Supplementary material to "Multi-Scale Sea Ice Kinematics Modeling with a Grid Hierarchy inCommunity Earth System Model (version 1.2.1)"

Shiming Xu, Jialiang Ma, Lu Zhou, Yan Zhang, Jiping Liu, Bin Wang2020 · DOI

Introduction

Sea ice is a dynamic and complex component of the Earth's climate system. Rather than a continuous solid sheet, the Arctic ice pack behaves as a collection of discrete floes that interact through grinding, cracking, and ridging. These kinematic processes occur across a vast range of scales, from individual leads (cracks) to basin-wide drift patterns. Accurately simulating these multi-scale kinematics is crucial for understanding heat and moisture exchange between the ocean and atmosphere. This supplementary material details the numerical validation and sensitivity analysis of a multi-scale sea ice modeling approach implemented within the Community Earth System Model (CESM) version 1.2.1.

Methods and Model Configuration

The study utilizes a grid hierarchy within CESM to bridge the gap between coarse climate scales and fine-scale ice fracturing. The sea ice component employs an Elastic-Viscous-Plastic (EVP) rheology to simulate the mechanical behavior of the ice.

A central focus of the numerical experiments is the NDTE parameter, which represents the number of sub-steps (sub-cycling) used to solve the ice momentum equation within a single global time step. Increasing NDTE allows the model more "numerical patience" to resolve the non-linear stresses associated with ice deformation. Three primary configurations—TS045, TS015, and TS005—are tested, representing a progression in spatial resolution (likely 0.45°, 0.15°, and 0.05° respectively).

Atmospheric Forcing and Validation

To ensure the model's results are grounded in physical reality, it is driven by historical atmospheric patterns. The primary driver of Arctic climate variability, the Arctic Oscillation (AO), was extracted from 50 years of NCEP Reanalysis wintertime Sea Level Pressure (SLP) data using Principal Component Analysis (PCA).

Spatial pattern and principal component time series of the Arctic Oscillation (AO) mode.

The spatial pattern (top panel) displays the characteristic dipole structure of the AO, accounting for 13.9% of the variance. The associated Principal Component time series (bottom panel) captures the decadal oscillations observed during the late 20th century, providing a robust forcing field for the sea ice simulations.

Numerical Convergence and Kinematics

The core of the analysis involves examining the statistical "fingerprints" of simulated sea ice motion. The researchers plotted the probability distributions of shear, divergence, and total deformation on log-log scales. Real-world ice fracturing typically follows a power-law distribution, appearing as a linear trend on these plots.

Statistical distributions of sea ice deformation rates for various model configurations.

The results indicate that the model successfully captures deformation rates spanning four orders of magnitude ($10^{-4}$ to $10^0$ per day). A key finding is the sensitivity to the NDTE parameter. As NDTE increases (e.g., from 120 to 960), the "tails" of the distributions extend further, meaning the model captures significantly more high-magnitude, localized cracking events. Low NDTE values tend to over-smooth the ice, missing the intense deformation critical for climate exchange. In the highest resolution configuration (TS005), the distributions for NDTE=480 and NDTE=960 begin to converge, suggesting that the model is approaching a numerically stable representation of the kinematics.

Conclusion

This validation study demonstrates that the multi-scale CESM configuration can realistically simulate the complex fracturing of Arctic sea ice when driven by accurate atmospheric forcing. The sensitivity tests highlight that capturing extreme deformation events—the "shattering" of the ice pack—requires sufficient numerical sub-cycling (high NDTE). While higher resolution and sub-cycling increase computational costs, they are essential for representing the small-scale features that dictate large-scale sea ice behavior. Future work will focus on optimizing these parameters for long-term climate simulations and verifying the spatial alignment of simulated features with satellite observations.

Summary updated

BibTeX
@misc{xu2020supplementary,
  author = {Shiming Xu and Jialiang Ma and Lu Zhou and Yan Zhang and Jiping Liu and Bin Wang},
  title = {{Supplementary material to "Multi-Scale Sea Ice Kinematics Modeling with a Grid Hierarchy inCommunity Earth System Model (version 1.2.1)"}},
  year = {2020},
  doi = {10.5194/gmd-2020-160-supplement},
}
Figure