Acoustic scattering from objects buried in water-sand sediment is the foundation of target detection and identification. It is also a research hotspot in areas of acoustic scattering while the domestic research on scattered field from buried targets is not deep. This paper deduces an approximate Green's function of acoustic scattering from targets buried in water-sand sediment, which describes clearly the whole physical process during the propagation of scattered waves. Next, on basis of geometric acoustics, the corresponding Helmholtz-Kirchhoff formula of integration is presented. Complicated integration of the full wave number spectral representation of the Green's function is avoided by employing approximate formula derived from the method of ray acoustics. As a result of neglecting the influence from lateral waves, the Helmholtz-Kirchhoff integral given applies to supercritical incidence case. The function of COMSOL Multiphysics software is expanded by writing this formula of integration into it. By means of finite-element method, numerical calculation models for two-dimensional axisymmetric targets are established on the software platform. The proposed model built in free field is verified through comparing numerical results obtained with the Rayleigh method which has been validated in previous research achievements of acoustics. The target strength of buried elastic solid sphere is calculated under different conditions in order to analyze the change regularity of buried scattered field. We provide a summary about the law of target strength of the elastic sphere varying with frequency, buried depth, and the attenuation of sand. Finally, we conduct acoustic scattering experiments in free space and shallow buried conditions and process the data with the method of isolation and identification of resonance to separate eliastic echoes from reverberation echo and specular echo. Results from the experiment of free field show that components of the scattered wave should include Rayleigh waves and whispering gallery waves. The processed data of objects buried inside layered fluid media indicate that characteristics of resonance spectra can be used to identify and detect the target effectively while echo signal is not available for identification of target. The proposed technique is verified through the comparison of data from total scattered field between experiment and theoretical prediction. This study has important guiding significance for detecting and identifying targets embedded within layered acoustic media in practical applications.