4:30 PM - 4:45 PM
[SCG55-34] Is the small shortening of the baseline length detected by Acoustic Distance Meter significant?
In the southwestern part of the Kuril Trench, several M8-class trench earthquakes have been recorded in the past 200 years, most recently the 1973 Nemuro-oki earthquake (M7.4) and the 2003 Tokachi-oki earthquake (M8.0). A tsunami deposit survey (Nanayama et al., 2003) revealed that a M9-class great earthquake occurred in this area in the 17th century, about 350 years ago. In 2017, the headquarters for Earthquake Research Promotion estimated that the probability of an Mw 8.8 or greater earthquake occurring in the southwestern Kuril Trench is 7-40% within the last 30 years, with an interval of 340-380 years. In addition, it was suggested that the last 17th-century great earthquake caused rupture near the trench axis (Ioki and Tanioka, 2016). However, no knowledge has been obtained on the state of the rupture because of a lack of observational data.
Based on this, Tohoku University and Hokkaido University established a seafloor geodetic observation network in 2019. In this study, we use a continuous observation data acquired by the seafloor acoustic ranging system, which were supposed to be installed across the trench axis, were obtained from July 2019-April 2021.
The seafloor acoustic ranging system measures the change of the baseline between stations on the seafloor. It can be obtained by measuring the round-trip travel time of sound between them and taking half of the product of that and sound velocity. Practically, the travel time varies depending on the attitude of the instrument, and the sound velocity varies depending on temperature and pressure, so temperature, pressure, and attitude data were obtained and used to correct. The equipment itself consists of three units, one on the Okhotsk Sea plate and two on the Pacific Ocean plate, across the trench axis, located off the coast of Nemuro.
Assuming steady rate movement during the survey period, we applied weighted linear regression to the obtained time-series of the corrected baseline changes to estimate rates in baseline changes and their error; -5.52±3.01 mm/yr for u1-u2, -5.52±3.01 mm/yr for u1-u3 and -0.85±1.04 mm/yr for u2-u3 (negative means shortening). Crossing angle of each baseline to the trench axis is 84.8° for u1-u2 and 54.6° for u1-u3. We calculate trench-normal and trench parallel component of the relative motion across the trench axis from these obliquely crossing two baselines. We also estimate error ellipse of this motion by multiplying individual error in above obtained errors in baseline length changes with error propagation matrix. Because of relatively ill-conditioned layout of the two baselines, error ellipse significantly elongates to trench-parallel direction if errors in the two baselines do not correlate to each other; 5.70±3.15 mm/yr shortening for trench-normal and 1.50±7.38 mm/yr right-lateral. In the realistic situation, the two baselines must have much of correlated error due to common disturbance of temperature hance sound speed variation. If correlation coefficients of the two error are assumed to be +0.75, ellipticity of error ellipse becomes better; 5.70±2.72 mm/yr shortening for trench-normal and 1.50±3.94 mm/yr right-lateral. Since the displacement rate for u2-u3 regards 0 within error range, we consider no effect of unmeasured salinity, ignoring the drift of temperature sensors, and removing drift in pressure. In any case, it is hard to say that trench-parallel movement is ongoing, but shortening in trench-normal component looks significant. GNSS-A measurement alone cannot discuss such small amount of difference, however, ADM measurement across the trench axis pointed out possibility to have slightly smaller coupling than full coupling.
Based on this, Tohoku University and Hokkaido University established a seafloor geodetic observation network in 2019. In this study, we use a continuous observation data acquired by the seafloor acoustic ranging system, which were supposed to be installed across the trench axis, were obtained from July 2019-April 2021.
The seafloor acoustic ranging system measures the change of the baseline between stations on the seafloor. It can be obtained by measuring the round-trip travel time of sound between them and taking half of the product of that and sound velocity. Practically, the travel time varies depending on the attitude of the instrument, and the sound velocity varies depending on temperature and pressure, so temperature, pressure, and attitude data were obtained and used to correct. The equipment itself consists of three units, one on the Okhotsk Sea plate and two on the Pacific Ocean plate, across the trench axis, located off the coast of Nemuro.
Assuming steady rate movement during the survey period, we applied weighted linear regression to the obtained time-series of the corrected baseline changes to estimate rates in baseline changes and their error; -5.52±3.01 mm/yr for u1-u2, -5.52±3.01 mm/yr for u1-u3 and -0.85±1.04 mm/yr for u2-u3 (negative means shortening). Crossing angle of each baseline to the trench axis is 84.8° for u1-u2 and 54.6° for u1-u3. We calculate trench-normal and trench parallel component of the relative motion across the trench axis from these obliquely crossing two baselines. We also estimate error ellipse of this motion by multiplying individual error in above obtained errors in baseline length changes with error propagation matrix. Because of relatively ill-conditioned layout of the two baselines, error ellipse significantly elongates to trench-parallel direction if errors in the two baselines do not correlate to each other; 5.70±3.15 mm/yr shortening for trench-normal and 1.50±7.38 mm/yr right-lateral. In the realistic situation, the two baselines must have much of correlated error due to common disturbance of temperature hance sound speed variation. If correlation coefficients of the two error are assumed to be +0.75, ellipticity of error ellipse becomes better; 5.70±2.72 mm/yr shortening for trench-normal and 1.50±3.94 mm/yr right-lateral. Since the displacement rate for u2-u3 regards 0 within error range, we consider no effect of unmeasured salinity, ignoring the drift of temperature sensors, and removing drift in pressure. In any case, it is hard to say that trench-parallel movement is ongoing, but shortening in trench-normal component looks significant. GNSS-A measurement alone cannot discuss such small amount of difference, however, ADM measurement across the trench axis pointed out possibility to have slightly smaller coupling than full coupling.