Geocheminar fall-2014

Climate models exhibit a persistent spread in their projected future changes in tropical precipitation. Can we use paleoclimatic proxies of past precipitation to constrain these projections? If a climate model simulates more realistically past precipitation changes, is it more credible for the future? To address this question, we analyze precipitation changes simulated for past and future climates by climate models from the CMIP5 archive. Over some regions like in South America, we can find encouraging statistical links between model behavior for past and future climates. Are the same physical mechanisms at play for past and future climates? And are the same mechanisms explaining the model spread? I'll present some work in progress on these issues. Then, to use paleoclimatic data to constrain precipitation projections, we need paleo precipitation records. Can we use archives of the past precipitation isotopic composition as proxies of precipitation? Or are they proxies of temperature? An isotopic general circulation model is used to try to resolve this long-standing controversy. Rather than resolving it, we find that the relative importance of precipitation and temperature in controlling the isotopic composition at paleo time scales depends on the model representation of cloud physics. New satellite measurements now allow us to better understand what control the isotopic composition at the daily scale. To what extent improved knowledge of isotopic controls at the daily time scale can help us better understand the controls at the paleo time scale? Finally, we look in more detail at why the relationship between precipitation and isotopic composition is so complex and model-dependent. We find that the relative importance of different cloud types and the shape of the vertical velocity profile affects this relationship. This complicates the interpretation of paleo records of isotopic composition. But in turn, measurements of water vapor isotopic composition could be used to better constrain convective and cloud processes. As an example, we present how measurements of water vapor isotopic composition could help us evaluate the relative timing of different convective and cloud processes simulated by models during intra-seasonal variations of tropical convection.