Simultaneous Multivariate Forecast of Space Weather Indices using Deep Neural Network Ensembles

Benson, Bernard; Brown, Edward M.; Bonasera, Stefano; Acciarini, Giacomo; Pérez-Hernández, Jorge A.; Sutton, E. K.; Jah, Moriba; Bridges, C.R.; Meng, J.; Baydin, Atılım Güneş

doi:10.48550/arxiv.2112.09051

Cited by 3 publications

(4 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The authors acknowledge that the SVR method is general and could use improvement, and that the addition of an input that represents general solar magnetic activity may help F 10 prediction, due to subtle changes in the solar magnetic field. Benson et al (2021) used an LSTM model to forecast various solar proxy and geomagnetic indices simultaneously. The authors considered the interaction between F 10 , F 30 , F 15 , geomagnetic indices, and solar imaging in forecasting of proxy values one solar rotation (27 days) in advance.…”

Section: Linear Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Probabilistic Solar Proxy Forecasting With Neural Network Ensembles

Daniell,

Mehta

2023

Space Weather

View full text Add to dashboard Cite

Space weather indices are used commonly to drive forecasts of thermosphere density, which affects objects in low‐Earth orbit (LEO) through atmospheric drag. One commonly used space weather proxy, F10.7cm, correlates well with solar extreme ultra‐violet (EUV) energy deposition into the thermosphere. Currently, the USAF contracts Space Environment Technologies (SET), which uses a linear algorithm to forecast F10.7cm. In this work, we introduce methods using neural network ensembles with multi‐layer perceptrons (MLPs) and long‐short term memory (LSTMs) to improve on the SET predictions. We make predictions only from historical F10.7cm values. We investigate data manipulation methods (backwards averaging and lookback) as well as multi step and dynamic forecasting. This work shows an improvement over the popular persistence and the operational SET model when using ensemble methods. The best models found in this work are ensemble approaches using multi step or a combination of multi step and dynamic predictions. Nearly all approaches offer an improvement, with the best models improving between 48% and 59% on relative MSE with respect to persistence. Other relative error metrics were shown to improve greatly when ensembles methods were used. We were also able to leverage the ensemble approach to provide a distribution of predicted values; allowing an investigation into forecast uncertainty. Our work found models that produced less biased predictions at elevated and high solar activity levels. Uncertainty was also investigated through the use of a calibration error score metric (CES), our best ensemble reached similar CES as other work.

show abstract

Section: Linear Methodsmentioning

confidence: 99%

“…There is only so much information that can be gathered from previous values alone, even using neural networks. The addition of solar disk images, suggested by Benson et al (2021) may be critical in improving longer term forecasts, due to the time lag between solar activity and thermosphere response.…”

Section: Limitations and Future Workmentioning

confidence: 99%

Probabilistic Solar Proxy Forecasting With Neural Network Ensembles

Daniell,

Mehta

2023

Space Weather

View full text Add to dashboard Cite

show abstract

“…An important model type for time-series forecasting is Long Short-Term Memory (LSTM), which was introduced by Hochreiter and Schmidhuber (1997). LSTM models have been used extensively in time-series forecasting problems such as stock market prediction (Bhandari et al, 2022), terrestrial weather forecasting (Karevan & Suykens, 2020), and the domain of space weather and forecasting; (Benson et al, 2021;Licata et al, 2021;Luo et al, 2022).…”

Section: Long Short-term Memory (Lstm)mentioning

confidence: 99%

Probabilistic Short‐Term Solar Driver Forecasting With Neural Network Ensembles

Daniell,

Mehta

2024

Space Weather

View full text Add to dashboard Cite

Space weather indices are used to drive forecasts of thermosphere density, which directly affects objects in low‐Earth orbit (LEO) through atmospheric drag force. A set of proxies and indices (drivers), F10.7, S10.7, M10.7, and Y10.7 are used as inputs by the JB2008, (https://doi.org/10.2514/6.2008‐6438) thermosphere density model. The United States Air Force (USAF) operational High Accuracy Satellite Drag Model (HASDM), relies on JB2008, (https://doi.org/10.2514/6.2008‐6438), and forecasts of solar drivers from a linear algorithm. We introduce methods using long short‐term memory (LSTM) model ensembles to improve over the current prediction method as well as a previous univariate approach. We investigate the usage of principal component analysis (PCA) to enhance multivariate forecasting. A novel method, referred to as striped sampling, is created to produce statistically consistent machine learning data sets. We also investigate forecasting performance and uncertainty estimation by varying the training loss function and by investigating novel weighting methods. Results show that stacked neural network model ensembles make multivariate driver forecasts which outperform the operational linear method. When using MV‐MLE (multivariate multi‐lookback ensemble), we see an improvement of RMSE for F10.7, S10.7, M10.7, and Y10.7 of 17.7%, 12.3%, 13.8%, 13.7% respectively, over the operational method. We provide the first probabilistic forecasting method for S10.7, M10.7, and Y10.7. Ensemble approaches are leveraged to provide a distribution of predicted values, allowing an investigation into robustness and reliability (R&R) of uncertainty estimates. Uncertainty was also investigated through the use of calibration error score (CES), with the MV‐MLE providing an average CES of 5.63%, across all drivers.

show abstract

“…The same group of researchers also developed a method to forecast with uncertainty estimation the input features for empirical models, combining solar data images with time-series information [105]. In 2020, SET (Space Enviroment Techonologies) made public 20 years of density data derived from HASDM, the model used by the United States Air Force [106].…”

Section: Thermospheric Density Mass Modelsmentioning

confidence: 99%

Machine Learning in Orbit Estimation: a Survey

Caldas¹,

Soares²

2022

Preprint

View full text Add to dashboard Cite

Since the late '50s, when the first artificial satellite was launched, the number of resident space objects (RSOs) has steadily increased. It is estimated that around 1 Million objects larger than 1 cm are currently orbiting the Earth, with only 30,000, larger than 10 cm, presently being tracked. To avert a chain reaction of collisions, termed Kessler Syndrome [1], it is indispensable to accurately track and predict space debris and satellites' orbit alike. Current physics-based methods have errors in the order of kilometres for 7 days predictions, which is insufficient when considering space debris that have mostly less than 1 meter. Typically, this failure is due to uncertainty around the state of the space object at the beginning of the trajectory, forecasting errors in environmental conditions such as atmospheric drag, as well as specific unknown characteristics such as mass or geometry of the RSO. Leveraging datadriven techniques, namely machine learning, the orbit prediction accuracy can be enhanced: by deriving unmeasured objects' characteristics, improving non-conservative forces' effects, and by the superior abstraction capacity that Deep Learning models have of modelling highly complex non-linear systems. In this survey, we provide an overview of the current work being done in this field.

show abstract

Simultaneous Multivariate Forecast of Space Weather Indices using Deep Neural Network Ensembles

Cited by 3 publications

References 0 publications

Probabilistic Solar Proxy Forecasting With Neural Network Ensembles

Probabilistic Solar Proxy Forecasting With Neural Network Ensembles

Probabilistic Short‐Term Solar Driver Forecasting With Neural Network Ensembles

Machine Learning in Orbit Estimation: a Survey

Contact Info

Product

Resources

About