Dear Hyeuk,
I understand the issues you raise and it could be interesting to
hear others' feedback on this first one. I can't speak on behalf
of the GEM development team, but personally I have always been
cautious about throwing even warnings related to issues with the
GMPEs when they are not actually a computational problem in
execution but rather a theoretical problem in the selection. One
reason is that there are simply so many combinations of GMPEs that
should not be used together, as well as various degrees of
incompatibility, that implementing the various checks and throwing
meaningful warnings can end up becoming a surprisingly large task.
Regarding the site model issue, there is work under way in this
direction, albeit in the early stages, to add more capability to
OQ to handle different user defined site amplification models and
their uncertainties. The one you highlight is noted, as are
others. At this point I can't pinpoint a particular date or
release for which we plan to have these features, as this will
depend a lot on the requirements and the complications/challenges
that arise, to say nothing of the time pressures on GEM staff and
OQ contributors in 2018, so for now all I can say is that it is
now it is on the table and that we're looking into it.
Thanks,
Graeme