array_aggregator

Code for assessing array accuracy on a set of earthquakes. Final project for GEOS694 Computational Geosciences (Instructor: Dr. Bryant Chow).

Version: 1.0.0

What

This code is for assessing the accuracy and precision of small-aperture seismic arrays at recording earthquakes or other seismicity.

How

This code does the following in the specified order:

1. Grab earthquakes/events in the vicinity of a seismic array for a set time period.
2. Compute the "real" backazimuth and slowness based on the catalog location and a given velocity model.
3. Use an STA/LTA scheme to idenfity whether the event "triggers" the array, following the parameters in the EPIC algorithm of ShakeAlert. This includes different handlings for events with multiple triggers and events with no triggers.
4. Compute array processing for a specified time window and waveform filtering around STA/LTA trigger. Current array processing algorithms incorporated are: Least Trimmed Squares (Bishop et al., 2020), Least Squares with cross-correlation (Bishop et al., 2020), and Frequency-Wavenumber (Obspy).
5. Plotting tools for comparing the catalog backazimuth and slowness to the array calculated backazimuth and slowness. These include:

• map of earthquakes in vicinity of array
• backazimuth error (catalog backazimuth - array backazimuth) as a function of the catalog backazimuth.
• slowness error (catalog slowness - array slowness) as a function of the catalog backazimuth.
• map of backazimuth error
• map of slowness error

The following graphic shows the general workflow of the code.

Why

Small aperture arrays have grown in interest in the earthquake monitoring and Earthquake Early Warning communities in recent years. This is largely due to an arrays ability to provide accurate backazimuth and slowness estimates of incoming seismic energy. This code provides the ability to assess how accurate these estimates are, check a number of input parameters, and identify systematic patterns in error and provide possible correction criteria.

Installation

Create a fork of this repository to your github. Then, create a local repository through your terminal. For example:

git clone <github_link_to_repo>

A conda environment with the needed dependencies are provided in environment.yml. This can be created with the following command in your local repository:

conda env create -f environment.yml

This will create a environment called arrayseis with the following dependencies:

• python3
• obspy
• pygmt
• pyproj
• numpy (2.14)
• lts_array
• numba

Usage

Once your python environment is installed, this can be activated as:

conda activate arrayseis

The input parameters for the code can be found in the input_parameters.yaml file. See the Input Parameters Github Wiki Page for documentation on different inputs.

In your local repository, the main code can be run as:

python array_aggregator.py input_parameters.yaml

Repository Contents

File	Description
array_aggregator.py	Main script for comparing USGS catalog eathquake locations to array processed locations. Includes some common plotting options.
input_parameters.yaml	Input parameters for the array_aggregator script. Explanation of possible inputs are in the Github Wiki.
array_functions.py	General functions used in the main script for array processing.
environment.yml	Conda environment file for creating a local environment.
array_figures.py	Aggregate of different plotting functions that can be called in the main array_aggregator.py script.
array_maps_pygmt.py	Different map plotting options using pygmt library.
POM_earthquakes_mseeds	Example dataset of earthquakes (.mseed) from the Aleutian Array of Arrays project. Dataset is from 2015-2016. See Github Wiki for more information.
green-purple.cpt	Cpt file used in one of the plotting scripts for array derived slowness.
./figures	Folder containing example figures of expected output from provided input_parameters.yaml.

Example

An example from an array in the Aleutian Islands is provided in input_parameters.yaml. Please see the Example Dataset Github Wiki page to learn more about the dataset and expected outputs.

Project Task 1

For project task 1, I will be implementing parallel/concurrency into my code.

Project Task 2

For project task 2, I will be doing a parameter input system (input_parameters.yaml). In addition, I will also doing Advanced Documentation (Github Wiki).

For class reviewers

The relevant python file for class reviews is array_aggregator.py. The other main file is array_functions.py, but I'll leave this up to your discretion.

Acknowledgments

This toolkit relies on the array processing toolkit LTS Array (Bishop et al., 2020). This toolkit also heavily relies on the Obspy package and the FK array processing algorithm within (Beyreuther et al., 2010).