400 km-1

Figure 6.18 (a) Initial development of a transient crater in thin oceanic lithosphere. (b) Simplistic representation of the effects of a major impact. (b1) Section showing the development of an impact crater and the modes of deformation that may operate for 1 Ma. (b2) Section showing break-up and foundering of oceanic lithosphere, thereby allowing molten asthenosphere to reach surface. (c) Plan view showing radiating vertical tension fractures and arcuate failure lines which permit cantilever sections of oceanic lithosphere to break away and founder.

10syr Boundary

400 km-1

Figure 6.18 (a) Initial development of a transient crater in thin oceanic lithosphere. (b) Simplistic representation of the effects of a major impact. (b1) Section showing the development of an impact crater and the modes of deformation that may operate for 1 Ma. (b2) Section showing break-up and foundering of oceanic lithosphere, thereby allowing molten asthenosphere to reach surface. (c) Plan view showing radiating vertical tension fractures and arcuate failure lines which permit cantilever sections of oceanic lithosphere to break away and founder.

explosive elimination. In a few tens of seconds the lithosphere, over an assumed diameter of 300 km, is vapourised, melted or comminuted and ejected into the atmosphere and stratosphere.

In the central zone of the crater, the asthenosphere will also be ejected. This part of the asthenosphere that has hitherto been subjected to a confining pressure of about 35 kb at a depth of 100 km, is suddenly subjected only to atmospheric pressures which, instantaneously, result in pressure-release melting in the upward cascade of blocks, blebs and particles that make up the ejecta.

The asthenosphere that remains beneath the transient crater floor also experiences instantaneous pressure-release melting, which is likely to develop to a depth of over 200 km, where it will be subjected to a release of vertical pressure from 35 to 70 kb, and to an ejecting pressure of at least an order of magnitude greater than that experienced in the most energetic of volcanological events. This will cause the crater floor to flow rapidly, even explosively, upward and the crater walls to collapse inward.

The crust and perhaps part of the upper 17.5 km forming the strongest part of the lithosphere will be blown upwards and outwards, possibly to form a recumbent isoclinal fold, the upper limb of which will be inverted (Figure 6.18a). Within the limits of this inverted limb, the thickness of the 'strong' layer may be doubled, and could, temporarily, stand as much as 8 km above the original base level of the oceanic crust.

In the walls of the crater, the weak asthenosphere below the remaining lithosphere, beyond the transient crater, will be eased sideways into the initial crater. The lithosphere adjacent to the crater walls loses its lateral cohesion and basal support and, being more dense than the asthenosphere, will tend to slide towards the crater and sink into the molten upward and sideways flowing asthenosphere, which is now rapidly rising towards the surface.

Soon, some, or even most, of the ejecta begins to return to the crater. That proportion of the ejecta that originated as 'cold' lithosphere and maintained its unmolten form will fall back into the melt and, because it is more dense, will sink. The returning blebs of molten material, whether they originated in the lithosphere or the asthenosphere, will return to the upwelling flood of molten material from the asthenosphere and will meld with it to enhance the upward flow of magma.

The molten material will occupy a slightly larger volume than the original unmolten material, so that the crater will eventually fill beyond its original level, and extrusive lava flows will begin to develop and extend over undeformed ocean floor. Soon, the former limits of the crater are hidden, or destroyed.

In addition (see Figure 6.18b(2)), melt extrudes up vertical fractures which develop between the sinking blocks, so that the 'igneous body' rapidly extends well beyond the crater's original bounds.

On impact, the pressure-wave blows away the ocean in a tsunami. However, the ocean returns in a matter of hours and rapidly re-establishes itself, so that the lava flows, being emplaced at and after this time, are rapidly cooled and develop into steep-faced plateau basalts.

White and McKenzie (1989) suggest that, in their p-model, the proportion of melt that is produced by sudden pressure-release is about 30 per cent of total volume experiencing pressure-release. One may infer that these authors tacitly assume that the melt gradually collects together and rises to the surface as magma. In the impact model proposed above, the situation is different. There is no retaining layer in the upper level of the plate, keeping the 'lid' on the molten material. In this model, the lid is blown away so that even if only 30 per cent of the volume of rock affected by pressure-release results in molten material, this will give rise to a weak solid/melt slurry that is unconfined at the interface, as defined by the transient crater; so that the slurry of solid/melt may be ejected, without substantial differentiation of the melt. The melt that is generated beneath the walls and floor of the transient crater migrates rapidly inwards and upward. In so doing, it causes the ambient pressure in the lower, and laterally more distant, areas to experience a smaller rate of flow. Progressively slower and slower migration of melt material will work its way towards the surface. Eventually, and it may take a million years, the inward and upward migration of molten material slows and eventually ceases. An indication of the mode of this evolution is shown in Figure 6.18b. If the migration zone extends as indicated in this diagram, the volume of erupted rock will be of the order of 35x106 km3.

A secondary mode of eruption will also come into play. This is related to the radial, vertical tensile fractures which develop in the oceanic lithosphere around the impact. These features, which are discussed in Chapter 5, where they were formed as the result of the Snowball 500 ton explosion, are also represented in Figure 6.18(b2) and (b3). These vertical fractures can be attributed to the tensile stresses generated normal to a vertical line passing through GZ. The tensile stresses only occur when the stress intensity of the compressive pulse falls to a level where an elastic response takes over. This response probably occurs when the radiating stress-pulse (Pr) has a magnitude up to 100-1000 kb. (We discuss the mode of development of these vertical fractures later in this chapter, where we will argue that transient open fissures can develop down to depths of about 140 km.)

The lithosphere is relatively cold and of higher density than the asthenosphere. Consequently, when the radiating fractures are generated, these planes form paths of easy movement for asthenospheric migration. The cold lithosphere, near the crater boundary, will begin to sink, so that a series of cantilevers develop. The tensile stresses that form at the upper surface of the cantilever can cause circumferential as well as radiating fractures, so that quite large areas of lithosphere will break away and founder, to be flooded by a corresponding volume of asthenosphere rising to the surface, where it forms part of the erupted mass.

Because the thickness of the lithosphere is about 50 km, it will take time for the temperature in the sinking body to equilibrate with that in the asthenosphere, so that it eventually ceases to sink. This mechanism, which helps the upward migration of the asthenosphere towards the surface, can give rise to a further volume of extruded asthenosphere which may be comparable with that generated by the transient crater.

Hence, we suggest that the Ontong-Java PB, and other similar features, result from impacts in extremely thin oceanic lithospheres and are caused by only modestly large impacting bodies, which give rise to a transient crater of perhaps 300 km diameter.

We have invoked the assumption that our model, which represents the Ontong-Java PB, results from a major impact event. From what we have already learned, it is to be expected that a record of this event will be manifest in the track of a known point (an island currently adjacent to the Mariana Trench) in the general vicinity of the erupted body. Such a track is shown in Figure 6.15a, where it will be seen that, from 123.0 to 119.3 Ma, the oceanic lithosphere moved westward (270°) at a uniform speed. From 119.3 to 115 Ma, the track runs at about 240° at a slightly slower rate. We can infer from this track that a major impact occurred at a modest distance from the island marker at 119.3 Ma. This figure is close to, but slightly older than, the age (as cited in the earlier literature) of samples taken from the Ontong-Java PB. The more up-to-date age puts the emplacement of the Ontong-Java PB at 121.6-123.2 Ma, which is over 2.6 Ma earlier than the inferred impact. However, the oldest biostratigraphic ages of basal sediments recorded on the Ontong-Java PB is 117.5 Ma. We suggest that the precise timing of the impact and, possibly, the initiation and development of the Ontong-Java PB has not yet been exactly established.

Coffin and Eldholm note that, in addition to the main development of the plateau basalts at about 120 Ma, there was also a later eruption at about 90 Ma; and this too, they suggest, resulted from the emplacement of a minor plume.

We have argued above that the main Ontong-Java PB was the result of a major cometary impact. Rather than assume that the later eruption was the result of a plume, it was reasonable to ascertain whether or not the 90 Ma eruption could have been induced by a relatively minor impact. Accordingly, we plotted the track of the NW tip of Bougainville Island (Figure 6.15b). This island lies immediately to the south of the Ontong-Java PB. It will be seen from Figure 6.15b that from 96 to 93 Ma the tip of Bougainville Island defined a linear track, at a constant velocity. At 93 Ma, however, there was a sudden 3° change in direction of track, which continued at least to 88 Ma. There was no detectable change in the velocity of motion.

We suggest that this behaviour is consistent with the results of a relatively minor impact which probably gave rise to a crater, possibly a little smaller than 100 km in diameter. Such a crater could require a significant period during which basalt was erupted at a relatively slow rate, compared with that associated with the earlier major eruptive event.

We therefore conclude that the evidence supports the thesis that both periods of eruption of the Ontong-Java PBs are the result of impacts. We further suggest that the existence of three basin flood basalts, referred to earlier, which developed in the vicinity of the Ontong-Java FB, lends further support to our argument.

As regards the size of the earlier impacting body, a large meteor, with a diameter approaching 40 km, would experience relatively modest retardation, even if it struck in deep ocean. It is sufficient to note here that the model we proposed for the generation of the Ontong-Java event not only required that the strike took place close to a spreading-ridge, but that it also occurred where the depth of ocean was relatively shallow. That is, the depth was only about 1-2 km. This is confirmed, to some extent, by the observation that the upper surface of the Ontong-Java PB has been eroded, indicating that the basalts once stood above sea-level.

Continental lithosphère

Pressure release / Melt zone

Continental lithosphère a


plume ?


Pressure release / Melt zone b plume ?

Figure 6.19 (a) Section showing transient crater with a diameter of 300 km which has developed in continental lithosphere with a thickness of 100 km. The volume of melt will be significantly less than that represented in Figure 6.18. Moreover, the melt produced will rise diapirically to produce relatively modest eruption of basaltic material. (b) The impact could induce the generation of a plume. As the plate moves away over the impact site a remnant hotspot may exist for many tens of millions of years. The black central area represents a 'plug' of the type inferred to exist off-shore of Bombay (after Negi et al., 1993).

It can readily be inferred from the arguments presented above that large volumes of melt can best develop in oceanic lithosphere if the impact occurs near the ridge, where the oceanic lithosphere is thin and where the depth of water is relatively shallow, so that the transient crater can extend well down into the asthenosphere.

Several of the large oceanic plateau basalts, such as the Ontong-Java, Shatsky and Minihiki PBs, are all surrounded by oceanic lithosphere only a little older than the various cited extrusive bodies, so it can be inferred that they formed when the ocean lithosphere on which they are set was relatively young, and therefore thin.

Mature oceanic lithosphere will approach 100 km in thickness, and so approximates more closely to the average thickness of continental lithosphere. It can readily be inferred from Figure 6.19a that it requires a transient crater diameter of at least 300 km to reach and breach the asthenosphere with a thickness of 100 km. The volume of melt produced by such an impact will be small, relative to that generated by the model shown in Figure 6.18. Also, it will be noted that the upwelling melt will mainly be restricted to the axial regions of the crater, so that the intrusion will be more likely to resemble diapiric intrusion (Figure 6.19b) than the broad upwelling from the base and the walls of the crater represented in Figure 6.19a.

One may infer from Figure 6.19, that impacts of somewhat smaller energy, which produce transient craters that do not completely penetrate the lithosphere, whether it be oceanic or continental, may result in a hotspot which produces a more modest volcanic activity at the surface, but which may possibly induce the development of a plume of modest vertical extent. Alternatively, as was noted in the previous chapter, larger impacts may generate antipodal hotspots.

However, even if an impact causes a transient crater of, say, 400 km diameter, so that the asthenosphere is exposed to a depth of about 33 km (assuming the continental lithospheric thickness to be 100 km), although a considerable volume of pressure release melt will be engendered, it will not give rise to the same degree of extrusion that would be associated with an oceanic event with comparable dimensions of crater diameter and lithospheric thickness. Overall, the density of continental lithosphere is significantly less than that of the melt formed in the asthenosphere, so, as indicated in Figure 6.19a, the walls will not sink in the

Was this article helpful?

0 0

Post a comment