Green’s functions, from model to waveform#
pygrnwang · computational seismology
Build Green's-function libraries with Wang's Fortran solvers, synthesize seismograms and static deformation, and keep model, source, time and coordinate conventions explicit.
Run a small, complete example: prepare AK135, compute a library, read three-component displacement and save a figure.
Compare static and dynamic workflows, spherical and layered models, and supported observables.
安装程序、运行同一套 QSEIS2025 示例，并理解输出文件和坐标约定。
A complete workflow#
Prepare the model. Define layered velocities, density, attenuation and the source/receiver geometry.
Compute the library. Preprocess the solver input, run the backend and convert its output where required.
Synthesize observables. Select a source mechanism, distance, sampling and output quantity.
Interpret the result. Check component order, amplitude normalization and the time origin before comparing observations.
The scientific conventions are part of the interface. In particular, vector and tensor orders, time reduction and source-duration units vary across backends.
Compare displacement and stress#
The fresh backend comparison shows displacement from all five dynamic backends and six-component QSEIS2025/QSSP2020 stress at 300, 600 and 900 km. It includes matched-source controls, downloadable arrays and a script to replot the results. 中文说明.
What is included#
The package wraps EDGRN/EDCMP, QSEIS06 (deprecated), QSEIS2025, SPGRN2012 (deprecated), SPGRN2020 and QSSP2020. Python provides preprocessing, serial and parallel orchestration, library lookup, source synthesis, coordinate transforms and signal processing. TauP travel times run through Java subprocesses or an ObsPy fallback.
Python 3.9 or newer is supported. Binary wheels cover the platforms listed in the installation guide. This site follows main; the package version appears in the header.
Getting started
Reference and development