From Barrow Strait to Barrow Alaska Sea ice gives the same interesting horizons.

After hitting play,  place mouse Cursor tip at sea ice
  horizon center, leave it alone and enjoy the
 shifting.   (2007-2008) 

U of Alaska Barrow webcam says it all.  Even a plain webcam easily demonstrates the shifting horizon between ice "seasons",  there are likely more than 11...   After watching the video above and or any other well done sequences.  You may realize the seasons namely:  1- Pre first fast ice ice ultra low horizon   2-Freeze up horizon ,  3- Thin ice ice horizon,  4-  Dark season ice,  5- Sunrise horizon 6-Great late winter horizon ice, 7-Great Diurnal shifting horizon ice,  8-"First Melt" astronomical horizon ice,  9- Melt Pond horizon ice,  10- Mixed water and ice horizons,11- open sea water horizons.  WD June 5,2015

 

Dedicated Sea ice model proofing, offering Horizon observations for correlations.

~First ever sea and ice temperature profiles extracted by refraction observations.  
~Already helped proving buoy top thermistors measuring wrong temperatures
~It is hoped to be a useful for Sea ice dedicated coupled models.  

   It has been known that GRIB  model can't duplicate exact Arctic Ocean sunset geometry,  or not calculate near surface inversions exactly by either failing to replicate an inversion temperature profile or missing to forecast them all together.  Mass Buoys also exaggerate the solar warming of its top thermistors,  I noticed this when a consistent isotherm was observed a few weeks after end of long night over NW passage sea ice horizon after every Local Apparent Noon.    When sea ice horizon elevation is identical to the true astronomical horizon it means that top of sea ice and the air right above have identical temperatures.  During Spring from Southern Cornwallis Island Nunavut Canada shores, a stable isotherm was seen lasting from a few minutes during first day it was observed  to nearly half a day several weeks later,  even  when quite cold temperatures persisted.  This contradicted buoy data suggesting an adiabatic profile with top  sea ice air warmer than surface air at 2 meters.  

There is substantial data to extract from horizon observations throughout the entire Arctic year.  The following figure encapsulates the main temperatures profiles of sea water, ice and air over the entire year:

    Temperature profiles determined by Horizon Elevations can be very useful to check Gridded GCM's.   Here is the summary of all phases of sea ice throughout the Arctic year.    Some temperature profiles are truncated in order to fit the sketch.   It would be desirable to see a similar figure created by a coupled model.

 From left to right there are about 9 recognizable horizon elevation changes which stem
from significant temperature changes on part of or the entire temperature profile comprising either two or three  different mediums.

Late Summer 74N 95W:

    High Arctic surface temperatures may vary between -5 to +5 C.   At +5 C there can be a temperature  inversion over a sea with sst's  not usually exceeding +4 C.

Early Autumn;

     Surface temperatures normally vary from -3 to -10 C,   it is common to observe the true astronomical horizon when sea surface temperature is equal to surface air.

Middle Autumn;

    Is  the time when sea water subsists despite colder -5 to -15 C air temps.  When the temperature is seasonal minimum cold,  the horizon drops to is lowest point of the year.

  Late fall;

    Fast ice forms usually after October 1,  especially during the last 10 years as opposed to early and mid September in the further away recent past.   Something spectacular
occurs with the horizon when grey ice starts to be prominent,  the horizon is seen very low with open water,  and jumps to above Astronomical Horizon in less a day after  sea ice completely covers all the way to the horizon,  the ice accretes till "first melt".

  Early Winter;

     With the beginning of long night (early November),  the sea ice horizon very slowly rises day by day until it vanishes in the night.

  Mid Winter;

     Is marked by temperatures between -25 and -40 C with only daily noon twilight for bright light.  The horizon is much higher at sunrise from the long night compared to when sea ice first set completely.   But there are some variances caused by advection of either cold or warmer air.   New ice usually is about 1 meter thick.   By mid February,  the effects of solar radiation increases horizon elevations even more, especially in the evening,  to highest levels until mid-April.

Late Winter;

   A surprising view occurs when the astronomical horizon is the same as the ice horizon for the first time since mid October, it is "first melt",   when the sun is high enough in the sky to eradicate the very persistent inversion giving horizon continuously above astronomical horizon.   "First melt" likely happens at bottom of the ice column, but the horizon rises a few minutes after  and the bottom refreezes.   Over the last 4 years ,  "first melt "  dates occurred on different dates,  2010 being the earliest.   After FM,  the sea bottom thaws and refreezes daily.  With a gradual progression of longer and longer horizons at same astronomical level,  only interrupted by snowfall, clouds or fog,  the ice bottom remains more less the same.   From this time the ice column warms slowly as well.

  Early Spring;

      Day by day the Astronomical horizon is achieved at longer and longer intervals.
Post Local Apparent Noon Isotherms subsist right above the ice.  The ice rots at bottom.
But in the evening with lower sun in sky,  the horizon rises, inversions occur diurnally, and exist from evening to early morning when sunny,  they even happen during cloudy periods.  The ice column warms more and more,  contributing to ice bottom rot.    But there is cold ice "middle" core which helps cool the adjoining air faster thus causing the horizon to rise. In time,  the cold ice core becomes quite insignificant,  the astronomical horizon period extends to nearly all day.

Late spring;

     The onset of melt ponds should lower the horizon below astronomical horizon, but the ponds must cover most of the ice all the way to the horizon.  If so,  an adiabatic profile subsists from top of ice surface upwards.  Both top and bottom ice starts to melt.  Sea water temperatures increase slowly, necessary to melt sea ice depleted of salt,  which has a point of fusion closer to 0 degrees C.  A sea ice column may subsist even with temperatures nearest or just above -1.8 C,  provided brine has been flushed away. WD May23, 2015

Sea Ice Thermal Flux Profiles II, as demonstrated by the horizon; thin ice

~A big surprise, very thin sea ice gives a similar horizon to much thicker ice, the instant it is set completely.
~The difference between the two ice profiles help explain thermodynamic action.

   Astounding as it sounds,  thin sea ice raises the horizon a lot as soon as it forms.  Historically, there has been well known weather the day sea ice forms,  the clouds clear along with a common impression of much colder temperatures all around coastal areas.  For obvious reasons thermal variances are difficult to study when sea water turns to sea ice,  although flux studies have been done mainly after sea ice is solid enough to put equipment on it.

   Just after Minima of 2013.  The apparent Arctic wide cooling caused by extraordinary dynamic Gyre stall of ice compaction over the entire Arctic Ocean (except for the Atlantic side) caused an earlier freeze up of McClure and Barrow Strait.   This reduced the usual cloudy autumn from masking the horizon.  Above left picture was on September 21, 2013, with surface temperature -6 C.  Some ice was already present but sea ice formed further afterwards.  Middle picture was September 23, 2013,  sea ice appeared to have formed completely and the horizon rose above true astronomical horizon.  This is simply an exhilarating discovery, it leaves long wave thermal transfer as the principal thermal contributor  of the near surface inversion causing the greater refraction looming.    Further to this (extreme right),   the horizon remained the same or even dropped a little 2 hours later in the evening.    This is another discovery.  Usually a thicker sea ice horizon rises in the evening,  even when cloudy, a stronger evening inversion did not happen when ice was seen bran new.   This lack of horizon rise was observed again multiple times with very thin ice.   I suggest an absent sea ice "cold" core which allows the air to cool faster above top of ice,  while with new ice the higher sun presence at about noon gave greater thermal flux upwards creating a slightly more visible inversion.

   A common problem with very new sea ice analysis is of course caused by fog or clouds.  Natural cooling during autumn causes a great deal of moisture obscurations.  Fortunately 2014 had a brief respite just about the right time,  although not perfect,  the photographic  repetition of 2013  freeze-up was achieved.

    Barrow Strait  can be very complicated by its tidal currents which change substantially during a moon cycle.    But here we can note the same thermal sceneries as witnessed in 2013, 2012, 2011 and 2009.  Capturing a freeze-up with very little clouds is rare.  This is why 2014 and 2013 are featured here,  the other years had some clouds making demonstrations possible with a lot more explanations.  2014 freeze-up took 3 days,  of which day 2 had grey ice which will be dealt with on a subsequent article. Picture of October 1 (extreme left) feature rolling water waves and sea ice bits in a fierce wind storm.  Temperatures ranged from -9 to -12 between Oct 1 and 3.  2014 freeze-up occurred at about -11 C which was a return to regular yearly feature.  except for 2013.    On October 3, Barrow horizon rose substantially despite similar temperatures.  20:07:54 UTC capture (second from left) had new thin ice slightly mangled by winds and tidal currents during its formation.  But at once the horizon rose when sea ice covered all of sea water up to the horizon.  A few hours later the horizon dropped (23:31:26) again likely due to air cooling faster than top of thin ice.  Well frozen with thicker sea ice,  the horizon rose most 15 days later (furtherest right).  the cold ice core started to to grow enough to affect the evening rise, accretion of ice goes in tandem with a more risen horizon.

  Barrow Strait Ice is usually more chaotic than McClure Strait , but concurring to the demo above  the Western view of the Northwest passage had similar refraction effects but on different days:

   October 2, 4 and 18 2014 (from left to right),  sea ice set a day later in McClure Strait and looking at NW passage .  October 3 had grey ice.  The much lowered water horizon (left) rose at freeze-up on (center),  the horizon rose further 2 weeks later indicating the build up of a cold ice core.  The repetition of this looming feature helps explain thermal fluxes instantly.  

  Thin sea ice main feature as giving as high or higher horizon compared with thicker spring time sea ice at noon must be due in large part to thermal long wave heat from the the summer warmed sea causing a weak inversion,    convection stops the moment the sea surface becomes crystalline, the insulation properties of very thin sea ice is simply spectacular.    Sea ice conduction in direct contact with surface air plays a different role in autumn,  likely slightly warming the interface (if at all)  as opposed to cooling the snow/ice with air interface in spring especially in the evening.  The missing core of cold ice is replaced by thermally "hot" sea water reducing evening inversion amplification.  But this feature has been observed to be short lived,  one week or so from onset the stable to slightly lowering horizon in evening changes to heightening as a top of ice cold core becomes more and more resilient and effectively cools the interface faster after warming from the autumnal noon Polar sun. Conduction from very thin sea ice appears to be very poor,  not powerful enough to warm interface air by causing and adiabatic profile which would lower the horizon below astronomical horizon height.   WD May17,2015

Reference:
 

Surface energy fluxes of Arctic winter sea ice in Barrow Strait

Sea ice thermal flux profiles , as demonstrated by the horizon.

~Part 1 ,  sea water to very thin sea ice
~ An attempt to explain what happens at the sea to air interface when sea ice forms

    I was going to demonstrate possible clashing between buoy data and refraction observations.  But it is now known that above sea ice Mass Buoys thermistors exaggerate
the snow or air layer temperatures a great deal.   Buoy 2015A had 20 cm of snow with temperatures in excess of +7 C,  while surface temperatures was +1.48 C (thanks to Jim Hunt for good detective work).   It is not impossible for snow surface to be warmer than the air when its sunny.  But it is impossible to have snow at +7 C.     This obvious flaw asserts questions and doubts  about for above ice string thermistors.  Moreover this data is lost to what should have been computer Model output verifications.  It is a loss which may be compensated by refraction observations.

Simple initial sea ice formation;
                                 
     Easy as it seems,  thermally speaking it is not, we let the horizon demonstrate,  but after several freeze-up seasons of observation,  the problem was to be sure if the sea ice has been totally covered all the way till the horizon.   Any open water affects the horizon line elevation.    Arctic sea water is magnificently stable in temperature,  so its horizon line varies almost exclusively with air temperature change:

   A much lowered horizon is due to "sinking",  the surface temperature is "about" 6.2 degrees warmer than air,  "about" because the horizon line changes position with the temperature differences between the entire light path which is never measured.    The few bits of ice are also miraged.  

     The most difficult analysis possible is to see what happens when sea ice forms over a wide expanse at once,  but what seems fairly certain,  there is no ice thickness limit by which the horizon always rises above true astronomical horizon.  This seems to be an astounding discovery:

   2014 October 3 (left) and the next day.  The horizon at left is about 17 km distant,  (at right) more than 25 Kilometers away.   Mostly water horizon turns to 100 % sea ice.  Similar freeze-ups were re- observed in 2013 and 2012.   It does not matter how thick the ice is,  it will set the horizon above astronomical line as long as the sea is completely covered with ice.  But the finer details,  from very low horizon of open water under a deep freezing air,  to above astronomical when sea ice covers everything,  need be carefully studied since it helps determine thermal flux  balance in one glimpse.  Usually the very low pre-freezing sea water horizon does not last long because the very cold air changes the state of sea surface.   Recent years freeze-ups happened because of colder Southern in provenance atmospheric advection , with the exception of 2013,  sea water was so warmed air temperatures needed to be below -11 C.

    High precision satellite pictures along with a careful analysis of a surface phase change would be quite revealing.   The reason for various horizon heights have everything to do with thermal fluxes.  As winter progresses,  the horizon changes until spring when the sun literally gives a similar horizon look,  but for entirely different  reasons than at freeze-up.   The very low sinking horizon exists because of a very steep near surface adiabatic profile,  the raised horizon on account of a completely ice covered sea,  is caused by looming  by profiles having stable inversions varying in strength based entirely on the current thermal balance.  This indirectly identifies ice thickness because
thicker sea ice has a cold or warm "core", significantly affecting thermodynamic action.   WD May 10,2015

It is shaping up like 2007

~Warmest NorthWest passage spring and the creation of a circulation Dipole

     Latest EH2r refraction Upper Air of NW passage data is warmest ever in 14 seasons.
Mid April projection of Cold Temperature North Pole locations is so far so good accurate.    But the greatest implication is a circulation pattern similar to 2007.

        2007 (extreme left), compared to 2012, 13 and 15 coldest air in the Northern Atmosphere locations.   2015 Refraction of sun disks over the NW passage is the least ever measured confirming NOAA summary above,  the more the upper air is cold the smaller the vertical diameter of the sun. Up to today,   2015 expanded sun disks  is an order of magnitude higher than warmest 2005 and 2010.   The obvious resulting  circulation pattern is favorable for High Pressure system over Beaufort Sea  very similar to 2007. WD May 7,2015