Site Characterization at Downhole Arrays by Joint Inversion of Dispersion Data and Acceleration Time Series

Abstract

We present a sequential data assimilation algorithm based on the ensemble Kalman inversion to estimate the near-surface shear-wave velocity profile and damping; this is applicable when heterogeneous data and a priori information that can be represented in forms of (physical) equality and inequality constraints in the inverse problem are available. Although noninvasive methods, such as surface-wave testing, are efficient and cost-effective methods for inferring an VS profile, one should acknowledge that site characterization using inverse analyses can yield erroneous results associated with the lack of inverse problem uniqueness. One viable solution to alleviate the unsuitability of the inverse problem is to enrich the prior knowledge and/or the data space with complementary observations. In the case of noninvasive methods, the pertinent data are the dispersion curve of surface waves, typically resolved by means of active source methods at high frequencies and passive methods at low frequencies. To improve the inverse problem suitability, horizontal-to-vertical spectral ratio data are commonly used jointly with the dispersion data in the inversion. In this article, we show that the joint inversion of dispersion and strong-motion downhole array data can also reduce the margins of uncertainty in the VS profile estimation. This is because acceleration time series recorded at downhole arrays include both body and surface waves and therefore can enrich the observational data space in the inverse problem setting. We also show how the proposed algorithm can be modified to systematically incorporate physical constraints that further enhance its suitability. We use both synthetic and real data to examine the performance of the proposed framework in estimation of the VS profile and damping at the Garner Valley downhole array and compare them against the VS estimations in previous studies.

Geotechnical Site Characterization via Deep Neural Networks: Recovering the Shear Wave Velocity Profile of Layered Soils

“…The training dataset is generated using 1D shear wave velocity profiles of California [7] and dispersion data is calculated using GEOPSY [8]. The model was evaluated on a test dataset in addition to the recently published work of Seylabi et al (2020) [1] where they used the ensemble Kalman inversion method. In both cases, the performance was shown to be satisfactory, while the model doesn't need any further fine-tuning or initialization as required by EKI.…”

“…Seylabi et al (2020) showed an application of ensemble Kalman inversion (EKI) to invert for soil shear wave velocity profile given dispersion data as input [1]. Figure 5 shows a comparison of predictions using the model trained in this study versus the results of EKI reported by [1].…”

“…The trained model gets the data on the left and returns the data on the right. ensemble Kalman inversion [1], stochastic direct search algorithm [2], uniform Monte Carlo method [3], and fully Bayesian Markov chain Monte Carlo method [4] to name but a few, to reconcile the issues. Such techniques mostly rely on ground surface acceleration time-series and dispersion data to infer the mechanical properties of soil layers.…”

“…Such techniques mostly rely on ground surface acceleration time-series and dispersion data to infer the mechanical properties of soil layers. A comprehensive review of inversion techniques for this application can be found in [1].…”

“…In general, dispersion data is more accessible than time-series, the reason that the majority of studies rely on them for inversion purposes [5,1]. However, due to the non-uniqueness of solutions, researchers have tried to better constrain the uncertain parameters by including time-series data in the inversion process [1]. In this article, we use only dispersion data to invert for the mechanical properties of a soil medium using a Deep Neural Network (DNN) model.…”