In shallow water, passive sonar usually has great difficulty in discriminating a surface acoustic source from an underwater one. To solve this problem, a supervised machine learning method using only one hydrophone is implemented in this paper. Firstly, simulated training data are generated by a normal mode model KRAKEN with the same environment setup as that in SACLANT 1993 experiment. Secondly, the k-nearest neighbor (kNN) classifiers are trained and evaluated using the scores of precision, recall, F1 and accuracy. Thirdly, the random subspace kNN classifiers are finely trained on three hyperparameters (the number of nearest neighbors, the number of predictors selected at random and the number of learners in the ensemble) to obtain the best model. Fourthly, a deep learning method called ResNet-18 is also applied, and it reaches the best balance between precision and recall, while the accuracies of both simulation and experimental data are all 1.0. Further, data from all 48 hydrophones of the vertical linear array (VLA) are analyzed using the three kinds of machine learning methods (kNN, random subspace kNN and ResNet-18) separately, and the results are compared. It is concluded that the performance of random subspace kNN is the best. Both the simulation and experimental results suggest the feasibility of machine learning as a surface and underwater acoustic source discrimination method even with only a single hydrophone.

The strong wind-induced current and sea level have influences on the wave distribution in a tropical cyclone (TC). In particular, the wave–current interaction is significant in the period in which the TC passed the mesoscale eddy. In this study, the wave fields of Typhoon Chan-hom (2015) are hindcastly simulated using a coupled oceanic model that utilizes a nested triangle grid, i.e., the finite-volume community ocean model-simulating waves nearshore (FVCOM-SWAVE) model. The forcing wind field is composited from the European Centre for Medium-Range Weather Forecasts (ECMWF) reanalysis data and the simulation using a parametric Holland model, denoted as H-E. The open boundary fields include tide data from TPOX.5 and the hybrid coordinate ocean model (HYCOM) global datasets, including sea surface temperature (SST), sea surface salinity, sea surface current, and sea level data. The simulated oceanic parameters (e.g., the significant wave height, SWH) are validated against the measurements from the Jason-2 altimeter, yielding a root mean square error (RMSE) of 0.58 m for the SWH, a correlation (COR) coefficient of 0.94, and a scatter index (SI) of 0.23. Similarly, the simulated SSTs are compared with the remote sensing products of the remote sensing system (REMSS) and the measurements from Argos, yielding an RMSE of 0.95, and an SI of 80% near the typhoon center, and the ratio is reduced to approximately <20% away from the typhoon center, indicating that Stokes transport is an essential aspect in the water mixing during a TC. The mesoscale eddies are detected by the sea level anomalies (SLA) fusion data from AVISO. It is found that the significant wave heights, Stokes drift, and Stokes transport inside the eddy area were higher than those outside the eddy area. These parameters inside the cold mesoscale eddies were higher than t inside the warm mesoscale eddies. Otherwise, the SST mainly increased within the cold mesoscale eddies area, while decreased within the warm mesoscale eddies area. The influence of mesoscale eddies on the SST was in proportion to the eddy radius and eddy EKE.

The sea surface drag coefficient plays an important role in momentum transmission between the atmosphere and the ocean, which is affected by ocean waves. The total air–sea momentum flux consists of effective momentum flux and sea spray momentum flux. Sea spray momentum flux involves sea surface drag, which is largely affected by the ocean wave state. Under strong winds, the sea surface drag coefficient ($C_D$) does not increase linearly with the increasing wind speed, namely, the increase of $C_D$ is inhibited by strong winds. In this study, a sea surface drag coefficient is constructed that can be applied to the calculation of the air–sea momentum flux under high wind speed. The sea surface drag coefficient also considers the influence of wave state and sea spray droplets generated by wave breaking. Specially, the wave-dependent sea spray generation function is employed to calculate sea spray momentum flux. This facilitates the analysis not only on the sensitivity of the sea spray momentum flux to wave age, but also on the effect of wave state on the effective $C_D$ ($C_{D, eff}$) under strong winds. Our results indicate that wave age plays an important role in determining $C_D$. When the wave age is >0.4, $C_D$ decreases with the wave age. However, when the wave age is ≤0.4, $C_D$ increases with the wave age at low and moderate wind speeds but tends to decrease with the wave age at high wind speeds.

Sound source recognition is a very important application of passive sonar. How to distinguish between surface and underwater acoustic sources has always been a challenge. Due to the mixing of underwater target radiated noise and marine environmental noise, especially in shallow water environments where multipath effects exist, it is difficult to distinguish them. To solve the surface and underwater acoustic source recognition problem, this paper proposes a multi-channel joint detection method based on machine learning. First, the simulation data are generated using the normal model KRAKEN setting in the same environment as the SACLANT 1993 experiment, which uses a vertical linear array of 48 hydrophones. Secondly, the GBDT classifier and LightGBM classifier are trained separately, and then the model is evaluated using precision, recall, F1, and accuracy. Finally, four ML models (kNN, random subspace kNN, GBDT, and LightGBM) are used to analyze all 48 channels of hydrophone data. For each model, two kinds of feature extraction methods (module features, real and imaginary features) are applied. Generally, the results show that both GBDT and LightGBM models have better performance than both kNN and random subspace kNN ones. For both GBDT and LightGBM models, the results using module features have better performance than using real and imaginary features.

Despite numerous studies, the impact of global warming on the tropical cyclone (TC) track and intensity by reasons of data inhomogeneity in remote sensing and large natural variability over a relatively short period of observation is still controversial. Three carbon-emission sensitivity experiments are conducted to investigate how TC track and intensity respond to changes in the oceanic and atmospheric environment under global warming. The results show a high sensitivity of the simulated TC track and intensity to global warming. On one hand, with increase in carbon emissions, the western Pacific subtropical high expands notably, increasing the poleward steering flow and eventually leading to a poleward shift of TC. On the other hand, the underlying sea-surface temperature and surface-entropy flux increase and, thus, favor the convections near the eyewall. Moreover, the TC structure becomes more upright, which is closely related to the larger pressure gradient near the eyewall. As a result, TC intensity increases with carbon emissions. However, this increase is notably smaller than the maximum potential intensity theory as the TC intensity can reach a threshold if carbon emission still increases in the future. The involved mechanisms on the changes of TC track and intensity are also revealed.

Wave-induced mixing can enhance vertical mixing in the upper ocean, facilitating the exchange of heat and momentum between the surface and deeper layers, thereby influencing ocean circulation and climate patterns. Building on previous research, this study proposes a wave-induced mixing parameterization scheme (referred to as EXP3) specifically designed for typhoon periods. This scheme was integrated into the fully coupled ocean–wave–atmosphere model COAWST and applied to analyze Typhoon In-Fa (2021) as a case study. The simulation results were validated against publicly available data, demonstrating a good overall match with observed phenomena. Subsequently, a comparative analysis was conducted between the EXP3 scheme, the previous scheme (EXP2) and the original model scheme (EXP1). Validation against Argo and Drifter buoy data revealed that both EXP2 and EXP3, which include wave-induced mixing effects, resulted in a decrease in the simulated mixed layer depth (MLD) and mixed layer temperature (MLT), with EXP3 showing closer alignment with the observed data. Compared to the other two experiments, EXP3 enhanced vertical motion in the ocean due to intensified wave-induced mixing, leading to increased upper-layer water divergence and upwelling, a decrease in sea surface temperature and accelerated rightward deflection of surface currents. This phenomenon not only altered the temperature structure of the ocean surface layer but also significantly impacted the regional ocean dynamics.