Statistical characterization of multispectral infrared properties of the GEO belt debris population

Advanced statistical characterization of multispectral infrared properties of the GEO belt debris population

“…The U.S. Air Force Research Lab created a debris catalog that includes measurements of over 2,000 new, previously un-cataloged, Debris objects near the GEO belt 2,4,5 . Since the original debris and satellite database was created, it has been reproduced by others in academia 6,3 . For the purposes of this work, we reproduced elements of the debris and satellite catalog based on the public domain sources.…”

“…In prior work Ref. [3], only 4,800 observations were reported, which are objects, specifically satellites, rocket bodies, and tracked debris that were associated with cataloged objects. There are many more observations in 2 bands, after the cryogenic hydrogen melted, however for the purpose of this paper, which is focused on multispectral aspects, the 4 band data is more interesting.…”

“…In prior work, we explored the utility of NASA-WISE data for infrared thermometry 2 and a preliminary investigation by applying standard techniques of data science to the database 3 . In this current work, we explore further the utility of machine learning (ML) and traditional statistical techniques to characterize earth orbiting satellites and space debris.…”

“…Space debris and satellites in the GEO belt orbits are in thermal equilibrium with the sun, however the exact nature of that thermal equilibrium depends on the nature of the space object. Based on first principle models developed in prior work 3 , debris and satellites are expected to undergo thermal cycles reflecting the distance of the space object relative to the sun. In addition to illumination differences, space objects and debris possess different material properties, thermal capacity, reflectivity, and emissivity.…”

“…Prior related work have hinted at the differences in thermal signature observable in the NASA-WISE data 6 . Unpublished work has also explored these differences in the thermal properties of Satellites and Space Debris 7,3 , suggesting that the thermal differences can be used to delineate, or classify the space objects based solely on their measured thermal properties. The present work builds on those results with machine learning classification.…”