Layered Velocity Model From Kinematic Attributes

In conventional processing, the classical algorithm<br>of Hubral and Krey is routinely applied to extract<br>an initial macro-velocity model that consists of a<br>stack of homogeneous layers bounded by curved<br>interfaces. Input for the algorithm are identified<br>primary reflections together with NMO velocities<br>derived from a previous velocity analysis conducted<br>on CMP data. This work presents a modified<br>version of the Hubral and Krey algorithm that is<br>adapted to advantageously use previously obtained<br>CRS attributes as its input. A simple synthetic<br>example is provided to illustrate and explain the<br>implementation of the method.