LATE MIOCENE GLOBAL COOLING DRIVEN BY A DECLINE IN PLATE TECTONIC CO2 RELEASE

Recent global compilations of marine surface temperatures reveal that global cooling commenced with the "refrigeration" of the Antarctic at circa 13.9 Ma, and accelerated in late Miocene (7-5.35) time to temperatures not much above Holocene conditions. Both hemispheres cooled in parallel, with the changes amplified at higher latitudes in comparison to the tropics. This cooling coincided with aridification of many subtropical regions on land and the expansion of C4 vegetation tolerant of lower CO₂ conditions. Despite the strong circumstantial case for CO₂ decline as the dominant cause of late Miocene climatic and evolutionary change, proxy indicators of CO₂ concentrations paint an equivocal picture at best. Here we provide evidence that global sea floor spreading (SFS) rates decelerated at exactly the times of major climatic coolings, linking a decline in tectonic degassing (at both subduction zones and mid-ocean ridges) to a fundamental shift in the global carbon cycle. Our work utilizes newly available global compilations of seafloor fabric and marine magnetic anomalies that are provided by the NSF-funded Global Seafloor Fabric and Magnetic Lineation Data Base Project. Previous global compilations of sea floor SFS typically binned estimates over 10 Myr increments, losing critical resolution on the timescale of late Neogene climate changes. We further improve the signal:noise of SFS estimates by incorporating recent advances in the astronomical calibration of the Miocene geomagnetic polarity timescale. We use two approaches to compile spreading rate estimates over the past 20 Myr at each spreading system: optimized finite rotation calculations, and averages of sea floor-spreading derived from the distances of magnetic lineations along flow lines in the sea floor. Weighted by ridge length, we find an ~25% reduction in global SFS since 15 Ma, with the strongest time of decline being the interval 8-5.5 Ma. Comparison of SFS to global temperature estimates suggest a very short time delay (unresolvable at the sample resolution) between tectonic forcing and climate response.