Python implementation of the upper ocean structure fitting algorithm described in González-Pola et al. (2007). The SHDR algorithm (Sharp Homogenization/Diffusive Retreat) performs a differential evolution search to fit an upper ocean profile to an idealized analytical form. This analytical form defines a constant region (mixed layer), followed by a region defined by a combination of exponential and gaussian decays (seasonal thermocline/pycnocline) and a linear decay (permanent thermocline/pycnocline). The algorithm allows for a physical characterization of the thermocline/pycnocline, but it can also be used as a robust MLD identifying method.
SHDR is a small package. To keep things simple, we decided to source it as a standalone module in a single file. To use it, download the file SHDR.py and place it in you working directory. This file contains all the routines needed to run SHDR, which should be imported in your preferred programming environment.
A sample call would be
from SHDR import fit_profile # SHDR.py file in working directoy
result = fit_profile(y, z, max_depth=400)The user manual contains extensive information on how to use the module. For a real use case, please see the example jupyter notebook.
python >= 3.6.numpy, pandas, tqdm.
Ensure these packages are installed on your system.
If using pip:
pip3 install --user numpy, pandas, tqdm
If using conda:
conda install numpy, pandas
conda install -c conda-forge tqdm
If you use SHDR in your project, please cite the original work González Pola et al. (2007). The code in this repository can be cited through (TODO: add DOI).