« PIOMAS June 2018 | Main | As we approach the final stretch... »


Feed You can follow this conversation by subscribing to the comment feed for this post.


We forget Hudson Bay and the average daily annual extent? Especially Hudson Bay, where there was a massive low temperature event during winter past... This ranking , 6th place is meaningless without that caveat:



There's a big caveat at the end of the post, showing volume and extent/area graphs without the periphery like Hudson Bay.


Hi Neven,

Yes, the only good thing is the Cyclone you cited, is there where it should be, hopefully likewise Lows will stay in the region.


Nice work, especially highlighting Oren's contribution.
Looks like Greenland Sea has flipped from being a negative to a positive for ice volume later in the season, unless Fram export picks up again.


At 80 degrees North, the sea ice pictures between 2012 and 18 can not be more different:


It is a new ice-scape, unfamiliar territory for all who analyze.

John Christensen

Thank you for another great PIOMAS update Neven - the comparison with prior years clearly shows how much weather slowed melting in June of '16 and '17 - and just how massive the 2012 melting momentum was already by June.

One comment puzzled me though:

"What makes this melting season interesting so far, is that melting has been progressing relatively slowly the past few weeks"

Melting has been progressing relatively slowly ever since the max volume was reached, as you have stated in the April, May, and June PIOMAS updates also, but I guess the understatement was not intended?


The real slowdown - relatively speaking - started around mid-June, so that's a couple of weeks. The point is 'that 2018 isn't far behind at all in those regions that determine the minimum'.

If it weren't for the current weather forecast, I would say that the melting season may have some tricks up its sleeve. But the clock is ticking.

Michael Hauber

I think the key difference between this event and the 2012 event is the size of the cyclone. Looking back at your blog post for the 2012 event the Canadian analysis chart shows the main circulation of the low covering almost the entire Arctic basin. The chart above shows the circulation mostly in the Canadian quadrant. The squeeze between the high pressure towards Siberia adds a bit of impact though. The ice in that direction had already been showing signs of weakness IMO so the impact could be interesting.

John Christensen

Hi Neven,

I am sorry, but I cannot detect any slowdown in melting in the second half of June. I will make just one more comment on this, as this is your space.

Subsequent to the SSW event late February, we saw a drop in DMI 80N temp and a consolidation of the Arctic circulation causing a reduction in heat influx.

The effect of the SSW event (As also observed in early 2013) was a switch to rapid sea ice gains followed by reduced melting - but with a waning effect, as we got further removed from the event.

Your monthly PIOMAS updates seem to reflect this:


“sea ice volume growth during March was the largest in the 2007-2018 period, well above the average of 1832 km3”
[About 450 km3 above average]


“sea ice volume grew by 373 km3 during April, which is the largest increase in the 2007-2018 period, almost 200 km3 above average.”

“After the maximum was reached, the trend line flattened, with sea ice volume going down by a meagre 91 km3”


“volume decrease for May 2018 has been below average: 2285 vs 2650 km3”


“After a slightly below average June volume decrease (6199 km3 vs 6217 km3 for the 2007-2017 period)”

The relative level of melting therefore has been accelerating from a very slow start in April, and seems forecast to continue with moderate acceleration in July also.

If you do not mind, please share by what metric you observe a relative slowdown in melting by mid-June.


JAXA SIE. It went from second lowest on June 10th, to 10th or 11th one month later. This doesn't apply as much to PIOMAS, I agree, but there too the gap with other years has been slowly increasing. So, relative to other years, melting has been on the whole below average these past few weeks.

But this seems to be mostly caused on the periphery, because 2018 still is among the lowest in the 'inner core'. This implies, like you say, that melt will accelerate again (relative to other years). That was the whole point.

2018 may be 10th or whatever on some charts, but it isn't telling the whole story. There's a discussion on the ASIF whether weather conditions need to be more conducive to melt for that to happen ostensibly, or whether there will be a cliff, regardless of the weather.

I'm inclined towards the former, but there seems to be a lot of (late) melt ponding and dispersion, so when the weather does switch, this melting season may go a lot lower than the numbers predict right now.

No signs of that in the weather forecast as of yet.

John Christensen

Got it, makes sense considering the change in SIE, thanks Neven.

BTW, following an impressive drop of 751 km3 on June 25-27, PIOMAS registered a massive 464 km3 drop for June 28 - the second largest daily volume drop in the PIOMAS record after June 19th 2007, but ahead of July 21st 2012 in third spot.

Robert S

Looks like the hottest spot in Canada today is northwest of Hudson's Bay, almost on the shores of the Arctic Ocean. That's somewhat anomalous, and if that heat starts getting sucked north it'll certainly accelerate the melting in the CAA, which would likely have knock-on effects.

John Christensen

For the 'inner core' sea ice:

I agree with you Neven that weather needs to be more conducive to melting for further significant volume drops in the most northern region.

As it is the weakened low, which seems to continue hovering over the CAB for the next days, has had its impact:

- DMI 80N temp ( http://ocean.dmi.dk/arctic/meant80n.uk.php ) has dropped both due to cloud cover, as well as below freezing temps extracted from northern Greenland into the Fram Strait and north of Svalbard and Frans Josef Lands ( http://ocean.dmi.dk/arctic/weather/arcticweather.uk.php )

DMI also translated a recent article on the cold summer in Greenland, exceptional precipitation, and the record temperature set in May at -46.5C:


With Greenland in completely reverse mode compared to 2012 and weather conducive to reduce sea ice melting for the next week, we are going to see the divide with 2012 widen, and the preconditioning will not be in place for a repeat of August 2012, where all indices kept plummeting due to the significant amount of heat that had been built up in that summer.


I wont be surprised if we get a 200K drop day, perhaps has happened yesterday.

Meantime back at the action on hand


2018 melt is complicated, rife for misinterpretations, nevertheless quite interesting, perhaps key area is Beaufort sea ice speed floes loosing sea ice opacity day by day.


It may sound counter intuitive, but Beaufort sea should be literally closed choked with ice at present, there is a massive amount of broken up sea ice heading its way:


Yet there is melting nevertheless, new Melt zones like Fram Strait popped up at 2 other locations as well..


From looking at DMI temps over the last week or two it's only got (relatively) colder... so the interesting thing will be just how much of a recovery year 2018 counts as in the overall narrative IMHO!

Greenland isn't out of whack either so the story of recovery can't be too far off,.. atleast this year.

John Christensen

Hi AJbT,

Agreed; by early July we are at 2014/15 levels for area, extent and volume from near or at record low by the end of February on all of these metrics.

At some point discussion must surface about the reasons for this relative recovery, which I see being closely linked with the SSW event by the end of February - exactly as we saw in January 2013 and subsequent months.

Wild speculation on my part:

- Could major wintertime SSW events in general be linked to subsequent Arctic circulation reinforcement?
- Secondly, and in particular if the above-mentioned hypothesis has validity, we have seen fewer major wintertime SSW events in the past decade compared to prior decades: Could the reduction in major SSW events be caused by AGW, and if not caused by AGW, could the randomly reduced occurrence of major SSW events have contributed to enabling more heat to enter the Arctic region in the past decade?


"Could major wintertime SSW events in general be linked to subsequent Arctic circulation reinforcement? "


What we are witnessing is a major summer time circulation change, coming in large part due to thinner sea ice. Which by the way is fitting exactly as I projected last April, so I do have some authority on this.

The cool spell of winter 17-18 came over the Southern CAA, the entire Arctic ocean was mostly warmer. It should be very interesting explanation on how sea ice can be reinforced by warmer air?

ECMWF calls for the return of clockwise gyre High pressure at month End, then it will be interesting, for those who predict correctly have done sound science work....

Written in April

"It is during this June-July period again when the sea ice melts slows, because cold air in Arctic summer is a great cloud saver, they don't evaporate as quick as with anticyclones. "


It does not mean that at end of July the melt season is over and we can speculate over mysterious means by which warmer air creates more sea ice...

" switch back to their more winter like modes. It is in early August when we will be able to evaluate if the Arctic sea ice will have a new all time low extent in September. "

Written same page. It is complicated folks, don't make it muddier.

John Christensen

wayne,you write:

"The cool spell of winter 17-18 came over the Southern CAA, the entire Arctic ocean was mostly warmer"

This is not a correct statement.

You see clearly on DMI 80N first the immediate heating caused by the SSW by the end of February, and next the drop by 20K by mid-March of the DMI 80N - bringing this measure below the 1958-2002 average:



"It should be very interesting explanation on how sea ice can be reinforced by warmer air?"

Sea ice volume growth in both March and April were highest of the 2007-2018 period, as reported by Neven, so what is really called for is an explanation of this reality, which is offered by the reinforcement of the Arctic circulation, not "warmer air".

In your forecast of April 27th, which was overall great, you place the primary cold temperature north pole in the CAA and then mention that the normal high north of Norway would make NW Europe cooler.

In fact the CTNP was primarily based over Greenland, pushing the Scandinavian high much further south causing record high temperatures across much of NW Europe - a situation that still persists mid-July.

Consequently, we have a situation opposite that seen in 2012:

2012: Greenland high/extreme melting and 'heat-dome' delivering substantial heat to Arctic region
2018: Greenland low/cold ice sheet delivering colder air to Arctic region


Hi John

"This is not a correct statement."

I let NOAA answer that:


Figure 1, latter part of winter had a huge positive surface temperature anomaly for the entire Arctic Ocean.

"primary cold temperature north pole in the CAA and then mention that the normal high north of Norway would make NW Europe cooler."

It was so in every way. please refer figure 2 and 3

I did not project for NW Europe after April and May period, I should have but did not do so, I am thrilled they had a nice summer so far.

"In fact the CTNP was primarily based over Greenland, pushing the Scandinavian high much further south causing record high temperatures across much of NW Europe - a situation that still persists mid-July."

The CTNP might have been centred over Greenland for a couple of days, but overall was primarily centred just North of Ellesmere Island.
Refer to figure 4.

"2018: Greenland low/cold ice sheet delivering colder air to Arctic region"

Nope, so far it was a dominant cloudy period mainly by 2 factors, 1 is the persistent yet expected Arctic Ocean Gyre cyclones increasing albedo not allowing the larger part of Arctic Ocean to warm up, and 2 the biggest factor, anti dumping of sea ice at Fram Strait, instead vast badly broken up sea ice constantly rushed towards Beaufort Sea barely handling the melt load, but melting sea ice by the ton loads nevertheless, the scattering of sea ice where there should be open water has decreased surface temperatures, the broken up sea ice spanning a huge section of the Arctic Basin has also stimulated cloud formation, but all this will change in a week or so.

We must remember, that was it not for Anthropogenically Enhanced Global Warming, we would be in the midst of a full fledged massive cooling La-Nina period, in response to 2016 Historically warm El-Nino.
Instead I see partial La-Nina effects, namely one smaller region cooled more than many. All this has changed in a mere 20 years, as we do remember the cooling post 1998.

My April projection is so far so good, now the last very hard to be correct real test is left , sea ice..... It is rather why I do this, weather is becoming to easy.


Too easy as opposed to very very hard getting our message through to the rest of the world, there are other challenges to surmount, all in the name of a better planet for our descendants and all living beings.

John Christensen

Hi wayne,

Thank you for sharing these metrics at this late hour of your day!

Figure 1 – Arctic temperature anomaly:

You measured the temperature anomaly from 2/15-4/30, which is not the period I am referring to above and includes both the warming part of the SSW as well as the subsequent cooling.

Try again with 3/15-4/30 to capture the relative cooling of the CAB, as the Arctic circulation stabilized subsequent to the SSW event.

Figure 2 (April/May pressure) and 3 (May temperature) – NW Europe:

Figure 2 clearly marks the high placed centrally in Scandinavia, and figure 3 shows the record heat in NW Europe. UK, Norway, Sweden, Denmark, and I believe Finland all had warmest May on record. People in Spain and NW Africa possibly felt a shiver..

Figure 4 – June air temperature:

Do you see more of the coldest air above Greenland or the CAA landmass? Seems like splitting hairs – and probably depends on the final status of the “Hans Island” in Nares Strait.. ;-)




"Do you see more of the coldest air above Greenland or the CAA landmass? Seems like splitting hairs – and probably depends on the final status of the “Hans Island” in Nares Strait.. ;-)"

The coldest air was consistently above Ellesmere as measured by Upper Air stations situated near sea level. I do recognize the models having problems with Greenland's height, as far as figuring out correct sea level pressures, always extrapolated and or which is the correct temperature of the entire atmosphere admitting Greenland's altitude and ice albedo always gives a cooler image.

Sea ice accretion depends in part on surface air temperature, if it is warmer as measured over a certain period, it can't accrete more, the days when was colder were vastly overwhelmed by the warmth of last winter. Nevertheless the said period you have suggested has no surface temperature anomaly over the Arctic Ocean. Definitely one for the South Central CAA. Refer to figure 5



For those who already are looking at 2019 for more action, 200,000 km2
is roughly the difference between 2018 and 2012:


There are other significant differences to be posted eventually....

John Christensen

"Nevertheless the said period you have suggested has no surface temperature anomaly over the Arctic Ocean."

Looking at the chart I will need to agree with you wayne: Maybe the higher than normal sea ice accretion towards the end of winter was only partly due to a slight drop in temperature, but primarily due to overall volume being very low and therefore enabling above-normal 'catch-up'.

"I do recognize the models having problems with Greenland's height"

Yes, the models in general have problems with the Greenland ice sheet, but DMI actually reports a 2meter temperature prognosis based on ECMWF forecast, where you see Greenland as the coldest area (Current day image):


DMI also takes physical temperature measurements at the Summit station during summer, but is often not reported due to the seasonal nature of this. DMI reported a record low for May this year at -46.5C.

Finally, the PROMICE weather stations on the slopes of the Greenland ice sheet seem to indicate some level of recovery or stabilization compared to 2010-2012, where temperatures both in winter and summer were higher than now:


I will check with DMI, why PROMICE data is only reported until 2016, as most stations are still operational..

John Christensen

PROMICE weather station data for Greenland ice sheet slopes can be found here:


See e.g. Kangerlussuaq (KAN_U) or Upernavik (UPE_U) for upper slope monthly temperature averages in the past decade.


There are very few reasons to be optimistic and expect a bonanza recovery of sea ice come September 2018, on three fronts the negatives outweigh the positives:


John Christensen

Yes, if these three fronts were the dominating factors determining Sept sea ice volume and extent, we have few reasons to be optimistic.

Why we should (still) be optimistic about 2018:

- Much more ice in the Arctic waters

- Daily volume drop stayed relatively high in July 2012 (See PIOMAS), whereas for 2018 the volume drop has slowed in line with 2004-13 average as seen on DMI:

- A larger sea ice cover in 2018 reflecting more radiation back into space

- Lower SST in peripheral seas in 2018 (Baffin, Kara, Laptev and Beaufort)

- Last, but certainly not least - GAC2012 appeared early August 2012 with significant mixing, resulting in unprecedented volume, extent and area drops in August. This could happen again, but if it does, there is less open water to mix with and lower average SST this year.

One of Jim Pettit's great trajectory projection charts show the likelihood of catching up with 2012 very well:



Of course if you believe that the Arctic seasons are moving in cycles, personally I've come to the conclusion of two, roughly, 5 year cycles moving in a 10-11 year cadence, which corresponds to the peaks and troughs of the solar sunspot cycle; then this all makes perfect sense.

If you think of it as two hotter years, two cooler years and one in-between year, then things match better.

In the cooler years it takes excessive weather and truly catastrophic ice conditions to create records. In the hotter years it takes moderately good weather to produce high melt and it takes exceptional weather to stop it melting.

What we saw in 2007 and 2012 was the two different scenario's.

2007 was disposed to melt and had exceptional melting weather. But it also had a fairly cohesive pack and a hell of a lot more ice to melt.

2012, on the other hand did not have really conducive melting weather but it just kept on melting. Then the GAC took over and it is the record we see today.

It is why I predicted that 2006 would not make a record. It was why I predicted 2017 would, but weather took over and damped it down.

It is also why I predicted that 2018 would be a poor melting season and talk of "recovery" would surface. Ditto 2019.

The thing is that the sheer weakness of the ice has taken over to some extent. We have more heat transfer in winter from the sea because the ice is thinner. Then again we have more heat transfer to the sea in summer due to the same thinness of ice.

Here, again, weather comes into play in a much bigger way.

If the ice is thin and fragile and the sun can heat the sea below, right through it, causing more bottom melt, then the weather is going to play a much larger part than it might have when it blocks the sun. Essentially because it is not just blocking the melting of the top ice, it is blocking the heating of the ocean through the much thinner ice.

Equally very cold temps in winter, especially late winter, with thin ice, will cause much faster than expected ice creation, for that time of year, as the thinner ice will not insulate the ocean so well.

Today the solar sunspots have been blank for nearly a month. It will be interesting to see how long that goes on. 2008/2009 had the lowest solar minimum recorded in the last 100 years. I'm sure nobody has forgotten, just yet, how cold the winters of 2008/9 and 2009/10 were in the northern hemisphere? It wasn't all jet stream.

So, after all that rambling. You probably don't need me to tell you what I think will happen in the "middling" year of 2020 and the "warmer" years of 2021/22, sitting right on top of the sunspots of solar cycle 25....

It is a _very_ long prediction, in terms of what we normally talk about here and the forum. However, as I have been watching for over 20 years now, I'm not that impatient.


A lesson may be learned with 2018: sea ice melting is a synergistic beast.

Absent all clouds it is all gone, no matter how thick, but since 2012, a new circulation system has been in place for summer, favouring more clouds. A melt forecast based on albedo alone is certainly going to fail big time at certain years. A melt forecast based on alleged ice thickness, not really severely tested, is also going to fail often. However, if the model includes everything, that model will be proven fiercely accurate.

In our case severe melting is occurring today exactly where sea ice is moving fastest, another important factor to consider:


There is also last winter Bering sea never having a good ice coverage,
its super warmed surface waters is now proven deadly for sea ice.


This current rate of sea ice extent decline does not come as a surprise,
was a long time in the making, and we can literally see where it is melting fast by "goodbye waves":


John Christensen

Zack Labe's take on the Arctic weather pattern this summer:



203,392 km2 extent drop on July 23 seems to be poised to be surpassed soon:


Hans Gunnstaddar


Looks like a whole lot of ice in the Beaufort ready to liquefy. Those melt day numbers you linked, Wayne for those recent days in July are huge!

Your early June prediction of this year's extent being similar to 2012 is much more likely now. You could be right, but more weeks ahead to see how it turns out. I have to admit, I'm getting a bit concerned. Record high temps in many areas of the NH and now Arctic melt speeding up this fast in July after a slow start. Seems like warming is now capable of shifting into higher gears.


Hi Hans

Absent Fram Strait, the melting continues never the less, pretty much where its moves most. I wont be surprised if a 250K day will happen...


Another WOW day, hang on to your hats , the roller coaster ride is about to begin going steep downwards, I am sure the coming days will push Arctic sea ice back in the headlines, rightfully so:


We have to wait a bit until at least 15% per grid coverage becomes 14%,
or less, it wont be long.

John Christensen

With much more sea ice this year compared to 2012 and 2018 SIE closer to 2012 SIE than is the case for SIA, we unquestionably have both a thicker and more compact ice pack in 2018.

The thin and very fractured ice pack in 2012 enabled GAC2012 to continue high melt rates far into August.


I can't log on to the ASIF. Anyone else experiencing trouble as well?

Account Deleted

Neither can I....

Account Deleted

Getting error message
Session verification failed. Please try logging out and back in again, and then try again.

John Christensen

No connection to ASIF from here either..


It will probably be solved this evening.


2018 is now #1 lowest extent in the Central Arctic Basin, which is what really matters after all said and done, the peripheral seas should be near 0 extent eventually.


Is very very unlikely that thicker sea causes more melting :(.....


Been thoroughly out of it, but for the side interest, it's interesting to so these 'competing' comments (but that may just be my perception, looked at this graph with select options.


Whatever ice extent there is, thick or thin, it's close to that poorest 2012 year, well, well, well outside 2 std deviation.


NSIDC shows the Extent pulling back away from 2012 now and running parallel.

From Wednesday 2012 recorded a straight 10 days of 100k breaks. Given the huge losses we have just seen, with peripheral seas and the Beaufort collapsing, it's going to be interesting to see if 2018 can emulate it.

Not sure about that but it will bear watching. The next two weeks will give us a clearer idea of just how bad the CAB ice is and how likely it is to melt.

DMI continues to show 2018 80N temps as lower than the melt years and we know it has been snowing up there.

This is going to be another interesting one to watch. I could understand it if 80N were at 1C, the ice was melting under clear skies and that the temp level was due to the ice. But we have seen constant cloudiness over the CAB all summer.

We shall watch and learn.

Seke Rob

"NSIDC shows the Extent pulling back away from 2012 now and running parallel."

As of August 2, there's 251KKm2 Extent difference with 2012... shrinking. Suppose we'll be seeing an update after the weekend.

Seke Rob

BTW, how is CRYOSAT-2 doing versus PIOMAS these days. The suggestion is that they're still accumulating data.


The comments to this entry are closed.