عنوان مقاله [English]
Volcanic eruptions are usually occurred by increasing the magma pressure. Monitoring this process in real time could provide useful information for predicting eruptions. The cross-correlation function of ambient seismic noises has been used so far to monitor the activity of volcanoes around the world. Still, this method is usually limited to volcanoes with a dense network and seismic broadband stations.
Here, we use a simple and robust method that can automatically calculate a cross-correlation function of seismic waves, without the need for advanced equipment or lots of data. Using the calculated cross-correlation function, it then analyzes and examines the temporary changes in the relative velocity of seismic waves as well as disrupts the change of the amplitude ratio of continuous data recorded in all station pairs as two eruption attributes.
The SARA (Seismic Amplitude Ratio Analysis) method is used to investigate the changes in the amplitude ratio, and the MWCS (Moving-Window Cross-Spectral) method is used to calculate the relative seismic velocity.
To validate these methods, the continuous data of 5 seismic stations near Etna volcano in Italy are used. Firs, a daily vertical record of all stations was divided into 30-minute segments. Then, these raw time series were demeaned and tapered and the traces are normalized to three times the root-mean-square (RMS). Next, the daily cross-correlation between all station pairs was calculated and the cross-correlation function was filtered in different frequency ranges, and range of 0/5-5/5 Hz had acceptable result. current cross-correlation functions were calculated for the period of 5, 10 and 30 days, and they were compared with reference cross-correlation function. Finally, the velocity variations are measured.
The results show a decrease of about 0/2 percent in relative velocity before the main eruption and an increase rapidly after the eruption.
In the second step, the amplitude ratio of all station pairs was calculated. three hours before the main eruption 2-5 times increase in the amplitude ratio was observed, compared to the normal state.
The results show a temporary increase in seismic amplitude ratio in 90 percent of station pairs, before the main eruption. To quantify the results, the Mann-Kendall trend analysis test was used for all pairs of stations, also by using Sen's slope estimation test, the slope value of each figure was calculated separately.
Automatic and continuous measurement of these attributes and combining their results can show the potential of this method to improve volcano monitoring and Eruption early warning for active volcanoes.