Current USGS probabilistic seismic hazard estimates (Frankel et al., 2002) show a higher seismic hazard in southwestern Indiana than for the rest of the state. This is expected based on past work documenting large pre-historic events in the Wabash Valley, and also based on isoseismal levels of shaking from the New Madrid 1811-1812 earthquakes. When assessing the hazard on a county by county basis, the local soil conditions produce amplification of shaking in regions with soft or unconsolidated sediments relative to firm-rock sites. We have reproduced the probabilistic seismic hazard calculation following the USGS methodology for the state of Indiana and provided details for the shaking level on each county of the state. Here we compare the 1996 and 2002 versions of the seismic hazard maps [Frankel et al., 1996; Frankel et al., 2002]. We also compare the results with an assumed maximum magnitude for the New Madrid earthquake of 7.3. This produces significantly lower acceleration levels. The USGS probabilistic maps were constructed assuming firmrock sites. Versions of the maps were produced assuming other NEHRP classification levels, however no information was provided specific to the state of Indiana that suggests which site classification is appropriate for a given location. More recent studies have attempted to provide more detailed soil amplification classifications for the central U.S. These are described by [Bauer et al., 2001] and [Street et al., 2001]. A new methodology has been developed to incorporate soil profile site effects into the probabilistic calculation [Cramer, 2003; Cramer et al., 2003]. We demonstrate the usefulness of this approach by applying it to a low resolution near-surface velocity model based on a limited amount of soil profile data in Indiana, and we report on the relative amplification and deamplification expected given the available data. The results show particular regions of amplification on the order of 2 or greater for both the 1 Hz and the 5 Hz spectral acceleration with 2% probability of being exceeded in 50 years, between the standard NEHRP B/C site and the probabilistic site effect calculation. For the 1 Hz maps, this occurs in the central and central northern part of the state including the Indianapolis metropolitan area, where the surficial units of glacial tills are particularly thick because of the sediments that both fill the bedrock Teays Valley and that overlie it in thickness up to 1,370 meters. Lower amplitudes are noted for much of the south central part of the state because bedrock crops out or is present near the surface. These results are useful because they give a first order estimate that illustrates the potential effect of geology. They are demonstration maps that lack the detailed data required for practical use, but serve as a proof-of-concept for the probabilistic site effect methodology. Even with these approximate maps it is clear that there is a significant variation in the number of counties that would be impacted by the engineering design criteria, depending on the level of approximation in the methodology chosen. We describe the datasets that we have collected and used to establish the input geology-based velocity model, including comparisons with independent datasets that allow us to estimate the true uncertainty of the measurements. Future versions of PSHA maps with site effects are planned that will use a database with higher resolution information on shear wave velocity structure. These future maps will provide information at the level necessary for planning and budgeting, though site specific studies will still be needed for engineering.

Report Number



seismic hazard, seismic design, bridges, earthquake, risk, acceleration, SPR-2812

SPR Number


Project Number


File Number


Performing Organization

Joint Transportation Research Program

Publisher Place

West Lafayette, IN

Date of this Version