Mean cloud distributions over tropical oceans

Mean cloud distributions over tropical oceans

By KSHUDIRAM SARA, IntJtitute of Tropical Meteorology, Poona-5, India

ABSTRACTThe paper suggests that mean cloud distributions over tropical coeans as revealedby satellites may be intimately related to distributions of mean ocean surface tem-perature. From a qualitative consideration of the atmosphere's response to the under-lying ocean surface temperature through energy exchanges, formation of clouds in amoving airmass over cold and warm oceans is discussed and illustrated with specificexamples from the Indian ocean. It is shown that when warm air moves over a coldocean, there is little or no formation of clouds. In the case of cold air moving over awarm ocean, there is gradual transformation of clouds from stratified to cellular typeand development of cloud cells and cloud groups in the warmer parts of the ocean.Randomly distributed large cloud masses or ensembles known as cloud clusters appearto form over warm equatorial oceans. The intertropical convergence zone which islocated in the region of maximum ocean surface temperature appears to be charac-terised by a band of cloud clusters. Occasionally, two separate cloud bands may beseen in near-equatorial waters, one in each hemisphere. Formation of two prominentcloud bands in the southern hemisphere is discussed. Finally, an examination of themean rainfall distribution over tropical oceans suggests that it may be related todistribution of mean ocean surface temperature in much the same way as the meancloud distribution.

1. IntroductionBjerknes (1966, 1969) has recently demon-

strated that equatorial anomalies of oceansurface temperature in the Pacific caused byvariation in the strength of the Peruvian cur-rent affects the intensity of the atmosphericHadley circulation of the northern hemisphereand through it the circulation features of higherlatitudes. Namias (1969) has sought to discovera systematic correlation between long-periodanomalies of ocean surface temperature andfluctuations of the general circulation of theatmosphere in middle and high latitudes. Re-cently, Saha (1970a) has indicated a possibilitythat zonal anomalies of sea surface temperaturein equatorial Indian ocean may produce amarked effect upon the intensity of the mon-soon circulation and rainfall distribution overIndia. In two subsequent papers, Saha (1970b,c) has suggested that equatorial anomalies ofthe ocean temperature in the major oceansmay be largely responsible for observed longi-tudinal asymmetry in the location and seasonaloscillation of the intertropical convergencezone as well as formation of a double inter-tropical convergence zone. In the present paper,it is proposed to study mean cloud distributionsover tropical oceans as revealed by satellites

Tellus XXIII (1971), 2

in relation to distributions of mean oceansurface temperature. A possible correlation be-tween the distributions of mean ocean surfacetemperature and mean annual rainfall is alsoexamined.

2. Distribution of mean ocean surfacetemperature

Fig. 1 shows the distribution of mean surfacetemperatures (Sverdrup et a1., 1942) and prin-cipal cold ocean currents in tropical oceansduring February and August. The cold oceancurrents are the Peruvian current south of theequator and the Californian current north ofthe equator in the Pacific ocean and theBenguela current south of the equator and theCanaries current north of the equator in theAtlantic ocean during both February andAugust and the cold Somali current in theIndian ocean during August. The contributionof the ocean currents to the observed distribu-tion of ocean surface temperature is qualitativel~revident from. Fig. 1. Upwelling along thecoast of Peru and the equatorward movingcold Peruvian current maintain an almostpermanent regime of cold equatorial watersfrom near Galapagos island (0054' S, 8937' W)to almost Canton island (248' S, 17143' W.)

184

FEBRUARY

KSHUDIRAM SARA

Fig. 1. Distribution of mean ocean surface temperature in tropical oceans. Thin continuous linesare isotherms (Oe). Thick continuous lines areridges of high ocean surface temperature. Thick lineswith arrow-heads show direction of cold ocean currents.

Between this wedge of cold equatorial watersand the cold waters of the south Pacific ocean,there appears a well-marked temperature ridgeextending from near Solomon islands (longitudeabout 160 E) eastsoutheastward into southPacific. The temperature ridge between thePeruvian current and the Californian currentin the eastern Pacific lies north of the equatorthroughout the year but in the western Pacificit lies north of the equator during August andsouth of the equator during February. In theIndian ocean, the temperature ridge lies a fewdegrees south of the equator during February,whereas it practically disappears during Au-gust. However, during the southwest mon-soon, upwelling along the Somali coast and thecold Somali current in the extreme westernIndian ocean maintains a strong zonal tem-perature anomaly with cold water in the westernIndian ocean (West of about 60 E) and warmwater in eastern Indian ocean (Saha, 1970a).

In the Atlantic, two temperature ridgesappear during February one north of the equa-tor, but very close to it, between the Benguelacurrent and the cold Canaries current and theother in extreme southwestern Atlantic betweenthe Benguela current and the cold waters ofthe southern Atlantic. During August, thetemperature ridge north of the equator persistsbut that in southwestern Atlantic weakensand remains confined near the coast of Brazil.

The seasonal movement of the ocean currentsand temperature ridges is evident from Fig. 1.

3. Ocean-atmosphere itneraction

Physical processes that may lead to cloudformation over an ocean must be very complexindeed, involving as they do, a two-way inter-action between the ocean surface and theatmosphere. A crude but somewhat realisticpicture of these processes may, however, beobtained from a qualitative and simplifiedconsideration of heat and water vapour ex-changes with the ocean surface. Normally,when an airmass after flowing over a cold sur-face enters a warm ocean, there is likely to beupward flux of both sensible heat and watervapour across the ocean-atmosphere interfaceaccording to the well-known flux-gradientrelationships (Priestley, 1959),

where H, E denote respectively the verticalfluxes of sensible heat and water vapour, e isdensity of air, 11 is specific heat of air at

TelIus XXIII (1971), 2

MEAN CLOUD DISTRIBUTIONS OVER TROPICAL OCEANS 185

constant pressure, T and q are respectivelytime-mean temperature and specific humidityof air, K H , K w are respectively the eddy diffusivities of heat and water vapour, r is dry.adiabatic lapse rate of temperature, and Z isvertical co-ordinate positive upward.

Evaporation and non-adiabatic warming ofthe lower boundary as the airmass moves overthe warm ocean and updrift due to boundary-layer frictional convergence may cause, throughturbulent mixing, a rapid modification of theairmass into a warm and humid atmosphere,the depth of the moist layer above the oceansurface increasing downstream. At some stage,condensation may begin with formation ofclouds, each cloud cell releasing its latent heatof condensation into the atmosphere. Thiscondensation heating together with surfaceheating may lead to a gradual fall of barometricpressure which in its turn may cause increasedfrictional convergence and upward motion anddevelopment of the cloud cells over the warmocean. The importance of both ocean surfaceheating and condensation heating in thedevelopment of tropical atmospheric circulationsystems has recently been demonstrated byCharney & Eliassen (1964) for hurricanes andby Pike (1968) and by Manabe, Holloway &Stone (1970) in the case of the intertropicalconvergence zone, using numerical models.The above processes are likely to be reversedwhen a warm airmass flows over a cold ocean.Rapid loss of sensible heat to the cold oceanmay occur across the lower boundary with thedepth of the cooled layer above the oceansurface increasing downstream. Taylor (1915)who first investigated this problem showedthat the height Z to which the effect of surfacecooling will proceed after a time t may begiven approximately by the relation Z = 2VKt,where K is the co-efficient of eddy diffusion.The net effect of the surface cooling may thusbe vertical stabilisation of the airmass andgeneral lack of updraft. In fact, cooling mayproduce a gener~l rise in barometric pressure,slow subsidence, and divergence of air. The netresult of these processes may be a general lackof cloudiness over a cold ocean when a warmairmass moves over it. Both the processesmentioned above may, however, be considerably influenced by the presence of convergenceor divergence in the moving airmass. Rapidcloud growth should occur when a system with


.

~''''O -

Fig. 2. Equatorial distribution of sea surfacetemperature (SST) and barometric pressure (P)in Indian ocean and parts of Western Pacific duringJuly, 1964. Values of SST east of 97.50 E referto 2.50 S latitude.

convergence of air such as a developing pressuretrough or front enters or lies over a warm ocean.Cloud growth is likely to be retarded or subdued when a system with divergence lies ormoves over a cold ocean. Intermediate cloudgrowth may be expected when a system withconvergence moves over a cold ocean or asystem with divergence moves over a warmocean.

The validity of the above analysis of ocean-atmosphere interaction may be qualitativelydemonstrated with reference to conditions inequatorial Indian ocean (Saha, 1970a). Fig. 2shows equatorial distribution of mean seasurface temperature and mean sea level baro-metric pressure over the Indian ocean andneighbouring Western Pacific during July,1964. It shows the equatorial eastern Indianocean which is warmer having a lower surfacebarometric pressure than equatorial westernIndian ocean which is colder. Fig. 3a-d showthe distribution of mean air temperature andhumidity mixing ratio at surface, 850, 700 and500 mb, respectively, over equatorial Indianocean during July 1964. It is qualitativelyevident that ocean-atmosphere interactionhas modified the temperature and humiditystructure of the overlying atmosphere upto atleast 500 mb.

4. Observed cloud distributions over coldand warm oceans

In a study of trade wind cloud formation overWestern Caribbean, Malkus (1957) relates thepresence and development of trade cumuluscloud groups to regions in the ocean which

186 KSHUDffiAM SARA

'3OE 40 50 6r! 70 80 90 100 110" 35"~'----,--r-"':";:"--,-----';':--r--=r=---r-":'=--.---=r-----,--""::':'------.--'-r--r-"':":":;35

.25

,JUL 19641000 IJlb

N . N

t5 i0~ 5 s ~S S

15 --- IS

~ ----14~ 24

. a 2525 10030E 70 80 9 110

30E 4r! 50 60 70 ar! 90" 100 110.3S 35

JULY 1964850 mb

2~

Fig. 3. Distribution of air temperature (OC) and humidity mixing-ration (g/kg) over equatorial andnorthern Indian ocean during July, 1964 at (a) 1000 mb, (b) 850 mb, (c) 700 mb, and (d) 500 mb.Continuous lines represent isotherms and dashed lines isohygries.


2t~s

2

Tellus XXIII (l971), 2


188 KSHunffiAM SARA

Fig. 4. Typical cloud pattern over the Arabianseadurig northernn summer as viewed by ESSA2satellite on 11 July, 1967 (orbit no. 6311).

are slightly warmer than the surroundings.Krueger & Fritz (1961) who studied the sat-ellite-observed distribution of cellular cloudsin the areas of north Atlantic and north Pacificanticyclones with relevant meteorological datafind that in the region of formation of theseclouds, ocean temperatures were on an averageabout 3C higher than the air temperatureswith the smallest difference on the westerndownwind side and the greatest on the easternupwind side. Hubert (1966) who made a de-tailed study of meso-scale cellular cloud forma-tions in the ocean finds that there are two typesof cellular clouds, namely open cells which areinvariably found when the ocean temperatureis higher than the air temperature and closedcells which are occasionally found when theocean temperature is actually colder than theair temperature. Hubert shows that closedtype cells which may have updraft at the cente:-and downdraft in the surrounding areas arcquite commonly found in the regions near theCanaries and off the California coast wherethe ocean temperatures are cold compared withother ocean temperatures. During the last fewyears, polar-orbiting satellites have madedaily passes over the Indian ocean and revealedthe cloud patterns relating to different circu-lation systems. Fig. 4 shows an ESSA-2 satellitaview of typical cloud patterns over the Arabiansea on 11 July 1967 during the southwestmonsoon season. It shows the sea area westof about 60 E which is generally colder thanthe air practically free of clouds. East of thislongitude, stratified clouds first appear which

Fig. 5. Cellular cloud patterns in the south-easttrade wind region of south Indian ocean as viewedby ITOS-l satellite on 25 August, 1970 (orbit nos.2676, 2677).

soon give way to cellular clouds of closed celltype. Cloud cells appear to grow rapidly asthe airmass moves over the warmer watersin eastern Arabian sea. It is not quite clearfrom the photograph whether the transforma-tion from closed cell type to open cell typetakes place as the airmass moves on to warmersea, although such transformation could beclearly seen on many other occasions, e.g. on18 and 21 Aug. 1966. Individual cloud cellsas well as cloud groups, however, appear todevelop rapidly over the warm eastern Arabian

Fig. 6. An ESSA-8 photograph of a typical cloudcluster in equatorial Indian ocean on II September,1970 (orbit nos. 7965, 7966).



Fig. 7. An ESSA2 satellite view of typical cloudpatterns over Arabian sea during northern winteron 4 January, 1967 (orbit no. 3927).

sea. Fig. 5 shows an ITOS-I satellite photo-graph of cellular cloud patterns on 25 Aug.1970 in southeast trade winds of the southernIndian ocean as they approach the wannequator. Small cloud cells which fonn overcold ocean in the field of the subtropicalanticyclone grow rapidly as the trades move

DEC.1967- FEB. 1968

on to warm equatorial sea. Near the equator,cloud cells develop both individually and ingroups and appear to fonn masses usuallyknown as clusters. Fig. 6 shows satellite photo.graph of a typical cloud cluster over the equatorduring the southwest monsoon season. Thecirculation pattern over the Indian ocean isreversed during the northern winter. In theArabian sea, in particular, cold northeast tradewinds blow at first over cold waters of northArabian sea but subsequently over wannerwaters of south Arabian sea. Fig. 7 shows anESSA2 satellite view of typical cloud patternsover the Arabian sea on 4 January 1967during the northern winter. It shows northernArabian sea practically clear of clouds. Cellularclouds appear to form and develop progres-sively over south Arabian sea where the oceansurface is wanner than the air.

In recent years, a number of workers (e.g.Winston & Taylor, 1967; Taylor & Winston,1968; Sadler, 1969; Kornfield & Hasler. 1969;Winston, 1969; Oliver & Anderson, 1969; An-derson et aI., 1969) have studied the meancloud distributions over tropical oceans, utilisingdata furnished by either polar-orbiting orgeostationary satellites or both. Most of theEeworkers give the average cloudiness in differentparts of the tropics. Fig. 8 shows the distribu-

60 90

150 180" 150" 3r! , I 0'--

/-- I

" -"'ISN,

,------- .[

o -

15~S

60 90

Fig. 8. Distribution of brightness from ESSA-3 and 5 digitized pictures for periods Dec. 1967 - Feb.1968 and Jun.-Aug., 1967. The dashed lines enclose areas of minimum brightness, while hatched areasshow regions of maximum brightness. The heavy dotted lines indicate the major axes of maximumbrightness (or cloudiness). (After Hubert, Krueger, and Winston, 1969.)Tellus XXIII (1971), 2

190 KSHUDIRAM SARA

30r;:----r'""'CT'--r---'4''---r--'-';n--,----'-'T''----,r--7f''-,.---7P'--,----'ir:''''r-:H".........>


Table 1. Number 01 days 01 occurrence 01 double cloud bands in equatoriel Indian oceen during July1966 through Sept. 1970

Year Jan. Feb. Mar. Apr. May Jun. Jul. Aug. Sep. Oct. Nov. Dec.

1966 a a a a a a 23 15 12 12 8 151967 4 6 4 4 3 3 3 6 4 2 0 31968 4 4 8 5 3 7 5 8 12 9 14 141969 3 3 1 0 6 4 6 7 7 2 3 11970 5 11 6 2 3 11 4 4 5

a APT pictures not available.

location as the near-equatorial ridge of oceansurface temperature. This is certaintly as itshould be, for convergence zones are theregions of ascending air motion and henceconversion of latent heat of water vapour intosensible heat energy by way of condensation.Riehl & Malkus (1958) have emphasised therole of the equatorial trough zone or the inter-tropical convergence zone in the heat balanceof the tropical atmosphere and pointed outthat although the tropics as a whole serves asa heat source for the general circulation of theatmosphere, it is mainly through the equaorialtrough zone that both sensible and latent heatenergy is transmitted upward for eventualtransfer to higher latitudes. Cloud distributionsare thus visible manifestations of this importantprocess of energy transfer and Figs. 8 and 9would seem to imply that bulk of this transfertakes place through the ITCZ. Occasionally,two separate cloud bands apparently associatedwith formation of double ITCZs, may be seenover limited regions of the tropical belt, oneon each side of the equator, though theexistence of these bands may not be reflectedin the average seasonal pictures. Hubert,Krueger & Winston (loc. cit.) who computedseasonally-averaged cloud distribution during1967-68 over the tropical belt conclude, interalia that the Indian ocean does not exhibitany double tropical cloud zones. Daily APTphotographs relating to Indian ocean, however,appear to reveal a different situation. Table 1presents the number of days of occurrence ofdouble cloud bands in the Indian ocean duringperiod July, 1966 through September, 1970, asrevealed by satellites. It is evident that doublecloud bands which no doubt are to be associatedwith double intertropical convergence zonesare of fairly frequent occurrence in the Indian


ocean, although Table 1 shows that there wereat least two months during the 51 monthperiod when no double ITCZs were observed.One may also note from TabJe 1 that the numberof days of occurrence of double cloud bandsduring period July through September, 1966which was a drought year in India is muchgreater than that during corresponding periodsin other years which were all good monsoonyears, Le. they had normal or more than normalamounts of rainfall. The life period of a doublecloud band appears highly variable with anaverage at about 5 to 6 days, and extremes at1 and 12 days. Fig. 10 shows satellite photo-graphs of two typical double cloud bands inthe Indian ocean, one an ITOS-l photographon 11 June, 1970 during the SW monsoon andthe other an ESSA-8 photograph on 19 Decem-ber, 1968 during the NE monsoon season.

The occurrence of a double cloud band inequatorial eastern Pacific ocean during northernlate winter and spring season has been re-ported by a numer of workers (see, for example,Kornfield & Hasler, 1967). Table 2 showsthe number of days of occurrence of doublecloud bands in equatorial Pacific ocean duringFebruary, 1967 through March, 1969.

A conspicuous difference between the twooceans appears to be that whereas in the Pacifica double cloud band may be observed duringnorthern later winter and spring only, it maybe observed in the Indian ocean at any time ofthe year.

In a numerical study of the interaction be-tween the ocean and the atmosphere usinga time-dependent two-dimensional primitive-equation model of a zonally-symmetric atmo-sphere, Pike (1968, 1970) shows that the ITCZstructure over the ocean is determined mainlyby the surface temperature profile, the major

192 KSHUDIRAM SARA

Fig. 10. A satellite view of two typical doublecloud bands in the Indian ocean, one (a) an ITOS-lphotograph on 11 June, 1970, (orbit nos. 1738,1739), and the other (b) an ESSA8 photographon 19 Decemcer, 1968 (orbit nos. 0043, 00(4).

updrafts being located, on the average at thelatitudes of maximum temperature. Withhemispherically symmetric conditions at the

surface and in the subtropical atmosphere, asingle equatorial convergence zone appears tobe the equilibrium state unless there is a markedsurface temperature minimum there when theITCZ forms several degrees of latitude awayfrom the equator. It may be double with abranch in each hemisphere or it may becomesingle if one of its branches reaches sufficientstrength to suppress the other by flank sub-sidence. Pike's finding that the ITCZ forms atthe latitude of maximum ocean surface tem-perature is, indeed, most interesting in thecontext of the present study. However, hismodel being two-dimensional cannot possiblytake account of the marked zonal asymmetrythat exists in equatorial ocean surface tempera-ture (Fig. 1) and that may have importantbearing on the formation of double ITCZs(Saha, 1970c). It is interesting to note thatManabe, Holloway & Stone (1970) who carriedout numerical experiments, with a globalgeneral circulation model conclude that acontinuous supply of energy from the warmsea surface is necessary for the location and in-tensity of the intertropical convergence zone aswell as for development of tropical cycloneswhich may form in it.

6. Southern hemispheric cloud handsAs shown in Fig. 8 and amply confirmed by

recent ATS photographs, two prominent andextensive cloud bands appear in the southernhemisphere, one in central Pacific and the otherin western Atlantic, both originating near theequator and extending deep into southernlatitudes in a NW-SE orientation. The cloudbands are prominent during February and fadeout considerably during August. In fact, duringAugust the band in southwest Atlantic appearsto be confined near the coast of Brazil only.One may ask: what is the origin of these hugecloud bands forming in the southern oceans.Riehl (1954) is of the view that a belt of heavyrainfall in south central Pacific (which more orless coincides with the position of the cloudband shown in Fig. 8) can be connected with alarge trough and convergence line emanatingfrom the polar zone and extending northwesttowards the equatorial trough. From Figs. 1and 8 it may be seen that the cloud bands inboth the Pacific and the Atlantic oceans appearin more or less the same positions as the re-


MEAN CLOUD DISTRffiUTIONS OVER TROPICAL OCEANS

Table 2. Number of days of occurence of double cloud bands in equatorial Eastern Pacific

193

Year

196719681969

Jan.

ooo

Feb. Mar. Apr. May

4 18 0 0o 6 7 0o 4 0 0

Jun. Jul.

o 0o 0o 0

Aug. Scp.

o 0o 0o 0

Oct.

ooo

Nov. Dec.

o 0o 0o 0

spective ridges of ocean surface temperature.One may, therefore, speculate that these bandsmay be generated in troughs or convergencezones that may form over warm waters whichmay be drawn into higher latitudes by theanticlockwise circulations of the subtropicalanticyclones. During February the equatorialbelt is warmest and there may be maximumpenetration of warm equatorial waters intohigher latitudes with the result that the bandsmay be very prominent and extensive. DuringAugust, when the equatorial ocean temperatureridge line moves well north of the equator andthe cold ocean currents of the southern oceansreach their northernmost positions, there maybe lesser penetration of warm water into thesouthern hemisphere with consequent dis-appearance or weakening of the convergencezones and hence the cloud bands. Satellitepictures appear to show tht the above conver-gence zones of the southern oceans interactfrequently with eastward-moving polar frontsthat appear to extend from the south polarzone into the tropics. The interaction, however,appears to be rather complex.

7. Mean rainfall distributions

It may seem relevant here to comment brieflyon distributions of mean annual rainfall overtropical oceans in relation to mean oceansurface temperature. Although clouds areessential for rainfall, mere presence of cloudsmay give no indication of the distribution ofrain. However, Sadler (1969) who compared2-year average cloud distribution as revealedby satellites with normal rainfall distributionover equatorial and southern Africa duringFebruary, 1965-1966, found a surprisinglyclose correlation between the two. Also, Pike(1968) in his numerical study has shown thata rainfall peak appears over equatorial oceanwhen it is warm and a rainfall minimum whenit is cold. The cold oceans of the subtropical


belts are generally dry. Almost desert condi-tions prevail in equatorial eastern Pacific wherePeru currents maintain an almost permanentregime of cold surface water and in equatorialAtlantic where the Benguela current keepsa minimum surface temperature. In equatorialIndian ocean, the region west of about 60 Eincluding the Arabian sea which is affectedby the cold somali current is practically dry.The equatorial belt comprising eastern Indianocean and western Pacific ocean is the wettestregion of the tropical oceans (Riehl, 1954).In south Pacific and south Atlantic oceans,the locations of cloud bands associated withthe ridge of ocean surface temperature discussedin the preceding section appear to be markedby high precipitation (Riehl, 1954). In short, thedistributions of mean annual rainfall appearto be related to mean ocean surface temperaturein much the same way as mean cloud distribu-tion.

8. Summary and conclusion

It has been suggested that mean cloud distri-butions over tropical oceans may be largelyrelated to distributions of mean ocean surfacetemperature. Stratified as well as cellular typesof clouds which form over cold oceans seldomappear to grow vertically. On the other hand,cellular clouds that form over warm oceansappear to grow to large heights. Large groupsfor ensembles of developed cellular clouds knownas cloud clusters appear to be characteristicfeatures of warm equatorial oceans. Cloudclusters which are usually randomly distributedappear to organise themselves in bands alongthe intertropical convergence zone and otherconvergence zones which form over the warmestparts of the oceans. Evidence is produced offormation of double cloud bands in the Indianocean and the Pacific ocean. Two prominentcloud bands which appear in southern oceansappear to be associated with ridges of ocean

194 KSHunmAM SARA

surface temperature in the respective regions.An examination of the distribution of meanannual rainfall suggests that it may be relatedto mean ocean surface temperature in muchthe same way as mean cloud distribution.

9. AcknowledgementThe author's grateful thanks are due to

reviewer whose comments on the original

manuscript were found very helpful. Acknowl.edgement is made to the Director General ofObservatories, India, for permission to publishthis paper.

REFERENCESAnderson, R. K. et al. 1969. Application of meteoro-

logical satellite data in analysis and forecasting.ESSA Tech. Report NESC-51.

Bjerknes, J. 1966. A possible response of theatmospheric Hadley circulation to equatorialanomalies of ocean temperature. Tellus 18, 820-829.

- 1969. Atmospheric teleconnections from theequatorial Pacific. Mon. Wea. Rev. 97, 163-172.

Charney, J. G. & Eliassen, A. 1964. On the growthof the hurricane depression. J. Atmos. Sc. 21,68-72.

Hubert, L. F. 1966. Mesoscale cellular convection.Meteorological Satellite Laboratory Report No.37, ESSA, U.S.A.

Hubert, L. F., Krueger, A. F. & Winston, J. S.1969. The double intertropical convergence zone-fact or fiction. J. Atmos. Sci. 26, 771-773.

Kornfield, J. & Hasler, A. F. 1969. A photographicsummary of the Earth's cloud cover for the year1967. J. Appl. Met. 8, 687-700.

Krueger, A. F. & Fritz, S. 1961. Cellular cloudpatterns revealed by TIROS I. Tel/us 13, 1-7.

Malkus, J. S. 1957. Trade Cumulus cloud groups:Some observations suggesting a mechanism oftheir origin. Tel/us 9, 33-44.

Manabe, S., Holloway, J. L. & Stone, H. M. 1970.Tropical circulation in a time-integration of aGlobal model of the atmosphere. J. Atmos. sci. 27,580-613.

Namias, J. 1969. Seasonal interactions between thenorth Pacific ocean and the atmosphere duringthe 1960's. Mon. Wea. Rev. 97, 173-192.

Oliver, V. J. & Anderson, R. K. 1969. Circulationin the tropics as revealed by satellite data.Bull. Amer. Met. Soc. 50, 702-707.

Pike, A. C. 1968. A numerical study of tropicalatmospheric circulation. Sc. Report No.1, ContractNo. F 19628-68-C-0144, Inst. of Atmos. Science,Univ. of Miami, 125 pp.

- 1970. The intertropical convergence zone studiedwith an interacting atmosphere and ocean model.

Sc. Report No.2, Contract No. F. 19628-68-C-0144. Inst. of Atmos. Science Univ. of Miami,25 pp.

Priestley, C. H. B. 1959. Turbulent transfer in thelower atmosphere. The University of ChicagoPress, Chicago.

Riehl, H. 1954. Tropical meteorology. McGraw-HillNew York.

Riehl, H. & Malkus, J. S. 1958. On the heat balanceof the equatorial trough zone. Geophysica 6, 503.

Sadler, J. C. 1969. Average cloudiness in the tropicsfrom Satellite observations. I IDE meteorologicalmonographs, no. 2. East-West Center Press,Honolulu, Hawaii, pp. 22.

Saha, K. R. 1970a. Zonal anomaly of sea surfacetemperature in equatorial Indian ocean and itspossible effect upon monsoon circulation. Tellus22, no. 4 (Accepted for publication.)

- 1970b. Longitudinal anomaly of the intertropicalconvergence zone as related to ocean tempera.tures. (To be published.)

- 1970c. On the origin and distribution of doubleintertropical convergence zone. (To be published.)

Sverdrup, H. U., Johnson, M. W. & Fleming, R. H.1942. The oceans: their physics, chemistry andgeneral biology. Asia Publishing House, Bombay.

Taylor, G. I. 1915. Eddy motion in the atmosphere.Phil. Trans. R. S., A., 215, 1-26.

Taylor, V. R. & Winston, J. S. 1968. Monthly andseasonal mean global charts of brightness fromESSA-3 and ESSA-5 digitised pictures. ESSATechnical report, N ESC-46.

Winston, J. S. & Taylor, V. R. 1967. Atlas of worldmaps of longwave radiation and albedo-forseasons and months based on measurements fromTIROS IV and TIROS VII. ESSA Technicalreport, N ESC-43.

Winston, J. S. 1969. Global distribution of cloudi-ness and radiation as measured from WeatherSatellites. Climate of the free atmosphere, (ed.D. F. Rex), 247-280.


MEAN OLOUD DISTRIBUTIONS OVER TROPICAL OCEANS

CPE;Il;HIIE P ACIlPE;Il;EJIEHIIH OBJIAHOB HA;Il; OHEAHAMII B TPOIlIIHAX

195

PeaYJIbTaThI CTaTbll IIORaahIBaIOT, 'ITO cpe~ml:epaCIIpe~eJIeHllH 06JIaROB Ha~ OReaHaMll B TpO.IIllRaX, IIOJIy'leHHhIe C IIOMO~blO CIIyTHllROB,MoryT 6hITb TeCHO CBHaaHhI C paCIIpe~eJIeHllHMll

cpe~Heit TeMIIepaTyphI IIOBepXHOCTll OReaHa.IICXOAH lla Ra'leCTBeHHOrO paCCMOTpeHllH peaR-~KK aTMOC~ephI Ha TeMIIepaTypy nOBepXHOCTllOReaHa IIyTeM IIpO~eCCOB 06MeHa ~lHeprKeit 06-

cym~aeTcH 06paaoBaHlle 06JIaROB B Boa~YllIHhIxMaccax, ~BlliRY~llXCH Ha~ XOJIOAHhIMll II TeII-JIhIMll 'IaCTHMll OReaHa. B Ra'leCTBe llJIJIIOCTpa.

~llH IIPllBO~HTCH CIIe~llaJIbHhIe npHMephI AJIHIIHAllitcRoro OReaHa. IloRaahIBaeTcH, 'ITO ROr~aTeIIJIhIit Boa~yx ABHiReTCH Ha~ XOJIO~HhIM OKeaHOM, TO 06paayeTcH MaJIO 06JIaROB HJIll OHllCOBceM He 06paaylOTCH. B cJIy'lae ABlliRemmXOJIOAHOrO Boa~yxa Ha~ TeIIJIhIM OKeaHOM npo.HCXO~llT nOCTeIIeHHoe npe06paaoBaHKe 06JIaROBHa CJIOllCThIX B Ry'leBhIe H paaBllTHe H'IeeK HrpyIIII 06JIaROB Ha~ 60JIee TeIIJIhIMll 'IaCTHMH

Tellua XXIII (1971), 2

OReaHa. CJIy'lafiHo paCIIpe~eJIeHHhIe 60JIbllIHe06JIa'lHhIe MaCChI llJIH aHCaM6JIll 06JIaROB,HaBeCTHhIe RaR ~RJIaCTepbU, npe~CTaBJIHIOTCH

06paaYIO~HMHCH Ha~ TenJIhIMH OReaHaMH BllRBaTOpllaJIbHhIX 06JIaCTHX. 30Ha MeiR~YTpOIIH'IecRofi ROHBepreH~HH,HaXO~H~aJICJIB 06JIaCTHMaRCHMaJIbHhlX TeMnepaTyp IIOBepXHOCTHOReaHa, xapaRTepHayeTcH IIOJIOCOtt 06JIa'lHblXRJIaCTepOB. IIHor~a MOiRHO BH~eTb ~Be OT~eJIbHhIX nOJIOCbl 06JIaROB B rrpH8RBaTOpHaJIbHott06JIaCTH, IIO o~Hofi B RaiR~OM IIOJIyllIapHH. 06-

cYiR~aeTCJI 06paaoBaHlle ~BYX 'IeTRO BhIpaiReH-HhIX IIOJIOC 06JIaROB B lOiRHOM IIOJIyllIapHH. Ha-

ROHe~, HCCJIe~OBaHlle cpe~HllX pacrrpe~eJIeHllitoca~ROB Ha~ TpOIIll'leCRHMll OReaHaMH IIORaahIBaeT, 'ITO OHO MOiReT 6hITb CBHaaHO C pacrrpe~eJIeHlleM cpe~Hett TeMIIepaTyphI rrOBepxHOCTH OReaHa BO MHorOM aHaJIOrll'lHhIM 06pa.aOM, RaR II cpe~Hee pacrrpeAeJIeHlle 06JIaROB.

Mean cloud distributions over tropical oceans

Documents

Transcript of Mean cloud distributions over tropical oceans