Neogene U.S. Southwest Temperatures Paced by Global Climate

Large uncertainty surrounds efforts to model the regional response to CO₂-driven warming in the southwestern U.S. The region hosts a seasonally variable hydroclimate and significant topography – much of which is tied to the region’s complex Cenozoic geologic history. These intricacies are difficult to reconcile in models, leading to disagreement even on the modelled sign of future precipitation change in the southwestern U.S. Previous and new stable isotope analyses of pedogenic carbonates from the well-dated Santa Fe Group of New Mexico suggest a potential shift from a middle Miocene winter-wet climate towards the dual-wet season regime seen in the region today, where annual precipitation is relatively low but delivered in both the summer and winter. This shift in regime might be spurred by either a strengthening of the North American Monsoon or a weakening of the westerlies. New clumped isotope analysis of these pedogenic carbonates documents a long-term cooling of as great as 18.5 ± 10.3°C between the Miocene Climatic Optimum (MCO) and the Pleistocene. This Neogene cooling trend in New Mexico tightly tracks the global benthic δ¹⁸O record over the same period, as well as Pacific sea-surface temperature records. This correlation suggests that paleotemperature change throughout the record is controlled by global climate rather than a potential shift in carbonate formation season driven by a shift in the precipitation regime. However, climate models, including both modern ocean-equilibrated LongRunMIP and Middle Miocene simulations, are unable to match the degree of continental MCO warmth in New Mexico indicated in our data even at CO₂ concentrations 8x higher than pre-industrial levels. Illustrating the magnitude of disagreement, Miocene and modern simulations respectively predict 4.6°C and 3.2°C of warming in New Mexico under a 560-ppm climate while the clumped isotope temperatures at the MCO are roughly 10°C warmer than modern mean annual temperatures in New Mexico. The disagreement between the magnitude of MCO warmth indicated by our clumped isotope record and that resulting from models can be explained either by (1) an underprediction of modelled temperature responses to CO₂-driven warming in the southwestern U.S. or (2) other factors that modify local temperature, such as changes in elevation associated with ongoing rifting along the Rio Grande rift and/or long-wavelength uplift associated with passage of the Farallon plate beneath the southwestern U.S. Whether the discrepancy in magnitude is an indication of extreme warmth at the MCO in New Mexico or a result of paleotemperatures encapsulating tectonic signals, our record demonstrates that global drivers pace temperature change in the U.S. Southwest.