It has been hypothesized that predecessors of today's bryophytes significantly increased global chemical weathering in the Late Ordovician, thus reducing atmospheric CO2 concentration and contributing to climate cooling and an interval of glaciations. Studies that try to quantify the enhancement of weathering by non-vascular vegetation, however, are usually limited to small areas and low numbers of species, which hampers extrapolating to the global scale and to past climatic conditions. Here we present a spatially explicit modelling approach to simulate global weathering by non-vascular vegetation in the Late Ordovician. We estimate a potential global weathering flux of 2.8 (km3 rock) yr−1, defined here as volume of primary minerals affected by chemical transformation. This is around three times larger than today's global chemical weathering flux. Moreover, we find that simulated weathering is highly sensitive to atmospheric CO2 concentration. This implies a strong negative feedback between weathering by non-vascular vegetation and Ordovician climate.

The Ordovician glaciation represents the acme of one of only three major icehouse periods in Earth's Phanerozoic history and is notorious for setting the scene for one of the “big five” mass extinction events. Nevertheless, the mechanisms that drove ice sheet growth remain poorly understood and the final extent of the ice sheet crudely constrained. Here using an Earth system model with an innovative coupling method between ocean, atmosphere, and land ice accounting for climate and ice sheet feedback processes, we report simulations portraying for the first time the detailed evolution of the Ordovician ice sheet. We show that the emergence of the ice sheet happened in two discrete phases. In a counterintuitive sequence of events, the continental ice sheet appeared suddenly in a warm climate. Only during the second act, and set against a background of decreasing atmospheric CO2, followed steeply dropping temperatures and extending sea ice. The comparison with abundant sedimentological, geochemical, and micropaleontological data suggests that glacial onset may have occurred as early as the Middle Ordovician Darriwilian, in agreement with recent studies reporting third‐order glacioeustatic cycles during the same period. The second step in ice sheet growth, typified by a sudden drop in tropical sea surface temperatures by ∼8°C and the further extension of a single, continental‐scale ice sheet over Gondwana, marked the onset of the Hirnantian glacial maximum. By suggesting the presence of an ice sheet over Gondwana throughout most of the Middle and Late Ordovician, our models embrace the emerging paradigm of an “early Paleozoic Ice Age.”

No abstract

Table of Contents Methods S1: Methodological details supporting main text analyses Tables S1: Values used to calculate dispersal distances in the simulations S2: Number of virtual species used for each global climate model (AOGCM) combination S3: Details of AOGCMs used in the simulations S4: Number of virtual species used for each AOGCM combination, shelf definition < 200 m S5: Post-hoc comparisons to assess statistical differences in proportional extinction between each of the greenhouse-icehouse transitions 200 m water depth S18: Definition of coastline orientation and small islands References References in support of the methods 2 S1. Methods S1.1. Simulation overview. We used a cellular automaton algorithm that linked a gridded geographic domain with a one-dimensional temperature landscape 1-3 to test the effect of paleogeography, sea level drop, and temperature change on extinction magnitude during three climate transitions: Late Ordovician, Eocene-Oligocene and Plio-Pleistocene. The geographic component of the model consisted of a global 1°x1° grid of shallow marine continental shelf for each of the three time periods. We generated virtual species that occupied grid cells in these continental margins as a function of their assigned temperature tolerances and dispersal abilities. The one-dimensional climate landscape was perturbed, and the geographic response of the virtual species recorded (Fig. 2). The framework builds on the model introduced by Qiao et al. 1 and Saupe et al. 3 , and is similar in concept to simulations explored by Rangel et al. 2 and Tomasovych et al. 4. S1.2. Paleogeography. We isolated the effect of continental configuration on expected extinction magnitude for the three target periods using paleogeographic reconstructions from 5 for the Late Ordovician (450 Ma), from Scotese 6 for the late Eocene (~37 Ma), and from Robertson Plc. for the Pliocene 7 (mid-Pliocene Warm Period, ~3.1 Ma). The palaeogeographic reconstructions needed to match those used in the AOGCMs (see S.1.4) and therefore derive from different sources (e.g., Blakey versus Scotese). For each paleogeographic reconstruction, we used the shallow marine areas around terrestrial continental margins (Fig. 1; Fig. S2), excluding Antartica for the Eocene and Pliocene. We considered both narrow and broad marine shelves: the former was generated by extending all terrestrial continental margins by one cell (1°), whereas the latter was generated by extending terrestrial continental margins by three cells (Fig. 1; Fig. S2). The broad margin in particular may be broader than most marine shelfs, given resolution of the climate model data (1° is approx. 100 km at the equator). S.1.2.1. Simplistic hypothetical climate gradient. Only paleogeography differed across the time periods of interest in these simulations. For the analyses in which the magnitude of climate change was held constant across all intervals, we generated hypothetical 'warm climate' and 'cold climate' temperature gradients by averaging interval-specific oceanatmosphere ge...

No abstract

No abstract

An early Cambrian greenhouse climate constrained by combining oxygen isotope and climate modeling data approaches.

Abstract. The Ordovician is a particular Period during Earth History highlighted by abundant evidence for continental-size polar ice-sheets. Modelling studies published so far require a sharp CO2 drawdown to initiate this glaciation. They mostly used non-dynamic slab mixed-layer ocean models. Here, we use a general circulation model with coupled components for ocean, atmosphere and sea ice to examine the response of Ordovician climate to changes in CO2 and paleogeography. We conduct experiments for a wide range of CO2 (from 16 to 2 times the preindustrial atmospheric CO2 level (PAL)) and for two continental configurations (at 470 Ma and at 450 Ma) mimicking the Middle and the Late Ordovician conditions. We find that the temperature–CO2 relationship is highly non-linear when ocean dynamics is taken into account. Two climatic modes are simulated as radiative forcing decreases. For high CO2 concentrations (≥ 12 PAL at 470 Ma and ≥ 8 PAL at 450 Ma), a relative hot climate with no sea ice characterises the warm mode. When CO2 is decreased to 8 PAL and 6 PAL at 470 and 450 Ma, a tipping-point is crossed and climate abruptly enters a runaway icehouse leading to a cold mode marked by the extension of the sea ice cover down to the mid-latitudes. At 450 Ma, the transition from the warm to the cold mode is reached for a decrease in atmospheric CO2 from 8 to 6 PAL and induces a ~ 9 °C global cooling. We show that the tipping-point is due to the existence of a quasi-oceanic Northern Hemisphere, which in turn induces a minimum in oceanic heat transport located around 40° N. The peculiar shape of the oceanic heat transport in the Northern Hemisphere explains the potential existence of the warm and of the cold climatic modes. This major climatic instability potentially brings a new explanation to the sudden Late Ordovician Hirnantian glacial pulse that does not require any large CO2 drawdown.