G005-0003
The IGS Repro3 Contribution by TU Graz: Benefits and Challenges of an Uncombined and Undifferenced Processing Approach
Abstract:
Graz University of Technology (TU Graz/TUG) contributes, for the first time, as an analysis center to repro3. We directly process uncombined and undifferenced (raw) GNSS observations, which enables us to include any code and phase signal on any frequency. Thus, we are able to provide observable-specific code and phase biases for all satellites and stations. The large increase of available signals due to the modernization of GPS and the inclusion of GLONASS and Galileo in repro3 makes this especially relevant. It provides users with the option to choose which signals to process.
For our repro3 contribution, we process observations from up to 850 stations per day to the three GNSS constellations at a 30-second sampling rate. This results in more than 200 million raw observations for a single day, from which we estimate almost 5 million parameters. Processing that many observations together in a least squares adjustment is computationally very challenging. A highly optimized processing chain and the usage of sizable computer clusters are requisite to making such a large processing effort feasible.
This contribution highlights the benefits our processing approach and the resulting products bring to users. It also focuses on how we have overcome the challenges that a reprocessing of this scale poses. Finally, it includes a first overview and analysis of the resulting products.