User guide

SWIIFT was developed with the goal of studying the impact of ocean waves on sea ice, through their shaping of the floe size distribution.

Sea ice forms in polar regions when seawater freezes, and an ice floe is a piece of floating sea ice. The floe size distribution is a statistical way to describe the sizes of a collection of floes, without having to describe each floe individually. The floe size distribution changes under the action of a collection of factors, including the lateral melt and growth of the ice, the rafting of existing floes, or the fracture of individual floes into smaller floes by the ice internal stress, ocean currents, or the bending stress imposed by waves. SWIIFT focuses on the latter.

We choose to model this phenomenon in isolation, which implies making simplifying assumptions.

Quickstart

SWIIFT can be installed from PyPI:

pip install SWIIFT

It can also be installed from GitHub:

pip install git+ssh://git@github.com/sasip-climate/swiift.git@$ver#egg=swiift

with $ver an optional release tag.

We strongly recommend installing SWIIFT in a dedicated virtual environment.

Overview

SWIIFT was designed to build time-domain simulations of wave fracture. It allows some flexibility in how these simulations are constructed. To achieve this, we rely on a collection of components. Physical objects—such as ice floes—or concepts—such as a fracture mechanism—are abstracted into objects of varying complexity, which can then be combined into higher-level objects.

  1. Experiment – the user-facing interface to simulations. It combines a Domain, a fracture Handler, and add a time dimension. It exposes convenience methods to step through a simulation, save its results, or load existing results from disk.

  2. Domain – the state of an experiment at a given time. It holds informations on the state of the ice (that is, a collection of WavesUnderFloe), the fluid that bears it (a FreeSurfaceWaves object), and forcing waves (DiscreteSpectrum).

  3. Fracture handler – the parametrisation of fracture. It determines whether floes in a domain should break.

  4. DiscreteSpectrum – SWIIFT handles monochromatic or polychromatic sine forcings. A polychromatic is a linear superposition of wave modes. These objects gather amplitudes, frequencies, and phases of these wave modes. These can be automatically built from spectral parametrisations exposed by swiift.api.spectra; attenuation is handled at the domain level. It might be moved to the floe level in the future, to allow for more flexibility.

  5. Floe – from the point of view of SWIIFT, floes are segments, localised in space, that carry information on ice properties (fixed for a given floe). When part of a domain, they are promoted to WavesUnderFloe, and get knowledge of the wave state (that varies along the floe). The latter have associated methods to compute physical quantities deriving from these properties and the wave forcing, such as the the vertical displacement or the elastic potential energy.

  6. Wave attenuation – the presence of floes may attenuate wave propagation. The module swiift.lib.att exposes builtin attenuation schemes, and users can define their own schemes as well.

Examples