A pronounced feature in atmospheric weather patterns is the onset and subsequent development of wintertime large-scale cyclogenesis (LSC) over the Northern Pacific Ocean, extending to North America. The project will use singular vector (SV) diagnostic techniques, a fairly new approach, to identify and understand the precursors for the establishment of cyclogenesis events through an examination of the spatial patterns of errors in weather prediction models for periods before, during, and after occurrences of the enhanced jet flow within the cyclogenesis lifecycles. The research will be conducted through the analysis of global observational data sets, simulations with numerical models, and ensemble forecasts from operational models. An analysis of the characteristic evolution of differences between members of ensemble forecasts will be used to understand the sensitivity of LSC forecasts to initial condition uncertainties. Other problems the PIs will investigate include: the role of dynamical instability on the transient growth in LSCs and possible degradation in predictability during LSC events.
Understanding the dynamics and predictability during LSC events will contribute to the improvement in the medium range weather prediction. The research will contribute to educating a new generation in a fairly new scientific area, namely, singular vectors and adjoint systems associated with numerical models. The principal investigators will organize two workshops that will be tailored for weather forecasters to share information on predictability and to educate the weather forecasting community on how to use SV analysis in daily forecasts. Presently, SV analysis remains a formidable topic to many. This project may provide an opportunity to transform a complicated mathematical formulation into practical applications.