Piyush Mehta, West Virginia University; Richard Linares, Massachusetts Institute of Technology
Keywords: Drag, Upper Atmosphere, Reduced Order Modeling, Data Assimilation, Radar Observation
Abstract:
Space flight is entering a period of renaissance, with considerable change in the perception of humanitys role in space. Recently, SpaceX and OneWeb have proposed large satellite constellations in Low Earth Orbit (LEO) that are expected to increase the number of satellites in LEO multiple-fold. These constellations could revolutionize the telecommunication industry by providing complete global internet coverage. The economic gains of completely connecting rural areas and developing nations cannot be understated, however, the current space infrastructure is not capable of handling such a dramatic increase in the number of active satellites. Therefore, there is a critical need for new solutions to the problem of Space Traffic Management (STM) and Space Situational Awareness (SSA).
Atmospheric drag remains the largest source of uncertainty in orbit prediction and conjunction assessments for collision probabilities in Low Earth Orbit. The uncertainty stems largely from inaccurate models of the upper atmosphere. Recently, a new framework that uses a Kalman filter for sequential assimilation of data into a dynamic quasi-physical reduced order model has been developed. The framework provides a transformative technology for real-time global state update of the upper atmosphere that will facilitate STM and SSA. In this work, we present a demonstration of the technology using radar satellite tracking data from LeoLabs commercial sensor network as measurements and validated with accurate GPS-derived ephemerides from Planet Labs in-orbit constellation of over 100 satellites.
Date of Conference: September 11-14, 2018
Track: Poster