PIPS 3.0 Statement

Igor Appel


Air Stress

Despite obvious achievements in theoretical study of interaction between ice cover and adjoining media, the parameterization of several processes should be based upon empirical observations. Primary of these processes concerns the relationship between the wind above the atmospheric boundary layer and surface wind determining tangential air stress.

Appended is a copy of a proposal to study "Relationship in the System Geostrophic Wind - Surface Wind - Ice Drift in the Arctic Ocean" using ARGOS buoy data and information from "North Pole" stations. The results of the study will immediately help to improve PIPS calculations.

Studying the relationships between surface wind, geostrophic wind, and ice drift is oriented toward the development of parameterization of dynamic interaction between the atmosphere and ice cover. But the results of investigation will also give valuable information on the influence of internal stress in ice cover.

Ice Dynamics

First priority - dynamical model of anisotropic ice. Until the present time, as far as is known, the only working anisotropic ice model has been developed by the author. Direct measurements of stress and deformation corresponding to specific phenomena of ice dynamics (SIMI and other experiments)along with analysis of Radarsat data has allowed significant improvement in the anisotropic model. The new model now directly brings in small scale ice mechanics and reflects all revealed features of ice cover behavior, some of which recently seemed mutually exclusive.

Attached is a copy of a proposal "Anisotropic Behavior of Ice Cover and its Modelling". The recommended approach allows one to calculate the following parameters of leads: relative area, orientation, spacing between leads, and their width.

We propose the simplest possible formulation of oriented constitutive law, keeping the main features of anisotropic ice behavior and dropping secondary ones. As a result, in a certain sense, the proposed anisotropic approach turns out even simpler than some isotropic models.

Allowing for Floe Size

The degree of ice anisotropy depends upon an aggregated state that determines the possibility of elements composing ice cover to be redistributed under influence of acting forces, especially in the event when signs of principal strains are different.

When ice floe size is included in the mathematical description, the model will successfully work for all range of ice conditions from multiyear pack ice to marginal sea ice zone.

An approach to determine the characteristic size of ice floes is also briefly described in the proposal "Anisotropic Behavior of Ice Cover and Its Modelling". Allowing for ice strength is proposed to be realized on the basis of modeling vertical profile of ice temperature.

Including ice floe dimensions as an internal parameter of the model will also help to improve mathematical description of lateral melting.

Ice Thermodynamics

I believe that the first priority in improving the mathematical model of thermal processes is to specify calculations of ice thickness changes on the basis of parameterizing vertical profile of ice temperature and taking into account the relationship between heat capacity and heat of fusion.

Suppression of Numerical Viscosity

One of main limitations of the present PIPS model is connected with "computational viscosity" that excessively smooths spatial distribution of calculated parameters and does not allow reliable simulation of the ice edge position.

The application of finite difference schemes and other methods to determine kinematic changes in ice cover parameters is connected with computational errors, comparable with calculated changes of the parameters. We propose to increase accuracy of computation by taking into account information on subgrid changes of the parameters: not only mean value of parameter, but also the first moments of its spatial distribution.

Simulation of Ice Edge Configuration

The pronounced features of mathematical modeling ice cover and a number of specific calculating problems are related to the varying configuration of the ice-covered zone.

We propose to consider an original approach to simulate varying configuration of the ice-covered zone. Eiler-Lagrangian representation of the ice edge is based on approximation of the edge by the segments of line. The location of the segment is defined by prescribing the coordinates of the point at the segment and the direction of the segment. This approach allows us to simulate motion of ice edge and take into account features of thermal processes in its vicinity.