Jovan Skuljan, Defence Technology Agency
Keywords: calibration satellites, data reduction, high-precision astrometry
Abstract:
The Defence Technology Agency (DTA) of the New Zealand Defence Force (NZDF) operates a space domain awareness (SDA) observatory located at Whangaparaoa peninsula just north of Auckland, New Zealand. The observatory is equipped with two 28-cm telescopes mounted side-by-side on a Paramount ME II robotic tracking mount. A range of cooled CCD cameras are used for imaging, and a specialized global positioning system (GPS) unit was developed for accurate timing.
Over the past several years, the observation program of the DTA observatory consisted of astrometric, photometric and polarimetric measurements of satellites in all orbital regimes. More recently, the focus had shifted to high-precision astrometric measurements of calibration satellites, mainly in geostationary Earth orbit (GEO) and medium earth orbit (MEO). It was demonstrated that small-aperture equipment can be used for obtaining astrometric data of high precision (low random error) and high accuracy (low systematic error). However, special attention must be paid to the image analysis and data reduction process. We have achieved this in-house by creating a dedicated SDA data analysis tool, StarView, optimised for images obtained with the DTA equipment.
The astrometric calibration in StarView is based on the European Space Agencys (ESA) GAIA catalogue (Data Release DR-2). For practical purposes, the catalogue was limited down to magnitude 16, in order to keep the total number of stars manageable, while providing enough data for reliable calibration. For the recognition of a stellar field and identification of individual stars in the image, we use our own algorithm called stellar fingerprints. After several years of application, this method proved to be extremely fast and reliable. The average random uncertainty of a typical astrometric solution is about 0.2 arc seconds, based on several hundred stars within the field of view.
In addition to the stellar fingerprint algorithm, which is only used on frames in sidereal mode, i.e. when the stars appear as point sources, a different technique is used when tracking a satellite, and stars leave traces. Instead of analysing each stellar trace, which can be difficult for faint objects, StarView applies a two-dimensional cross-correlation between the satellite image and the star catalogue to determine the shift in right ascension and declination. This ensures that every star in the image makes a contribution to the overall cross-correlation profile. The random errors in satellite astrometry are very low, between 0.2 and 0.5 arc seconds, depending on image quality.
A number of Tracking and Data Relay Satellites (TDRS) in GEO, as well as some NAVSTAR (GPS) satellites in MEO, were observed over the past several months. The astrometric coordinates were compared with the calibration ephemeris data provided by the 18th Space Defense Squadron (18 SDS) in Vandenberg, CA. The residuals showed a low scatter comparable with the random measurement errors, and a hint of a small absolute offset, typically within a fraction of an arc second.
Date of Conference: September 27-20, 2022
Track: Astrodynamics