PEER has just published Report No. 2020/05: "Conditional Ground-Motion Model for Peak Ground Velocity for Active Crustal Regions." It was authored by Norman A. Abrahamson, University of California, Berkeley and Sarabjot Bhasin, Duke University.
Conditional models for the horizontal and vertical peak ground velocity (PGV), given the pseudospectral acceleration [PSA(T)] values, are developed for active crustal regions. The period of the PSA(T) used in the conditional model, TPGV , is magnitude dependent, which captures the effect of the magnitude dependence of the earthquake source corner frequency on the PGV. Conditional models can be used to estimate the PGV given a design spectrum and are applicable for magnitudes between 3.0 and 8.5, and for distances up to 200 km. The conditional PGV models can also be combined with appropriate GMMs for PSA(T) to develop traditional GMMs for PGV that are consistent with the more complex scaling included in the PSA(T) model. Unlike previous conditional PGV models, the slope on the ln[PSA(T)] term is allowed to be different from unity. With this feature, an appropriate aleatory standard deviation of the resulting ln(PGV) can be computed, avoiding the over-prediction of the aleatory standard deviation of the PGV seen in previous conditional PGV models.