August 4, 2012 at 12:06 pm
The big picture. Hardly any change in global ice. 23.5 MM km in 2000, about the same now.
The link is to this graph:
The thick line is a 13-month running mean, and note that the figure is titled “Global sea ice area NSIDC.”
The problem with calling this the “big picture” is that it’s the small picture. And a misleading one at that.
Minor issue: this graph is not global sea ice area — it’s global sea ice extent. Major issue: the 13-month running mean, intended to remove the annual cycle from the data (better to isolate the trend from the seasonal variation), fails to do so — part of the annual cycle remains. Great big giant issue-zilla: NSIDC data begin in late 1978 — so why does it start in 2000?
Answer: because that’s the only way to make it look like global sea ice isn’t shrinking fast.
Here’s my version of the same graph (compare to the one above):
I left out July of this year because the linked graph only goes through June 2012. I’ve also indicated the level of the 1st 13-month running mean with a dashed line.
The 1st 13-month running mean value is indeed 23.5 million km^2, but the final value is not the same, it’s 23.24 million km^2. Hence the last value shown is less than the first, by 0.26 million km^2.
More important is the fact that 13-month running means don’t remove the annual cycle. Here is a “phase diagram” of the 13-month running means, showing the residual annual cycle which they fail to remove:
The first 13-month average is for data from Jan. 2000 through Jan. 2001, centered on July 2000. The last value covers June 2011 through June 2012, centered on December 2011. Since the first value is centered on July it’ll be artificially low (July is near the minimum of the residual annual cycle), but the last value, centered on December, will be artificially high. That will make it look — artificially — like the recent sea ice extent is only barely lower than it was in 2000.
If you really want to remove the annual cycle you should compute 12-month running means [note: you can also compute 13-month running means as long as you give only half weight to the first and last values]. That gives this:
The first 12-month average is 23.82 million km^2, the last is 23.11, less by 0.71 million km^2. That’s 2.7 times as much as indicated when one mistakenly characterizes the trend with 13-month running means. That much sea ice reduction amounts to an area larger than the state of Texas. And that’s just since the year 2000.
If you want a bigger picture, just include more data. Here’s all the sea ice extent data from NSIDC (through July of this year):
Note that the overall decline is plainly visible, even without the 12-month running mean added in. We can also remove the seasonal variation by computing anomaly as the difference between a given value and the average value for the same month during some “baseline” period. I used the entire data set as the baseline, giving these anomalies:
Again the decline is plainly visible. It’s also statistically significant. If we characterize it by a linear trend we get this:
There is evidence (but not conclusive) that the rate of decline of global sea ice extent has increased, as indicated by a smoothed version of the anomaly:
According to the smoothed estimate, the net decline since the NSIDC data begin has been more than twice the area of the state of Texas.
In my opinion, commenter “Smokey” believes he’s justified in calling the graph of a small time span the “big picture” because it includes the Antarctic as well as the Arctic. He utterly fails to note that it’s most certainly the small picture — a mere 12.5 years out of a nearly 34-year data record. I also believe the small time span was deliberately chosen to minimize the obvious visual impact of sea ice loss — i.e., to “hide the decline.” It was further (in my opinion) mischaracterized by a smoothing method (13-month averages) which fails to accomplish its purpose, removing the seasonal variation, further “muddying the waters” by leaving part of that seasonal variation intact. Lastly — in my opinion — the real reason “Smokey” thinks it’s the “big picture” is that is seems to (but doesn’t actually!) support his claim.
Yet another comment actually does try to get at the “big picture” by referring to a post detailing various stories about Arctic sea ice. But the entire thing is anecdotal, not quantitative, and I stronly suspect was deliberately chosen (“cherry-picked”) to conform to a pre-determined narrative. It quite ignores that there are indeed objective, quantitative estimates of Arctic sea ice extent long before the satellite era. And what do those quantitative estimates say? You can read about that here and here.
The only thing actually to be learned from the comments on the WUWT post is this: that those in denial will easily find an excuse to believe what they want to believe.