CHEN Yi-zhan (陳宜展), ZHAN Jie-min (詹杰民), LUO Ying-ying (羅瑩瑩)

Department of Applied Mechanics and Engineering, Sun Yat-sen University, Guangzhou 510275, China,

E-mail: chenyzh8@mail2.sysu.edu.cn

WAI Onyx W. H. (韋永康)

Department of Civil and Environmental Engineering, The Hong Kong Polytechnic University, Hong Kong, China TANG Ling (唐靈)

Department of Applied Mechanics and Engineering, Sun Yat-sen University, Guangzhou 510275, China

Responses of thermal structure and vertical dynamic structure of South China Sea to Typhoon Chanchu*

CHEN Yi-zhan (陳宜展), ZHAN Jie-min (詹杰民), LUO Ying-ying (羅瑩瑩)

Department of Applied Mechanics and Engineering, Sun Yat-sen University, Guangzhou 510275, China,

E-mail: chenyzh8@mail2.sysu.edu.cn

WAI Onyx W. H. (韋永康)

Department of Civil and Environmental Engineering, The Hong Kong Polytechnic University, Hong Kong, China TANG Ling (唐靈)

Department of Applied Mechanics and Engineering, Sun Yat-sen University, Guangzhou 510275, China

(Received December 26, 2013, Revised March 15, 2014)

The response of the South China Sea (SCS) to Typhoon Chanchu (2006) was examined using the MM5 and POM model. In the POM model, sea surface boundary conditions were forced by the simulation wind field from MM5, the velocity forcing was introduced in the eastern boundary and the computational schemes of heat fluxes at the surface were introduced. Comparison with the observation data shows that the simulated results are reliable. In the response process of the SCS to Typhoon Chanchu, the influence of the heat fluxes on thermal structure of the SCS was regionally different. Strong wind forcing would lead to upwelling phenomenon in the lateral boundary of deep water basin. Furthermore, the Ekman pumping theory was used to discuss subsurface upwelling and downwelling phenomenon in typhoon forced stage.

South China Sea (SCS), typhoon, thermal structure, vertical dynamic structure, Mesoscale Model 5 (MM5), princeton ocean model (POM)

Introduction

Typhoon is one of the most powerful mutant weather processes in the ocean, which results in great exchange of heat, energy and material between air and sea. The response of the ocean to typhoon has long been a focus for concern. At present, observation and numerical simulation are the most widely used methods for the research on this topic. The results of the historical observations show that sea surface temperature (SST) decreased in the upper ocean and the maximal cooling area of SST appeared in the right bias of the typhoon track[1-3]. The data could be accidentallycollected from buoys during typhoon, which means that the temporal and spatial resolutions are not high enough to throw a light on the process. On the other hand, the satellite remote sensing data would probably be lack of large area of clear image due to cloud cover. With the development of the computer technology, numerical simulation becomes an important way to study this problem.

The mechanism of the ocean response to typhoon still needs further study. Particularly, more attention should be paid to the SST and loss of energy from the mixed layer. On the basis of strong wind velocity profile and nonlinear ocean mixed layer current, the response of SST has cooler water on the right, entrainment dominates the ocean mixed layer heat budget and is also significant in the mixed layer moment balance. Another important mechanism is that inertial rotation of the current induces convergence and divergence, and hence vertical motion (pumping) in mixed layer[4]. However, the ocean vertical velocity response to typhoon is not yet system analysis. In recent years,the ocean response process to typhoon in the South China Sea (SCS) has been the burgeoning studies[1,5]. The SCS is an activity typhoon area in the western Pacific, which have been integrated the dynamic analysis of heat budget in the upper ocean. In this paper, we will focus on the thermal structure and vertical dynamic structure in the SCS response process.

第三，旅游危機事件網(wǎng)絡輿情本身在傳播的各個階段表現(xiàn)出不同的特點。旅游危機事件必須在不計其數(shù)的信息中“脫穎而出”，方能獲取公眾的廣泛關注。潛伏期的輿情本體通常是對旅游危機事件的始末進行描述，這一階段的輿情本體主要表現(xiàn)出信息醞釀的特點；爆發(fā)期的旅游危機事件網(wǎng)絡輿情已經(jīng)成功進入輿論場，并迅速吸引了大量的公眾關注，這一階段的輿情本體主要表現(xiàn)出信息擴散的特點；成熟期旅游危機事件的相關信息已經(jīng)大范圍擴散，公眾關注點開始轉移，這一階段的輿情本體主要表現(xiàn)出輿論集聚的特點；平息期公眾對事件關注熱度降低，參與各方逐步退出網(wǎng)絡輿論場域，這一階段的輿情本體主要表現(xiàn)出輿情消散的特點。

The Mesoscale Model 5 (MM5) and princeton ocean model (POM) have been adopted to calculate the SCS response to Typhoon Chanchu (2006) and monthly mean climatic forcing. In this study, we pay attention to the effects of heat flux on SST, water flow in the deep basin lateral boundary, and upwelling (downwelling) phenomenon in the upper SCS.

Fig.1 Track of Typhoon Chanchu in the SCS

1. Typhoon Chanchu

清政府第一次鴉片戰(zhàn)爭戰(zhàn)敗后，西方列強從協(xié)定關稅開始，參與清政府海關稅則的制定。繼而以維護對外貿(mào)易秩序為借口，借助上海小刀會起義導致江海關癱瘓的契機，強行成立英、法、美三國關稅管理委員會，取得江海關關稅的征收權，逐步形成了外籍稅務司制度。外籍稅務司制度不僅保障了不平等條約下所要求的稅收收入，并且還有效防止了貪污腐敗，所以它被中國政府所接納，從而使外籍稅務司制度在中國推廣開來。這樣，以英國為首的西方列強攫取了中國關稅的征收權。

2. Model description

Then, the net heat flux is given as

2.1Simulation of Typhoon Chanchu with MM5

The atmospheric model MM5(version V3) was developed by the National Center for Atmospheric Research (NCAR) at the Pennsylvania State University, details of which have been outlined by Tang et al.[3]30 s resolution global terrain data and 35 sigma levels in the vertical direction were used in MM5. The resolution of DOMAIN1 was 40.5 km, and the DOMAIN2 was 13.5 km. The NCEP data assimilation analysis of Newtonian relaxation was adopted in the surface and upper air in this simulation. Relaxation program was used as boundary conditions.

2.2Equilibrium mode in POM

The POM is a 2.0 turbulence closure model, in which the topography is smoothed by the following equation

whereH1andHi+1are depths in the neighboring grids, andαis a smoothing factor with a value of 0.4.

In this study, the POM domain covered all the SCS area between 98oE to 126oE, 3oS to 26oN with a 15 km horizontal resolution, and were divided into 50 sigma levels in the vertical scheme with a higher resolution in the upper layer. The model bottom topography was constructed from ETOPO 5＇5＇× geophysical data. The model could treat the cases with the minimum depth of 10m and the maximum depth of 4 000 m, respectively. The involved eastern boundary condition was open from 12oN to 26oN and forced with the flow field data calculated by Zhan and Yan[6], and other boundaries were closed. For initial conditions, the monthly Levitus sea temperature data and themonthly DASILVA wind field data were used. For calculating instant heat flux to force sea surface thermal variation, the correlative formulas will be discussed in the following sections.

2.3Response process of the SCS to Typhoon Chanchu by POM

術前準備:備好負壓吸引器、氧氣、氣管插管等搶救用品及藥物。患者術前需要排空膀胱,護士要給予指導,準備完成后進入取卵室,核對患者信息,確認簽字無誤,這部分工作由三方人員完成,巡回護士、手術醫(yī)師及實驗室人員,之后即進入手術過程。

花鳥畫最早可以追溯到原始社會時期，發(fā)展至兩漢時期，花鳥畫已經(jīng)初具雛形。東晉畫家謝赫憑借其作品《畫品》成為中國繪畫史上第一位有跡可循的花鳥畫家。隨著唐朝，宋朝的國家經(jīng)濟發(fā)展日趨繁榮，國泰民安，在藝術文化領域夜呈現(xiàn)一派欣欣向榮的景象，花鳥畫在這一時期也漸漸發(fā)展成熟。元朝時期出現(xiàn)了一些以“水墨梅竹”為主要繪畫對象的畫家。明朝則是以徐渭和陳道復的花鳥作品見著，清代則是出現(xiàn)了筆墨獨具風格的畫家，如“八大山人”等。發(fā)展至現(xiàn)代，花鳥畫在繪畫領域的總體發(fā)展平平，但是不可否認的是，在花鳥畫的不同風格都各有成者，他們?yōu)榛B畫的未來發(fā)展注入了新的力量。

In order to more clearly reflect the dynamic structure of the ocean, we examine the interception horizontal section geiven as the following formula in the model

whereKMis the vertical turbulent viscosity[7], andxτandyτare the tensor components of sea surface wind stress.

Fig.2 Simulated sea surface wind field in MM5. The typhoon center can be clearly seen in the pictures

Through comparing sea surface current results of typhoon conditions (Figs.2-3), it was found that the currents had larger sea surface velocity on the right side of the Typhoon Chanchu’s track, that is to say, the response process had an asymmetric bias to the right, which is consistent with Price’s opinion[4]. During the typhoon period, the internal water flew northeast through the Luzon Strait to the western Pacific from the SCS, and the water exchange in the Luzon Strait was significantly strengthened.

Fig.3 Simulated sea surface current field in POM

Fig.4 Comparison between the SST from AMSR remote sensing data and POM simulation results on 11 May 2006 (before Typhoon Chanchu)

Fig.5 Comparison between the SST from AMSR remote sensing data and POM simulation results on 20 May 2006 (after Typhoon Chanchu)

3. Case validation

The calculated SST is validated with the AMSR remote sensing data (NASR, http://wwwghcc.msfc. nasa.gov/AMSR). The AMSR data appears to be data absence during typhoon, thus we took those before (11 May) and after (20 May) Typhoon Chanchu’s SST for contrast and comparison. As are shown in Figs.4-5, the calculated SST results are consistent with the AMSR SST data in spatial distribution, and the POM’s SST is cooler than the AMSRs. Furthermore, the simulated SST has right-hand bias, which agrees with the former conclusions[8]. Another comparison is that there is a heating process in the southern SCS, so we can believe the simulated results are credible. In the last section, the simulated sea surface current results have been compared with the observed theory (see Figs.2-3).

4. Results and discussion

4.1Heat flux effect to SST

圍繞學校應用型人才培養(yǎng)目標，通過把教學內(nèi)容(使用Dreamweaver軟件制作網(wǎng)頁)、教學軟件(Dreamweaver)、教學網(wǎng)站(包含課件、大綱、案例、項目等內(nèi)容)等常規(guī)分離的教學項目合并到教師團隊構建的真實的網(wǎng)絡學習平臺中，精選教學內(nèi)容，改進教學方法，實現(xiàn)學校人才培養(yǎng)目標與學生的能力培養(yǎng)的統(tǒng)一，把知識學習與能力培養(yǎng)始終貫穿在真實的網(wǎng)絡學習平臺中，實現(xiàn)課堂內(nèi)外學習環(huán)境的統(tǒng)一，促進學生繼續(xù)深入學習相關知識.詳見圖1.

The heat flux under the influence of the typhoon has the characteristics of large spatial and temporal differences. In this paper, we use the heat flux attn-1to calculate that attn, thus the model can give prominence to the SST characters which are under the influence of heat flux. The heat flux can be constructed by the following equations:

Here, the shortwave radiationQsand the longwave radiationQlare calculated according to Bignami and Hopkins[9]QhandQeare respectively the sensible heat flux and the latent heat flux.I0is the solar radiation flux,Cis the total cloud amount,sαis the albedo of sea surface,εis the sea surface emissivity,σSBis the Stefan-Boltzmann constant,λis a blocking factor to account for occulted clouds,eis the vapor pressure,ρis the air density,CPis the specific heat of air,LPis the latent heat of evaporation of seawater,qsis the sea surface saturation specific humidity, andqAis the specific humidity. The sensible heat transfer coefficientCHand the latent heat transfer coefficientCEare given by Drennan et al.[10]andWis the wind velocity.

In this paper, we focus on the SCS thermal and dynamic structures generated by the typhoon, which has a close relation to the turbulent mixing. With excellent capabilities of simulating weather phenomena, the MM5 can provide high resolution dynamic coordination data and be propitious to study the complex physical mechanism. The POM is a three-dimensional baroclinic ocean model with 2.5-order turbulence closure sub-model, thus the vertical eddy diffusivity result can reduce human factors due to the vertical distribution pre-setting.

The net heat flux added to the POM model is

whereKHis the vertical diffusion coefficient, andCwis the specific heat of seawater.

In this paper, MM5 and POM have been employed to simulate the response process of SCS thermal structure and vertical dynamic structure under the influence of Typhoon Chanchu in May, 2006. We focus on the signature in SST of heat flux and the associated vertical water flow, and the main conclusions can be summarized as follow:

而上面的分析也提到過，有些時候，在現(xiàn)代漢語“讓”字句中，動作者有時被省略。此時雖然含義上沒有什么變化且在英語小句中不是很常見，但是在現(xiàn)代漢語的小句中還是較為常見的。例如：

The POM motion equation was transformed to Cartesian vertical velocity equation as followwhereωis surface velocity component in Sigma

To compare with the SST on the 11th of May before typhoon, the maximum SST difference was about -7oC. In the domains under Typhoon Chanchu’s influence, the SST was on average changing 0.2oC caused by the heat flux term.

Table 1 Comparison of Correlation betwen SST and net heat flux

Taking interval 6 h heat flux and SST calculated results from 11th to 18th of May for correlation analysis, and considering different typhoon impact domains to be an average research, as shown in Table 1, we found that the domain with greater cooling magnitude water had lower correlation coefficient, and the influence of heat flux to ocean during typhoon had significant local differences.

巖土工程勘察設計和施工過程中的水文地質問題分析………………………………………………………… 張衎（10-87）

4.2Response of deep water basin vertical dynamic st

ructure

Previous studies have indicated the SST development mechanism during typhoon[11,12]. The prevailing view is that typhoon wind stress can cause strong vertical mixing in the upper ocean, the cold water in the mixed layer can be brought to sea surface by entrainment and pumping with the effect of shear velocity, which is the main reason for the decrease in SST.

SCS is the largest marginal sea in the world, also it is one of the few semi-enclosed deep water basins, which topography is steep in the continental shelf. In recent years, several authors[12,13]have already researched the response mechanism in the upper SCS, but these papers did not involve the response mechanism in deep water basin, so we will analyze deep water response process by numerical analysis.

由于檔案本身沒有特征信息，首先需要對檔案數(shù)據(jù)進行處理，從每一個檔案中提取對應的特征數(shù)據(jù)，映射到對應的特征視圖中。不同類型的檔案數(shù)據(jù)需要采用不同的特征提取方法，除了傳統(tǒng)的人工著錄整理之外，還包括OCR文字識別、語義分析、視頻內(nèi)容提取等技術。本文不深入研究檔案實體的特征提取技術。所采用的實驗樣本均為經(jīng)過人工著錄，已經(jīng)包含所需要特征項的檔案樣本。

Fig.6 Vertical velocity (m/s) between (107oE, 16oN) and (120oE, 23oN) to research the vertical water flow of SCS basin boundary with time

On the 16th of May, the maximal net heat flux reached 1.1 kW/m2, and the calculated results were similar to the observation of Typhoon Lily[8]. It is obvious that the strong wind and the air-sea temperature difference are critical to large heat flux values (see Eqs.(3)-(7)). When Typhoon Chanchu initially came into SCS, the net heat flux was negative, SST was cooling, the typhoon intensity was strengthened and the maximum wind velocity increased to 45 m/s from 33 m/s. On the 16th of May, the net heat flux was positive, which restrained the SST cooling, and the typhoon intensity tended to be stable. That is to say, in strong wind velocity domain, the net heat flux transport direction depends on the air-sea temperature difference and had a direct relationship to typhoon intensity.

The POM is a three-dimensional primitive equation model. It takes the Boussinesq approximation and computes fluid motion from the Navier-Stocks equations. In the model,σcoordinate is used as the vertical coordinate, and bottom topography and sea surface undulation are considered. Explicit difference scheme and implicit difference scheme are used in the horizontal and vertical time differences, respectively. The latter can eliminate the time limitation to vertical coordinate, so the higher resolution can be reached at the surface and at the bottom.

Fig.7 Vertical velocity distribution (m/s) in SCS basin boundary along 111.5oE. Compared with the results (Fig.6) at different times produced by POM, the vertical water movement of basin boundary was more obvious than that in open ocean

The vertical velocity is mainly affected by horizontal velocity, terrain and sea surface undulation (Eq.(9)). Another deduction from Eq.(9) is that the horizontal velocity increases during typhoon, and the sea surface undulation is also enhanced, therefore, vertical velocity changes distinctly. In northern SCS, the shelf is gentle but the continental margin is very steep and close with basin. We selected a connector between (107oE, 16oN) and (120oE, 23oN) to investigate the dynamic structure in the slope section of SCS (see Fig.1). For the vertical velocity on the 14th to 17th of May as shown in Fig.6, the large values of water flow appears in different depths with the phase changing with time. It was obvious that the current violent reaction to Typhoon Chanchu is due to the shear velocity in the mixed layer, and it produced significant vertical water flow when horizontal velocity was reversed. There was a noticeable upwelling phenomenon in the basin boundary at 111.5oE (see Fig.6).

In Fig.7, the maximum vertical water flow of SCS appearing in the steep basin boundary, which is consistent with Spall’s theory[11]. The relative vorticity, which is produced by vertical movement stretching in the interior basin, only can be consumed near the boundary with large lateral eddy viscosity. Typhoon Chanchu passed through the SCS basin from the 14th to the 17 th of May, the maximal depth was about 2 000 m affected by the local movement of boundary water, there was upwelling generated in the upper lateral boundary, while downwelling intensity increased at the bottom of the boundary. After the typhoon crossed the basin, the subsurface upwelling intensity significantly decreased. The upwelling of basin lateral boundary influenced by Typhoon Chanchu was 10 times greater than in the normal state. So we can reach a conclusion that with the effect of strong wind forcing, the momentum and the energy disturbances in the middle of the basin can cause upwelling in the narrow viscous boundaries.

4.3Response of thermal and dynamic structure in sea subsurface

In recent years, the response of thermal and dynamic structure in sea subsurface during typhoon attracted the attention of physical oceanography scholars[14-16]. According to the Ekman pumping theory, we get the relationship between wind stress and sea vertical velocity (positive for upwelling and negative for downwelling in Eq.(10))

wherewE(0) is the sea surface vertical velocity,τthe wind stress,ρthe sea water density, andfthe Coriolis parameter. Therefore, under the influence of the northern hemisphere typhoon, counterclockwise vortices are formed in sea surface, and the upwelling is generated in the role of curl field at typhoon center.The mixed layer increases and SST decreases due to entrainment effect. Typhoon Chanchu was located in the center of SCS at 8 a.m. on May 15 through comparatively analyzing the vertical sections of calculated results as in Figs.8-10 (with the typhoon back section 12oN, typhoon center section 14.5oN and typhoon front section 17oN).

1.2.1 桿努盡煙水提液的制備 取桿努盡煙干樣2 kg，用蒸餾水浸泡，并于90℃的電熱恒溫水箱中連續(xù)浸提12 h，收集并過濾浸提液，旋蒸后合并浸提液，濃縮至藥物濃度2.5 kg/L(按生藥量計)，最后分裝滅菌，置于4℃冰箱保存?zhèn)溆谩?/p>

Fig.8 Change in temperature of the mixed layer at 8 a.m. on 15 May

The temperature difference of subsurface (Fig.8) was calculated for 8 a.m. on May 15 and 8 a.m. on May 11 (before typhoon), which was similar to Price’s observation distribution[4]. That is, with the effect of typhoon, sea surface water has low temperature and high salinity, while under-layer temperature rises. In the normal SCS condition, vertical flow magnitude is about 10-6m/s, which is very small compared with horizontal current. In the response process of SCS, subsurface vertical velocity increases 500 times. Typhoon Chanchu’s active domain in SCS was primary in the deep water region, and hence thermal structure and dynamic structure was little affected by terrain within 200 m depth, so it can distinctly reflect the subsurface response process. In the Typhoon Chanchu period, local sea water flowed spread from the center (Fig.9), upwelling and downwelling were staggered. That is due to the Ekman pumping phenomenon with a partial eddy structure under the geostrophic action.

Fig.9 Change in horizontal velocity of the mixed layer (positive for east direction) at 8 a.m. on 15 May

POM was started from January, when it could steadily calculate and simulate sea conditions in May, the MM5’ 1 h wind and air temperature fields in May 2006 were induced as sea surface boundary condition to drive the ocean model. The POM wind stress is calculated from the following equation

wherehn=Hσn(hn≤h＜hn+1),σnis the sigma vertical stratification factor, andfis the physical variables.

As is shown in Fig.11, the vertical motion of water was relatively strong at the depth of 120 m. At this layer horizontal water flow had been significantly weakened compared with sea surface flow, so it was prone to generate shear velocity. At the forced stage of typhoon, downwelling was caused by convergence current at the edge of typhoon, while upwelling was caused by divergence current in the center of Typhoon Chanchu, however, the maximum vertical flow position was not consistent with the maximum wind stress position, which was influenced by the topography and the geostrophic current.

Fig.10 Change in Vertical velocity of the mixed layer at 8 a.m. on 15 May

5. Conclusions

According to COADS net heat flux data, the net heat flux of SCS is in the range of 100 W/m2-200 W/m2in May. Equation (8) is in the condition for the balance of the vertical heat flux transport and the non-adiabatic heating, without considering the local variation of temperature and horizontal advection conditions, and the SST rises in May (positive for the downward direction and negative for the upward direction in Eq.(8)). In simulation, Typhoon Chanchu’s wind velocity was taken as 45 m/s2at 8 a.m..

Typhoon Chanchu started as a tropical storm, forming 550 km southeast of the Mindanao Island of Philippines on 9 May, 2006. It was upgraded to a typhoon with maximal wind velocity of 33 m/s when moving into SCS along northwest direction at 8 a.m. on 13 May. Typhoon Chanchu’s wind speed exceeded 45 m/s and suddenly turning 90onorthward at 2 a.m. on 15 May as it approached the strong typhoon level, at last it made landfall at 2 a.m. on 18 May on Guangdong province in China. Typhoon Chanchu was a typical typhoon generated in La Nina years. It initially formed in the western Pacific, and was supplied ample energy by high temperature sea water when entering SCS. Thus, the typhoon had strong wind velocity and a wide affecting range. Figure 1 shows the running track and strength of Typhoon Chanchu. Contour line denotes isobaths in meters. The left path is MM5 simulation results, while the right path is actual track of Typhoon Chanchu. The numbers are the dates in May, 2006. The colors of typhoon paths show the typhoon intensity, red line indicating typhoon, blue line strong tropical storm, green line tropical storm, and black line tropical depression, respectively. The brown line denotes a section plane, which will be described in the following description. It is found that after moving into the SCS from 12 to 18 of May, Typhoon Chanchu kept strong wind velocity motion in SCS for a few days, which provided a good condition to simulate the SCS response process to Typhoon Chanchu.

(1) At 8 a.m. on May 16, the maximum wind velocity reached 45m/s and the maximum net heat flux was 1.1 kW/m2in Typhoon Chanchu’s domain. The SST and the air-sea temperature difference indeed had significantly association with heat flux transport, and high temperature sea water was propitious to enhance typhoon intensity. When SST was reduced to a certain extent, the typhoon intensity no longer increased. The correlation analysis between SST and net heat flux had local difference. The more the SST decreased, the lower the correlation coefficient was.

(2) With strong influence of wind forcing, upwelling was caused by the internal vertical movement stretching action in the upper narrow viscous basin boundary, while downwelling had weakened in the under layer boundary. In Typhoon Chanchu period, the extreme depth affected by boundary water local vertical movement was about 2 000 m, and it was prone to appear vertical water flow in the critical state of mixed layer. Water exchange was clearly enhanced in the Luzon Strait, the internal water of SCS flew northeast through the Luzon Strait to the western Pacific.

Fig.11 Comparison of vertical velocity (m/s) at different horizontal sections at 8 a.m. on May 11

(3) The SST decreased in the typhoon center, while sea subsurface temperature increased around. Vertical velocity was relatively intense at 120 m depth.Another important circulation feature produced by POM is that, downwelling was caused by convergence current at the edge of typhoon, while upwelling was caused by divergence current in the center of Typhoon Chanchu, however, the maximum vertical flow position was not consistent with the maximum wind stress position.

[1] CHU P. C., VENEZIANO J. M. and FAN C. et al. Response of the South China Sea to tropical cyclone Ernie 1996[J].Journal of Geophysical Research,2000, 105(C6): 13991-14009.

[2] SRIVER S. L., HUBER M. Observational evidence for an ocean heat pump induced by tropical cyclones[J].Nature,2007, 447(7144): 577-580.

[3] TANG Ling, ZHAN Jie-min and CHEN Yi-zhan et al. Typhoon process and its impact on the surface circulation in the northern South China Sea[J].Journal of Hydrodynamics,2011, 23(1): 95-104.

該堆于2009年停堆。在完成重大升級改造后，新版Apsara于2018年9月10日實現(xiàn)臨界，使用低濃鈾燃料，最大輸出功率2 MWt。

對于給定的一個包含n個d維數(shù)據(jù)點的數(shù)據(jù)集X={ x1,x2,…,xi,…,xn}，其中xi∈Rd，K 表示整個數(shù)據(jù)集經(jīng)過K-means聚類算法被劃分為數(shù)據(jù)子集的數(shù)目，K個數(shù)據(jù)子集表示為C={ }ck,k=1,2,…,K ，其中每個劃分代表一個類ck，每個類ck有一個類中心μk。采用歐幾里得距離作為不同數(shù)據(jù)點間的相似度和距離判斷的標準，計算ck類內(nèi)各點到聚類中心μk的距離平方和DisSum( )ck，見公式（1）：

[4] PRICE J. F. Upper ocean response to a hurricane[J].Journal of Physical Oceanography,1981, 11(2): 153-175.

[5] ZHENG Z. W., HO C. R. and ZHENG Q. et al. Satellite observation and model simulation of upper ocean biophysical response to upper typhoon Nakri[J].Continental Shelf Research,2010, 30(13): 1450-1457.

[6] ZHAN Jie-min, YAN Xiao-hai. A large eddy simulation turbulence model and a local filtering assimilation method for ocean circulation problems[J].Acta Oceanologica. Sinica,2004, 26(3): 8-15(in Chinese).

[7] MELLOR G. L., BLUMBERG A. F. Wave breaking and ocean surface layer thermal response[J].Journal of Physical Oceanography,2004, 34(3): 693-698.

1.1 儀器 LC-20A HPLC(日本島津公司)；AB SCTEX QTRAP 5500 串聯(lián)質譜(美國AB公司)；T25 D高速勻漿機(德國IKA公司)；MS3 D S25渦旋混合儀(德國IKA公司)；BSA224S - CW 萬分之一電子天平(德國 Sartorius公司)；LDZS - 2 高速離心機(北京京立離心機公司)；0.22 μm 有機系微孔濾膜(北京科瑞?？茖W儀器有限公司)。

[8] SHAY L. K., UHLHORN E. W. Loop current response to hurricanes Isidore and Lili[J].Monthly Weather Research,2008, 136(9): 3248-3274.

[9] BIGNAMI F., HOPKINS T. S. Salt and heat trends in the shelf waters of the southern Middle-Atlantic Bight[J].Continental Shelf Research,2003, 23(6): 647-667.

[10] DRENNAN W. M., ZHANG J. A. and FRENCH J. R. et al. Turbulent fluxes in the hurricane boundary layer. Part II: Latent heat flux[J].Journal of the Atmospheric Sciences,2007, 64: 1103-1115.

[11] SPALL M. A. On the thermohaline circulation in flat bottom marginal seas[J].Journal of Marine Research,2003, 61(1): 1-25.

治療前，2組患者的FMA-UE、ARAT、FTHUE-HK、MAL-AOU、 MAL-QOM、MBI評分差異無統(tǒng)計學意義。治療后，2組患者FMA-UE、ARAT、FTHUE-HK、MAL-AOU、MAL-QOM、MBI評分與治療前比均有顯著提高(均P<0.05)，治療后2組間比較，gmCIMT組患者的FMA-UE、ARAT、FTHUE-HK、MAL-AOU及MAL-QOM評分的改變比常規(guī)組更加顯著(均P<0.05)，MBI評分2組間差異無統(tǒng)計學意義。見表1，2。

[12] JIANG Xiao-ping, ZHONG Zhong and JIANG Jing. Upper ocean response of the South China Sea to Typhoon Krovanh (2003)[J].Dynamics of Atmospheres and Oceans,2009, 47(1-3): 165-175.

[13] NIE Yu-hua, ZHAN Jie-min and CHEN Zhi-wu. Simulation and influence factor analysis of circulation and thermal structure of the surface layer of the South China Sea[J].Acta Scientiarum Naturalium Universitatis Sun Yat Sen,2011, 50(2): 134-138(in Chinese).

[14] LIU Jiu-liang, CAI Shu-qun and WANG Sheng-an. Current and mixing on the northern South China Sea[J].Chinese Journal of Oceanology and Limnology,2010, 28(5): 974-980.

[15] SUN Lu, ZHENG Quan-an and WANG Dong-xiao et al. A case study of near-inertial oscillation in the South China Sea using mooring observations and satellite altimeter data[J].Journal of Physical Oceanography,2011, 67(6): 677-687.

[16] ZHANG Shan-wu, FAN Zhi-song. Effects of high-order nonlinearity and rotation on the fission of internal solitary waves in the South China Sea[J].Journal of Hydrodynamics,2013, 25(2): 226-235.

10.1016/S1001-6058(14)60052-4

* Project supported by the National Nature Science Foundation of China (Grant No. 40906006), the Fundamental Research Funds for the Central Universities.

Biography: CHEN Yi-zhan (1984-), Male, Ph. D. Candidate

LUO Ying-ying,

E-mail: luoyy2@mail.sysu.edu.cn