Evaluation of Simulated Drop Size Distributions and Microphysical Processes Using Polarimetric Radar Observations for Landfalling Typhoon Matmo (2014)

Abstract

The Morrison (Morr), Thompson (Thom), Thompson aerosol‐aware (ThomA), and WRF double‐moment 6‐class (WDM6) microphysics schemes within the WRF model are used to simulate landfalling Typhoon Matmo (2014) and evaluated against polarimetric radar observations in terms of raindrop size distribution (RSD) and microphysical processes. Focus is placed on the inner rainband convection. Only ThomA is able to reproduce observed RSD characteristics having even smaller mean raindrop sizes and larger number concentrations than those of typical maritime convection, when measured in terms of mass‐weighted mean diameter and normalized RSD intercept parameter. The diagnoses of microphysical transfer terms show that warm rain processes are dominant in all experiments. The accretion and autoconversion processes dominate the production of rainwater content and raindrop number concentration, respectively. Examinations of vertical profiles of hydrometeor masses and number concentrations as well as rainwater‐related microphysical transfer rates suggest that raindrop total number concentration NTr near the freezing level, mainly produced by the autoconversion process, plays an important role in determining the near‐surface RSD characteristics. Morr and Thom assume fixed values of cloud droplet number concentration NTc that are much larger than those predicted by ThomA, leading to much smaller NTr near the freezing level produced by the autoconversion process. Sensitivity experiments using much reduced NTc substantially increase predicted NTr, mainly through invigoration of autoconversion. These results indicate that adequately predicting NTc is very important for capturing the right RSD characteristics for landfalling typhoons over East China where Typhoon Matmo is one of the representative examples.

Evaluating Simulated Raindrop Size Distributions and Ice Microphysical Processes With Polarimetric Radar Observations in a Meiyu Front Event Over Eastern China

“…The precipitation types (convection or stratiform) are identified using a robust method based on Steiner et al. (1995), which classify the convective and stratiform areas from the sharpness and intensity variations of the radar reflectivity values at 2‐km AGL height (Wang et al., 2016; Wen et al., 2018, 2020). Radar reflectivity from LSRD shows that the horizontal structure of the Meiyu system seems like the leading‐line trailing‐stratiform (TS) convective system type during its mature stage (Parker & Johnson, 2000).…”

“…Previous studies show that simulated radar variables from different types of precipitation systems (e.g., tropical cyclone, tropical maritime convection, and midlatitude continental storm) have similar statistical characteristics if the same microphysics scheme is used. For example, the simulated DR Z E of some specific microphysics schemes are much higher than radar observations near ground for all types of convective systems (Brown et al, 2016;Putnam et al, 2016;Wang et al, 2020). The biased values of these simulated polarimetric radar variables indicate the possible discrepancy of DSD distributions between model simulations and reality.…”

“…The simulated forward-and backward-scattering amplitudes of different hydrometeor species at each grid point were calculated independently based on T-matrix scattering and their particle size distributions, as described in Johnson et al (2015). The settings of particle shape and orientation are the same as Johnson et al (2015) and Wang et al (2020). For example, the axis ratio of raindrop is drawn from Brandes et al (2002) and rain canting angle is set to 0°.…”

“…Number concentration and size information of specific hydrometeors can be inferred from additional polarimetric radar variables such as differential reflectivity (; Seliga & Bringi, 1976). Recently, radar‐observed three‐dimensional microphysical structures have also been used to evaluate the simulated microphysical features in models (Brown et al., 2016, 2017; M. Han et al., 2013; Johnson et al., 2015; Matsui et al., 2020; Putnam et al., 2016; Wang et al., 2020). Simulated radar variables for certain microphysics schemes can be calculated using the so‐called polarimetric radar simulator and compared with polarimetric radar observations (Jung et al., 2008, 2010; Matsui et al., 2019).…”

Hurricane Dorian Outer Rain Band Observations and 1D Particle Model Simulations: A Case Study

“…In a subsequent article [9], it was shown that in stratiform rain from both hurricane rain bands and in MCSs, the treatment of melting snow to rain needed to be modified to predict the smaller values of Z DR consistent with observations. A similar explicit simulation of Typhoon Matmo comparing four different bulk microphysical schemes but focused on the inner rain band convection were used by [10]. They found that DSDs were dominated by warm rain processes, but the distributions were characterized by smaller values of mass-weighted mean diameter and much larger normalized intercept parameter relative to non-typhoon maritime convective rain [11,12].…”