Example Dataset

Example Dataset: Peace of Mind (POM) Array, Unalaska Island, Alaska

The example dataset is from the Peace of Mind (POM) Array, deployed on Unalaska Island in 2015-2016. The array consisted of 15 short-period seismometers spaced over an area of around 1 km. The array was deployed as part of the Aleutian Array of Arrays project (9C network; Ghosh, 2014). There were two other arrays deployed on Unalaska Island at the same time: 2A and 3A.

During the time of the deployment, there were over >800 M3.0+ earthquakes within 600 km of the three arrays. This includes a M6.9 that occurred on July 27th, 2015. Due to the high seismicity in the region, this acts as a good example to test array performance with earthquakes and identify systematic errors.

POM MSEEDS

Miniseed files for all M3.0 events within 600 km of the POM array are provided in the folder POM_earthquakes_mseeds. Each .mseed file uses the naming convention of <event_id>.mseed, where the event id is the one provided by the USGS (ex. us10002ngf.mseed). Each mseed file contains traces for all POM stations beginning 60 seconds before the expected p-arrival time calculated from a 1D velocity model and going for another 60 seconds. The data is in raw instrument counts with no filtering.

Parameter file

The file input_parameters.yaml is set up to run for this dataset and pulling the data locally. It is not set up to save any data, but it will print commands as it's running and return 5 plots. When initiated the first time, it should take less than a minute to run.

It can be run with the following commands:

conda activate arrayseis
python array_aggregator.py input_parameters.yaml

Plots

The file is set up to return 5 plots:

Map of earthquakes in vicinity of POM array.
Backazimuth error (catalog backazimuth - array backazimuth) as a function of catalog backazimuth. Includes the Niazi cosine fit.
Slowness error (1D velocity model slowness - array slowness) as a function of catalog backazimuth. Includes the Niazi cosine fit.
Backazimuth error plotted in map view. The circles are the true earthquake locations and the vectors are pointing in the direction of the array processed backazimuth. The length of arrow corresponds to the amount of error.
Slowness error plotted in map view. Vectors indicate whether it's a positive or negative slowness error (positive = up, negative = down). The length corresponds to the amount of error.

Below are what the plots should look like if using the provided parameter file.

Earthquake Map

alt