Haijin Cao
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
Haijin Cao has a total of 54 co-authors. Here are the top co-authors by number of shared publications:
Xiangzhou Song
6 shared publicationsZhiyou Jing
5 shared publications- Zhiyou Jing
State Key Laboratory of Tropical Oceanography South China Sea Institute of Oceanology Chinese Academy of Sciences Guangzhou China
5 shared publications - Baylor Fox-Kemper
Department of Earth, Environmental, and Planetary Sciences, Brown University, Providence, RI, USA
4 shared publications - Bo Qiu
Department of Oceanography, University of Hawaii at Manoa, Honolulu, HI, USA
3 shared publications Baylor Fox-Kemper
Department of Earth, Environmental, and Planetary Sciences, Brown University
2 shared publications
Latest Publications
- Open Access
- 23 October 2026
- Published by American Association for the Advancement of Science (AAAS) in Science Advances
- Vol. 12 (43) , eaef8067
Direct observations of submesoscale turbulence driving mesoscale eddy leakage
Understanding oceanic transport and material fate has long relied on observations that primarily resolve mesoscale, largely geostrophic flows, leaving the role of submesoscale turbulence comparatively unobserved and poorly understood. Here, using new high-resolution drifter-cluster experiments with closest initial spacings below 1 km, we provide direct Lagrangian observations of how submesoscale turbulence modifies mesoscale eddy coherence and transport pathways. The clustered drifters resolve intense, transient ageostrophic motions that are fundamentally absent from geostrophic descriptions of eddy behavior. Analyzed in an eddy-centered reference frame, the observations reveal pronounced submesoscale-driven radial dispersion that facilitates substantial water-mass exchange between the eddy core and its periphery in both the mixed layer and the subsurface. Quantitative estimates indicate submesoscale horizontal diffusivity of O (10 to 100) m 2 /s and a nonlocal dispersion regime across submesoscales. These results challenge the traditional view of mesoscale eddies as coherent structures containing a persistent water mass, with broad implications for interpreting and predicting the transport and fate of tracers in the ocean.
- 23 October 2026
- Published by American Association for the Advancement of Science (AAAS) in Science Advances
- Vol. 12 (43) , eaef8067
Open AccessDirect observations of submesoscale turbulence driving mesoscale eddy leakage
Understanding oceanic transport and material fate has long relied on observations that primarily resolve mesoscale, largely geostrophic flows, leaving the role of submesoscale turbulence comparatively unobserved and poorly understood. Here, using new high-resolution drifter-cluster experiments with closest initial spacings below 1 km, we provide direct Lagrangian observations of how submesoscale turbulence modifies mesoscale eddy coherence and transport pathways. The clustered drifters resolve intense, transient ageostrophic motions that are fundamentally absent from geostrophic descriptions of eddy behavior. Analyzed in an eddy-centered reference frame, the observations reveal pronounced submesoscale-driven radial dispersion that facilitates substantial water-mass exchange between the eddy core and its periphery in both the mixed layer and the subsurface. Quantitative estimates indicate submesoscale horizontal diffusivity of O (10 to 100) m 2 /s and a nonlocal dispersion regime across submesoscales. These results challenge the traditional view of mesoscale eddies as coherent structures containing a persistent water mass, with broad implications for interpreting and predicting the transport and fate of tracers in the ocean.
- Open Access
- 23 October 2026
- Published by American Geophysical Union (AGU) in Geophysical Research Letters
- Vol. 53 (20)
Isopycnal Pathway for Small‐Particle Export in a Decaying Cyclonic Eddy Revealed by a BGC‐Argo Float
Mesoscale eddies influence carbon export through production, subduction, mixing, and particle redistribution. In this study, a BGC‐Argo float observed a deep particle layer in the twilight zone during the decay of a cyclonic eddy in the Northwest Pacific. The layer contained pigment‐depleted particles, remained confined to a narrow density range near 26 kg m −3 , and often exhibited particulate backscattering comparable to or greater than that of the overlying subsurface biomass maximum. Hydrographic and oxygen properties showed no clear signatures of recent subduction associated with the layer. An idealized model based on a friction–Coriolis balance produces outward flow below the pycnocline, which can redirect slowly sinking particles along sloping isopycnals toward deeper peripheral waters. The results suggest that eddy relaxation provides a potentially important pathway for exporting slowly sinking particles into the twilight zone—a cyclonic counterpart to the anticyclonic “wineglass effect.”
- 23 October 2026
- Published by American Geophysical Union (AGU) in Geophysical Research Letters
- Vol. 53 (20)
Open AccessIsopycnal Pathway for Small‐Particle Export in a Decaying Cyclonic Eddy Revealed by a BGC‐Argo Float
Mesoscale eddies influence carbon export through production, subduction, mixing, and particle redistribution. In this study, a BGC‐Argo float observed a deep particle layer in the twilight zone during the decay of a cyclonic eddy in the Northwest Pacific. The layer contained pigment‐depleted particles, remained confined to a narrow density range near 26 kg m −3 , and often exhibited particulate backscattering comparable to or greater than that of the overlying subsurface biomass maximum. Hydrographic and oxygen properties showed no clear signatures of recent subduction associated with the layer. An idealized model based on a friction–Coriolis balance produces outward flow below the pycnocline, which can redirect slowly sinking particles along sloping isopycnals toward deeper peripheral waters. The results suggest that eddy relaxation provides a potentially important pathway for exporting slowly sinking particles into the twilight zone—a cyclonic counterpart to the anticyclonic “wineglass effect.”
- 1 June 2026
- Published by Elsevier in Ocean Modelling
- Vol. 203
Inferring upper-ocean submesoscale ageostrophic dynamics using a physics-informed constrained deep learning framework
- 1 June 2026
- Published by Elsevier in Ocean Modelling
- Vol. 203
Inferring upper-ocean submesoscale ageostrophic dynamics using a physics-informed constrained deep learning framework
- Open Access
- 16 March 2026
- Published by Elsevier in SSRN Electronic Journal
Inferring upper-ocean submesoscale ageostrophic dynamics using a physics-informed constrained deep learning framework
Recent advances in satellite technology have enabled high-resolution observations of sea surface height (SSH), creating an unprecedented opportunity to explore upper-ocean submesoscale motions, yet reconstructing submesoscale dynamics from SSH observations remains challenging. In this study, we develop a physics-informed deep learning framework to infer ageostrophic submesoscale dynamics, characterized by large relative vorticity ζ/f, from high-resolution SSH data. By incorporating oceanic flow strain as a physical constraint, this method enables more accurate reconstruction of ageostrophic features, yielding an approximately 40% reduction of in root-mean-square error (RMSE) compared with classical surface quasi-geostrophic reconstructions. We also assess the spatial and seasonal differences in reconstruction performance across five representative ocean regions. The results indicate that prediction skill depends strongly on the underlying dynamical regime, with high accuracy in regions dominated by submesoscale activity and degraded performance when internal gravity wave (IGW) disturbances are prominent. This dependence is further supported by a global evaluation of prediction skill. Additionally, discussion of geostrophic and ageostrophic momentum balance provides insight into the strengths and limitations of the deep learning approach in reconstruction of ageostrophic dynamics. Together, these results highlight the potential of physics-informed deep learning approaches for reconstructing upper-ocean submesoscale dynamics directly from observations.
- 16 March 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessInferring upper-ocean submesoscale ageostrophic dynamics using a physics-informed constrained deep learning framework
Recent advances in satellite technology have enabled high-resolution observations of sea surface height (SSH), creating an unprecedented opportunity to explore upper-ocean submesoscale motions, yet reconstructing submesoscale dynamics from SSH observations remains challenging. In this study, we develop a physics-informed deep learning framework to infer ageostrophic submesoscale dynamics, characterized by large relative vorticity ζ/f, from high-resolution SSH data. By incorporating oceanic flow strain as a physical constraint, this method enables more accurate reconstruction of ageostrophic features, yielding an approximately 40% reduction of in root-mean-square error (RMSE) compared with classical surface quasi-geostrophic reconstructions. We also assess the spatial and seasonal differences in reconstruction performance across five representative ocean regions. The results indicate that prediction skill depends strongly on the underlying dynamical regime, with high accuracy in regions dominated by submesoscale activity and degraded performance when internal gravity wave (IGW) disturbances are prominent. This dependence is further supported by a global evaluation of prediction skill. Additionally, discussion of geostrophic and ageostrophic momentum balance provides insight into the strengths and limitations of the deep learning approach in reconstruction of ageostrophic dynamics. Together, these results highlight the potential of physics-informed deep learning approaches for reconstructing upper-ocean submesoscale dynamics directly from observations.
- 2 January 2026
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research: Oceans
- Vol. 131 (1)
Submesoscale Stirring Drives Lateral Dispersion and Vertical Exchange in a Cyclonic Mesoscale Eddy
Identifying and quantifying submesoscale stirring in the ocean remains a significant challenge. This study investigates the lateral dispersion and vertical transport effects within a cyclonic mesoscale eddy, using a combination of submesoscale‐permitting in situ observations and model data. By analyzing water mass variations along cross‐eddy isopycnals, distinct dispersion behaviors are identified at different depth layers. The analysis reveals an equivalent submesoscale along‐isopycnal diffusivity larger than 100 m 2 s −1 in the eddy's subsurface layer. To further examine transport pathways, Lagrangian particle tracking experiments are conducted using LLC4320 model output, presenting the features of lateral spreading and vertical penetration at various depths and locations within the eddy. Our results show that in the mixed layer, submesoscale stirring associated with submesoscale instabilities drives efficient lateral dispersion away from the eddy core. While in the subsurface stratified layer, internal wave motions significantly enhance vertical shear, which drives isopycnal submesoscale stirring. This process disrupts the eddy's coherence and causes vertical tracer exchange. These findings emphasize the essential role of submesoscale stirring processes in driving both lateral and vertical tracer dispersion and consequently affecting the eddies' capacity to trap water masses.
- 2 January 2026
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research: Oceans
- Vol. 131 (1)
Submesoscale Stirring Drives Lateral Dispersion and Vertical Exchange in a Cyclonic Mesoscale Eddy
Identifying and quantifying submesoscale stirring in the ocean remains a significant challenge. This study investigates the lateral dispersion and vertical transport effects within a cyclonic mesoscale eddy, using a combination of submesoscale‐permitting in situ observations and model data. By analyzing water mass variations along cross‐eddy isopycnals, distinct dispersion behaviors are identified at different depth layers. The analysis reveals an equivalent submesoscale along‐isopycnal diffusivity larger than 100 m 2 s −1 in the eddy's subsurface layer. To further examine transport pathways, Lagrangian particle tracking experiments are conducted using LLC4320 model output, presenting the features of lateral spreading and vertical penetration at various depths and locations within the eddy. Our results show that in the mixed layer, submesoscale stirring associated with submesoscale instabilities drives efficient lateral dispersion away from the eddy core. While in the subsurface stratified layer, internal wave motions significantly enhance vertical shear, which drives isopycnal submesoscale stirring. This process disrupts the eddy's coherence and causes vertical tracer exchange. These findings emphasize the essential role of submesoscale stirring processes in driving both lateral and vertical tracer dispersion and consequently affecting the eddies' capacity to trap water masses.
- Open Access
- 24 December 2025
- Published by Wiley
Submesoscale stirring drives lateral dispersion and vertical exchange in a cyclonic mesoscale eddy
• Submesoscale stirring drives enhanced horizontal diffusivity exceeding 100 m² s⁻¹ within the eddy. • Submessocale stirring results in vertical exchange along the eddy's sloping isopycnals. • Distinct layer-dependent mechanisms governing tracer dispersion are identified within the eddy.
- 24 December 2025
- Published by Wiley
Open AccessSubmesoscale stirring drives lateral dispersion and vertical exchange in a cyclonic mesoscale eddy
• Submesoscale stirring drives enhanced horizontal diffusivity exceeding 100 m² s⁻¹ within the eddy. • Submessocale stirring results in vertical exchange along the eddy's sloping isopycnals. • Distinct layer-dependent mechanisms governing tracer dispersion are identified within the eddy.
- Open Access
- 6 August 2024
- Published by MDPI AG in Remote Sensing
- Vol. 16 (16) , 2871
Oceanic Precipitation Nowcasting Using a UNet-Based Residual and Attention Network and Real-Time Himawari-8 Images
Qualitative precipitation forecasting plays a vital role in marine operational services. However, predicting heavy precipitation over the open ocean presents a significant challenge due to the limited availability of ground-based radar observations far from coastal regions. Recent advancements in deep learning models offer potential for oceanic precipitation nowcasting using satellite images. This study implemented an enhanced UNet model with an attention mechanism and a residual architecture (RA-UNet) to predict the precipitation rate within a 90 min time frame. A comparative analysis with the standard UNet and UNet with an attention algorithm revealed that the RA-UNet method exhibited superior accuracy metrics, such as the critical ratio index and probability of detection, with fewer false alarms. Two typical cases demonstrated that RA-UNet had a better ability to forecast monsoon precipitation as well as intense precipitation in a tropical cyclone. These findings indicate the greater potential of the RA-UNet approach for nowcasting heavy precipitation over the ocean using satellite imagery.
- 6 August 2024
- Published by MDPI AG in Remote Sensing
- Vol. 16 (16) , 2871
Open AccessOceanic Precipitation Nowcasting Using a UNet-Based Residual and Attention Network and Real-Time Himawari-8 Images
Qualitative precipitation forecasting plays a vital role in marine operational services. However, predicting heavy precipitation over the open ocean presents a significant challenge due to the limited availability of ground-based radar observations far from coastal regions. Recent advancements in deep learning models offer potential for oceanic precipitation nowcasting using satellite images. This study implemented an enhanced UNet model with an attention mechanism and a residual architecture (RA-UNet) to predict the precipitation rate within a 90 min time frame. A comparative analysis with the standard UNet and UNet with an attention algorithm revealed that the RA-UNet method exhibited superior accuracy metrics, such as the critical ratio index and probability of detection, with fewer false alarms. Two typical cases demonstrated that RA-UNet had a better ability to forecast monsoon precipitation as well as intense precipitation in a tropical cyclone. These findings indicate the greater potential of the RA-UNet approach for nowcasting heavy precipitation over the ocean using satellite imagery.
- Open Access
- 16 June 2024
- Published by American Geophysical Union (AGU) in Geophysical Research Letters
- Vol. 51 (12)
Scale‐Dependent Vertical Heat Transport Inferred From Quasi‐Synoptic Submesoscale‐Resolving Observations
Oceanic motions across meso‐, submeso‐, and turbulent scales play distinct roles in vertical heat transport (VHT) between the ocean's surface and its interior. While it is commonly understood that during summertime the enhanced stratification due to increased solar radiation typically results in an reduced upper‐ocean vertical exchange, our study reveals a significant upward VHT associated with submesoscale fronts (−2 and extends to ∼150 m deep at the fronts between eddies. Combined with microstructure observations, this study demonstrates that mixing process can only partly offset the strong upward VHT by inducing a downward heat flux of 0.5–10 W m −2 . Thus, the submesoscale‐associated VHT is effectively heating the subsurface layer. These findings offer a quantitative perspective on the scale‐dependent nature of VHT, with crucial implications for the climate system.
- 16 June 2024
- Published by American Geophysical Union (AGU) in Geophysical Research Letters
- Vol. 51 (12)
Open AccessScale‐Dependent Vertical Heat Transport Inferred From Quasi‐Synoptic Submesoscale‐Resolving Observations
Oceanic motions across meso‐, submeso‐, and turbulent scales play distinct roles in vertical heat transport (VHT) between the ocean's surface and its interior. While it is commonly understood that during summertime the enhanced stratification due to increased solar radiation typically results in an reduced upper‐ocean vertical exchange, our study reveals a significant upward VHT associated with submesoscale fronts (−2 and extends to ∼150 m deep at the fronts between eddies. Combined with microstructure observations, this study demonstrates that mixing process can only partly offset the strong upward VHT by inducing a downward heat flux of 0.5–10 W m −2 . Thus, the submesoscale‐associated VHT is effectively heating the subsurface layer. These findings offer a quantitative perspective on the scale‐dependent nature of VHT, with crucial implications for the climate system.
- Open Access
- 8 March 2024
- Published by MDPI AG in Sustainability
- Vol. 16 (6) , 2274
Numerical Simulation of Storm Surge Inundation in Estuarine Area Considering Multiple Influencing Factors
With global climate change, the risk of extreme storms and storm surges in estuarine areas is increasing; thus, storm surge inundation research and prediction have become important issues to ensure sustainable development in estuarine areas. The Jitimen estuary in Guangdong Province, China, was chosen as our study area. In this study, a numerical model for simulating storm surge inundation in small regions based on unstructured triangular grids was established, and the model accuracy was validated. The typhoon characteristics in the study area were statistically analyzed based on historical data. Three experimental schemes, involving factors influencing storm surge inundation, such as typhoon landfall location, intensity, and direction, were used to evaluate the differences in the numerical results. The results showed that when the typhoon landfall direction remained unchanged and the highest tide levels at the Sanzao tide gauge station were similar, the differences between the numerical results for the typhoon landfall location and typhoon intensity schemes were less than 5%, and the inundation characteristics were similar. However, when the typhoon location and intensity were unchanged and the highest tide levels at the Sanzao tide gauge station were similar, the numerical results for the typhoon landfall direction scheme significantly differed; this result was caused by the difference in the duration of the high tide level (exceeding 3 m); these results indicated that the topographic characteristics and the typhoon landing direction had a greater impact on storm surge inundation. The results from this study can aid in the prediction of storm surge inundation information for the Jitimen estuary area when the typhoon landing direction and the maximum tide level are known.
- 8 March 2024
- Published by MDPI AG in Sustainability
- Vol. 16 (6) , 2274
Open AccessNumerical Simulation of Storm Surge Inundation in Estuarine Area Considering Multiple Influencing Factors
With global climate change, the risk of extreme storms and storm surges in estuarine areas is increasing; thus, storm surge inundation research and prediction have become important issues to ensure sustainable development in estuarine areas. The Jitimen estuary in Guangdong Province, China, was chosen as our study area. In this study, a numerical model for simulating storm surge inundation in small regions based on unstructured triangular grids was established, and the model accuracy was validated. The typhoon characteristics in the study area were statistically analyzed based on historical data. Three experimental schemes, involving factors influencing storm surge inundation, such as typhoon landfall location, intensity, and direction, were used to evaluate the differences in the numerical results. The results showed that when the typhoon landfall direction remained unchanged and the highest tide levels at the Sanzao tide gauge station were similar, the differences between the numerical results for the typhoon landfall location and typhoon intensity schemes were less than 5%, and the inundation characteristics were similar. However, when the typhoon location and intensity were unchanged and the highest tide levels at the Sanzao tide gauge station were similar, the numerical results for the typhoon landfall direction scheme significantly differed; this result was caused by the difference in the duration of the high tide level (exceeding 3 m); these results indicated that the topographic characteristics and the typhoon landing direction had a greater impact on storm surge inundation. The results from this study can aid in the prediction of storm surge inundation information for the Jitimen estuary area when the typhoon landing direction and the maximum tide level are known.