Regional Ensemble–Variational Data Assimilation Using Global Ensemble Forecasts

• DOI: 10.1175/waf-d-16-0045.1
• Publication Date: 2016-10-26T17:05:47Z
• Authors: Wan-Shu Wu; David F. Parrish; Eric Rogers; Ying Lin
• Keywords: Ensemble forecasting; Ensemble Learning; Interpolation; Forecast verification

Abstract At the National Centers for Environmental Prediction, global ensemble forecasts from Kalman filter scheme in Global Forecast System are applied a regional three-dimensional (3D) and four dimensional (4D) ensemble–variational (EnVar) data assimilation system. The application is one-way variational method using hybrid static error covariances. To enhance impact, three new features have been added to existing EnVar system Gridpoint Statistical Interpolation (GSI). First, constant coefficients that assign relative weight between background now allowed vary vertical. Second, formulation introduced contribution analysis surface pressure. Finally, order make use of information mean disregarded GSI, trajectory correction, novel approach, introduced. Relative 3D algorithm, clear positive impact on 1–3-day realized when applying 3DEnVar analyses North American Mesoscale (NAM). DA was operationally implemented NAM Data Assimilation August 2014. Application 4DEnVar algorithm shown further improve forecast accuracy 3DEnVar. approach described this paper effectively combines contributions both systems produce initial conditions