11:00 AM - 1:00 PM
[SVC34-P07] Current deformation rates of the Unzen lava dome and its physical mechanism: Application of InSAR time series analysis
Keywords:InSAR, Unzen, Lava dome, Finite element analysis
Unzen is an active volcano located in the Shimabara Peninsula, Japan. Following the eruption started in November 1990, the first lava dome was formed in May 1991 and its growth continued until February 1995. The lava dome consists of the summit area and the lava lobes formed at the top and on the eastern slope, respectively. This lava dome has been pointed out as a potential collapse hazard. The current deformation of the lava dome has been monitored by GB-SAR, EDM, GNSS, etc. The GB-SAR and EDM measure only the movement of the lower slope of the lava lobe, while the GNSS stations locate only at the summit area of the lava dome. Thus, these ground-based observations cannot monitor the entire deformation field. In this study, we apply InSAR time series analysis using ALOS-2 data to detect the current deformation rate of the entire lava dome and its spatiotemporal evolution with high spatial resolution. Then, we construct a physical model to explain the current movement of the lava dome.
First, a lot of InSAR images were produced from SAR data taken by ALOS-2 from ascending and descending orbits. Next, we applied an SBAS-based time series analysis (Berardino et al., 2002; Schmidt and Bürgmann, 2003) to the interferograms to reduce the effects of tropospheric and ionospheric disturbances. The LOS velocity fields estimated for both orbits commonly indicate the surface movements away from the satellite at an almost constant rate. The obtained displacement time series was consistent with the GNSS data in the summit area.
Next, the obtained velocity field was decomposed into quasi-vertical and quasi-east components (Fujiwara et al., 2000). The results indicate the subsidence and eastward movement of the lava dome, and that the velocity peaked at two points: the summit area and the lava lobe on the east flank. At the summit area, the rates of subsidence and eastward movement were 10 cm/yr and 4 cm/yr, respectively, while at the lava lobe, those were 8 cm/yr and 6 cm/yr, respectively. Thus, the motions at these two peaks are different from each other in both direction and magnitude.
Finally, we constructed a finite element model to explain the displacements of the lava dome detected above. The medium is modeled as a Maxwell viscoelastic body. We imported the DEHM data provided by GSI to consider the realistic topography. We also considered the rheological heterogeneity inside the volcanic edifice associated with the old landscape before the eruption. The current velocity field was obtained by the time integral over 10,000 days (~27 years) after initial gravitational loading to the entire medium. The calculation results reproduced the peaks in the east-west and vertical velocity components at two locations (i.e., the summit and the lobe). The difference in the direction of motion between the summit and the lobe of the lava dome can be attributed to the structural heterogeneity inside the edifice. By solving the heat conduction equation and the equation of motion simultaneously, we tentatively evaluated the subsidence rate due to the thermal stress as ~30 % of the total rate. However, we found that the effect of thermal stress depends on the temperature distribution in the depth direction immediately after the formation of the lava dome. Anyway, the overall eastward movements of the entire lava dome have been driven by gravity, not by thermal stress.
First, a lot of InSAR images were produced from SAR data taken by ALOS-2 from ascending and descending orbits. Next, we applied an SBAS-based time series analysis (Berardino et al., 2002; Schmidt and Bürgmann, 2003) to the interferograms to reduce the effects of tropospheric and ionospheric disturbances. The LOS velocity fields estimated for both orbits commonly indicate the surface movements away from the satellite at an almost constant rate. The obtained displacement time series was consistent with the GNSS data in the summit area.
Next, the obtained velocity field was decomposed into quasi-vertical and quasi-east components (Fujiwara et al., 2000). The results indicate the subsidence and eastward movement of the lava dome, and that the velocity peaked at two points: the summit area and the lava lobe on the east flank. At the summit area, the rates of subsidence and eastward movement were 10 cm/yr and 4 cm/yr, respectively, while at the lava lobe, those were 8 cm/yr and 6 cm/yr, respectively. Thus, the motions at these two peaks are different from each other in both direction and magnitude.
Finally, we constructed a finite element model to explain the displacements of the lava dome detected above. The medium is modeled as a Maxwell viscoelastic body. We imported the DEHM data provided by GSI to consider the realistic topography. We also considered the rheological heterogeneity inside the volcanic edifice associated with the old landscape before the eruption. The current velocity field was obtained by the time integral over 10,000 days (~27 years) after initial gravitational loading to the entire medium. The calculation results reproduced the peaks in the east-west and vertical velocity components at two locations (i.e., the summit and the lobe). The difference in the direction of motion between the summit and the lobe of the lava dome can be attributed to the structural heterogeneity inside the edifice. By solving the heat conduction equation and the equation of motion simultaneously, we tentatively evaluated the subsidence rate due to the thermal stress as ~30 % of the total rate. However, we found that the effect of thermal stress depends on the temperature distribution in the depth direction immediately after the formation of the lava dome. Anyway, the overall eastward movements of the entire lava dome have been driven by gravity, not by thermal stress.