1 / 28

Aerosol Data Assimilation in GEOS-5

Aerosol Data Assimilation in GEOS-5. Arlindo da Silva (1) Arlindo.daSilva@nasa.gov Peter Colarco (2), Dirceu Herdies (6) , Ricardo Todling (1) , Anton Darmenov (1,5) , Virginie Buchard-Marchant (1,3) , Cynthia Randles (2,3) , Ravi Govindaradju (1,4)

ewa
Download Presentation

Aerosol Data Assimilation in GEOS-5

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Aerosol Data Assimilation in GEOS-5 Arlindo da Silva(1) Arlindo.daSilva@nasa.gov Peter Colarco(2), Dirceu Herdies(6), Ricardo Todling(1), Anton Darmenov(1,5), Virginie Buchard-Marchant(1,3), Cynthia Randles(2,3), Ravi Govindaradju(1,4) Global Modeling and Assimilation Office, NASA/GSFC Atmospheric Chemistry and Dynamics Branch, NASA/GSFC GESTAR Science Applications International Corp. Earth Resource Technology CPTEC/INPE, Brazil JCSDA 9th Workshop College Park, MD, March 29, 2011

  2. Outline • Aerosols in GEOS-5 • QFED NRT Biomass Burning Emissions • Aerosol Optical Depth Assimilation • Aerosol impact on GSI radiances • Concluding Remarks

  3. GEOS-5 Earth-System Model ESMF From Weather to Seasonal to Decadal Time scales.

  4. GOCART Component Dust radius Seasalt radius Black Carbon hydrophobic hydrophilic Organic Carbon hydrophobic hydrophilic Sulfate Mass • Goddard Chemistry, Aerosol, Radiation, and Transport Model [Chin et al. 2002] • Sources and sinks for 5 aerosol species • Convective and large scale wet removal • Dry deposition (and sedimentation for dust and sea salt) • Optics based primarily on OPAC

  5. GEOS-5/GOCART Forecasts • Global 5-day chemical forecasts customized for each campaign • O3, aerosols, CO, CO2, SO2 • Resolution: Nomally 25 km • Driven by real-time biomass emissions from MODIS • MODIS aerosol assimilation to become operational by Summer 2011. CO Smoke SO4 http://gmao.gsfc.nasa.gov/projects/glopac/

  6. QFED: Quick Fire Emission Dataset • Near real time estimates based on MODIS Fire products (AQUA/TERRA) • Started as an attempt to use diurnal fire counts to better distribute in time the monthly GFED emissions • Current focus on MODIS to be followed by • Geostationary • VIIRS • Plume Rise model (Freitas et al.) • Driven by GEOS-5 meteorology • Under tuning/validation

  7. QFED v2 – Fire Radiative Power • Assumes simple linear relationship for each biome • E(x,y,t) = cbiome * FRP(x,y,t) • Only 4 global constants to estimate (boreal, tropical fires, savannah, grasslands) • Spatial structure determined by satellite data • Accounts for pixels obscured by clouds • In principle, tuning parameters can be determined using inverse modeling techniques.

  8. QFED 2.1: GFED-based Calibration GFED and QFED v2.1 led to anemic BB AOT in GEOS-5

  9. QFED 2.2: MODIS AOD Calibration MODIS AOD 8 9 G5 QFED 2.2 G5 QFED 2.1 aBB ~ 2.5 aAN ~ 0.4 aBB ~ 1.8 aAN ~ 1.3 10 aBB ~2.2 aAN ~ 2.6 Tuning relative to QFED 2.1

  10. Aerosol Data Assimilation • Focus on NASA EOS instruments • Global, high resolution (1/4 deg) AOD analysis • 3D increments by means of Lagrangian Displacement Ensembles (LDE) • Simultaneous estimates of background bias (Dee and da Silva 1998) • Adaptive Statistical Quality Control (Dee et al. 1999): • State dependent (adapts to the error of the day) • Background and Buddy checks based on log-transformed AOD innovation • Error covariance models (Dee and da Silva 1999): • Innovation based • Maximum likelihood

  11. Analysis Variable Quality control and Data Assimilation methodologies assumes Gaussian statistics AOD (and errors) is not normally distributed Log-transformed AOD has better statistical properties: This 0.01 factor is determined from goodness-of-fit considerations Log ( 0.01 + AOD)

  12. MODIS/TERRA Ocean AOD LAOD AOD O-F LAOD O-F

  13. AERONET-MODIS/MISR Joint PDF MODIS-Terra (Ocean) MISR (Ocean+Land)

  14. Neural Net forAOD Empirical Retrievals • Ocean Predictors • Multi-channel • TOA Reflectances • Retrieved AOD • Angles • Glint • Solar • Sensor • Cloud fraction (<85%) • Wind speed • Target: AERONET • Log(AOD+0.01) • Land Predictors • Multi-channel • TOA Reflectances • Retrieved AOD • Angles • Solar • Sensor • Cloud fraction (<85%) • Climatological albedo • < 0.25 • Target: AERONET • Log(AOD+0.01)

  15. OceanNeural Net Retrievals (Terra) Original NN Retrieved

  16. MODIS Aqua MODIS Aqua Annual Mean 2008 Neural Net Neural Net

  17. Bonanza Creek Aerosol Data Assimilation GEOS-5 AOD With MODIS Assimilation GEOS-5 AOD Without MODIS Assimilation Bonanza Creek Cabo Verde Comparison against independent AERONET ground stations.

  18. Better consistency between MODIS on AQUA/TERRA. MISR has not been assimilated

  19. Aerosols in GSI • CRTM allows for the inclusion of (GOCART) aerosols • The GEOS-5 GOCART aerosol species have been introduced as state variables in GSI • No aerosol increments for now • Aerosol effects included in the observation operators for AIRS, HIRS, IASI, etc • NCEP/STAR implementing AOD analysis in GSI

  20. Aerosol Contamination of GSI Radiances Control Experiment Aerosol Experiment Aerosols fully interactive in GEOS-5 No AOD assimilation GSI observation operators: 15 GOCART species Concentration Effective radius Optical parameters internally determined by CRTM • Aerosols fully interactive in GEOS-5 • No AOD assimilation • Standard GSI global analysis • ARCTAS Period • Summer 2008 • Resolution: • ½ degree

  21. AOD Validation MISR MODIS GEOS-5 GEOS-5 GEOS-5 overestimates dust

  22. Dust Distribution

  23. Temperature AnalysisDT = Taero - Tcontrol

  24. Temperature AnalysisDT = Taero - Tcontrol

  25. AIRS Q/C Control Aero effects About 3% more AIRS observations are accepted

  26. Innovation Statistics Neutral Impact on short-term forecast

  27. Concluding Remarks • GEOS-5 produces twice daily aerosols forecasts • AOD assimilation scheduled for parallel testing this summer • The GOCART/GSI interface is under active development in coordination with NCEP. • GOGART ESMF Grid Component has been implemented in GFS • Facilitates sharing of other aerosol modeling/data assimilation capabilities

  28. http://portal.nccs.nasa.gov/cgi-yotc/e540_fp_2d_chem.cgi

More Related