Mail 681 Friday July 1, 2011 – 2
Global Mean Temperature Anomalies
Message Body:
Dr. Pournelle:
I saw where you stated that you haven’t found a good reference for the method of how they make the Global Mean Temperature Anomaly. The first thing to realize is that there is no one method used. The three major surface air temperature datasets; National Climatic Data Center (NCDC), NASA GISS (GISS) and the Hadley/CRU (HadCRUT) all use different methods. However there is some basic similarities across them. For one they all divide the globe into 5° x 5° grid squares. Second they find an average for each grid square. Since the grid squares closet to the poles are smaller area wise compared to the ones closer to the equator when combining the grid they use a weighting system so that the larger squares do not become statistically dominant in the resulting Global Temperature Anomaly. To see how each dataset takes the quality controlled raw data and makes the grid averages you need to read the papers they are based on.
The easiest one IMHO to find and read is the NASA GISS documentation here:
http://data.giss.nasa.gov/gistemp/
It is based on the series of papers by Dr. Hansen and you will find the links to those papers at that webpage.
For HadCRUT you can find the names of the papers they are based:
http://www.cru.uea.ac.uk/cru/data/temperature/#filfor
The major papers to review are Jones et al 1999 (no link provided) and Brohan et al 2006 (link provided to PDF)
For NCDC there is their analysis here:
http://www.ncdc.noaa.gov/cmb-faq/anomalies.php#grid
However the reference section hasn’t been updated to the new version that the analysis is based on:
Menne and Williams 2009 ftp://ftp.ncdc.noaa.gov/pub/data/ushcn/v2/monthly/menne-williams2009.pdf
and Menne et al 2009 : ftp://ftp.ncdc.noaa.gov/pub/data/ushcn/v2/monthly/menne-etal2009.pdf
As to the Satellites they work completely different and I haven’t found yet how they do them exactly. For the University of Alabama at Huntsville (UAH) you would have to contact Dr. Roy Spencer of Dr. John Christy.
For the Remote Sensing Systems (RSS) dataset you would have to check on their website on who to contact: http://www.remss.com/contact_rss/contact_rss.html
Hope this is helpful
Sincerely Robert Allaband
Thank you. I am having trouble following some of the links, but that may be due to where I am. I still have seen no discussion of how these methods are chosen and how and why the weights are assigned to the various cells, nor have I seen any discussion of what methods are used to compensate for the loss of some of the data, such as when all of Siberia didn’t report for a month. By discussion I mean an explanation at the level of “scientifically educated tax payer who is expected to go without some discretionary luxury in order to pay for this” – ie, an explanation of why certain procedures are used, and why the results are often contrary to intuition.
For example, it may be obvious to NASA that last year was the second hottest year ever, but to the outside observer this is more a function of where you measured what, and what weights are given to the various measurements. It was pretty cold in many places. It was cold and wet in others. Surely out of the billions we pay for these models, they could spare a bright graduate student who can explain just what we are paying for? But perhaps I ask for too much.
