Some more horrific Perl, but it did the job…
rmw42@pandora:~/NOAA$ cat Average.txt | ./warmest 10 32 stations had 10 of their 10 warmest years post 1997 1360 stations did not have 10 of their 10 warmest years post 1997 11347 stations rejected for having insufficient data
So, I make that 98% of weather stations finding that the warmest ten years in their history were not post-1997. That’s quite shocking, really. I think it’s safe to say that the statement is completely and utterly false – if 98% of weather stations active for the last 24 years don’t show the last 12 as containing the ten warmest, by what measure can we claim they were the warmest years?
I want to test this to see if there’s any pattern to the stations used, whether requiring good data integrity skews the results, and whether rejecting so many stations (~90% of the total) was necessary – but I think the reasons I gave on Friday are sound. It doesn’t matter a damn to me if a station had good data during WW2 – if it hasn’t been active for 60 years, it can’t tell me how warm 2007 was! And surely a station giving only one temperature reading per year – yes, there are some like that – is hopeless?
As Adam and Jamie might say: Myth Busted!
rmw42@pandora:~/NOAA$ cat Average.txt | ./warmest 10 44 stations had 10 of their 10 warmest years post 1997 2460 stations did not have 10 of their 10 warmest years post 1997 13614 stations rejected for having insufficient data