Kenneth W. Hudnut U. S. Geological Survey Pasadena, California

Berkeley Seismo Lab SeminarU. C. Berkeley, CaliforniaMarch 11, 2003Measuring Fault Slip -Why and How?

Measuring fault slip - why and how?Why?What can details of slip variation tell us about earthquake source physics? (You may suggest or prefer other reasons )Variance and other statistical properties of slip distributions - physical relation to fundamental quantities (such as maximum slip, seismic and aseismic energy, stress drop, Dc)Large slip variations, if real, could constrain rupture process models and explain high-frequency seismic energy radiation

How? (first, my bias is that great earthquakes do differ from small ones)Attempt to observe slip pulses or other crack propagation phenomenaReal-time GPS Fault Slip Sensor - to complement inertial sensors Variation in slip along-strike and with depth - quantify the errorsAirborne Laser Swath Mapping - to observe near-fault deformation

earthquake scalingAre stress drops constant?e.g., Brune (1970, 71)What about L-model scaling for large events?e.g., Scholz (1982, 1994) and Romanowicz (1992)Bi-linear scaling modelHanks and Bakun (2002)Great continental strike-slip earthquakes are of special interest - their slip (and stress drops) may become larger with longer ruptures (up to 10 x W):1857 and 1906 SAF1905 Bulnay and 1957 Gobi-Altay1920 Haiyun1939 ErzincanKunlun & Denali

some questions How may friction drop co-seismically?Opening? What mode of crack propagation?Key observation - accurate 3D particle motions close to the fault --- inertial sensors are unable to discriminate between tilt & acceleration; GPS fault slip sensors can help to observe this for future large events

What impedes and stops rupture?Roughness and variation in slip - heterogeneity of stress, as well as fault interaction and dynamic effectsKey observation - objective systematic measurements of slip along-strike at surface for bedrock-on-bedrock fault with large amount of slip; ALSM for Hector Mine

GPS is now vital to earthquake monitoring(array technology and GPS satellites weredeveloped in Southern California)Measures buildup of strain on faults due to accumulating tectonic motion

Used to detect damage to large engineered structures such as dams, and effects of ground tilt and subsidence on water systems

Can be used in real-time to detect fault slip at surface, for early warning system

Science ReviewList of publications - updates on the main project web page at http://www.scign.org/45 papers from 1996 through April 2002 (includes several on Hector Mine co-seismic) - many important contributions that would have been impossible without SCIGN. BSSA Hector Mine special issue (May, 2002) contains 16 papers (out of a total of 36) that made direct or indirect use of SCIGN data. Nearly all 36 then also cited these 16.Fall AGU 2002 abstracts contain reports on several new projects using SCIGN data in innovative ways, as well as reports of the new results from the ongoing and coordinated SCIGN Analysis Committee effort, chaired by Nancy King. The use of continuous GPS data for various geophysical and other uses is increasing worldwide, and SCIGN has provided data used in a far wider range of studies than was ever imagined (e.g., Varner and Cannon, 2002). The open data policy has been a key to success.SCIGN has become an integral part of what the broader SCEC community does in southern California. These results influence peoples thinking well outside of our regional scope, just as we all learn greatly from studies of other regions. Others see the value inherent in having a state-of-the-art array like SCIGN available (hence the PBO part of the EarthScope initiative). New data products are making SCIGN data more easily accessible to a larger and broader base of earthquake and other earth scientists.

1st Year Combined timeseries (1996-2002)

3rd Year Real-time earth-quake response

5th Year Resolve rates onprimary LA basinfaults (and others)SCIGN Data Products

Hector Mine (Mw7.1)Photo by Paul Kip Otis-Diehl,USMC, 29 Palms Earthquake response rapidly assess the earthquake source

Pacoima Dam GPS structural health monitoring system (with LA County since Sept. 1995) Damage estimation rapidly assess damage, casualties, and lossesShakeMap/HAZUS

GPS can help withEarthquake response informationIdentify fault source, extent and amount of slipModel finite fault sourceMeasure and model deformation fieldProvide all of the above to the public, to emergency responders, and to other scientists

Damage estimationProvide data for use in HAZUS, etc.Support of remote sensing and positioning for accurate and timely collection, reporting and control of other data that require accurate position and/or timingMonitor large engineered structures and lifeline systems

Basic positioning infrastructure

Early WarningThe speed of light >> the speed of soundSeismicandGPS StationsUtilitiesEmergency ResponseTransport-ation

Mitigative ActionsStop trainsStop nuclear reactionsStop computersSecure hazardous materialsStop elevatorsIs 30 seconds of warning enough?

GPS Fault Slip SensorK. Hudnut, G. Anderson, A. Aspiotes,N. King, R. Moffitt, & K. Stark(all at USGS-SCIGN, Pasadena CA)APEC symposium Proc. Paper, Fall AGU poster,and paper in preparation forBulletin of the Seismological Society of America

http://pasadena.wr.usgs.gov/office/hudnut/slipsensor/

what we cannot dobecause the physical process is too chaoticweather turbulenceearthquakes - frictionRecognize a precursor for a particular event

Differentiate readily between the beginning of a M3 and a M7GPS slip sensorcan help with this!GPS Fault Slip Sensor is proposed to augment Earthquake Early Warning SystemsM3M7

GPS fault slip sensor

Lone Juniper Ranch and Frazier Park High SchoolFirst prototype GPS fault slip sensorSpans the San Andreas fault near Gorman, California

Real-Time GPS Network - Enhancing SCIGNOn 15 November 2002, first-ever GPS fault slip sensor deployed across San Andreas fault at Gorman, Calif.

Upgrade SCIGN telemetryDSL, frame relayRadio repeaters, WiGate and dedicated linksData buffering

Augment SCIGN real-time acquisition and processing systemImplemented sub-daily processing (4 hr) for ~100 SCIGN stations (down from 24 hr)Implementing multiple real-time streaming GPS processors (commercial software)

The Plate Boundary ObservatoryBARD & SCIGN are prototype deployments for PBO

PBO will extend the GPS & strain arrays throughout the Western U. S. A. and Alaska

With Canada and Mexico, we hope to cover the North American Pacific plate boundaryHas been funded !($30M for FY03)

before moving on State-of-the-art GPS networks and technology development have become vital infrastructure for earthquake research and response:Static deformation field data; source models & rapid tilt and strain mappingMonitoring of large engineered structures (e.g., dams, buildings)Understanding earthquake source physics - e.g., slip pulses (near-field accurate particle motions and static displacement field mapping)

New enhancements to telemetry support wider range of applicationsReal-time GPS sub-networks of SCIGNPrecise RTK positioning for surveying, AVL and GIS applicationsInSAR, Airborne Laser Swath Mapping (ALSM) and digital photographySCIGN provides ground control for airborne imaging and surveysMapping and imaging for rapid assessment of damage to buildings, lifeline infrastructure, etc. (The National Map; Homeland Security)

Collaboration between Scientific, Surveying, GIS, Engineering and Transportation communities has mutual benefits - this is expected to help with funding our projects in the future

ATC-20 building safety inspections(red, yellow and green tagging) Infrastructure for positioning supports data collection and mapping during disaster recovery effortsimagerybefore &afterCourtesy of City of GlendaleCourtesy of Ron Eguchi

High resolution topography along surface rupture of the October 16, 1999 Hector Mine, California Earthquake (Mw7.1) from Airborne Laser Swath Mapping Hudnut, K. W. (USGS), A. Borsa (UCSD),C. Glennie (Aerotec, LLC) and J.-B. Minster (UCSD)Bulletin of the Seismological Society of AmericaSpecial Issue on the Hector Mine earthquake (2002)

http://pasadena.wr.usgs.gov/office/hudnut/docs/

ALSM, InSAR and TM imagingand mapping before & afterLaser scan of landsurfaces and urbaninfrastructure buildings & lifelines- damage assessmentPre- and post-disaster images can be differenced to measure damage and track recoveryLANDSAT 7 & SRTM - damage in Bhuj, IndiaALSM 9/11/01 NYC

Active Faults in Southern California

Surface RupturePreviously mapped, but un-namedLavic Lake fault in recognition of breaks through dry lake bedUp to 5.7 meters of right-lateral motion48 km overall length of surface ruptureOnly ruptured once prior through ~50 ka alluvium

from Treiman et al. (BSSA 2002)

ALSM (Scanning LIDAR imaging)Slow, precise helicopter flight line data acquisition at 200-300 m AGL.6888 pps near infra-red (1064 nm) laser.

Scan Width: +/- 20 degrees. Nominally, 180 meters full-width.200 pulses across swath, ~ 80cm spacing.Footprint Diameter: Nominally 40cm. Half-meter posting, 15cm horizontal one-sigma absolute accuracy specified.

Integrated GPS & INS navigation and attitude determination.

Pitch Mirror Correction: maximum +3.5/-6.5 degrees (+ forward bias).

Geolocation Vectors and Error SourcesVector from CMearth to GPS phase centerMagnitude & directional errors both arestochastic, time and location variant. Vector from GPS phase center to laserMagnitude error is constant if no airframe flexing. Directional error due to constant and time-varying biases in INS.Vector from laser to ground footprintMagnitude error due to timing, instrument and atmospheric delays. Directional errorfrom constant mirror mounting offsets and time-varying biases in reporting of scanangles (both pitch and roll).Note: additional errors due to imperfectsynchronization of GPS, INS, mirror scanand laser firing times must be modeled and removed as well.

Flight Plan Two overlapping swaths 200-500m mapped width 70 km long GPS network 1 Hz Temporary GPS stations Cross-swath spurs Roll/Pitch/Yaw calibration maneuvers over dry lake Flights over well-mapped Hector Mine

GPS Sites at 1Hz During ALSM Mission1Hz2HzBMHLOPBLOPCPRDMTSIBETROYAGMTHCMNNBPSOPCLOPCXOPRDCalibration maneuvers at Hector Mine and Lavic Lake

New methods to explore, new synergies between data types (e.g., GPS & ALSM)Combinations of seismic, geologic, and geodetic data in new waysSource modellingHazard modellingCross-overs between fieldsGeology and geodesy with InSAR and ALSMSeismology and geodesy with high-rate GPS 1999Hector Mineearthquakesurfacerupture

ALSM-Derived Contours of Bullion Mountain Segment(blue lines are1-foot contours)

To assess geodetic capability of repeat-pass ALSM:calibration requirementsGeometry: mount angles, scan offsets, GPS-Laser vector, GPS antenna phase centerDelays: electronic, optical, atmosphericReference point on laser platformTiming of various components: e.g. INS vs. GPS vs. mirror attitude sensorsStabilization platforms (delays, accuracy)Detectors (thresholds, amplitude-range walk)

Lavic LakeRoll & PitchManeuversExplodedordnance(crater)pitch maneuvers

15 cm vert.10 cm vert.

Geological quantification and questionsTectonic interpretation of strain release in great earthquakes from their surface ruptureBasic documentation of surface rupturee.g., Kurushin et al. (1997) study of 1957 Gobi-Altay eq.

How does slip vary along-strike?e.g., need to assess variance and error in slip rate estimates from paleoseismic methodse.g., Barka et al. (2002) and Rockwell et al. (2002) extensive studies of 1999 eq.s in Turkey and similar studies of Hector Mine earthquake (BSSA 2002)is high-frequency energy radiated from fault?

Does slip vary from one earthquake to the next?can detailed topographic mapping of geomorphic features along the fault be modeled by repeats of exactly the same slip in successive earthquakes, or must slip vary in order to explain the topography?slip variation models for earthquake recurrence strongly influence seismic hazard analyses assumptions made in these analyses necessarily simplify faulting processes, with societal repurcussions

Southwest Corner of Lavic Lakedry lake bed for calibrations

Lavic Lake compressional step

New methods and data integrationprecise topographic mapping of surface ruptures and active fault scarpsrepresentation of actual fault ruptures recorded and preserved in unprecedented detail for use by future earthquake researchersAirborne platform navigationmust be highly precise andrequires high-rate GPS data

New methods and data integrationPrecise topographic mapping of surface ruptures and active fault scarpsslip models for prehistoric eventsrapid assessment of surface slip and damage patterns after large eventsRequires precise integration of GPS & INS for flight navigation1957Gobi-Altaiearthquakesurfacerupture

Estimating slip on max. slip segment ofthe fault

Estimating slip on max. slip segment ofthe fault

Is it a multiple-event offset?

Conclusions - and some open questionsAirborne LIDAR imaging (ALSM) offers remarkable promise for geomorphology and fault zone studies, even over inaccessible or vegetated areasCommercial operations are reliable and affordable on well-specified targets with carefully designed deployments (same as photogrammetry)CAL-VAL maneuvers are essential for geodetic-quality mapping of geomorphic featuresTurning ALSM into a geodetic-quality tool requires careful calibration and considerable analysis

Slip estimation:has been initially developed demonstratednew and improved methods are being developedsystematic measurement along the surface rupture will be done and thencompared with geologic estimates, InSAR and other methodsquantitative assessment of slip variation along-strikedynamic faulting models is high-frequency energy radiating from the fault?

Quantitative geomorphologyModel tectonic landform evolutionDid topographic features form as a result of exactly repeated slip distributions?Can topography be explained by only certain combinations of slip in past events?

Demonstrated GPS application to augment earthquake early warning systems

Understanding earthquake source physics - e.g., slip pulses & rupture mechanicsNear-field accurate 3D particle motions and static displacement field estimation