Support the Arctic Sea Ice Forum and Blog

Author Topic: Latest PIOMAS Volume update  (Read 1896850 times)

ChrisReynolds

  • Nilas ice
  • Posts: 1764
    • View Profile
    • Dosbat
  • Liked: 20
  • Likes Given: 9
Re: Latest PIOMAS update (October)
« Reply #400 on: October 12, 2014, 10:01:02 AM »
Jai,

Let's just get the IPCC issue out of the way first. When I was in the process of changing my mind from being a sceptic of AGW I downloaded AR3 and selected some random chapters from the WG1 scientific basis. I then went through those chapters checking to see if the IPCC was correctly reporting the science, this took some time as I didn't know very much. I've not done the same since, but on reading the section I referred you to I have read all the references (and more) and it reads like a fair assessment of the science. I'm just an amateur but I do know sea ice science pretty well. So unless you can point out problems with that section that will educate me further, I think the IPCC is making a fair assessment there.

You said:
Quote
...this one study alone, indicates that our current radiative forcing from 402% PPMV is actually acting like 460 PPMV according to our previous models.
I said, "The result doesn't seem so significant as to affect the 1.5 to 4.5degC confidence interval for climate sensitivity, which is based on multiple lines of evidence." Note the wide range of the pre-existing (basically Charney) estimate, and the multiple lines of evidence that have gone into that. 402 to 460 rounds up to a 15% increase, 15% extra on 3degC is about 3.5degC, within the range of 1.5 to 4.5degC, assuming RF tends to linear over such a short range. My opinion remains that until I see significant research substantially revising upwards outside the Charney range. I don't think Durack et. al is strong enough to overturn that range, but it is useful.

June/July insolation: What have you got, a 100W/m^2 variation with lattitude? Or some 20%. Agreed?

How much variation in insolation do you think different pressure systems with attendant cloud variation (thick low cloud to clear sky) will cause? I said "Tietsche removed the ice on 1 July, near enough to the solstice to make little difference from 1 June." Maybe I should have said "Tietsche removed the ice on 1 July, near enough to the solstice to make little difference from 1 June, bearing in mind the influence of weather"

I went on to refer to Schroeder & Connelly: "Schroeder & Connelly removed the ice on, 1 December, 1 March, 1 June, and 1 September, they found a very similar result." They also examined a warmed ocean due to open water, the results were that:

Quote
The impact is stronger when ocean temperature is
adjusted to ice-free conditions, but even then the differences
appearing between the sensitivity and Ctrl runs are clearly
smaller after a few years than the differences between
modelled and observed sea ice area."

You say:
Quote
How do you reconcile this with the ZHANG 2010 paper that states a largely summer ice free state with 4C regional (not NH) warming?
You seen to be getting mixed up. Actually Zhang et al 2010 shows virtually ice free with 2degC warming, but this is in the summer. Armour et al find a collapse of winter ice cover with between 4 and 6degC hemispheric annual average warming.

Sea ice formation is simply not inconsistent with an ocean warmed in the summer. Once the sun sets in winter temperatures drop. Over the ice free warmed ocean temperatures are held up lower in the atmosphere above the open water. But the net heat flux is upward, and as atmospheric temperatures cool it is massive.

On the slow transition thread I have shown calculations with a simple thermodynamic model of sea ice growth. The following uses a synthesised daily 2m temperature dataset derived from NCEP/NCAR monthly data over the ESS.



Heat flux is seen to be massive in the early autumn, with a delay to ice growth into mid October as the air temperatures exceed zero degC.

Consider the heat flux. The sum of heat flux for a 1m squared area from 6 October to 31 December is 4379W/m^2 over 92 days. 60x60x24x4379 = 378M joules for a 1m^2 area. 6 October being when the ice has formed.

The specific heat capacity of sea water near the surface is around 3985 J/kg K, the density is around 1027 kg/m^3, so to cool 1m cubed of sea water by 1degC (or Kelvin) you need to lose around 4M joules.

378MJ/m^3 K / 4MJ /m^3 = 94.5 metre depth cooled by 1 degC. And that is after the ice has formed, so temperature is already at -1.8degC for the surface. Before the ice forms heat loss is even greater, in winter over leads heat fluxes can exceed 400W/m^2 (Andreas 1979 via Alam & Curry). Bear in mind that the peak heat flux through ice is only 1/8 of such figures for open water in winter.
Alam & Curry.
http://curry.eas.gatech.edu/currydoc/Alam_JGR100.pdf

Once incoming solar radiation is gone the heat budget for the Arctic is upwards, cooling will occur. Over open water it is massive. Hence ice will form.

Using the same simple physical model it is possible to reformulate in terms of Freezing Degree Days (FDDs). Using such a formulation one can combine the impact of shortened melt season and winter warming.



Currently over the East Siberian Sea FDD is around 4000, in the 1980s it was about 5000, over the Central Arctic it is higher (over 5000). Over most of the pack you'd need to more than halve FDD to around 2000 to get out of the linear range of FDD/peak thickness. If winter ice only grows to around 1.5m across the pack then by that stage thickness will be such that summer thinning will largely wipe out the ice (apart from mechanically deformed ice off Canada). We have a long way to go before that happens and winter thickness falls enough such that regular total melt outs become the norm.

Note that in PIOMAS of grid boxes with an effective thickness of around 2m in April, only around 15% to 70% meltout totally to give open water. By 1.75m, this is 40% to 85%, which corresponds to 2000 FDDs. As mentioned above, the Central Arctic is currently around 5000 FDD and the ESS has fallen from 5000 FDD in the 1980s to 4000 FDD in recent years. There remains a lot of winter warming and later start to melting before we can expect to see winter thickness drop to levels where massive (85%) area loss is to be expected.

« Last Edit: October 12, 2014, 10:13:08 AM by ChrisReynolds »

jai mitchell

  • Nilas ice
  • Posts: 2357
    • View Profile
  • Liked: 207
  • Likes Given: 60
Re: Latest PIOMAS update (October)
« Reply #401 on: October 12, 2014, 06:10:32 PM »
Haiku of Futures Passed
My "burning embers"
are not tri-color bar graphs
+3C today

Phil.

  • Grease ice
  • Posts: 540
    • View Profile
  • Liked: 76
  • Likes Given: 11
Re: Latest PIOMAS update (October)
« Reply #402 on: October 13, 2014, 02:01:52 PM »
Once incoming solar radiation is gone the heat budget for the Arctic is upwards, cooling will occur. Over open water it is massive. Hence ice will form.


Chris have you accounted for incoming LW flux, into Nov that will still be ~200W/m^2?

ChrisReynolds

  • Nilas ice
  • Posts: 1764
    • View Profile
    • Dosbat
  • Liked: 20
  • Likes Given: 9
Re: Latest PIOMAS update (October)
« Reply #403 on: October 13, 2014, 04:40:52 PM »
Once incoming solar radiation is gone the heat budget for the Arctic is upwards, cooling will occur. Over open water it is massive. Hence ice will form.


Chris have you accounted for incoming LW flux, into Nov that will still be ~200W/m^2?

Downwelling longwave has typically been of such a order of magnitude in the past as the sample plots from ISCCP show on this page.
https://stratus.ssec.wisc.edu/projects/d1fluxes/d1fluxes.html
Under such a regime ice has formed, net flux is still upwards as evidenced by Autumn/Winter atmospheric cooling. The warming signature of recent years with large open water into the autumn is marked as a low level surface hugging warming, this warming will retard ice growth, however it is a warming above climatology, not always above freezing. So the process of ocean losing heat followed by ice growth still occurs. To halve the freezing degree days down to around 2000 and severely impact April thickness requires a substantial mix of late re-freeze and substantial winter warming.

In the 'even simpler model' from Semtner 1975, which I drew on in the above long comment, DLR is not included, the model depends on the temperature difference between the ice/ocean interface and the atmosphere/ice interface. See this thread:
http://forum.arctic-sea-ice.net/index.php/topic,933.150.html
Messages 150, 157, and 160 for details and comparison with PIOMAS over the East Siberian Sea.

jdallen

  • Young ice
  • Posts: 3410
    • View Profile
  • Liked: 650
  • Likes Given: 244
Re: Latest PIOMAS update (October)
« Reply #404 on: October 13, 2014, 09:15:59 PM »
Permit me to dovetail just a bit with Chris.

While I have spoken to heat levels, as important to the formation, or not, of ice is heat flow.

As has been demonstrated elsewhere, there already far and away exists enough heat in the Arctic Ocean to keep it ice free.  What as yet does not exist, is a mechanism by which it could be evenly distributed.

I will add also, while peak insolation at high latitudes may be as much as 200W/M2, total insolation being delivered is massively reduced over what it was just a few days ago. At 73N right now, total daily insolation amounts to barely 1KWH/ day. That's far too little to offset radiation out of the system.  The only things keeping temperatures high is heat from the ocean combined with the currently surprisingly high inputs of heat via weather from lower latitudes.

In fact, here's a pv calculator people can tinker with to get a sense of things.

http://pveducation.org/pvcdrom/properties-of-sunlight/calculation-of-solar-insolation

While intended to let one calculate power available for electrical generation, it is reasonable adaptable so one may get a "thumb nail sketch" of energy available as heat.  Multiply by 1-albedo, and you can get a sense of what is available to heat the system.
This space for Rent.

mark

  • New ice
  • Posts: 59
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: Latest PIOMAS update (October)
« Reply #405 on: October 13, 2014, 10:53:07 PM »
Its just a thought, but could it be that the slow warming of the seas and resultant reduction in ice volume over recent years has had an effect on the strength of the warm surface flow into the Arctic due to increased fresher water flow into the N Atlantic from the Labrador current. This should make a small change to the temperature and volume of the flow through the subpolar gyre. The fact that climatic temperature changes over Greenland are always substantially enhanced compared with other regions of the Arctic, it would seem that this flow is climatically very sensitive.

Over 30 years is there any evidence that this is the case (weakening of the sub polar gyre). As perhaps the recent recoveries are the result of long term sensitivity that is just catching up. Should this be the case then increased freezing in winter from a position of lower ice extent and volume (ie greater recovery) would lead to increasing saline flows into the Arctic Bottom Water and N Atlantic Deep water and a greater flow out of the Arctic between Scotland and Greenland. Eventually completing the cycle by pulling larger volumes of water back into the Arctic through the N Atlantic Drift, starting another warming period. Throw into this equation lower insolation due to lower solar energy, is this the reason for the signs of recovery now. If so how long is the cycle likely to last - if it is 20 years as some sources are implying, the next cycle is likely to come back with a bang
« Last Edit: October 13, 2014, 11:06:27 PM by mark »

jdallen

  • Young ice
  • Posts: 3410
    • View Profile
  • Liked: 650
  • Likes Given: 244
Re: Latest PIOMAS update (October)
« Reply #406 on: October 14, 2014, 10:12:30 AM »
Its just a thought, but could it be that the slow warming of the seas and resultant reduction in ice volume over recent years has had an effect on the strength of the warm surface flow into the Arctic due to increased fresher water flow into the N Atlantic from the Labrador current. This should make a small change to the temperature and volume of the flow through the subpolar gyre. The fact that climatic temperature changes over Greenland are always substantially enhanced compared with other regions of the Arctic, it would seem that this flow is climatically very sensitive.

Over 30 years is there any evidence that this is the case (weakening of the sub polar gyre). As perhaps the recent recoveries are the result of long term sensitivity that is just catching up. Should this be the case then increased freezing in winter from a position of lower ice extent and volume (ie greater recovery) would lead to increasing saline flows into the Arctic Bottom Water and N Atlantic Deep water and a greater flow out of the Arctic between Scotland and Greenland. Eventually completing the cycle by pulling larger volumes of water back into the Arctic through the N Atlantic Drift, starting another warming period. Throw into this equation lower insolation due to lower solar energy, is this the reason for the signs of recovery now. If so how long is the cycle likely to last - if it is 20 years as some sources are implying, the next cycle is likely to come back with a bang

Pondering what you are saying here, some purely qualitative thoughts.

Regarding fresh water flow - during the decline of the last glaciation there were some very distinct , and (possibly) in the case of the Dryas - catastrophic - changes in fresh water flow into the Arctic and North Atlantic by way of release of impounded melt water and considerable volume reduction melt from stagnant ice sheets.  The volumes to cause those changes was pretty extraordinary - thousands of KM3 over prompt periods.

The annual melt in the Arctic and off of Greenland is on that scale, but even though our endpoints - Volume at Max and Volume at Min - have driven lower, the average melt during the season remains reasonably constant.  We gain, and then lose reasonably close to the same volume each year, within about 20%.  That 20% while in scale of the ice is large, in terms of fresh water displacement is not.  We need to remember that additional 2-3K KM3 of meltwater is across several million square miles of arctic surface.  The net increase in surface fresh water is a matter of centimeters.  I don't think that is affecting the "conveyor" yet.  What *will* affect the conveyor most dramatically I think, is when the haline engine slows and stops because no ice is forming.

Compared to current flow also, the volume of exchange resulting from ice formation is trivial compared to over-all current volume.  Consider that the Gulf Stream/NA Drift system is moving 150,000,000 cubic meters a second past any given point near Newfoundland.  Farther on, the North Atlantic Drift drops to 10% of that figure - about 15,000,000 meters a second. That still adds up to a KM3 about every minute.  If my math and scale is correct, that's close to 1400 KM3/day moving steadily north - which annually adds up to 30+ times the volume of ice melted over an entire season. 

Not all of that makes it to the Arctic by any means, but it highlights this - the overall contribution of melt to drive this flow is quite small.  There are other mechanisms driving it completely independent of ice formation.

You make a reference to solar cycles as well.  I think it needs to be pointed out that the estimated variation in solar energy during those cycles is significantly less than the forcing caused by the additional CO2 added to the atmosphere by human activity over the last century and a half.  Numbers I see suggest a total variation of insolation during typical cycles being less than 1 watt/M2, generally less than 0.5/watt/M2.  As the change in forcing by CO2 is at least double that, it is highly doubtful a downswing in insolation will significantly impede the on-going increase in heat in the environment.

http://www.esrl.noaa.gov/gmd/aggi/aggi.fig3.png
This space for Rent.

mark

  • New ice
  • Posts: 59
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: Latest PIOMAS update (October)
« Reply #407 on: October 14, 2014, 12:57:11 PM »
Thanks very much for the reply - I wasnt alluding to a strong connection to solar cycles more that the reduced insolation may have a small effect. I perhaps come from a different viewpoint in that I like to look for evidence of balances within the system. I cant point to specific references but I do read as many of the references on this site (and others) as I can and Ive been looking at ASLR's very full thread as well as this one. There are just too many feedbacks to be able to narrow anything down. Each subject is fascinating and well supported with facts and figures, but as you quite rightly pointed out the ocean currents are absolutely collossal which is why It makes sense that very small changes to the flow can have big consequences, but all over a considerably longer period than just the last 18 years of levelling temperatures.

I havent got the depth of knowledge or the ability to 'see' the data as some on here do. But whether its freshwater (adequately rebuffed! thank you) or the reduced saline flows I have read too many times that the Sub Polar Gyre has slowed somewhat. So really I was asking if this is conjecture or whether it is fact based. If so is the present recovery purely down to meteorological coincidence or down to a shift in the way the gyres are transporting heat. Its a frustrating subject which is why I admire the persistence of the posters on this site.

I'm sorry if my response is a bit slow - I'm a 'newbie' so currently being moderated

ChrisReynolds

  • Nilas ice
  • Posts: 1764
    • View Profile
    • Dosbat
  • Liked: 20
  • Likes Given: 9
Re: Latest PIOMAS update (October)
« Reply #408 on: October 14, 2014, 10:27:04 PM »
I'll reply tomorrow night, it's now late and I'm whacked.

Actually nothing I feel I need to reply to.  :)

Night off.  8)
« Last Edit: October 15, 2014, 06:42:45 PM by ChrisReynolds »

viddaloo

  • Nilas ice
  • Posts: 1302
  • Hardanger Sometimes
    • View Profile
  • Liked: 1
  • Likes Given: 0
Re: Latest PIOMAS update (October)
« Reply #409 on: October 15, 2014, 05:29:09 AM »
I would like to see a statistical analysis of this data with a trend analysis for cumulative loss between June 1st and August 1st.  Then the deviation from the trend for 2014 can be more adequately addressed.

just by looking at the graph I am certain it surpasses the 2 sigma deviation from the trend (possibly 3!)

It would be absolutely foolhardy to imply that this extreme event can be anything indicative of a long-term trend.

however, it does indicate some kind of larger systemic shift, a "not-seen-before" event.  Either natural or unnatural it indicates a significant loss of boundary between the Ferrel and Arctic Cells  And/or some form of conscious manipulation of the atmospheric regimes.
This graph shows deviations from the 1979—2014 average for each day June 1st to August 1st, with deviations for first day, June 1st, leveled to zero and subtracted (for better comparison) for that year.

As you can see, there is a clear trend (2008, 2010, 2011, 2012, 2013, 2014) towards smaller losses after Solstice, which I've addressed in this thread. 2014 is the most extreme year in this trend so far. My personal theory is that of an increase in Boreal forest fires as the Arctic goes warmer, dryer and more snow–free, and that vast amounts of smoke from these fires reduce insolation (negative feedback).
[]

jai mitchell

  • Nilas ice
  • Posts: 2357
    • View Profile
  • Liked: 207
  • Likes Given: 60
Re: Latest PIOMAS update (October)
« Reply #410 on: October 15, 2014, 06:47:38 AM »
Thanks VID, but. . .

More specifically, I would like the SLOPES of the runs between June 1 and Aug 1 for the last 20 years.

Then a statistical analysis of the plots can be deduced with a standard deviation for the 2014 value to determine how far from the trend it is.

to do this I need the june 1 and aug 1 values of volume any idea where that data is?
Haiku of Futures Passed
My "burning embers"
are not tri-color bar graphs
+3C today

viddaloo

  • Nilas ice
  • Posts: 1302
  • Hardanger Sometimes
    • View Profile
  • Liked: 1
  • Likes Given: 0
Re: Latest PIOMAS update (October)
« Reply #411 on: October 15, 2014, 09:02:06 AM »
1979,31368,20179
1980,30925,19546
1981,28843,16610
1982,27330,16425
1983,29396,18585
1984,28872,17678
1985,29501,17471
1986,29533,18837
1987,30674,18510
1988,29074,17896
1989,28782,17882
1990,27484,16355
1991,28872,16722
1992,28464,17451
1993,28114,15016
1994,28464,16649
1995,26009,13999
1996,26464,16260
1997,27747,15974
1998,27611,15067
1999,26872,15090
2000,25675,13960
2001,25696,14919
2002,25722,13823
2003,24854,12992
2004,24474,13326
2005,23880,12125
2006,23076,11805
2007,21744,8964
2008,22787,11050
2009,22300,9698
2010,20100,7283
2011,19356,6882
2012,19416,6538
2013,20390,7705
2014,20081,9422
[]

crandles

  • Young ice
  • Posts: 3379
    • View Profile
  • Liked: 239
  • Likes Given: 81
Re: Latest PIOMAS update (October)
« Reply #412 on: October 15, 2014, 01:27:54 PM »
Losses:
11189
11379
12233
10905
10811
11194
12030
10696
12164
11178
10900
11129
12150
11013
13098
11815
12010
10204
11773
12544
11782
11715
10777
11899
11862
11148
11755
11271
12780
11737
12602
12817
12474
12878
12685
10659


Avg   11702
St deva   742

2014 from average 1043
This is just 1.4 standard deviations which could quite easily occur by chance and indeed there is a lower value 10204 for 1996.

Above just assumes no trend.

However there does seems an trend curving upwards. If you believe that pattern is as expected then average of 5 years 2009-2013 is 12691. 2014 differs from this by 2032 and a standard deviation from this trend would be less than 742 so it is looking like a 3SD event.

More precise calculation may well be possible but requires a decision on what type of rising pattern is justified.

jai mitchell

  • Nilas ice
  • Posts: 2357
    • View Profile
  • Liked: 207
  • Likes Given: 60
Re: Latest PIOMAS update (October)
« Reply #413 on: October 16, 2014, 07:28:53 AM »
Thanks JD and Crandals, I expect that we are dealing with a second order feedback analysis that includes a tipping point that was reached sometime around 2005 therefore the multi decadal analysis will hold too much deviation from the average (read: 1996 anomaly).  but The trend is definitely pushing 3-sig.

We may be operating in a dynamic environment that is so extreme that no trend analysis can be made, at least not for another decade or so.

Haiku of Futures Passed
My "burning embers"
are not tri-color bar graphs
+3C today

jdallen

  • Young ice
  • Posts: 3410
    • View Profile
  • Liked: 650
  • Likes Given: 244
Re: Latest PIOMAS update (October)
« Reply #414 on: October 16, 2014, 06:49:21 PM »
Looking at those volume numbers, what jumps out at me is the trend at maximum.  Since 2000, the volume at maximum has dropped 20%. I actually haven't been looking at Max trends that closely, so that surprised me.

That reflects the enthalpy increases in the system and is more key than minimum.  Less ice increases the pack vulnerability to weather,  and makes for greater volatility in system behavior.

What are our chances of working out regional volume changes at Max? That might give us some interesting clues.

Edit - quick and dirty purely statistical prediction from trend... Purely to provide basis for further discussion.

2015 volume max volume - 19500 +/- 500
2015 volume min volume  -   7825 +/- 350
2029 first "ice free" summer (sub 1000 KM3 pack)
2053 first "ice free winter (sub 1000 KM3 winter max pack)
« Last Edit: October 16, 2014, 07:30:28 PM by jdallen »
This space for Rent.

ChrisReynolds

  • Nilas ice
  • Posts: 1764
    • View Profile
    • Dosbat
  • Liked: 20
  • Likes Given: 9
Re: Latest PIOMAS update (October)
« Reply #415 on: October 16, 2014, 07:24:11 PM »
What are our chances of working out regional volume changes at Max? That might give us some interesting clues.

Regional PIOMAS volume data available via this page:
http://dosbat.blogspot.co.uk/2013/12/regional-piomas-volume-data.html

crandles

  • Young ice
  • Posts: 3379
    • View Profile
  • Liked: 239
  • Likes Given: 81
Re: Latest PIOMAS update (October)
« Reply #416 on: October 16, 2014, 07:26:56 PM »

jdallen

  • Young ice
  • Posts: 3410
    • View Profile
  • Liked: 650
  • Likes Given: 244
Re: Latest PIOMAS update (October)
« Reply #417 on: October 16, 2014, 07:59:31 PM »
More thoughts I can't resist posting.

Crandles, Looking at your annual volume variation numbers (losses), I'm struck by how consistent the annual arctic energy budget is.  Considering even the small size of the sample, there are no +2SD outliers.  Even more interesting is the SD range itself... Even the recent "big" melt years are inside the 1SD envelope, which is about 6%.

That confirms for me the inputs into the system are increasing incrementally, at very small values. The rate at which heat is shed during refreeze remains constant. The primary force driving the downward trend is the increasing residual heat in the system at the end of the melt season, not weather or increasing temperatures directly.
« Last Edit: October 16, 2014, 08:51:15 PM by jdallen »
This space for Rent.

crandles

  • Young ice
  • Posts: 3379
    • View Profile
  • Liked: 239
  • Likes Given: 81
Re: Latest PIOMAS update (October)
« Reply #418 on: October 16, 2014, 11:40:25 PM »
You do realise those are 1 June and 1 Aug volumes (and the change between those dates).

Also
2010,20100
2011,19356
2012,19416
2013,20390
2014,20081

doesn't seem to be reducing much recently.

Also
11702-742*2 = 10218 so 10204 is outside 2 sd

Expect 1 in 20 items to be outside 2 sd. So in a list of 36 would normally expect to see 2 such values but just 1 is not wildly different from the expectation.
« Last Edit: October 16, 2014, 11:49:43 PM by crandles »

jdallen

  • Young ice
  • Posts: 3410
    • View Profile
  • Liked: 650
  • Likes Given: 244
Re: Latest PIOMAS update (October)
« Reply #419 on: October 17, 2014, 03:00:34 AM »
You do realise those are 1 June and 1 Aug volumes (and the change between those dates).

Also
2010,20100
2011,19356
2012,19416
2013,20390
2014,20081

doesn't seem to be reducing much recently.

Also
11702-742*2 = 10218 so 10204 is outside 2 sd

Expect 1 in 20 items to be outside 2 sd. So in a list of 36 would normally expect to see 2 such values but just 1 is not wildly different from the expectation.
<wry look>  Thank you for the detail correction.

No, no significant  change over the last few years - oscillation mostly.

I think my point still stands - the drop over 14 years - 25% - I think is quite significant.  I think it still points at net energy in the system being what is driving the long term trend, rather than changes in weather.

That increase in net energy is eventually tied back to the forcing from CO2.  Now, to look at some regional numbers from the link to Chris's stuff you sent out earlier.
This space for Rent.

crandles

  • Young ice
  • Posts: 3379
    • View Profile
  • Liked: 239
  • Likes Given: 81
Re: Latest PIOMAS update (October)
« Reply #420 on: October 17, 2014, 01:06:10 PM »
I think my point still stands - the drop over 14 years - 25% - I think is quite significant.  I think it still points at net energy in the system being what is driving the long term trend, rather than changes in weather.

[point being]
That confirms for me the inputs into the system are increasing incrementally, at very small values. The rate at which heat is shed during refreeze remains constant. The primary force driving the downward trend is the increasing residual heat in the system at the end of the melt season, not weather or increasing temperatures directly.

In the 1 June figures, I think I would suggest the pattern is:
5k drop from 1980 to 2002 22 years
3k drop in 5 years
2k drop in 7 years

2002 to 2007 shows a fast drop between slower sections. The fast drop seems like MYI became FYI or changed thickness so little over 2m. As thickness reduces towards 2m, it can no longer drop at a fast rate as winter restores that thickness. So the rate of drop reduces and it becomes more like oscillation.

So while you may well be correct to deduce incremental change over the 36 year period, the validity of that in the last few years and going forward is much more questionable (in my opinion)

crandles

  • Young ice
  • Posts: 3379
    • View Profile
  • Liked: 239
  • Likes Given: 81
Re: Latest PIOMAS update (October)
« Reply #421 on: October 17, 2014, 02:02:32 PM »
https://sites.google.com/site/arctischepinguin/home/piomas/grf/piomas-trnd2.png

now suggests Sept 2019 for zero ice, (2018 for 1k km^3).

Wadhams at RS meeting on 22/23 September 2014 showed old image* with zero as soon as 2015. See page 10 onwards for discussion of this:
https://drive.google.com/file/d/0BzOmEySKhRVOc1prYUNLQXhoa3c/view

* couldn't be expected to include 2014 data at this was in Sept 14 but Sept 13 should have been included.

Quote
COMMENTARY This is the source of the data set around which Peter Wadhams has built his notable and controversial prediction of the disappearance of sea ice. Brandon tweeted that it is linked to data thus making the point that it is not just a computational model but has basis in measurements. However, note that the data originators – when questioned later in the week at the follow on Discussion Meeting – do not believe prediction such as Wadhams proceeds to do is appropriate.

Note that it is Wadhams that has made complaints about tweets during this meeting.
Also PIOMAS team have made that clear before.

One of the tweets was
Quote
Wadhams still using graphs with ridiculous projections with no basis in physics

It seems this is being robustly defended
Quote
There has been wide criticism of this methodology since no convincing physical basis for such behaviour has been presented.

Disagreement? Yes. Grounds for complaint? Clearly not AFAICS

ChrisReynolds

  • Nilas ice
  • Posts: 1764
    • View Profile
    • Dosbat
  • Liked: 20
  • Likes Given: 9
Re: Latest PIOMAS update (October)
« Reply #422 on: October 17, 2014, 07:33:50 PM »

jdallen

  • Young ice
  • Posts: 3410
    • View Profile
  • Liked: 650
  • Likes Given: 244
Re: Latest PIOMAS update (October)
« Reply #423 on: October 17, 2014, 09:17:26 PM »
So while you may well be correct to deduce incremental change over the 36 year period, the validity of that in the last few years and going forward is much more questionable (in my opinion)

Reasonable criticisms, worthy of further thought.

After spending time wit the regional 2000-2014 volume data Chris supplied, I'm struck by how asymmetric the changes are over time.  The change in behavior over the last 7-10 years seems very much tied to volatility in peripheral seas, which are melting out completely with increasing frequency.  However, given the right weather, they are still close enough to some systemic threshold that we observe good retention (relatively) and wild swings in metrics.

Each peripheral region in turn, save Greenland and the CAA, has dropped in peak volume to the point that the typical, not exceptional, spring and summer melt take them to zero volume.  Where ice remains is by way of reductions of normal energy input, or transport (in the case of the Barents specifically) from other regions.  I think this volatility may hide a trend, as each region year over year more often than not have experienced significant variations in weather, both over time with themselves season over season, and with each other, during the same season.   2014 may be our poster child for this.

So, while your criticisms require consideration, there may exist reasonable explanations for your questions.
This space for Rent.

viddaloo

  • Nilas ice
  • Posts: 1302
  • Hardanger Sometimes
    • View Profile
  • Liked: 1
  • Likes Given: 0
Re: Latest PIOMAS update (October)
« Reply #424 on: November 05, 2014, 08:42:23 PM »
Funny how PIOMAS want my email address every time I download data, but never use said address to notify of new data dumps. Nor for alerts that the data will be delayed. Let's hope the new Senate doesn't make them even more cash–strapped. Seem to be on a shoestring budget as it is.
[]

viddaloo

  • Nilas ice
  • Posts: 1302
  • Hardanger Sometimes
    • View Profile
  • Liked: 1
  • Likes Given: 0
Re: Latest PIOMAS update (October)
« Reply #425 on: November 07, 2014, 09:18:13 PM »
Looks like the plan is to call it a week, with no dump from PIOMAS till Monday the 10th. Just wondering whether PIOMAS data has ever been this late before? And same thing for IJIS extent: The update for November 6th is now 17 hours (and running) late, after their server crash.

Neither source seems to believe in the use of information architecture to inform about unforeseen events or delays. They don't use the web for this, they don't use Twitter for this and as mentioned above: Even when they harvest your email address, they don't use this to inform you once a month when new data is available. I wonder why.
[]

viddaloo

  • Nilas ice
  • Posts: 1302
  • Hardanger Sometimes
    • View Profile
  • Liked: 1
  • Likes Given: 0
Re: Latest PIOMAS update (October)
« Reply #426 on: November 08, 2014, 03:02:47 AM »
Fresh PIOMAS Volume numbers for October 2014:

Code: [Select]
2014 274   7.154
2014 275   7.191
2014 276   7.227
2014 277   7.284
2014 278   7.334
2014 279   7.382
2014 280   7.434
2014 281   7.486
2014 282   7.548
2014 283   7.614
2014 284   7.691
2014 285   7.773
2014 286   7.849
2014 287   7.939
2014 288   8.020
2014 289   8.090
2014 290   8.172
2014 291   8.278
2014 292   8.383
2014 293   8.481
2014 294   8.598
2014 295   8.691
2014 296   8.779
2014 297   8.850
2014 298   8.947
2014 299   9.027
2014 300   9.114
2014 301   9.224
2014 302   9.362
2014 303   9.520
2014 304   9.701
« Last Edit: November 11, 2014, 07:43:48 AM by viddaloo »
[]

viddaloo

  • Nilas ice
  • Posts: 1302
  • Hardanger Sometimes
    • View Profile
  • Liked: 1
  • Likes Given: 0
Re: Latest PIOMAS update (October)
« Reply #427 on: November 08, 2014, 03:42:33 AM »
2014 gained 0.4k km³ in October compared to the late 20th century average gain. That places 2014 ahead of the current trend of increased October gains, just above the orange 11/14 average.

« Last Edit: November 08, 2014, 03:59:16 AM by viddaloo »
[]

Wipneus

  • Citizen scientist
  • Young ice
  • Posts: 4220
    • View Profile
    • Arctische Pinguin
  • Liked: 1025
  • Likes Given: 0
Re: Latest PIOMAS update (November)
« Reply #428 on: November 08, 2014, 08:47:43 AM »
Updated, graphs are in the top post.

(gridded data updated also, I will post the graphs later)

Wipneus

  • Citizen scientist
  • Young ice
  • Posts: 4220
    • View Profile
    • Arctische Pinguin
  • Liked: 1025
  • Likes Given: 0
Re: Latest PIOMAS update (November)
« Reply #429 on: November 08, 2014, 05:13:10 PM »
Here are thickness maps for October 2013 and 2014 as well as their difference.

Wipneus

  • Citizen scientist
  • Young ice
  • Posts: 4220
    • View Profile
    • Arctische Pinguin
  • Liked: 1025
  • Likes Given: 0
Re: Latest PIOMAS update (November)
« Reply #430 on: November 08, 2014, 05:16:02 PM »
And difference in growth from September to October, red means more thickening/less thinning.

viddaloo

  • Nilas ice
  • Posts: 1302
  • Hardanger Sometimes
    • View Profile
  • Liked: 1
  • Likes Given: 0
Re: Latest PIOMAS update (November)
« Reply #431 on: November 10, 2014, 05:50:47 PM »
I'm quite happy with the accuracy of my PIOMAS estimate formula this month. 20 days out of 31 it was less than 2% off from the real thing. Here's how I did it, for those of you who speak PHP:

Code: [Select]
if($i>$currentmonthstart-1)
$piomas[$year][$i] = $ijis[$year][$i]/( ($ijis[$year-5][$i]/$piomas[$year-5][$i])
- (($ijis[$year-5][$currentmonthstart-1]/$piomas[$year-5][$currentmonthstart-1])
-($ijis[2014][$currentmonthstart-1]/$piomas[2014][$currentmonthstart-1])) );

In short, it uses the rate between extent and volume in October 5 years ago, which is known, to get volume (unknown) from another known variable; the current month extent data. It assumes the gap between the ext/vol rate that year and this year will be roughly the same through the month, and then follows that 2009 ext/vol rate graph up the peak and down again, as we move through October.

Why pick just 2009? Well, it's a spin-off from the Five Year Cycle hypothesis, that's why. Could another year have yielded even better estimates? Possibly, but 2009 is much more similar to 2014 when it comes to the rate between extent and volume, among many other things.

Will it work just as well in November? Only time will tell.
« Last Edit: November 17, 2014, 10:49:56 PM by viddaloo »
[]

ChrisReynolds

  • Nilas ice
  • Posts: 1764
    • View Profile
    • Dosbat
  • Liked: 20
  • Likes Given: 9
Re: Latest PIOMAS update (November)
« Reply #432 on: November 10, 2014, 09:28:49 PM »
Thanks Wipneus,

There seems to be a starting of Beaufort Gyre and Fram Strait export from the Central Arctic.

jdallen

  • Young ice
  • Posts: 3410
    • View Profile
  • Liked: 650
  • Likes Given: 244
Re: Latest PIOMAS update (November)
« Reply #433 on: November 10, 2014, 09:41:10 PM »
Thanks Wipneus,

There seems to be a starting of Beaufort Gyre and Fram Strait export from the Central Arctic.

... which may herald a halt in the volume increases we've seen.  It will be interesting to track the movement out of the Fram.  Lack of Fram export was pretty key to MYI retention last summer.
This space for Rent.

ChrisReynolds

  • Nilas ice
  • Posts: 1764
    • View Profile
    • Dosbat
  • Liked: 20
  • Likes Given: 9
Re: Latest PIOMAS update (November)
« Reply #434 on: November 11, 2014, 06:58:39 PM »
My thoughts exactly JD.

crandles

  • Young ice
  • Posts: 3379
    • View Profile
  • Liked: 239
  • Likes Given: 81
Re: Latest PIOMAS update (November)
« Reply #435 on: December 05, 2014, 10:47:54 PM »
30Nov graphs now out:

viddaloo

  • Nilas ice
  • Posts: 1302
  • Hardanger Sometimes
    • View Profile
  • Liked: 1
  • Likes Given: 0
Re: Latest PIOMAS update (November)
« Reply #436 on: December 06, 2014, 05:49:26 AM »
Code: [Select]
2014 305   9.860
2014 306  10.020
2014 307  10.161
2014 308  10.319
2014 309  10.450
2014 310  10.576
2014 311  10.661
2014 312  10.751
2014 313  10.837
2014 314  10.927
2014 315  10.999
2014 316  11.075
2014 317  11.164
2014 318  11.307
2014 319  11.402
2014 320  11.475
2014 321  11.600
2014 322  11.720
2014 323  11.850
2014 324  11.978
2014 325  12.113
2014 326  12.238
2014 327  12.361
2014 328  12.494
2014 329  12.635
2014 330  12.789
2014 331  12.916
2014 332  13.048
2014 333  13.184
2014 334  13.314
[]

Wipneus

  • Citizen scientist
  • Young ice
  • Posts: 4220
    • View Profile
    • Arctische Pinguin
  • Liked: 1025
  • Likes Given: 0
Re: Latest PIOMAS update (December)
« Reply #437 on: December 06, 2014, 08:24:33 AM »
Updated, graphs are in the top post.

(gridded data updated also, I will post the graphs later)

Wipneus

  • Citizen scientist
  • Young ice
  • Posts: 4220
    • View Profile
    • Arctische Pinguin
  • Liked: 1025
  • Likes Given: 0
Re: Latest PIOMAS update (December)
« Reply #438 on: December 06, 2014, 08:46:09 AM »
Here are thickness maps for November 2013 and 2014 as well as their difference.

Wipneus

  • Citizen scientist
  • Young ice
  • Posts: 4220
    • View Profile
    • Arctische Pinguin
  • Liked: 1025
  • Likes Given: 0
Re: Latest PIOMAS update (December)
« Reply #439 on: December 06, 2014, 08:47:44 AM »
And difference in growth from October to November, red means more thickening/less thinning.

viddaloo

  • Nilas ice
  • Posts: 1302
  • Hardanger Sometimes
    • View Profile
  • Liked: 1
  • Likes Given: 0
Re: Latest PIOMAS update (December)
« Reply #440 on: December 06, 2014, 11:03:33 AM »
Updated collapse estimate.
« Last Edit: December 09, 2014, 10:09:57 PM by viddaloo »
[]

OSweetMrMath

  • Frazil ice
  • Posts: 137
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: Latest PIOMAS update (December)
« Reply #441 on: December 06, 2014, 08:48:26 PM »
Updated collapse estimate.

Your graph does not predict that there will be no Arctic sea ice year round in 2031. Your graph predicts that the average ice volume in 2031 will be 0.

The graph has a built in assumption that for approximately half the year, the ice volume will be below the stated average, and for the other half, it will be above the stated average. This does not change as the average approaches zero.

Effectively, the graph predicts that in 2031, the Arctic ice volume will be negative for six months out of the year.

viddaloo

  • Nilas ice
  • Posts: 1302
  • Hardanger Sometimes
    • View Profile
  • Liked: 1
  • Likes Given: 0
Re: Latest PIOMAS update (December)
« Reply #442 on: December 06, 2014, 10:15:46 PM »
Updated collapse estimate.

Your graph does not predict that there will be no Arctic sea ice year round in 2031. Your graph predicts that the average ice volume in 2031 will be 0.

The graph has a built in assumption that for approximately half the year, the ice volume will be below the stated average, and for the other half, it will be above the stated average. This does not change as the average approaches zero.

Effectively, the graph predicts that in 2031, the Arctic ice volume will be negative for six months out of the year.

Well, no. Actually, an inch of ice on just half a day will keep the average from going to zero.
[]

OSweetMrMath

  • Frazil ice
  • Posts: 137
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: Latest PIOMAS update (December)
« Reply #443 on: December 06, 2014, 10:30:38 PM »
Well, no. Actually, an inch of ice on just half a day will keep the average from going to zero.

Then I don't understand how you are computing the prediction. It looks like you are computing the average each day over the previous 12 months, giving you the data in your graph. You are then fitting a curve to these average values. This curve reaches 0 in 2031.

If that is correct, then the 0 value for the curve means that the average over the 12 previous months is 0, not that every day for the previous 12 months is exactly 0. Unless you are directly compensating for the fact that the ice volume cannot be less than zero on any given day when you are fitting that curve, it builds in the idea that half the time, the ice volume will be below average.

When a large fraction of the days in the year have 0 ice, it becomes harder to reduce the average over the entire year. This should show up in a flattening of your curve as it approaches zero. I take the lack of flattening as evidence that you are averaging assuming negative ice rather than zero ice.

viddaloo

  • Nilas ice
  • Posts: 1302
  • Hardanger Sometimes
    • View Profile
  • Liked: 1
  • Likes Given: 0
Re: Latest PIOMAS update (December)
« Reply #444 on: December 06, 2014, 10:36:00 PM »
I don't know why many of you assume such things. IMO none of you have explained this sufficiently. In 2013 you had 365 days. Each of those days had their total ice volume figure. Add day 1 + day 2 ... + day 365 into a year total and divide by 365. Et voila!
[]

ChrisReynolds

  • Nilas ice
  • Posts: 1764
    • View Profile
    • Dosbat
  • Liked: 20
  • Likes Given: 9
Re: Latest PIOMAS update (December)
« Reply #445 on: December 06, 2014, 11:29:15 PM »

OSweetMrMath

  • Frazil ice
  • Posts: 137
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: Latest PIOMAS update (December)
« Reply #446 on: December 07, 2014, 02:07:38 AM »
I shouldn't have said anything. My mistake.

Unmex Chingon

  • New ice
  • Posts: 31
    • View Profile
  • Liked: 3
  • Likes Given: 0
Re: Latest PIOMAS update (December)
« Reply #447 on: December 07, 2014, 05:25:30 PM »
I don't know why many of you assume such things. IMO none of you have explained this sufficiently. In 2013 you had 365 days. Each of those days had their total ice volume figure. Add day 1 + day 2 ... + day 365 into a year total and divide by 365. Et voila!

And if you wanted to calculate Day 2 of 2014 - you would start with Day 2 of 2013 + day 3...+day 365 2013 + day 1 2014 and divide all that by 365 - Right?  If that is how you calculate then - I have a question for you though - in 2031 this calculation will result in Zero?

viddaloo

  • Nilas ice
  • Posts: 1302
  • Hardanger Sometimes
    • View Profile
  • Liked: 1
  • Likes Given: 0
Re: Latest PIOMAS update (December)
« Reply #448 on: December 08, 2014, 06:18:01 AM »
I don't know why many of you assume such things. IMO none of you have explained this sufficiently. In 2013 you had 365 days. Each of those days had their total ice volume figure. Add day 1 + day 2 ... + day 365 into a year total and divide by 365. Et voila!

And if you wanted to calculate Day 2 of 2014 - you would start with Day 2 of 2013 + day 3...+day 365 2013 + day 1 2014 and divide all that by 365 - Right?  If that is how you calculate then - I have a question for you though - in 2031 this calculation will result in Zero?

That is correct, well, almost: The AAV for day 2014/02 is (Day 3 of 2013 + day 4...+day 365 2013 + day 1 2014 + day 2 2014) divided by 365.

Of course I can't say if that calculation will result in zero in 2031, as the data for 2030 and 2031 aren't really in yet! What I can say is that the trend we see in AAV from 1980 to 2014 seems to end rather horribly in zero sometime during 2031 or the early thirties.

If you look at the purple part of the graph, that is the start of that move towards zero. And from what I can see in the purple real data part, we are almost half–way to zero in ice volume, and if the trend is correct, we're about 70% of the way to zero measured in time.
[]

Unmex Chingon

  • New ice
  • Posts: 31
    • View Profile
  • Liked: 3
  • Likes Given: 0
Re: Latest PIOMAS update (December)
« Reply #449 on: December 08, 2014, 06:24:21 AM »
I don't know why many of you assume such things. IMO none of you have explained this sufficiently. In 2013 you had 365 days. Each of those days had their total ice volume figure. Add day 1 + day 2 ... + day 365 into a year total and divide by 365. Et voila!

And if you wanted to calculate Day 2 of 2014 - you would start with Day 2 of 2013 + day 3...+day 365 2013 + day 1 2014 and divide all that by 365 - Right?  If that is how you calculate then - I have a question for you though - in 2031 this calculation will result in Zero?

That is correct, well, almost: The AAV for day 2014/02 is (Day 3 of 2013 + day 4...+day 365 2013 + day 1 2014 + day 2 2014) divided by 365.

Of course I can't say if that calculation will result in zero in 2031, as the data for 2030 and 2031 aren't really in yet! What I can say is that the trend we see in AAV from 1980 to 2014 seems to end rather horribly in zero sometime during 2031 or the early thirties.

If you look at the purple part of the graph, that is the start of that move towards zero. And from what I can see in the purple real data part, we are almost half–way to zero in ice volume, and if the trend is correct, we're about 70% of the way to zero measured in time.

So your projection is that the Artic will be with Zero ice for 12 consecutive months?  No refreeze at all during the winter seems to be a very bold prediction.