DEPTH-FOCUSING ANALYSIS USING A WAVE-FRONT-CURVATURE CRITERION

Depth-focusing analysis (DFA), a method of refining velocities for pre stack depth migration, relies on amplitude buildups at zero offset to determine the extrapolation depths that best focus the migrated data. Unfortunately, seismic energy from dipping interfaces, diffractions, a nd noise often produce spurious amplitude indications of focusing. To reduce possible ambiguity in the DFA interpretation process, we introd uce a new attribute for determining focusing that is relatively indepe ndent of amplitude. Our approach is based on estimates of the radius o f wavefront curvature. The estimates are derived from normal moveout a nalysis of nonzero-offset data saved during migration. By relating ste eper moveout to smaller radius of wavefront curvature, focusing is def ined by a wavefront curvature of zero radius. Additionally, we show th at applying inverse-radius weights to the amplitude data attenuates no nfocused events due to their large radius of curvature. Using the Marm ousi data set, our weighting scheme resulted in reduced spurious focus ing and enhanced velocity resolution in DFA.