Dynamics of Westerly Wind Bursts and Western US Climate over the Past 5 Million Years

Abstract

This dissertation covers investigations of the dynamics of Westerly Wind Bursts (WWBs), as well as the hydrological cycle, atmospheric circulation, and sea surface temperatures of the early-to mid-Pliocene (5.3-3 Ma) and last deglaciation (21-11 ka). Special attention is given to the past hydrological cycle over the western US, a region currently experiencing intense multidecadal drought that is expected to worsen as a result of anthropogenic climate change.

I investigate the mechanism of WWBs using an aquaplanet general circulation model (GCM), and find that eastward-propagating convective heating plays a role in the generation of these events. Using a more realistic model configuration, I study the nature of the atmospheric convection that leads to WWBs and the role of positive feedbacks involving surface evaporation. I find that disabling surface flux feedbacks a few days before a WWB peaks does not weaken the event, whereas directly suppressing convection by inhibiting latent heat release or eliminating surface evaporation rapidly weakens a WWB. By selectively suppressing convection near or away from the equator, I find off-equatorial cyclonic vortices are most important to WWB winds, while on-equator convection is unimportant. The conclusions are consistent with the idea that tropical cyclones, generally occurring more than 5° away from the equator, may be responsible for the majority of WWBs.

The next part of this dissertation concerns the climate of the Pliocene (5.3-2.6 Ma), the most recent period in Earth's history of prolonged global warmth. Curiously, geological evidence suggests many modern-day desert regions received higher levels of rainfall during the Pliocene, whereas simulations of the Pliocene and 21st century warming both predict drying over the subtropics. Furthermore, proxy reconstructions suggest anomalously warm sea surface temperatures (SSTs) 3-9 °C warmer than modern along midlatitude coastal upwelling sites. Using an atmospheric model, I show that introducing wetter conditions over subtropical continents weakens upwelling-favorable winds by reducing the land-sea surface pressure gradient, helping to sustain warm coastal SSTs. Furthermore, I show that underestimates of Pliocene rainfall along coastal regions may be related to insufficient SST warmth simulated over the upwelling zone. When SSTs on the California margin are raised to match proxy reconstructions, rainfall increases over much of adjacent western North America. The increased rainfall is maximal during the late boreal summer and can be interpreted as an expansion and intensification of the North American Monsoon. This suggests a positive feedback between land conditions, warm coastal SSTs, and weak upwelling-favorable coastal winds that could have sustained warm, wet conditions during the Pliocene and possibly in future climates.

The hydroclimate of the western US has varied dramatically in past climates. During the LGM (Last Glacial Maximum; 21 ka), the western United States was also covered by large lakes sustained by higher levels of precipitation. Increased rainfall was driven by the atmospheric circulation associated with the presence of large North American ice sheets, yet Pleistocene lakes generally reached their highstands not at LGM but during deglaciation. Previous work has given plausible explanations for the presence of lakes during the LGM, but few studies to date have explored why the lake highstands actually occurred around 16 ka when the planetary boundary conditions were different from LGM. Using GCM simulations run at time slices throughout the last deglaciation, I find that higher resolution simulations can produce a peak in winter rainfall over the Great Basin near 16 ka. The simulated peak is associated with a southward shift of the midlatitude storm track caused by changes in orbital configuration between 21 and 16 ka. The southward-shifted storm track, along with rising global temperatures, leads to increased rainfall over the southwestern US and helps explain the delayed lake highstands.