Next Article in Journal
An Image Fusion Algorithm for Sustainable Development Goals Satellite-1 Night-Time Light Images Based on Optimized Image Stretching and Dual-Domain Fusion
Previous Article in Journal
Structural Complexity Significantly Impacts Canopy Reflectance Simulations as Revealed from Reconstructed and Sentinel-2-Monitored Scenes in a Temperate Deciduous Forest
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Remote Sensing
Volume 16
Issue 22
10.3390/rs16224297
remotesensing-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

A Numerical Simulation of Convective Systems in Southeast China: A Comparison of Microphysical Schemes and Sensitivity Experiments on Raindrop Break and Evaporation

by
Zhaoqing Cheng
1,2 and
Xiaoli Liu
1,2,*
1
China Meteorological Administration Aerosol-Cloud and Precipitation Key Laboratory, Nanjing University of Information Science and Technology, Nanjing 210044, China
2
College of Atmospheric Physics, Nanjing University of Information Science and Technology, Nanjing 210044, China
*
Author to whom correspondence should be addressed.
Remote Sens. 2024, 16(22), 4297; https://doi.org/10.3390/rs16224297
Submission received: 13 August 2024 / Revised: 7 November 2024 / Accepted: 14 November 2024 / Published: 18 November 2024
Browse Figures
Versions Notes

Abstract

:
This study employed version 4.2.2 of the Weather Research and Forecasting (WRF) model for this simulation and applied two microphysics schemes, the Thompson scheme (THOM) and Milbrandt–Yau scheme (MY)—which are widely used in convective simulations—to simulate a mesoscale severe convective precipitation event that occurred in southeastern China on 8 May 2017. The simulations were then compared with dual-polarization radar observations using a radar simulator. It was found that THOM produced vertical structures of radar reflectivity (ZH) closer to radar observations and accumulated precipitation more consistent with ground-based observations. However, both schemes overestimated specific differential phase (KDP) and differential reflectivity (ZDR) below the 0 °C level. Further analysis indicated that THOM produced more rain with larger raindrop sizes below the 0 °C level. Due to the close connection between raindrop breakup, evaporation rate, and raindrop size, sensitivity experiments on the breakup threshold (Db) and the evaporation efficiency (EE) of the THOM scheme were carried out. It was found that adjusting Db significantly changed the simulated raindrop size distribution and had a certain impact on the strength of cold pool; whereas modifying EE not only significantly changed the intensity and scope of the cold pool, but also had great effect on the raindrop size distribution. At the same time, comparison with dual-polarization radar observations indicated that reducing the Db can improve the model’s simulation of polarimetric radar variables such as ZDR. This paper specifically analyzes a severe convective precipitation event in the Guangdong region under weak synoptic conditions and a humid climate. It demonstrates the feasibility of a method based on polarimetric radar data that modifies Db of THOM to achieve better consistency between simulations and observations in southeast China. Since the microphysical processes of different Mesoscale Convective Systems (MCSs) vary, the generalizability of this study needs to be validated through more cases and regions in the future.

1. Introduction

The Guangdong province is located in southeastern China and it has a complex terrain. It is also one of the regions with the highest rainfall in the world, with annual precipitation exceeding 2000 mm [1]. Approximately 40–50% of the yearly rainfall occurs during the pre-summer rainy season (PSRS) from April to June [1]. Extreme precipitation events cause severe disasters every year, posing serious threats to life safety and social stability [2,3]. Therefore, high-precision and high-resolution assessments of the timing and intensity of precipitation has become an urgent issue that needs to be addressed [4,5,6]. Currently, Numerical Weather Prediction (NWP) models remain the primary tool for forecasting convective weather. However, due to the inadequate understanding of cloud microphysical processes [7,8] and the imprecision in the parameterization of the microphysical schemes [9,10,11], there is still significant uncertainty in the microphysical schemes within the models when simulating cloud microphysical processes [12]. This uncertainty leads to poor performance of numerical models in simulating the accumulated precipitation amount and the location of heavy rainfall in urban areas [13,14].
During the development of convection, a relatively colder and denser air mass may form at the base of the convective system, which is referred to as a cold pool [15]. The intensity of the cold pool can influence the organization, dynamic structure, and thermodynamic structure of the ascending/descending airflows associated with convective systems, thereby affecting the formation and development of convection [9,16,17].
As the cold pool forms and expands, the cold air replaces the warmer air, causing the warm air to be lifted, which in turn leads to the formation of new clouds [18]. Furthermore, the downdrafts associated with the cold pool enhance the sensible and latent heat fluxes, thereby altering the thermodynamic properties and moisture structure below the clouds [19]. Convective cold pools play a significant role in various aspects of MCSs, including the maintenance of squall lines and the transition of tropical convection from weak to strong stages [20,21]. Additionally, it has also been shown that the self-sustaining mechanisms of cold pools can lead to the prolonged maintenance of rainbands, resulting in the persistence of heavy rainfall [22]. In the process of convective precipitation, raindrops undergo various microphysical processes such as coalescence, breakup, and evaporation as they fall from the cloud to the ground. Among these, the breakup of raindrops directly affects the raindrop size distribution, which in turn affects the evaporation rate, thus affecting the evolution of the cold pool [23,24,25]. The intense latent heat fluxes are brought by the evaporation of rainwater, playing a significant role in the formation of cold pools [19].
Studies have demonstrated that microphysical processes are crucial for accurately forecasting heavy rainfall in southeastern China. Qian et al. [17] pointed out that a lower rain evaporation rate and the resulting weaker cold pool were the reasons for the failure of simulating the convective system in southeastern China. By use of three microphysical schemes, Zhou et al. [26] pointed out that all simulations produced weaker cold pools compared to the observations, and improving the simulated cold pool intensity and drop size distributions (DSDs) led to a significant improvement in surface precipitation. Lompar et al. [27] incorporated the first gust front pulsation parameterization scheme into the WRF, which not only improved the intensity and distribution of the cold pool but also enhanced the overall simulation performance.
With the development of dual-polarization radar detection technology, polarimetric radar observations have become widely used to compare the microphysical characteristics of heavy rainfall between observations and the NWP models [28,29]. Polarimetric radar is capable of transmitting and receiving polarized electromagnetic waves in two orthogonal directions. Because the polarization of the echo is influenced by particle properties, it can provide rich microphysical characteristic information such as the phase state, density, and shape of hydrometeors [30,31]. In addition to the radar reflectivity (ZH) for horizontal polarization, there are two other commonly used polarimetric variables. One is the differential reflectivity (ZDR), which reflects the shape and size of hydrometeors, and the other is the specific differential phase (KDP), which primarily indicates the liquid water content within the sampling volume [31,32,33]. The combination of ZH and ZDR can reveal the characteristics of drop size distributions (DSDs) as well as the “fingerprints” of microphysical processes such as evaporation, coalescence, and breakup [32,34]. This paper will compare simulation results from NWP models with polarimetric radar observations by converting model output data into polarimetric radar data by the use of a radar simulator [35,36].
Meteorologists typically use NWP models with complex physical parameterization schemes to simulate and understand the mechanisms of weather systems [37]. Compared to bin microphysics schemes, bulk microphysics schemes offer superior computational efficiency, and are extensively utilized in both scientific research and operational contexts [10]. Two-moment (2M) bulk microphysical schemes generally forecast two parameters—the intercept parameter N0 and the slope parameter λ—while one moment (1M) bulk schemes prespecify N0 [7,38]. Due to the increased flexibility in parameter prediction, 2M schemes often produce better simulation results when evaluated against observational data than 1M schemes [8,11,39]. Therefore, this study employs two microphysics 2M schemes, the Thompson scheme (THOM) and Milbrandt–Yau scheme (MY), which are widely used in convective simulations.
The focus of this study is comparing polarimetric radar data with simulation results, evaluating and improving the parameterization scheme’s representation of macroscopic and microscopic features of the convective system. The structure of this paper is as follows: Section 2 introduces the radar data, simulation settings, and the radar simulator used in this study. Section 3 compares and analyzes the composite radar reflectivity, vertical cross-sections of radar polarimetric variables, and 18-hour accumulated precipitation from observations and simulations, and continues to compare and analyze different aspects of the simulation, including the hydrometeor mixing ratios, hydrometeor source and sink terms, cold pool, and the raindrop mass-weighted diameter (Dm,r) distribution. Section 4 modifies the parameters related to raindrop breakup and evaporation and assesses the results of these modifications. Finally, Section 5 discusses the results, and Section 6 draws conclusions and proposes possible measures for further improvement.

2. Materials and Methods

2.1. Radar Data

The radar data are derived from the Guangzhou S-band dual-polarization radar (Z9200; 23.00°N, 113.36°E), which was put into operation in May 2016. The radar operates in a dual-transmit dual-receive mode with a wavelength of 10 cm, an azimuthal resolution of 1°, a radial resolution of 250 m, and an observation radius of 230 km. This radar can provide a series of dual-polarization variables, including reflectivity (ZH), differential reflectivity factor (ZDR), correlation coefficient (CC), specific differential phase (KDP), and differential phase (φDP), which can characterize the microphysical structure of precipitation. The volume scan time of the dual-polarization radar is approximately 6 min, using the VCP21D scanning pattern, which includes nine elevation angles: 0.5°, 1.5°, 2.4°, 3.3°, 4.3°, 6.0°, 9.9°, 14.6°, and 19.5°.

2.2. Simulation Settings

On 8 May 2017, a significant convective precipitation event occurred in Guangdong, China, and the local development was rapid, with extreme precipitation intensity. On a spatial scale, the mature squall line was approximately 800 km long and 40 km wide, advancing at a speed of 30 to 40 km per h, sweeping across much of the Guangdong province [40]. In this study, we simulated the case using the generation mesoscale numerical model WRF and employed version 4.2.2 of the WRF-ARW. For the simulation, the THOM and MY microphysics schemes within the WRF model were selected. The simulation used two-layer nested domains with resolutions of 3 km for the outer and 1 km for the inner domain. The simulation period spanned from 18:00 UTC on 7 May 2017 to 18:00 UTC on 8 May 2017. The spin-up time of the model was set to 6 h, and the numerical simulation results following 00:00 UTC on 8 May 2017 were compared with radar observations. The initial and boundary field data were obtained from ERA5 with a resolution of 0.25° × 0.25°. The boundary layer scheme chosen was the Mellor–Yamada–Janjic (Eta) TKE scheme, and the surface layer scheme used was the Monin–Obukhov (Janjic Eta) scheme. The land-surface process scheme employed was the unified Noah land-surface model with the Rapid Radiative Transfer Model (RRTM) for the Global Climate Model (GCM) for shortwave and longwave radiation. The cumulus parameterization was turned off in two domains. This convective process occurred in Qingyuan City, Guangdong province. Therefore, two nested domains, d01 and d02, were set up, centered on Qingyuan City, with both domains consisting of 481 × 481 grid points. The setup of the simulation domain is shown in Figure 1.
This study delves into two extensively utilized 2M cloud microphysics schemes: THOM [41] and MY [42]. As shown in Table 1, THOM predicts the mixing ratios (q) for cloud water, rain, cloud ice, snow, and graupel, as well as the number concentrations (N) specifically for rain and cloud ice. Conversely, MY forecasts q and N for a broader spectrum of hydrometeors, including cloud water, rain, cloud ice, snow, graupel, and hail.

2.3. Radar Simulator

This paper utilizes the Polarimetric Radar Simulator (CAPS-PRS) developed by the Center for Analysis and Prediction of Storms (CAPS) at the University of Oklahoma, USA for non-hydrostatic weather forecasting models with explicit microphysics schemes. It includes calculations of reflectivity for horizontal and vertical polarizations, KDP, ZDR, ρhv. This simulator is capable of simulating polarimetric radar measurements in the weather radar frequency band and can take forecast variables simulated by NWP models using single-moment, two-moment, and three-moment microphysics schemes as inputs [43].

3. Results Analysis

3.1. Comparison Analysis of Observational and Simulated Results

3.1.1. Assessment of Convective System Simulations

The intense convective precipitation process in the Guangdong region was divided into three stages: development, maturity, and dissipation. The radar observations of composite reflectivity clearly illustrate the evolution of the MCS (Figure 2(a1–c1)). The convective cloud began to form outside of the Guangdong region, and then rapidly developed into a convective band by 0600 UTC, moving southeastward and continuing to develop (Figure 2(a1)). During the development stage (0600 UTC), both the convective and stratiform cloud areas expanded rapidly. By the time the system reached its maturity stage (0854 UTC), the front was characterized by a high and strong convective band associated with heavy rainfall, marking the peak of convective activity (the time with the highest hourly precipitation amount) (Figure 2(b1)). Other notable features included a transition zone with lower radar reflectivity directly following the convective zone. As the system entered the dissipation phase (1200 UTC), the convective cloud band began to weaken, and mesoscale organization gradually diminished (Figure 2(c1)).
Overall, by comparing the simulated composite reflectivity with the observation, both schemes simulate the development trends, intensities, and spatial positions of the convective process at different stages well. The three stages to the simulations by THOM correspond to 0630 UTC (Figure 2(a2)), 0930 UTC (Figure 2(b2)), and 1230 UTC (Figure 2(c2)), as well as to the simulations by MY at 0630 UTC (Figure 2(a3)), 0830 UTC (Figure 2(b3)), and 1200 UTC (Figure 2(c3)). The convective development simulated by THOM is slower compared to the observations, causing a delay in the timing of the mature and dissipation stages. In contrast, MY simulated faster convective development, resulting in an earlier timing for the mature stage. The reflectivity intensity in the convective region is stronger than the observed values, and the reflectivity intensity simulated by the MY scheme is weaker than that of the THOM scheme.

3.1.2. Analysis of Radar Polarimetric Parameters

Due to the similar vertical structures of observed and simulated characteristics in three stages of this MCS, this section will only focus on the mature stage. Figure 3 presents the vertical cross-sections of radar polarimetric parameters during the mature stage, with the transects indicated by the black lines shown in Figure 2. These transects are nearly perpendicular to the convective system, and their positions were selected to simultaneously observe the vertical structural characteristics of the convective region, stratiform region, and transition zone during the mature stage.
As the convective system enters the mature stage (Figure 3(a1)), the radar echo coverage expands significantly, the echo tops surpass 16 km in altitude, while the strong echo zones (ZH > 40 dBZ) ascend to around 9 km (Figure 3(a1)). THOM simulates radar echo tops reaching above 15 km, with the strong echo zones rising to around 10 km. Below the 0 °C layer, the THOM-simulated ZH slightly exceeds the observed values, with peak values exceeding 55 dBZ (Figure 3(b1)). In contrast, the MY scheme produces echo tops approximately at 14 km, with the strong echo zones ascending close to the −40 °C isotherm. Below the 0 °C layer, ZH reaches above 50 dBZ, and notably, there remains a vast area of strong echoes in the upper cloud area (Figure 3(c1)). Both THOM and MY generate extensive stratiform cloud regions and successfully simulate the transition zone. However, ZH in the stratiform areas generated by both schemes is lower than that which is observed by radar.
In the vertical cross-sections of the radar observations for ZH and ZDR (Figure 3(a1,a3)), the stratiform cloud region behind is separated from the forward convective region by a weak echo transition zone with lower ZH and ZDR values. The ZDR in the stratiform and convective regions shows a significant increase with decreasing altitude below the 0 °C level, with the ZDR maxima (~1.5–2.5 dB) existing below the 4 km altitude (Figure 3(a3)). In the convective region, KDP values exceeding 0.5°/km are observed in the upper part of the columnar ZDR, which may be due to the presence of supercooled raindrops adjacent to the updraft or hail coated with supercooled water [44,45]. The increase in KDP below the 0 °C level in the convective region is associated with high ZH (>45 dBZ). This may be related to ice particles melting too quickly below the 0 °C level, which need to be further analyzed.
It is noted that KDP primarily represents the content of liquid water per unit volume [26]. Although the structures of the observed and the simulated KDP are similar, both schemes overestimated the KDP values below the 0 °C layer in the convective region (Figure 3(a2–c2)). This situation is different from the simulations under North American climatic conditions, where both microphysical schemes produced weaker KDP values than observed [46].
Like KDP, both schemes simulate higher ZDR values below the 0 °C layer than what is observed (Figure 3(a3–b3)). Larger ZDR values indicate a greater difference in the horizontal and vertical dimensions of the particles, which is a significant microphysical factor for pure and large size rain [31]. This suggests that, in comparison to radar observations, the raindrop size within the convective area below the 0 °C layer is overestimated by both schemes.
It is shown that the MY’s microphysical scheme simulates stronger radar reflectivity zones compared to the observations, whereas the THOM’s simulated radar reflectivity is more closely aligned with the observations. Both schemes simulate a weak echo transition zone between the convective and stratiform regions, corresponding to the radar observations, with the ZDR and KDP simulated by both schemes being significantly greater than the observation below the 0 °C level in the convective region.

3.1.3. Comparison of the Accumulated Precipitation

Figure 4 shows the accumulated surface precipitation from the onset to the dissipation of the convective system (from 0000 to 1800 UTC, 8 May 2017). Generally, both schemes basically reproduced the distribution of surface precipitation, albeit with narrower precipitation areas compared to the ground observations. Both schemes simulated more heavy precipitation centers than observed with greater intensity and underpredicted intensity of precipitation in the southwest of the Guangzhou radar (Figure 4a–c). Considering the distribution and magnitude of several precipitation centers observed, THOM exhibited more concentrated and widespread precipitation centers in the areas north of the Guangzhou radar, leading to a better representation of precipitation (Figure 4b). The MY scheme simulated a similar precipitation pattern with respect to the THOM scheme, but with more scattered precipitation centers north of the Guangzhou radar (Figure 4c). This suggests that the inherent differences in microphysical schemes could significantly impact the precipitation patterns and rainfall rates.

3.2. Analysis of the Simulation Results

3.2.1. Vertical Distribution of Hydrometeors

Figure 5 presents the vertical profiles of the regional averaged hydrometeors mixing ratio (q) during the mature stages of the convective event. There is a significant difference in snow mixing ratio for the two schemes, which may relate to the different microphysical treatments of snow. Previous studies have noted that THOM has a broader distribution of snow due to its special treatment of snow [9,16,47], which helps THOM to better simulate convective and stratiform cloud regions [46,48,49]. It is noteworthy that the two schemes, which differ significantly in the treatment of ice-phase particles, produce similar surface precipitation (Figure 4) as well as mixing ratios of rain (Figure 5), which must be further analyzed.
For the MY scheme, it can be found that the strong radar reflectivity regions above the 0 °C level in the convective zone are primarily contributed by graupel and hail particles. Therefore, the differences in the heights of the strong radar reflectivity regions between the two schemes are due to the different vertical distribution of graupel and hail particles. Below the 0 °C level, the distribution characteristics of KDP is mainly contributed by rainwater.
The analysis in this section indicates that the differences between the two schemes, as well as the discrepancies between radar observations and simulation results, mainly arise from the variations in raindrops and graupel/hail particles. The next section will analyze the source and sink terms of raindrop particles and graupel/hail particles to elucidate the specific reasons for these differences.

3.2.2. Analysis of the Source and Sink Term for Rainwater and Graupel/Hail Particles

To compare the main microphysical processes related to rain and graupel/hail particles for the two schemes, Figure 6a,b display the average tendency of raindrops during the mature stage, while Figure 6c–e show the equivalent for graupel/hail particles. As can be inferred from Figure 6a,b, between the 0 °C level and 2 km in altitude, the raindrop evaporation process is weaker (qr_rev in the Figure 6a), and the production and growth of raindrops are mainly due to the melting of graupel and snow particles (qr_sml & qr_gml in the Figure 6a), as well as the collection of graupel and cloud water (qr_rcg & qr_rcc in the Figure 6a) by raindrops in THOM. In the same altitude range, the production and growth of raindrops in the MY scheme also rely on the melting of ice-phase particles such as graupel, hail, and snow (qr_gml & qr_hml & qr_sml in the Figure 6b), and the collection of cloud water by raindrops (qr_rcc in the Figure 6b). At the same time, the rate of hail collecting raindrops (qr_rch in the Figure 6b) and the evaporation of raindrops (qr_rev in the Figure 6b) also increase with decreasing altitude, but the source term rate of raindrops is significantly greater than the sink term. Between 2 km and the ground level, in the THOM scheme, graupel and snow particles are almost completely melted (qr_sml & qr_gml in the Figure 6a), and the rate of raindrops collecting graupel particles (qr_rcg in the Figure 6a) and the collection of cloud droplets by raindrops (qr_rcc in the Figure 6a) all decrease with altitude decreasing, while the evaporation rate of raindrops (qr_rev in the Figure 6a) first increases and then slowly decreases with decreasing altitude. The source term rate is less than the sink term rate, leading to a decrease in the rain mixing ratio with decreasing height. As a result, both schemes produced similar surface precipitation.
Similarly, in MY, between 2 km and the surface level, the main source terms for raindrops are the melting of hail and the collection of cloud water by raindrops (qr_hml & qr_rcc in the Figure 6b), while the main sink terms are the collection of hail particles by raindrops (qr_rch in the Figure 6b) and the evaporation of raindrops (qr_rev in the Figure 6b). In this case, the source term rate is also less than the sink term rate, resulting in a decrease in the rain mixing ratio with decreasing height. Consistent with Figure 5 and Figure 6, similar to the distribution trend of raindrops content with height, the KDP also shows a trend of increasing first and then decreasing with the decrease in height.
Above 8 km in altitude, the growth of THOM graupel particles is primarily due to the collection of cloud droplets (qg_gcc in the Figure 6c). The growth of MY graupel particles is mainly influenced by the combined action of collecting cloud droplets (qg_gcc in the Figure 6d) and the deposition of water vapor (qg_vvd in the Figure 6d). The growth of MY hail particles is mainly from the collection of cloud droplets (qh_hcc in the Figure 6e). Comparisons of the two schemes indicate that graupel/hail particles of MY has a significantly higher growth rate above 8 km in altitude than THOM, consistent with the more graupel/hail particles above 8 km in altitude (Figure 5) and the more intensified radar reflectivity above freezing level in MY schemes.

3.2.3. Characteristics of Cold Pool

Figure 7 shows the potential temperature at a 540 m height during the mature stage for both schemes. It is noteworthy that the spatial distributions of the cold pool simulated by the two schemes are similar, which contrasts with the usual outcome in previous studies where different microphysical schemes typically lead to distinct thermodynamic structures [17,50]. However, this does not imply that the impact of microphysical processes on the cold pool is negligible.

3.2.4. Analysis of Raindrop Mass-Weighted Diameter

In 2M microphysics schemes, the evaporation rate is closely tied to the raindrop size distribution [38,41], which is crucial for the formation of cold pools [24,39]. Zhou et al. [51] highlighted the importance of cold pool evolution for the development of bow echoes based on observational data analysis. This section will analyze the impact of Dm,r on the simulation of cold pools. As shown in Figure 8, Dm,r distribution in THOM is similar to that in MY, with the majority of raindrop diameter below the 0 °C level being between 0 and 2 mm, and a small fraction of raindrops with Dm,r larger than 2 mm. Below the 0 °C level, the ZDR in THOM and MY is greater than the observed values, indicating that THOM and MY overestimate the raindrop sizes (Figure 3(a3–c3)). Since THOM simulates the overall structure of the convective cloud evolvement and the accumulated precipitation more closely to observations, the next section will further explore the possible mechanisms and pathways to improve the simulation of Dm,r and radar polarimetric quantities in THOM.

4. Sensitivity Experiments

Previous studies have shown that the evaporation of raindrops, which can produce or intensify cold pools, have significant impacts on the lifecycle of convection [9,19]. Meanwhile, changes in the size distribution or spectrum caused by raindrop breakup may alter evaporation, thus affecting the cold pool [23,24,25]. The comparison of the simulated and the observed ZDR found that both schemes’ default settings overestimated raindrop sizes (Figure 3(a3–c3)).
It is found that the evaporation process rapidly depletes small raindrops, thereby increasing the Dm,r, whereas the breakup process, due to the transformation of large raindrops into smaller ones, tends to decrease the Dm,r [52]. From Figure 8, it can be seen that simulation results of MY exhibit a consistent increase in the Dm,r below the 0 °C level, where THOM shows a pattern of first decreasing and then increasing Dm,r below the 0 °C level.
Based on the aforementioned analysis, sensitivity experiments were conducted in THOM to specifically address the raindrop breakup and evaporation processes. For the raindrop breakup process, we modified the threshold diameter Db in the coalescence-breakup efficiency (Appendix B Formula (A1)) of the THOM scheme. Db is the threshold diameter for initiating breakup, modifying Db directly impacts the raindrop size distribution, which in turn affects the evaporation process, and for the raindrop evaporation processes, we directly modified the evaporation efficiency EE (Appendix B Formula (A2)) by multiplying it by different factors. The alteration in EE has a direct impact on the evaporation process of rainwater.
As shown in Table 2, for the raindrop breakup process, six simulations were conducted with Db set at 1.0, 1.2, 1.4, 1.6, 1.8, and 2.2 mm, respectively (THOM_BKP1000, THOM_BKP1200, THOM_BKP1400, THOM_BKP1600, THOM_BKP1800, THOM_BKP2200). For the raindrop evaporation process, similar to Qian et al. [17], five simulations were conducted by directly multiplying the evaporation efficiency EE by 0.5, 1.5, 3.0, 5.0, and 10.0, respectively (THOM_EVP0.5, THOM_EVP1.5, THOM_EVP3.0, THOM_EVP5.0, THOM_EVP10.0). For comparison, the original THOM simulation will be referred to as THOM_CTR in the following text.
Figure 9 presents a PDF of the difference in simulated Dm,r between sensitivity experiments and THOM_CTR during the mature stage. Overall, for the BKP tests, as Db decreases, the proportion of raindrops Dm,r larger than 2 mm decreases, while the proportion of smaller raindrops (Dm,r is 0–2 mm) increases, indicating an enhanced raindrop breakup process (Figure 9(a1–e1)); as Db increases, it has little impact on raindrops Dm,r larger than 2 mm (Figure 9(f1)). As for the EVP experiments, for the simulations with increased EE, due to the substantial evaporation of smaller raindrops (Dm,r is 0–1 mm), the proportion of raindrops Dm,r larger than 1mm below the freezing layer increases (Figure 9(b2–e2)); conversely, the situation with a decrease in EE is just the opposite (Figure 9(a2)). This phenomenon suggests that decreasing Db and decreasing EE can generate more smaller raindrops, while increasing EE will enhance the proportion of larger raindrops.
Figure 10(a1–f1,a2–e2) illustrate the difference of cold pools between two sets of sensitivity experiments and the THOM_CTR. As expected, the reduction of Db in the BKP experiments (enhancing breakup efficiency) led to the production of smaller raindrops (Figure 9(a1–e1)), resulting in higher rates of rain evaporation and stronger cold pools (Figure 10(a1–e1)). However, only in THOM_BKP1000, modifying the breakup parameters significantly affected the intensity and extent of the cold pool, and its particle size distribution already diverged significantly from THOM_CTR (Figure 8a). This suggests that modifying the raindrop breakup process not only improves the mass-weighted diameter distribution, but also leads to a substantial increase in cold pool intensity or extent. Regarding the EVP experiments, there were significant changes in both cold pool intensity and extent across all five simulations, particularly evident in THOM_EVP0.5 and THOM_EVP3.0 (Figure 10(a2–e2)). The transition from THOM_EVP3.0 to THOM_EVP5.0 and then to THOM_EVP10.0 showed only marginal changes in the cold pool.
As shown in Figure 5a, raindrops are primarily located below 4 km in altitude. The observation indicates that ZDR is mainly distributed between 0 and 1 dB at altitudes below 4 km (Figure 11a). Compared to THOM_CTR, where ZDR is mainly distributed between 0 and 3 dB (Figure 11b), THOM_BKP1000 shows ZDR values more closely aligned with observations (Figure 11c), concentrated between 0 and 1 dB below 4 km. Therefore, we believe that reducing Db can better improve the simulation results.

5. Discussion

In recent years, dual-polarization radar has been used to compare the microphysical characteristics of heavy rainfall between observations and NWP models in order to improve the microphysical schemes within the models [28,53,54,55]. Zhou et al. [26] pointed out that the default settings of microphysics schemes used in the WRF model that was primarily designed based on the environmental characteristics of North America produced poor simulation results in southeastern China. With three 2M schemes (THOM, Morrison, and WDM6) employed, it is shown that both THOM and WDM6 underestimated KDP and ZDR, while Morrison significantly overestimated them. Increasing the coalescence-breakup and evaporation efficiencies in THOM improved the simulation results. Sun et al. [56] compared the polarimetric radar variables and corresponding hydrometeor types derived from the WRF model and radar simulator, and conducted radar observations and retrievals of a strong squall line over central China. By comparing ZDR and KDP, it was found that the Morrison and WDM6 schemes simulated a lower proportion of large raindrops and lower liquid water content in the convective region [56]. Zhou et al. [57] simulated a heavy precipitation process in Guangdong, with the three 2M microphysics schemes (Morrison, THOM, and MY), overestimated KDP and ZDR, and raindrops’ mass-weighted mean diameter (Dm,r). Reducing the coalescence-break efficiency in the MY scheme led to better simulation results after the initial stage [57]. These studies indicate that dual-polarization radar plays a crucial role in evaluating and improving microphysics schemes in models.

6. Conclusions

In this study, we utilized the S-band dual-polarization radar parameters (ZH, KDP, ZDR) located in Guangzhou, and conducted a simulation analysis of a severe convective precipitation process in Guangdong on 8 May 2017 using the THOM and MY microphysical schemes in a WRF model. The outputs from WRFv4.2 were transformed into simulated radar polarization parameters by use of the CAPS-PRS radar simulator. The aim of this study was to deepen our understanding of the impact of microphysical processes on the development and evolution of convective systems in models through the investigation of parameterizations of raindrop breakup and evaporation.
Key conclusions obtained are as follows:
(1) The THOM scheme produced more snow crystals, while the MY scheme produced more graupel particles. The total ice water content of both schemes was very close. The ice-phase particles of both schemes melted rapidly below the 0 °C level and had similar evaporation efficiencies, resulting in comparable surface precipitation.
(2) The simulated vertical cross-section of ZH from THOM was closer to the observations than MY. It is shown that THOM’s simulation of the strong radar reflectivity region in the convective area was closer to the observations, which is due to the graupel/hail particles simulated by the MY scheme that exists at higher altitudes.
From the source and sink terms in two schemes, it is evident that the increase in the liquid water mixing ratio below the 0 °C level primarily originates from the melting of the ice-phase particle. Between the 0 °C level and 2 km in altitude, almost all ice-phase particles have melted. Below 2 km in altitude, the liquid water mixing ratio decreases primarily due to evaporation. Therefore, the differences between the simulated and the observed KDP between the melting layer and 2 km in altitude may be due to the default settings of the raindrop evaporation or breakup parameterization schemes.
(3) Compared to the observations of ZDR, the simulations produced overestimated the raindrop size. This discrepancy suggests that the relatively lower efficiency of raindrop breakup in the default settings of the THOM may be the contributing factors. To further investigate the impact of microphysical processes on raindrop size distributions and cold pool dynamics, two sets of numerical experiments were conducted, focusing on modifying the breakup diameter threshold (Db) and the evaporation efficiency (EE) to assess their effects on Dm,r and cold pool, respectively. It was found that adjusting Db significantly changed the simulated raindrop size distribution, and had a certain impact on the strength of cold pool, whereas modifying EE significantly not only changed the intensity and scope of the cold pool, but also had great effect on the Dm,r. We then compared the ZDR from the sensitivity experiments with those observed and found that reducing Db can cause ZDR to concentrate within the 0–1 dB range (consistent with observations) and THOM_BKP1000 most closely matched the observations. Therefore, we conclude that reducing Db can produce better simulation results of ZDR.
The schematic diagram of the potential impact of droplet breakup and evaporation on convective development can be shown in Figure 12. After the decrease/increase in Db, more/fewer small raindrops will be generated, and the intensity and extent of the cold pool will slightly increase/decrease, subsequently affecting the strengthening/weakening of downdrafts/updrafts. The increase/decrease in EE has little effect on raindrop size, but it significantly enhances/weakens the intensity and extent of the cold pool, leading to the strengthening of downdrafts/updrafts.
This paper specifically analyzed a severe convective precipitation event under weak synoptic conditions (the 0–6 km vertical wind shears of MCSs in southeast China during PSRS range from 10 to 20 m/s [51,58,59], which is weaker than the North America counterparts of 15–30 m/s [50,60]) and a humid climate in the Guangdong region. Future research should continue to utilize dual-polarization radar to explore the relationship between the formation of bow echoes and the environmental conditions in southeast China. This highlights the complexity of microphysical processes and the need for comprehensive parameter tuning. Our study demonstrates the feasibility of modifying the default microphysics settings of THOM based on polarimetric radar data and a radar simulator to improve the simulation of MCSs in southeast China. However, the effectiveness of this approach needs to be confirmed through additional case studies to refine the simulation and forecasting of MCSs in this region.

Author Contributions

Writing—original draft, Z.C.; Writing—review & editing, X.L. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the National Natural Science Foundation of China (Grant Nos. 41975176, 42061134009).

Data Availability Statement

The data presented in this study are available on request from the author.

Acknowledgments

We acknowledge the High Performance Computing Platform of Nanjing University of Information Science and Technology for their support of this work. Moreover, this study was supported by the National Key Scientific and Technological Infrastructure project Earth System Numerical Simulation Facility (EarthLab).

Conflicts of Interest

The authors declare no conflict of interest.

Appendix A

Table A1. THOM source and sink terms and full names of microphysical processes.
Table A1. THOM source and sink terms and full names of microphysical processes.
Self-Named TitlesOriginal TitlesMicrophysical Process
qg_gccprg_gcwGraupel collecting cloud water.
qr_gmlprr_gmlGraupel melting into rain.
qg_ihmprg_ihmIce multiplication from rime-splinters.
qg_rcgprg_rcgRain collecting graupel.
qg_gdeprg_gdeDeposition/sublimation of graupel.
qg_rfzprg_rfzRain freezing into graupel.
qg_rciprg_rciIce collecting rain into graupel.
qg_rcsprg_rcsSnow collecting rain into graupel.
qg_sccprg_scwSnow collecting cloud water into graupel.
qr_smlprr_smlSnow melting into rain.
qr_rciprr_rciRain collecting ice.
qr_rcsprr_rcsRain collecting snow.
qr_rcgprr_rcgRain collecting graupel.
qr_rccprr_rcwRain collecting cloud water.
qr_cauprr_wauAutoconversion.
qr_revprv_revRain evaporation.
qi_rfzpri_rfzRain freezing into ice.
Table A2. MY source and sink terms and full names of microphysical processes.
Table A2. MY source and sink terms and full names of microphysical processes.
Self-Named TitlesOriginal TitlesMicrophysical Process
qg_gccQCLcgGraupel collecting cloud water.
qr_gmlQMLgrGraupel melting into rain.
qg_iimQIMgiIce multiplication from rime-splinters.
qg_rcgQCLgrRain collecting graupel.
qg_vvdQVDvgDeposition/sublimation of graupel.
qg_gciQCLigGraupel collecting ice.
qh_gcnQCNghGraupel converting to hail.
qg_scnQCNsgSnow converting to graupel.
qg_icfDirg*(QCLir+QCLri)3-comp.freezing into graupel.
qg_gcfDgrg*(QCLgr+QCLrg)3-comp.freezing into graupel.
qg_scfDsrg*(QCLsr+QCLrs)3-comp.freezing into graupel.
qh_gcfDgrh*(QCLgr+QCLrg)3-comp.freezing into hail.
qh_scfDsrh*(QCLsr+QCLrs)3-comp.freezing into hail.
qh_icfDirh*(QCLir+QCLri)3-comp.freezing into hail.
qh_vvdQVDvhDeposition/sublimation of hail.
qr_hmlQMLhrHail melting into rain.
qh_hcrQCLrhHail collecting rain.
qh_rfzQFZrhRain freezing into hail.
qh_hcsQCLshHail collecting snow.
qh_hciQCLihHail collecting ice.
qh_hccQCLchHail collecting cloud water.
qr_smlQMLsrSnow melting into rain.
qr_gmlQMLgrGraupel melting into rain.
qr_rciQCLriRain collecting ice.
qr_icrQCLirIce collecting rain.
qr_rcsQCLrsRain collecting snow.
qr_scrQCLsrSnow collecting rain.
qr_rcgQCLrgRain collecting graupel.
qr_gcrQCLgrGraupel collecting rain.
qr_rccRCACCRRain collecting cloud water.
qr_cauRCAUTRAutoconversion.
qr_revQREVPRain evaporation.
qr_imlQMLirIce melting into rain.
qr_rchQCLrhRain collecting hail.
qh_rfzQFZrhFreezing water drops into hail.
*: multiplication sign; 3-comp.: three types of hydrometeors are involved.

Appendix B

For the raindrop breakup process, the coalescence-breakup efficiency in THOM was defined based on the formula proposed by Verlinde & Cotton [61]:
  E C = 2 − exp ⁡ 2.3 × 10 6 × D m − D b   ,   D m > 0.05   m m  
where EC represents the rain coalescence-breakup efficiency, and Db (mm) is the threshold diameter for initiating breakup. In THOM, the default value for Db is set at 1.95 mm.
For the raindrop evaporation process, the THOM model follows the Formula (16) from Srivastava & Coen [62], where the evaporation efficiency is given by:
  E E = 2 π S − 1 f N 0 f λ f ρ f V f a l l C P 92 ,
where S represents the air supersaturation, f(X) denotes the relationship between EE and X, ρ is the air density, Vfall is a parameter for the raindrop fall velocity, and CP92 refers to certain thermodynamic and nucleation dynamics parameters.

References

  1. Luo, Y. Advances in Understanding the Early-Summer Heavy Rainfall over South China. In World Scientific Series on Asia-Pacific Weather and Climate; World Scientific: Singapore, 2017; Volume 9, pp. 215–226. ISBN 978-981-320-090-6. [Google Scholar]
  2. Guo, J.; Feng, T.; Cai, Z.; Lian, X.; Tang, W. Vulnerability Assessment for Power Transmission Lines under Typhoon Weather Based on a Cascading Failure State Transition Diagram. Energies 2020, 13, 3681. [Google Scholar] [CrossRef]
  3. Wang, L.; Dong, Y.; Zhang, C.; Heng, Z. Extreme and Severe Convective Weather Disasters: A Dual-Polarization Radar Nowcasting Method Based on Physical Constraints and a Deep Neural Network Model. Atmos. Res. 2023, 289, 106750. [Google Scholar] [CrossRef]
  4. Cao, X.; Qi, Y.; Ni, G. X-Band Polarimetric Radar QPE for Urban Hydrology: The Increased Contribution of High-Resolution Rainfall Capturing. J. Hydrol. 2023, 617, 128905. [Google Scholar] [CrossRef]
  5. Chao, L.; Zhang, K.; Yang, Z.-L.; Wang, J.; Lin, P.; Liang, J.; Li, Z.; Gu, Z. Improving Flood Simulation Capability of the WRF-Hydro-RAPID Model Using a Multi-Source Precipitation Merging Method. J. Hydrol. 2021, 592, 125814. [Google Scholar] [CrossRef]
  6. Wang, S.; Zhang, K.; Chao, L.; Li, D.; Tian, X.; Bao, H.; Chen, G.; Xia, Y. Exploring the Utility of Radar and Satellite-Sensed Precipitation and Their Dynamic Bias Correction for Integrated Prediction of Flood and Landslide Hazards. J. Hydrol. 2021, 603, 126964. [Google Scholar] [CrossRef]
  7. Luo, L.; Xue, M.; Zhu, K.; Zhou, B. Explicit Prediction of Hail Using Multimoment Microphysics Schemes for a Hailstorm of 19 March 2014 in Eastern China. JGR Atmos. 2017, 122, 7560–7581. [Google Scholar] [CrossRef]
  8. Morrison, H.; Van Lier-Walqui, M.; Fridlind, A.M.; Grabowski, W.W.; Harrington, J.Y.; Hoose, C.; Korolev, A.; Kumjian, M.R.; Milbrandt, J.A.; Pawlowska, H.; et al. Confronting the Challenge of Modeling Cloud and Precipitation Microphysics. J. Adv. Model. Earth Syst. 2020, 12, e2019MS001689. [Google Scholar] [CrossRef]
  9. Bao, J.-W.; Michelson, S.A.; Grell, E.D. Microphysical Process Comparison of Three Microphysics Parameterization Schemes in the WRF Model for an Idealized Squall-Line Case Study. Mon. Weather Rev. 2019, 147, 3093–3120. [Google Scholar] [CrossRef]
  10. Khain, A.P.; Beheng, K.D.; Heymsfield, A.; Korolev, A.; Krichak, S.O.; Levin, Z.; Pinsky, M.; Phillips, V.; Prabhakaran, T.; Teller, A.; et al. Representation of Microphysical Processes in Cloud-resolving Models: Spectral (Bin) Microphysics versus Bulk Parameterization. Rev. Geophys. 2015, 53, 247–322. [Google Scholar] [CrossRef]
  11. Wu, D.; Dong, X.; Xi, B.; Feng, Z.; Kennedy, A.; Mullendore, G.; Gilmore, M.; Tao, W. Impacts of Microphysical Scheme on Convective and Stratiform Characteristics in Two High Precipitation Squall Line Events. JGR Atmos. 2013, 118. [Google Scholar] [CrossRef]
  12. Wang, H.; Auligné, T.; Morrison, H. Impact of Microphysics Scheme Complexity on the Propagation of Initial Perturbations. Mon. Weather Rev. 2012, 140, 2287–2296. [Google Scholar] [CrossRef]
  13. Zhang, Y.; Yu, H.; Zhang, M.; Yang, Y.; Meng, Z. Uncertainties and Error Growth in Forecasting the Record-Breaking Rainfall in Zhengzhou, Henan on 19–20 July 2021. Sci. China Earth Sci. 2022, 65, 1903–1920. [Google Scholar] [CrossRef]
  14. Zhu, K.; Zhang, C.; Xue, M.; Yang, N. Predictability and Skill of Convection-Permitting Ensemble Forecast Systems in Predicting the Record-Breaking “21·7” Extreme Rainfall Event in Henan Province, China. Sci. China Earth Sci. 2022, 65, 1879–1902. [Google Scholar] [CrossRef]
  15. Simpson, J.E. A Comparison between Laboratory and Atmospheric Density Currents. Quart. J. R. Meteoro Soc. 1969, 95, 758–765. [Google Scholar] [CrossRef]
  16. Morrison, H.; Milbrandt, J.A. Parameterization of Cloud Microphysics Based on the Prediction of Bulk Ice Particle Properties. Part I: Scheme Description and Idealized Tests. J. Atmos. Sci. 2015, 72, 287–311. [Google Scholar] [CrossRef]
  17. Qian, Q.; Lin, Y.; Luo, Y.; Zhao, X.; Zhao, Z.; Luo, Y.; Liu, X. Sensitivity of a Simulated Squall Line During Southern China Monsoon Rainfall Experiment to Parameterization of Microphysics. JGR Atmos. 2018, 123, 4197–4220. [Google Scholar] [CrossRef]
  18. Drager, A.J.; Van Den Heever, S.C. Characterizing Convective Cold Pools. J. Adv. Model. Earth Syst. 2017, 9, 1091–1115. [Google Scholar] [CrossRef]
  19. Langhans, W.; Romps, D.M. The Origin of Water Vapor Rings in Tropical Oceanic Cold Pools. Geophys. Res. Lett. 2015, 42, 7825–7834. [Google Scholar] [CrossRef]
  20. Khairoutdinov, M.; Randall, D. High-Resolution Simulation of Shallow-to-Deep Convection Transition over Land. J. Atmos. Sci. 2006, 63, 3421–3436. [Google Scholar] [CrossRef]
  21. Weisman, M.L.; Rotunno, R. “A Theory for Strong Long-Lived Squall Lines” Revisited. J. Atmos. Sci. 2004, 61, 361–382. [Google Scholar] [CrossRef]
  22. Davis, C.A.; Lee, W.-C. Mesoscale Analysis of Heavy Rainfall Episodes from SoWMEX/TiMREX. J. Atmos. Sci. 2012, 69, 521–537. [Google Scholar] [CrossRef]
  23. Morrison, H.; Milbrandt, J. Comparison of Two-Moment Bulk Microphysics Schemes in Idealized Supercell Thunderstorm Simulations. Mon. Wea. Rev. 2011, 139, 1103–1130. [Google Scholar] [CrossRef]
  24. Tridon, F.; Planche, C.; Mroz, K.; Banson, S.; Battaglia, A.; Van Baelen, J.; Wobrock, W. On the Realism of the Rain Microphysics Representation of a Squall Line in the WRF Model. Part I: Evaluation with Multifrequency Cloud Radar Doppler Spectra Observations. Mon. Weather Rev. 2019, 147, 2787–2810. [Google Scholar] [CrossRef]
  25. Van Weverberg, K.; Vogelmann, A.M.; Morrison, H.; Milbrandt, J.A. Sensitivity of Idealized Squall-Line Simulations to the Level of Complexity Used in Two-Moment Bulk Microphysics Schemes. Mon. Weather Rev. 2012, 140, 1883–1907. [Google Scholar] [CrossRef]
  26. Zhou, A.; Zhao, K.; Lee, W.; Ding, Z.; Lu, Y.; Huang, H. Evaluation and Modification of Microphysics Schemes on the Cold Pool Evolution for a Simulated Bow Echo in Southeast China. JGR Atmos. 2022, 127, e2021JD035262. [Google Scholar] [CrossRef]
  27. Lompar, M.; Ćurić, M.; Romanić, D. Implementation of a Gust Front Head Collapse Scheme in the WRF Numerical Model. Atmos. Res. 2018, 203, 231–245. [Google Scholar] [CrossRef]
  28. Johnson, M.; Jung, Y.; Dawson, D.; Supinie, T.; Xue, M.; Park, J.; Lee, Y.-H. Evaluation of Unified Model Microphysics in High-Resolution NWP Simulations Using Polarimetric Radar Observations. Adv. Atmos. Sci. 2018, 35, 771–784. [Google Scholar] [CrossRef]
  29. Wang, M.; Zhao, K.; Pan, Y.; Xue, M. Evaluation of Simulated Drop Size Distributions and Microphysical Processes Using Polarimetric Radar Observations for Landfalling Typhoon Matmo (2014). JGR Atmos. 2020, 125, e2019JD031527. [Google Scholar] [CrossRef]
  30. Bringi, V.N.; Chandrasekar, V.; Hubbert, J.; Gorgucci, E.; Randeu, W.L.; Schoenhuber, M. Raindrop Size Distribution in Different Climatic Regimes from Disdrometer and Dual-Polarized Radar Analysis. J. Atmos. Sci. 2003, 60, 354–365. [Google Scholar] [CrossRef]
  31. Ryzhkov, A.V.; Zrnic, D.S. Radar Polarimetry for Weather Observations; Springer Atmospheric Sciences; Springer International Publishing: Cham, Switzerland, 2019; ISBN 978-3-030-05092-4. [Google Scholar]
  32. Kumjian, M. Principles and Applications of Dual-Polarization Weather Radar. Part I: Description of the Polarimetric Radar Variables. J. Oper. Meteor. 2013, 1, 226–242. [Google Scholar] [CrossRef]
  33. Laroche, S. Polarimetric Doppler Weather Radar: Principles and Applications. Atmos. Res. 2002, 63, 159–160. [Google Scholar] [CrossRef]
  34. Carr, N.; Kirstetter, P.E.; Gourley, J.J.; Hong, Y. Polarimetric Signatures of Midlatitude Warm-Rain Precipitation Events. J. Appl. Meteorol. Climatol. 2017, 56, 697–711. [Google Scholar] [CrossRef]
  35. Ryzhkov, A.; Pinsky, M.; Pokrovsky, A.; Khain, A. Polarimetric Radar Observation Operator for a Cloud Model with Spectral Microphysics. J. Appl. Meteorol. Climatol. 2011, 50, 873–894. [Google Scholar] [CrossRef]
  36. Snyder, J.C.; Bluestein, H.B.; Dawson Ii, D.T.; Jung, Y. Simulations of Polarimetric, X-Band Radar Signatures in Supercells. Part II: ZDR Columns and Rings and KDP Columns. J. Appl. Meteorol. Climatol. 2017, 56, 2001–2026. [Google Scholar] [CrossRef]
  37. Gao, W.; Xue, L.; Liu, L.; Lu, C.; Yun, Y.; Zhou, W. A Study of the Fraction of Warm Rain in a Pre-Summer Rainfall Event over South China. Atmos. Res. 2021, 262, 105792. [Google Scholar] [CrossRef]
  38. Morrison, H.; Thompson, G.; Tatarskii, V. Impact of Cloud Microphysics on the Development of Trailing Stratiform Precipitation in a Simulated Squall Line: Comparison of One- and Two-Moment Schemes. Mon. Weather Rev. 2009, 137, 991–1007. [Google Scholar] [CrossRef]
  39. Dawson, D.T.; Xue, M.; Milbrandt, J.A.; Yau, M.K. Comparison of Evaporation and Cold Pool Development between Single-Moment and Multimoment Bulk Microphysics Schemes in Idealized Simulations of Tornadic Thunderstorms. Mon. Weather Rev. 2010, 138, 1152–1171. [Google Scholar] [CrossRef]
  40. Wang, H.; Kong, F.; Wu, N.; Lan, H.; Yin, J. An Investigation into Microphysical Structure of a Squall Line in South China Observed with a Polarimetric Radar and a Disdrometer. Atmos. Res. 2019, 226, 171–180. [Google Scholar] [CrossRef]
  41. Thompson, G.; Field, P.R.; Rasmussen, R.M.; Hall, W.D. Explicit Forecasts of Winter Precipitation Using an Improved Bulk Microphysics Scheme. Part II: Implementation of a New Snow Parameterization. Mon. Weather Rev. 2008, 136, 5095–5115. [Google Scholar] [CrossRef]
  42. Milbrandt, J.A.; Yau, M.K. A Multimoment Bulk Microphysics Parameterization. Part II: A Proposed Three-Moment Closure and Scheme Description. J. Atmos. Sci. 2005, 62, 3065–3081. [Google Scholar] [CrossRef]
  43. Jung, Y.; Xue, M.; Zhang, G. Simulations of Polarimetric Radar Signatures of a Supercell Storm Using a Two-Moment Bulk Microphysics Scheme. J. Appl. Meteorol. Climatol. 2010, 49, 146–163. [Google Scholar] [CrossRef]
  44. Hubbert, J.; Bringi, V.N.; Carey, L.D.; Bolen, S. CSU-CHILL Polarimetric Radar Measurements from a Severe Hail Storm in Eastern Colorado. J. Appl. Meteor. 1998, 37, 749–775. [Google Scholar] [CrossRef]
  45. Loney, M.L. Enhanced Polarimetric Radar Signatures above the Melting Level in a Supercell Storm. J. Appl. Meteorol. 2002, 41, 1179–1194. [Google Scholar] [CrossRef]
  46. Putnam, B.J.; Xue, M.; Jung, Y.; Zhang, G.; Kong, F. Simulation of Polarimetric Radar Variables from 2013 CAPS Spring Experiment Storm-Scale Ensemble Forecasts and Evaluation of Microphysics Schemes. Mon. Weather Rev. 2017, 145, 49–73. [Google Scholar] [CrossRef]
  47. Han, B.; Fan, J.; Varble, A.; Morrison, H.; Williams, C.R.; Chen, B.; Dong, X.; Giangrande, S.E.; Khain, A.; Mansell, E.; et al. Cloud-Resolving Model Intercomparison of an MC3E Squall Line Case: Part II. Stratiform Precipitation Properties. JGR Atmos. 2019, 124, 1090–1117. [Google Scholar] [CrossRef]
  48. Han, M.; Braun, S.A.; Matsui, T.; Williams, C.R. Evaluation of Cloud Microphysics Schemes in Simulations of a Winter Storm Using Radar and Radiometer Measurements. JGR Atmos. 2013, 118, 1401–1419. [Google Scholar] [CrossRef]
  49. Wheatley, D.M.; Yussouf, N.; Stensrud, D.J. Ensemble Kalman Filter Analyses and Forecasts of a Severe Mesoscale Convective System Using Different Choices of Microphysics Schemes. Mon. Weather Rev. 2014, 142, 3243–3263. [Google Scholar] [CrossRef]
  50. Fan, J.; Han, B.; Varble, A.; Morrison, H.; North, K.; Kollias, P.; Chen, B.; Dong, X.; Giangrande, S.E.; Khain, A.; et al. Cloud-resolving Model Intercomparison of an MC3E Squall Line Case: Part I—Convective Updrafts. JGR Atmos. 2017, 122, 9351–9378. [Google Scholar] [CrossRef]
  51. Zhou, A.; Zhao, K.; Lee, W.; Huang, H.; Hu, D.; Fu, P. VDRAS and Polarimetric Radar Investigation of a Bow Echo Formation After a Squall Line Merged With a Preline Convective Cell. JGR Atmos. 2020, 125, e2019JD031719. [Google Scholar] [CrossRef]
  52. Kumjian, M.R.; Lebo, Z.J.; Morrison, H.C. On the Mechanisms of Rain Formation in an Idealized Supercell Storm. Mon. Weather Rev. 2015, 143, 2754–2773. [Google Scholar] [CrossRef]
  53. Barnes, H.C.; Houze, R.A. Comparison of Observed and Simulated Spatial Patterns of Ice Microphysical Processes in Tropical Oceanic Mesoscale Convective Systems. JGR Atmos. 2016, 121, 8269–8296. [Google Scholar] [CrossRef]
  54. Kumjian, M.R.; Mishra, S.; Giangrande, S.E.; Toto, T.; Ryzhkov, A.V.; Bansemer, A. Polarimetric Radar and Aircraft Observations of Saggy Bright Bands during MC3E. J. Geophys. Res. Atmos. 2016, 121, 3584–3607. [Google Scholar] [CrossRef]
  55. Putnam, B.J.; Xue, M.; Jung, Y.; Snook, N.; Zhang, G. The Analysis and Prediction of Microphysical States and Polarimetric Radar Variables in a Mesoscale Convective System Using Double-Moment Microphysics, Multinetwork Radar Data, and the Ensemble Kalman Filter. Mon. Weather Rev. 2014, 142, 141–162. [Google Scholar] [CrossRef]
  56. Sun, Y.; Zhou, Z.; Gao, Q.; Li, H.; Wang, M. Evaluating Simulated Microphysics of Stratiform and Convective Precipitation in a Squall Line Event Using Polarimetric Radar Observations. Remote Sens. 2023, 15, 1507. [Google Scholar] [CrossRef]
  57. Zhou, Z.; Du, M.; Hu, Y.; Kang, Z.; Yu, R.; Guo, Y. An Evaluation and Improvement of Microphysical Parameterization for a Heavy Rainfall Process during the Meiyu Season. Remote Sens. 2024, 16, 1636. [Google Scholar] [CrossRef]
  58. Wan, Q.; Wang, B.; Wong, W.K.; Hu, Z.; Jou, B.J.-D.; Lin, Y.; Johnson, R.H.; Chang, C.-P.; Zhu, Y.; Zhang, X.; et al. The Southern China Monsoon Rainfall Experiment (SCMREX). Bull. Am. Meteorol. Soc. 2017, 98, 999–1013. [Google Scholar] [CrossRef]
  59. Meng, Z.; Yan, D.; Zhang, Y. General Features of Squall Lines in East China. Mon. Weather Rev. 2013, 141, 1629–1647. [Google Scholar] [CrossRef]
  60. Xu, X.; Xue, M.; Wang, Y. Mesovortices within the 8 May 2009 Bow Echo over the Central United States: Analyses of the Characteristics and Evolution Based on Doppler Radar Observations and a High-Resolution Model Simulation. Mon. Weather Rev. 2015, 143, 2266–2290. [Google Scholar] [CrossRef]
  61. Verlinde, J.; Cotton, W.R. Fitting Microphysical Observations of Nonsteady Convective Clouds to a Numerical Model: An Application of the Adjoint Technique of Data Assimilation to a Kinematic Model. Mon. Wea. Rev. 1993, 121, 2776–2793. [Google Scholar] [CrossRef]
  62. Srivastava, R.C.; Coen, J.L. New Explicit Equations for the Accurate Calculation of the Growth and Evaporation of Hydrometeors by the Diffusion of Water Vapor. J. Atmos. Sci. 1992, 49, 1643–1651. [Google Scholar] [CrossRef]
Remotesensing 16 04297 g001
Figure 1. The numerical simulation domain setup. The numerical simulation involved two-way nested domains, d01 and d02. The white triangle marks the location of Qingyuan City.
Figure 1. The numerical simulation domain setup. The numerical simulation involved two-way nested domains, d01 and d02. The white triangle marks the location of Qingyuan City.
Remotesensing 16 04297 g001
Remotesensing 16 04297 g002
Figure 2. The composite reflectivity (units: dBZ) from radar observations (a1–c1) at 0600, 0854, and 1200 UTC; the composite reflectivity from the THOM simulation (a2–c2) at 0630, 0930, and 1230 UTC and the MY simulation (a3–c3) at 0630, 0830, and 1200 UTC. The black lines represent the vertical cross-section transects.
Figure 2. The composite reflectivity (units: dBZ) from radar observations (a1–c1) at 0600, 0854, and 1200 UTC; the composite reflectivity from the THOM simulation (a2–c2) at 0630, 0930, and 1230 UTC and the MY simulation (a3–c3) at 0630, 0830, and 1200 UTC. The black lines represent the vertical cross-section transects.
Remotesensing 16 04297 g002
Remotesensing 16 04297 g003
Figure 3. Vertical cross-sections of polarimetric parameters during the mature stage of convections from radar observations and model simulations; (a1–a3,b1–b3,c1–c3) represent the ZH (units: dBZ), KDP (units: °/km), and ZDR (units: dB) from the radar observations of THOM, and MY, respectively. The red dashed and solid lines in the figures are the isotherms of −40 °C, −20 °C, and 0 °C, where the black arrows indicate the wind field.
Figure 3. Vertical cross-sections of polarimetric parameters during the mature stage of convections from radar observations and model simulations; (a1–a3,b1–b3,c1–c3) represent the ZH (units: dBZ), KDP (units: °/km), and ZDR (units: dB) from the radar observations of THOM, and MY, respectively. The red dashed and solid lines in the figures are the isotherms of −40 °C, −20 °C, and 0 °C, where the black arrows indicate the wind field.
Remotesensing 16 04297 g003
Remotesensing 16 04297 g004
Figure 4. The 18-h accumulated precipitation (units: mm) from 0000 to 1800 UTC on 8 May 2017, obtained from ground observation stations (a), THOM (b), and MY (c). The white triangle marks the location of the radar (GZRD).
Figure 4. The 18-h accumulated precipitation (units: mm) from 0000 to 1800 UTC on 8 May 2017, obtained from ground observation stations (a), THOM (b), and MY (c). The white triangle marks the location of the radar (GZRD).
Remotesensing 16 04297 g004
Remotesensing 16 04297 g005
Figure 5. Vertical distribution of the regional averaged hydrometeor mixing ratios (units: g kg−1) at the mature stage for the two schemes, with (a) for the THOM scheme and (b) for the MY scheme. The red solid line indicates the 0 °C isothermal layer.
Figure 5. Vertical distribution of the regional averaged hydrometeor mixing ratios (units: g kg−1) at the mature stage for the two schemes, with (a) for the THOM scheme and (b) for the MY scheme. The red solid line indicates the 0 °C isothermal layer.
Remotesensing 16 04297 g005
Remotesensing 16 04297 g006
Figure 6. Vertical distribution of the mean tendency of rain and graupel/hail mixing ratios (units: g kg−1) during the mature stage for the two schemes; (a) for raindrops in THOM, and (b) for raindrops in MY, (c) for graupel particles in THOM, (d) for graupel particles in MY, (e) for hail particles in MY. The red solid line indicates the 0 °C isothermal layer. The full names of microphysical processes are listed in the Appendix A Table A1 and Table A2.
Figure 6. Vertical distribution of the mean tendency of rain and graupel/hail mixing ratios (units: g kg−1) during the mature stage for the two schemes; (a) for raindrops in THOM, and (b) for raindrops in MY, (c) for graupel particles in THOM, (d) for graupel particles in MY, (e) for hail particles in MY. The red solid line indicates the 0 °C isothermal layer. The full names of microphysical processes are listed in the Appendix A Table A1 and Table A2.
Remotesensing 16 04297 g006
Remotesensing 16 04297 g007
Figure 7. Horizontal cross-sections of potential temperature (units: K) at 540 m simulated by the THOM (a) and MY (b) schemes, respectively. The black arrows indicate the horizontal wind field.
Figure 7. Horizontal cross-sections of potential temperature (units: K) at 540 m simulated by the THOM (a) and MY (b) schemes, respectively. The black arrows indicate the horizontal wind field.
Remotesensing 16 04297 g007
Remotesensing 16 04297 g008
Figure 8. Probability Density Function (PDF) distribution of the simulated Dm,r (units: mm) from the THOM (a) and MY (b) schemes. The solid black line indicates the vertical profile of the mean Dm,r, while the solid red line denotes the 0 °C isotherm.
Figure 8. Probability Density Function (PDF) distribution of the simulated Dm,r (units: mm) from the THOM (a) and MY (b) schemes. The solid black line indicates the vertical profile of the mean Dm,r, while the solid red line denotes the 0 °C isotherm.
Remotesensing 16 04297 g008
Remotesensing 16 04297 g009
Figure 9. A PDF of the difference in simulated Dm,r between sensitivity experiments and THOM_CTR during the mature stage of the convection. Panels (a1–f1) display the results of the BKP experiments, while panels (a2–e2) show the outcomes of the EVP experiments.
Figure 9. A PDF of the difference in simulated Dm,r between sensitivity experiments and THOM_CTR during the mature stage of the convection. Panels (a1–f1) display the results of the BKP experiments, while panels (a2–e2) show the outcomes of the EVP experiments.
Remotesensing 16 04297 g009
Remotesensing 16 04297 g010
Figure 10. Horizontal cross-sections of the difference in simulated potential temperature (units: K) between sensitivity experiments and THOM_CTR at an altitude of 540m during the mature stage, with the results of the BKP tests (a1–f1) and the results of the EVP tests (a2–e2). The black arrows represent the horizontal wind field.
Figure 10. Horizontal cross-sections of the difference in simulated potential temperature (units: K) between sensitivity experiments and THOM_CTR at an altitude of 540m during the mature stage, with the results of the BKP tests (a1–f1) and the results of the EVP tests (a2–e2). The black arrows represent the horizontal wind field.
Remotesensing 16 04297 g010
Remotesensing 16 04297 g011
Figure 11. A PDF of the ZDR (units: dB) by the OBS (a), THOM_CTR (b), and THOM_BKP1000 (c) during the mature stage. The solid red line denotes the 0 °C isotherm.
Figure 11. A PDF of the ZDR (units: dB) by the OBS (a), THOM_CTR (b), and THOM_BKP1000 (c) during the mature stage. The solid red line denotes the 0 °C isotherm.
Remotesensing 16 04297 g011
Remotesensing 16 04297 g012
Figure 12. A schematic diagram of the potential impact of droplet breakup and evaporation on convective development, divided into five parts: control group THOM_CTR, increased/decreased Db, and increased/decreased EE. In the figure, the depth of color and size of the grey ellipses represent the intensity and extent of the cold pool, the size of the blue/red arrows indicates the magnitude of the downdraft/updraft, and the red dashed line represents the zero-degree isotherm.
Figure 12. A schematic diagram of the potential impact of droplet breakup and evaporation on convective development, divided into five parts: control group THOM_CTR, increased/decreased Db, and increased/decreased EE. In the figure, the depth of color and size of the grey ellipses represent the intensity and extent of the cold pool, the size of the blue/red arrows indicates the magnitude of the downdraft/updraft, and the red dashed line represents the zero-degree isotherm.
Remotesensing 16 04297 g012
Table 1. Summary of microphysics options for THOM and MY schemes.
Table 1. Summary of microphysics options for THOM and MY schemes.
SchemeMass RatioNumber Concentration
THOMqc, qr, qi, qs, qgNr, Ni
MYqc, qr, qi, qs, qg, qhNc, Nr, Ni, Ns, Ng, Nh
c: cloud; r: rain; i: ice; s: snow; g: graupel; h: hail.
Table 2. Introduction to sensitivity experiments.
Table 2. Introduction to sensitivity experiments.
Modified VariablesThe Name of Sensitivity Experiments
Db (1.95 mm)THOM_BKP1000, THOM_BKP1200, THOM_BKP1400,
THOM_BKP1600, THOM_BKP1800, THOM_BKP2200
EETHOM_EVP0.5, THOM_EVP1.5, THOM_EVP3.0,
THOM_EVP5.0, THOM_EVP10.0
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Cheng, Z.; Liu, X. A Numerical Simulation of Convective Systems in Southeast China: A Comparison of Microphysical Schemes and Sensitivity Experiments on Raindrop Break and Evaporation. Remote Sens. 2024, 16, 4297. https://doi.org/10.3390/rs16224297

AMA Style

Cheng Z, Liu X. A Numerical Simulation of Convective Systems in Southeast China: A Comparison of Microphysical Schemes and Sensitivity Experiments on Raindrop Break and Evaporation. Remote Sensing. 2024; 16(22):4297. https://doi.org/10.3390/rs16224297

Chicago/Turabian Style

Cheng, Zhaoqing, and Xiaoli Liu. 2024. "A Numerical Simulation of Convective Systems in Southeast China: A Comparison of Microphysical Schemes and Sensitivity Experiments on Raindrop Break and Evaporation" Remote Sensing 16, no. 22: 4297. https://doi.org/10.3390/rs16224297

APA Style

Cheng, Z., & Liu, X. (2024). A Numerical Simulation of Convective Systems in Southeast China: A Comparison of Microphysical Schemes and Sensitivity Experiments on Raindrop Break and Evaporation. Remote Sensing, 16(22), 4297. https://doi.org/10.3390/rs16224297

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop