Support the Arctic Sea Ice Forum and Blog

Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Messages - Glen Koehler

Pages: [1]
Arctic sea ice / Re: The 2019 melting season
« on: June 24, 2019, 10:59:15 AM »
Here we have freshwater floating onto land-fast ice, land-fast ice breaking up, and a brief (?) reversal (?) of the Fram export.

Arctic sea ice / Re: The 2019 melting season
« on: June 23, 2019, 01:48:54 PM »
Jeez, these SSTs!

Indeed, here's a comparison with 2012, 2016 and last year. Mind you, my archive is far from complete, and so the 2012 map is for July 1st and the 2016 map is for June 28th. So, this year has 9 and 6 days to go, respectively, but 2019 is already leading basically everywhere, expect for the Atlantic side.

Which is interesting, as PIOMAS says that the ice is thicker there, compared to previous years. If that's true, transport rather than insolation will be needed to really damage the ice pack.

So, a Dipole would a) further attack the ice on the Pacific side, and b) transport all that (allegedly thicker) ice further south into the Atlantic. If July can provide this set-up for an extended period, 2019 will have a shot at the title. But as always, don't rule out surprises.

Arctic sea ice / Re: The 2019 melting season
« on: June 23, 2019, 11:55:26 AM »
The scale of melting significantly exceeded 2012.

Arctic sea ice / Re: The 2019 melting season
« on: June 22, 2019, 02:35:36 PM »
Fires south of the Laptev Sea are growing, and the high will draw smoke north over the next couple of days.

I've attached a gif (click to animate)of june 19 and 21(20th obscured by cloud) showing the northernmost fires, quite small on the 19th, but now producing a huge pall of smoke or smog. And the Copernicus/WindyTV PM2.5 particle forecast for tomorrow, which I'll assume is a proxy for smoke in the far north

Thank you for the animation. I would like to be clearer in my own mind as to what these fires mean for the ice.  My interpretation is that if the smoke drifts over the Arctic, then the smoke acts rather like clouds and shields the ice from solar radiation.  But if there is rain or snow or something else that brings the soot from the fires onto the ice, then that darkens the ice and top melting might then proceed more rapidly.  So a lot depends on whether the smoke stays aloft or settles  Am I missing something? 

If I have it right, then in these sunny conditions, then it would seem reasonable to assume that this smoke over the Laptev is helping preserve the ice.  Happy to be corrected!

Arctic sea ice / Re: The 2019 melting season
« on: June 13, 2019, 02:10:12 PM »
Slater 50 day probability model forecast taking somewhat of a vertical cliff stance. Certainly interesting modelling. Time will tell...

Arctic sea ice / Re: Arctic Ocean salinity, temperature and waves
« on: June 12, 2019, 12:39:16 PM »
Ascat sep17-jun11. Still experimenting with the circular mask. The swaths move a little over time.

Arctic sea ice / Re: The 2019 melting season
« on: June 12, 2019, 11:02:09 AM »
2012.6.12-2019.6.12 which one is worse

Arctic sea ice / Re: 2019 sea ice area and extent data
« on: June 10, 2019, 03:22:54 PM »

Recent daily area losses consistently lower than average, the last 3 days at  less than half the 2010's average area loss on those days.

Analysis of the last 16 days presented in gerontcrat's tables.
The seas are split in half:
There are four seas with almost no changes/even slight increases:
CAA, CAB, ESS and Grønland
There are seas with almost no ice at all, which cannot contribute anymore to total loss:
Okhotsk, St. Lawrence, Bering
And there are seas that constantly lose ice:
around 28% loss: Baffin and Barents,
around 20% loss: Hudson and Beaufort
around 14% loss: Chukchi, Kara and Laptev

This together gives the whole picture of much too low losses for all seas.

Arctic sea ice / Re: The 2019 melting season
« on: June 08, 2019, 06:53:05 PM »
The key to the existence of melt ponds is that fresh water from melted snow enters the pores and fractures and refreezes. The freezing point of fresh water is higher than the freezing point of salty first year ice or the basal ice/salt water mixture.

Quote below from the summary of the linked article:

Plain Language Summary

When meltwater pools atop Arctic sea ice it greatly increases how much sunlight the ice absorbs and how fast it melts. The formation of so‐called melt ponds on the sea ice is a major contributor to summer ice melt, but the fact that the ponds form at all is surprising. The sea ice is very porous and water should trickle right through it. We found the mechanism that allows ponds to form. When freshwater, from melting snow, percolates into the salty, cold ice interior, it re‐freezes, plugging the pores in the sea ice and making it so that ponds cannot drain. This very important process is not captured by current global climate models, and we suggest that working to include it in future models will improve the predictions the models produce of future sea ice.

Arctic sea ice / Re: The 2019 melting season
« on: June 08, 2019, 12:54:09 PM »
A couple of points about the lack of melt ponding and dry vs humid air.

Although the ice is quite fragmented in places, the individual ice floes are still mainly hundreds of metres or even tens of kilometres across. More than big enough for melt ponds, and more than big enough for the melt ponds to be big enough to be visible on one of the products available to us. 

The latent heat of vaporisation of water is roughly 7 times higher than the latent heat of fusion (melting). This means that 1 g of water condensing releases enough energy to the surroundings to melt 7 g of ice (assuming the ice temperature was close to freezing point). When warm humid air enters the Arctic, it cools, and some of the water vapour condenses, releasing energy. This can start surface melting and melt ponding. In past years we've frequently seen this happen in practice - a warm, cloudy weather system moves through part of the CAB, and when it leaves and you can see the area on Modis again, there is lots of surface melt/melt ponding. Obviously if the weather front brings rain as well, this will also contribute to melting. This effect has also sometimes been visible on the buoys.

Arctic sea ice / Re: The 2019 melting season
« on: June 08, 2019, 09:34:27 AM »
Imo it's premature to be surprised about the small fraction of ice surface showing melt ponds.

Attached is a gif of SMOS microwave images for each year from 2010 through 2019.

Specifically, the images are yyyy0607_hvnorth_rfi_l1c.png, where yyyy is the year, obtained from

These images are sensitive to melt ponds.
IGNORE THE COLOR LEGEND'S NUMERICAL SCALE & LABEL (the color order progression should be valid though) - DURING THE MELT SEASON THESE ARE NOT LEGITIMATE THICKNESS MEASUREMENTS. Instead, my understanding is that any color other than beige indicates ice that is:
a) thin, ~<50 cm; &/or
b) has concentration well below 100%; &/or
c) has surface liquid water.
In particular, colours other than beige in the ice pack interior are likely to indicate the presence of surface water.

& it is seen that only 4 of the 10 years have extensive melt ponding in the Arctic Basin on 7 June: 2012, 15, 16, and 18.

All of 2010, 11, 13, 14, 17, and now 2019, don't have extensive melt ponding by 7 June.

So the comparisons above with 2012 are not particularly surprising, given that 2012 is one of the 4 years in the data record that has extensive melt ponding on 7 June, while 5 of the 9 previous years on record are similar to 2019 in not displaying extensive melt ponds by 7 June.

P.S. Given the weather forecast, I expect SMOS to show extensive melt ponding, especially on the Russian side, within the next few days.

Arctic sea ice / Re: The 2019 melting season
« on: June 07, 2019, 05:23:41 PM »
Summary: despite record strong high pressure dominance in the Arctic and abundance of blue skies from late April to first week of June (that’s like 50 days) , record SATs as well, surface melting shows a delay of at least 1 week wrt 2012 or even 2016.
So something is missing from the “lots of insolation” formula to really get the surface melting season going.

Indeed, and as said before, I think it may have to do with melt onset (which starts earlier under cloudy, moist conditions). Here's that paper I referred to, called Melt onset over Arctic sea ice controlled by atmospheric moisture transport: 2016 Mortin et al. I still haven't looked at it properly, to see whether 2010 and 2012 were extra early.

But here's what it's about:

The timing of melt onset affects the surface energy uptake throughout the melt season. Yet the processes triggering melt and causing its large interannual variability are not well understood. Here we show that melt onset over Arctic sea ice is initiated by positive anomalies of water vapor, clouds, and air temperatures that increase the downwelling longwave radiation (LWD) to the surface. The earlier melt onset occurs; the stronger are these anomalies. Downwelling shortwave radiation (SWD) is smaller than usual at melt onset, indicating that melt is not triggered by SWD. When melt occurs early, an anomalously opaque atmosphere with positive LWD anomalies preconditions the surface for weeks preceding melt. In contrast, when melt begins late, clearer than usual conditions are evident prior to melt. Hence, atmospheric processes are imperative for melt onset. It is also found that spring LWD increased during recent decades, consistent with trends toward an earlier melt onset.

Here's a recent paper on how radar satellite images can be used to monitor melt onset, 2019 Howell et al. From the conclusion:

Given that the timing of melt onset influences the end of summer sea ice extent in the Arctic (Perovich et al., 2007) and that positive trends in downwelling longwave radiation are linked to positive melt onset trends across the Arctic (Mortin et al., 2016), continuing to provide melt onset estimates is important for understanding the response of sea ice to a warming Arctic.


In this study, we have shown excellent potential for the use of multi-sensor backscatter from SAR to provide high quality melt onset information over Arctic sea ice which would be of significant value to data assimilation systems. In anticipation of the availability of data from even more SAR satellites with the launch of the RADARSAT Constellation Mission, the multi-sensor γCo approach presented here may offer the most robust approach to estimate the timing of melt onset over sea ice across the Arctic.

Sorry, for the slightly off-topic, but given that May was so sunny and there was thus less atmospheric moisture transport, this could possibly explain why 2019 is lagging when it comes to melt ponding. Because the ice wasn't pre-preconditioned as much as in other years.

Arctic sea ice / Re: "Stupid" Questions :o
« on: June 05, 2019, 10:53:19 AM »
Off the top of my head (and take it with a grain of salt - I've only been following this since 2014):
2007 - warmest year in the Arctic, lots of thick multi-year ice exported into the Atlantic sector never to return. The year that crashed volume.
2012 - high pressure/clear skies, causing lots of melt ponds and an extreme June cliff, and then the Great Arctic Cyclone during August finished off large regions of thin ice. The year that crashed all records and cleared out every region in September except the CAA and the Greenland Sea (and the CAB of course). Had the advantage of following 2011 which broke previous extent and volume records.
2016 - early open water inside the Arctic ocean, soaking up energy in the spring. A cold June-July stalled it, but then August saw a GAC. The year that proved 2012 was not a statistical fluke, and came very close to it in the Central Arctic Basin (see attached area chart), though ice in other regions failed to clear completely, also due to a relatively early refreeze, and extent was higher due to low compaction.

I am surely doing the answer an injustice. For 2012 and 2016 it is best to read Neven's posts in the Arctic Sea Ice Blog at the time. Not sure if there is a similar summary of the 2007 season anywhere.
There is also a multi-year NSIDC animation on Youtube that can show some of the differences. I am sure someone can post the link.

Arctic sea ice / Re: Basic questions about melting physics
« on: May 29, 2019, 11:23:32 AM »
I am not certain I have calculations right so if I don’t let me know.

On the molecular scale freezing is an ordered stacking of water molecules so it makes sense that putting ions in the way makes that harder. On the macro scale we see evidence of that when the freezing point decreases.

It is common to see -1.8 C as freezing point of sea water but it is a bit more complicated. -1.8C corresponds to 29.52 g of salts/kg seawater. Molar ratios of  ions are listed below for 35g solute/kg seawater

0.546 moles Cl-
0.469 moles Na+
0.053 moles Mg2+
0.028 moles SO42-
0.0103 moles Ca2+
0.0102 moles K+

Anyway salinity in the artic near the surface varies from about 26 g of salts/kg seawater with a freezing point of -1.58C to 36 g of salts/kg seawater with a freezing point of -2.21C. For most purposes -1.8C is good enough.

But that is not really how sea ice freezes it starts at the temperature associated with the salinity of the water but the ice is nearly pure water. It rejects the salt when it freezes this increases the salinity of the water. So the water has to get a little bit colder to freeze the next bit. This process continues until the salty water gets trapped in ice. Eventually the last of the water freezes at about -21C.  This is first year Ice. It has pockets with high salt concentrations in it. Some of the pockets even most may not of frozen solid.

When the temperature climbs above -21C the pockets of high salt concentration melt first. The temperature is still too low to melt the pure ice.  Since Ice melts at the solid to liquid (or solid to gas interface but that is not relevant here) interface and as the salty water is in contact with pure ice and gets a little bit warmer it can melt a little bit more ice. This lowers the salinity and the ice can’t continue to melt until it gets warmer still.

The high concentration salt water can often burrow out of the pure ice before most of the pure ice melts. When the temperature drops again lower concentration salt water freezes inside. This is how multiyear ice is formed each freeze and thaw cycle of salty water can more and more salts out until it is pure ice with no salts. That makes the multiyear ice fresher and more melt resistant.

The melting temperature of the pure multiyear ice is dependent on the salinity of the surrounding liquid. When the pure multiyear ice melts the local salinity drops and the temperature must increase to melt more. That’s another reason it is more resistant to melt than first year ice.

So to succinctly answer the original question the purified ice melts at the temperature determined by the salinity of the surrounding water. That may be higher than -1.8C but it is not 0C.

Freezing point depression
 presentation on chemical composition in sea water

Arctic sea ice / Re: The 2019 melting season
« on: May 27, 2019, 04:36:52 PM »
Plotted the average skin surface temp from May 1-25 for the last 8 years. Sorry about the resolution, I was trying to save space.  2012 is upper left, 2019 lower right.

Arctic sea ice / Re: The 2019 melting season
« on: May 27, 2019, 11:11:13 AM »
I'd rather substitute the DMI temps with an insolation graph for the Arctic, but i do not even know if such data exist?

I'm not sure if I've understood your question correctly. But here's an insolation graph:

Rather more complex is Tealight's work on Arctic "Albedo Warming Potential"

I can't help but notice an apparent huge discrepancy between what PIOMAS reports as average arctic sea ice thickness and what Navy HYCOM+CICE shows. It seems that these two products are on opposite ends of the spectrum. Perhaps reality is somewhere in the middle?

Edit: i now realize these two products are a bit of an apples to oranges comparison with 2012 average PIOMAS thickness being lower because there is more extent to average the thickness across and some of that added extent is thinner ice.

As I've already mentioned, even comparing the US Navy's HYCOM+CICE thickness for 2012 versus 2019 is apples versus oranges.

By way of example, see ACNFS versus GOFS for May 25th 2016:

Arctic sea ice / Re: The 2019 melting season
« on: May 26, 2019, 11:26:42 PM »
     Bottom melting is a known driver but seems to get little monitoring.

Very difficult problem, as we don't have a lot of permanent stations for measuring/monitoring it, and it mostly has to be derived from satellite data.  This similarly translates into the uncertainty we have for ice thickness and volume. But to your point, it's actually monitored a lot, but the uncertainty of the measurement is not as low as we'd like.

    1)  In addition to temperature, cloud cover, insolation, wind, ice extent/area, ice thickness, ice age, ice condition, melt pond momentum, and land attachment, does the amount and temperature of Atlantic current inflow act as a driving factor for melt? 
It does, but we're still trying to understand it.  Mostly it's reflected in the increasing volatility of extent in the Barentz - looking at SST's over time, you can see "hot spots" evolving as we see here:

(sidebar: NOAA used to have very nice arctic view SST anomaly maps - they seem to have disappeared - anyone know what happened to them/have new links?)

    2)  Are there tracking and forecast maps/summaries for Atlantic inflow volume and temperature to use as predictors?  Known historical patterns or correlations with September minimum.     

Not really, or at least not that I'm aware of.  Again, we're still wrestling with a lack of sensor data, and mostly understanding it by inference from other measures, which have similarly limited sensor data such as salinity, SST and sub-surface temperature.

    3) Ditto Pacific inflow, but my understanding is that Atlantic inflow is a stronger influence.

See (2) answer above.  Same dance, different song.

    4)  Exploring my ignorance out to its edge, it seems like there might be a trend for cloudier Arctic summer weather with increased open water.   To the extent that there is skill, multi-month June-September Arctic weather outlook would be interesting.

It would, and there are others in forum who'd be better able to point you in the direction of resources.

A thread was started for that here:,2692.0.html

PS #1088 JDAllen - interesting point about melt ponds not just changing albedo but also serving to conduct air temperature into ice.  Never heard that before but it makes sense.

I was less thinking about transmission of heat from the atmosphere, and more about insolation and long-wave radiation trapped by the water being concentrated and directed to the ice via convection.  Having water on top of the ice transferring atmospheric heat I'm actually not so sure of; I actually think it may be a wash between the two.

Again, I think the major effect of meltponds is better capture of various forms of radiation rather than convective or evaporative heat exchange with atmosphere.

Test question "What is causing Arctic Sea Ice loss" - 3 letter acronym.  Stumped me.

Me too.  Context?  ;)

Pages: [1]