Thursday, 31 August 2023: 8:30 AM
Great Lakes BC (Hyatt Regency Minneapolis)
The current US operational weather radar network has dual-polarization capability, which provides more essential information on precipitation microphysics. Toward the goal of improving short-term severe weather forecasts, assimilating polarimetric radar data with accurate and efficient forward operators for polarimetric radar variables is required. Zhang et al. (2021) developed a set of parameterized forward operators for the above purpose. In this study, we plan to further evaluate and improve the accuracy of the parameterized operators.
First, we find that although the operators produce an unreasonably low melting layer and a positive bias for reflectivity in stratiform regions compared with the observations, they can provide reasonable simulated radar variables in convective regions. Examination of the contributions of different hydrometeor species indicates that the source of the discrepancy between the simulations and radar observations likely originates from an overestimation of the mean diameter of melting snow, graupel, and hail. Since most microphysical schemes do not predict the number concentration of wet species, current polarimetric radar operators based on two-moment microphysical schemes usually use the number concentration of the dry species to calculate the radar contribution of the melting species. The number concentration of dry snow decreases rapidly as snow falls past the melting layer, leading to an aberrant increase in wet snow mean diameter. This inconsistency is even more evident in the snow-dominated stratiform regions compared to convective regions. We introduce a way to calculate the wet species number concentration from these of dry species and rain. Two severe convective cases are evaluated to assess the impact of this new wet number concentration formulation on the simulation of the polarimetric radar variables using three double-moment microphysics schemes (i.e., Milbrant, Morrison, and NSSL). Compared with the operators that only use dry number concentration, the modified operators obtain radar variable simulations that are more reasonable and closer to radar observations. This improvement is also applicable to other existing polarimetric radar operators.
First, we find that although the operators produce an unreasonably low melting layer and a positive bias for reflectivity in stratiform regions compared with the observations, they can provide reasonable simulated radar variables in convective regions. Examination of the contributions of different hydrometeor species indicates that the source of the discrepancy between the simulations and radar observations likely originates from an overestimation of the mean diameter of melting snow, graupel, and hail. Since most microphysical schemes do not predict the number concentration of wet species, current polarimetric radar operators based on two-moment microphysical schemes usually use the number concentration of the dry species to calculate the radar contribution of the melting species. The number concentration of dry snow decreases rapidly as snow falls past the melting layer, leading to an aberrant increase in wet snow mean diameter. This inconsistency is even more evident in the snow-dominated stratiform regions compared to convective regions. We introduce a way to calculate the wet species number concentration from these of dry species and rain. Two severe convective cases are evaluated to assess the impact of this new wet number concentration formulation on the simulation of the polarimetric radar variables using three double-moment microphysics schemes (i.e., Milbrant, Morrison, and NSSL). Compared with the operators that only use dry number concentration, the modified operators obtain radar variable simulations that are more reasonable and closer to radar observations. This improvement is also applicable to other existing polarimetric radar operators.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner