Ionospheric Threats to Space-Based Augmentation System Development

The ionosphere contributes the largest and most unpredictable error to single frequency GPS users’ range measurements. The goal of a Space-Based Augmentation System (SBAS) in mitigating these ionospheric errors is two-fold. First, the SBAS broadcasts error corrections to its users for improved positioning accuracy. Moreover, the SBAS provides a service that GPS alone cannot: ensuring position estimate integrity, which is crucial to safety-of-life applications. The International Civil Aviation Organization (ICAO) has adopted a set of Standards and Recommended Practices (SARPs) for SBASs being developed worldwide. The SARPs are based on the Minimum Operational Performance Standards (MOPS) of the Wide Area Augmentation System (WAAS) currently operational in the United States. This paper surveys a range of ionospheric issues that an SBAS must consider if it is to comply with the ICAO SARPs. By examining observed ionospheric phenomena at a high level in a visually intuitive way, the author hopes to provide some insight as to why the SARPs are developed as they are and what additional issues are introduced by the constraints of the SARPs. This paper makes use of the following data: “supertruth” data collected from the WAAS network of receivers during several ionospheric storms as well as a nominal period for comparison; raw data from an individual WAAS network receiver during the 29-31 October 2003 ionospheric storm; and data from the same storm collected from nearly 400 stations in the Continuously Operating Reference Stations (CORS) and International GPS Service (IGS) networks and processed by the Jet Propulsion Laboratory (JPL). With these data sets the author illustrates the large absolute values of total electron content (TEC), to which GPS range errors are proportional, that may be seen during ionospheric storms. Large spatial and temporal gradients that have been observed are also shown. We discuss potential routes in bounding or mitigating the effect of these highly irregular periods of ionospheric activity by considering the approach WAAS has employed. In addition to bounding dangerous behavior that is not predicted by the SBAS choice of ionospheric model, the SBAS must also bound estimation and interpolation errors that exist both during nominal and stormy conditions. Such errors are introduced by modeling the ionosphere as a two-dimensional, infinitely thin shell. This error arises from the loss of altitudinal information in the collapse of the three spatial dimensions of the true ionosphere into a two-dimensional surface representation that can be easily broadcast. Finally, with an ionospheric model based on measurement and estimation of the real-time ionosphere, the SBAS runs a risk of undersampling the ionosphere over a geographic region for which it is providing service, as will be illustrated. When high spatial and temporal gradients are also highly localized, it is possible for them to remain undetected by the SBAS. For this reason bounding possible errors due to undersampling is crucial.