patient up to now. But I have no patience for troll-like persistence.
Pingback: Part 2 of “On Sallenger et al (2012) – Hotspot of Accelerated Sea Level Rise on the Atlantic Coast of North America” | Bob Tisdale – Climate Observations
Pingback: On Sallenger et al (2012) – Hotspot of Accelerated Sea Level Rise on the Atlantic Coast of North America | Bob Tisdale – Climate Observations