The Forecast Skills and Predictability Sources of Marine Heatwaves in the NUIST-CFS1.0 Hindcasts

Using monthly observations and ensemble hindcasts of the Nanjing University of Information Science and Technology Climate Forecast System(NUIST-CFS1.0) for the period 1983–2020, this study investigates the forecast skill of marine heatwaves(MHWs) over the globe and the predictability sources of the MHWs over the tropical oceans. The MHW forecasts are demonstrated to be skillful on seasonal-annual time scales, particularly in tropical oceans. The forecast skill of the MHWs over the tropical Pacific Ocean(TPO) remains high at lead times of 1–24 months, indicating a forecast better than random chance for up to two years. The forecast skill is subject to the spring predictability barrier of El Nino-Southern Oscillation(ENSO). The forecast skills for the MHWs over the tropical Indian Ocean(TIO), tropical Atlantic Ocean(TAO), and tropical Northwest Pacific(NWP) are lower than that in the TPO. A reliable forecast at lead times of up to two years is shown over the TIO, while a shorter reliable forecast window(less than 17 months) occurs for the TAO and NWP.Additionally, the forecast skills for the TIO, TAO, and NWP are seasonally dependent. Higher skills for the TIO and TAO appear in boreal spring, while a greater skill for the NWP emerges in late summer-early autumn. Further analyses suggest that ENSO serves as a critical source of predictability for MHWs over the TIO and TAO in spring and MHWs over the NWP in summer.

Study on the interannual variability of the Kerama Gap transport and its relation to the Kuroshio/Ryukyu Current system

An analysis of a 68-year monthly hindcast output from an eddy-resolving ocean general circulation model reveals the relationship between the interannual variability of the Kerama Gap transport(KGT)and the Kuroshio/Ryukyu Current system.The study found a significant difference in the interannual variability of the upstream and downstream transports of the East China Sea-(ECS-)Kuroshio and the Ryukyu Current.The interannual variability of the KGT was found to be of paramount importance in causing the differences between the upstream and downstream ECS-Kuroshio.Additionally,it contributed approximately 37%to the variability of the Ryukyu Current.The interannual variability of the KGT was well described by a two-layer rotating hydraulic theory.It was dominated by its subsurface-intensified flow core,and the upper layer transport made a weaker negative contribution to the total KGT.The subsurface flow core was found to be mainly driven by the subsurface pressure head across the Kerama Gap,and the pressure head was further dominated by the subsurface density anomalies on the Pacific side.These density anomalies could be traced back to the eastern open ocean,and their propagation speed was estimated to be about 7.4 km/d,which is consistent with the speed of the local first-order baroclinic Rossby wave.When the negative(positive)density anomaly signal reached the southern region of the Kerama Gap,it triggered the increase(decrease)of the KGT towards the Pacific side and the formation of an anticyclonic(cyclonic)vortex by baroclinic adjustment.Meanwhile,there is an increase(decrease)in the upstream transport of the entire Kuroshio/Ryukyu Current system and an offshore flow that decreases(increases)the downstream Ryukyu Current.

Frontogenesis and Frontolysis of a Cold Filament Driven by the Cross-Filament Wind and Wave Fields Simulated by a Large Eddy Simulation

The variations of the frontogenetic trend of a cold filament induced by the cross-filament wind and wave fields are studied by a non-hydrostatic large eddy simulation. Five cases with different strengths of wind and wave fields are studied.The results show that the intense wind and wave fields further break the symmetries of submesoscale flow fields and suppress the levels of filament frontogenesis. The changes of secondary circulation directions—that is, the conversion between the convergence and divergence of the surface cross-filament currents with the downwelling and upwelling jets in the filament center—are associated with the inertial oscillation. The filament frontogenesis and frontolysis caused by the changes of secondary circulation directions may periodically sharpen and smooth the gradient of submesoscale flow fields.The lifecycle of the cold filament may include multiple stages of filament frontogenesis and frontolysis.

Rotating tank experiments for the study of geophysical fluid dynamics

Geophysical fluid dynamics(GFD)is an interdisciplinary field that studies the large-scale motion of fluids in the natural world.With a wide range of applications such as weather forecasts and climate prediction,GFD employs various research approaches including in-situ observations,satellite measurements,numerical simulations,theoretical analysis,artificial intelligence,and physical model experiments in laboratory.Among these approaches,rotating tank experiments provide a valuable tool for simulating naturally-occurring fluid motions in laboratories.With proportional scaling and proper techniques,scientists can reproduce multi-scale physical processes of stratified fluids in the rotation system,which allows for the simulation of essential characteristics of fluid motions in the atmosphere and oceans.In this review,rotating tanks of various scales in the world are introduced,as these tanks have been actively used to explore fundamental scientific questions in ocean and atmosphere dynamics.To illustrate the GFD experiments,three representative cases are presented to demonstrate the frontier achievements in the the GFD study by using rotating tank experiments:mesoscale eddies in the ocean,convection processes,and plume dynamics.Detailed references for the experimental procedures are provided.Future studies are encouraged to further explore the utilization of rotating tanks with improvements in experimental design and integration of other research methods.This is a promising direction of GFD to help enhance our understanding of the complex nature of fluid motions in the natural world and to address the challenges posed by global environmental changes.

Influence of the upper mixed layer depth on Langmuir turbulence characteristics

The upper mixed layer depth(h)has a significant seasonal variation in the real ocean and the low-order statistics of Langmuir turbulence are dramatically influenced by the upper mixed layer depth.To explore the influence of the upper mixed layer depth on Langmuir turbulence under the condition of the wind and wave equilibrium,the changes of Langmuir turbulence characteristics with the idealized variation of the upper mixed layer depth from very shallow(h=5 m)to deep enough(h=40 m)are studied using a non-hydrostatic large eddy simulation model.The simulation results show that there is a direct entrainment depth induced by Langmuir turbulence(h_(LT))within the thermocline.The normalized depthaveraged vertical velocity variance is smaller and larger than the downwind velocity variance for the ratio of the upper mixed layer to a direct entrainment depth induced by Langmuir turbulence h/h_(LT)<1 and h/h_(LT)>1,respectively,indicating that turbulence characteristics have the essential change(i.e.,depth-averaged vertical velocity variance(DAVV)DADV for Langmuir turbulence)between h/h_(LT)<1 and h/h_(LT)>1.The rate of change of the normalized depth-averaged low-order statistics for h/h_(LT)<1 is much larger than that for h/h_(LT)>1.The reason is that the downward pressure perturbation induced by Langmuir cells is strongly inhibited by the upward reactive force of the strong stratified thermocline for h/h_(LT)<1 and the eff ect of upward reactive force on the downward pressure perturbation becomes weak for h/h_(LT)>1.Hence,the upper mixed layer depth has significant influences on Langmuir turbulence characteristics.

A machine learning approach to quality-control Argo temperature data

本文提出了一种基于机器学习的Argo浮标温度异常值检测方法.该方法采用机器学习无监督算法高斯混合模型对Argo浮标数据进行聚类分析,并构建包围所有数据点的最小多边形的凸包.基于射线投影算法实现点在多边形内分析,通过自动识别数据点位于凸包内外来判断该数据点数据质量的好坏.本文采用南海区域Argo浮标数据对该方法进行测试,结果表明该方法可以识别70%以上的包含异常值的温度剖面,同时自动标记出各异常值点.

Significant wave height forecasts integrating ensemble empirical mode decomposition with sequence-to-sequence model

As wave height is an important parameter in marine climate measurement,its accurate prediction is crucial in ocean engineering.It also plays an important role in marine disaster early warning and ship design,etc.However,challenges in the large demand for computing resources and the improvement of accuracy are currently encountered.To resolve the above mentioned problems,sequence-to-sequence deep learning model(Seq-to-Seq)is applied to intelligently explore the internal law between the continuous wave height data output by the model,so as to realize fast and accurate predictions on wave height data.Simultaneously,ensemble empirical mode decomposition(EEMD)is adopted to reduce the non-stationarity of wave height data and solve the problem of modal aliasing caused by empirical mode decomposition(EMD),and then improves the prediction accuracy.A significant wave height forecast method integrating EEMD with the Seq-to-Seq model(EEMD-Seq-to-Seq)is proposed in this paper,and the prediction models under different time spans are established.Compared with the long short-term memory model,the novel method demonstrates increased continuity for long-term prediction and reduces prediction errors.The experiments of wave height prediction on four buoys show that the EEMD-Seq-to-Seq algorithm effectively improves the prediction accuracy in short-term(3-h,6-h,12-h and 24-h forecast horizon)and long-term(48-h and 72-h forecast horizon)predictions.

Three-dimensional properties of mesoscale cyclonic warm-core and anticyclonic cold-core eddies in the South China Sea

In general,a mesoscale cyclonic(anticyclonic)eddy has a colder(warmer)core,and it is considered as a cold(warm)eddy.However,recently research found that there are a number of"abnormal"mesoscale cyclonic(anticyclonic)eddies associated with warm(cold)cores in the South China Sea(SCS).These"abnormal"eddies pose a challenge to previous works on eddy detection,characteristic analysis,eddy-induced heat and salt transports,and even on mesoscale eddy dynamics.Based on a 9-year(2000–2008)numerical modelling data,the cyclonic warm-core eddies(CWEs)and anticyclonic cold-core eddies(ACEs)in the SCS are analyzed.This study found that the highest incidence area of the"abnormal"eddies is the northwest of Luzon Strait.In terms of the eddy snapshot counting method,8620 CWEs and 9879 ACEs are detected,accounting for 14.6%and 15.8%of the total eddy number,respectively.The size of the"abnormal"eddies is usually smaller than that of the"normal"eddies,with the radius only around 50 km.In the generation time aspect,they usually appear within the 0.1–0.3 interval in the normalized eddy lifespan.The survival time of CWEs(ACEs)occupies 16.3%(17.1%)of the total eddy lifespan.Based on two case studies,the intrusion of Kuroshio warm water is considered as a key mechanism for the generation of these"abnormal"eddies near the northeastern SCS.

CNN-BiLSTM-Attention Model in Forecasting Wave Height over South-East China Seas

Though numerical wave models have been applied widely to significant wave height prediction,they consume massive computing memory and their accuracy needs to be further improved.In this paper,a two-dimensional(2D)significant wave height(SWH)prediction model is established for the South and East China Seas.The proposed model is trained by Wave Watch III(WW3)reanalysis data based on a convolutional neural network,the bidirectional long short-term memory and the attention mechanism(CNNBiLSTM-Attention).It adopts the convolutional neural network to extract spatial features of original wave height to reduce the redundant information input into the BiLSTM network.Meanwhile,the BiLSTM model is applied to fully extract the features of the associated information of time series data.Besides,the attention mechanism is used to assign probability weight to the output information of the BiLSTM layer units,and finally,a training model is constructed.Up to 24-h prediction experiments are conducted under normal and extreme conditions,respectively.Under the normal wave condition,for 3-,6-,12-and 24-h forecasting,the mean values of the correlation coefficients on the test set are 0.996,0.991,0.980,and 0.945,respectively.The corresponding mean values of the root mean square errors are measured at 0.063 m,0.105 m,0.172 m,and 0.281 m,respectively.Under the typhoon-forced extreme condition,the model based on CNN-BiLSTM-Attention is trained by typhooninduced SWH extracted from the WW3 reanalysis data.For 3-,6-,12-and 24-h forecasting,the mean values of correlation coefficients on the test set are respectively 0.993,0.983,0.958,and 0.921,and the averaged root mean square errors are 0.159 m,0.257 m,0.437 m,and 0.555 m,respectively.The model performs better than that trained by all the WW3 reanalysis data.The result suggests that the proposed algorithm can be applied to the 2D wave forecast with higher accuracy and efficiency.

Oceanic mesoscale eddy in the Kuroshio extension:Comparison of four datasets

Some relatively mature mesoscale eddy products have been released for scientific purposes in recent decades.However,the metrics used to identify eddies,the tracking methods,and the definition of the physical parameters are all different across the different datasets,so intercomparisons and validation of these datasets are badly needed.Here,the authors intercompare the basic features of ocean mesoscale eddies in the Kuroshio extension region from four eddy datasets—namely,Chelton,GEM-M,Faghmous,and Dong.In the case study,eddy numbers and locations as well as the eddy tracks identified by the four datasets are compared for a specific date.The authors find that all the datasets have different eddy numbers,but more than 50%of identified eddies coincide.GEM-M,with the so-called"segmentation"algorithm,can identify considerably more eddies than others,while Chelton identifies fewer eddies due to tracking errors,which also lead to a long lifespan.From the analysis of the probability distribution function of eddy features,GEM-M eddies tend to have a larger amplitude and radius and Chelton tends to have long-life eddies.It is further found that the geographic distributions and temporal variation of normalized eddy features are highly similar among the four datasets—particularly among Chelton,Fahgmous,and Dong.In addition,the mean trajectories of the four datasets are generally overlapped initially,and then spread after 245 days.The findings help toward better understanding the uncertainties of eddy features in the Kuroshio extension region.

Numerical study of the seasonal salinity budget of the upper ocean in the Bay of Bengal in 2014

Impact factors on the salinity budget, especially the eddy salt fluxes and smaller-scale diffusive salt fluxes for the upper 50 m of the Bay of Bengal(BoB) in 2014 are investigated using a box model based on the Regional Ocean Modeling System(ROMS) daily outputs. The model results reproduce that the precipitation and river runoff s are the dominant factors modulating the sharp salinity decrease during the summer monsoon season. The analysis shows that the salinity increase after the summer monsoon is mostly due to the meridional advective and diffusive salt fluxes. The vertical advective salt flux, which is sensitive to the different signals of the wind stress curl, plays an important role in balancing the salinity change induced by the meridional advective salt flux during both the summer and winter monsoon seasons. Distinctive spatial mesoscale structures are presented in the eddy salt flux throughout the year, and their contributions are sizeable(over 30% in the meridional direction and about 10%–30% in the vertical direction). The meridional eddy salt flux is larger in the monsoon seasons than that in the inter-monsoon seasons, and in a positive pattern near the western boundary during the winter monsoon and autumn inter-monsoon. The vertical eddy salt flux makes an important contribution to the salinity budget, especially along the coastal area and around the Andaman and Nicobar Islands. The vertical eddy salt flux becomes large when a tropical cyclone passes the area.

Eddy diffusivity and coherent mesoscale eddy analysis in the Southern Ocean

The spatial distribution of eddy diffusivity,basic characteristics of coherent mesoscale eddies and their relationship are analyzed from numerical model outputs in the Southern Ocean.Mesoscale fluctuation information is obtained by a temporal-spatial filtering method,and the eddy diffusivity is calculated using a linear regression analysis between isoneutral thickness flux and large-scale isoneutral thickness gradient.The eddy diffusivity is on the order of O(103 m2/s)with a significant spatial variation,and it is larger in the area with strong coherent mesoscale eddy activity.The mesoscale eddies are mainly located in the upper ocean layer,with the average intensity no larger than 0.2.The mean radius of the coherent mesoscale cyclonic(anticyclonic)eddy gradually decays from(121.2±10.4)km((117.8±9.6)km)at 30°S to(43.9±5.3)km((44.7±4.9)km)at 65°S.Their vertical penetration depths(lifespans)are deeper(longer)between the northern side of the Subpolar Antarctic Front and 48°S.The normalized eddy diffusivity and coherent mesoscale eddy activity show a significant positive correlation,indicating that coherent mesoscale eddy plays an important role in eddy diffusivity.

Submesoscale eddies in the East China Sea detected from SAR images

Seven-year(2005-2011)Synthetic Aperture Radar(SAR)images are applied to study oceanic eddies in the East China Sea.It is found that most of these eddies detected from the SAR images are less than 10 km,which are submesoscale eddies.Seasonal differences are evident in the distribution of eddies,with the highest and the lowest number of eddies noted in summer and winter,respectively.Since slick streaks in SAR images look dark,an eddy identified due to the slicks is referred to as“black eddy”.As a result of wave-current interactions in the zones of current shear,it can be seen that an eddy exhibits a bright curve,the eddy is called“white eddy”.During the seven years,95 black eddies and 50 white eddies are identified in the study area.Black eddies are found in the whole study area while white eddies are mainly distributed in the vicinity of the Kuroshio Current.This study suggests that the distribution of the white eddy is denser around the Kuroshio because of the strong shear in the Kuroshio region.In terms of the eddy sizes,white eddies are generally smaller than black eddies.

An approach to determine coeffi cients of logarithmic velocity vertical profile in the bottom boundary layer

Velocity vertical profiles in the bottom boundary layer are important to understand the oceanic circulation.The logarithmic vertical profile,u=A ln z+B,is the universal profile for the horizontal velocity in the boundary layer,in which two coefficients(A and B)need to be determined.The two coefficients are the functions of the friction velocity(u_(*))and the roughness length(z_(0)),and they are calculated using u_(*)and z_(0).However,the measurement of u_(*)and z_(0) is a challenge.In the present study,an approach is developed to estimate the two coefficients(A and B)by using a series of fl ume laboratory experiments with fl at boundary and regularly distributed cylinders as the rough boundaries.An acoustic doppler velocimeter(ADV)is used to measure the velocity vertical profiles of the steady flow.Using the measured velocity data,the regressed logarithmic profiles are obtained.Based on the series of the A and B values,the mathematical formula for A and B are statistically established as the function of the cylinder height,inflow velocity,and the water depth,which avoids the measurement of the friction velocity and the roughness length.

Sri Lanka seasonal warm pools

Sri Lanka seasonal warm pools(SSWPs)with high sea surface temperature(SST)(greater than 29.6℃in summer and greater than 28℃in winter)around the Sri Lanka Island are discovered using 37-year(1982–2018)satellite remote sensing SST data.It is identifi ed that during the onset of the summer(winter)monsoon,the maximum SST in the SSWPs can be 1.8℃(0.8℃)higher than the surrounding ocean.Due to orographic infl uences,the wind intensity on the leeward side of Sri Lanka is weakened,reducing evaporation and oceanic mixing,and resulting in higher SST.The analysis of the long-term data shows that SSWPs are related to El Niño-Southern Oscillation(ENSO)events.Background currents are also found to play important roles in governing the SSWPs intensities through the heat advection.

Eddy analysis in the Eastern China Sea using altimetry data

Statistical characteristics of mesoscale eddies in the Eastern China Sea （ECS） are analyzed using altimetry sea surface height anomaly （SSHA） data from 1993 to 2010. A velocity geometry-based automated eddy detection scheme is employed to detect eddies from the SSHA data to generate an eddy data set. About 1,096 eddies （one lifetime of eddies is counted as one eddy） with a lifetime longer than or equal to 4 weeks are identified in this region. The average lifetime and radius of eddies are 7 weeks and 55 km, respectively, and there is no significant difference between cyclonic eddies （CEs） and anticyclonic eddies （AEs） in this respect. Eddies＇ lifetimes are generally longer in deep water than in shallow water. Most eddies propagate northeastward along the Kuroshio （advected by the Kuroshio）, with more CEs generated on its western side and AEs on its eastern side. The variation of the Kuroshio transport is one of the major mechanisms for eddy genesis, however the generation of AEs on the eastern side of the Kuroshio （to the open ocean） is also subject to other factors, such as the wind stress curl due to the presence of the Ryukyu Islands and the disturbance from the open ocean.

Numerical research on evolvement of submarine sand waves in the Northern South China Sea

Submarine sand waves, vital to seabed stability, are an important consideration for oceanic engineering projects such as oil pipe lines and submarine cables. The properties of surface sediment and the evolvement of submarine sand waves in a specified area in the South China Sea are studied using both a hydrological model and field observational data. The bottom flow field data between 2010 and 2011 in the study area are simulated by the Regional Ocean Model System （ROMS）. The migration of submarine sand waves is calculated using Rubin＇s formula along with typhoon data and bottom flow field data, which allows for the analysis of sand wave response under the influence of typhoons. The migration direction calculated by Rubin＇s formula and bottom flow are very similar to collected data. The migration distance of different positions is between 0.0 m and 21.8 m, which reciprocates cumulatively. This shows that Rubin＇s formula can predict the progress of submarine sand waves with the bottom flow simulated by ROMS. The migration distances of 2 sites in the study area are 2.0 m and 2.9 m during the typhoon ＂Fanapi＂. The proportion of the calculated migration distance by the typhoon is 9.17% and 26.36% of the annual migration distance, respectively, which proves that the typhoon can make a significant impact on submarine sand waves.

Stimuli-responsive dual drugs-conjugated polydopamine nanoparticles for the combination photothermal-cocktail chemotherapy

We developed one-pot aqueous copolymerization of two dopamine prodrugs to prepare dual drugsconjugated polydopamine nanoparticles(PDOXCBs),which integrated near infrared(NIR)-mediated photothermal effect with cocktail chemotherapy into one copolymer nanoparticle.Upon a mild NIR irradiation(808 nm,1 W/cm^(2),10 min),PDOXCBs gradually heated aqueous solution over 12.8-13.9℃,which accordingly enhanced in vitro dual doxorubicin(DOX) and chlorambucil(CB) drug-release with assistance of the other stimuli of pH 5.0 and 10 mmol/L D,L-dithiothreitol(DTT).The combination photothermal-cocktail chemotherapy(PTT-CCT) treatment based on PDOXCB27 plus NIR irradiation gave a highly lowered half maximal inhibitory concentration(IC_(50)) of 2.23 μg/mL and a combination index of0.36,displaying a superior synergistic effect between PTT and CCT in vitro.

Seasonal variation of the global mixed layer depth： comparison between Argo data and FIO-ESM

The present study evaluates a simulation of the global ocean mixed layer depth （MLD） using the First Institute of Oceanography-Earth System Model （FIO- ESM）. The seasonal variation of the global MLD from the FIO-ESM simulation is compared to Argo observational data. The Argo data show that the global ocean MLD has a strong seasonal variation with a deep MLD in winter and a shallow MLD in summer, while the spring and fall seasons act as transitional periods. Overall, the FIO-ESM simula- tion accurately captures the seasonal variation in MLD in most areas. It exhibits a better performance during summer and fall than during winter and spring. The simulated MLD in the Southern Hemisphere is much closer to observations than that in the Northern Hemisphere. In general, the simulated MLD over the South Atlantic Ocean matches the observation best among the six areas. Additionally, the model slightly underestimates the MLD in parts of the North Atlantic Ocean, and slightly overestimates the MLD over the other ocean basins.

Recent Developments in Artificial Intelligence in Oceanography

With the availability of petabytes of oceanographic observations and numerical model simulations,artificial intelligence(AI)tools are being increasingly leveraged in a variety of applications.In this paper,these applications are reviewed from the perspectives of identifying,forecasting,and parameterizing ocean phenomena.Specifically,the usage of AI algorithms for the identification of mesoscale eddies,internal waves,oil spills,sea ice,and marine algae are discussed in this paper.Additionally,AI-based forecasting of surface waves,the El Niño Southern Oscillation,and storm surges is discussed.This is followed by a discussion on the usage of these schemes to parameterize oceanic turbulence and atmospheric moist physics.Moreover,physics-informed deep learning and neural networks are discussed within an oceanographic context,and further applications with ocean digital twins and physics-constrained AI algorithms are described.This review is meant to introduce beginners and experts in the marine sciences to AI methodologies and stimulate future research toward the usage of causality-adherent physics-informed neural networks and Fourier neural networks in oceanography.