New Posts
  • Hi there guest! Welcome to PoliticalJack.com. Register for free to join our community?

About that "Global Cooling" data....

Arkady

President
Anthony Watts is a TV weatherman with no known science education, but plenty of funding from the right-wing propaganda fund the Heartland Institute, who is famous for running a climate-change denialist blog. Recently he posted something suggesting that recent NOAA data showed a decade-long cooling trend in the US. Since this has been picked up in the conservative press, and thus will inevitably take hold among conservatives, I thought there were a few points to call attention to.

1) This doesn't say much of anything about global warming. The land area of the US is less than 2% of the Earth's surface. Proportionally, trying to figure out what's going on with global temperatures by measuring the land area of the US makes about as much sense as trying to figure out the win/loss record of an NFL team by reading the stats of a single player. It might give you some vague idea, but you're only assessing 2% of the team, and it may or may not be a representative 2%. He could have had a bad couple games, and the team could still be doing very well.

2) In spite of how this was presented in the conservative media, this isn't a case of new, improved data disrupting the previous understanding. The new data is very much in line with the old data. As even Watts admits, "USCRN and USHCN data align nearly perfectly." We already knew what the trend has been over the last decade, and the new "pristine" locations just confirmed what the old method was already showing.

3) The supposed "cooling" trend is just an accident of the time period used. If you start in a particularly warm year, you can use the least-squares method (which Watts uses) to plot a trend-line showing temperature decline. If you pick a cooler year, it'll show temperature increase. Over a period as short as a decade, the moment you start and finish your data set is highly important. Watts includes a link to the data, so you can have Excel do the calculation, to see what I mean. If, like Watts, you start on January 1, 2005, you see a cooling trend, exactly as he says. If you start a littl eearlier, you get a warming trend. If you start a little later, you get a warming trend. The existence of a cooling trend is dependent on a very specific start date. In fact, you need only shift your start date to May 2006 (or any month after that), and you get a warming trend. If you were to cherry-pick your data to show warming, in much the same way as Watts' cherry-picked date show cooling, you could show a very strong warming trend. For example, the trend-line for starting in January 2008 is y=0.0139x-0.359. That slope compares to a negative slope of -0.005x for the data series Watts uses. In other words, that warming trend is almost three times as rapid as the cooling trend he identifies.

This is the problem with getting your understanding of a topic like this from an uncredentialed TV weatherman on the payroll of a conservative think tank, rather than from actual experts. With a guy like Watts, what you'll get is someone who will scan the latest news, looking for isolated data (land temperatures sampling less than 2% of the global surface, over a relatively short period of time) that supports his politically-driven conclusion. With actual experts, you'll get science.
 
Last edited:
Back
Top