Numerical inversion of seismic surface wave dispersion data and crust‐mantle structure in the New York‐Pennsylvania area

A least-squares curve-fitting program in which calculations of surface wave dispersion are used has been developed in order to compute an interpretation of empirical dispersion data in terms of a layered model of the earth. Input may consist of dispersion data (phase velocity versus wave period) for both Love and Rayleigh waves in any modes of propagation. Output consists of successive approximations of the values of parameters such as layer thicknesses, shear velocities, and densities which minimize the mean square of residuals of the empirical data with respect to theoretical dispersion curves calculated from the parameters. Tests made by applying the method to precisely computed theoretical dispersion curves demonstrate the validity of the method. In these, as many as six parameters of the original theoretical models are calculated under a wide variety of conditions from the dispersion data only. In other cases, the amount of information obtainable may be greater or smaller, depending on the quality of the input data. The New York-Pennsylvania dispersion data of Oliver et al. [1961] can be successfully interpreted by using additional information from seismic refraction and studies of nearby earthquakes in the area. A solution including the important effect of the mantle low-velocity channel gives a crustal thickness of 38.6 km, a crustal shear velocity of 3.64 km/sec, a shear velocity of 4.685 km/sec below the M discontinuity, and a ratio of 2.86/3.30 between crustal and subcrustal density values. The density measurement is a new result. The shear velocity structure derived here is consistent with results obtained by Katz from seismic refraction profiles in the same area. Additional data are needed in order to derive more detailed information on crustal structure.