一种适用于噪声源非均匀分布的背景噪声面波相速度频散曲线测量方法

Based on the Frequency-Bessel (FJ) method, this paper proposes the Generalized Frequency-Bessel (GFJ) method to address non-uniform noise source distribution. Unlike the FJ method, which only considers the distribution of the noise cross-correlation function in the inter-station distance dimension, the GFJ method accounts for its two-dimensional distribution in both inter-station distance and azimuth. By utilizing the real and imaginary parts of the frequency-domain noise cross-correlation function, the GFJ method uses an azimuthal Fourier transform to separates the contributions of different Fourier series components from non-uniform noise sources in the ambient noise cross-correlation function. Subsequently, it calculates the dispersion spectrum using the corresponding order of Hankel transform. The GFJ method can obtain multiple sets of dispersion spectra from the same dataset for mutual validation and signal enhancement. The method is suitable for extracting higher-order surface waves in dense array data. Its accuracy and applicability are demonstrated through theoretical and numerical experiments, and its practical utility is validated using data from the USArray and ChinArray.