Accounting for Correlated Satellite Observation Error in NAVGEM

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner
Tuesday, 6 January 2015: 3:30 PM
230 (Phoenix Convention Center - West and North Buildings)
William F. Campbell, NRL, Monterey, CA; and E. A. Satterfield

We will show initial results from the inclusion of vertical (interchannel) correlation terms in the observation error covariance matrix (denoted R) for the Advanced Technology Microwave Satellite (ATMS) in the Navy's 4D-Var data assimilation system (NAVDAS-AR). Our dual formulation 4D-Var data assimilation scheme and observation sensitivity techniques make NRL uniquely well suited to explore correlated observation error, and the closely related problem of error of representation. Most operational NWP centers implement uncorrelated observation errors for data (a notable exception is the Met Office, which has been using vertically correlated observation error for IASI since January 2013 (Weston et al., 2014)). These centers compensate by thinning (discarding) or averaging data, and/or inflation of the assigned observation error variance. These suboptimal techniques can be rendered unnecessary by correct accounting for correlated error. The vertical observation error covariance for the ATMS was estimated using the Desroziers method (Desroziers et al., 2005) and an archive of historical satellite and NAVGEM model data. The results suggested lowering the error variance (diagonal of R) and introducing strong correlations (off-diagonal terms), especially in the moisture-sensitive channels. Because of the dual formulation of our data assimilation scheme, the inverse of the R matrix is not required, which has benefits both in reduced computation time and in solver convergence. In the future, we anticipate accounting for horizontally correlated error, which is infeasible for systems requiring the inverse of R. Preliminary results were evaluated with NRL's observation sensitivity tool to show the impact of correctly accounting for vertically correlated observation error in the ATMS instruments. Full cycling data assimilation experiments using standard forecast metrics are planned for later this year, along with implementation of correlated R for other instruments such as IASI, CrIS, AMSU-A, and high-resolution conventional obs.