580 California St., Suite 400
San Francisco, CA, 94104
Key research themes
1. How can atmospheric numerical weather data improve the accuracy of Zenith Tropospheric Delay estimation from GNSS observations?
This research area investigates the integration and performance of numerical weather prediction (NWP) or reanalysis datasets such as ERA5, GDAS, or NCEP in modeling and estimating Zenith Tropospheric Delay (ZTD) components. It matters because tropospheric delay modeling critically affects GNSS positioning accuracy, and traditional empirical models often lack spatial and temporal resolution to capture hydrostatic and wet delay variability precisely, especially in regions with high atmospheric variability. Using high-resolution numerical atmospheric data could refine a priori ZTD and Zenith Hydrostatic Delay (ZHD) estimates, improving the precision of GNSS tropospheric corrections and convergence of precise point positioning (PPP) approaches.
Key finding: This study demonstrates that ZHD estimates based on ERA5 surface pressure using a linear regression model (LRM) and the Saastamoinen model achieve identical accuracy with RMS error of 2.3 mm, outperforming the GPT3 model with... Read more
Key finding: This paper finds that ZTDs derived from the GDAS NWM have root mean square differences of about 3.0 cm against high-precision IGS tropospheric solutions, with biases up to 4.5 cm mainly due to relative humidity interpolation... Read more
Key finding: The study shows that PPP-derived ZTDs from Nigerian GNSS network data match closely with Numerical Weather Model (NCEP II) ray-traced ZTDs, with mean biases ranging from 6.4 mm to 23.9 mm and standard deviations of 33.1 to... Read more
Key finding: Using high-resolution NWP data with 2.5 km grid spacing over Scandinavia and surrounding regions, the study quantified differential ZTD values up to 18 cm over baselines up to 20 km, highlighting significant small-scale... Read more
Key finding: This work investigates PPP-RTK tropospheric delay modeling and finds that while Zenith Hydrostatic Delays (ZHD) can be modeled accurately a priori, Zenith Wet Delays (ZWD) require epoch-wise estimation. However, integrating... Read more
2. What is the impact of tropospheric horizontal gradients and small-scale spatial variability on Zenith Tropospheric Delay accuracy and GNSS positioning?
This theme explores the estimation, origin, and modeling of tropospheric horizontal delay gradients which represent azimuthal asymmetry and spatial decorrelation of tropospheric delay at GNSS stations. Since traditional ZTD models treat the atmosphere as azimuthally symmetric, understanding gradients is key for mitigating residual tropospheric errors impacting high-precision positioning such as PPP and PPP-RTK. Additionally, local topography and atmospheric dynamics causing variable lateral delay components are studied to inform gradient parameterization and improve the physical realism of tropospheric correction models.
Key finding: Analysis of horizontal tropospheric gradients from GNSS data across 52 stations globally reveals stable gradient directions over one year often aligning with local topographic features, such as pointing inland opposite large... Read more
Key finding: The study quantifies small-scale spatial decorrelation of tropospheric delay at baseline lengths under 20 km using high-resolution NWPs. They document significant differential ZTD values (up to 18 cm), indicating that... Read more
Key finding: Using one month of global GPS data from 10 diverse IGS stations, the study shows that neglecting tropospheric horizontal gradients biases total tropospheric delay estimates by up to ±2.14 mm on average and leads to degraded... Read more
Key finding: By analyzing 4 years of GNSS ZTD data over East Africa, including gradient estimates, the study finds that ZTD exhibits seasonal and spatial variability correlated with meteorological conditions, with horizontal gradients... Read more
Key finding: The research identifies that variations in Zenith Wet Delay are related to atmospheric water vapor variability, which is spatially and temporally heterogeneous and associated with gradients in ERA5 variables. The linear... Read more
3. How do higher-order ionospheric effects and relativistic atmospheric models influence the estimation of Zenith Tropospheric Delay in GNSS and VLBI?
This theme delves into effects beyond the classical first-order ionospheric delay corrections, including high-order ionospheric (HOI) delays and special relativistic effects (e.g., Fresnel-Fizeau effects) in signal propagation through the moving atmosphere. The significance lies in assessing whether neglecting HOI leads to biases in tropospheric parameters estimated from GNSS, and whether improvements in relativistic tropospheric delay modeling can refine time delay estimates in Very Long Baseline Interferometry (VLBI) and GNSS data processing. These subtle corrections become increasingly relevant for millimeter-level geodetic and atmospheric sensing applications.
Key finding: This regional case study shows that high-order ionospheric effects cause biases up to 6 mm in Zenith Tropospheric Delay (ZTD) and up to 12 mm in horizontal tropospheric gradients estimated from GPS data. HOI effects are... Read more
Key finding: This paper provides the first rigorous relativistic derivation of atmospheric delay including the Fresnel–Fizeau effect due to Earth's atmosphere motion through the barycentric celestial reference system. It theoretically... Read more
Key finding: While not focusing exclusively on ionospheric high-order effects, this comparative analysis shows that different PPP software estimate ZTD with millimeter-level consistency, implying that residual effects such as HOI or... Read more
Key finding: Focused on ionospheric delay estimation from dual-frequency Galileo E5 signals, this paper underscores the challenge that tropospheric delay (frequency-independent) remains uncorrected and must be modeled separately. Although... Read more