Semi-autonomous Target Selection for Orbital Remote Sensing Platforms as applied to the Hyper-Angular Rainbow Polarimeter (HARP) CubeSat

Abstract:

Small satellites for orbital remote sensing are a powerful new tool which will allow a wider variety of measurement techniques to be applied to the Earth system. One difficulty these sorts of instruments will face though is limited operation time due to small onboard memory, difficulty dispersing heat or maintaining power during continuous operation, and limited opportunities for data downloads. These are all problems which have affected the recently launched HARP CubeSat. HARP is a wide field-of-view polarimetric imager which seeks to determine cloud, aerosol, and surface properties of the Earth from space in the linearly polarized radiometric domain, but its operation is limited to, at best, a single 5 to 10 minute capture sequence each day. As such, we have developed an orbital propagator around the SGP4 model to predict the HARP orbit and have performed a virtual geolocation which provides all solar/viewing geometry angles of a potential capture over a week in advance. This is to prepare for, and determine potential targets from which we will pick the most scientifically applicable. The determination of these potential targets currently is done by a domain expert who must sit down to apply their knowledge to the region and scattering geometry for the entire orbital period of interest. We will present a method to improve this towards a semi-automated method producing a “Target Score” algorithm utilizing globally averaged level 2/3 products of aerosol thickness, cloud thickness/phase, land-type, land/cloud masking, and co-location with other orbital platforms. This score allows a potential target location to be ranked automatically and only needs to be finalized by the domain expert. Co-location events in particular will be explored, as they are relevant to HARP for validation and vicarious calibration of the payload. The score calculation, and general algorithm, will be presented so as to make a tool that is general enough to provide a framework for target selection of other small satellites, as well as to facilitate better the joining of multiple satellite platforms for validation and new science.