NGA-East: National Seismic Hazard Mapping Perspective
-
Upload
karly-mendez -
Category
Documents
-
view
32 -
download
0
description
Transcript of NGA-East: National Seismic Hazard Mapping Perspective
NGA-East: National Seismic Hazard Mapping Perspective
Mark Petersen
USGS
Golden, CO
New Madrid Ground Motions
• Tennessee: Lowered building codes• Arkansas: Legislative hearing to lower building
codes (Have these ground motions occurred historically?, How do they compare to California earthquake ground motions?)
• Kentucky: Ground motion equations are conservative, uncertainties high, PSHA not valid- mixing temporal
• Building code community: Can these ground motions occur? Are there physical limits that can be applied? Have we adequately accounted for the distribution of ground motions and truncation levels?
New Hazard Maps
Ratio of new hazard maps with 2002 maps
Figure 9: CEUS 0.2 s SA attenuation relations for M 7 earthquake on Vs30 760 m/s site conditions: AB95 AB05 (Atkinson and Boore, 1995, 2005; F96 (Frankel et al., 1996); T97 T02m (Toro, 1997, 2002); C03 (Campbell, 2003); S01 (Somerville 2001); SV02 (Silva et al., 2002); TP05 (TavakoliAnd Pezeshk, 2005) From Chris Cramer
Figure 10: CEUS 1 s SA attenuation relations for M 7 earthquake on Vs30 760 m/s site conditions: AB95 AB05 (Atkinson and Boore, 1995, 2005; F96 (Frankel et al., 1996); T97 T02m (Toro, 1997, 2002); C03 (Campbell, 2003); S01 (Somerville 2001); SV02 (Silva et al., 2002); TP05 (TavakoliAnd Pezeshk, 2005)
From Chris Cramer
FromAtkinson and Boore(2006; updated 2007)
Does this contradictIntensity data fromCEUS and WUS Earthquakes?
Differences
1. Geometrical spreading (r-1.3 vs r-1.0)2. Stress drop3. Site kappa (0.01 vs 0.02)
Amplitudes decay faster than 1/R at R<70 km. Amplitudes decay faster than 1/R at R<70 km. This has important implications for This has important implications for ENA ground motion relations.ENA ground motion relations.
From Gail Atkinson
Average Stress Drop Determinations
570 bars
290 bars
110 bars
≥410 bars
From Art Frankel
Some intriguing ground motion information from MMI, based on the USGS “Did You Feel It?” website database.Higher ENA MMI appears to persist to near-source, suggesting higher stress drops
Slide and caption from Gail Atkinson
From Art Frankel
From Frankel et al. (1990)
Conclusions
• Need more information on geometric spreading, stress drop, kappa
• Need to resolve the discrepancy between new attenuation relations and MMI data
• Need to determine whether high values over New Madrid are reasonable
• Need to involve CEUS ground motion modelers (Hermann, Chapman, Ebel, others)
Corrected for Atkinson Q model and B-C boundary site amplification
Acc
eler
atio
n S
pect
ral A
mp.
Frequency (Hz)
kappa= 0.01
kappa= 0.02
Vs30= 705 m/secVs30= 470 m/sec
We used kappa=0.01 for CEUS B-C sites
location dateMw
(From Herrmann)
MblgDepth
(km)
Corner Frequency
(hz)
Stress drop
(bars)
W. KY6/20/05
12:213.6 3.7 21 8-12 270-880
N.E. AR5/1/05
12:374.2 10 2-5 32-510
MO-AR border
10/21/99
8:183.7* 3.9 11 ≥ 8 ≥410
Fort Payne,
AL
4/27/03
8:594.6 4.9 12 2-3 130-440
S.W. Utah
2/23/01
21:434.2 10 2 32
S.W. Utah
10/22/99 19:06
4.0 6 3 55
Stress Drop Determinations from Spectral Ratios
* Determined from comparison of spectral levels with 5/1/05 event
From Art Frankel
USINVs30= 705 m/sec
OLILVs30= 470 m/sec
What is kappa (high-frequency spectral falloff) for CEUS B-C sites? Vs30= 760 m/sec
5/1/05 earthquake, NE Arkansas, M4.2, fc= 2 Hz (at OLIL)
Corrected for Atkinson Q model and B-C boundary site amplification
Acc
eler
atio
n sp
ectr
al A
mp.
Frequency (Hz)
USIN OLIL
PVMO
Mw= 3.6 (Herrmann), fc= 8-12 Hz, stress drop=270-880 bars
fc= 12 Hz fc= 12 Hz
fc= 9 Hz
8
UTMT
fc= 8 Hz
Western Kentucky 6/20/05 12:21 UTC, d= 21 km S
pect
ral R
atio
Spe
ctra
l Rat
io
Frequency (Hz) Frequency (Hz)
All spectral ratios wrt Md 2.7 foreshock
8
USIL
red: fc= 12 Hzblue: fc= 8 Hz
OLIL
red: fc= 12 Hzblue: fc= 8 Hz
uncorrected spectra (proportional to velocity 1-10 Hz)
6/20/05 W. Kentucky Station USINDist= 167 km
From Art Frankel
Brune spectra forearthquakes whose moments
differ by factor of 125 and have equal stress drops
Ratio of spectrum of larger eventdivided by spectrum of smaller event
fc= 5 Hz
fc= 25 Hz
fc= 5 Hz
fc= 25 Hz
Taking spectral ratio of collocated earthquakes removes path and site effects (and instrument response) and provides estimate of corner frequency of large event
(and the small event if you have sufficient bandwidth)
From Art Frankel
Red stars showearthquakes studied here, Mw 3.6-4.6
Triangles are broadband stationsof the New Madrid seismic network andother ANSS stations used inthis study
From Art Frankel