Premium
Some observations on the probabilistic interpretation of short‐term earthquake precursors
Author(s) -
Grandori G.,
Guagenti E.,
Perotti F.
Publication year - 1984
Publication title -
earthquake engineering and structural dynamics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.218
H-Index - 127
eISSN - 1096-9845
pISSN - 0098-8847
DOI - 10.1002/eqe.4290120604
Subject(s) - probabilistic logic , term (time) , interpretation (philosophy) , alarm , set (abstract data type) , false alarm , conditional probability , earthquake prediction , principal (computer security) , computer science , mathematics , geology , statistics , seismology , engineering , physics , computer security , quantum mechanics , programming language , aerospace engineering
This paper analyses the uncertainties in probabilistic interpretation of short‐term earthquake precursors, even when the statistical information commonly indicated in the literature as sufficient to define the characteristics of these precursors is assumed to be known. The wide margins for uncertainty in the interpretation of such data are pointed out. One of the principal causes of uncertainty, as an example, lies in the physical origin of false alarms. Depending on this physical origin, the conditional probability of an earthquake, other conditions being equal, may vary in certain cases from values around 0–1 to as much as 0–7 or even higher. The utility of an alarm system, defined as the ratio between the mean number of favourable cases (successful alarms) and the mean number of unfavourable cases (false and missed alarms) is analysed. It is shown that an alarm system based on monitoring n ≥ 2 precursors, if the alarm is set off only after a combination of all precursors, may be considerably less useful than a system based on monitoring only one of the considered precursors.