Parameterization of Melting Snow for Bulk Cloud Microphysics Schemes

Wet snow poses substantial threats to infrastructure during winter. However, bulk cloud microphysics schemes used in current operational weather prediction models cannot explicitly represent wet snow. This study modifies the Milbrandt–Yau (MY) microphysics scheme of the Weather Research and Forecasting Model to explicitly simulate melting snow. The modification introduces an additional prognostic variable for the snow liquid mixing ratio, representing the liquid component of melting snow, along with its source and sink terms. Furthermore, the formulation of fall speed for melting snow and the assumptions regarding meltwater distribution are designed to be consistent with observations. The modified scheme was evaluated during a heavy wet snow accretion event that occurred in Hokkaido, Japan, on 22–23 December 2022. Two numerical simulations with a horizontal resolution of 1.667 km, using the modified and original MY schemes, respectively, were compared with observations from the Kushiro test power line in eastern Hokkaido. The results showed slightly colder surface temperatures in the modified scheme due to meltwater evaporation. Disdrometer observations indicated that the precipitation particles included melting snow with low snow liquid fractions, as well as rain and snow, which were reasonably well reproduced by the modified scheme. The analysis of source–sink terms revealed that the partial melting of snowflakes along with refreezing and evaporation of meltwater are dominant processes governing the tendency equation. Sensitivity experiments further highlighted the influence of fall velocity formulation on precipitation characteristics. Overall, the modified scheme effectively represents melting snow, demonstrating its potential for improving the prediction of wet snow events.