Tuesday, June 20, 2006
Arctic Ice
Posted by at 11:17 AM
This is actually supposed to go with the previous post, but I haven't quite figured out how to hack blogger to let me put certain elements of this kind into comments (comments are HTML-restricted). Actually, as I just typed it, I figured out a method. But it would take an hour or so to code and I don't want to right now.
Anyway, the point is: This is a nifty little flash file I am leeching from the BBC News web site. There's a slider that lets you move years. Note that the 2002-2004 Arctic Ice levels are dramatically lower than the years before. But also note that the years before are considerably higher than the 1990 levels -- which was a low spike between the relatively high 1988 and 1992.
Without a corresponding data set by which to make a comparative analysis, this means nothing in particular (as far as causality), but it does show a fairly high regular fluctuation of arctic ice annually. If this is tied directly to human manufacturing (etc.), then it looks like we can control at least this phenomenon pretty easily. If not, then it seems that the trend is pretty vast fluctuation regardless of human mfg (etc.).
Anyway, the point is: This is a nifty little flash file I am leeching from the BBC News web site. There's a slider that lets you move years. Note that the 2002-2004 Arctic Ice levels are dramatically lower than the years before. But also note that the years before are considerably higher than the 1990 levels -- which was a low spike between the relatively high 1988 and 1992.
Without a corresponding data set by which to make a comparative analysis, this means nothing in particular (as far as causality), but it does show a fairly high regular fluctuation of arctic ice annually. If this is tied directly to human manufacturing (etc.), then it looks like we can control at least this phenomenon pretty easily. If not, then it seems that the trend is pretty vast fluctuation regardless of human mfg (etc.).
Satellites have made continual observations of Arctic sea ice extent since 1978, recording a general decline throughout that period. Since 2002, satellite records have revealed unusually early onsets of springtime melting in the areas north of Alaska and Siberia. In addition, the 2004-2005 winter season showed a smaller recovery of sea ice extent than any previous winter in the satellite record, and the earliest onset of melt throughout the Arctic.
NOTE: "sea ice extent" is a technical term that refers to waters with a surface covered at least 15% by ice.
Check out this CGI video from NASA.
That is to say, we have been observing it in this manner for λ years. We can detect trends within this time without problem. But this does set us up with a limit on the cycles we can detect. Specifically, the longest cycle we would be able to detect is one that ran in (λ / 2). So since we have been recording it for 26 years, we could detect a 13-year cycle at the longest. We could use ancillary data to speculate about further back (record high and low temperatures, sea levels, etc.), but this is specifically not data-driven. 13 years is an absurdly short expectation for a significant geo-meteorological cycle period. Even record high and low temp cycles are believed to move in about 400-800 year cycles (the best annual data we have goes back only to 1850, but we do have some records and interpreted data that indicate such (C02-cycles, etc).
This leaves us with having to rely upon interpreted data if we want to attempt to detect cycles. In this case, we do this by taking core samples of ice and interpreting their layers, and other such techniques. The problem with these is that there are some observational issues with a number of these (for example, a lost squadron of air force planes landed in the arctic during WWII. An expedition to recover them found them covered with several hundred years worth of ice layers.)
I'm hoping for some interpreted data that doesn't have such issues, but I haven't found any yet (I do some lay research on this topic for personal enrichment for the past few years). Anyway, I'll stop now =) It's very interesting to me.