With the rapid proliferation of the Internet, network security issues that threaten users have become increasingly severe, despite the widespread benefits of Internet access. Most existing intrusion detection systems (IDS) suffer from suboptimal performance due to data imbalance and feature redundancy, while also facing high computational complexity in areas such as feature selection and optimization. To address these challenges, this study proposes a novel network intrusion detection method based on an improved binary simulated annealing algorithm (IBSA) and TPE-FL-LightGBM. First, by integrating Focal Loss into the loss function of the LightGBM classifier, we introduce cost-sensitive learning, which effectively mitigates the impact of class imbalance on model performance and enhances the model’s ability to learn difficult-to-classify samples. Next, significant improvements are made to the simulated annealing algorithm, including adaptive adjustments of the initial temperature and Metropolis criterion, the incorporation of multi-neighborhood search strategies, and the integration of an S-shaped transfer function. These improvements enable the IBSA method to achieve efficient optimal feature selection with fewer iterations. Finally, the Tree-structured Parzen Estimator (TPE) algorithm is employed to optimize the structure of the FL-LightGBM classifier, further enhancing its performance. Through comprehensive visual analysis, ablation studies, and comparative experiments on the NSL-KDD and UNSW-NB15 datasets, the reliability of the proposed network intrusion detection method is validated.

The Water and Sediment Regulation Scheme (WSRS), implemented since 2002, has been essential for controlling water flow and mitigating sediment siltation in the lower Yellow River. However, WSRS was suspended for the first time in 2016 and 2017 due to extremely low water flow. The rapid floodwater discharge over roughly 20 days conducted by WSRS strongly impacts total suspended solids (TSS) distribution in the Yellow River Estuary (YRE). This study employs high-frequency Sentinel-3 OLCI satellite imagery to investigate intraday TSS variations in the YRE under new water-sediment regulation conditions from 2016 to 2023. TSS concentrations were generally low during the 2016 and 2017 flood seasons, but increased markedly after WSRS resumed in 2018. Peak TSS values occurred in July or August, sometimes extending into September and October during autumn floods. A moderately strong positive correlation was observed between TSS concentrations at the river mouth and sediment load at Lijin Station during the flood seasons. The 2018 WSRS event generated an extensive river plume, with average TSS concentrations at the river mouth exceeding 400 g·m⁻³. From 2018 to 2023, TSS concentrations exhibited a declining trend during flood seasons, attributed to reduced sediment discharge and ongoing sediment accretion in the Yellow River Delta. Our findings highlight Sentinel-3 OLCI as a powerful tool to resolve WSRS-driven sediment dynamics, offering critical guidance for estuarine management.

The estimated spatiotemporal characteristics of particulate matter in the ocean vary with the measurement method used. This variation introduces considerable uncertainty in our understanding of how particle scattering cross-section, particle size, and carbon content relate to one another at local, regional, and global scales. A more accurate and detailed characterization of the spatiotemporal variations of particles in the water column and of the contribution of different types of particles to the optical parameters of water are crucial for improving our understanding of the marine biogeochemical cycle. In this study, we investigated how composition, size, and particulate organic carbon (POC) content of particulate matter, along with their corresponding optical proxies, change in the upper 200 m of an oligotrophic region in the tropical Western Pacific Ocean. We estimated the contributions of various water components to the particle backscattering coefficient and to POC. Using newly collected, vertically resolved data, we derived depth-resolved net primary productivity (NPP) with the absorption-based production model (AbPM) and the carbon-based production model (CbPM); both models account for vertical variations in water column properties. Our results indicated that particles larger than 8 µm (especially minerals and aggregates) accounted for an increasing amount of POC at depths greater than 100 m, with a maximum at 500 m. In contrast, chlorophyll content decreased steadily with depth. Our comparison of the backscatter and absorption coefficients (optical proxies of POC) had the same trend, although the specific components that contributed to POC were different. Changes in parameters such as particle composition, size, POC content, and their optical proxies all corresponded to changes in the deep chlorophyll maximum (DCM) along the latitudinal gradient. When we compared the NPP estimates from the two approaches, the CbPM yielded higher values than the AbPM in surface waters, likely because of the way particles are distributed vertically. In areas where the DCM was deeper, the AbPM provided a better accounting of how individual components contributed to the NPP. Together, these findings clarify how particle composition and its vertical variability influence POC and inherent optical properties (IOPs) in this oligotrophic region. They also offer a basis for interpreting water column characteristics and assessing how changes in NPP may affect biogeochemical processes.