Thursday, 31 August 2023
Boundary Waters (Hyatt Regency Minneapolis)
Handout (2.6 MB)
Polarimetric radar measurements contain information about hydrometeor microphysical properties and processes. This can be used to evaluate and improve numerical weather models. Reliable forward operators (FO), which link physical cloud and precipitation properties to radar quantities, are key tools to exploit the polarimetric measurements in model evaluation, data assimilation, quantitative precipitation estimation, and interpretation of microphysical process fingerprints.
EMVORADO is the native radar FO of the ICON model putting strong emphasis on model-operator-consistency and is operationally applied in the German Weather Service's radar data assimilation. It was recently extended to the simulation of polarimetric radar quantities and, like many currently available polarimetric radar FO, applies the spheroidal T-matrix model for both liquid and frozen hydrometeor classes.
Comparing to radar observations, deficiencies of simulated polarimetric variables have been identified in the melting layers as well as in the dendritic growth and aggregation layer and could be attributed to ICON’s 2-moment microphysics and to assumptions in EMVORADO. In particular, the differential polarimetric signature of snowflakes was found difficult to reproduce when approximating ice hydrometeors as homogeneous effective-medium spheroids in the T-Matrix approach. This is in line with, e.g., Schrom and Kumjian (2018), who found the detailed structure of particles to be essential for the characterization of their scattering properties, especially for polarimetric quantities.
Hence, methods capable of capturing the detailed snow particle shapes are needed in order to accurately simulate the polarimetric radar signatures. However, the level of detail of the ice particle description used in operational weather models is not sufficient to fully characterize the scattering simulations, and the range of properties that are found for natural snowflakes are not covered by the few available databases of realistic ice and snow polarimetric scattering data.
Our key objective is to provide a more realistic polarimetric radar FO, which can also be more consistently linked with microphysical parameters used in numerical models. The core component of our project is the extension of the current T-Matrix based scattering tables in EMVORADO with DDA simulations of frozen hydrometeors with realistic particle properties. Here we present the results from the first steps, where a DDA-based scattering module is implemented into EMVORADO with primary focus on the consistency between the microphysical assumptions made in the ICON model and in the forward simulations. This allows direct comparison with the legacy T-matrix simulations, hence an immediate assessment of the impact of the scattering model. For later steps, we plan to model the particle structures using a combination of a Lagrangian super-particle model and a snow particle model that covers aggregation and riming, i.e. going beyond empirical and idealized habits as covered by most existing scattering databases.
EMVORADO is the native radar FO of the ICON model putting strong emphasis on model-operator-consistency and is operationally applied in the German Weather Service's radar data assimilation. It was recently extended to the simulation of polarimetric radar quantities and, like many currently available polarimetric radar FO, applies the spheroidal T-matrix model for both liquid and frozen hydrometeor classes.
Comparing to radar observations, deficiencies of simulated polarimetric variables have been identified in the melting layers as well as in the dendritic growth and aggregation layer and could be attributed to ICON’s 2-moment microphysics and to assumptions in EMVORADO. In particular, the differential polarimetric signature of snowflakes was found difficult to reproduce when approximating ice hydrometeors as homogeneous effective-medium spheroids in the T-Matrix approach. This is in line with, e.g., Schrom and Kumjian (2018), who found the detailed structure of particles to be essential for the characterization of their scattering properties, especially for polarimetric quantities.
Hence, methods capable of capturing the detailed snow particle shapes are needed in order to accurately simulate the polarimetric radar signatures. However, the level of detail of the ice particle description used in operational weather models is not sufficient to fully characterize the scattering simulations, and the range of properties that are found for natural snowflakes are not covered by the few available databases of realistic ice and snow polarimetric scattering data.
Our key objective is to provide a more realistic polarimetric radar FO, which can also be more consistently linked with microphysical parameters used in numerical models. The core component of our project is the extension of the current T-Matrix based scattering tables in EMVORADO with DDA simulations of frozen hydrometeors with realistic particle properties. Here we present the results from the first steps, where a DDA-based scattering module is implemented into EMVORADO with primary focus on the consistency between the microphysical assumptions made in the ICON model and in the forward simulations. This allows direct comparison with the legacy T-matrix simulations, hence an immediate assessment of the impact of the scattering model. For later steps, we plan to model the particle structures using a combination of a Lagrangian super-particle model and a snow particle model that covers aggregation and riming, i.e. going beyond empirical and idealized habits as covered by most existing scattering databases.

