17:15 〜 18:45
[SCG48-P05] MCMCを用いた数値シミュレーションによるGNSS-A海底地殻変動観測の音響測距誤差のモデリング評価
キーワード：GNSS-A、海底測地、音響測距、MCMC
Seafloor geodetic observation using the GNSS-Acoustic ranging combination technique (GNSS-A) is a complement of terrestrial GNSS observation, and is used to measure seafloor crustal deformation that cannot be detected onshore. GNSS-A observation provides us valuable data of the offshore regions of the subduction zones along the Japanese Islands, and has revealed numerous subseafloor tectonic phenomena (e.g. Yokota et al. 2016, Nature).
GNSS-A, which combines GNSS with underwater acoustic ranging to precisely measure the global coordinates of a point on the seafloor, is prone to errors emerging from both GNSS and underwater acoustic ranging. Recently, we have been largely focusing on the errors that originate from the acoustic ranging systems used in our GNSS-A observation. Similar to the GNSS cycle slip and phase center variation, the acoustic data obtained by GNSS-A observation contain signal reading errors and waveform deformation due to angular dependency. We have been conducting reanalysis of our acoustic data and water tank experiments to develop a correction method for such errors originating from the acoustic instruments (Nagae et al. 2024, JpGU).
However, there are limitations to the error source evaluation using actual GNSS-A observation data, due to the diversity of the error sources. To evaluate the individual effects of the error sources of GNSS-A, we have been conducting a series of numerical simulations. For example, the effects of the geometries of the survey line and the seafloor transponder array (Nakamura et al. 2021, FES) and the underwater sound speed gradient (Nakamura et al. 2023, JpGU) on the GNSS-A positioning accuracy have been evaluated using numerically simulated GNSS-A data and our GNSS-A data analysis software. Recently, we have conducted preliminary evaluations on the performance of different GNSS-A statistical models using MCMC and a simple one-dimensional GNSS-A data with a pseudo-bias representing acoustic signal reading error. (Nakamura et al. 2023, SSJ Fall Meeting). In this study, we will further evaluate the GNSS-A statistical models using synthetic data with more complex acoustic ranging errors and underwater sound speed structure.
GNSS-A, which combines GNSS with underwater acoustic ranging to precisely measure the global coordinates of a point on the seafloor, is prone to errors emerging from both GNSS and underwater acoustic ranging. Recently, we have been largely focusing on the errors that originate from the acoustic ranging systems used in our GNSS-A observation. Similar to the GNSS cycle slip and phase center variation, the acoustic data obtained by GNSS-A observation contain signal reading errors and waveform deformation due to angular dependency. We have been conducting reanalysis of our acoustic data and water tank experiments to develop a correction method for such errors originating from the acoustic instruments (Nagae et al. 2024, JpGU).
However, there are limitations to the error source evaluation using actual GNSS-A observation data, due to the diversity of the error sources. To evaluate the individual effects of the error sources of GNSS-A, we have been conducting a series of numerical simulations. For example, the effects of the geometries of the survey line and the seafloor transponder array (Nakamura et al. 2021, FES) and the underwater sound speed gradient (Nakamura et al. 2023, JpGU) on the GNSS-A positioning accuracy have been evaluated using numerically simulated GNSS-A data and our GNSS-A data analysis software. Recently, we have conducted preliminary evaluations on the performance of different GNSS-A statistical models using MCMC and a simple one-dimensional GNSS-A data with a pseudo-bias representing acoustic signal reading error. (Nakamura et al. 2023, SSJ Fall Meeting). In this study, we will further evaluate the GNSS-A statistical models using synthetic data with more complex acoustic ranging errors and underwater sound speed structure.