Professional Documents
Culture Documents
794 Seismological Research Letters Volume 81, Number 5 September/October 2010 doi: 10.1785/gssrl.81.5.794
TABLE 1
Correlation of Notation Differences for Components of the Variability of Ground-Motion Models from Previous Studies with
the Current Study
Atkinson
(2006) ,
Anderson & Chen Rodriguez-
Uchiyama Lin et al. & Tsai Marek et al. Morikawa et al.
Standard Deviation New Notation (2010) (2010) (2002) (2009) (2008)
Total on soil surface σG σI σT σreg δ
Total on baserock σB
Between-events τ σevents τ σE τ (no correction)
Within-event φ σIIE σ σ (no correction)
Site-to-site φS2S σstations σS σS
Site amplification φAmp
Path-to-path φP2P σpath σP
Baserock-to-baserock φB2B
Earthquake location- τL2L τSR
to-location
Between-events, τ0 τ0 τ (with correction)
single-path
Within-event, single-
baserock �P2 2P + �20,B
Single-site (total) σSS σIIS σSS σi
Single-baserock (total) σSB
Single-path on soil G
σSP σSS(IV) σSP σie
surface (total) σ 2 + τ2
(with correction)
Single-path on baser- B
σSP
ock (total)
ect and for the PEGASOS Refinement Project (Renault et al. GROUND-MOTION MODELS
2010), and it is the hope of the authors that they will be widely
adopted for use in this field. The acronyms and indices that we In this section, we return to fundamentals and discuss the basic
are adopting are designed to be more or less self-explanatory. A formulation of GMPEs in order to introduce our preferred
capital “B,” for example, stands for “between” and a capital “W” nomenclature and to clearly define each component of the
for “within,” and the use of “2” in the standard deviation nota- models and their associated variability. An empirical GMPE
tion suggests that the corresponding variability is treated as an generally has the form:
epistemic uncertainty if a non-ergodic assumption is applied.
Moreover, subscripts e and s are the earthquake and site indices, Y = f ( X es ,ϑ ) + ∆ (1)
respectively, subscript l is the earthquake location index, and
superscripts “B” and “G” refer to baserock and ground surface, where Y is the natural logarithm of the observed ground-
respectively. This way, we use only a minimum number of Greek motion parameter, f ( X es ,ϑ ) is the ground-motion model,
symbols. For clarity, a glossary of terms presenting our proposed X es is the vector of explanatory parameters (e.g., magnitude,
terminology and symbols is provided at the end of this paper. distance, style of faulting, site conditions), ϑ is the vector of
Table 1 provides a correspondence between the new terminol- model coefficients, and Δ is a random variable describing the
ogy and those that have been used in various other studies. total variability of the ground motion. Δ is usually decom-
The distinction between within-event and between-events vari- Given multiple recordings at an individual site (site-specific
ability is very useful for quantifying, understanding, and han- dataset on soil surface), the site-to-site residual, δ S2S s , repre-
dling the ground-motion variability and addressing the corre- sents the systematic deviation of the observed amplification at
lation of the residuals. To reduce the value of sigma, however, this site from the median amplification predicted by the model
it is first necessary to decompose the variability into smaller using simple site classification such as the average shear-wave
parts. The within-event residual of ground-motion models velocity in the uppermost 30 meters at the site, V S 30 . The site-
includes systematic baserock or site-specific effects and path- to-site residual results from the use of simple site parameters,
specific effects. Similarly, the between-events residual contains which do not provide a complete site characterization. For
systematic source-specific effects. Removing these systematic example, if V S 30 is used for site classification, two sites with
effects is a key to removing the ergodic assumption from the the same V S 30 can still have significantly different site profiles
seismic hazard and requires repeated sampling of the site, path, and therefore have different site amplifications. This differ-
and source in question. ence in the site amplification caused by having sites with the
According to Walling (2009), estimating the various com- same V S 30 but different profiles is reflected in the site-to-site
ponents of the variability of ground-motion models depends on residual. The remaining site amplification residual, δ Ampes ,
the type of dataset available for the ground-motion regression. describes the record-to-record variability of the amplification
Walling (2009) describes five types of ground-motion datasets. at site s for earthquake e. It is the misfit between an individual
The three most commonly used types of datasets are: observation of the amplification at site s due to earthquake e
1. A global dataset contains recordings of ground motion from the average site-specific amplification and is caused by
at multiple sites from earthquakes in multiple regions. effects such as variability in incidence angle, 3D structure, or
Such datasets were used in the development of the Next variability of input waveform (phasing) of rock motion. The
Generation Attenuation models for the western United site-to-site and site amplification standard deviations are ϕ S 2 S
States (NGA-West), assuming that ground motion is simi- and ϕ Amp , respectively; ϕ S 2 S represents site-to-site variability
lar across all regions and sites for one tectonic class. within a site class, while ϕ Amp represents unexplained variabil-
2. A site-specific dataset contains multiple recordings at one ity in the site amplification at an individual site.
site from earthquakes located in different source regions The within-event residual on baserock, δ W esB , includes
and can be used to estimate the systematic and repeatable systematic wave propagation effects. If we have multiple record-
site-specific effects. ings at one site from a small source region (single-path dataset),
3. A path-specific dataset contains multiple recordings at one then we can separate these systematic path effects from the
site from earthquakes located in a small source region. It within-event residuals as follows:
can be used to estimate the source region, site, and path δ W esB = δ P2Psl + δ W esl0 , (7)
effects on the ground-motion model.
If downhole data at the site are available (site-specific dataset where δ P2Psl is the path-to-path residual at site s for source
on baserock), we can estimate the baserock-specific effects on region l and δ W esl0 is the remaining within-event residual for
the ground-motion model. The different components of the earthquake e at site s and in source region l. The path-to-path
between-events and within-event residuals and standard devia- residual represents the average shift of the observed site-specific
tions of ground-motion models are described below. region-specific ground motion from the median site-specific
model prediction and has a standard deviation of ϕ P 2 P . In
Components of the Within-Event Variability other words, it represents how the specific characteristics of
The ground motion at the soil surface at point G is the prod- this travel path lead to motions that are systematically different
uct of the input rock ground motion at point B and the site from the average obtained from the model. The within-event
amplification factor. Assuming linear site response, the natural residual, δ W esl0 , represents the remaining unexplained path
logarithm of the ground motion at point G can be written as: and radiation pattern effects and has a standard deviation of
yesG (T ) = yesB (T ) + ln Ampes ( T , X s ) , (5) ϕ 0,G .
In this paper, we assume linear site response; the effects of
where Ampes ( T , X s ) is the site amplification factor at sta- soil nonlinearity on the standard deviations of ground-motion
tion s for earthquake e and X s is the vector of explanatory models are discussed in Al Atik and Abrahamson (2010).
site parameters. The observed site amplification at station s Combining the previous equations and assuming linear site
for earthquake e can be broken down into the median ampli- response, the natural logarithm of the median ground motion
fication factor for the simple site classification to which site s on soil at point G and the intra-event residual on soil surface
is assumed to belong, µ esAmp ( T , X s ) , a site-to-site residual, can be written as:
δ S2S s , and a site amplification residual, δ Ampes , as follows:
µ Ges ( M e , Res , Fe , ZtoRe , X sG ) = µ esB ( M e , Res , Fe , ZtoRe , X sB )
(8)
+µ esAmp ( T , X s )
where δ B2 Bsb is the baserock-to-baserock residual for baser- where δ L2 Ll is the location-to-location residual for earth-
0
ock b underlying site s and δ W eslb is the within-event residual quakes in source region l and has a standard deviation of
for earthquake e located in source region l and at site s over- τ L2 L . It represents the average shift in the ground motion for
lying baserock b. The baserock-to-baserock residual represents an earthquake in a single region as compared to the median
the average shift of the observed input baserock-specific ground predicted by the global ground-motion model. The reason for
motion from the median baserock ground motion predicted by such a shift could be, for example, a lower or higher than the
the global model (for that magnitude and distance) and has average stress drop for that particular source region. δ Bel0 is
a standard deviation of ϕ B2 B . The path-specific within-event the remaining between-events residual for earthquake e after
0
residual on baserock, δ W eslb , represents all remaining effects removing the earthquake location-specific effects and has a
not included in the ground-motion model and has a standard standard deviation of τ 0. Assuming linear site response and
deviation of ϕ 0,B . uncorrelated components of the between-events residual, the
We assume that the components of the within-event and between-events residual on baserock and ground surface, τ B
between-events residuals are uncorrelated. Additional work is and τ G, respectively, are:
needed to test the correlation of the various variability compo-
nents. The within-event standard deviations on baserock and τG = τ B = τ 2L2 L + τ 20 . (14)
on ground surface, ϕB and ϕG, respectively, can be written as:
The total standard deviation of the global ground-motion
ϕB = ϕ 2B2 B + ϕ 2P 2 P + ϕ 20,B (11) model on baserock, σB , and on ground surface, σG, when apply-
ing the ergodic assumption can be written as:
and
B
ϕG = ϕ 2S 2 S + ϕ 2Amp + ϕ 2P 2 P + ϕ 20,G . (12) σ = ϕ 2B2 B + ϕ 2P 2 P + ϕ 20,B + τ 2L2 L + τ 20 (15)
and
Seismic hazard studies incorporate two types of uncertainty: σ SB = ϕ 2P 2 P + ϕ 20,B + τ 2L2 L + τ 20 . (21)
aleatory variability and epistemic uncertainty. By definition,
aleatory variability is the natural randomness in a process that In deterministic and probabilistic hazard studies that apply the
cannot be reduced with increasing knowledge about the pro- ergodic assumption, the total standard deviation of the empiri-
cess. Epistemic uncertainty is the scientific uncertainty in the cal ground-motion model on baserock or on the soil surface, σB
model of the process; it is caused by limited data and knowledge or σG respectively, is assumed to be an aleatory variability. All
and is characterized by alternative models. With increased data components of the variability are classified as aleatory variabil-
and knowledge, the epistemic uncertainty can, in theory, be ity when applying the ergodic assumption.
reduced to zero.
In the context of the definitions presented here, the differ- APPLICATION TO HAZARD STUDIES
ent components of the variability of empirical ground-motion
models with and without the ergodic assumption can be char- Removing the ergodic assumption in seismic hazard analyses
acterized as aleatory variability or epistemic uncertainty as leads to a large reduction in the aleatory variability of ground-
shown in Table 2 for common application scenarios. motion models. Tables 3 and 4 present a comparison of the
As discussed in the previous section, the site-to-site, path- standard deviations of ground-motion models obtained in pre-
to-path, baserock-to-baserock, and earthquake location-to- vious studies with and without the ergodic assumption for peak
location variability represent systematic effects that can be ground acceleration, PGA, and spectral acceleration at a period
removed from the variability of the ground motion if we have of 1 second, SA(1.0). Tables 2 and 3 show significant reduction
multiple path-specific recordings on baserock or on soil surface. in the aleatory variability of the ground motion for PGA and
Under the fully non-ergodic assumption (single-path scenario), SA(1.0) with the partially non-ergodic assumption (single-site
the site-to-site, path-to-path, baserock-to-baserock, and earth- or single-baserock ground motions). A much larger reduction
quake location-to-location standard deviations are, therefore, is observed when applying the fully non-ergodic assumption.
epistemic uncertainties. If we develop more elaborate models The removal of the ergodic assumption also impacts
that better characterize the complex nature of these processes, the median of ground-motion models. Under a non-ergodic
these uncertainties can be reduced. The total aleatory variabil- assumption, we can no longer use median ground-motion val-
ity of single-path ground motion on soil surface, σ GSP , and on ues predicted by models developed from a global dataset for site-
B
baserock, σ SP , can be written as: specific or path-specific scenarios. The non-ergodic assumption
requires adding epistemic uncertainty in the median-ground
σ GSP = ϕ 2Amp + ϕ 20,G + τ 20 (17) motion for each site/source combination. For single-site
GMPEs, epistemic uncertainty in the site amplification is
and required. This is commonly treated in site-specific site-response
studies conducted in major projects. Applying the fully non-
B ergodic assumption is more complicated because it requires
σ SP = ϕ 20,B + τ 20 . (18)
epistemic uncertainty to be added for every site-path combi-
nation and because the epistemic uncertainty in the median
When the partially non-ergodic assumption that only removes
ground motion will be correlated for closely spaced sources.
the systematic site-specific effects is applied to the ground
The spatial correlation in the systematic and repeatable effects
motion on soil (single-site GMPE), the site-to-site variability
needs to be addressed if the fully non-ergodic assumption is to
is an epistemic uncertainty and the single-site within-event
be applied to probabilistic hazard studies (e.g., Walling 2009).
residual at station s for earthquake e, δWSes, can be defined as:
CONCLUSIONS
δWSes = δWes – δS2Ss. (19)
The identification of the various components of ground-motion
The standard deviation of δWSes is called single-station variability is promising for the improvement of PSHA since
within-event standard deviation and is referred to as ϕSS . The some of the apparent randomness can be transformed to epis-
aleatory variability of single-site ground motion, σSS , can be temic uncertainty, and with additional data this uncertainty
written as can be removed. However, if it is not removed through the
acquisition of additional information (recorded data or numer-
σ SS = ϕ 2SS + τ 2 = ϕ 2Amp + ϕ 2P 2 P + ϕ 20,G + τ 2L2 L + τ 20 . (20) ical simulations), the increased epistemic uncertainty must be
included in the logic tree. If there is no additional data to con-
Similarly, applying the partially non-ergodic assumption that strain the epistemic uncertainty, the consequence will be that
only removes the systematic baserock-specific effects from the the mean hazard curve remains unchanged and the fractile haz-
variability of the ground motion on baserock, the baserock-to- ard curves will be broadened. This approach is still useful for
TABLE 4
Total Standard Deviation (Natural Logarithms) of SA(1.0) from Previous Studies
On Ground Surface On Baserock
σG σSS G
σSP σB σSB B
σSP
Atkinson (2006) 0.668 0.617 0.414 — — —
Chen & Tsai (2002) — — — — — —
Morikawa et al. (2008) 0.909 — 0.418 — — —
Lin et al. (2010) 0.741 0.636 0.435 — — —
Rodriguez-Marek et al. (2009) 0.804 0.6245 — 0.728 0.637 —