An analytic model is developed for scattering from random inhomogeneities in range-dependent ocean waveguides using the Rayleigh-Born approximation to Green's theorem. The expected scattered intensity depends on statistical moments of fractional changes in compressibility and density, which scatter as monopoles and dipoles, respectively, and the coherence volume of the inhomogeneities. The model is calibrated for ocean bottom scattering using data acquired by instantaneous wide-area ocean acoustic waveguide remote sensing (OAWRS) and geophysical surveys of the ONR Geoclutter Program. The scattering strength of the seafloor on the New Jersey shelf, a typical continental shelf environment, is found to depend on wave number k, medium coherence volume V(c), and seabed depth penetration factor F(p) following a 10 log(10)(F(p)V(c)k(4)) dependence. A computationally efficient numerical approach is developed to rapidly compute bottom reverberation over wide areas using the parabolic equation by exploiting correlation between monopole and dipole scattering terms and introducing seafloor depth penetration factors. An approach is also developed for distinguishing moving clutter from statistically stationary background reverberation by tracking temporal and spatial fluctuations in OAWRS intensity images.