Frequency-dependent Polarization Analysis Research Articles

Glacial hydraulic tremor on Rhonegletscher, Switzerland

AbstractGlacial hydraulic tremor (GHT) can be monitored to observe changes in location and distribution of water flow beneath glacial ice, allowing the spatiotemporal evolution of subglacial hydrology to be studied continuously and remotely. We use frequency-dependent polarization analysis (FDPA) to classify types of GHT and assess its spatio-temporal extent beneath Rhonegletscher, Switzerland, in a continuous seismic record through 2018 and 2019 at three ice-proximal bedrock-based seismometers. We determine the frequency bands composing the GHT and calculate back azimuth angles pointing to a previously known subglacial channel. We also inspect the relationship between GHT seismic power and water discharge from the glacier to compare daily and seasonal shifts with the observed GHT. We observed the seasonal shift from a distributed system to a channelized system, and our dataset allowed comparison of channel locations within and across seasons, with implications for sediment evacuation and bed erosion. The successful use of this method to assess GHT previously on Taku glacier (the methods of which this project follows) and now Rhonegletscher shows that existing ice-proximal passive seismic installations can be used to easily and continuously monitor subglacial hydraulic activity.

Seismic signature of turbulence during the 2017 Oroville Dam spillway erosion crisis

Abstract. Knowing the location of large-scale turbulent eddies during catastrophic flooding events improves predictions of erosive scour. The erosion damage to the Oroville Dam flood control spillway in early 2017 is an example of the erosive power of turbulent flow. During this event, a defect in the simple concrete channel quickly eroded into a 47 m deep chasm. Erosion by turbulent flow is difficult to evaluate in real time, but near-channel seismic monitoring provides a tool to evaluate flow dynamics from a safe distance. Previous studies have had limited ability to identify source location or the type of surface wave (i.e., Love or Rayleigh wave) excited by different river processes. Here we use a single three-component seismometer method (frequency-dependent polarization analysis) to characterize the dominant seismic source location and seismic surface waves produced by the Oroville Dam flood control spillway, using the abrupt change in spillway geometry as a natural experiment. We find that the scaling exponent between seismic power and release discharge is greater following damage to the spillway, suggesting additional sources of turbulent energy dissipation excite more seismic energy. The mean azimuth in the 5–10 Hz frequency band was used to resolve the location of spillway damage. Observed polarization attributes deviate from those expected for a Rayleigh wave, though numerical modeling indicates these deviations may be explained by propagation up the uneven hillside topography. Our results suggest frequency-dependent polarization analysis is a promising approach for locating areas of increased flow turbulence. This method could be applied to other erosion problems near engineered structures as well as to understanding energy dissipation, erosion, and channel morphology development in natural rivers, particularly at high discharges.

The Characteristics of Microseisms in South China Sea: Results From a Combined Data Set of OBSs, Broadband Land Seismic Stations, and a Global Wave Height Model

AbstractIt has long been recognized that ocean gravity waves can generate microseisms through their interactions and coupling with the solid earth. Their generation mechanisms, wave types, and propagation have been studied and debated intensively. In this study, we focus on the characteristics of microseisms in South China Sea (SCS) and compare them with globally observed microseisms. We use data from six ocean bottom seismometers, 10 publicly available broadband land seismometers near SCS, and a reanalyzed global wave height model from 1 May 2012 to 20 August 2012 (UTC), the time span of ocean bottom seismometers' data. We apply three techniques including power spectral density, correlation, and frequency dependent polarization analysis. Our results show that (1) microseisms observed in SCS are dominated by sources from adjacent oceans, instead of the common global sources from the Pacific, Atlantic, or Indian Ocean; (2) the split of double‐frequency microseisms (DFMs) is observed in SCS, especially on land stations: the sources are located near the central sea basin for long‐period DFMs (0.1–0.25 Hz) and are local near stations for short‐period DFMs (0.25–0.5 Hz); (3) typhoons both strengthen microseisms and affect the source locations of microseisms with the biggest influence on the short‐period DFMs; and (4) microseisms in or near SCS are a mixture of Love and Rayleigh waves and the relative dominance of Love and Rayleigh waves changes with locations.

Microseisms from Superstorm Sandy

We analyzed and visualized the microseisms generated by Superstorm Sandy as recorded by the Earthscope Transportable Array (TA) during late October through early November of 2012. We applied continuous, frequency-dependent polarization analysis to the data and were able to track the course of Sandy as it approached the Florida coastline and, later, the northeastern coast of the U.S. The energy level of Sandy was roughly comparable to the background microseism level generated by wave–wave interactions in the North Atlantic and North Pacific oceans. The maximum microseismic power and degree of polarization were observed across the TA when Sandy sharply changed its direction to the west–northwest (specifically, towards Long Island, New York) on October 29. The westward turn also briefly changed the dominant microseism period from 5 s to 8 s. We identified three other microseismic source regions during the 18 day observation period. In particular, peak-splitting in the double frequency band and the orientation of the 5 s and 8 s polarization vectors revealed two contemporaneous microseism sources, one in the North Atlantic and one in the Northeast Pacific, for the dates of November 3–4. Predictions of microseismic excitation based on ocean wave models showed consistency with the observed microseismic energy generated by Sandy and other storms.

Time- and frequency-dependent polarization analysis: anomalous surface wave observations in Iberia

A time- and frequency-dependent polarization method is applied to synthetic and observational data in order to investigate effects of lateral heterogeneity on surface wave polarization. It is shown that interference of surface wave signals can produce artifacts in the polarization as determined for the dominant signal present in the data. The time- and frequency-dependent polarization technique is applied to surface waves recorded in Iberia to investigate the character of the recorded wavetrains. Rayleigh waves with a southwest backazimuth recorded in the south of Spain appear strongly attenuated relative to other locations in Iberia. By analysing their polarizations, it can be inferred that these surface waves are highly distorted, implying that scattering mechanisms must be operative in the south of Spain. Strong scattering is one of the mechanisms that might explain the observed amplitude reduction, although other mechanisms like intrinsic scattering or defocusing may also contribute. With the present data and the present technique it is not possible to determine the relative contributions of each of these mechanisms.