The performances of bulk microphysical schemes were evaluated through comparisons with observations, including a new volume scanning video disdrometer targeting one of a typical snowfall event in Hokkaido, Japan. For the evaluation, downscaling experiments using three bulk microphysical schemes were conducted: a two-moment bulk scheme, a one-moment bulk scheme, and an improved version of the one-moment bulk scheme coupled with an identical dynamical core. The two-moment scheme successfully simulated the measured relationship between particle size and terminal velocity distribution (PSVD). On the other hand, the one-moment scheme overestimated the graupel frequency, its terminal velocity, and underestimated the particle diameter. The improved version of the one-moment bulk scheme reduced the overestimation of the terminal velocity of hydrometeors, but still overestimated the graupel frequency. We improved the overestimation of terminal velocity and the frequency of graupel in the one-moment scheme by modifying the assumption of the velocity–diameter relationship and the intercept parameter of the size distribution of graupel based on the results of the new disdrometeor. The observation of the new disdrometer would give us hints to improve the microphysics schemes in snowfall cases.