Site Map


New PEER Report 2017/05: “Recommendations for Ergodic Nonlinear Site Amplification in Central and Eastern North America”

PEER has just published Report No. 2017/05: “Recommendations for Ergodic Nonlinear Site Amplification in Central and Eastern North America.” It was authored by Youssef M.A. Hashash, Joseph A. Harmon, Okan Ilhan, Grace A. Parker, and Jonathan P. Stewart.

Visit the PEER publications page to download a free color pdf of the document.


This document is a companion report to Expert Panel Recommendation for Ergodic Linear Site Amplification Models in central and eastern North America (PEER Report 2017/04, Stewart et al. 2017). This report describes the panel recommendations for ergodic median nonlinear site amplification models, which are meant to accompany linear models in the companion report. Nonlinear models for site amplification must represent the strength of the input ground motion in some manner, and peak acceleration for a reference condition (PGAr) is often used. The use of PGAr (and similar parameters) requires specification of a reference condition in the development of nonlinear models, and those provided here consider reference conditions of VS = 3000 m/sec and VS30 = 760 m/sec. One of the proposed models (the GWG-S nonlinear amplification model) is derived for a reference condition of VS = 3000 m/sec. A second is identical to the first except that PGAr is adjusted to a VS30 = 760 m/sec reference condition.

Nonlinear amplification models in this report are produced as functions of VS30 and (PGAr). Other models evaluated in this report are the PEA nonlinear amplification model and the GWG-S model with an alternative approach to convert GWG-S nonlinear amplification model estimations to a VS30 = 760 m/sec reference condition. A recommended epistemic uncertainty model on the GWG-S recommended median nonlinear amplification models is provided in piecewise functional form to generate reasonable variation of Fnl across the period and VS30 ranges of interest. Limitations on the recommended models are presented considering both the methodology of the recommended model derivation and limitations of nonlinear amplification
models in general.