A3 - Model error and uncertainty at the midlatitude tropopause

Project leaders: Dr. Andreas Schäfler, Prof. Dr. George C. Craig

Other researchers: Konstantin Krüger (PhD student), Dr. Martin Weissmann (W2W member)

The main aim of this project is to locate and quantify model errors and uncertainties at the mid-latitude tropopause and to investigate their potential impact on the downstream weather in a combined approach using campaign data and ensemble methods. Despite several studies addressing this topic, little is known about the spatial and temporal distribution of errors.

Profiting from the unique observational data set of the North Atlantic Waveguide and Downstream Impact Experiment (NAWDEX) that was obtained in Phase 1 of W2W, we address four topics. First, we will compare independent profile observations across the tropopause with model fields in synoptic situations that are representative for mid-latitude weather. A quantification of analysis errors at the tropopause is important as strong gradients in wind and temperature across the tropopause shape the sharp contrasts in potential vorticity (PV) that act as a waveguide for Rossby waves and greatly influences the evolution of our daily weather. Additionally, a correct representation of moisture across the tropopause is of central importance as tropopause sharpness is continuously modified by radiative diabatic heating. We extend work from Phase 1 that has shown insufficiently reproduced wind gradients in current numerical weather prediction (NWP) systems to obtain a coherent view of the analysis error distribution near the tropopause. Second, we will use data assimilation output of the ECMWF ensemble data assimilation system to investigate the impact of individual NAWDEX observations on the tropopause structure and sharpness, where previous data assimilation studies showed contradicting results. Here we refine and extend previous work with deterministic model output to ensemble output in model space and focus on the impact on the tropopause structure. Third, we will investigate the temporal and spatial evolution of tropopause sharpness in the full model-level ECMWF ensemble in a collaboration with the visualization project C9. Fourth, we will apply ensemble sensitivity analysis (ESA) in the ECMWF ensemble system to investigate the role of analysis uncertainties for the downstream weather evolution and to investigate physical processes that impact tropopause sharpness with increasing lead time.