Data and code for "Effect of initial source mixing on lock-release turbidity currents with application to deep-sea mining collector plumes"

James, Jellinek & Topf 2026 This repository contains the data used in the research paper: James C. B. G, Jellinek A. M. and Topf H. S. Effect of initial source mixing on lock-release turbidity currents with application to deep-sea mining collector plumes In review with Elementa: Science of the Anthropocene. Repository organization james-etal-2026-lock-release │ ├───analysis/ │ ├───matlab/ # pre-processing scripts to extract data from side-view videos │ │ └─── *.m │ └───python/ # further processing and analysis scripts to convert *.csv to *.npy and calculate mean shape │ └─── *.py │ ├───data/ │ ├───datafiles/ # .npy experiment files │ └─── *.np │ └───metadata/ # .csv files for own experiments and literature comparisons │ └─── *.csv │ └───paper/ ├───figure_scripts/ # scripts that read data/ and write to figures/ │ └─── *.ipynb └───figures/ # output figures └─── *.png Data organization The CSV file data/metadata/experiments_summary.csv offers a summary of all runs and corresponding experimental parameters. The folder data/datafiles contains 51 .npy files containing processed data from each lock release experiment captured via side-view video. Raw videos were processed in MATLAB to extract time-evolving front position, area, and height profiles. A second processing stage computes the mean and standard deviation of the current shape across time. Each file is a NumPy object array containing a single Python dictionary with the following fields: Time series time_s — Time vector, in seconds. front_m — Streamwise position of the current front, in meters. area_m2 — Cross-sectional area of the current, in m². height_m (n_time, n_x) — Height of the current at each time step and spatial position, in meters. Computing mean shape (1D, length n_x) xcenters — Streamwise positions of spatial grid cell centers (pixels) in meters height_stack (n_frames, n_x) — Stacked height profiles used to compute the shape statistics. Contains NaN values where data is absent. n_frames may differ from n_time as it represents a subset of time steps. In meters. av_shape — Time-averaged height profile (mean shape) of the current as a function of streamwise position, in meters av_shape_std — Standard deviation of the height profile across time at each streamwise position, in meters Getting Started Downloading the repository From Zenodo (recommended): Download and unzip — no additional steps needed. From GitHub: This repository uses Git LFS to store .npy data files. Install Git LFS before cloning: Mac: brew install git-lfs Linux: sudo apt install git-lfs (or equivalent) Windows: Download from https://git-lfs.com/ Then: git lfs install git clone https://github.com/CaraBGJames/james-etal-2026-lock-release.git cd james-etal-2026-lock-release git lfs pull Environment conda env create -f environment.yml conda activate env-james2026lock Loading the Data import numpy as np data = np.load('filename.npy', allow_pickle=True).item() time = np.abs(data['time_s']) # (n_time,) front = np.abs(data['front_m']) # (n_time,) area = np.abs(data['area_m2']) # (n_time,) height = np.abs(data['height_m']) # (n_time, n_x) x = np.abs(data['xcenters']) # (n_x,) av = np.abs(data['av_shape']) # (n_x,) av_std = np.abs(data['av_shape_std']) # (n_x,) stack = np.abs(data['height_stack']) # (n_frames, n_x)