Wednesday, December 30, 2009

Hot in herre

I'm just updating my old chart of annual average temperatures in the United States from 1895-2008.  Yeah, 2008 was relatively cool, but don't lose the steaming forest for the cold tree.

You can make that chart in under five minutes using Stata and downloading the data from the National Climatic Data Center.  They actually have their own graph generator, but the result is shockingly ugly:


I mean, you can sort of detect the trend, and you can see that the last decade was relatively hot, but all those red lines connecting the dots are meaningless, and it really doesn't give you a sense of the magnitude of the shift since the mid-20th century.

6 comments:

Kyle Baker said...

I'mma take off all a my clothes.

Seth Masket said...

Good gracious, carbon dioxide levels are bodacious.

Steve Balboni said...

Al Gore is fat and has a big house thus negating all of your facts and science.

If you want something amusing but that may make your ears bleed check this out,
http://www.youtube.com/watch?v=5sjiDlfB40E

and yes the lead singer is dressed as Cartman...

randomvariable said...

Yo,

Have you got the source code for that? I'm trying to do exactly the same charts for a bunch of weather stations in the UK, but my loess plots are going screwy.

Seth Masket said...

Sure. The two variables were "year" and "tempf."

twoway (scatter tempf year)(lowess tempf year, ylabels(, angle(0)) ytitle("Degrees" "Farenheit", orientation(horizontal)) xtitle(Year) legend(off) xsize(8) ysize(7) scheme(s1color))

Joe said...

It would be more legible if you used "Number of Pirates" to label the data points.