2020
Stevenson, Emma; Rodriguez-Fernandez, Victor; Minisci, Edmondo; Camacho, David
A deep learning approach to space weather proxy forecasting for orbital prediction Proceedings Article
In: 71st International Astronautical Congress (IAC), The CyberSpace Edition, 2020.
Abstract | Links | BibTeX | Tags: Deep Learning, Solar Radio Flux, Space Weather, Time Series Forecasting
@inproceedings{stevenson2020_iac,
title = {A deep learning approach to space weather proxy forecasting for orbital prediction},
author = {Emma Stevenson and Victor Rodriguez-Fernandez and Edmondo Minisci and David Camacho},
url = {http://oa.upm.es/64345/},
year = {2020},
date = {2020-10-01},
booktitle = {71st International Astronautical Congress (IAC)},
address = {The CyberSpace Edition},
abstract = {The effect of atmospheric drag on spacecraft dynamics is considered one of the predominant sources of uncertainty in Low Earth Orbit. These effects are characterised in part by the atmospheric density, a quantity highly correlated to space weather. Current atmosphere models typically account for this through proxy indices such as the F10.7, but with variations in solar radio flux forecasts leading to significant orbit differences over just a few days, prediction of these quantities is a limiting factor in the accurate estimation of future drag conditions, and consequently orbital prediction. This has fundamental implications both in the short term, in the day-to-day management of operational spacecraft, and in the mid-to-long term, in determining satellite orbital lifetime. In this work, a novel deep residual architecture for univariate time series forecasting, N-BEATS, is employed for the prediction of the F10.7 solar proxy on the days-ahead timescales relevant to space operations. This untailored, pure deep learning approach has recently achieved state-of-the-art performance in time series forecasting competitions, outperforming well-established statistical, as well as statistical hybrid models, across a range of domains. The approach was found to be effective in single point forecasting up to 27-days ahead, and was additionally extended to produce forecast uncertainty estimates using deep ensembles. These forecasts were then compared to a persistence baseline and two operationally available forecasts: one statistical (provided by BGS, ESA), and one multi-flux neural network (by CLS, CNES). It was found that the N-BEATS model systematically outperformed the baseline and statistical approaches, and achieved an improved or similar performance to the multi-flux neural network approach despite only learning from a single variable},
keywords = {Deep Learning, Solar Radio Flux, Space Weather, Time Series Forecasting},
pubstate = {published},
tppubtype = {inproceedings}
}
The effect of atmospheric drag on spacecraft dynamics is considered one of the predominant sources of uncertainty in Low Earth Orbit. These effects are characterised in part by the atmospheric density, a quantity highly correlated to space weather. Current atmosphere models typically account for this through proxy indices such as the F10.7, but with variations in solar radio flux forecasts leading to significant orbit differences over just a few days, prediction of these quantities is a limiting factor in the accurate estimation of future drag conditions, and consequently orbital prediction. This has fundamental implications both in the short term, in the day-to-day management of operational spacecraft, and in the mid-to-long term, in determining satellite orbital lifetime. In this work, a novel deep residual architecture for univariate time series forecasting, N-BEATS, is employed for the prediction of the F10.7 solar proxy on the days-ahead timescales relevant to space operations. This untailored, pure deep learning approach has recently achieved state-of-the-art performance in time series forecasting competitions, outperforming well-established statistical, as well as statistical hybrid models, across a range of domains. The approach was found to be effective in single point forecasting up to 27-days ahead, and was additionally extended to produce forecast uncertainty estimates using deep ensembles. These forecasts were then compared to a persistence baseline and two operationally available forecasts: one statistical (provided by BGS, ESA), and one multi-flux neural network (by CLS, CNES). It was found that the N-BEATS model systematically outperformed the baseline and statistical approaches, and achieved an improved or similar performance to the multi-flux neural network approach despite only learning from a single variable
2018
Stevenson, Emma; Virgili, Benjamin Bastida; Lemmens, Stijn; Reihs, Benedikt
Comparison of Atmosphere Models for Atmospheric Predictions Proceedings Article
In: 4th International Space Debris Re-entry Workshop, Darmstadt, Germany, 2018.
Links | BibTeX | Tags: Space Weather
@inproceedings{stevenson2018_reentry,
title = {Comparison of Atmosphere Models for Atmospheric Predictions},
author = {Emma Stevenson and Benjamin Bastida Virgili and Stijn Lemmens and Benedikt Reihs},
url = {https://conference.sdo.esoc.esa.int/proceedings/list},
year = {2018},
date = {2018-02-01},
urldate = {2018-02-01},
booktitle = {4th International Space Debris Re-entry Workshop},
address = {Darmstadt, Germany},
keywords = {Space Weather},
pubstate = {published},
tppubtype = {inproceedings}
}
2017
Virgili, Benjamin Bastida; Lemmens, Stijn; Stevenson, Emma; Reihs, Benedikt
Statistical comparison of ISO recommended thermosphere models and space weather proxy forecasting on re-entry predictions Proceedings Article
In: 68th International Astronautical Congress (IAC), Adelaide, Australia, 2017.
BibTeX | Tags: Space Weather
@inproceedings{stevenson2017_iac,
title = {Statistical comparison of ISO recommended thermosphere models and space weather proxy forecasting on re-entry predictions},
author = {Benjamin Bastida Virgili and Stijn Lemmens and Emma Stevenson and Benedikt Reihs},
year = {2017},
date = {2017-09-01},
urldate = {2017-09-01},
booktitle = {68th International Astronautical Congress (IAC)},
address = {Adelaide, Australia},
keywords = {Space Weather},
pubstate = {published},
tppubtype = {inproceedings}
}