Divergent constraints in background-error modelling for global atmospheric data assimilation

• Applicant: Professorin Dr. Nedjeljka Zagar, Ph.D.
• Subject Area: Atmospheric Science
• Term: since 2021
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 461186383

In data assimilation for numerical weather prediction (NWP), the short-range forecast (called background) is corrected by the most recent observations to provide initial conditions (analysis) for the next forecast. In this process, spatial properties of the forecast errors need to be modelled so that the impact of observations is propagated in a physically optimal way. The reduction of forecast errors at synoptic scales in midlatitudes, where divergence is an order of magnitude smaller than relative vorticity, have been successfully modelled using quasi-geostrophic theory. In this way, observations of temperature can correct background errors in the wind field and vice versa. However, the analysis uncertainties and short-range forecast errors in global NWP systems are largest in the tropics where divergence is comparable to vorticity or even greater. Tropical analysis uncertainties can have a dominant influence on the accuracy of medium and extended-range prediction in the mid-latitudes. It is argued here that the reduction of analysis uncertainties in the tropics is a crucial step required to extend the range of useful prediction in global NWP models. The goal of this project is to develop a new constraint for the background error models in data assimilation that take into account divergence-dominated relationships between the tropical temperature and wind variables across many scales. The proposed constraint employs the mass-wind relationships that describe equatorial inertio-gravity waves and was demonstrated to hold a great promise for improving tropical analyses. This project will extend these ideas to the global atmosphere by 1) developing a new data assimilation framework that couples the tropical and midlatitude aspects of atmospheric dynamics, 2) using the flow-dependent background-error information from the ensemble forecasts, and 3) estimating potential of improved tropical analyses on extratropics.

• DFG Programme: Research Grants