Pointing-Error Calibration of Balloon-Tracking Antennas using Satellite Signals

Abstract

The Taiki Aerospace Research Field in Hokkaido, Japan, is a crucial hub for the Japan Aerospace Exploration Agency’s scientific balloon operations, with three telemetry antennas playing a pivotal role in supporting these operations. The antennas can measure the balloon’s position and receive telemetry data. In contrast, GPS receivers carried in the balloon’s payload are the primary means of obtaining the balloon’s position via telemetry data. The antenna pointing direction obtained using electric-field-strength tracking provides essential backup for GPS balloon positioning since it does not require receiving and demodulating telemetry data. Our study has identified significant errors in the direction of the antenna pointing when analyzing GPS data via telemetry. For instance, previous balloon operations revealed systematic errors of approximately 0.4 degrees in azimuth, corresponding to a position error of 0.7 km at a distance of 100 km. We developed an automatic system to track satellite signals and correct pointing errors to address this inaccuracy. Traditional error prediction models proved challenging since the antenna-pointing error has a complex dependence on the pointing direction. Our study uniquely demonstrated that calibration models—such as Radial Basis Function interpolation and neural networks—combined with satellite tracking data can reduce the pointing error during balloon tracking for the first time. When applied to balloon-flight data, these models significantly reduced the systematic errors in both azimuthal and elevation angles, marking a novel advancement in balloon tracking systems.