3:00 PM - 3:15 PM
[SVC29-06] Ground deformations at Kuju volcano, Japan, revealed by PALSAR-2 InSAR
Keywords:Interferometric Synthetic Aperture Radar, Kuju volcano, volcanic deformation, Hattyobaru geothermal power plant
Kuju volcano is located in the western part of Oita prefecture in Kyushu Island, Japan. The latest eruption occurred in 1995 at Iwoyama, which is the most active area in Kuju volcano and has numerous active fumarolic vents. Post-eruptive ground deflation accompanying significant heat discharge started immediately after the eruption (Nakaboh et al., 2003), and continued until, at least, 2002 (Saito et al., 2003). From 2012, a continuous GNSS observation network installed in 2001 by Japan Metrological Agency (JMA), has recorded increases in the baseline length indicating ground inflation, which suggests a possibility of on-going magma intrusion or fluid migration into a shallow part of Kuju volcano.
To clarify the spatiotemporal evolution of this deformation in more detail, we analyzed PALSAR-2 data, which is L-band Interferometric Synthetic Aperture Radar (InSAR), operated by Japan Aerospace Exploration Agency (JAXA). We used two-track data; right-looking ascending and right-looking descending orbits, acquired from August 2014 to November 2019. We performed SAR interferometry using RINC software (Ozawa et al., 2016) and time series analysis using SBAS (Small Baseline Subset; Berardino et al., 2002) method implemented with GIAnT software (Agram et al., 2013).
SBAS results of both tracks show spatial patterns indicating ground deflation in Iwoyama and geothermal power plants of Hattyobaru and Otake (hereafter, Hattyobaru and Otake), which are located on the west side of Kuju volcano. The spatial extent of the deformation is less than 1 km at Iwoyama, while it is about 4 km in Hattyobaru and Otake. Quasi-eastward and vertical deformation maps show different spatial characteristics. At Iwoyama, there is a rather incomprehensible pattern; a combination of subsidence (1.4 cm/yr) and only eastward displacement (1.4 cm/yr), which is not likely to be modeled by only a simple deflation source. On the other hand, in Hattyobaru and Otake, a simple pattern indicating ground deflation can be recognized, and the maximum subsidence velocity is about 1.2 cm/yr at Hattyobaru and 0.4 cm/yr at Otake. These deformation characteristics are consistent with those of ALOS-derived deformation around Hattyobaru and Otake during 2007-2010 (Ishitsuka et al., 2016), suggesting that a common deflation source, such as the deflated geothermal reservoir just beneath Hattyobaru at 750-m depth (Nishijima et al., 2005), has consistently caused the deformation during 2007-2019.
Source modeling for the deformation at Iwoyama shows that a combination of a spherical deflation source and a right strike-slip fault with slight normal faulting better fits the observed deformation than only a single deflation source. The deflation source during 2014-2019, of which depth is 200-300 m, is clearly shallower than the deflation source during 1995-2002 at a depth of 500-600 m. On the other hand, the 2014-2019 deflation source is probably overlapped with the position of a demagnetized source from 2014, suggesting coeval progress of deflation and demagnetization. The combination of deflation and demagnetization is an unusual phenomenon because, in other volcanoes, only cases of coeval inflation-demagnetization and deflation-magnetization have frequently been reported (Hashimoto et al., 2019).
The ground deflation of Iwoyama observed in this study is opposed to the deformation observed by the GNSS network, i.e. ground inflation. In addition, GNSS baseline lengths (4-5 km) are longer than the spatial scale of the Iwoyama deformation (less than 1 km), which suggests the existence of a deeper inflation source. To disclose the real nature of the ground inflation observed from 2012, it is necessary to further investigate GNSS data; remove several non-volcanic factors, such as post-seismic deformation of the Kumamoto earthquake and subsidence of Hattyobaru and Otake geothermal power plants.
To clarify the spatiotemporal evolution of this deformation in more detail, we analyzed PALSAR-2 data, which is L-band Interferometric Synthetic Aperture Radar (InSAR), operated by Japan Aerospace Exploration Agency (JAXA). We used two-track data; right-looking ascending and right-looking descending orbits, acquired from August 2014 to November 2019. We performed SAR interferometry using RINC software (Ozawa et al., 2016) and time series analysis using SBAS (Small Baseline Subset; Berardino et al., 2002) method implemented with GIAnT software (Agram et al., 2013).
SBAS results of both tracks show spatial patterns indicating ground deflation in Iwoyama and geothermal power plants of Hattyobaru and Otake (hereafter, Hattyobaru and Otake), which are located on the west side of Kuju volcano. The spatial extent of the deformation is less than 1 km at Iwoyama, while it is about 4 km in Hattyobaru and Otake. Quasi-eastward and vertical deformation maps show different spatial characteristics. At Iwoyama, there is a rather incomprehensible pattern; a combination of subsidence (1.4 cm/yr) and only eastward displacement (1.4 cm/yr), which is not likely to be modeled by only a simple deflation source. On the other hand, in Hattyobaru and Otake, a simple pattern indicating ground deflation can be recognized, and the maximum subsidence velocity is about 1.2 cm/yr at Hattyobaru and 0.4 cm/yr at Otake. These deformation characteristics are consistent with those of ALOS-derived deformation around Hattyobaru and Otake during 2007-2010 (Ishitsuka et al., 2016), suggesting that a common deflation source, such as the deflated geothermal reservoir just beneath Hattyobaru at 750-m depth (Nishijima et al., 2005), has consistently caused the deformation during 2007-2019.
Source modeling for the deformation at Iwoyama shows that a combination of a spherical deflation source and a right strike-slip fault with slight normal faulting better fits the observed deformation than only a single deflation source. The deflation source during 2014-2019, of which depth is 200-300 m, is clearly shallower than the deflation source during 1995-2002 at a depth of 500-600 m. On the other hand, the 2014-2019 deflation source is probably overlapped with the position of a demagnetized source from 2014, suggesting coeval progress of deflation and demagnetization. The combination of deflation and demagnetization is an unusual phenomenon because, in other volcanoes, only cases of coeval inflation-demagnetization and deflation-magnetization have frequently been reported (Hashimoto et al., 2019).
The ground deflation of Iwoyama observed in this study is opposed to the deformation observed by the GNSS network, i.e. ground inflation. In addition, GNSS baseline lengths (4-5 km) are longer than the spatial scale of the Iwoyama deformation (less than 1 km), which suggests the existence of a deeper inflation source. To disclose the real nature of the ground inflation observed from 2012, it is necessary to further investigate GNSS data; remove several non-volcanic factors, such as post-seismic deformation of the Kumamoto earthquake and subsidence of Hattyobaru and Otake geothermal power plants.