Geoscience Reference
In-Depth Information
throughout the city. These profiles could be validated at two sites, where shallow bore-
holeshadbeenusedtomeasureshear-wavevelocitiesusingasuspendedloggingtool.The
amplificationcomputedfromthefinal1Dprofileswasingoodagreementwiththeprevi-
ous estimates of local amplification by Guti´errez et al. (1996). These results showed that
a credible microzonation map was not a viable alternative for the city of Colima. Due
to the large heterogeneity of the subsoil and the poor control on its properties the more
reasonable option was to propose a homogeneous amplification of a factor 6 throughout
thecityinthefrequencyband0.2to5Hz.Thisconclusionoffersnodetailsontheseismic
response, but has the merit of being consistent withall available data.
The case of Colima may be more representative of the usual problems faced to build
a model of the subsoil than the Los Angeles basin, to take a very contrasting example.
LosAngelesbasinhasbeenwellinstrumentedformanyyearsandhasbeentheobjectof
manydifferentstudies.Themodelthathasbeenderivedreflectsaconsensusamongmany
capablepeoplethathavecommittedtheireffortsinthistask.Asaresult,themodelisvery
reliable.Ithasallowedthetestingofmanydifferenthypothesesandthepredictionofuse-
ful scenarios. This effort cannot be reproduced easily elsewhere. Most site effect studies
must deal with very sketchy and incomplete information on the subsoil. Geotechnical
borehole data are very useful, but they seldom constrain the properties of the basement.
Thus,insiteeffectstudies,rough1Dprofilesaremorethenormthandetailed3Dmodels,
especially in developing countries.
4. Accounting for site effects in building codes
Asmentionedabove,ourabilitytoaccountforsiteeffectsinageneralframeworklikethat
of a building code lags well behind our ability to compute site response at specific sites.
Numerical simulation of ground motion can successfully tackle very complex models
these days. However, the results are difficult to use for general application. In the first
place, it is harder to verify the results of the model with the generality required by a
buildingcode.Inthesecondplace,themorecomplexthemodelisthelessgeneralwillbe
theresultscomputedforit,asbynatureeachspecificmodelisvalidforonlyonespecific
situation.Whenthemostsophisticatedmodelswere2D,itwaseasiertomakeparametric
studies(BardandBouchon,1980a,b).Whenweturnto3Dmodels,thispossibilityislost
because of the cost and the large number of parameters, in spite of interesting initiatives
like Sismovalp (http://www-lgit.obs.ujf-grenoble.fr/sismovalp/).
A very effective approach, one that is useful, is the classification of the soil response
using very simple parameters. The foremost example is the use of the average shear-
wave velocity for the topmost 30m, the Vs
30
value (e.g., Borcherdt, 1994). The reason
foritssuccessisitssimplicity.OnemayusetheVs
30
valuetoclassifythesiteofinterest.
Expectedamplificationhasbeencomputedforeachsoilclass.Lackingotherinformation,
if we are able to guess the average mechanical properties of the topmost 30m, this
approach is an honourable way out. There are, of course, some downsides. The largest,
in the author's opinion, is that it assumes that the properties of the materials below 30m
