Imaging seismic waves from space Rémi Michel 1,2, Sébastien Leprince 1, Serge Primet 3, S. Somala 1, Jean-Paul Ampuero 1, Nadia Lapusta 1, Jean-Philippe Avouac 1 1 California Institute of Technology, Pasadena, USA 2 Commissariat a l’Energie Atomique, Saclay, France 3 Institut d’Optique, Palaiseau, France

Motivation Source Model of the 1999, Mw 7.1 Duzce Earthquake. Data: SPOT offsets, inSAR, GPS, Seismograms (Konca et al, BSSA, 2010)

Motivation Source Model of the 1999, Mw 7.1 Duzce Earthquake: An initial crack-like rupture which evolve into a super-shear pulse (Konca et al, BSSA, 2010)

Simulation of a seismic rupture: horizontal strike-parallel component of the velocity field), computed for a theoretical Mw 7.0 earthquake simulated using a rate&state friction law and the Spectral Element Method (e.g., Kaneko, Lapusta and Ampuero, 2009) Motivation 50km t=15s

The determination of a kinematic source model from the inversion of static ground displacement and seismograms measured at a sparse set of seismic stations is an ill-posed problem. …but such models are needed to investigate Earthquake physics. Motivation

RequirementsComments Field of view# 550X250 kmMain Shocks, California SamplingSpatial# 100 mOk for Mw > 5.5 Poor for Mw < 5.5 ? Temporal# 1 Hz Accuracy# 1cm.s -1 Similar to accelerometers Requirements S -1 Mw7Mw7 Mw6Mw6 50km

Airborne/Space Systems InstrumentComments Drones Field Of View Airplanes FOV, temporal sampling ? Balloons Stability, FOV Earth Orbiter Low, Medium Temporal Sampling FOV # 1000km : Huge Constellations Geostationary Instrumentation?  {Radar GESS Geosynch Radar Study }  Optics Passive Imagery Limited to Clear Sky, Daylight ? A Large Geostationary Optical Telescope?

Photometry Essentials L:radiance,  : reflectance, S : atmosphere spherical reflectance Movie [  m] Wavelength (  m) Sun RadianceAtmosphere Transmittance Wavelength (  m) Ground Reflectance Wavelength (  m) Atmospheric Turbulence

SignalsTechniqueGeostationary Issues Horizontal Deformation Optical flow, Correlation  Telescope size (Resolution)  Detection  Data flow  1/100 pixel Slope (Radiance) VariationPhotoclinometry  Telescope size (Photometry)   I/I=  Detection  Stability Topographic LiDAR Differential topography  Laser Energy!  #100 photons=5cm Doppler Shift (atmosphere) Rayleigh=>Acoustic Wind Velocimeters  Small Doppler  10 cm.s -1  Complex Signal Measurable quantities with Optical Systems

Sun Poisson Incidence Atmosphere Steering (geometry) Kolmogorov Shift < few centimeters<6cm Blur < few centimeters<6cm Dancing (geometry) Scintillation (photometry) Rytov Ground Bidirectional Reflectance Distribution Function Natural, 1-100m Urban TBD (high level), lights, damages, etc. Vegetation TBD Shakened ground TBD Natural Light Fluctuations [10Hz-1Hz]

Horizontal offsets Measured by Correlation or Optical Flow 1/100 of the pixel size. SNR 1000 (reflectance 0.15, back-illuminated CCD) : 0.03 s per frame for  4m. Stability is an issue for data management and not for accuracy.

Photoclinometry Accuracy possible up to [ ] (number of detected photons [10 10 – ])  =0.15, pixel size : 100m, integration time 0.03s, most unfavorable incidence angle # 11 degrees

Simulations-Horizontal offsets Synthetic Telescope  4.0m Telescope  10.m Mw 7.1, t=15s Mw 6.0 t=4.22s Mw 6.0 t=4.22s Mw 6.0 t=4.22s Model (Mw 7.1, 30s, 2Hz, pixel:100m) and 2 critical telescope diameters (1Hz, pixel:100m)

Telescope  10.0m S -1 Accuracy 1/50 th pixel Telescope  4.0m Simulations- Horizontal offsets

Modele  7.0m  4.0m  10.0m

Simulations-Photoclinometry Synthetic Simulation (4m) rad.s -1 (Mw7)

Conclusions  Optics-Geostationary : a solution to measure seismic waves and other transients  40% of Earthquakes in Southern California, #1.2 per year with Mw 6 and above  Telescope size is critical, large field of view (up to 4 degrees?)  Less than 10% light budget, more applications possible  Payload sub-critical  Further investigations required Atmosphere, aliasing, moonlight, weather, wide field up to 4 degrees telescope design, geostationary environment, laser, etc.  Will require new analysis technique in seismology. Conclusions

 4.0m Ariane V Compatible  7.0m Challenging  10.0m Challenging  4.0m  7.0m  10.0m Horizontal Velocities [-0.4,+1.9] (m.s-1) ModelA A’ A Transect  Model of strike slip fault (Mw 7.1), Field of view 60*24km, 1Hz, ground resolution 100m Accuracy of Optical Correlation = f(  telescope, geostationary)