RIEFING ON THE M 6.0 SOUTH NAPA EARTHQUAKEThe recent surge in airborne laser scanning along active...

BRIEFING ON THE M 6.0 SOUTH NAPA EARTHQUAKE Ben Brooks

USGS Earthquake Science Center, Menlo Park

Berkeley'

San'Jose'

San'Francisco'Oakland'

Menlo'Park'

Napa'Sonoma'

Santa'Rosa'

•  3:20'AM,'Sunday'24'August''•  West'Napa'Fault'Mw'6.0,'11.3'km'depth''•  1898'M'6.3'near''Mare'Island''•  2000L09L03'M5.1''Yountville'earthquake''•  >'>'$100M'damage''•  InterLagency'response'cooperaSon/coordinaSon''

Green Valley!

source: Brad Aagaard!

Timeline Ezra%Brooks,%~3:24%am%

First&Human&

“We’re%a%li7le%ahead%of%the%curve%here%in%Northern%California”%%C%South%Park,%Ep.%1002%

EEW Performance

Shake Alert

• Hypocentral%Error:%<%~%3km%• Magnitude%Error:%<%0.2%within%11%s%of%origin%• Network%approach:%4%staPons%required%%for%alert%%• 3%of%4%triggering%staPons%would%have%been%faster%%if%upgraded%• Size%of%the%“zone%of%no%warning”%is%a%choice%based%%on%alert%thresholds.%

• VariaPon%in%data%transmission%latency%was%from%0.27s%(BK.CVS)%%to%2.62s%(NC.%NHC).%If%all%staPons%had%same%latency%as%BK.CVS%the%alert%could%have%been%out%in%half%the%Pme.%

source: Doug Given, Richard Allen!

Ground Motion

source: Brad Aagaard!

Shorter'duraSon'of'strong'shaking'at'N016'compared'to'NHC'is'consistent'with'northward'direcSvity'

DisSnct'pulses'with'separaSon'in'Sme'**independent'of'distance'from'the'rupture**'suggest'more'source'complexity'(i.e.,'two'asperiSes)'than'preliminary'rupture''models'indicate.'

Shake%Map%Fire%staPon%45%g%PGA%87%cm/s%PGV%

Main%street%Napa%64%g%PGA%34%cm/s%PGV%

Carquinez%bridge%98%g%PGA%22%cm/s%PGV%

Data%from%ShakeMap;%%ASK14=Abrahamson%et%al,%2014;%BSSA14=Boore%et%al,%2014%

1 10 100

10−4

10−3

10−2

0.1

1 PGA [g]

Distance [km]

PGA

[g]

East ComponentNorth ComponentAverage ComponentASK14 GMPEBSSA14 GMPE

Ground Motion

source: Annemarie Baltay!

'PGA'shows'small'residuals/typical'ground'moSons'at'close'distances,'implying'a'typical'stress'drop'consistent'with'the'NGAWest'data'(~'5'MPa).'N'components'larger'(true'at'all'periods),'consistent'with'the'direcSvity.'At'farther'distances,'a_enuaSon'in'the'napa/delta'region'is'stronger'than'GMPEs'predict.'Might'not'be'a'big'surprise,'since'the'NGA'west'GMPEs'are'created'with'data'that'is'sparser'in'NorCal'than'SoCal,'so'a_enuaSon'might'be'different.''

100 101 102

10−4

10−3

10−2

10−1

100

PGA [g]

Distance [km]

PGA

[g]

East ComponentNorth ComponentAverage ComponentASK14 GMPEBSSA14 GMPE

100 101 102

−3

−2

−1

0

1

2

ln residuals PGA

Distance [km]

ln re

sidu

al

BSSA residualASK residual

38.5° N

39.0° N

123.0° W 122.5° W 122.0° W 121.5° W 37.5° N

38.0° N

ln re

sidu

al

−2

−1.5

−1

−0.5

0

0.5

1

PGA%residuals%from%%ASK2014%GMPE%

Shake%Map%Fire%staPon%45%g%PGA%87%cm/s%PGV%

Main%street%Napa%64%g%PGA%34%cm/s%PGV%

Carquinez%bridge%98%g%PGA%22%cm/s%PGV%

Data%from%ShakeMap;%%ASK14=Abrahamson%et%al,%2014;%BSSA14=Boore%et%al,%2014%

1 10 100

10−4

10−3

10−2

0.1

1 PGA [g]

Distance [km]

PGA

[g]

East ComponentNorth ComponentAverage ComponentASK14 GMPEBSSA14 GMPE

Ground Motion

source: Annemarie Baltay!

'The'psa1.0'second'map'shows'strong'a_enuaSon'to'the'north,'but'also'a'possible'wave'guide'to'the'south'along'the'older'franklin/Southampton'fault'system.''In'both'the'PGA'and'psa1.0's'map,'and'the'distance'vs'PGA'plot,'you'can'see'the'few'staSons'pointed'out'on'the'shake'map'with'large'ground'moSons.'

100 101 102

10−4

10−3

10−2

10−1

100

PSA 1.0 seconds [g]

Distance [km]

PSA

[g]

East ComponentNorth ComponentAverage ComponentASK14 GMPEBSSA14 GMPE

100 101 102

−3

−2

−1

0

1

2

ln residuals PSA 1.0 seconds

Distance [km]

ln re

sidu

al

BSSA residualASK residual

38.5° N

39.0° N

123.0° W 122.5° W 122.0° W 121.5° W 37.5° N

38.0° N

ln re

sidu

al

−2

−1.5

−1

−0.5

0

0.5

1

PSA%1.0%s%residuals%from%ASK14%GMPE%

ï���Ý ï���Ý

��Ý

��Ý

0 50

km

.

���PP�������������2%6(59('���PP�������������2%6(59('Campaign sitescGPS sites

StaSc'displacements'nearly'double'ader'first'day'

P261%

source: Fred Pollitz!

GPS Geodesy

Fault Trace

InSAR: Cosmo-Skymed (x-band, ~3cm)

source: Andrea Donnellan!

•  Cosmo-Skymed: X (~3cm) •  Sentinel 1-A: C (~5cm) •  UAVSAR: L (~20 cm)

Finite Fault Models

source: Doug Dreger!

source: Fred Pollitz,!Jessica Murray!

source: Bill Barnhart!

Triggered slip on Green Valley fault?

source: Ross Stein & Jian Lin!

Surface Rupture

7'km'

source: Dan Ponti%

source: Tim Dawson% source: Tim Dawson%source: Dan Ponti%

Pre-existing Scarp

Ron%Rubin,%Tim%Dawson%(CGS)%in%the%trench%at%Alston%Park%~1%month%prior:%%“I%guess%we’ll%just%have%to%wait%for%an%earthquake%to%show%us%where%the%fault%is”%

%

%C%Napa%Watershed%Airborne%Lidar;%NCALM,%2003,%1%pt/m^2;%see%Open%Topo%C%New%acquisiPon,%DWR/CGS/USGS/PEER+GEER,%20%pr/m^2%–%ask%Hudnut%

Afterslip

8/24/2014 8/25/2014

source: Tim Dawson!

Afterslip: Saintsbury Winery

source: Dan Ponti!

8/24/2014

9/1/2014

1 2

3 4

1

2

3 4

Subsurface & Surface Slip Distribution

0

10

20

30

40

50

60

70

80

90

100

Offs

et (c

m)

source: Morelan, Trexler, Brooks, Hudnut, Lienkamper, Barnhart !

JPL/Caltech ARIA Rapid GPS

2 cm displacement

Epicenter

source: A. Donnellan!

UAVSAR

UAVSAR & Surface Rupture

Fault trace interpretation from Dan Ponti

Red dots show locations of mapped offsets identified by USGS

source: Andrea Donnellan!

400#m#

75#m#

Mobile Laser Scanning

Legacy Airborne LiDAR Datasets PLENARY PRESENTATIONS

16 | Southern California Earthquake Center

Differential L iDAR – a new tool for mapping coseismic fault-zone deformation , Edwin Nissen (CSM), J Ramon Arrowsmith (ASU), Adrian Borsa (UCSD), Craig Glennie (Terrapoint), Alejandro Hinojosa-Corona (CICESE), Tadashi Maruyama (JAMSTEC), and Michael Oskin (UC Davis) Tuesday, September 9, 2014 (14:00) The recent surge in airborne laser scanning along active faults in California and elsewhere provides high-resolution topographic baselines against which future, post-earthquake LiDAR datasets can be compared, equipping us with a powerful new tool for mapping coseismic rupture-zone deformation. Here, we showcase its rich potential using synthetic datasets derived from the southern California “B4” survey and real examples from earthquakes in Baja California and Japan to illustrate. Point-cloud differencing procedures based on the Iterative Closest Point algorithm or on cross-correlation techniques can determine the surface deformation field in 3 dimensions, complimenting conventional InSAR and pixel-tracking techniques which measure line-of-sight and horizontal displacement components only. A further advantage is that differential LiDAR retains coherence in the presence of steep displacement gradients, such as close to surface ruptures, and it is therefore well-suited for probing the slip distribution and mechanical properties of the shallow part of the fault zone. Challenges include the treatment of vegetation and the reliance on older, "legacy" LiDAR datasets – third party surveys which were not optimized for earthquake studies and for which important data acquisition and processing metrics may be unavailable. The 4 April 2010 El Mayor-Cucapah (Mexico) earthquake was the first complete rupture with both pre- and post-event LiDAR coverage, and its resultant 3-D displacement field is used to explore whether significant slip occurred on low-angle detachment faults, as has been postulated by some field geologists. The 11 April 2011 Fukushima-Hamadori (Japan) normal-faulting earthquake is the second such example, and was studied with support from the Virtual Institute for the Study of Earthquake Systems (VISES), a joint initiative between SCEC and the Earthquake Research Institute and Disaster Prevention Research Institute of Japan. Here, coherent differential LiDAR displacements within the interior part of the fault zone are used to help bridge a critical observational gap between surface faulting offsets (observed in the field) and slip occurring at depths of a few kilometers (inferred from InSAR modelling). Near-fault displacements and rotations are consistent with decreased primary fault slip at very shallow depths of a few 10s of meters, possibly accounting for the large, along-strike heterogeneity in field measurements of surface offset.

Plenary Talk Presentation Wednesday

The IDEA Model: A Practical Tool for Designing Effective Early Earthquake Warning Messages , Deanna Sellnow (U Kentucky) and Tim Sellnow (U Kentucky) Wednesday, September 10, 2014 (08:30) The right message at the right time saves lives. These words are certainly true for early earthquake warnings. In this session, we will describe how we derived the IDEA model from the principles of effective risk and crisis communication, highlight some of the research we have done to inform our work, and explain how we are now using the model to design early earthquake warning messages to be delivered in 10 seconds or less on a smart phone app. We will also summarize the studies we have employed and conclusions drawn from them since beginning our work with the USGS SAFFR group July 2013. We will then engage in a Q&A discussion regarding next steps in research and development for rolling out the early earthquake warning app., as well as questions focusing on risk and crisis communication in general.

2003%AIRBORNE:%1%pt/%m^2%MLS:%500C1000%pt/%m^2%ALSCMLS%

Structure From Motion source: Mike Oskin!

source: Kate Scharer!

RelocaPons%with%tomoDD%(Zhang'and'Thurber,%2003):%3D%velocity%model%(Hardebeck'et'al.,%2007)%%+%relaPve%relocaPon.%

Black%lines:%projecPon%of%planes%fit%to%locaPons,%OADC%algorithm%(Ouillon'et'al.,%2008).%%

Black%line:%plane%intersecPon.%%

Aftershock Locations

source: Jeanne Hardebeck!

Aftershock Relocation

source: David Shelly!

While%the%oneCweek%aoershock%counts%for%the%South%Napa%event%are%relaPvely%low%,%there%is%a%fair%amount%of%variability%even%among%just%the%N.%CA%mainshocks.%%Magnitude%distribuPons%(leo)%show%relaPvely%similar%bCvalues%among%the%N.%CA%earthquakes%,%which%are%low%compared%to%the%S.%CA%earthquakes.%

Event# M# N(M≥1.8)# N(M≥4)# Mmaxa6#

2014#South#Napa# 6# 70# 0# 3.9#

2004%Parkfield% 6% 243% 6% 5%

1984%Morgan%Hill% 6.2% 102% 0% 3.85%

1980%Livermore%Valley%

5.8% 300% 7% 5.4%

1979%Coyote%Lake% 5.8% 77% 2% 4.3%

2000%Yountville% 5% 9% 0% 2.57%

1992%Joshua%Tree% 6.1% 1880% 9% 4.47%

1986%Palm%Springs% 5.7% 1153% 6% 4.3%

Aftershock Comparison

source: Andrea Llenos, David Schwartz!

•  1947%Manix:%ML%6.5,%6km%rupture;%shallow%

•  1975%Galway%Lake;%ML%5.0%6.8%km%rupture,%5.8km%depth%•  1987%Elmore%Ranch;%Mw%6.2,%10%km%rupture,%10.6km%%

source: Andy Lutz!

Coulomb Grape Stress