The Use of Polarimetric Radar Data for Informing Numerical Weather Prediction Models
The explicit prediction of convective storms using storm-scale models has recently become feasible and is beginning to be incorporated operationally. Radar data is a crucial source of information about the microphysical and kinematic properties of convection at the storm-scale. Whereas assimilation studies have primarily focused on radial velocity and reflectivity, much less has been done to investigate how dual-polarization radar data, and the enhanced microphysical information it offers, may inform storm-scale models.
This work employs a suite of microphysical and numerical weather prediction models, coupled to a polarimetric radar operator, to study how dual-polarization radar data may be used to inform storm-scale models. The efficacy of reflectivity-based retrievals of hydrometeor mixing ratios in rain and rain/hail mixtures, and the potential benefits of dual-polarization radar, are briefly assessed. However, given the polarimetric “fingerprints” associated with distinct microphysical processes, examining the potential for these signatures to provide information about diabatic processes is a natural endeavor. Thus, the remainder of the work focuses on the potential for utilizing dual-polarization radar data for deriving information about diabatic cooling and heating rates in the melting layer and in deep moist convective updrafts, respectively.
A one-dimensional model of the melting layer is presented and used to study the impact of the environment on polarimetric brightband characteristics, the potential for retrieving the maximum cooling rate within the melting layer, and the potential microphysical causes of “sagging” brightband signatures. Predicated on a connection between differential reflectivity (ZDR) column characteristics and the latent heating rate within convective updrafts, a novel method for assimilating ZDR columns using a cloud analysis is developed, with results from two real data cases (19 May 2013 in central Oklahoma and 25 May 2016 in northeast Kansas) demonstrating positive impacts compared to reflectivity-based cloud analysis techniques. These results motivate future research into the potential for polarimetric thermodynamic retrievals and their assimilation into numerical weather prediction models.