Group seminar on 13. June, 14:15 CET

The benefits of wind-mass coupled data assimilation for tropical observations on extra-tropical forecasts

Chen Wang

Data assimilation uses the dominant coupling between the wind and mass (temperature, pressure) variables to get balanced increments. The coupling in the mid latitudes is geostrophic, but in the tropics there is no clear multivariate coupling. As a result, the tropical variabilities associated with gravity waves are inadequately represented in the analyses and the analyses uncertainties are the largest in the tropics. Furthermore, the errors in the tropics can propagate to higher latitudes and destroy longer-term forecasts in the mid latitudes. In this research, the mass-wind relationships of equatorial waves are explicitly represented in a variational assimilation system for a shallow water model on the sphere, where the Rossby and non-Rossby (inertia-gravity (IG), Kelvin, mixed-Rossby-gravity (MRG)) waves are used as prognostic variables and control variables. When assimilating tropical wind and mass observations, accounting for the multivariate correlation improves the analysis in the tropics and the error reduction can be seen propagating to higher latitudes.