Predict Summer 2011 Minimum Arctic Sea Ice Extent (millions sq km)

[ORH_wxman]

And because someone takes themselves way too seriously and complains to the mods, we can't have any fun in these threads. Guess I'll have to start complaining every time I want someone's posts deleted. That's skiier's go to now whenever he doesn't like what I'm posting - WHICH DID NOT INCLUDE ANY ATTACKS ON HIS CHARACTER.

Also love the uneven standards in moderation, as tons of conversations WAY more insulting and off-topic have been allowed. But let's reward the whiners instead! I'm a moderator over on the westernwx forum, so I know what it's like from that end.

No, I deleted it because both of you decided to engage in a 1 on 1 off topic discussion for about 20-25 posts. I'll let light hearted stuff go if it doesn't completely derail the thread. This thread was keeping track of average daily loss since August 14th which I thought was pretty useful info.

[ORH_wxman]

Tacoman: 72k

Friv: 8k

Skier: 30k

Through prelim in Aug 18th.

[The_Global_Warmer]

Tacoman: 72k

Friv: 8k

Skier: 30k

Through prelim in Aug 18th.

With the final it would be:

Tacoman: 75K

Friv: 5K

Skier: 33K

with August 19th prelim of -46K

Tacoman: 86K

Friv: 14K

Skier: 34K

If I got that wrong please correct it someone.

[ORH_wxman]

With the final it would be:

Tacoman: 75K

Friv: 5K

Skier: 33K

with August 19th prelim of -46K

Tacoman: 86K

Friv: 14K

Skier: 34K

If I got that wrong please correct it someone.

No, it looks right to me. I meant to update it last night but forgot...thanks.

But this morning it was revised to -38k loss...so its now:

Tacoman: 78k

Friv: 22k

Skier: 26k

[Mallow]

Didn't see this thread 'til now (I don't visit this forum very often anymore ), but just statistically speaking, I would have said about 4.8 +/- 0.7 million sq mi

[tacoman25]

No, it looks right to me. I meant to update it last night but forgot...thanks.

But this morning it was revised to -38k loss...so its now:

Tacoman: 78k

Friv: 22k

Skier: 26k

The slowdown has begun...should really decelerate over the next couple days and then stay considerably lower than the August average to date through at least the 26th.

This is about the time the melt really starts to slow down climatalogically, anyway. I like my chances.

[ORH_wxman]

No, it looks right to me. I meant to update it last night but forgot...thanks.

But this morning it was revised to -38k loss...so its now:

Tacoman: 78k

Friv: 22k

Skier: 26k

Prelim 20th:

Tacoman: +79k

Friv: -41k

Skier: +17k

I added the +/- in to keep better track of what side of the departure the contestant is currently on.

[The_Global_Warmer]

Prelim 20th:

Tacoman: +79k

Friv: -41k

Skier: +17k

I added the +/- in to keep better track of what side of the departure the contestant is currently on.

Skier is definitely going to be in good position for this.

Things are very chaotic to say which way it can go. Last night was revised to 38K loss after a 46K prelim after a day of 77K prelim revised to 80K loss.

Which means Yesterday was averaged into the day before. And was much lower and then revised even lower. Today was lower then yesterdays prelim but today's # was averaged with yesterdays final. Which means yesterdays number must have been much lower than the actual output. So today being averaged with yesterday and being 2K lower than yesterdays means it was probably a bit higher like 50-60k. The number 3 days ago was -56k then -80K then -38K now -36K prelim. basically knowing they use a 2 day average to add continuity to the maps. We had a wild swing where the one day number was massive and the the next day was almost nothing. That means today's Raw Number was probably 50-60K

In the end this helps skier.

If there system is different than this, someone please explain it to me.

[The_Global_Warmer]

Prelim 20th:

Tacoman: +79k

Friv: -41k

Skier: +17k

I added the +/- in to keep better track of what side of the departure the contestant is currently on.

Final was revised to:

5335469

Tacoman: +80k

Friv: -40k

Skier: +16k

No major changes. But a revision even that small probably means todays drop will be bigger based on math unless the arctic has other plans!!

[ORH_wxman]

Final was revised to:

5335469

No major changes. But a revision even that small probably means todays drop will be bigger based on math unless the arctic has other plans!!

skier is +18k with the revision up 1k, not +16k. He and tacoman are positive on their departure, so a revision up hurts them right now while it helps you.

[The_Global_Warmer]

skier is +18k with the revision up 1k, not +16k. He and tacoman are positive on their departure, so a revision up hurts them right now while it helps you.

If he predicted -45K per day wouldn't that take -1K off if his error was -8 instead of -9.

[ORH_wxman]

If he predicted -45K per day wouldn't that take -1K off if his error was -8 instead of -9.

His total error coming in was +26k...with a -9k error for yesterday, it was down to +17k, but it was revised to a -8k error, so its back to +18k. You can't go on absolutely value errors since we are doing a multiple day average error.

[The_Global_Warmer]

August 21st preliminary: -60km2

Tacoman: +105

frivolous: -35

Skier: +33

Ok, I think I got that right.

[ORH_wxman]

August 21st preliminary: -60km2

Tacoman: +105

frivolous: -35

Skier: +33

Ok, I think I got that right.

Tacoman: +106

Friv: -34

Skier: +34

Prelim was -61k right? I think you ran numbers for -60k.

[The_Global_Warmer]

Tacoman: +106

Friv: -34

Skier: +34

Prelim was -61k right? I think you ran numbers for -60k.

Going with 98K for prelim of August 22nd

Tacoman: +204

Friv: +9

Skier: +133

[ORH_wxman]

Going with 98K for prelim of August 22nd

We didn't revise last night's first (went from 61k to 59k)...and then your numbers are still a bit off....I get:

Tacoman: +167k

skier: +85k

Friv: +7k

[The_Global_Warmer]

We didn't revise last night's first (went from 61k to 59k)...and then your numbers are still a bit off....I get:

Tacoman: +167k

skier: +85k

Friv: +7k

Yeah I forgot to take out our averages to todays prelim.

[superjames1992]

Well, I blew my forecast. Not too much more to say about it.

Just a bad forecast on my part based on the incorrect assumption that the Arctic would continue to recover.

[skierinvermont]

yowzers.. what's been going on up there? I been away

[tacoman25]

Well, looks like I was definitely off on this one. The major -DA pattern that was modeled has not developed the same way it appeared it might...the big low pressure center didn't really move over the Beufort as expected and everything has shifted towards Asia somewhat, allowing warmer air to infiltrate on the North American side.

I still think the ice melt will slow as colder air continues to spread over the Asian side over the next few days. A lot of the thinnest ice on the outer edges of the pack has quickly melted over the past couple days, so it seems like the biggest melting should be over, but who knows.

[skierinvermont]

Well, looks like I was definitely off on this one. The major -DA pattern that was modeled has not developed the same way it appeared it might...the big low pressure center didn't really move over the Beufort as expected and everything has shifted towards Asia somewhat, allowing warmer air to infiltrate on the North American side.

I still think the ice melt will slow as colder air continues to spread over the Asian side over the next few days. A lot of the thinnest ice on the outer edges of the pack has quickly melted over the past couple days, so it seems like the biggest melting should be over, but who knows.

You are right that the big freezing beaufort low modeled 10-11 days ago didn't pan out exactly. But the pattern developed pretty much as modeled 3-7 days ago. Just 3 days ago (and for most of the last week) you were saying that the next few days would be extremely slow and we ended up with a 70k and 100k day. This isn't a case of model output being wrong.. it's a case of incorrect human interpretation. Moreover, on a 15 day forecast you should have been basing your reasoning more on the structural factors (record low concentration, record warm SSTs) rather that 10-day numerical model output.

The problem is that the ice is very low concentration and many people are taking an overly rosy view because they are still stuck in a mid-july climo mindset. It's actually continued to be pretty warm relative to late-August average. Combine that with record warm SSTs, extremely low concentration ice and you get above average melt.

I agree that it should start slowing down, if for no other reason than climo and the fact that there is less low concentration ice remaining now.

I believe the above average melt also largely corroborates my point in the discussion with Will regarding the importance of above average 850s and the unusually low concentration of the ice.

[tacoman25]

You are right that the big freezing beaufort low modeled 10-11 days ago didn't pan out exactly. But the pattern developed pretty much as modeled 3-7 days ago. Just 3 days ago (and for most of the last week) you were saying that the next few days would be extremely slow and we ended up with a 70k and 100k day. This isn't a case of model output being wrong.. it's a case of incorrect human interpretation. Moreover, on a 15 day forecast you should have been basing your reasoning more on the structural factors (record low concentration, record warm SSTs) rather that 10-day numerical model output.

I wasn't basing my forecast on 10-11 day model outputs. Look at when I made the forecast, 8 days ago. And I was looking at the pattern starting on the 19th, 4 days ago.

We have all had "incorrect human interpretation" of short term ice trends in response to weather patterns, yourself included. There are a lot of factors, and it is difficult to predict.

Your guess isn't looking too hot now either.

[skierinvermont]

I wasn't basing my forecast on 10-11 day model outputs. Look at when I made the forecast, 8 days ago. And I was looking at the pattern starting on the 19th, 4 days ago.

We have all had "incorrect human interpretation" of short term ice trends in response to weather patterns, yourself included. There are a lot of factors, and it is difficult to predict.

Your guess isn't looking too hot now either.

you made the guess on the 15th and then reiterated expectations of impending slow melt on the 20th. The 7 day model forecast from the 15th verified exceptionally well for the arctic. The models over the last 7-10 days have pretty much nailed the progression of the low from NE of greenland to the northern beaufort/chukchi and then to the asian side.

I'm just saying.. don't blame the models. It's a problem with your interpretation and the general difficulty of forecasting the ice.

[tacoman25]

you made the guess on the 15th and then reiterated expectations of impending slow melt on the 20th. The 7 day model forecast from the 15th verified exceptionally well for the arctic. The models over the last 7-10 days have pretty much nailed the progression of the low from NE of greenland to the northern beaufort/chukchi and then to the asian side.

I'm just saying.. don't blame the models. It's a problem with your interpretation and the general difficulty of forecasting the ice.

Why have you been off then? Are you trying to tell me you expected 160k of melt the last two days 7 days ago? No, you expected "near normal" melt for the forseeable future (as Will pointed out earlier in this thread, normal right now is around 40k/day).

We were both wrong in the short term.

[skierinvermont]

Why have you been off then? Are you trying to tell me you expected 160k of melt the last two days 7 days ago? No, you expected "near normal" melt for the forseeable future (as Will pointed out earlier in this thread, normal right now is around 40k/day).

We were both wrong in the short term.

My forecast of 45k/day is for the entire period and I believe it will be reasonably accurate (I haven't looked too closely at the modeled pattern but my guess is my per/day error will shrink), albeit a little too conservative. The effect of so many variables is difficult to predict and my forecast like all forecasts was no doubt flawed as well. But it is a matter of degrees.

I just saw (and see) little justification for a 15 day forecast of below average melt when the structural factors (very low concentration, high SSTs) pointed towards above average melt. The modeled pattern was only briefly favorable before the low moved more towards the asian side.

[Gray-Wolf]

I think the high SST's and the move into winter storms (no recurve 'canes though!!!) will just wash out a lot of the prone ice now. This is not 1996! we have no more 'old ice' and a lot of thin ice around the edges of the pack. Stick that on a spin cycle and the ice goes. Re-freeze will occur but only in areas of 15%+ so will not show on extent plots. Losses will!

How much of the older ice is now holding station above Fram? How much of that will be there in March 2012?

All the claims of a 'recovery in old ice'? Pah!

[tacoman25]

My forecast of 45k/day is for the entire period and I believe it will be reasonably accurate (I haven't looked too closely at the modeled pattern but my guess is my per/day error will shrink), albeit a little too conservative. The effect of so many variables is difficult to predict and my forecast like all forecasts was no doubt flawed as well. But it is a matter of degrees.

I just saw (and see) little justification for a 15 day forecast of below average melt when the structural factors (very low concentration, high SSTs) pointed towards above average melt. The modeled pattern was only briefly favorable before the low moved more towards the asian side.

At least when I'm wrong I can admit it. We BOTH expected the modeled pattern to result in lower ice losses than what actually occurred over the past 3 days. I have no problem with you saying I "misinterpreted" the modeled result in the short term, but you really should be honest and admit the obvious: you didn't expect it either.

My forecast was for slightly below average melt, yours was for slightly above. You will probably end up closer than me, but we both based our forecast partially on what we expected the pattern August 18-24 to deliver. And we were both off.

[skierinvermont]

At least when I'm wrong I can admit it. We BOTH expected the modeled pattern to result in lower ice losses than what actually occurred over the past 3 days.

My forecast was for slightly below average melt, yours was for slightly above. You will probably end up closer than me, but we both based our forecast partially on what we expected the pattern August 18-24 to deliver. And we were both off.

Can you read? I just said my forecast was probably wrong ("my forecast like all forecasts was flawed.") Although I do expect that the ice melt will not be too high due to the improved concentration which may put me in the ballpark. I've never claimed to have all the answers with regard to ice or to be able to predict it perfectly. But I do believe a certain level of skill which would have ruled out a forecast of 35k/day is possible.

I, unlike you, am not blaming any of my errors on the models. Even if my forecast does end up being correct (which I believe there is still a decent chance of), a lot of that will just be luck. On the other hand, I believe we could easily rule out below average forecasts because the structural factors (very low concentration, high SSTs) pointed to melt. Which is why I have continuously strongly objected to the overly-rosy pictures that have been painted over the last 10 days. I, unlike others, never expected a prolonged period of below average melt because 1) the modeled pattern for the first 7 days was not THAT favorable, more like briefly somewhat favorable, and when combined with the poor condition of the ice would at best produce near "average" melt 2) the weather outside of 7 days was a big unknown but the structural factors (above average SSTs, very low concentration) pointed to above average melt.

You'll also note that I did predict the slowest period would be 3-5 days ago followed by an acceleration the past two days. While you predicted the opposite stating on the 20th that you expected further slowing of melt the next few days. Did I expect a 70k and 100k day? No, and I don't think such specifics are predictable. But I did accurately forecast that the last 2 melt days would be greater than the previous three while you predicted the opposite.

[tacoman25]

It will certainly slow up or perhaps even stop over the coming week.. but that says almost nothing about what will occur after August 23rd.

Oh really?

All I did was come on here and admit I was wrong (without saying anything about you), and you felt the need to jump in and say I was wrong about how I was wrong. And then when I point out you were wrong for the same reasons I was, you still try to act like you had a much better idea of what would happen in the short term than I did.

You're falling off your high horse, cowboy.

[skierinvermont]

Oh really?

All I did was come on here and admit I was wrong (without saying anything about you), and you felt the need to jump in and say I was wrong about how I was wrong. And then when I point out you were wrong for the same reasons I was, you still try to act like you had a much better idea of what would happen in the short term than I did.

You're falling off your high horse, cowboy.

I've already admitted that initially I had an overly-rosy analysis but had corrected that by later in the day on the 15th. Initially I was sort of thrown off by how much colder the 850s were than a month ago.. but then I quickly learned that that is climo for late August and that melt can still occur with widespread -5C 850s.

You'll see that on the 18th I was predicting 30-40k losses on the 19th and 20th (37k and 38k actual) and that by the 21st and 22nd losses would accelerate (70k and 100k actual).

I agree your forecast and my first read on this was wrong for the same reasons. But at first you blamed the models.. which actually performed quite well. That's what I objected to. We were wrong because of the low concentration ice, high SSTs, and because the pattern was not actually that favorable. I figured this out by the 16th while you were still predicting impending low melt on the 20th. My only objection is that you initially tried to blame the models, which detracts from the learning experience that is possible here. I believe this is a valuable demonstration of late-August climatology (IE melt still occurs even with -5C 850s) and of the importance of low-concentration ice.