Bayesian joint inversion of controlled source electromagnetic and magnetotelluric data to infer presence of a freshwater aquifer offshore New Jersey. Issue 1 (1st December 2019)

• Record Type: Journal Article
• Authors: Blatter, Daniel; Key, Kerry; Ray, Anandaroop; Gustaffson, Chloe; Evans, Rob
• Is Part Of: ASEG Extended Abstracts (Online). Volume 2019:Issue 1(2019)
• Page Start: 1
• Page End: 5
• Publication Date: 2019-12-01
• Subjects: joint inversion -- marine electromagnetics -- Bayesian methods -- freshwater aquifers; Prospecting -- Geophysical methods -- Periodicals; Prospecting -- Geophysical methods; Periodicals
• DOI: 10.1080/22020586.2019.12072959
• Languages: English
• ISSNs: 2202-0586

Abstract

Summary: To evaluate the extent of a low salinity aquifer within the continental shelf offshore New Jersey, USA, we adapt a Bayesian trans-dimensional Markov chain Monte Carlo (McMC) algorithm for jointly inverting surface towed controlled source electromagnetic (CSEM) and seafloor magnetotelluric (MT) data. Owing to the relatively flat seafloor topography and layered underlying geology, our inversion is parametrized to produce an ensemble of 1D models at each site along the tow line. The results identify a high probability region of high resistivity of varying depth and thickness in the upper 500 m of the continental shelf, corroborating results from a previous study that used a regularized, gradient-based 2D inversion method. Using a combination of synthetic studies and real data we also evaluate the joint model parameter uncertainty in comparison to the uncertainty obtained from the individual data sets and demonstrate quantitatively that joint inversion offers reduced uncertainty. In addition, we show with a Monte Carlo scheme, how the Bayesian model ensemble can subsequently be used to derive quantitative uncertainty estimates of pore water salinity within the low salinity aquifer. The ability to produce meaningful non-linear uncertainty estimates on both conductivity as well as related parameters is a strong motivator for the use of Bayesian inversion methods when computationally feasible.