It would be great to perform the frequency-time analysis (FTAN) from within ObsPy. Is there any plan to implement this feature?
ObsPy can calculate misfits in the time-frequency domain: https://docs.obspy.org/tutorial/code_snippets/time_frequency_misfit.html
Thanks @krischer, that's indeed useful, but I'm more interested in plotting the dispersion curve of a signal and measuring travel times on several frequency (period) intervals...
Okay - that would have to be tacked onto that - we have nothing built-in in ObsPy to do that. But you can use the wavelet transform used in the misfits to construct a TF representation or use the same things that st.spectrogram() is using.
If I remember correctly @ThomasLecocq had something to automatically pick something in dispersion curves.
yep, in https://github.com/ThomasLecocq/msnoise-tomo we have an FTAN code that works OK, but the code is not ready for production yet (have to find some time). And yes, the FTAN should at some point be moved down in obspy :)
Thanks @ThomasLecocq, I'm looking forward to see your code integrated in ObsPy!
Most helpful comment
yep, in https://github.com/ThomasLecocq/msnoise-tomo we have an FTAN code that works OK, but the code is not ready for production yet (have to find some time). And yes, the FTAN should at some point be moved down in obspy :)