- Ground-motion model, based on Graizer and Kalkan (2015) (see Section 2.391), is: where c1 = 0.40, c2 = -6.25, c3 = 0.55, c4 = 2.237, c5 = -7.542, c11 = 3.9, c12 = -0.3445, F = 2.232 and
σ = 0.53.
- Uses V s,30 to characterise sites. States that model applies for 450 ≤ V s,30 ≤ 2800m∕s.
- Uses data from NGA-East database (both mid-continent and Gulf Coast regions). Also includes 6 records
from 12/11/2014 Mw4.8 Kansas earthquake.
- Notes that insufficient data, especially for Mw > 6 and close source-to-site distances, in NGA-East database
to derive purely empirical model.
- Includes the regional quality factor for 1Hz, Q0, to model anelastic attenuation (notes that Q0 for central
and eastern USA is between 650 and 1000).
- Adjusts coefficients from the western USA model (Graizer and Kalkan, 2015) by: a) ratios between
observations in NGA-East and NGA-West databases for 3.75 < Mw < 6, b) average stress-drop ratio
between western and central and eastern USA and c) by checking predictions w.r.t. recent ground-motion
simulations for central and eastern USA for Mw > 5.
- Presents equations (Equation 9.6) for adjusting model to other V s,30 through the terms G4 but it is not
clear how they are used as in the presentation of the model (Figure 9.2) lnG4 is set to zero. Similarly 4
additional coefficients (c6 = -0.125, c7 = 1.19, c8 = -6.13 and c9 = 0.60) are listed in Figure 9.2 but
they are not included within the presented model.
- Compares predictions and observations binned into magnitude ranges against rrup.
- Presents functions to correct for trends w.r.t. rrup in residuals. For PGA correction is: PGAcor =
PGA∕Residual where Residual = exp(-0.000257rrup + 0.270).