The prediction of oceanic features is always an important issue in oceanography, where deep learning has been proven to be a useful tool. In this study, we applied the improved U-net model to predict the monthly sea surface salinity (SSS) over the western Pacific (WP) Ocean, and the model was designed to use the SSSs from six consecutive months to predict the SSS in the next month. The monthly satellite-based SSSs in 2015–2020 were used for model training, and the data collected after January 2021 were used to evaluate the model’s predictive abilities. The results showed that the predicted SSSs represented the general patterns of SSSs over the WP region. However, the small-scale features were smoothed out in the model, and the temporal variations were also not well captured, especially over the East China Sea and Yellow Sea (ECS&YS) region. To further evaluate the potential of the U-net model, a more specific model was conducted for the ECS&YS region (Domain 2), which successfully predicted both spatial and temporal variations in the SSSs, including the spreading and retreating of the low-salinity tongue. Based on the comparison between the two domains and sensitivity experiments, we found that the prediction biases were contributed by the spatial distributions of the SSSs, the domain size, and the filter numbers. In addition, further multi-step prediction experiments suggested that our U-net model could also be used for long-time prediction, and we have examined up to five months. Overall, this study demonstrated the great ability and potential of the U-net model for predicting SSS, even though only a few trainable data are available.

Prolonged extremely warm ocean temperatures have great impacts on both natural ecosystems and human communities. These phenomena (i.e., marine heatwaves) could be easily monitored globally by satellite-based sea surface temperatures; however, the choice of datasets may lead to potential uncertainties in the marine heatwave assessment. Here we compared the marine heatwaves using three commonly used satellite products to illustrate the uncertainties over Asia and the Indo-Pacific. Distinct differences were found in the occurrence, duration, and long-term trend of marine heatwaves over both coastal and open oceans, while some discrepancies could become comparable with the obtained metrics themselves. Although differences in mean sea surface temperatures or their variances among datasets could not explain the abovementioned discrepancies, different sensors, procedures, and sea ice treatments in each dataset may contribute partially. Overall, our study suggests that the use of multiple datasets is crucial for evaluations of extreme events.

The frontogenesis and frontolysis processes of the subpolar front (SPF) in the surface mixed layer of the Japan Sea are investigated using state‐of‐the‐art oceanic reanalysis data. The SPF experiences a 9 month weakening period from January to September, which shifts to a strengthening period in October. Our analysis shows that horizontal advection consistently contributes to the intensification of the SPF. After September, as the weakening effect of surface heat flux diminishes, horizontal advection becomes the dominant factor that contributes to changes in the SPF strength. Thus, the SPF enters a 3 month strengthening period. The geostrophic component of horizontal advection provides the most important contribution to strengthening the SPF, acting to intensify the SPF year‐round. Ekman advection also promotes SPF strengthening with a smaller but still important contribution. During the weakening period, SPF strength is largely controlled by heat flux. The heat flux, especially the shortwave radiation component, is the primary cause of the surface front disappearance in the summer.

Satellite observations provide valuable information on the rapidly increasing marine heatwaves over the Arctic Ocean, yet current assessments of these events lack reliability because the satellite-based temperature in ice-covered regions heavily relies on ice conditions. Here we compared the marine heatwaves derived from two commonly used satellite products to evaluate the potential influences of sea ice in the pan-Arctic. We found inconsistent behaviors of marine heatwaves in both mean fields and temporal variations on interannual and seasonal timescales using two datasets, which became even larger in recent years when they showed opposite tendencies over regions with higher ice concentrations (>50%). Our findings revealed our knowledge of marine heatwaves in the Arctic is facing challenges due to the large influences of sea ice and datasets that were ignored in previous studies, highlighting an urgent need for in situ observations and consistent and reliable long-time datasets for the warming Arctic.