2020 Melting Season Predictions

[PragmaticAntithesis]

IMO, this is a low effort, no value added post. Someone should open a prediction thread for the non-science of guessing future weather events.

Your wish is my command!

Anyway, I still stand by my prediction from the start of melt season that this year's melt will be harsh, but not record-breaking. I'm calling a minimum of around 4.0 square megametres extent.

[Stephan]

I also think that a 4.0 M km² extent in September is likely. This value also derives from gerontocrat's plume graphs in the Sea Ice Extent Data thread. But this melting season probably contains enough surprises. Let's see where we will end up...

[blumenkraft]

I make a prediction about the predictions: No one will guess the real reason(s in correctly weighted combination) for the 2020 outcome.

That said, i would guess we will land around the 2012 numbers. And the reason for that will be the weather.

[Phoenix]

Your wish is my command!

Bless you Aladdin 

I'm not real big on the guessing, but I see barriers getting heat to the areas that did not melt out last year. 2012 was exceptional and while we could get something similar this year,  I wouldn't bet on that.

The diminished aerosol content due to Covid is a wildcard this year, so I'll guess 3.75M km2 at the minimum.

The big unknown for me is when does the Atlantic Ocean breach the 82N line that separates the shallow Atlantic from the area above Nansen Basin. When that happens in a substantial way, I'll be very bearish on the future of the ice. Until then, I bet it holds up.

[The Walrus]

With barely one month in the record books, I will quote Yogi Berra, "it is tough to make a prediction, especially about the future."  That said, this year is starting out similarly to 2007 and 2011.  Those two years had minimum extents of 4.16 and 4.34.  Therefore, I will split the difference and guess 4.25.

[Niall Dollard]

Here is the assessment report for 2019 from the Sea Ice Prediction Network.

 https://www.arcus.org/sipn/sea-ice-outlook/2019/post-season

Our own poster Rob Dekker did very well with his predictions.

Just to note, some sites use different metrics. Here are a few:

1) September Monthly Mean  (NSIDC)

2) September Monthly Mean  (Jaxa)

3) Lowest Daily Min (Jaxa)

4) Lowest Daily Min (NSIDC)

5) Lowest 5 day trailing average  (NSIDC)

Of the predictions made already, upthread I don't know which metric people are referring to.

The one I am most used to is number 5, the lowest NSIDC 5 day trailing average. Last year it came in at 4.15 million km².

I think this year will be lower but not enough to beat 2012. I'll say 4.0 

[Stephan]

My guess of today (4.0 M km²) refers to the daily minimum in September according to JAXA.

[Freegrass]

A prediction like this completely depends on the weather this season, but I'm gonna gamble on the worst case scenario, which is to beat 2012, and end up below 3 million km2.

First of all the clean air, and lack of contrails. That will not only influence the arctic directly, but also indirectly as Eurasian heatwaves move north. Like it happened last year, with record Greenland melt.

So Eurasian heatwaves will cause temperatures in the arctic to go up, above it's own extra insolation that it will get because of clean air.

The big question is if this will produce more clouds and storms. I'm guessing that more heat means more energy, and thus bigger storms. If we're unlucky, and we get storms that push out the ice through fram and the garlic press, then we'll surely break the record.

But what I'll be watching is hot pacific water entering deep into the CAB. I think that this will be a major problem this year. Pacific hot water will heat up even more with clean air and higher insolation. And more of that hot water is flowing into the arctic because of the slowdown of the AMOC. Which gives that water more speed to penetrate deeper into the CAB every year.

The ice in Chukchi sea is already very thin, with the Bering sea seemingly losing ice fast. So that's the side of the arctic I'll be watching this season. Together with the weather.

If it all turns against us, I can see us dropping below 3 million km2.

[Juan C. García]

My guess of today (4.0 M km²) refers to the daily minimum in September according to JAXA.

I bet below 4M km², according to JAXA. And hope not have it under 2012 min. So that is the range.

[oren]

I don't have a proper "feel" for this year's melting season, so I'll not hazard even a wild guess as to the Sept. min. Extent and area are stubbornly tracking low which could be a sign of things to come, as both 2016 and 2019 showed the same stubbornness, and both gave a good fight and finished 2nd. However, the current anomalies are mostly in the peripheral Okhotsk and Baffin (with some Kara thrown in the mix) so not necessarily meaningful.
I would expect the Siberian side to be the easier victim, with the Beaufort hopefully showing more resilience, and the rate of export from the Pole to the Fram what could dictate the fate of the season.
We will wait and see.

[pearscot]

It's so funny that the longer that I have been here, and the more I know about the arctic, the less certainty I feel like I have in terms of overall projections (each year has a seemingly new facet/implication). That said, I don't think this year will beat 2012, but this is the year I'm really going to look if there are more systematic changes.

I define systematic changes as an anomalously warm Bering Sea in conjunction with Atlantification and increased bottom melt. Beyond that, the crack above Greenland is wild; there's a lot to be said about changing ocean currents, but my studies in grad school never got to that point.

Overall, I find it oddly strange it has been so clear in certain areas of the arctic for so long. I just don't recall it on WorldView being like that. I feel like the changing arctic is more cloudy. I don't think 2020 will break 2012, but if there are more signs of significant, more common patterns my interest will be piqued 

[johnm33]

I'm thinking the major consequence of increasing tidal action is greater recycling of Arctic waters south, if the flow of water south through Fram forces inflows from the Pacific then Beaufort will cycle anti[counter] clockwise and free up the channels of the CAA allowing the release of much of the freshwater lens. If the forcing is of Atlantic waters then these will cycle through to the ESS and a river of weak ice will move towards the pole from there, but this will induce more water from the Atlantic to flow south towards Banks is. again risking opening up the channels through the CAA. I'm hoping that this will occur too late to be important.

[interstitial]

I don't consider unattached areas like st Lawrence or Okhotsk (Hudson bay sort of is but I don't pay it much attention either) but connected things like the Bering sea I think are important even if they melt out completely every year. If they are iced over they slow down melting of more central seas. I like to compare the same day for different years I usually look at 2000 to 2020 to see what is normal. This is highly subjective but I find it helpful. The pacific side looks pretty normal to the recent past except for the last two year where winter freezing was abnormally low. Its been a few years since the Atlantic side has seemed abnormally low. It looks typical to me and I wasn't expecting that because the pacific side is less melted this year. Baffin bay seems like the only anomaly this year it seems a bit lower ice cover than normal. The Laptev and the Kara seas seem to be getting an early start at melting. Looking at the overall extent isn't a useful predictor at this time of year though its near the lowest. The central arctic seas are at the 2010 average piomas volume. So that is an indication of typical 2010 melt season. The peripheral seas are near 2000 decade average indicating a probable slower melt out of those areas. The aerosol effect and of course the weather are wildcard in this analysis. If I remember correctly lack of aerosol will add as much as a degree and a half centigrade more heat. However some aerosol takes longer to be removed from the atmosphere than others and not all sources will be shut down and not for the entire melt season. That will help melt the extra volume in peripheral seas. Even if you can predict a general temperature or precipitation for the coming season timing can make a significant impact on outcomes. So have no idea what this will do.


It doesn't look like a record breaker year to me for Vishop arctic sea ice extent. Maybe second place or fourth or fifth.

[Richard Rathbone]

2nd place for extent is on, 1st place isn't, and won't be for at least 10 years.
1st place for volume is on, but if it happens it won't be beaten by much.

The downward pressure on summer ice is much lower now than it was from roughly 1980-2010 because the long term variability has switched into a phase that increasingly favors ice in summer over ice in winter and is acting against the pressure from increasing CO2 rather than with it.

September 2012 ice was really freaky in how well organised it was to minimise extent for the amount of ice there was, so while I think records in other measures are threatened every year, that one is going to last for a long time, and decarbonisation fast enough that it lasts for ever is still a possibility.

[misfratz]

As others have said it's really hard to say because so much is down to the weather. I recall that tamino had a great post about this years ago, where he showed there was no predictability using anomalies before June for the September minimum.

However, I think it's interesting to think about what we can say with some certainty. So, on the NSIDC numbers it looks certain that ice area and extent will be lower than the lowest values before 2007 (5.5 & 4.01), since every year since 2007 has been lower than those values.

I also feel quite confident that we will end up below 5 & 3.5 on the basis of the decadal trend applied to the last years with values above those thresholds. This would mean the lowest this year could be ranked would be 11th. This might seem like a weak statement, but I think it's pretty remarkable in the larger context.

Last year was 3rd, behind 2012 and 2007. It's interesting how 2007 begins to look like a paradigm shift, rather than simply a random minimum on a steady trend. I don't know whether the Arctic is yet ready for the next paradigm shift to consistently lower levels of Arctic Sea Ice, but the aerosol/cloud changes due to Covid-19 may well provide a big enough jolt to achieve that.

Oh, and looking at the detail of the sea ice thickness estimates, I think it's almost certain that the Northern Sea Route will be open again for a long while this summer.

[Aluminium]

Recently I found a correlation between low SIE in September and early low maximum discharge in two rivers. Maybe, it's way to predict melting season a little better. According to this rule, 2013-2018 had no chance to break the record. 2019 had low chance. The forecast by Hydrometcenter seems worse compared to 2019. If it works, the most probable minimum is about 3.2 M km² (JAXA SIE).

[bbr2315]

I will guess 2M KM^2 extent and 1.2M KM^2 area.

I think 2021 and 2022 will rebound a la 2013 and 2014. But this summer is going to be terrible. The worst impacts of the aerosol and contrails will result in worst-ever melt ponding this spring, and worst-ever melt this summer. However, they will also result in an excellent refreeze come winter 2020-21, and a brutal winter across the mid-latitude and well-populated areas of the continents.

The ice will survive this summer in a belt close to the CAA and Beaufort. Hudson Bay will melt out very late. Everything else will go early and quickly.

[JNap]

It seems that melt ponds and melting momentum created in May and early June were important indicators of the possible / likely range of September minimum -- if I am correctly remembering Neven's posts over the years. 

Thus too early to have much of an indicator in that regard.

But is seems with overall volume likely to reach an all-time low, I am predicting that extent will be 3.8 million Km2 and thus end up 2nd behind 2012.

[jdallen]

It's so funny that the longer that I have been here, and the more I know about the arctic, the less certainty I feel like I have in terms of overall projections (each year has a seemingly new facet/implication).

I've learned and arrived at much the same conclusion.  Way too soon for any projections.

What happens hinges on spring weather, especially loss of high latitude snow pack.  Not just on the ice, but on land as well.

In the melt season thread, there is discussion conditions are ramping up in serious ways to do just that.  I'll be watching for "May Melt Ponds".  If we see them, along with loss of snow cover, that will push September's potential down seriously.  Without that, (and continued assault on the ice in June)  we're looking at another year somewhere between 3.75 and 4.0 million km2 extent and 2.5-2.75 million km2 area.  Volume is another matter, but dropping under 4,000 KM3 is not out of the question.

[bluice]

Early season absolute extent/area figures are not that important.

The data thread tells us year 2012 has 15th lowest ice extent in early May. Difference to leading years 2016 and 2019 is over 800 000 km2. Yet 2012 holds the Sept minimum record by a significant margin.

[Pavel]

GFS supposes snow will be gone in the Chuckchi sea in 10 days, this mean ice turning blue. But the snow in some parts of North America will continue to refuse melting

[Phoenix]

Early season absolute extent/area figures are not that important.

The data thread tells us year 2012 has 15th lowest ice extent in early May. Difference to leading years 2016 and 2019 is over 800 000 km2. Yet 2012 holds the Sept minimum record by a significant margin.

The data thread also tells us what a huge outlier 2012 was in terms of ice loss. In terms of remaining melt from this point in the season forward, 2012 beats every other year in the last 13  years listed on the data thread by over 1M km2.

If we were to match the second highest melt year of the last 13 (2007) from this point forward, we still won't be close to a new record. Something well out of the ordinary must happen.

[PragmaticAntithesis]

Some fairly unimpressive losses in the past month have made me think my 4.0sqMm prediction was a bit on the low side...

[Niall Dollard]

Some fairly unimpressive losses in the past month have made me think my 4.0sqMm prediction was a bit on the low side...

Going by NSIDC 1981-2010 median data, July and June are the two biggest melt months (for extent).

It's only beginning now. A long way to go yet.

[Aluminium]

3.2 M km² (JAXA SIE).

It seems, I should correct this preliminary prediction to 3.1 M km² or even lower. The Yenisei River far exceeded my expectations.

[Freegrass]

A prediction like this completely depends on the weather this season, but I'm gonna gamble on the worst case scenario, which is to beat 2012, and end up below 3 million km2.

First of all the clean air, and lack of contrails. That will not only influence the arctic directly, but also indirectly as Eurasian heatwaves move north. Like it happened last year, with record Greenland melt.

So Eurasian heatwaves will cause temperatures in the arctic to go up, above it's own extra insolation that it will get because of clean air.

The big question is if this will produce more clouds and storms. I'm guessing that more heat means more energy, and thus bigger storms. If we're unlucky, and we get storms that push out the ice through fram and the garlic press, then we'll surely break the record.

But what I'll be watching is hot pacific water entering deep into the CAB. I think that this will be a major problem this year. Pacific hot water will heat up even more with clean air and higher insolation. And more of that hot water is flowing into the arctic because of the slowdown of the AMOC. Which gives that water more speed to penetrate deeper into the CAB every year.

The ice in Chukchi sea is already very thin, with the Bering sea seemingly losing ice fast. So that's the side of the arctic I'll be watching this season. Together with the weather.

If it all turns against us, I can see us dropping below 3 million km2.

It sure looks like the first part of my scenario, and your river data, are turning out to be right Aluminium... Let's hope we don't drop too much below 3 million km2...

[grixm]

I'm placing my bet at 3.67 Mkm^2 (JAXA)

[HapHazard]

What, no poll?!

[gerontocrat]

What, no poll?!

Ask him nicely & maybe Juan will do the biz like last year.

https://forum.arctic-sea-ice.net/index.php/topic,2719.msg202499.html#msg202499

[PragmaticAntithesis]

What, no poll?!

I'm new here, and didn't know how to make a poll until now. Next year's thread will have one!

[gerontocrat]

What, no poll?!

I'm new here, and didn't know how to make a poll until now. Next year's thread will have one!

If I remember rightly the polls are setup as follows.....

The first poll(s) is (are) usually set up in late May & called the June Poll, as by then at least one month of solid melt has happened. 2 polls, a poll for NSIDC Sept Average extent , and another for JAXA daily extent minimum.

Then 2 new polls for use in July, and 2 more  for use in August.

More kudos goes to the person with the closest call in the earliest poll.

[PragmaticAntithesis]

What, no poll?!

I'm new here, and didn't know how to make a poll until now. Next year's thread will have one!

If I remember rightly the polls are setup as follows.....

The first poll(s) is (are) usually set up in late May & called the June Poll, as by then at least one month of solid melt has happened. 2 polls, a poll for NSIDC Sept Average extent , and another for JAXA daily extent minimum.

Then 2 new polls for use in July, and 2 more  for use in August.

More kudos goes to the person with the closest call in the earliest poll.

Nice! JAXA June poll added.

[charles_oil]

Suggestion - these seasonal - or other - polls should have a title that starts with the word POLL  - so its easy to spot them and zoom in (and se how well or badly we are doing).  Thanks

[oren]

P.A., I would much prefer that the person who opened the polls last year (JCG) will do so this year as well. Both because they deserve to continue the tradition, but also because things should happen at the same seasonal timing and with similar format and title as has been found to work in the past, along with lessons learned. So I'd rather you remove the poll from this thread.
If JCG is not interested (which I doubt) then the poll(s) should be opened at a similar time and with similar expiration as previous years, in a separate thread titled accordingly, to enable wide participation and enable comparison of poll results between years.
I thought this thread was quite useful in and of itself, by venting away some of the long term speculation off the melting season thread, and also giving people a place to discuss their thoughts freely for the coming season, even if they are quantitative or necessarily based on data.

[Juan C. García]

P.A., I would much prefer that the person who opened the polls last year (JCG) will do so this year as well. Both because they deserve to continue the tradition, but also because things should happen at the same seasonal timing and with similar format and title as has been found to work in the past, along with lessons learned. So I'd rather you remove the poll from this thread.
If JCG is not interested (which I doubt) then the poll(s) should be opened at a similar time and with similar expiration as previous years, in a separate thread titled accordingly, to enable wide participation and enable comparison of poll results between years.
I thought this thread was quite useful in and of itself, by venting away some of the long term speculation off the melting season thread, and also giving people a place to discuss their thoughts freely for the coming season, even if they are quantitative or necessarily based on data.

I am waiting for June 1st., to make the poll. This poll can be May poll.

[oren]

That's fine. But in that case this poll's expiration should be set much earlier than its current date of July 1st.

[Juan C. García]

That's fine. But in that case this poll's expiration should be set much earlier than its current date of July 1st.

It can be closed on June 1st. For now, it is a good place to forecast 2020. And with today's drop, this topic can become a hot one. 

[ADS NIPR VISHOP (JAXA)] Arctic Sea Ice Extent.

May 14^th, 2020:
     11,744,488 km², a century drop of -152,324 km². 
     Source: https://ads.nipr.ac.jp/vishop/#/extent

[Juan C. García]

It is starting Arctic sea ice melt on several places...   
https://seaice.uni-bremen.de/sea-ice-concentration/amsre-amsr2/

[gerontocrat]

I'm going to wait to see if the GFS forecast holds true - it's done very well so far.

Because if it does ........

[RikW]

I'd say between 4m and 4.25m +/- 250k, following the trend we have been on for a long time; Anything else would require good or bad weather in my opinion. And with the current drop in aerosols I'd rather expect weather that's bad for the ice than good weather.

[ArcticMelt2]

There is a great danger of disaster this summer in the Arctic. This is due to the almost total ban on air travel this year. Apart from volcanoes, airplanes are the only way to throw sulfur dioxide into the stratosphere.

[El Cid]

There is a great danger of disaster this summer in the Arctic. This is due to the almost total ban on air travel this year. Apart from volcanoes, airplanes are the only way to throw sulfur dioxide into the stratosphere.

I agree. I expect a new record low this September. The COVID induced recession will have a very strong effect on the ice which was already on the brink of collapse...

[Stephan]

I wouldn't go for a new minimum record in September. I guess the minimum JAXA value will be something around 4.1 ± 0.15 M km².

[kiwichick16]

September daily minimum  2.5 - 2.75  million sq kms

International commerce …..and particularly aviation ….looks like it will struggle to ramp up again for several months yet.

The current pandemic could cause lasting climatic effects long-term if it accelerates the transition to a low carbon economy

[jdallen]

I'm going to couch my estimates in probabilities, based roughly on how the weather plays out.

With good weather for ice retention, I see us having a 90% chance of going under 4 million km2 extent.  I think under 4.1 million is a near certainty.

With conditions like some of the hotter recent years, I see us having about a 40% chance of beating 2012.

If we have conditions like that of 2007, 2011 or 2012, 95% chance of beating 2012, and 50% chance of going under 3 million km2.

[johnm33]

Further to ^11, the high has been pushing near surface waters through Fram fairly rapidly, the easiest route for flow is aligned with Amundsen basin and from the other side of Lomonosv between 140E-180 so a remote possibility of a +[near] pole hole. Come tuesday we should see Amundsen gulf begin to clear and that forcing in turn push the ice towards Chukchi, then any thick ice along the CAA sucked into Beaufort over the following days.
ATM the incoming AW looks like it'll head towards both Kara and Ellesmere

[PragmaticAntithesis]

Of course 3 severe daily drops, a nasty storm brewing, very early albedo shifts and alarmingly early volume drops all happen immediately after I say my prediction was too alarmist! I bloody jinxed it.

Alright, let's see if this counter-jinx works...

AAAHHH THESE CENTURY LOSSES ARE TERRIBLE AND WILL CARRY US ALL THE WAY TO BOE, WE'RE ALL GOING TO DIE!!!

[gerontocrat]

Hopefully at the end of this month Juan will open 2 polls, one of which will be the NSIDC September monthly average.

Attached are two graphs

- the extent plume from now to end of September - using daily change from now of the last 10 years, which shows how in September extent staying close to and just below 4.5 million km2.

- the September monthly averages from 1979 + the deviations from the trend. The trend value for September 2020 is 4.40 million km2.

Both graphs show the extent to which 2012 was the outlier.

The first graph also shows that for the remainder of the melt season 2016 melt was average - most of the damage was done earlier in the melt season.

[Aluminium]

3.2 M km² (JAXA SIE).

3.1 M km²

3.0 M km². Dates are almost determined this year. I have no exact data about discharges but assume +30(+10...+50)k m³/s from 2019, based on very low value in 2019 and relatively weak floods this year. However, the Lena River may surprise. Anyway, observed minimum highly unlikely to be above 4.0 M km². Current weather conditions seem to support it too.

[bluice]

Thanks to ongoing background warming we should expect every season to go lower than the previous ones. However seasonal variability, a.k.a. weather, obviously plays a major part.

I predict 2020 Sep minimum to end up 2nd in JAXA SIE. Figure will be somewhere between 3.4 and 3.6 km2.