The Lying Statistics of Climate Change Denial

When it comes to opinions on climate change, people don’t like to take them in moderation. In the course of some very exciting discussions on climate change elsewhere, I got linked to the “Real Science” blog published by fellow WordPresser stevengoddard.  I was shocked to find out that, according to his blog, we’re breaking records for sea ice growth in the north polar region.

http://stevengoddard.wordpress.com/2013/03/14/arctic-blows-away-the-old-record-for-winter-ice-growth/

Included on his blog entry is the following chart:

http://stevengoddard.files.wordpress.com/2013/03/screenhunter_192-mar-14-06-15.jpg

The chart shows what appears to be a trend of increasing growth in arctic sea ice.  Sounds good, right?  That’s what I thought, except it flies in the face of everything I’ve heard about sea ice.  So I got curious, and had to dig into the numbers a bit to satisfy my own curiosity.

According to the link pasted below the chart, it is based on data from the University of Illinois Cryosphere Today website.  If you click on the link you get a page of ASCII formatted data that looks something like this:

1979.0027   0.5883108  12.9048309  12.3165197
1979.0054   0.5693423  12.9448051  12.3754625
1979.0082   0.5900520  13.0047798  12.4147282
1979.0110   0.5933599  13.0592566  12.4658966
1979.0137   0.6045471  13.1295013  12.5249538
....

The data continues like this for thousands of entries, each line a daily record of sea ice extents in the northern hemisphere, logged since 1979 through the present.  The columns are interpreted as follows:

Column 1 – Date, expressed as decimal year.  This has the strange result of whole numbered years meaning December 31 of the preceeding year, i.e. 1980.0000 = December 31, 1979.

The following columns are all related to measurements of sea ice in the northern hemisphere on any particular day and are expressed as millions of square kilometers.

Column 2 – This is the anomaly from the 1979-2008 mean, which is also documented in this chart elsewhere on the site.

Column 3 – This is the actual measured extent of sea ice.

Column 4 – This is the mean extent of sea ice in the northern hemisphere, averaged for the given day of the year across years 1979 through 2008.

I’m a data geek, so when I get a data set I tend to slice and dice it to see what I can do.  My first objective was to duplicate the chart published by stevengoddard.  This involved reducing the dataset to a tabulation of the following calculations for each year:

  • minimum sea ice measurement that year (obviously, this tends to occur in the later part of the year in late Summer or early Fall).
  • maximum sea ice measurement that year (similarly, this occurs in the early part of the year towards the end of Winter or early Spring).
  • growth in sea ice since the preceeding summer minimum – this measure was calculated to duplicate the data used in stevengoddard’s chart, and involved taking the maximum extent from the current year and subtracting the minimum extent from the preceeding year (hence, there is no calculation of “growth” in my chart for 1979 since there is no “minimum” data from 1978).
  • loss of sea ice from the winter maximum to the summer minimum – calculated by subtracting the summer minimum from the winter maximum from earlier that year.

A little monkeying with the chart settings and it looks like I was able to duplicate stevengoddard’s results quite closely.

growth

Good.  So having validated stevengoddard’s analysis, I am comfortable that we are playing from the same sheet of music, so to speak. But still I suspect a bit of statistical chicanery on the part of my fellow blogger if his chart purports to demonstrate a “non warming” effect on the northern polar ice cap. So I only did what was fair, and flipped the parameter. Take a look at a similar chart, based on the same set of raw data, except this time showing the trend in artic sea ice loss during the same period.

loss

WHOA!  Now to understand my reaction when I first saw this chart you have to understand that I’ve been been told that the stevengoddard chart is the ultimate rebuttal against global warming alarmists. So imagine my surprise when I see the above chart that clearly shows we’re also breaking records by the rate of ice we’re losing every year. What’s more, if you compare the two charts, the loss is clearly outpacing the growth. The “record” 10 million square kilometers of growth in north arctic sea icethe climate change denialists are so proud of? Well, it follows a record melt of 11.5 million kilometers. In a record year for sea ice growth, the net effect is to exceed last year’s maximum by a paltry 0.7%. We have roughly 6 months yet to see whether or not the 2013 minimum represents a 23% drop from the prior years’ minimum, like what we saw happen between 2011 and 2012.

So what does this tell us?  I suggest it tells us that looking at annual growth in ice (or loss of ice, for that matter) following summer minimums doesn’t tell us much of anything at all relative to what’s going on with the total, semipermanent sea ice cover; that which has basically been a part of the polar landscape for millinea.  So I present one final chart, with a nice Y axis that begins at zero like all good Y axes should, and shows the trend in both the minimum and maximum extents of northern hemisphere sea ice:

trend

This final chart confirms what is consensus within the scientific community – that there is a dramatic drop in total sea ice occurring, particularly as it pertains to annual minimums.  Despite the relative ease by which one might conjure statistical hackery to convey a misleading picture of climate change, this final graph shows what is really happening – the collapse of the northern ice cap is a geological blink of an eye.

Advertisements
Posted in Science, Uncategorized | Tagged , , | 6 Comments