Special Geosciences Dept. Seminar: Reconstructing the global atmosphere-ocean dynamics of hydroclimate extremes with data assimilation

Abstract:

Data assimilation (DA)-based reconstruction methods are statistical tools that optimally combine proxy data with climate models. DA-based paleoclimate reconstructions have the benefit of being physically-consistent across the reconstructed climate variables and of being able to provide dynamical explanations for past climate phenomena. For the first time, DA is used to reconstruct global hydroclimate on seasonal and annual timescales. Through a series of pseudo and real proxy experiments based on the PAGES 2k proxy network, it is found that hydroclimate variables like the Palmer-drought severity index can be skillfully reconstructed using DA. These reconstructions show skill during the boreal and austral growing seasons as well as at annual averages. Some dynamically-relevant drivers of hydroclimate included in the DA-based reconstructions, such as ENSO and ITCZ variability, are also skillfully reconstructed. These driving variables are then used to elucidate the mechanisms behind megadroughts over the last millennium in the American Southwest and hydroclimate extremes in the tropics.

Seminar Organizer: Prof. Eyal Haifetz