How dual-polarization radar observations can be used to verify model representation of secondary ice

In this paper it is discussed how dual-polarization radar observations can be used to verify model representations of secondary ice production. An event where enhanced specific differential phase, K_dp, signatures in snow occur at the altitudes where temperatures lie in the range between −8 and −3°C is investigated. By combining radar and surface-based precipitation observations it is shown that these dual-polarization radar signatures are most likely caused by ice with concentrations exceeding those expected from primary ice parameterizations. It is also shown that the newly formed ice particles readily aggregate, which may explain why K_dp values seem to be capped at 0.2–0.3°/km for a C band radar. For the event of interest, multiple high-resolution (1 km) Weather Research and Forecasting (WRF) model simulations are conducted. When the default versions of the Morrison microphysics schemes were used, the simulated number concentration of frozen hydrometeors is much lower than observed and the simulated ice particle concentrations are comparable with values expected from primary ice parameterizations. Higher ice concentrations, which exceed values expected from primary ice parameterizations, were simulated when ad hoc thresholds for rain and cloud water mixing ratio in the Hallett-Mossop part of the Morrison scheme were removed. These results suggest that the parameterization of secondary ice production in operational weather prediction models needs to be revisited and that dual-polarization radar observations, in conjunction with ancillary observations, can be used to verify them.