Snow variability in Oklahoma and Colorado characterized by 2D-Video Disdrometer and dual-polarization radar measurements

Accurate observations of snowfall are difficult to achieve. As opposed to radar rainfall estimation uncertainty caused by the variability of raindrop size distributions, radar measurements of snow are even more challenging due to variability of snow density. Thus Z – S relations are even more variable than in case of rain. Information about the snow size distribution variability and variation in snow density can be obtained from 2D video disdrometer measurements. In this paper, the impact of snow size distribution and density variability on the Z – S relations is examined. The influence of differential reflectivity ZDR on snowfall quantification is investigated using extensive data set of snow disdrometer measurements from Oklahoma and Colorado combined with the polarimetric radar measurements. It has been found that the relation Z = aS^b has to be parameterized by the intercept parameter of exponential size distribution, N_0s. We introduce a new algorithm for 2D video distrometer particle (snow) matching, which mitigates the effects of mismatching by previous approaches. The influence of individual big snowflakes on snow water equivalent measurement and reflectivity factor calculation is addressed. Combined snow measurements with a 2D video disdrometer and dual-polarization S-band KOUN and KGJX radars were performed in central Oklahoma from 2009 to 2015 winter seasons and west Colorado during the winter of 2012/2013.