As I understand, HVSR is commonly computed from ambient noise and earthquake records using velocity sensors. We now aim to compute HVSR from strong motion data, i.e., accelerograms. Is this approach reliable for inverting HVSR to estimate shear wave velocity?