Changes in the Rate of Ocean Crust Production over the Past 20 Myr

The rate of ocean-crust production exerts control over mantle heat loss, sea level, seawater chemistry, and climate through the global carbon cycle. Reconstructing ocean-crust production rates back in time relies heavily on the distribution of present-day seafloor age. How best to account for the incomplete preservation of older seafloor has led to differing conclusions about whether production rates have (Conrad and Lithgow-Bertelloni, 2007) or have not (Rowan and Rowley, 2017) changed significantly since the Cretaceous, with concomitant implications for the importance of seafloor spreading to global climate change. Here, we test the hypothesis that the global cooling of the sea surface that occurred in the late Miocene has a tectonic origin. We have constructed a new global synthesis of ocean-crust production rates for the past 20 Myr at high temporal resolution (~2 Myr). We use newly available plate reconstructions, global compilations of seafloor fabric and magnetic anomalies, and the recently derived astronomical timescale for magnetic polarity reversals. We also estimate changes in the length of each ridge system in the past 20 Myr. Investigating tectonic forcing of climate over this time period is advantageous for several reasons, most notably that it encompasses major and well-dated transitions in climate, there has been little destruction of ocean crust, and the ages of magnetic reversals are known with high accuracy. We find that the global ocean-crust production rate has decreased by ~30% since its maximum during 19-15 Ma. Although spreading-rate changes (mostly slowing) and length changes (mostly shortening) on eastern Pacific spreading centers dominate changes in the global sum, spreading rates along many mid-ocean ridges have decreased from the mid-Miocene to present. Using a simple model of the carbon cycle, we show that these ocean-crust production changes could have triggered the global cooling.