ADVANCING TERRESTRIAL PALEOCLIMATE WITH A PROCESS-BASED UNDERSTANDING OF THE SEASONAL BIAS OF THE CLUMPED AND STABLE ISOTOPIC COMPOSITIONS OF SOIL CARBONATES (Invited Presentation)

Soil carbonates are commonly used to investigate terrestrial paleoclimates, but do their seasonal biases betray scientists’ best efforts to interpret their stable isotope compositions? We seek to quantify and explain the varied seasonal biases of soil carbonate formation and dissolution. In a global compilation of clumped isotope data from modern soil carbonates, formation temperatures range from equaling that of local mean annual air temperature (MAAT) to exceeding MAAT by 24 °C. This apparent range in seasonal biases presents ambiguity for paleoclimate workers who wish to interpret clumped isotope temperatures of ancient soil carbonates, and it implies that reconstructions of meteoric water and soil respiration (through δ¹⁸O and δ¹³C) also could be seasonally biased. Empirical data suggests that the timing of rainfall and plant growth each explain some of the apparent variation in seasonal bias in clumped isotope temperatures of soil carbonates. We isolate the importance of these same environmental variables by modeling the formation of calcite in a one-dimensional soil profile (with HYDRUS 1-D), using meteorological and other input parameters based on calcic soils in New Mexico. The model results confirm that the timing of maximum plant growth (C3 vs C4) and rainfall (summer vs winter wet) each shift the timing of calcite formation and dissolution. In the model, calcite forms and dissolves episodically. This cycle of dissolution and re-precipitation generates a need to identify which calcite (and associated growth temperatures) would be preserved in the geologic record. We determine calcite preservation in two ways: 1) assuming that the most recently precipitated calcite dissolves first, and 2) assuming that existing calcite dissolves evenly. Amongst the simulations that vary rainfall and plant growth, and depending on how we determine calcite preservation, we simulate preserved carbonate formation temperatures of 13 - 22 °C. These temperatures are 1 - 10 °C hotter than local MAAT (12 °C), suggesting that the variables that we examined are responsible for about half of the varied biases in temperature observed in the global empirical data. We examine published paleosol data to show how these results can be leveraged to improve paleoclimate reconstructions.