MTG-IRS OSSE on regional scales

Huang et al: MTG-IRS OSSE MMT, 17-18 June 2008. 1

MTG-IRS OSSE on regional scales

Xiang-Yu Huang, Hongli Wang, Yongsheng Chen and Xin Zhang

National Center for Atmospheric Research, Boulder, Colorado, U.S.A.

Stephen A. Tjemkes, Rolf Stuhlmann

EUMETSAT, Darmstadt, Germany


Contents

• The OSSE configuration

• The nature run

• Simulated (conventional and MTG) observations

• Calibration experiments

• Data assimilation and forecast results

• 4D-Var results

• Summary


The OSSE Configuration

• Nature run model: MM5

• Assimilation model: WRF

• Data assimilation scheme: WRF 3D-Var (4D-Var)

• Selected cases:

– International H2O Project (IHOP, 13 May - 25 June 2002)

– Southern Great Plains of US

http://catalog.eol.ucar.edu/ihop/catalog/missions.html









OSSE setup


IHOP Case

Three convection cases are

selected from 11 June 2002

to 16 June 2002:

• 11 June: Dryline and Storm

• 12 June: Dryline and Storm

• 15 June: Severe MCSMap illustrating the operational instrumentation within the IHOP_2002 domain (Weckwerth et al. 2004)


Nature run configuration

• Nature model: MM5• Grid points: 505X505X35• Horizontal resolution: 4Km• Time step: 20s• Physics parameterizations:

– Reisner 2 microphysics– No cumulus parameterization– MRF boundary layer

• Initial and Lateral boundary condition:

– 6-hourly ETA model 40-km analyses (phase I: GFS analyses)

• ~ 220 minutes with 256 CPUs

model domain


observed 6-h rainfallobserved 6-h rainfall simulated 6-h rainfallsimulated 6-h rainfall

The observation is on The observation is on Polar Stereographic Projection Grid. Grid.The simulated rainfall is on The simulated rainfall is on Lambert Projection Grid. Grid.The color scales are different.The color scales are different.

0600 UTC 12 Jun0600 UTC 12 Jun 0600 UTC 12 Jun0600 UTC 12 Jun

Case A: 11 June CaseCase A: 11 June Case


observed 6h-rainfallobserved 6h-rainfall simulated 6h-rainfallsimulated 6h-rainfall


Case B: 12 June CaseCase B: 12 June Case


Case C: 15 June CaseCase C: 15 June Case

observed 6h-rainfallobserved 6h-rainfall simulated 6h-rainfallsimulated 6h-rainfall



Simulated Dataset

• WRF-Var is employed to produce simulated conventional observations

(NCEP ADP Upper Air sounding, Surface Observation, AIRCAR/AIRCFT and Satellite wind data )

– Simulated observations use the actual locations and

times

– Use realistic observation errors


Simulated Dataset• NCEP ADP Upper Air sounding • NCEP ADP Surface Observation

501 Surface33 Sounding

Example of Simulated Data distribution

within the time window: 2300 UTC 11 June to 0100 UTC 12 June 2002


Simulated Dataset• NCEP ADP AIRCAR/CRAFT • NCEP ADP Satellite wind

748 ACAR/CFT



717 SATOB


MTG-IRS retrieved profiles

11679 MTG-IRS RP




T q

Error statistics for MTG-IRS


Calibration runs

• Generate pseudo radiosonde observations from the

nature run.

• Exp 1. No-obs.

• Exp 2. Pseudo-obs. Data assimilation experiment using the pseudo observations.

• Exp 3. Real-obs. Data assimilation experiment using real (radiosonde) observations.


SOP: Simulated Observation Profiles; ROP: Real Observation Profiles

Difference in T (K), 12 km results, averaged over 1800 UTC 11 to 1200 UTC 15 June 2002

At analysis time At 12 h FCST


Difference in q (g/kg), 12 km results, averaged over 1800 UTC 11 to 1200 UTC 15 June 2002


SOP: Simulated Observation Profiles; ROP: Real Observation Profiles


Difference in u (m/s), 12 km results, averaged over 1800 UTC 11 to 1200 UTC 15 June 2002



Experiments design (3D-Var)• Forecast model: WRF V3.0• Grid points: 169X169X35• Horizontal resolution: 12 Km• Time step: 60s• Physics parameterizations:

• WSM6 microphysics• Grell cumulus paramerization• MRF boundary layer

• Cases: 2002-06-11 12Z to 2002-06-15 12Z• Data: MTG-IRS retrieved profiles + SOP• Verification against truth


Lists of Experiments


RMSE: 6hourly cycling resultsAnalysis Time 12 h FCST


RMSE: 6hourly & 1hourly cycling results

Analysis Time 12 h FCST


Averaged ETS and BIAS at 18 h FCST

ETS BIAS


• Forecast model: WRF• Data assimilation system: WRF 4D-Var• Grid points: 169X169X23• Horizontal resolution: 12Km• Time step: 60s• Assimilation window: 3 hours• Physics parameterizations:

– WSM6 microphysics– New Grell cumulus paramerization– MRF boundary layer

• Case A: 2002-06-11 18Z to 2002-06-12 12Z• Case B: 2002-06-12 18Z to 2002-06-13 12Z• Background: GFS analysis and forecasts• Data: Retrieved profiles + SOP or “Truth”• Verification against truth

Experiments design (4D-Var)

D1


RMS error of Case A12 h FCST


Case A at 18H FCST

ETS

BIAS


RMS error of Case B12 h FCST


Case B at 18H FCST

ETS

BIAS


Summary

• Three storms are well reproduced in the 5 day nature run.

• The calibration experiment shows that the real and simulated

observations have the similar impacts on the analyses increments

and forecasts differences.

• The quality of the retrievals has been improved significantly.

• The forecast skill of wind, temperature and moisture forecasts is

improved when MTG-IRS T and/or q profiles are assimilated.

Rainfall forecast skill is improved also.

• In two short periods of experiments, assimilating MTG-IRS data

using 4D-Var gives slightly better forecast skills over 3D-Var.


Current and planned work• Continue the 4D-Var experiments

• Other nature runs

• Try other cases – Europe– Tropics– Over oceans– …

• Assimilate wind data

Documents

MTG-IRS OSSE on regional scales