Has anyone filed a demand under the Freedom Of Information Act for the adjustment rules, system specifications, parameters and test data / expected results that were used for the system(s) that made these adjustments?

Of special interest would be (#18) “The GISS adjustments consist of data quality control and a homogeneity adjustment applied to urban stations. The data quality control, including comparison of each station with its several nearest neighbors, is the same in the current GISS analysis as described by Hansen et al. [1999]. The urban adjustment is improved in the current GISS analysis. The urban adjustment of Hansen et al. [1999] consisted of a two-legged linear adjustment such that the linear trend of temperature before and after 1950 was the same as the mean trend of rural neighboring stations. In the new GISS analysis the hinge year is a variable chosen to be that which allows the adjusted urban record to fit the mean of its neighbors most precisely.”

Further in #18:
“… uses satellite measurements of nightlights to identify urban areas and remote stations in the United States (and southern Canada and northern Mexico); only unlit stations are used to define homogeneity adjustments.”

Anyone who has driven from a mall parking lot to a suburban home with trees knows that the external temperature drops several degrees.

To discriminate between urban and rural by “lit” and “unlit” probably involves a parameter on the amount of light observed. A poor choice of parameter could well “smear” the urban heat dome into the suburbs.

Further, where is a reference to Dr. Watts work on the quality of the climate monitoring stations that are part of the US Historical Climatological Network (USHCN)?

Lastly, where is the reference to a time series of what is classified as urban, suburban and rural?

As for me, I worked over 40 years in systems, process and data architecture. This included quality assurance (prevention of error) and quality control (elimination of error).

One of the most essential aspects of this work was identification of stakeholders and their agendas.

To find broad changes to data, made by undocumented processes, followed by a statement “As we predicted last year….” gives me the willies.

Surely an audit is necessary. James Hansen is quoted:
“As we predicted last year, 2007 was warmer than 2006, continuing the strong warming trend of the past 30 years that has been confidently attributed to the effect of increasing human-made greenhouse gases,” said James Hansen, director of NASA GISS.”

“It is unlikely that 2008 will be a year with truly exceptional global mean temperature,” said Hansen. “Barring a large volcanic eruption, a record global temperature clearly exceeding that of 2005 can be expected within the next few years, at the time of the next El Nino, because of the background warming trend attributable to continuing increases of greenhouse gases.”

Nasa News Stories Archive
January 16, 2008
2007 WAS TIED AS EARTH’S SECOND WARMEST YEAR
http://earthobservatory.nasa.gov/Newsroom/NasaNews/2008/2008011626028.html

The loss of pre-1958 data (“Gone with The Wind”, like the rest of Port Gibson’s antebellum past) is disappointing but I suppose the data still exists somewhere. The thing that baffles me are the temperature changes in some of the years from 1958-1980 (I circled several examples).

Seems that by now the 60s and 70s would be settled. I don’t see a pattern or anything material and I don’t suspect anything untoward, but I do find it odd.

I checked several other towns nearby and saw changes in those, too, in scattered years in the 60s and 70s.

I guess the science is settled but the temperature is not.

As stated in many earlier posts, the net result of these “adjustments” is to lower the NYC temps in the 1961-1990 period and then raise the temps as the adjusments are reduced post 1990. And guess what, you’ve created a mini hockey stick! This would be so even if the raw annual mean data from 1961-2006 were isothermal. So far, all I see is a blank when it comes to defending this absurd data by NCDC/GISS. Is this stonewalling?

[snip - feeding]

USHCN adjustments:

Areal Edit for outlier and suspect data.

TOBS for time of observation changes.

MMTS for liquid-in-glass thermometer bias.

SHAP for a Karl rendition of a homogeneity adjustment.

FilNET for filling in missing data.

Urban for urban warming.

From the excerpt below I understand that GISS does its own homogeneity adjustments (to urban sites only??) and quality control (like the Areal Edit that USHCN does ??) and applies its own urban warming adjustments (with the night light data).

I need help from Steve M and JerryB in understanding if my view of the data adjustments are correct. Also, Steve M, what does USCHN (and GISS) raw and adjusted imply when you are looking at differences.

If I can better understand the above process, perhaps I can became less confused on what it was the GISS did starting in the year 2000.

The source of the monthly mean station temperatures for the GISS analysis is the Global Historical Climatology Network (GHCN) of Peterson and Vose [1997] and updates, available electronically, from the National Climatic Data Center (NCDC). This is a compilation of 31 data sets, which include data from more than 7200 independent stations. One of the 31 data sets is the U.S. Historical Climatology Network (USHCN), which includes about 1200 stations in the United States. The USHCN [Karl et al., 1990; Easterling et al., 1996a] is composed of stations with nearly complete records in the 20th century and with metadata that aid homogeneity adjustments. The GISS analysis uses the version of the GHCN without homogeneity adjustments, as adjustments are carried out independently in the GISS analysis. The GISS adjustments consist of data quality control and a homogeneity adjustment applied to urban stations. The data quality control, including comparison of each station with its several nearest neighbors, is the same in the current GISS analysis as described by Hansen et al. [1999]. The urban adjustment is improved in the current GISS analysis. The urban adjustment of Hansen et al. [1999] consisted of a two-legged linear adjustment such that the linear trend of temperature before and after 1950 was the same as the mean trend of rural neighboring stations. In the new GISS analysis the hinge year is a variable chosen to be that which allows the adjusted urban record to fit the mean of its neighbors most precisely. The current GISS analysis also uses satellite measurements of nightlights to identify urban areas and remote stations in the United States (and southern Canada and northern Mexico); only “unlit” stations are used to define homogeneity adjustments. For USHCN stations the time-of-observation and station history adjustments of Karl et al. [1990] are applied before the urban adjustment is made.

In a paper which can be found via
http://pubs.giss.nasa.gov/abstracts/2001/Hansen_etal.html
Hansen et al discussed using USHCN adjusted numbers instead
of the raw numbers they had previously used. However, the
USHCN “adjustment” for missing data was not going to be used;
GISS would use its own approach for missing data. It now
seems that a special set of USHCN adjusted data, absent the
missing data adjustment, was prepared for GISS as a one time
service by the USHCN folks. That’s my current guess. YMMV