This text is slightly adapted from a lecture I gave at Oxford in June 2022.
In PSHA, we are seeking to characterize the behavior of a system. In regional
PSHA, we are looking at a system of seismic sources (typically faults and
background sources), as well as a heterogeneous lithosphere and its irregular
surface, where we want to calculate the effect of seismic waves generated by the
sources. For a site-specific or scenario-based analysis, we are cutting down on
the dimensionality of the problem in some capacity but we may be adding more
depth or complexity of analysis for whatever parts of the system that we are
analyzing.
The different components of the system are interactive, and to do the best job
possible we have to be able to describe each component and understand how it
interacts with other components. Working in this manner is different than
considering the components in isolation, as is often done, but this presents
opportunities in addition to challenges, because we can use system-level
constraints such as budgets to reduce uncertainties and to make inferences where
measurements are difficult or otherwise lacking.