P2.6 DART/WRF: A community mesoscale ensemble data assimilation facility

Wednesday, 19 August 2009
Arches/Deer Valley (Sheraton Salt Lake City Hotel)
Jeffrey Anderson, NCAR, Boulder, CO; and N. Collins, J. Hacker, G. S. Romine, C. Snyder, H. Liu, T. J. Hoar, D. C. Dowell, and R. Torn

The Data Assimilation Research Testbed (DART) provides open-source,

state-of-the-art software tools for ensemble data assimilation.

DART includes interfaces that allow it to be used with regional,

global, nested, and single-column domains with the latest releases of the

Advanced Research WRF model. DART can assimilate in situ observations

as well as radar reflectivity and radial velocity, GPS radio occultation,

tropical storm position and intensity, and QuikSCAT winds. DART includes

adaptive inflation, localization, and observation thinning tools that can

produce high-quality, computationally efficient ensemble assimilations on

a variety of parallel computing platforms. DART comes with an assortment of

analysis tools that facilitate evaluating forecast quality, model

systematic error, and the sensitivity of forecasts to initial conditions.

Sample results are shown for ensemble analysis and prediction of convection

over the United States and tropical storms in the Pacific and Atlantic.

- Indicates paper has been withdrawn from meeting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