A comparison of hybrid variational data assimilation methods for global NWP. (22nd November 2018)

• Record Type: Journal Article
• Authors: Lorenc, Andrew C.; Jardak, Mohamed
• Is Part Of: Quarterly journal of the Royal Meteorological Society. Volume 144:Number 717(2018)
• Page Start: 2748
• Page End: 2760
• Publication Date: 2018-11-22
• Subjects: 4DVar -- 4DEnVar -- hybrid -- NWP; Meteorology -- Periodicals; 551.5
• DOI: 10.1002/qj.3401
• Languages: English
• ISSNs: 0035-9009

Abstract

Abstract : Variational data assimilation methods are reviewed and compared in the Met Office global numerical weather prediction system. This supports hybrid background‐error covariances which are a weighted combination of modelled static "climatological" covariances with covariances calculated from a current ensemble of forecasts, in both three‐ and four‐dimensional methods. For the latter, we compare the use of linear and adjoint models (hybrid‐4DVar) with the direct use of ensemble forecast trajectories (hybrid‐4DEnVar). Earlier studies had shown that hybrid‐4DVar outperforms hybrid‐4DEnVar, and 4DVar outperforms 3DVar. Improvements in the processing of ensemble covariances and computer enhancements mean we are now able to explore these comparisons for the full range of hybrid weights. We find that, using our operational 44‐member ensemble, the static covariance is still beneficial in hybrid‐4DVar, so that it significantly outperforms hybrid‐4DEnVar. In schemes not using linear and adjoint models, the static covariance is less beneficial. It is shown that the time‐propagated static covariance is the main cause of the better performance of 4DVar; when using pure ensemble covariances, 4DVar and 4DEnVar show similar skill. These results are consistent with nonlinear dynamics theory about assimilation in the unstable subspace. Abstract : The relative skill of NWP forecasts produced using different variational DA methods, plotted against the fraction of the assumed covariance which is derived from a current ensemble of forecasts. The results show that hybrid methods, combining static and ensemble covariances, beat pure static or ensemble covariances, and that hybrid‐4DVar (using a linear and adjoint model) beats hybrid‐4DEnVar (using the time‐evolution of the ensemble). This is because the models improve the static covariance; they perform best used together.