Technical Updates: 2012, First Quarter

Data Sets (Task A)

  • The CENA database is going through an independent quality assurance (QA) review by the project team. The effort is led by Dr Christine Goulet (Assistant Researcher, PEER) in collaboration with the Database Working Group members who were not part of the final database development. Dr Albert Kottke (Post-Doctoral Researcher, PEER) has started a systematic review of the processed data using NGA-West2 software and Dr Dave Boore (USGS) independently processed selected earthquake records for comparison with Prof. Chris Cramer’s results (CERI, University of Memphis). Other QA activities include a cross-check for metadata, a review of station codes for duplicate Canada-US stations, and a comparison of processed spectral acceleration to existing GMPEs. Feedback is provided to Prof. Cramer and his team as solutions are developed for identified issues (time series and flatfile).

Reference rock and site amplification models (Task B)

All updates in this section reflect the work of the Geotechnical Working Group. The group is very active with many tasks underway and meetings every month.

  • Reference Rock: The group has previously presented a memo with the recommended reference rock definition and velocities. The group is refining the statistics further by including additional data sources, as well as estimating the epistemic uncertainty in the methodology. It does not appear the recommendation will change. This work should conclude at the beginning of May 2012.
  • Corrections at Recording Stations: The group provided recommended correction functions at the most recent SSHAC workshop. These correction functions have been shared with Simulation Working Group. The group continues to update the underlying data and procedure for these corrections with a particular focus on constraining Vs30 proxy calculations. The group introduced an additional method for Vs estimate based on proposed approach by Drs Paul Somerville and Sidao Ni (URS Corp.) that uses ratios of the amplitude of the radial and vertical components of the P-wave. A recent collaboration with EPRI should lead to additional measurements at recording sites with special focus on rock sites. This activity is expected to be concluded in midsummer. An update of the recommended site correction factors is also expected by mid-late summer 2012.
  • Profile Database: Continued efforts are ongoing to expand the shear-wave velocity database that will be used to develop idealized soil profiles for the simulations of site response.
  • Site Response Analysis Approaches: The group is in the process of developing protocols of the large scale site response simulations that will be conducted starting this summer. After numerous iterations, looking at existing simulations, the group has identified two data sets for use as input ground motions time series. One data set was developed by the Nuclear Regulatory Commission and the other by Prof. Youssef Hashash (University of Illinois). The group is also defining the threshold for conducting equivalent linear and nonlinear analyses which is based on studies conducted by Prof. Assimaki (Georgia Institute of Technology).

Regionalization (Task C)

  • Dr Walter Mooney (USGS) and his team started a new set of tasks in December 2011 to support the NGA-East Source/Path WG. A major component in the work is to compile an updated catalog of North American crustal structure (both shallow and deep, both P-wave and S-wave velocity and Q) for the CEUS region. The catalog, which contains models from the published literature, is currently being updated to include the most recent data available for both shallow and deep seismic structure.
  • Dr Mooney and his team are trying to anticipate the needs of end-users of their data compilation. For this, they have created an Arcmap module for the NRC CEUS hazards project. The Arcmap module allows users to have interactive access to relevant geophysics data. Arcmap allows multiple data layers to be analyzed separately or in sets, and permits the convenient and rapid analysis of our geophysical data. When completed, this effort will allow the user to select data from the index in map-view format and/or data-table format. The map and data tables can be used at either a large or small scale (i.e., regional, state, county). Geo-referenced topographic base-maps allows the user to access attributes such as elevation data and/or rock type and lithology. Major cities, population sizes, interstate highways, state boundaries, and US nuclear power plants may also be plotted.

Regionalization & Source/Path Studies (Tasks C and D)

  • Dr Boore published a new paper on stress parameters for nine well-recorded ENA earthquakes. Ref.: Boore, D. M. (2012). Updated determination of stress parameters for several well-recorded earthquakes in Eastern North America, Seismol. Res. Letters 83, 190-199.
  • Dr Boore and Prof. Gail Atkinson (University of Western Ontario, Canada) are working on empirical Green’s functions (eGFs) analysis to estimate the stress parameter from ratios of Fourier acceleration spectra. Their analyses are based on ratios for the mainshock and aftershocks of several ENA earthquakes, including the Mineral, VA, Saguenay, Riviere-du-Loup and Val des Bois earthquakes, from which mainshock stress parameters are estimated. They compare T= 0.1 s and 0.2 s PSA observations with simulations using these derived stress parameters, using a number of proposed geometrical spreading and Q models. Their preliminary finding indicate that a geometrical spreading steeper than 1/R1.3 in the first 60 or 70 km is required to provide stress parameters that are consistent for both methods (eGF analysis and direct inversion of short period PSA). This fast decay of motions is inconsistent with the PSA observations for longer periods (1 or 2 s), thus suggesting the need for a frequency-dependent geometrical spreading. Dave Boore has revised his stochastic-method programs (SMSIM) to handle a new frequency dependent geometric spreading model.
  • Dr Jack Boatwright (USGS) is making progress on the preliminary work he presented at the NGA-East SSHAC workshop in October and at AGU in December. He has finalized the regional spectral analysis of the August 23, 2011, M5.8 earthquake in Mineral, Virginia, and the regional spectral analysis of 14 additional 3.1 ≤ M ≤ 5.0 earthquakes in Canada and New England. The analysis was updated to include the Lake Anna Nuclear Power Station records, which were obtained in January. The addition of this record had a minor effect on the conclusions. The attenuation in the Mid-Atlantic region obtained from the 23 regional accelerographs and broadband recordings of the Mineral earthquake is Q = 466 f 0.48, is very similar to the average attenuation obtained previously for the northeastern part of the Appalachian Province by Boatwright and Seekins (2011), that is, Q = 410 f 0.5.

Finite Fault Simulations (Task E)

  • The finite fault simulations validation exercise is being rebooted! All the validations will be redone for a set of common velocity structures that are selected to minimize nonlinear site response effects. New rules for defining the near-surface Vs are currently being developed by members of the TI team. The 2011 Mineral, VA earthquake has also been added to the existing 7 scenarios.
  • Dr Somerville submitted his report on Magnitude-Area relationships. The TI team comments are being compiled after their initial review.

Inputs to stochastic ground motion simulation models (Task F)

  • Dr Boore and Dr Eric Thompson (Research Assistant Professor, Tufts University) finalized a paper in which they provide an improved equation for computing duration for root-mean-square oscillator response in the random vibration simulations for the stochastic model. This model will be used in NGA-East to simulate some of the ground motions for GMPE development. Ref.: Boore, D. M. and E. M. Thompson (2012). Empirical improvements for estimating earthquake response spectra with random-vibration theory, Bull. Seismol. Soc. Am., In press.

Sigma (Standard Deviation) (Task J)

Sigma and single-station sigma using CENA dataset:

  • Data exploration of the latest release of the CENA dataset dated September 30, 2011 was performed. Recorded response spectra for each earthquake were compared to average predictions for the same earthquake using ground motion prediction equations (GMPEs) such as Toro 2002 and Atkinson and Boore 2006.
  • A QA of the CENA dataset was performed by Dr Linda Al Atik (Researcher, Linda Alatik Consulting). Dataset issues such as the presence of 458 co-located stations (pairs of recordings with different IDs but corresponding to the same earthquake and station), different station IDs for stations having the same longitude and latitude, and Vs30 problems (same stations with different Vs30 values reported) were observed and reported.
  • Preliminary regression analysis using the CENA dataset (after removing the potentially problematic recordings) was performed with respect to the Atkinson and Boore 2006 functional form. Issues still need to be resolved regarding the regression results.

Sigma and single-station sigma using NGA West-2 dataset:

  • Data exploration and residual analyses were performed using the large magnitude NGA West-2 dataset with respect to NGA 2008 GMPEs.
  • Small-to-moderate magnitude NGA West-2 dataset was distributed recently and it is going through a QA process.