MHVSR ‘Big Data’ to Outline Standard Procedures for Data Acquisition, Processing, Interpretation and Presentation
Session: Earthquake Science, Hazards and Policy in Cascadia [Poster]
Type: Poster
Date: 4/20/2021
Presentation Time: 11:30 AM Pacific
Description:
The microtremor horizontal-to-vertical spectral ratio (MHVSR) method has introduced as a cost-effective and non-invasive way to estimate shear-wave resonant frequencies at a site. For the MHVSR to become a standard tool in site classification, stricter standards for acquisition, processing and interpretation need to be established. Using a database of ~1700 measurements collected throughout the Metro Vancouver region from 2018-20, the effects of acquisition parameters, processing choices, and peak detection strategies were investigated. We initially analysed each measurement individually using the opensource software Geopsy. All peaks identified were manually inspected to assess clarity and reliability. The simple GUI allows easy, real-time observation of the effect of altering processing parameters on the computed MHVSR. Though time consuming, this process provides information about how to optimally design an automatic routine to batch process MHVSRs, that makes intelligent decisions regarding processing inputs, and picks peaks in a statistically robust manner. We utilized recent MHVSR algorithms (OpenHVSR, HVSRPy, in-house software) to batch process the database of measurements, and automatically identify peaks. Processing decisions and peak definition were guided by logic determined from initial analysis. These automatic processing algorithms reduce workload but must be used with care if measurement quality is low. A database of the measurements was created to allow access to not only the interpreted peak frequency values and amplitudes, but also the individual spectra for each individual time window and component, and the meta-data related to acquisition and processing choices. Finally, a map of the lowest resonant frequency interpreted from each measurement was produced. Future work involves understanding if the shape of the MHVSR can be related to earthquake site effects experienced at a site, as well as if it is possible to invert MHVSRs for soil profile models.
Presenting Author: Aamna Sirohey
Student Presenter: Yes
Authors
Aamna Sirohey Presenting Author Corresponding Author asirohey@uwo.ca University of Western Ontario |
Sheri Molnar smolnar8@uwo.ca University of Western Ontario |
|
|
|
|
|
|
|
MHVSR ‘Big Data’ to Outline Standard Procedures for Data Acquisition, Processing, Interpretation and Presentation
Category
Earthquake Science, Hazards and Policy in Cascadia