Integration of radiometer data to improve tropospheric correction in GNSS-PPP processing.

All geodetic space techniques used today are based on measuring the propagation time of electromagnetic waves between a transmitter and a receiver. The atmosphere is one of the most limiting factors for the achievable accuracy of the observation techniques. The wet part of the troposphere, particularly water vapour, has a strong and highly variable influence on signals in the microwave range, as is the case with GNSS. By using current models, this high variability makes it difficult to sufficiently correct the influence of the troposphere on the signal propagation time. Neither weather models nor estimation of parameters can accurately capture the high tropospheric variability. An alternative method for determining the wet propagation delay is a water vapour radiometer.
With the HATPRO-G5, the Geodetic Observatory Wettzell has a modern radiometer that is capable of measuring the wet part of the troposphere above the station not only at the zenith but also for various azimuth-elevation combinations. This allows the wet delay to be recorded as a function of azimuth, elevation, and time. The resulting data set will first be analyzed and then used in combination with GNSS observations. To make this data available for GNSS evaluation, the water vapour radiometer data must be smoothed and interpolated both spatially and temporally. To assess the impact of radiometer-based tropospheric correction, the estimated station height component is compared with a standard GNSS processing and with the height component of a station coordinate product. We performed a PPP to estimate the station coordinates. However, the results do not show a clear picture. On the one hand, the estimation of the station height component using radiometer-based correction appears to deliver better results than the classic approach for certain time periods. On the other hand, however, these improvements cannot be reproduced for the entire time period of the data set. The data set still exhibits systematic errors whose origin has not yet been clarified, which in turn negatively affect the accuracy of the height component estimate. Possible reasons for this could include rapid weather changes, rain events or also electronic-specific systematics. Nevertheless, radiometer-based corrections have the potential to positively influence the accuracy of parameter estimation in GNSS processing.