2B.1
An Overview of the NRL Atmospheric Variational Data Assimilation (NAVDAS) and NAVDAS-AR (Accelerated Representer) Systems
Nancy L. Baker, NRL, Monterey, CA; and J. Goerss, K. Sashegyi, P. Pauley, R. Langland, L. Xu, C. Blankenship, B. Campbell, B. Ruston, T. Rosmond, and R. L. Pauley
The U.S. Navy, by virtue of operating primarily over coastal and open ocean areas, is extremely concerned about the accurate analysis and forecast of high impact weather systems that affect the safety of personnel, ships, and aircraft. Specific high impact weather events of interest include tropical cyclone forecasting, topographically enhanced high wind events, high winds and seas, dust storms and other obscurations to visibility.
NAVDAS is the U.S. Navy's three-dimensional variational data assimilation system used for the global model NOGAPS1, the relocatable mesoscale model COAMPSŪ2, and its on-scene counterpart, COAMPS-OS. NAVDAS became operational at Fleet Numerical Meteorology and Oceanography Center (FNMOC) on October 1, 2003 for the global NWP model NOGAPS, and on Dec 12, 2006 for COAMPSŪ. NAVDAS is being adapted to provide an aerosol optical thickness analysis for the Navy Operational Aerosol Analysis and Prediction System (NAAPS), and to replace the multivariate optimum interpolation analysis core of the ocean data assimilation system NCODA3.
NAVDAS is formulated in observation space, so that the computational costs are proportional to the observation count rather than the NWP model grid resolution (number of grid points). This makes NAVDAS an ideal choice for applications with high resolution data assimilation in data-sparse regions. NAVDAS was designed to be a precursor to U.S. Navy's next-generation four-dimensional variational data assimilation system, NAVDAS-AR (Accelerated Representer). NAVDAS-AR is being targeted for operational implementation for NOGAPS in late 2008.
An adjoint-based method to provide near real-time monitoring of the impact of observations assimilated operationally by NAVDAS (or NAVDAS-AR) on the short-range NWP forecast error been developed and tested at NRL-Monterey over the last several years. This technique has proven valuable for identify observing system issues relevant to the NAVDAS operational data assimilation, and several specific examples will be presented. The NAVDAS-AR adjoint will allow the temporal aspects of the satellite observation impact to be evaluated.
This presentation will give an overview of NAVDAS, its observation usage and applications. The NAVDAS-AR system will be briefly described, with emphasis on the differences in observation usage, observation impact, and forecast skill.
1 Navy Operational Global Atmospheric Prediction System (NOGAPS)
2 COAMPSŪ is a registered trademark of the Naval Research Laboratory
3 Navy Coupled Ocean Data Assimilation
Session 2B, Data Assimilation
Tuesday, 26 June 2007, 10:30 AM-12:00 PM, Summit B
Previous paper Next paper