Long-Period Transition for Subduction Earthquake Spectra

The long-period transition period ($T_L$) is an important parameter in both seismological theory and seismic design. Current estimates of the parameter used for design are based on seismological model data from 2006, and presumably are for nonsubduction earthquakes. ASCE 7-16, which contains standards for seismic design, provides mapped estimates for $T_L$ based on the maximum considered seismic ground shaking. To expand the data considered for $T_L$, this study used the NGA-Sub flatfile obtained from the NGA Subduction Portal at the B. John Garrick Institute for the Risk Sciences. For each record, $T_L$ was interpreted as the period at which spectral displacement is a maximum, an interpretation which becomes evident upon close examination of the two-point design response spectra in modern codes. The rationale for this interpretation is the very definition of $T_L$ in ASCE 7-16. As performance-based design for multiple hazard levels has evolved and continues to gain acceptance by the engineering community, there is a need for a method to estimate $T_L$ for multiple levels of ground shaking at a given site. The primary purposes of this study were to develop estimates for $T_L$ for subduction zones and to facilitate the estimation of $T_L$ for lower-magnitude, more-frequent subduction events in subduction zones for performance-based seismic design, as opposed to using a single mapped value for a particular region. In this study, other methods for estimating $T_L$ also were explored. The long-period transition parameter, $T_L$, for subduction zones was found to vary with moment magnitude as the primary influence. A model in tabular form, similar to that used to develop maps in ASCE 7, was developed for subduction zones.