PrintAugust 15, 2007
NASA's global warming misinformation needs full retraction
The news blackout on the erroneous NASA temperature data has been partially lifted by the Toronto Star. More than anything their story was driven by home pride in a local man, Steve McIntyre, who single-handedly exposed the vaunted American Space agency's mistakes and errors.
Drudge is linking this story so more pressure is building on the media to lift the blackout.
NASA quietly corrected the data last week. The media blackout has been going on for the better part of 5 days. Expect the blackout to continue for sometime to come. Many reporters and editors are on vacation. Global warming stories written before they left to take advantage of the heat of August have probably been shelved for the time being.
On their return, reporters and editors will have to face a tricky dilemma: how to row-back on their coverage admitting the error and mistake while protecting the reputation of the their biggest global warming cheerleader, NASA's James Hansen. In October 2005 Hansen embarked on yet another publicity blitz breathlessly claiming 2005 was "on track" to surpass 1998 as the warmest year. (In the revised data 2005 doesn't even make the top 10)
To reinforce the point, the WaPo reporter Juliet Eilperin threw in another uncorroborated study claiming the Arctic ice cap is shrinking dramatically and the Gulf of Mexico's temperature was the highest ever, well at least since 1890 and this was causing the hurricanes (think Katrina).
Using the occasion in early December 2005 of an announcement that the U.S. was joining some global warming talk fest, Eilperin repeated the erroneous temperature information along with the previously cited unsubstantiated reports of melting glaciers, rising sea levels and more intense hurricanes to hammer home her views.
A few days later Hansen was spinning again. Using the release of an unscrutinized British report that affirmed Hansen's (now-discredited) temperature data.
Presumably, this was intended to create the false impression that their is broad "international" agreement about the temperature data.
In late January 2006, Hansen was back at it. He must be on Eilperin's speed dial. Repeating for the third time in four months his erroneous claim that 2005 was the warmest year on record suggesting that this amounted to to a "tipping point"
Waxing apocalyptically, Eilperin wrote:
"The debate has been intensifying because Earth is warming much faster than some researchers had predicted. James E. Hansen, who directs NASA's Goddard Institute of Space Studies, last week confirmed that 2005 was the warmest year on record, surpassing 1998. Earth's average temperature has risen nearly 1 degree Fahrenheit over the past 30 years, he noted, and another increase of about 4 degrees over the next century would "imply changes that constitute practically a different planet."
Hansen has admitted he exaggerated in the past. This sure looks like another exaggeration to me.
In June 2006 the Washington Post repeated the error again in a short item column called "Findings."
In June 2006 the Washington Post repeated the error again in a short item column called "Findings."
These stories received wide circulation, appearing in literally thousands of newspapers and television reports. The Toronto Star tries to minimize the error, citing McIntyre's description of his finding as a "a micro-change." I don't want quarrel with McIntyre, but his analysis caused NASA to revise their data down 0.15 degrees Celsius. Prior to the correction of the error, NASA claimed that global temperatures for 2005 were 0.75 degrees Celsius above the average between 1950 and 1980. This is a 20% error, hardly insignificant. In fact, from a statistician's point of view this is a huge mistake. And why was the 1950 - 1980 period used to compare the average? What would the difference be if we used another time period to compare, say 1940-2004? Did NASA use the 30 year period to exaggerate the deviation?
Moreover, the nominal annual deviation of a selected period is not necessarily meaningful. To make any valid inferences, you have to compare the annual deviation to the average deviation. This is crucial because mean absolute deviation is the average absolute deviation from the mean and is a common measure of forecast error in time series analysis, the method NASA is using to analyze their data.
Don't rely on my unlearned judgement as to what is statistically significant. Here is what Eilperin said in October 2005:
"Climate experts say such seemingly small shifts are significant because they involve average readings based on measurements taken at thousands of sites. To put it in perspective, the planet's temperature rose by just 1 to 1.5 degrees Fahrenheit over the past century." [emphasis added]
Reading these stories made my B.S. detector was go crazy. The condescension and cocksure attitude of the "experts" is astounding. This error is significant. McIntryre suspects he may find more if NASA stops hiding their data and methods from the public.
Frankly, I don't know how to handle this correction. The error was repeated so widely and so often it is probably impossible to pull it back. Perhaps NASA can make spokespeople available to all major media. The best way is probably having the President do a national press conference with the NASA Administrator to correct the error. In the end, he is responsible.
