This study is based on the meteorological, oceanic, terrain and other physical quantity data covered by the observation points in the southern Zhoushan Islands from January 1, 2019 to December 31, 2021, and uses long short-term memory neural network (LSTM) to build deep learning wave forecast model. We explore the impact of the input-output sequence ratio and the number of input elements on the prediction performance of the model, realize the short-term forecast of the three elements of waves in the Zhoushan sea area, that is the significant wave height, the significant wave period and the propagation direction, and use the data during the 2022 typhoons “Hinnamnor” and “Muifa” to test the model’s prediction ability for extreme sea conditions. The research results show that the multi-element deep learning wave forecast model trained based on measured data has good prediction accuracy and stability, and can realize the prediction of extreme sea conditions. When the input-output sequence ratio is 1∶1, the model accuracy is higher. In non-extreme sea conditions, the three-element model with a prediction time of 1 hour accurately predicts significant wave height, significant wave period and direction, with Root Mean Squared Errors (RMSE) of 0.116 m, 0.569 s, and 24.583° respectively. In extreme sea conditions, the prediction RMSE for the significant wave height is 0.191 m. The increase in the number of input elements can further improve the model accuracy but also increase the training cost when the prediction time is long.

The Multidisciplinary drifting Observatory for the Study of Arctic Climate (MOSAiC) was conducted from October 2019 to September 2020, yielding complete observations of atmosphere, ocean, sea ice thickness (SIT), and snow thickness. These observations provide new opportunities for the development of sea ice models. In this study, the seasonal evolution of SIT during MOSAiC was simulated using the ICEPACK sea ice model and atmospheric and oceanic forcing observations from two periods without missing data (from November 1, 2019 to May 7, 2020; from June 26 to July 27, 2020). The simulation was compared with SIT observation and the reasons for SIT simulation errors were diagnosed. The results show that, in the winter and spring seasons, the model can reproduce the increase in SIT, but overestimates the transition from submerged snow to sea ice and its contribution to sea ice mass balance. This causes the overestimation of SIT in spring. During the summer season, the combination of two thermodynamic schemes and three melt pond schemes indicates that the model overestimates the sea ice surface melting, resulting in thinner SIT at the end of simulation period. Our research demonstrates that the MOSAiC atmospheric and oceanic observation with all variables needed to force ICEPACK can be used to diagnose current sea ice models and very useful for their future improvements.

Oil and plastic pollutants are a serious threat to marine ecosystems and have even been found in the Ross Sea of the Southern Ocean. In order to obtain low-temperature alkane degrading bacteria and plastic-degrading bacteria in the region, a total of twelve sediment samples were collected in the Ross Sea area for enrichment and isolation of alkane-degrading bacteria at low-temperature, and the diversity analyses of the tetradecane-enriched communities showed that the most dominant genera were Pseudomonas, Alcanivorax, Marinomonas, Pseudoalteromonas. The polyethylene terephthalate(PET)and polyethylene(PE)was further validated using the dominant alkane-degrading bacteria. Scanning electron microscopy (SEM) and Fourier transform infrared spectroscopy (ATR-FTIR) demonstrated that the four pure cultures of Pseudomonas pelagia R1-05-CR3, Pseudomonas taeanensis A11-04-CA4, Halomonas titanicae A11-02-7C2 and Rhodococcus cerastii R1-05-7C3 could degrade PE effectively. The results of UPLC-MS and SEM confirmed the PET degradation by isolates of R. cerastii R1-05-7C3, Microbacterium maritypicum RA1-00-CA1, and H. titanicae A11-02-7C2. In conclusion, this study reports the diversity of the tetradecane-enriched consortia at low-temperature and plastic degrading bacteria in the offshore sediments of the Ross Sea in the Southern Ocean, which play a selfpurifying role in in-situ environmental contamination, and also provided strain resources for the biodegradation of hydrocarbon and plastic contaminants at low temperature.

Mangroves are a crucial ecological barrier for coastal zones, and sensitive areas to climate change and human activities, where benthic fauna responds directly to the sedimentary environment because of their intimate contact and relatively fixed settlement habitats. In-situ observations and sample collections were carried out at 7 stations in the mangrove area off the Chaoshan coastal zone, in April 2021, for further analysis of the sedimentary environment, meiofaunal communities, and eco-environment quality evaluation. A total of 15 meiofauna groups were identified. Free-living marine nematodes were the most dominant group, accounting for 90.32% of the total abundance of meiofauna; while polychaetes were the first in biomass (58.44%). The average abundance of meiofauna was slightly higher than that of the other mangrove forests. The spatial pattern of the meiofaunal abundance exhibited a similar distribution trend to that of the abundance of nematodes and the content of chlorophyll a (Chl a), phaeophorbide, organic carbon, and heavy metals (Cd, Zn, Cu, Cr, and Hg): all the parameters showed the highest values in the northern of the Lianyang River, and decreased towards the south and north. Meiofaunal communities in different patches shared a high similarity of 70%, and the best explanation factor for the differences in the meiofauna community was the content of heavy metal Pb. The abundance ratio of marine nematodes to copepods (N/C), the possible ecological risk index, and the sediment quality grading all indicated a poor environmental state in the area investigated.

To clarify the composition and distribution characteristics of macrobenthic communities in Yantai Changdao Island marine ranch and evaluate the impact of marine ranching on these macrobenthos, sampling stations were set up inside and outside the marine ranch in October 2022. Surveys of macrobenthic animals and the characteristics of macrobenthic communities were analyzed. A total of 88 species of macrobenthic animals were collected and identified during this voyage. Although the number of species in the pasture and control area was similar, the dominant groups differed. In the pasture, 70 species were identified, with mollusks being the dominant group; whereas, in the control area, 69 species were identified, with annelids as the dominant group. Eight dominant species were found, including 3 mollusk species in the pasture and 6 species in the control area, comprising 2 mollusk species, 1 echinoderm species, and 3 annelid species. The average abundance and biomass of macrobenthic animals in the pasture were significantly higher than those in the control area. However, Margalef species richness index (d), Pielou evenness index (J'), and Shannon Wiener diversity index (H') values showed little difference between the pasture and the outside. The results of Cluster analysis (CLUSTER) and non-metric multidimensional scaling (NMDS) analysis indicated that relatively low similarity among each station inside and outside the marine ranch. The AMBI and m-AMBI analyses revealed that the overall pollution disturbance in the studied water area was relatively small, indicating good benthic ecological health. Combined with historical data, the analysis revealed a significant increase in species abundance and biomass of macrobenthic communities in the surveyed area. These results suggest that the development of marine pastures has a certain degree of impact on the growth and development of macrobenthic communities.

Sea surface temperature (SST) is one of the most important physical variables of the ocean, which provides the basic information of the climate system. Accurately SST forecasting system has a comprehensive and essential application. In recent years, AI-based SST forecasting methods have become popular and shown great potential. Based on the convolutional long and short-term memory neural network (ConvLSTM), this paper studies the influence of multi-scale input fields on SST prediction in the northern South China Sea. Multi-dimensional ensemble empirical mode decomposition method (MEEMD) is used to decompose the average daily SST into the spatial eigenmodes of differentiated scales. Input different combinations of eigenmodes into ConvLSTM for training and prediction experiments. Results show that when using all four SST eigenmodes, the RMSE of the predicted SST in 1−7 days is 0.4−0.8℃, decrease 0.2−1.2℃ compared with the original SST alone; the MAPE is 1%−6%, decrease 0.5%−10%; the spatial correlation coefficient is 99.5%−96.5%, improve 0.5%−3.5%. Moreover, the randomized experiments also further proved the method has a high universality. The prediction model based on deep learning needs to select the appropriate training data in order to further improve its prediction accuracy. This paper preliminarily explores the integration of artificial intelligence methods and physical concepts in SST prediction, which can provide some reference for future research.

Utilizing the molecular biomarkers of organic matter in marine sediments from the Antarctic Peninsula (D1-7) and adjacent waters of the South Orkney Islands (D5-6), the ecological relationships implicit in the reconstructed variations of planktonic productivity and population structure are examined in relation to the Biological Pump (BP)/Microbial Carbon Pump (MCP), as well as the efficiency of marine carbon sinks and storage. Over the past century, a series of molecular biomarkers in sediment cores has exhibited significant changes, reflecting substantial spatiotemporal evolution in upper ocean planktonic productivity/community structure and sedimentary carbon reservoirs. These changes are indeed linked to global climate variability. The research findings are as follows: (1) Based on the characteristics of molecular composition and chromatographic peak patterns of biomarker compounds, as well as parameters such as Main Peak Carbon (MH), Light Hydrocarbons/Heavy Hydrocarbons (L/H), Bacterial-Algal Ratio (nC₁₅ + nC₁₇ + nC₁₉), Large Phytoplankton Ratio (nC₂₁ + nC₂₃ + nC₂₅), and carbon preference index (CPI), it is evident that the primary source of sedimentary carbon is marine-derived organic carbon. Marine organisms serve as natural carbon sinks for carbon fixation and storage. (2) The sediments from the D5-6 region exhibit high organic matter enrichment, primarily influenced by factors such as higher surface water productivity, higher sedimentation rates (average of 0.19 cm/a), shallower water depths (385 m), and a reducing sedimentary environment (average Pr/Ph value of 0.95). These conditions favor the transport of Particulate Organic Carbon (POC) from the ocean surface to the deep sea via the Biological Pump (BP) process, facilitating rapid burial and storage. In contrast, sediments from the D1-7 region, characterized by greater water depths (1 100 m) and lower sedimentation rates (0.07 cm/a, experience degradation of carbonaceous compounds during sedimentation processes and subsequent oxidative degradation in an oxic environment (average Pr/Ph value of 1.22). Both processes are unfavorable for carbon sequestration in sediments. However, the control factor determining carbon preservation in sediments may predominantly be sedimentation rate. (3) Over the past century, the total abundance of zooplankton, primary productivity of phytoplankton, and biomass of diatoms and dinoflagellates in the waters near the Antarctic Peninsula and the South Orkney Islands have shown an increasing trend, while the biomass and proportion of coccolithophores have decreased (particularly evident near the Antarctic Peninsula). This indicates a declining trend in the effectiveness of the calcium carbonate pump while the silica pump dominated by diatoms is strengthening. The relative strengths of these two processes largely determine the structure and efficiency of the biological pump, as well as the proportion of organic and inorganic carbon transported to marine sediments. (4) The trends in molecular biomarker variations in the two sediment cores show certain comparability overall, with distinct stages. Following interdecadal shifts (since 1972), the waters near the South Orkney Islands experienced a significant increase in zooplankton abundance from a depth of 5-6 cm beginning in 1982. Particularly, during the periods of 1997 and 2012, zooplankton abundance witnessed a dramatic increase, indicating rapid changes in planktonic community structure under the backdrop of global warming. Variations in both decreased primary productivity of phytoplankton and increased zooplankton abundance contribute to significant uncertainties in the changes in the strength of the biological pump (enhancement/weakening). (5) In contrast, over the past century, the productivity of phytoplankton/diatoms and dinoflagellates in the waters near the Antarctic Peninsula has gradually increased, while microbial productivity/ancient archaeal biomass has decreased. This suggests a weakening of microbial carbon sequestration intensity, indicating a decrease in the efficiency of the microbial carbon pump (MCP). This underscores the crucial role of global warming in the fluctuations of phytoplankton productivity/biomass in marine waters. The biomass and composition characteristics of planktonic communities directly affect the transport of organic carbon in the upper water column and the efficiency of carbon sequestration in the MCP. As the largest carbon sink in the global ocean, the carbon sequestration capacity of the Antarctic may be diminishing.

In this paper, a high spatial-temporal resolution sea ice concentration estimation method for the Arctic melting season is proposed, aiming to improve the overestimation of sea ice concentration in seawater by the Global Navigation Satellite System-Reflectometry (GNSS-R). The method utilizes machine learning algorithms to extract feature parameters from the Delay Doppler Maps (DDM) obtained through GNSS-R and combines them with sea surface temperature data to establish a LightGBM model. The inversion results are then subjected to correlation analysis and evaluation against reference sea ice concentration values. The model’s performance is compared with the sea ice concentration product from OSI SAF, demonstrating good consistency, with correlation coefficient, mean absolute error, and root mean square error being 0.965, 0.061, and 0.090, respectively. This approach enables high-precision estimation of sea ice concentration in the Arctic marginal ice zone.

The “fairy circle” represents a unique form of spatial self-organization found within coastal salt marsh ecosystems, profoundly influencing the productivity, stability, and resilience of these wetlands. Unmanned Aerial Vehicle (UAV) imagery plays a pivotal role in precisely pinpointing the “fairy circle” locations and deciphering their temporal and spatial development trends. However, identifying “fairy circle” pixels within two-dimensional images poses a considerable technical challenge due to the subtle differences in color and shape characteristics between these pixels and their surroundings. Therefore, intelligently and accurately identify “fairy circle” pixels from two-dimensional images and form individual “fairy circle” for the identified pixels were the current technical difficulties. This paper introduced an innovative approach to extract “fairy circle” from UAV images by integrating the SAM (Segment Anything Model) visual segmentation model with random forest machine learning. This novel method accomplished the recognition and extraction of individual “fairy circle” through a two-step process: segmentation followed by classification. Initially, we established Dice (Sørensen-Dice coefficient) and IOU (Intersection Over Union) evaluation metrics, and optimize SAM’s pre-trained model parameters, which produced segmentation mask devoid of attribute information by fully automated image segmentation. Subsequently, we aligned the segmentation mask with the original image, and utilized RGB (red, green, and blue) color channels and spatial coordinates to construct a feature index for the segmentation mask. These features underwent analysis and selection based on Out-of-Bag (OOB) error reduction and feature distribution patterns. Ultimately, the refined features were employed to train a random forest model, enabling the automatic identification and classification of “fairy circle” vegetation, common vegetation, and bare flat areas. The experimental results show that the average correct extraction rate of “fairy circle” is 96.1%, and the average wrong extraction rate is 9.5%, which provides methodological and technological support for the accurate depiction of the spatial and temporal pattern of “fairy circle” as well as the processing of coastal remote sensing images by UAVs.

Sediment oxygen consumption (SOC) is an important parameter of marine sediments and an important characterization parameter of the rate of organic carbon mineralization in seafloor sediments, and the study of SOC can help us to understand the carbon cycling process in the whole ocean. As one of the most important and active sites for organic carbon mineralization and burial, marginal seas have received widespread attention and research around the world, but there is still a lack of relevant attention to the Chinese marginal sea region with typical seasonal variations of the marine environment, especially the Yellow Sea and Bohai Sea. In this paper, the intact core incubation was used to study the SOC in the Yellow Sea and Bohai Sea in April, July and October 2022, and the results showed that the rates of SOC ranged from 7.11 mmol/(m²·d) to 17.33 mmol/(m²·d). There was no significant difference between the SOC of the Yellow Sea and the Bohai Sea in spring (ANOVA, p > 0.05), and the SOC of the Yellow Sea was lower than Bohai Sea in summer (ANOVA, p < 0.01) and autumn (ANOVA, p < 0.01); the SOC of the Yellow Sea was the largest in spring and the smallest in summer, and there was no significant difference between the SOC of the Bohai Sea in summer and autumn, which were significantly higher than that of spring (ANOVA, p < 0.05). Temperature and sediment Chl a concentration were the influencing factors. Meanwhile, the SOC was used to assess the rate of benthic organic carbon mineralization. When compared with the primary productivity, the results indicated that the contribution of benthic organic carbon mineralization to primary productivity in the Bohai Sea ranged from 42.8% to 74.5%, which was one of the key links in the carbon cycle of the Bohai Sea, while the benthic organic carbon mineralization in Yellow Sea plays a less significant role in the carbon cycle of the Yellow Sea carbon cycle than Bohai Sea. This paper systematically studied the SOC in the Yellow Sea and Bohai Sea and its spatial and temporal distribution characteristics, exploring the contribution of organic carbon mineralization to primary productivity in the Yellow Sea and Bohai Sea, which provided theoretical support for the understanding of organic carbon mineralization and burial in the Yellow Sea and Bohai Sea.