Lake Ice

Ice Formation

Lake ice occurs primarily in the Northern Hemisphere, where most of the ice is seasonal: it forms in the autumn, thickens during the winter and melts in the spring.

Meteorological factors such as air temperature, precipitation, wind speed and radiation balance coupled with physical characteristics of the lakes and ice (lake area, depth, volume and fetch; snow depth; ice thickness, type and albedo) lead to complex interactions and feedbacks that affect the timing of freeze-up and break-up (ice cover duration) each year.

In general, there are two types of lake ice. They are:


MST Pond ice cover illustrating black ice (right) and white ice (left). MST Pond is on the Poker Flat Research Range ~50 km northeast of Fairbanks.

Congelation Ice (black ice) forms as water freezes on the bottom of the ice cover and the latent heat of crystallization is conducted upwards through the ice and snow to the atmosphere. Its growth rate is proportional to the rate at which energy is transferred from the bottom surface of the ice layer to the air above.

Congelation ice is often referred to as black ice because it has a high optical depth that permits significant light transmission to the underlying water.

Snow Ice (white ice) forms when the weight of a snow cover is sufficient to overcome the buoyancy of the ice supporting it, the ice surface becomes submerged and water flows through cracks in the ice and saturates the snow, which then freezes.


The processes leading to congelation ice (1) and snow ice (2-4).

Snow ice is often referred to as white ice because it contains a large number of densely packed air bubbles and small ice crystals that cause strong light scattering and a high albedo.

.



Vertical thin section showing granular snow ice (top) and columnar congelation ice (bottom).

Vertical thin section showing 15-31 cm of ice core. Horizontal thin sections at 15 cm (far left) and 31 cm (near left).


Thin Sections
Thin sections are made by cutting ice cores vertically or horizontally into very thin layers. These layers allow light to pass through them. When thin sections are placed between cross-polarizing filters on a light table the crystallography of the ice is revealed.


Phenology

The phenology of an ice cover is the freeze-up date, break-up date and ice cover duration.

New ice cover

Freeze-up (FU) defines the period between initial ice formation and the establishment of a complete ice cover. The FU date is the day that the pond is completely ice covered.

After the lake cools to 4°C (maximum density of water), the water surface cools to the freezing point and ice begins to form.

The first ice to form (initial ice skim) often appears first as border ice in shallow, protected areas. However, the entire lake surface can reach the freezing point simultaneously and a continuous ice cover will appear in a matter of a hours.

Break-up (BU) defines the period between the onset of snow melt and the complete disappearance of the ice. The BU date is the day when the lake is completely ice free.


A moat forms around 35.5 Mile Pond during spring break up.

When average daily air temperatures rise above the freezing point, ice begins to decay by:

  • dimensional thinning and
  • deterioration of the ice crystal grains at their boundaries.

Thinning of the ice layer is caused by heat transfer and by melting at the top or bottom surface (or both) or by contact with inflowing warm water.

Taken together freeze-up and break-up denote the endpoints of the ice cover duration.


Phenology at Poker Flat Research Range

Since Fall 1999, ice and snow data have been acquired at Poker Flat Research Range approximately 50 km northeast of Fairbanks, AK.


The freeze-up patterns at the 11 PFRR ponds are not systematic. In 2003, the ponds freeze over only to completely melt out again before the seasonal ice cover is established.

The break-up patterns of the 11 PFRR ponds is orderly compared to freeze-up. The ponds at PFRR have broken-up at increasingly earlier dates in the 4 years for which there are data.


Ice Growth Season FU date
(100% ice)
MM/DD/YY
BU date
(0% ice)
MM/DD/YY
Ice duration (days)
2000-2001 10/02/00 05/16/01 227
2001-2002 10/13/01 05/20/02 220
2002-2003 10/11/02 05/12/03 214
2003-2004 10/16/03 05/07/04 207
2004-2005 10/16/04 05/03/05 200

At MST Pond, freeze-up was about 2 weeks later and break-up was about 2 weeks earlier in the 2004-2005 season than in 2000-2001. This leads to a 27 day reduction in the ice cover duration.

QUESTION: What environmental factors might lead to such a reduction of the ice cover duration in Interior Alaska?


Variability

Between the beginning of freeze-up and the beginning of break-up the lake ice thickness increases primarily by thermodynamic processes (freezing of water) which can be understood relatively easily in terms of heat conduction driven by the energy balance at the top surface. On larger lakes dynamic processes (deformation of the ice due to motion) may come into play.

The thermodynamic processes are dependent on snow (depth and density) and air temperature. Since these vary from place-to-place (spatial) and year-to-year (temporal), it is not surprising that lake ice thickness (and composition) varies too.

LOCAL VARIATION IN SNOW DEPTH AND ICE THICKNESS AT PFRR (2004-2005)Snow and ice thickness at PFRR 2004-2005
The distance between MST Pond and 33.5 Mile Pond is 2.5 miles. Maximum ice thickness was 0.68m at MST Pond, 0.71m at 31.6 Mile Pond and 0.69m at 33.5 Mile Pond.

QUESTIONS: What could account for the differences in the PFRR 2004-2005 data? Are they significant?

Snow depths and ice thicknesses vary more dramatically over larger areas due to differing climatology:
  • Arctic Coast: Winters are long and have low wind chills, low precipitation and frequent blowing snow (Barrow).
  • Bering Strait Coast: Warmer (longer) summers than Arctic coast. Moderately windy with capes and passes very windy (Nome).
  • Interior: Coldest zone in winter with light winds (Amos Lakes and Fairbanks).
  • Southcentral: Transition between Interior and Pacific Coast (Wasilla).
  • Pacific Coast: Rain and snow in winter with moderate to strong winds (Seward).

REGIONAL VARIATIONS IN SNOW DEPTH AND LAKE THICKNESS IN AT ALISON SITES IN 2003-2004

Variation in snow depth and ice thickness in AK during 2003-2004

QUESTION: Why is the wind an important factor in lake ice thickness?


Temporal Variability

TEMPORAL VARIATIONS IN SNOW DEPTH AND ICE THICKNESS AT MST POND (PFRR)

Temporal variations in lake at MST Pond PFRR

Data have been collected at MST Pond (PFRR) since November 1999.

The temporal variability of the snow depth and ice thickness (right) is significant. Maximum ice thickness values range from 0.62m in 2001 to 0.91m in 2003 (mean for the 6 years is 0.76m ± 0.10).

Ice composition also varies from year-to-year (below). In 2002, snow ice did not contribute anything to the total ice thickness at MST Pond. In 2005, over 50% of the ice thickness was made up of snow ice.

QUESTIONS: What early winter phenomenon produced the very thick ice cover in 2002-2003?

Is there some connection between the amount of snow ice in the 2005 ice cover and its relative thinness?

LOCAL TEMPORAL VARIATIONS IN ICE COMPOSITION AT PFRR
Variations in lake ice compostion at 3 ponds at PFRR

Project Details

ALISON
Alaska Lake Ice and Snow Observatory Network

Support From

National Science Foundation
International Arctic Research Center
UAF Geophysical Institute
University of Alaska in Fairbanks