README

This repository accompanies the publication of ’Extreme weather magnifies the effects of forest structure on wildfire, driving increased severity in industrial forests’ in [insert journal name after publication] by Jacob I. Levine, Brandon M. Collins, Michelle Coppoletta, and Scott L. Stephens. link

Repository Structure

The repository is organized as follows:

• data/: Contains raw and processed datasets used in the study (though see note below).
• code/: Includes fall scripts used for data processing, analysis, and visualization.
• plots/: Provides all main text and supplementary figure files, including intermediate, unpublished, and illustrator files
• Manifest.toml/Project.toml: julia package versioning and session information for reproducability.

Note on data availability

The data were too large to upload and version control using Git and Github. Instead, we provide a .zip repository of the data as a separate Zenodo submission at doi:10.5281/zenodo.15848845. To facilitate data transfer, the .zip file was split into four subfiles, each less than or equal to 2.1 Gb.

To download and merge these components, either navigate to the Zenodo submission directly, or using bash:

#!/bin/bash

# Download and reassemble split archive from Zenodo
for part in aa ab ac ad; do
  wget -c https://zenodo.org/record/15848846/files/data_part_$part
done
cat data_part_aa data_part_ab data_part_ac data_part_ad > data.zip

# Unzip into the main directory, so that the scripts will function as intended.
unzip data.zip -d severity_and_silviculture/data

To use the directory, simply copy the dataset into this repository and label it “data/”. The scripts will then reference individual files within this directory.

While the repository on Zenodo contains most of the raw data used in the analyses described by the paper, we were unable to provide some raw data for direct download. In particular, we could not provide raw LiDAR or forest ownership data. However, we do provide all intermediate data, meaning the processed LiDAR data and ownership data are available within our study area (defined as the boundaries of the Plumas National Forest and adjacent private land). The absence of this data means that some of the data cleaning scripts will not function properly. However, we retain these scripts as documentation of the processing methodology, and all data analysis and visualization scripts should work as intended.

Requirements

Running the code in this repository requires functioning distributions of both R and julia. The analyses reported in the accompanying manuscript were run using R version 4.4.1 and julia version 1.11.1. All session and package versioning information for julia is available in the Manifest.toml and Project.toml files.

Usage

To reproduce the analyses and figures from the paper, run the scripts contained in the code/ repository in the order indicated by the file names (noting that some of the scripts in code/01_data_cleaning/ will not run because of raw data availability). Re-running the scripts will overwrite some of the output files in both the data/ and plots/ directories. Unless the user makes major changes, these overwritten files should be similar (in most cases identical) to existing files.

Citation

If you use the original data or methods in your research, please cite the associated paper:

Jacob. I Levine, Brandon M. Collins, Michelle Coppoletta, Scott L. Stephens. (2025). Extreme weather magnifies the effects of forest structure on wildfire, driving increased severity in industrial forests. [insert journal name after publication]. DOI: [insert DOI after publication]

Contact

For questions or issues, either message Jacob Levine through Github (@jsilevine), or contact:

• Jacob I. Levine
• jacob.levine@utah.edu
• Wilkes Center for Climate Science and Policy, University of Utah, Salt Lake City UT
• levine-ecology.com