The terrestrial record of Late Heavy Bombardment

Abstract

Until recently, the known impact record of the early Solar System lay exclusively on the surfaces of the Moon, Mars, and other bodies where it has not been erased by later weathering, erosion, impact gardening, and/or tectonism. Study of the cratered surfaces of these bodies led to the concept of the Late Heavy Bombardment (LHB), an interval from about 4.1 to 3.8 billion years ago (Ga) during which the surfaces of the planets and moons in the inner Solar System were subject to unusually high rates of bombardment followed by a decline to present low impact rates by about 3.5 Ga. Over the past 30 years, however, it has become apparent that there is a terrestrial record of large impacts from at least 3.47 to 3.22 Ga and from 2.63 to 2.49 Ga. The present paper explores the earlier of these impact records, providing details about the nature of the 8 known ejecta layers that constitute the evidence for large terrestrial impacts during the earlier of these intervals, the inferred size of the impactors, and the potential effects of these impacts on crustal development and life. The existence of this record implies that LHB did not end abruptly at 3.8–3.7 Ga but rather that high impact rates, either continuous or as impact clusters, persisted until at least the close of the Archean at 2.5 Ga. It implies that the shift from external, impact-related controls on the long-term development of the surface system on the Earth to more internal, geodynamic controls may have occurred much later in geologic history than has been supposed previously.

Introduction

One of the many outcomes of the NASA Apollo program and the 6 manned lunar landings between 1969 and 1972 was the direct sampling and return of lunar rocks and the establishment of a more exact, quantitative lunar chronology. The results include recognition that the heavily cratered lunar highlands formed about 4.0 billion years before the present (= 4.0 Ga) or slightly earlier and the more lightly cratered lunar maria representing vast, low-relief basaltic plains with much fewer, mainly smaller craters, formed 3.9–3.1 Ga but mostly between 3.7–3.9 Ga. Abundant radiometric evidence for widespread impact melt formation 4.1–3.8 Ga and the subsequent abrupt termination of large-scale cratering at about 3.8 Ga led to the concept of a Late Heavy Bombardment (LHB) or lunar cataclysm (Tera et al., 1974) during which the Earth, Moon, and presumably other planetary surfaces were bombarded by large bolides at rates substantially higher than those that prevailed before about 4.1 Ga (Fig. 1(A)). Based on this lunar chronology and studies of lunar cratering, the LHB was thought to have extended from about 4.1 or 4.0 Ga until 3.8 Ga (Fig. 1(A)). After 3.8 Ga, the impact rate was interpreted to have dropped abruptly until, somewhere between 3.6 and 3.0 Ga, depending on the cratering rate curve one prefers, it reached a value that differed little from the very low impact rate of today (Neukum et al., 2001, Valley et al., 2002, Koeberl, 2006). A number of investigators have questioned the concept of a Late Heavy Bombardment, suggesting instead a more continuous decline in the impact flux (Fig. 1(B)) (e.g. Neukum et al., 2001, Boehnke and Harrison, 2016).

The paucity of terrestrial rocks older than about 3.8–3.9 Ga is consistent with this concept of profound crustal modification and disruption by large impacts during LHB, and the apparent absence of terrestrial craters and impact features in preserved early Archean terrestrial rocks, which range from about 3.55 to 3.0 Ga, seemed for many years to be consistent with the inferred low impact rates at this time. However, it should be noted that Green (1972) and others compared the thick mafic volcanic sequences in early Archean terrestrial greenstone belts to the basaltic lunar maria and hypothesized that they may reflect large terrestrial impacts that accompanied the large maria-forming lunar impacts. These and other possible effects of large impacts on early terrestrial tectonics will be considered in Section 4.3.

Evidence for ancient terrestrial impacts comes in two main forms: craters and ejecta layers. Craters, both terrestrial and lunar, have been regarded by some as possible meteor impact structures since they were first studied but the wide acceptance of the role of impacts in generating ancient terrestrial structures is a more recent event. Many more-or-less circular zones of surface deformation and rock brecciation were commonly termed cryptovolcanic or cryptoexplosion structures before about 1965 because there was often little or no evidence for an associated igneous event and because of the general reluctance to accept a major role for impacts in later Earth history. Work by Robert Dietz on the Vredefort structure, South Africa (Dietz, 1961), and the Sudbury structure, Ontario, Canada (Dietz, 1964) pioneered the modern advocacy of a major role for impacts in Earth history.

An ejecta layer represents materials thrown out from an impact site at the time of and as a result of the impact (Fig. 2). Ejecta takes three forms:

• Discrete particles, either liquid or solid, that are thrown from the impact site, follow more-or-less ballistic trajectories, and fall either as coarse ejecta close to the impact site or at more distant sites as finer, distal ejecta (Fig. 2). Simonson and Glass (2004) and Glass and Simonson (2012) suggest that ejecta deposited >2.5 crater diameters from the crater should be called distal ejecta and closer deposits should be termed proximal ejecta. Proximal ejecta includes blocks and boulders of target rock, mineral grains, melted and partially melted materials, accretionary grains called accretionary lapilli, and some spherical melt droplets.

• More distal ejecta deposited >10 crater diameters from the crater consists primarily of glassy particles. These include rare larger (> 1 cm) particles called tektites representing melted masses of the target rock and abundant smaller grains, most <1 mm in diameter, representing ballistic liquid silicate droplets. Many ejecta layers, especially >10 crater diameters from the impact site, include spherules representing either melted or vaporized target and bolide material. Large impacts generate large amounts of rock vapor, up to four or five times the mass of the impactor, that is ejected from the crater, often into or above the upper atmosphere, spread around the Earth as a rock vapor cloud, and condensed to form liquid silicate droplets that chill and fall to the earth as solid, spherical to sub-spherical bodies termed spherules. Spherules are commonly <0.1 to about 1 or 2 mm in diameter. They include glassy particles termed microtektites and partially or fully crystalline grains termed microkrystites (Glass and Burns, 1988). Between 2.5 and 10 crater diameters from the impact site, ejecta may include both spherules and small unmelted or partially melted ballistic particles.

• Impacts also generate a great deal of fine dust that is ejected into the atmosphere and subsequently carried by air currents as non-ballistic particles to locations potentially around the globe. These are analogous to volcanic plumes that can cloud the atmosphere with dust for months. It was dust clouds that are thought to have contributed to global cooling and the severe climatic changes that drove collapse of the Late Cretaceous ecosystems and extinction of the dinosaurs about 65 million years ago. General discussions of ejecta layers are provided by Smit, 1999, Simonson and Glass, 2004, Glass and Simonson, 2012, and Glass and Simonson (2013).

In the present report, we provide a review of the distribution, ages, and characteristics of the Archean ejecta layers and the Barberton Greenstone Belt layers in particular, discuss the possible influence of large impacts on early Archean crustal development, and consider the implications of this preserved terrestrial impact history on the nature of Late Heavy Bombardment and the late Solar System impact history. All of the Archean ejecta layers reported to date are distal layers composed largely of spherules and fine ash and dust. Coarser ejecta layers have not yet been recognized in Archean sequences.