This contains files supporting our paper on building a global statistical burned area model for seamless integration into Dynamic Global Vegetation Models (DGVMs).

Introduction

Our paper, A Statistical Global Burned Area Model for Seamless Integration into Dynamic Global Vegetation Models, will be submitted to Biogeosciences.

Authors:

• Blessing Kavhu
• Matthew Forrest
• Thomas Hickler

Workflow and Code Components

1. Download the Data

Download the large data table from the Zenodo repository and place it in your working directory: https://doi.org/10.5281/zenodo.14110150

2. Plot the Correlation Matrix

Run the following script to generate the correlation matrix plot:

Model_fitting_&_plots/Correlation_matrix_plot_v1.R

3. Fit the Models

Run this script to fit various models and evaluate their performance based on deviance explained and normalized mean error:

Model_fitting_&_plots/Sensitivity_models_v1.R

4. Generate Final Model and Plots

Fit the final model and generate associated plots, including the spatial distribution map and partial residual plots, by running:

Model_fitting_&_plots/Final_model_&_plot_v1.R

5. Plot Interannual Variability

To visualize interannual variability, run:

Model_fitting_&_plots/Interannual_variability_plot_v1.R

Options for handling HDI:

• Exclude HDI
• Keep HDI constant based on initial year values
• Use HDI values for matching periods

6. Plot Seasonal Variability

Run the following script to plot seasonal variability:

Model_fitting_&_plots/Seasonal_variability_plot_v1.R

7. Test for Trends in Burned Area Extent

Analyze trends in burned area extent across different GFED regions by running:

Model_fitting_&_plots/Mankendall_trend_test_v1.R

The required dataset is located at:

Datasets/GFED_trends.csv

8. Plot Spatial Variation in Trends and R-Squared Values

Visualize the following using the GeoJSON file provided:

• Variation in trends of interannual variability by GFED region
• Spatial distribution of R-squared values for annual and seasonal cycles

The file to use is:

Datasets/GFED_regions_&_results.geojson

These plots can be generated in any GIS software. Relevant fields for each parameter are included in the dataset.

Citation

If you use this repository, please cite:

Authors:

• Blessing Kavhu
• Matthew Forrest
• Thomas Hickler

Title: A Statistical Global Burned Area Model for Seamless Integration into Dynamic Global Vegetation Models
Journal: Biogeosciences

Contact

For any questions or issues, please contact: [kavhublessing@gmail.com]