A dataset of the high water lines extracted from 113 Landsat images from 1986 to 2019 and the measured profile data from 2015 to 2019 were used to examine the middle-term to long-term shoreline process and driver at the embayment scale in this paper. The results show that the western and eastern beaches of the Qiwang Bay, which is separated by one small bedrock headland, have four and three different spatial characteristics, respectively. More than half of the shorelines behaved nonlinear in their variation trends. Thus, we use the Mann-Kendall method to solve the problem of the lack of basis for the division of time periods. In addition, the east breakwater resulted in the unstable embayment planform due to changing the position of the controlling “headland” and therefore the longshore sediment transport from west to east is the main driver of the most recent shoreline. And the intervening small bedrock headland also influenced the spatial variability of erosion and accretion at the Qiwang Bay. These findings will have important theoretical and practical significance for predicting further shoreline position and reducing the risk of shoreline erosion.

The internal solitary wave and mesoscale eddy are common mesoscale dynamic processes in the northern South China Sea. In this paper, we use the Terra/Aqua-MODIS, ENVISAT ASAR and multi-source satellite altimeter data from 2010 to 2015 to realize remote sensing of isolated waves and mesoscale eddy in the South China Sea, and analyze the influence of mesoscale eddy on the propagation direction of internal solitary wave. The results show that the mesoscale eddy and the internal solitary wave coexist mainly in the northeastern part of the South China Sea. When the two coexisted, the cyclone (cold eddy) caused the internal solitary wave to deviate from the original propagation direction and propagation and spread to the west-south direction. The anticyclonic (warm eddy) causes the internal solitary wave to spread westward to the north, and the cyclone and the anticyclone change the direction of the internal solitary wave propagation is just opposite. The coexistence time of internal solitary wave and mesoscale eddy is mainly concentrated from March to September, and the propagation direction of internal solitary wave is almost unchanged in March due to the interaction of cyclone and anticyclone. In April and May, the internal solitary wave deviated from its original propagation direction and propagated southward mainly due to cyclone. From June to September, the internal solitary wave deviated from its original direction and propagated northward, mainly under the influence of anticyclone. In April and May, the internal solitary wave deviated from its original propagation direction and propagated southward mainly due to cyclone. From June to September, the internal solitary wave deviated from its original direction and propagated northward, mainly under the influence of anticyclone. The effect of mesoscale eddy on the propagation direction of internal solitary wave is investigated by remote sensing, and the results are in agreement with the field observation.

In order to quantitatively study the sheltering effect between multiple ridges keel on sea ice drift, the laboratory experiment is carried out in a tank, which is 0.45 m deep. The shape of keel models is a triangle with 45° slope angle, 4 keel depths and 9 keel spacings are selected in the experiments. The variations of the front and back keel drag force and its ratio under wake effect is investigated. The drag force on the front keel is not affected by the back keel and keeps a linear relationship with the square of keel velocity; however, the drag force of back keel appears negative value (opposite direction) when the keel spacing is small. With the increase of the spacing the drag coefficient of the back keel first decreases and then increases to a constant. The variation of the ratio of drag forces between the front and back can be described by an exponential sheltering function, which is related to keel spacings and keel depths, and independent of keel velocity. Compared with the sheltering functions which are used in present sea ice models, the exponential formula is given and improves our understanding about sheltering function in sea ice dynamic model.

Current CMIP6 climate models (such as CESM2 and NESM3) use constant snow density, while those models that focus on snow depth and density changes (such as SnowModel-LG) use empirical snow density formulas. Comparing the modeled snow depth with those observed by the CryoSat-2 satellite, it is found that from the perspective of the spatial distribution and average value of the snow depth, it is difficult to detect the effects of varying snow density on the simulation of snow depth in the Arctic Ocean. The model improvement and its mechanism from varying snow depth is still to be further studied. Here an empirical snow density model considering meteorological factors such as air temperature, wind etc., is applied to the SNOTEL observational site to carry out the following sensitivity experiments for different factors: A. snow density model considering all meteorological factors; B. constant snow density model; C. same as A but the influence of wind on the densification is not considered and D. same as A but the influence of temperature on the densification is not considered. The root mean square error of snow depth simulated by experiments A, B, C and D from November 1, 2018 to May 10, 2019 are 4.2 cm, 4.8 cm, 25.9 cm, and 4.2 cm, respectively. The results show that the mean snow density and depth simulated by the varying snow density model are close to the results using constant snow density, but the root mean square error of the simulated snow depth from Case A is the smallest, and the Case A simulation can reproduce the high frequency variations of snow depth on the time scale of several days to ten days. In the meantime, the relative errors in the period with high-frequency snow depth variations are also reduced as they are found to be related. In addition, it is also found that the influence of temperature on snow densification is much smaller than that of wind.

In this study, four samples of Gymnothorax mucifer were collected from the fishery market of Xiamen City, Fujian Province during the year 2020 to 2021, which were newly recorded in the coastal waters of China. Previously, the species had only been recorded in Australia and Hawaii, and was considered to be a synonym of Gymnothorax kidako. Detailed morphological characteristics of four G. mucifer species were analyzed, and the molecular identifications as well as phylogenetic constructions were also carried out basing on DNA barcode COI gene in this study. The main distinguishing characteristics of G. mucifer were as follow: the colour of the body was yellowish brown, the front of the head was slightly purple, the whole body was covered with slender, sparse, irregular branch-liked brown marking and the markings became darker and thicker near the tail, forming clear net-liked patterns; the margin on the anal fin was white, and became serial pale blotches on posterior part of the tail; total length was 1.01 times of standard length and 8.00−8.39 times of head length; the maxillary teeth were 8−10 and dentary teeth were 14−20 on each side, both teeth were uniserial; the median inter maxillary teeth were slender and uniserial; the total vertebrae were 117−139 and mean vertebral formula was 6-47-130. Basing on the COI gene analysis, the genetic distance between G. mucifer and G. kidako was 0.074, which was greater than the value 2% (0.020) suggested by Herbert as minimum genetic distance value to distinguishing different species, revealing that the two species might be two independent species. Morphologically, G. mucifer could also be distinguished from G. kidako by certain external characteristic: the body markings of G. mucifer were slender, sparse and inconspicuous, the front of the head was slightly purple, and the white margin of the anal fin broke into series of pale blotches on posterior part of the tail; the markings of G. kidako were obvious and thick with darker colour, the front of the head was yellow-white, the white margin on the anal fin continuous to the tip of the tail. The results of the study provided a taxonomic basis for the systematic classification and the species list revision of the Gymnothorax fish in our country.

Based on the sea surface temperature (SST), which is the most dominant environmental climate factor affecting the distribution of squid, the potential habitat changes of Ommastrephes bartramii in July to October in 1996−2005, 2021−2030, 2051−2060 and 2090−2100 are analyzed using maximum entropy (MaxEnt) model with the historical climate data from 1996 to 2005 and the projection climate data from RCP4.5 and RCP8.5 scenarios. The results show that the fishing grounds of O. bartramii performs a seasonal north-south migration. Meanwhile, as the seasonal north-south migration of O. bartramii may be affected by the suitable SST range in fishing season, with the feature climate change the potential habitat distribution of O. bartramii from July to September in 2021−2030, 2051−2060 and 2090−2100 will move northward and the suitable habitat area will increase compare to 1996−2005 under both scenarios of RCP4.5 and RCP8.0. Under scenario of RCP4.5, the potential most suitable habitat for O. bartramii will move northward by 1°−2° and the suitable habitat area will increase by 3%−13% by the end of the 21st century. Under scenario of RCP 8.5, the potential most suitable habitat for O. bartramii will move northward by 3°−5° and the suitable habitat area will increase by 42%−80% by the end of the 21st century.

Sillago sihama is an important fishery species in China and plays an important role in the marine ecosystem of the Yellow Sea. Species distribution models can be used to predict its distribution by establishing the relationships between its abundance and environmental factors. However, due to high mobility of the marine animals, the relationship between their distribution and environmental factors is often nonlinear and variable with spatial locations. Based on data collected from bottom trawl survey in the Shandong coastal waters in autumn of 2016, both generalized additive model (GAM) and geographically weighted regression (GWR) model were used to analyze nonlinear and spatial nonstationary relationships between distribution of the species and environmental factors, and results from the two models were compared. Results from the GAM indicated that the main environmental factors were depth, sea bottom temperature and salinity, and depth had the largest deviance explained (23.50%). GWR model results showed that there were spatial non-stationary relationships between distribution of the species and depth and sea bottom temperature. GWR model results indicated a negative correlation between depth and biomass of the species, and a positive correlation between sea bottom temperature and biomass of species. Regarding performance of the models, GWR model showed advantages over GAM in identifying influencing factors and predicting distribution, and GWR model was recommended for use in similar applications.

The changes of shoreline caused by human activities affect the kinematic characteristics of tidal wave in the propagation process. Based on the hourly tidal level data from five tidal stations in the Wenzhou Bay from 1984 to 2019, the temporal and spatial variation of tidal wave patterns in this area and deconstructed the contribution of major tidal clusters to the tidal asymmetry is analyzed in this study . The results show that the tidal patterns in Oujiang Estuary and Yueqing Bay, two semi-enclosed embayments of the Wenzhou Bay, are obviously different. The tidal symmetry of Oujiang Estuary is flood-dominant tide and the tidal asymmetry increases continuously in the upstream direction, whilst tide is ebb-dominant in Yueqing Bay. Moreover, the tidal asymmetry shows distinct seasonal variation. The skewness ($ \gamma $) reaches maximum in June to July and December to January in the Wenzhou Bay. The tidal asymmetry in this area is mainly controlled by the component groups such as M₂/M₄, M₂/S₂/MS₄ and M₂/N₂/MN₄. The skewness caused by nonlinear interactions from M₂/S₂/MS₄, M₂/N₂/MN₄, O₁/K₁/M₂ shows obvious seasonal variation. Since 2000, the tidal asymmetry of Wenzhou Bay has been decreasing, which is related to the frequent reclamation surrounding the Oujiang Estuary.

With the intensification of global warming, the source sink process of carbon in the Arctic shelf-edge sea is becoming more and more important in the study of global carbon cycle. As a typical continental shelf marginal sea in the Arctic Ocean, the source, transportation and burial of sedimentary organic carbon in this area are unique under the influence of rivers, sea ice, marine primary productivity and coastal erosion. Based on the sampling of suspended particulate matter (SPM) and hydrological data obtained from the second Sino-Russian Arctic joint expedition during late summer and early fall in 2018, we foucus on the distribution characteristics, sources and influencing factors of particulate organic carbon (POC) in the Laptev Sea. The results show that POC ranges from 35.27 μg/L to 1 185.58 μg/L, with an average of 172.65 μg/L. Under the effect of river input, coastal erosion and marine primary productivity, the distribution of surface POC shows a decreased trend from near shore towards offshore; the bottom POC is mainly controlled by sediments resuspension, and the high content of POC appears in the east of Lena River Delta. There was a significant positive correlation between SPM concentration and POC concentration, indicating its direct impact on the occurrence of POC; a more positive relation is found among the bottom layer samples, which may indicate the varied origin of POC in different layers. The value of δ¹³C_POC in study area value is between −31.03‰ and −25.79‰, and the value of δ¹³C in surface layers is obviously depleted compared with the bottom layer, which is even lower than the end-member of the surrounding terrestrial contributor, suggesting that these POC is not derived from land-based origin. The utilization of the terrestrial POC degraded dissolved inorganic carbon by offshore phytoplankton maybe responsible for this depletion of δ¹³C offshore, which could also be an important process on the supply and source apportionment of POC in this Arctic coastal area.

A multivariate long short term memory (LSTM) neural network model was developed for the Pacific decadal oscillation (PDO) index time series prediction using sea level pressure, sea level height, ocean heat content data and sea ice concentration from 1921 to 2020 as forecast elements of the PDO index. The PDO index prediction results of different time series from 2011 to 2020 were compared and analyzed, and finally the PDO index forecasting from 2021 to 2030 is realized by using the multivariate LSTM neural network model. The results show that the average correlation coefficient and root mean square error of the predicted value and the observed value of the multivariate LSTM model after cross-validation are 0.70 and 0.62, respectively. PDO will remain in the cold phase in the next ten years, and the PDO index will fluctuate twice, there will be a minimum in 2025. Compared with other time series forecasting models, the multivariate LSTM neural network model used in this paper has less error in forecasting results and good fitting effect, which can be used as a new method of predicting PDO index.