Task I – Test of GMPEs


  • To develop simple relationships to relate paleoliquefaction and intensity data to ground motion intensity measures.
  • To evaluate the GMPEs developed in tasks H.1 and H.3 and J.5-6 against finite-fault simulations results, CENA and other SCRs data and the intensity measures associated with paleoliquefaction and intensity data.

Four main sets of data will be used to evaluate the GMPEs developed in task H: 1) ground motion data from CENA and other SCRs (for large magnitudes not represented in the CENA database), 2) finite-fault simulation results and 3) liquefaction and 4) intensity data from past large CENA earthquakes. Given that some of the data has large inherent uncertainty, and that some data come from simulations and not observations, the goal of these tests is not to calibrate the GMPEs, but rather to check if the different data sets are consistent with the GMPEs. The TI team will also use finite-fault simulation ground motions (I.2) to ensure that their properties have been carried through the GMPE development process. The paleoliquefaction data are expected to come directly from the CEUS SSC project, but will need to be associated with ground motion intensity measures. A simple database of earthquake intensity will be developed by a contracted researcher, possibly from the USGS, who will also need to associate the intensity with ground motion intensity measures . Finally, a member of the TI team will conduct the test of the GMPEs against these data to identify any inconsistency. The tests will be conducted at the logic tree level, but on single branches that combine the median and standard deviation models. The TI team will adjust the weights for the GMPEs as needed based on the test results (task L).