(This was originally posted as a comment -- thought I'd promote it to "diary" status)
The animated GIF file can be viewed/downloaded here: http://tinyurl.com/...
Below is some information that explains what is going on in that animation.
First, a bit of background info -- off and on, I've been playing around with the global temperature data that NASA and NOAA use to compute global-average temperature estimates. (I wrote software to compute my own global-average temperature results, implementing a seriously "dumbed down" version of the NOAA gridding/averaging routine.)
I've found that it's amazingly easy to replicate the NASA global-warming trend from that data. I've also found that it's amazingly easy to debunk all of the major denier claims about the global temperature data.
Warming caused by UHI ("Urban Heat Island")? Crunch rural and urban temperature data separately, and you get nearly identical results. So UHI is not responsible for the warming.
Warming the result of data "homogenization"? Raw and adjusted (aka "homogenized") data produce very similar results -- once you've averaged together more than a handful of stations, the data adjustments largely cancel each other out.
Warming due to the "dropped temperature stations" effect? (If you are unfamiliar with this particular claim, just Google up "dropped temperature stations" to see how bogus it is). Watts and other deniers claim that the global warming trend is skewed because only 1500 of the 6000+ stations used by NASA/NOAA are currently reporting data in "real time". Well, I found that you can replicate the NASA/NOAA global warming trends with just a few dozen stations. So going from 6000 stations to 1500 is a non-issue -- 1500 stations is way more than you need, anyway.
More details after the jump.
(Edited to add -- someone correctly pointed out in the comments that I shouldn't expect to convince those who have always denied reality. But my intent isn't to convince deniers so much as it is to try to isolate and quarantine them. To the extent that I can provide sensible folks with "ammo" to help do this, I'll keep giving it my best shots ;) )
After seeing how easy it is for me (an engineer with a "good enough to get by" programming background) to debunk denier claims about the global temperature record, I had been trying to come up with a quickie way to knock down those denier claims in one crack, and in a way that wouldn't leave "non-Big-Bang-Theory types" wondering what I'm going on about.
So after thinking about it a bit, I came up a simple graphic that basically debunks all of the above claims in a single shot.
What I did was to compute global-average temperature results from randomly-selected rural stations (scattered around the globe), starting with one station, and then updating the results one random station at a time. I did this for both raw and "homogenized" data. I plotted up my results (global-avg temps for 1 station, then for two stations averaged together, then for 3 stations averaged together, then for 4, .... up to 40 stations). I then merged the plots into a single animated GIF file.
Stations were pre-screened only on the basis of adequate data record length (i.e. stations with data histories that are too short were excluded from the selection pool). Other than that, station selection was entirely random.
Once again, the file can be viewed/downloaded here: http://tinyurl.com/...
Here's an explanation what the animation shows.
There are two data plots in the animation.
The upper plot shows my global average temperature results computed from raw and "homogenized" data, starting with 1 station and going up to 40. (Each animation frame shows the results updated with data from 1 new random station.)
My own raw results are shown in red. My own "homogenized" results are shown in green. The official NASA results are included for comparison purposes, and are shown in blue.
The lower plot shows how many of the selected stations actually reported data for any given year. For example, in my "40 stations" results, 15 or fewer of the selected stations actually reported data prior to 1900. (Note: Stations that report data for a given year may not have data for all months of that year. I take that into account -- a station that reports data for 6 months out of a year gets counted as "half a station" in my calculations.)
I found this plot to be important because most stations don't have a continuous data record for the entire 1880-present period. This plot is intended to allow folks to "mentally correlate" the number of stations with the noisiness/quality of the global-average results. On average, more stations will give you better warming estimates with less "random jitter".
As you watch the animation, note how quickly the global-warming signal emerges from the data. Note how quickly the raw and homogenized results converge to each other and to the official NASA results.
My animation debunks the denier UHI claims because I use only rural stations. It also debunks the claims about data "homogenization" because I show the results for both raw and homogenized data. And it debunks the "dropped stations" claim because I use at most only a few dozen stations.
I should emphasize that the results were not "cherry picked" in any way. I didn't go through this procedure a bunch of times in order to cherry pick the best results. This really is the result of my first attempt to do this. I did this just once and then uploaded what I got.
Please feel free to share the animated graphic with co-workers/friends/family who have bought into denier claims about the global-temperature results published by climate scientists.