Abstract. The CO2 liberated along subduction zones through intrusive/extrusive magmatic activity and the resulting active and diffuse outgassing influences global atmospheric CO2. However, when melts derived from subduction zones intersect buried carbonate platforms, decarbonation reactions may cause the contribution to atmospheric CO2 to be far greater than segments of the active margin that lacks buried carbon-rich rocks and carbonate platforms. This study investigates the contribution of carbonate-intersecting subduction zones (CISZs) to palaeo-atmospheric CO2 levels over the past 410 million years by integrating a plate motion and plate boundary evolution model with carbonate platform development through time. Our model of carbonate platform development has the potential to capture a broader range of degassing mechanisms than approaches that only account for continental arcs. Continuous and cross-wavelet analyses as well as wavelet coherence are used to evaluate trends between the evolving lengths of carbonate-intersecting subduction zones, non-carbonate-intersecting subduction zones and global subduction zones, and are examined for periodic, linked behaviour with the proxy CO2 record between 410 Ma and the present. Wavelet analysis reveals significant linked periodic behaviour between 60 and 40 Ma, when CISZ lengths are relatively high and are correlated with peaks in palaeo-atmospheric CO2, characterised by a 32–48 Myr periodicity and a ∼ 8–12 Myr lag of CO2 peaks following CISZ length peaks. The linked behaviour suggests that the relative abundance of CISZs played a role in affecting global climate during the Palaeogene. In the 200–100 Ma period, peaks in CISZ lengths align with peaks in palaeo-atmospheric CO2, but CISZ lengths alone cannot be determined as the cause of a warmer Cretaceous–Jurassic climate. Nevertheless, across the majority of the Phanerozoic, feedback mechanisms between the geosphere, atmosphere and biosphere likely played dominant roles in modulating climate. Our modelled subduction zone lengths and carbonate-intersecting subduction zone lengths approximate magmatic activity through time, and can be used as input into fully coupled models of CO2 flux between deep and shallow carbon reservoirs.

Plate tectonics, as the unifying theory in Earth sciences, controls the functioning of important planetary processes on geological timescales. Here, we present an open‐source workflow that interrogates community digital plate tectonic reconstructions, primarily in the context of the planetary deep carbon cycle. We present an updated plate tectonic reconstruction covering the last 400 million years of Earth evolution and explore components of the plate–mantle system that is involved in the exchange and storage of carbon. First, the workflow enables us to estimate subduction zone lengths through time, which represent the “tap” of carbon that is released at convergent tectonic margins. Second, we explore the role of Andean‐style versus intra‐oceanic subduction regimes during Pangea assembly and breakup. Third, we provide an improved model for carbonate platform evolution since the Devonian and evaluate the interaction of subduction zones and buried carbonate platforms. Last, we present a new model for estimating oceanic age, carbon content in the upper oceanic crust, and estimated (carbon‐containing) sediment thicknesses through time and present methods to track the subduction of this material through time. These components of the deep carbon cycle are key mechanisms controlling, or at least modulating, atmospheric CO2 on geological timescales and hence strongly influencing long‐term climate. We find that the mid to Late Cretaceous greenhouse climates were likely driven by increased subduction fluxes of volatiles and increased subduction zone interactions with carbonate platforms in the Tethyan tectonic domain. Our work highlights the importance of community digital plate tectonic reconstructions as a framework for studying key systems, such as the deep carbon cycle, that influence the life‐support mechanisms on our planet.

Carbon dioxide (CO2) liberated at arc volcanoes that intersect buried carbonate platforms plays a larger role in influencing atmospheric CO2 than those active margins lacking buried carbonate platforms. This study investigates the contribution of carbonate-intersecting arc activity on palaeo-atmospheric CO2 15 levels over the past 410 million years by integrating a plate motion model with an evolving carbonate platform development model. Our modelled subduction zone lengths and carbonate-intersecting arc lengths approximate arc activity with time, and can be used as input into fully-coupled models of CO2 flux between deep and shallow reservoirs.Continuous and cross-wavelet as well as wavelet coherence analyses were used to evaluate trends between 20 carbonate-intersecting arc activity, non-carbonate-intersecting arc activity and total global subduction zone lengths and the proxy-CO2 record between 410 Ma and the present. Wavelet analysis revealed significant linked periodic behaviour between 75-50 Ma, where global carbonate-intersecting arc activity is relatively high and where peaks in palaeo-atmospheric CO2 is correlated with peaks in global carbonateintersecting arc activity, characterised by a ~32 Myr periodicity and a 10 Myr lag of CO2 peaks after 25 carbonate-intersecting arc length peaks. The linked behaviour may suggest that the relative abundance of carbonate-intersecting arcs played a role in affecting global climate during the Late Cretaceous to Early 2Paleogene greenhouse. At all other times, atmospheric CO2 emissions from carbonate-intersecting arcs were not correlated with the proxy-CO2 record. Our analysis did not support the idea that carbonateintersecting arc activity is more important than non-carbonate intersecting arc activity in driving changes in palaeo-atmospheric CO2 levels. This suggests that tectonic controls are more elaborate than the subduction-related volcanic emissions component or that other feedback mechanisms between the 5 geosphere, atmosphere and biosphere played larger roles in modulating climate in the Phanerozoic.

Filtering and resampling of raw proxy-CO2 data (purple crosses). One median value was taken for each time step that displayed multiple observations. The result (blue) has one atmospheric CO2 (ppm) value per time step. The proxy-CO2 data was subsequently resampled using the resample function in MATLAB (black circles).

No abstract