Using Airborne Measurements to Evaluate Forecasts of Freezing Drizzle Aloft: Results from the WINTRE-MIX Field Campaign

Hazardous conditions resulting from icing associated with the accretion of supercooled large droplets (SLD) can significantly impact aviation activities. Complexities exist in forecasting SLD events due to the limitations of numerical weather prediction models. Observations collected as part of the Winter Precipitation Type Research Multiscale Experiment (WINTRE-MIX) provide a detailed dataset for evaluation of numerical forecasts of precipitation type, including SLD aloft. In particular, data from research flights orchestrated with the National Research Council of Canada Convair-580 aircraft provide observations from a suite of in-situ sensors and a profiling cloud radar. This study uses this data to examine a WINTRE-MIX intensive observing period on 07–08 March 2022 that coincided with a warm frontal passage through northern New York and southern Quebec. As the Convair ascended, widespread freezing drizzle (FZDZ) associated with SLD icing was observed with abnormally cool cloud-top temperatures as low as -15°C.

This study evaluates the operational High-Resolution Rapid Refresh (HRRR) model forecasts of FZDZ against Convair-580 observations. Additional Weather Research and Forecasting (WRF) model simulations, including sensitivity experiments, were conducted and examined to isolate and diagnose the sources of biases documented in the HRRR forecasts.

HRRR and the control (CTRL) WRF simulation disagree with observations regarding the vertical structure of clouds and precipitation. In-situ radar observations taken by the W-band radar depict a lower-layer cloud structure. Conversely, HRRR and CTRL simulate snow and ice hydrometeors throughout the vertical column failing to reproduce the FZDZ and SLD icing conditions observed at flight level. The presence of snow hydrometeors throughout the vertical column suggests that seeding from the upper-level cloud inhibits collision-coalescence. Frozen hydrometeors will serve as ice nuclei and can foster the evaporation of supercooled liquid water through the Bergeron-Wegener-Findeisen process. A sensitivity experiment is performed by decreasing the relative humidity of the initial conditions aloft to explore the influence of seeding from the upper-layer cloud. This results in a reduction of snow mixing ratios and a better representation of rain mixing ratios at locations of observed icing, suggesting that biases in the HRRR forecast are due to unrealistic seeding from ice aloft. Results from an additional sensitivity experiment will be discussed that explores how modifications to the snow terminal fall speed formula alter how seeding is simulated aloft.