Chunxin Yuan
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Chunxin Yuan has a total of 50 co-authors. Here are the top co-authors by number of shared publications:
- Edward R. Johnson
Department of Mathematics, University College London, London WC1E 6BT, UK
6 shared publications - Roger Grimshaw
Department of Mathematics, University College London, London WC1E 6BT, UK
6 shared publications - Xueen Chen
State Key Laboratory of Physical Oceanography (POL) Ocean University of China Qingdao China
6 shared publications - Zhan Wang
Institute of Mechanics, Chinese Academy of Sciences, Beijing 100190, People’s Republic of China, and School of Engineering Sciences, University of Chinese Academy of Sciences, Beijing 100049, People’s Republic of China
5 shared publications - Zhen Gao
School of Mathematical Sciences, Ocean University of China, Qingdao, People’s Republic of China
4 shared publications Yankun Gong
3 shared publications
Latest Publications
- 27 April 2026
- Published by American Physical Society (APS) in Physical Review Fluids
- Vol. 11 (4)
Generation and propagation of mode-1 and mode-2 internal waves over bottom topography in a three-layer system
- 27 April 2026
- Published by American Physical Society (APS) in Physical Review Fluids
- Vol. 11 (4)
Generation and propagation of mode-1 and mode-2 internal waves over bottom topography in a three-layer system
- 6 February 2026
- Published by Elsevier in Ocean Modelling
- Vol. 201
InWaveSR: Topography-aware super-resolution network for internal solitary waves
- 6 February 2026
- Published by Elsevier in Ocean Modelling
- Vol. 201
InWaveSR: Topography-aware super-resolution network for internal solitary waves
- 1 January 2026
- Published by AIP Publishing in Physics of Fluids
- Vol. 38 (1)
Ocean density reconstruction and forecasting using the Hankel dynamic mode decomposition-bidirectional long short-term memory model
Ocean density is a key physical variable in marine science. Integrating satellite remote sensing data with in situ observations enables comprehensive three-dimensional ocean density reconstruction and prediction. However, this task remains challenging for traditional interpolation techniques. To address this issue, we propose a reduced-order model that combines Hankel dynamic mode decomposition (HDMD) with a bidirectional long short-term memory (BiLSTM) network, thereby integrating dynamical systems theory with deep learning. HDMD captures the vertical evolutionary patterns of ocean density anomalies, enabling physically consistent interpolation and extrapolation across unobserved depths. The BiLSTM then integrates sea surface observations with HDMD-reconstructed subsurface profiles to achieve continuous three-dimensional reconstruction and temporal forecasting. Furthermore, the reduced-order formulation substantially improves computational efficiency. The proposed method is evaluated using simulation data from the Max Planck Institute Ocean Model, and results demonstrate that it achieves high estimation accuracy and consistently outperforms baseline models such as multilayer perceptrons. Moreover, the method preserves physical plausibility in vertical structure predictions by combining dynamical modeling with deep learning, offering a robust and interpretable framework for ocean state reconstruction.
- 1 January 2026
- Published by AIP Publishing in Physics of Fluids
- Vol. 38 (1)
Ocean density reconstruction and forecasting using the Hankel dynamic mode decomposition-bidirectional long short-term memory model
Ocean density is a key physical variable in marine science. Integrating satellite remote sensing data with in situ observations enables comprehensive three-dimensional ocean density reconstruction and prediction. However, this task remains challenging for traditional interpolation techniques. To address this issue, we propose a reduced-order model that combines Hankel dynamic mode decomposition (HDMD) with a bidirectional long short-term memory (BiLSTM) network, thereby integrating dynamical systems theory with deep learning. HDMD captures the vertical evolutionary patterns of ocean density anomalies, enabling physically consistent interpolation and extrapolation across unobserved depths. The BiLSTM then integrates sea surface observations with HDMD-reconstructed subsurface profiles to achieve continuous three-dimensional reconstruction and temporal forecasting. Furthermore, the reduced-order formulation substantially improves computational efficiency. The proposed method is evaluated using simulation data from the Max Planck Institute Ocean Model, and results demonstrate that it achieves high estimation accuracy and consistently outperforms baseline models such as multilayer perceptrons. Moreover, the method preserves physical plausibility in vertical structure predictions by combining dynamical modeling with deep learning, offering a robust and interpretable framework for ocean state reconstruction.
- 14 July 2025
- Published by Cambridge University Press (CUP) in Journal of Fluid Mechanics
- Vol. 1015, A3
A novel Kadomtsev–Petviashvili type model for nonlinear internal waves with horizontally two-dimensional shear currents and Earth’s rotation
Inspired by the need to theoretically understand the naturally occurring interactions between internal waves and mesoscale phenomena in the ocean, we derive a novel model equation from the primitive rotational Euler equations using the multi-scale asymptotic expansion method. By applying the classic balance $\epsilon =\mu ^2$ between nonlinearity (measured by $\epsilon$ ) and dispersion (measured by $\mu$ ), along with the assumption that variations in the transverse direction are of order $\mu$ , which is smaller than those in the propagation direction, we arrive at terms from the classic Kadomtsev–Petviashvili equation. However, when incorporating background shear currents in two horizontal dimensions and accounting for Earth’s rotation, we introduce three additional terms that, to the best of the authors’ knowledge, have not been addressed in the previous literature. Theoretical analyses and numerical results indicate that these three terms contribute to a tendency for propagation in the transverse direction and an overall variation in wave amplitudes. The specific effects of these terms can be estimated qualitatively based on the signs of the coefficients for each term and the characteristics of the initial waves. Finally, the potential shortcomings of this proposed equation are illuminated.- 14 July 2025
- Published by Cambridge University Press (CUP) in Journal of Fluid Mechanics
- Vol. 1015, A3
A novel Kadomtsev–Petviashvili type model for nonlinear internal waves with horizontally two-dimensional shear currents and Earth’s rotation
Inspired by the need to theoretically understand the naturally occurring interactions between internal waves and mesoscale phenomena in the ocean, we derive a novel model equation from the primitive rotational Euler equations using the multi-scale asymptotic expansion method. By applying the classic balance $\epsilon =\mu ^2$ between nonlinearity (measured by $\epsilon$ ) and dispersion (measured by $\mu$ ), along with the assumption that variations in the transverse direction are of order $\mu$ , which is smaller than those in the propagation direction, we arrive at terms from the classic Kadomtsev–Petviashvili equation. However, when incorporating background shear currents in two horizontal dimensions and accounting for Earth’s rotation, we introduce three additional terms that, to the best of the authors’ knowledge, have not been addressed in the previous literature. Theoretical analyses and numerical results indicate that these three terms contribute to a tendency for propagation in the transverse direction and an overall variation in wave amplitudes. The specific effects of these terms can be estimated qualitatively based on the signs of the coefficients for each term and the characteristics of the initial waves. Finally, the potential shortcomings of this proposed equation are illuminated.- 1 May 2025
- Published by Elsevier in Ocean Modelling
- Vol. 196
Oblique interactions of mode-1 and mode-2 internal solitary waves in the presence of variable topography
- 1 May 2025
- Published by Elsevier in Ocean Modelling
- Vol. 196
Oblique interactions of mode-1 and mode-2 internal solitary waves in the presence of variable topography
- 26 February 2025
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research: Oceans
- Vol. 130 (3)
Dynamics of Surface Cold Patch off the Korean Peninsula Cape: Driven by Upwelling
Surface cold patches (SCPs) are important hydrological phenomena in the Yellow Sea (YS). We investigate the dynamics behind the appearance of the SCP off the Korean Peninsula Cape during summer. The numerical model results revealed that upwelling is the primary mechanism of SCP formation in this region. On one hand, the 3D vorticity balance of the dynamic processes of the SCP indicated that the combined effect of the tides and the baroclinic pressure gradient enhances dissipation in the bottom boundary layer, leading to seawater convergence in the bottom layer west of the tidal front and divergence east of the tidal front, forming a stable frontal secondary circulation and resulting in the formation of upwelling. On the other hand, the offshore transport, which is triggered by the Jeju Warm Current and persists throughout the year in the strait, leads to a divergence of seawater within the region. The joint effect of the above two processes ultimately contributes to the generation of upwelling. The presence of upwelling forces the thermocline to be changed. Strong tidal upwelling triggers the upward ventilating of the thermocline, which can bring cold water to the surface, and the dome‐like upwarp of the thermocline isotherms caused by the current‐induced upwelling, together lead to the formation of the SCP.
- 26 February 2025
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research: Oceans
- Vol. 130 (3)
Dynamics of Surface Cold Patch off the Korean Peninsula Cape: Driven by Upwelling
Surface cold patches (SCPs) are important hydrological phenomena in the Yellow Sea (YS). We investigate the dynamics behind the appearance of the SCP off the Korean Peninsula Cape during summer. The numerical model results revealed that upwelling is the primary mechanism of SCP formation in this region. On one hand, the 3D vorticity balance of the dynamic processes of the SCP indicated that the combined effect of the tides and the baroclinic pressure gradient enhances dissipation in the bottom boundary layer, leading to seawater convergence in the bottom layer west of the tidal front and divergence east of the tidal front, forming a stable frontal secondary circulation and resulting in the formation of upwelling. On the other hand, the offshore transport, which is triggered by the Jeju Warm Current and persists throughout the year in the strait, leads to a divergence of seawater within the region. The joint effect of the above two processes ultimately contributes to the generation of upwelling. The presence of upwelling forces the thermocline to be changed. Strong tidal upwelling triggers the upward ventilating of the thermocline, which can bring cold water to the surface, and the dome‐like upwarp of the thermocline isotherms caused by the current‐induced upwelling, together lead to the formation of the SCP.
- Open Access
- 30 December 2024
- Published by American Geophysical Union (AGU) in Geophysical Research Letters
- Vol. 52 (1)
Large‐Scale Climate Patterns Modulate Intraseasonal Mesoscale Activities in the Bay of Bengal
By focusing on the intraseasonal mesoscale eddy kinetic energy, this study comprehensively distinguishes the influences of El Niño‐Southern Oscillation (ENSO) and Indian Ocean Dipole (IOD) on eddy activities in the Bay of Bengal. Altimetry records since 1993 show that intraseasonal mesoscale activities are enhanced in the middle (N) and southern bay approximately 6 months and 8 months after El Niño peaks, respectively, through the wave‐dominated approach. Meanwhile, influences from IOD manifest more rapidly and concentrate in the western bay, suggesting distinct mechanisms compared to those associated with ENSO. Further investigations reveal asymmetric effects linked to the phase of ENSO and IOD: positive ENSO events (El Niño) and negative IOD events are followed by enhanced intraseasonal mesoscale activities in the BoB, potentially triggered by the warmer background in the tropical Indian Ocean and the bay during these events, respectively. However, influences from the opposite phases are neglectable.
- 30 December 2024
- Published by American Geophysical Union (AGU) in Geophysical Research Letters
- Vol. 52 (1)
Open AccessLarge‐Scale Climate Patterns Modulate Intraseasonal Mesoscale Activities in the Bay of Bengal
By focusing on the intraseasonal mesoscale eddy kinetic energy, this study comprehensively distinguishes the influences of El Niño‐Southern Oscillation (ENSO) and Indian Ocean Dipole (IOD) on eddy activities in the Bay of Bengal. Altimetry records since 1993 show that intraseasonal mesoscale activities are enhanced in the middle (N) and southern bay approximately 6 months and 8 months after El Niño peaks, respectively, through the wave‐dominated approach. Meanwhile, influences from IOD manifest more rapidly and concentrate in the western bay, suggesting distinct mechanisms compared to those associated with ENSO. Further investigations reveal asymmetric effects linked to the phase of ENSO and IOD: positive ENSO events (El Niño) and negative IOD events are followed by enhanced intraseasonal mesoscale activities in the BoB, potentially triggered by the warmer background in the tropical Indian Ocean and the bay during these events, respectively. However, influences from the opposite phases are neglectable.
- Open Access
- 8 July 2024
- Published by Taylor & Francis in International Journal of Digital Earth
- Vol. 17 (1)
AttGAN: attention gated generative adversarial network for spatio-temporal super-resolution of ocean phenomena
- 8 July 2024
- Published by Taylor & Francis in International Journal of Digital Earth
- Vol. 17 (1)
Open AccessAttGAN: attention gated generative adversarial network for spatio-temporal super-resolution of ocean phenomena
- 27 June 2024
- Published by Elsevier in Ocean Modelling
- Vol. 190
Interaction of two oblique internal solitary waves in the presence of shoaling topography and rotation
- 27 June 2024
- Published by Elsevier in Ocean Modelling
- Vol. 190
Interaction of two oblique internal solitary waves in the presence of shoaling topography and rotation
- 20 February 2024
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research: Oceans
- Vol. 129 (2)
Physics‐Informed Deep Operator Learning Based on Reduced‐Order Modeling for Retrieving the Ocean Interior Density From the Surface
Exploring methods to reconstruct the ocean interior from surface data is a crucial focus in the study of ocean processes and phenomena due to the shortage of subsurface and deep‐sea data. Nonetheless, the existing methods predominantly concentrate on either data‐driven or dynamical methodologies, with limited exploration of integrating the strengths of both approaches. To combine the advantages of these two methods for reconstructing the subsurface density field from surface data, a novel dynamics‐constrained deep operator learning network based on reduced‐order model is proposed. Encoding the mean‐squared residuals of the reduced‐order equation along with the mean‐squared error between the network outputs and targets into the loss function effectively merges the dynamical and data constraints during the training process. This integration makes the network outputs and inputs approximately satisfy a specific form of the equation, allowing for interpretability, and once the network is well‐trained, rapid reconstruction evaluation can be performed. The reduced‐order equation is established by the Galerkin projection of quasi‐geostrophic equation onto the low‐dimensional subspace identified via reduced‐basis, which explains the vertical variation of ocean density. The developed model can tackle the challenge of directly measuring subsurface potential vorticity and predicting subsurface density. Evaluation is conducted using simulation data from the Max‐Planck‐Institute ocean model, indicating that it can offer precise estimations, outperforms the purely data‐driven algorithm presented in the paper and the interior plus surface quasi‐geostrophic method, and enables model sharing across different regions.
- 20 February 2024
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research: Oceans
- Vol. 129 (2)
Physics‐Informed Deep Operator Learning Based on Reduced‐Order Modeling for Retrieving the Ocean Interior Density From the Surface
Exploring methods to reconstruct the ocean interior from surface data is a crucial focus in the study of ocean processes and phenomena due to the shortage of subsurface and deep‐sea data. Nonetheless, the existing methods predominantly concentrate on either data‐driven or dynamical methodologies, with limited exploration of integrating the strengths of both approaches. To combine the advantages of these two methods for reconstructing the subsurface density field from surface data, a novel dynamics‐constrained deep operator learning network based on reduced‐order model is proposed. Encoding the mean‐squared residuals of the reduced‐order equation along with the mean‐squared error between the network outputs and targets into the loss function effectively merges the dynamical and data constraints during the training process. This integration makes the network outputs and inputs approximately satisfy a specific form of the equation, allowing for interpretability, and once the network is well‐trained, rapid reconstruction evaluation can be performed. The reduced‐order equation is established by the Galerkin projection of quasi‐geostrophic equation onto the low‐dimensional subspace identified via reduced‐basis, which explains the vertical variation of ocean density. The developed model can tackle the challenge of directly measuring subsurface potential vorticity and predicting subsurface density. Evaluation is conducted using simulation data from the Max‐Planck‐Institute ocean model, indicating that it can offer precise estimations, outperforms the purely data‐driven algorithm presented in the paper and the interior plus surface quasi‐geostrophic method, and enables model sharing across different regions.