[7] Multi-scale & multi-component treatments 2


[7-1] February 28, 11:10-11:40

[Invited]

Estimating and sampling error variances in coupled models

Malaquias Pena (NOAA), Zoltan Toth and Jie Feng (NOAA-ESRL); Suranjana Saha, Xingren Wu, Jiande Wang, Jack Woolen, Patrick Tripp, Shrinivas Moorthi, Dave Behringer, H-C Lee (NCEP); Steve Penny (UMD)

 
Abstract

EMC is undertaking the development of the Unified Global Coupled System (UGCS) for weather to seasonal predictions. While the system is under development, the prospects for multi-scale data assimilation and model parameter optimization are being considered. In parallel with the UGCS's development, a series of diagnostic tools are being prepared to systematically assess its performance on each new model revision. In this talk, I will describe the characteristics and the status of the UGCS, and then will focus on one of the tools to diagnose uncertainty. The tool estimates analysis and short-range forecast error variances based on assumptions about the growth and correlation of errors, and is independent of the data assimilation scheme utilized. In a resent study where the truth is known, it is shown that these estimates are accurate within determined confidence intervals.

  Presentation file: 07_1_M.Pena.pdf