Search

SEIS, the seismometer of the InSight mission, which landed on Mars on 26 November 2018, is monitoring the seismic activity of the planet. The goal of the Mars Structure Service (MSS) is to provide, as a mission product, the first average 1-D velocity model of Mars from the recorded InSight data. Prior to the mission, methodologies have been developed and tested to allow the location of the seismic events and estimation of the radial structure, using surface waves and body waves arrival times, and receiver functions. The paper describes these validation tests and compares the performance of the different algorithms to constrain the velocity model below the InSight station and estimate the 1-D average model over the great circle path between source and receiver. These tests were performed in the frame of a blind test, during which synthetic data were inverted. In order to propagate the data uncertainties on the output model distribution, Bayesian inversion techniques are mainly used. The limitations and strengths of the methods are assessed. The results show the potential of the MSS approach to retrieve the structure of the crust and underlying mantle. However, at this time, large quakes with clear surface waves have not yet been recorded by SEIS, which makes the estimation of the 1-D average seismic velocity model challenging. Additional locatable events, especially at large epicentral distances, and development of new techniques to fully investigate the data, will ultimately provide more constraints on the crust and mantle of Mars.

Over a period of less than a decade, distributed acoustic sensing (DAS) has become a well-established technology in seismology. For historical and practical reasons, DAS manufacturers usually provide instruments that natively record strain (rate) as the principal measurement. While at first glance strain recordings seem related to ground motion waveforms (displacement, velocity and acceleration), not all the seismological tools developed over the past century (e.g. magnitude estimation, seismic beamforming, etc.) can be readily applied to strain data. Notably, the directional sensitivity of DAS is more limited than conventional particle motion sensors, and DAS experiences an increased sensitivity to slow waves, often highly scattered by the subsurface structure and challenging to analyse. To address these issues, several strategies have been already proposed to convert strain rate measurements to particle motion. In this study, we focus on strategies based on a quantity we refer to as 'deformation'. Deformation is defined as the change in length of the cable and is closely related to displacement, yet both quantities differ from one another: deformation is a relative displacement measurement along a curvilinear path. We show that if the geometry of the DAS deployment is made of sufficiently long rectilinear sections, deformation can be used to recover the displacement without the need of additional instruments. We validate this theoretical result using full-waveform simulations and by comparing, on a real data set, the seismic velocity recovered from DAS with that recorded by collocated seismometers. The limitations of this approach are discussed, and two applications are shown: enhancing direct P -wave arrivals and simplifying the magnitude estimation of seismic events. Converted displacement provides better sensitivity to high velocity phases, improves broadside response and permits the direct application of conventional seismological tools that are less effective when applied to strain (rate) data. [ABSTRACT FROM AUTHOR]

The seismic noise recorded by the Interior Exploration using Seismic Investigations, Geodesy, and Heat Transport (InSight) seismometer (Seismic Experiment for Interior Structure [SEIS]) has a strong daily quasi-periodicity and numerous transient microevents, associated mostly with an active Martian environment with wind bursts, pressure drops, in addition to thermally induced lander and instrument cracks. That noise is far from the Earth's microseismic noise. Quantifying the importance of nonstochasticity and identifying these microevents is mandatory for improving continuous data quality and noise analysis techniques, including autocorrelation. Cataloging these events has so far been made with specific algorithms and operator's visual inspection. We investigate here the continuous data with an unsupervised deep-learning approach built on a deep scattering network. This leads to the successful detection and clustering of these microevents as well as better determination of daily cycles associated with changes in the intensity and color of the background noise. We first provide a description of our approach, and then present the learned clusters followed by a study of their origin and associated physical phenomena. We show that the clustering is robust over several Martian days, showing distinct types of glitches that repeat at a rate of several tens per sol with stable time differences. We show that the clustering and detection efficiency for pressure drops and glitches is comparable to or better than manual or targeted detection techniques proposed to date, noticeably with an unsupervised approach. Finally, we discuss the origin of other clusters found, especially glitch sequences with stable time offsets that might generate artifacts in autocorrelation analyses. We conclude with presenting the potential of unsupervised learning for long-term space mission operations, in particular, for geophysical and environmental observatories. [ABSTRACT FROM AUTHOR]

We compile and analyze a data set of secondary microseismic P wave spectra that were observed by North American seismic arrays. Two distinct frequency bands, 0.13-0.15 Hz and 0.19-0.21 Hz, with enhanced P wave energy characterize the data set. Cluster analysis allows to classify the spectra and to associate typical spectral shapes with geographical regions: Low-frequency-dominated spectra (0.13-0.15 Hz) are mostly detected in shallower regions of the North Atlantic and the South Pacific, as well as along the Central and South American Pacific coast. High-frequency-dominated spectra (0.19-0.21 Hz) are mostly detected in deeper regions of the northwestern Pacific and the South Pacific. For a selected subset of high-quality sources, we compute synthetic spectra from an ocean wave hindcast. These synthetic spectra are able to reproduce amplitude and shape of the observed spectra, but only if P wave resonance in the water column at the source site is included in the model. Our data sets therefore indicate that the spectral peaks at 0.13-0.15 Hz and 0.19-0.21 Hz correspond to the first and second harmonics of P wave resonance in the water column that occur in shallower ocean depths (<3,000 m) and in the deep ocean (∼5,000 m), respectively. This article demonstrates the important effect of water column resonance on the amplitude and frequency of P waves that are generated by secondary microseisms and that the amplitude of high-quality sources can be predicted from ocean wave hindcasts within a factor of 0.4-6. [ABSTRACT FROM AUTHOR]

The most exciting initiative for the recent polar studies was the International Polar Year (IPY) in 2007–2008. The IPY has witnessed a growing community of seismologists who have made considerable efforts to acquire high-quality data in polar regions. It also provided an excellent opportunity to make significant advances in seismic instrumentation of the polar regions to achieve scientific targets involving global issues. Taking these aspects into account, we organize and publish a special issue in Polar Science on the recent advance in polar seismology and cryoseismology as fruitful achievements of the IPY. [ABSTRACT FROM AUTHOR]

An elastic wavefield propagating in an inhomogeneous elastic medium is only sensitive in an effective way to inhomogeneities much smaller than its minimum wavelength. The corresponding effective medium, or homogenized medium, can be computed thanks to the non-periodic homogenization technique. In the seismic imaging context, limiting ourselves to layered media, we numerically show that a tomographic elastic model which results of the inversion of limited frequency band seismic data is an homogenized model. Moreover, we show that this homogenized model is the same as the model that can be computed with the non-periodic homogenization technique. We first introduce the notion of residual homogenization, which is computing the effective properties of the difference between a reference model and a ‘real’ model. This is necessary because most imaging technique parametrizations use a reference model that often contains small scales, such as elastic discontinuities. We then use a full-waveform inversion method to numerically show that the result of the inversion is indeed the homogenized residual model. The full-waveform inversion method used here has been specifically developed for that purpose. It is based on the iterative Gauss–Newton least-square non-linear optimization technique, using full normal mode coupling to compute the partial Hessian and gradient. The parametrization has been designed according to the residual homogenized parameters allowing to build a real multiscale inversion with progressive frequency band enrichment along the Gauss–Newton iterations. [ABSTRACT FROM AUTHOR]