Thursday, August 16, 2012

Probability of New Arctic Ice Record: 72% (Est.)

time series
Arctic sea ice extent as of 8/16/12
Will Arctic sea ice extent reach a record low this year?

It looks like it probably will. It's currently at 5.04 million square-kilometers (Mkm2), while 2007's low was 4.25 Mkm2. From this point in the year, the average remaining melt to minimum over the last decade is 0.92 Mkm2, with a standard deviation of 0.22 Mkm2.

So to break the low, this year only needs to be more than -0.60 standard deviations above the mean. If you assume the annual remaining melts are distributed normally, the probability of that happening is 72%.

Moreover, this probability has been growing over the last few days by about 2 percentage points a day. Like all such calculations, it's moderated by the fact that the Arctic is a place that usually defies simple-minded examinations like this one (except that over time, ice will always melt on a planet that is out of energy balance).

(Update, 8/17 a.m.: Overnight numbers have raised this probability to 82%.)

Why does this matter? Jennifer Francis of Rutgers gave a good talk about it:


Dan Satterfield said...

I was at the meeting where that talk was taped and talked with Dr. Francis several times. It's some pretty amazing research. It is ridiculous to think that the loss of all that ice would NOT have an effect on the seasonal weather when you think about it.

Steve said...

David, Sea Ice Area (SIA)is the first of the dominoes to fall, and that has happened today with Cryosphere Today SIA now 27280 sq kms below the previous record minimum.