ibrium with the surface water, it would have about the same density in August, but would be less dense (by 2.2 or units) in February. Therefore, any mixture of deep and surface water will produce a neutrally or positively buoyant water mass, provided its temperature is raised to that of the surface. Comparison of temperature measurements made on the discharged water (Table 1) with the prevailing in situ deep-water temperatures (Fig. 1b) indicates that some warming occurred in the pipeline during the airlift pumping process, probably due to frictional effects and heat conduction.

TABLE 1. PROPERTIES OF SURFACE WATER, OF DEEP WATER, AND OF
FINE PARTICULATE MATTER SUSPENDED IN IT

All analyses are the mean of duplicate determinations. All results are given in μg at/liter, except temperature, salinity and o *Corrected by in-field calibration error.

To obtain a visual record of the fate of the surface-discharged water, a 3-minute injection of a 15-percent solution of Rhodamine dye was introduced into the ship's outboard discharge. Observations of the behavior of the dye were made from aboard ship and by a series of photographs taken from an aircraft at 200 to 300 m altitude. Some 9,500 liters of dyed deep water were discharged in 3 min: Figure 2 shows the resulting dye patch 8 min later, estimated to be an area approximately 10,000 m2, in which the main body of dye remained in the top few meters of the sea. Heavier sedimentary material appeared to sink, and lighter, air-filled pteropod shell material floated at the surface. The latter is suggestive of Langmuir cells, frequently seen during dye diffusion experiments and considered to be an important mechanism in surface-mixing processes (Assaf et al., 1971). Four hours later, the dye patch appeared to cover about twice the area and its depth was estimated to be approximately 10 m by visual observation from the deck.

The properties of the surface and discharged water measured at the time of the dye experiment are the basis for the information on mixing given in Figure 3 which shows the degree of warming needed to bring different mixtures of discharged and surface water to the same density as the surface water. Sinking of a mixture containing 20 percent deep- and 80 percent surface-water would be limited by its density of 23.1 to the upper mixed layer (about 5 to 20 m at this season; below these depths the density gradient increases rapidly).

31-659 O 74 16

Continued mixing with surface water, spreading, heating by solar radiation and by conduction from the surrounding water, generation of convection cells within the discharged water mass, wind and wave action, are some factors complicating this simplistic model. There would be less heat conduction in a larger diameter suction-dredge pipe, resulting in a denser discharge which would be more likely to sink unless it were sprayed over the sea surface to ensure rapid mixing and warming. This technique could be used in other parts of the world's oceans where manganese nodules are found (Horn et al., 1972) and where the vertical temperature, salinity and density profiles differ considerably from those on the Blake Plateau.

We concluded that under the experimental conditions of the Blake Plateau mining test, the discharged water will remain at or near the surface during August, when the sea-surface density is at its minimum, and will have an even greater tendency to do so in other seasons.

Because the most likely areas of manganese nodule mining in the deep sea have oceanographic characteristics different from those prevailing in the region of the Blake Plateau where the pilot-scale experiment was conducted, we undertook model studies to predict the amount of warming that would be required in different seasons to bring bottom water to the same density as surface water for the North Atlantic. The same model can, of course, also be applied to other oceanic areas. To do this, we profiled all data filed at the National Oceanographic Data Center for North Atlantic hydrographic stations with water column data to within 300 m of the bottom. This pointed out the extreme lack of data in the Atlantic, from which only 1600 stations meeting our criterion (sampling within 300 m of the bottom) were available.

We calculated the amount of warming that would be required in summer (Fig. 4) and in winter (Fig. 5) to bring such bottom water to the same density as the surface water. A large body of the ocean would have to be warmed by 20° or more, assuming no mixing with the surface and no heating in the airlift pumping process. Since bottom conditions do not vary much seasonally, the differences are due to seasonal variations in surface conditions.

In some areas of the ocean, however, bottom water is more saline than surface water; therefore its temperature would have to be raised higher than that of the surface water to keep it there. Figures 6 (July-Aug-Sept) and 7 (Jan-Feb-Mar) chart the differences between bottom and surface waters in σ units. Where this difference is negative, the water would sink unless there was warming and mixing with surface water. In summer, large areas of very low salinity water, originating from the Amazon discharge, sometimes become isolated within the equatorial flow. These layers of low salinity water largely disappear in winter.