Saturday, 27 August 2022

The Nadir Crater: A possible second End Cretaceous Impact Structure off the coast of West Africa.

High velocity impacts by large asteroids and comets are a known, but poorly understood, threat to life on Earth. Impactors with diameters of about 50 m (the estimated size of the object which caused the Tunguska explosion) are thought to hit the Earth roughly once every 900 years, and objects with diameters of 1 km or greater, likely to cause devastation on a regional or global scale, to impact the Earth roughly every million years. The largest impacts cause devastation on a global scale, with the best known of these being the Chicxulub impact 66 million years ago, an event thought to have caused the End Cretaceous Extinction Event. Despite the importance of such events, very little evidence of their having occurred is known, with only about 200 large impact craters known on Earth, of which only 15-20 are interpreted as having happened in a marine environment, despite 71% of the Earth's surface being covered by water. This lack of impact craters, and in particular well preserved impact craters, severely hampers our ability to understand the effects of the events which cause them.

In addition to this general lack of impact craters in the rock record, there is a proportional lack of double impact craters. About 15% of the asteroids in near-Earth orbits are thought to be binaries (i.e. pairs of gravitationally locked asteroids, whereas only 2-4% of known impact craters on Earth have ever been suggested as potentially being parts of pairs. Furthermore, most of these suggestions have either been disproven or are disputed. On Venus, where meteor impacts are generally recorded as 'dark splotches' rather than true craters, due to the much more destructive nature of the Venusian atmosphere, about 14% of impacts appear to be binary. Even allowing for a considerable degree of inaccuracy in these numbers, binary impacts seem to be significantly under-represented on Earth.

Another phenomenon seen elsewhere in the Solar System but largely absent on Earth is impact clustering, i.e. the formation of a series of impact craters over a relatively short period of time. This is typically caused by the breakup of a large object due to a gravitational encounter with a planet, followed by the debris from this breakup remaining on an orbit which intersects that of the planet, resulting in a series of impacts over a drawn out period of time. The only observed instance of such an event was the breakup of Comet Shoemaker-Levy 9 in 1992, which was followed by a series of impacts in 1994 as the debris from this breakup collided with Jupiter, but it is thought that the breakup of larger bodies can result in strings of impacts which could continue for several million years. Only a single such event has been recorded on Earth; a series of linked craters in the Ordovician, which is combined in a shift in the ratio of argon isotopes in the Earth's atmosphere at that time, and evidence for an increase in the amount of extraterrestrial dust raining down on the Earth at that time. Other proposals have been made for clusters of impacts in the Cretaceous and Eocene, but the evidence in support of such events is sparse. This is despite the Moon showing clear evidence of a number of periods of impact clustering, events which must surely also have affected the Earth.

In a paper published in the journal Science Advances on 17 August 2022, Uisdean Nicholson of the School of Energy, Geoscience, Infrastructure and Society at Heriot-Watt UniversityVeronica Bray of the Lunar and Planetary Laboratory at the University of ArizonaSean Gulick of the Institute for GeophysicsDepartment of Geological Sciences and Center for Planetary Systems Habitability at the University of Texas at Austin, and Benedict Aduomahor, also of the School of Energy, Geoscience, Infrastructure and Society at Heriot-Watt University, announce the possible discovery of a second End-Cretaceous impact structure on the Guinea Plateau of the coast of West Africa, based upon evidence from two-dimensional seismic surveys.

The Guinea Plateau is an area of submerged continental crust extending southwest from the coasts of Guinea and Guinea Bissau for about 400 km, where it meets with the Guinea Fracture Zone, beyond which lies the deep ocean. The plateau can be divided roughly into two parts, an inner zone which is about 400 m deep across most of its extent, and an outer zone, known as the Guinea Terrace, which deepens gradually from about 400 m to about 1200 m. The Guinea Fracture Zone is home to a series of seamounts (inactive submarine volcanoes), including the Nadir Seamount.

Map and regional seismic sections showing location of Nadir Crater. (A) Regional bathymetry map of the Guinea Plateau and Guinea Terrace showing location of 2D seismic reflection and well data used in the study. JS, Jane Seamount; NS, Nadir Seamount; PS, Porter Seamount. The white dashed line shows the NE extent of high-amplitude discontinuous seismic facies at the top Maastrichtian interpreted as ejecta deposits and associated tsunami deposits. The north-east limit of this facies closely corresponds with the Maastrichtian shelf-slope break at the landward margin of the Guinea Terrace. Inset map shows a palaeogeographic reconstruction of the Atlantic near the end of the Cretaceous, about 66 Ma ago. Ch, Chicxulub Crater; Nd, Nadir Crater; Bo, Boltysh Crater. (B). Regional composite 2D seismic reflection profile extending from the GU-2B-1 well in the east to the deep Atlantic basin in the west, showing the structural and stratigraphic character of the Guinea Plateau and Guinea Terrace. (C) North-South seismic profile from the salt basin in the north to the Nadir Seamount, south of the Guinea Fracture Zone. Nicholson et al. (2022).

The Guinea Plateau formed by extensional rifting in the Triassic-Jurassic, as the opening of the Atlantic Ocean pulled the North American landmass away from the west coast of Africa. The plateau remained part of the passive (i.e. seismically inactive) margin of the Central Atlantic until about 100 million years ago, when the its southern margin underwent further rifting, as South America separated from Africa.

The crater which Nicholson et al. propose is located on the southwestern part of the Guinea Terrace, roughly 60 km to the north of the Guinea Fracture Zone and 100 km northwest of the Nadir Seamount. The crater is covered by about 400 m of sediment, and a water column about 900 m deep.

The stratigraphy of the area has been determined from boreholes and seismic surveys. Cretaceous deposits here can be divided into a Lower Cretaceous sequence, roughly 145-100 million years old, and an Upper Cretaceous sequence, 100-66 million years old, separated by an unconformity known as the Top Albian. The deposits making up the Upper Cretaceous sequence are largely undeformed, and extend laterally across the Guinea Terrace. These deposits can be further subdivided by a series of (seismically) reflective surfaces, numbered KU1-KU4, which are interpreted as representing major flooding events. The Upper Cretaceous sequence can be divided into a Lower Unit, which is about 300 m thick and contains a series of high amplitude reflective surfaces, and an Upper Unit, which is about 500 m thick and contains moderate-to-low amplitude reflective surfaces.

Seismic characteristics of the Nadir Crater. (A) Seabed depth map of crater showing seismic line locations and the mapped extent of the crater rim and damage zone.  (B)  West-east  seismic  section  (pre-stack  depth  migration  –  depth  domain)  across  the  crater,  highlighting  the  crater  morphology  and  damage  zone,  and  the  extent  of  subsurface deformation. (C) Detailed seismic stratigraphic and structural elements of the crater. KP, Cretaceous-Paleogene sequence (KP1 equivalent to Top Maastrichtian); KU, Upper Cretaceous seismic horizons. KU1 and KP1 'regionals' are schematic reconstructions of these seismic horizons before formation of the crater at the end of the Cretaceous and have been used to reconstruct a conceptual model of crater formation. (D) Southwest-northeast seismic section (pre-stack time migration – time domain) across the crater, showing crater morphology and seismic facies outside the crater, including high-amplitude seismic facies sitting above a roughly 100-ms-thick unit of chaotic reflections, interpreted to have formed as a result of seismic shaking following the impact event. Nicholson et al. (2022).

The end of the Cretaceous sequence here, identified as the Top Maastrichtian, is clearly identifiable, and well defined, but the age of other reflective surfaces in the sequence is less certain. The KU-1 reflective surface may mark the Top Turonian, making it about 90 million years old, and the high amplitude surfaces between this and the underlying Top Albian surface may represent black shale deposits, with elevated levels of organic carbon. The lower amplitude reflective surfaces above KU-1 probably represent marine shales and marls.

The Top Maastrichtian Surface lies at the top of another set of distinctive, high amplitude reflective surfaces, and is marked by an erosional unconformity on the southwest part on the Guinea Terrace, which is not present to the north and east. The reflective surfaces at this level extend laterally across the Guinea Plateau, although they are eroded away on both the continental slope and the inshore shallows. On the Guinea Terrace these surfaces overlie a chaotic seismic surface about 100 m deep.

Overlying the Top Maastrichtian, the Cainozoic deposits begin with a thick sequence of carbonates and clastic sediments, although these are considerably thinner on the Guinea Terrace than elsewhere on main Guinea Plateau. Precise dating of these sediments is again difficult, though sediments younger than about 23 million years old are apparently absent on the Guinea Terrace; strat here comprise a Palaeocene-Eocene sequence about 300 m thick, overlain by a series of Oligocene mass transport deposits (debris from submarine landslides).

The feature which Nicholson et al. identify as the Nadir Crater is a depression at least 8.5 km wide in the Top Maastrichtian. This can be seen where two separate seismic profiles cross one-another, suggesting a rounded or elliptical shape. Since there is no particular reason to believe that the intersection of the two seismic profiles lies at the centre of the depression, 8.5 km wide must be taken as a minimum measurement. The crater bed is 200 m below the 'seafloor' represented by the Top Maastrichtian, while the crater rim is raised 20-40 m above this level. There appears to be a slight uplift in the centre of the crater at the Top Maastrichtian level; this is more pronounced at the Top Albian level, where it rises 350 m above the surrounding terrain. On either side of this central uplift the strata below the crater floor and rime are intensely deformed, with a column of extensively folded and fractured material extending downwards 700 m from the Maastrichtian surface. A wider area of faulting extends 10-12 km from the crater in all directions, with all faults dipping towards the crater.

All of these phenomena are consistent with large impact craters elsewhere on Earth and other Solar System bodies. The rim is raised  above the surrounding terrain approximately one tenth as much as the crater floor is depressed beneath it. The crater is approximately one twentieth as deep as it is wide. The damage zone around the crater is approximately twice as wide as the crater itself. The crater has a terraces structure and a raised central peak.

The deformation beneath the crater is also consistent with the predictions for large craters of this sort. The central uplift seen in such craters is caused by the elastic rebound of shock waves passing through the rock, causing material beneath the crater to flow upwards, initially flowing freely the brecciating as the rock regains its brittleness. The deformations seen in the zone 10-12 km from the crater are consistent with these areas being subjected to a significant shockwave, then flowing back towards the crater to fill a partial void.

Based upon examination of other impact craters on Earth, the zone of uplift beneath a crater 8.5 km wide would be expected to extend downwards to between 630 and 780 m beneath the impact surface. That of the Nadir Crater extends down about 800 m, with some deformation extending downwards into underlying Jurassic Sediments. However, this does not necessarily imply a much larger crater; rather it may be a consequence of a bolide impacting soft, unconsoidated marine sediments. The Mjølnir Crater, in the Barents Sea off the coast of Norway, shows a similar pattern of deformation beneath the crater, although here there is less deformation outside the crater rim, as is the case at Chicxulub. This deeper pattern of damage at sites where large, high velocity objects impact shallow marine sediments may be caused by the sudden decrease in the porosity of the strata impacted, resulting greater brecciation, and the formation of hydrothermal systems which could potentially last for millions of years.

The Top Maastrichtian layer within the crater is overlain by a series of high amplitude reflective layers identified as KP-1 to KP-3, with KP-1 marking the boundary between the Maastrichtian and the Palaeocene. Between the Top Maastrichtian and KP-1 is a layer about 100 m thick and relatively transparent to seismic waves, which Nicholson et al. interpret as a possible suevite layer (i.e melt materia), similar to that seen as the Chicxulub Crater. Above this, between KP-1 and KP-2 is a unit made up of a series of low-amplitude reflective surfaces onlapping onto the inner crater walls, which Nicholson et al. interpret as reworked ejecta material. 

The crater is surrounded by a blanket of material made up of a series of high amplitude reflective surfaces, which Nicholson et al. also interpret as reworked ejecta, combined with material deposited by tsunamis triggered by the impact. Beneath this is a layer of chaotic material, which may represent older strata re-organised by the shock wave from the impact.

The Nadir Crater a deep inner crater with in inner peak and a flat outer zone, surrounded by a raised rim, features also seen in confirmed marine craters such as the Chesapeake Crater (off the east coast of North America), Lockne Crater (in northern Sweden), and Flynn Creek Crater (in Tennessee), strongly supporting the idea that this is an impact crater, although this hypothesis will only be confirmed by drilling into the structure.

There are other processes which can form structures resembling impact craters, such as the dissolution of salt deposits, escape of gas or fluids, volcanic activity, deformation due to tectonic stresses, or any combination of these. However, none of these possible alternatives appears consistent with the geology of the Guinea Terrace,

Salt diapers (domes formed when salt intrudes into overlying rocks) can collapse as the salt within them dissolves into the water column, leaving a circular depression surrounded by radial faults. However, such structures do not have uplifted central zones, nor would they be surrounded by a ring of high amplitude material of the type Nicholson et al. interpret as ejecta at the Nadir Crater. Furthermore, while there are salt diapers on the Guinea Terrace, these are in a zone far to the northwest of the Nadir Crater, with the closest diaper being over 250 km from the crater. 

Craters left by fluids escaping from mud layers as they are buried and then subjected to pressure, are common features on the seafloor, being particularly common on the continental margins, where large amounts of methane are generated. Most of these are a few metres to a few hundred metres across, but the largest can exceed 10 km, making ruling our such an origin for the Nadir Crater impossible on the basis of size alone. However, such structures are almost always found in clusters, and no structures similar to the Nadir Crater can be seen anywhere else on the Guinea Plateau, nor are there any other signs of gas or fluid escape in this region. Furthermore, craters formed by fluid escape do not typically produce crater rings, central uplift, or ejecta, all of which can be seen at the Nadir Crater. 

A variety of depressions, typically bounded by fault zones, can be formed when slip-strike deformation leads to areas of seafloor being pulled apart. However, such depressions are rarely circular, typically being more than twice as long as they are wide. Nor do such structures have rims or centrally uplifted zones. More importantly, the Nadir Crater is located on an area of the African continental margin where slip strike activity is thought to have ceased about 44 million years before it formed, making such an origin highly unlikely.

The most plausible alternative hypothesis for the origin of the Nadir Crater is volcanism. Phreatic explosions, which occur when hot magma comes into contact with water, can form maars (broad, low relief volcanic craters) similar in form to the Nadir Crater. Furthermore, there clearly was volcanic activity nearly contemporary to the formation of the crater in the area; the Nadir Seamont is only 100 km away from the Nadir Crater, and 7.4 million years younger, i.e. close enough to be part of the same volcanic complex, along with the nearby Grimaldi Seamounts. However, maar structures seldom exceed 2-3 km in diameter, and their three dimensional structure has a distinctive funnel-shape unlike anything revealed by the seismic surveys at the Nadir Crater. Furthermore, maars do not have a central uplift structure, making it highly unlikely that the Nadir Crater is a volcanic maar.

Next Nicholson et al. used a hydrocode model to simulate the impact of a 400 m wide object (the estimated size of the Nadir impactor) into soft sediments covered by a range of water depths. All of the models produced a crater between seven and nine kilometres in diameter, with elevated rims and significant central uplift. The amount of subsequent rim collapse varied with water depth, with impacts into deeper water showing more collapse. This lead to more infill into the crater, resulting in a flatter profile and a less prominent central uplift zone. At depths of greater than 1 km, the rim completely collapsed into the crater, creating a very flat profile. In all cases, regardless of depth, the impact resulted in a crater surrounded by a network of faults.

Numerical model results of iSALE hydrocode simulations of final crater architecture for different water depths (200, 500, 800, 1100, and 1500 m, as indicated on the figure) Models assume a 400 m diameter impactor with an impact angle of 90°. The water layer has been removed from each image to better highlight the final crater morphology. Model outcomes show total plastic strain on the left and deformed lithologies on the right - the gray unit represents the assumed Cenomanian-Turonian black shale deposits as a marker horizon. Nicholson et al. (2022).

All models show the development of a central uplift, although the extent of this varies with depth, with larger central uplift zones in impacts at greater depth, possibly representing the rocks rebounding more when the weight of the overlying water is removed by the flash-vaporization cause by the impact. These simulations showed material from as deep as 2 km below the surface being moves upwards, with material from up to 1 km deep reaching the subsurface within the central peak; based upon the available data Nicholson et al. estimate that material from as deep as 750 m beneath the Nadir Crater is present within its central peak.

Snapshots of numerical model results of an iSALE hydrocode simulation for 800 m target water depth. Considered this to be the most likely water depth for the impact, based on crater morphology. Model snapshots show transient crater formation and generation of a rim-wave tsunami after 3 seconds; rebound of central uplift and propagation of a rim-wave tsunami away from crater at 38 seconds; crater collapse, resurge, and central jet formation at 85 seconds; and further resurge at 245 seconds. Nicholson et al. (2022).

All of the simulated impacts produced zones of brecciation beneath the crater, with the depth to which this extended ranging from about 1 km up to about 2.5 km. Areas of damage were also present around all the simulated craters, with faults and fractures extending at least 5 km from the craters and reaching depths of at least 1 km.

The models are consistent with the observed features at the Nadir Crater, insofar as all of the craters forming at depths shallower than 1 km produced raised rims, uplifted central areas, terraces made up of debris, and areas of damage outside the crater. The models which produced craters most similar to the Nadir Crater had water depths of about 800 m.

Based upon the data gained by running these simulations, Nicholson et al. developed a conceptual model for the chain of events around the impact. It is assumed that the initial stratigraphy of the area comprised a series of flat, horizontal bedding plains, overlain by about 800 m of seawater. The initial impact would have formed a transient crater about 1 km deep, followed rapidly by the formation of the central uplifted zone, and the collapse of the initial crater rims. This would produce the current observed structure; a crater with a marked central uplift and observable rims surrounded by terraces of collapsed material, and a wider area of fallen ejecta, a hypothesis which they propose could by tested by drilling in the area, in order to confirm the actual nature of the deposits observed in the seismic profiles.

Conceptual model of the impact sequence at the Nadir impact site, based on seismic observations and analog models.  (A) At time zero, the impactor hits the water surface at a velocity of about 20 km per second, initiating a rim-wave tsunami in its wake. (B) Several seconds later, the transient crater forms, as the impactor and a substantial body of water are vaporised. Impactites (melt rock and breccias) line the transient cavity. Tsunami waves propagate away from the evacuated crater. Shock waves cause substantial damage below and around the impact site, and seismic waves propagate across the plateau; (C) major uplift (roughly 400 m relative to pre-impact regional) occurs in the 'rebound' crater modification phase, resulting in the formation of a raised crater peak; (D) radial collapse of the subsurface damage zone results in further modification of the crater, including the formation of terraces at the surface. Resurge of water transports substantial volume of ejecta and other sediment into the crater, deposited above the impactites. Nicholson et al. (2022).

Taking the data from the simulations, Nicholson et al. used the Earth Impacts Effect Program to assess the environmental damage caused by the event which formed the Nadir Crater. The initial impact would vaporise vast amounts of seawater and seafloor sediment and release an amount of energy roughly equivalent to 5000 megatons of TNT. This in turn would lead to the formation of a fireball over 5 km in diameter, an an air blast which would be travelling at about 470 km per hour at a distance of 50 km from the impact site. 

The event would release seismic waves equivalent to a Magnitude 7.0 Earthquake, fluidising seafloor sediments for hundreds of kilometres in all directions. Following fluidisation of the sediments they would likely be reworked extensively by the tsunami waves caused by the impact. The seismic waves would also be likely to trigger massive landslides along the Guinea Terrace.

The initial impact would cause an 'ejecta curtain', made up of water and sediment, over 2 km high, which would then collapse back onto the sea surface, triggering the formation of a tsunami wave over 500 m high, which would spread at about 400 m per second. As well as spreading out from the rim of the crater, this tsunami would spread inwards, forming a central water jet rising to over 2 km. This would then collapse back into the crater, forming a new tsunami, a process which might repeat itself several times. These repeated tsunamis would cause extensive reworking of sediments along the Guinea Terrace, at depths of up to 800 m. They would also cause extensive scouring of the West African coastline, and to a lesser extent that of South America, which, despite being about 1000 km away, would still be hit by waves about 5 m high. 

The impact would lead to large volumes of water entering the atmosphere, as well as black carbon dust derived from the black shale deposits laid down on the Guinea Plateau during the Cenomanian  and  Turonian.  The high temperatures and pressures unleashed by the impact would turn much of the organic material present in the area directly into methane, with global climatic implications, albeit short lived ones. The seismic shock wave could potentially release a lot more methane from gas hydrate deposits on the Guinea Plateau, and adjacent areas of the deep sea floor.

The impact which caused the Nadir Crater appears to have happened at, or very close to, the Cretaceous-Palaeocene boundary. This places the Nadir Impact very close to the Chicxulub Impact, chronologically speaking, raising the question as to whether the two events were connected. Potentially the Nadir and Chicxulub impactors might have been two parts of a binary object, two fragments of an object which broke up as it came within the Earth's gravitational field, two objects from a longer impact cluster, or two unrelated objects which happened to fall at about the same time. 

Impacts by binary asteroids have previously been suggested as the cause of other crater pairs, including the Lockne and Målingen craters in  Sweden, the East and West Clearwater craters in Quebec, and the Suvasvesi craters in Finland, although more refined dating has shown that neither the Clearwater and Suvasvesi craters are in fact binaries, and no there is no evidence in favour of the Lockne and Målingen craters being related, and therefore no evidence of a binary impact anywhere on Earth. 

Reconstructions show that the Nadir Crater site would have been about 5500 km from the Chicxulub Crater site at the end of the Cretaceous; significantly less than the 8000 km that separates them today, but still to far apart to have been caused by the impacts of two parts of a single object which broke up within the Earth's Roche limit (the minimum distance to which a large satellite can approach a larger body without tidal forces overcoming the internal gravity holding the satellite together), or within the Earth's atmosphere. 

This does not, however, rule out two fragments of a single object which was broken into fragments by an earlier encounter with the Earth's tidal field, then impacted the Earth on a subsequent encounter, in the way that Comet Shoemaker-Levy 9 was torn apart by Jupiter's gravitational field, then impacted Jupiter in several fragments on a subsequent pass several years later. There is some evidence of other large impacts around the End of the Cretaceous, most notably at Boltysh in Ukraine (currently dated to 65.4 million years ago, or 650 000 years younger than the Chixulub impact), as well as preserved fossil meteorites from Poland and the North Pacific. This could imply that the Earth was not just hit by a single large impact, but a protracted asteroid shower, lasting from months to one or two million years, with each impact triggering earthquakes and tsunamis, and contributing to rising atmospheric levels of greenhouse gasses and aerosols.

Estimates as to how often objects the size of the Nadir impactor hit the Earth vary from about once every 100 000 years to about once every 700 000 years. Thus even if it can be demonstrated that the Nadir Impactor fell within a million years of the Chicxulub impactor, it would be impossible to demonstrate that the two events were part of a single asteroid shower (Near Earth Asteroid (101995) 1999 RQ36 is roughly equivalent in size to the Nadir impactor, and is estimated to have a 1 in 1750 chance of hitting the Earth within the next few hundred years). However, if it can be demonstrated that the Chicxulub, Nadir, and Boltysh events all happened within a relatively short period of time, then it would present a much stronger case for the events being related, with important implications for our understanding of events at the End of the Cretaceous.

See also...

Follow Sciency Thoughts on Facebook.

Follow Sciency Thoughts on Twitter.