TIMING OF PEDOGENIC CARBONATE FORMATION IN ARID SOILS IN RELATION TO CLUMPED ISOTOPE TEMPERATURE RECORDS

The clumped isotope geothermometer uses pedogenic (soil) carbonates to provide near-surface formation temperature (T(Δ ₄₇)) estimates important to paleoaltimetry and paleoenvironmental studies. It is expected that strong seasonal biases impact carbonate formation and ensuing T(Δ ₄₇) estimates, though the timing and conditions required for soil carbonate formation are as yet poorly understood. Pedogenic carbonates form over hundreds to thousands of years in response to calcite supersaturation of soil water caused by soil dewatering due to evapotranspiration, resulting in long-term time-averaged formation temperatures. The timing and frequency of these formation events vary with depth, impacting the values and interpretation of T(Δ ₄₇) profiles. This study develops a 1000 year stochastic weather simulation coupled with soil temperature and moisture models in order to identify carbonate forming wetting and drying events between 10 and 100 cm depth, based on conditions recorded in the semi-arid eastern Andean piedmont of Argentina under a summer precipitation regime. In this model, potential carbonate forming soil drying events occur in 100, 50, and 15% of simulated years at 10, 50, and 100 cm depth, respectively, suggesting that carbonate formation at depth is driven by large precipitation events in particularly wet years. We find that hot, isothermal with depth T(Δ ₄₇) profiles observed from carbonate nodules collected in a summer-only precipitation regime may be reproduced by simulated, time-averaged soil temperatures occurring during the initial stages of soil drying after large precipitation events. The results of this study suggest that if pedogenic carbonate forms in equilibrium with soil conditions such that T(D₄₇) values reflect soil temperature at the time of formation, carbonate formation is primarily biased by seasonal precipitation regimes that drive soil wetting and drying conditions. This conclusion is important in applying clumped isotope formation temperature estimates to paleoenvironmental studies, particularly in regions where precipitation regime may have changed or is poorly understood over time.