Case Studies of Low‐Visibility Forecasting in Falling Snow With WRF Model

Accurate low-visibility forecasts in falling snow are critical to the safety and efficiency of air traffic. The Weather and Research Forecast (WRF) model successfully captured two unusual snowstorms occurred in Urumqi. On this basis, the quality of 15 parameterizations for predicting visibility in snow is evaluated, using both observations and forecasts of the meteorological variables from WRF model. The parameterizations are mainly based on the relations between the extinction efficient () or visibility (Vis) and the snowfall rate (S). Comprehensive evaluations show that most of these parameterizations (13 of 15) are skillful as well as having the ability to predict low visibilities to some extent. Among them, the parameterization performs the best, followed by the approach of Stoelinga and Warner. It is also found that the visibility forecasts based on the observations always have considerably higher quality than the visibility forecasts from WRF model. The results suggest that more than one parameterization is promising if the WRF model is able to provide accurate predictions of the relevant meteorological variables. Furthermore, the forecast accuracy of low visibilities strongly depends on the accurate predictions of snowfall rate of greater than or equal to 1.0mm h(-1). Plain Language Summary In previous studies, visibility parameterizations as a function of snowfall rate or hydrometer mass concentration have been used in various operating systems. However, the knowledge of how well those parameterizations perform in predicting low visibilities based on WRF model is still lacking, so the topic is something worthy of exploring.