Large earthquakes in stable continental regions (SCRs). The need for a new paradigm

Présentation d'Eric Calais (ENS) aux 2èmes Rencontres Scientifiques et Techniques Résif | 12-14 octobre 2015, La Grande Motte

Transcript

1. Large  earthquakes  in  stable   con3nental  regions  (SCRs)     The  need  for  a  new  paradigm   Eric  Calais,  Ecole  normale  supérieure,  Paris,  France     Contribu6ons  from  T.  Craig  (ENS),  T.  Camelbeeck  (ROB),  S.   Stein  (NWU),  M.  Liu  (U.  Missouri),  A.  Friedrich  (LMU),  L.   Fleitout  (ENS)  
2. The  paradigm   •  Stress  builds  up  on  faults  over  6me,  un6l   their  fric6onal  strength  is  exceeded  –  at   which  point  they  rupture  in  an  earthquake   with  a  stress  drop.   •  Once  unloaded  by  the  earthquake,  it  takes  a   variable  amount  of  6me  for  the  fault  to  be   reloaded  to  the  point  of  rupture.   •  But  eventually  the  fault  will  rupture  again   and  the  cycle  will  repeat,  regularly  or   irregularly.   •  Over  several  such  “cycles”,  a  balance  is   achieved  between  the  rates  at  which  strain   accrues  and  is  released  on  faults.   •  Paleoseismic  “slip  rates”  (=  strain  release   rates)  agree  with  geode6c  “slip  rates”  (=   strain  accumula6on  rates).   •  As  a  corollary,  past  earthquakes,  strain   accrual,  and  fault  segmenta6on  have  some   (complex)  predic6ve  power  for  future   events.   “Earthquakes  occur  as  a  result  of  global   plate  mo6on”  (Kanamori  and  Brodsky,   Rep.  Prog.  Phys.,  2004)     with  variable  strength,  loading  rate,  and  stress  drop    
3. Applica6on  to  a  slow  fault:   The  Wasatch  fault,  Utah   Paleoseismology  =  1.7+-­‐0.5  mm/yr  over  past  10  Ka   (Friedrich  et  al.,  2003)   GPS  =  1.6+-­‐0.4  mm/yr   (Chang  et  al.  JGR  2006)   Pechmann  and  Arabasz,  1995   Stein  et  al.,    2005   2.0 3.0 8.0 7.0 6.0 5.0 4.0 10-4 101 100 10-1 10-2 10-3 MAGNITUDE EARTHQUAKES PER YEAR Wasatch Front Paleoseismology GPS, seismicity, paleoseismology consistent with one M7 per ~1000 yr => slow, steady-state system, loaded by far-field motions
4. Large  earthquakes  in  stable   con6nental  regions  (SCR)   Widespread   Spa6al  and  temporal  paderns  unclear   Challenge  quan6fying  hazard     M>6  -­‐  NEIC  catalog,   historical  +   recorded    
5. •  16  December  1811,  M7.3,  Codonwood  Grove  fault   •  16  December  1811,  M7.0,  Codonwood?  Reelfoot?   •  23  January  1812,  M7.0,  S.  Illinois?   •  7  February  1812,  M7.5,  SW-­‐dipping  Reelfoot  thrust   •  Current  seismicity  =  ajershocks?     The New Madrid Earthquake Sequence Na6onal  Seismic  Hazard  Map,  USGS   CEUS PGA 10%/50 years, 2008 100˚W 95˚W 90˚W 85˚W 80˚W 75˚W 70˚W 25˚N 30˚N 35˚N 40˚N 45˚N 50˚N 0.05 0.05 0.05 0.1 0.1 0.1 0.2 0.3 0 500 km 0.01 0.02 0.03 0.04 0.06 0.07 0.10 0.12 0.14 0.16 0.20 0.25 0.30 0.35 R o c k P G A g CEUS PGA 10%/50 years, 2008 http://earthquake.usgs.gov/hazards/products/conterminous/2008/maps/ •  PSHA  =>  up  to  0.3  g  for  10%/50  years.   •  The  occurrence  of  similar  earthquakes  today  would   cons6tute  a  significant  hazard.   •  Reducing  vulnerability  is  costly.  
6. No  measurable  strain  in  the  NMSZ   (<  0.2  mm/yr  over  200  km)   Craig  and  Calais,  JGR  2014   0.5  mm/yr   Mean  residuals  and  confidence  limits  in  the  NMSZ  as  a  func6on  of  observa6on  period   dura6on  (11  best  sites).  Dark  grey  bars  =  1-­‐sigma  uncertainty  interval,  light  grey  bars  =   95%  uncertainty  interval.  
7. 1450±150  A.D.   900  ±  100  A.D.   300  ±  200  A.D.   1811-­‐1812   Mapping  and  da6ng  of   paléo-­‐liquefac6on  features   =>  ~500  year  average   recurrence  over  past  2,000   years.   Tudle  et  al.  (2002)   New  Madrid   paleoseismology  
8. ⇒  Strain  accumula6on  rate  in  the  NMSZ  cannot  sustain  the  ~2,000  yr  seismicity  rate   ⇒  Earthquakes  are  releasing  elas6c  energy  stored  in  the  crust  over  >  2,000  yr    –  “strain   reservoir”   GPS  (upper  bound)   From  Calais  et  al.,  Science,  2009   Adapted  from  Newman  et  al.,  Science,  1999   strain  rate  accumula6on   Evidence  for  non-­‐steady  state  behavior   500  yr  average  repeat   6me  over  ~2,000  yr  =>   2  mm/yr  for  low  M7   0.2  mm/yr  =>   minimum  repeat   6me  ~5,000  years   for  low  M7