Can WAAS Availability Be Inferred From Geomagnetic Data? An Analysis

The Federal Aviation Administration’s Wide Area Augmentation System (WAAS) provides high integrity GPS-based precision navigation service to users in the conterminous United States (CONUS). User integrity is ensured by bounding all possible GPS error sources, of which the ionospheric delay is the largest and most variable. In particular, during severe ionospheric storms, the WAAS Irregularity Detector alerts that the errors cannot be bounded tightly enough to allow for precision navigation service, and protects users in realtime by lowering precision navigation service availability. As a result, a question of interest often arises: can we know in advance that WAAS availability will be reduced and if so, what are the reliable indicators? A current heuristic method based on observed magnetosphere-ionosphere coupling is to be on alert when the planetary index of geomagnetic activity Kp reaches the higher range of its possible values. However, the Kp index is anecdotally known to have a high false alarm rate for WAAS availability. Kp has been known to reach its highest value on days for which there has been a negligible decrease in WAAS precision service availability. This paper is an exploration of how Kp and Dst perform as indicators of WAAS LPV service availability. We examine this question by computing the correlation between the WAAS Irregularity Detector’s chi-squared values, representing WAAS availability, for the known WAAS storm (i.e. low-availability) days and the geomagnetic indices of Kp and Dst. We show that Kp shows moderate correlation with the observation of WAAS-effective storms, and that Dst has a strong negative correlation with the log of the chi-squared ratio. We find that Kp also yields a higher false alarm rate than Dst. This work provides preliminary evidence that Dst is slightly more indicative of WAAS availability than Kp.