Accurate prediction of towing resistance is essential for selecting tug capacity and controlling energy demand during the wet-towing of floating offshore wind turbines (FOWTs). This study established a high-precision CFD numerical wave tank to simulate the towing hydrodynamic performance of the IEA-15 MW semi-submersible FOWT under calm water and regular wave conditions, covering two typical towing layouts (single-column-forward and double-column-forward) and three representative towing speeds. Based on CFD free-decay results, viscous damping corrections in heave, roll, and pitch were applied to the AQWA, while the drag coefficients were introduced through Morison elements. On this basis, an integrated viscous-potential flow coupled numerical model for the tugboat-FOWT system was then established. Frequency-domain analysis showed that wave excitation loads increase with towing speed under following-wave conditions, while the reverse trend occurs under head-wave conditions. Time-domain validation demonstrated that the viscosity-corrected potential flow model achieves a deviation below 2% for calm water resistance and less than 5% for mean wave resistance compared with high-fidelity CFD benchmarks. Parametric analysis under multiple regular wave conditions revealed that wave-added resistance first increases and then falls as wave period increases, consistent with the variation in surge wave excitation forces, and exhibits a nonlinear growth pattern with wave height. The double-column forward layout yields overall higher towing resistance than its single-column alternative, while main towline tension can accurately reflect the total towing resistance level. These technical outcomes provide actionable engineering guidance for low-carbon, cost-effective wet-towing scheme formulation, facilitating the sustainable, large-scale exploitation of deep-sea wind resources and advancing the long-term decarbonization of the marine energy sector.

Frequent flooding induced by climate change has been posing growing threats to human lives and property worldwide. Most of 2D and 3D physics-based hydrodynamic (PBH) models are generally too time-consuming to support decision-makers to react in time during an urgent flood event. Therefore, highly efficient and accurate flood inundation prediction models are urgently needed to improve emergency response capabilities. In this paper, a novel model based on a 1D deep residual neural network with Squeeze-and-Excitation (ResNet-SE) attention is proposed for fast forecasting dynamic evolution of 2D fluvial flood inundation. Different from widely-used 2D image-oriented DL models, our model directly consumes original 1D nodal hydraulic output vectors from TELEMAC-2D, avoiding interpolation errors introduced by mesh rasterization. The results of the well-trained ResNet-SE model compare with those from the TELEMAC-2D and a convolutional neural network (CNN) model. As indicated by several quantitative assessment matrices, the present model outperforms CNN model and conforms closely to the TELEMAC-2D in high precision. Furthermore, we demonstrate that ResNet34-SE (34 layers) stands out distinctly among ResNet-SEs of 18 and 50 layers in flood inundation prediction. The average $MAE$ and $RMSE$ metrics of ResNet34-SE under three unseen flood events reach 0.04623 m and 0.08052 m in water depth prediction. As for spatial flood extent assessment, this model also delivers impressive performance, with the average $Precision$, $Recall$, $F1$, and $Accuracy$ scores reaching 96.92%, 99.69%, 98.28%, and 98.55%, respectively. Furthermore, current deep-learning (DL) model achieve a speedup of 300 to 400 times over traditional PBH models, enabling the generation of 24-h flood evolution maps within seconds. The proposed ResNet-SE model demonstrates good simulating accuracy of fluvial flood dynamics for proactive emergency warning systems and flood risk mitigation planning within the Lushui River basin.