Estimating forecast error covariances for strongly coupled atmosphere-ocean 4D-Var data assimilation
Strongly coupled atmosphere-ocean data assimilation solves the assimilation problem in terms of a single combined atmosphere-ocean state. A significant challenge is a priori specification of the cross-domain forecast error covariances. These covariances must capture the correct physical structure of interactions across the air-sea interface as well as the different scales of evolution in the atmosphere and ocean. We investigate the nature and structure of cross-domain forecast error correlations using an idealised single-column coupled incremental 4D-Var assimilation system. We present results from identical twin experiments that use ensembles of cycled strongly coupled 4D-Var assimilations to estimate the forecast error correlations. Our results show significant variation in the strength and structure of the error cross-correlations in the boundary-layer between summer and winter and between day and night. These differences are explained by the underlying model physics, forcing and known atmosphere-ocean feedback mechanisms. We also explore the effect of including the ensemble derived atmosphere-ocean forecast error correlation information within our simple 4D-Var assimilation system. Introducing improved cross-domain error covariance information enables greater use of near surface observations and should in turn produce more accurate and balanced atmosphere-ocean analysis states and more reliable coupled model forecasts and reanalyses.
Prof. Nancy K Nichols University of Reading