Next Article in Journal
Orbital Uncertainty Propagation Based on Adaptive Gaussian Mixture Model under Generalized Equinoctial Orbital Elements
Next Article in Special Issue
METEO-DLNet: Quantitative Precipitation Nowcasting Net Based on Meteorological Features and Deep Learning
Previous Article in Journal
A Framework for the Construction of a Heritage Corridor System: A Case Study of the Shu Road in China
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Remote Sensing
Volume 15
Issue 19
10.3390/rs15194651
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

An Evaluation of Simulated Cloud Microphysical Characteristics of Three Mei-Yu Rainfall Systems in Taiwan by a Cloud-Resolving Model Using Dual-Polarimetric Radar Observations

by
Chung-Chieh Wang
1,
Yu-Han Chen
2,*,
Yu-Yao Lan
1 and
Wei-Yu Chang
3
1
Department of Earth Sciences, National Taiwan Normal University, Taipei 116325, Taiwan
2
Department of Atmospheric Sciences, National Taiwan University, Taipei 106319, Taiwan
3
Department of Atmospheric Sciences, National Central University, Taoyuan 320317, Taiwan
*
Author to whom correspondence should be addressed.
Remote Sens. 2023, 15(19), 4651; https://doi.org/10.3390/rs15194651
Submission received: 30 June 2023 / Revised: 4 September 2023 / Accepted: 19 September 2023 / Published: 22 September 2023
(This article belongs to the Special Issue Recent Advances in Precipitation Radar)
Browse Figures
Versions Notes

Abstract

:
This study selected three heavy-rainfall events of different types in Taiwan’s Mei-yu season for high-resolution simulations at a grid size of 1 km and assessed the model’s capability to reproduce their morphology and characteristics. The three cases include a pre-frontal squall line, a mesoscale convective system (MCS) embedded in southwesterly flow, and a local convection near the front in southern Taiwan during the South-West Monsoon Experiment (SoWMEX) in 2008, chosen mainly because of the availability of the S-band polarimetric (S-Pol) radar observations, and especially the particle identification results. The simulations using the Cloud-Resolving Storm Simulator (CReSS) could reproduce all three corresponding rainfall systems at roughly the correct time and location, including their kinematic structures such as system-relative flows with minor differences, although the cells appeared to be coarser and wider than the S-Pol observations. The double-moment cold-rain microphysics scheme of the model could also capture the general distributions of hydrometeors, such as heavy rainfall below the updraft core with lighter rainfall farther away below the melting level, and graupel and mixed-phase particles in the upper part of the updraft with snow and ice crystals in stratiform areas between updrafts above the melting level. Near the melting level, the coexistence of rain and snow corresponds to wet snow in the observations. Differences in cloud characteristics in the events are also reflected in the model results to some extent. Overall, the model’s performance in the simulation of hydrometeors exhibits good agreement with the observation and appears reasonable.

1. Introduction

Located in the East Asian monsoon region, Taiwan (Figure 1) receives its rainfall and freshwater resources mainly during the warm season, especially from typhoons and mesoscale convective systems (MCSs) associated with fronts in the Mei-yu season (May–June). In the latter season, organized MCSs, often fed by the low-level jet (LLJ) with enhanced transport of warmth and moisture in the southwesterly flow [1], can bring prolonged heavy rainfall. The steep and complex topography of Taiwan, reaching over 3 km in height within 100 km of the coast, also acts to enhance rainfall. Owing to the above factors, heavy to extreme rainfall can occur in Taiwan and lead to disasters, and presents challenges to not only weather forecasting but also hazard prevention and management by the government.
In the past, when the models used for numerical weather prediction (NWP) were lower in resolution (generally with a grid spacing of ≥10 km), convective rainfall was produced by the cumulus parameterization scheme (CPS), e.g., [2,3,4]. As computers increase in speed and capacity, model grids become finer, and convective clouds can be simulated explicitly through cloud microphysics (CMP) schemes that gradually increase in complexity [5,6,7,8,9]. Thus, convective clouds are more realistic in the model and the rainfall produced by them is also more accurate nowadays, e.g., [6,10,11,12,13].
Many previous studies focused on the development of various CMP schemes and their applications to study convection and rainfall processes, including heavy rainfall events. Ice species and their transformation processes were developed in the 1980s [8,14,15,16], and the effects of these processes in deep convection on rainfall were then explored and discussed [17,18,19]. Later, different kinds of ice species were also considered in CMP schemes to simulate the cloud structure and rainfall of tropical cyclones more realistically [20,21]. As more types of water/ice species in clouds are included, the CMP schemes become more sophisticated and more complete, and the simulated convection increases in scope and agrees better with the observations [9]. Near the surface, the cooling effect from the evaporation of raindrops was also found to be important in extending the lifespan of convective systems [22].
For observations, dual polarimetric radars give not only the reflectivity (ZH) and Doppler velocity fields as conventional Doppler radars, but also the differential reflectivity (ZDR), linear depolarization ratio (LDR), specific differential phase (KDP), and cross-correlation coefficient (ρHV) information. As detailed in [23], ZDR is the ratio of horizontal to vertical returns and thus measures the oblateness of particles, while ρHV is their correlation coefficient and can also identify hail and mixed-phase particles [24,25,26]. Large values of LDR (the ratio between vertically polarized power backscattered for a horizontally polarized transmission and the copolar backscattered power) and KDP (the difference in phase per km of the received horizontal and vertical polarized waves) are used to identify non-spherical particles [23,27,28,29]. For estimating rain rates, the most commonly used parameters are ZH, ZDR, and KDP. In Aydin et al. [30], ZDR and KDP are shown to be able to identify heavy raindrops and hail by using the National Center of Atmospheric Research (NCAR) S-band (dual) polarimetric radar (S-Pol). As valuable tools to study cloud physics and the dynamics of convective systems, dual polarimetric radars can observe particle size and the shape of water/ice species and thus distinguish the types of the particles [23,24,25,26,31,32,33]. More recently, works have also been published on clarifying the limitations of polarimetric radars and improving their observational performance. For example, the work of [23] used fuzzy logic to distinguish the characteristics of rain particles observed by the NCAR S-Pol. They pointed out that different wavelengths of polarimetric radars may give different values of some parameters, and ZH and ZDR of the S-Pol are more easily affected. May and Keenan [34] indicated a bright band issue in the C-band polarimetric radar (C-Pol) that needed to be resolved.
The Southwest Monsoon Experiment (SoWMEX)/Terrain-Influenced Monsoon Rainfall Experiment (TiMREX) was a joint field campaign between Taiwan and the United States and was held over the plains and mountain slopes in southwestern Taiwan (see Figure 1) from 15 May to 30 June 2008. The purpose of this field experiment was to understand the behavior of the Mei-yu front and southwesterly monsoon under the influence of the topography [35]. In this field experiment, the NCAR S-Pol was also deployed on the southwestern coast of Taiwan and provided not only the traditional Doppler variables but also information on the microphysical characteristics of Mei-yu rainfall systems in Taiwan. Such detailed information on water species can be used to improve the quantitative precipitation estimate (QPE) and also to verify model results. Zhou [36] carried out a detailed analysis of the event on 14–15 June during the eighth intensive observing period (IOP-8) of SoWMEX/TiMREX and evaluated the simulations by the WRF model using the S-Pol data. Four microphysics schemes, including the Morrison, Thompson, WRF single-moment 6-class (WSM6), and WRF double-moment 6-class (WDM6) schemes were chosen for comparison, and among them, the WDM6 was shown to produce the most realistic results for the simulated characteristics of the squall line in IOP-8 near Taiwan [36].
Developed at Nagoya University [5,37], the Cloud-Resolving Storm Simulator (CReSS) has been used to simulate and study many heavy rainfall events in Taiwan, including Typhoons Morakot (2009) [11] and Kong-Rey (2013) [38], and Mei-yu events such as the SoWMEX IOP-8 case [12,39], those over 11–12 June 2012, and on 2 June 2017 [40,41,42]. Its more updated version (3.4.2) includes a double-moment bulk cold-rain scheme in both liquid and ice (e.g., [42,43,44]). Meanwhile, a retrieval method from CReSS outputs to radar reflectivity has also been developed (S. Tsujino, personal communication) through Equation (54) of [45]. Therefore, with a more sophisticated CMP scheme, the development, evolution, and structure of MCSs simulated by CReSS can be expected to become more realistic and improved, and the model results more helpful in understanding the mechanisms leading to heavy rainfall in Taiwan. However, while some evaluation of the performance of cloud microphysics in the WRF model in the Mei-yu season in Taiwan has been carried out, as reviewed above, similar studies on the CReSS model are still lacking. Since a comparison between CReSS simulations and S-Pol observations has not been performed, we cannot be certain of the model’s capability, particularly in cloud microphysics. Filling this gap in knowledge is the purpose of the present work.
In this study, three heavy-rainfall events associated with the front in the Mei-yu season are selected for high-resolution simulation by the CReSS model, at the finest grid size (Δx) of 1 km, with a goal of assessing the model’s ability to reproduce the cloud microphysical characteristics of the observed MCSs. These cases all occurred during the field phase of SoWMEX/TiMREX in 2008, in order to take advantage and make use of the detailed observations by the NCAR S-Pol and its particle identification (PID) results. The three events belong to different types of MCSs and include (1) a pre-frontal squall line (PFSL) and (2) MCSs embedded in southwesterly flow (SWMCS), both during IOP-8 (14–17 June), and (3) frontal convection in IOP-3 (29–31 May). Detailed observational studies using the S-Pol data have been conducted for both events, by [46] on IOP-8 and [47] on IOP-3, and both studies include the PID results for the target MCS. Therefore, by making comparisons with these earlier studies, not only the kinematic and rainfall structures of the three MCSs but also their microphysical characteristics in the model simulations will be evaluated. As mentioned, the performance of the WRF model in the PFSL case was studied by [36], but not yet for the CReSS model in any Mei-yu event in Taiwan.
The remaining parts of this paper are arranged as follows: The data and methodology, including the CReSS model and our experiments, are described in Section 2. In Section 3, an overview of the SoWMEX IOP-8 case is first given, followed by S-Pol observations and our simulation results and evaluation of the PFSL (event A). Similar analyses and model assessments are carried out for the SWMCS (event B) in Section 4, and then for the frontal convection (event C) in IOP-3 in Section 5. Finally, further discussion and conclusions are given in Section 6.

2. Data and Methodology

2.1. The CReSS Model

As mentioned, the 3.4.2 version of the CReSS model [5,37] is used and its simulation results are evaluated in this study. In this three-dimensional non-hydrostatic and compressible model, a terrain-following vertical coordinate is used with no nesting (single domain). The basic configuration is summarized in Table 1. Using its double-moment bulk cold-rain scheme [8,16,45,48,49] with six water species (i.e., vapor, cloud water, cloud ice, rain, snow, and graupel), all clouds are simulated explicitly without cumulus parameterization. Subgrid-scale turbulent mixing in the boundary layer is parameterized using 1.5-order closure with turbulent kinetic energy prediction [37,50]. Momentum and energy fluxes and radiation at the surface are also considered with a substrate model [51,52,53], but both shortwave and longwave cloud radiation in the air are neglected [5,37]. At the lower boundary, real topography data are also provided on a fine grid. More details regarding CReSS, including the full equation set in model dynamics and in cloud microphysics, are available in the user guide for further reference (http://www.rain.hyarc.nagoya-u.ac.jp/~tsuboki/cress_html/index_cress_jpn.html (accessed on 1 February 2019)).

2.2. Model Experiments

As the CReSS model has no nesting function, two experiments were first conducted at a coarser grid of Δx = 2.5 km to cover the three SoWMEX cases studied: experiment R2.5AB to cover the first two cases during IOP-8 (PFSL on 14 June and SWMCS on 16 June, referred to as events A and B) together and experiment R2.5C for the third one in IOP-3 (frontal convection on 31 May, referred to as event C), as shown in Table 2 (top half). The National Centers for Environmental Prediction (NCEP) Global Forecast System (GFS) final analyses, on a 1° × 1° latitude–longitude grid at 26 levels, provided the initial and boundary conditions (IC/BCs) for these 2.5 km runs (including the sea surface temperature). In [12], both events A and B were successfully simulated using an earlier version of CReSS with a similar configuration (Δx = 2.5 km) and the same IC/BCs, including rainfall distributions in Taiwan on days 3 and 5 that compare favorably with the observations (their Figures 7 and 12). In addition, the same IC/BCs (produced in real time) were used operationally by the 2.5 km CReSS for years [41,44,54,55,56].
The outputs of the two 2.5 km experiments are then used as IC/BCs to drive three 1 km experiments of shorter integration lengths: R1.0A, R1.0B, and R1.0C, for each of the three cases, respectively (Table 2, bottom half). As southwesterly flow prevailed at low levels during IOP-8 (discussed below), Taiwan is placed closer to the eastern boundary in R1.0A and R1.0B to allow for a longer distance from the upstream (Figure 1). By contrast, Taiwan in R1.0C is placed close to the domain center without such a condition. In later sections, to verify the model results for cloud structure and microphysical characteristics, only the 1 km results are used.

2.3. Data and Methodology for Model Evaluation

The data used in this study include the following: For a synoptic overview, East Asia surface weather charts from the Central Weather Bureau (CWB) of Taiwan and the NCEP GFS gridded final analyses (1° × 1°, see Table 2) at selected pressure levels, both at 6 h intervals, are used. For mesoscale analysis on the convection and MCSs, rain gauge observations and radar reflectivity mosaic (every 30 min) from the CWB, as well as infrared cloud imagery from the geostationary Multifunctional Transport Satellite (MTSAT, every 30 min), are used. To evaluate the CReSS model’s ability to simulate the structure of rainfall systems and their cloud microphysical characteristics, the NCAR S-Pol observations are used. These results are mainly adapted from earlier studies, including the PID results. The model results from the three 1 km experiments are then compared with the S-Pol data.

3. The PFSL Case on 14 June 2008

3.1. Synoptic Overview of Events A and B

In this section, an overview of the synoptic backgrounds of PFSL (event A) and SWMCS (event B) during SoWMEX IOP-8 (14–17 June) is first given. The daily surface weather charts from the CWB (Figure 2) show that a quasi-stationary Mei-yu front existed in South China from 13 June during IOP-8, and did not approach northern Taiwan until around 16 June. Thus, Taiwan was located on the warm side ahead of the front, where south-southwesterly to southwesterly flow prevailed during much of IOP-8. At 850 hPa, as the front approached, strong southwesterly flow prevailed over the prefrontal area near Taiwan on 14 June (Figure 3a, also [12,39]), with LLJ peaking at about 20 m s−1 in response to the increased horizontal geopotential height gradients, e.g., [57]. This Mei-yu frontal system was associated with a trough at 500 hPa (Figure 4a), which provided positive vorticity advection aloft and favorable conditions to trigger convection [58]. Thus, a PFSL developed on 14 June over the Taiwan Strait, with an orientation roughly parallel to the low-level flow (northeast–southwest). The mesoscale aspects of this PFSL will be further described.
With the cyclone migrated to the east, the Mei-yu front also advanced south to approach northern Taiwan on 16 June (Figure 2d). While the 500 hPa trough had moved off and weakened by this time (Figure 4b), the southwesterly and LLJ at low levels remained strong in the northern South China Sea (SCS) and the southern Taiwan Strait (Figure 3b). Under such conditions, an MCS developed upstream from Taiwan over the northern SCS on 16 June, then moved onshore to cause heavy rainfall over southwestern Taiwan. This is referred to as the SWMCS (and event B) in the present work. Since the SoWMEX IOP-8 case was studied quite extensively, similar synoptic analyses can be found in [12,39,58,59,60], with emphases on different aspects of the event.

3.2. Mesoscale Aspects of Event A

On 13–14 June during SoWMEX IOP-8, a well-organized, northeast–southwest-aligned squall line developed along the southeastern coast of China and then propagated eastward across the Taiwan Strait to make landfall in western Taiwan (Figure 5a–f). After the line made landfall around 0100 UTC 14 June, however, its segment on land gradually dissipated and only the southern portion over the water remained intact (Figure 5b; also [57]). Later, this squall line evolved into several convective lines with a general orientation from north-northeast to south-southwest by 0900 UTC (Figure 5c), and their northern portions made landfall in southwestern Taiwan in succession (Figure 5c–e), until after 1600 UTC 14 June (Figure 5f, also [46]). As a whole, this PFSL event brought heavy rainfall to southwestern and central Taiwan on 14 June, with a peak daily amount reaching 200 mm on 14 June (Figure 6a, in LST).

3.3. S-Pol Radar Observations of Event A

Both events A and B during SoWMEX IOP-8 were studied in detail at selected times using the observations from the S-Pol and other land-based radars reported in [46]. For the PFSL, the time chosen for the analyses of the kinematic and microphysical structure of the system was 1000 UTC 14 June, as depicted in Figure 7a–c for plain views and Figure 8a,b for the vertical cross-sections (taken from [46], their Figures 7a and 8). As mentioned earlier, at this time, the PFSL had already evolved into several north–south-aligned narrow lines (see Figure 6c,d) and one of them was approximately 60 km to the west of the S-Pol site and about to make landfall (Figure 7a). The mosaic reflectivity from the S-Pol shows that the line was less than 10 km in width and was embedded in a strong southwesterly flow of almost 20 m s−1, as revealed from wind synthesis at 2.5 km (Figure 7a). Dual-Doppler analyses (inside the rectangular area) also show that relative to the system (moving northeastward at cx = 10.3 m s−1 and cy = 12.0 m s−1, see [46]), there was low-level inflow from the east at the height of 2 km (Figure 7b) and rear inflow from the west at 5 km (Figure 7c). Individual convective cells, roughly 10 km apart, were embedded inside the line.
In the east–west vertical cross-section through the strongest cell in the analysis region at 1000 UTC 14 June (Figure 8a), the range-height indicator (RHI) from the S-Pol indicated a roughly vertical updraft without too much tilt, with a maximum ascending speed of over 10 m s−1 near 6.5 km and reflectivity near 55 dBZ below the melting level of 4.7 km. Ahead (east) of the line, the low-level inflow and upper-level outflow (from dual-Doppler analysis) were both visible, whereas the rear inflow, reaching about 6 m s−1 near 5–6 km in height was also clearly depicted over the area roughly 10–35 km west of the updraft. Just below it, there was stratiform echo reaching 35 dBZ near 5 km height, and a similar feature also existed roughly 10 km ahead of the updraft in the cross-section (Figure 8a). In Figure 8b, the particle identification (PID) results on the cross-section from [46] (their Figure 14a) are shown, and will be further discussed when compared to the model results.

3.4. Model Results of Event A and Comparison with Radar Observations

As described in Section 2, a 1 km CReSS simulation was conducted (Exp. R1.0A) from the initial time of 0000 UTC 13 June for 48 h and driven by the 2.5 km outputs, which are similar to those used in [12,39]. In this section, the model results are compared with the observations mainly from radars, so as to evaluate its performance in reproducing the kinematic structure and microphysical characteristics of the PFSL in event A. To facilitate comparison, most of the plots from the model results are presented in a way similar to the observations.
In Figure 5g–i, the model simulation results of the PFSL at 0930, 1030, and 1130 UTC 14 June are shown when the system approaches southwestern Taiwan and can be compared with Figure 5c–e. First of all, the 1 km model successfully reproduced the system and its landfall time was also roughly correct. Around this time, when the entire system exhibited a northeast–southwest alignment, some narrow convective lines were oriented in a more north–south direction, and the model was able to reproduce this characteristic as well. From outputs every 15 min (Table 2), the results at 1100 UTC 14 June are selected and presented in Figure 7d–f, as a convective line is approaching southwestern Taiwan at this time in the model (its motion determined to be cx = 11.2 and cy = 3.8 m s−1) and characteristics similar to those in Figure 7a–c. At the elevations of 2 and 2.5 km, winds ahead (behind) of the line are from the southwest (west-southwest) in the model (Figure 5g–i and Figure 7d), also in general agreement with the radar synthesis (Figure 7a). However, even at Δx = 1 km, the model rainband appears about 20 km in width with strong echoes exceeding 50 dBZ in Figure 7e,f, wider and stronger than captured by the S-Pol. Also, the system-relative low-level inflow at 2 km east of the line in the model seems to be stronger and more from the south than in the observations (Figure 7b,e), yet nonetheless in agreement with its slower northward motion in the model. Near 5 km, however, the rear inflow was much too weak and almost non-existent (Figure 7c,f). Finally, despite generally stronger reflectivity, the model updrafts embedded in the line also appear somewhat weaker (in vertical velocity), at least at heights of both 2 and 5 km.
The structure of the convection along an east–west vertical cross-section also through a strong cell in the model (see Figure 7e,f for location) is depicted in Figure 8c,d, and can be compared to Figure 8a,b. Following the method of [48] described earlier, the retrieved reflectivity in the model tends to be stronger (Figure 8c) as the rainband also appears wider compared to the S-Pol observation, but also reaches peak values of >55 dBZ below the updraft maximum, which is over 16 m s−1 at around 9 km. Thus, the model updraft in the selected cell is in fact stronger, rather than weaker, than the observation, and is associated with higher reflectivity values underneath, at the lower levels. East of the line, the low-level inflow and upper-level outflow bear close resemblance to the S-Pol data and are quite reasonable. On the other hand, strong real-to-front (RTF) inflow appears higher up, near 8 km, somewhat different from the observation of around 5–6 km. However, mid-level inflow can be found to the rear of another developing cell about 25 km farther to the west (near 119.4°E), which might affect the flow around the selected cell in R1.0A. Thus, some disagreement is perhaps not surprising. On the other hand, just below the freezing level, near 3–5 km, stronger reflectivity is also found to accompany the aforementioned cell to the west, and also about 12 km east of the main updraft in Figure 8c, in good agreement with Figure 8a. Like the main updraft, the echoes of these features are stronger than the observed ones (especially for the one to the west), and this result might be due to the apparently stronger convergence below 2 km associated with the squall line in the model.
The PID results of S-Pol observation in the cross-section as obtained by [46] using the fuzzy-logic method of [23] are shown in Figure 8b, with a total of 17 different types (14 types of hydrometeors). In general, liquid water was observed near and below the freezing level (at 4.7 km) with solid-phase particles above, from about 40 km east to 90 km west of the main updraft. The main updraft was associated with heavy rainfall (HR) near and below freezing level, and graupel mixed with rain (GR) slightly above (~5–7 km) and with small hails (GSH) further up (about 7–11 km). Farther away from the updraft, the rainfall intensity (below 5 km) decreased from HR to moderate rain (MR) and then to light rain (LR) and drizzle (DZ), while the solid particles aloft also became smaller, from dry snow (DS) to irregular ice crystals (IIC) to ice crystals (IC). Near the freezing level, wet snow (WS) could be found at the backside of the main updraft and just above the two regions of higher reflectivity.
The model results of the mixing ratios of hydrometeor species in R1.0A are plotted in Figure 8d using colors similar to those in Figure 8b. Near and below the freezing level at about 5.1 km, raindrops (warm colors) are simulated, with higher values (≥2.5 g kg−1), which can be interpreted as heavier rainfall, located below the updraft maximum as well as in the second high-reflectivity area near 119.4°E. Above 5.1 km, graupels (green contours) are produced in the main updraft and advected eastward by the RTF outflow, and also in the updraft to the western cell (but over a smaller area). In Figure 8d, graupels can reach about 14 km due to the stronger updraft. Farther away from the main cell, snow particles (blue) are also produced, with ice crystals (pink) farther up and to the west at high levels (typically ≥ 9 km). Near the freezing level where both liquid and solid phases coexist in abundance, the presence of mixed-phase particles can be assumed, such as GR in the updrafts and WS. In general, while the CReSS model does not produce as many types of hydrometeors as the S-Pol observations, the rain (liquid), snow and ice (solid), and mixed-phase particles in Figure 8d show good agreement with the observations in Figure 8b at the selected time for the PFSL case.

4. The SWMCS Case on 16 June 2008

4.1. Mesoscale Aspects of Event B

In this subsection, the mesoscale aspects of event B are reviewed, since the synoptic overview was already given earlier for the entire SoWMEX IOP-8 period. Event B of the SWMCS was observed on 16 June, roughly 30 h after the passage of the PFSL, when the southwesterly LLJ occupied the northern SCS and nearby ocean southwest and east of Taiwan (Figure 3d). The LLJ advected warm and unstable air from upstream toward southwestern Taiwan and provided favorable conditions for the development of convection [12,58,59]. From late 15 June, active but scattered convections developed over a wide area near southwestern Taiwan and the upstream ocean (Figure 9a–f), constituting our second event of SWMCS (i.e., event B), but they were not as well organized as the PFSL. After initiation, many of these convective cells moved onshore into the southwestern plains and led to heavy rainfall on 16 June (in LST), with a peak amount of over 200 mm (Figure 6b). Later, as the subtropical high strengthened on 17 June, the areas of strong low-level winds and more active convection also moved westward, gradually away from Taiwan (also [12]).
During the period of SWMCS, the CReSS model is seen to be able to reproduce the active but scattered convection embedded in the confluent southwesterly low-level flow near southwestern Taiwan in the experiment R1.0B (Table 2), in general agreement with the observations, particularly from 0000 to 0800 UTC 16 June (Figure 9g–i).

4.2. S-Pol Radar Observations of Event B

As shown earlier and demonstrated in several previous studies [12,39,46,58], the SWMCS event consisted of scattered convection over a wide area surrounding southwestern Taiwan and the upstream ocean. The S-Pol observations selected for analysis in [46] were at 0507 UTC 16 June, as depicted in Figure 10. At this time, a roughly north–south-oriented rainband developed over the ocean southwest of Taiwan and the S-Pol site (Figure 10a), along a confluent zone between the southerly flow and southwesterly flow at low levels (near 2 km, Figure 10b). This rainband was determined to move at cx = 6.6 m s−1 and cy = 8.8 m s−1 [46]. Further up, at 5 km, the updrafts embedded in the band typically exhibited stronger upward motion (Figure 10c).
The east–west vertical cross-section through the convective cell near the southern end of this band (Figure 11a, also from [46]) reveals that the updraft core reached a peak ascending velocity of >8 m s−1 near 5–6 km where the reflectivity was about 50 dBZ, while the maximum reflectivity was >55 dBZ but located below the core at lower levels (below about 3 km). Compared to event A (Figure 8a), this updraft in event B is weaker to some extent. Widespread reflectivity of 30–40 dBZ can also be seen near and below the freezing level (at 5 km), mostly from stratiform clouds, both east and west of the main updraft (Figure 11a). In the latter region, both moderate system-relative inflow (from the west) below 5 km and outflow near 8–15 km, but strongest near 10–11 km in height, can be observed.

4.3. Model Results of Event B and Comparison with Radar Observations

In experiment R1.0B, a system along the low-level confluent zone similar to the one depicted in Figure 10a–c is reproduced at 0930 UTC (Figure 10d), but it also appears wider than the observation. With a motion vector of cx = 5.6 and cy = 11.0 m s−1, the structure of this rainband in the model at 2 and 5 km also resembles the observed one (Figure 10e,f), except that most of the mid-level flow at 5 km is RTF in the model (Figure 10f), but FTR in Figure 10c.
Figure 11c presents the east–west cross-section through a similar cell near the southern end of the rainband at 0930 UTC in R1.0B, and it can be compared with Figure 11a. While their general characteristics are quite similar, some minor differences can still be seen. First, the model updraft reaches only about 8.5 km in height, compared to 12 km in the S-Pol observations, although the peak velocity of both is around 8.5 m s−1. In fact, in the entire cross-section, the echo areas in R1.0B do not extend upward to levels as high as those in the observation. Second, the strong-echo area in the model also appears wider and can reach higher reflectivity values below the freezing level, with a maximum of over 55 dBZ. Third, to the east of the rainband, the FTR inflow below 2 km and RTF outflow near 6 km in Figure 11c are not evident in Figure 11a. On the other hand, west of the band, the RTF inflow at 2–5 km and FTR outflow aloft above 9 km are both well reproduced in R1.0B, but they are both somewhat weaker than the observed ones.
The PID results from the S-Pol along the cross-section and the mixing ratios of the hydrometeors in R1.0B are shown in Figure 11b and 11d, respectively. Again, below the freezing level, liquid water (rain) was seen in the observation, with the heaviest rainfall below the updraft core, and this is well reproduced in the model. Outside of the main updraft, the rainfall was mostly moderate and light in both the model and observation. At 4–5 km, just below the freezing level, the S-Pol results also indicate some heavier rainfall (red, Figure 11b), but this is not readily identifiable in the model results, although the mixing ratios of rain do appear to decrease with height in some regions just below the freezing level, which is also near 5 km. Above the freezing level, snow typically appeared up to about 9 km while ice particles were higher up (9–15 km), except inside the main updraft (where snow could reach higher levels, Figure 11b). These characteristics are also reproduced by the model reasonably well (Figure 11d). Near and above the updraft maximum, graupels are produced with higher mixing ratios (around 2.5 g kg−1) near 5–8 km in R1.0B, also in agreement with the observation, while their appearance in other regions near the freezing level and slightly above, mostly coupled with snow, likely corresponds to WS in the observation. In general, most of the hydrometeors are distributed in the model (Figure 11d) in a similar way compared to the observations (Figure 11c) in event B.

5. The Frontal Convection in IOP-3

5.1. Synoptic Overview of Event C

The IOP-3 of SoWMEX occurred on 29–31 May 2008, when a Mei-yu front approached and passed through Taiwan. In the later part of IOP-3, the convection on 31 May was studied by [47] using the S-Pol data, and this is our target for comparison in event C. The synoptic condition of this event is first discussed here. The surface weather map at 0000 UTC 31 May (Figure 12a) shows that the Mei-yu front was moving slowly across southern Taiwan, pushed by the northerly flow behind the front driven by the two anticyclones near 30°N. The surface front was associated with a wind-shift line at 850 hPa, with a cyclonic vortex and larger temperature gradient over northern to central Taiwan (Figure 12b). Further aloft, at 700 hPa, the prevailing wind surrounding Taiwan was southwesterly to westerly and the region was very humid, with relative humidity reaching 90% (Figure 12c). At 500 hPa, westerly winds prevailed, with small relative vorticity near southwestern Taiwan (Figure 12d). The presence of the front and high moisture content provided favorable conditions for local heavy rainfall on 31 May [59], and event C developed subsequently.

5.2. Mesoscale Aspects of Event C

Caused by the frontal convection, the rainfall in Taiwan on 30 May was mostly along the coastal plains from northern to central Taiwan (in LST, Figure 6c). Afterward, an MCS developed near the coast of central Taiwan and moved eastward, and then dissipated after 2200 UTC 30 (i.e., 0600 UTC 31) May (Figure 13a). A new cell developed at 0000 UTC 31 May, again near the coast of central Taiwan (Figure 13b). During the subsequent hours, it grew upscale into an MCS and moved toward the southeast into the mountain regions (Figure 13c–f). At 0800 UTC, the coldest clouds were over the southern CMR. Later, this MCS continued to move south, toward the southern tip of Taiwan (Figure 13g,h), before eventually weakening and dissipating (not shown). Thus, caused by this MCS near the Mei-yu front, the rainfall in Taiwan on 31 May (in LST) was mostly in the mountain regions in southern Taiwan, with a peak amount of over 110 mm (Figure 6d).

5.3. S-Pol Radar Observations of Event C

The S-Pol observations of reflectivity in southern Taiwan at selected times from 0752 to 0845 UTC are depicted in Figure 14a–e, adapted from [47]. During this period, an arc-shaped rainband was developing near the foothills of the southern CMR, some 20 km to the northeast of the S-Pol. Inside this band, cell mergers were observed. Another east–west band appeared just to the north of the first band, and the two bands gradually evolved into an elliptical shape at 0815 UTC (Figure 14d). At 0845 UTC (Figure 14e), the two bands became parallel and about 10–15 km apart, and the RHI from the S-Pol was chosen and is shown in Figure 15a,b (source: [47]). In Figure 15a, the main updraft roughly 6–8 km in width was located right at the foothills of the CMR (see Figure 14e) and about 22 km from the S-Pol, and a new updraft to its southwest could also be seen closer to the site, about 15–16 km away. In the RHI, the updrafts slanted toward the northeast with height. While the echoes below the freezing level were in general stronger than above, peak reflectivity greater than 55 dBZ associated with the main updraft was found aloft, near 5–7 km (Figure 15a), a characteristic different from that in both events A and B. The echoes of 25 dBZ could reach 14 km above the main updraft, but generally below 9 km elsewhere.

5.4. Model Results and Comparison with Radar Observations

The winds at 3 km and retrieved reflectivity from the model run in R1.0C (Table 2) around 0500 UTC 31 May are displayed in Figure 14f–j. While this time period is about 3 h earlier than the observation, similar convection can be seen to develop near the coastline of southwestern Taiwan with mergers taking place, but again with less detail and at locations slightly to the south compared to the observation. Embedded in the westerly flow at low levels, the main cell moved close to the foothills of the southern CMR at 0515 and 0530 UTC after merging (Figure 14i,j). For this latter time, a vertical cross-section at a similar orientation (running southwest–northeast) is constructed (Figure 15c), and shows features comparable to the observations. Among such features, the main updraft also tilts toward the northeast with height, and the echoes ≥ 25 and 50 dBZ associated with it can reach about 13 and 5 km in height, respectively. While echoes ≥ 40 dBZ are typically limited to below 6 km in Figure 15c, the strongest echoes, reaching 60 dZB, are found at low levels near 1–2 km and against the terrain. Thus, they are not elevated aloft as observed by the S-Pol. Near 12 km, the upward velocity in the model updraft can reach almost 20 m s−1, and this does not seem to be supported by the observation either.
The PID results from S-Pol in the RHI [47] are shown in Figure 15b, and can be compared with the model outputs in Figure 15d. Likely due to the high moisture content in event C, heavy rain was observed to occupy a wide area of 7–8 km below the melting level, which was close to 5 km, in association with the main updraft, but was quite narrow accompanying the new cell (Figure 15b). Between the cells, rainfall was mostly moderate, with light rain farther away. Above the melting level, the main updraft was associated with graupel up to about 9 km, then snow and ice crystals further up and away. In Figure 15b, one interesting and unique feature that was not seen in events A or B is the downward extension of mixed graupel with raindrops (GR) from the elevated area of peak reflectivity along the southwestern edge of the updraft, almost all the way down to the surface (green pixels). Near the melting level, WS was also found outside the updrafts.
When Figure 15b,d are compared, one can see that most of the main features in the PID results mentioned above, with liquid phase below and solid phase above the melting level, are reasonably reproduced in R1.0C for event C. Heavy rainfall (high mixing ratio of rain) is also found produced by the main updraft, and the graupel is also seen to extend downward at the southwestern edge of the heavy rainfall area (Figure 15d), although perhaps not as apparent as in the observation. Away from the main updraft, the coexistence of snow (blue) and rain (warm color) near the melting level also corresponds to WS. Above the melting level, graupel can be produced at levels as high as 14 km by the main updraft, and such excessive production is linked to the overly strong updraft at 9–15 km in the model.

6. Conclusions and Summary

In this study, three heavy-rainfall events of different types in Taiwan’s Mei-yu season, all during the field campaign of SoWMEX/TiMREX in 2008, were selected for simulations using the CReSS model at a grid size (Δx) of 1 km, to assess the model’s ability to reproduce the kinematic structure and cloud microphysical characteristics. Event A is a prefrontal squall line (PFSL) on 14 June and event B is an MCS embedded in prevailing southwesterly flow (SWMCS) on 16 June, both during IOP-8 and over the ocean upstream from Taiwan. On the other hand, a frontal convection with cell mergers, event C occurred near the foothills of the SMR earlier, on 31 May, during IOP-3. During SoWMEX/TiMREX, the NCAR S-Pol was deployed near the coast of southwestern Taiwan, and detailed analyses of its high-resolution observations were carried out for all three events in previous studies. Thus, the present study could take advantage of the S-Pol observations to compare with the model outputs, with particular emphasis on the derived particle identification (PID) results.
At Δx = 1 km, the three simulations could reproduce the rainfall systems of interest reasonably well at comparable locations and times, and with morphology and characteristics similar to the observed ones, except that the convective cells still appear coarser and wider in the model. In each case, the convective updraft reached the maximum speed near 5–8 k and could extend well into the upper troposphere, around 12–14 km in height. The north–south-oriented PFSL in event A was also associated with FTR inflow at low levels, RTF outflow aloft on the front (eastern) side, and rear inflow near 5 km on the back side. Most of the above-mentioned features were also reproduced by the model, except that the rear inflow in the PFSL appears near 7–8 km and higher up. In the observations, the peak radar reflectivity was typically > 50 dBZ and appeared below the melting level, except in event C where an elevated echo maximum was observed near 5–7 km and slightly above the melting level. These characteristics were also captured by the model, albeit with a tendency toward higher peak reflectivity values (by about 5 dBZ). In addition, the elevated echo maximum in event C was not captured.
Comparisons between the S-Pol PID results and the model-simulated mixing ratios of hydrometeors using the double-moment cold-rain scheme in cross-sections were also performed, although the latter have five such species (cloud liquid, rain, cloud ice, snow, and graupel), and fewer than the S-Pol. In all three events, heavier rainfall appeared below the core of the main updraft with moderate and lighter rainfall farther away below the melting level (near 5 km in height). Above the melting level, graupels and mixed-phase particles were produced by the updraft and could extend into the upper troposphere, with snow and then ice crystals farther away and higher up (to above 14 km), including stratiform regions. Such general distributions of particles were also reasonably captured in all three model runs, where the existence of rain and snow near the melting level corresponds to wet snow in the S-Pol PIDs. In event B, the graupel region in the model appears too big, likely due to downwind advection by the more evident RTF outflow (westerly winds) at upper levels. In event C, graupels also extended to lower levels after formation in the observation, and this feature was also captured by the model to some extent.
It is perhaps worth mentioning that the three events were not captured to the same degree of realism in R2.5AB and R2.5C with a Δx of 2.5 km. More detailed structures in events A and B, and in particular, the formation of event C were obtained only when the grid size was further reduced to 1 km and the outputs from the 2.5 km experiments were used as IC/BCs. As the model results still appear coarser than the S-Pol observation, even a Δx of 1 km may not be sufficient for some cases, e.g., [6], and further improvement might be attainable with even higher resolutions. Nonetheless, based on such results, IC/BCs with high spatiotemporal resolution and a finer model grid may be the key to model performance, at least for convective systems in Mei-yu seasons in Taiwan. With the continuous deployment of more rainfall radars in Taiwan in recent years, more detailed observations of cloud microphysics will become available. Thus, more studies that assess the model’s ability to simulate convective clouds in Taiwan, such as the present one, can be anticipated.

Author Contributions

Conceptualization, C.-C.W.; formal analysis, C.-C.W., Y.-H.C., Y.-Y.L. and W.-Y.C.; funding acquisition, C.-C.W.; investigation, C.-C.W., Y.-Y.L. and W.-Y.C.; data curation, Y.-Y.L. and W.-Y.C.; methodology, Y.-Y.L. and C.-C.W.; project administration, C.-C.W.; software, Y.-Y.L.; supervision, C.-C.W.; visualization, Y.-Y.L. and Y.-H.C.; writing—original draft, C.-C.W. and Y.-H.C.; writing—review and editing, all authors. All authors have read and agreed to the published version of the manuscript.

Funding

This study was supported by the National Science and Technology Council (NSTC) of Taiwan under grants MOST 111-2111-M-003-005, MOST 111-2625-M-003-001, and NSTC 112-2111-M-003-005.

Data Availability Statement

The CReSS model and its user’s guide are open to researchers and available at http://www.rain.hyarc.nagoya-u.ac.jp/~tsuboki/cress_html/index_cress_eng.html (accessed on 1 February 2019). The NCEP GFS analysis and forecast data are available at http://rda.ucar.edu/datasets/ds335.0/#!description (accessed on 14 September 2018).

Acknowledgments

The CWB of Taiwan is acknowledged for providing most of the observational data used in this study, and the original plots used in Figure 2, Figure 5a–f, Figure 6, Figure 9a–f, Figure 12a and Figure 13. Thanks also go to the SoWMEX/TiMREX project and NCAR S-Pol team, as well as [46] for providing Figure 7a–c, Figure 8a,b, Figure 10a–c and Figure 11a,b, and [47] for providing Figure 14a–e and Figure 15a,b.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Chen, G.T.-J. Research on the phenomena of Meiyu during the past quarter century: An overview. In East Asian Monsoon; Chang, C.-P., Ed.; World Scientific Series on Asia-Pacific Weather and Climate: Hackensack, NJ, USA, 2004; Volume 3, pp. 357–403. [Google Scholar]
  2. Anthes, R.A. A cumulus parameterization scheme utilizing a one-dimensional cloud model. Mon. Weather Rev. 1977, 105, 270–286. [Google Scholar] [CrossRef]
  3. Anthes, R.A.; Warner, T.T. Development of hydrodynamic models suitable for air pollution and other meteorological studies. Mon. Weather Rev. 1978, 106, 1045–1078. [Google Scholar] [CrossRef]
  4. Molinari, J.; Dudek, M. Parameterization of convective precipitation in mesoscale numerical models. A critical review. Mon. Weather Rev. 1992, 120, 326–344. [Google Scholar] [CrossRef]
  5. Tsuboki, K.; Sakakibara, A. Large-scale parallel computing of Cloud Resolving Storm Simulator. In High Performance Computing; Zima, H.P., Joe, K., Sato, M., Seo, Y., Shimasaki, M., Eds.; Springer: Berlin/Heidelberg, Germany, 2002; pp. 243–259. [Google Scholar]
  6. Bryan, G.H.; Wyngaard, J.C.; Fritsch, J.M. Resolution requirements for the simulation of deep moist convection. Mon. Weather Rev. 2003, 131, 2394–2416. [Google Scholar] [CrossRef]
  7. Skamarock, W.C.; Klemp, J.B.; Dudhia, J.; Gill, D.O.; Barker, D.M.; Wang, W.; Powers, J.G. A Description of the Advanced Research WRF Version 2; University Corporation for Atmospheric Research: Boulder, CO, USA, 2005; 100p. [Google Scholar]
  8. Lin, Y.-L.; Farley, R.D.; Orville, H.D. Bulk parameterization of the snow field in a cloud model. J. Appl. Meteorol. Climatol. 1983, 22, 1065–1092. [Google Scholar] [CrossRef]
  9. Straka, J.M.; Mansell, E.R. A bulk microphysics parameterization with multiple ice precipitation categories. J. Appl. Meteorol. Climatol. 2005, 44, 445–466. [Google Scholar] [CrossRef]
  10. Gentry, M.S.; Lackmann, G.M. Sensitivity of simulated tropical cyclone structure and intensity to horizontal resolution. Mon. Weather Rev. 2010, 138, 688–704. [Google Scholar] [CrossRef]
  11. Wang, C.-C.; Kuo, H.-C.; Chen, Y.-H.; Huang, H.-L.; Chung, C.-H.; Tsuboki, K. Effects of asymmetric latent heating on typhoon movement crossing Taiwan: The case of Morakot (2009) with extreme rainfall. J. Atmos. Sci. 2012, 69, 3172–3196. [Google Scholar] [CrossRef]
  12. Wang, C.-C.; Hsu, J.C.-S.; Chen, G.T.-J.; Lee, D.-I. A study of two propagating heavy-rainfall episodes near Taiwan during SoWMEX/TiMREX IOP-8 in June 2008. Part I: Synoptic evolution, episode propagation, and model control simulation. Mon. Weather Rev. 2014, 142, 2619–2643. [Google Scholar] [CrossRef]
  13. Luo, X.; Fei, J.; Huang, X.; Ding, J.; Ma, Z. Relative roles of dry intrusion, latent heat and instabilities in the Mei-yu rainband life cycle: A case study. Atmos. Res. 2018, 214, 10–20. [Google Scholar] [CrossRef]
  14. Knupp, K.R.; Cotton, W.R. An Intense, Quasi-steady thunderstorm over mountainous terrain. Part II: Doppler radar observations of the storm morphological structure. J. Atmos. Sci. 1982, 39, 343–358. [Google Scholar] [CrossRef]
  15. Rutledge, S.A.; Hobbs, P.V. The mesoscale and microscale structure and organization of clouds and precipitation in mid latitude clouds. Part XII: A diagnostic modeling study of precipitation development in narrow cold frontal rainbands. J. Atmos. Sci. 1984, 41, 2949–2972. [Google Scholar] [CrossRef]
  16. Cotton, W.R.; Tripoli, G.J.; Rauber, R.M.; Mulvihill, E.A. Numerical simulation of the effects of varying ice crystal nucleation rates and aggregation processes on orographic snowfall. J. Appl. Meteorol. Climatol. 1986, 25, 1658–1680. [Google Scholar] [CrossRef]
  17. Yoshizaki, M. Numerical simulations of tropical squall-line clusters: Two-dimensional model. J. Meteor. Soc. Jpn. Ser. II 1986, 64, 469–491. [Google Scholar] [CrossRef]
  18. Fovell, R.G.; Ogura, Y. Numerical simulation of a midlatitude squall line in two dimensions. J. Atmos. Sci. 1988, 45, 3846–3879. [Google Scholar] [CrossRef]
  19. Tao, W.-K.; Simpson, J. Modeling study of a tropical squall-type convective line. J. Atmos. Sci. 1989, 46, 177–202. [Google Scholar] [CrossRef]
  20. Wang, Y. An explicit simulation of tropical cyclones with a triply nested movable mesh primitive equation model: TCM3. Part II: Model refinements and sensitivity to cloud microphysics parameterization. Mon. Weather Rev. 2002, 130, 3022–3036. [Google Scholar] [CrossRef]
  21. Zhu, T.; Zhang, D.-L. Numerical simulation of Hurricane Bonnie (1998). Part II: Sensitivity to varying cloud microphysical processes. J. Atmos. Sci. 2006, 63, 109–126. [Google Scholar] [CrossRef]
  22. Tao, W.-K.; Scale, J.R.; Ferrier, B.; Simpson, J. The effect of melting processes on the development of a tropical and a midlatitude squall line. J. Atmos. Sci. 1995, 52, 1934–1948. [Google Scholar] [CrossRef]
  23. Vivekanandan, J.; Ellis, S.M.; Oye, R.; Zrnic, D.S.; Ryzhkov, A.V.; Straka, J. Cloud microphysics retrieval using S-band dual-polarization radar measurements. Bull. Am. Meteorol. Soc. 1999, 80, 381–388. [Google Scholar] [CrossRef]
  24. Aydin, K.; Seliga, T.A.; Balaji, V. Remote sensing of hail with a dual linear polarization radar. J. Appl. Meteorol. Climatol. 1986, 25, 1475–1484. [Google Scholar] [CrossRef]
  25. Doviak, R.J.; Zrnic, D.S. Doppler Radar and Weather Observations; Academic Press: Cambridge, MA, USA, 1993; p. 562. [Google Scholar]
  26. Straka, J.M.; Zrnic, D.S. An algorithm to deduce hydrometeor types and contents from multiparameter radar data. Preprints. In Proceedings of the 26th International Conference on Radar Meteorology, Norman, OK, USA, 24–28 May 1993; American Meteorological Society: Boston, MA, USA, 1993; pp. 513–515. [Google Scholar]
  27. Vivekanandan, J.; Bringi, V.N.; Raghavan, R. Multiparameter radar modeling and observations of melting ice. J. Atmos. Sci. 1990, 47, 549–564. [Google Scholar] [CrossRef]
  28. Zrnic, D.S.; Balakrishnan, N.; Ziegler, C.L.; Bringi, V.N.; Aydin, K.; Matejka, T. Polarimetric signatures in the stratiform region of a mesoscale convective system. J. Appl. Meteor. 1993, 32, 678–693. [Google Scholar] [CrossRef]
  29. Vivekanandan, J.; Bringi, V.N.; Hagen, M.; Meischner, P. Polarimetric radar studies of atmospheric ice particles. IEEE Trans. Geosci. Remote Sens. 1994, 32, 1–10. [Google Scholar] [CrossRef]
  30. Aydin, K.V.; Bringi, N.; Liu, L. Rain-rate estimation in the presence of hail using S-band specific differential phase and other radar parameters. J. Appl. Meteorol. Climatol. 1995, 34, 404–410. [Google Scholar] [CrossRef]
  31. Hall, M.P.M.; Goddard, J.W.F.; Cherry, S.M. Identification of hydrometeors and other targets by dual-polarization radar. Radio Sci. 1984, 19, 132–140. [Google Scholar] [CrossRef]
  32. Hendry, A.; Antar, Y.M.M. Precipitation particle identification with centimeter wavelength dual-polarization radar. Radio Sci. 1984, 19, 115–122. [Google Scholar] [CrossRef]
  33. Lopez, R.E.; Aubagnac, J.P. The lightning activity of a hailstorm as a function of changes in its microphysical characteristics inferred from polarimetric radar observations. J. Geophys. Res. 1997, 102, 16799–16813. [Google Scholar] [CrossRef]
  34. May, P.T.; Keenan, T.D. Four-dimensional microphysical data from Darwin. In Proceedings of the 13th ARM Science Team Meeting Proceedings, Broomfield, CO, USA, 31 March–4 April 2003. [Google Scholar]
  35. Jou, B.J.-D.; Lee, W.-C.; Johnson, R.H. An overview of SoWMEX/TiMREX. In The Global Monsoon System: Research and Forecast, 2nd ed.; Chang, C.-P., Ding, Y., Lau, N.-C., Johnson, R.H., Wang, B., Eds.; World Scientific Series on Asia-Pacific Weather and Climate: Hackensack, NJ, USA, 2011; Volume 5, pp. 303–318. [Google Scholar]
  36. Zhou, J.-Y. Observational Analysis and Numerical Simulation of SoWMEX (IOP-8): The Characteristics of the Microphysical Structure and Compared to Microphysics Schemes. Master’s Thesis, Dept. of Atmospheric Sciences, National Central Univ., Taoyuan, Taiwan, 2012. (In Chinese with English abstract). [Google Scholar]
  37. Tsuboki, K.; Sakakibara, A. Numerical Prediction of High-Impact Weather Systems: The Textbook for the Seventeenth IHP Training Course in 2007; Hydrospheric Atmospheric Research Center, Nagoya University, and UNESCO: Nagoya, Japan, 2007; 273p. [Google Scholar]
  38. Wang, C.-C.; Chen, Y.-H.; Li, M.-C.; Kuo, H.-C.; Tsuboki, K. On the separation of upper and low-level centres of tropical storm Kong-Rey (2013) near Taiwan in association with asymmetric latent heating. Q. J. R. Meteorol. Soc. 2021, 147, 1135–1149. [Google Scholar] [CrossRef]
  39. Wang, C.-C.; Hsu, J.C.-S.; Chen, G.T.-J.; Lee, D.-I. A study of two propagating heavy-rainfall episodes near Taiwan during SoWMEX/TiMREX IOP-8 in June 2008. Part II: Sensitivity tests on the roles of synoptic conditions and topographic effects. Mon. Weather Rev. 2014, 142, 2644–2664. [Google Scholar] [CrossRef]
  40. Wang, C.-C.; Chiou, B.-K.; Chen, G.T.-J.; Kuo, H.-C.; Liu, C.-H. A numerical study of back-building process in a quasistationary rainband with extreme rainfall over northern Taiwan during 11–12 June 2012. Atmos. Chem. Phys. 2016, 16, 12359–12382. [Google Scholar] [CrossRef]
  41. Wang, C.-C.; Kuo, H.-C.; Yeh, T.-C.; Chung, C.-H.; Chen, Y.-H.; Huang, S.-Y.; Wang, Y.-W.; Liu, C.-H. High-resolution quantitative precipitation forecasts and simulations by the Cloud-Resolving Storm Simulator (CReSS) for Typhoon Morakot (2009). J. Hydrol. 2013, 506, 26–41. [Google Scholar] [CrossRef]
  42. Wang, C.-C.; Li, M.-S.; Chang, C.-S.; Chuang, P.-Y.; Chen, S.-H.; Tsuboki, K. Ensemble-based sensitivity analysis and predictability of an extreme rainfall event over northern Taiwan in the Mei-Yu season: The 2 June 2017 case. Atmos. Res. 2021, 259, 105684. [Google Scholar] [CrossRef]
  43. Wang, C.-C.; Yeh, T.-Y.; Chang, C.-S.; Li, M.-S.; Tsuboki, K.; Liu, C.-H. A modeling study of an extreme rainfall event along the northern coast of Taiwan on 2 June 2017. Atmos. Chem. Phys. 2023, 23, 501–521. [Google Scholar] [CrossRef]
  44. Wang, C.-C.; Tsai, C.-H.; Jou, B.J.-D.; David, S.J. Time-lagged ensemble quantitative precipitation forecasts for three landfalling typhoons in the Philippines using the CReSS Model, Part I: Description and verification against rain-gauge observations. Atmosphere 2022, 13, 1193. [Google Scholar] [CrossRef]
  45. Murakami, M.; Clark, T.L.; Hall, W.D. Numerical simulations of convective snow clouds over the sea of Japan; Two-dimensional simulations of mixed layer development and convective snow cloud formation. J. Meteor. Soc. Japan 1994, 72, 43–62. [Google Scholar] [CrossRef]
  46. Chang, W.-Y.; Lee, W.-C.; Liou, Y.-C. The kinematic and microphysical characteristics and associated precipitation efficiency of subtropical convection during SoWMEX/TiMREX. Mon. Weather Rev. 2015, 143, 317–340. [Google Scholar] [CrossRef]
  47. Tseng, Y.-S. Frontal Convection Analysis Using SPOL Radar during SoWMEX. Master’s Thesis, Department of Environmental Information and Engineering, Chung-Cheng Institute Technology, National Defense University, Taoyuan, Taiwan, 2013. [Google Scholar]
  48. Murakami, M. Numerical modeling of dynamical and microphysical evolution of an isolated convective cloud: The 19 July 1981 CCOPE cloud. J. Meteor. Soc. Jpn. 1990, 68, 107–128. [Google Scholar] [CrossRef]
  49. Ikawa, M.; Saito, K. Description of a nonhydrostatic model developed at the forecast research department of the MRI. Tech. Rep. MRI 1991, 28, 238. [Google Scholar]
  50. Deardorff, J.W. Stratocumulus-capped mixed layers derived from a three-dimensional model. Bound.-Layer Meteor. 1980, 18, 495–527. [Google Scholar] [CrossRef]
  51. Kondo, J. Heat balance of the China Sea during the air mass transformation experiment. J. Meteor. Soc. Jpn. 1976, 54, 382–398. [Google Scholar] [CrossRef]
  52. Louis, J.F.; Tiedtke, M.; Geleyn, J.F. A Short History of the Operational PBL Parameterization at ECMWF. In Workshop on Planetary Boundary Layer Parameterization; ECMWF: Reading, UK, 1982; pp. 59–79. [Google Scholar]
  53. Segami, A.; Kurihara, K.; Nakamura, H.; Ueno, M.; Takano, I.; Tatsumi, Y. Operational mesoscale weather prediction with Japan spectral model. J. Meteor. Soc. Jpn. 1989, 67, 907–924. [Google Scholar] [CrossRef]
  54. Wang, C.-C.; Huang, S.-Y.; Chen, S.-H.; Chang, C.-S.; Tsuboki, K. Cloud-resolving typhoon rainfall ensemble forecasts for Taiwan with large domain and extended range through time-lagged approach. Wea. Forecast. 2016, 31, 151–172. [Google Scholar] [CrossRef]
  55. Wang, C.-C.; Chang, C.-S.; Wang, Y.-W.; Huang, C.-C.; Wang, S.-C.; Chen, Y.-S.; Tsuboki, K.; Huang, S.-Y.; Chen, S.-H.; Chuang, P.-Y.; et al. Evaluating quantitative precipitation forecasts using the 2.5 km CReSS model for typhoons in Taiwan: An update through the 2015 season. Atmosphere 2021, 12, 1501. [Google Scholar] [CrossRef]
  56. Wang, C.-C.; Chuang, P.-Y.; Chang, C.-S.; Tsuboki, K.; Huang, S.-Y.; Leu, G.-C. Evaluation of Mei-yu heavy-rainfall quantitative precipitation forecasts in Taiwan by a cloud-resolving model for three seasons of 2012-2014. Nat. Hazards Earth Syst. Sci. 2022, 22, 23–40. [Google Scholar] [CrossRef]
  57. Chen, X.A.; Chen, Y.-L. Development of low-level jets during TAMEX. Mon. Weather Rev. 1995, 123, 1695–1719. [Google Scholar] [CrossRef]
  58. Xu, W.; Zipser, E.J.; Chen, Y.-L.; Liu, C.; Liou, Y.-C.; Lee, W.-C.; Jou, B.J.-D. An orography-associated extreme rainfall event during TiMREX: Initiation, storm evolution, and maintenance. Mon. Weather Rev. 2012, 140, 2555–2574. [Google Scholar] [CrossRef]
  59. 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]
  60. Tu, C.-C.; Chen, Y.-L.; Chen, C.-S.; Lin, P.-L.; Lin, P.-H. A comparison of two heavy rainfall events during the terrain-influenced monsoon rainfall experiment (TiMREX) 2008. Mon. Weather Rev. 2014, 142, 2436–2463. [Google Scholar] [CrossRef]
Remotesensing 15 04651 g001
Figure 1. The topography of Taiwan and the surrounding regions (m, color) and the model domains used for the three events (A, B, and C) in this study (see text for details). Dashed and dotted areas depict 2.5 and 1 km domains for events A and B (red; R2.5AB and R1.0A/R1.0B) and event C (blue; R2.5C and R1.0C), respectively.
Figure 1. The topography of Taiwan and the surrounding regions (m, color) and the model domains used for the three events (A, B, and C) in this study (see text for details). Dashed and dotted areas depict 2.5 and 1 km domains for events A and B (red; R2.5AB and R1.0A/R1.0B) and event C (blue; R2.5C and R1.0C), respectively.
Remotesensing 15 04651 g001
Remotesensing 15 04651 g002
Figure 2. The CWB surface weather maps at 0000 UTC of (a) 13, (b) 14, (c) 15, and (d) 16 June 2008, respectively. Parameters shown include mean sea-level pressure (hPa, isobars every 4 hPa), surface front, and high-/low-pressure systems with station plots. (source: CWB.).
Figure 2. The CWB surface weather maps at 0000 UTC of (a) 13, (b) 14, (c) 15, and (d) 16 June 2008, respectively. Parameters shown include mean sea-level pressure (hPa, isobars every 4 hPa), surface front, and high-/low-pressure systems with station plots. (source: CWB.).
Remotesensing 15 04651 g002
Remotesensing 15 04651 g003
Figure 3. NCEP (1° × 1°) 850 hPa final analyses of geopotential height (gpm, black contours every 10 gpm) and horizontal wind (kt, barbs, one full/half barb = 10/5 kt with isotachs (m s−1, color, scale at bottom) at 0000 UTC on (a) 14 and (b) 16 Jun 2008, respectively. An additional isotach at 12 m s−1 is also plotted in red.
Figure 3. NCEP (1° × 1°) 850 hPa final analyses of geopotential height (gpm, black contours every 10 gpm) and horizontal wind (kt, barbs, one full/half barb = 10/5 kt with isotachs (m s−1, color, scale at bottom) at 0000 UTC on (a) 14 and (b) 16 Jun 2008, respectively. An additional isotach at 12 m s−1 is also plotted in red.
Remotesensing 15 04651 g003
Remotesensing 15 04651 g004
Figure 4. As in Figure 3, except for geopotential height (gpm) and horizontal winds (kt, barbs, one full/half barb = 10/5 kt) without isotachs, plus relative vorticity (10−4 s−1, color, scale at bottom) at 0000 UTC on (a) 14 and (b) 16 Jun 2008, respectively. Troughs are depicted by thick dashed lines and the 5880 gpm isopleths are plotted in green.
Figure 4. As in Figure 3, except for geopotential height (gpm) and horizontal winds (kt, barbs, one full/half barb = 10/5 kt) without isotachs, plus relative vorticity (10−4 s−1, color, scale at bottom) at 0000 UTC on (a) 14 and (b) 16 Jun 2008, respectively. Troughs are depicted by thick dashed lines and the 5880 gpm isopleths are plotted in green.
Remotesensing 15 04651 g004
Remotesensing 15 04651 g005
Figure 5. (a–f) The CWB radar reflectivity mosaic (dBZ, color, scale at bottom) and NCEP 800 hPa wind analysis (m s−1, vectors) at (a) 2000 UTC 13 June, and (b) 0100, (c) 0900, (d) 1100, (e) 1300, and (f) 1600 UTC 14 Jun 2008, respectively. (g–i) As in (a–f) except from model-simulated reflectivity (scale on the right) and wind vectors at 2 km height at (g) 0930, (h) 1030, and (i) 1130 UTC 14 Jun in R1.0AB. Reference vectors are shown at the bottom.
Figure 5. (a–f) The CWB radar reflectivity mosaic (dBZ, color, scale at bottom) and NCEP 800 hPa wind analysis (m s−1, vectors) at (a) 2000 UTC 13 June, and (b) 0100, (c) 0900, (d) 1100, (e) 1300, and (f) 1600 UTC 14 Jun 2008, respectively. (g–i) As in (a–f) except from model-simulated reflectivity (scale on the right) and wind vectors at 2 km height at (g) 0930, (h) 1030, and (i) 1130 UTC 14 Jun in R1.0AB. Reference vectors are shown at the bottom.
Remotesensing 15 04651 g005
Remotesensing 15 04651 g006
Figure 6. Daily accumulated rainfall observation (mm, color) in Taiwan (in LST) on (a) 14 and (b) 16 Jun during IOP-8, and on (c) 30 and (d) 31 May during IOP-3 in 2008, respectively. In Taiwan, 0000 LST corresponds to 1600 UTC.
Figure 6. Daily accumulated rainfall observation (mm, color) in Taiwan (in LST) on (a) 14 and (b) 16 Jun during IOP-8, and on (c) 30 and (d) 31 May during IOP-3 in 2008, respectively. In Taiwan, 0000 LST corresponds to 1600 UTC.
Remotesensing 15 04651 g006
Remotesensing 15 04651 g007
Figure 7. (a) The reflectivity mosaic (dBZ) from S-Pol and winds at 2.5 km from radar syntheses (m s−1, blue vectors), and interpolated reflectivity (dBZ) and dual-Doppler-analyzed system-relative winds (m s−1, vectors) at (b) 2 and (c) 5 km at 1000 UTC 14 Jun 2008, respectively. In (b,c), red (blue) contours depict upward (downward) motion every 2 m s−1. Adapted from [46] and used with permission. In (a), the topography of 0.5 km is also plotted and the box depicts the regions shown in (b,c), where the dashed line AB shows the cross-section used in Figure 8a,b. (d–f) As in (a–c), except from model simulation and retrieved reflectivity in R1.0A at 1100 UTC 14 Jun 2008 at the same scale. In (d), the topography of 2.5 km is plotted (blue contour) instead. In (e,f), only upward motion is plotted (also every 2 m s−1) and the dashed line A′B′ shows the cross-section used in Figure 8c,d.
Figure 7. (a) The reflectivity mosaic (dBZ) from S-Pol and winds at 2.5 km from radar syntheses (m s−1, blue vectors), and interpolated reflectivity (dBZ) and dual-Doppler-analyzed system-relative winds (m s−1, vectors) at (b) 2 and (c) 5 km at 1000 UTC 14 Jun 2008, respectively. In (b,c), red (blue) contours depict upward (downward) motion every 2 m s−1. Adapted from [46] and used with permission. In (a), the topography of 0.5 km is also plotted and the box depicts the regions shown in (b,c), where the dashed line AB shows the cross-section used in Figure 8a,b. (d–f) As in (a–c), except from model simulation and retrieved reflectivity in R1.0A at 1100 UTC 14 Jun 2008 at the same scale. In (d), the topography of 2.5 km is plotted (blue contour) instead. In (e,f), only upward motion is plotted (also every 2 m s−1) and the dashed line A′B′ shows the cross-section used in Figure 8c,d.
Remotesensing 15 04651 g007
Remotesensing 15 04651 g008
Figure 8. East–west vertical cross-section of system-relative wind vectors and upward motion (contours, both in m s−1), plus (a) reflectivity (dBZ) and (b) PID results from S-Pol along the line AB (Figure 7b) at 1000 UTC 14 Jun 2008. Adapted from [46] and used with permission. The abbreviations for particle types are CL (cloud), DZ (drizzle), LR (light rain), MR (moderate rain), HR (heavy rain), HA (hail), RH (rain–hail mixture), GSH (graupel–small hail mixture), GR (graupel–rain mixture), DS (dry snow), WS (wet snow), IC (ice crystals), IIC (irregular ice crystals), SLD (supercooled liquid droplets), BGS (biological signals), ST (second trip), and GC (ground clutter), respectively. (c) As in (a), except from model simulation in R1.0A along the line A′B′ (Figure 7e) at 1100 UTC 14 Jun 2008. (d) As in (c), but showing the mixing ratios (g kg−1) of hydrometeors (color for rain, and green, blue, and pink contours for graupel, snow, and ice, respectively). All panels have the same scale.
Figure 8. East–west vertical cross-section of system-relative wind vectors and upward motion (contours, both in m s−1), plus (a) reflectivity (dBZ) and (b) PID results from S-Pol along the line AB (Figure 7b) at 1000 UTC 14 Jun 2008. Adapted from [46] and used with permission. The abbreviations for particle types are CL (cloud), DZ (drizzle), LR (light rain), MR (moderate rain), HR (heavy rain), HA (hail), RH (rain–hail mixture), GSH (graupel–small hail mixture), GR (graupel–rain mixture), DS (dry snow), WS (wet snow), IC (ice crystals), IIC (irregular ice crystals), SLD (supercooled liquid droplets), BGS (biological signals), ST (second trip), and GC (ground clutter), respectively. (c) As in (a), except from model simulation in R1.0A along the line A′B′ (Figure 7e) at 1100 UTC 14 Jun 2008. (d) As in (c), but showing the mixing ratios (g kg−1) of hydrometeors (color for rain, and green, blue, and pink contours for graupel, snow, and ice, respectively). All panels have the same scale.
Remotesensing 15 04651 g008
Remotesensing 15 04651 g009
Figure 9. As in Figure 5, except for (a–f) radar reflectivity mosaic (dBZ) and NCEP 800 hPa winds (m s−1) every 4 h from (a) 2000 UTC 15 June to (f) 1600 UTC 16 Jun 2008, respectively, and model-simulated reflectivity and winds at 2 km height at (g) 0400, (h) 0800, and (i) 1200 UTC 16 Jun in R1.0B. The areas of interest are circled.
Figure 9. As in Figure 5, except for (a–f) radar reflectivity mosaic (dBZ) and NCEP 800 hPa winds (m s−1) every 4 h from (a) 2000 UTC 15 June to (f) 1600 UTC 16 Jun 2008, respectively, and model-simulated reflectivity and winds at 2 km height at (g) 0400, (h) 0800, and (i) 1200 UTC 16 Jun in R1.0B. The areas of interest are circled.
Remotesensing 15 04651 g009
Remotesensing 15 04651 g010
Figure 10. (a–c) As in Figure 7a–c, except for reflectivity from S-Pol, radar-synthesized winds, and dual-Doppler-analyzed system-relative winds at 0507 UTC 16 Jun 2008. Adapted from [46] and used with permission. In (a), the box depicts the regions shown in (b,c), where the dashed line CD shows the cross-section used in Figure 11a,b. (d–f) As in Figure 7d–f, except from model results in R1.0B at 0930 UTC 16 Jun 2008. The dashed line C′D′ in (e,f) shows the cross-section used in Figure 11c,d.
Figure 10. (a–c) As in Figure 7a–c, except for reflectivity from S-Pol, radar-synthesized winds, and dual-Doppler-analyzed system-relative winds at 0507 UTC 16 Jun 2008. Adapted from [46] and used with permission. In (a), the box depicts the regions shown in (b,c), where the dashed line CD shows the cross-section used in Figure 11a,b. (d–f) As in Figure 7d–f, except from model results in R1.0B at 0930 UTC 16 Jun 2008. The dashed line C′D′ in (e,f) shows the cross-section used in Figure 11c,d.
Remotesensing 15 04651 g010
Remotesensing 15 04651 g011
Figure 11. As in Figure 8, except for east–west vertical cross-sections (a,b) from S-Pol along the line CD (Figure 10b) at 0507 UTC 16 Jun, and (c,d) from the model simulation in R1.0B along the line C′D′ (Figure 10e) at 0930 UTC 16 Jun 2008, respectively. (a,b) Adapted from [46] and used with permission.
Figure 11. As in Figure 8, except for east–west vertical cross-sections (a,b) from S-Pol along the line CD (Figure 10b) at 0507 UTC 16 Jun, and (c,d) from the model simulation in R1.0B along the line C′D′ (Figure 10e) at 0930 UTC 16 Jun 2008, respectively. (a,b) Adapted from [46] and used with permission.
Remotesensing 15 04651 g011
Remotesensing 15 04651 g012
Figure 12. (a) The CWB surface weather map (isobars every 4 hPa, surface front and high/low systems with station plots) and (b–d) NCEP (1° × 1°) final analyses of geopotential height (gpm, black contours every 10 gpm), horizontal winds (kt, barbs, one full/half barb = 10/5 kt), plus (b) temperature (°C, color) at 850 hPa, (c) relative humidity (%, color) at 700 hPa, and (d) relative vorticity (10−4 s−1, color) at 500 hPa, respectively, all at 0000 UTC 31 May 2008.
Figure 12. (a) The CWB surface weather map (isobars every 4 hPa, surface front and high/low systems with station plots) and (b–d) NCEP (1° × 1°) final analyses of geopotential height (gpm, black contours every 10 gpm), horizontal winds (kt, barbs, one full/half barb = 10/5 kt), plus (b) temperature (°C, color) at 850 hPa, (c) relative humidity (%, color) at 700 hPa, and (d) relative vorticity (10−4 s−1, color) at 500 hPa, respectively, all at 0000 UTC 31 May 2008.
Remotesensing 15 04651 g012
Remotesensing 15 04651 g013
Figure 13. Enhanced infrared cloud images (brightness temperature, °C, scale at bottom) surrounding Taiwan by the geostationary Multifunctional Transport Satellite (MTSAT) every 2 h from 2200 UTC 30 to 1200 UTC 31 May 2008 (a–h).
Figure 13. Enhanced infrared cloud images (brightness temperature, °C, scale at bottom) surrounding Taiwan by the geostationary Multifunctional Transport Satellite (MTSAT) every 2 h from 2200 UTC 30 to 1200 UTC 31 May 2008 (a–h).
Remotesensing 15 04651 g013
Remotesensing 15 04651 g014
Figure 14. (a–e) As in Figure 7a, except for reflectivity (dBZ) from S-Pol at the height of 3 km at (a) 0752, (b) 0800, (c) 0807, (d) 0815, and (e) 0845 UTC 31 May 2008, respectively. Adapted from [47] and used with permission. The location of S-Pol (red triangle) and convection of interest (circles) are marked. In (e), the dashed line EF shows the cross-section used in Figure 15a,b. (f–j) As in Figure 7d, except for reflectivity and wind vectors (m s−1) at a height of 3 km from model results in R1.0C every 15 min from (f) 0430 to (j) 0530 UTC 31 May 2008, respectively, at the same scale. The dashed line E′F′ in (j) shows the cross-section used in Figure 15c,d.
Figure 14. (a–e) As in Figure 7a, except for reflectivity (dBZ) from S-Pol at the height of 3 km at (a) 0752, (b) 0800, (c) 0807, (d) 0815, and (e) 0845 UTC 31 May 2008, respectively. Adapted from [47] and used with permission. The location of S-Pol (red triangle) and convection of interest (circles) are marked. In (e), the dashed line EF shows the cross-section used in Figure 15a,b. (f–j) As in Figure 7d, except for reflectivity and wind vectors (m s−1) at a height of 3 km from model results in R1.0C every 15 min from (f) 0430 to (j) 0530 UTC 31 May 2008, respectively, at the same scale. The dashed line E′F′ in (j) shows the cross-section used in Figure 15c,d.
Remotesensing 15 04651 g014
Remotesensing 15 04651 g015
Figure 15. Similar to Figure 8, except for the vertical cross-sections (a,b) from S-Pol along the line EF (Figure 14e) at 0844 UTC 31 May (without kinematic field), and (c,d) from the model simulation in R1.0C along the line E′F′ (Figure 14j) at 0530 UTC 31 May 2008, respectively. The dashed line depicts the 0 °C isotherm. (a,b) Adapted from [47] and used with permission. All panels have the same scale.
Figure 15. Similar to Figure 8, except for the vertical cross-sections (a,b) from S-Pol along the line EF (Figure 14e) at 0844 UTC 31 May (without kinematic field), and (c,d) from the model simulation in R1.0C along the line E′F′ (Figure 14j) at 0530 UTC 31 May 2008, respectively. The dashed line depicts the 0 °C isotherm. (a,b) Adapted from [47] and used with permission. All panels have the same scale.
Remotesensing 15 04651 g015
Table 1. Physical options and configurations of CReSS common to all experiments in this study.
Table 1. Physical options and configurations of CReSS common to all experiments in this study.
ProjectionLambert Conformal, Center at 120°E, Secant at 10°N and 40°N
Topography dataDigital elevation model at intervals of (1/120)°
Cloud microphysicsDouble-moment bulk cold-rain scheme
PBL turbulence1.5-order closure with turbulent kinetic energy prediction
Surface processesMomentum/energy fluxes and shortwave and longwave radiation
Substrate model43 levels, every 5 cm to 2.1 m in depth
Table 2. Domain configuration and experimental setup of the CReSS model for the three events in 2008 considered in this study.
Table 2. Domain configuration and experimental setup of the CReSS model for the three events in 2008 considered in this study.
Event NameEvent AEvent BEvent C
Experiment name (2.5 km)R2.5ABR2.5C
Grid spacing (km)2.5 × 2.5 × 0.5 *
Grid dimension (x × y × z)600 × 480 × 40585 × 520 × 60
Domain size (km)1500 × 1200 × 201455 × 1305 × 30
Initial time0000 UTC 12 June0000 UTC 30 May
Integration length and output frequency144 h, 30 min48 h, 15 min
IC/BCsNCEP GFS final analyses (1° × 1°, 26 levels, every 6 h)
Experiment name (1 km)R1.0AR1.0BR1.0C
Grid spacing (km)1.0 × 1.0 × 0.5 *
Grid dimension (x × y × z)900 × 720 × 37528 × 549 × 56
Domain size (km)900 × 720 × 18.5528 × 549 × 28
Initial time0000 UTC 13 June1800 UTC 15 June2100 UTC 30 May
Forecast range and interval48 h, 15 min22 h, 15 min18 h, 5 min
IC/BCsOutputs of R2.5ABOutputs of R2.5C
* The vertical grid spacing (Δz) of CReSS is stretched (smallest at the bottom), and the averaged spacing is given in parentheses.
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

Wang, C.-C.; Chen, Y.-H.; Lan, Y.-Y.; Chang, W.-Y. An Evaluation of Simulated Cloud Microphysical Characteristics of Three Mei-Yu Rainfall Systems in Taiwan by a Cloud-Resolving Model Using Dual-Polarimetric Radar Observations. Remote Sens. 2023, 15, 4651. https://doi.org/10.3390/rs15194651

AMA Style

Wang C-C, Chen Y-H, Lan Y-Y, Chang W-Y. An Evaluation of Simulated Cloud Microphysical Characteristics of Three Mei-Yu Rainfall Systems in Taiwan by a Cloud-Resolving Model Using Dual-Polarimetric Radar Observations. Remote Sensing. 2023; 15(19):4651. https://doi.org/10.3390/rs15194651

Chicago/Turabian Style

Wang, Chung-Chieh, Yu-Han Chen, Yu-Yao Lan, and Wei-Yu Chang. 2023. "An Evaluation of Simulated Cloud Microphysical Characteristics of Three Mei-Yu Rainfall Systems in Taiwan by a Cloud-Resolving Model Using Dual-Polarimetric Radar Observations" Remote Sensing 15, no. 19: 4651. https://doi.org/10.3390/rs15194651

APA Style

Wang, C. -C., Chen, Y. -H., Lan, Y. -Y., & Chang, W. -Y. (2023). An Evaluation of Simulated Cloud Microphysical Characteristics of Three Mei-Yu Rainfall Systems in Taiwan by a Cloud-Resolving Model Using Dual-Polarimetric Radar Observations. Remote Sensing, 15(19), 4651. https://doi.org/10.3390/rs15194651

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