Weather forecasts on timescales of less than two weeks have steadily improved over the past several decades. These forecasts have saved lives, led to better informed decisions, and helped prepare the public for upcoming extreme events. However forecast accuracy rapidly decreases as longer timescales are projected; those longer than ten days are especially limited. For certain applications like water management, agriculture, and energy usage, extended outlooks on longer timescales can provide vital data to inform advance preparations by the emergency management community and the general public. The aim of this project is to improve the quality of these forecasts by clarifying the role of the atmospheric layer known as the stratosphere, which lies 10km-50km above the surface.
The traditional approach to weather forecasting on one- to two-week timescales utilizes weather forecasting models. As computing power has increased over time, forecast skills have also improved. However, on timescales longer than two weeks, current forecasts are generally useless. Yet certain modes of variability in the climate system have timescales longer than this two-week threshold, and the key to longer scale prediction is to take advantage of these modes. By understanding the impacts of these modes of variability on surface weather, the potential for improved forecasts on a monthly timescale can be demonstrated and eventually realized.
Several such persistent modes of variability lie above in the stratosphere, and the goal of our research is to improve our fundamental understanding of how the stratosphere influences weather system, with the hope of better predicting surface climate. Specifically, we will make progress towards answering the questions: To what extent is variability in the stratosphere predictable? What are the key factors controlling the downward influence of stratospheric variability? What is mechanism whereby stratospheric variability affects the surface? Are models currently used operationally capturing these processes, and if not, what leads to biases?
