Director, Center for Advanced Data Assimilation and Predictability Techniques (ADAPT)
And, Professor, Department of Meteorology and Department of Statistics
Pennsylvania State University
Predictability limits, data assimilation, and simultaneous state and parameter estimation
I will first present our recent progresses in using state-of-the-art high-resolution global numerical weather prediction models in identifying the predictability limits of the multi-scale atmosphere. We show that an intrinsic predictability limit of about 2 weeks may indeed exist for the prediction of the mid-latitude day-to-day weather and is intrinsic to the underlying dynamical system and instabilities even if the forecast model and the initial conditions are nearly perfect. Currently, a skillful forecast lead time of midlatitude instantaneous weather is around 10 days, which serves as the practical predictability limit. Reducing the current-day initial-condition uncertainty by an order of magnitude extends the deterministic forecast lead times of day-to-day weather by up to 5 days, with much less scope for improving prediction of small-scale phenomena like thunderstorms. We further explain the intrinsic predictability limit by deriving a simple analytical error model applicable to the real atmosphere that has an approximate -3 energy spectral slope at the synoptic scales and an approximate -5/3 slope at smaller scales. This model shows surprisingly good consistency in successfully capturing the error growth characteristics of mid-latitude weather simulated by the operational global prediction models. This simple analytical framework signifies the joint effects of large-scale baroclinicity and small-scale convective instability in dictating the error growth and saturation, and ultimately limiting multiscale atmospheric predictability. In the second part of my talk, I will first show some of our recent advances in improving the practical predictability of severe weather and hurricanes through cloud-allowing assimilation of high-resolution Doppler radar and/or all-sky satellite radiance observations. I will further advocate for a generalized data assimilation software framework on Ensemble-based Simultaneous State and Parameter Estimation (ESSPE) that will facilitate data-model integration, uncertainty quantification and improved understanding and modeling of physical processes for the broad earth and environmental science communities. I will show example of its applications for improving atmospheric boundary and air-sea flux parameterizations, coupled atmosphere-CO2 data assimilation, coupled hydrology and land surface modeling, and palaeclimate analysis. Through augmenting uncertain model parameters as part of the state vector, the ESSPE framework will allow for simultaneous state and parameter estimation through assimilating in-situ and/or remotely sensed observations.