Imaging near-surface heterogeneities by natural migration of backscattered surface waves

We present a migration method that does not require a velocity model to migrate backscattered surface waves to their projected locations on the surface. This migration method, denoted as natural migration, uses recorded Green's functions along the surface instead of simulated Green's functions. The key assumptions are that the scattering bodies are within the depth interrogated by the surface waves, and the Green's functions are recorded with dense receiver sampling along the free surface. This natural migration takes into account all orders of multiples, mode conversions and non-linear effects of surface waves in the data. The natural imaging formulae are derived for both active source and ambient-noise data, and computer simulations show that natural migration can effectively image near-surface heterogeneities with typical ambient-noise sources and geophone distributions.