NOGAPS

1. Year-to-year variations of short-scale wintertime waveluence in NH polar regions have been discussed by Siskind et al. (GRL-2007)

2. NOGAPS

3. 2005-2006: Zonal mean GW-drag in NOGAPS-best simulations

4. 75-80 N: GEOS5 and Multi-Instrument Satellite Limb Viewing T-Observations (TIMED/SABER Aura/HIRDLS-MLS) 2006-SSW 2005-RW

5. Examples of the broad wave-spectra identification with S-Transform:

6. Spectra of HIRDLS (2007-2005, 70o-80oN, Jan 20-31) and GEOS5 Temperature -data performed by S-transform

7. Simulations of GW-rms in T-fields between 70-80 N with ensemble of waves launched between 8-16 km (right) with GEOS-5 Jan 2005 and 2006 background atmosphere (left). Jan 2006 HIRDLS short-wave T-rms (bottom plot).

8. Both procedures => to shift model simulations towards reliable observations to produce well-established climate signatures. Both overall modify momentum and heat tendencies. GWP makes it directly at every model grid and time step, while DA modifies variables incrementally. Both procedures establish non-local response of models to local adjustment of tendencies through the mass-wind balances. Stochastic GW-rms of wind and T can in principle represent uncertainties of forecast (error covariance in DA). Current GWP are formulated in “vertical column physics” framework, while DA employs horizontal correlations to spread analysis increments. GWPs are mainly solicited in the adjustment of momentum sources, while operational DA systems provide mainly the wind adjustment through calculated temperature analysis increments; Foundation of DA is error metrics of data and forecast uncertainties; Current GWPs are relatively deterministic although uncertainties of GW sources are large and waves are stochastic. in nature. GWP and Data Assimilation (DA): similarities and differences

9. On DA language, Generalized Inverse related to effects of GWs and possible cost functions

10. NOGAPS sensitivity studies suggest GW control mechanisms during SW events to reproduce high elevation of the stratopause

11. Jan WACCM (Base & GWPD) simulations and HRDI/UARS + UKMO (93 & 94) wind data

12. Possible inversion (balanced bias propagator) schemes for ZMF with global Temperature-data • Scheme 1 /extratropical balance, HSEq-scheme/: Temperature OmF => geopotential increment, restoring dU-increment and dAx-guess /parameterization dependent/. Spectral iterative solutions of zonal mean vorticity-divergence equations with updated GW momentum deposition without explicit vertical layer coupling. • Scheme 2 /HSEq +XiEq-scheme/ adds vertical layer coupling through explicit adjustment of meridional streamfunction (Xi) and “time-dependent” U-T iterations with inluence of meridional advection terms (layer coupling => elliptical equation for Xi, iterations => time integrations of U and T equations with observed composition).

13. U-balances in WACCM simulations (Base & GWPD) /fV* ~ Ax, leading MLT terms are forced/ SF=5

14. WACCM twins: HSEq, HSEq+XIEq wind inversions through mass-wind balances Setup: 2 WACCM runs Results: Compare 1 & 3 colums Sensitivity V-bar to momentum forcing terms

15. Temperature structures produced by GEOS5 and by simple KF mapping with HIRDLS T data (2006-01-20, top and 20006-01-27, bottom, at Z=40km, 20 km) GEOS5 KF-HIRDLS

16. 2005 (strong vortex) and 2006 (major warming) HIRDLS short-scale T-oscillations in polar NH latitudes /70N-80N, Jan/ 2005 2006