Overview

Objective

The goal of this project is to update, restructure, and consolidate existing USGS ground-motion processing software to incorporate recent advances from researchers at the USGS, PEER, and others. It will standardize tools for multiple USGS ground-motion products and enable scientists within the USGS and the external community to develop and expand ground-motion datasets used in many different applications. Additionally, it will provide a standard interface for local storage of recorded and synthetic ground-motion waveforms and intensity metrics, as well as event and station metadata, in one container that can be easily distributed. The resulting open-source software will provide customizable processed ground-motion waveforms and intensity metrics, while adhering to USGS software standards, including documentation, peer review, testing, and continuous integration.

Diagram of workspace

Motivation

  • Facilitate creation of ground-motion data sets for multiple types of analysis.

  • Leverage best practices from the community to standardize processing algorithms used in ShakeMap ground-motion processing software.

  • Disentangle ground-motion processing (broad range of uses) from specific applications, e.g., ShakeMap generation.

Target Use Cases

  • Standardize processing of ground-motion waveform data, including ground-motion simulations, for deriving a wide variety of Intensity Measure Types (IMTs) for real-time and simulation-based scenario ShakeMap production.

  • Develop uniformly processed ground-motion data sets leveraging data from a variety of sources for use in development and analysis of ground-motion prediction equations, shaking duration, site response, testing of earthquake early warning algorithms, and testing seismic velocity models.

  • Facilitate access to and routine processing of waveform and parametric data from a wide variety of data sources in many different formats, including COSMOS, PEER/NGA, CESMD/VDC, and international strong motion data sets. These parametric data can be formatted per-event or in the style of the NGA “flat” files.

  • Convert files from any of the various strong-motion formats into a standard capable of being read by modern seismological processing software (i.e., Obspy).

  • Facilitate the creation of relational databases containing relevant waveform metadata, stream/station metrics, etc.

Features

  • gmprocess is written in Python and builds upon ObsPy, PyASDF, and SEIS-PROV. These libraries provide fundamental processing functionality, an HDF specification for seismic data, and a seismic standard for tracking data provenance.

  • The functionality can be accessed via Python libraries or through command line programs.

  • We currently support Mac and Linux, and Windows systems.

  • We provide file readers for many strong motion data formats that are not otherwise supported in ObsPy.

  • We provide subclasses of ObsPy’s Trace and Stream classes, which are designed to aid analysis and metadata storage and validation specifically for ground motion data that is organized by event.

  • We use the ASDF format to store earthquake metadata, station metadata, raw ground-motion time histories, processed ground-motion time histories, waveform and station metrics, and provenance information in a single, portable file.

  • Import data from local filesystem using a wide variety of formats or retrieve data using web services from FDSN data centers.

  • “Plug-and-play” architecture for efficiently evaluating data reprocessed with new or alternative algorithms.