Dissolved oxygen (DO) is a crucial indicator for characterizing water quality and ecosystem status in freshwater lakes, and its concentration is closely correlated with the surrounding aquatic environment, particularly meteorological conditions. However, traditional DO prediction methods struggle to effectively capture the intricate coupling relationships between multi-station meteorological factors and DO concentration time series, limiting the prediction accuracy. This study proposes a multi-scale spatio-temporal graph neural network with integrated multi-meteorological factors. Taking Taihu Lake and its surrounding cities as the study area, a meteorological graph is constructed based on the geographic proximity between meteorological stations, and a dual-stage “local–global” modeling strategy is adopted to capture the spatio-temporal dependencies of DO concentration under meteorological forcing. Using R², RMSE, MAE and MAPE as evaluation metrics, we conducted single-step and multi-step DO prediction experiments on the 2023–2024 Taihu Tuoshan water quality dataset and compared the proposed model with commonly used prediction models. In the single-step prediction task, the proposed model improved R² by 2.12–20.84% and reduced RMSE, MAE, and MAPE by 3.05–40.80%, 14.97–53.26%, and 6.91–55.62%, respectively. In the 6-step-ahead and 12-step-ahead prediction tasks, RMSE and MAE were reduced by 3.79–15.75% and 6.68–23.09%, and by 5.03–10.39% and 7.13–16.46%, respectively. The experimental results provide quantitative evidence for the superiority of the proposed model in single-step and multi-step DO prediction. This study offers a novel data-driven tool for lake water quality early warning and drinking water safety, and the proposed framework can serve as a reference for water quality prediction studies driven by multi-source environmental factors.

Against the backdrop of intensifying global climate change, reducing carbon emissions has become a shared global objective. Blue carbon, as a significant carbon sink type, still lacks a mature assessment framework. Monitoring carbon fluxes in marine salt marsh wetlands is a core technology for accurately evaluating blue carbon potential. In response, this study independently developed a spatiotemporally coupled carbon flux monitoring system for marine salt marsh wetlands. The system consists of real-time monitoring equipment, a cloud-based intelligent storage and visualization analysis platform, and a terminal assessment system. It enables the real-time monitoring of carbon fluxes across multiple spatial scales and integrates time-series patterns to assess carbon sequestration potential from multiple dimensions. To address the bottleneck of sensor accuracy, a multi-algorithm fusion technology was innovatively developed, significantly enhancing the accuracy of monitoring data. A modular integrated design was employed to construct a land–air integrated monitoring architecture, which is adaptable to the complex environments of salt marsh wetlands. This facilitates long-term automated monitoring while reducing the need for manual intervention. The terminal assessment system processes spatial-scale data using the DeNitrification-DeComposition model (DNDC 9.5) and integrates time-series carbon flux patterns, enabling precise quantification of marine carbon sink potential through spatiotemporal comprehensive analysis. The system first completed performance verification during the experimental phase, acquiring a total of 5760 sets of valid monitoring data, with a data qualification rate of 99.72%. The proposed multi-algorithm fusion method kept monitoring data fluctuations within 0.5%, and the relative error of the spatiotemporal integrated prediction was as low as 0.31%, thereby ensuring the stability and accuracy of long-term in situ monitoring. Based on this, a one-year field validation was conducted in a 100-hectare coastal salt marsh wetland in Dafeng, Yancheng. Using a spatiotemporal coupling assessment, the annual total carbon sequestration of this area was estimated at 1498.4 tons of carbon, with an assessment error of only 5.1%, achieving precise quantification of the blue carbon sink in the salt marsh wetland. This study provides reliable technical support for evaluating the carbon sequestration capacity of coastal salt marsh wetlands, contributing to the implementation of carbon emission reduction strategies. It also offers a scientific basis for global carbon cycle research and carbon sink management decision-making.