Calibration of the clumped isotope geothermometer in soil carbonate in Wyoming and Nebraska, USA: Implications for paleoelevation and paleoclimate reconstruction

Abstract

Recent work has shown that soil carbonates typically form as the soil dries after seasonal rainfall, and therefore record seasonal aspects of climate rather than mean annual conditions. Because soil carbonate formation is closely related to the timing of local rainfall and drying, it is necessary to understand the formation seasonality and temperature, and soil water δO18 values recorded in modern soil carbonate before accurate estimates of stable isotope-based paleoelevation and paleoclimate can be made. Here we study carbonate clumped-isotope (Δ47) and oxygen isotope (δO18) compositions of modern soil carbonates and seasonal variations of soil moisture and temperature of nearby climate stations in Wyoming and western Nebraska, USA, to understand the seasonality of soil carbonate formation in semi-arid to arid temperate montane settings. We find that soil carbonate clumped-isotope temperatures (T(Δ47)) are 3–5 °C higher than mean summer air temperature and are similar to or higher than average summer soil temperature. At depths >40 cm, soil moisture dramatically decreases in early summer following the cessation of spring rains and snowmelt and shows only brief increases after major mid–late summer rain events. Soil water δO18 values calculated using carbonate δO18 and T(Δ47) values are similar to the δO18 values of local mean summer precipitation. These lines of evidence suggest that soil carbonates in our study area formed during times of soil dewatering in early summer and after major summer storm events in mid–late summer, and in or near equilibrium with mean summer precipitation δO18 values.T(Δ47) values of modern soil carbonates are inversely correlated with elevation, with a lapse rate, −4.0 °C/km, similar to the modern air and soil temperature lapse rates. Calculated soil water δO18 values also are inversely correlated with elevation, with a lapse rate of −3.7‰/km. Sample elevations can be reconstructed using both T(Δ47) values and calculated soil water δO18 values between 900 m and 2600 m in our studied area, with an accuracy of ±0.5 km or less. However, δO18 values of precipitation and soil water reflect the complex interaction of climatic, environmental, and topographic conditions prior to carbonate formation, whereas T(Δ47) values may be influenced by site-specific conditions. We suggest that both soil carbonate T(Δ47) values and calculated soil water δO18 values should be used to corroborate paleoelevation reconstruction, and the seasonal nature of soil carbonate formation should be considered when soil carbonate T(Δ47) values and δO18 values are applied for paleoclimate reconstruction.