3.3 Ensemble and variational data assimilation of surface particulate matter and MODIS aerosol optical depth

Monday, 7 January 2013: 4:30 PM
Room 9C (Austin Convention Center)
Craig S. Schwartz, NCAR, Boulder, CO; and Z. Liu and H. C. Lin

MODIS total aerosol optical depth (AOD) retrieval products and surface PM2.5 (particulate matter with diameters less than 2.5 microns) observations from the AIRNow network were assimilated in ensemble Kalman filter (EnKF) and three-dimensional variational (3DVAR) data assimilation (DA) frameworks. The forward operators to calculate model-simulated AOD and PM2.5 used 15 aerosol variables (hydrophobic and hydrophilic organic and black carbon; sulfate; sea salt in four particle-size bins; dust in five particle-size bins; and unspeciated contributions to PM2.5) output from the GOCART aerosol module implemented within the Weather Research and Forecast-Chemistry (WRF-Chem) model. Specifically, the observation operator for PM2.5 was a weighted sum of the 15 GOCART variables, while the Community Radiative Transfer Model (CRTM) was used to compute model-simulated AOD.

These DA systems were applied over a computational domain with 20-km horizontal grid spacing spanning the continental United States. Between 2 June and 14 July 2010, parallel 3DVAR and EnKF experiments assimilated both AOD and surface PM2.5 observations every 6-hrs in a “full-cycling” framework, and 48-hr WRF-Chem model forecasts were initialized from the analyses. As both DA systems used identical forward operators and assimilated the same observations, this experimental design permits a clean assessment of the utility of EnKF and 3DVAR techniques for aerosol forecasting.

This presentation will discuss results of the 48-hr PM2.5 and AOD forecasts. Challenges of assimilating AOD with an ensemble DA algorithm will also be detailed.

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner