Water Volume Transport Necessary to Remove This Heat Is at Least

AN ABSTRACT OF THE THESIS OF

Boyd Ellertson Olson for the Ph. D. (Name) (Degree) in Oceanography presented on August 2,1967 (Major) (Date) Title: On the Abyssal Temperatures of the World Oceans

Redacted for Privacy Abstract approved: June G. Pattullo In comparison with solar radiation, the energy ofgeothermal heat flowing through the sea bottom is extremely small; nevertheles;, this energy is not insignificant in the circulation of the bottom water. Calculations indicate that in the deep basins of the South Atlantic the water volume transport necessary to remove this heat is atleast one-tenth of the total northward flow of Antarctic Bottom Water. Plots of mean values of near bottom salinity and oxygen versus mean potential temperatures help to trace the movement of the bottom water. Geothermal and adiabatic warming associated with downslope flow combine to produce a deep temperature (in situ) minimum in portions of most of the deep basins of the world.Adiabatic or near adiabatic temperature gradients have been measured near the bottom in many of these basins.Evidence of superadiabatic gradients from temperature measurements made with reversing thermometersis inconclusive; however, careful measurements with closely spaced thermometers suggest that such gradients do exist over vertical distances of a few hundred meters in some of the deepest basins. Decreasing potential density with depth, as found in some of the Atlantic Basins in association with sharp temperature and salinity gradients, is not necessarily an indication of unstable equilibrium. This is demonstrated by the results of stability calculations in the manner prescribed by Hesselberg and Sverdrup(1915). On the Abyssal Temperatures of the World Oceans

Boyd Ellertson Olson

A THESIS submitted to Oregon State University

in partial fulfillment of the requirements for the degree of Doctor of Philosophy June 1968 APPROVED: Redacted for Privacy

'rofessor of Oceanography In Charge of Major

Redacted for Privacy

Crm7tepartmentof Oceanography

Redacted for Privacy

Dean of Graduate School

Date thesis is presentedAugust 2, 1967 Typed by Marcia Ten Eyck for Boyd Ellertson Olson ACKNOWLEDGMENTS Chronologically, my first expression of appreciation should be to Dr. Wayne V. Burt, Chairman of the Department of Oceanography, who provided the opportunity for my return to school after a lapse of many years.I am indebted to him for his realistic appraisal of course requirements and his assistance in developing my doctoral program. For hours of stimulating discussions and written suggestions on my dissertation, I am indebted to my two major professors, Dr. Peter K. Weyl, now Professor of Oceanography at State University of New York at Stony Brook, and Dr. June G. Pattullo.To these and the other members of my Committee--Dr. Norman H. Anderson, Dr. Byron L. Newton, Captain John F. Tatom and Dr. Robert L. Smith--I extend my sincere appreciation.

Arranging in absentia for typing, printing and proofreadingwas a formidable obstacle until Dan Panshin offered to undertake this for me,I am deeply grateful for his generous assistance in making these arrangements and in resolving the innumerable small problems that arose. To the many others who offered assistance and encouragement-- faculty members, fellow students, friends, my understanding wife and tolerant children--I am also grateful. TABLE OF CONTENTS Page

L INTRODUCTION 1

IL DISTRIBUTION OF TEMPERATURE, SALINITY, POTENTIAL DENSITY AND OXYGEN IN THE DEEP OCEAN BASINS 12 Atlantic-Indian Basin 12 South Indian Basin 17 Southeast Pacific Basin 22 Argentine Basin 24 Brazil Basin 28 North American and Contiguous Basins 36 Sierra Leone Basin 41 Canary and Iberian Basins 42 Guinea and Angola Basin 49 Cape, Agulhas and Natal Basins 53 Crozet Basin 59 Madagascar-Mascarene Basins 61 Somali Basin 66 Mid-Indian Basin 67 Wharton Basin 73 South Australian Basin 77 Tasman Basin 79 Southwest Pacific Basin 83 Central Pacific Basin 817 Northeast Pacific Basin 91 Northwest Pacific and Mariana Basins 98 Philippine Basin 101 Ill. DISCUSSION AND CONCLUSION 108

IV. BIBLIOGRAPHY 126

V. APPENDICES 130 LIST OF FIGURES

Figure Page

1 Comparison of scale used for temperatureprofiles in this study with conventional scale. 11

2 Deep temperature (in situ) and salinityprofiles, Atlantic-Indian Basin. 14

3 Deep sigma-Q and oxygen profiles,Atlantic-Indian Basin. 16

4 Deep temperature (in situ) and salinityprofiles from small basins within the Scotia Sea. 18

5 Deep temperature (in situ) and salinityprofiles, South Indian Basin. 20

6 Deep sigma-Q and oxygen profiles,South Indian Basin.21

7 Deep temperature (in situ) and salinityprofiles, Southeast Pacific Basin. 25

8 Deep sigma-Q and oxygen profiles,Southeast Pacific Basin. 26

9 Deep temperature (in situ) and salinity piofiles, Argentine Basin. 29

10 Deep sigma-0 and oxygen profiles,Argentine Basin. 30

11 Deep temperature (in situ) and salinityprofiles, Brazil Basin. 33

12 Deep sigma-.0 and oxygen profiles, BrazilBasin. 35

13 Deep temperature (in situ) and salinityprofiles, North American Basin. 38

14 Deep sigma-Q and oxygen profiles,North American Basin. 40

15 Deep temperature (in situ) and salinityprofiles, Sierra Leone Basin. 43 LIST OF FIGURES Continued Figure Page

16 Deep sigma-.0 and oxygen profiles, Sierra Leone Basin. 44

17 Deep temperature (in situ) and salinity profiles, Canary and Iberian Basin. 47

18 Deep sigma-Q and oxygen profiles, Canary and Iberian Basin. 48

19 Deep temperature (in situ) and salinity profiles, Guinea and Angola Basin. 52

20 Deep sigma-0 and oxygen profiles, Guinea and Angola Basin. 54

21 Deep temperature (in situ) and salinity profiles, Cape. Agulhas and Natal Basin. 56

22 Deep sigma-Q and oxygen profiles, Cape, Agulhas, and Natal Basins. 57

23 Deep temperature (in situ) and salinity profiles, Crozet Basin. 60

24 Deep sigma-0 and oxygen profiles, Crozet Basin. 62

25 Deep temperature (in situ) and salinity profiles, Madagascar-Mascarene Basin. 64

26 Deep sigma.0 and oxygen profiles, Madagascar- Mascarene Basin. 65

27 Deep temperature (in situ) and salinity profiles, Somali Basin. 68

28 Deep sigma.-0 and oxygen profiles, Somali Basin. 69

29 Deep temperature (in situ) and salinity profiles, Mid-Indian Basin. 71

30 Deep sigma-0 and oxygen profiles, Mid-Indian Basin. 72 LIST OF FIGURES Continued Figure Page

31 Deep temperature (in situ) and salinity profiles, Wharton Basin. 75

32 Deep sigma-Q and oxygen profiles, Wharton Basin. 76

33 Deep temperature (in situ) and salinity profiles, South Australian Basin. 78

34 Deep sigma-.Q and oxygen profiles, South Australian Basin. 80

35 Deep temperature (in situ) and salinity profiles, Tasman Basin. 82

36 Deep sigma-Q and oxygen profiles, Tasman Basin. 84

37 Deep temperature (in situ) and salinity profiles, Southwest Pacific Basin. 86

38 Deep sigma -Q and oxygen profiles, Southwest Pacific Basin. 88

39 Deep temperature (in situ) and salinity profiles, Central Pacific Basin. 90

40 Deep sigma-Q and oxygen profiles, Central Pacific Basin. 92

41 Deep temperature (in situ) and salinity profiles, Northeast Pacific Basin. 95

42 Deep sigma-0 profiles, Northeast Pacific Basin. 96

43 Mean profiles of deep temperature (in situ), salinity and sigma-Q from 13 repeated casts, Northeast Pacific Basin. 97

44 Mean profiles of deep temperature and salinity, Northwest Pacific Basin. 100

45 Mean profiles of deep sigma-Q, Northwest Pacific Basin. 102 LIST OF FIGURES Continued Figure Page

46 Deep oxygen profiles, Northwest Pacific Basin. 103

47 Deep temperature (in situ) and salinity profiles, Philippine Basin. 105

48 Deep sigma-0 and oxygen profiles, Philippine Basin. 107

49 Adiabatic processes. 111

50 Potential temperature gradient vs. depth. 114

51 Potential temperature vs. salinity, major deep ocean basins. 119

52 Dissolved oxygen vs. potential temperature, major deep ocean basins. 120 LIST OF TABLES

Table Page

1 Temperature and Salinity Increases Between CRAWFORD Stations 1335 and 251. 49

2 Stability Function, F, for the Near-Bottom Water in the Major Basins of the World Oceans. 116

3 Stability Function, F, at Various Depths in the Brazil Basin. 118

4 Minimum Transport Required to Remove Geo- thermal Heat by Advection. 124

LIST OF PLATES Plate Page

I Major Deep Basins of the World Oceans. 145

II Temperature (In Situ) Distribution at the Deep Temperature Minimum. 146

III Areas in Which the Deep Temperature Minimum and Adiabatic Lapse Rates Have Been Observed. 147

IV Bottom Potential Temperature Distribution in the Major Deep Ocean Basins. 148

V Salinity Near the Bottom in the Major Deep Ocean Basins. 149

VI Sigma-O [1000 x (potential density-1)11Near the Bottom in the Major Deep Ocean Basins. 150

VII Dissolved Oxygen Content (ml/l) Near the Bottom in the Major Deep Ocean Basins. 151 ON THE ABYSSAL TEMPERATURES OF THE WORLD OCEANS

I.INTRODUCTION

The object of this study is to examine deep oceanographic station data for evidences of terrestrial heating and to explore how this heat affects the abyssal circulation.The inadequacy of information on ocean bottom depths, the inaccuracy of temperature and salinity measurement and the complexity of processes largely limit this examination to qualitative considerations.Because the density, hence the stability, of a water column is determined by the vertical distribution of temperature and salinity, these properties will be considered in some detail.In addition, dissolved oxygen content will be examined superficially for indications of the relative length of residence of the water in the different basins. In the process, charts (Plates I through VII) of the major basins (generally deeper than 5000 meters), the deep minimum temperature (in situ), the salinity, the bottom potential temperature, the potential density and the dissolved oxygen content of the bottom water in the major basins of the world have been prepared.These plates together with vertical profiles of temperature, salinity, potential density and dissolved oxygen are used in describing the abyssal conditions. Charts of the global distribution of potential temperature and salinity at the bottom have been prepared by Wust (1938), Dietrich 2

(1957), and Defant (1961).The first thorough analysis was made by Wust in the Atlantic Ocean.It is not the purpose here to perform an extensive analysis to improve upon these charts, butrather to look more closely at the implications ofthe vertical temperature structure near the bottom.Nevertheless, potential temperatures were required in this analysis, and a bottom potential temperaturechart has been included. By comparison with solar radiation, the energy of terrestrial heat flow is indeed small; however, it does not follow thatterrestrial heat is inconsequential to the abyssal circulation.The average inso- lation at the top of the atmosphere is one-fourth of the solar constant, or about 8 x 10 cal/cmsec, of which about three-fourths(6 x 10 cal/cm2sec) is absorbed by the earth's surface (Budyko, 1956). By comparison, the mean terrestrial heat flow (about 1. 4 x1o6 cal/cm2sec) is very small; still, this comparison is not too meaning- ful in that it does not represent in any way the amount of heat trans-. mitted to great depths in the ocean.This latter is largely a function of the vertical temperature gradients, vertical current components, and eddy coefficients.These last two factors are difficult to measure but are undoubtedly small in abyssal regions.Thus, the terrestrial heat flow cannot be dismissed as unimportant in the scheme ofthe abyssal circulation merely because it does not rank in magnitude with solar radiation. 3

The question of the effect of variations in the temperature of the bottom water on terrestrial heat flow and associated temperature gradients in the sea bottom has been raised frequently in recent years (Taylor and Langseth, in press; and Langseth, LePichon and Ewing, 1966).Such variations have been documented to depths of about 1000 meters (Gordon, Grim and Langseth, 1966).As early as 1922, Tydeman attributed the renewal of water in the deepbasins to the flow of heat from the earth's crust (Tydeman, 1922 and Potsma,

1958).Kuenen (1943) recognized heat flow and tidal currents as the combined cause of renewal of this water.A. rather extensive study of the water movement and renewal in the eastern basins of the East Indian Archipelago was based on the data collected by the Snellius Expedition of 1929-1930 Potsma, 1958).In his study of the bottom water in the Atlantic Ocean, Wust (1933) concluded that even in the more stagnant basins (not trenches) such as the Cape Basin (A-8, Plate I), the evidence of terrestrial heat is completely obscured by the heat of mixing and vertical components of flow.Evidence of the warming of the bottom water by crustal heat has been reported by both physical oceanographers and geophysicists in recent years (Gerard, Langseth and Ewing, 1962; Wooster and Volkman, 1960). The first successful determinations of heat flow through the sea floor were made in 1950 (Revelle and Maxwell).This was done by measuring the temperature gradient in the sediment and the thermal 4 conductivity of a core sample.Lee and Uyeda (1965) estimate that as of the end of 1964, about 2000 measurements of heat flow had been made, 89% of these in oceanic areas.Von Herzen and Langseth (1965) estimate that for an average geothermal gradient (6 x i0 °C/m.) anda uniform sediment layer, the error in heat flow measurements is not more than 10%; otherwise the error may be as high as 20%. The most comprehensive summary and statistical analysis of heat flow measurements is that by Lee and tJyeda (1965).Such measurements give values ranging from less than 0. 1 to greater than 8 micro-calories percm2per second(i.Lcal/cm2sec.).Average values appear to be highest along the oceanic rises and ridges (2. 13 iJ.cal/cm2sec. for the East Pacific Rise) and least for the trenches (0. 94ical/cm2sec. for 16 Pacific trench measurements). Mean values for the oceanic basins are 1. 18,1. 13 and 1. 34ical/cm2sec. for the Pacific, Atlantic and Indian oceans, respectively.The global mean heat flow found by Lee and Uyeda (1965) by spherical harmonic analysis is 1. 5±10% cal/cmsec. at the 95% confidence level.The means found from averages computed for grids having an area of

9 x1O4square nautical miles are: Atlantic,1. 21; Indian,1. 35; Pacific,1. 53; and all oceans,1. 42ical/cm2sec.The last figure compares with 1. 47cal/cm2sec. given by Von Herzen and Langseth

(1965).Nearly all of the extremely high and low values are found within a few hundred kilometers of the crests of the oceanic rises and ridges, the lower values occuring on the flanks of these features. For calculating the geothermal heating of bottom water circu- lating freely over large distances, a mean value of 1. 4I.Lcal/cm2Sec. will be used.This lower value, rather than 1. 5I.Lcal/cm2sec. is preferred inasmuch as the deeper basins are being considered in this study.A. larger, more convenient unit of about 44cal/cm2year is obtained by multiplying by the number of seconds in a year (31, 536, 000). A description of the most commonly measured and derived parameters reflecting the deep circulation patterns will be given for each of the major deep basins of the world in the next section.These basins are outlined at a depth of 2500 fathoms (about 4570 meters) in Plate I.The base chart for this and the other plates is an adaptation (rescaled, cropped and reprinted) of one of a series of depth charts prepared by Anderson (1964).A.lthough it was initially intended to study only those basins 5000 meters and deeper, scarcity of data and the importance in some regions of depths less than 5000 meters have prompted consideration of selected shallower depths, only slightly deeper than 4500 meters in some instances. Oceanographic data were obtained primarily from the U. S. National Oceanographic Data Center.The basic data source was a listing of all stations in the NODC files which extended to 5000 meters or deeper (over 700 stations).This set was later supplemented selectively by more than 300 shallower stations mostly to depths greater than 4500 meters from the NODC files plus stations taken by the author, and others gleaned from the literature or obtained by correspondence.Altogether more than 1000 stations were examined. A few of these were not used in the analysis because of the wide gap between bottom measurements and the next deepest measurements. The deepest measurements for each station ranged from the bottom to several hundred meters above the bottom. The parameter of greatest interest in this study is temperature. Plate II shows the horizontal distribution of temperatures at the deep temperature minimum (in situ).This feature is not discernible on the deep temperature traces in all basins (Plate III).In many cases the lowest temperature is obtained from the deepest thermometer so that the minimum cannot be determined.A. minimum must exist everywhere in order for geothermal heat to flow across the sea- bottom interface, but in regions of high horizontal convection the distance from the bottom to the temperature minimum is small compared to normal thermometer spacing.Detection then requires con- tinuous sensing such as is provided by the thermograd or closely spaced, accurate reversing thermometers.The temperature minimum has little physical significance except as the boundarybetween upward and downward molecular heat conduction; however, it is an easily identifiable feature in many basins which can be used in comparison of independentl y taken measurements of temperature, salinity, and other oceanographic variables. Potential temperatures were computed from the new tables of adiabatic cooling prepared by Crease and Catton (1961).The horizontal distribution of potential temperature near the bottom, which was used in the computation of potential density, is shown in Plate IV. It was initially intended to include a chart showing the horizontal distribution of salinity at the level of the deep temperature minimum; however, lack of precision and comparability in salinity measurements, coupled with the small vertical changes generally found at great depths, make this impractical.In place of this, the approximate salinity distribution near the bottom is shown in Plate V. Where significant differences between this and the salinity at the temperature minimum occur, these are discussed in the text.Variations in salinity attributable to errors in measurement were found to be disap- pointingly large and constituted the major handicap in this study.

Potential densities were computed from Knudsen1 s hydrographical tables (1901).Eckart (1958) has pointed out that random errors in experimental specific volume determinations (hence density) are not less than ±2 x ml/gm (±2 x 10 gm/mi for density) and that the systematic errors, which increase with pressure, are not less than ±p x 10ml/gm (wherep is the pressure in atmospheres).Accord- ingly, he questions the necessity and realism of computing the density to 0.001 or even 0.01 sigrna-t units.Any examination of the stability at depths as great as 5000 meters must, however, consider density variations of this order.Although discrepancies of the order of 0.01-0.03 sigma-t units are found at this depth between different water masses (discussed in Section II) there appears to be a consistency greater than this within a single water mass.Thus, some useful conclusions can be drawn even though the absolute value of the density is subject to considerable error.Sign-ia-O, the abbreviated expression of potential density i(potential density - 1.0) x1000] was used in plotting graphs and charts and is used interchangeably with potential density in the text.Its horizontal distribution in the major deep basins is shown in Plate VI.The potential density was used in reconciling divergent salinity observations and postulated water movement. Although the potential density is a conservative property and is thus useful in following water masses, its vertical distribution is not a measure of stability as has sometimes been assumed.Hessel- berg and Sverdrup pointed out as early as 1915 that the vertical dci- distribution of potential density (_.Q) only indicates qualitatively whether the equilibrium is stable, indifferent or unstable under conditions of constant salinity.As will be discussed later, this fails even as a qualitative indicator in the presence of a vertical salinity gradient. Plate VII gives the mean value of the dissolved oxygen content near the bottom in each of the major basins.Profiles for most of the major basins have also been included in the descriptive section. Of the properties covered in this report, oxygen appears to be most subject to error.This has been noted by several authors, particularly regarding some of the earlier determinations.Wooster and

Volkman(1960)noted that oxygen values at 5 km. from the DISCOVERY II expedition were lower than others in the western

South Pacific.Wust(1964)discusses the differences in oxygen determinations of various cruises and concludes that the systematic errors may be as much as 15%(too low) in some instances..

Gordon(1966)states that DISCOVERY II oxygen values used by him are about 10% too low.Although the oxygen values from DISCOVERY II used in this report are generally low, they are not uniformly low. Excluding two unreasonably high values, the mean of near-bottom oxygen values reported by DISCOVERY II in the Cape, Agulhas and Natal basins is about 0. 4 ml/l lower than the mean of all other values reported in this region, whereas the mean of the bottom values of representative profiles used in this study shows adifference of 0. 7 ml/l between DISCOVERY II data and all other reports.The depar. ture of the DISCOVERY II values from other reported values generally increases with depth. 10 A brief description of the distribution of temperature, salinity, potential density, and dissolved oxygen content for each of the major deep basins of the world is given in the following section.Represen- tative profiles of these variables are included.Scale expansion of these profiles, illustrated in Figure 1,facilitates examination of finer structures. 1 2 4

CONVENTIONALEXPANDED 5.00100 COVEIT!±! TEMPERATURE SCALE (°C) 10.00 1.50 15.00 2.00 Temperature scale 5 expanded 10 times 20.00 2 50 for this portion in plot at right. Figure 1. Comparison of scale used for temperature profiles in this study with conventional scale. I- 12

II. DISTRIBUTION OF TEMPERATURE, SALINITY, POTENTIAL DENSITY AND OXYGEN IN THE DEEP OCEAN BASINS

Atlantic-Indian Basin

This southernmost basin, whose east-west elongation is exag- gerated in the modified mercator projection, is not well defined by bathymetric surveys.It includes the Weddell Sea and the South Sand- wich Trench in its western extremity and measures nearly 1000 by 4000 kilometers at a depth of 4500 meters.Exclusive of the trench, the deepest portion of this basin appears to be in the vicinity of 58° 5. latitude and350F. longitude.Early oceanographic measurements were made in this basin by the VALDIVIA (1898-1899), the

DEUTSCHLAND (1911- 1912), the NOR VEGIA (1927- 1931), the METEOR (1925-1926), the VIKINGEN (1929-1930), the WILLIAM

SCORESBY (1931) and DISCOVERY II (1933-1935).In more recent years, the ELTA.NIN (1963-1965) and the OB (1957) have made measurements in this area.The hydrographic stations taken by the ELTANIN have been reported by the Lamont Geological Observatory (Jacobs, 1965 and 1966).Descriptions of the physical oceanography of the Atlantic Antarctic Ocean have been published by Brennecke (1921), Schott (1905), Mosby (1934), Wust (1928 and 1933), Deacon (1933 and 1937), and others. It is now widely accepted that the Antarctic Bottom Water has its 13 origin in the Weddell Sea. Brennecke(1921) first suggested that the water formed by cooling on the shelf bordering the Weddell Sea on the southwest formed a dense bottom current.This was substantiated by Wust (1928) and Deacon (1937).Mosby (1934) demonstrated from the data collected by Brennecke, that the bottom water could be a mixture of Atlantic deep water and shelf water of about equal proportions. Although the mechanism by which it is formed is not well understood, the source of this bottom water which spreads so extensively eastward and northward is not in serious dispute. Although not evident in all of the deep temperature profiles for this basin, a layer of minimum temperature extends throughout most of it at a depth of about 4500 meters.This minimum ranges from about -0. 55 to -0. 15 °C.In an elongated region extending west from 150E. longitude, the temperature profiles are very similar below a depth of 1000 meters.East of 15° F. longitude the temperatures increase gradually in such a way that the slopes of the profiles between the minimum and a depth of 1000 meters become progressively steeper.Below the minimum, the slopes appear to approach the adiabatic lapse rate rather quickly, particularly in the colder part of the basin.The top of the adiabatic layer is estimated to lie between 4800 and 5000 meters.Temperature and salinity profiles for this basin are shown in Figure 2. The salinity below about 3000 meters is between 34. 66 and 0 1

34.40 34.50 34.60 34.70 34.80 I I 0.00 I I 0.50 1.00 1.50 - i I -0.50 LI Il I IT__Ill] Il 2

6 T S 7 x ELTANIN DISCOVERY II 145 2014 - 610 58046'S. 34'S. 00 35'S.23° 49'W. 580 23'S. 57° 10'S. (Depth: 5175e) (Depth: 6657m)5394a) 2113I - i I I i o DISCOVERY II I i t i i FigureI 2. Deep temperature (in situ) I i i I and salinity profiles, i AUanticTT1t2n Basin. A 15 and 34. 67%o .Above 4000 meters depth there is some intrusion of higher salinity water into the northeast sector of the basin, where the North Atlantic Deep Water gradually sinks into this basin as it flows southeastward around the tip of Africa.This basin is essentially isohaline below a depth of 3000 meters. Throughout the isohaline layer, the density and potential density are functions only of the temperature and potential temperature, respectively, within the limits established by the constancy of the constituents of the water, the dissolved gasses and the salts.Using a salinity of 34. 66%o below 3000 meters and the reported salinities above this depth together with the observed potential temperatures, the potential densities were computed and are shown in Figure 3. In the vicinity of the Weddell Sea, the potential deisity increases very slightly with depth and at a nearly constant rate below about 500 meters.Farther east the vertical increase in potential density below 500 meters is greater and extends nearly to the bottom. Four stations were taken by the ELTANIN in the South Sandwich Trench in 1963.Profiles for one of these have been included in Figures 2 and 3.The temperature and salinity profiles are similar to others in the western part of this basin suggesting that the water is essentially of the same origin.At and below the oxygen minimum, the oxygen values are much higher in the Trench than at the other stations shown (Figure 3).This suggests that this water has had a 1

02 (mI/I) _. '& 27.60 4.0 rri 27.70 5.0 IHJ ''L27.80 6.0 li' 27.90 7.0 *T-W 28. 8.0 .00 2 0 14 4 6 7 x ELT&NIN DISCOVERY II 145 - 58° 34'S. 230 49'J 2014 - 61° 46'S 0° 35's I 10'E. II I I I - I 2113 - 58° 23'S.II 57 I II i I i I I i I t 0 DISCOVERY II 1 i I i 1 I I I I I I i FigureI 3. Deep slgma-o and oxygen profiles, I t 1 1 Atlantic-Indian Basin. '7 shorter residence time since its formation. Two small basins to the west of the South Sandwich Trench (A-la and A-lb, Plate I) are separated from this trench and the Atlantic- Indian Basin by the Scotia Ridge and the South Sandwich Islands. Profiles from these small basins show clear evidence of the relatively shallow sill depths (Figure 4).The easternmost basin has a minimum temperature of -0. 43 °C. between about 3500 and 3800 meters.The temperature gradient is essentially adiabatic below 3800 meters.In the westernmost basin temperature measurements show a minimum of -0. 14 °C.at about 2600 meters, indicating a much shallower sill depth.

South Indian Basin

East of the Atlantic-Indian Basin neither the bathymetry nor the physical oceanography is well known.The GAUSS traversed this area in 1902-1903 (Drygalski, 1903).Deep oceanographic stations were also taken by DISCOVERY II in this southernmost part of the Indian Ocean in 1932, 1936 and 1938.One deep station (3981 meters) was taken by the OB in 1956. A single observation by the GAUSS at 58° 29' S. latitude and 89° 58' F. longitude reporteda temperature of -0. 05°C. (potential temperature of ..0. 40 °C. ) at 4661 meters.The associated salinity (34. 40%o) is much too low to be plausible.Salinities in this area are 0 2 .90. 4 ri 6 T(°C) - lit i't lII 34.40 0.00 III! 'I 34.50 0.50 I I 1111111 34.60 1.00 lilIlillilill iii 34.70 1.50 34.80 2.00 - T - 7 - x ELTANIN ELTANIN 287286 - 60°600 21'S.55'S. 47°41° 40'W.28'W. - Basin A-lbA-la (Depth: 5199m)5238m) - ilii 111)1! till Figure 4. Deep temperature (In situ) and salinity profiles from small thu 11111111 I I basins within the ScotiaI Sea. I I thu 1 19 expected to be higher than in the Atlantic-indian Basin because of the addition of higher salinity water flowing downslope southeastward around the tip of Africa, mixing with the Antarctic Circurnpolar and Bottom waters.Farther east this is borne out by stations taken in the South Indian Basin. The South Indian Basin is also in a region that has been poorly charted.Its dimensions are estimated at about 700 by 1400 kilometers at a depth of 4500 meters.It is bounded on the north and east by the Southeast Indian Rise and on the south by the continental rise of Antarctica.Its western boundary is not well defined. The deepest known part of the South Indian Basin lies between about 110° F. and 140° F. longitude with its east-west axis at about 60° S. latitude.Maximum reported depths are between 4500 and 4800 meters.Representative temperature, salinity, potential density and dissolved oxygen profiles for this basin are shown in Figures 5 and 6.Striking features of these profiles are:1) the steepened salinity gradient below 1500 meters; 2) the high bottom potential density associated with the relatively low bottom temperature and high bottom salinity; and 3) the steep, nearly constant temperature gradient.Below 1000 meters the potential density gradient is steeper than anywhere in the Atlantic-Indian Basin. In this relatively shallow basin, there is no observational eviden of a deep temperature inversion in the data examined.In fact, the 01

Fl a 0 4

34.40 I 34.50 I I 34.60 34.70 34.80 T(°C) -'. I-_i I IjII1 0.00 'I lIIi III 0.50 I III1.00 I I 'I I I 1.50 I I I I' I' 2.00 - T S 7 DISCOVERY II 890 - 59° 5'S. 133° 19'E. (Depth: 4771m) - - I I I -t I I I I Figure I5. Deep temperature I I I (in situ) and salinity profi1es I I I SouthI Indian Basin. I i I i I I I C 1 E 0 4 6 8.0 02(mh/1)_.. I I 2'1.90 I ±0 'I II '1 11 21.3O IIII ill II 28. 00

7 DISCOVERY II 890 - 59° 5'5 1330 19'E -I I i II I Figure 6. Deep sigma-G andI oxygen profiles, SouthI Indian Basin. I I I I I I I I I III I I 22 temperature gradient of the deepest station in this basin (Figure 5) is nearly linear between 2500 meters and its maximum depth (4430 meters).The in situ bottom temperature ranges between about -0. 10 and 0. 0° C. The dissolved oxygen content in the deep waters is slightly less (4. 4 to 4. 6 ml/l) in the eastern part of this basin than in the eastern part of the Atlantic-Indian Basir, indicating a slightly longer residence time.In the western part of the basin, the oxygen content is comparable to that in the central part of the Atlantic-Indian Basin. This suggests the possibility of this water being formed on the Antarctic Shelf far eastward of the Weddell Sea.

Southeast Pacific Basin

The Southeast Pacific Basin is bounded by the continental rises of Antarctica and South America, by the Chile Rise on the north and the Albatross Cordillera on the northwest.Its connection with the South Indian Basin far to the west is uncertain. From the existing bathymetry, it appears that there are no passageway connecting with this basin at depths greater than 4000 meters.Southward extensions of the Southeast Indian and South Tasmania rises and the Albatross Cordillera appear to confine the cold core of deep circumpolar water so that it does not enter the Southwest Pacific Basin to the north.At a depth of 4500 meters, this basin has a maximum length of about 23 3200 kilometers and a width of about 1900 kilometers.Deep stations have been taken in this basin by DISCOVERY II (1931-1934 and 1951), the OB (1958), the BURTON ISLAND (1960) and the ELTANIN (1963-

1965). Although it has been observed throughout most of the basin, the deep temperature minimum is not confirmed by observations everywhere (Figure 7).The temperature at the deep minimum ranges from about 0. 20° C. in the southwest to greater than 0.70° C. in the northeast part of the basin (Plate II).Shallow adiabatic gradients are evident in several profiles.These are best documented at station 408 of the OB and station 231 of the ELTANIN. The potential temperature near the bottom ranges from about -0. 20 to +0. 35° C.This is somewhat less than the range given by Dietrich (1957).He indicates bottom potential temperatures ranging from -0. 3 to +0.4° C.Both Dietrich (1957) and Defant (1961) show a pronounced potential temperature minimum extending along the axis of the basin.This feature is less conspicuous in the data considered here, particularly in the northeastern half of the basin. A salinity maximum of about 34. 74%o at a depth of 1200 meters in the western end of the basin decreases and deepens to about 3000 meters in the northeastern end of the basin.This maximum reflects the high salinity water which flows southeastward from the Indian Ocean via the South Australian Basin.Below this maximum, the 24 salinity decreases gradually to values between about 34. 70 and 34. 71%o (Figure 7 and Plate V). The relatively high salinity water entering the basin from the northwest gradually mixes with colder, somewhat fresher water entering the area of the basin farther south.As it flows eastward, the high salinity water moves downward along constant density surfaces having values between about 1. 02788 and 1. 02789 near the bottom of this basin (Figure 8 and Plate VI). The decrease in oxygen during transit of the high salinity water from west to east appears to be more than compensated by oxygen contained in the cold, lower salinity water intermixing along these surfaces from the south.The dissolved oxygen content of the water entering the western end of this basin is high (5. 1 ml/l) compared to the oxygen content of the water in the eastern end of the Atlantic- Indian Basin (4. 85 ml/I); however, in view of the uncertainty of these values, particularly that reported by DISCOVERY II in the eastern end of the Atlantic-Indian Basin, this difference cannot be regarded as significant.

Argentine Basin

The Argentine Basin is bounded by the contirental rise of South America on the west, by the Scotia Ridge and South Georgia Rise on the south and southwest, by the MidAtlantic Ridge on the east and by T (°C) - 34.40 0.50 34.50 1.00 34.60 1.50 34.70 2.00 34.80 2.50 14 30E T 0 ELTANINELTMIN 225444411 - 56°62063° 4'S.1'S. 104°135° 57'I. (Depth: 5072n) 7'W.4'W. (Depth: 5064m)4821m) Figure 7. Deep temperature (in situ) and salinity profiles, Southeast Pacific Basin. Lu 02 (mi/i) .__ 4.0 5.0 6.0 7.0 8.0 0j 27.60 27.70 27.S0 27.90 2.IDo

1 I I I 1 I I I I I I I 1' I j'I 1

- 6 ELTANIN411 - 63° 1'S. 135° 4'W. x ELTANIN444 - 620 4'S. 104° 57'W. 0 ELTANIN 225 - 56 4'S. 79 7'W.

7 lIIIIIIIIIIIItIIiIII!IIIIIIIIIIIIIIIIIIJ_i_II!;IIIi Figure 8. Deep slgma-e and oxygen profiles, Southeast Pasific Basin. 27 the Rio Grande Rise on the north.Wust's assumption (1933) of a relatively narrow passageway between theScotia Ridge and the Mid- Atlantic Ridge is supported by the deep minimumtemperature chart of this study (Plate II).LePichon (in press) shows a narrow erosional passageway nearly4600meters deep through the Rio GrandeRise at the northwest end of this basin.He estimates that the current velocity may be as high as half a knot, or more,with transport comparable to that of the Gulf Stream. Deep hydrographic casts were first madein this basin by the

DEtJTSCHLAND (1911), the METEOR(1925),and the DISCOVERY

(1926). More recent deep casts have been made by theCAPITAN CANEPA (1957-1959) and the ELTANIN (1963). Temperature measurements do not disclosethe presence of the deep temperature minimum in this basin exceptpossibly in the extreme southwest.The maximum horizontal gradient in the bottom water is found in the southeast sector of thebasin where the minimum temperature is less than0.20° C.in the southeast extremity and about +0.10° C. in the central part of the basin.Elsewhere the minimum temperature lies between +0.10° C. and about +0.20° C. There is an observable horizontal gradient in bottomsalinity across this basin, although it issomewhat obscured by errors in the salinity determinations.The salinity near the bottom increases from less than 34.66%oin the southeast to near34. 68%oin the 28 northwest, reflecting the varying mixture of water from the Antarctic with that from the North Atlantic and, possibly, from the southeast Pacific. Below 2000 to 2500 meters there is a sharp decrease in both temperature and salinity with depth throughout the basin (Figure 9). These steep vertical gradients are maintained by the southward flow of warm, saline North Atlantic Deep Water above the northward spreading Antarctic Bottom Water.The potential density increases downward to about 4000 meters and is nearly constant below this depth. The very steep temperature gradient near the bottom at ELTANIN station 170 appears suspect at first, especially since it is supported by but a single observation; however, this is found to be characteris- tic of the entrance to these deep basins where a deep passageway connects with a cold water source. From the oxygen minimum between 1000 and 2000 meters, the dissolved oxygen content increases all the way to the bottom (Figure 10), reflecting the predominance of water of Antarctic origin at the greater depths.

Brazil Basin

The Brazil Basin is enclosed by the continental rise of South America on the west, the Rio Grande Rise on the south, the 0 1 T (°C) -- 34.60 0.50 34.70 1.00 34.80 1.50 34.90 2.00 35.00 2.50 2 E 0 4 6 _T S - 7 - x METEOR ELTANIN 51 - 45° 58'S.170 -550 46° 22'W. 50'S. 35 5'W. (Depth: 5614m) (Depth: 5563ni) - I I I I I I I I I I 1 I I I _1_,_J___.L._ Figure 9. Deep temperature (In situ) and salinity profiles, Argentine Basin. NJ 1 2 0 4 6 02 (mI/I) _-. 4.0 5.0 6.0 7.0 a .o 27.60' I I I 27.70 ' I 27.80I I I 1 I I 27.901 I I I I I I j 28, .00

7 g METEOR ELTARIN 51 - 45° 58's.170 - 55046° 22W. 50's. 35 5W. - Figure 10. Deep sigma-O and oxygen prafiles, Argentine Basin. C 31 Mid-Atlantic Ridge on the east and interrupted ridges on the north. The basin is roughly rectangular in shape with dimensions of abot.it 1600 by 3000 kilometers at a depth of 4500 meters.The long axis is oriented north-south.The deepest connection is with the Argentine Basin through a narrow passageway entering the southwestern part of this basin.The sill depth connecting these two basins is between 4400 and 4500 meters (LePichon, in press).The deepest exitis into the Sierra Leone Basin through the Romanche Gap (0° 10 'S. 18015'W., Plate I) or, possibly, a somewhat deeper, meandering connection slightly to the south (Metcalf, 1961).The sill depth between the Brazil Basin and the Romanche Gap is about 4300 meters; that between the Romanche Gap and the Sierra Leone Basin is estimated to be about 3500 meters. Deep hydrographic casts have been made in the Brazil Basin by the METEOR (1926), the ALBATROSS (1948), the SAN PABLO (1956), the CRAWFORD (1957, 1958, and 1963), the LOMONOSOV (1959), the ATLANTIS (1959), the CHAIN (1961), and the JOHN E. PILLSBURY (1963).Temperature and salinity measurements have been taken to depths greater than 5000 meters at more than 30 stations in the basin proper.In addition, measurements at depths greater than 7000 meters have been taken in the Romanche Gap. The relatively warm North Atlantic Deep Water entering from the northeast is confined along the eastern slopes of the basin by the 32 leftward deflection of Coriolis force.Similarly, the cold, less saline (but denser) Antarctic Bottom Water flowing in from the west and southwest is deflected toward the left along the western slopes beneath the North Atlantic Deep Water producing east-west bottom gradients in both temperature and salinity.Because of the steep vertical gradients in temperature and salinity, their values at the bottom vary appreciably with the basin depths, higher values being found in the more shallow parts. A deep temperature inversion is found in the northern half of this basin except along the weStern edge, where the Antarctic Bottom Water enters.The associated minimum temperature is at a depth of 5400 meters at CRAWFORD station 1364 and shoals eastward to about 4700 meters in the Romanche Gap.The minimum temperature increases (from about 0. 650C. to about 1. 100C. ) as its depth decreases.Below the minimum, the temperature gradient rapidly approaches the adiabatic gradient (Figure 11). Over much of the deep basin, the bottom salinity is between 34. 70 and 34. 71%; along the western extremity, however, it approaches 34. 69%o, and in the northeast corner it increases rapidly to 34. 76%o. Vertically above the bottom, the salinity increases about 0. 2%c in 2000 meters (Figure 11), which is probably the steepest gradient to be found anywhere at this depth.The temperature increases nearly 2° C.(upward) over this same depth interval (Figure 11).Associated 0 1 2 a 6 34.60 34.70 34.80 34.90 I I 1.00 I III I 1.50 I 1 2.00 I I I I I 2.50 I I I T(°C) - I I I I I I I' I I ' p I [I I I T T 7 x ATLANTIS CRAWFORD 578].1364 - -18° 0'S. 29° O'W. 60 59S. 25° 8'W. (Depth: 5755u*) (Depth: 5077m) I Figure 11. Deep temperatureI (In situ) and salinityi_1i_i profiles, Brazil Basin. III II I i I I I I I II I I III I I I I II III II 34 with this vertical distribution of temperature and salinity is a potential density profile suggesting a deep unstable layer (Figure 12). This suggestion is misleading, however, and results from the failure of sigma-O to take into account differences in the compressibility of water at different temperatures and at different salinities.It will be shown in the next section that the equilibrium is stable in this basin and in the North Atlantic basins in which the sigma-e decreases or is nearly constant with depth near the bottom. There is a striking difference in the deep oxygen profiles of the Argentine and Brazil basins (Figures 10 and 12).Although the bottom

values range between about5. 0and5. 5ml/l in both basins, the gradients above the bottom are of opposite signs.The oxygen decreases upward in the Argentine Basin, whereas it increases from the deep temperature minimum to about3500meters in the Brazil Basin.This increase is a reflection of the shorter residence time of the North Atlantic Deep Water in the lower latitudes nearer its source. The coldest water from the eastern Brazil Basin is excluded from the Romanche Gap by the intervening sill, which has a depth of about4300meters; thus, the water in the deepest part of the Romanche

Gap is nearly 0.5° C.(potential temperature) warmer and 0.04%o saltier than that in the deepest part of the Brazil Basin.Above the sill depth, the temperatures, salinities and potential densities are 21 .9a gOs 4 M 6 - I iii iiiii z'r.bu 11111111 z.,.iu 11111111 IIII I 1111111111 III 9.0 00 7 -

- ATLANTIScRAWFORD 57811364 - 18° 6° 59'S. 25° 8'W. 0'S. 290 O'W. -t I i I i I t I i I I I i I i I i I I I I II I i I I I I I II I III II I II FIgure 12. Deep slgma-O and oxygen profiles, Brazil Basin. Ui nearly identical.

North American and Contiguous Basins

The North American Basin occupies a major part of the western North Atlantic Ocean.Crescent shaped with north-south elongation, it has overall dimensions of about 2500 by 3500 kilometers ata depth of 5000 meters.This large basin connects with the smaller and somewhat shallower Newfoundland Basin to the north.This and the Venezuelan Basin, which is separated below about 2000 meters from the North American Basin by the Antilles and their connecting ridges, will also be discussed in this section.The North American Basin is bounded on the west and south by the Antilles and the continental rise of North America; it is bounded on the east by the Mid-Atlantic Ridge. This ridge and the continental rise converge to formnarrow, shoaling passageways to the northeast and southeast. In 1933 Wust noted the lack of deep measurements in the North American Basin.Since then numerous oceanographic surveys have been conducted in this area,and more than 250 hydrographic casts have now been made to depths exceeding 5000 meters.The great majority of these have been from ships of the U. S. Naval Oceano-

graphic Office and the Woods Hole Oceanographic Institution.Other ships that have made casts to these depths include: the DANA (1922), the ALBATROSS (1948), the DISCOVERY II (1957), the EXPLORER 37

(1962 and 1964), and the LOMONOSOV (1959). It is evident that the bottom waters in this basin are of two dis- tinct types and sources: one of Antarctic origin flowing into the basin from the southeast with an initial potential temperature of less than 1. 3° C. and a salinity of about 34. 83%o; the other entering the basin from the north with a potential temperature near 1.80C. and a salinity of about 34. 90%o, probably of Arctic origin.As in the Brazil Basin, although the computed potential density is higher for the water entering from the north, the fact that this water overflows the Antarctic Bottom Water at points of contact clearly demonstrates that it is actually less dense than the latter. The existence of a deep temperature minimum in the eastern half of this basin is substantiated by observations wherever stations extend below its level (Figure 13).This minimum has also been observed in the deeper regions elsewhere, except in the southeastern and south central part of the basin, that is, along the axis of the Antarctic Bottom Water.Actually there are two minimum layers in this basin, the most extensive one associated with the Antarctic Bottom Water at a depth of about 5200 meters, the other associated with the Arctic Bottom Water at a depth of about 4500-4600 meters. In some areas (A.J. MYER station at 22° N., 69° W., for example), these two minima are found superposed, one at or near the bottom, associated with the Antarctic Bottom Water, and the other near 0

2 6

' I I I I 34.70 I I I I ' ' T(°C) - I J 2.00 I I I34.80 2.50 I I I I34.90 3.00 I I I I 35.00 3.50 I I' I I 35.10 4.00

T S 7 - x CHAIN ATLANTIS 350 5220 - 17° 30'N.- 29° 5615'N. 57° 30'W. 47'W. (Depth: 6065a)5500m) - I 864 - 37° 35'N. 58° 28'W. 0 CRAWFORD I (Depth: 5180m) I I I II I -- - 22°N. I 69°W. I I ii KYER I I (Depth: Figure 13. Deep temperature (In situ) and salinity profiles, North American Basin. I I I 5312m) I 39 4500 meters, associated with the overlying Arctic water. The bottom water entering this basin from the southeast has a 70 minimum temperature between 1. 6 and 1. C. This is nearly 1.00C. higher than the temperature of the bottom water in the northwest end of the Brazil Basin at about the same depth.Similarly, the bottom salinity (34. 83%o) of the entering water is more than 0. l%o higher than that in the northwest end of the Brazil Basin.Mixing of the two bottom water types produces large north-south horizontal gradients in the abyssal temperature and salinity, both increasing from the southeast end of the basin to about 30° N. (Plates II,IV, and V). North of this latitude, the temperature gradient becomes very flat, but the salinity gradient continues.Vestiges of the Antarctic Bottom

Water are evident at least as far north as360N. Below about 4500 meters the potential density (Figure 14) is nearly constant with depth.Horizontally, the potential density in-

creases from about1. 02790in the south to1. 02792in the north. There is a large increase in the dissolved oxygen content of the bottom water between 20° N. and 40° N. (Figure 14), reflecting the rapid increase in the proportion of water of North Atlantic origin in this region.Wust (1933) estimates that the proportion of Antarc- tic Bottom Water changes from 32% to 5% between these latitudes. None of the casts in the Newfoundland Basin show depths as great as 5000 meters, although depths slightly exceeding this have been 0

33S.

Figure 14. Deep sigma-e and oxygen profiles, North American Basin. 41 reported in the deepest part of the Basin.Two stations show a well defined temperature minimum at about 4300 meters in this basin. Deep water of the Venezuelan Basin has about the same potential temperature (3.84° C.) and salinity (34. 97%o)as that at a depth of 1700 meters in the Brazil Basin.The effective sill depth between the Venezuelan Basin and the North American Basin appears to be about 2100 meters.

Sierra Leone Basin

This small triangular basin is only about 900 kilometers across its largest dimension at a depth of 4500 meters.It is bounded on the northeast by the continental rise of Africa, on the northwest by the

Sierra Leone Rise and on the south by the Mid-Atlantic Ridge.In the southeast, a narrow passageway conwects with the Guinea Basin. The maximum depth in the Sierra Leone Basin appears to be about 5000 meters. Deep hydrographic casts have been made in this basin by the CHAIN (1961), the CRAWFORD (1957 and 1963), and the EXPLORER

(1963).Only one station extends deeper than 5000 meters; therefore, all stations deeper than 4500 meters have been considered. No deep minimum temperature is disclosed by the hydrographic casts in this basin, although the vertical temperature gradient decreases near the bottom.The bottom temperature lies between 2. 20 42

and 2.300c.(Figure 15), with the colder temperatures in the deeper portions in the southern part of the basin.The bottom salinity is uniformly about 34. 86%o with sharply increasing salinity above the bottom (Figure 15).The negative temperature and salinity gradients combine to give slightly decreasing potential density below about 3400 meters (Figure 16).This reflects different proportions of Antarctic water at different levels.As in the Brazil and North American basins, the water is in stable equilibrium; nevertheless, the combination of cold Antarctic Bottom Water flowing into this basin from the south beneath the high salinity water from the North Atlantic does facilitate mixing.The downward flux of salt and heat is balanced by the horizontal advection of cold, lower salinity water toward the northwest.The potential density near the bottom is between 1. 02789 and 1. 02790. The mean value of the dissolved oxygen near the bottom is higher for this basin than for the surrounding basins (Plate VII).This probably reflects sampling errors rather than real differences. The oxygen values show considerable variation.Representative oxygen profiles are shown in Figure 16.

Canary and Iberian Basins

These two basins together form an elongated, irregular channel (about 5000 kilometers long and 2000 kilometers wide ata depth of 0 T (°C) - 8(°/ao) 34.60 2.50 34.70 3.00 34.80 3.50 34.90 4.00 35.00 4.50 2 .9zE 4 6 x PILLSBURY CRAIN 317 - 00 49'N. 16° 45'W. 4 - 12° O'N. 20° ?'W. (Depth: 4841m) (Depth: SOlOm) S 7 Figure 15. Deep temperature (in situ) and salinity profiles, Sierra Leona Basin. Basin. Leona Sierra profiles, oxygen and sigma-O Deep 16. FIgure

i i i I i I I I I I I I I I I I I I I I I I I I 1 I I I I I I

7' 20° N. 0 12° - 4 PILLSBURY x - - 45'W. 16° 49'N 00 317 CHAIN - 6

0-s - 02 -

J I J I I j I I I I I I I I I I I Ii II I I I I I I I I I I I I I 0 LW 9.0 8.0 7.0 6.0 - (iuh/I) 02 45 4500 meters) between the continental rise of Europe and Africa and the Mid-Atlantic Ridge.The greatest depths are found near 24° N. and 34° W.The gradual shoaling toward the northeast is interrupted frequently by ridges and other irregular features rising above the basin floor.Thus, there are not just two, or even three, basins, but a series of basins of varying sizes and shapes.Nevertheless, there are no sharp discontinuities in the characteristics of the bottom water, but, rather, a gradual northward transition.For this reason, these and the southern, deeper part of the West European Basin are considered together in this section. The earliest deep measurements in these basins were by the PLANET (1906), the MICHAEL SARS (1910), and the DEUTSCHLAND (1911).Later more extensive measurements were made by the METEOR (1927).A single station to 5352 meters was taken in the Canary Basin by the ALBATROSS in 1948.The balance of the deep stations, and by far the most, have been taken since 1956.These include stations taken by the CRAWFORD (1957 and 1963), DISCOVERY II (1931, 1957 and 1958), the ATLANTIS (1962), and the JOHN E. PILLSBURY (1963).In all, over 70 stations were included in this analysis. A temperature minimum exists throughout most of these basins at depths between about 4500 and 5000 meters.Several of the stations show near-adiabatic lapse rates within 300 or 400 meters of the bottom.The salinity decreases sharply with depth above thetemper- ature minimum but is constant through the adiabatic layer.Temper- ature and salinity combine to give potential densities thatare nearly constant with depth below about 3200 meters.

The horizontal temperature gradient at the levelof the deep minimum is rather weak with the minimumtemperature ranging from 30 about 2. C. in the south to nearly 2. 6° C. in theextreme north. The bottom salinity ranges from 34. 87%o in the southto 34. 91%c in the

north.The salinity at the temperature minimum is about0. 0l% higher than at the bottom.The potential density of the bottom water is near 1.02791, increasing slightly towardthe north. From Figure 17 and Plates II and V,it is evident that both heat and salt are mixing downward in this basin,at least to the top of the adiabatic layer.The temperature and salinity increases at various levels between CRAWFORD stations 1335 (10°N., 25° 5' W. ) and 251 (40° 13' N., 14° W.)are shown in Table 1. The potential density (Figure 18) decreasesrather sharply upward above 3200 meters at both of these stations;thus, we see that the greatest horizontal changes take placein regions where the positive curvature (curvature conducive to downwardeddy transfer) is least and where the stability implied by thepotential density gradient is greatest.The oxygen profiles (Figure 18) indicatea rather large northward depletion below 3000meters.We might 0 T(°C) - 34.70 2.50 34.80 3.00 34.90 3.50 35.00 4.00 35.10 4.50 21 0 4 6 0 DISCOVERYX CRAWFORD II CRAWFORD 2511335 - 400- 100 13'N. O'N. 14° 25° O'W. 5'W. 3539 - 48° 34'N. 16° 44'W.(Depth: 4821m) (Depth: 5405in)5125m) FIgure 17. Deep temperature (In situ) and salinity profiles, Canary and Iberian Basins. -J 21 O2(m1/1)_.

E I I hI.UU I I I I I 1 i i h0. 8.0 0 Ue I I 1* 4 II 6 7 ci o DISCOVERY* CRAWFORD II CRAWFORD 2511335 -- 400100 13'N.6111. 14°25° O'W.5'W 3539 - 48° 34'N. 16° 44'W. FIgure 18. Deep sigma-G and oxygen profiles, Canary and Iberian Basins. Table 1.Temperature and Salinity Increases Between CRAWFORD Stations 1335 and 251.

Stat.1335 Stat.251 Increases Depth (m) T (°C) S (%o) T (°C) S (%o) T (°C) S (%o)

2000 3.38 34.96 4.04 35.07 0.66 0.11 3000 2. 67 34. 925 2. 87 34. 96 0. 20 0. 035 4000 2. 37 34. 90 2. 56 34. 925 0. 19 0. 025 4650 (mm.)2. 30 34. 885 2. 52 34. 91 0. 22 0. 025 5100 2.31 34.88 2.56 34.91 0.25 0.03 speculate that the bottom water at station 251 has arrived by downslope motion from a depth of about 3800 meters in the latitude of station 1335, has gained 0. 01% in salinity by downward mixing across the density surfaces, and has lost about 0. 3 ml/l of oxygen by oxidation processes during transit.

Guinea and Angola Basins

The Guinea Basin is here considered a shallower extension of the Angola Basin and is discussed with it because of the similarities and the spars ity of data in the former.Both basins are confined by the continental rise to the east and north and the Mid-Atlantic Ridge to the west.Whale Ridge separates the Angola Basin from the Cape Basin to the southeast.The Guinea Basin is relatively small (600 by 1600 kilometers) and irregular with its long axis orientedeast-west. 50 It connects with the larger (1500 by 2700 kilometers) Angola Basin to the south through a narrow passageway constricted by the Guinea Rise.

From the temperature profiles of CRAWFORD stations 472 and 95,it is estimated that the sill depth between these two basins is about 4100 meters.The water is warmed adiabatically as it descends from here to the greater depths of the Angola Basin. The earliest deep oceanographic stations in these basins were taken by the GAUSS (1903) and PLANET (1906).Stations were also taken by the METEOR (1925 and 1926), the DISCOVERY LI (1931), and more recently by the CRAWFORD (1952, 1957 and 1958), the SA.N PABLO (1956), the CHAIN (1961), the ATLANTIS (1963), and the GERONIMO (1963).Perhaps the earliest oceanographic description of this area was by Buchanan in 1888.Discussions of this area have been included in the writings of Wust (1933), Metcalf (1961) and others. Flowing through the Romanche Gap, the coldest, deepest current appears in the westward extremity of the Guinea Basin.On the basis of both temperature and salinity profiles, the sill depth between the Sierra Leone and Guinea basins is estimated to be about 3900 meters (2130 fathoms).This agrees well with the latest manuscript bathymetric chart on file at the U. S. Naval Oceanographic Office, which shows a sill depth between 2000 and 2300 fathoms at about00SOT N.,

16° 20'W. The deep temperature minimum is not well documented in the 51 Guinea Basin, where it nevertheless is probably present at least in the eastern part of the Basin.The minimum is observed throughout the Angola Basin between depths of 3700 and 4300 meters (Figure 19), the shallower depths occurring in the southern part of the basin. Near-adiabatic temperature gradients are found below about 4900 meters.The thickness of this adiabatic or super-adiabatic layer is probably as much as 500 meters in the deeper parts of the basin. Vertically, the salinity is nearly constant below 4500 meters (Figure 19) except in the western Guinea Basin where the influence of the Antarctic Bottom Water entering through the Romanche Gap is greatest.Horizontally, both the temperature and salinity gradients are extremely weak in these basins (Plates hand V).The salinity at the bottom is about 34. 87 throughout the Guinea Basin and near

34. 88o in the Angola Basin.The salinity averages about0. Ol%o higher at the depth of the deep temperature minimum than at the bottom. Except for some indication of slight stratification in the western part of the Guinea Basin, the potential density shows little variation below 4000 meters throughout these two basins.The slight indication of decreasing potential density with depth in the deeper parts of the Angola Basin where the salinity is constant falls within the limits of error and cannot be considered significant. 34.60 34.70 34.80 34.90 35.00 I 3.00 3.50 4,00 0 T(°C) - 'I'l'I' l'I'I'I 2.00 I IIj2.50 11111111111111 i!lI' I

T s - 0 CRAWFORDx CRAWFORD CRAWFORD 448472 - 24° 14'S.95 - 2° 30'E. (Depth: 5250i) 8°00 15'S.18'S. 4° 19'E. (Depth: 5400m) O 42'W. (Depth: 5155m) - -I I I I I I I I I I I I I I I I I I I I 1 I I I I I I I i I I Figure 19. Deep temperature (In situ) and salinity profiles, Guinea and Angola Basins. u-I N) 53 Of particular interest in the Angola Basin is the increase in temperature below 3700 meters between CRAWFORD stations 95 in the north and 448 in the south.This increase averages about 0. 05° C. through a 1500-meter layer.Profiles of dissolved oxygen (Figure 20) shows evidence of its depletion between the Guinea and Angola basins.

Cape, Agulhas and Natal Basins

The three contiguous basins which are discussed together in this section form an irregular crescent shaped depression around the southern tip of Africa.The inside arc is formed by the continental rise of Africa; the outside arc is formed by the continuum of the Mid- Atlantic Ridge and the Atlantic-Indian Rise.Whale Ridge separates the Cape Basin from the Angola Basin, and the Madagascar Rise separates Natal Basin from the Mascarene Basin to the northeast. Within this large depression, Cape Rise separates the Cape Basin and the Agulhas Basin, whereas the Mozambique Rise separates the latter from the Natal Basin.The overall length of the arc at a depth of 4500 meters is about 5000 kilometers. The earliest bottom temperature measurements used in this study in these basins were taken by the GAUSS in 1901 and the PLANET in 1906.Subsequently, deep oceanographic stations were taken by the METEOR (1925), the DISCOVERY II (1926, 1930, 1932, 1933, 1935, 1936, 1938, 1939, and 1951), the GALATHEA (1951), 02 (mi/i) 27.60 4.0 01L ______I

2 1 a 0 I. 4 6 27.70 5.0 27.80 6.0 27.90 7.0 28.00 8.0

"2 a-a 7 O CRAWFORD* CRAWFORD CRAWFORD 448472 - 24° 14'S.95 - 2° 30'S. 80 018'S. 4° 19'S. 15'S. 50 42'W. Figure 20. Deep sigma-e and oxygen profiles, Guinea and Angola Basins. Ui 55 the CAPITAN CANEPA(1957),the CRAWFORD(1958),the VEMA

(1958),the ATLANTIS (1959 and1963)and the NATAL(1963). In this region of strong horizontal gradients in temperature and salinity, it is difficult to reconcile all of the observations taken in these basins. The strongest gradients are in the Agulhas Basin.The sample profiles of Figures 21 and 22 do not represent the full range of conditions found in these basins. This is a region of intermixing of the relatively warm saline water overflowing the Whale Ridge from the Angola Basin and the cold, less saline water entering the Aguihas Basin from the Atlantic- Indian Basin to the south.Water having a potential temperature of about 2. 20° C. and a salinity of about34. 88%ofrom the Angola Basin flows over the sill at a depth of about3000meters.The sill depth between the Aguihas Basin and the Atlantic-Indian Basin is about3400 meters.Water flowing across it has a potential temperature of about -0. 30° C. and a salinity of about 34. 68%o. These source waters mix in varying proportions to produce the widely varying characteristics found in these basins. No deep temperature minimum is found in the oceanographic stations in this region, although the temperature traces in both the Cape and Natal Basin rapidly steepen near the bottom.The salinity reaches a maximum near 3000 meters and decreases rapidly below this depth.Horizontally, the minimum temperature and salinity 01 S El

' I I 34.50 ' 34.60' F F F 34.70 F F I I F 34.80 34.90 T (°C) - t 1.00 1.50 I 2.00 I I 11 2.50 1 F I I 3.0 a 0 0. 4 - T GALATREA 200 - 29° 39'S. 37° l'E. - Natal Basin (Depth: 5108m) s - - Iii liii lit iFil ililili liii I ilit tlt 0 DISCOVERYx DISCOVERY II II 26251167 - 41°36° 50'S. 18° 50'E. - Aguihas Basin (Depth: 5303m) 1'S. 6° 32'E. - Cape Basin (Depth: 5290m) liii lilt - Figure 21. Deep temperature (in situ) and salinity profiles, Cape, Aguihas and Natal Basins. lililit u-I 02 (mI/I) - 4.0 5.0 6.0 7.0 8.0 27.60 27.70 27.80 27.90 28.00

I I I I I I 1 I I I I I 1 I I II i--I j II II 1

a 0

Q5 2

6 GALATREA 200 - 29° 39'S. 37° 1'E. - Natal Basin - x DISCOVERY II 1167 - 36° 1'S. 6° 32'E. - Cape Basin O DISCOVERY II2625 - 410 50'S. 18° 50'E. - Aguihas Basin 1

I I I I I I I I I I I I I I i I -I I I I i I i I I I I I I I I I

Figure 22. Deep sigma-O and oxygen profiles, Cape, Aguihas and NatalBasins.

U-I decrease from about 1.100C. and 34. 74%&in the northeastern part of the Cape Basin to about 0.10° C. and 34. 69%o in the southern extremity of the Agulhas Basin.The salinity is about 34. 69%o in the Natal Basin and the temperature is between about 0. 50 and 0.60° C. Following the method of Wust (1933), the proportions of water from the Atlantic-Indian Basin have been computed for several points. These were found to be 60% in the northern Cape Basin, 80%in the Natal Basin, and 92% in the central part of the Agulhas Basin. Smooth curves drawn through the computed sigma-Q values of Figure 22 would show increasing density and thus stable stratification all the way to the bottom.This is consistent with the strong vertical temperature and salinity gradients.Although the computed potential density decreases near the bottom at the representative station in the Natal Basin (GALATHEA 200), stability calculations show this water to be in stable equilibrium. The oxygen profiles (Figure 22) indicate a shorter residence time for the near-bottom water of the Natal Basin than for that of the other two basins; however, this interpretation must be accepted with reservation because of inaccuracies in oxygen determinations and im- perfect understanding of its depletion processes.The oxygen content near the bottom ranges between about 4. 4 and 5. 0 ml/l in these basins.As has been noted previously, the DISCOVERY II oxygens appear to be low at great depths. 59 Crozet Basin

This somewhat diamond shaped basin is about 2000 kilometers in its north-south dimension by 1200 kilometers in the east-west axis at a depth of 4500 meters.It is bounded by the Atlantic-Indian Rise and the Mid-Indian Rise, which converge to form the northern apex, and

ridges connecting the Kerguelen and Crozet Islands to the south.It appears that the deepest passage into this basin connects with the Atlantic-Indian Basin through these two groups of islands.The sill depth of this passage is estimated from thepotential temperature and salinity to be about 2700 meters. Although the deepest part of this basin is slightly greater than 5000 meters, no oceanographic stations were found extending to this depth and only two reached below 4500 meters.These were taken by the VEMA (1960) and the ARGO (1962).In addition, one station taken by the OB (1956) reached to a maximum depth of 4455 meters. No deep temperature minimum is found on the temperature profiles constructed from data from these three stations (Figure 23) although there is a marked decrease in the vertical temperature gradient below 4000 meters in the northern end of the basin.The bottom temperature increases from near o.4 C.in the southern ° part of the basin to nearly 0. 7C. in the northern extremity.The bottom salinity increases from about 34. 68%o in the south to near 0 21 U.9za 4 0. 6

' ' I I I [ I I 34.60 I I I I 34.80 T(°C) - I - 34.40 1.00 34.50 1.50 2.00 34.70 2.50 3.00 S - T x VEMA ARGO LUS 111 - 62 - 300 21 S. 67° 15'S. ?° 45'S. 63° 58'E. (Depth: 4817m) (Depth: 4892m) -

7 I I -t Figure 23. Deep temperature (in situ) and salinity profiles,I Crozet Basin. I I I I I I I I i i t I I I I 1 C 61

34. 70%o in the north (Plate V).Vertically, the salinity increases continuously with depth below3000meters (Figure23).

The potential density near the bottom is about 1.02786(Figure24).

The mean dissolved oxygen content near the bottom is about4. 9ml/l

(Figure24). This basin shows only slight stability below4000meters.

The stability below3000meters is greater than that indicated by the potential density distribution for the same reasons advanced for the Brazil Basin and amplified in the next section.

Madagascar-Mascarene Basins

This T-shaped complex of basins has maximum dimensions of about1600kilometers (north-south) by about 1800 kilometers (east- west) at a depth of4500meters. Mascarene is the name given to the northward extension of the larger, deeper Madagascar Basin. These basins are bounded by the Atlantic-Indian Rise on the south and east, Madagascar and the Madagascar Rise on the west, and a series of islands and connecting minor ridges to the north.From the temperature and salinity profiles,it is concluded that the deepest exchange is with the Crozet Basin and takes place at a maximum depth of about4000to4200meters.

The deepest stations in these basins were takenby the VITYA.Z

(1960),the ARGO(1962)and the ATLANTIS(1963). None of these stations extend to a depth of5000meters and only five were found 02 5.0 6.0 7.0 - - 00 8.0

FIgure 24. Deep sigma-e and oxygen profiles, Crozet Basin. N.) 63 deeper than 4500 meters. To a depth of about 3800 meters, the vertical temperature gradient is very steep in these basins.Below 3800 meters it decreases considerably, then reverses itself rather suddenly below about 4700 meters.Two of the Atlantis stations show superadiabatic gradients between the bottom two observation levels in the Mascarene Basin. Because such a gradient is supported by only one measurement in each instance and because the gradient would be reduced to adiabatic or less by allowing for possible measurement error, its validity is uncertain. The deep minimum temperatureis about 0. 90° C. in the southern end of the basin and increases but slightly to less than 1. 00C. in its northern end (Figure 25).The salinity below this depth is about 34. 7l%o; it shows slight change vertically but increases somewhat northward.The mean potential density is about 1. 02786 and the mean dissolved oxygen content is about 4. 85 ml/l at these depths (Figure 26). It appears that the deep water in these basins arrives along constant density surfaces from the Crozet Basin with slightly increased potential temperature and salinity caused by vertical mixing. 01 a 0 4

T(°C) - 34.40 1.00 34.50 1.50 34.60 2.00 34.70 2.50 34.80 3.00

6 - T x ATLANTIS ARGO LUS 109 - 26° 51'S. 58° 15'S. (Depth: 5450m)219 - 23° 32'S. 50° 13'S. (Depth: 4896m) S - -I I I FigureI 25. Deep temperature (In situ) and salinity profiles, Madagascar-Mascarene Basin. II I I I I I I I I I I I I I I I I I II 0 02 (ii/ __ ,p7 17fl 5.0 6.0 7.0 8.0 ,.uu I ....'-I 'i g --.--27 9fl E 0 II' Iii III I I I -..- II 1 I I 11111 2800 0 F. 4 02 0e - x ATLANTIS ARGO LUS 109 - 26° 51'S. 58° 15'E. 219 - 230 32'S. 500 13'E. -

I i I I I I I I I I I I I I I I i I I I I I I I Figure 26. Deep sigma-O and oxygen profiles, Madagascar-MasCarene Basin. u-I Somali Basin

This small(300by1600kilometers) elongated basin is distinctly separated from the basins to the south by Madagascar, the Seychelles- Mauritius island groups and connecting ridges.From the temperature profiles, it is estimated that the maximum effective connecting depth between the Somali and Mascarene Basins is about3800meters.

This is near the maximum connecting depth(3750meters) through the passage between the Farquhar Island Group and Madagascar (about

110S and50° W)as indicated on manuscript charts on file at the U. S. Naval Oceanographic Office. Deep oceanographic stations have been taken in the Somali Basin by the ALBATROSS(1948),the ARGO(1962and1964)and the

ATLANTIS(1963). The deep temperature minimum is clearly evident in the Somali

Basin at a depth of about4700meters.The temperature gradient becomes adiabatic below about4900meters.The salinity below a depth of 5000 meters is nearly uniform throughout the basin at

34. 725%o. The average salinity of the deepest measurements at15 stations within the basin is34. 723%o. This water appears to come almost entirely from the Mascarene Basin or other regions to the south.Its deepest potential temperatures and salinities are nearly the same as those at about 3700 meters in the Mascarene Basin, 67 which suggests that this water has moved along a constant density surface from that level.A marked increase in the salinity above 2400 meters suggests that this intermediate water comes from the Arabian Sea and the gulfs emptying into it. The mean values for the deepest potential density and oxygen determinations are 1. 027855 and 4. 43 ml/l, respectively.Profiles of temperature (in situ) salinity, potential density(Cr0)and oxygen are shown in Figures 27 and 28.

Mid-Indian Basin

The Mid-Indian Basin is well defined by the Mid-Indian Rise to the west, the Amsterdam-Naturaliste Ridge to the south and the Ninety East Ridge to the east.The latter separates the Mid-Indian Basin from the Wharton Basin.Recent bathymetric soundings indicate that the shape of this basin is more rectangular than is shown on the base charts used in this study.At a depth of 4500 meters it is nearly 3000 kilometers in its longest dimension by about 1600 kilometers in width. The earliest deep oceanographic station found for this area was taken by the SNELLIUS in 1929.Since then deep stations have been taken by the ALBATROSS (1948), the OB (1957), the VITYA2 (1960 and 1962), the ARGO (1960 and 1962), the DISCOVERY (1964) and the PIONEER (1964).Several of the deep salinities reported by the 0

34.60 34.70 34.80 34.90 35.00 I ' ' I 1.50 I 2.00 I I 2.50 I I I ' I I I 3.00 I 3.50 T(°C) -- I I I I 'I I 2 0 4 111 T S 6 - 0 ATLANTISx ATLANTIS ATLANTIS 140129144 -- 00530 57'S.435 49051° 22'E.23'E. O'N. 54 S.5'E.(Depth: 5141m) (Depth: 5024m)5106m) - I II I I Figure 27. Deep temperature (in situ)I and salinity profiles, Somali Basin. I I I II I I I I I I II I I I I I I I I Basin. Somali profiles, oxygen and sigma.o Deep 28. Figure 0 4 0

28.00 8.0 27.90 7,0 27.80 6.0 27.70 5.0 27.60 4.0 (mI/i) 02 70 VITYAZ appear to be unreasonably low compared to measurements by other ships in the area. The deep temperature minimum is found throughout this basin, generally at depths between about 4000 and 4500 meters (Figure 29). An adiabatic temperature gradient is evident at several stations below about 4300 meters.There is little horizontal change in the temperature at the depth of the minimum with the temperature ranging 450 between 1. 30 and C. The Vema Trench to the west has a minimum temperature of 1.5 9° C. (reported by the DISCOVERY) at a depth of about 3550 meters (Figure 29).It is separated from the Mid-Indian Basin below a depth of 3200 meters. The salinity below the deep temperature minimum in the Mid- Indian Basin is nearly constant at about 34. 72%o (Plate V and Figure 29).The mean of the deepest measurements excluding the extremely low values reported by the VITYAZ is 34. 7l8%o.Because of the variability of the reported salinities and small number of observations, this is not considered significant beyond the second decimal place. In view of the slight vertical salinity gradient in the deep water of this basin, the potential density profiles probably reflect the stability rather accurately.These profiles show little or no displacement stability below aEout 3600 meters (Figure 30).The dissolved oxygen content (Figure 30) increases continuously with depth 0 T(0C) - 34.60 1.50 34.70 2.00 34.80 2.50 34.90 3.00 35.00 3.50 2 Q50 4 x PIONEER DISCOVERYARGO MON II 111-15 - 12° 58'S. 75° S'E. 442-020 - 5341 - 9° 10.3'S. 67° 35.7'E. (Depth: 6180m)-Vema Trench20 28.3'S. 84° 3.8'E. (Depth:(Depth: 5139in) 4544m) Figure 29. Deep temperature (in situ) and salinity0 profiles, Mid-Indian Basin. 0 1 2 0 0. 4

Izi 6

02 (mi/i) _.... a I ULI )'7'7fl4.0 . . . )7Qn 5.0 I s."ynfl I .OJ 6.0 a',)2 flA 7.0

I 7 - 0 DISCOVERYx PIONEER II 442-020ARGO - MON 111-15 - 12° 58'S. 75° 5'E. 5341 - 9° 10.3'S. 67° 35.7'E.-Vema Trench 2° 84° 3.8'E. - - - I liii iii Figure 30. Deep sigma-GI and oxygen profiles,It Mid-Indian Basin. nil I Iii iii iii I I I ii ii ii it C'.) 73 in the Mid-Indian Basin, although this increase is small below 4000 meters.This reflects the shorter residence time plus, perhaps, a lower oxidation rate in the bottom water.

Wharton Basin

The Wharton Basin connects with the Java Trench and a smaller basin to the southeast, sometimes referred to as the West Australian Basin.This latter is often regarded as an extension of the Wharton Basin as it will be in this study.The Wharton Basin is rectangular shaped and has dimensions of about 1400 by 1700 kilometers at a depth of 4500 meters, elongated to the southeast by another 1400 kilometer oval comprising the West Australian extension.Thus this system of basins and trench is bounded on the east and north by the continental rise of Australia and Indonesia, on the south by the Southeast Indian Rise, and on the west by the Ninety East Ridge. Manuscript bathymetric charts at the U. S. Naval Oceanographic Office show passageways interconnecting these basins at depths greater than 5000 meters.They also indicate that the sill depth between this system and the South Australian Basin is about 5000 meters.Because of the variability of depths within the Wharton complex, tlaere is not the consistency of temperature profiles found in some basins and the effective depth of interchange as determined from the temperature profiles is less than 5000 meters in most cases, 74 and as shallow as 4000 meters in some. The earliest deep oceanographic stations reported for this area were taken by the SNFLLITJS and DANA. in 1929.More recent deep stations have been taken by the DISCOVERY 11(1936, 1937, 1950 and 1951), the ALBATROSS (1948), the GA.LATHEA. (1951), the OB (1957), the DIAMANTINA (1957 to 1963), the VITYAZ (1959 and 1962), the VEMA (1960), the GASCOYNE (1962 and 1963), the ARGO (1962), the KOYO MARU (1963 and 1964) and the PIONEER (1964). The depth of the deep temperature minimum in the Wharton Basin is quite variable.It is present nearly everywhere between about 4000 and 4700 meters, with an adiabatic gradient below 4500 meters in part of the basin (Figure 31), somewhat deeper in other parts.Horizontally, the minimum temperature ranges between about 0. 90 and 1. 20° C.The salinity is nearly constant below 4500 meters at about 34. 71%o, or slightly higher.The potential density is also nearly constant below 4500 meters at about 1. 02785 to 1. 02786 (Figure 32).The 1. 02785 constant potential density surface slopes upward from 5000 meters to about 4200 meters in the South Australian Basin.The oxygen increases continuously with depth to about 4. 4-4. 6 ml/1. 0 1

T(°C) - - 34.50 1.50 34.60 2.00 34.'TO 2.50 34.80 3.00 34.90 3.50 2 E 0 F.. 4 6 ODIAMA14TINAx PIONEER 3-427 DIAMANTINA- 9° 0'S. 104° 16 E. 56 - 2° 58.3'S. 92° 0.9'E3-441 - 26° 41'S. 108 2'E. (Depth: 5852m) S (Depth: 5669m)4673m) Figure 31. Deep temperature (In situ) and salinity profiles, Wharton Basin. U, 02 (mi/i) -_ 3.00 I I I _.9?I I I ' I I 27.60 Z1.1U &I.ou 1 I [1I a0 4 -- 6 7

i. ., 7.00 An - 02 - ODIAMANTINAx. PIONEERDIAMANTINA 3-427 - 9° 56 - 2° 58.3'S. 92° O.9'E.3-441 - 26° 41'S. 108° 2'E. 0'S. 104° 16'E. 08 II I i I I Figure 32. Deep sigma-9 and oxygen profiles, I I I I I I I I I Wharton Basin. I C' 77 South Australian Basin

The mabasin of this system is oval shaped with a long dimension of about 2500 kilometers.It is bounded on the north by the Australian continental slope, on the south by the southeast Indian Rise and on the east by the South Tasmania Rise.At its westward end, a series of smaller closely related basins extend northwestward for about 1000 kilometers farther.Thus, the South Australian Basin as considered here has overall dimensions of about 3500 by 1000 kilometers at a depth of 4500 meters.The smaller basins have depths in excess of 5500 meters, and the main basin hasan extensive area of about this depth.These basins are interconnected at 5000 meters or deeper. The five stations from this basin used in this studywere all taken by the DIAMANTINA in 1960.The salinities of three of these stations appear to be 0. 02 to 0.03%otoo high judging from the salin-. ities of adjacent basins and the probable potential density. The deep temperature minimum is recorded at all of the stations between about 4200 and 4700 meters, with the greatest depth in the eastern part of the basin (Figure33). Adiabatic or slightly superadiabatic gradients exist below 4500 to 5200 meterson all but one of the five profiles.The lowest reported minimum (0. 82° C.) is at the western end of the main basin; the highest reported minimum is in 0 34.50 34.60 34.70 34.130 34.90 1 I 1.00 I 1.50 2.0 2.50 3.00 2 I I ( F 4..0a --- 0 Os 4 - T S - x DIMLANTINA 84 -DIAMANTINA 21 40°340 59'S. 102° 42'E. (Depth: 6857m) 7'S. 134° 11'L (Depth: 5486nt) -I I I I I Figure 33. Deep temperatureI (in situ) and salinity profiles, SouthI I I I I I I I I I I I I I I I t AustralianI Basin. I I I I I I I I 79 the most northwesterly of the smaller basins.The coldest water spreads along the southern part of the basin producing a weak north- south gradient.The salinity below 4800 meters is near 34. 7l%o throughout the basin (Figure 33) complex.The potential density below 4800 meters appears to be just less than 1. 02786 (Figure 34). In the main basin, the dissolved oxygen content increases continuously with depth to about 4. 6 to4. 7 ml/l (Figure 34).The most northwesterly station shows a deep oxygen minimum at the top of the adiabatic layer (5200 meters).

Tasman Basin

The Tasman Basin is a diamond shaped depression bounded on the northwest by the Australian continental rise, on the northeast by the Lord Howe Rise, on the southeast by New Zealand and the MacQuarie Rise and on the southwest by the South Tasmania Rise. It has a maximum length of about 3300 kilometers and a maximum width of about 1200 kilometers at a depth of 4500 meters.Although this basin has a maximum depth of about 5200 meters, it is generally shallower than the other basins considered in this study. The earliest deep stations in this basin were taken by the DANA. in 1929.Later stations were taken by DISCOVERY II (1936, 1950, 1951, and 1958), the OB (1958), GASCOYNE (1960) and other uniden- tified ships of the Commonwealth Scientific and Industrial Organizaticn. 0 1

02 (mi/i) ._ 4.0 5.0 6.0 7.0 O 'U 2 I I I I I I I I.OU I I I 4i.'JI ' I I I j I 4.OU hl.IU 1 I 1 I E -4 0 4 6 - x DIAMANTINA DIAMANTINA 8421 - 34040° 59'S. 102° 42'E. 7'S. 134° 11'E. 02 oe - -I I I II II FigureII 34. Deep sigma-O and oxygen profiles, South I I I I I I I I I I II I Australian Basin. I II I Data were made available by this organization for a series of 17

stations taken near 34° S. and 153° F. between 1962 and 1965.No stations extend as deep as 5000 meters. Only one deep station was available for this basin south of 46° S. This station shows no deep temperature minimum; however, its maximum depth is only 4380 meters.A comparison of this station with those farther north suggests that a rise extending from north- 440 west to southeast between about and460S. separates the water into two separate depressions below about 4000 meters.The minimum temperature is near the bottom and ranges between about 0. 90 and 1.000C. in the southern depression.Although not observed at all stations, the deep temperature minimum appears to be present throughout the deeper parts of the northern depression.This minimum temperature ranges from about 1.10C. in the southeast to about 1. 2° C. in the extreme north end of the depression. The bottom salinity is very near 34. 72%o throughout this basin, probably slightly less in the south and slightly more in the northern end.Above about 4000 meters, the salinity increases with height to a maximum of about 34. 73%c near a depth of 2900 meters.The bottom salinity is greater than that in any of the surrounding basins. This is a result of the shallower depth of the basin and progressive downward mixing ofthe higher salinity water.Representative temperature and salinity profiles are shown in Figure 35. 0 1

- 34.50 1 flfl 34.60 34.70 34.80 34.90 3.00 I I I II I I I I I 1 !O I I I 2.0 I I 2.50 I I I I I 0 2

- 0 DISCOVERYx II CSIROCSIRO Station Station G G2/54/62 3/62/65 - - 340330 52'S. 153° 27'E. (Depth: 4837m) 1'S. 153° 5'E. (Depth: 4837m) (Depth: 4961m) - 1683 - 46° 59'S. 155° 39'E. I I I I I I I II I I I II I -I I I I I I I I I FIgure I 35. I I Deep temperatureI (in situ) and salinityI profiles, t I I II Tasman Basin. The potential density below the temperature minimum is nearly constant at about 1. 02785, decreasing rather sharply upward, particularly above 3500 meters.Observations of the dissolved oxygen near the bottom range from 3. 15 to 4. 64 ml/l; however, it is believed that the oxygen content actually ranges more narrowly around 4. 50 ml/l.Profiles of o and dissolved oxygen are shown in Figure 36. The potential temperature, salinity, and oxygen near the bottom in this basin are nearly the same as those at about 4000 meters in the South Australian Basin.

Southwest Pacific Basin

This spacious, angular basin extends from 15° to610S.latitude and from 130° W. to 164° F. longitude.Its maximum dimensions are 3200 by 5000 kilometers at a depth of 4500 meters.The South- west Pacific Basin is bounded on the west by the Friendly and Kermadec Islands and the ridges and plateaus associated with these islands and New Zealand.The Albatross Cordillera forms its eastern and southern boundaries; the Cook Islands, Tuamotu Archi- pelago andAustral Seamount Chain interrupt its northward extension into the Central Pacific Basin.The bottom water is presumed to enter the Pacific through the passage between New Zealand and the

Antarctic continent (Sverdrup, Johnson and Fleming, 1942).The 0 1

n., nfl no nn 'V I I 5.0g.uu I 6.0'.o" I 7.0 ' ______I I 3.0 4.0 I I ... I E E;I 0 0. 4 - 02 6 - 0 DISCOVERYx CSIRO IIStation GCSIRO 3/62/65 Station - G 2/54/62 1683 - 46° 59'S. 155° 39'S. 330340 52'S. 153° 27'S. 1'S. 153° 5'E. - - -I I II I I I I Figure 36. Deep sigma-e and oxygen profiles, I I I I 1 I I II I I Tasman Basin.III I I earliest deep stations i.n this area were taken by the CARNEGIE (1929) and the DANA (1928).A large number of deep stations were takenbyDISCOVERY II (1932, 1938, 1950 and 1951).Other deep stations in this area have been taken by the GALATHEA (1952), the HORIZON (1957), the VITYAZ (1957 and 1958), the OB (1958), the

GASCOYNE (1961) and the ARGO (1961). The deep temperature minimum is well documented in the western part of the Southwest Pacific Basin (Figure 37).Stations in the eastern part of the basin are generally too shallow to sub- stantiate its existence there; however, there is a strong indication of such in the temperature profile of the most southeasterly station in this area (DISCOVERY II,station 964).The depth of the minimum is between 4500 and 4800 meters.Horizontally, the temperature increases from west to east, ranging from just under 0.60C. to just over 1.30C., with little rorthsouth gradient (Plate II). The deep salinity measurements in this basin show considerable scatter, so that only the gross features of this parameter can be discerned.Although reported salinities below 4500 meters range from 34. 59 to 34. 76%o, the actual range is undoubtedly much narrower than this.The mean of all observed values at these depths is 34. 696%o, and it is concluded that the saiirAty ranges between about 34. 69 and 34. 7l%o, with the lower values in the southwestern part of the basin (Plate V).The associated potential density is calculated 01 .9a 4 14 6 T(°C) 34.40 100 j4.50 1.50 34.60 2.00 34.70 2.50 34.80 3.00

T S x DISCOVERY II VITYAZ 3812 - 14 964 - 49° 42'S. 135°33'W.(Depth: 4734m) 49'S. 172° 56'W. (Depth: 7203m) 7 Figure 37. Deep temperature (in situ) and salinity profiles, 0 ARGO 25 - 54° 31.5'S. 177° 10'W. Southwest Pacific Basin. (Depth: 5302m) to range between about 27. 855 in the southwest and 27. 845 in the eastern end of the basin.Except in the deeper part of the Tonga. and Kérmadec trenches, the potential density shows a rather sharp increase all the way to the bottom. Near the bottom the dissolved oxygen content is between 4. 60 and 4. 8 ml/l (Figure 38) except in the eastern part of the basin where it decreases somewhat.The older DISCOVERY II measurements indicate that it falls as low as 3. 70 ml/l; however, a more recent determination by the HORIZON gives a value of 4. 55 ml/l, which seems more probable.

Central Pacific Basin

The Central Pacific Basin is bounded by ridges associated with island chains - - on the west, the Gilbert and Marshall Islands; on the east, the Line Islands; and on the south, the Society Is lands.On the north it is bounded by the Mid-Pacific Mountains lying between Wake Island and the Hawaiian Islands near 20° N.This basin is irregular in shape with a maximum dimension of 5500 kilometers in the northwest-southeast direction by about 2000 kilometers in width at a depth of 4500 meters.There appears to be free interchange of water below about 4400 meters with evidence of a barrier b e low this depth separating the southwestern sector from the rest. The earliest deep station from this basin which was used in this 1

I I I I I I I z.1.0I I I I I I Jo 27.60 I Z7.i0 I I 1 7.0 I 1 1 8.0 I 4.0 I I I 5.0 I 6.0 1* 4 02 (mi/I) _- ' I I 0. ni 02E? 67 - x DISCOVERY II VITYAZ 3812 - 14° 49'S. 172° 56'W. 964 - 49° 42'S. 135° 33'W. I I I I II Figure 38. Deep sigma-O and oxygen profiles, Southwest Pacific Basin. I o ARGO 25 - 540 31.5'S. 177° 1O'W. I I I I I I I I I I I I study was taken by the CARNEGIE in 1929.Subsequently, such stations have been taken by the ALBATROSS (1947), the STRANGER (1956), the VITYAZ (1957), the ARGO (1960), the REHOBOTH (1961), the GASCOYNE (1961) and the DAVIS (1964). Most of the stations in this basin which adequately sample the depth to 5000 meters show a temperature minimum at about 4600 to 4800 meters (Figure 39).The temperature at this depth ranges between 1. 2 and 1. 3° C.Below 5000 meters, the gradient steepens but remains slightly less than adiabatic at all stations, even showing signs of decreasing somewhat near the bottom at two stations (REHOBOTH stations 5 and 16).The temperature below 4700 meters at these two stations is about 0. 05° C. warmer than that in the southwestern part ofthe basin.This is believed to be caused by adiabatic warming rather than geothermal heating, although the latter could not be positively ruled out on the basis of the observations at hand. The observed salinity values in this basin are extremely van- able, ranging from 34. 69 to 34.93%°,the latter and other anomalously high values rather obviously being in error.It is concluded (somewhat arbitrarily) from the statistical distribution of observations that the actual salinity falls between 34. 70 and 34. 71%o below 4500 meters in this basin. From these values of salinity and the corresponding potential temperatures, the potential density was then computed to be about 1. 02784 below about 4700 meters.The dissolved oxygen E 0 0 0, El ri

T(°C).8(7,,,) - 34.60 1.50 34.70 2.00 34.80 2.50 34.90 3.00 35.00 3.50

6 o REBOBOTUx REHOBOTH ALBATROSS 16 - 70 9o'5 -133 3°160° 48'S. O'W. 179° 55'W. (Depth: 5650m) 50 O'N. 172° 2'W. (Depth: 5390m) (Depth: 5879m) Figure 39. Deep temperature (in situ) and salinity profiles, Central Pacific Basin. 91 content below 5000 meters ranges from about 4. 5 ml/l in the southwest sector to about 4. 1 ml/l in the nortIast corner of the basin (Figure 40). Potential density, potential temperature, and salinity of the bottom water entering the southwest corner of this basin are almost identical with those at a depth of 4300 or 4400 meters in the northern part of the Southwest Pacific Basin (compare REHOBOTH station 9 and ARGO station 14).It appears that this water moves downslope along the 27. 84 sigma-O surface with little modification, including only a slight decrease in dissolved oxygen content.

Northeast Pacific Basin

This extensive basin ranges from near the equator to the Aleutian Trench at about 50° N. latitude.It is bounded on the east and north by the continental rise bordering the North American Continent and the Aleutian Islands.The broken western boundary is formed by the ridges associated with the Emperor Seamount Chain and the Hawaiian Rise in the southwest.The southeastern boundary is rather indefinitely formed by the gradually southeastward decrease in depth in the central Pacific.The basin roughly has the shape of a parallelogram with sides 3500 and 6000 kilometers in length at a depth of 4500 meters. 5.00 600 28.1 7 .00 DO 27.70 I I I 27.S0 I 27,90I ' 1 1 I I I O2(m1fl).. I I I ' I I I lj [I J 3:aI. 0 - - - Ø'k02 A - - 0 REBOBOTHx REHOBOTH ALBATROSS 16 - 70 0'S.9133 1600 0'!. 30 48'S. 179° 55'!5° O'N. 172° 2'!. tI II I 1 I I t FigureI iii 40. I Deep slgma-e and oxygen profiles, I I I I I I I I II i I Central Paelflc Basin. t I I I I ii I I I i 0 93 The earliest deep station in this area was taken by the DANA (1928) followed closely by the CARNEGIE (1929).Other deep stations were taken later by the ALBATROSS (1947), the BROWN BEAR (1954, 1957 and 1958), the HORIZON and SPENCER F. BAIRD (1955, 1956, 1960 and 1961), the REHOBOTH (1956, 1959, 1961 and 1963), the VITYAZ (1957-1959), the ARGO (1961), the PIONEER (1961 amd 1962), the AGASSIZ (1963), the SURVEYOR (1963 and 1964), the DAVIS (1964-1966) and the YAQUINA (1966).As is evident from the foregoing list of ships, the basin has been well sampled within the accuracy of existing techniques and practices. The deep temperature minimum is documented throughout this large basin at a depth of about 3900 meters.The temperature at the minimum is between 1.45 and 1. 50° C. throughout most of the basin, falling to near 1.350C. south of the Hawaiian Islands, where water from the Central Pacific Basin first flows into this basin. Example profiles (Figures 41 and 42) were selected from ten deep stations taken by the author from the YAQUINA during June of 1966. Nansen bottles and reversing thermometers were spaced 100 meters apart near the bottom of these casts and 250 meters apart over the remainder of the deep casts.Smooth curves could be drawn to within 0.020C. of nearly all temperature values.The two stations chosen as examples are the most southerly and northerly; theyare representatives of the other stations of this series.As can be seen from these examples, the temperature gradient decreases gradually and continuously with depth to produce a sweeping broad curve through the temperature minimum where it reverses and approximates the adiabatic gradient below about 4700 to 5000 meters.The decrease in curvature with latitude should be noted.This largely results from the latitudinal decrease in temperature at the shallower depths. Average temperatures, salinities, and potential densities have been computed from the values interpolated at standard depths for 13 stations taken at nearly the same location by the REHOBOTH. These stations are between 28° and 29° N. and 176° to 177° W.The result- ing profiles together with the ranges are shown in Figure 43.The temperature profile agrees well with those from the YAQUINA. The salinity increases continuously with depth to about 4500 meters, below which it remains nearly constant between 34. 69 and

34. 70%o.By comparison with other salinity measurements in the area, the salinities for the Hawaiian-Adak leg of the YAQUINA (1966) cruise were found to be low by about 0.0Z%o. The salinity profiles in Figure 41 have been adjusted by this amount. The potential density increases rather sharply to about 4500 meters, below which there is little change (Figure 42).The deep potential density ranges from about 1. 02782 to 1. 02783 in the southern part of this basin and from about 1. 02781 to 1. 02782 in the northern part.The mean oxygen T(°C) - 34.40 1.50 34.50 2.00 34.60 2.50 34.70 3.00 34.80 3.50 2 .2I..E 4 6 Figure 41. Deep temperature (In situ) and salinity profiles, Northeast Pacific Basin. Ui 0 02 (m1/) 1

2 E z..Du Z-I.-w Z1.ft) 0 p1 4 ru . 6 YAQUINA HAH-56H&H-23 - 5O26° 37.8'N. 161° 31.4'W. 27.S'N. 176° 13.8'W. r Figure 42. Deep slgma-O profiles, Northeast Pacific Basin. 27.50 27.60 27.70 27.80 27.90 34.30 34.40 34.50 3460 34.70 T(°C) . 2.00 2.50 3.00 3.50 0 I I I I I I i 1 I j i I ' i 'I' II I I ' I 1i 1!' 'I I 'I 'I I I

1 I - _____1______.______1______I -

2- I- -

I 4

6 x REHOBOTH (13 casts, 1959-1961) - 28°-29°N. 176°-177°W. (Depth: 530O)

-I III I I I II II I I I I I I I IIII III I I i I I I IIIII I III III I I

FIgure 43. Mean profiles of deep temperature (In situ), salinity and sigma-a from 13 repeated casts, Northeast Pacific Basin. '0 below 5000 meters in this basin is about 3. 6 mi/i; at the temperature minimum the oxygen content averages about 3. 35 mi/i.After elimi- nation of the 10% highest and iowest values, the range in observations at the greater depths is from 3. 35 to 3. 90 ml/i.The dissoived oxygen content is slightly iower in the central part of the basin; however, the horizontal variation is small.

Northwest Pacific and Mariana Basins

Inasmuch as the deep water in the Northwest Pacific Basin has essentially the same characteristics as that in the Mariana Basin, these two basins are discussed together, even though they are par- tially separated by the Mid-Pacific Mountains.These basins are bounded by rises associated with the Emperor Seamount Chain on the east and with Kamchatka, the Kuril Islands, Japan, and the Mariana Islands on the west,These rises converge northward forming an apex southwest of the Komandorskiye Ostrova, which separates the Kuril and Aleutian trenches.The Mariana Basin is bounded on the south and east by the islands and associated rises of Micronesia.These combined basins are roughly in the shape of an arrowhead with a length of about 5200 kilometers and a width of about 4000 kilometers at a depth of 4500 meters, The earliest deep stations available for this area were taken by the CARNEGIE in 1929.Ships of the Japan Maritime Safety Board took several deep stations in the l930's and again in the late 1950's and early l960s.Ships of the Japan Meteorological Agency took several deep stations in the Northwest Pacific Basin in 1960 and 1961. Deep stations have also been taken by the VITYA.Z (1957 and 1959), the REHOBOTH (1958 through 1963), the SPENCER F. BAIRD (1962), the BERING STRAIT (1964) and the DAVIS (1964). The deep temperature minimum is found throughout these basins except in the southeast corner where the water appears to enter from the Central Pacific Basii.As illustrated in Figure 44 the temperature profiles are represented by broad sweeping curves below 2000 meters.The temperature minimum is found between 4000 and 4500 meters.Below 5500 meters the vertical temperature gradient is generally adiabatic.The horizontal temperature gradient is very flat throughout these basins with the coldest water of slightly less than 1. 40° C. entering from the Central Pacific Basin to the Southeast.The highest deep minimum temperature is about 1. 55° C. The bottom salinity varies from just less than 34. 69%o in the Kuril,

Japan, and Mariana trenches to about 34.70%oin the south central and southeastern part of the basin.The salinity decreases upward and is as much as 0. 01 to 0. O2%o less at the depth of the temperature minimum than it is at the bottom.Because of the great variability in the salinity observations, attributed largely to measurement error, profiles (Figure 44) are averages for several stations.The 34.40 34.50 34.60 34.70 34.8r! I I I 3.00 3.50 1.50 I I I 2.50 I I ' I 'I I I j I I 2.00 I I I T(°C) - I 'I' I 'I I I I TI I I 2 - - 0 x 6 - 0 REHOBOTHx BERING (7 STRAIT casts, (3 1961-1963) REHOBOrEcasts, 1964) (8 - casts,- -52°N. '-.34° 1958-1960)161°E. 5'N. 164° - 30'E.T'% (Depth: 5850m) 18°-20°N. 186°L (Depth: 5320n)6-7000a)0\\x S I I i I I I i Figure 44. Mean profiles of deep temperatureI and salinity, NorthwestI Pacific I i I I I II I III III I II i I i 1 i I I II I Basin. i I i I C 101 possibility that some of these differences reflect temporal variations cannot be entirely discounted; however, the vertical density gradient does not support arguments in favor of such large variations. The bottom potential density ranges from less than 1. 02781 in northern end of the Kuril Trench to over 1. 02782 in the south central and southeast sector of this basin complex.The potential density decreases continuously with height above the bottom (Figure 45). Reported observations of dissolved oxygen content near the bottom range from 2. 91 to 5. 11 ml/l; however, the range is probably much narrower, probably between about 3. 5 and 4. 0 ml/l. The oxygen content appears to increase downward below the minimum between 500 and 1500 meters all the way to the bottom (Figure 46). Even though there is a rather large number of deep observations in these basins, the accuracy of many is questionable, particularly regarding the salinity values; thus, the confi4ence level in the values derived for these basinsisnot high.Additional, carefully controlled measurements are needed.

Philippine Basin

The Philippine Basin lies between the Ryukyu and Philippine Islands on the west and the Mariana and Volcano Islands on the east. The ridges of these islandchains converge to the north and south forming a diamond.-shaped basin about 2000 kilometers in length 1 S 6

02 (uui1/1) I I I I I I I I Z7. 90 I II I Ij II Ii 'I' liii - 0 - - x BERING STRAIT (3REHOBOTH (8 1958-1960) - 1964) - - 18°-20°N. 166°E.340 5'N. 164° 30'E. 0 x - - 7 - 0 REHOBOTH (7 casts, casts, 1961-1963) - - 52°N. 161°E. 0 - I iii Iii I Figure 45. Mean profiles of deep I I I I I ii ii I slgma- I I liii Northwest Pacific Basin. Ii I! I I I I liii 0 01 a 0 04 4

05 6

7 Figure 46. Deep oxygen profiles, Northwest Pacific Basin. 0 (J 104 and 1600 kilometers across. The first deep stations available for this area were taken by the PLANET (1922-1924); however, each of these three stations (one in each year) sampled only one level (the bottom) deeper than 1000 meters.The SNELLIUS obtained several deep stations in this area in 1929 and 1930.Ships of the Japan Hydrographic Department and the Japan Maritime Safety Board took deep stations in the 1930Ts. Ships of the latter took one deep stations here in 1959.Other deep stations have been taken by the ALBATROSS (1948), the GALA.THEA (1951), the VITYAZ (1957), the SPENCER F. BAIRD (1962), ai the

DAVIS (1964). A temperature minimum is found throughout the basin at a depth of about 4000 meters.The temperature at the minimum ranges between about 1. 5 and 1. 6° C.Thus, there is only a slight horizontal gradient within this basin at this level.Below the temperature minimum, the gradient steepens quickly and appears to be adiabatic throughout most of the basin below about 5000 meters (Figure 47). Near the bottom the salinity is about 34, 69%.It decreases very gradually with height above the bottom (Figure 46) and shows no clear pattern of change horizontally within the basin.The potential density is nearly constant at about 1. 02780 below 5000 meters, decreasing with height above that depth (Figure 47).The dissolved oxygen content shows little change with depth below 4000 meters but 0 1 34.60 34.70 34.80 34.90 I 35.00 2 T(°C)-- I11111111!I 1.50 I I 2.00 It I jT 1111(11 2.50 11111! 3.00 liii II 3.50 B 0 gi 4 6 x VITYAZ 3746 - 190 15'N. 134° 15'E. (Depth: 5827w) i I i I i I FigureI 47. Deep temperature (In situ) I I I III I I I and salinity I I I I profiles, Philippine Basin. I I II 0 U.' 106 decreases rather rapidly with height above that depth (Figure 48). Near the bottom it probably ranges between about 3. 3 and 3. 6 ml/l although the reported values show a much greater range than this. 1

2 .2a 4 6 02 (mi/i) __ 27.60 1.0 27.70 2.0 27.80 3.0 27.90 4.0 28. 5.0 00

7 - z VITYAZ 3746 - 19° 15'N. 134° 15'E. - -1 I i 1 i I I I t I Figurei 48. Deep sigma-e andI oxygen profiles, PhilippineI Basin. I I I I I I I i I 002 I I I i I I II C III.DISCUSSION AND CONCLUSIONS

It is rather tempting to attribute the near bottom curvature of the in situ temperature profiles to geothermal heating; however, except in the trenches and the more stagnant deep basins, the evidence of geothermal heat flow in the circulation of the bottom water is over-shadowed by other processes.Primarily, these are: 1) adiabatic warming accompanying downslope flow; 2) mixing and advection approximately along constant potential density surfaces; and 3) vertical eddy flux of heat and salt. Adiabatic processes are those which take place without loss or gain of heat or, more precisely, without change in entropy.Pres- sure and temperature relationships for such processes in ideal fluids are derived by expressing the total change in entropy by partial derivatives and setting the sum of these equal to zero.The derived equation is:

8T av avT C 3PEC/Tp p where: F = entropy; T = temperature; P = pressure; v = specific volume; acoefficient of isopiestic thermal expansion; and

C= isopiestic heat constant. p In order to eliminate temperature changes due to adiabatic heat-

ing,it is convenient to introduce the potential temperature, 0.The 109 potential temperature is the temperature water would have if it were decompressed adiabatically to a pressure of one atmosphere. Using the new values for the specific heat of seawater at atmospheric pressure determined by Cox and Smith (1959) and Fkman's (1908) data on the compressibility of seawater, Crease and Catton (1961) have prepared new tables of the adiabatic cooling of seawater. These tables or an equation derived from them have been used in adiabatic computations in this study.The following polynomial expression was derived by least squares fit to the data of the tables:

AQ= 3.49 x JO TZ + 9. 89 x Z2- 2. 13 Z3+ 7. 79

x l0 T2Z2+ 9. 28 x l0 SZ

where:/Q= T - 0; T = in situ temperature(°C); Z = depth of temperature measurement (kilometers);

S = salinity(%o);and 0 = potential temperature (°C).

A super-adiabatic temperature gradient accompanied by constant salinity does not indicate absolute instability but rather potential instability released by vertical, adiabatic displacement of the parcel under consideration.Critical instability, which is associated with an autoconvective lapse rate (terms used in meteorology) is a condition in which the in situ density decreases with depth (pressure). Expressed in terms of specific volume, the limiting case for critical instability is 0.By partial differentiation of the specific 110 volume (v) with respect to temperature (T) and pressure (P) and introduction of the known coefficients of isothermal compressibility and isopiestic thermal expansion through Maxwell's equations, this 50 lapse rate is found to be about 5. c/loometers.In contrast, the adiabatic lapse rate doesnt exceed 0. 2° C/bOO meters in the deep ocean basins. The important point here is that, even with constant salinity, lapse rates between the adiabatic and autoconvective are unstable only in the presence of vertical displacements.Such lapse rates might be expected in the absence of vertical motion where bottom heating is present.They are found in the lower atmosphere under special conditions and are thought to give rise to dust devils (Byers, 1944). Adiabatic processes are illustrated in Figure 49.From the example shown, it can be seen that where the lapse rate is sub- adiabatic, mixing causes a transfer of heat downward.Mixing or overturning of a layer of water having a superadiabatic lapse rate transfers heat upward. Molecular conduction takes place without displacement of the molecules to levels of higher or lower pressure; thus, molecular conduction is a function of the in situ rather than the potential temperature gradient.In the case of eddy conduction, the displacement of the interacting parcels is finite.That conductive heat ii

Ad.iaba.ts

U t LEVELS\ \ LAPSE RATE DISPLACNT RESULTS \ \Temperature mint____ minn A - - -- From A to B Particle is warmer than Sub-adiabatic its surroundings; particle gives up heat (cools).

I From B to A Particle 18 colder than its surroundings; particle - - - - B - - -- absorbs heat (warms). \ Adiabatic From B to C Particle is the same temper- 0 or C to B ature as its surroundings \ for either displacement.

- C I From C to D Particle is colder than its surroundings; particle ab- Super adiabatic sorba heat (warms). II --- D - - From D to C Particle is warmer than \ its surroundings; particle T \ gives up heat (cools).

I-

Figure 9. Adiabatic processes. 112 transfer (molecular) in the ocean is not significantcan be seen from the following example.Since the thermal conductivity ofsea water is about 1. 45 x cal/°Ccm sec at 6000 meters, to remove the geothermal heat from the sea floor by conductionwould require a temperature gradient of about1000C/km. In summary then, for a condition ofconstant salinity the adiabatic lapse rate is the boundary condition betweenupward and downward heat transfer by eddy conduction andoverturning, whereas the in situ temperature minimum (or maximum) isthe boundary between upward and downward molecular conduction.Eddy conduction and overturning will convey heat downward above levelB and upward below level C in the example.If these are the only processes operat- ing, geothermal heat can be carried upwardonly along superadiabatic paths.

Although superadiabatic gradientsmust exist at the bottom in order for geothermal heat to flow out of thesediments, verification of such gradients is difficult with presentmeasurement capabilities. Kuenen (1942) concluded that the temperaturegradient in the depths of the Celebes Sea and elsewherewas less than adiabatic; however the adiabatic tables of Ekman (1914) inuse at that time gave somewhat larger adiabatic gradients than thenewer tables of Crease and Catton (1961).Many of the deep temperaturemeasurements yield gradients greater than the adiabatic, but themajority of these are 113 greater by less than the probable error of measurement.In an effort to detect such superadiabatic gradients, data from the1966 YA.LOC cruise of the YAQIJINA have been averaged.For these measurements thermometers were spaced at 100 meter intervals near the bottom and at 250 meters above 300 or 400 meters.Great care was taken in reading the thermometers.Gradients were computed at reasonable intervals below 4000 meters and plotted against a curve of adiabatic gradients constructed from the tables of Crease and Catton (1961).The results, shown in Figure 50, indicate superadiabatic gradients below5500meters; however, because of the small number of observations at the greater depths, this cannot be considered conclusive.There are undoubtedly temporal variations in such gradients but the magnitude and time scale of these cannot be determined from the limited number of suitable measurements available.Many more such observations are needed, including thermograd measurements with good depth indicators. Another conservative property which takes into account differences in salinity is the potential density.This is the density that a parcel of water would have if it were decompressed adiabatically to a pressure of one atmosphere (Wust,1933). Aspreviously mentioned, Hesselberg and Sverdrup(1915)pointed out that thep0- tential density only indicates qualitatively whether the equilibrium is positive, negative, or indifferent in the absence of a vertical 114

TEMPERATURE GRADIENT (°C/km) 0.05 0.10 0.15

I 8

[1 3

7 ® Means for YAQUINA-HAH stations (Numerals give number of measuremen 10 and 20curves based on tables by 8 Crease and Catton (1961).

Figure 50. Potential temperature gradient vsdepth. 115 salinity gradient.Because of the variation of the compressibility of water with temperature and, to a lesser extent, with salinity, the equilibrium of a water column is not reflected in the potential density distribution in the presence of vertical salinity gradients. In place of the potential density,Hesselberg and Sverdrup (19 15) proposed the use of a stability function (F) given by the following

equation:

F ---( L\+ - Q dZ aT \dZdZ) asdZ T dZ 8S dZ

where:Z = depthinmeters (positive downward) S = salinity in %

T = temperature in0C. p = density d= the adiabatic lapse rate inC/meter potential temperature in°C.

Hesselberg and Sverdrup provide tables for obtaining values of

and as functions of depth, temperature and salinity.Using these and gradients measured at depths of interest from temperature and salinity profiles, the stability parameter, F, can be computed. This was done by Schubert (1935) for several stations in the Atlantic and has been done in this study for near-bottom portions of selected profiles in the major deep basins.The results are shown in Table 2. 116

Table 2.Stability Function, F, for the Near-Bottom Water in the Major Basins of the World Oceans.

Approximate depth 8 Basin (meters) F x 10 Atlantic-Indian 5000 0. 5 South Indian 4500 7. 0 Southeast Pacific 5000 -1.0 to2.0 Argentine 5500 5. 0 Brazil 5000 3.0 to10.0 North American 5000 0 to 4. 0 Sierra Leone 5000 3.0 Canary and Iberian 5500 0. 5 Angola 5000 -0. 5 Cape 5000 3. 5 Natal 5000 5.0 Crozet 4700 3.0 to11.0 Madagascar-Mascarene 5000 5.5 Somali 5000 1.0 Mid-Indian 4800 -1. 5 Wharton 5000 0.0 South Australian 6000 0. 0 Tasman 4500 2.0 Southwest Pacific 5000 0. 5 to1. 5 Central Pacific 5500 1.0 Northeast Pacific 5-7000 0. 0 to-0. 5 Northwest Pacific 5-6000 0. 0 to2. 0 Philippine 5000 1. 0 117 In addition, F values have been computed for several depths below 2500 meters for CRAWFORD station 1364 in the Brazil Basin (Table 3) in order to show that this water is in stable equilibrium even though the potential density distribution suggests otherwise.Values of the stability function, F, (Table 2) also show that the near-bottom water in the basins of the North Atlantic Ocean is in stable equilibrium even where the potential density distribution is constant or decreases with depth.Negative values appear in this table only where the salinity is essentially constant with depth. In order to examine the gross movement and modifications of the bottom water for the oceans as a whole salinity and oxygen values have been plotted against potential temperatures for each of the major basins (Figures 51 and 52).It is observed that in general the potential temperature increases with distance from the Weddell

Sea.Thus, the temperature of the bottom water in the Antarctic basinsisprogressively warmer proceeding from the Atlantic-Indian Basin (A-i) eastward to the Southeast Pacific Basin (P-l).In each of the three oceanic extensions, the bottom water is increasingly warmer with distance from the Antarctic.The warmest Antarctic Bottom Water is found in the North Pacific (P-5, P-6 and P7) with the next warmest in the northern basins of the Indian Ocean (N-6 and N-7).Still warmer water of Arctic origin is found in the North Atlantic and Angola basins (A-4, A-5, A-6 and A-7). 118

Table 3.Stability Function, F, at Various Depths in the Brazil Basin.

Depth (meters) F x108

2900 6. 1

3395 8. 9

3892 17.9

4392 9. 1

4891 4.6

5391 2.5

5740 2.8 TEMPERATURE (°C) 0.00 1.00 2.00 J42'

34.70

34.80

.Oo

Z 34.90- ATLANTIC

34.65

34.70 ___ INDIAN

34 70 PAC IF IC 3475 I I I I Figure 51. Potential temperature vs Salinity, major deep ocean basins. PO2NTIAL TN2URE (°c) -1.00 0.00 1.00 2.00

I I I

Letter-number ebinations A-6 designate basins (see Plate I). A-7 A-i

A-5b$A-5a -S A-3 A-2 5.0 .N-2, N-3 , A-8 N-i . P-i N-9. N-8b P-2, N-8a. N-6 S P-3 I

S p-6 P-i.

-I I I

Figure 52. Dissolved oxygen vs temperature, major deep ocean basins. 121 The dissolved oxygen content shows somewhat more scatter and an inverse relationship, decreasing with increasing potential temperature and distance from the Weddell Sea and Antarctic generally. Except in the Brazil, Angola and North Atlantic basins, the dissolved oxygen content generally increases at depth essentially all the way to the bottom.Thus, in the Pacific, Indian and most of the South Atlan- tic basins there is a decrease in oxygen in the direction of flow caused by downward mixing of water with a lower oxygen content coupled with depletion by oxidation of organic material.Processes con- trolling the oxygen content are more complex north of300S. latitude in the Atlantic Ocean. Mean values of the dissolved oxygen content near the bottom are shown for the major deep basins in Plate VII.Oxygen values reported by DISCOVERY II in the Cape-Aguihas-Natal basin complex are sig- nificantly lower than values reported by other ships and were omitted in the computation of the mean value shown (4. 9 ml/1).By including the DISCOVERY II data, the mean oxygen content is reduced to 4. 7 ml/l. By constructing mean profiles and horizontal gradients of oxygen, salinity and potential temperature, we can theoretically approximate the rate of flow, the rate of oxidation and the vertical mixing coefficient.This assumes that this coefficient is constant throughout the layer considered, that it is the same for oxygen, 122 salinity and temperature, and that vertical mixing is balanced by horizontal advection and depletion (oxidation) or generation (heat flow).This would be a very gross approximation and will not be carried out here. It should be noted that w}reas the salinity near the bottom increases with distance from the Antarctic in the Atlantic and to a lesser degree in the Indian Ocean,it decreases northward in the Pacific. As previously stated, the anomalously high potential temperature, salinity and oxygen in the Angola and North Atlantic basins are associated with bottom water of an entirely different origin than that in the other basins. The minimum volume of water that must be transported in order to remove the geothermal heat added below 4600 meters has been computed for several basins.This was done by multiplying the area of the basin below 4600 meters by the rate of geothermal heat flow (1. 4 x 1O cal/cm2sec) and dividing by the maximum temperature increase across the basin.That is, AxHT where, A = area of the basin incm2 H = geothermal heat flow (1. 4 x1O6 cal/cm2sec) 123 = maximum potential temperature difference acrossthe basin in °C. T = transport in 12Sverdrup (1 Sverdrup = cubic meters/s e c) This assumes that the water enters the basin at the coldest temperature and leaves at the warmest temperature and that the only heating is through the bottom.Downward flux of heat from above, flow paralleling the isotherms, entrance of water at a higher temperature and exit at lower temperature all increase the transport required to maintain a heat balance.The results of these computations are given in Table 4. In the South Atlantic the computations give a required transport ranging from 0. 05 to 0. 3 Sverdrup.This compares with an estimated transport of bottom water of 1 to 3 Sverdrup between the equator and 30° S. latitude in the Atlantic as given by Sverdrup, Johnson and Fleming (1942).It is seen that in this area where the volume of transport of bottom water is best known, the proportion required to remove the geothermal heat under the conditions assumed is at least 10% of the total. It is readily apparent from Table 4 that the largest minimum transports are found in regions where the deep temperature minimum is highest above the bottom(in the Central and Northeast Pacific, for example).In these regions flow is sluggish and the 124 goethermal heat is transported upward through a thicker layer.

TabJe 4.Minimum Transport Required to Remove Geothermal Heat by Advection. Minimum Area Transport Basin (1016 cm2) (Sverdrup)

AtlanticIndian 5. 9 0. 3

Southeast Pacific 3. 8 0. 1

Argentine 3. 1 0. 1

Brazil 3. 1 0. 1

Guinea and Angola 4. 1 0. 3

Cape and Agulhas 3. 6 0. 1

Mid.Indian 3. 3 0. 2

Wharton 5. 2 0. 2

Southwest Pacific 14. 0 0. 3

Central Pacific 8. 1 0. 8

Northeast Pacific 18. 0 1. 3

It is evident that from existing measurements of temperature and salinity only general conclusions can be drawn concerning the part that geothermal heat flow plays in the abyssal circulation.Spacing of bottles and thermometers in most deep casts is generally too great to determine the vertical temperature distribution with the precision required.Closer spacing of bottles and the use of thermograd 125 recorders with depth indicators are required to document the small but significant temperature variations found in the deep basins. Errors in tne determinations of salinity and dissolved oxygen content must be substantially reduced, if temporal fluctuations in these properties at great depths are to be studied.Substantial improvements are possible by consistent use of rigorous techniques now available. 126

IV. BIBLIOGRAPHY

Anderson, E. R.1964.Single-depth charts of the world's ocean basins at depths to 3500 fathoms, set 2.San Diego.90 charts. (U. S. Navy Electronics Laboratory.Research and development report 1252)

Brennecke, Wilhelm.1921.Die ozeanographischen Arbeiten der Deutschen Antarktischen Expedition 1911-1912.In:Archiv der Deutschen Sweewarte.Vol. 39.Hamburg, Hammerich and Lesser.216p.

Buchanan, J.Y. 1888.The exploration of the Gulf of Guinea. The Scottish Geographical Magazine 4:177-200 and 233-251.

Budyko, M. I.1956.The heat balance of the earth's surface. Leningrad, Gidrometeorologicheskoe izdatel'stvo.259 p. (English translation by Nina A. Stepanova, U.S. Weather Bureau)

Byers, Horace R.1944.General meteorology.New York, McGraw-Hill Book Company.645p. Cox, R. A. and N. D. Smith.1959.The specific heat of sea water. Proceedings of the Royal Society, Ser. A, 252:51-62.

Crease, J. and Diana Catton.1961.Tables of adiabatic cooling of seawater.Wormley, England, National Institute of Oceanography. 11p.(National Institute of Oceanography.Internal report no. A 15) Deacon, C. F. R 1933.A general account of the hydrology of the South Atlantic Ocean.Discovery Reports 7:l71-38.

1937.The hydrology of the Southern Ocean. Discovery Reports 15:1-124.

Defant, Albert.1961.Physical oceanography.Vol.1. New York, Pergamon Press, Inc. 729p. Dietrich, Gunter.1957.Aligemeine Meereskunde, eine Einfuhrung in die Oceanographie.Berlin, Gebruder Borntraeger.492p. 127

Drygaiski, Erich von.1903.Deutsche Sudpolar-Expedition auf dem Schiff Gauss.Heft 5.Berlin, Instituts fur Meerskunde und des Geographischen Institus an der Universitat Berlin. 181p. Eckart, Carl.1958.Properties of water, part II.The equation of state of water and sea water at low temperatures and pressures.American Journal of Science 256:225-240.

Ekman, V. W.1908.Die Zusammendrueckbarkeit des Meerwassers nebst einigen Werten fuer Wasser und Quecksilber.Publications de Circonstance du Cons eil Permanent International pour l'Exploration de la Mer 43:5-47.(U. S. Naval Oceanographic Office.Translation 334.1966.43p. 1914.Der adiabatische Temperaturgradient in Meere.Annalen der Hydrographie and Maritimen Meteorologie 42(6):340-344.(U. S. Naval Oceanographic Office.Translation 330.1966.7p.) Fox, Charles J.J..1909.On the coefficients of absorption of nitrogen and oxygen in distilled water and sea-water, and of atmospheric carbonic acid in sea-water.Transactions of the Faraday Society 5:68-87. Gerard, Robert, Marcus G. Langseth, Jr. and Maurice Ewing. 1962. Thermal gradient measurements in the water and bottom sediment of the western Atlantic.Journal of Geophysical Research 67:785-803,

Gordon, Arnold L.1966.Potential temperature, oxygen and circulation of bottom water in the Southern Ocean.Deep-Sea Research 13:1125-1138.

Gordon, Arnold L., Paul J. Grim and Marcus Langseth.1966. Layer of abnormally cold bottom water over southern Ayes Ridge.Science 151:1525-1526.

Hesselberg, Th. and H. U. Sverdrup.1915.Die Stabilitatsver- haltnisse des Seewassers bei vertikalen Verschiebungen. Bergens Museums Aarbok 1914-1915, no.15,p.1-16. (U. S. Navy Hydrographic Office.Translation 148.1962. 16p Jacobs, Stanley 5.1965 and 1966. Physical and chemical oceanographic observations in the southern oceans.Palisades. 2 vols.(Lamont Geological Observatory.Technical report no. l-CU-l-66 on NSF Grant GA-305). 128

Knudsen, M.1901.Hydrographical tables.Copenhagen, G. F. C. Gad.63 p.(Photolithographic reprint by G. M. Manufacturing Company, New York, 1962).

Kuenen, Ph. H.1942.Collecting of samples and some general aspects.In:The Snellius expedition in the eastern part of the East Indian Archipelago 1929-1930 under leadership of P. M. van Riel.Vol.5, part 3,section 1.Leiden, F. J.Brill. 268p. Langseth, M. G., X. LePichon and M. Ewing.1966.Heat flow through the Atlantic floor and convection cells.Journal of Geo- physical Research 7l(22):532l-5355.

Lee, William H. K. and Seiya TJyeda.1965.Review of heat flow data.In:Terrestrialheatflow.Washington, D. C. p. 87-190. (American Geophysical Union.Geophysical monograph series no.8).

LePichon, Xavier.n.d. Exit Antarctic bottom current through the Rio Grande Rise.(In press, Lamont Geological Observa- tory)

Metcalf, W. G.1961.Chain-17 in the Romanche Trench.Oceanus 8(l):Z-7.

Mosby, Hakon.1934.The waters of the Atlantic Antarctic Ocean. Oslo, Kommisjon Hos Jacob Dybwad.117 p.

Potsma, H.1958.Chemical results and a survey of water masses and currents.In:The Snelliusexpedition in the eastern part of the East Indian Archipelago 1929-1930 under leadership of P. M. van Riel.Vol.2,Oceanographic results, part 8. Leiden, F.J.Brill.64p. Revelle, Roger and Arthur E. Maxwell.1952.Heat flow through the floor of the eastern North Pacific Ocean.Nature 170:199-ZOO.

Schott, Gerhard.1905.Die Bodenformen und Bodentemperaturen des sudlichen Eismeeres.Dr. A. Petermanns Mitteilungen aus Justus Perthes Geographischen Anstalt 51:241-247.

Schubert, 0.V. 1935.Die Stabilitatsverhaltnisse des Atlantischen Ozeans.In: Meteor reports, Band 6,Teil 2, Lieferung 1. Berlin, Walter de Gruyter and Co.54 p. 129 Sverdrup, H. U., Martin W. Johnson and Richard H. Fleming.1942. The oceans, their physics, chemistry, and general biology. New York, Prentice-Hall, Inc.1087 p.

Taylor, P. and M. Langseth. n.d.Recent heat-flow measurements in the Indian Ocean.(In press, Lamont Geological Observatory) Tydeman, G. F.1922.Opmerkingen betreffende de temperatuur van bet Zeewater.In: De zeeen van Nederlandsch Oost-Indie. Leiden, E. J.Brill.p.102-124. Von Herzen, R. P. and M. G. Langseth.1965.Present status of oceanic heat-flow measurements.In:Physics and chemistry of the earth, ed. by L. H. Ahrens etal.Vol. 6.London, Pergamon Press.p.365-40 7. Wooster, Warren S. and Gordon H. Volkmann.1960.Indications of deep Pacific circulation from the distribution of properties at 5 kilometers.Journal of Geophysical Research 65(4):1239- 1249.

Wust, Georg.1928.Der Ursprung der Atlantischen Tiefenwasser. Zeitschrift der Gesellschaft fur Frdkunde zu Berlin; Sonderband zur Hundertjarfeier der Gesellschaft, p.506-534.

1933.Das Bodenwasser und die Gliederung der Atlantischen Tiefsee.In: Meteor reports, Band 6, Teil 1. Berlin, Walter de Gruyter and Co.107 p.

1964.Stratification and circulation in the Antillean- Caribbean basins; part one, spreading and mixing of the water types with an oceanographic atlas.New York, Columbia University Press.201 p. APPENDICES 130

APPENDIX A SOURCES OF OCEANOGRAPHIC STATIONS

(Cruise numbers are those of the National Oceanographic Data Center unless prefaced by letter abbreviation of the ship's name.Station numbers are originator's station numbers; where these are not available, the station number assigned by the National Oceanographic Data Center appears in parenthesis.)

Basin and ShiD Cruise Stations (A-i) Atlantic-Indian Basin DEUTSCHLAND 060044 116 and 117 DISCOVERY II 740038 852, 853, 858, 1147, 1148, and 1158 74039 1363, 1539, 1547, 1549 and 1551 74040 2014,2093, 2094,2113, 2318,2320, 2322,2340, 2541,2565, 2567,2592, 2594,and 2616 ELTANIN EL-8 131,132, 144 and145 OB 90004 210,242, 243 and246 90029 491 (A-la) Small basin to NW of and Atlantic-Indian Basin (A-lb) ELTANIN FL-l2 286 and 287 (A-2) Argentine Basin CAPITAN CANEPA 08691 C-344, C-345, C-348, C-349, C-350, C-359, C-360, and C-361 08692 230 and 246 08694 324 and 327 DEUTSCHLAND 060044 94,99, 100 and 110 DISCOVERY 74035 D-7l, D-72, D-74 and D.-77 131

Basin and Shi Cruise Stations ELTANIN EL-8 121 EL-9 166, 168 and 170 METEOR 06004 51,54, 56, 57 and 58 (A-.3)Brazil Basin ALBATROSS 770418 337 ATLANTIS 310837 5777, 5779, 5789, 5781 and 5846 CHAIN 31035 341 CRAWFORD 310583 110,111, 112,113, 114, 115, 116,128,129, and 130 31831 425, 426, 427, 428, 429, 432 and 482 31993 1358, 1360, 1362, 1363, 1364, 1365, 1367, 1368, 1369, 1370, 1371, 1372, 1374, 1430, 1431, 1432, 1433 and 1434 L OMONOSOV 90008 380 METEOR 06004 156 and 171 JOHNE. PILLSBURY 31996 30,36, 37 and 38 SAN PABLO 31065 (51) and 84 (A-4) North American Basin ALBATROSS 770418 367 ATLANTIS 310302 1581,2639(a), 2639(b), 2639(c),2809, 4750and 4755 310566 5205,5206, 5207,5208, 5209,5217, 5219,5220, 5221,5261, 5308,5309, 5310,5315, 5317,5319, 5338,5359, 5360,5361, 5362and 5363 132

Basin and Ship Cruise Stations

310974 147,149,151,152, 153, 154,155,156,157,158, 159,160,161, 162,163, 164,165,166,167,168, 169,170,171,189,191, 192,193,194,195,196, 198,199,200,201and 202 31209 449 CRAWFORD 310583 73 and 74 310623 226,228,304 and 311 310816 325,373,and 385 310974 821, 837, 838, 840, 858, 860, 861, 862, 863, 864, 868 and 889 DANA 260008 1356 and 1357 DISCOVERY II 740622 3611, 3612, 3613, 3614, 3620, 3621, 3622 and 3625 EXPLORER 31098 E13, E23,E24,E28, E32, £33,and £35 31195 36 and 47 31596 60 and 61 GILLIS 31185 NC2, NC3and NC4 31622 4 and 5 HORIZON 31183 10 KRUZENSHTERN 90040 7 LOMONOSOV 90014 435 and 443 ALBERT J. MEYER 31258 (2),(3),(4),(5), (7),(9), (18),(19),(20),(22),(23), (24), (25),and (27) 31260 (1),(3),(4),(8),(11), (12),(13),(14),(15),(16), (17),(18),(19),and (21) 133

Basin and Ship Cruise Stations

31280 (1),(2),(3),(4),(5),(6), (7),(8) and (9) PREVAIL 310037 6,8,11,17,23, 24 and 25 31053 C3, C6, Gb, C14, G15, J5, J6 and J9 REQUISITE 310001 8,11, 27, 30, 31, 32, 37, 40 and 46 REHOBOTH 310059 (1),(2),(3), (7),(23), (31), (33),(39), and (48) 31061 (14) SAN PABLO 31057 (4),(6),(7),(14), and (15) 31058 (3),(41),(43),(51),(52) and (53) 310060 (32),(48),(49), (52),(54), (72),(83),(85),(93), (104), (106) and (111) 310062 (32) 310068 (36) and (38) 310159 (23) and (24) 310648 (1) 310656 (5), (6), (8),(9),(10), (11),(19), (21),(22) and (23) 310662 (8),(20),(21),(22),(23), (24),(26), (27),(30),(31), (32), (33),(34),(35), (36) and (41) 310932 (2) 310972 SP38 SEDOV 90036 19 90037 10,. 22, 812 and 1425 90038 7,10, 13and 18 134

Basin and Ship Cruise Stations SHELDRAKE 31282 (8),(16),(27),(28),(29), (30),(31), (32) and (33) THOR 310647 (8) (A-5a) Iberian Basin CRAWFORD 310623 248, 249, 250, 251 and 252 DISCOVERY II 740596 3870, 3871, 3872, 3891, 3895, 3896, 3905, 3907 and 3908 740622 3539 900961 L221 METEOR 06004 283 MICHAEL SARS 580007 91 PLANET 060049 1 and 10 (A-5b) Canary Basin ALBATROSS 770418 328 and 342 ATLANTIS 310837 5763, 5764, 5765, and 5759 CARNEGIE 310002 19 CHAIN 310035 347, 348 and 349 CRAWFORD 310583 78 and 79 310623 260, 265, 284, 285, 288, 289, 290, 291, 293, 294 31993 1333, 1334, 1335, 1405 DEUTSCHLAND 060044 39 DISCOVERY II 74037 699 740622 3596, 3597, 3598, 3599, 3600, 3601, 3639, 3640 and 3641 HORIZON 31183 21 LOMONOSOV 90008 365, 366, 369, 370, 371 and 372 135

Basin and Ship Cruise Stations 90014 482 METEOR 06004 264, 265, 281, 282, 305, 306 and 307 PILLSBURY 31996 4 (A-6) Sierra Leone Basin CHAIN 310035 317 and 340 CRAWFORD 583 157 90993 1138 EXPLORER 31990 87 and 94 LOMONOSOV 1142 (A-7a) Guinea Basin CHAIN 310035 302, 304, 310, 311 and 312 CRAWFORD 31831 467, 472, 473 and 476 METEOR 06004 224 MOWE 06043 29 PLANET 060049 21 SAN PABLO 31065 67 (A-7b) Angola Basin ATLANTIS 31197 234 CRAWFORD 310583 94, 95, 96, 98, 143,145, 146, 148, 149, 150, 443, 447, 448 and 449 DISCOVERY II 74037 D696 GAUSS 06220 G106 GERONIMO 31995 123 and 135 METEOR 06004 179 and 190 PLANET 060049 30 (A-8) Cape Basin ATLANTIS 31197 232 and 233 310837 5840, 5841 and 5842 136

Basin and Ship Cruise Stations

CAPITAN CANEPA 08692 219 CRAWFORD 31831 452 DISCOVERY II 74035 D87 74037 D416 740038 1167 74040 1801 and 2343 GAUSS 06220 G26 and G104 METEOR 06004 24, 25, 75 and 76 NATA.L 91044 174 (N-i) South Indian Basin DISCOVERY II 740038 884, 885 and 890 74040 1705, 2164, 2166 and 2180 OB 90830 48 N-Z) Aguihas Basin CAPITAN CANEPA 08692 248 DISCOVERY II 740038 848, 849 and 850 74039 1569 74040 2378, 2574 and 2625 PLANET 06049 54 and 60 (N-3)Natal Basin DISCOVERY II 74037 D431 74039 1567 74702 2868 GA.LATHEA 26014 181 and ZOO NATAL 91044 150 and 151 VEMA 31834 48 137 Basin and Ship Cruise Stations (N-4) Crozet Basin ARGO 31184 111 DISCOVERY II 74040 1756 OB 90830 129 VEMA 31834 62 (N-5) Madagascar-Mascarene Basins ATLANTIS 31197 217, 218 and 219 ARGO 31184 109 DISCOVERY II 74040 1758 VITYAZ 90010 4655 (N-6) Somali Basin ALBATROSS 770418 232 and 235 ARGO 31184 21 and 23 31269 14 and 16 ATLANTIS 31197 129, 130, 131, 138, 139, 140, 141, 142 and 144 (N-7) Mid-Indian Basin ALBATROSS 770418 205 ARGO 31181 15 31184 72 and 78 DISCOVERY II 74030 5341 OB 90004 316 PIONEER 31201 20, 23 and 26 SNELLItJS 64198 23 VITYAZ 90010 4594 and 4599 90034 5262, 5267, 5269, 5272 and 5278 138

Basin and Ship Cruise Stations (N-8a) Wharton Basin ALBATROSS 770418 183 ARGO 31184 2,121 and 128 DANA 260009 3813 DIA.MANTINA 09001 24, 32, 48, 50, 52, 55 and 59 09003 425, 427 and 432 09004 143,145, 148, 323, 329, 336 and 339 09029 48,53, 58,63,64, 87, 91 and 95 09030 6,7,33 and 35

DISCOVERYII 74702 2888, 2891, 2892 and 2893 GALATHEA 26014 463 GASCOYNE 09012 199 and 205 09034 18,19 and 22 KOYOMARtJ 49703 11 and 19 OB 90004 309, 312 and 314 PIONEER 31201 52 and 56 SNFLLITJS 64198 24, 145 and 146 VITYAZ 90010 4504, 4523, 4530, 4535 and 4553 90034 5205 and 5212 (N-8b) West Australian Basin DIAMANTINA 09001 121 and 131 09002 8,114 and 23 09003 350, 357, 397, 438 and 441 09004 139 and 307 09007 8 and 48 09029 44 and 45 139

Basin and Ship Cruise Stations 09032 193 09036 5,14 and 46 09038 90 and 121

DISCOVERYII 74040 1738 and 2149 74702 2698 GASCOYNE 09012 213 and 215 09034 2,4,31 and 33 VEMA 31834 69 VITYAZ 90010 4567 (N-9)So. Australia Basin DIAMANTINA 09002 21, 33, 58, 61 and 84 (P-l) Southeast Pacific Basin BURTON ISLAND 31650 10 and 12 DISCOVERY II 740038 728, 730, 731, 972, 974, 975, 988 and 990 74039 1224, 1228, 1241, 1253, 1255, 1316, 1419, 1445, 1447, 1453, 1455, 1459, 1464 and 1469 74702 2837 and 2839 ELTANIN EL-b 195, 197, 221 and 225 EL-il 231, 236, 238, 239,and 240 EL-13 293 EL-iS 361, 363 and 376 EL-17 411 EL-l9 441, 443, 444and445 EL-20 484 EL-Zi 507 OB 90005 408, 409, 411 and 413 140

Basin and Ship Cruise Stations (P-2) Southwest Pacific Basin ARGO 31181 11,13, 14 and 25 CARNEGIE 310002 162 DANA 260009 3580 and 3628

DISCOVERYII 740038 944, 947, 949, 950, 964, 965 and 966 74040 2206, 2217, 2218 and 2219 74702 2734, 2735, 2736, 2738, 2739, 2769, 2770, 2823 and 2825 GALATHEA 26014 677 GASCOYNE 09033 242, 244, 246, 248, 250, 252 and 254 HORIZON 31802 D018 OB 90005 361, 363, 365, 381, 389, 393 and 396 VITYA.Z 900862 3812, 3818, 3823, 3827 and 3831 (P-3) Tasman Basin CSIRO Cruises 09785 182 and 193 17 casts at station G (from CSIRO list) DA.NA 260009 3656, 3663 and 3665 DISCOVERY 74040 1683 74702 2715, 2726, 2817, 2819, and 2820 GASCOYNE 09005 2 and 91 OB 90005 347, 349, 350 and 351 (P-4) Central Pacific Basin ALBATROSS 770418 105, 128, 133 and 138 ARGO 31181 MONN CARNEGIE 310002 159 141 Basin and Ship Cruise Stations DAVIS 31203 9 GASCOYNE 09033 217 and 222 REHOBOTH 31884 5,6,8,9 and 16 STRANGER 31724 17 and 28 VITYAZ 900862 3795 and 3805 (P-5) Philippine Basin ALBATROSS 770418 162 DAVIS 31203 12 GALATHEA 26014 412, 433, 440 and 449 JAPAN MARITIME SAFETY BOARD SHIPS490003 61 490004 33 490005 24 490037 T18 JAPAN HYDROGRAPHIC DEPARTMENT 490002 OLA7 PLANET 060049 276, 293 and 296 SNELLItJS 64198 47,76, 260, 261, 264, 275 and 301 SPENCER F. BAIRD 31182 H14 VITYAZ 900861 3707, 3726, 3728, 3746 and 3759 (P-6) Northwest Pacific Basin BERING STRAIT 31267 26 and 62 31397 6 CARNEGIE 310002 15, 116 and 117 DAVIS 31203 10, 41, 47 and 49 JAPAN MARITIME SAFETY BOARD SHIPS 490036 T96, T99 and T100 490037 TiOl 490003 15,18, 28,31, 35, 38,68, 83, B77 and B84 142

Basin and Ship Cruise Stations 490004 30, 37, 46, 51,78, 92 and 143 490005 94 JAPAN METEOR- OLOGICAL AGENCY SHIPS 490327 1358, 1360, 1361 and 1362 49330 1506, 1507, 1509, 1510 and 1511 REHOBOTH 310156 (6), (7),(13),(14),(20), (22),(34), and (35) 31161 (23) and (24) 31165 (15) 31180 17, 21, 22, 27,30, 35, 38, 45, 50 and 51 310617 (ii),(30) and (77) 310661 (19), (21),(22), (23), (31) and (37) SPENCER F. BAIRD 31182 H4 and H92 VITYAZ 900589 4343, 4355, 4369, 4371, 4373 and 4377 900861 3625, 3631, 3634, 3686 and 3689 (P-7) Northeast Pacific Basin AGASSIZ 31418 60180 and 80200 ALBATROSS 770418 80,87, 93 and 123 ARGO 31181 21 and 29 BROWN BEAR 31493 6426 31558 14 and 21 31576 1,8,10, 15 and 24 CARNEGIE 310002 137, 138, 139, 141, 142, 146, 151, 155 and 156 DANA 260009 3563 143

Basin and Ship Cruise Stations DAVIS 31203 4,5, 52 and 53 31589 7 31929 3 HORIZON/SPENCER F. BAIRD 31802 C4, Gb, Gil, C12, C13, G14, D4, D6, D7, D8, D9, D10, MiD, M4D, and M1OD 31346 7306 and 821234 PIONEER 31099 P72, P75, P91, P94 and P97 31172 120 310945 P11, P12, P13, P14, 54 and 63 REHOBOTH 31052 H4, H5, HiS and H23 31063 (38) 31161 (17 and (40) 310617 (3),(5),(6),(55),(56), (57),(59),(60),(61), (62), (74),(75), (76),(78),(88) and (89) 310641 (11),(12),(39),(40),(42), (47) and (49) 31665 (4) and (8) 31685 (3),(18),(22),(23),(24) 31884 24 SURVEYOR 31172 S5 and S19 31204 37, 41, 44, 46, 49, 50, 53 and 54 VITYAZ 900589 4066, 4096, 4120, 4199, 4239, 4249, 4289, 4295, 4311, and 4320 900862 3778 144

Basin and Ship Cruise Stations

YAQUINA YALOC HAH23, HAHZ8, HAH3O, HAH33, HAH34, HAH37, HAH5Z, HAH54, HAHS6, and HAH57 APPENDIX B PLATES 145 70 0900 105° 120° 135° 150° 155°( 9° $50 1500 135° 120° 105° .00e 750 60° 450 30° 15°W 0° 15°E. 30° 45° 10° 750 900700

60° 600 I II I 45° - 4 (A- b) ------N(THWEST NT1tEAST - r 300 PACIFIC a) BASIN PACIFIC * ArTCAN NTJ{ and (p_6) ' - BASIN CANARY 15°N 14 ttLIPPINEBASIN BASIN I 1 BASIS>(A-5) 15°N 00 5) CENTRAl PACIFIC (P-7) 4. 4 - GtJINEA. -. - 00 I . 15°S -t--* ** j * ---4*rn- I A-3) A ANGOLA(- BASIN * F 30°--(N_8) I APE - 450 (b) AS SOUTHWEST crpic AEGENTI SI BASIN(A-8) AS4 wrAxIBASIN C1OZ1BASIN \ 45e * BASIN(P-2) - * EN-- I (N2 0 ° 1 PACIFICBASfl'1 SOUTHEAST (P-i) -D A-1aJ , ATlANTIC-INDIAN BASIN (A-i) a 700 900 1050 1200 1350 1500 165°E 180° 165°W 150° 135° 120° 105° 90° 75° 600 45° 30° 15°W 0° 15°E 300I 450 I 60° - 15° 900 10° PLATE I. Major deep basins of the world oceans, o90 105° 120° 1350 150° 165°E 180° 165°W 150° 135° 120° 105° 90° 15° 60° 45° 30° 15°W 00 15°E 30° 45° 600 75° 900 146 60° 990 - 150 50 .. 4, 2.20 I 450 300 1.50 . 1 >150 ' .__._.___:>-__._:. i so 300 15°N I 1 601 50 to 11.50 55 to 1 40 1 30 L91 15°N - 0 O j * <150 150 1 1.40 4 1,8S.1 7- t ' I . 0° 1 / \ 15°S\ > 1 15 >115 .. P ' 1 LIII IJ\\ " oo >240 1.60 30 1.10 S12 100 \ \ 050 / 060 0.70/11 \ 1 080 ) --.o.s' ('-, 130 30° 450 J : 1.10 0$0(oo 45° \S_ :°° 0 1.20 1 0.10'oio0.00 0 600 00o' 1,10 H 070 014 026 ) o.io . 80° : 0.25 4O 020 100 °'° 0 30 040 _..___..-040 0.40 0 30 0° 900 105° 120° 135° 150° 165°E PLATE II. 180° 165°W TemperatureContour interval (in situ) is distribution generally 0.1150° at °C.the deep temperature minimum. 135° 120° 105° 90° 150 600 450 30° 15°W 0° 15°E 30° 450 600 150 900 147 105° 120° 135° 150° 165°E 180° 165°W 135° 120° 105° 900 15° 45° 30° 15°W 0° 15°E 30° 45° 50° 15° 10090° . r 6° 900700 650 I. 600 450 - // /////// ///// //// //// '////' ,/////. 7/, , : -4. 30° // '/// ////, ::. /// - 30° 15°N // //////////////////////// // / 15° 0° 0 / - /////// ////, //// //, 15°S _#, - - p ///// -; I 30° /' - 30° / /1 -. // /// - // / - / // : //,//i /// // //// /////' ////// /// /77/, ///'/ // /////,// b 0 60° : /7/ / - ///////////// / /////// ///////////////// ////////Z - 7' / /////////////_// /' I / z///////////// z// 10° 900 105° Z1 120° Deep temperature minimum observed 135°- 50° 165°E 180° 165°W 150° 135° 120° 105° 90° 15° 60° 45° 30° 150* 0° 15°E 30° 450 60° 75° 90° 70°

Adiabatic lapse rate observedPLATE iii. Areas in which the deep temperature minimum and adiabatic lapse rates have been observect. l'i 8

750 450 450 90° 105° 120° 135° 150° 16S°E 180° 165°W 150° 135° 120° 105° 00° 60° 30° 15°W 0° 15°E 300 50° 15° 90°

II' I,

45° I 45° >110 / >210 1 .0 I/ z10.-___ * I 7 7 -

30° - 30° 1i70

2 110 _

095 30 1,10 0 0° 0° 2

070 150S.\çF. / I .l(' / L

020 035

----H-'N -'s.., _o j / 060 070

10° - - IIi. 10° 90° 105° 120° 135° 150° 450 165°E 180° 165°W 150° 135° 120° 105° 90° 75° 60° 30° 15°W 0° 15°E 30° 45 60° 15° 9J0

PLATE N. Bottom potential temperature distributionin the major deep ocean basins. Contour interval is generally 0.1 °C. 149

1200 450 300 450 750 10090° 105° 135° 150° 195°F 180° 155°W 150° 135° 120° 105° 90° 15° 60° 15°W 0° 15°E 300 611°

60e 60°

450 3490

1 34 69 to 34 70 * 30° 70 _&' 30° 1 I \\ I \ ..., 3469 34.8___._488 I 15°N ) 15°N 3469

3470 34 70 3486 3472 347 0° 3471

15°S 3488 r 15S 3471

3471 30° 3470 30° >34 70 3473 3471 3469 >34.12 3469 34 71 3474

45° '....3447 to < 34 72 3470 3466 3470

34673466

3468 to 3469 - 34.69 34 66 to 34.67 60° 34 70 to 34 71 60° -

100 790

90° 105° 120° 135° 150° 165°F 180° 165°W 150° 135° 120° 105° 90° 15° 80° 45° 30° 15°W 0° 15°F 30° 45° 600 15° 900

PLATE V.Salinity near the bottom in the major deep ocean basins. Contour interval is generally O.Ol%. 150

750 450 150 10090° 1J5° 120° 135° 150° 165°[ 5° 150° 135° 120° 105° 90° 60° 30° 15°W 0° 15°t 30° 45° 60°

600 60°

27 81 S

450 45° 2781 to2782 ,. 4 27 93 '-2792 30° 2782 2791 300 to 27.91 92 27 80 15°N 27 82 to 27 83 15°N

2784 2790 27.85 00 2790 00

27 2785 I 27.87 15°S 2785 2790 27 83 27 85 * 27 88 2786 300 30° 2786 ; 1 2788 27.87 27851 2786 1 2786 tO 0 <2785 ¶ 45 276 ç 27 87 to 27 88 27 88 >2785k

27.87 27.

60° 2789 27 88to 27 89 2788to 27.89

1 .

700 10° 450 90° 1050 1200 1350 1500 165°F 180° 165°W 150° 1350 120° 105° 90° 15° 600 300 15°W 00 150F 30° 45° 60° 150 900

PLATE VI.Sigma-O [i000 x (potentialdensity - 1)] near the bottom in the major deep ocean basins. 151 70090° 105° 120° 135° 150° 165°E 180° 165°W 150° 135° 120° 105° .00° 75° 600 450 30° 15°W 0° 15°E. 30° 45° 60° 750 60° 1 60° I j 0 450 5.3 45° 30° 36 L. 300 5 5 2 15°N o 4.3 J j. 15°S 1I 4 I 5.1 L 53 41 --15°S 300 44 4.6 4.8 1 46 5.0 I I I I I J I I - 49I 45°.+4 I 11 4.8 47 I I I L 1 53 100 i 100 90° 105° 120° 135° PLATE VII.150° 165°E 180°. Dissolved oxygen content (mi/i) near the.165°W bottom in the major deep ocean basins0 150° 135° 120° 105° 900 75 600 45° 30° *QJfljjflg DISCOVERY15°W II data (otherwise 11.7) 0° 15°E 30° 45° 600 750 900