In this short note we cover two questions concerning the inclusion in seismic FWI of an attempt to solve simultaneously for elastic and anelastic geological properties, i.e., Q P and Q S as well as (say) V P , V S and ρ . The first question is whether or not to do so: anecdotally (and incorrectly), attenuation parameters have been suggested to be unimportant to FWI, since they dominate at high frequencies and FWI is primarily concerned with low. The second is really several questions: where in the mechanisms of FWI (e.g., within the machinery of the gradient and the various approximate Hessians), are the tasks we normally associate with Q contained? Is there a component of the Hessian operator tasked with Q compensation, for instance.
View full article as PDF (1.83 Mb)