Research Projects
Marine Clouds
My primary research project focuses on how fog and marine clouds might change in the next 100 years as climate changes. Changes in large-scale circulation of air in the atmosphere, changes in the ocean temperature, and changes in the amount of moisture in the air will all impact marine clouds and coastal fog--the trouble is, each of these factors influences these delicate clouds in a different way, and so it is not immediately clear how they might change in the future. In order to deal with all these interacting factors, I am employing state-of-the art computer models of the climate system to simulate how marine clouds might change in the future.
For more information on marine clouds and coastal fog, visit the marine stratocumulus cloud primer.
Dust Bowl
The Dust Bowl disaster of the 1930's was one of the most expensive natural disasters (until Hurricane Katrina) to hit the United States. The disaster was started by a major drought, which was caused by La Nina-like cooling of sea surface temperatures. Poor farming practices (over-tilling of soil, and planting of drought-susceptible crops) compounded the disaster by allowing major dust storms to wreak havoc on the nation.
While the cause of the drought is generally agreed upon, it is not clear why the drought lasted so long, or why it was so severe. We are using high-resolution climate modles to investigate possible additional factors, such as land surface modification and dust-climate feedbacks, that may have aggravated drought conditions in the United States.
Climate Noise
Due to the chaotic nature of complex climate models, simulated climatologies can be quite sensistive to initial conditions. Because of this, controlled experiments that compare climatologies between two climate simulations have a certain amount of 'climate noise' that has no physical significance. In our type of experiments, this noise is generally small compared to whatever physical signal we are trying to detect. However our research shows that there are some cases where this climate noise is totally dominant, which makes interpretation of results from these controlled experiments nearly impossible.
Using a newly-developed control system for the climate model that we use for these experiments, we are investigating whether using an ensemble of carefully perturbed experiments can help to reduce the level of this noise relative to whatever physical signal we hope to detect