Comparison of Different Snowflake Models and In-situ Measurements in Finland During the Winter of 2014

The availability of higher-frequency radars both on-ground and in space provides an opportunity to study the microphysical properties of snow formation and precipitation in more detail than before. Complexity of natural snowflakes is a challenge when making retrievals of cloud and precipitation properties using remote-sensing measurements. In radar volumes, the largest particles tend to dominate the backscattering signals, which means that particle shape becomes an important factor in the retrievals. However, the uncertainty in the single-particle properties can be constrained by in-situ measurements using particle imagers on ground and on aircrafts. Also, since the link between physical and scattering properties is fairly well understood for snow particles, it is crucial to know the assumptions and choices that have been made when generating the articifial snowflake shapes and how much these can affect the scattering properties.

In this study, we are comparing various snowflake models that are publicly available to particle-imager measurements that have been obtained in the Hyytiala forestry station in Finland during the winter of 2014. Our goal is to test, how much certain physical properties, such as fall speeds, area ratios, aspect ratios, and masses vary between the models and how well they can cover the measurement regime. We also check how sensitive certain scattering properties are to these differences. It should be noted that the usability of snowflake models should be tested in different environments and tuned to those in order to provide a more reliable forward method.