In order to understand and forecast the sea ice retroactions on climate evolution, we need to simulate its dynamic. We present here the modeling process and discretization strategy developed to build a simulation tool: FloeDyn. This computation code manage the dynamic simulation of sea ice at floes scale. The goal was to build a tool enable to simulate sea ice at elementary scale (floes) and be able to link these simulations with large scale simulations via an averaging process to obtain large scale rheologic parameters but also to obtain characterization of packs (floes groups) from large scale simulation in order to perform elementary scale simulations to take into account local detail in a large scale framework.
In this part, first we will explain the modeling choices and mathematical bottlenecks. After, we will present the discretization process based upon non-regular mechanic to manage contacts and Griffith approach to take into account floes fracturation. Finally, we will detail the strategies developed to optimize computation performances in order to enable the largest number of floes in simulation and give some examples of computations.
Romain Duboscq, Ariane Trescases
