Syngine - On-Demand Synthetic Seismograms
The IRIS DMC recently released the Synthetics Engine (Syngine), an on-demand synthetic seismogram service (http://ds.iris.edu/ds/products/syngine) that complements the time series data IRIS has traditionally distributed. The highly customizable synthetics allow for any source-receiver geometry for a variety of source mechanisms.
Currently nine global models are in Syngine with durations ranging from 30 minutes to 5 hours and resolutions ranging from 1-100 to 20-100 s periods. Synthetic seismograms are available for any source-receiver path from any source depth down to 700 km and include attenuation, radiation pattern, and honors all boundaries within the Earth model. Support for using network and station codes make accessing the synthetics service similar to accessing data from the DMC via our FDSN dataselect service.
Under the hood
Syngine relies on pre-computed global TB-scale databases of Green’s functions using the 3D axi-symmetric spectral-element method AxiSEM. Behind the scenes, the web service runs Instaseis, a system that rapidly calculates broadband synthetic seismograms from the pre-calculated Green’s functions.
Syngine can help users quickly:
- plan array configurations (see figure)
- explore the effects of source parameters (e.g. strike, depth, mechanism)
- perform data quality assessment (e.g. polarity, approximate phase amplitude, etc.)
- compare synthetics from our reference models with your own synthetics
Web service interface
Help making requests
Command line client: FetchSyn
Together with the Instaseis development team, we are currently adding support for finite-fault models, variable width source-time functions, and custom source-time functions. We are also exploring including 3D regional models to add to our model database in addition to global models with continental and oceanic (with water!) crusts. Finally, we’re working to add more post-processing features using ObsPy in our FetchSyn command line client which will enable users to perform common processing steps, such as filtering, calculating envelopes, spectra, figures, etc.
by Alex Hutko (IRIS Data Management Center) and Chad Trabant (IRIS DMC)