Session E3b: Advanced Technologies in High Precision GNSS Positioning

Comprehensive Performance Evaluation of QZSS CLAS Over Four Years (2021-2024) Spanning the Solar Maximum
Hayato Shiono, Nobuaki Kubo, Tokyo University of Marine Science and Technology
Location: Holiday 1 (Second Floor)
Alternate Number 2

The Quasi-Zenith Satellite System (QZSS) Centimeter Level Augmentation Service (CLAS) is the world’s first satellite-based open service for PPP-RTK. While its nominal performance is well-documented, a comprehensive, multi-year evaluation during a period of high solar activity is essential for fully understanding its capabilities and limitations under challenging ionospheric conditions. This paper presents a comprehensive performance evaluation of CLAS from 2021 to 2024, a period covering the rising phase of Solar Cycle 25. Using 30-second data from over 1,000 stations of GEONET, we analyzed key performance metrics including positioning accuracy (horizontal/vertical), Time-To-First-Fix (TTFF), and fix rate for both static and kinematic modes. The analysis of degradation factors was performed by correlating positioning performance with various geophysical data and models. Our four-year analysis reveals a clear trend of performance degradation across all metrics, which strongly correlates with the increase in solar activity. The primary cause for this degradation was overwhelmingly identified as ionospheric disturbances, accounting for 96.9% of all non-compliant days. We show that the nature of these disturbances evolved, with Medium-Scale Traveling Ionospheric Disturbances (MSTIDs) being a factor in earlier years, while Plasma Bubbles (PBs) and other complex disturbances became the dominant cause of degradation in 2023-2024. This analysis also revealed a significant spatial dependence on performance. Specifically, positioning accuracy and TTFF were consistently and substantially worse for users located ”outside” the core network area compared to those ”inside”. Case studies also revealed specific system vulnerabilities, such as the cross-contamination between error models during severe events. The findings demonstrate that while CLAS is a highly capable service, its performance is significantly challenged by the ionospheric conditions of the solar maximum. This evidence-based characterization provides valuable insights for both end-users in setting realistic performance expectations and for the service provider in identifying areas for future algorithmic and system improvements.