Noise Correlations of Wavefield Gradients to Improve Sensitivity to (near surface, time-dependent) Structural Heterogeneity

We assess the potential of rotational ground motions to resolve time-dependent near surface structural heterogeneities using noise correlations. Recent studies reveal an increased sensitivity of gradient related observations to near surface structural heterogeneities (e.g., material contrast, cavities) compared to directly measured wavefields (and their time derivatives). The development of new sensing technologies, such as rotational ground motion sensors and distributing acoustic sensing (DAS), enable measurements of strain and rotations and motivate this study. Combining gradient related observations with ambient noise-based monitoring methods has the potential to increase both spatial and temporal resolution. In order to investigate the suggested benefits, we perform a numerical study in 2D, where we simulate seismic noise with random sources at random locations. We apply interferometric principles and calculate cross-correlations of the resulting noise traces recorded at different receiver locations for multiple realizations of the noise field. After analysing the convergence of the correlation functions in terms of simulation length and number of simulations, we compare noise correlations of acceleration and rotation rate for a homogenous reference and a perturbed model. Ultimately, we establish that noise correlations of wavefield gradients are more sensitive than noise correlations of wavefields to small-scale heterogeneity.