28 July 2010

Revisiting a sea ice prediction

Regular readers will recall that in late March to early April, there was a fair amount of excitement in some parts of the web about the Arctic sea ice cover having reached almost to climatology.  That was rather exciting given that for much of the last several years the Arctic extent had been from moderately to extremely below climatology.  I wrote up my take in Arctic Sea Ice Updates.  Some of the excitable sources on the web were talking about sea ice recovering and the like. My comment back then (April 7, 2010) was:

So my guess for where we are in the Arctic: The ice formed by late season freezing and conveyor belt is thin.  There has not been time for it to freeze thickly, nor for it to get mechanically piled up to be thick.  The expansive winds that lead to the increase in extent also mean driving the ice towards warmer water.  If the current pattern of blowing the ice out towards the edge were to be sustained, it points to a temporary high value for extent, and then a rapid drop in extent as the ice melts, or as winds reverse and compact the ice pack.


It's now almost 4 months later.  What happened to the ice pack?  Did it continue to hang near climatology?  Go above climatology?  Or did it sink rather rapidly back below climatology, as I'd suggested it would?  The NSIDC report for July 6th notes that June saw the fastest recorded decline in June Arctic sea ice extent, and the lowest June Arctic sea ice extent. 

5 comments:

jg said...

I believe you would have also reported if your prediction had been off, so it's nice to see your estimate come true.

Could you clarify what "close to climatology" means? Does it mean within the 30 year trend?

thanks,
jg

Jesús said...

Your guess was well founded.
Do you have a guess for this year's minimum?

Robert Grumbine said...

jg
Absolutely. The point of making predictions, to me as a scientist, is to look back afterwards and see how they did.

The 'climatology' involved is the mean for 1979-2000. No trends. The idea of including a trend in your climatology is not yet a standard thing. I may start arguing for it in professional settings.

Jesús:
I don't have anything new relative to my Sea Ice Estimations post from June 1. It was a little unsettling to see the ice extend immediately plummet right after that. But in July, the decline was not so extreme, and my estimates for September don't look so unreasonable now. We'll see.

Kate said...

Something interesting regarding sea ice is a recently discovered limitation of the NSIDC satellites. Smaller, separate chunks of multi-year ice, covered by a thin layer of first-year ice, are easily mistaken to be a solid chunk of multi-year ice. It's referred to as "rotten ice" and disappears far more quickly than the satellites lead us to believe it will.

See a recent paper by Barber et al in GRL: http://web.mac.com/barber1818/D.G.Barber/Sea_Ice_Research_files/Barber_etal_GRL'09.pdf

Kate
http://climatesight.org

Robert Grumbine said...

kate:

Thanks for the paper link. I mentioned a Barber radio interview last December, and discussed the remote sensing issue in Fake Ice.

side note: The NSIDC uses the satellites, but they're not NSIDC satellites. I use them too, and for the same reasons. But what we're using are actually US Department of Defense satellites -- the DMSP (Defense Meteorological Satellite Program) satellites, carrying SSMI (for F-13, 14, 15) or SSMI-S (F16, 17) instruments.