Autocorrelation Reflectivity of the Gulf Coast from USArray Data

Abstract:

We are calculating autocorrelation reflectivity (Claerbout,1968) for ~250 Transportable and Reference Array stations from the Gulf of Mexico north to ~37^oN, and from -101^o to -87^oW. This subarray covers the Gulf Coast Basin, Gulf Coastal Plain, extends across the Ouachita Fold and Thrust Belt and into the Mississippi Valley graben. The same subarray was recently used for determining Ps receiver functions, tele-seismic P-wave cross-correlations, and ambient noise Rayleigh wave tomography (Miao et al, 2020, this session and submitted).

The autocorrelation functions were calculated using all of the available TA data, and 2.5 years of Reference Array data. The data were preprocessed using a number of statistical metrics to remove data windows with undesirable noise characteristics. The Autocorrelations were made from 600s long time windows using six different amplitude normalization and balancing methods (Bensen et al, 2007; Oren and Nowack, 2017). The six methods produce almost identical autocorrelation functions. Post-processing included band-pass filtering and trace editing.

Preliminary examination of the data shows that we can identify the base of the sedimentary column in high-pass filtered (0.3-7.5Hz) autocorrelations. The Moho and coherent lower crustal reflections are apparent in broadband autocorrelations (0.02-7.5Hz). In low-pass filtered (0.02-1.0Hz) autocorrelations we can identify weak reflections in the mantle at the appropriate time for the lithosphere-asthenosphere boundary (Ainsworth et al, 2014). Reflections from the Moho and lower crust increase in complexity passing northward from the Gulf Coastal Plain across the boundary with the Ouachita Fold and Thrust belt.