Accurate extraction of marine raft aquaculture area information is of great significance for marine resource management and environmental monitoring. But the raft culture area is often submerged in water with weak data signal areas, resulting in low extraction accuracy based on optical images alone. Therefore, this paper takes Weihai Rongcheng Bay as the research area, and improves the U-Net neural network by adding channel attention mechanism to combine the spectral information of GF-2 optical image and the texture information of GF-3 radar image, trying to improve the extraction accuracy of raft aquaculture area. The results show that: （1) Whether it is a single optical image or a fusion image of optical and radar images, the overall accuracy of the prediction results of the U-Net neural network with channel attention mechanism will be improved, with an increase of 2.21%−4.12%. （2) Using the improved U-Net neural network to process the fusion data, the overall accuracy is 95.75%, which is 4.3% higher than that of only using GF-2 image. （3) For weak signal region, the overall accuracy and Kappa coefficient of extraction based on improved network and data fusion are 91.61% and 0.827 7, respectively. This method can effectively extract the weak signal area of marine raft aquaculture area, and can provide technical support for marine aquaculture area statistics and marine environment detection.

The virtual ocean platform constructed by digital twin technology can further realize the 3D visualization and analysis of seabed environmental monitoring data. Based on the long-term in-situ monitoring data of the deep seabed, this paper uses Unity3D technology to build a virtual marine environment with multi-model fusion, and initially establishes a digital twin system for marine engineering geological environment; combines MATLAB and ArcGIS data analysis technology to realize intelligent monitoring, data analysis, human-computer interaction and auxiliary decision-making; the study further constructs a virtual environment particle system to conduct a 3D visualization and analysis of the near-bottom suspended sediment concentration elevation event in the northern part of the South China Sea. The results show that there are large spatial and temporal differences in suspended matter concentration, particle number and aggregation degree in the virtual environmental particle system. In particular, when the suspended matter concentration is maintained at a high level, particles are found to collide and overlap with each other, and the denser microclusters are derived in space. When the suspended matter concentration rises further to the peak, the number of microclusters increases and occupies most of the volume of space, forming a highly aggregated suspended matter aggregate with a wider coverage of suspended particles. This paper is based on image analysis techniques and compares the visualization results with real seafloor cameras with relative errors in the range of 0.16%–2.80%, which is highly feasible.

The Huanghe River Delta is an important ecological function area in China. It is of guiding significance to explore the distribution of suspended sediment concentration (SSC) at the estuary and its influencing factors for sediment erosion and re-suspension, and the ecological process in the estuary and coastal zone. Based on the relationship between chromaticity angle and SSC, an SSC inversion model (R² = 0.80, MRE = 11.0%, RMSE = 1.35 mg/L) suitable for the Huanghe River Estuary and its adjacent sea areas is established in this paper. With the help of the GEE (Google Earth Engine) platform, the spatial and temporal distribution characteristics and changing rules of SSC in the Huanghe River Estuary and its adjacent sea area during the 22 years from 2000 to 2021 were studied, and the influencing factors were analyzed from two aspects of nature and human activities. The annual average value of SSC in the study area showed a downward trend (−1.83 mg/(L·a)). The spatial distribution shows that SSC gradually decreases from near shore to far shore; the diffusion interval (SSC > 20 mg/(L·a)) is only 4.8−14.6 km away from the estuary, and the influence of sediment from the Yellow River on the current sediment diffusion at the estuary is limited. The wave and suspended sediment concentration have the same seasonal characteristics. During the non-water and sediment regulation period, there is a positive correlation between the effective wave height and the SSC interdecadal monthly average (r = 0.66, p < 0.01). During the period of water and sediment regulation, the current SSC at the mouth of the Huanghe River and Laizhou Bay is limited by wind speed and effective wave height, and water and sediment regulation is dominant. During the period of water and sediment regulation, there is a positive correlation between the incoming sediment coefficient and the area change rate of high-concentration areas. After the water and sediment regulation (within 16 days), the boundary of the high concentration area (SSC > 200 mg/L) is expanded from about 1.3 km to about 2.5 km from the coast.

As a frontier subject of nonlinear science, wave-current interaction has been studied by many scholars. This paper will explore the wave-current interaction under the serious sea conditions in the Northwest Pacific during the influenced by No.1 super Typhoon “Nepartak” in 2016. The results indicate that wave-current coupling model can effectively improve the simulation accuracy of significant wave height (SWH) under serious sea conditions. The influence of wave-current interaction on SWH is closely related to the angle between wave direction and sea surface current direction: when wave direction is close to current direction, wave-current interaction will decrease the SWH; in those areas where the wave and the current have the opposite direction, wave-current interaction will increase the SWH; the closer the angle between the wave and the current is to 90°, the less influence wave-current interaction has on the SWH. The maximum SWH difference whether including wave-current interaction or not is about 1.5 m.

Based on the tide level data along the tidal reach of Xijiang River from 2015 to 2017, we applied the nonstationary tidal harmonic analysis model NS-TIDE to study the spatial and temporal variation characteristics of the amplitude and phase of different cycles of tidal constituent, the change of tidal dynamics composition and tidal wave deformation characteristics, and discussed the causes of tidal level deformation and the reversal of the lowest low waters along the tidal reach. The results show that, unlike the diurnal constituent and semi-diurnal constituent along the tidal reach that have amplitude decay and phase increment, the shallow water component amplitude increases and then decreases, the amplitude of MS_f is increasing, and the phase is changing alternately; the decay rate of each tidal amplitude is higher in the flood season than in the dry season, and the phase changes more in the flood season, indicating that the tidal wave propagates upstream more slowly under the top support of river discharge. Due to the influence of river discharge and topography, the amplitude of MS_f is larger in the flood season and upstream section of the river, and the tidal dynamic composition changes from the main tidal constituent to the secondary tidal constituent; the change of amplitude ratio of M₄ and M₂ indicates significant tidal deformation, and the reversal of the lowest low waters when the amplitude ratio of MS_f to M₂ and S₂ is relatively large.

Under the background of global climate change, the extreme storm surge events caused by tropical cyclones in the Changjiang River Estuary and adjacent coastal area present non-stationary feature. In this study, a storm surge model for the Changjiang River Estuary was constructed using the ADCIRC model to reproduce the storm surges during 241 tropical cyclones affecting the Changjiang River Estuary from 1979 to 2019. By combining the non-stationary generalized extreme value distribution with the state space approach, a statistical model for capturing the non-stationarity of extreme storm surges was built to investigate the spatiotemporal variability of the extreme storm surges in the Changjiang River Estuary and its adjacent coastal area. The statistical model can well reproduce the non-stationary feature of extreme storm surges, which was mainly represented by the time-dependent location parameter. The time-dependent location parameters at the tidal gauge stations were stationary before 2008 and presented increasing trends afterwards, which was mainly caused by the increase of the annual second- and third-largest storm surges. The reoccurrence period of storm surge event with 100-year return period under the stationary assumption was reduced to around 40–80 years, indicating an increased flood risk in the Changjiang River Estuary. Combined with the changes in the intensity and path of the tropical cyclones that caused the annual second- and third-largest storm surges, it was concluded that the increasing trends of extreme storm surges were mainly caused by the increase in the intensity of the tropical cyclone that tracking northward to the offshore of the Changjiang River Estuary and veering eastwards.

Phytoplankton blooms in polar regions with seasonal sea ice cover show a unimodal seasonality. However, the bloom processes are controlled by multiple physical and biogeochemical factors, including sea ice, light availability, mixed layer depth, and nutrients; those may result in great uncertainties in simulating phytoplankton bloom by the Earth System Models (ESMs). In this study, the results of 11 Coupled Model Intercomparison Phase-6 (CMIP6) ESMs were analyzed and evaluated with various types of observational products in order to determine whether those ESMs can correctly model the phytoplankton blooms in three Arctic shelf seas, Barents Sea, Chukchi Sea, and Bering Sea. By calculating multiple indices that represent light and nutrient limitations, the error sources of simulated surface chlorophyll a concentrations were comprehensively analyzed. Our results show that the 11 ESMs can be divided into three groups based on ice-adjusted photoperiod, rate of change of mixed layer depth, and surface nitrate concentration. Some groups are characterized by the smallest bias between modeled indices and observation-based reference, and those ESMs perform best in simulating phytoplankton bloom characteristics. The other groups of ESMs differ significantly from the reference values in terms of surface nitrate and/or rate of change of mixed layer depth, resulting in delayed occurrences of annual chlorophyll a peak concentration and greater differences in corresponding peak values. In general, in addition to the two primary constraints of light and nutrients, the ESMs should also well represent the upper mixed layer controlled by temperature and salinity distributions, so as to accurately simulate the seasonal variation of surface chlorophyll a concentration. The above analyses indicate ESMs can be used in assessing polar planktonic ecosystems, and there is room for improving ecosystem-related parametrization in future ESM development.

Rare earth elements (REE) and their relationships with biogenic silica (BSiO₂), magnetic susceptibility, Al₂O₃ and Fe₂O₃ in Core DC-11 were analyzed to reveal sediment provenances and transport history by iceberg-current-atmosphere since 34 ka BP in the southeastern Scotia Sea, Antarctica. Temporal variation of REE is similar to that of Al₂O₃, indicating they mainly occur in terrigenous detritus and BSiO₂ has obvious dilution effect on them. Sediments with high REE concentration, flat shale-normalized pattern, weak positive Eu anomaly, and high La_N/Yb_N ratio during the last glacial period indicated they are transferred from the Weddell Sea and eroded from the bordering lands with relatively old crust. The increases in magnetic susceptibility, ΔAl₂O₃, TFe₂O₃/Eu ratio indicated an enhanced input of dust from South America during this period. In early Deglaciation (19.6−14.1 ka BP), increasing Eu positive anomaly and lower La_N/Yb_N ratio indicated the southern branch of Antarctic Circumpolar Current (ACC) strengthened and contributed more sediments from the South Shetland Islands and Antarctic Peninsula due to the southward shifts of oceanic fronts, while decreasing magnetic susceptibility, ΔAl₂O₃, TFe₂O₃/Eu ratios showed rapid decrease in dust supply from South America. During the Antarctic Cold Reversal period (ACR, 14.1−12.9 ka BP), sediments from the South Shetland Islands and Antarctic Peninsula decreased sharply due to cold condition and weakened ACC branch, the weakest Eu positive anomaly and highest La_N/Yb_N ratio indicated that the sediments from the Weddell Sea dominated in the core again, and the peak of ice raft debris indicated ice rafting is vital or dominant agent. In the late Deglaciation (12.9−11.7 ka BP), the return of ACC branch to the South Shetland Islands and Antarctic Peninsula contributed more to the sediments in Core DC-11; in Holocene (11.7−0 ka BP), the ACC branch in the area between the South Shetland Islands and Antarctic Peninsula was generally enhanced, and its contribution to core sediments increased to be roughly equivalent to the amount of sediments from the Weddell Sea.

Occasional species are vulnerable to external threats such as environmental changes and human activities and have important values in biodiversity conservation. However, due to their limited availability of data and associated difficulties in statistical analysis, there are few studies on the spatial distribution and their relationships with environmental factors. In this study, based on the fishery resource surveys in the Haizhou Bay conducted from 2013 to 2019, we analyzed the relationships between the distribution and environmental factors for three occasional species, Coilia mystus, Odontamblyopus rubicundus and Erisphex pottii, using generalized additive model (GAM) and random forest (RF) model. The models were compared according to their goodness of fit and the predictive performances were evaluated using cross-validation. The results showed that depth was the most significant factor affecting the distribution of C. mystus and O. rubicundus in spring and autumn, the sea bottom temperature was the most important environmental factor influencing the distribution of E. pottii in autumn. The distribution model of C. mystus had the highest deviance explanation, followed by O. rubicundus, and E. pottii had the lowest deviance explanation. The deviance explanation by the distribution models of C. mystus, O. rubicundus and E. pottii were all lower in spring than in autumn. The cross-validation showed that the area under the curve (AUC) of the three species ranged from 0.70 to 0.85, and only the AUC of C. mystus reached 0.9 in autumn; meanwhile, the AUC of the GAM prediction results were larger than those of the RF model, indicating that the prediction performance of the GAM was better than that of the RF model for the occasional species. This study would provide a reference for the selection of models for future studies of occasional species, and have guiding significance for the conservation of the occasional species.

In the real sea state, the differences in directions of component waves result in three-dimensional short-crested waves, which have many different characteristics compared to two-dimensional long-crested waves. Therefore, in present study, the whole generation and evolution processes of short-crested freak waves are experimentally simulated by dispersive and directional focusing of component waves, in order to examine the external features and relationship of the abnormal large waves occurring during the processes. The results indicate that the abnormal large wave evolves symmetrically along the peak wave direction and undergoes a three- or four-stage process, each stage is indicated by the characteristic parameters of the abnormal large wave, it implies that the characteristic parameters can be used to identify the stage which the large abnormal wave is in and predict the variation tendency; and that compared to the two-dimensional cases, the “wave group” and “deep trough” stages are likely to be skipped, as a result, the spatio-temporal spans of the generation and evolution processes are smaller for the three-dimensional cases, it implies that in three-dimensional wave fields, the deep troughs and successive large waves (wave group) associated with freak waves may have a low probability of occurrence.