Environmental Science and Engineering Seminar

Weather systems span a range of scales, from small (e.g., individual cumulus clouds) to large (e.g., mid-latitude winter storms), and each develops in a specific context. The details of the thermodynamic structure of the atmosphere, the three dimensional distribution of wind, and characteristics of the underlying surface pre-determine the types and outcomes of possible weather events. Changes in the climate system are synonymous with changes in the development context for weather, and as such it is important to understand how contextual changes influence the outcomes of various weather systems.

We have conducted a set of experiments designed to assess how weather systems respond to changes in climate. The weather events of interest include: tropical thunderstoms, mountain precipitation, winter storms, and hurricanes. The fundamental challenge is to quantify the manner in which simultaneous changes in environmental factors influence weather-related outcomes (e.g., precipitation). The difficulty arises from the fact that weather systems are complex: they are influenced by multiple different control factors, and exhibit nonlinear responses. We show how ensembles of simulations, machine learning techniques, and data assimilation theory can be used as effective tools for examining multivariate weather-climate interactions. We also highlight the fact that many types of weather system exhibit strong transitions in behavior as the climate system warms.