Jennifer Kowalczyk, Postdoctoral Research Associate, Earth, Environmental, and Planetary Sciences
Abstract Title: Wetter past, drier future? How vegetation and geography affect early Eocene North American precipitation
Abstract: As anthropogenic carbon emissions continue, climate model projections are critical for understanding how the Earth system will respond to this unprecedented forcing. Past high CO2 paleoclimates offer the best testing grounds to increase confidence in model performance under atmospheric CO2levels far outside the instrumental record. In particular, the early Eocene (~56-48 Ma) has been identified as the best analog for our long-term future climate under high emission scenarios. Proxy and fossil evidence indicates that North America was lushly vegetated with high mean annual precipitation in the early Eocene, in contrast to projections of drying and browning in the continental interior under high emission scenarios. Specifically, climate models predict drying for the American south and southwest in the annual mean, with more widespread drying in the summer. Understanding this discrepancy between a wetter high-CO2 past and projections for a drier high-CO2 future is critically important due to the impact of precipitation change on natural and human systems. Here we use the National Center for Atmospheric Research Community Earth System Model to investigate this discrepancy by conducting simulations under pre-industrial (PI) and 4xPI CO2levels with present-day and early Eocene geography. Our simulations show divergent responses of North American vegetation and precipitation to increasing atmospheric CO2 depending on geography. With present day geography, forest coverage decreases in the American west with increasing CO2, but with early Eocene geography, forested area increases. Similarly, precipitation decreases in the south with present-day geography, but increases with early Eocene geography, and the whole continental interior is much wetter in the early Eocene than with present-day geography under high CO2. We identify that this divergent response is driven by differences in sea surface temperature and geography. Namely, warmer water in the Gulf of Mexico increases atmospheric water vapor in the early Eocene case, and a taller, longer west coast mountain range allows moisture from the Gulf to be advected farther north in the early Eocene. The difference in forest area also plays a key role, as we show that transpiration supplies up to 50% of the moisture for summer precipitation in the continental interior in the early Eocene case. By showing how differences in geography and vegetation between the early Eocene and present day lead to different responses of continental precipitation to increased CO2, our study resolves the discrepancy between evidence for a wetter past and predictions for a drier future in North America.