AS KUROSHIO TURNS: (II) THE OCEANIC FRONT NORTH OF TAIWANt

Transcription

1 ACTA OCEANOGRAPHICA T AIWANICA NO. 18, PP , 9 FIGS., SEPTEMBER, 1987 AS KUROSHIO TURNS: (II) THE OCEANIC FRONT NORTH OF TAIWANt CHO-TENG LIU 2 and SU-CHENG P AI 2 ABSTRACT As the Kuroshio turns northeastward towards Japan from Taiwan, part of it spins off to the north of Taiwan and up to the continental shelf of East China Sea. Since the Kuroshio is a current with high salinity and high temperature whereas the shelf water is relatively cold and fresh, an oceanic front is formed at the' East China Sea. Satellite thermal images were used to visualize the pattern, the scales, and the locations of oceanic thermal fonts, and to assist cruise planning In situ observations were collected to verify the existence of the front and to find the vertical structure of the cross-front distributsn of both chemical and physical properties. In the spring of 1987, the observed oceanic front north of Taiwan appears to be the result of topographically forced upwelling of Kuroshio water from about 100 m depth. INTRODUCTION The Kuroshio, ongmating east of the Philippines, flows past the east coast of Taiwan, then follows the East China Sea continental slope and the Pacific coast of Japan, and terminates in the sub-polar region of the North Pacific ocean. While turning from Taiwan to Japan, part of the Kuroshio water will cross the shelf break into the East China Sea (Liu, 1984). Since the Kuroshio is a warm and salty current with the surface layer nearly nutrient-depleted, a conspicuous band of property gradient will be formed at the confluence of Kuroshio and shelf waters which differ from the open ocean water in every aspect. This band may be named as a front if the change of water property across the band is much larger than the local variations. The surface thermal gradient of this front is easily detectable in the winter time, but not necessarily in the summer time when the cold coastal water is held to the north by the warm current passing the Taiwan Strait, and the surface layer is almost uniformly heated by the solar radiation. Surface fronts of the salinity, nutrients, and dissolved oxygen have much less seasonal variability because they do not have strong seasonal dependent surface sources/sinks, and because their intensities are mostly determined by the local water properties (like surface nutrient concentration) which has a lot smaller intra-annual variability than the temperature has. If there are oceanic fronts close to Taiwan, then why we could not study these fronts using historical data? The reason is that the spatial scale across this front is much smaller than that along the front, either the wide station spacing (e. g. 50 km) adopted by earlier hydrographic surveys, or the large spatial scale averaging, or the long-term averaging used in processing hydrographic data, prohibits one from finding, or even noticing the existence of fronts. Some examples can be found in the earlier publications of Institute of Oceanography. Lien and Chen (1977) showed the analyzed historical ( ) hydrographic data around Taiwan. From the winter time distributions of temperature (T) and salinity (S), one may conclude that most of the property changes occurred in 1. Contribution No. 200, Institute of Oceanography, ational Taiwan University. 2. Institute of Oceanography, National Taiwan University, Taipei, Taiwan, R. O. C. 49

2 50 Cho-Teng Liu and Su-Cheng Pai regions either north or west of Taiwan where Kuroshio meets the shelf water. Fan (1985) showed the T and S distributions in the Taiwan Strait with finer station spacing, and therefore sharper T and S gradient with time-dependent locations. In either case, one can not determine frontal features from sparsely distributed ship data. In the satellite images that will be shown in this paper, thermal fronts may be found all year around, either to the north or to the southwest of Taiwan. A preliminary cruise for studying fronts, especially the northern front, resulted from Kuroshio edge exchange processes (KEEP) was named KEEP-O. It was carried out during March 29-1, This paper presents results derived from hydrographic data of KEEP-O. SATELLITE IMAGES The thermal radiation from the sea surface can be measured remotely by an infrared radiometer either on board a ship, an aircraft, or the meteorological satellites, like the polar orbiting NOAA series satellites and the GOES series geosynchronous satellites. In the 8-12 /lm band of wavelength, the atmosphere behaves nearly transparent to the thermal radiation from the sea surface. The AVHRR (A.dvanced Very High Resolution Radiometer) flown on the NOAA series satellites have one or two channels operated in this band. If there is only one thermal channel (channel 4 of AVHRR) available for measuring the thermal radiation from the earth surface, the sea surface temperature (SST) may be derived with 1 to 4 C absolute accuracy, which depends on the sensor calibration and the total moisture content in an atmospheric column. The relative accuracy Fig. I. Satellite image of sea surface temperature for January 7, It is enhanced with light color representing the lower temperature and dark color representing the higher temperature. The center of the cloud is always white because the cloud top temperature is usually much lower than the lowest sea surface temperature in this latitute. (reprinted from Weaks, 1986)

3 As Kuroshio turns: (II) The Oceanic Front North of Taiwan 51 in a small area is better than O.5 C. This is to say that, in a region which is free of clouds and atmospheric fronts, the thermal images derived from AVHRR data can reveal thermal fronts of temperature difference no more than O. SOc. In cases where two or three thermal channels are available for deriving the SST, then the absolute accuracy is better than O.6 C (McClain, 1985; Uu and Liu, 1986) with temperature precision being O.I C or better (Lauritson, 1979). Fig. 1 to Fig. 4 show the satellite thermal images of East China Sea and Taiwan Strait. The time of the images spans from July 1983 to March 1986, and they cover three seasons from winter to summer. Fig. 1 is the NOAA-9 thermal image for January 7, 1986 (Weaks, 1986). It is derived from the HRPT (High Resolution Picture Transmission) mode of AVHRR data. The ground resolution varys from 1.1 km along the subsatellite track, to 4 km near the edges of each scanline. In this image, the mainland China is on the left, Taiwan is at the southwest portion, Shanghai and Yangtze River are marked with Sand Y respectively. The intensity of shading increases with the earth surface temperature. The lighter portion to the northeast (indicated with double arrows) represents a cold eddy formed by the cold water from the Yellow Sea, Along the coast of Ghina is the China Coastal Current Fig. 2. Satellite thermal image of January 13, This channel 4 (band of wavelength: pm) dat~ from the HRPT (High Resolution Picture Transmission) mode of AVHRR are enhanced the same way as Fig. 1. (Courtesy or: Chung-Jen Shyu) I

4 52 Cho- Teng Liu and Su-Cheng Pai (CCC), which is a southward coastal flowing current carrying the cold and fresh coastal water. Offshore from the Ming River mouth (near (26 N, E)), a small branch of CCC splits off towards the northwestern coast of Taiwan (Wu, 1984). The CCC and its branch forms an axil which takes up the warm water from the Taiwan Strait branch of the Kuroshio. To the north of Taiwan, there is a thermal front which indicates sharp turning or strong meandering motion of a branch of the Kuroshio water. Further north is a warm eddy that forms a sharpest thermal gradient in the image. During the winter monsoon from November to March, the northeasterly brings cold air mass from the north and extracts large amount of heat from the East China Sea. From Hsueh and Tinsman (1986), the southerly component of the buoy wind at (28 20'N, 126 5'E) during January 20-March 6 of 1986, is estimated to be about 5 m/s. Since this warm eddy and the sharp thermal front was observed on January 3 and on March 6 (Fig. 3) of 1986, fast and continuous supply of warm water was required to maintain this thermal front and this warm eddy. Therefore, we conclude that a large amount of Kuroshio water was spun off into the East China Sea as the Kuroshio was making turns to the northeast of Taiwan. Fig. 2 is the NOAA-8 thermal image for January 13, It is the channel 4 data of HRPT mode of AVHRR data. Like in Fig. 1, there are fronts that separate the warm oceanic water and the cold shelf/coastal water. But the oceanic fronts north of Taiwan are formed by a band of cold water that extended northward from north of Taiwan to the CCC, and the sharpest thermal gradient is at the continental edge of the Kuroshio, rather than at the offshore edge of the CCC as in Fig. 1. Fig. 3 is the NOAA-9 thermal image for March 6, It is the channel 4 data of the APT (Automatic Picture Transmission) mode of A VHRR data. The grayness of the satellite image increases with the earth surface temperature. Eastward and southward from Taiwan, the image is contaminated by the clouds (lighter shades). The white portion to the north is the cold water mass from Yellow Sea. Compared to Fig. 1 of January 7 image, the Taiwan Strait branch of the warm Kuroshio water still forms sharp thermal front with the CCC in early March but is restricted further southward Fig. 3. Thermal image of March 6, It is derived from the APT (Automatic Picture Transmission) mode of AVHRR data with the ground resolution degraded to 4 m. The temperature decreases with lighter shades.

5 As Kuroshio turns: (II) The Oceanic Front North 0/ Taiwan 53 :;; Fig. 4. Like Fig. 2 for July 29, 1983 Only the upper portion of the photo is cloud free region which is applicable for deriving the SST. Each level of grayness represents about O.37 C change in SST. by the CCC water. Since March 6 is nearly at the end of winter monsoon of 1986, this thermal image may represent the beginning state for the warm Kuroshio water to steadily flow through the strait in the late winter (Chuang, 1986). Fig. 4 is the NOAA-7 thermal image for July 29, It is the channel 4 data of HRPT mode of AVHRR data like Fig. 2. Here every change in grayness represents about 0.37 C change in SST. The oceanic front that was found in the winter (Fig. 1 & 2) and early spring (Fig. 3) to the north of Taiwan, still can be found in the summer, only the cross front temperature change is reduced from about SOC to about 1 C. The spatial scale for the interleafing of water masses across the front is larger than those in Fig. 1 & 2. METHODS OF IN SITU OBSERVATION In order to truely reveal the sharp-gradient structure of the front north of Taiwan, we have carried out continuous measurements of the sea surface temperature, accompanied with a section of CTD casts and water samples for analyses of dissolved oxygen, phosphate and silicate. The spacing of the stations was fixed at 18 km as shown in Fig. 5. To observe the oceanic chemical front, fast speed analysis with high analytical precision are required for verifying the horizontal chemical gradient. On this cruise, a working group of four was arranged to carry out the on-board manual analyses of dissolved reactive phosphate, silicate, and oxygen. An automated analyzer was installed to monitor the surface silicate concentration, as well as to aid the choice of sampling location. The samples were collected by the Niskin bottles (General Oceanic Inc., USA) at various depths and were transferred to 250 ml DO and PE bottles immediately after their surfacing. The analysis was performed within one hour after sampling. Dissolved oxygen was measured by the traditional Winkler titration method, while phosphate and silicate were determined calorimetrically following the procedures suggested by Grasshoff et al. (1983). These methods were calibrated on shore before the cruise with relative precisions better than 5%, 1% and 2% for oxygen, phosphate and silicate respectively.

6 54 Cho- Teng Liu and Su-Cheng Pai 250 i--~'---~------~'-~~~ _... EUJ 0:: ::J +!«0:: UJ Cl. :;: 20-. _ UJ j- U Cl. o «ci j u _ /,,/~~~ i,!! I ~ I '"'....,'// ",); PRT AOCP -- CTO looo _ :,J -;i 150 (;Jl < oc:; l> Cl m 3 S,5L,0~~-'-~~~~~~--: ----':-~~~~--~---':-~~~-: :---'100 STATION Fig. 5. The CTD stations and the spatial changes of SSTs measured in various ways. PRT-5 is an infrared radiometer operating in the band 8-14,um, it measures the thermal radiation of sea surface from the deck of a ship. ADCP represents the 2-min average of water temperature at 3.5 m depth. CTD represents the water temperature measured by the NBIS CTD/IR at 2 m depth. NO. In Fig. 5, the curve labeled CTD is the water temperature at 2 m depth measured by the NBIS CTD/IR which has I Hz sampling rate. The curve labeled ADCP is the water temperature measured by the thermometer of ADCP (Acoustic Doppler Current Profiler) at 3.5 m depth with 2-min averaging. The discrepency between these two curves is partly due to the difference in their sampling times. As for the curve labeled PRT-5, it is the voltage reading given by the ship-borne infrared radiometer PRT-5. This voltage increases with the thermal radiation received by PRT-5; and it increases with the temperature of the top 1 mm sea surface layer. KEEP-O DATA From station 3 to 5, we observed a 6 C-change of SST in 36 km distance (Fig. 5). We also observed that the SST did not decrease further morthwestward to station 10, i. e. towards the cold CCc. This interleafing of cold and warm water masses is somewhat like those in the satellite thermal images in Fig To get a closer look of the thermal front, we have included the time series of the ADCP SST and PRT-5 voltage data in Fig. 6. The ADCP data shows that most of the SST changes occurred in 5-10 min of cruising time. For a 12-knot ship speed, this amounts to 3-4 C SST change in a 4 km band between st. 3 and 4, and another 2 kn band between st. 4 and 5. Similar 'SST fronts were also recorded in the PRT-5 data. Because of the'sensor drift which started when

7 As Kuroshio turns: (II) The Oceanic Front North 0/ Taiwan 55 S' 3 DISTANCE (km) ", 50 i-"-l", ADCP PRT-5 I -;< (».. <: o c:; Gl :;:OOm 3' <: 10 TIME (MAR 30) Fig. 6. Time series of raw data from the PRT-5 voltage output and the ADCP measured surface layer temperature at 3.5 m depth. the ship is arriving a CTD station, we consider that only those PRT-5 data 10 min away from any CTD station are reliable. Since the PRT-5 measures the thermal radiation from the top 1 mm of the sea surface only, PRT-5 data are much more susceptible of the weather modification (e. g. solar heating and changing wind) and the surface condition (e. g. wave breaking) than the ADCP data which measures SST at 3.5 m depth. In addition to this intrinsic difference between the bucket SST (by ADCP) and the skin temperature (by PRT-5), the PRT-5 data show more wiggles because the PkT-5 radiometer has a wide band frequency response which may follow the fluctuating skin temperature closely. So far we have only reviewed the surface temperature, nothing is said about the frontal structures beneath the surface layer. For example, how deep is the thermal front? Does the vertical structure of the front reveal the sources of water masses? Fig. 7a and 7b show the sections of temperature T and salinity S from CTD stations 1 to 10. The thermal front between s1. 3 and 4 is also a transition zone of salinity distribution, that separates the salinity-stratified Kuroshio water at s1. 1 to 3 from the water at s1. 4 to 7 which has little change of salinity over the 100 m deep water column. Between s1. 7 and 8, there is a salinity front which is sharp and deep. If we may classify the water masses according to their distributions of T and S, then we will divide the water of this section into three types: the warm and salty Kuroshio water to the east (s1. 1 to 3), the coldest water at the center (st. 4 to 7), and the least salty water to the west (s1. 8 to 10). We believe the last two types of water are the upwelled Kuroshio water and the Taiwan Strait water, and we shall prove it in the following. The water to the east is considered to be Kuroshio water because it has the same vertical distribution of T and S as the Kuroshio water has (Liu, 1983, Fig. 3). Liu also showed that the mean depth of salinity maximum layer is about 80 m to the west of Kuroshio and is about 200 m to the east of Kuroshio. For the layer between S-max and 150 m depth near the shelf break, named S-layer (Fig 7a), T varys from 18 to 16"C while S varys from 34.7 to These characteristics fit well with the cold water between st. 4 and s1. 7. Can we determine whether this cold water is (1) a surface water that is the natural extention of S-layer? or (2) due to topographically driven upwelling of S-layer? or (3) the cold water from Taiwan Strait? This cold water can not come from the north nor from the south, which

8 - 56 Cho- Teng Liu "and Su-Cheng Pai TEMP'::RRTURE 2: 10e. "- j ~ J ț (a) SriLINlTY :e ~/--,-,T'-';''''1'''--;/-+-'\-~~--''-\-~--&'-'~>-3,-.6~5--/-~-<'34--,.t~~ '3,"'3'11/ J' \... / 346 l- :3-t4 I -II I \ t... _v/ ----J f i / /1 l-://ff.,1/ / \ <34.55 }.,-~ I I I '"... / t 1, + 13~.4 // ///,/, 34.5 : \ \,' /... ///./Il't ~_ 50 /!.../34.4 I I \ 1 /' '...:_... t:;:,e j, v / / I : " j I 0! I I I \ J /--\ \ 1 '... _.// \ ~ E. \ 1102 I- \ -! c... \. 1 w \1 : 50, 1150' I 20~ j~--+---i-,,---f----i---'--f- 4--_-+-.;-L_~L00 (b) Fig. 7~ Section plots of (a) water temperature T, and (b) salinity S. Stations were occupied at 18 kmspacing. K, C and Ware three water type, and S-layer is the source of upwelling water (see text for detail). is evident in Fig. 8 where l8'c SST contour should be closed to all direction except to the west where the data gap prohibits SST contouring. Chemical tracers of the water may provide us another type of clues for classifying the water types. If the cold water between st. 4 and st. 7 is a natural extension of S layer as in hypothesis (l), then this water is of surface water type and it has been exposed to the sunlight for a long time. As a. general rule, the phytoplanktons in the surface layer will quickly utilize, and finally deplete the nutrients in the euphotic layer, unless the mixed layer depth is much deeper than the eutrophic layer, like the situation in polar ocean before the spring bloom. Fig. 9a and 9b show that this cold water is rich in both silicates and phosphates. This is contrary to the characteristics of surface water and therefore the hypothesis (l) is rejected. Hypothesis (3) assumes that the cold water came from the west. Because the cold water has the lowest temperature among the three, it must contain some of the CCC t e

9 As Kuroshio turns: (II) The Oceanic Front North of Taiwan 57 [ I' I I I,, _ 60 -_, \ I, \ , / / / / /, / \ TAIWAN, 1, 20 '", I \ 200. Fig. 8. Contour of sea surface temperature at 3.5 m depth. The oceanic front was at s and the cold water at S came from upwelling of Kuroshio water at 100 m depth. Depth (in fathoms) contours are the dashed lines with small size numbers. water if this hypothesis holds{bu-t there are three arguments that are contradictory to this hypothesis. Firstly, even in April when the winter monsoon is tapering off, the strait water is still either weakly stratified or well mixed (Fan, 1987) and it is of surface water type with nutrients nearly depleted. A good example may be found from CTD data and nutrient data of st. 10. Secondly, this wa~er's salinity is about 34.6 which is much closer to Kuroshio water's 34.7 than CCC water's 31 or less (Wu, 1984), whereas its temperature is closer to that of the Taiwan Strait water which is in between those of the CCC water and the Kuroshio water. Thirdly, because the cold water has high salinity value, it can not be a mixture of the CCC water, Taiwan Strait water and Kuroshio water. Therefore, hypothesis (3) is rejected. We consider this cold water band being the result of topographically forcyd upwelling of the S-layer, based on the following arguments (a) both of its T and S match those of the S-layer; (b) it is a nutrient (silicate Si and phosphate P in Fig. 9) rich water in the euphotic zone; (c) its surface concentration of dissolved oxygen is of deep water type, unlike the surface waters on each side of it; (d) all the isolines of T, S, Si, P, and DO (Fig. 7 and 9) hay~ the symptoms of upwelling; (e).it located right over a "b<\rrier" to the Kuroshio, like' the case of shelf ~reak upwelling. AHj1o,ui?h ther.~ was no counter evidence found against thfs hypothesis,' we still may nlisequestions on this upwelled

10 58 Cho- Teng Liu and Su-Cheng Pai SILICRTE [i-ig ",t. 11) (a) RER=T I V E FHCJ S~ H~ i E ()J gat. /1 J (b) DISSOLVED OXYGE~ (mill) (c) 150 Fig. 9. Section plots of chemical tracers: (a) silicate (Si in p.g-at.1g), (b) phosphate (P in p.g-at.1 ) and (c) dissolved oxygen (DO in milg).

11 As Kuroshio t-urns: (II) The Oceanic Front North 0/ Taiwan S9 water: (a) what is the. source of its low salinity core? (b) where is its destination? (c) what are the implications to the marine ecosystem? and (d) what are the dynamical balances that maintain the upwelling centered at st. 6? Question (a) on the low salinity core causes little difficulty because this core is too weak to reject the hypothesis. It takes about 70 mm of fresh water input for the water column to generate such a low salinity core of 90 m deep and 0.1 maximal dip in salinity. In this region, there are frequent passages of atmospheric fronts, accompanied with strong wind and heavy shower. According to weather data of March (one week before the CTD section), there was up to 15 mls wind with total 124 mm rain fall at Peng-chia Yu ( 'N, 'E), and up to 10 mls wind with total mm rain fall at Keelung (25 8'N, 'E). The post weather analysis (Ho Lin, personal communication) shows that even when there was no rain recorded on March 29, the chance of rain is still there. Question (b) may be answered implicitly by the satellite thermal image in Fig. 2 where a band of cold water extends from the north of Taiwan to the CCC further north. The front between st. 3 and 5 corresponds to the thermal front in Fig. 2, while its cold water band implys the northward flow of the upwelled water. The answer for question (c) is typical to all upwelling regions. In the upstream region (e. g. the Equator for the equatorial upwelling) or at the beginning stage of an upwelling, the water is rich in nutrient, but the phytoplankton growth, the zooplankton growth, and the aboundant fish catch all happen either further downstream or at a later stage. The local fishing ground and the winter fishing ground near the northern end of the cold water band in Fig. 2 may have some contribution from the upwelled water. There are three mechanisms that may contribute to the topographically driven upwelling near st. 6. The bottom friction will result cross-stream pressure gradient flow which creeps up along the shelf bottom, the inertia of a current may force itself to climb onto a slope or even over a bathymetric barrier, and the turning of a geostrophic current requires centripedal force which may be provided by the pressure gradient force resulted from the local adjustments of isopycnals. As for the relative importance of these and other possible controlling mechanisms, further field works and studies are required. DISCUSSION The topography north of Taiwan is far from smooth and the currents in this region are relatively strong and mostly tidal. Liang (1981) has observed tidal current of amplitude up to 4 knots at two closely spaced YAF current meter mooring station A (25 25'22"N, '21"E) and station B (2Y25'47"N, '27"E). As indicated in the time series of the measured current velocity, the water velocity at these two stations often differ in amplitude by 20%. This kind of fine scale phenomenon is also found in the satellite thermal image where interleaving of water masses and shingles spinning off from the Kuroshio (Fig. 1-4) are often found. Both physical and chemical data of cruise KEEP-O demonstrate the complexity of the front, or more properly the complexity of the confluence of Kuroshio and East China Sea water masses. Beyond these small scale irregularities, we may summarize the major phenomena revealed in these satellite thermal images: (1) there is a permanent oceanic thermal front which extends from the north of Taiwan to about halfway to the Yangtze River mouth. The temperature difference across the front varys with the season, and it is about 6 C in the winter as compared to about 1 C in the summer, (2) the China Coastal

12 60 Cho-Teng Liu and Su-Cheng Pai Current is a current of cold water and it generates a thermal front on its oceanic boundary; (3) there is a branch of CCC that spills across the Taiwan Strait, and in the winter time it may even block the branch of Kuroshio water from passing the Taiwan Strait. Since fronts were indicated in all satellite images of years 1983 and 1986, we expected to locate an oceanic thermal front during the KEEP-O cruise in March of The surface observations from the cruise not only verified the existence 'a'nd the intensity of the thermal front north of Taiwan, they also revealed that: (1) to the west of this year's front, there, was a band of cold water which appears to be the Kuroshio water upwelled from about 100 m depth, (2) this band of cold water is rich in nutrients but low in dissolved oxygens, and henceforth (3) there is a nutrient front on each side of this cold water band. The CCC carrys cold and low salinity water down the coast of China, forms interleaving fronts with the Kuroshio water over the shelf, and spills a brach of its water across the northern Taiwan strait. The Kuroshio water enters the strait in two ways: the one entering the strait from the south forms thermal front while mixing with both the main stream and the branch of the CCC, and disappears into the axil formed by them; the other approaching from the north merges with the CCC over the East China Sea and forms the oceanic front north of Taiwan. For the deep currents in the Taiwan Strait, it is likely to be different from the surface currents, otherwise it will be difficult to explain the destination of the water after the mixing process. Direct measurement of the deep currents may be the best method to clarify this point. ACKNOWLEDGEMENTS We are grateful to Ms Ray-Jen Liu who processed the physical oceanographical data and prepared the satellite images and draftings, and to Tsai-Chu Chen and Tien-Hsi Fang who did the on-board analyses of dissolved oxygen concentration. We are also grateful to Captain Wang, the crew, and the marine technicians of R/V Ocean Researcher I, for their whole hearted support of the cruise KEEP-O. This research is sponsored by the National Science Council through grants NSC M002a-05 and NSC M002a-07. REFERENCES CITED CHUANG, W. S. (1985) Dynamics of subtidal flow in the Taiwan Strait. J. Oceanogr. Soc. Japan 41: (1986) A note on the driving mechanisms of current in the Taiwan Strait. J. of Oceanogr. Soc. of Japan 42(5): FAN, K. L. (1985) STD measurements in the seas around Taiwan during Special Pub. No. 44, lnst. of Oceanogr., National Taiwan Univ. 337 pp. (1987) CTD measurements in the seas around Taiwan during Special Pub. No. 51, lnst. of Oceanogr., ational Toiwan Univ., 183 pp. GRASSHOFF, K., M. EHRHARDT and K. KREMLING (ed.) (1983) Methods of seawater analysis Verlag Chemie GmbH, Weinheim, p & HSUEH, Y. and J. H. TINSMAN, III (1986) A comparison between geostrophic and observed winds at the Japan Meteorological Asency buoy No.4 in the East China Sea. Submitted to the J. of Geophys. Res. LAURITSON, L. (1979) Data extraction and calibration of TIROS- / OAA radiometers. NOAA Tech. Memo. NESS-I07, 58 pp. LIANG, N. K. (1981) Marine environmental investigations in the zone YAF. Special Pub. No. 34, lnst. of Oceanogr., National Taiwan Univ., 219 pp.

13 As Kuroshio turns: (II) The Oceanic Front North of Taiwan 61 LIEN, S. L. and W. C. CHEN (1977) Oceanographic data of the sea sea surrounding Taiwan. Special Pub. No. 15, Inst. of Oceanog r., National Taiwan Univ. LIU, C. T. (1983) As the Kuroshio turns: (I) characteristics of the current. Acta Oceanographica Taiwanica 14: LIU, R. J. and C. T. LIU (1986) The comparison between satellite derived and in situ measured sea surface temperature. Meteorological Bulletin, 32(1) : (in Chinese). MCCLAIN, E. P. (1985) Comparative performance AVHRR-based multichannel sea surface temperatures. Ame r. Geophys. Union Trans. EOS, 66(18): 284. WEAKS, M. L. (1986) SST structure in the East China Sea. NOAA jnwsjnesdis Oceanogr. Monthly Sum mary VI (I): 3. WU, Boyu (1984) Some problems on circulation stu dy in the Taiwan Straits. Collected Oceanic Works, China Ocean Press 7(1) : 當黑潮轉向時 ( 11) 臺灣北部的海洋鋒面 劉伴騰 臼書禎 摘 要 當黑潮流過臺灣轉向日本方向時, 部卦的黑潮水轉向基灣北部海域流入東海大陸棚, 在高溫高鹽的黑潮水與低溫低鹽的東海海 7Jz i 區流處, 產生海洋鋒面 衛星熱幅射影像可顯示此鋒面的形狀 大小與位置, 有助於計劃海上探視 l 工作 民國 7 6 年春季的海上質詢 [IJ 資料額示, 該鋒面兩側的物理與化學性質分布截然不同 ; 經由水圈特性分析得知該鋒面是由 100 公尺深處的黑潮海水受地形影響而上昇至表層, 與東海海 7Jz 對峙所形成的