The right strategy wins the war Gifts, gadgets, weather stations, software and here!\
The Blogosphere
Wednesday, April 23, 2008
As Earth Cools, Data Centers Busy Re-inventing the Past

By Joseph D’Aleo, CCM

Tom Peterson of NCDC in the next Bulletin of the AMS will be co-authoring a paper Study: Global cooling a 1970s myth. In it he tries to downplay the cooling and the coverage and hype it received. Why the efforts to downplay the cooling? Like the Medieval warm period and Little Ice Age, the mid-20th century cold period has been a thorn in the side of alarmists because it implies natural factors at play. And remember a lot of the cooling came during the post World War II boom.

Mann did his best to do away with the Medieval Warm Period and his band of merry men are still trying to find proof it didn’t exist despite the steady stream of peer review papers that suggest it did and was not confined to Europe but was worldwide. CO2Science has done an excellent job collating those studies here while World Climate Report has done due diligence finding evidence not only that the Little Ice Age was real but was global (see latest here).


In 2007, NCDC announced their latest “upgrade” to their USHCN data base that eliminates the Karl (1988) urban adjustment in the first version in favor of a change point algorithm that is supposed to detect changes in siting and urbanization. It is totally unclear how this method which looks for inflection points (sudden changes) in data trends would be able to detect and adjust for a town or city which slowly grows to expand its heat island to the airport often built on the outskirts of a town. Their new algorithms warmed the early 1900s, cooled slightly the middle 1900s and warmed the last decade. The result of this is that the peak in 2000 is higher relative to the peak in 1930 in the initial analysis. The changes made the US data look more like the global data with less of a dip down in the mid century and more recent warming. A little while back, NASA, announced that 2007 was the second warmest year on record tied with 1998 based on alterations to their adjustments (a very regular process it appears based on attempted analyses by Steve McIntyre) and their increased emphasis on polar regions compared to other data sets. Other data centers and the satellite data sets suggested it ranked much lower.

The national centers with their GISS, GHCN and CRU data bases do little to adjust for urbanization and local factors. The IPCC has bought into this despite the fact their have been numerous peer-reviewed papers that suggested contamination by urbanization, land use changes, station dropout (2/3rds of the world’s stations, mostly rural), missing data (again smaller stations), changes in siting, poor siting (a bigger problem than anyone imagined based on the good work of Anthony Watts and his volunteers) may be responsible for 30-50% of the warming of the last century. Tom Peterson’s 2003 work is responsible for most of the data centers downplaying urbanization. This study concluded: “Contrary to generally accepted wisdom, no statistically significant impact of urbanization could be found in annual temperatures.” It along with papers by Hadley’s David Parker in 2004 and 2006, which stated “globally, temperatures over land have risen as much on windy nights as on calm nights, indicating that the observed overall warming is not a consequence of urban development” were highly influential in the IPCC’s attribution analysis and modeling in AR4 and the decisions on how to handle urban data by NCDC (GHCN, USHCN), NASA (GISS) and Hadley CRU).

In this analysis in August 2007 by Steve McIntyre on Climate Audit, he accessed and worked with the data Peterson used in his analysis to show how urban differs markedly from rural data and is undoubtedly contaminating attempts to determine climate trends.

See full size image here

Read much more on this issue including more detail on Steve McIntyre’s analysis, a post by Roger Pielke Sr. on this same issue with a focus on Parker’s work and additional thoughts and comments and charts in this PDF here

Posted on 04/23 at 01:17 AM
(1) TrackbacksPermalink

Page 1 of 1 pages