SUMMARY: The wide-swath altimeter satellite Surface Water and Ocean Topography (SWOT) will provide high spatiotemporal resolution sea surface heights (SSHs), which is crucial for studying the impact of observation errors on marine gravity recovery. This study uses simulated SWOT data to derive deflection of the vertical (DOV) and gravity anomalies in the northern South China Sea. We quantified the impact of SWOT errors on DOV and gravity anomalies, and analysed the contributions from different directions of geoid gradient. The results show that the geoid gradient in the cross-track direction significantly improves gravity field recovery by enhancing the precision of east component of DOV. For one-cycle SWOT observations, phase errors emerge as the most impactful error affecting both DOV and gravity anomalies, followed by random errors. 2-D Gaussian filtering and the tilt correction proposed in this study could effectively mitigate their impact. Using the corrected data for DOV computation, the precision in the east and north components improves by 75.32 and 46.80 per cent, respectively, while enhancing the accuracy of the gravity field by 70.23 per cent. For 17-cycle data, phase errors and random errors remain the predominant factors affecting DOV and gravity anomalies, but their impact diminishes with an increase in SWOT observations. Our results indicate that marine gravity accuracy improves by approximately 70 per cent compared to a single cycle. Whether for single-cycle or multicycle data, the impact of phase errors is roughly twice that of random errors. These data processing strategies can serve as valuable references for wide-swath altimeter data processing, aiming to advance the precision and resolution of marine gravity field recovery.

Accurate seabed topography is crucial for marine research, resource exploration, and engineering applications. While deep learning techniques have been widely applied in seabed inversion, existing methods often overlook the multi-scale influence of gravity anomalies, particularly the critical role of short-wavelength gravity anomalies in resolving fine-scale bathymetric features. In this study, we propose a novel Fully Connected Deep Neural Network (FCDNN) approach that systematically integrates long-wavelength, short-wavelength, and residual gravity anomaly components for seabed topography inversion. Using multi-satellite altimetry-derived gravity anomaly data (SIO V32.1) and shipborne bathymetric data (NCEI), we constructed a high-resolution (1′ × 1′) seabed topography model for the South China Sea (108°E–121°E, 6°N–23°N), termed FCD_Depth_SCS. The workflow included multi-scale decomposition of gravity anomalies, linear regression-based residual calculation, and FCDNN-based nonlinear training to capture the complex relationships between gravity anomalies and water depth. The FCD_Depth_SCS model achieved a difference standard deviation (STD) of 44.755 m and a mean absolute percentage error (MAPE) of 2.903% when validated against 160,476 shipborne control points. This performance significantly outperformed existing models, including GEBCO_2024, SIOv25.1, DTU18, and GGM_Depth (derived from the Gravity–Geologic Method), whose STDs were 82.234 m, 108.241 m, 186.967 m, and 58.874 m, respectively. Notably, the inclusion of short-wavelength gravity anomalies enabled the model to capture fine-scale bathymetric variations, particularly in open-sea regions. However, challenges remain near coastlines and complex terrains, highlighting the need for further model partitioning to address localized nonlinearity. This study highlights the benefits of integrating multi-scale gravity anomaly data with a fully connected deep neural network. Employing this innovative and robust approach enables high-resolution inversion of seabed topography with enhanced precision. The proposed method provides significant advancements in accuracy and resolution, contributing valuable insights for marine environmental research, resource management, and oceanographic studies.

Eddy identification and tracking are essential for understanding ocean dynamics. This study employed the elliptical Gaussian function (EGF) simulations and the py-eddy-tracker (PET) algorithm, validated by Surface Velocity Program (SVP) drifter data, to track eddies in the western North Pacific Ocean. The PET method effectively identified large- and mesoscale eddies but struggled with submesoscale features, indicating areas for improvement. Simulated satellite altimetry by EGF, mirroring Surface Water and Ocean Topography (SWOT)’s high-resolution observations, confirmed PET’s capability in processing fine-scale data, though accuracy declined for submesoscale eddies. Over 22 years, 1,188,649 eddies were identified, mainly concentrated east of Taiwan. Temporal analysis showed interannual variability, more cyclonic than anticyclonic eddies, and a seasonal peak in spring, likely influenced by marine conditions. Short-lived eddies were uniformly distributed, while long-lived ones followed major currents, validating PET’s robustness with SVP drifters. The launch of the SWOT satellite in 2022 has enhanced fine-scale ocean studies, enabling the detection of submesoscale eddies previously unresolved by conventional altimetry. SWOT observations reveal intricate eddy structures, including small cyclonic features in the northwestern Pacific, demonstrating its potential for improving eddy tracking. Future work should refine the PET algorithm for SWOT’s swath altimetry, addressing data gaps and unclosed contours. Integrating SWOT with in situ drifters, numerical models, and machine learning will further enhance eddy classification, benefiting ocean circulation studies and climate modeling.

SUMMARY: The Surface Water and Ocean Topography mission (SWOT), equipped with the Ka-band Radar Interferometer (KaRIn), provides groundbreaking 2-D sea surface heights (SSHs), bringing new potential for optimizing the deflection of the vertical (DOVs). However, conventional DOV modeling—combining along- and cross-track geoid gradients with equal weights—fail to fully exploit the potential of SWOT/KaRIn observations and overlook the spatial variability in precision. We present a tailored method for optimizing DOVs estimation. The method combines geoid gradients in the along-track, cross-track, diagonal (forward and backward) directions with adaptive weighting. The refined weights are employed to exploit the potential of each geoid gradient based on the relationship between the standard deviation of SSHs and the significant wave height. To mitigate data gaps, prior and locally averaged geoid gradients are incorporated in the gaps and overlapping regions. SWOT/KaRIn-derived DOVs and gravity anomalies from the science-phase observations are validated against shipborne gravity in the Philippine Sea. Results indicate that the DOV model derived by the tailored method—particularly by combining triple-directional (along, cross and diagonally forward) geoid gradients with refined weights—achieves a 7.3 percent improvement in accuracy over the conventional method. The supplement of additional geoid gradients is critical for mitigating leakage errors caused by missing or reduced observations in the gap regions. Furthermore, the gravity anomaly model recovered from DOVs by stacking 17-cycle observations achieved an accuracy of 2.97 mGal, representing a 7.2 percent improvement over single-cycle observations. The clear advantages of SWOT/KaRIn observations are gradually emerging in marine gravity recovery.

Next Point-of-Interest recommendation, which aims to forecast a user's future destination based on historical trajectories, plays a pivotal role in enhancing user experiences and facilitating precision marketing. The key to this task lies in effectively capturing the underlying patterns of user behavior. These behavioral patterns are inherently shaped by the complex interplay between semantic intent and long- and short-term preferences. First, the substantial heterogeneity and sparsity existing between semantic intents and entity selections render their accurate coupling significantly challenging. Second, existing approaches for modeling long- and short-term preferences rely excessively on trajectory session segmentation, which hinders the model's ability to capture latent behavioral rhythms and periodic evolving patterns. Third, Semantic features are often treated as static attributes, overlooking their role as endogenous drivers in the dynamic evolution of spatio-temporal topology. To address these limitations, we propose a novel framework named FreqSDG-poi. First, to achieve a tight coupling between semantic intents and entity selections, we design a Time-Anchored Dual-Layer Coupling Representation framework. Second, utilizing Fast Fourier Transform to isolate and capture user periodic behavioral rhythms from a frequency-domain perspective, we construct a Frequency-Domain Enhanced long- and short-term Preference Decoupling model. Finally, to explicitly model the dynamic driving mechanism of semantic features on spatio-temporal evolution, we propose a Semantic-Driven Dynamic Graph Convolution method. Extensive experimental results on two real-world datasets demonstrate that our proposed method outperforms state-of-the-art methods.