Multidisciplinary Center for Earthquake Engineering Research logo google logo
navigation bar
bridgesmall.gif (4301 bytes)MCEER/NCEER Bulletin Articles: Workshop/Conference Reviews...

National Representation of Seismic Ground Motion for Highway Facilities

by Ian Friedland and Maurice Power

This article summarizes discussions from a recent FHWA/NCEER workshop conducted through NCEER's Highway Project, task 106-F-5.4.1. More detailed information is available in Proceedings of the FHWA/NCEER Workshop on the National Representation of Seismic Ground Motion for New and Existing Highways, NCEER-97-0010. Comments and questions should be directed to Ian Friedland, NCEER, at (716) 645-3391; email:

A significant amount of research has been conducted within the FHWA/NCEER Highway Project between 1993 and 1997 on how to adequately portray the national seismic hazard in highway design specifications and guidelines. This work included a review of existing national, state, and regional seismic hazard maps, an evaluation of alternative strategies for the future portrayal of the national seismic hazard, and the development of alternative recommendations for presentation to AASHTO and other highway design and specification authorities. During this time, the USGS published new national seismic ground motion maps in 1996 which appear significantly different than those currently in AASHTO specifications.

In order to ensure that the key issues related to the development of national seismic hazard portrayals were adequately addressed, NCEER organized and conducted the FHWA/NCEER Workshop on the National Representation of Seismic Ground Motion for New and Existing Highway Facilities on May 29 and 30, 1997, in San Francisco, California. The workshop provided a forum under which more than 50 earth scientists, geotechnical engineers, and structural engineers were brought together to discuss a number of these key issues and develop consensus recommendations with respect to their implementation in new highway facility design specifications (see review of the workshop in the July 1997 issue of the NCEER Bulletin, Vol. 11, No. 3).

The ground motion issues that have emerged in recent years as potentially important to highway facilities design and that were considered at the workshop were:

These issues were considered sequentially at the workshop. For each issue, selected workshop participants prepared papers and made presentations illuminating the issues and proposing a course of action in terms of design criteria and procedures and/or further development. Papers covering the presentations by each speaker are contained in the workshop proceedings (report number NCEER-97-0010).

Following the presentations on each issue, the workshop participants as a whole discussed the issues and developed conclusions and consensus recommendations. A final set of recommendations will be presented to AASHTO in the spring of 1998. The following summarizes the key elements in the discussion of each issue and the conclusions and consensus recommendations resulting from the workshop.

Using the New USGS Maps
In 1996, the U.S. Geological Survey (USGS) developed new seismic ground shaking maps for the contiguous United States (maps for Alaska and Hawaii are under development). These maps depict contours of peak ground acceleration (PGA) and spectral accelerations (SA) at 0.2, 0.3, and 1.0 second (for 5% damping) of ground motions on rock for probabilities of exceedance (PE) of 10%, 5%, and 2% in 50 years, corresponding to return periods of approximately 500, 1000, and 2500 years, respectively. These maps for the contiguous United States and separately for California and Nevada are available from and can be viewed or downloaded from the USGS World Wide Web site at

97OctFig1pg5.gif (11169 bytes)The workshop considered whether the new USGS maps should replace or update the maps currently incorporated in AASHTO specifications, which were developed by the USGS in 1990. The key issue regarding whether the new USGS maps should provide a basis for the national seismic hazard portrayal for highway facilities is the degree to which they provide a scientifically improved representation of seismic ground motion in the United States. Based on an analysis of the process of developing the maps, the inputs to the mapping, and the resulting map values, the workshop concluded that these new maps represent a major step forward in the characterization of national seismic ground motion. The maps are in substantially better agreement with current scientific understanding of seismic sources and ground motion attenuation throughout the United States than are the current AASHTO maps. The workshop therefore concluded that the new USGS maps should provide the basis for a new national seismic hazard portrayal for highway facilities.

The workshop also examined the issue of an appropriate probability level or return period for design ground motions based on the new USGS maps. Analyses were presented showing the effect of probability level or return period on ground motions and comparisons of ground motions from the new USGS maps and the current AASHTO maps. The workshop recommended that for design of highway facilities to prevent collapse, consideration should be given to adopting probability levels for design ground motions that are lower than the 10% probability of exceedance in 50 years that is currently in AASHTO (i.e., ground motion return periods longer than 500 years should be considered). This recommended direction is consistent with proposed revisions to the 1997 NEHRP Provisions for buildings, in which the new USGS maps for a probability of exceedance of 2% in 50 years (an approximate 2500 year return period) have been adopted as a collapse-prevention design basis. (The NEHRP proposal for buildings was described at the workshop and is summarized in the Proceedings.) Figure 1 provides an example of how Maximum Considered Earthquake (MCE) ground motion maps were developed for use in design (see paper by Hamburger and Hunt in Friedland et al., editors, 1997).

Using Energy or Duration in Design Procedures
At the present time, the energy or duration of ground motions is not explicitly recognized in the design process for bridges or buildings, yet many engineers are of the opinion that the performance of a structure may be importantly affected by these parameters, in addition to the response spectral characteristics of the ground motion. Based on the presentations and discussions at the workshop, the participants concluded that some measure of the energy of ground motions is important to the response of a bridge, but, at present, we do not have an accepted design procedure to account for energy. Research in this area should be continued to develop energy-based design methods that can supplement current elastic response-spectrum-based design methods. The workshop also concluded that energy, rather than duration, is the fundamental parameter affecting structural behavior.

Characterizing Site Effects
At the Workshop on Site Response During Earthquakes and Seismic Code Provisions, held in 1992 at the University of Southern California (USC), a revised quantification of site effects on response spectra and revised definitions of site categories was proposed. Subsequently, these revised site factors and site categories were adopted into the 1994 NEHRP Provisions and the 1997 Uniform Building Code (UBC). Since the development of these revised site factors, two significant earthquakes occurred, Northridge in 1994 and Kobe in 1995, which provided substantial additional data for evaluating site effects on ground motions, and research using these data has been conducted.

The site factors and site categories in the current AASHTO specifications are those that were superseded by the USC workshop recommendations for the NEHRP Provisions and the UBC (see Table 1, taken from Dobry et al. in Friedland et al., editors, 1997). The questions for consideration at this workshop were whether the USC workshop recommendations should be utilized in characterizing ground motions for highway facilities design and whether they should be modified to reflect new data and new knowledge since the 1992 workshop. The most significant differences in the USC workshop recommendations and the previous site factors (those currently in AASHTO) are: (1) the revised site factors include separate sets of factors for the short-period and long-period parts of the response spectrum, whereas the previous site factors were only for the long-period part; (2) the revised site factors are dependent on, rather than independent of, intensity of ground shaking, reflecting soil nonlinear response; and (3) the revised site factors are larger (i.e., show a greater soil response amplification) than the previous factors at low levels of shaking, which is important for the lower-seismicity regions in the United States.

The workshop found that the post-Northridge and post-Kobe earthquake research conducted to date generally was supportive of the site factors derived during the 1992 USC workshop, although revisions to these factors might be considered as further research on site effects is completed. The workshop therefore recommended that the factors developed at the USC workshop and adopted by the NEHRP Provisions and the UBC be proposed as part of a new national representation of seismic ground motion for highway facilities design.

Table 1: Site Categories in New Building Codes (NEHRP 1994, UBC 1997)

Soil Profile Type



top 100 ft (fps)

S1 { A


Hard rock


> 5000

2500 - 5000

S1 and S2 { C


Very dense soil/soft rock

Stiff soil

1200 - 2500

600 - 1200

S3 and S4 { E


Soft soil

Special soils requiring site-
specific evaluation

< 600

Additionally, current AASHTO specifications incorporate a conservatively slow decay of long-period response spectra with increasing period (response spectral accelerations proportional to 1/T 2/3 , where T = period, rather than 1/T as is typically seen in ground motions). The workshop recommended that the new higher site factors not be coupled with the conservative long-period response spectral decay currently in AASHTO. Rather, long-period ground motions should be permitted to decay in a more natural fashion, i.e., approximately proportional to 1/T rather than 1/T 2/3 .

Specifying Vertical Ground Motions
97OctFig2pg5.gif (16406 bytes)At present, the AASHTO specifications do not contain explicit requirements to design for vertical accelerations. Ground motion data from many earthquakes in the past 20 years have shown that, in the near-source region, very high short-period vertical spectral accelerations can occur. For near-source moderate- to large-magnitude earthquakes, the rule-of-thumb ratio of two-thirds between vertical and horizontal spectra is a poor descriptor of vertical ground motions. At short periods, the vertical-to-horizontal spectral ratios can substantially exceed unity, whereas at long periods, a ratio of two-thirds may be conservative (see Figure 2 in paper by Foutch in Friedland et al., editors, 1997). The workshop demonstrated that our current understanding and ability to characterize near-source vertical ground motions is good, especially in the western United States where the near-source region is relatively well defined (i.e., near mapped active faults).

The workshop also demonstrated that high vertical accelerations as may be experienced in the near-source region can significantly impact bridge response and design requirements in some cases. On the basis of these findings, it was concluded that vertical ground motions should be considered in bridge design in higher seismic zones for certain types of bridge construction. It was recommended that specific design criteria and procedures be developed for identified bridge types.

Specifying Near-source Ground Motions
The workshop examined both the characteristics of near-source horizontal ground motions and the effects of near-source ground motions on bridge response. As the distance to an earthquake source decreases, the intensity of ground motions increases, and this increase in ground motion intensity is incorporated in new USGS maps. However, in addition to their higher intensity, near-source ground motions have certain unique characteristics that are not found at greater distances. The most significant characteristic appears to be a large pulse of long-period motions when an earthquake rupture propagates to-ward a site. Furthermore, this pulse is larger in the direction perpendicular to the strike of the fault than in the direction parallel to the strike. This characteristic of near-source ground motions has been observed in many earthquakes, including most recently in the Northridge and Kobe earthquakes. Preliminary analyses of bridge response presented at the workshop indicate that near-source ground motions may impose unusually large displacement demands on bridge structures. The workshop concluded that traditional ground motion characterizations (i.e., response spectra) may not be adequate in describing near-source ground motions, because the pulsive character of these motions may be more damaging than indicated by the response spectra of the motions. The workshop recommended that additional research be carried out to evaluate more fully the effects of near-source ground motions on bridge response and to incorporate these effects in code design procedures. Until adequate procedures are developed, consideration should be given to evaluating bridge response using site-specific analyses with representative near-source acceleration time histories.

Specifying Spatial Variations of Ground Motions
Spatial variations of ground motions along a horizontally-extended structure such as a bridge include (1) spatial incoherency in ground motions due to scattering of the propagating seismic waves by the geologic media as well as spatial variations in wave superposition from seismic waves arriving from an extended earthquake source; (2) wave passage effects, in which non-vertically incident seismic waves arrive at different locations along the structure at different times (time-lag effects); (3) attenuation effects, in which ground motion amplitudes decrease with increasing distance from the earthquake source; and (4) differential site response due to variations in the geologic conditions along the structure (which can include two- and three-dimensional site response effects in basin environments, as well as simple one-dimensional site response effects). For important long-span bridges, procedures are available and have been employed in many cases which take these effects into account in relatively sophisticated site-specific analyses. The issue addressed at the workshop was whether there are classes of structures (e.g., related to bridge span length and other characteristics) for which spatial variations of ground motions may safely be neglected in design or the effects of these variations incorporated using simplified code-type design procedures.

Results of analyses were presented through which the effects of spatial variations of ground motion were systematically examined. Generally, these analyses indicated that in the absence of strong differential site response effects, the response of "ordinary" highway bridges was not greatly affected by spatial variations of ground motion. However, the workshop concluded that we cannot yet adequately define those categories of bridges for which spatial variations of ground motions can be neglected, even for the case of relatively uniform soil conditions along the bridge. Further research is needed to define the importance of spatial variations of ground motions as a function of bridge characteristics and to develop simplified procedures for incorporating the effects of these variations in design.

NEHRP, (1994), Recommended Provisions for Seismic Regulations for New Buildings, FEMA 222A/223A, May, Vol. 1 (Provisions) and Vol. 2 (Commentary).

Friedland, I.M., Power, M.S., and Mayes, R.L., editors, (1997), "Proceedings of the FWHA/NCEER Workshop on the National Representation of Seismic Ground Motion for New and Existing Highway Facilities," Burlingame, California, May 29-30, 1997, NCEER-97- 0010.

UBC, (1997), Uniform Building Code.


NCEER Bulletin, Vol. 11, No. 4, October 1997

top.gif (1113 bytes)

  Contact Us  |  Acknowledgements   |  Disclaimer  |  Copyright© 2007 by the Research Foundation of the State of New York. All rights reserved.