Stephen Catsamas, The University of Melbourne; Sarah Caddy, The University of Melbourne; Michele Trenti, The University of Melbourne
Keywords: catagorization, infrared, classification, thermal, modelling, space telescope
Abstract:
Current methods of monitoring satellites with remote thermal observations assume a single characteristic temperature. However, this simplification results in information being lost. In this work, we explore the ability to derive information such as power flows, thruster firing and spacecraft health from a more complex two component thermal model.
All objects in space emit thermal infrared radiation. Infrared imaging at multiple wavelengths can be used to sample this radiation (or infrared spectrum) from spatially unresolved remote observations. In this work, to understand the health and activity of a satellite from these observations, we compare them to multi-component thermal models. We derive analytic thermal and infrared spectral models for satellites and present how the spectrum changes with variables such as observation perspective and internal power usage. We motivate and introduce a two-component satellite model where both components have their own temperature and emissive properties and correspond to a component of the satellite such as a chassis or solar panels.
First, we demonstrate that non-linear least squares regression can be used to fit this two-component model from idealised unresolved thermal infrared space telescope observations. Furthermore, assuming observations with a small (~0.5 m aperture) space telescope, uncertainties of a few Kelvin can be achieved on both satellite components. We remark that a temperature difference well above 10 K is expected due to radiative geometry for a chassis and solar panel of an idle spacecraft, and hence that temperature of both these components would be able to be individually inferred using these observations. We discuss the implications our results hold for monitoring solar panel to chassis power flows in operational spacecraft.
Next, given that a) space telescope design increases in cost and complexity for longer wavelength observations as significant telescope baffle cooling is required to prevent excessive foreground noise and b) solar radiation begins to dominate the spectra at shorter wavelengths, we determine the necessary wavelength range of observation filters to attain a given temperature inference uncertainty. We show that the needed wavelength range is within current technological capabilities and that of existing infrared space telescopes.
Overall, our results show that an infrared space telescope holds clear prospects for RSO classification and categorization, and thus would be a valuable space domain awareness asset.
Date of Conference: September 16-19, 2025
Track: Satellite Characterization