Information on gravitational mass movements obtained from the spectrograms of their seismic signals generated

Abstract

<p>An understanding of the characteristics of a mass movement descending a slope enables us to obtain a better control through models and also to reduce its associated risks. The seismic signals generated by the mass movement are mainly caused by friction of the moving mass on the ground. Most of the studies of the seismic signals use the spectrograms as a complementary information of the signals. Our study seeks to expand the current applications of the spectrograms using the information contained in them. A spectrogram represents the evolution in time of the frequency content of a time series. It can also be read as a 3D representation of amplitude, frequency and time of the seismic signal. The spectrograms of the seismic signals generated by a mass movement that descend a slope and approach a seismic sensor can be divided into sections: SON (Signal ONset), SOV (Signal Over) and SEN (Signal End), depending on whether the gravitational mass movement is approaching the sensor, is on it or is moving away from it.</p><p>The method presented here consist of analyzing the spectrogram as an image, applying image processing techniques as “Hough Transform”. This method allows us to obtain quantitative information from the spectrograms. Our aim is to obtain the parameters of the shape of the spectrograms, focused on SON section, to create indicators linked to the evolution of the mass movement, for example the speed. The method is applied to spectrograms of three types of gravitational mass movements: snow avalanches (7), lahars (4), and debris flows (1). The results indicate similarities in the shape of the spectrograms of the different types of mass movement, prevailing, however, the specific characteristics of each type.</p>