Todd Walter and Juan Blanch, Stanford University; Lance de Groot and Laura Norman, Hexagon, Canada

View Abstract Sign in for premium content


Stanford and Hexagon performed this analysis to assess the accuracy and behavior of ionospheric corrections provided for use with Precise Point Positioning (PPP). In particular we are interested in the achieved level of precision as a function of ionospheric state, reference correction message latency, and network location and coverage. In this study we analyze 28 days’ worth of data from three different locations within North America. The data is from April 2020 and the locations include Priddis in Alberta, and Stanford and Vandenburg in California. Correction data is applied to IGS reference station data from these three locations. Double difference evaluations were formed by comparing the measurements minus the estimated ionosphere for one satellite against the measurements minus the estimated ionosphere for another. These double differences demonstrated precision to the level of 2 cm 95% across the entire period with little day to day variation. The accuracy degraded very slowly with latency and did not exhibit noticeable increase in the error until 120 seconds of latency was applied. By 240 seconds the 95% containment had increased to 3 cm. This weak dependency on latency is most likely due to the fact that 2020 is near the very lowest point of solar minimum conditions and the ionospheric variability is also minimized. We are also very interested in bounding these ionospheric errors. A confidence sigma was provided along with the outliers of the corrections and we evaluate the normalized distribution as well. We examined the full distribution of both the corrections errors and normalized correction errors in order to assess the trustworthiness of the corrections and to determine the probability that the provided sigma value will not sufficiently describe the expected error distribution.