Wright mdis2019 shared
- 1. What have we learned about faults from three decades of Tectonic InSAR? Tim J Wright COMET, University of Leeds, European Union @NERC_COMET @timwright_leeds
- 2. The beginnings of tectonic InSAR July 1993: Massonnet et al., first coseismic interferogram July 1991: Launch of ESA’s ERS-1 satellite 1988: Goldstein et al., 3-day Interferogram from Seasat 1986: Zebker and Goldstein, topography from airborne InSAR
- 3. The beginnings of tectonic InSAR July 1993: Massonnet et al., first coseismic interferogram July 1991: Launch of ESA’s ERS-1 satellite 1988: Goldstein et al., 3-day Interferogram from Seasat 1986: Zebker and Goldstein, topography from airborne InSAR Athens, September 1997What has Tectonic InSAR done for us?
- 4. The Earthquake Cycle 200 yrs 20 secs 100km 200 yrs 5m Note: Numbers vary for different faults COSEISMICINTERSEISMIC
- 5. The Earthquake Cycle 200 yrs 20 secs 100km 200 yrs 5m Note: Numbers vary for different faults COSEISMIC TUBI (IGS08) from Nevada Geodetic Lab 80mm Pre-quake velocity removed (Ergintav et al., 2009) POSTSEISMIC COSEISMICINTERSEISMIC POSTSEISMIC
- 6. The Earthquake Cycle 20 secs 100km Note: Numbers vary for different faults COSEISMICINTERSEISMIC POSTSEISMIC Part 1 Part 3 Part 2
- 7. The Earthquake Cycle 20 secs 100km Note: Numbers vary for different faults COSEISMICINTERSEISMIC POSTSEISMIC Part 1 Part 3 Part 2
- 8. Part 1: What have learned about earthquakes . Before InSAR: 11 earthquakes with *any* geodetic observations Now: ~150-170 earthquakes with dense displacement measurements Stein and Barrientos, JGR 1985
- 9. ~10 m M7.8 New Zealand 13 November 2016
- 11. First Sentinel-1 result posted online 4.5 hours after satellite acquisition, on 15 November
- 16. 100 km 3D Slip Model Hamling et al., Science 2017
- 17. 100 km 3D Slip Model “Rules” broken by the Kaikoura earthquake: • Earthquake broke a complex network of faults (mapped and unmapped) in several tectonic zones • Jumps of > 15 km (standard models have 5 km limit) • Subduction zone and crustal faults moving together Hamling et al., Science 2017
- 18. Part 1: What have learned about earthquakes: (a) Ruptures are more complex than we thought 2016 M7.8 Kaikoura (NZ) earthquake. (Hamling et al., 2017) 2019 M6.4/7.1 Ridgecrest (California) earthquake sequence. (Barnhart et al., 2019) 2010 M7.1 Darfield; 2011 M6.0 Christchurch (NZ) earthquakes. (Elliott et al., 2012)
- 19. Part 1: What have learned about earthquakes: (b) Surface slip is poor guide to slip at depth 10 0 cm 1995 M6.5 Dinar (Turkey). Wright et al., 1999; Fukahata and Wright, 2008 >1.2 m slip at depth <0.3 m slip at surface
- 20. 2003 M6.5 Bam (Iran). Talebian et al., 2004; Funning et al., 2006 25 cm (max) at surface; 2-3 m at depth Part 1: What have learned about earthquakes: (b) Surface slip is poor guide to slip at depth
- 21. 2003 M6.5 Bam (Iran). Talebian et al., 2004; Funning et al., 2006 25 cm (max) at surface; 2-3 m at depth Kaneko and Fialko, 2011 “Shallow Slip Deficit” Part 1: What have learned about earthquakes: (b) Surface slip is poor guide to slip at depth
- 22. Part 1: What have learned about earthquakes: (c) Earthquakes can be triggered dynamically 1997 M7.1 Pakistan Earthquake Doublet (separated by 19 s); Nissen et al., 2016 2000 M5.8 triggered by M6.6, 100 km away (not detected by seismology); Pagli et al., 2003
- 23. Part 1: What have learned about earthquakes: (d) Earthquakes can be structurally controlled 2001 M7.8 Kokoxili (Tibet); Lasserre et al., 2005
- 24. 2015 M7.8 Nepal; Qiu et al., 20162008/2009 M6.3/6.3 Qaidam; Elliott et al., 2011 2009 M6.3 (shallow) 2008 M6.3 (deep) Part 1: What have learned about earthquakes: (d) Earthquakes can be structurally controlled
- 25. The Earthquake Cycle 20 secs 100km Note: Numbers vary for different faults COSEISMICINTERSEISMIC POSTSEISMIC Part 1 Part 3 Part 2
- 26. Reproduced with permission from UN Part 2: Seismic Hazard and Interseismic Strain Accumulation What’s the seismic hazard in Mongolia?
- 27. Walker et al., Geology 2007; Reproduced with permission What’s the seismic hazard in Mongolia? Records of earthquakes are too short for the conventional approach in many areas of the continents
- 28. If we can measure strain, it should be causally linked to seismic hazard West Coast USA Strain-rate ‘Quake rate Elliott, Walters & Wright, 2016 Strain-rate ‘Quake rate
- 29. First measurements of interseismic strain with InSAR North Anatolian Fault; Wright, Parsons and Fielding, GRL 2001 East California Shear Zone; Peltzer et al., Geology 2001
- 30. • 1 PhD 1997-2000 (Wright): 2 ERS Frames / 100 km of fault / ~20,000 km2 Progress…and the future
- 31. • 1 PhD 1997-2000 (Wright): 2 ERS Frames / 100 km of fault / ~20,000 km2 • 1 PhD 2009-2012 (Walters): 5 Envisat Tracks / 200 km of fault / ~250,000 km2 • 1 PhD 2012-2016 (Hussain): 23 Envisat Tracks / entire fault / ~750,000 km2 Hussain et al., 2018 Progress…and the future
- 32. Velocity Field for Turkey from Sentinel-1 (Weiss et al., in prep)
- 33. Velocity Field for Turkey from Sentinel-1 (Weiss et al., in prep)
- 34. Velocity Field for Turkey from Sentinel-1 (Weiss et al., in prep)
- 35. Velocity Field for Turkey from Sentinel-1 (Weiss et al., in prep)
- 36. 4 years of COMET-LiCS Sentinel-1 processing for Anatolia ~500 ifgs/frame x 40 frames = ~20,000 interferograms ~1200 frames for the AHB = ~450,000* interferograms to process *east of Turkey only 12-day revisit Progress…
- 37. Download data from http://comet.nerc.ac.uk/COMET-LiCS-portal • Currently ~150,000 interferograms (and counting) available over ~1200 frames for Alpine Himalayan Belt, plus recent data from global volcanoes • Time series and velocity fields being produced – their accuracies will improve rapidly as observation period increases.
- 38. Focused strain throughout earthquake cycle Hussain et al., Nat. Comm. 2018
- 39. Focused strain throughout earthquake cycle Hussain et al., Nat. Comm. 2018 Time since earthquake (years) log10(strainrate) Derived from InSAR & GNSS velocity field Derived from pre-1999 GNSS (McCluskey et al., 2000) Derived from post-1999 GNSS (Ergintav et al., 2009) 0.5 Strainrate(mstrain/year) 5 50 Result: Strain rate along the entire North Anatolian Fault is independent of time since the last earthquake, except in decade following a major earthquake. Implications: (1) Short term strain → long-term hazard (2) Relaxation time of lower crust > inter-event time (viscosities > 1020 Pa s).
- 40. Xianshuihe Karakoram ATF Hunter Mountain Haiyuan Zagros–Makran Khazar NAF DSF Alps Alpine SISZ Fairweather & Denali SAF Imperial Saging Sumatra Tierra del Fuego Husavik Flatey EVZ Yammouneh Carmel PMF Riganpei Co Manyi Is focused strain around faults ubiquitous? (187 examples in Wright et al., 2013 review)
- 41. Extension in South Central Tibet (Wang et al., 2019)
- 42. Extension in South Central Tibet (Wang et al., 2019)
- 43. Extension in South Central Tibet (Wang et al., 2019)
- 44. Part 2: Seismic Hazard and Interseismic Strain Accumulation (Key Points) • Measuring Interseismic strain accumulation is challenging with InSAR, but can be done with large data stacks • Interseismic strain is usually focussed around major faults. • Strain rate is approx. constant throughout the cycle on the North Anatolian Fault. May not be true elsewhere? • Uncertainties in strain estimates will reduce as data improves – this will lead to improved seismic hazard models.
- 45. The Earthquake Cycle 20 secs 100km Note: Numbers vary for different faults COSEISMICINTERSEISMIC POSTSEISMIC Part 1 Part 3 Part 2
- 46. Part 3: Postseismic deformation and aseismic slip transients Manyi (Tibet) postseismic from Ryder et al, GJI 2007 Interseismic: 3±2 mm/yr (Bell et al., GRL 2011)
- 47. From: Stephane Baize blog Elliottetal.,EOS2015 31 Aug – 12 Sept 2014 http://stephaneonblogger.blogspot.co.uk/2015/11/those-faults-that-move-without-quaking.html
- 48. Elliottetal.,EOS2015 Floyd et al., GRL 2016
- 49. Maximum postseismic velocities follow power law (𝑣 ∝ 𝑡−1 ) Ingleby and Wright, GRL 2017
- 50. Aseismic Creep controlled by lithology Taiwan; Thomas et al., Tectonophysics 2014 Turkey; Cetin et al., G-cubed 2014
- 51. Aseismic Creep Events North Anatolian Fault, Turkey Rousset et al., GRL 2016
- 52. Aseismic Creep Events Khoshmanesh and Shirzaei et al., Nat. Geosci., 2018
- 53. Aseismic Creep Events Slow earthquake in Chaman Fault Zone (Pakistan); Furuya and Satyabala, 2008 Aseismic movement on fold and thrust belt during 1998 Fandoqa earthquake Fielding et al., Geology 2004
- 54. Part 3: Postseismic deformation and aseismic slip transients (key points) • Details of postseismic transient behaviour can be spatially complex • But overall, postseismic deformation may be remarkably simple: 𝑣 ∝ 𝑡−1 • InSAR has helped identify a range of aseismic creep behaviour including time-varying shallow creep, slow earthquakes and triggered slip.
- 55. EE10 Candidate (2028??) Harmony (formerly STEREOID) Radar remote sensing in the next decade Freeand OpenData! EE10 Candidate (2028??) Hydroterra (formerly G-CLASS) InSAR (nearly) everywhere (nearly) all the time Elliott, Walters & Wright, 2016
- 56. Big Unsolved Questions in (Continental) Tectonics How can we link short-term earthquake-cycle deformation with long-term tectonics? How can we incorporate strain estimates in seismic hazard models? What causes time-varying deformation? What are the dynamics of continental tectonics? How does fault friction vary in space and time, and how is it controlled by geology? What is the role of fluids in fault zones? What do the deep roots of faults look like and do they control what we see at the surface? How does magmatism influence tectonics? …?
- 57. Take Home Messages Tectonic InSAR is living up to the potential identified by early pioneers. Earthquakes continue to surprise us and we continue to learn from them Measuring slow, long-wavelength deformation is more challenging but data from long-duration missions like Sentinel-1 will lead to exciting new discoveries Trans-national partnerships, collaborations, discussions essential for success @NERC_COMET @timwright_leeds
- 58. With thanks to colleagues in NERC COMET