REVISION OF THE PALEOSOL PCO2 PROXY THROUGH DIRECT MEASUREMENT OF SOIL CARBONATE-HOSTED FLUID INCLUSIONS

The paleosol diffusion model is a commonly used atmospheric pCO₂ proxy due to the abundance and high frequency of paleosols in the terrestrial rock record. The model is based on equilibrium mixing between atmospheric CO₂ and soil-respired CO₂, which is tracked by means of carbon isotope mass balance on paleosol carbon phases. However, the concentration of CO₂ in soil is a crucial parameter that cannot be directly measured on paleosols, and represents the single largest source of inaccuracy in the model. Using a modern vertisol from Dallas, TX, we show that soil carbonate-hosted fluid inclusions retain geochemical signatures that are most likely indicative of soil conditions at the time of carbonate precipitation. Analysis of CO₂ from fluid inclusions permits the direct measurement of two commonly assumed components of the paleosol diffusion model: 1) the carbon isotopic composition of soil CO₂; and 2) total soil CO₂ concentration, both of which have a strong soil depth dependency. We develop a new methodology to make high precision isotope ratio and concentration measurements on small amounts of CO₂ released from pedogenic carbonate inclusions by online crushing into both an isotope ratio and a quadrupole mass spectrometer. When the full paleosol gas diffusion model is applied to these new soil-CO₂ constraints from a detailed depth transect in our test soil, it predicts near modern atmospheric CO₂ concentrations. While a more rigorous modern validation of the proxy using contemporaneous soil gas measurements is still in development, we apply the technique to Late Triassic age soil carbonates as a proof of concept. The good agreement between our new method and modern atmospheric conditions along with promising results from paleosol carbonates demonstrates that our fluid-inclusion approach results in a substantial revision of both the accuracy and precision of the paleosol pCO₂ proxy.