Ground motion selection for seismic risk analysis of civil infrastructure - Seismic Risk Analysis and Management of Civil Infrastructure Systems - page 107

this algorithm in OpenSHA upon further investigation into user-interface
preferences and interfacing with as-recorded ground motion databases (e.g.
the PEER NGA database), as well as further investigation on incorporation
of simulated ground motions.
As was noted with respect to Table 4.1, in order for ground motion inten-
sity measures to be utilized in this methodology, empirical prediction equa-
tions must be available. While the examples in this chapter made use of 17
different ground motion intensity measures, there is still scope for the con-
sideration of other intensity measures which may be important for particu-
lar problems. Furthermore, it is well known that GMPEs for different
tectonic regions vary signifi cantly, such that a GMPE for an active shallow
crustal tectonic region is not applicable for a subduction slab tectonic
region. Empirical ground motion and correlation equations are most readily
available for active shallow crustal tectonic regions, but are lacking for
other tectonic regions (Bradley, 2012b). Hence, there is a clear potential to
improve the application of the proposed ground motion selection method-
ology for subduction interface, subduction slab and stable continental
regions, by developing empirical prediction equations for various intensity
measures due to earthquakes in such regions. It is envisaged that the imple-
mentation of the GCIM method in OpenSHA will facilitate the application
of the ground motion selection method to other tectonic regions because
of its open-source nature.
For regions in which the necessary empirical ground motion and correla-
tion equations currently exist (e.g. active shallow crustal tectonic regions),
the benefi ts provided by the proposed ground motion selection methodol-
ogy can be more widely distributed by providing 'default' ground motion
sets for common seismic response problems (e.g. mid-rise structures) in
large population centres. This will enable projects which do not possess the
budget for 'site-specifi c' ground motion selection, to at least employ 'region-
specifi c' ground motion selection, and will further facilitate the uptake of
the methodology by a wider audience.
Finally, it was noted at the beginning of the chapter that the methodology
described here is oriented toward the selection of ground motions consis-
tent with PSHA outputs (i.e. to quantify seismic performance for one or
more exceedance probabilities, including computation of the seismic
demand curve). However, it was also mentioned that quantifying the seismic
performance of a structure given a particular seismic rupture (i.e. obtained
from a deterministic seismic hazard analysis) provides complementary
information to that of the probabilistic case. The ground motion selection
methodology presented can easily be adapted for a particular scenario
event; in such a case, the conditional distribution f IM | IM J ( im | im j ) based on
which ground motions were selected is simply replaced by the uncondi-
tional distribution, f IM ( im ).