The Cryogenian Ghaub Formation of Namibia – New insights into Neoproterozoic glaciations

Abstract The Neoproterozoic Cryogenian ('Marinoan') Ghaub Formation of northwestern Namibia represents an important founding pillar of the Snowball Earth hypothesis and its derivative, the Panglacial Earth hypothesis. These hypotheses assume oceans and continents covered by thick ice, even in the tropics, which caused a very distinct drop in eustatic sea-level. Over time, strongly increased CO2 contents of the atmosphere led to sudden ice melting, very substantial sea-level rise, and strong weathering on the continents associated with the deposition of cap carbonates in the newly ice-free oceans. The ongoing controversy about Snowball-type glaciations in Namibia and elsewhere is reviewed, and other hypotheses (Slushball Earth, Waterbelt Earth, Jormungand state of the Earth, Thin Ice state of the Earth, Zipper-Rift Earth, High-Obliquity Earth) are discussed. We prefer the term 'Waterbelt Earth' instead of the originally proposed 'Waterbelt state' because of the clearer contrast with 'Snowball Earth'. Because a great deal of information related to Cryogenian glaciations comes from the Ghaub Formation of northwestern Namibia, these hypotheses should be tested independently based on a time-equivalent depositional system. This analogue was found in the carbonate-dominated successions of the Otavi Mountainland (OML), northeastern Namibia, and is highly comparable with the successions in the well-investigated northwest of the country. An extreme eustatic sea-level drop caused by a global glaciation of oceans and continents and imposed on a carbonate platform or ramp such as the one in the OML would have led either to glacial cover or widespread subaerial exposure and extensive erosion, including deeply incised valleys. The presence of such features would strongly support the Snowball Earth hypotheses if tectonic effects did not play a major role. During the postglacial transgression, distinct reworking of the carbonate platform/ramp surface would have occurred, leaving behind lag deposits, as well as infills of incised valleys with fluvial, reworked glacial, and marine deposits. The main objective of our research was to weigh and investigate the strengths and weaknesses of the proposed Snowball Earth model of glacially induced large-amplitude sea-level changes during Ghaub time, and to compare different models to obtain a rough estimate of the amount of glaciation. The study area in the OML includes two different, age-equivalent facies realms: platform sedimentation in the Southern area without diamictites, and slope deposits, including Ghaub diamictites, in the Northern area. The southern, continuously shallow-marine area shows a shallowing-upward succession from the pre-glacial lower Auros Formation, often varve-like laminated shales formed below wave base, to metre-high columnar stromatolites and microbial mat-related carbonates with intervals of vertical tubes (degassing features) of the upper Auros Formation, overlain by cap carbonates of the Maieberg Formation. The columnar stromatolites and the microbial tubestone lithotypes were clearly deposited in the euphotic zone. Indications for tidal conditions or subaerial exposure were not recorded in this platform succession without unconformities. Neither dropstones, nor incised channels, nor transgressive lag deposits were observed. The facies changes from below storm wave base to the photic zone and finally a shallow subtidal zone is explained by a prolonged, modest sea-level fall, partly counterbalanced by subsidence, followed by a slow transgression. In contrast, coarse-grained sedimentary rocks (e.g., oolites, debrites) characterise the time-equivalent successions in the Northern area. Starting with laminated shales at the base, similar to the Southern area, the overlying redeposited oolites and breccias of the Auros Formation show distinct lateral and vertical inhomogeneities and thickness changes, which indicate long-lasting synsedimentary tectonism. The same phenomenon is observed in the overlying diamictites of the Ghaub Formation. Their variable clast content indicates erosion of a strongly uplifted local source area formerly covered by a thick carbonate succession, which was downstripped to the crystalline basement. The prograding diamictite succession with repeatedly intercalated silt-stringers is interpreted as periglacial debris flows into a marine environment. Sparse striated clasts in the diamictites and very rare dropstones (much less common than in northwestern Namibia) are indicators of glaciations somewhere in the area. However, compared with other glacial sequences, e.g. Quaternary periglacial sediments at the forefront of continental ice, dropstones and striated clasts would be expected to be much more common and more uniformly distributed if the entire area was covered by melting continental ice, as proposed in the Snowball/Panglacial Earth scenario. In the Southern area, dropstones would be expected to occur on the flooded platforms/ramps as well, even when diamictites are absent. Both the relatively moderate sea-level change and the less common, irregular distribution of locally concentrated glacial rainouts provide strong evidence against the presence of a thick, laterally continuous ice cover over oceans and continents extending to equatorial areas. The oceans possibly corresponded to the scenario of a Waterbelt Earth or High-Obliquity Earth; evidence of open oceanic water exists, which would have enabled the continued evolution of biota. Glacial ice was present on tropical continents, but its occurrences may have been regional in patches, sourced from mountainous areas, and ice streams would have reached the oceans only locally, unrelated to a thick continental ice cover