Searching the comments of the Archibald thread from 2009 ,,,,found this...
**********************************EXCERPTR*********************************
E.M.Smith says:
Is the ap index still dropping like this ‘a bad thing’, a ‘really bad thing’ or a ‘gee, intersting…’ thing? Just eyeballing the graph and with the statement that it’s a proxy for solar output, I feel like buying some longjohns…
Then there is this head scratcher:
Well I’ll be. James Lovelock, the greens green and creator of Gaia mythology, agrees that carbon trading is a waste of time! I’ve softened a couple of his words a bit (my edits are in [square brackets] in the quote) to save the moderator a ‘snip’…
From:
http://www.newscientist.com/article/mg20126921.500-one-last-chance-to-save-mankind.html
Not a hope in [heck]. Most of the “green” stuff is verging on a gigantic scam. Carbon trading, with its huge government subsidies, is just what finance and industry wanted. It’s not going to do a [darn] thing about climate change, but it’ll make a lot of money for a lot of people and postpone the moment of reckoning. I am not against renewable energy, but to spoil all the decent countryside in the UK with wind farms is driving me mad. It’s absolutely unnecessary, and it takes 2500 square kilometres to produce a gigawatt – that’s an awful lot of countryside.
NOAA & GISS:
And finally, I’ve done a first pass through the NOAA data and the GISS code. I’m still figuring out what it all means (table of variables with description? You’ve got to be dreaming. Comments? OK, how about one cryptic one per program?) At this point though, my ‘first blush’ is that NOAA has the false precision problem. They hand over ‘monthly mean’ data in 1/100 degree C precision. I don’t see how that is even remotely possible.
It also looks (per the terse readme) like GISS uses the UHI unadjusted NOAA data set rather than the adjusted one (though it is a manual download – easy ftp! – so anyone could use any dataset at the time of running the code. In the ‘readme’ the GHCN and HCN station description files have the .Z ending confounded. The readme for one said to use it (when it was missing) the other said not (when it was there). Hope this isn’t a trend.
Finally, it looks like all GISS does is glue together the HCN, GHCN, and antarctic data (plus some small bits) with some removal of dups and ‘preening’ then does the magic UHI homogenization dance, and some final formatting/cleaning. So that would lead me to believe that a simple cross check dataset can be made by taking the NOAA UHI adjusted data directly and doing station to station comparison graphs.
From the GISS Readme:
GISS Temperature Analysis
=========================
Sources
——-
GHCN = Global Historical Climate Network (NOAA)
USHCN = US Historical Climate Network (NOAA)
SCAR = Scientific Committee on Arctic Research
Basic data set: GHCN – ftp://ftp.ncdc.noaa.gov/pub/data/ghcn/v2
v2.mean.Z (data file)
v2.temperature.inv.Z (station information file)
For US: USHCN – ftp://ftp.ncdc.noaa.gov/pub/data/ushcn
hcn_doe_mean_data.Z
station_inventory
For Antarctica: SCAR – http://www.antarctica.ac.uk/met/READER/surface/stationpt.html
so a simple ftp window on the data location, in your browser, and you can get the UHI adjusted data (though the fahr implies F) and compare to GISS to see what he’s doing. Or just use the same dataset, he uses, not UHI adjusted by anyone…
From the HCN Readme:
urban_max_fahr.Z Urban Heat Adjusted Maximum Monthly Temperature
urban_calc_mean_fahr.Z Urban Heat Adjusted Mean Monthly Temperature
(Calculated from urban.max.Z and urban.min.Z) urban_mean_fahr.Z Urban Heat Adjusted Mean Monthly Temperature
urban_min_fahr.Z Urban Heat Adjusted Minimum Monthly Temperature
For some unknown reason, GISS break the processing down into steps 0,1,2,3,4-5. The start and end are FORTRAN, but step 1 is python (with some C bits to compile and install on your machine). Go figure… The good news is that the PApars.f chunk in Step2 that does the pasteurizing process is the one bit of code that does have decent comments in it.
The code is not particularly complex. It has oodles of undocumented variables and many scratch files, especially between steps, so decoding it will take a bit of work. My estimate is that the code could be shrunk by about 60% with no loss of function.
The Grapj at post #3 is of the Archibald AP Index....
Accuracy and precision are foreign concepts to climate scientists.
If weather stations are only reporting whole degrees, then any average will only be accurate to whole degrees.