Quality of Orbit Predictions for Satellites Tracked by SLR Stations

Abstract

This study aims to evaluate and analyze the orbit predictions of selected satellites: geodetic, Global Navigational Satellite Systems (GNSS), and scientific low-orbiting, which are tracked by laser stations. The possibility of conducting satellite laser ranging (SLR) to artificial satellites depends on the access to high-quality predictions of satellite orbits. The predictions provide information to laser stations where to aim the telescope in search of a satellite to get the returns from the retroreflectors installed onboard. If the orbit predictions are very imprecise, SLR stations must spend more time to correct the telescope pointing, and thus the number of collected observations is small or, in an extreme case, there are none of them at all. Currently, there are about 120 satellites equipped with laser retroreflectors orbiting the Earth. Therefore, the necessity to determine the quality of predictions provided by various analysis centers is important in the context of the increasing number of satellites tracked by SLR stations. We compare the orbit predictions to final GNSS orbits, precise orbits of geodetic satellites based on SLR measurements determined in postprocessing, and kinematic orbits of low-orbiting satellites based on GPS data. We assess the quality degradation of the orbit predictions over time depending on the type of orbit and the satellite being analyzed. We estimate the time of usefulness of prediction files, and indicate those centers which publish most accurate predictions of the satellites’ trajectories. The best-quality predictions for geodetic satellites and Galileo reach the mean error of 0.5–1 m for the whole 5-day prediction file (for all three components), while the worst ones can reach values of up to several thousand meters during the first day of the prediction.

Initial Tracking, Fast Identification in a Swarm and Combined SLR and GNSS Orbit Determination of the TUBIN Small Satellite

“…In this case, the position error after one day compared with SGF's prediction reached approximately 10 cm in the radial direction, 1 m in the along-track direction, and 5 m in the cross-track direction. This was nearly an order of magnitude worse than the SGF's prediction error of LAGEOS-2 shown by Najder and Sośnica in [2]. Nevertheless, no further efforts were attempted to try to improve the prediction accuracy, as the focus of this work was on LEO satellites where atmospheric drag is the dominant source of uncertainty.…”

“…This consisted of a limited verification of the orbit determination program against other orbit determination software used by prediction centers within the ILRS network, although this verification was less systematic than the cross-validation carried out by Schutz and Tapley on simulated orbit data in [30]. The orbit prediction of LAGEOS-2 corresponding to the example shown in Figure 7 above was compared to a prediction from the same day generated by the Natural Environment Research Council (NERC)'s Space Geodesy Facility (SGF), which is the ILRS prediction center considered to provide the highest quality predictions for LAGEOS-1 and LAGEOS-2 [2]. In this case, the position error after one day compared with SGF's prediction reached approximately 10 cm in the radial direction, 1 m in the along-track direction, and 5 m in the cross-track direction.…”

“…The low laser beam divergence of SLR ground stations involves stringent accuracy requirements for orbit predictions. However, because of the different geographic locations, technologies, and level of automation of SLR stations, there is no universal rule on the exact level of orbit prediction accuracy required for a successful station pass [2]. Furthermore, once a given prediction led to successful passes over SLR stations, the time bias service from DiGOS and GFZ Potsdam was able to predict the time bias of this prediction for future passes, which amounts to compensating part of the along-track drift due to atmospheric drag [3].…”

“…While dense, geodetic satellites in medium-altitude orbits such as LAGEOS have well-determined orbits with centimeter-level residuals and sub-meter-level prediction accuracy [4], this is not always the case for spacecraft in low-Earth orbits (LEOs). The lower the orbit altitude, the higher the area-to-mass ratio of a satellite, or the higher the solar activity, the faster the orbit prediction can drift because of uncertainty in atmospheric drag modeling, reaching in some cases several kilometers per day [2]. TUBIN, with a relatively high area-to-mass ratio (cross-section from 0.13 m 2 to 0.22 m 2 for a 23 kg mass) and low-altitude orbit (approximately 530 km) and without a controlled attitude most of the time, represents one of most challenging cases for orbit predictions within the ILRS network to the authors' knowledge.…”

Satellite Attitude Identification and Prediction Based on Neural Network Compensation

“…are used to filter out the observation noise tangled in the data. These data filtering algorithms have been successfully validated on multiple aircraft platforms [13][14][15][16].…”