Data Assimilation Meetings at Reading
Date | Meeting type | Speakers |
15 May 2013 | Invited Speaker |
Glenn Shutts (Met Office) Latest thoughts on stochastic kinetic energy backscatter - good and bad. (pdf) A brief historical review of the motivation for, and use of stochastic KE backscatter in NWP is presented. Emphasis will be placed on the need for physical justification of the ad hoc formulations used whose evolution has been dictated by their efficacy in medium-range ensemble prediction systems. Results from coarse-graining high resolution 'truth' simulations are discussed with respect to the view that backscatter is an upscale energy transfer process associated with quasi-two-dimensional motions. Given the well-known deficiencies in convection parametrization, it is suggested that convective vorticity generation is a likely source of the lack of spread in ensemble prediction systems. |