Automatic fuzzy-logic recognition of anomalous activity on long geophysical records: Application to electric signals associated with the volcanic activity of La Fournaise volcano (Reunion Island)статья
Статья опубликована в высокорейтинговом журнале
Информация о цитировании статьи получена из
Web of Science,
Scopus
Статья опубликована в журнале из списка Web of Science и/или Scopus
Дата последнего поиска статьи во внешних источниках: 22 декабря 2014 г.
Аннотация:For the mitigation of natural hazards, geophysicists install more and more sensors in the field to enlarge the monitoring networks. In volcanic and seismogenic areas unambiguous results can be obtained only if many parameters are continuously recorded and processed over very long time series (several years) overlapping an eruption or an earthquake. On the other hand recent research has led to a sharp increase in the sampling rate of the data acquisition systems. Therefore, manual data processing with a first-step visual expertise becomes more and more difficult, time consuming and at the same time less objective. This paper introduces an alternative to manual signal recognition. It includes the development of a specific anomaly recognition algorithm called Difference Recognition Algorithm for Signals (DRAS) and its application to self-potential (SP) records obtained on the 10 channels of the electric stations located on La Fournaise volcano before, during and after the eruption of March 9, 1998. The algorithm starts with the construction of "rectification functionals" (examples are energy, length, zero crossing rate) from the data over a running characteristic time-window. Application of fuzzy set measures over the calculated functionals allows DRAS to identify, in particular, well spaced and time-organised SP oscillations observed on the volcano up to 2 weeks before the March 9, 1998 eruption. Based on the results obtained one can conclude that electric signals in the ULF band (frequency < 10 Hz) can be generated by the volcanic activity. The morphology and the distribution with time before the eruption can give some information on the location of the future vents. (c) 2005 Elsevier B.V. All rights reserved.