Detection of seismic signals on noisy seismograms

• Autorzy: Niewiadomski J.
• Języki publikacji: EN

Abstrakty

EN

In seismologic observational practice seismic signals are very often obscured by a strong seismic noise. Sometimes one needs to have only a very limited knowledge on them, and sole information on the presence of a seismic signal is quite important. Some of such situations include, e.g., seismic source location, where only an onset time is needed, or the situation of monitoring the seismic activity. Below we will concentrate on the really difficult cases, when a visual inspection of a trained seismologist fails to find any hint of the presence of seismic signal on the noisy seismograms. The difficult task of signal detection can be done by a neural network. As the input data we will use the autoregressive parameters which model the seismogram.

Słowa kluczowe

EN

neural network; signal parameterization; autoregressive process

• Wydawca: Instytut Geofizyki PAN
• Czasopismo: Acta Geophysica Polonica
• Rocznik: 2002
• Tom: Vol. 50, nr 2
• Strony: 159--166
• Opis fizyczny: Bibliogr. 9 poz.

Twórcy

autor

Niewiadomski J.

• Institute of Geophysics, Polish Academy of Sciences, Warszawa, Poland,