Yellow Sea Region Research Articles

Spatial and temporal distribution patterns and conservation status of seagrasses in the Yellow Sea and Bohai Sea.

Seagrasses represent a significant class of marine foundation species, yet the distribution of seagrasses in the Yellow Sea and Bohai Sea remains uncertain, thereby impeding the efficacy of conservation and restoration practices. In this study, the spatial and temporal distribution pattern of seagrasses was simulated by the MaxEnt model based on the construction of marine environment and human activity datasets. The main controlling factors affecting seagrass potential distribution were analyzed, and the response of seagrass distribution to global change was clarified. Additionally, the current status of protected and disturbed seagrass meadows was determined. The results indicate the accuracy of the MaxEnt model was enhanced (AUC=0.987) through the integration of a human activity index, pinpointing a highly suitable seagrass area of 867.64km2. The primary factors dictating seagrass distribution were identified as human activities index (18.4%), water quality indicators (including nitrate 20.9%, silicate 11.3%, and dissolved iron 6.4%), and topographic elements (such as bathymetry 15.1% and slope 12.0%). The highly suitable areas for seagrasses showed a gradual expansion in the future, with a projected increase of 12% in 2100 under the SSP585 scenario. A binary overlay analysis showed that only 2.61% of seagrass potential habitats (highly and moderately suitable areas) were adequately protected, highlighting a significant conservation gap in the Yellow Sea and Bohai Sea regions. On the contrary, seagrass potential habitats exposed to mariculture activities were as much as 49.76%. These findings underscore the urgent need for monitoring and protecting seagrasses in the study area, with the establishment of MPAs emerging as a viable conservation strategy. This research provides a scientific basis for understanding the degradation and conservation status of seagrass meadows and has important practical implications for targeted conservation and restoration efforts of these critical marine foundation species.

Precise Relocation and Source Characterization of Two Earthquake Sequences in Yellow Sea Using Regional Lg-Wave Observations

ABSTRACT Two significant earthquakes of magnitude ML 4.6 and 4.5 occurred on 18 January and 3 December 2021 in the central region of the Yellow Sea, respectively. The earthquakes occurred beneath the Gunsan sedimentary basin at about 10 km depth with a strike-slip faulting mechanism on nodal planes striking northwest–southeast (NW–SE) and north-northeast–south-southwest (NNE–SSW). Despite a lack of close-by seismographic stations, we successfully utilized regional Lg-wave observations on both coasts of the sea—the Korean peninsula on the east and eastern China on the west. For nine earthquakes in two event sequences, the Lg-wave differential travel times of the nearby event pairs at the common station are carefully measured using the waveform cross-correlation technique. The double-difference earthquake relocation method is employed to obtain precise relative epicentral locations using the Lg correlation measurements. Relocated epicenters align along the NW–SE direction, indicating that the nodal plane striking the same direction is the likely fault plane on which both sequences occurred. This is the first case reported in the literature in which the causative fault plane has been identified for earthquakes in the central Yellow Sea region. It has an important implication for current regional tectonics; it favors neither old tectonic features trending NE–SW (Qianliyan uplift) nor the north–south alignment of significant earthquakes in the region along the Amur plate boundary. The Lg waves from the earthquake sequences are dominant seismic signals on all three-component records at stations in 160–550 km and allowed us to analyze source properties of the two largest earthquakes using the empirical Green’s function approach. Azimuthal variations of the source corner frequencies suggest that earthquake rupture likely propagated toward southeast (125°) along the fault plane, supporting the aftershock relocation results.

Assessment of high-resolution regional ocean reanalysis K-ORA22 for the Northwest Pacific

The Korea Institute of Ocean Science and Technology developed the Korea Operational Oceanographic System-Ocean Predictability Experiment for Marine Environment (KOOS-OPEM), a high-resolution (1/24°, 51 vertical levels) ocean prediction model for the Northwest Pacific Ocean that incorporates ensemble optimal interpolation. In this study, we present KOOS-OPEM ReAnalysis version 2022 (K-ORA22), which covers the period from 2011 to 2022. We conducted a comparative analysis between K-ORA22 and other high-resolution (1/10°–1/12°) global reanalyses, including the Hybrid Coordinate Ocean Model, Global Ocean Reanalysis and Simulation (GLORYS), and Bluelink ReAnalysis (BRAN), to demonstrate the reproducibility and reliability of regional characteristics. Statistical comparisons revealed that while K-ORA22 exhibited some warm biases, its sea surface temperature (SST) anomaly correlation after removing the seasonal cycle was approximately 0.87, comparable to other reanalyses. Additionally, K-ORA22 effectively reproduced coastal upwelling, which is characterized by a sharp decrease in SST, as observed by marine meteorological buoys in the Southwest of the East/Japan Sea. K-ORA22 exhibits a warm bias of approximately 0.50 °C around 200 m, slightly higher than those of GLORYS and BRAN, while maintaining a low salinity bias in the subsurface. Notably, K-ORA22 outperformed the other reanalyses in accurately reproducing the unique characteristics of North Pacific and East Sea intermediate waters, characterized by a salinity minimum layer. In addition, K-ORA22 stands out in its ability to accurately reproduce the Yellow Sea Cold Water Mass with a low-temperature root-mean-square error (RMSE) of 0.76 °C in the Yellow Sea (YS) region. However, it exhibited the highest RMSE for salinity in the YS region and Korea/Tsushima Strait, indicating a potential overestimation of river discharge from Korea and China. While the sea surface height (SSH) anomaly correlation of K-ORA22 did not surpass 0.80 in the entire region because of limitations in the background error covariance used, its ability to reproduce the Kuroshio path was comparable to those of other reanalysis datasets. In conclusion, K-ORA22 excels in reproducing the unique characteristics of Korean marginal seas. Still, it exhibits weaknesses, such as the overestimation of river discharge and a somewhat limited ability to simulate SSH variability, compared with other global reanalyses. We plan to enhance K-ORA22 by updating background error covariance, addressing biases related to river discharge and assimilating the best available in situ observations and satellite data.

Statistical Analysis of Atmospheric Ducts in the Yellow and Bohai Seas of China and Study of Their Electromagnetic Wave Propagation Characteristics

Atmospheric ducts play a critical role in the propagation of electromagnetic waves by minimizing signal loss and extending transmission distances, which is essential for radar, communication, and navigation systems. This study leverages meteorological sounding data and reanalysis data to analyze the distribution of atmospheric ducts in the Bohai Sea and Yellow Sea regions of China. The parabolic equation method was employed to simulate the propagation characteristics of electromagnetic waves in evaporation ducts, surface ducts, and mixed duct environments, focusing on the effects of electromagnetic wave frequency and antenna height. In the Bohai Sea region, the height of evaporation ducts peaks at 13 m in spring and autumn, decreasing to 6 m in winter. In the Yellow Sea region, the height reaches 12 m in autumn and drops to 7 m in summer, indicating a heterogeneous distribution. The monthly mean occurrence rate of atmospheric ducts is defined as the number of atmospheric duct events in a given month divided by the total number of samples for that month. Influenced by the summer and winter monsoons, the occurrence rate of surface ducts is higher from May to September and lower from October to April of the following year. In contrast, elevated ducts reach their peak occurrence rate of 60% in October. In an evaporation duct environment, propagation loss gradually increases with distance, and the loss is more pronounced in non-uniform environments. In surface ducts, propagation loss exhibits periodic fluctuations with distance, exceeding 47 dB. The mixed duct environment integrates the characteristics of both evaporation and surface ducts, effectively filling the shadow zone between 10 m and 70 m.

Review of Allelopathy in Green Tides: The Case of Ulva prolifera in the South Yellow Sea.

The proliferation of large green macroalgae in marine environments has led to the occurrence of green tides, particularly in the South Yellow Sea region of China, where Ulva prolifera has been identified as the primary species responsible for the world's largest green tide events. Allelopathy among plants is a critical factor influencing the dynamics of green tides. This review synthesizes previous research on allelopathic interactions within green tides, categorizing four extensively studied allelochemicals: fatty acids, aldehydes, phenols, and terpenes. The mechanisms by which these compounds regulate the physiological processes of green tide algae are examined in depth. Additionally, recent advancements in the rapid detection of allelochemicals are summarized, and their potential applications in monitoring green tide events are discussed. The integration of advanced monitoring technologies, such as satellite observation and environmental DNA (eDNA) analysis, with allelopathic substance detection is also explored. This combined approach addresses gaps in understanding the dynamic processes of green tide formation and provides a more comprehensive insight into the mechanisms driving these phenomena. The findings and new perspectives presented in this review aim to offer valuable insights and inspiration for researchers and policymakers.

Dynamic Thresholding Fully Automated sea ice extraction and classification methods based on multi-source remote-sensing data in the Yellow sea and Bohai sea regions

This paper presents the extraction and classification of sea ice in the Yellow and Bohai seas using multi-source remote sensing products, including Landsat 8, HY-1C, and MODIS. We propose a Dynamic Thresholding Fully Automated (DTFA) method for sea ice extraction and classification. The accuracy of sea ice extraction and classification is evaluated using four metrics: confusion matrix accuracy rate, overall accuracy, user accuracy, and kappa coefficient. The DTFA method showed higher extraction and classification accuracy compared to the SVM and k-means++ methods. HY-1C has the most effective extraction, followed by Landsat 8. Landsat 8 is able to extract new ice, grey ice, and grey-white ice with the best effect, while HY-1C has a better effect in extracting white ice. MODIS data has advantages in coverage and data updating. Finally, the fusion of multi-source remote sensing images was used to obtain the sea ice extent of the entire Chinese offshore. This method is also applicable to remote sensing monitoring of sea ice in other sea areas.

Population connectivity of small yellow croaker (Larmichthys polyactis) in the southern Yellow Sea: Implications from multiple otolith-based approaches

Connectivity is a fundamental aspect governing the viability of marine species populations, particularly during their early life stage. However, a dearth of comprehensive studies exists regarding connectivity patterns in Larmichthys polyactis within the southern Yellow Sea region. To address this gap, we conducted a thorough investigation into the population connectivity of L. polyactis collected from Haizhou Bay (HZB) and Lvsi (LS) fishing grounds, employing a suite of otolith-based methodologies. Initially, our analysis of otolith microstructure revealed a notable incubation peak occurring between early and mid-April for juvenile L. polyactis sampled in HZB, despite the sea surface temperature (SST) during that period not aligning with the known spawning requirements of L. polyactis. By juxtaposing the spawning period and associated SST conditions of L. polyactis across different fishing grounds, we deduced a high likelihood that the HZB samples originated from the LS. Subsequently, we investigated the stable isotope ratios (δ13C and δ18O) of L. polyactis collected in HZB, LS, and offshore locations, focusing on the core zone indicative of the early life stage. Statistical analysis using Dunn’s multiple comparison test revealed that no significant disparity in otolith δ13C and δ18O values between HZB and LS samples (P > 0.01). However, these values were notably lower than those observed I n offshore samples (P < 0.01). Furthermore, through otolith morphological analysis, we identified two distinct aggregations within the HZB samples, one aligning with the entirety of the LS samples and another representing the remaining HZB sample subset. This finding provides additional support for the population connectivity between HZB and LS stocks. Drawing upon these findings, we posit the existence of a a mixed population of L. polyactis within HZB, likely resulting from the transportation of eggs or larvae originating from LS to HZB via prevailing northward currents. An enhanced understanding of connectivity dynamics among L. polyactis holds significant implications for the formulation of effective management strategies aimed at bolstering stock recovery efforts.

Multi-Time-Scale Optimal Scheduling Strategy for Marine Renewable Energy Based on Deep Reinforcement Learning Algorithm.

Surrounded by the Shandong Peninsula, the Bohai Sea and Yellow Sea possess vast marine energy resources. An analysis of actual meteorological data from these regions indicates significant seasonality and intra-day uncertainty in wind and photovoltaic power generation. The challenge of scheduling to leverage the complementary characteristics of various renewable energy sources for maintaining grid stability is substantial. In response, we have integrated wave energy with offshore photovoltaic and wind power generation and propose a day-ahead and intra-day multi-time-scale rolling optimization scheduling strategy for the complementary dispatch of these three energy sources. Using real meteorological data from this maritime area, we employed a CNN-LSTM neural network to predict the power generation and load demand of the area on both day-ahead 24 h and intra-day 1 h time scales, with the DDPG algorithm applied for refined electricity management through rolling optimization scheduling of the forecast data. Simulation results demonstrate that the proposed strategy effectively meets load demands through complementary scheduling of wave power, wind power, and photovoltaic power generation based on the climatic characteristics of the Bohai and Yellow Sea regions, reducing the negative impacts of the seasonality and intra-day uncertainty of these three energy sources on the grid. Additionally, compared to the day-ahead scheduling strategy alone, the day-ahead and intra-day rolling optimization scheduling strategy achieved a reduction in system costs by 16.1% and 22% for a typical winter day and a typical summer day, respectively.

Spatiotemporal dynamics of bacterioplankton communities in the estuaries of two differently contaminated coastal areas: Composition, driving factors and ecological process

Seasonal variations of environmental parameters usually lead to considerable changes in microbial communities. Nevertheless, the specific response patterns of these communities in coastal areas subjected to different levels of contamination remain unclear. Our results revealed notable fluctuations in the bacterioplankton community both seasonally and spatially, with seasonal variations being particularly significant. The diversity and composition of bacterioplankton communities in the estuaries varied significantly across seasons and between seas. Some bacterial phyla that were highly abundant in the dry season (e.g., Patescibacteria and Epsilonbacteraeota) were almost absent in the wet season. Furthermore, the network analysis revealed that the bacterioplankton networks were more complex during the wet season than in the dry season. In the wet season, the estuarine bacterioplankton network in the Yellow Sea region was more complex and stable, while the opposite was true in the dry season. According to the neutral community model, stochastic processes played a more significant role in the formation of bacterioplankton communities during the wet season than during the dry season. Estuarine bacterioplankton communities in the Yellow Sea region were more affected by stochastic processes compared to those in the Bohai Sea. In summary, in the estuaries of two differently contaminated coastal areas, the seasonal increase in nutrient levels enhanced the deterministic processes and network complexity of the bacterioplankton communities.

An evaluation of microplastic contamination in the marine waters and species in the coastal region of the South Yellow Sea, China

Microplastics (MPs) contamination of marine environments poses a significant ecological risk, although impacts on species’ realized niche spaces remain unclear. The current study investigates MPs distribution across pelagic habitats, benthic sediments, and key biota in the South Yellow Sea, China. Samples were collected via trawling across estuarine transects, and tissues were digested to extract MPs. Density gradient separations and vacuum-filtrations prepared particle extracts for ATR-FTIR and Micro-Raman spectroscopic characterization. Sampling along industrialized river transects reveals ubiquitous plastic particle presence, with concentrations ranging from 0 to 51.68 item/L seawater. Contamination levels reach their peak at station estuaries before dispersing offshore, indicating significant waste stream inputs. Importantly, MPs detected in demersal and pelagic fish species, as well as in bivalves, confirm exposure across trophic niches. Gastrointestinal tract and gill concentrations reached 0.6 items/g fresh tissue, reflecting significant biological uptake and in vivo retention. The greatest population of organisms occurred adjacent to polluted areas. Overall, distribution of MPs from polluted rivers to coastal food webs was evident, suggesting potential negative impacts on key ecological functions in this system. These findings underscore the need to develop upstream mitigation efforts so as to minimize MPs contamination in areas where nearshore and offshore niches intersect.

Multi-Module Fusion Model for Submarine Pipeline Identification Based on YOLOv5

In recent years, the surge in marine activities has increased the frequency of submarine pipeline failures. Detecting and identifying the buried conditions of submarine pipelines has become critical. Sub-bottom profilers (SBPs) are widely employed for pipeline detection, yet manual data interpretation hampers efficiency. The present study proposes an automated detection method for submarine pipelines using deep learning models. The approach enhances the YOLOv5s model by integrating Squeeze and Excitation Networks (SE-Net) and S2-MLPv2 attention modules into the backbone network structure. The Slicing Aided Hyper Inference (SAHI) module is subsequently introduced to recognize original large-image data. Experimental results conducted in the Yellow Sea region demonstrate that the refined model achieves a precision of 82.5%, recall of 99.2%, and harmonic mean (F1 score) of 90.0% on actual submarine pipeline data detected using an SBP. These results demonstrate the efficiency of the proposed method and applicability in real-world scenarios.

Framing seascape connectivity modeling to prioritize marine conservation effort in China’s coastal sea

Anthropogenic disturbances and climate change are projected to become leading drivers of biodiversity loss and ecological connectivity degradation in marine ecosystems. However, the lack of quantitative understanding for seascape connectivity modeling hinders our ability in providing large-scale marine conservation guidance. By applying well-established theories and tools in landscape connectivity study and marine-specific indicators, we proposed a framework to evaluate the resistance-based seascape connectivity among marine habitats and Marine Protected Areas (MPAs) across China’s coastal sea. The spatial vulnerability of marine habitats was further assessed to prioritize conservation effort. Our results showed that 82% of China’s coastal seas were covered with middle to high level of resistance for migratory marine species, mainly due to concentrated inshore anthropogenic disturbances such as ship traffic and ocean pollution. With the modeled migration distances of different species guilds from short-range (25 km) to long-range (100 km), the area percentage of connected corridors increased from 12.02% to 44.68% in the study area. Vulnerable areas were identified as high resistance (high exposure) and abundant threatened species (high sensitivity) but with small number of connected corridors (low adaptive capacity), primarily distributed in offshore regions of Yellow Sea and East China Sea. Collectively, inshore regions with high anthropogenic disturbances warrant regulation and mitigation in major coastal cities and ports. While the lack of interconnected networks for offshore regions prioritized efforts to enhance seascape connectivity through the establishment of MPA network. Sensitive species groups require more attention in future marine conservation, including threatened populations, climate refugees of marine species, and species with limited movement ranges. This study highlights the potential of developing seascape connectivity model based on landscape theories, and the importance of seascape connectivity study in guiding evidence-based marine conservation strategies.

Assessing the impact of shipping emissions on ozone concentrations in East Asia: Insights from KORUS-AQ and SIJAQ 2021 campaign periods

This study analyzed the impact of shipping emissions on ozone (O3) concentrations in East Asia during the Korea-United States Air Quality (KORUS-AQ) and the Satellite Integrated Joint monitoring for Air Quality (SIJAQ) 2021 campaign period. Numerical simulations were performed during the KORUS-AQ (May 2016, the KORUS case) and the SIJAQ 2021 (October–November 2021, the SIJAQ case) using Community Multi-scale Air Quality (CMAQ), the three-dimensional chemical transport model. We additionally simulated an experiment without shipping emissions to compare the impact with and without shipping emissions. Using the validated simulation results, the average O3 concentrations in inland regions were found to be 35.63 ppb and 20.62 ppb in the KORUS and SIJAQ cases, respectively. The average O3 concentrations were higher in the KORUS case owing to the difference in weather conditions between the two cases in various seasons. The results of examining the impact of shipping emissions on O3 concentrations in the ocean area indicate that both the KORUS and SIJAQ cases resulted in a decrease in O3 concentrations in the Yellow Sea region, while the O3 concentrations increased in the East Sea and North Pacific regions. In particular, in the Yellow Sea and East China Sea, where the O3 concentrations decreased in the KORUS case, O3 sensitivity results showed volatile organic compounds (VOC)-sensitive regime when the shipping emissions were removed but changed to nitrogen oxide (NOx)-sensitive regime when the shipping emissions were added. Meanwhile, in the inland regions, the SIJAQ case decreased O3 concentrations in most areas, whereas the KORUS case showed an increase in O3 concentrations in the Korean Peninsula and Japan. Analysis of O3 sensitivity revealed that the regions of the Korean Peninsula and Japan, where O3 concentrations increased, exhibited a NOx-sensitive regime. Particularly, in the case of the Korean Peninsula, the initial transport of shipping emissions in the western part led to a decrease in O3 concentrations. However, as shipping emissions were further transported from west to east, they became more diluted, and additional NOx generated by O3 titration reactions was also transported, gradually increasing O3 concentrations. Furthermore, we found that during seasons with abundant BVOC emissions, the transport of shipping emissions could contribute to additional O3 production. These results confirmed that shipping emissions in East Asia primarily reduce O3 concentrations, however, differences in seasonal and regional O3 sensitivity and transportation patterns can contribute to O3 production.

Importance of habitat heterogeneity in tidal flats to the conservation of migratory shorebirds.

Understanding species distribution patterns and what determines them is critical for effective conservation planning and management. In the case of shorebirds migrating along the East Asian-Australasian Flyway (EAAF), the loss of stopover habitat in the Yellow Sea region is thought to be the primary reason for the precipitous population declines. However, the rates of decline vary considerably among species, and it remains unclear how such differences could arise within a group of closely related species using apparently similar habitats at the same locales. We mapped the spatial distributions of foraging shorebirds, as well as biotic (benthic invertebrates consumed by migrating shorebirds) and abiotic (sediment characteristics) environmental factors, at a key stopover site in eastern China. Five of the six sediment characteristics showed significant spatial variation with respect to distance along the shoreline or distance from the seawall in the same tidal flat. The biomasses of four of the six most abundant benthic invertebrates were concentrated in the upper or middle zones of the tidal flat. The distribution patterns of all three focal shorebird species on the tidal flat were best explained jointly by this heterogeneity of sediment characteristics and invertebrate prey. These results suggest that the loss of tidal flats along the Yellow Sea, which is typically concentrated at the upper and middle zones, may not only reduce the overall amount of staging habitat, but also disproportionately affect the most resource-rich portions for the birds. Effective conservation of shorebird staging areas along the EAAF and likely elsewhere must consider the subtle habitat heterogeneity that characterizes these tidal flats, prioritizing the protection of those portions richest in food resources, most frequently used by focal bird species, and most vulnerable to anthropogenic threats. Article impact statement: Heterogeneity of tidal flats with respect to biotic and abiotic factors must be considered in shorebird conservation planning.

TCUNet: A Lightweight Dual-Branch Parallel Network for Sea–Land Segmentation in Remote Sensing Images

Remote sensing techniques for shoreline extraction are crucial for monitoring changes in erosion rates, surface hydrology, and ecosystem structure. In recent years, Convolutional neural networks (CNNs) have developed as a cutting-edge deep learning technique that has been extensively used in shoreline extraction from remote sensing images, owing to their exceptional feature extraction capabilities. They are progressively replacing traditional methods in this field. However, most CNN models only focus on the features in local receptive fields, and overlook the consideration of global contextual information, which will hamper the model’s ability to perform a precise segmentation of boundaries and small objects, consequently leading to unsatisfactory segmentation results. To solve this problem, we propose a parallel semantic segmentation network (TCU-Net) combining CNN and Transformer, to extract shorelines from multispectral remote sensing images, and improve the extraction accuracy. Firstly, TCU-Net imports the Pyramid Vision Transformer V2 (PVT V2) network and ResNet, which serve as backbones for the Transformer branch and CNN branch, respectively, forming a parallel dual-encoder structure for the extraction of both global and local features. Furthermore, a feature interaction module is designed to achieve information exchange, and complementary advantages of features, between the two branches. Secondly, for the decoder part, we propose a cross-scale multi-source feature fusion module to replace the original UNet decoder block, to aggregate multi-scale semantic features more effectively. In addition, a sea–land segmentation dataset covering the Yellow Sea region (GF Dataset) is constructed through the processing of three scenes from Gaofen-6 remote sensing images. We perform a comprehensive experiment with the GF dataset to compare the proposed method with mainstream semantic segmentation models, and the results demonstrate that TCU-Net outperforms the competing models in all three evaluation indices: the PA (pixel accuracy), F1-score, and MIoU (mean intersection over union), while requiring significantly fewer parameters and computational resources compared to other models. These results indicate that the TCU-Net model proposed in this article can extract the shoreline from remote sensing images more effectively, with a shorter time, and lower computational overhead.

Evaluation of HY-2C and CFOSAT Satellite Retrieval Offshore Wind Energy Using Weather Research and Forecasting (WRF) Simulations

This study simulated the spatial distribution of wind speeds and wind energy density by using the WRF model, and we used the WRF-simulated results to evaluate the sea surface wind speeds retrieved from the HY-2C and CFOSAT satellite-borne microwave scatterometers over the Yellow Sea region. The main conclusions were as follows: (1) The combination of the MRF boundary layer parameterization scheme, the MM5 near-surface parameterization scheme, and the Global Data Assimilation System (GDAS) initial field demonstrated the best performance in simulating the 10 m wind speed in the Yellow Sea region, with a root-mean-square error (RMSE) of 1.57, bias of 1.24 m/s, and mean absolute percentage error (MAPE) of 17%. (2) The MAPE of the HY-2C inversion data was 9%, while the CFOSAT inversion data had an MAPE of 6%. The sea surface wind speeds derived from the HY-2C and CFOSAT satellite scatterometer inversions demonstrated high accuracy and applicability in this region. (3) The wind speed was found to increase with altitude over the Yellow Sea, with higher wind speeds observed in the southern region compared to the northern region. The wind power density increased with altitude, and the wind power density in the southern area of the Yellow Sea was higher than in the northern region. (4) The CFOSAT satellite inversion products were in good agreement with the WRF simulation results under low wind speed conditions. In contrast, the HY-2C satellite inversion products showed better agreement under moderate wind speed conditions. Under high wind speed conditions, both satellite inversion products exhibited minor deviations, but the HY-2C product had an overall overestimation, while the CFOSAT product remained within the range of −1 to 1 m/s. (6) The wind power density increased with the satellite-inverted 10 m wind speed. When the 10 m wind speed was less than 9 m/s, the wind power density exhibited a roughly cubic trend of increase. However, when the 10 m wind speed exceeded 9 m/s, the wind power density no longer increased with the rise in 10 m wind speed. These findings provide valuable insights into wind energy resources in the Yellow Sea region and demonstrate the effectiveness of satellite scatterometer inversions for wind speed estimation. The results have implications for renewable energy planning and management in the area.

Holocene relative sea-level changes on the southern east coast of the Yellow Sea

Projections indicate increasing coastal hazards in the Yellow Sea region owing to relative sea-level rise. Accordingly, regional studies of such change over the Holocene timescale could be helpful in mitigating these threats. Here, we describe the Holocene record of relative sea-level change in the Gochang-Yeonggwang area of South Korea. We compared this record with that of other sites around the coast of the Yellow Sea, including Bohai Bay (China), the Tando Estuary and Gunsan Bay (South Korea), and sites on the east coast of the South Sea and Toyo-oka Basin, Japan. The findings are based on a multi-proxy study, including lithostratigraphy combined with radiocarbon dating and analyses of grain size distribution, magnetic susceptibility, and geochemistry, which facilitated the generation of 89 paleo-sea-level points reconstructed from 28 borehole cores. The early Holocene transgression was found to start later at Gochang-Yeonggwang than at Gunsan Bay, although the distance between them is only 150 km. This result was inferred from substantial differences in sediment delivery owing to Holocene differences in river basin size, topography, and tectonic context. Before ∼7 ka, the slope and highstands from the predicted mean sea-level models at Bohai Bay, Tando Estuary, and Gochang-Yeonggwang were steeper and lower, respectively, from north to south in the Yellow Sea, approaching the slope of the glacio-eustatic sea-level curve. Therefore, the levering effect of transgression in the early Holocene decreased southward in the Yellow Sea. The slight fluctuation in mean sea-level points of the late Holocene resulted from the formation of storm deposit layers in the coastal area, erosion and redeposition in the channel, and geomorphic effects. A partial difference occurred between the relative sea levels at Gochang-Yeonggwang and Toyo-oka. Future relative sea-level research at the southernmost part of the east coast of the Yellow Sea would further improve understanding of the coastal evolution of the region.

SeaMAE: Masked Pre-Training with Meteorological Satellite Imagery for Sea Fog Detection

Sea fog detection (SFD) presents a significant challenge in the field of intelligent Earth observation, particularly in analyzing meteorological satellite imagery. Akin to various vision tasks, ImageNet pre-training is commonly used for pre-training SFD. However, in the context of multi-spectral meteorological satellite imagery, the initial step of deep learning has received limited attention. Recently, pre-training with Very High-Resolution (VHR) satellite imagery has gained increased popularity in remote-sensing vision tasks, showing the potential to replace ImageNet pre-training. However, it is worth noting that the meteorological satellite imagery applied in SFD, despite being an application of computer vision in remote sensing, differs greatly from VHR satellite imagery. To address the limitation of pre-training for SFD, this paper introduces a novel deep-learning paradigm to the meteorological domain driven by Masked Image Modeling (MIM). Our research reveals two key insights: (1) Pre-training with meteorological satellite imagery yields superior SFD performance compared to pre-training with nature imagery and VHR satellite imagery. (2) Incorporating the architectural characteristics of SFD models into a vanilla masked autoencoder (MAE) can augment the effectiveness of meteorological pre-training. To facilitate this research, we curate a pre-training dataset comprising 514,655 temporal multi-spectral meteorological satellite images, covering the Bohai Sea and Yellow Sea regions, which have the most sea fog occurrence. The longitude ranges from 115.00E to 128.75E, and the latitude ranges from 27.60N to 41.35N. Moreover, we introduce SeaMAE, a novel MAE that utilizes a Vision Transformer as the encoder and a convolutional hierarchical decoder, to learn meteorological representations. SeaMAE is pre-trained on this dataset and fine-tuned for SFD, resulting in state-of-the-art performance. For instance, using the ViT-Base as the backbone, SeaMAE pre-training which achieves 64.18% surpasses from-scratch learning, natural imagery pre-training, and VRH satellite imagery pre-training by 5.53%, 2.49%, and 2.21%, respectively, in terms of Intersection over Union of SFD.

A Modified U-Net Model for Predicting the Sea Surface Salinity over the Western Pacific Ocean

The prediction of oceanic features is always an important issue in oceanography, where deep learning has been proven to be a useful tool. In this study, we applied the improved U-net model to predict the monthly sea surface salinity (SSS) over the western Pacific (WP) Ocean, and the model was designed to use the SSSs from six consecutive months to predict the SSS in the next month. The monthly satellite-based SSSs in 2015–2020 were used for model training, and the data collected after January 2021 were used to evaluate the model’s predictive abilities. The results showed that the predicted SSSs represented the general patterns of SSSs over the WP region. However, the small-scale features were smoothed out in the model, and the temporal variations were also not well captured, especially over the East China Sea and Yellow Sea (ECS&amp;YS) region. To further evaluate the potential of the U-net model, a more specific model was conducted for the ECS&amp;YS region (Domain 2), which successfully predicted both spatial and temporal variations in the SSSs, including the spreading and retreating of the low-salinity tongue. Based on the comparison between the two domains and sensitivity experiments, we found that the prediction biases were contributed by the spatial distributions of the SSSs, the domain size, and the filter numbers. In addition, further multi-step prediction experiments suggested that our U-net model could also be used for long-time prediction, and we have examined up to five months. Overall, this study demonstrated the great ability and potential of the U-net model for predicting SSS, even though only a few trainable data are available.

Dissolved organic matter in the northwestern Pacific marginal seas: Insight into the distribution of its optical properties

Dissolved organic carbon (DOC) and the optical properties of dissolved organic matter (chromophoric- and fluorescent dissolved organic matter; CDOM and FDOM) were measured to determine the distributions and drivers of DOM in the northwestern Pacific marginal seas, including the East China Sea, the Yellow Sea, and the East/Japan Sea, in August 2020. In this study, the concentrations of DOC and CDOM/FDOM in surface water showed good correlations with salinity, indicating a predominant contribution from the Changjiang River. However, significantly high concentrations of DOC and FDOM were also observed in the central Yellow Sea region, which seems to be produced mainly from the continental shelf-water and enriched over the water residence time of the Yellow Sea, while the lowest concentrations of DOC and CDOM/FDOM were found in high-salinity waters near the southern sea of Korea. In addition, the East/Japan Sea showed relatively low DOC concentrations and high FDOM values. To distinguish the physical mixing and biogeochemical processes of DOM, we estimated the water mass fractions using an optimum multi-parameter analysis with hydrological and DOM parameters of the major water masses in this region. Our results showed that five primary external drivers of DOM distribution were intrusions of (1) the Changjiang diluted water and (2) the Kuroshio Current water in the surface water, the mixing of (3) the Yellow Sea bottom cold water and (4) the Kuroshio sub-surface water in the bottom layer, and (5) upwelling of the East/Japan Sea water. We conclude that the distribution of DOM is significantly influenced by the mixing of multiple water masses, and the optical signature of DOM can be an effective proxy for tracing the origins and characteristics of DOM in this region.