Current Issue
Display Method:
2026, 48(2): 1-10.
doi: 10.12284/hyxb20260020
Abstract:
In 2025, China’s manned submersible, the Jiaolong, successfully completed its first manned deep-sea dive in the ice-covered waters of the Arctic, marking a pivotal first step in advancing China’s deep-sea exploration from ‘full ocean depth’ to ‘all ocean regions’. The expedition faced numerous challenges, including extremely low temperatures, shifting sea ice, geomagnetic anomalies and intricate underwater acoustic conditions. We made systematic adaptations to the Jiaolong submersible, achieving key technological breakthroughs in deployment and recovery, under-ice ascent guidance, and ice-zone navigation and positioning. This paper outlines the core challenges of manned deep-sea operations in polar environments. It focuses on the submersible’s tailored modification plan and the key technologies implemented during the world’s first coordinated dual-submersible mission in polar waters. This paper aim to provide technical references and engineering paradigms for China’s future routine polar deep-sea exploration, habitat studies and resource surveys.
In 2025, China’s manned submersible, the Jiaolong, successfully completed its first manned deep-sea dive in the ice-covered waters of the Arctic, marking a pivotal first step in advancing China’s deep-sea exploration from ‘full ocean depth’ to ‘all ocean regions’. The expedition faced numerous challenges, including extremely low temperatures, shifting sea ice, geomagnetic anomalies and intricate underwater acoustic conditions. We made systematic adaptations to the Jiaolong submersible, achieving key technological breakthroughs in deployment and recovery, under-ice ascent guidance, and ice-zone navigation and positioning. This paper outlines the core challenges of manned deep-sea operations in polar environments. It focuses on the submersible’s tailored modification plan and the key technologies implemented during the world’s first coordinated dual-submersible mission in polar waters. This paper aim to provide technical references and engineering paradigms for China’s future routine polar deep-sea exploration, habitat studies and resource surveys.
2026, 48(2): 79-94.
doi: 10.12284/hyxb20260010
Abstract:
High-quality underwater optical images are crucial for tasks such as digital twins of seabed scenes, benthic habitat protection, seabed mineral resource detection, and understanding unknown underwater phenomena. However, due to factors such as complex aquatic environments and lighting conditions, underwater optical images suffer from degradation issues including color distortion, blurred details, and low contrast. Existing underwater image enhancement methods often focus on optimizing enhancement algorithms themselves, lacking systematic analytical mechanisms for tracing, classifying, and grading different types of degradation. To address this, considering the complexity and heterogeneity of underwater optical imaging environments, this paper proposes an image quality enhancement strategy that takes degradation types into account. First, a degradation-type-aware network is constructed to identify underwater hazy and blurred images, achieving an accuracy of 97%, and also demonstrating a high distinguishing capability for illumination degradation types. Second, for the identified underwater hazy images, an adaptive color correction method is designed based on the statistical distribution of color bias values in real underwater images, effectively restoring color attenuation in varying degrees. Finally, a block indexing strategy is introduced to obtain more precise background light estimates, further addressing the hazy blur issue in underwater images in conjunction with the underwater dark channel prior. Experimental results on various real underwater image datasets, including UIEB and RUIE, indicate that compared to representative underwater image enhancement methods, the PSNR and SSIM metrics are improved by 22.17% and 4.5%, respectively.
High-quality underwater optical images are crucial for tasks such as digital twins of seabed scenes, benthic habitat protection, seabed mineral resource detection, and understanding unknown underwater phenomena. However, due to factors such as complex aquatic environments and lighting conditions, underwater optical images suffer from degradation issues including color distortion, blurred details, and low contrast. Existing underwater image enhancement methods often focus on optimizing enhancement algorithms themselves, lacking systematic analytical mechanisms for tracing, classifying, and grading different types of degradation. To address this, considering the complexity and heterogeneity of underwater optical imaging environments, this paper proposes an image quality enhancement strategy that takes degradation types into account. First, a degradation-type-aware network is constructed to identify underwater hazy and blurred images, achieving an accuracy of 97%, and also demonstrating a high distinguishing capability for illumination degradation types. Second, for the identified underwater hazy images, an adaptive color correction method is designed based on the statistical distribution of color bias values in real underwater images, effectively restoring color attenuation in varying degrees. Finally, a block indexing strategy is introduced to obtain more precise background light estimates, further addressing the hazy blur issue in underwater images in conjunction with the underwater dark channel prior. Experimental results on various real underwater image datasets, including UIEB and RUIE, indicate that compared to representative underwater image enhancement methods, the PSNR and SSIM metrics are improved by 22.17% and 4.5%, respectively.
2026, 48(2): 11-28.
doi: 10.12284/hyxb20260002
Abstract:
Infragravity waves (with periods of 25−250 s) are critical components of nearshore hydrodynamic processes and have significant influence on coastal geomorphological evolution and engineering safety. Based on the conservation equations of mass, momentum, and energy, this paper systematically reviews the latest research progress on the generation mechanisms and evolution characteristics of infragravity waves. Regarding generation mechanisms, the review elaborates on four primary mechanisms: bound long waves, moving breakpoint forcing, bore merging, and wind gusts. Particular attention is given to the theoretical development from the classical equilibrium solution to non-equilibrium solutions for bound long waves, along with the recently proposed unified Green’s function approach. In terms of propagation and evolution, the phase variation and energy transfer, nonlinear shoaling, nearshore dissipation, and shoreline reflection of infragravity waves on sloping beaches are introduced. Then, the amplification of infragravity waves over offshore raised topographies and coral reefs is also examined. The article further points out the inherent randomness present during the evolution of infragravity waves. Finally, future research directions are outlined, providing a theoretical reference for further study and application in terms of infragravity waves.
Infragravity waves (with periods of 25−250 s) are critical components of nearshore hydrodynamic processes and have significant influence on coastal geomorphological evolution and engineering safety. Based on the conservation equations of mass, momentum, and energy, this paper systematically reviews the latest research progress on the generation mechanisms and evolution characteristics of infragravity waves. Regarding generation mechanisms, the review elaborates on four primary mechanisms: bound long waves, moving breakpoint forcing, bore merging, and wind gusts. Particular attention is given to the theoretical development from the classical equilibrium solution to non-equilibrium solutions for bound long waves, along with the recently proposed unified Green’s function approach. In terms of propagation and evolution, the phase variation and energy transfer, nonlinear shoaling, nearshore dissipation, and shoreline reflection of infragravity waves on sloping beaches are introduced. Then, the amplification of infragravity waves over offshore raised topographies and coral reefs is also examined. The article further points out the inherent randomness present during the evolution of infragravity waves. Finally, future research directions are outlined, providing a theoretical reference for further study and application in terms of infragravity waves.
2026, 48(2): 29-39.
doi: 10.12284/hyxb20260008
Abstract:
Based on the observations of 3 wave buoys deployed in Liaodong Bay in winter of 2023−2025 and wave numerical model, this paper analyzes the characteristics of waves during freezing winter of Bohai Sea. According to the statistic results, the mean significant wave heights (mean significant wave periods) observed by 2 buoys in the center of Bohai Sea are about 1 m (4−5 s). Observations from the located buoy in Liaodong Bay indicate that waves are great affected by sea ice. When the buoy locates in a freezing region, the observed mean significant wave heights (mean significant periods) are 0.2 m (9 s), indicating a 54% reduction (98% increase) compared to the measurements without sea ice. The existence of sea ice may also lead the peak wave direction differ from the dominant wind direction. On the perspective of numeric modeling, the error of simulated wave heights can be reduced by 33% via adding ice-wave terms compared to the model without ice terms. By comparing two wind input terms (Komen and ST6), this paper finds that the simulations match the observations well and the error is comparable. Based on observations of buoys, the results rich the acknowledgements of people in the wave features in Bohai Sea during the freezing winter.
Based on the observations of 3 wave buoys deployed in Liaodong Bay in winter of 2023−2025 and wave numerical model, this paper analyzes the characteristics of waves during freezing winter of Bohai Sea. According to the statistic results, the mean significant wave heights (mean significant wave periods) observed by 2 buoys in the center of Bohai Sea are about 1 m (4−5 s). Observations from the located buoy in Liaodong Bay indicate that waves are great affected by sea ice. When the buoy locates in a freezing region, the observed mean significant wave heights (mean significant periods) are 0.2 m (9 s), indicating a 54% reduction (98% increase) compared to the measurements without sea ice. The existence of sea ice may also lead the peak wave direction differ from the dominant wind direction. On the perspective of numeric modeling, the error of simulated wave heights can be reduced by 33% via adding ice-wave terms compared to the model without ice terms. By comparing two wind input terms (Komen and ST6), this paper finds that the simulations match the observations well and the error is comparable. Based on observations of buoys, the results rich the acknowledgements of people in the wave features in Bohai Sea during the freezing winter.
2026, 48(2): 40-57.
doi: 10.12284/hyxb20260016
Abstract:
Shallow-buried gas hydrates are distributed along the continental slope margin of the northern South China Sea. These hydrates are characterized by shallow burial depths and thin overburden layers, rendering them sensitive to changes in seabed temperature and pressure and prone to dissociation. Focusing on internal solitary waves, internal tides, and seasonal bottom-water temperature variations in the northern South China Sea, this study employs a one-dimensional heat conduction model to simulate their effects on shallow subsurface hydrate dissociation and conducts a parameter sensitivity analysis. Results indicate that temperature-pressure perturbations induced by a single internal solitary wave propagate less than a few centimeters into the sediments, falling short of reaching the top of the Hydrate Occurrence Zone (HOZ) located approximately 0.078 m below the seabed, and are thus unlikely to trigger dissociation. In contrast, an internal-tide-induced temperature increase of 1.72℃ lasting 18 hours transfers heat to the HOZ top within 60 d, potentially leading to the dissociation of approximately 4 cm of hydrate. Seasonal bottom-water warming with an amplitude of 1.76℃ persisting for five months drives the dissociation front downward continuously over one year, resulting in a cumulative dissociation thickness of up to 14 cm. This significant impact demonstrates that the effect of sustained warming is substantially stronger than that of transient perturbations. Furthermore, parameter sensitivity analysis reveals that temperature amplitude and effective thermal diffusivity jointly control the propagation depth of thermal perturbations and the dissociation rate. The initial distribution characteristics of hydrates also significantly influence the dissociation process; specifically, the geothermal gradient, methane flux, and permeability determine the positions of the HOZ upper and lower boundaries, whereas porosity regulates the initial saturation and dissociation sensitivity. This study provides a critical basis for evaluating and predicting the stability of shallow subsurface hydrates and the associated risks of methane release.
Shallow-buried gas hydrates are distributed along the continental slope margin of the northern South China Sea. These hydrates are characterized by shallow burial depths and thin overburden layers, rendering them sensitive to changes in seabed temperature and pressure and prone to dissociation. Focusing on internal solitary waves, internal tides, and seasonal bottom-water temperature variations in the northern South China Sea, this study employs a one-dimensional heat conduction model to simulate their effects on shallow subsurface hydrate dissociation and conducts a parameter sensitivity analysis. Results indicate that temperature-pressure perturbations induced by a single internal solitary wave propagate less than a few centimeters into the sediments, falling short of reaching the top of the Hydrate Occurrence Zone (HOZ) located approximately 0.078 m below the seabed, and are thus unlikely to trigger dissociation. In contrast, an internal-tide-induced temperature increase of 1.72℃ lasting 18 hours transfers heat to the HOZ top within 60 d, potentially leading to the dissociation of approximately 4 cm of hydrate. Seasonal bottom-water warming with an amplitude of 1.76℃ persisting for five months drives the dissociation front downward continuously over one year, resulting in a cumulative dissociation thickness of up to 14 cm. This significant impact demonstrates that the effect of sustained warming is substantially stronger than that of transient perturbations. Furthermore, parameter sensitivity analysis reveals that temperature amplitude and effective thermal diffusivity jointly control the propagation depth of thermal perturbations and the dissociation rate. The initial distribution characteristics of hydrates also significantly influence the dissociation process; specifically, the geothermal gradient, methane flux, and permeability determine the positions of the HOZ upper and lower boundaries, whereas porosity regulates the initial saturation and dissociation sensitivity. This study provides a critical basis for evaluating and predicting the stability of shallow subsurface hydrates and the associated risks of methane release.
2026, 48(2): 58-69.
doi: 10.12284/hyxb20260004
Abstract:
To investigate the differences in wave attenuation characteristics among rigid, flexible, and rigid-flexible composite vegetation under regular waves, a series of physical model tests were conducted in a laboratory flume. The wave attenuation effects of these three vegetation types were quantitatively analyzed, and the relationships between the drag coefficient (CD) and Reynolds number (Re), Keulegan-Carpenter number (KC), and Ursell number (Ur) were determined. Results show that all three configurations induce a progressive along-flume reduction in wave height. Increasing incident wave period or vegetation submergence ratio consistently weakens wave dissipation for all vegetation types. The response to wave height differs by configuration: dissipation by rigid vegetation increases markedly and continuously with wave height, whereas flexible vegetation exhibits a nonlinear behavior, strengthening at first and then weakening as wave height further increases. The rigid-flexible combined configuration integrates these advantages and also shows enhanced dissipation with increasing wave height. Moreover, CD for the three vegetation types can be represented using a unified theoretical expression; the primary distinction among configurations is the value of the influence factor γ, which accounts for the effect of vegetation swaying on wave-height attenuation. Statistically significant dependencies of CD on Re, KC, and Ur are observed and can be parameterized by a unified empirical formulation. These results provide a theoretical basis and design reference for optimizing vegetation configurations in coastal ecological protection and restoration engineering.
To investigate the differences in wave attenuation characteristics among rigid, flexible, and rigid-flexible composite vegetation under regular waves, a series of physical model tests were conducted in a laboratory flume. The wave attenuation effects of these three vegetation types were quantitatively analyzed, and the relationships between the drag coefficient (CD) and Reynolds number (Re), Keulegan-Carpenter number (KC), and Ursell number (Ur) were determined. Results show that all three configurations induce a progressive along-flume reduction in wave height. Increasing incident wave period or vegetation submergence ratio consistently weakens wave dissipation for all vegetation types. The response to wave height differs by configuration: dissipation by rigid vegetation increases markedly and continuously with wave height, whereas flexible vegetation exhibits a nonlinear behavior, strengthening at first and then weakening as wave height further increases. The rigid-flexible combined configuration integrates these advantages and also shows enhanced dissipation with increasing wave height. Moreover, CD for the three vegetation types can be represented using a unified theoretical expression; the primary distinction among configurations is the value of the influence factor γ, which accounts for the effect of vegetation swaying on wave-height attenuation. Statistically significant dependencies of CD on Re, KC, and Ur are observed and can be parameterized by a unified empirical formulation. These results provide a theoretical basis and design reference for optimizing vegetation configurations in coastal ecological protection and restoration engineering.
2026, 48(2): 70-78.
doi: 10.12284/hyxb20260014
Abstract:
Settling velocity is a key parameter in the dynamic characteristics of fine-grained cohesive sediments, holding significant importance for understanding the movement patterns of fine sediments and predicting the evolution of scour and fill in port channels. Due to their small particle size, large specific surface area, and tendency for flocculation between particles, the settling velocity of fine-grained cohesive sediments is influenced by multiple factors, with salinity and sediment concentration exerting particularly significant effects. Using in situ sediment samples collected from the Jiaxing Port channel, 42 sets of hydrostatic settling velocity tests were conducted in a sedimentation tank. These tests covered a salinity range of 0–15‰ and a sediment concentration range of 1–20 kg/m3, examining the combined effects of varying salinity and sediment concentration. The results indicate that when salinity is below 7–9‰, settling velocity increases gradually with rising salinity; beyond this range, settling velocity gradually decreased and stabilised with further salinity increases. For sediment concentrations below 8–10 kg/m3, velocity increased with concentration; above this threshold, velocity progressively decreased. Furthermore, compared to the influence of individual factors, salinity and sediment concentration exhibit synergistic effects, with their combined impact exerting a greater influence on sedimentation velocity. A comparative analysis was conducted on the effects of salinity and sediment concentration on sedimentation velocity under varying conditions. A formula for the hydrostatic sedimentation velocity of fine-grained sediments under different salinity and sediment concentration conditions was established through fitting, and validated against previous research findings. These results may provide relevant reference for studies on sediment transport patterns within the Jiaxing Port channel.
Settling velocity is a key parameter in the dynamic characteristics of fine-grained cohesive sediments, holding significant importance for understanding the movement patterns of fine sediments and predicting the evolution of scour and fill in port channels. Due to their small particle size, large specific surface area, and tendency for flocculation between particles, the settling velocity of fine-grained cohesive sediments is influenced by multiple factors, with salinity and sediment concentration exerting particularly significant effects. Using in situ sediment samples collected from the Jiaxing Port channel, 42 sets of hydrostatic settling velocity tests were conducted in a sedimentation tank. These tests covered a salinity range of 0–15‰ and a sediment concentration range of 1–20 kg/m3, examining the combined effects of varying salinity and sediment concentration. The results indicate that when salinity is below 7–9‰, settling velocity increases gradually with rising salinity; beyond this range, settling velocity gradually decreased and stabilised with further salinity increases. For sediment concentrations below 8–10 kg/m3, velocity increased with concentration; above this threshold, velocity progressively decreased. Furthermore, compared to the influence of individual factors, salinity and sediment concentration exhibit synergistic effects, with their combined impact exerting a greater influence on sedimentation velocity. A comparative analysis was conducted on the effects of salinity and sediment concentration on sedimentation velocity under varying conditions. A formula for the hydrostatic sedimentation velocity of fine-grained sediments under different salinity and sediment concentration conditions was established through fitting, and validated against previous research findings. These results may provide relevant reference for studies on sediment transport patterns within the Jiaxing Port channel.
2026, 48(2): 95-113.
doi: 10.12284/hyxb20260018
Abstract:
Constructing continuous spatiotemporal sequences of sea ice is a prerequisite for achieving more accurate, timely, and high-resolution sea ice predictions in the Bohai Sea. To address the inherent limitations of visible light and passive microwave data in sea ice monitoring, this paper proposes a technical approach based on multi-source synergy and super-resolution fusion. First, the DT-ASI algorithm is optimized and local tie points for the Bohai Sea are established to obtain time-series AMSR data at 6.25 km resolution. Subsequently, a sub-pixel convolutional neural network (Pixel Shuffle) is employed for super-resolution reconstruction, identifying the multi-stage Pixel Shuffle strategy as optimal. This approach reduces the mean absolute error by 8.79% and increases the correlation coefficient by 0.19. By integrating the super-resolution AMSR results with MODIS data, a highly continuous spatiotemporal sequence at 1 km resolution from 2002 to 2025 is constructed, with ice coverage during the ice season rising from less than 28.31% to over 95.86%. The dataset reveals preliminary trends of sea ice area reduction, a shortened ice season, and enhanced interannual variability over the past decade. During the 2024–2025 ice season, the Bohai Sea exhibited a distinctive cyclic pattern of “developing-almost completely melting-redeveloping”. This study, through a synergistic and fusion-based approach, overcomes the limitations of single data sources and provides a spatiotemporally continuous data foundation for refined sea ice monitoring and prediction in the Bohai Sea.
Constructing continuous spatiotemporal sequences of sea ice is a prerequisite for achieving more accurate, timely, and high-resolution sea ice predictions in the Bohai Sea. To address the inherent limitations of visible light and passive microwave data in sea ice monitoring, this paper proposes a technical approach based on multi-source synergy and super-resolution fusion. First, the DT-ASI algorithm is optimized and local tie points for the Bohai Sea are established to obtain time-series AMSR data at 6.25 km resolution. Subsequently, a sub-pixel convolutional neural network (Pixel Shuffle) is employed for super-resolution reconstruction, identifying the multi-stage Pixel Shuffle strategy as optimal. This approach reduces the mean absolute error by 8.79% and increases the correlation coefficient by 0.19. By integrating the super-resolution AMSR results with MODIS data, a highly continuous spatiotemporal sequence at 1 km resolution from 2002 to 2025 is constructed, with ice coverage during the ice season rising from less than 28.31% to over 95.86%. The dataset reveals preliminary trends of sea ice area reduction, a shortened ice season, and enhanced interannual variability over the past decade. During the 2024–2025 ice season, the Bohai Sea exhibited a distinctive cyclic pattern of “developing-almost completely melting-redeveloping”. This study, through a synergistic and fusion-based approach, overcomes the limitations of single data sources and provides a spatiotemporally continuous data foundation for refined sea ice monitoring and prediction in the Bohai Sea.
2026, 48(2): 114-130.
doi: 10.12284/hyxb20260012
Abstract:
The Arctic is a climate-sensitive region where Arctic Amplification is influenced by aerosol radiative forcing. Aerosol Optical Depth (AOD) as key parameter characterizing the extinction properties of atmospheric aerosols, plays a critical role in understanding the influence of aerosols on environmental and climate systems. Satellite remote sensing provides an important means of deriving AOD at global and regional scales, but existing satellite products still suffer from substantial data gaps due to retrieval limitations and the complex Arctic surface environment. The Bayesian Maximum Entropy (BME) method is commonly used for AOD data fusion, yet the traditional BME approach, which employs least squares to model covariance, struggles to effectively handle the complexity and non-stationarity of high-dimensional parameter spaces. Based on AOD products from the Moderate Resolution Imaging Spectroradiometer (MODIS) and the Multi-angle Imaging Spectro Radiometer (MISR), this study introduces a Particle Swarm Optimization (PSO) algorithm with global search capability to improve the covariance modeling process, resulting in a PSO-BME fusion algorithm that enhances the stability and accuracy of data integration. The results demonstrate that the PSO-BME method effectively integrates MODIS and MISR AOD data and successfully fills data gaps. In regions covered by both sources, the fused AOD achieves an RMSE of 0.055, an EE of 78%, an MAE of 0.04, and a correlation coefficient of 0.7, while maintaining acceptable accuracy in unobserved areas. The annual spatial coverage increased from 15.45% (MODIS) and 1.45% (MISR) to 32.7%. Spatiotemporal distribution analysis shows that the fusion product significantly improves spatial continuity and more accurately reflects overall AOD variations. Furthermore, the spatiotemporal evolution patterns reveal that aerosol distribution in the Arctic is influenced by both local meteorological conditions and cross-border transport of pollutants from mid- and low-latitudes.
The Arctic is a climate-sensitive region where Arctic Amplification is influenced by aerosol radiative forcing. Aerosol Optical Depth (AOD) as key parameter characterizing the extinction properties of atmospheric aerosols, plays a critical role in understanding the influence of aerosols on environmental and climate systems. Satellite remote sensing provides an important means of deriving AOD at global and regional scales, but existing satellite products still suffer from substantial data gaps due to retrieval limitations and the complex Arctic surface environment. The Bayesian Maximum Entropy (BME) method is commonly used for AOD data fusion, yet the traditional BME approach, which employs least squares to model covariance, struggles to effectively handle the complexity and non-stationarity of high-dimensional parameter spaces. Based on AOD products from the Moderate Resolution Imaging Spectroradiometer (MODIS) and the Multi-angle Imaging Spectro Radiometer (MISR), this study introduces a Particle Swarm Optimization (PSO) algorithm with global search capability to improve the covariance modeling process, resulting in a PSO-BME fusion algorithm that enhances the stability and accuracy of data integration. The results demonstrate that the PSO-BME method effectively integrates MODIS and MISR AOD data and successfully fills data gaps. In regions covered by both sources, the fused AOD achieves an RMSE of 0.055, an EE of 78%, an MAE of 0.04, and a correlation coefficient of 0.7, while maintaining acceptable accuracy in unobserved areas. The annual spatial coverage increased from 15.45% (MODIS) and 1.45% (MISR) to 32.7%. Spatiotemporal distribution analysis shows that the fusion product significantly improves spatial continuity and more accurately reflects overall AOD variations. Furthermore, the spatiotemporal evolution patterns reveal that aerosol distribution in the Arctic is influenced by both local meteorological conditions and cross-border transport of pollutants from mid- and low-latitudes.
2026, 48(2): 131-143.
doi: 10.12284/hyxb20260006
Abstract:
This study is based on sea surface temperature (SST) data from MODIS and AMSR2 satellite observations. Three machine learning models—backpropagation neural network (BPNN), random forest (RF), and convolutional neural network (CNN)—were constructed to conduct research on SST data fusion technology. In terms of model input design, two differentiated schemes are proposed: the basic scheme includes only latitude, longitude, and raw SST data, while the enhanced scheme introduces a time parameter, resulting in six fusion schemes—BP_without_time, BP_with_time, RF_without_time, RF_with_time, CNN_without_time, and CNN_with_time. Experimental test results show that among the three machine learning models, CNN demonstrates the most outstanding performance, while the RF model performs relatively weakly. In comparative tests of the three models, the enhanced schemes incorporating time parameters significantly outperform the basic schemes without time parameters. Validation results based on 2023–2024 buoy measurement data indicate that the accuracy of the fused SST data is slightly lower than that of AMSR2_SST, but shows a significant improvement compared to MODIS_SST. The monthly coverage of the fused data has been significantly improved compared to the original data. The minimum coverage for 2023–2024 increased from 19.79% of MODIS and 32.10% of AMSR2 to over 49.56%. Additionally, the high-resolution fused results can capture more detailed temperature distribution characteristics, providing richer spatial details compared to the 10 km resolution AMSR2 data.
This study is based on sea surface temperature (SST) data from MODIS and AMSR2 satellite observations. Three machine learning models—backpropagation neural network (BPNN), random forest (RF), and convolutional neural network (CNN)—were constructed to conduct research on SST data fusion technology. In terms of model input design, two differentiated schemes are proposed: the basic scheme includes only latitude, longitude, and raw SST data, while the enhanced scheme introduces a time parameter, resulting in six fusion schemes—BP_without_time, BP_with_time, RF_without_time, RF_with_time, CNN_without_time, and CNN_with_time. Experimental test results show that among the three machine learning models, CNN demonstrates the most outstanding performance, while the RF model performs relatively weakly. In comparative tests of the three models, the enhanced schemes incorporating time parameters significantly outperform the basic schemes without time parameters. Validation results based on 2023–2024 buoy measurement data indicate that the accuracy of the fused SST data is slightly lower than that of AMSR2_SST, but shows a significant improvement compared to MODIS_SST. The monthly coverage of the fused data has been significantly improved compared to the original data. The minimum coverage for 2023–2024 increased from 19.79% of MODIS and 32.10% of AMSR2 to over 49.56%. Additionally, the high-resolution fused results can capture more detailed temperature distribution characteristics, providing richer spatial details compared to the 10 km resolution AMSR2 data.
Email alert
RSS