Next Generation Attenuation (NGA)

Next Generation Attenuation (NGA)
Name	Next Generation Attenuation
Abbreviation	NGA
Developed by	Pacific Earthquake Engineering Research Center; United States Geological Survey; European Commission participants
First release	2003–2008
Application	Seismic hazard analysis; ground motion prediction

Next Generation Attenuation (NGA) is a suite of empirical ground motion models and associated methodologies developed for seismic hazard analysis to improve prediction of earthquake shaking for engineering and risk applications. The project synthesized strong‑motion records, paleoseismic and instrumental observations, and advanced statistical techniques to produce ground motion prediction equations used by Federal Emergency Management Agency, California Geological Survey, and other institutions. NGA influenced revisions to regional hazard maps, building codes, and performance‑based engineering practices in jurisdictions such as California, Japan, and the European Union.

Overview

The NGA effort produced collections of ground motion models that relate earthquake source, path, and site descriptors to intensity measures used by American Society of Civil Engineers, International Code Council, and engineering practitioners. Key outputs include predictive equations for spectral acceleration, peak ground acceleration, and velocity used by Federal Highway Administration, Federal Aviation Administration, and urban planners in municipalities including Los Angeles, San Francisco, and Tokyo. NGA frameworks integrate data from networks such as the Instrumental Seismic Recordings Network, national seismic systems in the United States Geological Survey, and academic arrays managed by California Institute of Technology and Massachusetts Institute of Technology researchers.

Development and History

The NGA initiative began as collaborative research among academic groups, national agencies, and professional organizations following major events like the 1994 Northridge earthquake, the 1995 Kobe earthquake, and the 1999 İzmit earthquake. Funding and coordination involved entities such as the National Science Foundation, the Pacific Earthquake Engineering Research Center, and the United States Geological Survey. Key contributors included research teams from University of California, Berkeley, University of Southern California, Stanford University, and international partners from University of Tokyo and Swiss Seismological Service, who compiled datasets, developed regression frameworks, and proposed site characterization metrics that informed subsequent editions and regional adaptations.

Ground Motion Models and Methodology

NGA ground motion models (GMMs) employ mixed‑effects regression, Bayesian updating, and nonstationary stochastic approaches advanced by groups at Cornell University, Columbia University, Imperial College London, and ETH Zurich. Predictor variables include rupture magnitude, distance metrics tied to rupture geometry used by USGS scenarios, hanging‑wall effects recognized after the 1999 Chi‑Chi earthquake, and site amplification described by shear‑wave velocity (Vs30) as referenced by American Society of Civil Engineers guidance. Models from the NGA suite—developed by teams led by researchers at University of California, Los Angeles, Purdue University, and University of Washington—were statistically compared using information criteria applied in studies cited by National Institute of Standards and Technology.

Implementation and Applications

Practitioners in structural engineering and risk assessment implement NGA models within seismic hazard software and workflows developed by vendors and institutions such as Applied Technology Council, OpenSHA developers, and consultants serving agencies like Port Authority of New York and New Jersey. NGA outputs are embedded in performance‑based earthquake engineering design processes adopted by the American Institute of Architects and building officials in regions impacted by earthquakes including Chile, New Zealand, and Mexico. Transportation infrastructure projects funded or reviewed by Federal Highway Administration and World Bank teams often specify NGA‑consistent GMMs for design and retrofit analyses.

Validation and Uncertainty Analysis

Validation of NGA models relied on out‑of‑sample prediction tests against datasets from significant earthquakes recorded by networks including Japanese Meteorological Agency arrays, the Kinemetrics strong‑motion instrumentation deployments, and regional catalogs curated by Geoscience Australia. Researchers from Seismological Society of America meetings and journals evaluated aleatory and epistemic uncertainty quantification, comparing intra‑model variability and logic‑tree approaches promoted by NRC panels. Sensitivity studies by teams at University of California, San Diego and National Institute of Geophysics and Volcanology addressed residual trends, magnitude scaling, and site response scatter.

Impact on Seismic Hazard Assessment and Building Codes

NGA informed revisions to seismic hazard maps produced by United States Geological Survey and influenced code changes adopted by the International Building Code, American Society of Civil Engineers 7‑16 and later editions, and regional standards used in jurisdictions such as California Building Standards Commission. Its predictive improvements affected risk evaluations for critical facilities overseen by organizations like FEMA and the Department of Energy, and were incorporated into loss estimation models used by insurers including multinational firms operating in Europe and Asia.

Criticisms and Limitations

Critiques of NGA center on dataset biases toward certain tectonic regimes represented in records from California and Japan, limitations in extrapolating GMMs to low‑probability large‑magnitude ruptures like those studied after the 2011 Tōhoku earthquake and tsunami, and challenges in capturing complex site and basin effects noted in case studies from Mexico City and Istanbul. Additional concerns involve the portability of Vs30 as a universal site proxy debated by researchers at Lawrence Livermore National Laboratory and the need for improved physics‑based simulation integration promoted by groups at Los Alamos National Laboratory and Pacific Northwest National Laboratory.

Category:Seismology