Earthquake Hazard Mitigation with Passive Seismic Arrays


      Download Documents:       Proposal       Presentation-2019/01/31

Motivation

The three main factors controling the level of shaking due to an earthquake are:

  1. the earthquake itself,
  2. the path effects, which in Los Angeles primarily means the basin structure, and
  3. the site response.

The last factor refers to the local subsurface and soil conditions, and is the least well determined of the three. This is due to the extreme lack of measurements - primarily the shallow shear velocity. The following example shows that to determine the site response, we will need several velocity measurments per kilometer to capture the variability.


Click to play movie.
Example of a small earthquake in Carson recorded on a dense array. The magnitude 2.5 earthqake was recored on a dense oil-company network located in Long Beach. The center panel shows the ground motion (click to see the movie> with clear disrpution of the wavefield by the Newport-Inglewood Fault (crossing the center of the panel on an angle). The right panel show the peak acceleration from this earthquake. Note the lateral variations of a factor of 5 over a scale length 1 km.

The LA Basin Passive Survey (LABPS)

LABPS is a survey in the Los Angeles Basin designed to produce a better model of the subsurface for the purposes of earthquake hazard mitigation.

Objectives of the Survey
  • Near-surface S-wave velocity for the site response component of seismic hazard
  • Structural model for the basement and intermediary layers for path effects
  • S-wave and P-wave velocity model for the basin for path effects
  • Location of active faults
  • Models for the whole crust down to the Moho
The size and scope of the survey is shown in the following figure (click to enlarge)

Proposed Survey The survey consists of approximately 18,000 land node positions in the Los Angeles Basin area. Each position will be occupied for 30 days and recorded 24-7 to produce a continuous data stream that can be correlated. We are also considering approximately 1000 off-shore nodes in a 1 km strip along the coast (shown in red). There is a fixed backbone array of 200 node stations that will record for the duration of the survey (blue dots).

The deployment method is shown in the following movie:

Deployment Plan The density part of the array moves across the region in novel roll-a-roll fashion using a pool of 5000 land nodes. The survey will start at one end of the basin with an intial deployment of 5000 stations, and when each station has recored 30 days of ambient noise, it is retrieved, downloaded and recharged and redeployed on the front edge of group. This way the array rolls across the basin and in approximately 4 months it will complete the survey by coverging the entire basin. On the right side the coverage in the form of a midpoint fold count is shown. This plot show the number of virtual seismograms for each midpoint and indicates the entire basin is well covered.


The Analysis

The first step in the analysis is to create the correlated seismograms. Each pair of receiver points that have simultaneous occupancy for at least 20 days will be correlated to produce approximately 100 million virtual seismograms. The will then be analyzed by techniques listed below. In addition, we will use earthquakes to do receiver function to determine structure from converted waves.

Analysis Techniques to Be Applied to the Data
  • Surface Waves - shallow shear velocity.
  • P-wave Refraction - basin P-wave velocity
  • Machine-learning microseismity locations
  • Autocorrelations - (zero-offset reflections) - structure.
  • Multi-offset Reflection? - structure + velocity.

Products

The products that will be creaed by the LABPS survey are: