4:00 PM - 4:15 PM
[SCG55-32] Effects of internal tides on sound speed field modeling of GNSS-A positioning
Keywords:GNSS-A, Seafloor geodesy, Internal tide
GNSS-A seafloor geodesy enables centimeter accuracy measurement of the position of seafloor acoustic transponders with respect to global coordinate system. This positioning technique enables us to detect precise crustal deformation on the seafloor just above the plate boundary in subduction zones and plays an important role in understanding earthquakes at plate boundaries.
The main error factor in GNSS-A, which uses the round-trip travel time of acoustic waves, is the variation of the sound speed field caused by the variation of water temperature and salinity of seawater. Due to the limitation of understanding the complex spatio-temporal variation of the sound speed field through observation, GNSS-A analysis reduce the effect by modeling the variation of the sound speed field. The free GNSS-A analysis software “GARPOS” (Watanabe et al., 2020) parametrizes the spatial variation of the sound speed field as a linear function of the positions of the sea surface and sea floor instruments. Analyses using actual observation data have yielded results consistent with a Kuroshio-induced density gradient (e.g., Yokota et al., 2024). The gradient field due to the geostrophic current is stable in the range of a few hours to a day, which is the observation time of GNSS-A. However, a gradient field with a period of about 12 hours is often estimated. This time-varying gradient field is caused by internal waves (internal tides) caused by semidiurnal tide.
Because GNSS-A is a combination of multiple positioning techniques, the actual observed data is affected with many error factors of different origins. Therefore, it is difficult to investigate each factor in detail. We examine the effect of internal tides on sound speed field modeling in GNSS-A analysis through numerical simulations using theoretically generated internal tides.
Internal tides are represented by the superposition of oscillation modes with boundary conditions with fixed sea surface and seafloor. For example, single mode (mode 1) oscillation can be approximated as a simple linear gradient field (Fig. 1). For higher-order modes, the situation is more complex, with multiple overlapping gradient fields that have various directions, and it is not obvious whether the simple linear function modeling is a good approximation. It is also important to evaluate whether the temporal and spatial variations can be adequately separated, since the tidal wave progression causes the temporal variation of gradient fields.
The main error factor in GNSS-A, which uses the round-trip travel time of acoustic waves, is the variation of the sound speed field caused by the variation of water temperature and salinity of seawater. Due to the limitation of understanding the complex spatio-temporal variation of the sound speed field through observation, GNSS-A analysis reduce the effect by modeling the variation of the sound speed field. The free GNSS-A analysis software “GARPOS” (Watanabe et al., 2020) parametrizes the spatial variation of the sound speed field as a linear function of the positions of the sea surface and sea floor instruments. Analyses using actual observation data have yielded results consistent with a Kuroshio-induced density gradient (e.g., Yokota et al., 2024). The gradient field due to the geostrophic current is stable in the range of a few hours to a day, which is the observation time of GNSS-A. However, a gradient field with a period of about 12 hours is often estimated. This time-varying gradient field is caused by internal waves (internal tides) caused by semidiurnal tide.
Because GNSS-A is a combination of multiple positioning techniques, the actual observed data is affected with many error factors of different origins. Therefore, it is difficult to investigate each factor in detail. We examine the effect of internal tides on sound speed field modeling in GNSS-A analysis through numerical simulations using theoretically generated internal tides.
Internal tides are represented by the superposition of oscillation modes with boundary conditions with fixed sea surface and seafloor. For example, single mode (mode 1) oscillation can be approximated as a simple linear gradient field (Fig. 1). For higher-order modes, the situation is more complex, with multiple overlapping gradient fields that have various directions, and it is not obvious whether the simple linear function modeling is a good approximation. It is also important to evaluate whether the temporal and spatial variations can be adequately separated, since the tidal wave progression causes the temporal variation of gradient fields.