MASW is a common processing method for civil/geotechnical and even for oil/gas exploration (usually weathering static corrections) applications. In my experience, it seems to work well for most environments and various subsurface conditions.

The more I learn about FWI and MASW (for either Raleigh- or Love-waves), the more I realize that perhaps one day FWI methods will surpass any MASW processing because it attempts to overcome the limitations naturally imposed by MASW because of its inherent assumptions of subsurface conditions.

To date, I cannot find a case study that compares both methods in a single environment. Also, I realize that deriving a reliable source function, recording the 3-D seismic wavefield, and overcoming the stability constraints associated with multiparameterization of FWI (especially for elastic wave phenomenon) is task to be handled with in the future.

My questions are: (1) Has anyone seen a paper dealing with comparing both methods - if so, where? (2) Will some form of FWI possibly surpass MASW workflows (e.g. ParkSeis or SurfSeis) in the near future?

  • $\begingroup$ I'm not on the geology side and can't answer this, but I want to point out that you should write out fully what the acronyms MASW and FWI are at least once. $\endgroup$ – Jareth Holt May 4 '19 at 14:53

I think there are some initial exploratory abstracts out there on the topic, e.g., http://www.earthdoc.org/publication/publicationdetails/?publication=75396 . But not much yet.

My view is that that MASW (multichannel analysis of surface waves) is simply so much more robust in practice compared to FWI (full waveform inversion, which really doesn't work as spectacularly well as some papers may suggest; the method requires a LOT of iterative and hands-on work, and is furthermore extremely computationally demanding, without always giving dramatically better results compared to travel-time tomographies...)... hence, I think that MASW will remain the standard in the coming future, until FWI truly becomes a 'cheap' and 'hands-off' method, which could easily take another 15+ years, I imagine...

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.