I put up my "sparse stations" results on my home-town newspaper's on-line message-board, along with attached .kml files that allow people to view the station locations in GoogleEarth. Linky here: http://forums.utsandiego.com/...
The basic computational algorithm is quite simple and straightforward (most of the work implementing it involved basic "programming housekeeping" chores -- checking for missing data, accounting for variable station record temperature record lengths, etc.)
For those who are interested, here's a quick summary of the procedure:
1) For each station with enough data in the 1951-1980 period to compute baseline averages, compute the average temperature for each month. That is -- for, say, station 11235, calculate the 1951-1980 average temperatures for January, then February, March, etc.
Do this for every station.
2) Take those baseline temperatures and subtract them from all of that particular station's data. I.e. subtract the January baseline temp from all that station's Jan temps, the station's Feb baseline from all that station's Feb temps, etc.
Do this for every station. These are the "seasonally adjusted temperature anomalies" for all stations. "Temperature anomaly" is just a fancy $10 term for "difference between the actual temperature and the temperature baseline".
3) For each year, just average all those "seasonally adjusted tempeature anomalies" together. That's it.
Basically, just a straightforward average of raw data from a few dozen rural stations scattered around the world is all you need to confirm NASA/GISS and prove all the deniers (like Anthony Watts and his minions) completely wrong.
UHI contamination? None here -- I used just rural stations. Warming the result of "recently dropped stations" as Anthony Watts has been claiming? I "dropped" 98 percent of the stations going all the way back to 1880 and was still able to confirm the NASA/GISS results.
Let me repeat that -- proving that NASA/GISS is right and deniers are wrong involves nothing more than coding up a basic anomaly-averaging program and running publicly-available raw temperature data through it. This is something that I could teach college-freshman compsci students to do, and something that deniers have been unable to do in all the years that they have spent attacking NASA/GISS.
If you use all of the stations in the GHCN, then the processing's a bit more complicated (but not that much) -- you need to perform "area-weighting" (i.e. gridding and merging of station data) before computing the global averages. But that's for another day. For a small number of stations scattered roughly uniformly around the Earth, the simple averaging procedure described above will do just fine.
Once again, here's the link to the results and Google Earth files: http://forums.utsandiego.com/...
The .kml files allow for very nice visualization of the station locations
One thing nice about the utsandiego.com message-board is that you don't have to register to download message-post attachments. So it's been a good place for me to stash results and files to share with others.
Feel free to snag the plots and .kml files to pass around to skeptical family-members/friends/co-workers.