Effects of Air Turbulence on Snowfall

A predictive understanding of snow settling is necessary for reliably forecasting snowfall. Laboratory and field measurements indicate that air turbulence can significantly enhance the settling velocity of inertial particles in general, and snowflakes in particular. We test whether this knowledge can improve the accuracy of numerical weather prediction tools. The Predicted Particle Properties (P3) cloud microphysics scheme is employed. Laboratory data obtained in a zero-mean flow turbulence chamber are incorporated into the model, and simulations are run with and without the influence of turbulence on the settling of snow crystals for a wintertime cyclone over the Mountain West of the United States. Contrary to expectations, the average precipitation decreases when including the enhancement in snowfall speed due to turbulence, the largest decreases coming from regions of heavier snowfall. Because the baseline simulation overestimates snowfall compared to observations, such decrease results in an improved forecast. Two mechanisms are explored to explain the resulting dichotomy: reduced cloud depths and enhanced entrainment in the simulations with. Both mechanisms act to limit the total condensate in the clouds, thus reducing the amount of precipitation that is generated.