IMPACT OF CARBONATE DEPOSITIONAL SETTING AND SEASONALITY ON CLUMPED ISOTOPE RECORDS OF TOPOGRAPHY AND CLIMATE

Reconstructing continental paleotopography and paleoclimate depends on quantifying differences in earth-surface temperature through time and space. For instance, temperature changes through time can constrain the magnitude of glacial-interglacial climate change or cooling associated with uplift of a plateau surface, while spatial differences in temperature can map out latitudinal climate gradients in the past or the elevation difference between two contemporaneous paleosurfaces. Investigations of these topics have benefitted from the development of carbonate clumped isotope thermometry (T(D₄₇)), which circumvents some of the difficulties of other paleotemperature proxies by providing a direct measure of temperature at the time and place of carbonate formation (e.g., in lakes or soils). However, seasonal bias in carbonate formation poses a challenge to reconstructing temperature differences, specifically, for how to compare measured carbonate T(D₄₇) values to temperature estimates from other proxies, to modern climate data or climate model output, and to the T(D₄₇) values of carbonates formed in different depositional and/or climate settings. T(D₄₇) data for modern lacustrine carbonates from the southwestern USA and Tibet and from pedogenic (soil) carbonates from the Andes of Argentina and Chile and from the northwestern USA point to a warm-season bias in carbonate formation in most, but not all environments. Basic understanding of carbonate proxy seasonality can reduce uncertainties in paleoclimate and paleoelevation reconstructions and account for systematic offsets between T(Δ₄₇) and convenient climate variables such as mean annual air temperature (MAAT). The ability to translate T(Δ₄₇) into MAAT also affords the opportunity to compare clumped isotope records from different types of carbonates with reference MAAT estimates derived from other proxies. But while using such methods to estimate paleotemperature differences is simple in theory, in practice it is easier said than done. Examples from the Tibetan plateau and western US illustrate how reasonable choices of reference temperature can lead to vastly different estimates of temperature differences—and therefore paleoelevation and paleoclimate—interpreted from equivalent T(Δ₄₇) data.