Multicomponent seismic recordings are currently being analyzed in an attempt to improve conventional P‐wave sections and to find and use rock properties associated with shear waves (e.g. Dohr, 1985; Danbom and Dominico, 1986). Mode‐converted (P-SV) waves hold a special interest for several reasons: They are generated by conventional P‐wave sources and have only a one‐way travel path as a shear wave through the typically low velocity and attenuative near surface. For a given frequency, they will have a shorter wavelength than the original P wave, and thus offer higher spatial resolution; this has been observed in several vertical seismic profiling (VSP) cases (e.g., Geis et al., 1990). However, for surface seismic data, converted waves are often found to be of lower frequency than P-P waves (e.g., Eaton et al., 1991).