{"id":16745,"date":"2011-08-31T21:46:21","date_gmt":"2011-09-01T03:46:21","guid":{"rendered":"http:\/\/rankexploits.com\/musings\/?p=16745"},"modified":"2011-09-01T12:18:29","modified_gmt":"2011-09-01T18:18:29","slug":"nh-ice-update-aug-31","status":"publish","type":"post","link":"https:\/\/rankexploits.com\/musings\/2011\/nh-ice-update-aug-31\/","title":{"rendered":"NH Ice Update: Aug 31."},"content":{"rendered":"

Time for the Wednesday update when you can learn who is on track to win the quatloos. But first, let’s look at this weeks events in the Arctic.<\/p>\n

Extent Loss<\/b>
\nThis week was the second week of rapid extent loss in the arctic. Similar to last week, the 7 day average extent loss rate fell inside the bounds previously observed at this time of year, but it was high. The 7 day average extent loss rate is shown below left with this is shown in black, with this years rate in black. The final “dot” indicating the rate if the most recent daily extent loss is maintained 7 days. Below right is the daily extent loss rate showing only years when JAXA was fully operational. Some daily records were set for JAXA, but this is expected to occur rather frequently as there are 365 days in a year, and many fewer years of JAXA operation:<\/p>\n\n\n\n
\"\"<\/a><\/td>\n\"\"<\/a><\/td>\n<\/tr>\n
Click graphs to enlarge.<\/td>\n<\/tr>\n<\/table>\n

Nevertheless, as I mentioned last week: We don’t need to see extra ordinarily daily or even weekly extent losses to beat the 2007 minimum. All we’ll need to see is sustained losses at the high end seen previously or a melt season on the long end of what has been seen. Both remain possible.
\n<\/p>\n

Current Extent<\/b>
\nDespite higher than averages losses for this time of year, the current extent remains above the minimum for this day set in 2007:
\n\"\"<\/a><\/p>\n

Current Area<\/b>
\nThe qualtoo bet is on extent, but discussion often turns to area. The 7 day smooth area is at a record low for this day of year, but we have not yet reached the absolute minimum. The 7 day smooth areas for various years are shown below left. <\/p>\n\n\n
\"\"<\/a><\/td>\n\"\"<\/a><\/td>\n<\/tr>\n<\/table>\n

The unsmoothed area is shown above right. I focused on 100 days around the current day to let you see how close to setting an all time daily record low we came this week. I thought we’d reach it when the black trace indicating 2011 reached a relative minimum and then rose back up. Coincidentally, this happened at nearly the same time of year in 2007. <\/p>\n

Current Prediction<\/b>
\nAs a result of the rapid loss, I tweaked my prediction method again! Previously, the regressors included ( 7-day smooth extent, 7-day smooth areas, PIOMAS July 31 volume, blah, blah, blah…). I’d liked using smoothed extents because it prevented my forecasts from oscillating too much from day to day.<\/p>\n

However, it seemed to me that as we near the final minimum with so little time available for changes it might be useful to include the most recent extent<\/I> without smoothing. This seemed particularly important as the extent loss is on the high range or anticipated values– using the smoothed extent meant that the recent losses only fully register 4 days after they are first reported. <\/p>\n

So, I added the unsmoothed extent to the regressors; it turned out this gave a superior corrected Akaike coefficient (AICc) for predicting the minimum 7-day smooth extent. (Note: When I add a regressor to the mix, I don’t remove any. The “method” automatically selects the best for the current date using the AICc values. The weighted method lets me do include many regressors with little risk of over fitting.)<\/p>\n

Owing to the fairly rapid recent extent losses, and my decision to switch to a method that fully incorporates the recent extent loss now<\/I> rather than four days from now, best estimate for the minimum 7 day extent declined to 4.4 millions square kilometers. This is above the minimum 7 day extent recorded in 2007. Assuming the errors in my methods are normally distributed the chance of breaking the 2007 extent is estimated at 21%. This estimate says the record is well within the possibility of happening. On the other hand, according to the model, the chance of this years melt exceeding last years is 0.1%. That’s pretty small! <\/p>\n

\"\"<\/a><\/p>\n

(For those wondering: Oddly, when I apply the exact same method for the NSDIC September mean, the AICc criterion indicates that the smoothed extent gives a better prediction of the NSDIC extents. So, there isn’t a big change in forecast for that. Go figure.) <\/p>\n

Histogram<\/b>
\nNow to see who is leading the pack– according to the forecast. If you click to read the text, you’ll see the front runner is BobZ!<\/p>\n

\"\"<\/a>
\n(Note: Current value is smoothed extent.)<\/p>\n

Other things to notice:<\/p>\n

    \n
  1. My bet (in yellow) looks like it’s going to be high. I used my incipient statistical method based on extent only, and that was a high value. Oh. Well. My estimates did drop after CRandals suggested I look at area, and I started hunting down more data to add. But it looks like I was a day short, and I’ll be five quatloos shy! <\/li>\n
  2. The Blackboard mean (purple) is not only high, but outside the \u00c2\u00b195% confidence bounds (blue). I haven’t recomputed taking out the math-enabled ‘bot or suzanne, but they bet, so they count in the mean.<\/li>\n<\/ol>\n

    On the other thread, (I think) Nick Stokes tells me JAXA’s values for today are in. They must have come in while I was writing– and I’m not going to rewrite. Bad timing on my part. If you like, maybe tomorrow I’ll rerun and show a new histogram in comments. Meanwhile those of you who know you are out of the running: you’re in good company. Each week a different person has been the closest contender. No one knows who will win! <\/p>\n","protected":false},"excerpt":{"rendered":"

    Time for the Wednesday update when you can learn who is on track to win the quatloos. But first, let’s look at this weeks events in the Arctic. Extent Loss This week was the second week of rapid extent loss in the arctic. Similar to last week, the 7 day average extent loss rate fell … Continue reading NH Ice Update: Aug 31.<\/span> →<\/span><\/a><\/p>\n","protected":false},"author":2,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[49],"tags":[334],"class_list":["post-16745","post","type-post","status-publish","format-standard","hentry","category-betting","tag-nh-ice"],"_links":{"self":[{"href":"https:\/\/rankexploits.com\/musings\/wp-json\/wp\/v2\/posts\/16745","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/rankexploits.com\/musings\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/rankexploits.com\/musings\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/rankexploits.com\/musings\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/rankexploits.com\/musings\/wp-json\/wp\/v2\/comments?post=16745"}],"version-history":[{"count":0,"href":"https:\/\/rankexploits.com\/musings\/wp-json\/wp\/v2\/posts\/16745\/revisions"}],"wp:attachment":[{"href":"https:\/\/rankexploits.com\/musings\/wp-json\/wp\/v2\/media?parent=16745"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/rankexploits.com\/musings\/wp-json\/wp\/v2\/categories?post=16745"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/rankexploits.com\/musings\/wp-json\/wp\/v2\/tags?post=16745"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}