Uncertainty Quantification for Ice-Characteristics Parameters in WDM6

The characteristics of hydrometeors, pre-defined in cloud microphysics parameterizations, can alter the simulated convection, cloud morphology, surface precipitation, and so on. The ice microphysics processes in the Weather Research and Forecasting (WRF) Double-Moment (WDM) bulk-type cloud microphysics parameterizations adopts the fall velocity-diameter and mass-diameter relationships suggested by Heymsfield and Iaquinta (2000, HI00 hereafter), and mean mass-weighted terminal velocity-mixing ratio relationship suggested by Heymsfield and Donner (1990, HD90 hereafter). In our study, five parameters defining the cloud ice characteristics were sampled through the Latin Hypercube Sampling (LHS) method. We created 50 samples, in which parameters defining cloud ice characteristics vary within the range proposed by HI00 and HD90. To figure out the impact of the parameters, the numerical experiments were conducted for three different winter precipitation types (convection, low pressure, and orographic case) over Korean peninsula using the WDM 6-class (WDM6) cloud microphysics scheme. The control experiment using the original parameters in the WDM6 overestimates the precipitation regardless of the winter precipitation types. From the inspection of the spatial distribution of the correlation coefficient between each parameter and precipitation for the convection case, we can conclude that the parameters defining the relationship between the mass and diameter of cloud ice have a significant impact on the simulated precipitation distribution and intensity over the land. For the case of low-pressure, the parameters defining the relationship between the fall velocity and diameter of cloud ice are analyzed as the important parameters affecting the simulated precipitation. For the orographic case, it has been analyzed that the correlation coefficients between each parameter and precipitation show the opposite sign according to the locations, that is windward or leeward sides. The revealed precipitation bias of the existing model was improved by adjusting parameters, identified as the important ones for each winter precipitation type.

Acknowledgement: This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. 2021R1A4A1032646).