Application of deep learning technique to the sea surface height prediction in the South China Sea

A deep-learning-based method,called ConvLSTMP3,is developed to predict the sea surface heights(SSHs).ConvLSTMP3 is data-driven by treating the SSH prediction problem as the one of extracting the spatial-temporal features of SSHs,in which the spatial features are“learned”by convolutional operations while the temporal features are tracked by long short term memory(LSTM).Trained by a reanalysis dataset of the South China Sea(SCS),ConvLSTMP3 is applied to the SSH prediction in a region of the SCS east off Vietnam coast featured with eddied and offshore currents in summer.Experimental results show that ConvLSTMP3 achieves a good prediction skill with a mean RMSE of 0.057 m and accuracy of 93.4%averaged over a 15-d prediction period.In particular,ConvLSTMP3 shows a better performance in predicting the temporal evolution of mesoscale eddies in the region than a full-dynamics ocean model.Given the much less computation in the prediction required by ConvLSTMP3,our study suggests that the deep learning technique is very useful and effective in the SSH prediction,and could be an alternative way in the operational prediction for ocean environments in the future.

Upper-Tropospheric Environment–Tropical Cyclone Interactions over the Western North Pacific:A Statistical Study

Based on 25-year（1987–2011） tropical cyclone（TC） best track data, a statistical study was carried out to investigate the basic features of upper-tropospheric TC–environment interactions over the western North Pacific. Interaction was defined as the absolute value of eddy momentum flux convergence（EFC） exceeding 10 m s^（-1）d^（-1）. Based on this definition, it was found that 18% of all six-hourly TC samples experienced interaction. Extreme interaction cases showed that EFC can reach^120 m s^（-1）d^（-1） during the extratropical-cyclone（EC） stage, an order of magnitude larger than reported in previous studies.Composite analysis showed that positive interactions are characterized by a double-jet flow pattern, rather than the traditional trough pattern, because it is the jets that bring in large EFC from the upper-level environment to the TC center. The role of the outflow jet is also enhanced by relatively low inertial stability, as compared to the inflow jet. Among several environmental factors, it was found that extremely large EFC is usually accompanied by high inertial stability, low SST and strong vertical wind shear（VWS）. Thus, the positive effect of EFC is cancelled by their negative effects. Only those samples during the EC stage, whose intensities were less dependent on VWS and the underlying SST, could survive in extremely large EFC environments, or even re-intensify. For classical TCs（not in the EC stage）, it was found that environments with a moderate EFC value generally below ~25 m s^（-1）d^（-1） are more favorable for a TC＇s intensification than those with extremely large EFC.

On the heat budget and water mass exchange in the Andaman Sea

The characteristics of the T/S structures,water mass exchange and deep circulation in the Andaman Sea are investigated based on the simulation from a high-resolution general circulation model(MITgcm).The results show that,below 1000 m,the water mass is saltier,warmer and more homogeneous in the Andaman Sea than that in the Bay of Bengal,attributing to the strong vertical mixing at the depth of^1800 m.The water mass exchange between the Andaman Sea and the Bay of Bengal goes through three major channels,which manifests itself as follows:the northern channel(Preparis Channel)is the main passage of water mass transport from the Bay of Bengal to the Andaman Sea,whereas the Middle Channel(the south of Andaman Islands and the north of Nicobar Islands)has an opposite transport;the southern channel(Great Channel)features with a four-layer water exchange which results in the least net transport among the three channels;all the transports through the three channels have an intra-annual variation with a period of half a year.At 1000-m depth,the entire Andaman Sea is occupied by a cyclonic circulation in January and July while by an anticyclonic one in April and October.The semiannual cycle found in both the deep circulation and water mass exchange is likely associated with the downwelling eastward-propagating Kelvin waves induced by the semiannual westerly component in the equatorial Indian Ocean during intermonsoon seasons.

Advances in research of the mid-deep South China Sea circulation

The South China Sea(SCS)is a large marginal sea connecting the Indian and Pacific oceans.Under the factors of monsoons,strait transport,and varied bathymetry,the SCS presents a three-layer structure and strong diapycnal mixing which is far greater than that in the open ocean.Theoretical analysis and observations reveal that internal tides,internal solitary waves,and strong winds are the sources of the strong mixing in the northern SCS.A major consequence of the strong mixing is an active mid-deep circulation system.This system promotes exchange of water between the SCS and adjacent oceans,and also regulates the upper layer of wind-driven circulation,making the 3 dimensional SCS circulation clearly different from that in other tropical and subtropical marginal seas.The mass transport capacity of the mid-deep circulation has a substantial impact on marine sedimentation,the biogeochemical cycle,and other processes in the SCS.This paper summarizes the recent advances in middeep sea circulation dynamics of the SCS,and discusses the opportunities and challenges in this area.