EQ Prediction

Story of two cities:

Haicheng, 1975: world's first successful prediction of an eq

saved thousands of lives

Tangshan, 1976: one of worst natural disasters in recorded history

—250,000 killed

Requirements for EQ Prediction

1. Time period

2. Location of epicenter

3. Magnitude estimate

4. Probability

Long-term EQ forecasting

Goal—identify areas at high risk

Strengthen critical structures

Begin eq preparedness programs

Intensify short-term EQ prediction research

Methods

Historical seismicity:

Cycles of large earthquakes

Paleoseismology:

geological, archeological methods: dating of prehistoric eqs

Seismic gap theory

rupture areas of large eqs define 'broken' segments of plate boundaries

seismic gaps = unbroken parts of plate boundary: areas of highest seismic risk

long-term strain accumulation:

measure rate of strain accumulation along fault, estimate time to recurrence of large event

eq statistics

rely on pattern of eq occurrence: 'magnitude-frequency relation'

decrease of one unit of magnitude à factor of ten increase in number

use statistics of small and moderate sized events to predict recurrence rate of larger events

Short-term EQ prediction

empirical approach

search for anomalies—deviations from steady-state

Observations

(1) seismicity changes: 'Quiescence' around future eq location surrounded by areas of higher activity ('Mogi doughnut'); precursory swarms of activity (foreshocks)

(2) changes in velocity of seismic waves

(3) changes in ground deformation (tilt, strain, ground level)

(4) changes in ground water level and chemistry

(5) changes in electrical, magnetic fields, radio signals

(6) animal behavior

Model to explain eq precursors: Dilatancy-diffusion model

from laboratory experiments with rock specimens

Stage I. Gradual buildup of strain

Stage II. Development of open cracks ('dilatancy')

Stage III. Influx of water into dilatant zone ('diffusion')

Stage IV. Sudden rupture (Earthquake!)

Stage V. Recovery, start cycle over again