29th May, 2015 eruption of Kuchinoerabu-jima volcano – Earthquake Research Institute, the University of Tokyo

launched: 1st June, 2015

29th May, 2015, The Kuchinoerabu-jima volcano in Kagoshima prefecture, erupted and the volcanic alert level 5(evacuate) was issued by JMA.

Observation at the Kuchinoerabu-jima volcano using the unmanned helicopter (Tentative conclusions)

(Earthquake Research Institute, The University of Tokyo / Disaster Prevention Research Institute Kyoto University)

We installed 4 seismic observation modules using an unmanned helicopter in the close vicinity of the summit crater of Kuchinoerabu-jima volcano, southern Kyushu. We successfully obtained seismic data from 17 April to just before the summit eruption 09:59, 29 May. Figure 1 shows the locations of the helicopter’s takeoff and landing position (■) and the newly installed seismic modules (●) .

Fig.1: Helicopter's takeoff and landing position and the locations of seismic sensors. — Fig.1: Helicopter’s takeoff and landing position and the locations of seismic sensors.

The seismic observation modules are powered by solar panels, and the seismic sensors are short-period vertical velocity sensors with corner frequency of 4.5Hz (Fig.2). The unmanned helicopter carries the module above the installation position (Fig.3), and then the module is slowly lowered down to the ground surface by winch.

The seismic data are transmitted every 10 minutes through the cellular-phone data network. The final data time ranges from 09:47 to 09:52, but this timing is not when the modules were destroyed by the eruption. This is when final data transmission was done before the eruption.

Fig.3 The unmanned helicopter heading to install the module toward the summit.

The installation of the modules is done just by descending it from the helicopter using a winch, therefore, the coupling between the observation modules and the ground surface is not perfect. Moreover, being affected by the resonance of the module- frame and by the noise caused by wind, the signal-to-noise ratio is not good especially in the high frequency band.

Comparing the unfiltered raw records of the two stations EV.E1 and EV.E2, it looks as though there are huge differences in numbers of the earthquakes (Fig.4). However, after applying low-pass filter at 8Hz, most of the noises are removed and the overall appearances in earthquake occurrence at the two stations are almost the same (Fig.5).

Fig.4 seismic record taken at the 2 observation stations: EV.E1, EV.E2 on 28th May, 2015, 23:00-23:59 (unfiltered).

Fig.5 seismic record taken at the 2 observation stations: EV.E1, EV.E2 on 28th May, 2015, 23:00 – 23:59 (low pass filtered at 8Hz).

Fig.6 shows the number of events. They were extracted from the data from 1 May to 29 May via an automated detection procedure assuming that the ratio between short time average (STA) and the long time average (LTA) is greater than 5.

This histogram shows that the number of earthquakes around the summit area slightly increased from 4 May to 7 May and from 19 May to 22 May. On 23 May, a felt earthquake in JMA seismic intensity 3 occurred, but the our measurements do not show any significant increase in the number of earthquakes in the summit area. Number of earthquakes is rapidly increasing from 25 May to 28 May.

The decrease on 29 May may be due to the destruction of the observation modules by the eruption, and may be due to the absence of sudden increase in number of earthquakes before the eruption.

The seismic activity at summit area had not changed significantly till 24 May, but it escalated toward the eruption on 29 May. This fact may reflect the upward migration of the thermal source toward the shallow part of the volcano.

Note, however, that this is a tentative conclusion based on the automated event detection, therefore, the result may change by further investigations.

Fig.6 Daily number of events which was automatically detected from 1 May to 29 May.