Quantifying dynamical predictability is a fundamental problem in dynamical systems, having enormous social and economic implications. The task is especially challenging in nonlinear multiscale systems involving a vast-range of interacting spatial-temporal scales. Outstanding examples of multiscale phenomena include electrical power outages, earthquakes, tsunamis, and tropical cyclones (TCs). TCs in particular, are responsible for billions of dollars of damage annually in the United States alone.
To take into account the growth of inevitable uncertainties in the initial conditions and/or the uncertainties in the computational model formulation, ensemble forecasting has been developed and found important applications, especially in weather and climate forecasting, including TC forecasting. Currently, with simple models, one would choose as many ensembles as possible, with each ensemble containing a large number of members. When the forecast models become increasingly complicated, however, one would only be able to afford a small number of ensembles, each with limited number of members, thus sacrificing estimation accuracy of the forecasts. An even more serious limitation with the ensemble technique is that it cannot be applied at all, when the model equations are not available, which is often the case in practice.
The major objective of this project is to develop a new theoretical framework and a practical technique, called ?pseudo-ensemble? technique, for quantifying dynamical predictability, so that understanding of ensemble forecasting can be fundamentally advanced, forecast accuracy from weather to climate scales can be drastically improved, and computational complexity and data storage in forecasting can be tremendously reduced. These shall be achieved by unifying statistical, dynamical, and different information theoretic approached used in ensemble forecasting.
The ?pseudo-ensemble? technique in particular, shall be applicable to the important situation that observational data are available but the exact dynamical model equations are unknown. This thus shall allow determination of practical predictability of important weather and climate systems including floods, TCs, winter storms, and monsoons. The PI will recruit students from underrepresented minority and women groups.
The broader impacts of the proposal include establishment of an international collaboration with the National Center for Typhoon and Flooding Research (NCTFR) in Taiwan. The project will directly train both graduate and undergraduate students. The research results will be disseminated through various channels including the Purdue Climate Change Research Center and the NSF TeraGrid.
