Python script accompanying:
W.Edeling, D. Crommelin, "Reducing data-driven dynamical subgrid scale models by physical constraints", (submitted) 2019.
Recent years have seen a growing interest in using data-driven (machine-learning) techniques for the construction of cheap surrogate models of turbulent subgrid scale stresses. These stresses display complex spatio-temporal structures, and constitute a difficult surrogate target. In this paper we propose a data-preprocessing step, in which we derive alternative subgrid scale models which are virtually exact for a user-specified set of spatially integrated quantities of interest, i.e. for time series. The unclosed component of these new subgrid scale models is of the same size as this set of integrated quantities of interest. As a result, the corresponding training data is massively reduced in size, decreasing the complexity of the subsequent surrogate construction.
This research is funded by the Netherlands Organization for Scientific Research (NWO) through the Vidi project "Stochastic models for unresolved scales in geophysical flows", and from the European Union Horizon 2020 research and innovation programme under grant agreement #800925 (VECMA project).
- Python 3
- Numpy
- Scipy
- Matplotlib
- h5py
The main file reduced_sgs.py contains the entire script. As input it takes json files generated by generate_input.py. Two input files have been pregenerated. To execute, simply type
python3 ./inputs/EZ_track.json
in order to execute a simulation which tracks the reference energy and enstrophy for a simulated time of 10 years. The command
python3 ./inputs/E_TK_E_track.json
launches a simulation where the full energy and just the energy close to the cutoff scale are tracked seperately.