HR: 13:55h
AN: NG32A-02 INVITED [PDF]
TI: Hazard Forecasting by MRI: A Prediction Algorithm of the First Kind
AU: * Lomnitz, C
EM: cinna@prodigy.net.mx
AF: National University of Mexico, Institute of Geophysics UNAM
Ciudad Universitaria, Mexico, 04510
Mexico
AB:
Seismic gaps do not tell us when and where the next earthquake is due. We present new results on limited earthquake hazard
prediction at plate boundaries. Our algorithm quantifies earthquake hazard in seismic gaps. The prediction window found for
M7 is on the order of 50 km by 20 years (Lomnitz, 1996a).
The earth is unstable with respect to small perturbations of the initial conditions. A prediction of the first kind is an
estimate of the time evolution of a complex system with fixed boundary conditions in response to changes in the initial
state, for example, weather prediction (Edward Lorenz, 1975; Hasselmann, 2002). We use the catalog of large world
earthquakes as a proxy for the initial conditions. The MRI algorithm simulates the response of the system to updating the
catalog. After a local stress transient dP the entropy decays as (grad dP)2 due to transient flows directed toward the
epicenter. Healing is the thermodynamic process which resets the state of stress. It proceeds as a power law from the
rupture boundary inwards, as in a wound. The half-life of a rupture is defined as the healing time which shrinks the size of
a scar by half. Healed segments of plate boundary can rupture again.
From observations in Chile, Mexico and Japan we find that the half-life of a seismic rupture is about 20 years, in agreement
with seismic gap observations. The moment ratio MR is defined as the contrast between the cumulative regional moment
release and the local moment deficiency at time t along the plate boundary. The procedure is called MRI.
The findings: (1) MRI works; (2) major earthquakes match prominent peaks in the MRI graph; (3) important events (Central
Chile 1985; Mexico 1985; Kobe 1995) match MRI peaks which began to emerge 10 to 20 years before the earthquake; (4) The
emergence of peaks in MRI depends on earlier ruptures that occurred, not adjacent to but at 10 to 20 fault lengths from the
epicentral region, in agreement with triggering effects.
The hazard enhancement in space is shaped like a Mexican hat function. The central part is the aftershock region, separated
by a ring of quiescence from an outer region of increased rupture probability(Lomnitz, 1996b). In conclusion, we may speak
of seismic weather prediction using MRI.
Hasselmann, K. (2002). Is climate predictable? In The Science of Disasters, A. Bunde, J. Kropp and H.J. Schellnhuber, eds.
(Springer, Berlin, 140-169).
Lomnitz, C. (1996a). Predicting earthquakes with the MRI algorithm, Seismol. Res. Letters, 67, 40-46.
Lomnitz, C. (1996b). Search of a worldwide catalog for earthquakes triggered at intermediate distances, Bull. Seismol. Soc.
Am., 86, 293-298.
Lorenz, E. (1975). Climate predictability: The physical basis of climate and climate modeling. World Meteorol. Org.,
Geneva, Report 16, 132.
DE: 3220 Nonlinear dynamics
DE: 7209 Earthquake dynamics and mechanics
DE: 7223 Seismic hazard assessment and prediction
SC: Nonlinear Geophysics [NG]
MN: 2003 Fall Meeting