Background

Sea Ice Primer

Arctic sea ice cover

Sea ice is a thin, fragile, solid layer that forms in the Polar Oceans. Sea ice floats on the surface of the ocean and forms a boundary between the relatively warm ocean and the cooler atmosphere. Sea ice moves in response to winds and ocean currents. In the Arctic, sea ice is, on average, about 3 meters thick. The ice thickness ranges from as little as a few millimeters in ice that is just forming to as much as tens of meters in places where pieces of ice have crunched together to form ridges. At its minimum extent in summer, the Arctic ice pack covers an area around the size of the continental United States. In winter, the area of the ice cover increases by about 3 or 4 times the size of Alaska!

Sea ice is generally divided into two categories:

First-year ice floes: Floes that grow during one winter. This ice is usually smooth and is commonly 0.5 to 1.0 m thick.
Multi-year ice floes: Ice floes that have survived at least one summer melt season. Generally multi-year ice floes are 3 to 5 m thick, although ice floes thicker than 5 m are quite common. Multi-year ice floes are not smooth like first-year ice, but contain both old and new ridges.

Snow covered multi-year ice. The bumpy looking areas are pressure ridges where the ice has crunched together.
This is a view of Ice Station SHEBA in April 1997.

At the fundamental level, an understanding of sea ice in the polar regions continues to elude scientists. Because sea ice covers a great extent of the polar seas but exists in an inhospitable environment, science projects directly studying sea ice are few. In the global context, sea ice currently covers roughly 6% of the earth's surface at any given time. It is also the fastest global-scale solid material moving upon the earth's surface (blowing sand and snow being airborne materials) making sea ice the most difficult global-scale solid material to track on earth. Even in this era of space borne platforms, the polar regions still retain much of their mystery because satellites cannot access these regions through any form of geosynchronous orbit (an orbital pattern which allows a satellite to target the same point on the planet continuously). This constraint inhibits the visualization of sea ice movement that can occur on a time scale of days.

Sea ice is in constant motion. As stated earlier, floating sea ice is pushed by both ocean currents and wind. Ice floes are constantly being pushed against each other. When pushed together, with enough force, the edges of the ice floes will start giving way and ice will break into pieces and pile up along the floe boundary forming ridges (pictured below, left). These ridges can contain many ice blocks and be very thick, extending into the air, above the ice floes, and into the ocean, below the ice floes, for many meters. A combination of dynamic and thermodynamic processes continuously alter the composition of sea ice at all scales, from the basin scale down to the crystalline level.

The wind and ocean currents, and other stresses in the ice, can cause both first year and multi-year floes can also break apart, forming areas of open water through the middle of the ice floe. These areas of open water are called leads (pictured below, right). In summer, melting snow may form fresh water ponds on top of the sea ice, further altering its composition.

This ridge formed under the tent in less than an hour. There was no time to remove the contents from inside.

This lead, through the middle of the ice camp formed without warning.

This CD (these web pages) concerns the study of dynamic forces and stress in the ice pack; forces that can lead to the development of ridges and leads. To better understand the material presented in this CD, it is useful to provide some general understanding of sea ice mechanics.

General Information on Sea-Ice Mechanics

At spatial scales of 10-300 km sea ice consists of a collection of plates with differential motion that takes place along the edges of these plates. It is equivalent to the oceanographic mesoscale (10-100 km) which is rich in high energy dissipation processes such as eddies and other geostrophic density-driven processes. At this scale, differential sea-ice motion plays an analogous dissipative role resulting in the development of features such as leads, slip lines, cracks, and pressure ridges. Within the sea-ice community there is, at present, no formal definition of this scale. For most sea-ice models small-scale discontinuous behavior is accounted for by a plastic yield curve in an average description suitable for a continuum modeling approach. Within the literature definitions of discontinuous behavior such as linear kinematic features (LKF's) and piece-wise rigid motion and aggregate scale are beginning to emerge.

From a human perspective, when navigating in polar waters or maintaining offshore structures, sea-ice discontinuities are either useful conduits (leads) or impediments (ridges). Thus, mobility at the human scale is affected by changes in discontinuities of the sea-ice pack. These discontinuities are also fundamental regulators of heat, mass, and momentum transfer at the air-sea interface of one of the world's most sensitive climatic regions.

Relationship Between Ice Thickness and Dynamics relevant to Dynamic processes

Field measurements of stress and strain-rate (through drifting buoy arrays) are essential to investigations and validations of the horizontal meso- to large-scale material behavior (rehology) of sea ice. The three different scales of the experiments contained herein provide unique Lagrangian representations of stress and strain-rate from which to test theoretical rheological formulations. Thickness information is crucial to this type of analysis, as demonstrated in the original Hibler (1979) formulation for maximum ice strength ( P max ),

P_max = P* h exp[-C(1-A)]

where P* is an ice strength coefficient used often by modelers, h is the ice thickness, C is an empirical constant, and A is the compactness or area of ice within a given location. This relationship is significant because it defines the limit at which ice begins to yield (i.e., has reached the limits of its material strength).

Dynamic vs. thermodynamic pack ice evolution

The thickness of ice cover changes as a result of thermodynamic (formation and ablation) and dynamics (ridge and lead formation) processes. The thermodynamic contribution plays a more significant role in strengthening the sea ice cover early in the season, when there is a significant amount of young ice. By 1 January, much of the sea ice cover has nearly reached its maximum thickness. Beyond this point, mechanical thickening is left to account for the significantly observed increase in the compressive failure of the ice cover.

Thermal vs. dynamic stresses

Tucker and Perovich (1992) describe three sources of stress in the drifting pack ice in the Eastern Arctic: 1) changes in the ice temperature, 2) ice motion due to winds and currents (low frequency forcing), and 3) ice motion due to tidal or inertial oscillations (high frequency forcing). These sources can also be categorized as either thermal (1) or dynamic in nature (2 and 3). Tidal oscillations (3) were not evident in analyses of these data, and we speculate that this may be due to the high concentration and strength of the ice pack during the winter period, when our experiments took place.

Thermal stresses are caused by the differential response of ice to changes in air temperature. A change in air temperature induces variations in the ice temperature with the top surface of the ice being the first affected. The top ice cover attempts to expand or contract in response to the temperature change, but is restricted by lower portions of the ice which have yet to experience the thermal load. Stresses therefore develop throughout the thickness of the ice sheet even though an actual change in ice temperature may only have penetrated the top portion of the ice cover. Variations in air temperature can occur rapidly and are often extreme. The snow and ice cover acts to dampen these rapid variations, producing an in-ice variation of temperature that has as a frequency response on the order of days rather than hours. These data, and that of other investigations, shows that thermal stress measured at the sensor is isotropic in the plane of an ice sheet.

Ice motion-induced stresses occur in response to wind and current forcing and can result in the formation of ridges, leads and rubble fields. These data, and observations by others, show that ice deformation events are associated with large, and high-frequency changes in the ice stress, with the stress exhibiting strong directionality during periods of rapid activity.

Instrumentation

Stress Sensor Instrumentation

The sensors used in the study were specifically designed for insitu measurement of ice stress (Cox & Johnson, 1983) and have been successfully used on a number of Arctic field experiments.

Image of a stress sensor. This steel cylinder is placed vertically into the ice. The deformation of the cylinder is measured by a rosette of vibrating wires. The circle drawn on the sensor indicates a location where a wire is attached to the inside wall of the cylinder.

Image looking down the inside of the stress sensor showing a 6-wire rosette crossing through the center of the steel cylinder. (Note: A newer Stress Sensor, containing a 3 wire rosette was used for our projects.)

Each sensor (above, left) provides information on the stress acting at a point in the horizontal plane of the ice cover. Stresses are determined by measuring changes in the radial deformation of a cylindrical steel annulus, using a 120°, three-wire rosette (similar to the sensor pictured above, right). The wires in the rosette stretch across the hollow center of the annulus at the midpoint along its length. Periodically these wires are magnetically plucked to measure the frequency of vibration of each wire. Changes in the diameter of the annulus cause a change in the length of the wires and hence a change in the their frequency of vibration. Knowing the material properties of the steel annulus, we can directly relate this change in frequency to the stress applied to the sensor.

Typically we use the data from the stress sensors to establish the primary and secondary principal stresses and the direction of the primary principal stress. We use the convention that compressive stresses are positive and that primary stresses are greater than secondary stresses. Laboratory calibration indicates that, over a loading range of 0 to 2 MPa, the measured principal stress is within 15% of the applied stress, the principal stress direction is typically correct to within 5º, and their resolution is 20 kPa.

In addition to measuring the ice stress, each sensor is equipped with a thermistor, providing the ice temperature at the point of the stress measurement.

Installation is accomplished by supporting the sensor in a 10-cm-diameter hole drilled in the ice using standard coring equipment, backfilling the hole with water, and allowing it to refreeze. All of the sensors in this experiment were located at approximately the same horizontal level in the ice sheet, with the wire rosette an average of 22.1 cm below the ice surface. Field experience indicates that the compressive stresses associated with freezing the sensors into the ice cover are typically 100 kPa, which dissipate within 2-4 days.

Buoy Drifter Processing and Grid Configuration

Hourly buoy drifter data was received from the NOAA Pacific Marine Environmental Laboratory (PMEL). The hourly time and position (decimal degree latitude and longitude) were read in and transformed to x, y positions on the SSM/I grid from global latitude, longitude positions using the polar stereographic special Sensor Microwave Imager (SSM/I) grid transformation with 70^o N chosen as the reference latitude (plane of no distortion), orientation relative longitude 45W (also known as the RGPS grid), and an eccentricity for the earth's surface shape (e=0.08181615). Data gaps of three days or less were linearly interpolated to hourly positions with longer gaps flagged as undefined. Data quality control was performed using a simple forward difference of positions to compute velocity. A threshold of 50 and 20 cm s^-1 in velocity and velocity change (surrogate acceleration), respectively, was used to flag suspect results. Centered differencing was computed from the interpolated, filtered x, y positions to obtain “cleaned” velocity. Quality control (QC) flags were applied based on pre-filtered QC results and checks of extreme velocities from the final centered differencing results.

The resulting data set provided herein consists of a one line record for each reading with values of time (year, day-of-year, hour, month, day, decimal year), original hourly positions (decimal degrees), "cleaned" position (km) and velocity (cm s^-1). Flagged results are declared undefined and labeled 9999.000 in the data fields. Tables, included in each of the project archive directories, provide an overview of the buoy names and duration during each experiment with an additional table in each archive directory describing any data gaps larger than 3 days. The data (excluding time) are written to significant digits starting with a position accuracy (0.0001 decimal degrees). This propagates to 2 and 4 digits after the decimal in units of km and cm s^-1, respectively. Updates in forthcoming editions of the CD will include background information on the raw data (prior to hourly format), data quality flags on the cleaned data, and calibration of buoy position uncertainty using buoys on the same floes as available.

To access the data from each project, Click on the following links:

SIMI Project

SHEBA Project

Beaufort Project