Observational Studies of how Atmospheric Vertical Motions Influence Frozen Hydrometeor Fallspeed

The vertical velocity of hydrometeors is one of the most important parameters to get right in weather and climate models. Many fallspeed parameterizations have been developed as a function of hydrometeor habit, but in all cases, the measurements or theory that has been used as their basis has either constructed or assumed conditions of still air. But clearly snow swirls as it gets caught up in local eddies, yet the importance of this has been almost entirely neglected in the atmospheric sciences literature. There is however guidance from the engineering literature. Theoretical and numerical modeling calculations suggest particles can become concentrated in the downward flow of turbulent eddies, and this accelerates their mean settling velocity by up to 50% relative to the particle terminal fallspeed. This phenomenon of preferential acceleration has been shown numerically to occur when the Stokes number is nearly equal to unity, where the Stokes number is the ratio of the timescale for adjustment to the terminal fallspeed in still air to the Kolmogorov eddy turnover time at the microscale of turbulence. Indeed, snowfall composed of aggregate flakes is associated with more moderate turbulence and fallspeeds, in which case the Stokes number might plausibly near unity, allowing for aggregate acceleration. The effect of turbulence on precipitation fallspeeds is being tested using recent measurements from the Multi-Angle Snowflake Camera, a new instrument that takes high resolution photographs of hydrometeors in freefall while simultaneously measuring their fallspeed. In situ MASC measurements show a remarkable lack of correlation between particle size or shape and fallspeed. While the measured range of fallspeeds measured spans more than two orders of magnitude, there appears to be a nearly equal preference for particles to fall at about 1 m/s independent of whether the particles are compact graupel or aggregate flakes, or whether the particles are small or large. If snowflake aggregates are accelerated by turbulence, then snowflakes and graupel might actually be expected to end up with fallspeed distributions that are more similar than would be anticipated from their terminal velocities in still-air. To evaluate this problem further, fallspeed relationships are currently being compared with measurements of turbulent winds, and for MASC measurements obtained simultaneously in both sheltered and unsheltered conditions. Images of snowflakes captured by the Multi Angle Snowflake Camera. Each snowflake has an associated measurement of fallspeed