Investigating the Microphysics of a Simulated Convective Event Using EMVORADO Polarimetric Forward Operator

This work is part of an effort to improve the representation of cloud and precipitation processes in weather and climate models, by identifying inconsistencies between model and observations. In particular we focus on the microphysical and dynamical processes within convective storms. For that purpose, simulations with the Terrestrial Systems Modelling Platform were performed for three convective events in northwestern Germany. The evaluation approach is done in the radar observation space using a polarimetric forward operator and radar-derived 3D wind fields. Thus, both simulated polarimetric radar variables and 3D winds can be directly compared to the radar observations. Discrepancies can arise from the actual model simulation or from the assumptions in the forward operator.
The model simulates several polarimetric signatures of convective storms like Zdr columns, rings of Zdr and RhoHV, as well as Zdr arcs. However, the shape (structure) and intensity of the Zdr columns are different, with the model underestimating its width and intensity.
Other shortcomings identified include the absence of a Kdp column in the simulation, along with a general underestimation of the convective area and rainfall.
Additionally, the dynamics of the simulated and observed cell show differences as the complexity of the observed storm is not adequately reproduced in the model. In the observation the storm has one main updraft and a few secondary updrafts, with complex interactions between them, while in the model only one updraft is simulated throughout the whole period. The intensity of the simulated updraft is also systematically stronger than the observed ones, with peak velocities often reaching 40 m/s in the model vs 25 m/s in the observation.
Analyzing the relationship between the updrafts and the Zdr columns revealed a delayed development of Zdr columns in the simulations compared to the observations.
The discrepancies found are analyzed and hypothesis are tested in order to understand what the mechanisms responsible for the differences are. These findings will help to tweak the microphysical parameterizations in the model and/or the assumptions required for the calculations in the forward operator.