Marine heatwaves (MHWs) are extreme events characterized by abnormally high sea surface temperatures, and they have significant impacts on marine ecosystems and human society. The rapid and accurate forecasting of MHWs is crucial for preventing and responding to the impacts they can lead to. However, the research on relevant forecasting methods is limited, and a dedicated forecasting system specifically tailored for the South China Sea (SCS) region has yet to be reported. This study proposes a novel forecasting system utilizing U-Net and ConvLSTM models to predict MHWs in the SCS. Specifically, the U-Net model is used to forecast the intensity of MHWs, while the ConvLSTM model is employed to predict the probability of their occurrence. The indication of an MHW relies on both the intensity forecasted by the U-Net model exceeding threshold T and the occurrence probability predicted by the ConvLSTM model surpassing threshold P. Incorporating sensitivity analysis, optimal thresholds for T are determined as 0.9 °C, 0.8 °C, 1.0 °C, and 1.0 °C for 1-, 3-, 5-, and 7-day forecast lead times, respectively. Similarly, optimal thresholds for P are identified as 0.29, 0.30, 0.20, and 0.28. Employing these thresholds yields the highest forecast accuracy rates of 0.92, 0.89, 0.88, and 0.87 for the corresponding forecast lead times. This innovative approach gives better predictions of MHWs in the SCS, providing invaluable reference information for marine management authorities to make well-informed decisions and issue timely MHW warnings.

Using satellite remote sensing data and re-analysis products for the period of 1993–2015, the variation of a large anticyclonic eddy, the so-called Great Whirl (GW), located in the Northwest Indian Ocean off the coast of Somali, and its impacts on atmosphere were investigated. The GW is generated in early summer and decays in late fall every year. The center of the GW is located at 7.73°N, 53.20°E. The mean lifetime, sea level anomaly (SLA) difference, sea surface temperature anomaly (SSTA), radius, normalized vorticity, eddy kinetic energy (EKE), and deformation rate are 169 days, 0.07 m, 0.83 °C, 116.86 km, −0.53, 0.08 ${\mathrm{m}}^2·{\mathrm{s}}^{-2}$ , and 0.58, respectively. All these variables exhibit interannual variations. Composite analyses show that the maximum values of sea surface temperature (SST), wind, and water vapor anomalies occur in the northwest of the GW center. The fitting coefficient between the SST and wind speed anomaly is 1.1, indicating that, corresponding to 1 °C increases of the SST, the wind speed increases by about 1.1 $\mathrm{m}·{\mathrm{s}}^{-1}$ , and the fitting coefficient between the SST and water vapor anomaly is 0.45, indicating that water vapor increases by about 0.45 mm in response to 1 °C increases in the SST. In the vertical direction, the maximum and minimum values of vertical velocity anomalies and vertical transport of transient zonal momentum occur over the GW at about 900 hPa, and wind speed anomalies occur at about 950 hPa. Both the positive transport anomalies of transient zonal momentum and the positive vertical velocity anomalies on the west side of the GW can accelerate the wind speed in the lower level.

Marine cold spells (MCSs) are extreme ocean temperature events impacting marine organisms, yet their characteristics and trends in the South China Sea (SCS) historical period remain unclear. This study systematically analyzes sea surface temperature (SST) and MCSs in the SCS using satellite observation data (OISSTv2.1) from 1982 to 2022. The climatological mean SST ranges from 22 °C near the Taiwan Strait to 29 °C near the Nansha Islands, showing notable variations. Annual SST anomalies demonstrate a heterogeneous spatial trend of approximately 0.21 ± 0.16 °C/decade (p < 0.01) across the SCS, indicating an increase in SST over time. MCS analysis uncovers spatial non-uniformity in frequency, with higher values near the Beibu Gulf and Hainan Island, and longer durations in the northeastern coastal areas. Statistical analysis indicates normal distributions for frequency and duration trends but skewness for intensity and cumulative intensity, reflecting extreme values. Winter months exhibit larger MCS occurrence areas and higher mean intensities, illustrating seasonal variability. Anticipated changes will significantly impact the ecological structure and functioning of the SCS.

Up to now, the literature has shown that the relative wind stress does negative work on ocean mesoscale eddies. In other words, the relative wind stress inhibits the development of the eddies. However, based on a newly derived simplified theoretical model, the present study finds that under the action of a steady and uniform wind field, eddies can rapidly obtain kinetic energy from the wind field following several hours of adaption and adjustment, in which the wind stress transitions from doing negative to positive work. The finding is supported by the fact that the relative wind stress work on oceanic eddies over the northeastern tropical Pacific ocean is positive with the nearly constant gap wind. This implies that energy input from the wind is sensitive to eddy velocity structure, and hence, wind stress is not always a killer of eddies.

This study proposes a novel causality-structured matrix long short-term memory (C-mLSTM) model for significant wave height (SWH) forecasting. The framework incorporates a two-stage causal feature selection methodology using cointegration testing and Granger causality testing to identify long-term stable causal relationships among variables. These relationships are embedded within the C-mLSTM architecture, enabling the model to effectively capture both temporal dependencies and causal information within the data. Furthermore, the model integrates Bayesian optimization (BO) and twin delayed deep deterministic policy gradient (TD3) algorithms for synergistic optimization. This combined TD3-BO approach achieves an 11.11% improvement in the mean absolute percentage error (MAPE) on average compared to the base model without optimization. For 1–24 h SWH forecasts, the proposed TD3-BO-C-mLSTM outperforms the benchmark models TD3-BO-LSTM and TD3-BO-mLSTM in prediction accuracy. Finally, a Shapley additive explanations (SHAP) analysis was conducted on the input features of the BO-C-mLSTM model, which reveals interpretability patterns consistent with the two-stage causal feature selection methodology. This research demonstrates that integrating causal modeling with optimization strategies significantly enhances time-series forecasting performance.