Global Ionospheric TEC Forecasting for Geomagnetic Storm Time Using a Deep Learning‐Based Multi‐Model Ensemble Method

Ren, Xiaodong; Yang, Pengxin; Mei, Dengkui; Liu, Huan; Xu, Guozhen; Dong, Yue

doi:10.1029/2022sw003231

Cited by 8 publications

(3 citation statements)

References 30 publications

(36 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Training the model with imbalanced EIC data, the mean square error (MSE) and a focal loss (L4) are utilized as loss functions to be minimized, for comparatively studying the improvement of the imbalanced regression by different loss functions. Many other machine learning models dealt with the imbalance problem by manually selecting only extreme events to improve the performance, for example, (Hu et al., 2023; Ren et al., 2023). We avoid providing such a priori knowledge to the predictive model, such that our model is robust to provide continuous predictions/forecasts in the real operational environment.…”

Section: Model Descriptionmentioning

confidence: 99%

The Response of Ionospheric Currents to External Drivers Investigated Using a Neural Network‐Based Model

Cao,

Chu,

Bortnik

et al. 2023

Space Weather

View full text Add to dashboard Cite

A predictive model for the variation of ionospheric currents is of great scientific and practical importance to our modern industrial society. To study the response of ionospheric currents to external drivers including geomagnetic indices and solar radiation, we developed a feedforward neural network model trained on the Equivalent Ionospheric Current (EIC) data from 1st January 2007 to 31st December 2019. Due to the highly imbalanced nature of the ionospheric currents data, which means that the data of extreme events are much less than those of quiet times, we utilized different loss functions to improve the model performance. Our model demonstrates the potential to predict the active events of ionospheric currents reasonably well (e.g., EICs during substorms) within a timescale of a few minutes. Although the data used for training are measurements over the North American and Greenland sectors, our model is not only able to predict EICs within this region, but is also able to provide a promising out‐of‐sample prediction on a global scale.

show abstract

Section: Model Descriptionmentioning

confidence: 99%

The Response of Ionospheric Currents to External Drivers Investigated Using a Neural Network‐Based Model

Cao,

Chu,

Bortnik

et al. 2023

Space Weather

View full text Add to dashboard Cite

show abstract

“…They utilized seven different features to model 170 geomagnetic storm events. The results of their study showed that the DLMEM model was able to reduce the RMSE by an average of 43.6% compared to their previously presented model, Ion-LSTM [29]. LSTM is a powerful tool for handling temporal variations due to its memory retention capabilities, but it has limitations when dealing with problems that have both temporal and spatial characteristics.…”

Section: Introductionmentioning

confidence: 99%

Operational Forecasting of Global Ionospheric TEC Maps 1-, 2-, and 3-Day in Advance by ConvLSTM Model

Yang,

Huang,

Xia

et al. 2024

Remote Sensing

View full text Add to dashboard Cite

In this paper, we propose a global ionospheric total electron content (TEC) maps (GIM) prediction model based on deep learning methods that is both straightforward and practical, meeting the requirements of various applications. The proposed model utilizes an encoder-decoder structure with a Convolution Long Short-Term Memory (ConvLSTM) network and has a spatial resolution of 5° longitude and 2.5° latitude, with a time resolution of 1 h. We utilized the Center for Orbit Determination in Europe (CODE) GIM dataset for 18 years from 2002 to 2019, without requiring any other external input parameters, to train the ConvLSTM models for forecasting GIM 1, 2, and 3 days in advance. Using the CODE GIM data from 1 January 2020 to 31 December 2023 as the test dataset, the performance evaluation results show that the average root mean square errors (RMSE) for 1, 2 and 3 days of forecasts are 2.81 TECU, 3.16 TECU, and 3.41 TECU, respectively. These results show improved performance compared to the IRI-Plas model and CODE’s 1-day forecast product c1pg, and comparable to CODE’s 2-day forecast c2pg. The model’s predictions get worse as the intensity of the storm increases, and the prediction error of the model increases with the lead time.

show abstract

“…The Universitat Politècnica de Catalunya (UPC-ionSAT) GIMs forecasts for 6, 12, 18, 24 and 48 h in advance present another family of the Nearest-Neighbor (NN) model [ 15 ] used for the comparisons. There are other methods of GIMs forecast published in literature which are not used in the present analysis [ 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 ].…”

Section: Introductionmentioning

confidence: 99%

Ionospheric Weather at Two Starlink Launches during Two-Phase Geomagnetic Storms

Gulyaeva,

Hernández-Pajares,

Stanislawska

2023

Sensors

View full text Add to dashboard Cite

The launch of a series of Starlink internet satellites on 3 February 2022 (S-36), and 7 July 2022 (S-49), coincided with the development of two-phase geomagnetic storms. The first launch S-36 took place in the middle of the moderate two-phase space weather storm, which induced significant technological consequences. After liftoff on 3 February at 18:13 UT, all Starlink satellites reached an initial altitude of 350 km in perigee and had to reach an altitude of ~550 km after the maneuver. However, 38 of 49 launched spacecrafts did not reach the planned altitude, left orbit due to increased drag and reentered the atmosphere on 8 February. A geomagnetic storm on 3–4 February 2022 has increased the density of the neutral atmosphere up to 50%, increasing drag of the satellites and dooming most of them. The second launch of S-49 at 13:11 UT on 7 July 2022 was successful at the peak of the two-phase geomagnetic storm. The global ionospheric maps of the total electron content (GIM-TEC) have been used to produce the ionospheric weather GIM-W index maps and Global Electron Content (GEC). We observed a GEC increment from 10 to 24% for the storm peak after the Starlink launch at both storms, accompanying the neutral density increase identified earlier. GIM-TEC maps are available with a lag (delay) of 1–2 days (real-time GIMs have a lag less than 15 min), so the GIMs forecast is required by the time of the launch. Comparisons of different GIMs forecast techniques are provided including the Center for Orbit Determination in Europe (CODE), Beijing (BADG and CASG) and IZMIRAN (JPRG) 1- and 2-day forecasts, and the Universitat Politecnica de Catalunya (UPC-ionSAT) forecast for 6, 12, 18, 24 and 48 h in advance. We present the results of the analysis of evolution of the ionospheric parameters during both events. The poor correspondence between observed and predicted GIM-TEC and GEC confirms an urgent need for the industry–science awareness of now-casting/forecasting/accessibility of GIM-TECs during the space weather events.

show abstract

Global Ionospheric TEC Forecasting for Geomagnetic Storm Time Using a Deep Learning‐Based Multi‐Model Ensemble Method

Cited by 8 publications

References 30 publications

The Response of Ionospheric Currents to External Drivers Investigated Using a Neural Network‐Based Model

The Response of Ionospheric Currents to External Drivers Investigated Using a Neural Network‐Based Model

Operational Forecasting of Global Ionospheric TEC Maps 1-, 2-, and 3-Day in Advance by ConvLSTM Model

Ionospheric Weather at Two Starlink Launches during Two-Phase Geomagnetic Storms

Contact Info

Product

Resources

About