recurrence within ICR: a seismicity-based model (given a

1 second, and that of Atkinson and Boore extends to a

weight of 0.25); and a geology-based model (given a weight

period of 2 seconds. The EPRI (1993) relationship was

of 0.75). The seismicity-based model divides ICR into cells

extrapolated to a period of 2 seconds. This was

of one-half degree latitude and longitude and calculates

accomplished by extrapolating the coefficients of the

recurrence rates based on the historical seismicity in the

attenuation relationship and examining the reasonableness

cell. Different degrees of smoothing of seismicity rates and

of the resulting spectral prediction. The smooth quadratic

b-values among adjacent cells is accomplished using the

form of the relationship of Atkinson andBoore (1995)

methodology developed by EPRI (1988). In the geology-

underestimates their simulations of longer period ground

based model, Zone ICR is divided into subzones as

motions at distances beyond 100 km (62 miles). Therefore,

indicated in Figure E-23. Different combinations of

their relationships were modified at periods greater than

subzones are defined in a logic tree approach. The possible

0.5 second to result in ground motion estimates closer to

combinations are controlled in part by the presence or

the simulation results. Plots of the attenuation

absence of four possible tectonic boundaries within the ICR

relationships of EPRI (1993) and Atkinson andBoore

(Figure E-23) and the assessed likelihood that these

(1995) for peak ground acceleration and response spectral

features represent fundamental boundaries that control the

accelerations at 1.0 second are presented in Figure E-26.

distribution, rate, and maximum magnitudes of seismicity.

The modifications to the 1-second motion at distances

The logic tree for weights assigned to these boundaries is

greater than 100 km (62 miles) can be seen in the figure.

shown on Figure E-25. Thirty alternativesubzonations

The plots in Figure E-26 clearly indicate the distinctive

(not shown herein) of ICR result from the logic tree of

differences between the two eastern United States

Figure E-25. Within each subzone of each alternative,

attenuation relationships: the Atkinson andBoore (1995)

seismicity rates are determined based on the seismicity

relationships result in higher spectral values than those of

within the subzone and assuming the rate is uniform within

EPRI (1993) for peak ground acceleration and for short-

the subzone.

period response spectral accelerations (less than about

0.2 second period), but lower values than those of EPRI

(c) Probabilistic distributions of maximum

(1993) at longer periods.

earthquake magnitudes are also part of the source model

logic tree. These probabilistic distributions were

(c) In the hazard analysis, the relationship of

determined using the methodology developed by EPRI

EPRI (1993) was given a higher weight (0.67) than that of

(Johnston et al., 1994) that utilized worldwide data bases to

Atkinson and Boore (1995) (0.33). The reason for this

assess maximum earthquake magnitudes in stable

judgment was that the EPRI (1993) relationship resulted

continental regions (like the eastern United States (EUS))

from an EPRI study that involved input from a number of

where active faults have not been identified and therefore

ground motion experts and thus could be viewed as having

maximum magnitude cannot be estimated on the basis of

achieved a certain degree of consensus regarding the

fault dimensions (as is done in the western United States

model. The practical effect of higher weighting on the

(WUS)). However, for the New Madrid zone, maximum

EPRI (1993) model is to increase longer period ground

earthquake magnitudes were estimated on the basis of both

motions and reduce short-period ground motions.

(1) estimated rupture models by Johnston (1996) and

Gomberg and Ellis (1994) and correlations of magnitude

(3) PSHA Results. Hazard curves obtained from

with rupture dimensions, and (2)estimates of magnitudes

the analysis for peak ground acceleration and response

of the 1811-1812 earthquakes by Johnston (1996).

spectral acceleration at two periods of vibration are shown

in Figure E-27. The uncertainty bands around the mean

(2) Ground Motion Attenuation Characterization.

curves, reflecting the alternative seismic source models and

attenuation relationships incorporated into the logic tree,

(a) It was desired to estimate ground motions

are shown in the figure. The contributions to the hazard

on rock at the site. Two attenuation relationships

are almost entirely from Zone ICR. Figure E-28 shows

applicable to hard rock in the EUS for horizontal peak

contributions within ICR from large New Madrid Zone

ground acceleration and response spectral accelerations of

earthquakes with rates defined by paleoseismic data

ground motions at different periods of vibration were used.

(dashed-dotted line) and smaller earthquakes defined by

The relationships are those of EPRI (1993), (later published

seismicity (dashed line). It can be seen that the smaller

as Toro et al., 1997) and Atkinson and Boore (1995) (later

earthquakes dominate hazard at higher frequencies

published as Atkinson andBoore, 1997).

(probabilities) of exceedance and the larger, 1811-1812-

type earthquakes dominate at lower frequencies

(b) The relationship for response spectral

(probabilities) of exceedance. Figure E-29 compares the

acceleration of EPRI (1993) extends to periods as long as

hazard obtained from geology-based and seismicity-based

E-31

Integrated Publishing, Inc. |