Ice Stars:  Fantastic flow patterns in snow ice crated by water flowing from under the ice into snow covering an ice sheet.



Vicki Gluek sent this picture from Canyon Ferry Reservoir in Montana.

 This formation is an ice star (Note 1)

The ice star and associated infiltration features create a domain for each star.  The domain starts as the initial pore or small hole in the ice.  Water flowing through the hole to the dry snow.  The flow rate is determined by the size of the hole, the thickness of the ice sheet, the submergence of the plate, the density of the snow, and the thickness of the dry snow. Each ice star occupies a domain that includes the hole, the dark arms, and the streamline like rays that branch our from the dark arms.   The dark arms around the hole are infiltration conduits that carry water away from the hole. Water flow appears to be turbulent in the dark arms.  The streamline like pattern seen in the pictures below are where water flow is laminar. Water flows flows through the partially saturated snow. If you look at the pictures carefully you can where the how the domains form boundaries as the streamlines approach approach other other expanding domains. 

are partial melting of the slush as i t infiltrates the snow.  They appear to be smaller conduits that occur near the surface so they have grey snow ice underneath them. The streamlines grow outward through the dry snow until they approach 

One plausible sequence is:

  • A new black ice sheet forms and thickens.
  • Just before a snow storm the sheet thaws enough to develop sporadic drain holes.  They may be tripple junctions between ice grains and they may have been enlarged by surface melt/rain water flowing down through them. Any passage that allows water to wick into the snow will do.
  • Alternatively, the the insulation effect of the snow and moderate temperatures should allow the ice to become isothermal at 32 degrees.  Triple junctions between ice crystals have a very slightly lower melting point.  A few of them may melt enough to form the first very narrow passages that then erode into larger ones as water flows up through them. 
  • It snows which submerges the ice sheet with a combination of the weight of the snow and the weight of water wicked into the snow as water comes through the ice holes.
  • Water infiltrated from the inflow hole melts conduits in the slush layer that are clear and appear as dark, water ice 'octopus arms'. 
  • The streamline pattern fans out from the conduits to the edge of the area fed by each hole.  The streamlines appear to only be in the top 1/4" of the slush layer (this layer often 1/2 to a couple inches thick).
  • The slush freezes into snow ice and clear ice in subsequent cold weather.
  • The ice surface gets Zambonied in a thaw to remove the layer of white ice enough to make the patterns visible.
  • We come along and enjoy the show.





The roundish hole in the center is about 8" in diameter

The streamline like infiltration pattern shown in these two pictures is seen, at least to some degree, on about half the ice stars I have looked at although rarely as pronounced as this. In this case there was roughly a 1/4 square mile patch of ice with this pattern (Lake Champlain, 2009).


This ice star is about a foot from end to end.

This large octopus like feature is about 20 feet across. The pattern in this case is serving as drain channels eroded by melt or rain water headed for the hole. It is hard to say for sure if it started as an inflow hole or is just a drain feature that started as a fishing hole (the center hole is round and the same size as a nearby drilled hole).



 In this case a crack (instead of a hole) let lake water infiltrate into the snow pack.

 Click here for an annotated gallery of the probable formation process.

 First published in January 2011.  The text has been significantly modified since, most recently on 11/14/12.





1) Ice: The Nature, The History, And The Uses Of An Astonishing Substance  Mariana Gosnell, Published by Alfred A. Knopf.  Chapter 1.  Comment: This is an excellent and broad look at ice. 

2) Charles A Knight wrote a paper on Slush on Lakes in 1987 that explains the formation process.  See CRREL record ID 51000661

3) Jan 15, 2013: Dave Dermott just sent a reference to a technical paper on these formations.  It gives a general description and provides a mathmatical model for their behavior.