Automated determination of transport and depositional environments in sand and sandstones

As sand moves across Earth's landscapes, the shapes of individual grains evolve, and microscopic textures accumulate on their surfaces. Because transport processes vary between environments, the shape and suite of microtextures etched on sand grains provide insights into their transport histories. For example, previous efforts to link microtextures to transport environments have demonstrated that they can provide important information about the depositional environments of rocks with few other indicators. However, such analyses rely on 1) subjective human description of microtextures, which can yield biased, error-prone results; 2) nonstandard lists of microtextures; and 3) relatively large sample sizes (>20 grains) to obtain reliable results, the manual documentation of which is extremely labor intensive. These drawbacks have hindered broad adoption of the technique. We address these limitations by developing a deep neural network model, SandAI, that classifies scanning electron microscope images of modern sand grains by transport environment with high accuracy. The SandAI model was developed using images of sand grains from modern environments around the globe. Training data encompass the four most common terrestrial environments: fluvial, eolian, glacial, and beach. We validate the model on quartz grains from modern sites unknown to it, and Jurassic-Pliocene sandstones of known depositional environments. Next, the model is applied to two samples of the Cryogenian Bråvika Member (of contested origin), yielding insights into periglacial systems associated with Snowball Earth. Our results demonstrate the robustness and versatility of the model in quickly and automatically constraining the transport histories recorded in individual grains of quartz sand.

Neoproterozoic syn-glacial carbonate precipitation and implications for a snowball Earth

The Neoproterozoic 'snowball Earth' hypothesis suggests that a runaway ice-albedo feedback led to two intense glaciations around 717-635 million years ago, and this global ice cover would have drastically impacted biogeochemical cycles. Testing the predictions of this hypothesis against the rock record is key to understanding Earth's surface evolution in the Neoproterozoic. A central tenet of the snowball Earth hypothesis is that extremely high atmospheric CO₂  levels-supplied by volcanic degassing over millions of years-would be required to overcome a strong ice-albedo feedback and trigger deglaciation. This requires severely diminished continental weathering (and associated CO₂ drawdown) during glaciation, and implies that carbonate minerals would not precipitate from syn-glacial seawater due to a lack of alkalinity influxes into ice-covered oceans. In this scenario, syn-glacial seawater chemistry should instead be dominated by chemical exchange with the oceanic crust and volcanic systems, developing low pH and low Mg/Ca ratios. However, sedimentary rocks deposited during the Sturtian glaciation from the Adelaide Fold Belt-and contemporaneous successions globally-show evidence for syn-sedimentary dolomite precipitation in glaciomarine environments. The dolomitic composition of these syn-glacial sediments and post-glacial 'cap carbonates' implies that carbonate precipitation and Mg cycling must have remained active during the ~50 million-year Sturtian glaciation. These syn-glacial carbonates highlight a gap in our understanding of continental weathering-and therefore, the carbon cycle-during snowball Earth. In light of these observations, a Precambrian global biogeochemical model (PreCOSCIOUS) was modified to explore scenarios of syn-glacial chemical weathering, ocean chemistry and Sturtian carbonate mineralogy. Modelling results suggest that a small degree of chemical weathering during glaciation would have been capable of maintaining high seawater Mg/Ca ratios and carbonate precipitation throughout the Sturtian glaciation. This is consistent with a severe ice age during the Sturtian, but challenges predictions of biogeochemical cycling during the endmember 'hard snowball' models. A small degree of continental weathering might also help explain the extreme duration of the Sturtian glaciation, which appears to have been the longest ice age in Earth history.

Environmental insights from high-resolution (SIMS) sulfur isotope analyses of sulfides in Proterozoic microbialites with diverse mat textures

In modern microbial mats, hydrogen sulfide shows pronounced sulfur isotope (δ³⁴ S) variability over small spatial scales (~50‰ over <4 mm), providing information about microbial sulfur cycling within different ecological niches in the mat. In the geological record, the location of pyrite formation, overprinting from mat accretion, and post-depositional alteration also affect both fine-scale δ³⁴ S patterns and bulk δ³⁴ S_pyrite values. We report μm-scale δ³⁴ S patterns in Proterozoic samples with well-preserved microbial mat textures. We show a well-defined relationship between δ³⁴ S values and sulfide mineral grain size and type. Small pyrite grains (<25 μm) span a large range, tending toward high δ³⁴ S values (-54.5‰ to 11.7‰, mean: -14.4‰). Larger pyrite grains (>25 μm) have low but equally variable δ³⁴ S values (-61.0‰ to -10.5‰, mean: -44.4‰). In one sample, larger sphalerite grains (>35 μm) have intermediate and essentially invariant δ³⁴ S values (-22.6‰ to -15.6‰, mean: -19.4‰). We suggest that different sulfide mineral populations reflect separate stages of formation. In the first stage, small pyrite grains form near the mat surface along a redox boundary where high rates of sulfate reduction, partial closed-system sulfate consumption in microenvironments, and/or sulfide oxidation lead to high δ³⁴ S values. In another stage, large sphalerite grains with low δ³⁴ S values grow along the edges of pore spaces formed from desiccation of the mat. Large pyrite grains form deeper in the mat at slower sulfate reduction rates, leading to low δ³⁴ S_sulfide values. We do not see evidence for significant ³⁴ S-enrichment in bulk pore water sulfide at depth in the mat due to closed-system Rayleigh fractionation effects. On a local scale, Rayleigh fractionation influences the range of δ³⁴ S values measured for individual pyrite grains. Fine-scale analyses of δ³⁴ S_pyrite patterns can thus be used to extract environmental information from ancient microbial mats and aid in the interpretation of bulk δ³⁴ S_pyrite records.