THE CLUMPED ISOTOPE GEOTHERMOMETER IN SOIL AND PALEOSOL CARBONATE

In this talk we review the state-of-the-art of clumped isotopes in soil carbonate. We and others have studied both modern soils and buried paleosols in order to understand the relationship of clumped isotope temperatures to actual surface and burial temperatures. Most studies show that carbonates from modern soils in the western US, Africa, Wyoming Tibet, and India return clumped isotope temperatures (T°C_clumped) >> mean annual temperature (MAT). This suggests that soil carbonate generally forms in the very warmest months of the year, probably in response to intense soil dewatering. On average, T°C_clumped significantly exceeds mean annual temperature by 10-15°C due to (1) summertime bias in soil carbonate formation, and (2) sensible heating of soil. However, a few modern soils from South America have returned T°C_clumped close to MAT, apparently because the soils there do not dry sufficiently in the summer to form soil carbonate.

Aside from this exception, site mean annual temperature (MAT) across 0-30°C is highly correlated (r²=0.92) with T°C_clumped in modern soils. However, there are two challenges for reconstructing mean annual temperature in the past using these modern relationships. One is that multiple samples from simple and complete paleosol profiles should be sampled for best estimates of paleotemperature. The other issue is that clumped isotopes appear to be reordered rather easily, after ~3-4 km burial depth, or ≥125°C. Deeper than this and above this temperature, T°C_clumped in paleosol carbonate are reset to temperatures >40°C.