Development of nonlinear dynamics-based techniques for sea ice diagnosis

By incorporating the latest nonlinear Lagrangian methods to describe sea ice dynamics, we will gain new insights into the Arctic sea ice system and a better understanding of the interaction between the dynamics and the sea ice conditions. To ensure that we identify the underlying flow structures responsible for the material deformation of sea ice, we rely on objective (i.e., Euclidean frame-indifferent) metrics. Without the need to identify an appropriate inertial frame, we can achieve similar insights from gridded and sparse flow data (e.g., remote-sensing, model, and buoy data). Our research will utilize a three-pronged approach. First, we will develop a network-topology based method to assess the similarity of two dynamical systems (such as sea ice or ocean velocity fields). By comparing the network of hyperbolic Lagrangian Coherent Structures (hLCS), we quantify the complexity and connection of the underlying flow skeleton through the topology of nodes and edges. Second, we will use trajectory clustering to develop the first dynamics-based map of the Pan-Arctic. We will develop this dynamic map using two types of Lagrangian trajectory data: IABP buoys, and synthetic trajectories from the novel NeXtSIM model. This clustering will help define dynamically coherent regions, and diagnose changes in regional dynamics as ice thickness changes over yearly and decadal timescales. Finally, we plan to advance the analysis of experimental buoy campaigns and long-term IABP trajectory data by conducting an exhaustive analysis of the newly developed sparse Lagrangian trajectory diagnostics, the relative Trajectory Stretching Exponent and the rotation angle. This will be conducted with NeXtSIM sea ice model data where sea ice drift can be compared with ice characteristics, as well as a focused experimental SAR campaign.