The westward expansion of the Beaufort Gyre (BG) is essential for the transport and redistribution of heat, freshwater, and nutrients in the western Arctic Ocean. During 2003–2014, the core and edge of BG expended westward significantly, as indicated by satellite dataset and reanalysis data. An idealized model with spatial and temporal variation of the anticyclonic surface stress also demonstrates the BG westward expansion. Based on the potential vorticity (PV) budget analysis of the idealized model results, the westward advection of thickness perturbation near the continental slope transports the negative PV anomalies generated by variations in surface stress, which contributes to the westward expansion of the BG. Idealized model case studies are conducted to investigate how displacement of surface stress field location, enlargement of surface stress field, and asymmetry enhancement of surface stress field affect the westward expansion and asymmetry of the BG via advective processes. The enlargement of surface stress led to the most significant enhancement in advection of thickness perturbation and a pronounced westward expansion of BG, with both the core and edge of BG westward expansion over 300 km. The asymmetry enhancement of surface stress significantly contributes to the pronounced zonal asymmetry of BG. Advection of thickness perturbation exerts a more pronounced influence on the western edge of BG, while advection of vorticity perturbation has a greater impact on the core of BG. The beta effect is essential for the westward expansion of the BG, a more pronounced beta effect leads to a more rapid westward expansion.
The rapid warming of global oceans challenges traditional fixed baseline approaches for identifying marine heatwaves (MHWs), particularly in rapidly warming regions. We systematically compared three baseline approaches—fixed (FBA), linearly detrended fixed (LDFBA), and shifting (SBA) baselines approach—to evaluate MHW characteristics and their biological impacts in the South China Sea (SCS). Under FBA, MHWs showed significant increasing trends during 1982–2022, with rates of 15.6 d/decade in duration and 1.0 events/decade in frequency, while LDFBA showed no significant temporal trend and SBA exhibited moderate increases (9.9 d/decade). Despite substantial differences in MHW detection, all three approaches revealed similar spatial patterns of instantaneous chlorophyll-a (Chl-a) responses, with mean decreases of 9
In the field of water color remote sensing and environmental monitoring, smartphone-acquired water imagery has emerged as a powerful tool for crowdsourced surveillance. However, the widely used Joint Photographic Experts Group (JPEG) format presents significant challenges for accurate water-leaving reflectance (remote sensing reflectance) correction and water quality monitoring. To address these issues, this study puts forward a high-precision method for water-leaving reflectance calibration and transparency inversion using RAW-format photos captured by smartphones. The digital number (DN) values in RAW-format photos exhibit a linear relationship with reflectance, with the regression line closely passing through the origin. This allows for the conversion of DN values of water and sky photos into reflectance using a single gray reference card with known reflectance, enabling more precise calculation of water-leaving reflectance. Validation with data from 267 sampling points across 15 study regions in China showed an average R2 of 0.93 in the blue, green, and red spectral bands. When the calibrated reflectance was applied to transparency inversion, the modified QAA-RGB semi-analytical model outperformed the semi-empirical model, achieving an R2 of 0.78. These results provide a solid technical foundation for using smartphones to achieve high-precision, crowdsourced water quality monitoring in the context of water color remote sensing. This study focuses on the linear response characteristic of RAW format data. Through large-scale, cross-regional validation, smartphone-based water quality monitoring is extended from reflectance correction to parameter inversion. Consequently, threefold breakthroughs are achieved in methodological simplification, validation robustness, and application scope.
Green tide caused by floating macroalgae blooms of Ulva prolifera, is a typical ecological disaster in the Yellow Sea of China, severely impacting the marine environment. This study, based on Moderate Resolution Imaging Spectroradiometer (MODIS) standard products, analyzes the impact of residual mixed pixels of Ulva prolifera on chlorophyll-a (Chl-a) concentration and the spectral slope of particle backscattering (η). Results indicate that such mixed pixels lead to overestimated Chl a and underestimated η. To address this, a gradient-based mixed-pixel removal method was developed, reconstructing daily Chl a and η datasets from June to July for 2003–2022. By integrating data from the remaining months, the analysis revealed the impacts of green tide outbreaks on these two parameters and their interrelationship. Compared with 2003–2007, Chl a showed no significant change in June from 2008 to 2022, but increased significantly in July, with the monthly trend shifting to higher values in July than June. The change can be preliminarily explained by nutrient competition between Ulva prolifera and phytoplankton: rapid growth of Ulva prolifera in June limits phytoplankton growth, whereas reduced competition in July allows rapid proliferation of phytoplankton. There were no significant changes in η in both June and July after green tide outbreaks; however, the trend of η changed from significantly increasing to significantly decreasing in certain areas from June to July, indicating that green tide outbreaks altered the proportion of different particle sizes in the water. The relationship between Chl a and η suggests that the change in particle size distribution caused by green tide may be related to the succession of dominant phytoplankton species. This study provides key data support and a scientific reference for understanding the optical response mechanisms of marine ecological parameters under green tide influence.
Water transparency is a key indicator of marine optical properties and ecological conditions, and its spatiotemporal variability in the Pacific Ocean is closely associated with the El Niño-Southern Oscillation (ENSO). In this study, water transparency across the Pacific Ocean was retrieved using the semi-analytical algorithm, and its spatial and temporal variations were examined in relation to ENSO variability. The results show that the basin-averaged annual mean transparency exhibits relatively limited variability, with persistently high transparency observed in the eastern South Pacific and the western North Pacific. In contrast, the equatorial low-latitude region (10°N–10°S) displays the most pronounced interannual and intra-annual fluctuations. Furthermore, water transparency shows a significant positive correlation with the Oceanic Niño Index (ONI), suggesting a close linkage between transparency variability and ENSO phases. This relationship may be associated with ENSO-related changes in upwelling intensity and nutrient transport, which can influence phytoplankton biomass and suspended particulate concentrations. These findings highlight a pronounced sensitivity of water transparency in the equatorial Pacific to ENSO variability and underscore its potential value for marine ecosystem monitoring and climate-related assessments.
A global sensitivity analysis of critical factors in radiometric benchmark transfer is conducted in this study using oceanic pseudo-invariant sites to quantify and control transfer uncertainty under nonideal matching conditions. Prior to the sensitivity analysis, oceanic pseudo-invariant sites are objectively identified through a screening procedure to ensure long-term radiometric stability under varying environmental conditions. Single-parameter sensitivity pre-screening is first performed to identify environmental variables that significantly affect the transfer quality at the optimally selected pseudoinvariant sites. The resulting multidimensional parameter space is defined by the following parameters: viewing zenith angle, solar zenith angle, relative azimuth angle, aerosol optical thickness at 550 nm, wind speed, and discrete spectral wavelength. A variance-based Sobol global sensitivity analysis driven by low-discrepancy sequence sampling was then conducted at 40 discrete wavelengths spanning 412–945 nm. Results indicate that, across all bands, differences in top-of-atmosphere (TOA) reflectance are dominated by angular-domain parameters. At 443 nm, for example, the sum of first-order sensitivity indices exceeds 0.99, while the normalized contributions from non-angular uncertainty sources are below 0.01; total-effect indices closely match the first-order indices, indicating weak higher-order interactions. Under these conditions, contributions from different uncertainty domains can be combined through the addition of the standard uncertainties within each domain. The proposed framework enables a transition from traditional cross-matching, focusing on spatio-temporal matching, toward sensitivity-guided geometric matching constraints and correction strategies, providing quantitative support for uncertainty budgeting and subsequent correction scheme design in ocean-satellite radiometric benchmark transfer and cross-calibration.
The high-resolution weather research forecasting (WRF) downscale simulations over China’s offshore regions are used to produce high accuracy wind data, which is evaluated by 19 wind measurement towers located across the coastal waters of China. Wind data from WRF exhibit an excellent consistency compared to observations via various validations, including statistical analysis, score computation and extreme event analysis. Then a comprehensive evaluation of exploitable wind energy resources and wind speed extreme value analysis based on regional frequency analysis are investigated. These systematic analyses provide insights for optimizing offshore wind farm siting and ensuring structural reliability. Based on comprehensive assessment of annual mean wind speeds (6–10 m/s), wind power density (400–1 100 W/m2), and capacity factors (0.3–0.8), the Taiwan Strait to offshore Jiangsu waters region (117°E–128°E, 22°N–34°N) is identified as having optimal wind power development potential. The interannual standard deviation of wind speed indicates higher variability in the northern Bohai and Yellow seas (0.8–1.0 m/s), while the Taiwan Strait and East China Sea show relatively stable conditions (0.2–0.6 m/s). The 50-year return period extreme wind speeds are relatively small (<30 m/s) in the shallow water region (<25 m). The coastal areas of Jiangsu, Zhejiang, and Fujian, with shallow water, and the lower wind speed extreme, are very suitable for development of offshore wind farms.
Marine fungi, inhabiting unique ecological niches, have evolved distinct metabolic profiles compared to terrestrial fungi, leading to the production of diverse secondary metabolites with novel structures and significant bioactivities. In this study, we employed the “Species-NPAtlas-Activity” strategy to conduct a comprehensive screening of 32 marine fungal strains collected from the Xisha Islands. Among these strains, Westerdykella aurantiaca SYSU-MS9520 was selected for further investigation due to its high potential for secondary metabolite biosynthesis. Genomic analysis revealed numerous biosynthetic gene clusters (BGCs) in this strain. From its culture broth, twelve polyketides were isolated, including three new isochromenes (1–3), three new isochromanones (4–6) along with one known analogue (7), a new methylbenzoic acid derivative (8), one known isobenzofuranone (9) and three known depsidones (10–12). Their structures were fully elucidated using a combination of analytical techniques, including ultraviolet-visible (UV) spectroscopy, infrared (IR) spectroscopy, high-resolution mass spectrometry (HRMS), nuclear magnetic resonance (NMR) spectroscopy, electronic circular dichroism (ECD) spectroscopy, and X-ray crystallography. Notably, compounds 1–7 were identified as isochromane derivatives. Moreover, compound 1 was isolated as a pair of isochromenone enantiomers, whose absolute configurations were established by X-ray crystallography analysis, chiral separation, and ECD calculation. Antibacterial assay showed that depsidone compound 10 exhibited attractive antibacterial effects against the Gram-positive strains Staphylococcus aureus and Bacillus subtilis with minimum inhibitory concentration (MIC) values of 12.5 and 50 µg/mL, respectively.
The Xisha Islands in the South China Sea is a key area for Cenozoic hydrocarbon exploration, featured by widespread reef carbonate systems. Among them, dolomite reservoirs play a significant role in carbonate hydrocarbon exploration. To explore the dolomitization processes in this region, an integrated study was conducted on reef dolomites from the Sanya Formation (Member 1) in Well Xike 1. The approaches taken in this study include core description, petrography, scanning electron microscopy-energy dispersive X-ray spectroscopy (SEM-EDS), geochemical analysis on major and trace elements, rare earth elements (REEs), stable isotope (C, O, Mg) and fluid inclusion analyses. Petrographic observations reveal that the dolomites are primarily composed of euhedral to subhedral crystals, ranging in size from silt to fine crystals with local presence of microcrystal. Common features include cloudy cores with clear rims, point and planar contacts, and occasional dedolomitization. Cathodoluminescence (CL) imaging presents characteristics of the dolomites, predominantly orange luminescence with minor purple and reddish-orange zones, clear oscillatory zoning, and local saddle dolomite, collectively demonstrate a multi-phase diagenetic evolution. Geochemical data of dolomites show low concentration of Fe, Mn, and Sr, with δ13C values consistent with modern marine carbonates and δ18O values slightly depleted. Meanwhile, all Z-values are larger than 120 (mean 131), proving a seawater-dominated, semi-restricted to semi-open diagenetic setting. Furthermore, hydrothermal influence is also clearly detected with negative Ce and positive Eu anomalies in the left-leaning REE patterns and high homogenization temperature (mean 188 °C) from fluid inclusions. With the vertical variation trend of Mg isotopes, the current study proposes the two-stage genetic model regarding dolomitization of the studied region: the initial stage began with shallow reflux of hypersaline brines, which infiltrated through the limestone stratum and formed micritic to silt-crystalline dolomite; in the later stage, recrystallization and more pervasive dolomitization occurred with hotter fluids during deeper burial. This study contributes critical insights to the complex diagenetic history and fluid evolution of Cenozoic carbonate reservoirs in the region.
Cyanobacterial blooms and aquatic vegetation exhibit similar spectral characteristics, which limits the discriminative capability of traditional spectral index-based methods. Moreover, existing deep learning models still suffer from omission and misclassification errors in fragmented boundary regions. This study proposes a semantic segmentation network, termed global-local multi-scale feature extraction network (GLMFENet), which is designed to jointly exploit global and local multi-scale features. Based on an encoder-decoder architecture, GLMFENet incorporates a global feature extraction module to model long-range semantic dependencies, together with a spatial attention-guided local multi-scale feature extraction module to enhance fine-grained discriminability. By synergistically integrating global contextual information with local structural features, the proposed network enables accurate differentiation between cyanobacterial blooms and aquatic vegetation. A dataset comprising 467 Sentinel-2 MSI images collected from multiple lakes across China was established for model training and validation. GLMFENet achieved a mean IoU (mloU) of 84.6
Water clarity, a straightforward and intuitive indicator of water quality, provides critical insights into the dynamics of aquatic and lake ecosystems and is commonly quantified using Secchi disk depth (ZSD). To address the optical complexity of turbid waters in Taihu Lake, we developed a Random Forest Regression-based ZSD (RFR-ZSD) algorithm, leveraging in situ measurements for ZSD estimation. Following the validation of atmospheric correction performance and the RFR-ZSD model using Sentinel-2 Multispectral Instrument (MSI) imagery, the algorithm was employed to derive ZSD from MSI images acquired between 2016 and 2023. The results indicate a slight overall improvement in water quality in Taihu Lake, with increasing water clarity and consistently higher ZSD values observed during summer and autumn compared to spring and winter. Analysis of meteorological influences suggests that variations in ZSD are primarily associated with internal hydrodynamic conditions, where wind-driven disturbance directly regulates water transparency, while air temperature and wind direction reflect broader seasonal and atmospheric stability; short-term precipitation exhibits no discernible influence during the observation period. Additionally, the impact of underwater tunnel construction on the spatial distribution of ZSD in Zhushan Bay and Meiliang Bay was assessed, revealing pronounced disparities in water clarity on opposite sides of the tunnel. This study underscores the effectiveness of integrating Sentinel-2 MSI imagery with the RFR algorithm for monitoring lake water quality dynamics. The high spatial resolution of Sentinel-2 data enables the detection of both natural variability and localized anthropogenic impacts on water clarity, providing robust support for long-term ecological assessment and management of large shallow lakes.
This study reports the discovery of Vittina lambda sp. nov., a new neritid species from Hainan Island, South China Sea, alongside three new records: Nerita antiquata, Clithon eudeli, and C. peguense. V. lambda sp. nov., distinguished by unique Λ-shaped shell color patterns and radular morphology (narrower central teeth, sharper marginal teeth), was historically misidentified within V. coromandeliana populations. Integrative taxonomy combining mitochondrial COI sequences (revealing>5
The Qingshui River Basin is the largest tributary entering the Huanghe (Yellow) River within Ningxia of China, contributing about 49
The chlorophyll-a (Chl-a) concentration is a core indicator for assessing the health of marine ecosystems. With single-photon-sensitive detection capability, satellite-borne lidar technology can penetrate the sea surface to obtain subsurface optical signals in polar nights, overcoming the constraints of passive remote sensing. In this study, the backward scattering coefficient of particulate matter (bbp) was inverted based on the lidar equation, and a trinomial fitting model was developed to convert bbp(532) to Chl a using the Biogeochemical-Argo (BGC-Argo) measured data in the Norwegian Sea. The retrieval accuracy was verified using corresponding BGC-Argo buoy measurements and Moderate Resolution Imaging Spectroradiometer (MODIS) water color data. Four Ice, Cloud, and Land Elevation Satellite-2 (ICESat-2) orbital photon counting lidar tracks data (2020–2024) of the Norwegian Sea were selected to retrieve the spatial and temporal distribution characteristics of the vertical Chl-a profile. Horizontally, high Chl-a concentration (up to >2 mg/m3 owing to warm-warm convergence) was observed at the eastern edge and in the central sea, followed by at turbid nearshore waters at low latitudes, and low concentrations (<0.5 mg/m3) were observed in the open sea. Vertically, significant stratification was observed in the depth range of 3–14 m. We verified the feasibility of ICESat-2 to invert the vertical profiles of Chl a under complex sea conditions at high latitudes, and filled the observation gap of passive remote sensing in polar nights.
The South China Sea (SCS), due to its unique geographical location and marine environment, harbors a diverse and representative coral reef ecosystem. This study examines the trophic structure of major marine organisms within the coral reef ecosystem of the southern SCS through stable isotope analysis (δ13C and δ15N). The analysis is based on field samples collected during the spring of 2023 and 2024. The δ13C values of consumers spanned from −21.72‰ to −3.26‰, and the δ15N values ranged from 2.42‰ to 10.97‰, which indicates a well-defined and continuous trophic spectrum. The trophic levels (TLs) of fish varied from 1.80 to 3.35, whereas those of invertebrates ranged from 1.08 to 2.76, suggesting a structured ecosystem in the study area. MixSIAR modeling revealed that macroalgae, plankton, and organic matter on coral surfaces (OMCs) were the predominant basal food sources, supporting five major trophic groups: carnivores, corallivores, planktivores, macroalgae-consuming herbivores and omnivores. Notably, several species that traditionally feed on coral, such as Chaetodon ornatissimus and Heniochus chrysostomus, demonstrated an increased reliance on algal resources, implying a possible change in feeding strategy. Additionally, the diet preference of Acanthaster planci showed a strong reliance on crustose coralline algae (CCA), with juveniles experiencing growth stagnation or adults suffering from long-term starvation being the predominant individuals. This highlights the importance of food-source monitoring for understanding the population dynamics of A. planci. This study provides basic data and novel perspectives for coral reef conservation and sustainable management in the coral reef ecosystems of the western Pacific Ocean.
Phytoplankton primary productivity (PP) is a key indicator of carbon fixation and ecosystem functioning in inland waters. Using high-resolution Sentinel-2 imagery, this study developed a regionalized Vertically Generalized Production Model (VGPM) optimized for highly turbid, shallow lakes to estimate modeled PP dynamics in Taihu Lake from 2019 to 2025. The multi-year mean daily PP in Taihu Lake ranged from 820.68 to 1 527.11 mg C/(m2·d), with a lake-wide average of ∼1 232.28 mg C/(m2∼d). Spatially, modeled PP decreased from the bays to the central basin and from west to east, with Zhushan and Meiliang bays maintaining the highest productivity (up to ∼1 667.93 mg C/(m2·d)). At the same time, the eastern region dominated by aquatic vegetation showed lower values (<950 mg C/(m2·d)). Temporally, modeled PP exhibited a strong seasonal cycle, with higher values in summer and autumn and lower values in winter and spring, forming a bimodal pattern with peaks in May and September (>1 800 mg C/(m2·d)). Moderate water temperature and sufficient light availability promoted productivity, whereas turbidity (light attenuation) limited phytoplankton growth. By integrating corrections for phytoplankton vertical distribution, algal bloom classification, and the removal of aquatic vegetation signals, the model’s performance was greatly improved. The Google Earth Engine (GEE)-based workflow enabled efficient long-term modeled PP monitoring and provides a transferable framework for quantifying carbon cycling, guiding eutrophic lake restoration, and assessing climate responses.
Muddy tidal flats (hereafter referred to as mudflats) dominate China’s coastal tidal wetlands. A comprehensive assessment of their carbon stock characteristics is thus essential for refining the understanding of China’s marine carbon sink capacity. In this study, sediment core sampling was conducted to systematically analyze sedimentary organic carbon (SOC) content, grain-size parameters, clay mineral composition, and carbon stock characteristics of mudflat sediments from different bays in Zhejiang Province. The results indicate that muddy sediments predominate in all three studied bays. The semi-enclosed Yanpu Bay (YP Bay), with stable hydrodynamics and small-river input, has significantly higher SOC content (0.63
Particulate organic carbon (POC) is critical to the coastal carbon cycle of the Yellow Sea (YS), a typical marginal sea, but its interannual-to-multiyear variability and driving mechanisms remain unclear due to the limitations of traditional linear or single-factor analyses. To address this gap, we employed model-derived surface POC data from 2003 to 2023 and wavelet coherence analysis to investigate POC variability and regulating mechanisms at multi-scales in the YS. The dominant spatiotemporal mode of surface POC exhibits pronounced seasonality, a consistent coastal-offshore gradient (higher in coastal waters), and a spring maximum, which is regulated by the seasonal alternation controlled by coastal production or input and offshore stratification or ventilation. Wavelet analysis reveals distinct subregional differences in driving mechanisms of POC variability at interannual-to-multiyear scales. Colored dissolved organic matter (CDOM) is the optimal single driver in most subregions, while sea surface temperature (SST) dominates in the Southern Yellow Sea Cold Water Mass (SYSCWM). Specific optimal multi-factor combinations include CDOM+sea surface wind speed (SSW) in the Northern Yellow Sea Cold Water Mass (NYSCWM), chlorophyll a (Chl a)+Photosynthetically Available Radiation (PAR)+partial pressure of carbon dioxide (pCO2) in the SYSCWM, CDOM+Chl a+suspended matter (SPM) in the Jiangsu Shoal (JSS) and CDOM+sea surface salinity (SSS) in the Changjiang River estuary (CRE). These findings clarify the subregional heterogeneity of POC variability and its driving mechanisms in the YS at interannual-to-multiyear scales, and provide a robust scientific basis for accurate regional carbon budget assessments and the optimization of marine numerical models.
Taihu Lake is one of the most severely eutrophic large lakes worldwide, characterized by a complex ecosystem where algal blooms and aquatic vegetation coexist, and it serves as a critical regional water source. Identifying the channels through which algal blooms invade vegetated areas and understanding their underlying mechanisms is essential for effective lake management. In this study, multi-source remote-sensing data were used to extract the spatial patterns of algal bloom frequency and aquatic vegetation over the past 22 years. We detected the major encroachment routes of algal blooms into vegetated zones and revealed the long-term interaction mechanisms between algae and vegetation. Algal blooms were persistently concentrated in the western and northern bays, exhibiting an overall fluctuating upward trend. Aquatic vegetation was mainly distributed in four eastern bays, remained relatively stable before 2015, declined markedly in 2016 due to dredging and shoreline modifications, and then gradually recovered. A stable negative response relationship was observed between algal blooms and aquatic vegetation, jointly regulated by seasonal climate and nutrient structure. High temperatures and intense precipitation in summer facilitated bloom expansion, whereas low temperatures in winter favored vegetation recovery. Meanwhile, lower TN to TP ratios suppressed vegetation growth and enhanced bloom encroachment during summer. Three primary channels of algal bloom encroachment into vegetated areas were identified: the northern shoreline of Gonghu Bay, the surroundings of Dongshan Island, and the southern marginal zone. Among them, the water intake in northern Gonghu Bay represents the most vulnerable zone requiring priority control. Although bloom risk is relatively low for the central and southern water intakes, maintaining the existing hydrodynamic regime and vegetation structure remains essential to prevent vegetation degradation, and enhanced monitoring is needed to identify in situ bloom occurrences within vegetated regions.
The Zonag Lake basin is situated in the hinterlands of the Qinghai-Xizang Plateau (QXP) and serves as a renowned calving ground for Tibetan antelopes. Zonag Lake is also a typical large inland lake on the QXP. Over the past three decades, the lake areas within the Zonag Lake basin have generally shown an expanding trend. However, different lakes therein have exhibited varying trend over time. Between 1988 and 1995, Zonag Lake, Kusai Lake, Haidingnor Lake, and Salt Lake all experienced shrinkage. After the 2011 breach of Zonag Lake, it continued to shrink, the downstream lakes, such as Kusai Lake, Haidingnor Lake, and Salt Lake consistently expanded. After the 2019 implementation of the Salt Lake drainage project, the water levels began to decline in 2020. The Zonag Lake basin has experienced an average annual precipitation of 308.5 mm over the past six decades, with a general upward trend in rainfall amounting to an average increase rate of 15.1
