The Late Cenozoic has been marked by large and rapid fluctuations in temperature.This cooling has been attributed to accelerated erosion, with concomitant increased chemical-weathering rates and CO 2 drawdown from the atmosphere. At the same time, much of the supporting evidence appears to be affected by a sampling bias, implying that global erosion and weathering have remained largely constant over the past millions of years. We suggest that sedimentary archives of geomorphic activity, such as grain size and the ratio of terrestrial to oceanic sedimentation, which show accelerated erosion are not subject to these biases. Furthermore, the active tectonic settings where these erosion increases were likely to have taken place are exactly those locations where chemical-weathering signals are least likely to faithfully follow physical erosion rates. A lack of evidence for an increase in chemical weathering does not necessarily preclude an increase in physical erosion.In this contribution, we suggest an alternative interpretation in which erosion rates have increased in the Late Cenozoic but without significantly increased silicate weathering, which can explain the meagre response of chemical-weathering proxies.
| INTRODUCTIONMany recent papers have highlighted a lack of consensus on our understanding of how, and even if, climate impacts landscape change over geological timescales (e.g. Herman and Champagnac, 2016;Willenbring and Jerolmack, 2016). Over the past few million years, climate cooled significantly (e.g. Lisiecki & Raymo, 2005) such that glaciers began to grow globally. Initially, geological records were interpreted to indicate that sedimentation rates (and by proxy erosion rates) increased at the same time (Hay et al. 1988;Kuhlemann et al., 2002;Zhang et al., 2001;Herman et al., 2013). These datasets seemed to support positive feedbacks between mountain building, climate and erosion (Molnar and England, 1990; Raymo and Ruddiman, 1992 among others). The most accepted version of this argument is that global cooling drove glacial advance, which in turn resulted in increased erosion and weathering rates (e.g. Molnar and England, 1990;Herman et al., 2013;Herman and Champagnac, 2016). These increased weathering rates would then result in CO 2 drawdown through silicate-mineral weathering (e.g. Raymo and Ruddiman, 1992;Berner & Kothaval, 2001). This scenario was consistent with the available data until it was suggested that a measurement bias existed in the geological record (Sadler, 1999;Schumer & Jerolmack, 2009;Sadler & Jerolmack, 2014;Willenbring and Jerolmack, 2016