16
2011/12/14 CCC @ CWB 震震震震震 ( 震震震震震Pattern Informatics) 震震震震震震震震震 震震震 Chien-chih Chen 震震震震震震震震震震震 Inst Geophysics, Natl Central Univ

震模資訊學 ( 地震資訊學, Pattern Informatics) 與地震預測之可行性

  • Upload
    tuan

  • View
    61

  • Download
    0

Embed Size (px)

DESCRIPTION

震模資訊學 ( 地震資訊學, Pattern Informatics) 與地震預測之可行性. 陳建志 Chien-chih Chen 中央大學地球物理研究所 Inst Geophysics, Natl Central Univ. Computing a PI (Pattern Informatics) Earthquake Forecast. Schematic: Spatial Cross Section Of Intensity Map along a Linear Track. Intensity = I. Intensity Change =  I. - PowerPoint PPT Presentation

Citation preview

PowerPoint Presentation2011/12/14 CCC @ CWB
P = Probability = {< I >}2
2011/12/14 CCC @ CWB
Early Middle Late
Theory (Self-Organizing Spinodal Model) predicts 3 distinct temporal regimes for Gutenberg-Richter statistical laws prior to a large earthquake (“Early, Middle, Late”; Rundle et al., PAGEOPH, 2000)
Data from 1999 Chi-Chi, Taiwan earthquake (C.C. Chen, GJI, 2003)
2011/12/14 CCC @ CWB
2011/12/14 CCC @ CWB
Data assimilation in the Pattern Informatics of Seismicity can proceed by analysis cycles, which possibly reflect “earthquake cycles.” In each analysis cycle, observations of the past state in the data space are combined with the results from a PI model to produce a PI hotspot map, which is considered as 'the best' estimate of the current state of the earthquake fault system. This is called the analysis step. Essentially, the analysis step tries to balance the uncertainty in the seismicity data and in the PI parameters. ‘The best' PI model/hotspot map is then advanced in time and its result may become the forecast for the next analysis/earthquake cycle.
2011/12/14 CCC @ CWB
2011/12/14 CCC @ CWB
2011/12/14 CCC @ CWB
What we have observed is some kind of spatial cooperation between main shock and aftershocks existing actually before the occurrence of main shock, since we have derived such correlation (PI) patterns from the fluctuation in seismicity before great earthquakes.
2011/12/14 CCC @ CWB