Non-ergodic ground motion models for central and… (G25AP00433) – Project Grant to University Of Memphis

Activities: Due to the unavailability of regional datasets, some gmms are developed based on the ergodic assumption (anderson and brune, 1999). The ergodic assumption suggests that the variability of ground motion in time and space is equal, enabling researchers to combine data from similar tectonic regimes around the world. Partially non-ergodic gmms (e.g., stewart ET AL., 2017) are models that include mean regional differences, such as differences because of source and or site location. The categorization of systematic site amplification effects (site-specific) and their elimination from the aleatory terms could be a primary objective in partially non-ergodic ground motion modeling (lavrentiadis ET AL., 2022). In recent years, with the collection of more data, IT has become evident that there is a significant systematic difference in ground motion based on the location of the site and the seismic source (lavrentiadis ET AL., 2022). Gmms developed based on ergodic assumptions might not function effectively for specific site and source locations. To address this issue, various researchers have proposed non-ergodic gmms (ngmms), where site and source location effects are explicitly modeled. The advantage of developing ngmms is that aleatory variability is reduced, and uncertainties in site and source location effects are considered as epistemic uncertainty. In this study, we aim to develop ngmms based on backbone models for the central and eastern united states (ceus). Developing ngmms based on backbone GMM (ergodic GMM) is a common way that has been mentioned in previous studies (e.g., lavrentiadis ET AL. 2022) and has several advantages, especially for the region where the dataset is sparse (e.g., ceus). In this framework, the median scaling is borrowed from the backbone GMM (ergodic GMM), and the non-ergodic part is estimated from the residuals. We divided ceus into two separate subregions: outside the coastal plain and inside the coastal plain. We use the GMM developed by pezeshk ET AL. (2018) as a backbone model for outside the coastal plain and the GMM developed by akhani ET AL. (2024) as a backbone modelfor inside the coastal plain. We aim to use gaussian processes (GPS) regression to develop varying coefficient models (VCM) for ceus. The PI of this proposal is a member of the coastal plain amplification working group (CPA-WG). This WG is looking at issues related to site amplification factors within the coastal plains of the united states of america. This research proposal fits well with the goals of CPA-WG. Our proposed ngmms development framework includes four major phases: a comprehensive review of the published NGA-east gmms, bayesian regression, and gaussian processes (phase i), strong ground motion databasecollection and processing (phase II), developing ngmms based on backbone models using gaussian processes(GPS) regression to develop varying coefficient models (VCM) (phase III), and comparing the developed model with existing models in terms of standard deviations (phase IV).deliverables and expected outcomes: the final product of this study will be two non-ergodic gmms: one for areas outside the coastal plain and one for areas inside the coastal plain. This project can be used in seismic hazard analysis and will improve the seismic risk hazard assessment for central and eastern united states.
Intended beneficiaries: The intended beneficiary of this award will be USGS for the development of seismic hazard maps and hazard analysts that work with ground motion models.

Non-ergodic ground motion models for central and eastern united statesproject period: 9 1 2025 through 8 31 2026 award purpose: this research aims to propose a non-ergodic ground motion model (varying…

Award facts

Description

Quick answers

Who received award G25AP00433?

How much is award G25AP00433 worth?

See this award's full record