This study analyzed the horizontal and vertical distribution characteristics of temperature and salinity in the central and eastern regions of the Beibu Gulf, based on conductivity measurements in summer 2022, temperature, and depth (CTD) measurement data from the Sino–Vietnamese cooperative project “Demonstration Study on Ecological Protection and Management in Typical Bays: Seasonal Survey of the Beibu Gulf”. Furthermore, the study utilized the computational results from the numerical Finite-Volume Coastal Ocean Model (FVCOM) to elucidate the intrinsic patterns that formed the temperature and salinity distribution characteristics in August 2022 from both thermodynamic and dynamic perspectives. The circulation in the Beibu Gulf drives external seawater to move northward from the bay mouth. During this movement, numerous upwelling areas are created by lateral Ekman transport. The formation of different scales of cyclonic and anticyclonic vortices and current convergence zones is influenced by topography, runoff, and the water flux from the Qiongzhou Strait, which are key factors in the formation of upwelling and downwelling. The surface circulation in August 2022 significantly differed from the 20-year average surface circulation, with an influx of 1.15 × 104 m3/s more water entering the Beibu Gulf compared to the multi-year average. The water flux from the Qiongzhou Strait is a critical factor affecting the circulation patterns in the Beibu Gulf. The northeastern waters of the Beibu Gulf are characterized by current convergence zones, where extensive upwelling occurs. The rich nutrient salts in these areas promote the reproduction and growth of phytoplankton and zooplankton, making this the most favorable ecological environment in the Beibu Gulf and serving as a natural reserve for fisheries, coral reefs, dugongs, and Bryde’s whales.
Hangzhou Bay has long experienced excessive nitrogen loading coupled with limited hydrodynamic exchange, leading to some of the highest nitrogen concentrations in China’s coastal waters. As critical land-sea ecotones, intertidal wetlands play a crucial role in mitigating nitrogen pollution across the bay. However, rapid urbanization and extensive reclamation since 1990 have led to a loss of over 50% of the intertidal wetlands in southern Hangzhou Bay. In this study we measured sediment denitrification and anammox potentials across key habitats: salt marshes (vegetated by Spartina alterniflora, Phragmites australis, and Scirpus mariqueter), bare mudflats, and shellfish aquaculture zones. We used 15N isotope tracing techniques coupled with slurry incubation experiments. Analysis of sediment physicochemical properties was conducted to elucidate the driving mechanisms of nitrogen removal. By integrating wetland landscape evolution with regional nitrogen budgets, we evaluated the nitrogen sink function of these intertidal wetlands. Our results revealed a distinct spatial hierarchy in denitrification potential, decreasing in the order: S. alterniflora (13.02 ± 3.54 μmol·N·kg−1·h−1) > shellfish aquaculture zones (12.86 ± 7.50 μmol·N·kg−1·h−1) > P. australis (11.54 ± 1.80 μmol·N·kg−1·h−1) > S. mariqueter (7.33 ± 2.08 μmol·N·kg−1·h−1) > bare mudflats (5.99 ± 1.62 μmol·N·kg−1·h−1). S. alterniflora has higher primary productivity, biomass accumulation, and a more robust root system structure. It regulates the content and availability of sediment organic carbon, the supply of nitrate nitrogen, pH, and water content. These regulations subsequently enhance denitrification. In contrast, shellfish aquaculture zones enhance denitrification potential primarily through bioturbation, which increases water content and lowers pH conditions. An integrated assessment of denitrification potential and landscape patterns revealed that, despite ongoing habitat loss, the remaining intertidal wetlands in southern Hangzhou Bay still remove about 30.65% of exogenous inorganic nitrogen. This finding underscores their critical role as effective pollution buffers under high nitrogen loading.
Automatic building extraction from high-resolution remote sensing imagery is vital for urban applications. Although the Segment Anything Model (SAM) offers powerful semantic capabilities, its direct application faces severe challenges. The inherent low-pass filtering of Vision Transformers typically causes blurred building edges, distorted right angles, and severe adhesion among dense clusters. To address these bottlenecks, we propose a Boundary-Aware Segment Anything Model via High-Frequency Enhanced and Residual Injection (HFER-SAM). First, a Multi-scale Semantic Adaptation Module (MSAM) uses domain adapters and coordinate attention to recover spatial priors for precise domain alignment. Second, a High-Frequency Residual Injection Module (HFRIM) employs a lightweight CNN to extract roof textures and edge gradients. These are injected as residual terms into the semantic stream to explicitly sharpen physical boundaries. Finally, a Boundary-Constrained Dual-Task Head (BC-DTH) utilizes morphological supervision to sever erroneous connections between adjacent buildings and resolve target adhesion. Experiments on the WHU and Inria datasets reveal that HFER-SAM achieves an Intersection over Union of 91.53% and 82.06%, respectively. Our method significantly outperforms existing state-of-the-art models in both accuracy and boundary fidelity, providing a robust avenue for high-precision building mapping.