Hansen et al 1999, 2001
The principal argument at realclimate and in Hansen's jeremiads against the materiality of the U.S. errors is that the U.S. is only 2% of the earth's surface. Hansen also argued that in 2001, he stated that 1934 was slightly warmer than 1998; so that, even though NASA online figures as of July 2007 showed that 1998 was warmer than 1934, Hansen claimed that he was for 1934 before he was against it, a claim that I do not dispute.
One of the realclimate posters picked up this trivialization of an interest in U.S. history as follows:
Again, if somebody is really interested in improving worldwide data, spending most of their effort chasing 2% USA Lower-48 is very weird. Reasonable, well-informed people can and do disagree about good procedures, but much of this thread seems right out of the Philip Morris GEP or Data Quality Act playbooks, and it does not help good science…
It is interesting to contrast their present view on the immateriality of U.S. temperature history with their position in Hansen et al 1999 and Hansen et al 2001. The two figures are shown on a common scale in the re-plot below (courtesy of digitization by Hans Erren). As you see immediately, in 1999, Hansen reported that 1934 was 0.6 deg C warmer than 1998. Further, there was actually a negative trend in U.S. temperatures since the 1920s. However, this negative trend for the period since the 1920s had been replaced by a flat trend in the 2001 figures. For the period since 1880, the earlier report had a flat trend, the later report a temperature trend of about 0.32 deg C per century (with, as noted above, a flat trend since the 1920s.)
Figure 1. Digitized version of U.S. temperature history from Hansen et al 1999 Figure 6 (left) and Hansen et al 2001 page 22 (right). 1934 and 1998 values highlighted.
While Hansen and Schmidt now decry the immateriality of U.S. temperature history, this was definitely not the position of Hansen et al 2001. The entire purpose of Hansen et al 2001 was to provide a re-statement of U.S. temperature history, eliminating the inconvenient negative trend since the 1920s in the earlier publication. The heavy lifting for this re-statement had been done by co-author Tom Karl at NOAA, whose adjustments for time-of-observation and "station history" led to the re-statement of U.S. results reported by Hansen et al. The first sentence of the abstract stated:
The purpose of the present paper is to document the changes that have been made in the GISS analysis of surface temperature change subsequent to the documentation of Hansen et al. [1999] and to use this new analysis for a closer look at the United States and global temperature change.
In the running text, they observe ( a quote cited by a poster at realclimate):
Although the contiguous U.S. represents only about 2% of the world area, it is important that the analyzed temperature change there be quantitatively accurate for several reasons. Analyses of climate change with global climate models are beginning to try to simulate the patterns of climate change, including the cooling in the southeastern U.S. [Hansen et al., 2000]. Also, perceptions of the reality and significance of greenhouse warming by the public and public officials are influenced by reports of climate change within the United States.
Another reason is, of course, that there are far more temperature histories reaching back to the 1930s from the U.S. than anywhere else (more on this on another occasion). The adjustments did not include allowance for HO-83 thermometers known to have a positive bias on readings for a number of stations in the 1990s.
In any event, a few years ago, Hansen et al thought that a re-statement of U.S. results, amounting only to an upward re-statement of about 0.32 deg C per century was worth publishing. However, it seems that an error amounting to a downward step of 0.15 deg C is immaterial. I'm sure that Jor-El's engineering was more precise than this.
"Error Bars"
Here's another interesting graphic showing the difference between Hansen 2001 and Hansen 1999 (from the digitized versions - I'll try to get original versions to ensure that the comparison here has not been affected by digitization artifacts: however the graphics is a vector graphic and Hans Erren's digitization is probably pretty good.)
There are several very interesting aspects to this graphic. The total increase in the smoothed version is 0.35 deg C: thus whatever trend exists in U.S. temperature since the 1930s exists entirely because of Karl's adjustments, adopted here by Hansen (Karl) et al. Second, aside from this inserted trend, look at the size of the differences for individual years: the 95% range in differences is -0.38 deg C to +0.39 deg C.
If the U.S. temperature history of Hansen et al 1999 is regarded as an estimate of the history in Hansen et al 2001, it is obviously not a very good estimate: despite having histories from over 1200 stations, the 2-sigma confidence was only about plus-minus 0.38 deg C. It's worth pondering what type of error distribution for 1200 stations would give rise to such high error. Any i.i.d. assumption for errors in the 1999 estimates would give a pretty tight error interval. Since the "true" errors were much higher than such estimates would have been, this would show that the actual error distribution in 1999 was tremendously far from i.i.d. Indeed, this serves as rather a pretty illustration of non-i.i.d. errors in a pretty simple situation.
climateaudit.org
------------------------------
Um, Whatever
James Hansen, NASA climate scientist and lead singer in the climate apocalypse choir, responded to his temperature data revisions a week ago:
What we have here is a case of dogged contrarians who present results in ways intended to deceive the public into believing that the changes have greater significance than reality. They aim to make a mountain out of a mole hill. I believe that these people are not stupid, instead they seek to create a brouhaha and muddy the waters in the climate change story. They seem to know exactly what they are doing and believe they can get away with it, because the public does not have the time, inclination, and training to discern what is a significant change with regard to the global warming issue.
The proclamations of the contrarians are a deceit
Um, whatever. Remember, this is the man who had large errors in his data set, used by nearly every climate scientist in the world, for years, and which were only recently discovered by Steven McIntyre (whom Hansen refuses to even name in his letter). These errors persisted for years because Mr. Hansen refuses to allow the software and algorithms he uses to "correct" and adjust the data to be scrutinized by anyone else. He keeps critical methodologies that are paid for by we taxpayers a secret. But it is his critics who are deceitful?
In particular, he is bent out of shape that critics' first presented the new data as a revised ranking of the hottest years rather than as a revised line graph. But it was Hansen and his folks who made a big deal in the press that 1998 was the hottest year in history. It was he that originally went for this sound byte rather than the more meaningful and data-rich graph when communicating with the press. But then he calls foul when his critics mimic his actions? (Oh, and by the way, I showed it both ways).
Hansen has completely ignored the important lessons from this experience, while focusing like a laser on the trivial. I explained in detail why this event mattered, and it was not mainly because of the new numbers. In short, finding this mistake was pure accident -- it was a bit like inferring that the furniture in a house is uncomfortable solely by watching the posture of visitors leaving the house. That's quite an deductive achievement, but how much more would you learn if the homeowners would actually let you in the house to inspect the furniture. Maybe its ugly too.
So why does Hansen feel he should be able to shield himself from scrutiny and keep the details of his database adjustments and aggregation methodology a secret? Because he thinks he is the king. Just read his letter:
The contrarians will be remembered as court jesters. There is no point to joust with court jesters. … Court jesters serve as a distraction, a distraction from usufruct. Usufruct is the matter that the captains wish to deny, the matter that they do not want their children to know about.
Why do we allow this kind of secrecy and spurning of scrutiny in science? Is it tolerated in any other discipline?
Steve McIntyre has his response here. McIntyre still has my favorite comment ever about Hansen and his gang:
While acolytes may call these guys "professionals", the process of data adjustment is really a matter of statistics and even accounting. In these fields, Hansen and Mann are not "professionals" - Mann admitted this to the NAS panel explaining that he was "not a statistician". As someone who has read their works closely, I do not regard any of these people as "professional". Much of their reluctance to provide source code for their methodology arises, in my opinion, because the methods are essentially trivial and they derive a certain satisfaction out of making things appear more complicated than they are, a little like the Wizard of Oz. And like the Wizard of Oz, they are not necessarily bad men, just not very good wizards.
Update: If you have a minute, read Hansen's letter, and then ask yourself: Does this sound like what I would expect of scientific discourse? Does he sound more like a politician or a scientist?
Posted on August 20, 2007 at 02:44 PM | Permalink
Comments
To me the real story is the continued refusal to publicly disclose his methods. Doesn't this by it's very nature render it non-science? How does this secrecy conform with the scientific method???
Posted by: Allen | Aug 20, 2007 3:16:40 PM
coyoteblog.com |
