Upload
joseph-marley-abdul-jabbar
View
217
Download
0
Embed Size (px)
Citation preview
7/23/2019 Expo Jamg Axa 2015
1/19
On the computation of H/V and
its application to micro zonation and seismic
Jos Antonio
MARTNEZ
-
GONZLEZ
Mathieu
Perton,
Javier
Lermo
, and F
. J.
Snchez
-
SesmInstituto
de Ingenieria,
UNAM
2nd AXA UNAM Workshop
7/23/2019 Expo Jamg Axa 2015
2/19
Traditionally H/V:
- allows good estimation of fundamental frequency of
- has a bad recovery of amplification factor
- Is not reliable
- INTRODUCTION
Objectives
Methodology
Data acquisition
Data processing
Validation of results
Synthetic time-
series
- SHAKEMAPS
- CONCLUSIONS
Introduction
Why?
7/23/2019 Expo Jamg Axa 2015
3/19
Subduction
Introduction Seismic Hazard in Mexic Sources Paths Site effects
1D effect
3D effec
The earthquakes in Mexico are of 3 types
Subduction Zonemotion mainly composed of surface waves
Intraplate Zonemotions are composed of surface & body wa
Local Events ground shaking mainly composed of body wave
De
intra
- INTRODUCTION
Objectives
Methodology
Data acquisition
Data processing
Validation of results
Synthetic time-
series
- SHAKEMAPS
- CONCLUSIONS
7/23/2019 Expo Jamg Axa 2015
4/19
We also distinguish the events inrecording site. Mexico City has tygeotechnical zones :
Former Lake surface waves (loamplification,
Transition body waves (high framplification,
Hill Zone response sensitive to
Introduction: path and site ef- INTRODUCTION
Objectives
Methodology
Data acquisition
Data processing
Validation of results
Synthetic time-
series
- SHAKEMAPS
- CONCLUSIONS
7/23/2019 Expo Jamg Axa 2015
5/19
Introduction
Objectives
- Compare H/V from earthquakes (HVSR) with H/V from noise
- Establish shakemaps for average peak ground accelerations for the maximum acceleration structural period and 5% dam
Outline
- Data acquisition for seismic noise
- Data processing applied to seismic noise
- Data processing applied to earthquakes
- INTRODUCTION
Object ives
Methodology
Data acquisition
Data processing
Validation of results
Synthetic time-
series
- SHAKEMAPS
- CONCLUSIONS
Introduction
7/23/2019 Expo Jamg Axa 2015
6/19
We record seismic noise at 8 accelerometric stationsCity
One at Hill Zone,
Another at Transition Zone,
Six more at the former Lake Zone
Use of broad-band seismograph
Two months of continuous recording
- INTRODUCTION
Objectives
Methodology
Data acquis it ion
Data processing
Validation of results
Synthetic time-
series
- SHAKEMAPS
- CONCLUSIONS
Data acquisition for noise
7/23/2019 Expo Jamg Axa 2015
7/19
Noise processing:Classical Method (i.e. Geopsy)
This formula is appropriate
to remove source effectsin any window
Processing Parameters
Baseline correction
Filter Passband: 0.1 30 Hz,
Window length: 81.92 s
Disable tapper and
smoothing options
Z
N
E
1
1[ / ] [ / ]( )n
H Hn
V V i
Window length
[ / ]H V i
H
- INTRODUCTION
Objectives
Methodology
Data acquisition
Data processing
Validation of results
Synthetic time-
series
- SHAKEMAPS
- CONCLUSIONS
7/23/2019 Expo Jamg Axa 2015
8/19
Noise processing:Diffuse-Field Method [ / ] NS EW
ZZ
E EH V
E
Based on Greens function retrieval
so we apply usual data processingAdvanced filters (Bensen, 2007)
Processing Parameters
Baseline correction
Passband filter: 0.1 30 Hz
NO tapper, NO smoothing.
Window length: 81.92 s
11ImG NN SE U
Window length
- INTRODUCTION
Objectives
Methodology
Data acquisition
Data processing
Validation of results
Synthetic time-
series
- SHAKEMAPS
- CONCLUSIONS
ImGE W
E
Z ZE
7/23/2019 Expo Jamg Axa 2015
9/19
Applying advanced filters (Bensen, 2007)
- Spectral Whitening:spectral source deconvolutionin each time window
- 1 bit: this filter removestransitory effects (mainlybody
body waves)
- Earthquake has the signature ofbody waves
- INTRODUCTION
Objectives
Methodology
Data acquisition
Data processing
Validation of results
Synthetic time-
series
- SHAKEMAPS
- CONCLUSIONS
Typical stati
Noise processing
b l l
7/23/2019 Expo Jamg Axa 2015
10/19
* Spectral whitening
0.1 1 1
1
10
20
30
SPECTRALRATIO
FREQUENCY [Hz]
DIFFUSSE-FIELD PROCESSING
0.1 1 10
1
10
20
30
SPECTRALRATIO
FREQUENCY [Hz]
CLASSICAL PROCESSING
1
2
3
- INTRODUCTION
Objectives
Methodology
Data acquisition
Data processing
Validation of results
Synthetic time-
series
- SHAKEMAPS
- CONCLUSIONS
*
Noise processing: Variability along
h k
7/23/2019 Expo Jamg Axa 2015
11/19
The coda is a representation of a diffuse field
For each earthquake type we apply the DFM processing
Signals are cut into windows
We add the density energy for each horizontal and vertic
Finally we compute the H/V spectral ratio
Processing Parameters:
Baseline correction
Filter Passband: 0.1 30 Hz,
Window length: 40.96, 81.92 s
Diffuse Field Method (DFM)
Z
N
E
- INTRODUCTION
Objectives
Methodology
Data acquisition
Data processing
Validation of results
Synthetic time-
series
- SHAKEMAPS
- CONCLUSIONS
Earthquake processing
HVSR i M i Ci f i
7/23/2019 Expo Jamg Axa 2015
12/19
Eq type Site FORMER LAKE ZONE TRANSITION ZONE
SUBDUCTION
seismnoise
(3d scatteringmainly surface
waves)
INTERPLATE
(surface & body
waves)
LOCAL
(mainly body
waves)
- INTRODUCTION
Objectives
Methodology
Data acquisition
Data processing
Validat ion of
results
Synthetic time-
series
- SHAKEMAPS
- CONCLUSIONS
HVSR in Mexico City as function o
0.1 1 10
1
10
2030
S
PECTRALRATIO
CJ03
0.1 1 10
1
10
2030
SPECTRALRATIO
FREQUENCY [Hz]
CJ03
0.1 1 10
1
10
2030
SPECTRALRATIO
CJ03
0.1 1 10
1
10
2030
SPECTR
ALRATIO
ES57
0.1 1 10
1
10
2030
SPECTRALRATIO
ES57
0.1 1 10
1
10
2030
SPECTRALRATIO
FREQUENCY [Hz]
ES57
0.1
100
101
SPECTRALRATIO
0.
100
101
S
PECTRALRATIO
0.
100
101
SPECTRALRATIO
0.1 1 10
1
10
2030
S
PECTRALRATIO
CJ03
0.1 1 10
1
10
2030
SPECTRALRATIO
FREQUENCY [Hz]
CJ03
0.1 1 10
1
10
2030
SPECTRALRATIO
CJ03
0.1 1 10
1
10
2030
SPECTRALRATIO
ES57
0.1 1 10
1
10
2030
S
PECTRALRATIO
ES57
0.1 1 10
1
10
2030
SPECTRALRATIO
FREQUENCY [Hz]
ES57
0.
100
101
SPECTRALRATIO
0.
100
101
SPECTRALRATIO
0.
100
101
SPECTRALRATIO
R d Vib i Th
7/23/2019 Expo Jamg Axa 2015
13/19
Fourier + Amplification + Duration = Response spe- INTRODUCTION
Objectives
Methodology
Data acquisition
Data processing
Validation of results
Synthet ic t ime-
series
- SHAKEMAPS
- CONCLUSIONS
Random Vibration Theory
100
200
300CJ03
Sa[cm/s
2]
CO47 CO56 CP28
100
200
300CS78
Sa[cm/s
2]
CT64 DM12 DR16
100
200
300DX37
Sa[cm/s
2]
EO30 ES57 FJ74
1 2 3 4
100
200
300GC38
PERIOD [s]
Sa[cm/s
2]
1 2 3 4
GR27
PERIOD [s]
1 2 3 4
HJ72
PERIOD [s]
1 2 3 4
IB22
PERIOD [s]
A
E
L li ti f H/V M i
7/23/2019 Expo Jamg Axa 2015
14/19
140 H/V spectralratios to generate
shakemaps
- INTRODUCTION
Objectives
Methodology
Data acquisition
Data processing
Validation of results
Synthetic time-
series
- SHAKEMAPS
- CONCLUSIONS
Localization of H/V on Mexico
7/23/2019 Expo Jamg Axa 2015
15/19
Subductionearthquake:Mw8.1
background- INTRODUCTION
Objectives
Methodology
Data acquisition
Data processing
Validation of results
Synthetic time-
series
- SHAKEMAPS
- CONCLUSIONS
Downtown
1990 - 2013
H, To are two variables that
become with time
Downtown
Subsidence veloci
1992 - 2000
4o
HT
Vs
(s)
Co
S bd ti
th k
M 8 1
7/23/2019 Expo Jamg Axa 2015
16/19
Subductionearthquake:Mw8.1Spectral acelerations (T=2s ~ 16-20 floors)- INTRODUCTION
Objectives
Methodology
Data acquisition
Data processing
Validation of results
Synthetic time-
series
- SHAKEMAPS
- CONCLUSIONS
1990 2013
L l
E th k
M4 5
7/23/2019 Expo Jamg Axa 2015
17/19
LocalEarthquake:M4.5- INTRODUCTION
Objectives
Methodology
Data acquisition
Data processing
Validation of results
Synthetic time-
series
- SHAKEMAPS
- CONCLUSIONS
Spectral acelerations (T=0.1s ~ 1-2 floors)
Interplate
earthq ake
M 7 1
7/23/2019 Expo Jamg Axa 2015
18/19
Interplateearthquake:Mw7.1- INTRODUCTION
Objectives
Methodology
Data acquisition
Data processing
Validation of results
Synthetic time-
series
- SHAKEMAPS
- CONCLUSIONS
Spectral acelerations (T=1s ~ 8-10 floors)
7/23/2019 Expo Jamg Axa 2015
19/19
CONCLUSIONS
Significant variability of H/V is observed in terms of ehypocenter location (subduction, interplate, local) wdifferent illumination regime
The H/V from noise (HVNR) can almost represent allearthquakes (HVSR) in Mexico City
It is possible to use microtremores to get trustworth
amplification factor (aproximately an Empirical Trans
In sites with low seismicity, microtremors now are uestimate the amplification factor obtained with eart(under certains considerations)
- INTRODUCTION
Objectives
Methodology
Data acquisition
Data processing
Validation of results
Synthetic time-
series
- SHAKEMAPS
- CONCLUSIONS