Tracking volcanic explosions using Shannon entropy at Volcán de Colima

Abstract

The main objective of this work is to show that Shannon Entropy (SE) calculated on continuous seismic signals can be used in a volcanic eruption monitoring system. We analysed three years of volcanic activity of Volcán de Colima, México, recorded between January 2015 and May 2017. This period includes two large explosions, with pyroclastic and lava flows, and intense activity of less energetic explosion, culminating with a period of quiescence. In order to confirm the success of our results, we used images of the Visual Monitoring system of Colima Volcano Observatory. Another of the objectives of this work is to show how the decrease in SE values can be used to track minor explosive activity, helping Machine Learning algorithms to work more efficiently in the complex problem of distinguishing the explosion signals in the seismograms. We show that the two big eruptions selected were forecasted successfully (6 and 2 days respectively) using the decay of SE. We conclude that SE could be used as a complementary tool in seismic volcano monitoring, showing its successful behaviour prior to energetic eruptions, giving time enough to alert the population and prepare for the consequences of an imminent and well predicted moment of the eruption.

Figure 1

(a) Map of the seismic stations (squares) and visual cameras (cameras) monitoring Volcán de Colima, Mexico. Black triangles show the old Nevado volcano and the active Volcán de Colima. Red squares are the representing seismic stations in this work. Map made with Surfer 16 (https://www.goldensoftware.com/products/surfer). (b) Pictures of eruptive episodes analyzed; from upper to lower: path of the pyroclastic flow occurred on July 11th 2015; vulcanian eruption started on October 1st 2016. Pictures of Raúl Arámbula-Mendoza.

Figure 2

Temporal evolution of the SE between January 2015 and May 2017 analysed at SOMA (blue) and INCA (green) seismic stations using window lengths of 10 min overlaped 50%. We represented the envelope of the SE values. Vertical red line represents the two selected eruptive episodes occurred on July 11th 2015 and October 1st 2016. Shadow red areas represent the two intervals selected to analyzed smaller volcanic explosions. White spaces represent periods without data. Pictures from left to right show three explosive episodes recorded by the CUEV cameras occurred on 11 July 2015, 1 October 2016 and 3 February 2017 respectively.

Figure 3

(a) SE during June and July 2015. Red lines show the moment of the two pyroclastic flows occurred in July 11th. Red shadow areas are the periods used to evaluate how SE can be used to monitor volcanic explosions. Green area is the confirmed short term forecasting period (5 days) obtained from the decay of the SE. (b) Plot of the STA/LTA ratio during June and July 2015. Period in which values are over 70% of decay are highlighted in red. (c) Zoom of 11 of July showing as the SE reached zero when the pyroclastic flows happened.

Figure 4

Temporal evolution of the SE values, obtained for the seismic station SOMA, associated to the high explosivity period of 3–20 June 2015. In this period, we identified local minima and compared them (if available) with the images obtained by the visual monitoring Network. As observed all identified minima are linked low energy explosive activity.

Figure 5

Temporal evolution of the SE values, obtained for the seismic station INCA, associated to period of 3 January-5 February 2015. In this with period we identified local minima and compared them (if available) with the images obtained by the visual monitoring Network. As observed all identified minima are linked to explosive activity.