Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
CHAPTER 4 TRANSPORT AND DISPERSAL OF RADIOACTIVE ELEMENTS IN THE SEA ' WARREN S. WOOSTER, Scripps Institution of Oceanography, La Jolla, California and BOSTWICK H. KETCHUM, Woods Hole Oceanographic Institution, Woods Hole, Massachusetts THE FATE of radioactive elements in the sea differs from that of non-radioactive elements since they are subject to radioactive decay. Otherwise, concentrations of radioactive ele- ments are changed by the same physical and biological processes as are those of other iso- topes in the same physical state. Thus the fate of radioactive material introduced into the sea depends on: 1. What is introduced â the nuclide, its radio- active properties (half-life, nuclear reaction, kind and energy of radiation), its physical state in sea water (whether particulate, colloidal or ionic) and its chemical properties (including its role in biological processes). 2. Where it is introduced â position and depth with respect to the density and velocity struc- ture of the sea. This paper describes the physical processes whereby radioactive elements in true solution are diluted by mixing and are carried from one part of the ocean to another. Although all parts of the open ocean appear to be in continuous motion and in communication with each other, the rates of this motion and exchange cover such a wide range that it is convenient to con- sider separately the questions of near-surface vertical and horizontal exchange, intermediate and deep circulation, and the exchange between the deep sea, coastal areas and enclosed basins. Near-surface circulation In middle and low latitudes the surface layer of the ocean, from 10 to 200 meters thick, is 1 Contribution from the Scripps Institution of Oceanography, New Series, no. 903. Contribution no. 870 from the Woods Hole Oceanographic Institution. This paper, in part, represents results of research car- ried out by the University of California under con- tract with the Office of Naval Research. Reproduction in whole or in part is permitted for any purpose of the United States Government. separated from the colder deep waters by a layer of rapid density change and great sta- bility, the pycnocline or thermocline. This intermediate layer varies in depth and stability from time to time and from place to place. At times there are two such layers, the seasonal thermocline and a deeper main thermocline. The surface layer is often called the "stirred" or "mixed" layer - because of its relative uni- formity in temperature and in concentrations of dissolved substances. It is believed that radioactive material in- troduced into this surface layer will be rapidly distributed vertically throughout the layer. The general uniformity of concentrations within this layer suggests that forces are present which tend to bring it about. Because density increases only slightly with depth through the layer, little energy is required for vertical stirring. Some evidence of the rapidity of vertical mixing in the upper layer is given by Folsom (Revelle, Folsom, Goldberg and Isaacs, 1955), who observed that when fission products were introduced at the surface in an area where the surface layer was about 100 meters thick, the lower boundary of the radioactive water reached the bottom of this layer in about 28 hours. Within this period of time radioactivity had be- come uniformly distributed vertically through- out the layer. Rapid vertical mixing in the upper layer is brought about primarily by the following two processes: 1. Convection: When the density of surface water is sufficiently increased, owing to either a A distinction is made here between stirring and mixing. In stirring, one causes relative motion of different parts of the liquid, and the average value of the gradient is increased. Mixing then takes place, the gradients disappearing and the liquid becoming homogeneous (Eckart, 1948). 43
44 Atomic Radiation and Oceanography and Fisheries a decrease in temperature or an increase in salinity, the surface water sinks and mixes with deeper water. Convection is maintained by (a) surface cooling due to long wave radiation and heat conduction to the atmosphere, (b) the loss of latent heat and water vapor in evapora- tion, or (c) the increase of surface salinity from freezing of surface water. 2. Wind stirring: Vertical turbulence in the upper layer results from wind action on the sea surface. The extent of wind stirring depends both on the magnitude and uniformity of wind stress and on the vertical density gradient. Stir- ring is effective only to a depth where there is sufficient energy to overcome the effect of sta- bility. Both the homogeneity and the depth of the upper layer are affected by wind. Single gales have been observed to deepen the surface layer on the average by about 20-30 feet, (Francis and Stommel, 1953). Rapid vertical mixing may be brought about by other processes. Thus in shallow coastal areas stirring by strong tidal currents is im- portant. Stirring may also be accomplished by the vertical component of currents, particularly in regions of upwelling and sinking. It should be noted that even above a well- developed pycnocline there is not complete ho- mogeneity within the so called "mixed" layer. Concentrations of those elements affected by biological activity (such as oxygen and phos- phorus) may show significant variation within the euphotic zone. Even so-called "conserva- tive" concentrations (temperature and salinity) may not be uniform within the surface layer. Such heterogeneity may be attributed to in- complete vertical mixing or to vertical shear in the surface layer. When radioactive materials are introduced into the near-surface layer, they are transported away from the area of introduction by surface currents. These currents extend, in general, through the entire depth of the upper layer and seem to be driven, directly or indirectly, by the wind. The average locations and velocities of the important surface currents of the world ocean have been studied for many years and are well known (see, for example, Deutsche Seewarte, 1942; U. S. Navy Hydrographic Office, 1947 a and b, 1950; Sverdrup, Johnson, and Flem- ing, 1942, ch. 15). This knowledge comes primarily from averages of countless ship-drift observations, and from computations based on the observed subsurface distribution of density. These calculations give mean speeds as high as 193 cm/sec (90 miles per day) in the Florida Current (Montgomery, 1938a) and 89 cm/sec (41 miles per day) in the Kuroshio (Koenuma, 1939). The volume of water flow- ing through the Florida Straits in 15 years is about equal to that of the upper 500 meters of the whole North Atlantic. Similarly, between the northern Ryukyus and Kyushu, the Kuro- shio transports a volume equivalent to that of the upper 500 meters of the North Pacific in about 50 years. It seems likely that there is no area of surface water in the ocean that can be considered as isolated from the remaining surface waters. Recent intensive studies of the Gulf Stream and other surface currents, using such modern instruments as the bathythermograph, electronic navigational aids, and geomagnetic electro-kine- tograph (GEK), have revealed complicated fine structures, with filamentous jets and counter- currents not apparent in the average picture (Fuglister, 1951). Characteristic maximum sur- face velocities measured by GEK and Loran dead reckoning in the Gulf Stream were found to fluctuate between 150 and 300 cm/sec or 70 to 140 miles per day (von Arx, Bumpus and Richardson, 1955). Thus, in estimating the time at which radioactive materials will be found at various distances from the area of introduction, one must be cautious in the use of average surface current speeds. Direct evidence of the transport of radio- active materials by surface currents in the western Pacific is given by "Shunkotsu-Maru" survey (Miyake, Sugiura and Kameda, 1955) and the "Taney" survey (U. S. Atomic Energy Commission, 1956) four months and thirteen months respectively after nuclear weapons tests in the Marshall Islands in March, 1954. The earlier survey found significant levels of radio- activity at a distance of 2000 kilometers from Bikini, suggesting a westward drift of more than 9 miles per day (about 20 cm/sec). The later survey found significant levels of radio- activity at least 7000 kilometers downstream from Bikini; this gives about the same minimum westward drift. In addition to being drifted away from the area of introduction, radio-active materials are
Chapter 4 45 Transport and Dispersal dispersed by diffusion. Diffusion in the ocean is caused by turbulence or eddies, and the coefficient of eddy diffusivity is usually more than a million times the corresponding molecu- lar coefficient. The rate of eddy diffusion de- pends on wind speed, current shear, density gradient, gradient of the diffusing concentra- tion, direction of diffusion, and the dimensions of the phenomenon. The calculated rates de- pend upon the magnitudes of eddy diffusivity coefficients used, and they have been estimated by a number of methods (Sverdrup et al., 1942, p. 484â485; Munk, Ewing and Revelle, 1949). Because of both the large number of variables concerned and the present unsatisfactory state of our quantitative knowledge of turbulence in the ocean, it is difficult to predict the diffusion of radioactive materials under any given cir- cumstances. The most satisfactory approach at present is to conduct diffusion studies and ex- periments at the place and under the conditions of contemplated release. The results are only applicable to the particular areas. During the 1946 preliminary survey in Bikini Lagoon, the state of turbulence was determined by a variety of measurements, and the subse- quent observed distribution of radioactivity was in close agreement with the predicted values (Munk, Ewing and Revelle, 1949). A mean value for the radius of the contaminated area was 3 km., which approximately doubled be- tween the first and second days after the burst. The initial distribution of radioactivity as de- posited by the atomic bomb was patchy, and the turbulent eddies, which spread the con- tamination over a larger area, did not appreci- ably reduce this patchiness during the first three days. Another pertinent study was made by Ketchum and Ford (1952) who examined the rate of dispersion of acid-iron wastes in the wake of a barge at sea. Computed mixing coefficients showed a tendency to increase with increasing time, and thus with the dimensions of the mix- ing field, and the radius of the contaminated area was observed to double in time periods ranging from 0.5 minutes to 35 minutes. It should be noted that the scale of this phenom- enon was about 10~2 that of Munk, Ewing and Revelle (1949) ; they show that the ratio of lateral eddy diffusivity coefficient to the radius of the area considered is relatively constant over a range of radius between 102 and 109 cm. A large scale tracer experiment was carried out in the Irish Sea prior to the discharge of radioactive effluent (Seligman, 1955). During each experiment, 10 tons of 6.7 percent fluores- cein solution were introduced near the surface during a 20-minute period, and the sensitivity of subsequent detection was believed to be of the order of 1 part in 109. Maximum concen- trations detected directly after release were 10~4 of the original concentration; 12 hours after release, they were down to 5 x 10~T of the original concentration. The trial area was prob- ably part of an eddy and was subject to tidal mixing, so the results may not be generally applicable. Exchange between near-surface and intermediate waters Since the surface layer is separated from deeper waters by a layer of rapid density in- crease, and hence of great stability, vertical transfer of materials across this layer by eddy diffusion must be much less rapid than is ver- tical diffusion in the upper layer. Thus radio- activity introduced at the surface by fallout may remain in the upper layer for a long time and be diluted by only a small part of the total volume of the sea. Conversely, radioactive ma- terials introduced below the pycnocline should only slowly contaminate the upper layer where they are most likely to endanger human ac- tivities. However, organisms and particles of sufficient density may readily cross the pyc- nocline, due both to gravity and to vertical migrations. There are few observations which show di- rectly the existence of cross-pycnocline exchange on a local scale. In the western Pacific, both the "Shunkotsu-Maru" survey (Japanese Fishery Agency, 1955) and the "Taney" survey (U. S. Atomic Energy Commission, 1956) reported patches with significant concentrations of radio- activity below the thermocline four months and thirteen months, respectively, after mixed fis- sion products were introduced at the surface in the Marshall Island area. It is not known, how- ever, whether this exchange was effected by mixing processes, or by particulate or ecological processes. Exchange of properties between the near-
46 Atomic Radiation and Oceanography and Fisheries surface and deeper waters is most likely to take place under the following conditions: 1. In regions where the pycnocline is suffi- ciently shallow to be eroded at the top by wind stirring. In coastal waters the pycnocline is usually shoaler than in midocean, and shallow pycnoclines may also be found in high lati- tudes, at the equator, along the north edge of the Equatorial Countercurrent, and at the cen- ter of strong cyclonic eddies. This process is not effective to great depths, but could serve to bring radioactive materials into the surface wa- ters from the pycnocline layer. 2. In regions of upwelling, where the pycno- cline is relatively weak and where vertical cur- rents not only carry water toward the surface but also stir surface and deeper waters. It is unlikely that water from depths of more than 500 meters is ever brought to the surface by this process. Upwelling is common along west- em coasts of continents in the trade wind belt, such as the coasts of Peru and Northern Africa. In a simple sense, the persistent trade winds blowing parallel to or offshore develop an off- shore component of transport in the surface waters, and deeper waters upwell to maintain the volume continuity. Upwelling may also occur along other coasts when the winds are suitable. The process has been extensively stud- ied along the coast of California where it is not continuous because of the variability of the winds (Sverdrup et al., 1942, p. 725). The speed of coastal upwelling has been variously estimated as 0.6 m/day (McEwen, 1934), 2.25 m/day (Saito, 1951) and 2.7 m/day (Hidaka, 1954). However, since these estimates are theo- retical mean values, they may differ significantly from actual instantaneous upwelling rates. Midocean upwelling, associated with diver- gence of the surface currents, occurs in a band along the equator in the eastern and central Pacific Ocean (Cromwell, 1953). Observations indicate that the effects of this upwelling ex- tend to 50 meters in the eastern Pacific and to 100-150 meters in the central Pacific (Wooster and Jennings, 1955). Similar but less pro- nounced upwelling has been observed in the equatorial Atlantic (Bohnecke, 1936). 3. In regions of surface convergence, where sinking waters may fill the depths of the ocean, or may spread at intermediate depths according to their density. In tropical and temperate latitudes such sinking is confined to the surface layer. In such regions mixing in the upper layer may be facilitated but exchange across the pycnocline probably is not, since the sinking water tends to increase the density gradient in the pycnocline. In high latitudes, on the other hand, sinking waters may reach great depths, and it is in such regions that most of the intermediate and deeper water masses of the ocean are formed. The most extensive and pronounced of these convergences is the Antarctic Convergence which occurs at 50 to 60° S in a band around the entire Antarctic Continent. The cold, low-salinity water which sinks there forms an identifiable water mass, the Antarctic Intermediate Water, which spreads at depths between 800 and 1200 meters in all southern oceans. This water can be identified everywhere in the South Atlantic and extends across the equator as far as 22 °N in the North Atlantic (Deacon, 1933; Iselin, 1936). In the Irminger Sea, between Iceland and Greenland, and in the Labrador Sea, warm high salinity water of the Gulf Stream is partly mixed with cold low-salinity water flowing out of the Arctic Ocean. The resulting mixture may spread in small quantities as Arctic Intermediate Water, or when sufficiently dense may form the deep and bottom water of the North Atlantic (the possibility that the formation of this deep water is not a continuous process is discussed later). Intermediate waters of the North Pacific are probably formed in winter at the convergence between the Kuroshio Extension and the Oya- shio (Sverdrup et al, ch. 15). There is ap- parently no deep or bottom water formed by this process in the Pacific. 4. In regions where the density of surface waters is so increased by evaporation, cooling or freezing, that they sink to intermediate or greater depths. Active formation of Antarctic Bottom Water takes place in the Weddell Sea due to the freezing of high salinity surface waters. In the Mediterranean and Red Seas, bottom water is formed by winter cooling of waters whose salinity has been greatly increased by evaporation. Mediterranean water flows out into the North Atlantic at depths of 1000 to 1500 meters and can readily be identified near Bermuda, 2500 miles from its source. In summary, exchange between near-surface and deeper waters takes place most commonly (1) in high latitudes, (2) along the equator,
Chapter 4 47 Transport and Dispersal and (3) in coastal regions, particularly along the western coasts of continents. Conversely, such exchange is least likely in temperate and tropical latitudes in the vast central regions of the northern and southern oceans. Exchange between the open sea and coastal areas In coastal areas or enclosed basins where precipitation exceeds evaporation, there is a seaward surface drift of diluted water and a landward subsurface drift of water derived from the open sea. If radioactive materials were released in such a coastal area, the ma- terial which remained in the surface layer would be carried seaward, but the part of the material which mixed or settled to the deeper water would move toward shore and the estuaries of rivers. Conversely, if radioisotopes were lib- erated in the open sea, some would eventually be carried inshore as a result of the coastal and estuarine circulation. It is clear that the ultimate distribution in coastal areas of radioactive materials added to the sea would depend on the location of the release, the vertical distribution of radioactivity and density in the area of release, the length of time required for the transport to the coastal area or estuary, and the location of the source sea water which provides for the counter drift. The number of variables involved makes it difficult to discuss the effects in general terms, but it is worthwhile to note that the circulation in coastal areas is rapid, and water bathing the North Atlantic beaches is not uncommonly 90 per cent sea water even off large rivers such as the Hudson and Delaware. An idea of the lengths of time involved in the coastal circulation can be obtained from the mean age of waters in various parts of the At- lantic seacoast. Such mean ages are computed from the volume of water contained in the region and the estimated transport of water through the region. The waters of the con- tinental shelf from Cape Hatteras to Cape Cod have a mean age of about 2J years, those of the Bay of Fundy about 3 months, and those of Delaware Bay from the ocean to the height of tide about 3-4 months (Ketchum and Keen, 1953, 1955). The source sea water for all of these circulations is the "slope water" which is formed between the Gulf Stream and the edge of the continental shelf. A few data are available for confined basins and seas from which estimates of the mean age of the water can be derived. In most cases, however, the sources of water entering into the circulation are uncertain, and it should be em- phasized that in all cases some of the waters within the basin will be older or younger than the mean age. The source waters of the Florida Current are funnelled through the Caribbean Sea. The mass transport is 26 million cubic meters a second (Sverdrup et al., 1942, p. 638), so that this current carries annually a volume of water equivalent to one-sixth of the total volume of the Caribbean. However, there is evidence that the renewal of the deep water of the Caribbean proceeds at a much slower rate than the six year mean age that this ratio implies. Wor- thington (1955) has calculated, on the basis of loss of oxygen from this deep water during the last 30 years, that the age of the deep water in the various parts of the Caribbean may range from 93-142 years. The mean age of the waters above 2000 meters would be reduced to about 5 years if the deepest £ of the volume of the basin is isolated from the present circulation. The same current passes through the Yucatan channel into the Gulf of Mexico, before emerg- ing as the Florida Current. No estimate of the mean age of the waters of the Gulf of Mexico is possible, however, since the current data in the Gulf indicate an anticyclonic eddy in the western portion, and suggest that the waters of the Gulf of Mexico are drawn into the Florida Current to only a slight extent (Die- trich, 1939, Sverdrup et al., 1942, p. 642). The Black Sea probably contains the most isolated and the oldest deep water to be found anywhere in the oceans. Precipitation and run- off exceed evaporation, and the surface waters are dilute (salinities less than 18 per cent) and isolated from the deep water by an intense density gradient. The deep waters are anaero- bic; hydrogen sulfide reaches large concentra- tions below about 200 meters. The sill at the Bosporus is only 90 meters below the surface so that this deeper water is isolated from the more rapid surface circulation. The inflow of sea water is so small that it would take about 2500 years to replace the deep water in the basin (Sverdrup et al., 1942, p. 651). The mean replacement time for the surface layers
48 Atomic Radiation and Oceanography and Fisheries to a depth of 200 meters is equivalent to about 200 years. Gololobov (1949) has computed the mean age of the deep water on the basis of the annual contribution of phosphorus in the river inflow and the quantity accumulated in the depths. This computation indicates an accumulation time of 5600 years. The Arctic Basin receives its major inflow north of Scotland and a much smaller inflow through the Bering Strait. Additional sources are from the river runoff and excess of precipi- tation over evaporation. The outflow is pri- marily through the Denmark Strait (Sverdrup et al., 1942, p. 655). These flows would pro- vide a volume equal to that of the Arctic Ocean in about 160 years. The Arctic is also stratified because of the addition of fresh water from rivers and melting ice, and it is not known how isolated some of the waters in the deeper basins may be. However, recent analyses have shown that the deeper water in the Arctic Ocean is far from anaerobic, so that it seems unlikely that this water can be considered as isolated from the circulation. The Mediterranean is a basin in which evaporation exceeds precipitation and runoff. Through the Strait of Gibralter there is an inflow of oceanic surface water and a sub- surface outflow of high salinity Mediterranean water. The exchange is sufficiently rapid to replace the entire Mediterranean in about 75 years (Sverdrup et al., 1942, p. 647). The Mediterranean is divided into eastern and west- ern basins by a 500-meter sill between Sicily and Tunisia, and it is not know to what extent the deep waters of these basins are involved in the over-all exchange. Deep circulation Most of our present knowledge of the inter- mediate and deep circulation (see Sverdrup et al., 1942, ch. 15) has been obtained in- directly from the observed distribution of prop- erties. The general uniformity of temperature and dissolved substances in deep water suggests that deep currents are very slow, perhaps at most a few centimeters per second. But deep currents cannot be computed by the geostrophic method because only relative velocities can be thus obtained. Furthermore, small errors in the measurement of salinity or temperature produce uncertainties in velocity of the same magnitude as the currents being computed. The direction of movement in the deep and bottom water has been deduced from the observed distribution of properties such as salinity and potential temperature, but little can be learned about current speeds from such observations. Existing direct measurements of subsurface currents have been summarized by Bowden (1954). Such measurements have been made since the time of the CHALLENGER Expedi- tion (1873-76), but because of practical diffi- culties (such as the problem in the open sea of referring observations to a fixed frame of reference) they have taught us little about the deep oceanic circulation. The few successful measurements at depths greater than 1000 me- ters reported by Bowden showed mean speeds ranging from "negligible" to about 13 cm/sec. At nearly all stations and depths at which current measurements have been made, semi- diurnal tidal currents of the order of 10 cm/sec. have been recorded. Recently measurements of subsurface currents have been made in the North Atlantic by track- ing for three days a neutral-buoyant float sta- bilized at a given depth (Swallow, 1955 and unpublished). These measurements show small resultant speeds (1.7 to 9.1 cm/sec or 0.8 to 4.2 miles/day at depths from 600 to 1900 meters), tidal components of about 10 cm/sec, and in two successive three-day measurements at 1900 meters, a change in direction of 124°. Thus it seems likely that motion below the pycnocline is characterized by more variation, periodic or otherwise, than previously supposed and indeed that the mean drift may represent only a small part of the total motion. Little is known about the nature and extent of lateral and vertical mixing in the deep sea. It is generally believed, however, that flow and mixing take place along surfaces of constant potential density (isentropic surfaces) and that below the upper layer vertical mixing is very slow except near coastlines and areas where upwelling may occur (Montgomery, 1938). An observation supporting this belief was reported by Revelle, et al. (1955). Introduction of mixed fission products below the pycnocline led to the formation of a lamina of high radio- activity about one meter thick and 100 or more square kilometers in area. The radioactive water apparently spread out along an isentropic sur- face and resisted destruction by vertical mixing for at least three days.
Chapter 4 49 Transport and Dispersal Age of intermediate and deep waters It is generally accepted that intermediate and deep waters in most parts of the oceans acquired their characteristics while at or near the surface. Thus the low temperature and relatively high oxygen content of deep water can only be ex- plained by assuming an exchange between deep and surface waters. The problem of the dis- posal of radioactive wastes in the deep sea has stimulated the oceanographer's natural curiosity as to the rate of this exchange. The North Atlantic receives surface waters from the South Atlantic and loses deep water to the South Atlantic. Assuming a surface flow from the South to the North Atlantic of 6 million cubic meters per second (Sverdrup et a!., 1942, p. 685), and considering only the upper kilometer of the North Atlantic to be affected, the mean replacement time is about 140 years. The gyral in the North Atlantic, which includes the Gulf Stream, carries about ten times the volume of water exchanged be- tween the South and North Atlantic, so that the mean circulation time is only about one- tenth the replacement time. This surface exchange between the North and South Atlantic is balanced by a deep current from North to South. The mean displacement time for the deep water of the North Atlantic (2000-4000 meters) is calculated as about 250 years. This time is in reasonable agreement with more recent estimates of the age of the deep water discussed below. Between these surface and deep layers are the intermediate waters which appear to circu- late even more rapidly. Deacon (1933) calcu- lated rapid rates of northward flow of the Antarctic intermediate water in the South At- lantic, based upon alternate maxima and minima in the concentrations of oxygen in the oxygen minimum layer. These were interpreted as rep- resenting annual cycles when the waters were formed at the surface. He estimated a transit time of about 4£ years between the Antarctic convergence and the equator. Seiwell (1934) has similarly computed rapid flows and a mean transport time of 7-8 years for the drift of the oxygen minimum layer of the North Atlantic Ocean. Deacon's and Seiwell's interpretations have been questioned (see Riley, 1951, p. 77) on various grounds. However, their rates of flow agree with direct current measurements at comparable depths (see earlier) which also indicate rapid rates of circulation. The deep outflow from the Mediterranean sinks from sill depth to 1000-1500 meters in the North Atlantic Ocean. This water, although much diluted by Atlantic water, is characterized by relatively high salinity and temperature, and spreads out in a sheet which may be identified in most of the temperate North Atlantic, and some spreads into the South Atlantic. It can be readily identified near Bermuda, 2500 miles from its source. Iselin (1936) computed that sufficient excess salt would be produced by the Mediterranean outflow to produce the observed anomaly in 12-15 years. He pointed out that the actual replacement would be more rapid because he neglected admixture of Atlantic water in the immediate vicinity of the Straits of Gibraltar. Defant (1955) has evaluated the mixing processes involved in dissipating the Mediterranean water within the Atlantic Ocean, and has concluded that the total accumulation in the Atlantic Ocean represents the contribu- tion resulting from six years of flow through the Straits of Gibraltar. The rapid dissipation of this large water mass at mid depths suggests a more rapid circulation than had been gen- erally accepted for intermediate waters. During recent years other lines of investiga- tion have led to the belief that the overturn of water in the ocean basin takes place in less than a thousand years and probably in 200 years or less. Evidence supporting this belief follows. (Carbon-14 and carbon dioxide ex- change estimations are discussed in greater detail by Craig elsewhere in this report.) 1. Heat flow measurements: Measurements re- ported by Revelle and Maxwell (1952) have shown a heat flow through the floor of the Pacific Ocean of 1.2x10-' calories per square centimeter per second, or 38 calories per square centimeter per year. If not dissipated by circu- lation and mixing, this heat flow would lead to warming of the deep and bottom water during its passage from the Antarctic to the equator. From considerations of meridional circulation, observed temperature gradients and mixing in the deep sea, Revelle and Maxwell estimate that the deep water is replenished in less than 1000 years. 2. Secular change of oxygen: Worthington (1954) has shown that the North Atlantic Deep Water has suffered a loss of dissolved
50 Atomic Radiation and Oceanography and Fisheries oxygen of about 0.3 ml/L over the last twenty years. Assuming a steady rate of attrition he computes that the date at which this water was saturated, presumably while at the surface, was about 1810. A further study (Worthington, 1955) suggests that the Caribbean Deep Water was formed at the same time. Thus it seems possible that formation of the North Atlantic Deep Water, which composes about half of the contents of the Atlantic, is not continuous but sporadic. 3. Carbon-14 dating: In recent years the tech- niques of carbon-14 age determination have been applied to deep sea water samples. The most reliable measurements (Rubin, unpub- lished), of samples from east of the Lesser Antilles, show the carbon at 1750 meters to be about 200 years older than the surface carbon. Present estimates of the age of deep waters are based primarily on measurements in the North Atlantic and on geochemical calculations for the entire world ocean. That the deep cir- culation of the Pacific is significantly slower than that of the Atlantic is suggested by the apparent absence of regions of deep and bottom water formation in the Pacific and the rela- tively high nutrient salt content and low dis- solved oxygen content of deep Pacific waters. In order to determine whether the deep waters of the Pacific would provide a longer period of isolation for radioactive wastes than elsewhere, deep Pacific oceanographic data must be care- fully scrutinized. REFERENCES BOHNECKE, G. 1936. Atlas: Temperatur, Salz- gehalt und Dichte an der Oberflache des Atlantischen Ozeans. Deutsche Atlantische Exped. Meteor, 1925-27, Wiss. Erg. 5: vii + 76 pp. BOWDEN, K. F. 1954. The direct measurement of subsurface currents in the oceans. Deep- Sea Res. 2:33-47. CROMWELL, T. 1953. Circulation in a meri- dional plane in the central equatorial Pa- cific. /. Marine Res. 12:196-213. DEACON, G. E. R. 1933. A general account of the hydrology of the South Atlantic Ocean. Discovery Rep. 7:171-238. DEFANT, A. 1955. Die Ausbreitung des Mit- telmeerwassers im Nordatlantischen Ozean. Pap. Mar. Biol. and Oceanogr., Deep-Sea Res., suppl. to 3:465-470. DEUTSCHEN SEEWARTE. 1942. Weltkarte zur Ubersicht der Meeresstromungen. Deutschen Seewarte No. 2802. DIETRICH, G. 1939. Das Amerikansiche Mit- telmeer. Gesellsch. Erdkunde zu Berlin, Zeitschr., 108-130. ECKART, C. 1948. An analysis of the stirring and mixing processes in incompressible fluids. /. Marine Res. 7:265-275. FRANCIS, J. R. D., and H. STOMMEL. 1953. How much does a gale mix the surface layers of the ocean. Quart. J. Roy. Me- teorol. Soc. 79:534-536. FUGLISTER, F. C. 1951. Multiple currents in the Gulf Stream System. Tellus 3(4): 230-233. GOLOLOBOV, Y. K. 1949. Contribution to the problem of determining the age of the present stage of the Black Sea (in Rus- sian). Dokl. Akad. Nauk SSSR 66:451- 454. HIDAKA, K. 1954. A contribution to the theory of upwelling and coastal currents. Trans. Am. Geophys. Union 35(3) =431-444. ISELIN, C. O'D. 1936. A study of the circula- tion of the western North Atlantic. Pap. Phys. Oceanogr. Meteorol. 4(4): 1-101. JAPANESE FISHERY AGENCY. 1955. Report on the investigations of the effects of radia- tion in the Bikini region. Res. Dept, Japanese Fishery Agency, Tokyo, 191 p. KETCHUM, B. H., and W. L. FORD. 1952. Rate of dispersion in the wake of a barge at sea. Trans. Am. Geophys. Union 33 (5) :680-684. KETCHUM, B. H., and D. J. KEEN. 1953. The exchanges of fresh and salt waters in the Bay of Fundy and in Passamaquoddy Bay. /. Fish. Res. Bd. Can. 10:97-124. 1955. The accumulation of river water over the continental shelf between Cape Cod and Chesapeake Bay. Pap. Mar. Biol. and Oceanogr., Deep-Sea Res., suppl. to vol. 3: 346-357. KOENUMA, K. 1939. On the hydrography of south-western part of the North Pacific and the Kuroshio. Kobe Imper. Marine Observ., Memoirs 7:41-114. McEwEN, G. F. 1934. Rate of upwelling in the region of San Diego computed from
Chapter 4 51 Transport and Dispersal serial temperatures. Fifth Pac. Sci. Congr., Toronto 3:1763. MIYAKE, Y., Y. SUGIURA, and K. KAMEDA. 1955. On the distribution of radioactivity in the sea around Bikini Atoll in June, 1954. Pap. Meteorol. Geophys., Tokyo 5(3,4):253-362. MONTGOMERY, R. B. 1938a. Fluctuations in monthly sea level on eastern U. S. coast as related to dynamics of western North Atlantic Ocean. /. Marine Res. 1:165- 185. 1938b. Circulation in upper layers of south- ern North Atlantic deduced with use of isentropic analysis. Pap. Phys. Oceanogr. Meteorol. 6 (2): 1-5 5. MUNK, W. H., G. C. EWING, and R. R. RE- VELLE. 1949. Diffusion in Bikini Lagoon. Trans. Am. Geophys. Union 30(1): 59- 66. REVELLE, R., and A. E. MAXWELL. 1952. Heat flow through the floor on the eastern North Pacific Ocean. Nature 170:199. REVELLE, R., T. R. FOLSOM, E. D. GOLDBERG, and J. D. ISAACS. 1955. Nuclear science and oceanography. United Nations Inter- national Conference on the Peaceful Uses of Atomic Energy, Geneva, Paper no. 277: 22 pp. RILEY, G. A. 1951. Oxygen, phosphate and nitrate in the Atlantic Ocean. Bull. Bing- ham Oceanogr. Coll. 13, Art. 1:1-126. SAITO, Y. 1951. On the velocity of the vertical flow in the ocean. /. Inst. Polytech., Osaka City Univ. 2(1) Ser. B.:1-4. SEIWELL, H. R. 1934. The distribution of oxygen in the western basin of the North Atlantic. Pap. Phys. Oceanogr. Meteorol. SELIGMAN, H. 1955. The discharge of radio- active waste products into the Irish Sea. Part 1. First experiments for the study of movement and dilution of released dye in the sea. Proceedings of the International Conference on the Peaceful Uses of Atomic Energy, Geneva. Paper 9:701-711. SVERDRUP, H. U., M. W. JOHNSON, and R. H. FLEMING. 1942. The oceans, their physics, chemistry and general biology. Prentice- Hall, N. Y, x+1087 pp. SWALLOW, J. C. 1955. A neutral-buoyancy float for measuring deep currents. Deep- Sea Res. 3:74-81. U. S. ATOMIC ENERGY COMMISSION. 1956. Operation Troll. U. S. Atomic Energy Commission, New York Operations Office, NYO 4656, Ed. by J. H. Harley, 37 pp. U. S. NAVY HYDROGRAPHIC OFFICE. 1947a. Atlas of surface currents, north Atlantic Ocean. H. O. Pub. 571, First Ed., re- printed 1947. 1947b. Atlas of surface currents, northeast- ern Pacific Ocean. H. O. Pub. 570, First Ed. 1950. Atlas of surface currents, northwestern Pacific Ocean. H. O. Pub. 569, First Ed., reprinted 1950. VON ARX, W. S., D. F. BUMPUS, and W. S. RICHARDSON. 1955. On the fine structure of the Gulf Stream front. Deep-Sea Res. 3:46-65. WOOSTER, W. S., and F. JENNINGS. 1955. Exploratory oceanographic observations in the eastern tropical Pacific, January to March, 1953. Calif. Fish and Game 41: 79-90. WORTHINGTON, L. V. 1954. A preliminary note on the time scale in North Atlantic circulation. Deep-Sea Res. 1:244-251. 1955. A new theory of Caribbean bottom- water formation. Deep-Sea Res. 3:82-87.