Relationship between weathering, mountain uplift, and climate during the Cenozoic as deduced from the global carbon–strontium cycle model

A global carbon–strontium coupled model over the Cenozoic is presented. The carbon cycle includes improvements on the uplift parameter in the Himalayan and Tibetan regions and the hydrothermal flux at back-arc basins from the carbon cycle model of Kashiwagi and Shikazono [Kashiwagi, H., Shikazono, N., 2003. Climate change during the Cenozoic inferred from global carbon cycle model including igneous and hydrothermal activities. Palaeogeography, Palaeoclimatology, Palaeoecology 199, 167–185]. The strontium cycle incorporates groundwater flux, basalt weathering, and Himalayan weathering. The model result indicates that the Paleocene to the late Eocene is characterized by decreasing strontium weathering and gradually increasing isotopic value ( 87Sr/ 86Sr), which is a result of the decrease in radiogenic silicate weathering and less radiogenic carbonate weathering caused by decline of the atmospheric temperature. Since then, the 87Sr/ 86Sr continuously increases towards the present. The increase in the Sr isotopic value from 40 Ma to 35 Ma is attributed to appearance of hydrothermal calcite in the Himalayan region. The increasing patterns of the 87Sr/ 86Sr from that time to the early Miocene roughly correspond to the continental glaciations in high latitudes such as those at the Eocene/Oligocene boundary, in the middle Miocene and at the Oligocene/Miocene boundary. Radiogenic weathering accelerated by physical erosion due to glaciations and the subsequent deglaciations as suggested by previous studies might be related. However, there are time lags between the 87Sr/ 86Sr and δ 18O signals, which might have resulted from long residence time of Sr in the ocean. Reactivity of minerals under such low temperature and other climatic conditions is still unclear. More supportive works to assure a favourable environment for sufficient weathering and to constrain the timings and amplitudes of the glacial events during these periods would be necessary. The presented model shows the increase in the Sr weathering flux since the late Miocene, which would have induced the elevation of seawater 87Sr/ 86Sr whereas the evolution of the glaciations partially might have contributed to that. Some disagreement between other studies and the model is confirmed. The evaluation of the palaeo sealevel variation and carbonate compensation depth might produce the uncertainties in the estimate of weathering flux. Moreover, the basaltic province, through which several major rivers drain, and