A neuro-fuzzy approach to the reliable recognition of electric earthquake precursors

Abstract

Abstract. Electric Earthquake Precursor (EEP) recognition is essentially a problem of weak signal detection. An EEP signal, according to the theory of propagating cracks, is usually a very weak electric potential anomaly appearing on the Earth's electric field prior to an earthquake, often unobservable within the electric background, which is significantly stronger and embedded in noise. Furthermore, EEP signals vary in terms of duration and size making reliable recognition even more difficult. An average model for EEP signals has been identified based on a time function describing the evolution of the number of propagating cracks. This paper describes the use of neuro-fuzzy networks (Neural Networks with intrinsic fuzzy logic abilities) for the reliable recognition of EEP signals within the electric field. Pattern recognition is performed by the neural network to identify the average EEP model from within the electric field. Use of the neuro-fuzzy model enables classification of signals that are not exactly the same, but do approximate the average EEP model, as EEPs. On the other hand, signals that look like EEPs but do not approximate enough the average model are suppressed, preventing false classification. The effectiveness of the proposed network is demonstrated using electrotelluric data recorded in NW Greece.

On the electric field transient anomaly observed at the time of the Kythira <i>M</i>=6.9 earthquake on January 2006

“…2 and 3 indicates the time of occurrence of the main seismic event. The observed characteristics of the recorded possible electric earthquake precursor support previous results (Konstantaras et al, 2002(Konstantaras et al, , 2004(Konstantaras et al, , 2006aVarotsos, 2005) regarding the features of EEP signals (Lighthill, 1996;Vallianatos and Tzanis, 1998;Tzanis and Vallianatos, 2001;Varotsos, 2005). It is well accepted that EEP signals are transient electric potential anomalies external to the natural electromagnetic field of the Earth (Hayakawa and Molchanov, 2002;Vallianatos and Tzanis, 1998;Konstantaras et al, 2002).…”

Stochastic analysis of geo-electric field singularities as seismically correlated candidates

“…To train and evaluate the reaction of the neuro-fuzzy model, 4096 data samples of electric field recordings have been selected Konstantaras et al (2004), corresponding approximately to the time-period starting at 9 p.m. on the 29 December 2005 and ending at 1 p.m. on 11 January 2006, which include the possible electric earthquake precursor. Although the initial sampling frequency of the recorded data is f s =5 Hz, the overall data set has been decimated by a factor of 1280 as it is very costly in terms of processing time Kosko (1991) to train a neural network with such a heavy workload.…”

“…Previous studies Konstantaras et al, 2007Konstantaras et al, , 2002Konstantaras et al, , 2004Konstantaras et al, , 2006a provide information regarding the nature of EEP signals. These results indicate that EEP signals are transient electric potential anomalies external to the natural (of ionospheric origin) electromagnetic field of the Earth.…”

Hybrid Adaptive Filter development for the minimisation of transient fluctuations superimposed on electrotelluric field recordings mainly by magnetic storms

“…In layer 2, every node i in this layer is an adaptive node with a node function: node i in this layer is an adaptive node using an output membership functions to compute the weighed output of the equivalent rule, according to the following node function: O 3,i =w i f i where f i =p i +q i +m i +n i +r i , and {p i , q i , m i , n i , r i } are the consequent parameters (Konstantaras et al, 2004) of the network that specify the rules of the fuzzy inference system. In layer 5, the single node in this layer is a fixed node, which converts the weighted fuzzy outputs of all rules in the system into a single crisp output, as described by the following node function:O 4,1 = i w i f i .…”