We introduce a new partial prestack-migration operator called ''azimut
h moveout'' (AMO) that rotates the azimuth and modifies the offset of
3-D prestack data. Followed by partial stacking, AMO can reduce the co
mputational cost of 3-D prestack imaging. We have successfully applied
AMO to the partial stacking of a 3-D marine data set over a range of
offsets and azimuths. When AMO is included in the partial-stacking pro
cedure, high-frequency steeply dipping energy is better preserved than
when conventional partial-stacking methodologies are used. Because th
e test data set requires 3-D prestack depth migration to handle strong
lateral variations in velocity, the results of our tests support the
applicability of BMO to prestack depth-imaging problems. AMO is a part
ial prestack-migration operator defined by chaining a 3-D prestack ima
ging operator with a 3-D prestack modeling operator. The analytical ex
pression for the AMO impulse response is derived by chaining constant-
velocity DMO with its inverse. Equivalently, it can be derived by chai
ning constant-velocity prestack migration and modeling. Because 3-D pr
estack data are typically irregularly sampled in the surface coordinat
es, AMO is naturally applied as an integral operator in the time-space
domain. The AMO impulse response is a skewed saddle surface in the ti
me-midpoint space. Its shape depends on the amount of azimuth rotation
and offset continuation to be applied to the: data. The shape of the
AMO saddle is velocity independent, whereas its spatial aperture is de
pendent on the minimum velocity. When the azimuth rotation is small (l
ess than or equal to 20 degrees), the AMO impulse response is compact,
and its application as an integral operator is inexpensive. Implement
ing AMO as an integral operator is not straightforward because the AMO
saddle may have a strong curvature when it is expressed in the midpoi
nt coordinates. An appropriate transformation of the midpoint axes to
regularize the AMO saddle leads to an effective implementation.