Understanding Astronomically Forced Carbon Cycle Feedbacks Through the Lens of an Earth System Model

Milankovitch cycles recorded in marine sediments demonstrate the influence of astronomical forcing on Earth’s climate-carbon dynamics. Proxies suggest that during greenhouse climates, isotopically light carbon is released during episodic warm intervals (at eccentricity maxima) and re-sequestered during the following cooling (at eccentricity minima). However, the dominant carbon sources and sinks at play on orbital timescales remain unclear-- particularly when large dynamic ice sheets are absent as during the early Cenozoic. Methods: In an Earth system model (ESM), we apply 4-Myr-long transient astronomical forcing to examine how various climate-sensitive physical and (bio)geochemical processes respond and how this forcing is recorded in key oceanographic variables (temperature, pCO₂, δ¹³C of DIC, and wt% CaCO₃). Among others, we assess the impact of marine productivity, CaCO₃ compensation, terrestrial weathering, organic matter burial, and phosphorus cycling. Results: Most processes are driven by changes in local conditions -controlled by obliquity and precession, but these high-frequency changes are converted to low-frequency eccentricity cycles expressed in pCO₂, benthic δ¹³C, and wt% CaCO₃ as a result of the lowpass filtering effect of the ocean reservoir. While the magnitude of early Cenozoic δ¹³C variability can be explained by astronomically forced input and burial fluxes of marine organic carbon alone, the dominant frequency and relative phasing of proxies highly depend on the geographic distribution of landmasses that control organic carbon fluxes. For example, only short eccentricity cycles of 100 kyr periodicity (as opposed to long 400 kyr cycles) are simulated in benthic δ¹³C under favorable paleogeographic configurations. In our model, the pCO₂ and temperature response to orbital forcing is minimal, and eccentricity maxima coincide with enhanced preservation of CaCO₃. In contrast, early Cenozoic proxies suggest a stronger temperature response and reduced CaCO₃ preservation during warm intervals. Implication: Our results support the hypothesis that additional feedbacks that are not yet included here (e.g., terrestrial carbon or methane) were likely important controls during orbital-scale climate variability in greenhouse climates.