This tutorial shows the basic functionality of the coarse-graining codebase.

Tutorial

It includes:

• a python script to generate a sample dataset: generate_data.py
• a python script to interpret the coarse-graining results: process_results.py
• this tutorial outline

About the sample data

• generate_data.py script creates a file (velocity_sample.nc) that contains x- and y- velocity components
• doubly-periodic Cartesian grid
• domain is 400km in x by 150km in y
• velocity field is a 'wavy jet' with noise

What to do

1. Compile Case_Files/coarse_grain.x (see notes below) and copy into this directory.
2. Create the sample dataset (python generate_data.py)
3. Run coarse_grain.x (./coarse_grain.x --input_file velocity_sample.nc --filter_scales "1e3 15e3 50e3 100e3").
4. Run the python plotting routine (python process_results.py)

When specifying filtering scales, only pick a few. Other tutorials will cover how to handle outputs with many filter scales. A suggestion is: 1e3, 15e3, 50e3, 100e3. While the codebase is heavily parallelized, it unnecessary for this example and will be covered in later tutorials, and so we simply run in serial here.

Notes when compiling

When compiling, use the constants.hpp file provided in this directory.