Where is the minimum of gravity

July 1981: Diving expedition to the East Pacific threshold

The biological community now discovered by us turned out to be very similar to the one that had emerged in 1977 on the Galapagos spreading axis. A certain type of characteristic brown mussel was missing, but otherwise the local sea anemones, the worms from the family, seemed to be Serpulidaewho have favourited crabs from the family of Galatheidae and out of submission Brachyura, as well as the large mussels and giant tube worms to be identical with their conspecifics off the Galapagos. Each colony covered an area about thirty by one hundred meters. It was not the warmth of the hydrothermal solutions that attracted the animals, but the abundant food. In the immediate vicinity of the chimneys there is a hundred times greater supply of nutrients than in the surrounding waters.

Robert R. Hessler of the Scripps Institution and J. Frederick Grassle of the Woods Hole Oceanographic Institution and others have studied the food chain of this extraordinary ecosystem. In the beginning there are chemosynthetic bacteria, which oxidize hydrogen sulphide, which escapes from the openings in the ground, to sulfur and various sulphates. They use the energy released during oxidation to build up the body's own organic substances from carbon dioxide and water. Most of the larger organisms feed on the bacteria. They filter them out of the water or live in symbiosis with them. Some of the animals are scavengers or predators. The communities do not depend in any way on photosynthesis (i.e. solar energy); the energy from which they draw comes from the interior of the earth alone. The fact that such communities have been discovered both on the Galapagos splitting axis and around three thousand kilometers away on the ridge of the East Pacific Sill suggests that they may populate a large part of the Rift system worldwide. Of course, their existence is always threatened, as the activity of the hydrothermal vents is linked to the sporadic volcanic cycles. Isolated heaps of empty mussel shells of the same size are silent witnesses of local mass extinctions.

Geochemical inferences

The discovery of hydrothermal vents on the two splitting axes in the Pacific overturned the old theories about the chemical balance of the oceans. It used to be the opinion that the amounts of substance that are fed into the sea (mainly from the rivers) and withdrawn (through the deposition of sediments and through low-temperature reactions between seawater and the seabed) are balanced. When it became clearer how common the individual minerals are and which low-temperature reactions take place in detail between the seawater and the sediments or volcanic rocks, there were deficits in the balance for certain elements. For example, the rivers wash larger amounts of magnesium and sulphate ions into the oceans than this basalt weathering can get rid of. Likewise, much more manganese appears to accumulate on the sea floor than the rivers as a whole bring in.

The hydrothermal seawater cycle along the submarine rift systems offers a way out of this dilemma in that it brings a completely new component into play: exchange processes between liquid and solid substances at high temperatures. According to John M. Edmond of the Massachusetts Institute of Technology, reactions between hot seawater and basalt rock can convert dissolved sulfates into solid sulfate and sulfide minerals. In similar reactions, magnesium and hydroxide ions (OH-) are removed from the seawater and incorporated into hydrothermal clays. During these reactions, the hot sea water is transformed into a reducing, acidic solution that leaches calcium, silicon, manganese, iron, lithium and other positively charged ions from the rock and flushes them into the oceans. In this way, hydrothermal processes can balance the budget for the most important mineral components of seawater and also provide an explanation for the observed concentration and distribution of numerous minor components and trace elements.

As Edmond noted, the hydrothermal solutions that emerge at the Galapagos Threshold mix with normal seawater as they rise through the volcanic rock. At the same time, their temperature drops and minerals are already deposited within the rocks. In contrast, the high temperatures and the chemical composition of the springs on the East Pacific Sill indicate that the hydrothermal solutions there do not noticeably mix with cold sea water on their way to the sea floor. These solutions therefore represent the true hydrothermal contribution to the marine chemical cycle. The water that flows out of the chimneys naturally also originates from the sea. However, it descends almost to the magma chamber before flowing back to the sea floor.

As soon as the undiluted hydrothermal solutions come into contact with cold, alkaline seawater on the sea floor, fine-grained iron and zinc sulphide precipitates appear, which color the rising water black. Preliminary analyzes by Rachel Haymon and Miriam Kästner of the Scripps Institution indicate that the elevations and the walls of the chimneys protruding from the ground consist mainly of zinc, iron and copper sulphide, as well as calcium and magnesium sulphate. How the minerals are formed in detail, with what speed they are deposited and in what proportions water and rock are at the individual points of the system, is the subject of hot debates among geochemists today. One thing is beyond doubt, however: hydrothermal systems will play a central role in all future models of ocean chemistry.

Geophysical experiments

Our geophysical research program was primarily geared towards telling us more about the relationship between the presumed axial magma chamber and the tectonic, volcanic and hydrothermal processes on the sea floor. We were fortunate enough to be able to conduct our experiments in an area where hydrothermal activity was in full swing. So we pulled out all the stops of modern experimental geology to explore the structures below the sea surface. We measured the speed of propagation of seismic waves, the frequency and strength of earthquakes, gravity anomalies, electrical conductivity and the direction of magnetization in the rock. Theoretically, it is possible to calculate how quickly newly formed crustal rock cools down through heat conduction. The calculation shows the heat flow within the mid-ocean ridges due to heat conduction. When measuring the actual heat flow on the ridge of such a ridge, however, one obtains values ​​that are almost an order of magnitude below the theoretically predicted. Are the models incorrect, or does the circulation of the ice-cold seawater within the hot, newly formed crust accelerate the cooling process by dissipating heat through convection? How deep does the seawater penetrate the oceanic crust, and how big is the area in which the circulation takes place? What is the chemical composition of the crust involved in this cycle, and which minerals are deposited? The answer to all of these questions depends crucially on how deep the cracks and fissures in the rocks of the splitting axis extend. A limit value is obtained by determining the speed of earthquake waves, the electrical conductivity and the local gravity anomalies.

The large discrepancy between the heat flux measured in the sea floor and the value that results from the models of the cooling of the lithosphere due to heat conduction suggests that at least one third of the heat loss at the mid-ocean ridges is not due to heat conduction, but rather due to hydrothermal circulation comes about. At the Galapagos spreading axis, where hydrothermal activity was observed for the first time, the heat flow was unfortunately difficult to estimate due to unclear circulation conditions. The conditions for such measurements were much more favorable on the East Pacific Sill.

Using film and video recordings from the chimneys, which we looked at very closely, we estimated the flow speed of the exiting water streams. With an average exit speed of two or three meters per second, a single vent provides a heat flow of around sixty million calories per second. That is three to six times the amount of heat that a one-kilometer-long section of the mid-ocean ridge with an extension of thirty kilometers on each side theoretically emits. We found at least twelve larger chimneys in the southwest part of our investigation area. The total heat loss is therefore obviously very high - so high that such a chimney cannot be active for long. It will probably go out after a few years.

The vents seem to be concentrated in a narrow strip that is six kilometers long but only a few hundred meters wide and lies within the youngest volcanic zone. In this strip we discovered a temperature deviation in twelve places and made sure by taking photographs that there were actually hydrothermal vents at these places. We visited and explored eight of them with the Alvin.

Seismic measurements

John A. Orcutt and one of us (Macdonald) devised a particularly nifty seismic experiment to determine the depth of the cracks and crevices in the rocks along the ridge line. We wanted to measure the speed of propagation of seismic waves as a function of depth with high resolution. If an explosive charge is ignited on the surface of the water, the explosion waves spread out spherically and are mainly thrown back from the closest topographical surfaces. They only provide a rather blurred picture of the inner structure of the top thousand meters of the earth's crust.

In order to circumvent these difficulties, we had to come up with a method of placing both the source and the receiver of the seismic vibrations on the sea floor. It was also necessary to measure the propagation times of the vibrations to an accuracy of about a thousandth of a second. With the Alvin both problems could be solved. Since explosives are practically useless as seismic sources in the deep sea due to the high water pressure, we attached to the Alm a jackhammer to generate the desired seismic vibrations with it. In order to calibrate the seismometers anchored earlier on the seabed, we drove with the Alvin to within two meters of each one and gave a blow with the jackhammer; the shock was both from a receiver on board the Alvin as well as recorded by the seismometers. At the end of the dive we turned around again to calibrate each seismometer a second time.

During four dives we created two complete seismic refraction profiles: one thousand meters long parallel and one eight hundred meters long across the ridge line. The analysis of the data obtained is not yet complete, but for the waves that ran along the surface of the ridge line, a provisional value for the propagation speed can simply be calculated from the transit times of the signals. It is 3.3 kilometers per second and is thus far below the laboratory value for basalt rock at the same pressure (approximately 5.5 kilometers per second). The reason for this is obviously the strong fissures and porosity of the rock. Although there were neither large trenches nor crevices on our way, there were numerous hairline cracks and pores in the lava pillows. Before we can give more precise information about the degree of fissure and porosity, we have to wait until the physical properties of the rock samples have been determined and the results of the seismic analyzes over greater distances are available. In particular, we are excited to find out at what depth the speed of the seismic waves increases over five kilometers per second. That would be a sign that most of the cracks and crevices have closed there.

The results of an earlier seismic refraction experiment had already indicated that there was a magma chamber two to three kilometers below the hydrothermal field. In this less precise, but larger-scale experiment, explosives were detonated from a ship on the surface of the water up to sixty kilometers away from the triangular seismometers on the sea floor. At that time, the scientists registered an area just two thousand meters below the sea floor in which seismic compression waves were slowly spreading - a sign that partially molten rock had accumulated here. At points ten kilometers from the center line of the spreading axis, however, the speed of seismic waves showed values ​​that were normal or even somewhat too high for sea basalt. The magma chamber thus appears to be limited to a twenty kilometers wide zone centered below the ridge line of the East Pacific Ridge.

Earthquakes and volcanism

A second clue to the existence of such a magma chamber emerged when we examined the propagation of shear waves such as those that occur in earthquakes. Such waves are dampened by the partially molten rock inside a magma chamber. In fact, John A. Orcutt, Ian Reid and William A. Prothero Jr. found that shear waves that ran directly along the spreading axis were greatly weakened, whereas those ten kilometers away were hardly any.

James W. Hawkins of the Scripps Institution received from the Alvin collected rock samples analyzed. His results also indicate that a narrow magma chamber extends below the expansion zone at a shallow depth. The basalt samples, which all came from a six-kilometer strip across the young volcanic zone, did not differ greatly in their composition. Obviously they all came from the same magma from which the minerals olivine and plagioclase crystallized one after the other at relatively low pressure. This is compatible with the assumption that the magma chamber is less than six kilometers deep, which means that its ceiling can also be only a few kilometers thick. Apparently this thin "skin" is so fissured that seawater can penetrate deep enough to heat up to at least 350 degrees.

But how far does the water circulation extend into the depths? Last summer we returned to our place of work to investigate tiny earthquakes in the hydrothermal field as a supplement to our seismic measurements on the sea floor. We brought seven seismometers into position with the aid of acoustic signal answering devices that were fixed to the ground and which we had left behind to mark the hydrothermal vents. If the hydrothermal circulation had a telltale seismic signature, it might be possible to determine its depth. The results so far are encouraging. The earthquake foci in this area seem to be very close to the surface - at most two to three kilometers deep. This shows once more how thin the ceiling of the magma chamber must be. In addition, we know that the cracks also extend to this depth at the most.

Among the recorded seismic events, we noticed weak, very characteristic tremors, which are known as harmonic tremors designated. They are an unmistakable sign of imminent or past volcanic eruptions and occurred, for example, in connection with the devastating eruption of Mount St. Helens a good year ago (see "The eruptions of Mount St. Helens" in Spectrum of Science, May 1981) . They became more and more common just before the outbreak and eventually grew to an almost continuous tremor. We observed something similar on the East Pacific Sill. Up to several hundred vibrations per hour occurred in the examined section. It may well be that a volcanically active phase is just beginning there - or that it is coming to an end.

Gravity anomalies

Closely related to the seismic experiments were a series of gravity measurements, the Spiess and one of us (Luyendyk) with the Alvin carried out. Fluctuations in the density of the earth's crust, which result from the fissuring of the ground or a magma chamber close to the surface, should not only change the speed of seismic waves, but also the local gravitational field. The expected gravity anomalies are of course small and difficult to measure from the sea surface , since the source and sensor are far apart in this case and sensors on board a ship also register a lot of spurious acceleration signals. Here, too, offered the Alvin a way out.If you sat quietly with it on the seabed and measured the force of gravity from there, there were practically no disturbing accelerations. Because of the closer proximity to the source, the signals were also stronger.

Gravitational measurements along a seven-kilometer profile extending from Zone 1 to Zone 3 revealed a pronounced negative gravity deviation over the young volcanic zone. It appears that the anomaly has its maximum above the central ridge line and extends almost to the edge of Zone 2. The sub-par density expressed in it could be caused by either a fracture in the ground or a near-surface magma chamber. In fact, as the appearance shows, the fissure is greatest in zone 2, while the gravity deviation is at its maximum in the center of zone 1, which has relatively few cracks. This leads to the conclusion (without proving it) that the gravity anomaly originates from a magma chamber near the surface.

Suppose the magma chamber had the shape of a horizontal cylinder, the axis of which runs parallel to the ridge line and coincides with the minimum of gravity. Then the center of the cylinder should be about a thousand meters below the sea floor. If one also assumes that the chamber is filled with molten basalt, its density should be 0.21 grams per cubic centimeter below that of the surrounding rock. It would follow that the top of the cylinder would be about six hundred meters below the sea floor. On the other hand, assuming that the density difference between magma and the surrounding rock is smaller, the result would be a larger magma body that would come even closer to the surface of the sea crust.

However, according to the seismic results, the magma chamber is much larger and deeper. Geological reasons suggest that the permanent main part of the chamber lies two to six kilometers below the sea floor and is two to three times as wide as it is deep. What is expressed in the gravity data should not be the main chamber itself, but a small, temporary, elongated dome or bulge on the apex of the main chamber. It could fill the subsurface of the entire zone 1 and feed the lava flows discovered there with its magma.

The amount of the negative gravity deviation shows that on every meter of the threshold a mass of ninety million kilograms is missing below the ridge line of the spreading axis. If the crust were in isostatic equilibrium at this point (gravity and buoyancy were therefore in balance), there would have to be an excess of mass somewhere on the ocean floor, which compensates for the deficit in the depth. Zone 1 is a clod about one kilometer wide and about twenty to thirty centimeters above the neighboring soil. In order for isostatic equilibrium to prevail at the ridge line, it would have to be twice as high. Either the mass deficit is compensated for by other topographical peculiarities at a greater distance from the ridge line, or the friction on the fracture surfaces is so great that the buoyancy forces do not manage to raise the central block any further.

The mean rock density down to a depth of one hundred meters can also be estimated from the gravity measurements. Along the ridge line it is approximately 2.6 grams per cubic centimeter, while around ninety rock samples taken from the bottom had a density of 2.9 grams per cubic centimeter. The difference results in a probable pore content of the soil of around fifteen percent.

Electrical measurements

The pores are filled with water. Since water conducts electrical current, it should be possible to detect it directly through conductivity measurements in the crustal rock. Conductivity measurements at greater depths could also indicate the existence of a magma chamber, since magma has a much higher conductivity than solid basalt. For these investigations, Charles S. Cox of the Scripps Institution developed a new technique for conducting electrical probes. They should give us more information about the seepage of seawater into the crust, the depth of the cracks and crevices, and the lateral extent of the magma chamber. No one had even attempted such measurements before, and no other techniques had ever succeeded in even approximating the conductivity of the deep-sea crust.

The research vessel Melville, operated by the Scripps Institution, pulled an eight hundred meter-long electrical antenna behind it just above the ocean floor. The antenna sent out electrical signals, the frequency of which was chosen so that they were quickly absorbed in the ocean, but could penetrate a considerable distance into the sea crust. Three receivers were set up on the sea floor near the spread axis. Since it was the first time we had traced a long, fragile antenna above the ocean floor, we avoided the uneven surface of Zone 1 and preferred electrical probes in the area ten to fifteen kilometers west of the ridge line in 300,000 to 400,000 years ago Crustal rock in front.

The soundings reached about eight kilometers below the sea floor. The conductivity pattern obtained suggests that the magma chamber is not even two hundred meters thick at a distance of ten to fifteen kilometers from the ridge line, an emphatic confirmation of the seismic findings, which at the time indicated a narrow magma chamber directly below the ridge line. The one from the Cyana Interestingly, from observations made it can be seen that the active distortion of the crust also gradually subsides at a distance of ten to twelve kilometers from the ridge line of the spreading axis. Perhaps the width of the magma chamber also determines the width of the rupture zone on the mid-ocean ridges with a medium to high expansion speed. As the electrical probes also showed, the conductivity is low even relatively close to the sea floor. Apparently the seawater penetrates no more than two to four kilometers deep into the crust.

Magnetic border lines

For another series of dives, we moved our investigation site from the ridge line a little towards the northwest to the point where the last major polarity reversal of the earth's magnetic field was reflected in a magnetic boundary line in the sea floor. We wanted to see whether the geometric course of such a boundary line tells us something about how new sea crust forms at an axis of spread. Because even almost two decades after the Vine-Matthews model was set up, one still does not fully understand how the magnetic stripes are formed. There was a time when it was believed that uniform magnetization within a strip only extended to about five hundred meters. Because when deep-sea drilling into deeper crustal layers a chaotic mess of the most different polarities and field strengths came to light, which no longer bore the slightest resemblance to the regular sequence of magnetic strips as measured from the surface of the sea. On the basis of previous measurements with a magnetometer that was pulled across the ground on the towline of a ship, we first created a three-dimensional mathematical model of the magnetic interface. According to these calculations, it was remarkably straight and narrow: no wider than 1.4 kilometers. Of course, in order to get a consistent solution to the math equations, we had to filter the data. Hence it seemed questionable whether the boundary line was really that sharply defined. To find out, we put a sensitive magnetometer on the Alvin at. It was not only able to measure the magnetic field in the three spatial directions, but also the vertical field gradient: the extent to which the magnetic field changes with the height above the sea floor. Loren Shure from the Scripps Institution and Stephen P. Miller and Tanya M. Atwater from the University of Santa Barbara also took part in this first attempt to explore a boundary on the ocean floor between two strips with reversed direction of magnetization.

In five dives across the borderline area, we were able to clearly determine the direction of magnetization in the basalt at over 250 points. The result was impressive. On both sides, the magnetic field had the correct orientation at each measured point, i.e. its polarity matched the polarity that the tow magnetometer had determined for the respective strip as a whole. On the younger side of the borderline, this observation was not too sensational, because here the newer crust with positive polarity (crust whose polarity corresponds to that of the current earth's magnetic field) should in any case overlay any existing older crust with negative polarity. What surprised us, however, was the fact that the new crust had nowhere penetrated the boundary line to the older side.

We made with the help of the Alvin found that the magnetic boundary line measured directly on the ocean floor runs about five hundred meters further from the axis of spread than the line that we calculated from the data collected by the tow magnetometer. The reason for this is that the tow magnetometer measured the magnetization down to a certain depth and averaged it over the depth; that is, it determined the mean position of the boundary line for a crust cross-section. From the fact that the boundary line directly on the sea floor runs five hundred meters further from the spreading axis than the boundary line averaged over a certain depth, one can conclude that basaltic lava from the volcanic vents poured laterally over the older, negatively polarized crust.

The previous readings and the one now with the Alvin