Overview

Pyatoa is a waveform assessment package designed to compare one set of wiggles to another set of (similar) wiggles. Core functionalities of the package are:

• Waveform standardization and preprocesing (with Obspy)

• Time windowing (with Pyflex) and adjoint source generation (with Pyadjoint)

• Hierarchical data storage of waveforms, metadata, and measurements (with PyASDF)

• Bulk measurement aggregation and analysis (with Pandas)

• Waveform and measurement plotting

• Built-in interface with the automated workflow tool: SeisFlows

In a nutshell: Provide two similar waveforms and some configuration parameters, receive misfit windows and adjoint sources. Do this in bulk and analyze all returned misfit in aggregate.

Why is Pyatoa necessary?

If Pyatoa simply wraps a number of already-working dependencies, why do we need it?

Pyatoa provides an easy-to-use tool that acts in liue of the inevitable collection scripts a User would have to write to link the data collection capabilities of ObsPy with the windowing functions of Flexwin/Pyflex with the adjoint source generation of Pyadjoint.

It also provides structure for these processes, to ensure that data-synthetic misfit is calculated properly. Similarly, useful routines within the package provide assistance during seismic inversions.

What doesn’t Pyatoa do?

Pyatoa does not: generate sythetic waveforms, interface with HPC systems, gather seismic data. These capabilities are supplemented by other packages, some of which are hosted within the adjTomo ecosystem.

Recommended tools to use with Pyatoa:

• Gather seismic data: PySEP

• Generate synthetics: SPECFEM

• Automate inversions on HPCs: SeisFlows

How do (I use) Pyatoa?

Pyatoa was written following the design philosophy of ObsPy i.e., called through a Python interface or within Python scripts, Jupyter notebooks or other Python packages. That is:

from pyatoa import Manager

mgmt = Manager()

Check out First Glance for a short code-based introduction to Pyatoa.