wg2 t6

Assimilation of space and air borne measurements in a tropospheric chemistry transport model by 4D-var: 

Assimilation of space and air borne measurements in a tropospheric chemistry transport model by 4D-var H. Elbern, A. Strunk Rhenish Institute for Environmental Research at the University of Cologne

Contents: 

Contents Introduction: why 4D-var Assimilation of flight campaign data Assimilation of satellite data

Design of the case study: 

Design of the case study CTM and adjoint CTM (symmetric operator split): RADM2 gas phase: 61 species 4th order Bott advection, horiz. & vert. Implicit diffusion (Thomas algorithm) Grid: 125/54 km horiz. spacing, 100 hPa large grid: 77 x 67 x 26 (refined tropopause: Dz ~1 km) Meteorological fields by MM5 Case studies: CONTRACE: 14. November 2001 SPURT: 17.-19. January 2002

Design of the assimilation experiment : 

Design of the assimilation experiment assimilation interval 00:00-16:00 UTC( max), optimisation: SPURT: chemical state variables CONTRACE: chemical state variables + emission rates isotrop. background error covariance matrix (BECM) L-BFGS (quasi-Newton) minimisation Preconditioning by square root (BECM)

Mesoscale EURAD 4D-var data assimilation system: 

Mesoscale EURAD 4D-var data assimilation system direct CTM observations analysis gradient fore- cast

SPURT: Trace gas Transport in the Tropopause RegionCoord.: A. Engel, Univ Frankfurt: 

SPURT: Trace gas Transport in the Tropopause Region Coord.: A. Engel, Univ Frankfurt 17.-19. Jan. 2002 a.m. 17. p.m. 17. a.m. 18. p.m. 18. a.m. 19. p.m. 19. Original project objectives: to analyse transport signatures in terms of chemical ages in the mixing region of troposphere and stratosphere However: not suitable for a limited area data assimilation set-up. Hence, find chemical states, correlation length. Find chem state at 00:00 UTC to best match daytime observations 00 10 UTC 12 assimilation interval

SPURT Jan 17, 2002: 

SPURT Jan 17, 2002 1. guess assimilation result observations flight height [km] O3 CO

SPURT Jan 18, 2002: 

SPURT Jan 18, 2002 1. guess assimilation result observations flight height [km] am: Canary Isl.  Casablanca pm: Casablanca  Hamburg O3 CO

SPURT Jan 19, 2002: 

SPURT Jan 19, 2002 O3 CO 1. guess assimilation result observations flight height [km] am: Hamburg North Cape pm: North Cape  Hamburg

SPURT Jan 19, 2002: 

SPURT Jan 19, 2002

CONTRACEConvective Transport of Trace Gases into the upper Troposphere over Europe: Budget and Impact of ChemistryCoord.: H. Huntrieser, DLRflight path Nov. 14, 2001: 

CONTRACE Convective Transport of Trace Gases into the upper Troposphere over Europe: Budget and Impact of Chemistry Coord.: H. Huntrieser, DLR flight path Nov. 14, 2001

CONTRACE Nov. 14, 2001, southbound: 

CONTRACE Nov. 14, 2001, southbound 1. guess assimilation result observations flight height [km] O3 H2O2 CO NO HCHO

CONTRACE Nov. 14, 2001 north (= home) bound: 

CONTRACE Nov. 14, 2001 north (= home) bound O3 H2O2 CO NO 1. guess assimilation result observations flight height [km]

Assimilated data types: 

Assimilated data types in situ surface observations radio sondes aircraft tethered balloons Satellites: GOME tropospheric NO2 columns tropospheric NN derived ozone profiles (Dh=1km)

Assimilation of GOME NO2 tropospheric columns, 3.8.1997: 

Assimilation of GOME NO2 tropospheric columns, 3.8.1997 GOME NO2 columns: Courtesy of A. Richter, IFE, U. Bremen # molec./cm2 NOAA ch 3 (near infrared) GOME NO2 columns: Courtesy of A. Richter, IFE, U. Bremen

GOME ozone profile assimilation (3.8.1993)Data: Neuronal Network retrieval (Müller et al., 2003) : 

GOME ozone profile assimilation (3.8.1993) Data: Neuronal Network retrieval (Müller et al., 2003) + NN ozone retrieval

GOME ozone profile assimilation II (3.8.1993)Data: Neuronal Network retrieval (Müller et al., 2003): 

GOME ozone profile assimilation II (3.8.1993) Data: Neuronal Network retrieval (Müller et al., 2003) + NN ozone retrieval

Assimilation of GOME NO2 tropospheric columns, 5.8.1997: 

Assimilation of GOME NO2 tropospheric columns, 5.8.1997 GOME NO2 columns: Courtesy of A. Richter, IFE, U. Bremen # molec NO2/cm2 NOAA ch 3 (near infrared) # molec NO2/cm2 # molec NO2/cm2 post assimilation forecast started: 3.8.97 06:00 UTC forecast without assimilation

Conclusions:: 

Conclusions: successful UTLS data assimilation in a mixing regime requires readiness to identify different chemical regimes locally: strat. intrusions, free troposphere, updrafts problem of first guess with phase shifts (e.g. streamers must be captured correctly) smoother algorithms give qualitative indication on the ability to simulate observations

Acknowledgments: 

Acknowledgments Data provision CONTRACE: Dr. H. Huntrieser and project, SPURT: Dr. C. Schiller and group, ICG-I, FZ Jülich; Dr. H. Fischer and group, MPI-C, Mainz Funding by BMBF ZAM, FZ Jülich for computational facilities