Geometry of real and 'model' craters

The morphology of craters has been studied in natural examples on the Earth, the Moon, other planets and satellites, and also by means of theoretical and physical models. The theoretical models and many of the

Figure 5.7 Superposition of simple craters, by adjusting the scales of a number of natural impact events (after Grieve and Pilkington, 1996).
Figure 5.8 Flat-lying recumbent fold outside the crater formed by a 100 ton TNT explosion (after G.H.S.Jones).

physical models (especially those using small energy sources) have used a target material which is usually homogeneous and isotropic.

The results of such calculations or experiments are usually in the form of a basin crater (Figure 5.5a). Most natural craters on Earth with a diameter of less than about 3.5 km (e.g. Meteor Crater, Arizona, and Brent Crater, Canada; Figures 5.5 and 5.7) exhibit this geometry. Such craters, whether they be natural, experimental or theoretical, represent the geometry of the initial phase of excavation and development of large real craters. Hence, understanding this initial phase is of critical importance in interpreting the mechanisms involved in cratering, which will be considered later in this chapter.

To understand how natural, circular, impact features can be detected, even after deep erosion, and how they are related to folds and faults which are also generated during impact, let us consider the morphology that develops as the result of explosions of known energy, and the structural development at deeper levels within and outside the crater, which were revealed by detailed survey coupled with quite deep trench excavations across the experimental area.

We will be mainly concerned with the model TNT explosive experiments conducted at Suffield, Alberta, which gave rise to complex craters and various forms of structures which were produced by 100-500 ton TNT explosions. These craters and associated structures produced in explosive experiments constituted a breakthrough, that was recognised by Dence, Dietz, Grieve, Shoemaker, Roddy and others as exhibiting important diagnostic features which could be seen in many tens of natural impact structures on Earth, as well as on the Moon.

Cratering experiments, which were being conducted in the USA as well as in Canada, were designed to establish the probable behaviour of ground movement when subjected to megaton nuclear explosions. Some of the tests in the USA were in excess of 500 tons of TNT (or TNT equivalent). In December 1997, the US government released information regarding their various tests using nuclear weapons. One such explosive test, in Nevada, with an energy of 104 kilotons TNT equivalent (approximately seven times more powerful that the Hiroshima bomb) gave rise to a 'simple' crater with a diameter of 1280 ft (390 m) and a depth of 320 ft (98 m).

Those tests in the USA that had the explosive charge on or close to the surface of the test site, produced only relatively simple basin craters. In Canada, explosions of 500 tons of TNT produced completely different results.

King-Hubbert (1937) published an extremely important paper on the theory of scale models, as applied to the study of geological structures. He showed that if one wishes to imitate large geological structures by using scale models, then the model material must be proportionately weaker than the rock in which the full-scale structure has, or will, develop. Thus, it can readily be inferred from this scale-model theory, that the overall size and form of the crater depends critically upon the magnitude of the explosion and the strength of the target rock.

The tests carried out in the USA between 1950 and 1960, in which the explosive charge was configured to rest on the surface, were conducted mainly in desert areas, where the surface rocks were generally indurated sediments and/or volcanics which were dry, and in some instances approached or even exceeded the strength of the rocks in which the full-scale structures would develop, as the result of a multi-megaton explosion. Also, in these desert test areas, the water-table sometimes tended to be at a depth comparable to the diameter of the crater. Hence, as the majority of the US tests were scale-models probably in the range of 1:1 to 1:10, it is not surprising that the craters that developed in these tests were simple basins.

There is no implied criticism of the tests conducted in the desert areas in the USA for the sites were, of necessity, mainly determined by social factors. Indeed, small-scale experiments were conducted by Davis (1967), of the US Army Engineers, to establish the effects of a near-surface water-table on the dimensions of the crater which developed. However, in these experiments, the charges were limited to 250 lb wt, so that, again, only simple craters were formed.

The explosive tests conducted at Suffieid, however, were carried out on the site of a dried-out postglacial lake, where the sediments were poorly, or non-indurated, and the water-table was at no great depth beneath the surface. The strength of such sediments is of the order of one hundredth to one thousandth that of well-indurated sediments or metamorphic or igneous rock. The resulting structures that developed in the Suffieid explosive tests deviated sharply from what, at that time, was regarded as the normal pattern. The larger of the Suffieid craters did not conform to the normal, simple basin, but instead they appeared remarkably similar to structures which were then only known on the lunar surface and a few terrestrial structures which were, at that time, attributed to 'crypto-volcanic' agencies.

It is, of course, difficult to infer precisely the scaling factor of these Suffieid explosive experiments. Tests in the 20-100 ton TNT range probably gave good modelling of megaton, or multi-megaton, nuclear explosions. However, the 500 ton tests, which produced a craterrim diameter of about 87 m, probably represented real large-scale impact structures with a diameter of 20-100 km, or possibly even larger. Thus, the nature of the test site, which comprised a sequence of flat-lying, lacustrine sediments, with the water-table at a depth of about 8-9 m, was sufficiently uncompacted and weak that, at the scale of the detonation energy, the modelling was more accurate for large-scale natural impacts than for the results one could expect from megaton nuclear tests conducted on strong rock.

Fortunately, members of the scientific community beyond the military, were alerted to the production of 'lunar-type' craters at Suffieid, so that this work had an important influence and enabled the pioneers in this field to identify a number of natural structures which, at that time, had not been firmly identified as impact structures. Indeed, these Suffieid craters were closely studied by the astronauts destined to land on the Moon.

Figure 5.9 Panoramic view of Snowball crater showing central uplift in a moat of groundwater. Diameter of crater is 87 m (after G.H.S.Jones).

The first 500 ton TNT experiment was code-named Snowball. In this experiment, the TNT blocks were constructed to form a hemispherical configuration, so that Half-Snowball would have been a more accurate description. The diameter of the hemisphere of explosives was 34 ft (10.46 m).

The Snowball explosion (see Frontispiece) gave rise to a crater with a diameter of 87 m. The overall view of the water-filled crater (Figure 5.9) is, of course, instructive, but tends to diminish the magnitude of the structure. For this reason we have included Figure 5.10, which shows a 6-foot-tall man (Gareth Jones) on the internal slope of the crater, from which one can infer its depth and also the arcuate form of the crater rim and the main depression. This photograph was taken soon after the explosion, well before the upward migrating water buried most of the crater area.

Over a period of several days, water poured into the crater from a pronounced central, vented uplift, which, eventually, just cleared the surface of the lake which developed in the crater, as can be seen in Figure 5.9.

It was immediately realised that the extra energy of the 500 ton TNT explosion had given rise to a dramatic and unforeseen change in the pattern of crater-floor movement. This central uplift morphology had been seen on the Moon. Also, it was soon realised that the central uplift morphology was matched by several of the impact structures in the Canadian Shield, hitherto considered by some to be crypto-explosive features. This was the first Suffieid cratering experiment that enabled moderate size, natural impact craters to be recognised as such, and greatly stimulated the pioneers in this field of study.

The reader will perhaps be surprised by the degree of 'contrast' in the reproduction shown in Figure 5.9 and in other photographs shown later in this chapter. This can be attributed in part to the fact that the original photographs were taken 50 or so years ago, when the standard 'black and white' film emulsion was orthochrome, which inherently tended to produce high-contrast images. However, there was a second and more important cause for the high-contrast photographs. This was the intensity and extent of the fire-ball set

Figure 5.10 Internal shot of portion of crater taken soon after the explosion when ground water was beginning to seep into the base of the structure (after G.H.S.Jones).

off by the 500 ton TNT explosions. For a distance from ground zero, a zone of black 'scorched-earth' extended out beyound the crater rim for a distance of 3-4 times the diameter of the crater (i.e. about 250320 m). The temperature of this 'fire-ball' was not only responsible for the scorching, the pressure wave and the high temperatures generated by the 500 TNT explosions were sufficient to fuse grains of quartz and produce small hollow spheres which Jones termed 'silica pop-corn'.

In his list of certain impact structures, Hodge (1994) cites several natural examples of such craters which exhibit the central uplift, e.g. Gow Lake, Saskatchewan (4 km diameter), Haughton, Northwest Territories (20.5 km diameter) and Gosses Bluff, Australia (22 km diameter). It is interesting to note that the Gow Lake structure is just outside the diameter limit for simple basins, suggested earlier, of about 3.5 km. Such a limit is better considered as marking a transition zone between simple and more complex craters. This follows from looking at the Suffield cratering experiments with hindsight.

The first recognised central 'bump' was noticed after a 100 ton explosion. However, it was not until the Snowball experiment took place that it was realised that the central 'bump' in the floor of the 100 ton crater was an embryonic central uplift. It is interesting to note that Meteor Crater (Figure 5.5b) also exhibits such an embryonic central uplift.

Military dissatisfaction with the observed structural form exhibited by the Snowball crater had most illuminating consequences; it resulted in the configuration of the charges used to produce subsequent craters to be changed from the hemisphere to that of a sphere which was tangential to the ground (see Frontispiece). The detonation with this 'Prairie Flat' configuration produced the remarkable ringed structure shown in Figure 5.11. Later detonations, such as Dial Pack, extended and confirmed the pattern of ringed structures.

In the second of these 500 ton explosive experiments, the blocks of explosive were arranged to form a sphere which was in contact with the surface at ground zero. The lower half of the sphere was supported by

Figure 5.11 Oblique aerial view of crater with concentric ringed uplifts which developed after the 500 ton TNT Prairie Flat explosion. Crater diameter is 65 m (after G.H.S.Jones).

slabs of polystyrene foam. The sphere of explosives had a diameter of 24 ft (10.46 m). The explosive gave rise to a crater with a diameter of 64 m and a disposition of structures within the crater which was completely different from that of Snowball. As will be seen in Figure 5.11, the crater floor contained multiple, circular uplifts, with associated 'volcanic cones'.

A comparable 50 km diameter circular feature is shown in Figure 5.12, in the Sahara Desert area of Mauritania. This ringed structure has been photographed from a space capsule and also from a Space Shuttle, but we are not aware that the feature has, so far, been studied on the ground. Consequently, it has not yet been classified as a certain impact structure.

The reasons for the change from central uplift to ringed uplift, which resulted from the new configuration, is not completely understood (Melosh, 1989). However, it can be inferred that, because of the spherical charge, Prairie Flat primarily resulted in an air-blast. This developed a smaller diameter crater than did the same tonnage of TNT of hemispherical form for Snowball. As we shall see in later chapters, Snowball can possibly be likened to an impacting comet.

While there is much that is common to the Snowball and Prairie Flat craters, the details vary enormously as regards the location of the secondary pseudo-volcanism. For example, in Snowball, most of the observed pseudo-volcanic cones lay beyond the crater rim, along the circumferential and radial fissures. In Prairie Flat, there was an intense concentration of such cones on what appear to have been fractures associated with an inner circumferential structure on the inner floor of the crater.

A detail of the turn-back of strata, which occurs beneath the rim of the Prairie Flat Crater is shown in Figure 5.13. This structure leads to an inverted limb of the flap outside the crater, as indicated in the section of the 100 ton TNT experiment shown in Figure 5.8.

Such inversion of strata is now well documented for natural craters, e.g. Barringer/Meteor (Shoemaker, 1960), Flynn Creek (Roddy, 1968) and Vredefort (Dietz, 1963). It required only a few years, before this

Figure 5.12 A view from space of concentric rings in a 50 km diameter impact structure in the Sahara of Mauritania, caused by a 'direct' impact or 'low altitude' explosion (c.f. Prairie Flat 'bomb' (Frontispiece) and aerial view (figure 5.11)). The picture was taken by a member of NASA prior to 1979.

Figure 5.12 A view from space of concentric rings in a 50 km diameter impact structure in the Sahara of Mauritania, caused by a 'direct' impact or 'low altitude' explosion (c.f. Prairie Flat 'bomb' (Frontispiece) and aerial view (figure 5.11)). The picture was taken by a member of NASA prior to 1979.

type of feature had become sufficiently well-documented to become a diagnostic feature of moderate to large impacts in sedimentary and even basement rocks.

Very small-scale experiments, not necessarily involving explosives, have also been conducted in laboratories. These usually involved firing a small metal sphere at a weaker target area. However, even a lead pellet fired from a powerful air-pistol, when the target is weak clay, will provide comparable results, as indicated in Figure 5.14. In such experiments, the target material often takes the form of a 'frozen' incomplete recumbent 'flap'. Such flaps in model work represent 'arrested' development because (a) the energy of the impact is not sufficient to spread the flap further and (b) the model is of such a scale that the flap material is sufficiently strong to support itself without collapsing to form a completely recumbent feature. Flaps with comparable geometry may develop in basement rock covered by sediments, as the result of a major impact. In such instances, the flap is rooted in the basement and spreads to cover the sediments around the crater, which, of course, supports the weight of the flap.

Dietz (1963) realised that such model profiles (Figure 5.14) were reproduced on a far grander scale in the Vredefort structure in S Africa. Moreover, he recognised that minor conical shear features, now termed shatter-cones (which we shall discuss later in this chapter) were almost certainly a manifestation of an impact event. This enlightened insight was altogether too much for the vast majority of geologists. Indeed, it is only in the last decade of this 20th century that the majority of S African and other geologists have been persuaded by his thesis.

Was this article helpful?

0 0

Post a comment