Quality assessment of daily GNSS coordinate time series from the SoftBank proprietary reference network

Abstract In recent years, earth science disciplines have begun to rely heavily on geospatial data to determine crustal deformation associated with earthquakes and volcanic activity. Therefore, having adequately high-resolution data is essential. In this study, we evaluated the quality of the daily coordinate time series from the proprietary nationwide Global Navigation Satellite System (GNSS) reference network operated by SoftBank Corp. (over 3300 stations) relative to the GNSS Earth Observation Network System (GEONET) to assess its suitability for crustal deformation monitoring. Both networks were processed using an identical single-constellation precise point positioning strategy with ambiguity resolution, and the data from the months immediately prior to the 2024 Noto Peninsula Earthquake, from November 2019 to December 2023, were analyzed. To address non-tectonic offsets at the SoftBank stations, we employed a semi-automated processing scheme and retained corrections only when they reduced the root-mean-square error of the linear fits to the horizontal components. Short-term scatter was nearly identical between networks: histogram modes of 1.0–1.5 mm (horizontal) and 4.5–5.0 mm (vertical), with median standard deviations for 300 randomly sampled stations of 1.3/1.3/4.5 mm (east–west/north–south/up–down) for GEONET and 1.2/1.3/4.3 mm for SoftBank. Both networks show a seasonal increase in scatter during summer and a weaker increase during mid-winter, consistent with tropospheric and snow effects. Harmonic modeling with Akaike Information Criterion-based order selection yielded similar distributions of seasonal amplitudes and orders, with slightly larger vertical annual amplitudes for GEONET. Steady velocities from an approximately 4-year time series achieved uncertainties of ≤0.1 mm/yr (horizontal) and ~0.1–0.2 mm/yr (vertical) for both networks. These two networks provide ultra-dense national coverage, substantially enhancing the capability of detecting both tectonic and volcanic deformations. Graphical Abstract