Jim Overland Jackie Richter Menge Lyn McNutt PMEL/NOAA CRREL Univ. of Alaska
Goal: Provide an observational basis for formulation and interpretation of a high resolution 0(10km) PIPS 3.0 sea ice model.
1. Sea Ice as a Hardening Plastic
Based on SIMI observations, we have had success in validating that sea ice behaves as a granular material (Overland et al, in press). Our study is based on analysis of AVHRR imagery, 11 drifting buoys with GPS navigation, and SAR motion vectors in an 80x500 km swath north of Alaska. At moderate atmospheric forcing, i.e. windstress multiplied by fetch, the ice appears to fail along sliplines which occur at an acute angle to each other and to the wind forcing. With longer fetch, the ice appears to fail in compression perpendicular to the wind direction. Beaufort sea ice tends to move in 20-150 km rigid aggregates, separated by linear deformation zones or sliplines. In one case of onshore winds, a compressional wave could be tracked from the coast to the SIMI camp, a distance of 450 km. In a separate case, the lateral confining stress was insufficient to contain the accumulated onshore stress.
The continuum model for a granular material is a hardening plastic. The mathematics was developed in the nineteenth century and the body force was gravity. In our case, the body force is wind stress. We an credit the AIDJEX and Hibler models for a correct mathematical formulation. The concept of a yield curve models slipline behavior, P* suggests the correct aggregate size and compressional wave speed, and changes in the thickness distribution suggest P* as a hardening parameter. The zero order issue for PIPS 3.0 is sufficient numerical resolution 0(5 km) to solve the continuum equations.
Our analysis suggest that the continuum length scale is 0(10 km). This is based on the width of sliplines and that the CRREL stress measurements were better correlated with buoy measurements at the 10 km scale than the 5 km scale (Richter Menge et al, 1996). At 5 km, the floes had more of an independent jostling motion. Plastic-like behavior is already suggested in the high resolution modeling by the Semtner group at NPS.
There are still some conceptual difficulties. For example, the angle between sliplines is less than that predicted by theory. There is work to be done on the P* - thickness distribution relationship. This issue involves modeling the ridging process due to shear. Energetics arguments are one approach (Ukita and Moritz, 1995; and others).
2. Atmospheric Forcing
Systems are strongly coupled when their space, time and velocity scales match. The atmosphere tends to have large space scales 0(500 km) and rapid time scales (days). Ice velocities tend to be measured in kilometers per day. If velocities are the main variable to be predicted, the present PIPS model is OK. However, there is a scale matching between the atmosphere and ice deformation to be resolved by PIPS 3.0 (Overland et al, 1995) .
Our view is that the continuum model would run for the full basin or partial basin 0(500-1000 km). If one wanted, one could embed a discrete granular model 0(100 km) within the continuum model.
Aggregates are of the same size as sectors of large storms. Our results from LEADEX and SIMI suggest that fronts and small rapidly moving storms such as polar lows move too fast to load the stress field; what is required is larger air mass regions with time scales of 3 days or greater. Thus, projected atmospheric forecast models (resolution 50 km) should be sufficient to drive sea ice models at 10 km resolution.
To calculate air-ice stress, one needs a geostrophic or "surface" wind, and a drag coefficient which is a function of surface roughness and stratification. These are often calculated internally to atmospheric forecast model and may not be tuned for the Arctic. Over the central Arctic pack, increased roughness is offset by increased stratification. Near marginal ice zones stratification can be reduced and drag coefficients can increase dramatically.
3. Our perspective on developing algorithms for PIPS 3.0.
We propose to work at the interface between understanding and interpreting sea ice as a granular material and the modeling formulation of sea ice as a hardening plastic. We thus wish to be involved with validation. We propose a validation experiment for winter 1999-2000. We would deploy six GPS navigation buoys at the continuum scale 0(10 km) and five buoys at the deformation scale 0(40 km). These buoys would be equipped with real time reporting stress sensors and supporting meteorology. We would interpret the deformations and stresses in relation to SAR motion vectors from RADARSAT (500x500 km) and atmospheric forcing from Navy models. We would make use of the Ice Centre of Environment Canada SAR floe Tracker software and coordinate with the National Ice Center.
References:
Overland, J.E., S.L. McNutt, S. Salo, J. Groves and S. Li, 1998: Arctic sea ice as a granular plastic. J. Geophys. Res. In press.
Overland, J.E., B.A. Walter, T.B. Curtain and P. Turet, 1995: Hierarchy and sea-ice mechanics: a case study from the Beaufort Sea. J. Geophys. Res., 100, 4559-4571.
Richter-Menge, J.A., B.C. Elder, J.E. Overland, and S. Salo, 1996: Relating Arctic pack ice stress and strain at the 10 km scale. In Proceedings of the ACSYS Conference on the Dynamics of the Arctic Climate System WMO/TP No. 760, 327-331.
Ukita, J. and R.E. Moritz, 1995: Yield curves and floe rules of pack ice. J. Geophys. Res., 100,
4545-4557.