Joint Inversion of High-Frequency Receiver Functions and Surface-Wave Dispersion: Case Study in the Parnaíba Basin of Northeast Brazil

Abstract

ABSTRACTWe assess the performance of the joint inversion of receiver functions (RF) and surface-wave dispersion in the characterization of the sedimentary package comprising the Parnaíba basin. This procedure is routinely utilized in passive-source crustal studies to retrieve S-wave velocity variations with depth, and has seldom been used with higher-frequency datasets to investigate fine sedimentary structure. The Parnaíba basin is a Paleozoic cratonic basin composed of five supersequences, accumulating ∼3.5 km of sedimentary rocks interbedded by Late Cretaceous diabase sills. The dataset used for this research was acquired between 2015 and 2017 through deployment of 10 short-period and one broadband seismic stations distributed along an approximately 100-kilometer-long linear array in the center of the basin. The deployment was carried out under the Parnaíba Basin Analysis Project, a multi-institutional and multidisciplinary effort funded by BP Energy do Brasil. High-frequency RFs (f<4.8 Hz) were calculated from deconvolution of teleseismic P waveforms (30°<Δ<90°) after rotation into the great-circle path, whereas high-frequency dispersion curves (0.25–2 Hz) were obtained through multiple filter analysis of empirical Green’s functions developed from cross-correlation (ZZ component) and stacking (six months) of time–frequency-normalized ambient seismic noise recordings. S-wave velocity–depth profiles down to ∼5 km depth were developed through an iterative, linearized joint inversion approach. Comparison to independent active-source seismic profiles overlapping with our passive-source seismic line reveals the inverted velocity models successfully retrieve sedimentary thickness (top of the Cambrian), sedimentary velocity structure, and depth to the Cenozoic sedimentary sequence. In addition, high-velocity zones at depths ranging from 1.5 to 2.5 km are observed in the inverted velocity–depth profiles, which are interpreted as due to the Late Cretaceous sills interbedding the basin’s sedimentary rocks. The relative low cost of our approach makes it ideal for basic characterization of relatively unknown sedimentary basins.