Messages

1

Arctic sea ice / Re: Northern Sea Route thread

« on: October 22, 2020, 06:37:38 PM »

So when do we think it'll close this year? Before or after the US election?

2

Arctic sea ice / Re: The 2020/2021 freezing season

« on: October 10, 2020, 07:04:18 PM »

Interesting figures, S.Pansa!  Is there a particular reason why, in Figure 6-5, the incoming and outgoing radiation don't seem to net to zero globally?  (Just eyeballing it, it looks like there's no way the areas under the curves match).

Remember that the total area from (say) 0-10 degrees north is MUCH larger than the area from 80-90 degrees north.

3

Arctic sea ice / Re: "Smart" and "Stupid" Questions - Feel Free To Ask

« on: October 04, 2019, 12:15:38 AM »

Pumice is also foamy with a lot of trapped air, which makes it an excellent insulator.

4

Arctic sea ice / Re: The 2019 melting season

« on: August 15, 2019, 11:55:22 AM »

I found these graphs, and after 2012, the sea ice extent in the summer months has been relatively steady above trend. You'd expect at least one month to have an extreme, but I don't see it. Could bathymetry have something to do with this? Has a threshold been reached?

http://ocean.dmi.dk/arctic/icecover_30y.uk.php

Er, no?  Pretty much all the months are exactly on the trend, with 2007 and 2012 being below it.  The only point where 2019 is off-trend is the one for August, with a ? beside it. As you may or may not be aware, we are only half way through August, so this figure is inaccurate.

5

Arctic sea ice / Re: When will the Northwest Passage "open" in 2019?

« on: August 12, 2019, 02:37:27 PM »

1. Don't care what the definition is bc I do not now, and never have, agreed with it, thus stated I was stating my perspective from my very first post. Ergo, I cannot be wrong within that context. I am not attempting to conform.

6

Arctic sea ice / Re: The caa-greenland mega crack

« on: July 30, 2019, 10:12:21 AM »

The phrase you're looking for is "flaw lead".

7

Arctic sea ice / Re: When will the Arctic Go Ice Free?

« on: July 29, 2019, 01:23:18 PM »

Not free. I followed your link, had to log in ( give information) and then directly request access to the paper. Haven't received it yet.

Here is the abstract of the paper:

During recent decades, there has been dramatic Arctic sea ice retreat. This has reduced the top‐of‐atmosphere albedo, adding more solar energy to the climate system. There is substantial uncertainty regarding how much ice retreat and associated solar heating will occur in the future. This is relevant to future climate projections, including the timescale for reaching global warming stabilization targets. Here we use satellite observations to estimate the amount of solar energy that would be added in the worst‐case scenario of a complete disappearance of Arctic sea ice throughout the sunlit part of the year. Assuming constant cloudiness, we calculate a global radiative heating of 0.71 W/m2 relative to the 1979 baseline state. This is equivalent to the effect of one trillion tons of CO2 emissions. These results suggest that the additional heating due to complete Arctic sea ice loss would hasten global warming by an estimated 25 years.

From here:

https://agupubs.onlinelibrary.wiley.com/doi/abs/10.1029/2019GL082914

... or just use Sci-Hub to find papers <cough>

Looking at it...  this is a self evidently ridiculous paper.  It's a simple albedo calculation, based on the complete absence of ice from March through to September.

From their methods:

In the calculations of albedo and radiative heating presented here, we use data from the National Aeronautics and Space Administration's CERES Terra SSF Edition 4 monthly averaged 1 × 1-degree product, between March 2000 and October 2016, available online (https://ceres.larc.nasa.gov/order_data.php). The Arctic Ocean is defined here as the land-free area poleward of 60◦N. Due to issues concerning polar night, we only consider the months of March to September of each year.

i.e. they account for extra incoming energy during an ice-free summer, but do not account for any extra outgoing energy during the winter.  This is effectively equivalent to assuming that the Arctic freezes over as normal each winter (i.e. insulating the surface and restricting heat loss) and then the ice magically disappears overnight some time in March.

It's completely unphysical and the only utility it has is giving a ballpark figure for HALF the energy equation resulting from an ice-free Arctic. Don't waste any time trying to interpret this one any further than that.

8

Arctic sea ice / Re: When will the Arctic Go Ice Free?

« on: July 15, 2019, 02:42:53 PM »

BZZT - ambiguous pronoun referents detected!

"It is worse than we thought" = the "we" refers to the climate science community.

"It isn't as bad as we feared" = the "we" refers to members of this board.

Both statements are thus true, because this board is much more alarmist that the community in general.

9

Arctic sea ice / Re: "Smart" and "Stupid" Questions - Feel Free To Ask

« on: July 14, 2019, 10:57:35 PM »

It's worth appreciating that even natural convection generates lateral movement as well as vertical - how could it not?  Although the "engine" for a convection current is the vertical motion of fluid due to thermal buoyancy, something has to come in from the sides to replace the rising/falling fluid.

So (e.g.) in a cyclone, air is rising in the centre by convection, forcing air at low levels to flow inwards (given a spiral twist by Coriolis effects) and air at higher levels to flow outwards (ditto).  All of these elements of the motion, whether vertical or horizontal, transfer heat.

10

Arctic sea ice / Re: "Smart" and "Stupid" Questions - Feel Free To Ask

« on: July 14, 2019, 10:08:58 PM »

The clue is in the etymology: con+vection = with+movement. The molecules travel with the heat and vice versa.  This is also why a fan oven is called a convection oven.

11

Arctic sea ice / Re: "Smart" and "Stupid" Questions - Feel Free To Ask

« on: July 14, 2019, 10:03:43 PM »

Convection is when heat is transported by fluid flow.

If the fluid flow is itself caused by the heat (e.g. warm water rising from the bottom of a saucepan, cold water sinking from melting ice) then it is called "free heat convection".  If the fluid flow is caused by something else (e.g. I squirt a warm stream of liquid into a cold pool) then it is called "forced heat convection".

In both cases, heat is transported by moving hot molecules from one location to another, rather than by transferring energy from one molecule to another.

12

Arctic sea ice / Re: 2019 sea ice area and extent data

« on: July 04, 2019, 10:14:45 PM »

am i the only one who thinks that i cannot be a coincidence that several time each year, around special events like extent falling below 2012/2016, minima, maxima or other extremes, some of the data providers stop delivering.

Probably not, there's a lot of nutters out there.

https://i.kym-cdn.com/entries/icons/facebook/000/016/146/eird-al-yankovic-tin-foil-hat-160x160.jpg

13

Arctic sea ice / Re: Freeform season chatter and light commentary

« on: July 03, 2019, 02:34:55 PM »

A-Team posted this back on November 24, 2017 on the Ice Apocalypse thread.  I think it is a little funny yet worth another read considering the shape of the Arctic Ice today and should cause us to pause a moment and consider the implications.  It also parallels gerontocrat's "Perils of Projections."

Another parable I find of use in my own field of molecular biology is to consider how your measurement protocols affect the thing you're measuring.

When the blind men studied an elephant, they could not agree what it was like.  They said:
"An elephant is like a rope"
"An elephant is like a treetrunk"
"An elephant is like a wall"
"An elephant is like a snake"
"An elephant is like a sail"
Nevertheless, by synthesising their points of view, they were able to approximate the truth.


In contrast, when the blind elephants studied a man, they were unanimous:
"Men are flat"

14

Arctic sea ice / Re: Global sea ice area and extent data

« on: June 17, 2019, 09:03:41 PM »

patiencia, calma...

Hmm i don't think humans are ever going to see 18m sq km of seaice area again.

This is a bit silly.  Apart from one exceptional year, the Nov peak is always higher than the June peak, and the June peak was over 18m.

I think it is silly to assume we are only going to see one exceptional year.

True, but for your "never again" to be true, you're expecting every year from now on to be exceptional....

15

Arctic sea ice / Re: 2019 sea ice area and extent data

« on: June 17, 2019, 01:08:36 PM »

So why the sudden, weird gain in extent?  A rapid flash refreezing of this thin layer of surface water on the fast ice (which might occur simply if temps drop overnight [...]

Agreed that this could be local temperature variations (or sun angle, or cloud interference).  However it's not day/night effects. It's almost Midsummer Day and you're looking north of the Arctic Circle.  There IS no "overnight", the sun is up 24/7.

16

Arctic sea ice / Re: The 2019 melting season

« on: June 17, 2019, 01:00:37 PM »

My question concerned the other blue line on that plot, the LIGHT blue line labelled "Anom. Forecast", which may well be short for "Anomaly Forecast". I have no idea how this line was derived.

The "anomaly forecast" simply assumes that the anomaly stays constant, i.e. if we are currently 2 million below average, then in 50 days' time we will still be 2 million below average. Note that the anomaly in this case measured relative to the 1988-2013 average.

From the "About the Plots" section:

http://cires1.colorado.edu/~aslater/SEAICE/

"Anomaly Persistence uses data from 1988-2013 as the mean state."

17

Arctic sea ice / Re: Global sea ice area and extent data

« on: June 17, 2019, 12:51:35 PM »

patiencia, calma...

Hmm i don't think humans are ever going to see 18m sq km of seaice area again.

This is a bit silly.  Apart from one exceptional year, the Nov peak is always higher than the June peak, and the June peak was over 18m.

18

Arctic sea ice / Re: Poll: Beaufort Polynya connecting to Bering Strait open water

« on: June 14, 2019, 10:02:34 PM »

so my meaning applies to all kinds of boats an ships that dare to cruise the arctic in the first place.
hope that's more precise

Well quite.  You wouldn't find anyone paddling a one-man craft in those waters, they'd freeze to death.  After all, you can't have your kayak and heat it.

19

Arctic sea ice / Re: "Smart" and "Stupid" Questions - Feel Free To Ask

« on: April 25, 2019, 10:05:52 AM »

Do you know if such a chemical reaction is possible binntho? If you are out at sea on a platform whose details are irrelevant for now, getting electricity from a source that is irrelevant for now, grabbing pure concentrated CO2 with a technology that is irrelevant for now, could that CO2 be made solid and dumped over board and not dissolve in the sea water or anything like that?

Sure.  Grow a tree.

20

Arctic sea ice / Re: "Smart" and "Stupid" Questions - Feel Free To Ask

« on: April 24, 2019, 01:11:00 PM »

If you want to remove heat from the ocean, that heat has to go somewhere - and it has to go somewhere that is colder than the ocean. You could melt the ice caps, using Greenland and Antarctica as the heat sinks. This is a bad idea. Or you can let the energy radiate out to space, using the cosmic microwave background as a heat sink. This is a better idea, but requires us to stop adding insulating gases that get in the way of transferring heat from our planet into space.

21

Arctic sea ice / Re: "Smart" and "Stupid" Questions - Feel Free To Ask

« on: April 24, 2019, 01:07:51 PM »

If by "excess energy" you mean the small temperature increase, then no - there is no known or exptected future technology that can harvest that energy in any meaningful way.

Yes i mean that the majority of heat being trapped is in our oceans. Wave energy conversion is surely not an impossible technology. https://en.wikipedia.org/wiki/Salter%27s_duck#Energy_efficiency that's a link to one option.

It is an impossible technology.  Specifically, it breaks the second law of thermodynamics.

All energy ultimately ends up as heat.  Whenever you convert heat into some other form of energy (kinetic energy or motion, chemical potential energy, or whatever), you inevitable generate more heat in the process.

What you are asking for is a way of converting the heat energy of the oceans into stored chemical energy of bound CO2, with no further heat loss to the surroundings such that you end up with a net loss of heat.  This is impossible.

22

Arctic sea ice / Re: Latest PIOMAS update (April 2019)

« on: April 16, 2019, 12:52:55 PM »

I think an exponential was the perfect fit from about 2002 to 2012. After and before that period linear fits were better. Sometime in the near future exponential fits will become the better fit again.

That suggests to me that the "bound" the exponential trend ran into (and bounced off of) was 3-4000KM2 rather than zero.

Let me once again point to Tietsche et al 2011.
https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2010GL045698

DO NOT LOOK AT THE TIME FRAME
DO NOT LOOK AT THE SPECIFIC "ICE FREE" THOUGHT EXPERIMENT

The point to make here is that Figure 1 shows the "shape" of the decline towards zero, as predicted by a full climate model with atmosphere, ocean, geography and bathymetry included.  It is not linear, exponential, Gompertz or any such simple shape.  It is conditioned by the ocean currents, land masses and airflow patterns.  These conspire to produce a "stepped" decline with plateaus at ~4.5 million km^2 and ~2 million km^2 mean September extent.

The first plateau corresponds to an Arctic where (roughly speaking) the "shallow rim" seas melt out, but ice cover is retained over the deep central triangle.  The second plateau corresponds to an Arctic which is mostly ice-free but has a remnant core of ice crammed up against north Greenland/Ellesmere. Since 2007, we have been basically bumping along the first of these plateaus, as is clear from the September record.
http://nsidc.org/arcticseaicenews/2018/10/

It's anybody's guess as to when we fall off and plummet down to the second, but the Tietsche paper suggests it will be some time in the next decade. Given the rather wide standard deviation, the "ice free" threshold of 1 million could be hit at any point thereafter.

23

Arctic sea ice / Re: When will the Arctic Go Ice Free?

« on: March 12, 2019, 03:00:25 PM »

Discussions of fast crashes when thickness gets below a given threshold always strike me as conflating two UTTERLY different things.

The first is the kinetics of how the thickness of a given piece of ice melts during the course of a season.  As it gets thinner, eventually there comes to be a point where it's too thin to maintain structural integrity, fragments and rapidly melts.

The second is the kinetics of how the average statistical distribution of ice thickness changes from year to year.

24

Arctic sea ice / Re: Winter Temperatures in the Arctic

« on: December 16, 2018, 12:30:00 PM »

Easiest way to visualise this is to appreciate that DMI are effectively calculating the average temperature across a Mercator projection of the globe.The further North you go, the bigger the distortion.

25

Arctic sea ice / Re: "Smart" and "Stupid" Questions - Feel Free To Ask

« on: November 06, 2018, 10:18:18 AM »

https://en.wikipedia.org/wiki/Extratropical_cyclone#Extratropical_transition

26

Arctic sea ice / Re: Latest PIOMAS update (October)

« on: October 08, 2018, 01:45:27 PM »

27

The forum / Re: Poll predictions are stupid

« on: September 23, 2018, 12:41:50 PM »

They're useful, but not for anything they tell us about the ice.  They're useful as a barometer for how "connected" this forum is with reality.

An interesting exercise - if technically feasible, which I assume it isn't - would be to have two versions of the results graph.  One where everybody's vote is weighted evenly, and one where the votes are voted by how prolific each poster is.

The first of these tells us whether the readership of the forum has any insight into what's actually going on.  The second tells us who's driving the forum content, and whether they're more or less barking mad than the rest of us :-)

28

Arctic sea ice / Re: The 2018 melting season

« on: August 23, 2018, 06:03:13 PM »

But the -11 degrees figure was from memory, and it is supposedly the temperature that experience tells is needed for ice-free sea water to freeze. I'm not actually able to find any good references

By my recollection, it was first mentioned on this forum (or possibly the ASI blog) by Wayne Davidson, and was referring to his personal observations of what is necessary for ice formation on or around his home location, which IIRC is somewhere in the Canadian Arctic Archipelago.

It's not a strict physical constant.  In an system with no further heat input, then ice will form - eventually - whenever the air temperature has been below the freezing point of sea water for long enough.  However, if the air temperature is only 0.1 degrees below freezing, then the heat transfer is slow enough that it will take weeks for the ice to form. 

Of course, given that the temperature continues to drop during the autumn / winter transition, and that there may be ongoing heat input from ocean currents etc, this means that by the time ice starts to form, the actual air temperature is always substantially below freezing.  The precise details will vary massively from place to place.  Shallow seas will freeze quicker (and ice will appear earlier in the season, at relatively higher air temperatures), because there is less of a heat store.  Fast flowing water will freeze slower because there's a constant heat influx as warmer waters get brought into the cooling zone.

Where the -11 comes from is that in Wayne Davidson's home town, the conditions are such that in an average year, the temperature is roughly -11 degrees by the time the ice starts to form.  No more.  This forum shouldn't be treating it as a magic number.

29

Arctic sea ice / Re: When will the Arctic Ocean (CAB) go Ice Free in Winter?

« on: August 12, 2018, 10:28:17 PM »

However, as A-Team remarked earlier this melt season, the time of peak solar activity does not overlap well with the time of minimum ice. 

Why would you expect it to?  Even in a world where sunlight is the main cause of ice melt, then the peak solar activity would not coincide with the point of minimum ice but with the peak point of ice loss - i.e. the steepest part of the yearly ice loss curve during the melt season.

As it happens, there's a little bit of extra lag beyond that, because it takes a while for various short-term summer feedbacks (snow melt lowering albedo, melt ponds and leads likewise, and the sheer time taken to melt through the ice layer) to kick in. Not much of a lag though - the point of fastest melt rate typically occurs within 2-4 weeks after the solstice.

The inflection point between melting and freezing (i.e. the annual minimum) happens at or around the equinox, more or less exactly as you'd expect from that being the point where solar input (or rather radiation balance) starts to turn negative.

If you approximate ice extent as a yearly sine wave, it's offset by 1/4 of a cycle from the solar input cycle, exactly as you'd expect if insolation was the main factor affecting melt and freeze-up.  Sure, there are other interesting dynamics from warm air and warm water intrusion - and we discuss them to death every year - but they're second order effects.

30

Arctic sea ice / Re: When will the Arctic Go Ice Free?

« on: July 30, 2018, 10:28:41 AM »

The amount of heat that will be in the Arctic after it goes ice free is truly scary. All that energy that melts the ice.....once there is no more ice to melt it will make things hot, really hot, and that will make it hard for the ice to refreeze.

This applies to every seasonal ice zone in the world, and yet (to give but one example) Hudson Bay refreezes every winter.

31

Arctic sea ice / Re: JAXA 2018 Arctic SIE September daily minimum: August poll

« on: July 30, 2018, 12:19:51 AM »

Since I doubt the min will be in September I see no reason to vote.

what

32

Arctic sea ice / Re: The 2018 melting season

« on: July 25, 2018, 07:38:24 PM »

Here's what I have been considering and thinking a lot about lately. First off, I do not think this is going to be a record breaking year, rather it will be similar to 2016.

Actually, if you flip through the years on the NIPR site (=JAXA), then 2015 is closest.  2018 and 2015 have followed a virtually identical trajectory from March right up to mid July.  I don't know what that means in the grand scale of things, but for now that's where I'd put my bet for the rest of the season.  What was 2015 like weather-wise?

33

Arctic sea ice / Re: SMOS

« on: July 22, 2018, 02:44:12 AM »

Bottom melt happens primarily near the edges of floes, and the heat source is from radiation absorbed by the open water between the floes.  The onset is generally quite late in the season, but because of the huge thermal mass of the ocean, it can continue even after atmospheric temperatures have dropped below freezing at the transition from summer to autumn.

Anything other than the above is a mere detail in the small decimal places. This forum used to know stuff like this!

34

Arctic sea ice / Re: SMOS

« on: July 21, 2018, 10:38:50 AM »

...the factual claims you keep making are large, strange and without evidence, and I see it as my duty to point this out before this forum becomes a fantasy free-for-all.

Quite. As far as I can tell from Hyperion's ramblings, warm air currents are supposed to magically cause bottom melt without top melt, so the ice surface stays dry.  A contention that's trivially disproved with a large dry martini and a hairdryer.

This magical bottom-only melt arises because "Ice conducts heat better than water", and so it will conduct the atmospheric energy through to the water under the ice, allowing the ice to melt from the bottom, but not the top.  This is drivel.  Heat cannot of itself pass from one body to a hotter body. i.e. if the ice is conducting heat downwards, then the top of the ice is hotter than the bottom, so the top will melt first.

35

Arctic sea ice / Re: The 2018 melting season

« on: July 21, 2018, 10:27:01 AM »

A-Team, if you look at the compactness curves you may notice that compactness drops every time a storm moves over the central Arctic towards the CAA this summer.

Isn't this mainly due to cloudiness interfering with the sensors and giving falsely low concentration readings?  As we discuss several times every year?

36

Arctic sea ice / Re: SMOS

« on: July 20, 2018, 12:09:05 PM »

So massive thickening in the CAB, so I presume melt season somehow ended already?

It's not thickening.  SMOS doesn't measure thickness accurately during the melt season.
(This could be posted as a followup to every other post in this thread, so let's just assume it was)

37

Arctic sea ice / Re: "Smart" and "Stupid" Questions - Feel Free To Ask

« on: July 18, 2018, 02:35:01 PM »

I think it would be difficult to reach -21 degrees using this method, although I guess it could be done with enough salt.

"Enough" here is around 1/3 sate to ice by weight. Unless the airborne salt is a layer several centimetres thick (hint: no) then it's not relevant in this context.

38

Arctic sea ice / Re: 2018 sea ice area and extent data

« on: July 14, 2018, 09:40:00 PM »

Worldview shows thick cloud over the basin, I expect that's confusing the sensors and a significant amount of it will reappear in a few days. The images from 11th/12th didn't look anything like as low concentration as Uni Bremen suggests, so it's probably also getting confused by melt ponds.

39

Arctic sea ice / Re: The 2018 melting season

« on: July 14, 2018, 09:34:47 PM »

Well they have plenty of sub's and drones both above and below.

... no they don't.  They integrate what data is available, which is very very little.  It's predominantly a mathematical model driven by weather data, with assimilation of ice concentration data from NSIDC.

40

Arctic sea ice / Re: SMOS

« on: July 13, 2018, 12:11:00 PM »

It near as dammit is a blue ocean event already Neven. I bet my left nut that anything near the periphery, extending now to a Chukchi to Barents strip right across the pole where meltponding is being reported is actually open water, and satellites being fooled by wave action.

Blue is ice, white is clouds.  Why even bother to post something so easily disproven? The people that built and run SMOS say that during the summer it doesn't accurately measure ice thickness, and that is the only factual report here.
https://go.nasa.gov/2NKVaF7

41

Arctic sea ice / Re: Land snow cover effect on sea ice

« on: July 13, 2018, 12:04:35 PM »

Now let's rewrite that to estimate ice loss between June and September :

  June-area - Sept_extent = -alpha - beta * (June_formula - June-area/beta)

See, it's still the same equation.
The factor "June-area/beta" is a constant, not a parameter.

Depends whether June-area is also included as a variable in June_formula.  If it is, then it's a parameter. Note that when you do use area as a parameter, you get exactly the same result whether you're predicting Sept_extent or predicting (June-area - Sept_extent_.

I've added some of these permutations to the table :

Code: [Select]

              1979 - 2017           1992 - 2018
              Adjusted SD           Adjusted SD

k=2 (one variable) :
[...]
Area                  420               396
[...]

So let's change the regression formula so that it tries to predict the "June-area minus Sept-extent" variable instead of "Sept-extent" in absolute numbers.

Code: [Select]

              1979 - 2017           1992 - 2018
              Adjusted SD           Adjusted SD

k=2 (one variable) :
[...]
Area                  420               396
[...]


In cases where you're not actually using June-area as a parameter, it's justifiable to include it as a constant. 

42

Arctic sea ice / Re: Land snow cover effect on sea ice

« on: July 12, 2018, 03:28:36 PM »

So let's change the regression formula so that it tries to predict the "June-area minus Sept-extent" variable instead of "Sept-extent" in absolute numbers.

That's mathematically meaningless.  All you're doing is including area as one of your parameters while pretending you aren't. 

43

Arctic sea ice / Re: 2018 sea ice area and extent data

« on: July 11, 2018, 09:06:51 PM »

OK, I will buy into that.  Now....how close are we to an average size being about 100M?

A very long way away.  Any floe you can see on Worldview is a minimum of kilometres across, given that the pixel size is 250 metres.

The centre of this picture has some faint milky swirls. That's what floes smaller than 100m look like, because each one isn't large enough to turn a whole pixel white.  Not a lot of the Arctic looks like this yet, thank God.
https://go.nasa.gov/2NI3qW8

44

Arctic sea ice / Re: Land snow cover effect on sea ice

« on: July 11, 2018, 08:59:37 PM »

Further observation - adding snow cover does little to improve predictions over the whole time course, but is a bit more use in recent years.  That's consistent with the idea that snow cover matters more when the ice pack as a whole is thinner and has less thermal inertia.

It would be interesting to look at extent as an individual variable - is it better or worse than area?  Snow as a single variable is unlikely to be much use :-)

For completeness' sake, the other permutations of the two-variable models might be fun to look at. Formally, there are another 5 pairs that could be checked, namely extent+area,  extent+snow, extent+time or area+time.

Similarly, there are two more three-variable models: snow+extent+time, and area+extent+time, both of which I suspect would be worse; and a four-variable model which would likely be slightly better, but which will fit an elephant if you want it to :-)

45

Arctic sea ice / Re: Land snow cover effect on sea ice

« on: July 09, 2018, 12:56:26 PM »

2) The 'year' variable is adding one more variable, which will increase the risk of "over-fitting". Remember the famous saying by John Von Neumann :
With four parameters I can fit an elephant, and with five I can make him wiggle his trunk.

This is an unfair comment. Your model has three parameters (June ice area, June ice extent and June snow extent) while cesium62's model also has three parameters (June area, June snow extent and year).  Using standard methods of regression analysis, cesium62 found that there's virtually no benefit to including June extent as a parameter, while including year as a parameter improved the fit.  It's worth looking at why.

46

Arctic sea ice / Re: Land snow cover effect on sea ice

« on: July 08, 2018, 12:24:03 PM »

Normally for a hind-cast, that would include ALL the prior years available.

No.  To be a hind-cast you have to use the algorithm to predict data points that ARE NOT PART of the dataset used for the regression.

Forecasting is predicting future data.  Hindcasting is "predicting" data that already exists, which you do by not looking at it before you make your prediction. That graph contains three forecast points (2016-2018) and no hindcast points.

47

Arctic sea ice / Re: Land snow cover effect on sea ice

« on: July 07, 2018, 11:56:37 PM »

Here is what this hind-cast method did for the past 26 years :

Presumably at least some of those years were the ones used to fit the regression line, rather than being true hind-cast years.  Which of those years were not part of the data set used to fit the regression?

48

Arctic sea ice / Re: The 2018 melting season

« on: July 05, 2018, 03:48:42 PM »

Quite solid ice bridge on jun29 (the last relatively clear day).https://tinyurl.com/yabrcqvu
edit:Beaufort, should have added

Ice bridges don't form in open ocean, what on Earth are you talking about?  An ice bridge forms between the banks of a strait or river, and prevents the ice flowing with the current. It has to be anchored to both sides of the channel (i.e. at both ends of the bridge) in order to provide the necessary force.  Nothing in the picture you posted remotely resembles that.

49

Arctic sea ice / Re: Arctic Image of the Day

« on: June 25, 2018, 10:35:10 AM »

"What are you kids doing with my best blanket?"
"We're, um... watching for whales, Mum!"

50

Arctic sea ice / Re: September Predictions Challenge

« on: June 24, 2018, 01:13:27 PM »

(Side note - next year it might be interesting to try the "distribute 100 tokens" approach since that lets people give a more nuanced description of their confidence intervals, which could be non-normal, asymmetric or even bimodal.)