Recorded presentation9D.5
Studying the MJO with models that don't have one
Brian E. Mapes, Univ. of Miami/RSMAS, Miami, FL
Data assimilation drags models through observed states, using artificial "analysis tendencies." Diagnosis of these tendencies tells us a lot about model errors - and aslo, I will argue, something about the MJO in nature. I will describe studies of analysis tendency data sets in the NASA MERRA atmospheric re-analysis, composited according to the phase of the MJO. Also, the tactic of nudging any model to any existing set of analyses will be shown to be a low-budget approach to similar diagnosis for any model of choice (demonstrated with NCAR CAM).
The larger hope is that, from multiple models nudged to multiple analyses, we can begin to use the totality of global observations (which tell us about the planetary-scale MJO) to bring forward a more robust understanding the MJO in nature, while the concomintant lessons about model errors are relegated to their proper role as guidance for model-specific model improvement.
Session 9D, Intraseasonal Variability II
Wednesday, 12 May 2010, 10:15 AM-12:00 PM, Tucson Salon A-C
Previous paper Next paper