The objective of this research is to improve the ability to track the orbits of space debris and thereby reduce the frequency of collisions. The approach is based on two scientific advances: 1) optimizing the scheduling of data transmission from a future constellation of orbiting Cubesats to ground stations located worldwide, and 2) using satellite data to improve models of the ionosphere and thermosphere, which in turn are used to improve estimates of atmospheric density. Intellectual Merit Robust capacity-constrained scheduling depends on fundamental research on optimization algorithms for nonlinear problems involving both discrete and continuous variables. This objective depends on advances in optimization theory and computational techniques. Model refinement depends on adaptive control algorithms, and can lead to fundamental advances for automatic control systems. These contributions provide new ideas and techniques that are broadly applicable to diverse areas of science and engineering. Broader Impacts Improving the ability to predict the trajectories of space debris can render the space environment safer in both the near term---by enhancing astronaut safety and satellite reliability---and the long term---by suppressing cascading collisions that could have a devastating impact on the usage of space. This project will impact real-world practice by developing techniques that are applicable to large-scale modeling and data collection, from weather prediction to Homeland Security. The research results will impact education through graduate and undergraduate research as well as through interdisciplinary modules developed for courses in space science, satellite engineering, optimization, and data-based modeling taught across multiple disciplines.
The near Earth space environment is increasingly crowded with satellites and space debris. To avoid collisions and the creation of new debris fields, it is important that active satellites be able to avoid collisions. One key aspect of doing this is being able to predict where satellites will be in the future. Since drag from the atmosphere on orbiting satellites plays a key role in predicting where satellites will be in the future, an ability to forecast the thin upper atmosphere accurately is important. Data assimilation is the tool that combines a prediction model with measurements to make a forecast. Ensemble data assimilation is a state-of-the-art method that produces an ensemble of forecasts that helps to estimate the effect of prediction model and measurement errors. In order to predict the state of the upper atmosphere, it is necessary to combine a data assimilation system with a prediction model and observations of the upper atmosphere, mostly taken from satellites. This project constructed an ensemble data assimilation and prediction system using the Data Assimilation Research Testbed (a comprehensive facility for ensemble data assimilation) and the GITM upper atmosphere model. The prediction system was tested and shown to be able to make improved predictions of the upper atmosphere that can help forecast satellite orbits. All the tools needed to do this are now publically available so that other scientists can use and improve them in order to better safeguard satellites. At the same time, scientists can learn more about the physics of the upper atmosphere so that better models can be developed in the future.
