PIOMAS 2020 September Prediction

[Phoenix]

There is no poll thread for PIOMAS. Would be happy if Juan would start one as this is arguably the most important metric.

Last year was around 4.1k km3. Based on CAB volume being in 9th position at 5.31, I would project volume at the minimum as 4.7k km3 +/- 0.3k km3. 

[Phoenix]

Phoenix, this has to be said:

It would do you well with a bit more humility and a bit less confidence. You have latched onto your theory like it's the gospel. I have taken the time to patiently explain some of its shortcomings in the DHACSOO thread, to no avail it seems.

Specifically I have explained the DMI N 80 data is heavily weighted around the pole itself, and is not a true measure of temperatures north of 80. And that the added energy from AGW gets soaked up by the ice and is not showing in temperature readings, this does not mean AGW is irrelevant. And that the data shows Inner Basin volume during the melting season does matter, and the CAB is not the only thing we should care about, due to melt progress, ice mobility and other factors.
I am fine with people expecting crashes and with people expecting recoveries. However I am not fine with your excessive preaching that can intimidate others from posting, others who may dislike confrontation, dislike harsh criticism and feel less sure of their insights and contributions. Be warned I am losing my patience. And the numerous moderator reports I have received say my instincts are justified.

BTW, 2020 could be a recovery year, this wiill not mean your theory was sound.

I am not preaching anything about DACHSOO here, I am trying to steer people in the direction of simple facts.

1) That the CAB should contain ~ 90% of the sea ice at the minimum is based upon facts.
2) That current CAB volume is ~ 700 km3 > 2019 at 5.31 is a fact.
3) That no years are demonstrating material upside performance in CAB volume reductions from day 165 to 210 seems to be a fact. You present 20 years of data on this. The absence of any years demonstrating an exception to this is strong supporting evidence.
4) That 2012 is the only year in the last 20 to demonstrate the necessary post day 210 performance necessary to make up the 700km3 gap in the CAB is a fact.

These facts give rise to very high confidence that 2020 is a rebound year (to use Juan's term).

Why would posting factually supported inference stop any reasonable person from wanting to post here? I can see that these facts might not be welcome to those who want to maintain some "hope" that 2020 will set or approach a record.

If the truth is bad for traffic and we decide to restrict solid rational argument in the service of maintaining user interest, that is nothing to be proud of. Don't create straw man arguments to try and discredit me. Argue the four points raised above and explain the reasonable argument against a rebound outcome.

[oren]

Argue here to your heart's content. Do not clog the melting season thread with repeated long term predictions. Everyone does it on a one time basis, but the repetition and level of confidence do not suit that thread.
And certainly do not post insinuations on traffic and truth.

[Phoenix]

The argument is made and empirically supported. The four points which give rise to high confidence are given above and presently there is an absence of counter argument.

"Insinuation" refers to something bad. There is nothing bad about noting that people will be less interested in the melt season when it becomes clear that a low minimum is highly unlikely.

The idea that we should not strive for a level of understanding which enables one to apply a high confidence interval to projections should be examined. My level of confidence rests in historical data distribution. You don't argue with the underlying logic I present here which supports the confidence, You say that high confidence is a bad thing. I disagree.

This is not DACHSOO. This is about projecting the 2020 minimum which is a function of a) the present ice level and b) the expected variability of CAB volume in the final 3 months of the melt season.

We have data for these things which allows us to create a likely range of outcomes with some level of confidence. Statistics involves confidence. We shouldn't knock confidence.

[pleun]

Phoenix, I have a hypothetical question for you :
What would happen if we were to start the melting season in March with no peripheral sea ice, just the Cab volume.
In your opinion, what would the outcome be in September?

I think we should begin by noting that "melting season" as defined by the agencies or ASIF conforms to the 2D paradigm of ice. The 2D and 3D max do not arrive simultaneously and the core of the Arctic is still gaining volume in Mid-March.

The next question involves whether this hypothetical ice configuration takes place with all other variables being similar to 2020 or are we describing the future where AGW has advanced to arrive at this scenario organically and ocean / surface temperatures are at a much higher baseline than today ?

Your hypothetical scenario is not so hypothetical. It's where we are heading. This thread aims to present the case that the trajectory to your scenario is slower than the linear assumption.

Since your question is off-topic on this thread and I don't want to dilute the message here with a tangential conversation, I would invite you to find a more suitable thread or open a new one.

I sense a lot of evading tactics here.  The question was : what would be the September outcome if there were no peripheral ice in March, all else being equal to last year. I don't ask for a specific area or volume, just a general idea of what the status of the ice would be.

[Phoenix]

Phoenix, I have a hypothetical question for you :
What would happen if we were to start the melting season in March with no peripheral sea ice, just the Cab volume.
In your opinion, what would the outcome be in September?

I think we should begin by noting that "melting season" as defined by the agencies or ASIF conforms to the 2D paradigm of ice. The 2D and 3D max do not arrive simultaneously and the core of the Arctic is still gaining volume in Mid-March.

The next question involves whether this hypothetical ice configuration takes place with all other variables being similar to 2020 or are we describing the future where AGW has advanced to arrive at this scenario organically and ocean / surface temperatures are at a much higher baseline than today ?

Your hypothetical scenario is not so hypothetical. It's where we are heading. This thread aims to present the case that the trajectory to your scenario is slower than the linear assumption.

Since your question is off-topic on this thread and I don't want to dilute the message here with a tangential conversation, I would invite you to find a more suitable thread or open a new one.

I sense a lot of evading tactics here.  The question was : what would be the September outcome if there were no peripheral ice in March, all else being equal to last year. I don't ask for a specific area or volume, just a general idea of what the status of the ice would be.

And what do you propose that I am evading? I'm always curious to understand how's people's imagination's operate. I'll remind you once again that you're of-topic in this thread as well. This thread is related to predicting the PIOMAS 2020 volume.

You may notice that thread's here are organized by topic, not by the identity of the person who you are seeking to discredit.

[oren]

As 2020 is higher in the CAB and relatively low in the Siberian Inner Basin seas, I take it that you believe these seas don't matter to the Sept outcome. I believe otherwise. Ice is mobile and moves quite frequently between them and the CAB. In addition that ice soaks up energy from the air, and reflects back energy from the sun, thus protecting the CAB from that energy. Thus I believe the volume statistic that matters at this stage is the Inner Basin+CAA volume.
Pleun's hypothetical question is directly relevant to this discussion. Claiming that the CAB is the only thing that matters, and that the CAB outcome depends only on the CAB now, is in essence saying the other seas don't matter, including directly connected seas such as the ESS and Laptev. Pleun has taken that to the extreme to help you explain your implied position and/or realize the problem with this position. I would also be interested in hearing your response to this hypothetical question.

[Phoenix]

As 2020 is higher in the CAB and relatively low in the Siberian Inner Basin seas, I take it that you believe these seas don't matter to the Sept outcome. I believe otherwise. Ice is mobile and moves quite frequently between them and the CAB. In addition that ice soaks up energy from the air, and reflects back energy from the sun, thus protecting the CAB from that energy. Thus I believe the volume statistic that matters at this stage is the Inner Basin+CAA volume.
Pleun's hypothetical question is directly relevant to this discussion. Claiming that the CAB is the only thing that matters, and that the CAB outcome depends only on the CAB now, is in essence saying the other seas don't matter, including directly connected seas such as the ESS and Laptev. Pleun has taken that to the extreme to help you explain your implied position and/or realize the problem with this position. I would also be interested in hearing your response to this hypothetical question.

[/b]

I am happy to engage. I have nothing to evade.

But I was first with identifying what I represent as four basic facts. I would first like you to comment as to whether you have any disagreement with those points as being accurate.

To reiterate, they are:

1) Historical data supports an assumption that CAB volume will be ~ 90% of final volume.
2) CAB volume this year is ~ 700K > last year.
3) Review of your PIOMAS CAB volume chart shows no material upside performing years in CAB loss from mid-June to end July although there are a few which were downside outliers. This chart feature almost 20 years of data.
4) The only year which outperformed 2019 by enough to close the 700k CAB gap was 2012.

This historical data is the basis for my confidence. Before, I comment on the interpretation of the data, I think it's best that we establish whether there is any disagreement as to these four points. 

[Juan C. García]

There is no poll thread for PIOMAS. Would be happy if Juan would start one as this is arguably the most important metric.

Done. 
https://forum.arctic-sea-ice.net/index.php/topic,3121.0.html#new

[oren]

1) Historical data supports an assumption that CAB volume will be ~ 90% of final volume.
2) CAB volume this year is ~ 700K > last year.
3) Review of your PIOMAS CAB volume chart shows no material upside performing years in CAB loss from mid-June to end July although there are a few which were downside outliers. This chart features almost 20 15 years of data.
4) The only year which outperformed 2019 by enough to close the 700k CAB gap was 2012.

I haven't checked 1 numerically, but it seems correct.
I note the gap to 2012 is 570k.
4 years have lost more CAB volume than 2019 in the period from day 151 to day 255 (try to find out which). None of them lost enough to close a 700k gap nor a 570k gap (try to find out what gap they could close).

I also note that in the Inner Basin seas Beaufort+Chukchi+ESS+Laptev+Kara 2020 is at record low, with a gap of 910k above 2012, and 192k above 2019. Does that mean anything? I think it does. When the ice runs out in the ESS, melt continues in the adjacent CAB, and doesn't magically stop at the border between them.

Will 2020 break 2012's volume record or 2019's 2nd place? I can't say. Currently the Inner Basin sports relatively little heat, little sun and little export. This way it's hard to break records despite the early preconditioning. Surely if the sun fails to come out in the very near future, with some favorable weather pattern, probability of extreme melt year will strongly recede. But if the weather pattern changes to a favorable one for ice melt and export, I don't think the current volume level and distribution will be a good enough insurance. The season is still early enough for surprises.

[Phoenix]

1) Historical data supports an assumption that CAB volume will be ~ 90% of final volume.
2) CAB volume this year is ~ 700K > last year.
3) Review of your PIOMAS CAB volume chart shows no material upside performing years in CAB loss from mid-June to end July although there are a few which were downside outliers. This chart features almost 20 15 years of data.
4) The only year which outperformed 2019 by enough to close the 700k CAB gap was 2012.

I haven't checked 1 numerically, but it seems correct.
I note the gap to 2012 is 570k.
4 years have lost more CAB volume than 2019 in the period from day 151 to day 255 (try to find out which). None of them lost enough to close a 700k gap nor a 570k gap (try to find out what gap they could close).

I also note that in the Inner Basin seas Beaufort+Chukchi+ESS+Laptev+Kara 2020 is at record low, with a gap of 910k above 2012, and 192k above 2019. Does that mean anything? I think it does. When the ice runs out in the ESS, melt continues in the adjacent CAB, and doesn't magically stop at the border between them.

Will 2020 break 2012's volume record or 2019's 2nd place? I can't say. Currently the Inner Basin sports relatively little heat, little sun and little export. This way it's hard to break records despite the early preconditioning. Surely if the sun fails to come out in the very near future, with some favorable weather pattern, probability of extreme melt year will strongly recede. But if the weather pattern changes to a favorable one for ice melt and export, I don't think the current volume level and distribution will be a good enough insurance. The season is still early enough for surprises.

Thank you for providing some QA on those points Oren. I note that Point #3 isn't specifically addressed as it speaks to consistent results from day 165 - 210 and your response addresses days 151-255.

The relative advanced state of the Siberian and Atlantic Seas is noted as is the slower state of the Pacific Seas. In the absence of a way to quantify these things, it seems like these are offsetting. I do have the question in my mind about how much the ice melting accelerates as a result of the adjacent open water (vs. the ongoing melting process from insolation and heat). I don't think there is math available for that, so I won't trouble you with asking.

I don't know which of the four years exceeded 2019 in volume loss from day 151. The chart is tough to read with so many lines scrunched close together and my eyes are struggling enough as is. I would guess 2012 and 2016 are two of them. Perhaps 2011. I wish the PIOMAS numbers were given as they are with the 2D measures, so I don't have to check by eyeballing. But it's not like I'm a paying customer or anything, so grateful for what I'm getting for free. If you want to share the four years and the maximum variation vs. 2019, that's your choice.

I have raised a question about how we quantify the materiality of potential export in the PIOMAS thread, so I won't raise it again here. It is the potential to assist in answering questions like these which made the observation of consistent performance in days 165-210 seem a bit exciting for me. There is interest in more numbers (vs. Wip's graphs) of Fram export over time, but again, I pay nothing here so I don't have any cause to expect more information.

I'm promised to do my best to answer the hypothetical question of what would happen if we began the season with only CAB ice in March. The answer is that I don't know. In March, this ice might still expand some into the periphery before the contraction begins because March is still freezing weather. Certainly the peripheral ocean surface would warm faster with no ice albedo and attack from all sides, including through CAA. i would wonder what impact this hypothetical has on the prevailing winds. if there is still a gyre with no ice, the lack of buttressing would make the pack infinitely more mobile and perhaps the entire cab ice would rotate toward Fram and be sucked away. I'm not a mind reader so i don't what attribute you and pleun had in mind.

[oren]

Here's some CAB daily volume data for selected days of the year, courtesy of my PIOMAS spreadsheet. Why not open an online spreadsheet, such as google docs, and play around with the data a bit? You can do that with a tablet or phone, no need for a computer.
If this is too difficult, why not take a math paper and do some analysis by hand and calculator?
Wipneus provides the regional file every time it updates. Maybe you can find a way to get it into the online spreadsheet, and become an active contributor of numerical insights. If you so wish.

As to whether your insight holds and it is so extremely difficult to exceed 2012's CAB loss from days 166 to 212 or from days 151 to 255, you may be right, you may be wrong. Time will tell. I personally think export and losses in adjacent seas play a large role, but I could be wrong.
BTW 2019 exceeded the 166 to 212 loss of 2012 by a little.

Losses   Losses   Year   151 End May   166 Mid Jun   196 Mid Jul   212 End Jul   227 Mid Aug   243 End Aug   243 End Aug
166 to 212   151 to 255   Year   151   166   196   212   227   243   255
3.291   4.247   2000   12.657   12.567   10.377   9.276   8.674   8.371   8.41
2.488   3.595   2001   12.421   12.175   10.617   9.687   9.141   8.913   8.826
2.944   4.425   2002   13.151   12.794   10.903   9.85   9.125   8.745   8.726
3.023   4.306   2003   12.596   12.271   10.228   9.248   8.633   8.326   8.29
2.976   3.975   2004   11.442   11.344   9.372   8.368   7.811   7.436   7.467
3.189   4.677   2005   12.064   11.642   9.619   8.453   7.86   7.434   7.387
2.906   3.953   2006   11.123   10.852   8.89   7.946   7.397   7.136   7.17
3.438   5.773   2007   11.375   11.001   8.905   7.563   6.496   5.739   5.602
3.271   5.374   2008   11.525   11.123   8.935   7.852   6.97   6.488   6.151
3.464   5.504   2009   11.535   11.063   8.883   7.599   6.775   6.229   6.031
3.787   5.717   2010   9.826   9.341   6.683   5.554   4.947   4.253   4.109
3.882   6.361   2011   10.337   9.639   6.95   5.757   4.683   4.201   3.976
3.908   6.685   2012   10.115   9.432   6.773   5.524   4.388   3.67   3.43
3.402   5.067   2013   9.755   9.371   7.161   5.969   5.245   4.77   4.688
2.6   4.656   2014   10.772   10.36   8.875   7.76   6.92   6.349   6.116
3.827   6.348   2015   11.676   10.937   8.422   7.11   6.257   5.597   5.328
3.76   6.306   2016   10.321   9.836   7.339   6.076   5.157   4.239   4.015
3.18   5.114   2017   9.268   8.828   6.744   5.648   4.752   4.257   4.154
3.401   5.491   2018   9.97   9.42   7.057   6.019   5.238   4.625   4.479
3.953   6.218   2019   9.98   9.313   6.605   5.36   4.405   3.876   3.762

[Phoenix]

Thank you for sending this.

[Phoenix]

What do I see?

Re: day 166 to 212

There is a real big step up in the system with 2010 and also one in 2007.

Clearly, the numbers do not represent the tight connection with DMI 80N temperatures that I saw in the graph and CAB is much more sensitive during the peak than I inferred from the graph. The root cause of the excitement is gone. This is no long term tight range.

The last 10 years have one dog year, three middling years and six high performers in the 3.8 - 3.9 range. So for sure I would be less confident about excluding the possibility of 2020 having a breakout peak season if we had a step up 10 years ago. But 10 years is still a trend worth considering.

Re: 2020 prediction

Well, 2020 needs to beat 2012 by ~ 250 km3 to catch 2019. I consider this extremely unlikely as 2020 missed the opportunity for making a longer peak season on the front end. Area measures now are the same as two weeks ago in some key areas. Even to match 2012 which was a freak year is very unlikely.

Summary, there is no replacement for numbers. I have more reason to pay attention to the details of the peak season so I can try to determine cause and effect of CAB declines. If there is a daily data dump in the future, then I can try and correlate the daily volume changes to weather and export condition observations. The devil is in the details.

Even a historic dump of recent daily data is useful to get a sense of the volatility.

I have mixed intrapersonal feelings about this episode. I think presenting and discussing volume data should be higher priority than 2D data in terms of understanding what is going on. I'm very happy to see more data, but not so happy with myself for being so high maintenance in order to get it. Not a sustainable model at all 

Perhaps there is a possibility to collaborate going forward with a shared google spreadsheet with daily data when there is a place to put in observations related to weather conditions which attempt to explain the fluctuations in volume loss. You can possibly crowd source the responsibility of recording observations and then link to the data file when it comes in after the fact. You can turn your ASIF user base into an observation team.

Just a thought.....

[Freegrass]

Stop looking at the data Phoenix! Just look at the weather. One big bad storm, and all that data becomes irrelevant... JMHO.

[Tom_Mazanec]

Stop looking at the data Phoenix! Just look at the weather. One big bad storm, and all that data becomes irrelevant... JMHO.

Does AGW make “big bad storms” more likely?

[The Walrus]

Stop looking at the data Phoenix! Just look at the weather. One big bad storm, and all that data becomes irrelevant... JMHO.

Hmmm.  We have had exactly one big, bad storm during the satellite era.  I think the data is a better indicator.

[Freegrass]

Stop looking at the data Phoenix! Just look at the weather. One big bad storm, and all that data becomes irrelevant... JMHO.

Hmmm.  We have had exactly one big, bad storm during the satellite era.  I think the data is a better indicator.

I think the data is useful, but I also think that people can get lost in it if they don't know how to interpret it. That's why I prefer to look at the weather and the ice. But maybe that's just me... I'm not really a data nerd...   

I don't know if AGW makes “big bad storms” more likely Tom. You'd have to look at the data for that...