Support the Arctic Sea Ice Forum and Blog

Author Topic: Slater's thread  (Read 62126 times)

blumenkraft

  • Guest
Re: Slater's thread
« Reply #150 on: June 20, 2020, 03:12:25 PM »
Hey, OTG, welcome to the forum.

That link (https://cires1.colorado.edu/~aslater/SEAICE/this_year_map.png) is updated daily. If you open this link (which happens when you open this thread) it would only see the recent uploaded file at that link.

If you want to keep the original, you have to download it from the Slater site, and then upload it to the forum as an attachment to your post.

Freegrass

  • Nilas ice
  • Posts: 1640
  • None but ourselves can free our minds...
    • View Profile
  • Liked: 522
  • Likes Given: 765
Re: Slater's thread
« Reply #151 on: June 20, 2020, 03:23:00 PM »
Oops... My bad. It is indeed updating.
Now let's pray...

If the science don't fit our beliefs, we pray to God and cuddle up in our own delusional fantasy where everything makes sense again...

blumenkraft

  • Guest
Re: Slater's thread
« Reply #152 on: June 20, 2020, 03:26:18 PM »
Think about it, FG. There is no date in the URL. How would the server know which picture you mean without that? So logically it can only be the most recent one at this URL.

OffTheGrid

  • New ice
  • Posts: 64
    • View Profile
  • Liked: 35
  • Likes Given: 31
Re: Slater's thread
« Reply #153 on: June 22, 2020, 02:08:59 AM »
 Heres the tenth of August as a reference to return to.
Does this model usually return such jagged prediction curve? Seems weird. Can't predict weather 50 days ahead.

Freegrass

  • Nilas ice
  • Posts: 1640
  • None but ourselves can free our minds...
    • View Profile
  • Liked: 522
  • Likes Given: 765
Re: Slater's thread
« Reply #154 on: June 22, 2020, 02:32:01 AM »
I still haven't figured out what the light blue line indicates. What does anomaly forecast mean?
Now let's pray...

If the science don't fit our beliefs, we pray to God and cuddle up in our own delusional fantasy where everything makes sense again...

oren

  • Moderator
  • First-year ice
  • Posts: 6414
    • View Profile
  • Liked: 2386
  • Likes Given: 2034
Re: Slater's thread
« Reply #155 on: June 22, 2020, 08:57:43 AM »
Heres the tenth of August as a reference to return to.
Does this model usually return such jagged prediction curve? Seems weird. Can't predict weather 50 days ahead.
I am almost certain the jagged prediction curve is a result of the abnormally jagged NSIDC area curve at the time the forecast was made. And in general I doubt this model's output.

binntho

  • Nilas ice
  • Posts: 1534
    • View Profile
  • Liked: 544
  • Likes Given: 122
Re: Slater's thread
« Reply #156 on: June 22, 2020, 09:12:47 AM »
And in general I doubt this model's output.

As do I. But in an earlier discussion, it was pointed out that Slater's model apparently has the best track record at predicting the average September extent.
because a thing is eloquently expressed it should not be taken to be as necessarily true
St. Augustine, Confessions V, 6

oren

  • Moderator
  • First-year ice
  • Posts: 6414
    • View Profile
  • Liked: 2386
  • Likes Given: 2034
Re: Slater's thread
« Reply #157 on: June 22, 2020, 10:01:04 AM »
OTG, bear in mind the model does not predict the path of extent, just a point 50 days ahead based on the currently available data. This prediction is generated every day, and you can connect the resulting points using the thick blue line, and it looks like a path, but it's not. Just a collection of predictions made at different times for different times.
Caveat: this was discussed last year and many people said it was a predicted path. Just be aware of the various opinions.

Freegrass

  • Nilas ice
  • Posts: 1640
  • None but ourselves can free our minds...
    • View Profile
  • Liked: 522
  • Likes Given: 765
Re: Slater's thread
« Reply #158 on: July 06, 2020, 10:48:08 AM »
Slater is going mental. 4,05 on August 24.
That red over the CAB is where the ice got "stretched out" as I said it back then.
Now let's pray...

If the science don't fit our beliefs, we pray to God and cuddle up in our own delusional fantasy where everything makes sense again...

Richard Rathbone

  • Grease ice
  • Posts: 831
    • View Profile
  • Liked: 139
  • Likes Given: 11
Re: Slater's thread
« Reply #159 on: July 06, 2020, 11:18:58 AM »
What does anomaly forecast mean?

That if its 100k below the norm for the time of year now, it will be 100k below the norm for the time of year in future.


Freegrass

  • Nilas ice
  • Posts: 1640
  • None but ourselves can free our minds...
    • View Profile
  • Liked: 522
  • Likes Given: 765
Re: Slater's thread
« Reply #160 on: July 12, 2020, 02:10:26 AM »
Is anyone still maintaining the Slater website? It looks like their security certificate needs to get an update, because the site is getting a warning now.

Your connection is not private
This server could not prove that it is cires1.colorado.edu; its security certificate expired 3 days ago.

https://cires1.colorado.edu/~aslater/SEAICE/
Now let's pray...

If the science don't fit our beliefs, we pray to God and cuddle up in our own delusional fantasy where everything makes sense again...

Freegrass

  • Nilas ice
  • Posts: 1640
  • None but ourselves can free our minds...
    • View Profile
  • Liked: 522
  • Likes Given: 765
Re: Slater's thread
« Reply #161 on: July 12, 2020, 02:11:22 AM »
What does anomaly forecast mean?

That if its 100k below the norm for the time of year now, it will be 100k below the norm for the time of year in future.
I kinda get it. It's basically not important, right?
Now let's pray...

If the science don't fit our beliefs, we pray to God and cuddle up in our own delusional fantasy where everything makes sense again...

Richard Rathbone

  • Grease ice
  • Posts: 831
    • View Profile
  • Liked: 139
  • Likes Given: 11
Re: Slater's thread
« Reply #162 on: July 12, 2020, 03:06:53 AM »
What does anomaly forecast mean?

That if its 100k below the norm for the time of year now, it will be 100k below the norm for the time of year in future.
I kinda get it. It's basically not important, right?

Its an indication of how good (or not) the Slater forecast is. The anomaly forecast  is easy to make, so a complex model has to beat it to be considered of any use.

Ktb

  • Frazil ice
  • Posts: 297
    • View Profile
  • Liked: 127
  • Likes Given: 16
Re: Slater's thread
« Reply #163 on: September 17, 2020, 02:27:47 AM »
Slater's model deserves more credit than it gets. Once again.
I have amazing news for you. Man is not alone on this planet. He is part of a community, upon which he depends absolutely.
- Ishmael

binntho

  • Nilas ice
  • Posts: 1534
    • View Profile
  • Liked: 544
  • Likes Given: 122
Re: Slater's thread
« Reply #164 on: September 17, 2020, 06:31:15 AM »
To my eyes it seems to be wildly off ... so not sure what there is to credit.
because a thing is eloquently expressed it should not be taken to be as necessarily true
St. Augustine, Confessions V, 6

Ktb

  • Frazil ice
  • Posts: 297
    • View Profile
  • Liked: 127
  • Likes Given: 16
Re: Slater's thread
« Reply #165 on: September 17, 2020, 07:39:50 AM »
Anomaly forecast is spot on?
I have amazing news for you. Man is not alone on this planet. He is part of a community, upon which he depends absolutely.
- Ishmael

binntho

  • Nilas ice
  • Posts: 1534
    • View Profile
  • Liked: 544
  • Likes Given: 122
Re: Slater's thread
« Reply #166 on: September 17, 2020, 08:48:43 AM »
Anomaly forecast is spot on?
"Spot on" is apt - the red and the light-blue lines only touch that one time! Not a good match in my not so humble opinion.

Besides, what is meant by "Anomaly foreacast" ...
because a thing is eloquently expressed it should not be taken to be as necessarily true
St. Augustine, Confessions V, 6

Ktb

  • Frazil ice
  • Posts: 297
    • View Profile
  • Liked: 127
  • Likes Given: 16
Re: Slater's thread
« Reply #167 on: September 17, 2020, 09:32:22 AM »
Clearly by spot on I meant that their anomaly forecast for minimum was accurate. And there is ample discussion in this thread about what Anomaly forecast means, what the 50 day lead is, and the comparisons between the two.

Peruse at your pleasure.
I have amazing news for you. Man is not alone on this planet. He is part of a community, upon which he depends absolutely.
- Ishmael

binntho

  • Nilas ice
  • Posts: 1534
    • View Profile
  • Liked: 544
  • Likes Given: 122
Re: Slater's thread
« Reply #168 on: September 17, 2020, 09:46:39 AM »
Clearly by spot on I meant that their anomaly forecast for minimum was accurate. And there is ample discussion in this thread about what Anomaly forecast means, what the 50 day lead is, and the comparisons between the two.

Peruse at your pleasure.

So you mean an "explanation" like this one here:

What does anomaly forecast mean?

That if its 100k below the norm for the time of year now, it will be 100k below the norm for the time of year in future.

Which is not very helpful. But of course if nobody understands what Anomaly forecast is, then the lack of understandable explanations become clear. The Slater website makes no attempt at explanation.

As long as nobody can explain what the Anomaly line means, there is no way to claim that the Slater model has been anywhere near being able to make any real predictions this year.

Just look at the graph. The dark blue line jumps all over the place with very little correlation with the red line, predictive valu nil. The light blue line is practically never comes close to the red line except at the very end.  And since nobody knows what the blue line is showing, no conclusions can be drawn from it!

The Slater website has 2016 as "last year", shown below. That year the predictions were actually quite good, notice the close correlation between the red and dark blue lines towards main melting season, while the the light blue line stays well away from both.
because a thing is eloquently expressed it should not be taken to be as necessarily true
St. Augustine, Confessions V, 6

El Cid

  • Nilas ice
  • Posts: 1312
    • View Profile
  • Liked: 567
  • Likes Given: 105
Re: Slater's thread
« Reply #169 on: September 17, 2020, 10:12:41 AM »
The anomaly forecast (light blue line) is NOT the Slater forecast. It is simply there to show what a naive forecast approach would be, ie. looking at the current anomaly and saying that the size of the anomaly would stay the same in the future.

The Slater forecast is the dark blue line.

This year the "naive" approach was spot on, while the "official" Slater model was behind events (overestimating final extent by cca 0,5 m sq km). Also, the final distribution of ice was wrong.

So once again, the light blue line is just a guide to highlight the model's difference from a naive forecast approach.

binntho

  • Nilas ice
  • Posts: 1534
    • View Profile
  • Liked: 544
  • Likes Given: 122
Re: Slater's thread
« Reply #170 on: September 17, 2020, 10:17:50 AM »
The anomaly forecast (light blue line) is NOT the Slater forecast. It is simply there to show what a naive forecast approach would be, ie. looking at the current anomaly and saying that the size of the anomaly would stay the same in the future.
A bit clearer, but still more needed: How is the anomaly calculated (i.e. anomaly of what from what) and is "the future" the same 50 days as in the Slater forecast?
because a thing is eloquently expressed it should not be taken to be as necessarily true
St. Augustine, Confessions V, 6

El Cid

  • Nilas ice
  • Posts: 1312
    • View Profile
  • Liked: 567
  • Likes Given: 105
Re: Slater's thread
« Reply #171 on: September 17, 2020, 01:29:30 PM »
As far as I remember it is like this:

Both forecasts are made 50 days in advance.

The anomaly ("naive") forecast (which is there for reference only!) simply takes a look at the current extent anomaly vs some sort of longer term average and adds that anomaly to the average of T+50, eg. if July 1 ice extent average for the past 10 years is 10 M sqkm and this year it is 9 m sqkm, then the anomaly is -1 M sqkm. 50 days from now is August 19. We take a look at the Aug 19 average, which is let's say 6 M sqkm, so the forecast for Aug 19 (made on July 1) is 6-1= 5 M sqkm.

The official model breaks down the Arctic into many plots and takes a look at each plot's survival probablities based on past years' data and makes its forecast for 50 days from now for each plot. Then the plots are summarized.


binntho

  • Nilas ice
  • Posts: 1534
    • View Profile
  • Liked: 544
  • Likes Given: 122
Re: Slater's thread
« Reply #172 on: September 17, 2020, 01:53:12 PM »
Thanks El Cid. I'll try to remember that til next time! But the "naive" forecast seems to have predicted the minimum pretty well, while Slater's predictions are off the mark by an unusual amount.

Which tells us ... about the Artic and the changes it is undergoing (fill in blank with your thoughts).
because a thing is eloquently expressed it should not be taken to be as necessarily true
St. Augustine, Confessions V, 6