For more information about this meeting, contact Robin Enderle.
| Title: | Reduced Stochastic Models for Filtering Turbulent Dynamical Systems |
|---|
| Seminar: | Job Candidate Talk |
|---|
| Speaker: | John Harlim |
|---|
| Abstract: |
|---|
| Fundamental issues in advancing real-time filtering problems are model errors even when expensive high-dimensional global weather forecasting models are used. For example, the modern operational weather models poorly reproduce the tropical observational records even with 10^9 state variables; this long-standing issue hinders the global weather model forecasting skill to improve from weekly to monthly, as reported in a recent article in the World Meteorological Organization bulletin.
In this talk, I will discuss a class of computationally fast reduced stochastic models to help filtering. I will try to convince you that such a filtering approach can provide practical guidelines to improve estimation of moisture coupled tropical waves. Furthermore, I'd like to use mathematical theory to convince you that one can achieve ``optimal filtering" with judicious choice of reduced stochastic models. |
Room Reservation Information
| Room Number: | MB114 |
|---|
| Date: | 02 / 12 / 2013 |
|---|
| Time: | 02:00pm - 03:00pm |
|---|
