What will be the Arctic sea ice minimum extent for 2023?
12
641
770
resolved Sep 27
100%99.1%
4.0-4.249 million square km
0.1%
Above 5 million square km
0.1%
4.75-4.999 million square km
0.1%
4.5-4.749 million square km
0.4%
4.25-4.499 million square km
0.1%
Below 4.0 million square km

According to the National Snow and Ice Data Center, the lowest daily value for the Arctic Sea Ice extent as it appears in the 2023 column here (in millions of square kilometers):

https://masie_web.apps.nsidc.org/pub/DATASETS/NOAA/G02135/seaice_analysis/Sea_Ice_Index_Daily_Extent_G02135_v3.0.xlsx

Graph here (5-day trailing average):

https://nsidc.org/arcticseaicenews/charctic-interactive-sea-ice-graph/

Get Ṁ200 play money

🏅 Top traders

#NameTotal profit
1Ṁ297
2Ṁ171
3Ṁ22
4Ṁ14
5Ṁ11
Sort by:

Resolved. 4.23 million square kilometers.

Edit: New daily data from NSIDC was late today:

NSIDC shows a rise:

2023-09-20 4.251

2023-09-21 4.333

ViSHOP showed a rise:

2023-09-20 4.19038

2023-09-21 4.206245

Tomorrow will be last update for me since it seems its on the rise now...

Last update I will do:

NSIDC update:

2023-09-21 4.333

2023-09-22 4.327

ViSHOP:

2023-09-21 4.206245

2023-09-22 4.227554

NSIDC shows a rise:

2023-09-19 4.239

2023-09-20 4.251

ViSHOP shows a rise:

2023-09-19 4.181598

2023-09-20 4.19038

I'll post the historical probabilities for just two more days (when then it's out of the window)

New NSIDC data shows a tiny drop:

2023-09-18 4.241 2023-09-19 4.239

ViSHOP shows a rise:

2023-09-18 4.166193

2023-09-19 4.181598

From the trends still looks very likely 17th will be the minimum (30% without even considering the trends or values)

@parhizj You say 30% (or is that 2/7?) is "very likely", seems a strange description of 30%..

I say 21 of last 25 years do not show large enough drops from 19th to cause a new minimum. ~84% sounds more like "very likely" to me.

Vishop (JAXA) has looked very smooth this year so that showing a rise might be additional evidence that the % should be high like ~84% rather than 30%. OTOH perhaps my 84% is giving too much weight to years that have been showing a rise for several days by 19th Sept.

Only looking at the 7 year out of 44 years that do show a fall to say that a further fall of more than 0.026 is 5/7 ~70% likely seems like very bad methodology to me.

@ChristopherRandles Yes 30 percent even without considering the trends at all does it make it likely. Put another way, intuitively when adding in the trends and that there is only 3 days left makes that 30 percent jump much higher

@parhizj What do you mean by "considering the trends"?

2007 was lower on 24th than it was on 19th. that is more than 3 days but yes it is pretty much over in 3 to 5 days time. But I don't understand at all what you see in that 30% figure or what you mean by "without considering the trends at all".

@ChristopherRandles Basically I just meant what you wrote in your first reply about drops being unlikely. Sorry if I was unclear.

Chartic showing 4.234 suggests a value close to 4.239 (later shown as 4.241) about 0.026 up on yesterdays data minimum.

Only 4 of last 25 years show a subsequent drop of 0.026 (0.028) or more. Is that enough to say there is ~84% chance minimum was 17th? Or is that biased in some way?

Assuming it is later than 18th would seem a strange assumption?

Edit: Both rising....

NSIDC:

2023-09-17 4.213

2023-09-18 4.241

ViSHOP:

2023-09-17 4.141911

2023-09-18 4.166193

My point forecast also shows only rising again (to 4.215)

Historically its still likely the minimum might come later if you conditionalize a certain way, but with just that small bump increase from yesterday it becomes more difficult.

bought Ṁ2 of 4.25-4.499 million s... NO

@parhizj
9/17 4.213

9/18 4.241

That's probably all she wrote. I'll make an arctic Max market soon.

New NSIDC data with another small drop:

2023-09-16 4.224

2023-09-17 4.213

ViSHOP shows a small increase:

2023-09-16 4.13596

2023-09-17 4.141911

bought Ṁ0 of Below 4.0 million sq... NO

My scatter plots do seem as if they may have changed round so that now larger recent drops seem more associated with larger further falls to minimum.

New NSIDC data with another drop:

2023-09-15 4.235

2023-09-16 4.224

ViSHOP also shows a drop:

2023-09-15 4.154314

2023-09-16 4.13596

I don't see the older comments ... My log of my predictions... 😭

@parhizj have you tried sorting different order then going back to newest?

@ChristopherRandles Nope nothing. Looks like the load more comments on scrolling to bottom isn't working anymore (checked on another device too)

Looks like a GUI bug.. the comments are still there it's just not loading them

https://manifold.markets/api/v0/comments?contractId=CYPrZbZhJQKViMRgVVFZ

Reported to discord

@parhizj I'll post here temporarily:

Sept. 16 4.216

Sept. 17 4.210

Sept. 18 4.197

Sept. 19 4.215 *(the minimum for future days past the last observation)

Sept. 20 4.224

Sept. 21 4.229

sold Ṁ11 of 4.25-4.499 million s... YES

@ChristopherRandles You beat me in the race 😭

bought Ṁ1,000 of 4.0-4.249 million sq... YES

@parhizj just hoping it doesn't get revised ;)

@ChristopherRandles Nah.... yesterday they did update the modification time (overwrite the file) but didn't change any of the data.

@parhizj Only yesterday did I figure out the spreadsheet was not masie but the NSIDC product and found the ftp site for it, got set up for downloading files from ftp site and then today chartic graph updated first. lol

@ChristopherRandles Weird. Well I usually have a script to watch for file size changes to all of the markets I'm interested in so I don't have to put too much effort.

More related questions