1I,. ,

h..

ICES 1990 -PAPER C.M. 1990 / C:15 Sess. P

•

Remotely sensed water mass distributions and large scale seasurface circulations in the Skagerrak.

by

Bertil HäkanssonSwedish Meteoroligical and Hydrological InstituteS-601 76 Norrköping, Sweden

ABSTRACf

The sea sutface temperature remotely sensed from the weather satellites in the NOAA-series is

used to identif~ water masses in the Skagerrak area. Data from May 17 and 18, 1989 are

compared with almost simultaneous in situ data of hydrography, showing good correspondance

between sea surface salinity and temperature distributions. The case demonstrates that the sea

sutface temperature can be used as a tracer for the water masses in the area, determined

otherwise wi~ only in situ salinity or a combination of in situ salinity and temperature:' The

remotely sensed sea sutface temperature have a geometrie and radiometrie resolution of 1 km

and 0.12 K, respectively. The satellite data shows the synoptie view of the upper layer of the

oeean to a much beuer extent than any ship data, allowing detailed analysis of even relatively

small seale events such as eddies in the Skagerrak. Coneeeutive images with about 12 hours

interval are used to calculate sea sutfaee drift velocities with a method tested by Svejkovsky

(1988).

Validated sea sutface temperatures from May 17, 1989 indicates that an anticyclonie eddie takes

place in the NO Skagerrak. This may be caused by the large seaIe eyclonie circulation in the

main part of the Skagerrak interaeting with the boundaries of the ~ea. This circulation is

probably driven by westerly and southwesterly winds. In this case the Baltie current is blocked

at the Swedish coast, forcing the current to move in a northwesterly direction. This circulation

has not been observed ealier with traditional methods of current measurements.

0"" ...

._-- -----------_._-----_._-----

•

•

Remotely sensed rind in silu observations of water mass distribution ariddynamics in the Skagerrak.

1. IritroduCtion

SkageITak is a traIlsitiona1 area betWeen the Baltle imd the North Sea. it reeeives in the stiiface

layer low salinity water froIll the Kattegat arid the Baltie Sea with the Raltie ctirrent arid high

salinity water from the North Sea with the Jutland and Dooley curreri~s. In the lower layer high

salinity water is also entering along the southem boundary with inflows from the North

Atlaritie. These eurrents inake up the large seale eyclonie circuhitiori often found in theSkagerrak (Svansson, 1975). The Dooley eurrent is less known and is only occassionally

diseussed in the litterature. The Jutland and the Baltie eurrent are variable in magnitude, theformer appears to be winddriven and the latter density driven.

• Outflows iri the sunace iayer geriei-aIiy takes phice along the NOrWegian coast with the

Norwegiari Coastal Ctirrent (Saetie and Mork, 1980) , which occasionally can be blocked by

westerly wind events (Mork. 1981). There is also an outnow iri the lower layer along the

Norwegian eoast. Detiiled current measurements along the triinsect between Norway arid

Derimark investigated by Rodhe (1987). imllcate that the eyclonie gyre is most pronoUnced irithe lower layer with maximum mean flows along the bonom. lrithe'sllrface layer the cYclonie .

circuiation of the mean flow is weak with the largest variability in eurrent strenghi arid direetion

found alorig the rigid boundaries of Skagerrak.

The surfaee curreriis entering the Skagerrak originates in areas where the anthropogenie

influence is considernble. The transport of this material to the SkageITak may be of irnponance, ,

.to the eeosystem in the area. where exceptional blooms ofphytoplankton arid oxygen defieit

bottom layers in some 'coastal sites has 1?6en observed iri receni years. It is also put forward thai

the geographie variation of the stratification contribute to the patchiness in phytoplankton

distributiori (Richardsson. 1989). Hence. an increased understaridirig of the hyetrographie

processes in SkageiTäk. and especially in the surfaee layer is necessary. \Vith this in rrimd the

airri of the preserit investigation is to g~ve a qualitative deseription of a near synoptic event ofcircuiation arid hydrography in the upper part of the Skageri-ak:. Deiailed horizontal distributionsof the wäter masses arid drift velocities Me detennined by rerrioteiy sensed images of the seasurfaee ternpecirii!e arid surve)ririg sWp measriremerits of the hydrographY. The results fromthese different rrieasUrement techruques are qu3.litativeiy integratoo iri order to increase the

accuraey of the interPretation. For exarnple. in the northeastem part of the Skagerrak an

anrlcycioriic eddie was fOUnd, whieh had a pronoimced irifluence on the coastal waterdistribution arid on the patbway of the Baltic current. Details which eouid riot be obtained from

. the ship transect meäsmeinents alone.

1

,,':

•

----------------------------

•

2

Remote sensirig have previously been used in some studies of water masses and withphytopiankton ciistributions. Pingree et. al. (1982) used infra-red satellite images imdhyclrography durlng a late summer situation (August 1981) to interpretate tbe planktonpatchiness. They fourid ~ relatively colder sUrface regiori in the middle of the Skagerrak,reflecrlng the area where the pycnocline was shaiIow. The relatively colder water was

circumferencoo byw~er water in which tbe currents were in geostrophic balance, muting the

doming of the pycnocline fourid in the cold spot. This rloIning was f6und to favour a rlch~, '- ,

biomass Withiri the pYcnocline. Remotely sensed sea sunace temperature as a trilcer for watermasses in th6 area was also used during the Chrysochromulina Polylepis blooin iri 1988. The

idea was that the bloom was advected with the wann and low salinity surface water of the Baiticcurient from the Kattegatt along the Swedish and Norwegian coasts into the North Sea. ntis

drift observoo from satellite images corresponded with in situ measureinenis of chlorophyllconient (Johannessen et. al., 1989). These stuclles indicate that it is possible to use remoiely

serised sea surfac6 temperatiire as a tracer of surface water masses, which ein indirectly berelated to phytoplankton distribution when iri situ measuiments are avaiiable.

Water masses may also b6 irac:oo, using the optical conditions in the sunace layer obserired

with satellite images. Totill suspended matter ai1d chlorophyll content are retrievable arid well

understOOd in oeean water ( More! case one water), whereas in cOaStal seas the interpretation ofsatellite images for quantitative evahiation of these prirameters are less kriown. Here

nonbiological pamcles contribute to the total amount of suspended matter, which may create

ambiguities in the image analysis of chlorophyll content. Also the amo.:mt of yellow substancesis knOwrl to b6 high in the Skagerrak and especially in the low saiiriity parts of the watermasses. How this contlibute to the optical conditions and influencies the analysis tecimiques inremote sensing is not yet clear. There~ nevertheless studies made iri the area by Aarup et al(i987, 1989) using CZCS images. They.conclude thai so far it is possible to obtain seml­quantitative assessrrientS oe chlorophyll which is a valuable indicilior of dynarn1c processesfrom time senes of images. For more accurate estimates of optical proPerties there is an urgentneed for a whoie senes of bio-oprlcal measurements to establish a comprehensive set of

; ,. .. - ,. ""

calibration data. A dewcatoo airbOrne experiment for this purpose was conducted by Norwegianinstitutions dUiing the spring i989 in Skageri-ak (pettersson, 1989 and 1990). ether studies inthe field and coverlrig the SbgeiTalc area are reported by Sörensen et al (1989) ancl iiäkänssori(1989). These mainly focus on coastal areas afirl the Glomma river ouiflow. Herice, there is aneed for furtber investigatlons with satellite arid simultaneoüs in situ dara ofoprlcai properrles to

obtain calibrarlon data iri offshore mas. From the pOint of View ofoceanogrnphy, however, th6mentloned siuclles iridiciues that the iritegratlon of ship and satellite ciatii weIl benetits arid make

progress to th6 unclerstandirig of tbe physical arid biological oceanography in the Sk:lgerrak,

·e

•

although tbe chlorophyll content is not yet accUrately retreivabie.

in the foiiowing Sectlon of this paper"tbe results from the validation of the reirieverl sea surface

iemperattire is presented. The interpretation of remotely sensed arid in situ gathered

measurements is discussed in Section 3, whereas the condusions are suiriinarlzed in the final

Seetion.

2. Calibration of sea surface temperritures from AVHRR sensor

The methodology of retrieving sea sulface temperiltUre from sateIDteborne sensors in thethermal dectromagnetic band orten make use of multlspectril.1 techniques to eliminate the

contributions öf atmospheric water vapour (Robinson et.alI984). This so called split window

technique is used in the present study on three images which are compared with in sini

measurements of bulk sea surface temperatures in the Skagerrak. The expression userl forretrleVing this teniperature has the following form:

3

"

MCSST = T4 + 2.5 (T4 - T5) - 0.96 (2.1)

The second term which represent the difference between the brlghtriess temperaiure in the

AVHRR channel four and five are smoothed using a3*3 median rilter to supress the noi~e

generated by lhe algebraic calculatiori (Barton, 1989). In this way surface temperature

strucrures are tnaktaineci in the MCSST ('MUlti Chaririel Se~ Surface TemperatiIre) image.

The ship trarisects are schemaiically shown in Figure 2.1, the east-west transect en20mpass 12

hydrographie stations which were ~ove~ed within 18 hours on May 17, whereas the main pari

of the non-south transect was covered within 10 hOufs on May 18, 1989 (the three stations

dosest to the Deninark coast was sampled during May 19). DUring these surveys three AVHRR

(Advariced Very High Resolution Radiometer) images have been used for calibriition of the

algorithrl1 ofMCSST. In the beginning ofthe east-west transect an image from May 17,0226

GMT was obtained, whereas the other was from May 17, 1221 GMT when the major part of

the trimsect was endecl. in order to comparc' image and in situ data as similltaneous as possible

the values of MCSST were selected froni both images from M~y 17 and the image data fromMay 18 were seleded to be as dose as possible With lhe iri' sÜu data. The result from thecomparison is preserited iri Figure 2.2 from' which the correlation coefflcient was foimd to be

0.91, iridicating the validity of lh6 MCSST. .

The tempocil variation of bulk sea surface iemperilture at tbe cmistal station of Väderöarna have. '..' " . ' ,"., ~ -': ';, " . .been used for companson With the Satdlite data. The cOaStal station meastirem~nts of the ri.iriemperariire and the water terriperatuie at 5meters depth rife shöwn in Figure 2.3, indicatlng ä

~ _ , .,. _ , • . , 1

dose relatioriship bdween cUr arid water temperatures. Both variables demor1stnites a dailyI

. ' ' . ·.~·~·~",.f .' - .

•

< ••

•..

4

hearirig and coolirig cycle arid an almost linear increase iri tempernnrre on the time sc3J.e of a

week. The diily variation durlng May 17 was 1.3 K and the mean iriciease was aootit 0.7 Kbetween May i7 arid i8. The corresponding v3J.ues obtained from the horizolliaÜy averaged

MCSST coveririg an area of 87*87 km2 in the central Skagerralc were 0.2 and 0.5 K(thlstemperature difference was calculated using the night images from Maj 17 and 18,1989). ThecmlstiiI site is apparently more influenced by the daily variation of the air temperature thän theopen sea off the coast. The trend in the increase of the mean daily sea.sulface terriperature is,however, of the same magnitude.

3~ Resuits and discussion3.1 Hydrography a1uJ wind

The wind speed arid direction the period before 17 and 18 May, i989 (corresponding (6 theiulian days 137 and 138) was characierized hy SW winds with daily mean speeds of 6 to 9 iri/saccording to observations at the coasiaI sites Mäseskär and Koiier (see Figure 2.1). Thissituation was persistent during four days before th6 studied event During westerly winds

uppweUing along the Norwegian coast inay talce place and will be weU cleveloped on the tiine. .sc3J.e of a week. The hydrographic survey along the seetion across the Skagerrak indeed

indicate that uppwelling is occurring. The salinity distribution pres~nted iri. Figure 3.1.1 showatilting from ten metres depth to the sUlface and is clearly shown in the horizontal temperature

distribution of Figure 3.2.1.

In generiU the pycriocline, which iri Figure i 1.2 and 3.1.3 is iridicated by the halocllrie, isfound between 10 and 20 mettes clepth, ~hereas the halocline becomes thicker arid tilted alongthe cOaStai bouridaries iridicatirig goostrophic currents. Alollg the Danish coast the stllface

ctirrent <JtitIaricl current) moves easiVIard arid along the Swedish coast the naitic ciirieni isO

fouod abOve ten metres depth exteoding about 10 km offshore. The tippwelling aloog the

Norwegiari coast appears to block any suiface outflow along the coast Beneath the rimin

halocline, however, the isohaline tilting iridicates a cyclonic circulation in the whole Skagerrak.

32 Sea surjace iemperaiure and salimty,The~~,~blliion of the, ~ea ~u;race temperature shown in Fig. 3.2.1 is obtained by averngillg theMCSST unages from May 17~ This unage represent the meari. tempernnire 'distribution aridshould be comparnble with the less synoprlc shit> data. The map of sllifacc' saÜnity<5 metersd.epth) showri. in Fig. 3~2.2 is obtained from the ship data. The distribution of these tW6

I " ,,", "",;,' "', " ';'. '.' .." ! ..' >

variables shows sirmlar featUres, regarding the location of tWo frontal zones. One is the front of

tbe Baitic cirrrerii offshore the S~edish coast iri. wWch the magnitUde of the horizontal salinityarid temPerature gradieniS (S-l * fj.s/tix.) and (T-1 * fj.T/tu ) are approxim~lteIy ...12.8*10-3arid -

i2.S*lO-3 iari-i, respectively. The second is the front betweeri the inii~r änd'otiter Skagerciic

... . ..' ; ~ ..~. .'

....

sllIface water alorig the transect from Aiericiai to Hirtshals. In this frontthe saIinity gradient is

weaker on the Danish side of the trarisect than on the Norwegian. 1n6 iargest griidieoisofs:ilinitY arid temperature fouod in this region are +13.8*10-3 arid -4.8*10-3 kIll-I, respecrlveIy. The

appearance of this front iridicates that the inner Skagerrak sllIface water is blocked and cannot

leave the rirea along the Norwegian coast.

The scarcity of ship data in other areas of Skagerrak limits the analys~s to this basinwide scale,

whereas more detailed information about the distribution of surface water masses may beobtained from the remotely sensed temperature data shown in Fig 3.2.1 and 3.2.3. Note

especially that the Baltic current (temperature ~1O.5 oe) does oot extend along the Swedish

coast north of N580 40' as the surface salinity distribution appeaci to indicate, but is forcedoffshore. The river Glomma plym is exiencllng from the eastern to the western side (image iromMay 1'7, GMT 1221, 1989) of tbe mouth of the Oslofjord from where it makes a sharp },erid to

southeast. These two features indicate that an anticyclonic eddie is formed in the NO Skagerrak,

forcing the Baltic current offshore ancl the river plym to southeast FUrthermore a cold iunge of

surface water appears to circumvent awater mass chamcterized by higher teinperature arid

salinity <T ~ 10.5 oe and 28 ~ S~ 29 PSU) than the surrounding water mas~es in the iriner

part of Skagefrak. Close to the NOrWegiari coast in tbe NO SkagerTak this "cold;' water mass

separates in two brariches· one to the east probably forming the ariticyclonie eddie and one tothe southwest. Along the shallow areas ( ~ 50 m) offshore the Danish eoast a warm water mass .is located. This temperature distribution is fourid in the tlu-ee images from May 17 and 18,indieating a persistent hydrographie stnicture to which a tentative sllIface circulation has beeri

discussed. In the foUowing section this eirculatiori wiri be quatiiatively verifiecl to some extent .

33 LtiTge seale sea sUr/ace cireulationTo study frontS, water masses and velocities it is not riecessary to calibrnte the images to

absolute temperatures. Instead, the single channel brightness temperature with the relative

accuracy of the 0.12 K (NOAA-AVHRR) is used. The velocity estitriates are made by

idenrlfying arid foUoWirig temperature patterns in images with at least 6 hOUfS interVäl

(Svejkovsky, 1988). He found an accuracy of 0.06 inJs at veiocities lower than 0.4 iriJs,comparing surface drOgue data and estimated flow velocity from sateUite images with 12 to 24hours interval. This technique is bas&! on a manual identification of featUres arid may thus be

subjecrlvely influenced by the interpretator to some degree. Iri th6 present work the e~tiiruited

veiocities should be recognized as an iridication of a probable circulatiori.The estimated sllIface

flow velocitles from May 17 are shown in Fig 3.3.1. These are not evenly cllstrlbuted since it is

not possible to ruid features in all area.s withiIl the images.

Nevertbeiess. the circulaclon is in agreemerit \\'ith the water mass distribution in general. ItI

..

5

....

shows thai the cold water off the Swedish coast is moVing nolthwarci arid ai N580 40' turns

northwestward, where the Baitic current separates from the coast Also the anticyc10nic eddie in

the NE Skagerrak is indicatCd aithough the magnitude ofthe current is weak (S 10 cIn!s). Iri the

south of Skagerrak twO cuirents meet; one from the SW part of the Norwegian coast tiaced by

the cold water (see Fig 3.2.1), which may be driven by the Ekrriari drift arid one followirig the

Danish west cmist (Jutland current). The Jutland cirrrent is bro3rl and exteods over th~ shallow

areä from the coast to the 100 m isobath, congruous with the comparatively warm water InaSSfouod in this region. The Jutland current and the uppwelled water originating from the

Norwegian coast converge north of the Damsh coast

Th6 surface current offshore the Swedish coastline at the isobath of 100 m was measured

duririg May 1989. The vector plot is shown in Fig 3.3.2 and indicate that the velocity on May

• 17 was ciirected NNW with an amplitude of 0.5 rn/s at 15 tri depth, in agreement with theestirnated cirirt velocities from the images. The advection velocity of the uppwelIed water may

also be compared with the Ekman drirt, wWch most likely is the driving mechariism in this

case. The stancIard equation relating the Ekman transport, 11le. to the obserVoo wind at 10 m,

WlO, is:

iTIe = Ci (Pa i Pw) f-t (W1O)2 (3.3.1) ,

H~~e Ctt (". 2.5 10-3) is the dI:ag ce>efficien~ (Pa/pw) (". 1.3 10-3) is the density ratio betWeen air

and water, f (". 1.2 10-4) is the Coriolis pararrieter. With the mean windspeCd of 7 in/s the

Eicirian transport amounts to 0.85 m2/s, and the typical Ekrrian drift is of the order of 10 crit/s

'when the sufface layer depth is 10m. This is indicating that the velocity calculatlori from the

image cIata ofthe uppwelled water is consistent with what can be expected from EkriJ.an theory.

6

• ~owirig the surface cuirent and the striiti!ication it is possible to calculate the geostrophic

velocity iri the d~eper layers, takirig mto'account th6 c1assical fonnula of Marguel. This

expression relates the geostrophic veiocity difference, AV, to the tilting of isopycrials, Ah lAx.yielding:

AV:= g; f-t Lih I~ (3.3.2)

Heie g' is the reduced coristant of gravitY. g* (P2-PI' P:z). Iri a two layer stnitified basin the

. lower layer velocity (V2) can be estimatCd ir the sUrface layer velocity (vI> is kIlown, hence:

. V2 = VI - g' f-t ~h I Ai (3.3.3)

In the jutland current the upper layer veioeity waS 25 crn/s, the tilt cari be esrlrriated froIll Figure

3.2.3 to 1.3 10-3 arid g'". 2.7 iO-3.riils2• In thi~ case th6 lower iayer vetocit}T is -:5 ciO/s which

is irideed lower than the surface layer velocity and of opposite sign, demonstratirig tbät the .

cwrent is baroclinic. This simple analysis rnay also be appüed t6 the currentS offshore the Balticemerit, which hai a suifac~ velocitY of 50 errvs, a tilt ofO.1 10-3 and a rectuc'ea gdvitY of .5.3

..

i0·3 rn/S2. Hence, the lower layer velocity is 45 crn/s, indicating that the ciurent is aJrilost

barotropic.

4. ConclusionsThis investigation covers a near synoptic situation of the Skagerrak surface water mainly driven

by winds from SW. This is a typical wind direction occuring in the area and, hence, the given

synoptic case may be of some importance for the understanding of the general surface water

circulation. It is believed that the following scenario conc1udes the results. The southwesterly

wind is driving the upwelled water along the the Norwegian coast by the Ekman transport

offshore, blocking the surface water in the inner part of Skagerrak. The Jutland and the Baltic

current add surface water to the inner Skagerrak. A complex convergens of these currents and

. the cold surface water from the uppwelling area is taking place in the southeastern part of

Skagerrak. Strong northward currents are found offshore the Swedish coast, which splits in

two parts when reaching the Norwegian coast. One part turns east, supponing the anticyc10nic

gyre in the NE Skagerrak which is forcing the river Glomma plym to move southeastward and

the Baltic current to move nonhwestward and thereby separating from the coast. The second

part presumably turns southwest and thereby circUriuerences the warm and slightly more saline

inner Skagerrak surface water.

The blocking of surface water in the Skagerrak during westerly winds has been noted earlier,

but the antieyclonie gyre in the NE corner and the separation of the Baltic current are features

not previously investigated in detail. For example it is not known under what meteorological

and oceanographic conditions the gyre takes place. Hence, further studies are needed to

investigate the frequency of these features.

Acknowledgements

In this investigation the remote sensing data was obtained by the Swedish Meteorologieal and

Hydrological Institute during a monitoring program of the Skagerrak area dunng 1989 and

1990, initiated by the Swedish Enviromental Board. The hydrographie survey data were nicely

delivered by the Institute of Marine Research at Flödevigen Biological Station.

7

•

..4. References

Aarup, T., GrooIn, S.and Holligan, P. (1987): CZCS Scenes from the North Sea. ICES,CM.19871C;31

Barton, I. (i989): Digitization effects in AVHRR and MCSST data. RemOte Sens. Environ., 29, 87­89.

. .

Holligan, P, Aarup, T. and S. Groom (1987): The North Sea satellite colour atlas. Cont. SheljRes.,vol 9, No. 8, 667-765.

Häkansson, B. (1989):Remote sensing oftotal suspended matter from the Glomma river in the.• Skagemik. Vatten, 45, Nr. 4, 271-277.

Johannesseen, J., Johannessen, 0., arid P. Haugan (1989): Remote sensing and model simulationstudies of the Norrwegian coastal current during the algal bloom in May 1988. Int. J. RemoteSensing, 1989, vol. 10, no. 12.

Merk, M., and R., Saetre (1980): The Norwegian coastal current. Proceedingsjrom the Non't'egiancoastal current symposium, Geilo, vol1+2.

Mork, M. (1981): Circulation phenomena and frontal dynamics of the Norwegian coastal cirrrent .Royal Soc. London, Philosophical transactions, ServA, mathematical and physical sciences,302(1472): 635-647.

Pettersson, H. (1989): Norwegian rerriote sensing spectrometry for mapping and rnonitoring of algalblooms and polurion - NORSMAP'89. Summary of field campaign. NRSC, Techizical report No.22.

Pettersson, H: (1990): NORSMAP'89 Project report & recominendaiions. NRSC, Technical reportNo. 28. '.

Pingree, R., Holligan, P., Mardell, G. and R. Harns (1982): Vertical distributiOll ofplankton in theSkagerrak in relation to doming of the seasonal thennocline. Cont.SheljRes., vol. 1, no. 2.

Richardson,K (1989): Phytoplankton distribution and activity in the Skagerrak: a reView. ICESpaper CM19891L:24 Sess Q.

RObinson,I. och N. Ward (1989): Co~parison between satellite and ship measurements of seasurface temperature in the north-east Atlantic Ocean. IntJRemote Sensing,Vol.10,NoA and 5,787-799.' .

Rodhe. J. (1987): The large-scale circulation in the Skagerrak; interpretation of some observations.Tellus 39A, 245-253.

Svarisson, A. (1975): Physical anrl chemiCal oceanography of the Skagerrak and the Kattegatt Inst.Marine Res., Report No. 1.

Svejkovsky, J.,(1988): Sea sunace flowestimation from advanced very high resolution radiometerand coastal zone color scanner satellite imagery. J.GeophysRes.,Vol.93,No.C6,6735-6743.

Sörenseri,K., Lfnddl, T. and J. Nisell (1989): The informatiori content of AVimR, TM and SPOTdata in the Skagerrak sea. Proceedings ofthe IGARRS'89-12th Canadian SymposiUm on RemoteSensing. Vancouver, Canada, July 10-14,1989. .

.Tassan,S. (1987): Evahiation of th~ pOtential of the thematic mapper for manne applicarlon.Int.JRemote Sensing,Vol.8,No.lO,1455-1478.

..Figure captions

Figure 2.1: Topography, ship transects and some coastal meterological and

oceanographic stations along the Swedish coast.

I'I

•

Figure 2.2:

Figure 2.3:

Figure 3.1.1:

Figure 3.1.2:

Figure 3.2.1:

Figure 3.2.2:

Figure 3.2.3:

Figure 3.3.1:

Figure 3.3.2:

The satellite derived sea surface temperature shows good correlation with the

in situ measured temperature at 5 m depth' The correlation coefficient is 0.91

and data were obtained from May 17 and 18, 1989.

Air temperature from 10 m height and water temperature from 5 m depth

measured hourly at the coastal station Koster.

Wind speed and direction from the two coastal stations Koster and

Mäseskär. The presented observations are daily average values, beginning on

May 10, 1989 and ending on May 20, corresponding to the Julian days

130 and 140.

The upper figure shows the salinity distribution in the surface layer along the

transect between Norway Oeft) and Denmark (right). The bottom

configuration on the Danish side of the seetion is schematically drawn in

dark. The lower figure shows the salinity distribution of the surface layer in

the section from Sweden (left) and westward.

Schematically shown MeSST distribution in the Skagerrak.

Th6 horizontal salinity distribution at 5 m depth, obtained from the in situ

measurements.

Enhanced sea surface temperature schematically obtained from MCSST

covering the eastem part of Skagerrak. Notice the Baltic current separation

from the coast and the warm plym from the river Glomma tuming sharply to

southeast.

The surface velocity drift obtained from feature identification in the satellite

images from May 17,1989.

Vector diagram of the current offshore the Swedish west coast at N 58014' E

11003'. The current meter was placed at a depth of 15 m.

9E llE 13 E

S9N

S8N

S7N

11

10.5

10-u0

9.5-E-VJ• rI'.lU

9~

8.5

8.5 9 9.5 10In situ (OC)

10.5

---------------1

11

•12.5..----"---....&...- -.....1...-.............-..............- ...........---'--""'---'--+

(lC) 12 • VII., I.nI..• AirttaJe ..

11oS

11

10.5

10

9.5

9

80S

8~--r-..,....- .........----.--........-..---""""T""-~---r-~~ .........o 25 SO 7S 100 125 lSO 17S 200 225 ~

15i' 111' 1715 . 1115 1115

•

•

•

14

12

10

8

6

4

2 • Wind Speed m/s (MAseskär)o Wind Speed m/s (Koster)

142140138136134132130O+------.,...---.----r-----...----.----.----+128

JulianDays

350• Wind Direction (Mäseskär)

300 0 Wind Direction (Koster)

250

== 200.gj~ 150o"0.5 100~

50

142140138134 136JulianDays

132130O.J----.---..---__--.---r-----.----Io128

•

•

0

·10

·20

[;] ·30

·40

·50

·60

o 10 20 30 40 50 60 70 80

, 36,35

:..:

'. 34

t33

~{32

{31

,.:.,30

29

90 100 110 28

27

o 10 20 30 40 50 60 70 80 eo 100 110 120 130 140 150

36

35..34

;?33

..32

~31

~ 30-:~.

::29

28

27

26

25

24

•

•

•

59N

saN

57N

- ---- ------ ------1

~11.0°C

10.510.0

9.59.08.58.07.5

FI'S .3. J.. 1.

•

---------------------------------

•

9E UE 13 E

59N

S8N

57N

•

""""""""'"""""""""'"""""""""'"""""""""'"""""""""'"""""""""'"""""""""""""""""'"

" """""'"""""""""""""""""""""'""""'"'" """'""""""'""""""'""""""'":--..""""""""""""""""""""""""""""""""""'"""""'"""""'"""""'"""""'"""""'""""""", "'", "'"

"""""'""""""""'""""'"" ", '"" '":-... '"

"""'""'""'""""""'"'"" ,

10.8 oe10.410.0

9.69.2

•

17 maj 1989

1m/s

•

N

Velocity scale

O.5m/s

lat: 58-14'long:1r 03'depth:15 m