Refinement of Eocene lapse rates, fossil-leaf altimetry, and North American Cordilleran surface elevation estimates

• We simulate Eocene climate of western North America with a global climate model.
• Eocene temperature and moist enthalpy lapse rates differ from present-day.
• Eocene temperature and moist enthalpy vary non-linearly with increasing elevation.
• Proxies agree with simulations when Cordilleran elevation is prescribed to be 2–3 km.
• Model suggests foreland elevations are significantly lower than the Cordillera hinterland.

 Estimates of continental paleoelevation using proxy methods are essential for understanding the geodynamic, climatic, and geomorphoric evolution of ancient orogens. Fossil-leaf paleoaltimetry, one of the few quantitative proxy approaches, uses fossil-leaf traits to quantify differences in temperature or moist enthalpy between coeval coastal and inland sites along latitudes. These environmental differences are converted to elevation differences using their rates of change with elevation (lapse rate). Here, we evaluate the uncertainty associated with this method using the Eocene North American Cordillera as a case study. To do so, we develop a series of paleoclimate simulations for the Early (∼55–49 Ma) and Middle Eocene (49–40 Ma) period using a range of elevation scenarios for the western North American Cordillera. Simulated Eocene lapse rates over western North America are ∼5 °C/km and 9.8 kJ/km, close to moist adiabatic rates but significantly different from modern rates. Further, using linear lapse rates underestimates high-altitude (>3 km) temperature variability and loss of moist enthalpy induced by non-linear circulation changes in response to increasing surface elevation. Ignoring these changes leads to kilometer-scale biases in elevation estimates. In addition to these biases, we demonstrate that previous elevation estimates of the western Cordillera are affected by local climate variability at coastal fossil-leaf sites of up to ∼8 °C in temperature and ∼20 kJ in moist enthalpy, a factor which further contributes to elevation overestimates of ∼1 km for Early Eocene floras located in the Laramide foreland basins and underestimates of ∼1 km for late Middle Eocene floras in the southern Cordillera. We suggest a new approach for estimating past elevations by comparing proxy reconstructions directly with simulated distributions of temperature and moist enthalpy under a range of elevation scenarios. Using this method, we estimate mean elevations for the North American Cordillera of ∼2 km using proxy temperatures and ∼3 km using proxy moist enthalpy. This discrepancy is likely related to an inconsistency between the proxy data for temperature and moist enthalpy. The combination of temperature and moist enthalpy estimates implies a warm and dry environment, which is inconsistent with geological evidence for humid conditions. Our study emphasizes that lapse rates in warm climate were likely much different than today's, and highlights the advantages of using climate models to understand past climate states and regional circulation patterns to derive more accurate paleoelevation estimates using proxy temperature and moist enthalpy.