Challenges in Selection of Ground Motions and Time Histories: A Geotechnical Engineer’s Perspective

September 21, 2022

In the early days, peak ground acceleration(PGA) is the only key parameter used in the seismic design of new structures or the seismic performance assessment of the existing structures. Earthquake records used in dynamic analyses were scaled to the PGA. But the advent of newer ground motion prediction equations (GMPEs) facilitate the development of spectra over a period range up to ten seconds. Probabilistic and deterministic seismic hazard assessments are usually performed to derive ground motions. Either uniform hazard response spectra or deterministic spectra are developed. Seismic hazard models to derive probabilistic based ground motions up to the 2,475 year return period are normally available for most places in seismically active regions. These are regional models and may not be appropriate for developing ground motions at lower probability levels and site specific seismic hazard assessment are normally required.

The scope of the site specific seismic hazard assessments, however, vary depending on the level of uncertainties in the source characterization and the GMPEs. It is also sometimes dictated by how the ground motions are used in the design, whether they govern the overall design of the structure, and the risk and cost associated with accepting greater uncertainties in ground motions. The choice of acceptable level ground motion estimate, whether mean, median or 84th percentile from a probabilistic assessment or median or 84th percentile from a deterministic assessment or how to reconcile the differences between the probabilistic and deterministic assessments still remain a challenge as these assessments depend on how appropriate or good these assessments are.

With increased reliance on performance-based design, how to handle the aftershocks has also become a challenge without adequate guidelines to assess the hazard due to aftershocks and to develop spectra or time histories to represent them. A key challenge is how to incorporate the aftershocks into the probabilistic frame work to evaluate the acceptable hazard. Vertical ground motions in the geotechnical seismic design is another area which remains a challenge due to lack of understanding of its impact on geo-structures and inadequate or evolving guidance to develop the vertical spectra or the time histories.

While the spectra alone is adequate for most of the routine structural design or assessment, geotechnical design require time histories for most of the analyses. Although earthquake records form the basis for developing the spectra through seismic hazard assessment with the use of GMPEs, most seismic hazard assessments stop at developing the spectra. Earthquake records are normally developed subsequent to the development of spectra by scaling or matching the spectra, through a reverse process. Selection and scaling of records has been and continue to be a challenge for geotechnical engineers as there is no clear or consensus guidelines to select or scale time histories. An overview of the various methods used in the current practice for the seismic hazard assessment and for the selection and scaling of earthquake records will be presented and various challenges faced by the geotechnical engineers and their impact on the seismic design of structures will be described using examples. Recommendations for seismic hazard assessment and for selection and scaling of earthquake records will also be made.



Thavaraj, T. 2022. “Challenges in Selection of Ground Motions and Time Histories: A Geotechnical Engineer’s Perspective.” in Proceedings of the 28th Symposium of the Vancouver Geotechnical Society, September 16, 2022. [Vancouver, BC: Vancouver Geotechnical Society.] [Abstract and Presentation only.]