Shunzu Gao
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Shunzu Gao has a total of 19 co-authors. Here are the top co-authors by number of shared publications:
- Chao Xiong
School of Earth and Space Science and Technology Wuhan University Wuhan China
6 shared publications Qian Pan
4 shared publications- Hong Zhang
School of Earth and Space Science and Technology Wuhan University Wuhan China
3 shared publications Weijia Zhan
Space Weather Technology Research and Education Center University of Colorado Boulder Boulder CO USA
3 shared publications- Yuhao Zheng
School of Earth and Space Science and Technology Wuhan University Wuhan China
3 shared publications - Yuyang Huang
School of Earth and Space Science and Technology Wuhan University Wuhan China
3 shared publications
Latest Publications
- 9 January 2026
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research: Space Physics
- Vol. 131 (1)
Merging and Dissipation of Intense Equatorial Plasma Bubbles With Large‐Scale Mid‐Latitude Low Electron Density Regions During the 23–24 April 2023 Geomagnetic Storm
The extension of the post‐sunset equatorial plasma bubbles (EPBs) to mid‐latitudes during geomagnetic storms has been widely reported. However, previous research has primarily focused on the initiation and spatial distribution of these EPBs, while their decay processes remain underexplored. In this study, we provide observations of the evolution and dissipation of intense EPBs during the geomagnetic storm on 23–24 April 2023, using measurements from the Global‐scale Observations of the Limb and Disk (GOLD), Swarm, and COSMIC‐2 satellites. During the storm main phase, intense EPBs with zonal extents of ∼2–3° and latitudinal extents reaching up to ±35° magnetic latitude (Mlat) were observed between 15°W and 5°W. Simultaneously, very low electron density regions, spanning approximately from ±30° to ±45° Mlat and 420–500 km in altitude, were observed in the mid‐latitudes. During the storm, the EPBs were observed to extend gradually poleward and subsequently encountered the low electron density regions. This encounter process, along with the storm‐caused electric field dynamics, led to a gradual weakening and diffusion of the EPBs. The airglow images from GOLD showed that EPBs evolved from coherent depletions into structurally diffused forms with embedded filamentary features and eventually merged into the background. Notably, a hemispheric asymmetry in EPB dissipation was observed. In the Northern Hemisphere, EPBs tended to retain their morphology for a longer time, while in the Southern Hemisphere, more rapid dissipation occurred, likely due to differing background plasma conditions and magnetic field configurations in the two hemispheres. These findings reveal a previously uncharacterized feature of EPB dissipation during geomagnetic storms.
- 9 January 2026
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research: Space Physics
- Vol. 131 (1)
Merging and Dissipation of Intense Equatorial Plasma Bubbles With Large‐Scale Mid‐Latitude Low Electron Density Regions During the 23–24 April 2023 Geomagnetic Storm
The extension of the post‐sunset equatorial plasma bubbles (EPBs) to mid‐latitudes during geomagnetic storms has been widely reported. However, previous research has primarily focused on the initiation and spatial distribution of these EPBs, while their decay processes remain underexplored. In this study, we provide observations of the evolution and dissipation of intense EPBs during the geomagnetic storm on 23–24 April 2023, using measurements from the Global‐scale Observations of the Limb and Disk (GOLD), Swarm, and COSMIC‐2 satellites. During the storm main phase, intense EPBs with zonal extents of ∼2–3° and latitudinal extents reaching up to ±35° magnetic latitude (Mlat) were observed between 15°W and 5°W. Simultaneously, very low electron density regions, spanning approximately from ±30° to ±45° Mlat and 420–500 km in altitude, were observed in the mid‐latitudes. During the storm, the EPBs were observed to extend gradually poleward and subsequently encountered the low electron density regions. This encounter process, along with the storm‐caused electric field dynamics, led to a gradual weakening and diffusion of the EPBs. The airglow images from GOLD showed that EPBs evolved from coherent depletions into structurally diffused forms with embedded filamentary features and eventually merged into the background. Notably, a hemispheric asymmetry in EPB dissipation was observed. In the Northern Hemisphere, EPBs tended to retain their morphology for a longer time, while in the Southern Hemisphere, more rapid dissipation occurred, likely due to differing background plasma conditions and magnetic field configurations in the two hemispheres. These findings reveal a previously uncharacterized feature of EPB dissipation during geomagnetic storms.
- 18 December 2025
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research: Space Physics
- Vol. 130 (12)
The Retrieving of Ionospheric Upper Transition Height Over Low Latitude From C/NOFS Ion Composition Measurements
The upper transition height (UTH), defined as the altitude where the dominant ion changes from O + to light‐ions, is a key parameter in the topside ionosphere. In this study, we propose an innovative approach to retrieve ionospheric UTH over low latitudes by integrating the ion composition measurements from the C/NOFS satellite with the deep learning ionosphere model (DLIM). Compared to the direct UTH observations when the O + composition is 0.5, this approach can retrieve UTH when the O + composition ranges from 0.3 to 0.7. Our approach increases the number of available UTH measurements from approximately 92,100 to 1,075,000, which allows more comprehensive local time (LT) and magnetic latitude coverage. The proposed approach was validated using observations from the incoherent scatter radar (ISR) located in Jicamarca. The UTH between retrieved and measured from ISR shows good agreement, with the average value and standard deviation of discrepancy being −0.32 and 5.81 km, respectively. The retrieved UTH exhibits typical local time, seasonal and magnetic latitudinal dependence. Comparison results show that during low solar flux conditions, the retrieved UTH showed excellent consistency with the measured UTH with a relative error of less than 3%. Further analysis revealed a positive correlation between the UTH and the F10.7 index, which is stronger than results from directly measured UTH and ISR observations. The rate of change of retrieved UTH with respect to the F10.7 index varies with local time and can reach up 9.15 km/sfu at noontime (12:00 LT), which could be linked to the enhancement of E×B drifts.
- 18 December 2025
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research: Space Physics
- Vol. 130 (12)
The Retrieving of Ionospheric Upper Transition Height Over Low Latitude From C/NOFS Ion Composition Measurements
The upper transition height (UTH), defined as the altitude where the dominant ion changes from O + to light‐ions, is a key parameter in the topside ionosphere. In this study, we propose an innovative approach to retrieve ionospheric UTH over low latitudes by integrating the ion composition measurements from the C/NOFS satellite with the deep learning ionosphere model (DLIM). Compared to the direct UTH observations when the O + composition is 0.5, this approach can retrieve UTH when the O + composition ranges from 0.3 to 0.7. Our approach increases the number of available UTH measurements from approximately 92,100 to 1,075,000, which allows more comprehensive local time (LT) and magnetic latitude coverage. The proposed approach was validated using observations from the incoherent scatter radar (ISR) located in Jicamarca. The UTH between retrieved and measured from ISR shows good agreement, with the average value and standard deviation of discrepancy being −0.32 and 5.81 km, respectively. The retrieved UTH exhibits typical local time, seasonal and magnetic latitudinal dependence. Comparison results show that during low solar flux conditions, the retrieved UTH showed excellent consistency with the measured UTH with a relative error of less than 3%. Further analysis revealed a positive correlation between the UTH and the F10.7 index, which is stronger than results from directly measured UTH and ISR observations. The rate of change of retrieved UTH with respect to the F10.7 index varies with local time and can reach up 9.15 km/sfu at noontime (12:00 LT), which could be linked to the enhancement of E×B drifts.
- 9 December 2025
- Published by Elsevier in Advances In Space Research
- Vol. 78 (1) , 344-360
A global neural network model of 3D ionospheric electron density based on COSMIC radio occultation measurements
- 9 December 2025
- Published by Elsevier in Advances In Space Research
- Vol. 78 (1) , 344-360
A global neural network model of 3D ionospheric electron density based on COSMIC radio occultation measurements
- 20 November 2025
- Published by Elsevier in Advances In Space Research
- Vol. 78 (1) , 361-373
Forecasting ionospheric foF2 based on a machine learning CNN-BiLSTM-MHA model
- 20 November 2025
- Published by Elsevier in Advances In Space Research
- Vol. 78 (1) , 361-373
Forecasting ionospheric foF2 based on a machine learning CNN-BiLSTM-MHA model
- Open Access
- 13 March 2025
- Published by American Geophysical Union (AGU) in Space Weather
- Vol. 23 (3)
Interpretable Machine Learning for Thermospheric Mass Density Modeling Using GRACE/GRACE‐FO Satellite Data
With rapid development of artificial intelligence technology, machine learning has been widely applied to the thermospheric mass density (TMD) modeling. In this study we propose a machine‐learning approach, the bidirectional gated recurrent unit with multi‐head attention mechanism (BGMA), for modeling and predicting the TMD, based on the Gravity Recovery and Climate Experiment (GRACE) satellite data. GRACE data spanning over one solar cycle provide a valuable opportunity to explore altitude and solar activity dependencies in TMD. We selected 11 key parameters to construct the model, the correlation coefficient (R2) between the model's predictions and satellite observations is 0.944, significantly outperforming the 0.893 of the latest version of the Naval Research Laboratory Mass Spectrometer and Incoherent Scatter Radar Extended Model (NRLMSIS‐2.0). In addition, the TMD measurements from GRACE Follow‐On satellite served as an independent test to evaluate the BGMA model's generalization, yielding an R2 of 0.924, underscoring the model's robustness. A critical aspect of our work is minimizing the number of input parameters while maintaining high prediction accuracy. And the interpretability analysis of the input parameters using the Shapley additive explanation algorithm has been applied, revealing that the altitude, solar activity index P10.7 and solar zenith angle are the three most influential parameters affecting TMD variations at GRACE satellite. When using just these three parameters, the R2 still reaches 0.842. Additionally, our model demonstrated robust performance over varying prediction durations, with R2 exceeding 0.800 for predictions extending up to 1 hr. These results highlight the BGMA model's effectiveness in accurately predicting TMD.
- 13 March 2025
- Published by American Geophysical Union (AGU) in Space Weather
- Vol. 23 (3)
Open AccessInterpretable Machine Learning for Thermospheric Mass Density Modeling Using GRACE/GRACE‐FO Satellite Data
With rapid development of artificial intelligence technology, machine learning has been widely applied to the thermospheric mass density (TMD) modeling. In this study we propose a machine‐learning approach, the bidirectional gated recurrent unit with multi‐head attention mechanism (BGMA), for modeling and predicting the TMD, based on the Gravity Recovery and Climate Experiment (GRACE) satellite data. GRACE data spanning over one solar cycle provide a valuable opportunity to explore altitude and solar activity dependencies in TMD. We selected 11 key parameters to construct the model, the correlation coefficient (R2) between the model's predictions and satellite observations is 0.944, significantly outperforming the 0.893 of the latest version of the Naval Research Laboratory Mass Spectrometer and Incoherent Scatter Radar Extended Model (NRLMSIS‐2.0). In addition, the TMD measurements from GRACE Follow‐On satellite served as an independent test to evaluate the BGMA model's generalization, yielding an R2 of 0.924, underscoring the model's robustness. A critical aspect of our work is minimizing the number of input parameters while maintaining high prediction accuracy. And the interpretability analysis of the input parameters using the Shapley additive explanation algorithm has been applied, revealing that the altitude, solar activity index P10.7 and solar zenith angle are the three most influential parameters affecting TMD variations at GRACE satellite. When using just these three parameters, the R2 still reaches 0.842. Additionally, our model demonstrated robust performance over varying prediction durations, with R2 exceeding 0.800 for predictions extending up to 1 hr. These results highlight the BGMA model's effectiveness in accurately predicting TMD.
- Open Access
- 11 March 2025
- Published by American Geophysical Union (AGU) in Space Weather
- Vol. 23 (3)
Predicting Equatorial Spread F at JICAMARCA Sector Via Supervised Machine Learning
The nighttime equatorial spread‐F (ESF) occurrence is essential to cause disturbances of the trans‐ionospheric radio wave signals. Previous research has successfully modeled the probability or growth rate of ESF, but remains deficient in distinguishing their typology and duration, which are also critical for predicting their influences on radio wave propagation. In this paper, an ensemble learning approach is proposed to predict the occurrence and typology of ESF, based on over one solar cycle observations of coherent scattering radar (CSR) located at Jicamarca. The training data set cover from 2000 to 2016, and the observations in 2017 are chosen as testing data set. The model inputs include local time (LT), day of year (DoY), height, solar flux index (P10.7), geomagnetic activity related indices (Kp, SYMH, AE), the true height of the ionospheric F layer (HF), the vertical plasma drift velocity of the ionospheric F layer, and the occurrence rate of ESF from the previous 3 hr. By using the Shapley Additive exPlanations (SHAP) tool analysis, it shows that the LT is the most influential parameter, and the P10.7 contributes more than the geomagnetic indices, to predict the occurrence of ESF. An important advantage of our model is that it can present different types of ESF by reconstructing their duration versus height distributions. Comparison results between prediction and observation suggest that the model predicts all bins with an accuracy reaching 81.2%, and it is able to predict the daily occurrence of ESF with accuracy of about 90.1%.
- 11 March 2025
- Published by American Geophysical Union (AGU) in Space Weather
- Vol. 23 (3)
Open AccessPredicting Equatorial Spread F at JICAMARCA Sector Via Supervised Machine Learning
The nighttime equatorial spread‐F (ESF) occurrence is essential to cause disturbances of the trans‐ionospheric radio wave signals. Previous research has successfully modeled the probability or growth rate of ESF, but remains deficient in distinguishing their typology and duration, which are also critical for predicting their influences on radio wave propagation. In this paper, an ensemble learning approach is proposed to predict the occurrence and typology of ESF, based on over one solar cycle observations of coherent scattering radar (CSR) located at Jicamarca. The training data set cover from 2000 to 2016, and the observations in 2017 are chosen as testing data set. The model inputs include local time (LT), day of year (DoY), height, solar flux index (P10.7), geomagnetic activity related indices (Kp, SYMH, AE), the true height of the ionospheric F layer (HF), the vertical plasma drift velocity of the ionospheric F layer, and the occurrence rate of ESF from the previous 3 hr. By using the Shapley Additive exPlanations (SHAP) tool analysis, it shows that the LT is the most influential parameter, and the P10.7 contributes more than the geomagnetic indices, to predict the occurrence of ESF. An important advantage of our model is that it can present different types of ESF by reconstructing their duration versus height distributions. Comparison results between prediction and observation suggest that the model predicts all bins with an accuracy reaching 81.2%, and it is able to predict the daily occurrence of ESF with accuracy of about 90.1%.
- 4 December 2024
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research: Space Physics
- Vol. 129 (12)
Altitudinal Variation of O+ Scale Height at the Equatorial Topside Ionosphere
Altitude variation of the topside ionospheric electron density or its scale height has been widely investigated in the past. However, as the oxygen ion (O+) is an important indicator for separating the topside ionosphere and plasmasphere, the altitude variation of O+, which has not been well investigated, is crucial to understand the topside ionosphere. In this study, we provided analysis on how the O+ scale height varies with altitude under different solar and geomagnetic activities, by using 12‐year measurements from the incoherent scatter radar (ISR) located at Jicamarca. Constant scale height Chapman (CSC) as well as Linearly Varying Chapman (LVC) functions are used to reconstruct the O+ profile. The corresponding scale heights of O+ based on both approaches have been compared. The O+ profile derived from LVC function shows better agreement with the ISR measurements than that from CSC function. We found that the O+ scale height increases with increasing solar/geomagnetic activity, and its height gradient varies significantly with local time, reaching a maximum of 0.05 at sunrise (around 06:00 local time) and a minimum of about −0.08 at noon (around 12:00 LT). We further investigated possible drivers causing the O+ scale height variations at topside ionosphere, based on simulations from the SAMI2 physics‐based model. The model results show that the solar extreme ultra‐violet (EUV) radiation plays a key role in the positive gradient of O+ scale height observed around sunrise, while the vertical plasma drift caused by E × B significantly contributes to the negative gradient observed around noon.
- 4 December 2024
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research: Space Physics
- Vol. 129 (12)
Altitudinal Variation of O+ Scale Height at the Equatorial Topside Ionosphere
Altitude variation of the topside ionospheric electron density or its scale height has been widely investigated in the past. However, as the oxygen ion (O+) is an important indicator for separating the topside ionosphere and plasmasphere, the altitude variation of O+, which has not been well investigated, is crucial to understand the topside ionosphere. In this study, we provided analysis on how the O+ scale height varies with altitude under different solar and geomagnetic activities, by using 12‐year measurements from the incoherent scatter radar (ISR) located at Jicamarca. Constant scale height Chapman (CSC) as well as Linearly Varying Chapman (LVC) functions are used to reconstruct the O+ profile. The corresponding scale heights of O+ based on both approaches have been compared. The O+ profile derived from LVC function shows better agreement with the ISR measurements than that from CSC function. We found that the O+ scale height increases with increasing solar/geomagnetic activity, and its height gradient varies significantly with local time, reaching a maximum of 0.05 at sunrise (around 06:00 local time) and a minimum of about −0.08 at noon (around 12:00 LT). We further investigated possible drivers causing the O+ scale height variations at topside ionosphere, based on simulations from the SAMI2 physics‐based model. The model results show that the solar extreme ultra‐violet (EUV) radiation plays a key role in the positive gradient of O+ scale height observed around sunrise, while the vertical plasma drift caused by E × B significantly contributes to the negative gradient observed around noon.