Supplementary material for "Physics-Based Hindcasting of an Induced Earthquake Sequence"

Physics-Based Hindcasting of an Induced Earthquake Sequence

This repository contains the supplementary files, input data, and post-processing scripts for the AGU Journal of Geophysical Research: Solid Earth submission:

“Physics-Based Hindcasting of an Induced Earthquake Sequence”by V. van der Heiden ([e-mail]), L. Buijze, T. Ulrich, J.-D. van Wees, M. van Isselt, L. Matenco, J. Spetzler, A.-A. Gabriel, and Y. van Dinther.

The repository provides MATLAB-based workflows to reproduce the figures and results presented in the manuscript, focusing on physics-based modeling of induced seismicity using a 2.5D on-fault stress framework.

[Code availability]

The open-source software PANTHER is available at:https://github.com/TNO/PANTHER

The geological parameterization framework PhysMmax is available at:https://github.com/loes-buijze/PhysMmax

This repository contains scripts and data required to reproduce the published results:https://github.com/VvanderHeiden/InducedEarthquakeSequence

Specific commits used in the study should be archived via Zenodo and cited accordingly.

[Repository structure]

run_fault_loading_reference_model.mMain script to reproduce the reference model simulations and figures.

/input/Contains input data for the Wirdum fault case:

wirdum_geology.mat: Geological model parameters

200pillars_constoffset_*.mat: Fault discretization and pillar geometry

/functions/MATLAB functions for model setup, stress computation, and visualization:

Stress computation and processing:write_stresses_to_file.m, calculate_nucleation_depths.m, calculate_depth_average.m

Geometry and data handling:rotate_pillars.m, resample_data.m, extract_regularize_filter_data.m

Model adjustments:apply_smoothing_within_offset_regions.m, apply_scaling_within_reservoir.m, update_modelname.m

Visualization:plot_results_2D_continuous.m

LICENSELicensing information for reuse.

[Model description]

This repository implements a 2.5D on-fault stress calculation for the Wirdum fault case using a pillar-based representation. The model applies a plane-strain approximation per pillar to simulate stress evolution and earthquake nucleation.

The workflow couples:

Fault mechanics (PANTHER)

Geological parameterization (PhysMmax)

Stress evolution and nucleation analysis (Day3D criterion)

[Workflow overview]

Load fault geometry and geological parameters

Construct a pillar-based fault representation

Define initial stress state and hydro-mechanical parameters

Assign depth-dependent and depth-independent properties

Apply smoothing and scaling (reservoir vs. full fault)

Run quasi-static stress evolution simulations

Evaluate nucleation and reactivation conditions

Post-process and visualize stress fields and seismic response

[Outputs]

On-fault stress distributions per pillar

Temporal evolution of stress and slip

Nucleation depth and timing estimates

Filtered reservoir-scale stress changes

Publication-quality figures reproducing the manuscript results

[Setup instructions]

Clone this repository:

git clone https://github.com/VvanderHeiden/InducedEarthquakeSequence.git

Clone required dependencies:

git clone https://github.com/TNO/PANTHER.gitgit clone https://github.com/loes-buijze/PhysMmax.git

Open MATLAB (version > 2021 recommended)

Open the PhysMmax.prj project file

Add PANTHER as a referenced project:

Go to Project Tab → References → Add Reference

Select the Panther.prj file

After setup, PANTHER functionality is available within the PhysMmax environment.

[Notes]

The model assumes plane-strain conditions per pillar (2.5D approximation)

Depth-dependent parameters are vertically interpolated

Diffusion and aseismic slip processes can be toggled in the simulation settings