Organizational Modes of Spring and Summer Convective Storms and Associated Severe Weather over Southern China during 2015–19

Chenbin Xue aKey Laboratory of Meteorological Disaster, Ministry of Education/Joint International Research Laboratory of Climate and Environment Change/Collaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters, Nanjing University of Information Science and Technology, Nanjing, China
bGuangdong Provincial Key Laboratory of Regional Numerical Weather Prediction, Institute of Tropical and Marine Meteorology, China Meteorological Administration, Guangzhou, China
cJiangxi Institute of Meteorological Sciences, Nanchang, China

Search for other papers by Chenbin Xue in
Current site
Google Scholar
PubMed
Close
,
Xinyong Shen aKey Laboratory of Meteorological Disaster, Ministry of Education/Joint International Research Laboratory of Climate and Environment Change/Collaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters, Nanjing University of Information Science and Technology, Nanjing, China
dSouthern Marine Science and Engineering Guangdong Laboratory (Zhuhai), Zhuhai, China

Search for other papers by Xinyong Shen in
Current site
Google Scholar
PubMed
Close
,
Zhiying Ding aKey Laboratory of Meteorological Disaster, Ministry of Education/Joint International Research Laboratory of Climate and Environment Change/Collaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters, Nanjing University of Information Science and Technology, Nanjing, China

Search for other papers by Zhiying Ding in
Current site
Google Scholar
PubMed
Close
,
Naigeng Wu aKey Laboratory of Meteorological Disaster, Ministry of Education/Joint International Research Laboratory of Climate and Environment Change/Collaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters, Nanjing University of Information Science and Technology, Nanjing, China
bGuangdong Provincial Key Laboratory of Regional Numerical Weather Prediction, Institute of Tropical and Marine Meteorology, China Meteorological Administration, Guangzhou, China

Search for other papers by Naigeng Wu in
Current site
Google Scholar
PubMed
Close
,
Yizhi Zhang cJiangxi Institute of Meteorological Sciences, Nanchang, China

Search for other papers by Yizhi Zhang in
Current site
Google Scholar
PubMed
Close
,
Xian Chen eJiangxi Institute of Land and Space Survey and Planning/Jiangxi Geomatics Center, Nanchang, China

Search for other papers by Xian Chen in
Current site
Google Scholar
PubMed
Close
, and
Chunyan Guo fInner Mongolia Meteorological Service Center, Hohhot, China

Search for other papers by Chunyan Guo in
Current site
Google Scholar
PubMed
Close
Free access

Abstract

This study investigates the organizational modes of convective storms and associated severe weather in spring and summer (March–August) of 2015–19 over southern China. These storms are classified into three major organizational structures (cellular, linear, and nonlinear), including 10 dominant morphologies. In general, cellular systems are most frequent, followed by linear systems. Convective storms are common in spring, increasing markedly from April to June, and peak in June. Convective storm cases are usually longer lived in spring, while shorter lived in summer. They also present pronounced diurnal variations, with a primary peak in the afternoon and several secondary peaks during the night to the morning. Approximately 79.7% of initial convection clearly exhibits a dominant eastward movement, with a faster moving speed in spring. Convective storms frequently evolve among organizational modes during their life spans. Linear systems produce the most severe weather observations, in which convective lines with trailing stratiform rain are most prolific. Bow echoes are most efficient in producing severe weather events among all systems, despite their rare occurrences. In spring, lines with parallel stratiform rain are abundant producers of severe wind events, ranking the second highest probability. In summer, embedded lines produce the second largest proportion of intense rainfall events, whereas lines with leading stratiform rain are most efficient in generating extremely intense rainfall and thus pose a distinct flooding threat. Broken lines produce the largest proportion of severe weather events among cellular storms. In contrast, nonlinear systems possess the weakest capability to produce severe weather events.

Significance Statement

Under the influence of the East Asian summer monsoon, severe weather events produced by convective storms occur frequently in China, leading to serious natural disasters. Numerous studies have demonstrated that the morphologies of convective storms are helpful to improve our understanding and prediction of convective storms. However, fewer attempts have been made to examine the convective morphologies over southern China. We aim to reveal the general features of convective organizational modes (e.g., frequencies, durations, variations, etc.) and determine which particular types of severe weather are more or less likely to be associated with particular convective morphologies. These results are of benefit to local forecasters for better anticipating the storm types and issuing warnings for related hazardous weather.

© 2022 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding authors: Xinyong Shen, shenxy@nuist.edu.cn; Zhiying Ding, dingzhiying@nuist.edu.cn

Abstract

This study investigates the organizational modes of convective storms and associated severe weather in spring and summer (March–August) of 2015–19 over southern China. These storms are classified into three major organizational structures (cellular, linear, and nonlinear), including 10 dominant morphologies. In general, cellular systems are most frequent, followed by linear systems. Convective storms are common in spring, increasing markedly from April to June, and peak in June. Convective storm cases are usually longer lived in spring, while shorter lived in summer. They also present pronounced diurnal variations, with a primary peak in the afternoon and several secondary peaks during the night to the morning. Approximately 79.7% of initial convection clearly exhibits a dominant eastward movement, with a faster moving speed in spring. Convective storms frequently evolve among organizational modes during their life spans. Linear systems produce the most severe weather observations, in which convective lines with trailing stratiform rain are most prolific. Bow echoes are most efficient in producing severe weather events among all systems, despite their rare occurrences. In spring, lines with parallel stratiform rain are abundant producers of severe wind events, ranking the second highest probability. In summer, embedded lines produce the second largest proportion of intense rainfall events, whereas lines with leading stratiform rain are most efficient in generating extremely intense rainfall and thus pose a distinct flooding threat. Broken lines produce the largest proportion of severe weather events among cellular storms. In contrast, nonlinear systems possess the weakest capability to produce severe weather events.

Significance Statement

Under the influence of the East Asian summer monsoon, severe weather events produced by convective storms occur frequently in China, leading to serious natural disasters. Numerous studies have demonstrated that the morphologies of convective storms are helpful to improve our understanding and prediction of convective storms. However, fewer attempts have been made to examine the convective morphologies over southern China. We aim to reveal the general features of convective organizational modes (e.g., frequencies, durations, variations, etc.) and determine which particular types of severe weather are more or less likely to be associated with particular convective morphologies. These results are of benefit to local forecasters for better anticipating the storm types and issuing warnings for related hazardous weather.

© 2022 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding authors: Xinyong Shen, shenxy@nuist.edu.cn; Zhiying Ding, dingzhiying@nuist.edu.cn

1. Introduction

In recent decades, extreme climate and weather events have received widespread attention, as they can lead to serious societal impacts, causing huge economic damages and human losses (Schoen and Ashley 2011; Wang and Cui 2011; Pacey et al. 2021). Numerous studies have demonstrated that severe convective storms are most likely to produce short- duration heavy rainfall, severe convective wind gusts, hail and sometimes even tornadoes, accounting for a large portion of midlatitude and tropical severe weather (Maddox et al. 1979; Maddox 1981; Cotton et al. 1983; Li et al. 2012; Chen et al. 2013; Meng et al. 2013; Zheng et al. 2013; Luo et al. 2014; L. Luo et al. 2020). The storm clusters, storm lines or storm complexes formed by the interaction of individual convective storms are collectively termed mesoscale convective systems (MCSs), which are the largest form of organized convective storms usually with a length scale of >100 km in space and temporal maintenance for at least 3 h (Schumacher and Rasmussen 2020). The recognition of the global importance of MCSs should be attributed to the rapid developments in satellite meteorology. Since the occurrence and frequency of various MCSs can be analyzed by satellites from a global perspective, their convective–stratiform structures can be quantitatively described (Anderson 2010). According to Orlanski’s (1975) definitions of meteorological scales with subclassifications of meso-α, meso-β, and meso-γ scales, MCSs vary greatly in scale and shape, sometimes as mesoscale convective complexes (MCCs; e.g., Maddox 1980; Tao et al. 1998), and sometimes as linear or persistent elongated convective systems (PECSs; e.g., Anderson and Arritt 1998; Jirak et al. 2003; Liu et al. 2021). Different morphologies of convective storms may be responsible for different types of severe weather due to different dynamic, thermodynamic, cloud microphysics and atmospheric radiation processes (Bluestein and Jain 1985; Houze 1989; Parker and Johnson 2000). Therefore, a detailed classification of convective storms is very important for forecasters of severe weather to understand the organization and evolution of convective storms.

Although geostationary satellites are advantageous to identify convective storms and monitor their activities over a wide range, they are still unable to provide the patterns underlying the cloud top of the convective systems from satellite imagery alone. On the contrary, Doppler radar data provide detailed information of convective-scale systems with high temporal and spatial resolution at the mid- and low levels, such as the spatial distribution and development of convective and stratiform regions, three-dimensional cloud microphysical characteristics, and dynamic structures of convective storms. Therefore, radar data have been widely applied to study the classification and dynamical mechanism of convective storms. Different classification schemes have been used in previous studies depending on the aims of the research. For example, based on an 11-yr period of radar data, Bluestein and Jain (1985) identified four distinct classes of severe convective-line formation in Oklahoma during the spring: broken line, back building, broken areal, and embedded areal. They focused strictly on storms that exhibited linear types of cumulonimbus. In contrast, Houze et al. (1990) concentrated not just on squall lines but on all organization types associated with significant rainfall, and put forward the symmetrical and asymmetric structures of leading-line/trailing-stratiform mesoscale precipitation systems (hereafter TS). In addition to the TS structure, Schiesser et al. (1995) documented the existence of severe precipitation systems with leading stratiform rain (hereafter LS) in Switzerland. Based on the above research, Parker and Johnson (2000) devised three major morphologies according to the stratiform precipitation arrangement in linear MCSs: those with trailing stratiform precipitation, leading stratiform precipitation, and parallel stratiform precipitation (hereafter PS). Trapp et al. (2005) considered a relatively simple designation of the convective modes for 3828 tornadoes in the contiguous United States, and attempted to classify the type of storm associated with each tornado as cell, quasi-linear convective system, or other. Gallus et al. (2008) proposed a more detailed classification scheme with nine distinct convective structures to classify 949 convective storms occurring in the central United States and analyzed the risk of severe weather (hail, severe wind, flooding, and tornado events) posed by each. Due to the comprehensive constitution of their classification scheme, subsequently, it was widely adopted to identify various convective storms in many documents (Duda and Gallus 2010; Lombardo and Colle 2010; Snively and Gallus 2014; Carlberg et al. 2018; Ma et al. 2021; Pacey et al. 2021).

The morphologies of convective storms can help toward indirectly reflecting the mesoscale dynamical structures of convective storms. And some patterns have been shown to be responsible for producing one or more types of severe weather. For instance, Fujita (1978) provided deep insights into the development of downburst, and pointed out that there were two major kinds of radar echo patterns associated with the production of severe wind events, which were hook echoes and bow echoes. Houze et al. (1990) found that in terms of tornadoes and hail production, asymmetric TS squall lines are the most prolific types, while symmetrical TS squall lines are more likely to be associated with floods. Then, Pettet and Johnson (2003) found that storms with LS and PS systems were prone to producing more flooding events. Gallus et al. (2008) found that linear systems having bow echoes and trailing stratiform precipitation produced far more severe wind reports than any other morphology; flooding events were most common in broken lines, squall lines with trailing and line-parallel stratiform (PS and TS systems); severe hail and tornado events were most likely to occur in cellular storms, especially in broken squall lines. Duda and Gallus (2010) indicated that supercell storms produced severe wind and flooding much more frequently and severely than non-supercell cellular morphologies. Although the above studies are only a small proportion of numerous investigations of severe weather, these achievements have broadened our knowledge of various mesoscale convective morphologies.

Extensive research has been conducted to study MCSs based on satellite and radar data in the central United States. However, considering that MCSs are a global weather phenomenon, some of the features of MCSs summarized in these particular regions may not apply universally over the globe. Convective storms developed in different regions present different organizational structures. One of the important factors is the large-scale circulation, which plays an important role in the initial, developing and decaying stages of convective storms. In North America, the initiation and evolution of severe convective storms are closely related to the large-scale systems, such as upper-level troughs, midlevel shortwave troughs, low-level jet streams, Atlantic subtropical anticyclone, fronts and drylines (Maddox 1983; Parker and Johnson 2000). In contrast, the activities of the western Pacific subtropical high, the monsoon moisture surge from the South China Sea, the cold air outbreaks over mid- and high latitudes, and the eastward moving convective storms from the Tibetan Plateau are generally considered to be responsible for heavy rainfall during the warm season over East Asia (Zhang et al. 2002; Y. Luo et al. 2020). And in recent years, many studies have shown that the double low-level jets (LLJs) in the southwesterly moist flow are a key factor in controlling the early-summer rainfall over southern China (Du and Chen 2018, 2019a,b). These differences thus lead to the distinct convective organizational features over southern China, posing a considerable challenge to severe storm forecasting. However, the organizational morphologies of convective storms in China have received little attention compared to the substantial amount of research in the United States (Houze et al. 1990; Parker and Johnson 2000; Trapp et al. 2005; Smith et al. 2012). Previous studies just focused on the morphologies of convective storms over North China and central East China: for example, 7 morphologies during June–September of 2007–10 over central East China (Zheng et al. 2013), 6 organizational modes from 2010 to 2014 over North China (Yang and Sun 2018), 10 morphologies during May–September of 2011–18 (except for 2014) over North China (Ma et al. 2021), and 5 organization modes of MCSs at their mature stages during May–June of 2013–17 over South China (Chen et al. 2022). Fewer attempts have been made yet to investigate the general features of convective modes in spring and summer (March–August) over southern China.

The present study is the first to perform a 5-yr identification and classification of convective organizational morphologies and related severe weather events over southern China. Obviously, the frequency, temporal, and spatial evolution of mesoscale convective modes will provide some useful information on the likelihood for the development of a particular morphology, allowing forecasters to better predict these types of organized convective storms. Therefore, this study will focus on the following issues:

  • What are the types, frequencies, and spatiotemporal distributions of convective organizational morphologies over southern China? How do these features compare with central East China and North China?

  • How do these convective morphologies evolve in the initial, mature and decaying stages during the life spans of convective storms?

  • Which type of organizational morphologies plays a dominant role in a particular type of severe convective weather?

The rest of this study is organized as follows. Section 2 introduces the data sources and methodology used in this study. Section 3 highlights the temporal and spatial characteristics of convective modes. Section 4 presents the relationship between convective modes and severe weather. Finally, conclusions and future work are discussed in section 5.

2. Data and methodology

As shown in Fig. 1a, the present work focuses on southern China, from 17° to 33°N and from 103° to 124°E, covering the entirety of South China and parts of the middle and lower reaches of the Yangtze River. This region is chosen because of its active convection, with a significant number of severe thunderstorms occurring during both the spring and summer seasons each year and thus producing various types of severe weather. The observation data are provided by the National Meteorological Information Center of the China Meteorological Administration (CMA). A total of 1073 national surface stations and 104 operational radars (Fig. 1b) are used to analyze the activities of convective storms over southern China.

Fig. 1.
Fig. 1.

Maps of southern China: (a) terrain height (shaded) and names of related provinces and (b) locations of national surface stations (blue crosses) and operational radar stations (green triangles). NLM and WYM in (a) denote the Nanling Mountains and Wuyi Mountains, respectively. The gray shading in (b) denotes the region within 460 km from a radar station where convection can be detected.

Citation: Monthly Weather Review 150, 11; 10.1175/MWR-D-22-0061.1

For the purposes of this research, the historical radar mosaic data should be of high quality. In recent years, with the great efforts of CMA, numerous advanced methods have been applied to improve the quality of radar mosaics. For example, by implementing a fuzzy logic method (Liu et al. 2008) and a series of strict quality control procedures, the ground clutters and abnormal echoes under anomalous propagation conditions of single-site radar base data have been effectively removed. Surface rain gauge data and observations from geostationary satellites have also been utilized to remove nonmeteorological echoes in the clear sky. As a result, the publicly accessible radar mosaics have achieved quite good quality (Bai et al. 2020). Although the algorithms used to make radar mosaics have been developing, the general features of convective storms can be well studied in terms of the morphological characteristics from radar echoes as long as the nonmeteorological echoes are reasonably filtered. Radar mosaics present the composite reflectivity factors from nine elevation volume scans (i.e., 0.5°, 1.5°, 2.4°, 3.4°, 4.3°, 6.0°, 9.9°, 14.6°, and 19.5°) of multiple individual radars with an update time of approximately 6 min. In southern China, the majority of Doppler radars are S-band radars and similar to the Weather Surveillance Radar-1988 Doppler (WSR-88D) in the United States in both hardware and software (Meng et al. 2013; Bai et al. 2020).

a. Classification of convective organizational modes

This investigation is based on the national radar mosaics of composite reflectivity for a 5-yr period of 2015–19 in spring and summer (March–August), with a temporal interval of 60 min before 28 October 2016 and 6 min after that date, 2 km × 2 km horizontal resolution, and 89% of time coverage. All convective systems formed within this domain (Fig. 1) and time period are included as long as they meet the following criteria depicted by the radar echoes as in Gallus et al. (2008):

  1. the contiguous radar echoes above the noise level of the data (10 dBZ) covering an area greater than 6 km × 6 km,

  2. a maximum radar echo intensity of at least 30 dBZ within the area in criterion 1, and

  3. temporal maintenance of criteria 1 and 2 for at least 1 h.

The radar echoes meeting the above criteria thus identify the convection over southern China well, especially those isolated convective storms. In contrast, Zheng et al. (2013) adopted more stringent criteria to study MCSs over central East China: the first and last appearances of 30-dBZ contiguous radar echoes with a peak intensity greater than 45 dBZ reaching an area of at least 30 km × 30 km. In the present study, storms are classified according to the dominant morphology over time from each radar image. The lifetime of a storm is determined by the start (end) time being the first (last) radar image. If one particular mode persists for only 1 h or less, usually the case in the newly formed or decaying stages, hence, that mode is ignored. As in previous studies (Gallus et al. 2008; Duda and Gallus 2010; Zheng et al. 2013), all convective storms meeting the above criteria are classified by visual inspection according to their dominant morphology. Although it has been widely acknowledged that classification of convective systems through mere visual inspection of radar is inherently subjective, the quantitative criteria employed for classification should reduce the subjectivity.

Therefore, the present study integrates the classification schemes from Gallus et al. (2008) and Zheng et al. (2013) to classify storm morphologies into 10 categories (Fig. 2). This scheme possesses three types of cellular storms, which are isolated cells (IC), clusters of cells (CC), and broken lines (BL); six types of linear systems, which are squall lines with no stratiform rain (NS), squall lines with trailing stratiform rain (TS), squall lines with leading stratiform rain (LS), squall lines with parallel stratiform rain (PS), bow echoes (BE), as well as embedded lines (EL); and nonlinear systems (NL). Convective and stratiform echoes are defined as those with reflectivity factors greater than 40 and 20–40 dBZ respectively from radar images (Parker and Johnson 2000). This scheme is selected for this study since it demonstrates 10 storm morphologies that can be easily identified within the capabilities of the radar datasets over southern China. It should be noted that the visual inspection utilized to classify convective storms is inherently subjective and difficult at times. Even if strict objective criteria are used to define each morphology, there are still storms that sometimes are likely to present two or more structures, making identification and classification challenging. However, the potential advantage of visual inspection is that it is capable of identifying and tracking multiple convective storm cases on one radar mosaic image according to the characteristics of convective storms in the initial, mature and decaying stages during their lifetimes, respectively. Approximately 2% of the systems proved difficult to classify. Figure 3 exhibits in detail the four convective cases with different organizational modes in South China during 18–19 April 2019. Furthermore, for a complex convective system that is closely related to the forcings of synoptic-scale weather systems, especially fronts, shear lines, and the activities of the accompanying MCSs, and usually covers a horizontal scale of hundreds of kilometers, the convection at different parts generally exhibits different organizational structures and characteristics, and thus it remains challenging to depict such a convective system with only one morphology (Luo et al. 2014; Wang et al. 2014; Chen et al. 2022). Considering that defining or introducing additional morphologies will make the problem more complicated, in this study, these complex convective storms are divided into two or more parts, and then tracked and classified according to the primary behaviors of convective storms in different stages.

Fig. 2.
Fig. 2.

Schematic radar echoes demonstrating each of the 10 morphologies are adapted from the classification schemes used by Gallus et al. (2008) and Zheng et al. (2013). Morphologies are abbreviated as follows: isolated cells (IC), clusters of cells (CC), broken line (BL), squall line with no stratiform rain (NS), squall line with trailing stratiform rain (TS), squall line with leading stratiform rain (LS), squall line with parallel stratiform rain (PS), bow echo (BE), embedded line (EL), and nonlinear system (NL).

Citation: Monthly Weather Review 150, 11; 10.1175/MWR-D-22-0061.1

Fig. 3.
Fig. 3.

The observed composite radar reflectivity factors (shaded; dBZ) of four convective storm cases with different organizational morphologies (notation as in Fig. 2) in South China during 18–19 Apr 2019. The purple ellipses together with the labels denote the dominant modes tracked in each case. Times are local standard time (LST), where LST = UTC + 8 h.

Citation: Monthly Weather Review 150, 11; 10.1175/MWR-D-22-0061.1

In addition, in the present work, in order to track convective storms and quantify their morphology changes during their life spans, only convective storms triggered in the study domain are included. From an integrity perspective of morphology statistics, if storms extend beyond the edge of the study domain or at sea during their lifetimes, they are also included in the morphology sample, but only severe convective weather observations from within the study domain are included. If these events were ignored, statistical uncertainties would have been introduced into the results. Meanwhile, the events propagated from the upstream western area beyond the study domain are excluded from this study, despite low occurrences of such events. It is recognized that very few events originated from the western boundary of the study domain are likely related to the topography of the Yunnan–Guizhou Plateau. Moreover, the convective activities in the northwest region (i.e., the Sichuan basin and the western Guizhou Province) are mainly concentrated from September to November (i.e., autumn rain in West China), which is beyond the scope of this study in time and space.

b. Definition and identification of severe convective weather

It has been acknowledged that various types of severe convective weather (such as short-duration heavy rainfall, severe convective wind gusts, hail, tornadoes, etc.) are closely associated with severe convective storms. However, worldwide, there is still no unified definition of severe convective weather. In China, it is usually identified as one of the following convective events (Yu and Zheng 2020): 1) short-duration heavy rainfall of no less than 20 mm h−1, 2) severe convective wind gusts of at least 17 m s−1, 3) hail with a diameter of 2 cm or greater on the ground, or 4) a tornado with intensity rated by the enhanced Fujita (EF) scale on land. As a contrast, severe convective weather is defined more strictly by the National Weather Service in the United States, especially for severe convective wind and hail (Gallus et al. 2008). These differences are likely related to the great contrasts in geography and climate that exist between the two regions. Considering that hail and tornadoes usually relied on manual observation, there is still a large uncertainty in determining their formation time, location, and intensity. Therefore, this study only focuses on short-duration heavy rainfall and severe convective wind events, which are extracted from the hourly data of national surface stations.

After performing a series of quality-control procedures (Feng et al. 2004; Ren et al. 2015), including climatic high-low extremes check, internal consistency check, temporal outliers check, spatial outliers check, data gaps check, etc., the surface rain gauge data and wind gusts data have achieved fairly good quality. It is worth mentioning that, by definition, severe convective wind events should be accompanied by intense thunderstorms, which are defined as those associated with more active cloud-to-ground lightning than the background climatology. Therefore, according to the method proposed by Yang et al. (2017), this study additionally employs the cloud-to-ground lightning data to remove some severe wind records associated with weak thunderstorms by comparing the intensities of cloud-to-ground lightning between each severe wind event and the climatic background in the same month in which the severe wind observations acquired. After excluding the events with invalid or poor quality data, a dataset of severe weather events in relation to each convective mode is established.

To match the radar images with the hourly surface observations, the dominant mode within 1 h exhibited in the radar data is considered in this study. It should be noted that the hourly radar mosaic images before 28 October 2016 may introduce some uncertainties into monitoring those rapidly evolving convective storms, whose morphology and intensity may change within one hour. Since we emphasize the dominant mode of the storm within one hour, the potential influence of radar mosaics with coarse temporal resolution on the results can be effectively minimized. For instance, if a TS system is present over a period of more than 1 h, the severe weather observations will be attributed to the TS category, and it is not necessary to consider whether the TS mode present at a later time is the same as that seen at a slightly earlier time. Moreover, considering some fast moving and developing storms, as well as gust fronts (and high winds) sometimes located away from the storms, the associated severe weather observations within 1 h in time and within 20 km around the area covered by the continuous radar echoes in space are assigned to the corresponding convective mode in the present study.

3. Characteristics of convective modes

According to the classification criteria defined in section 2, a total of 1287 convective storm cases associated with 12 564 morphology samples are identified during a 5-yr period of 2015–19 over southern China. These numbers can be compared to a 7-yr dataset over North China (Ma et al. 2021): 1429 morphology samples from 371 severe convective storms; a 4-yr dataset over central East China (Zheng et al. 2013): 159 morphology samples from 47 MCS cases; and a 5-yr dataset over South China (Chen et al. 2022): 148 morphology samples during the lifetimes of 98 MCSs. It is shown that the inclusion of southern China in this study results in a sample size that is an order of magnitude larger than that in previous studies, leading to a great challenge to identify and classify these storms.

a. Frequencies and durations of convective modes

Percentage contribution of each convective mode to the total morphological samples is shown in Fig. 4. It is shown that nonlinear systems are most frequent, accounting for 29.6%, followed by clusters of cells (19.1%). Among the linear systems, TS systems are most common, accounting for 42.7%, and EL (32.7%) and NS (14.6%) systems are ranked the second and third most frequent, respectively. In total, cellular storms account for 36.2% (19.1% CC, 15.8% IC, and 1.3% BL). Linear systems account for 34.2% (14.6% TS, 11.2% EL, 5.0% NS, 1.4% PS, 1.1% LS, and 0.9% BE). It should be noted that PS, LS, and BE modes contribute very little to the whole population, which are consistent with the results from Gallus et al. (2008), Duda and Gallus (2010), and Lombardo and Colle (2010).

Fig. 4.
Fig. 4.

Percentages of the 10 convective modes (notation as in Fig. 2) examined among the 12 564 morphological samples which are identified from the 1287 convective storm cases. Light gray represents cellular storms, medium gray represents nonlinear systems, and dark gray represents linear systems.

Citation: Monthly Weather Review 150, 11; 10.1175/MWR-D-22-0061.1

As shown in Table 1, these percentages can be compared to a 7-yr dataset over North China: 30.6% for cellular storms, 26.3% for linear systems, and 37.8% for nonlinear systems. Linear systems and cellular storms over southern China are 30% and 18% more than those over North China, respectively, while nonlinear systems over North China are more dominant, with 28% more than those over southern China. Linear systems account for 53.6% of the total number of nonlinear and linear systems, similar to the results of Zheng et al. (2013), although they paid more attention to large convective systems rather than cellular storms over central East China. Another apparent distinction between this study and Zheng et al. (2013) is the absence of a primary linear organizational mode over central East China. TS and EL modes dominate the majority of the linear systems over southern China, while TS and NS modes dominate over northern China (Ma et al. 2021). Overall, the proportion of cellular storms (36.2%) herein is comparable to that of linear systems (34.2%), whereas cells in the United States are most prevalent (Gallus et al. 2008; Duda and Gallus 2010). These percentages of linear systems in China are also comparable to that in Europe (33%), but both are more than that in the United States. In contrast, the cellular storms in the United States and Europe are more than 1.5 times as much as those in China. This could explain why hail and tornadoes occurred more frequently in the United States than in China during the past few decades, because cellular storms are more likely to produce such severe weather as demonstrated in previous studies (Gallus et al. 2008; Duda and Gallus 2010).

Table 1

A comparison between the percentages of the three types of convective systems (cellular, nonlinear systems, and linear systems) over southern China and those in previous studies. A dash (—) indicates that the corresponding data are unavailable.

Table 1

As shown in Table 2, the convective storms are fairly common in spring, mainly in summer, and attain their peak occurrences in June, after the onset of the South China Sea monsoon. In spring, the storms are longer lived (10–12 h on average), while in summer, the storms tend to be shorter lived (8–10 h). On average, the lifetime of convective storms in spring is 2.8 h longer than that in summer. In June, larger and longer-lived systems (such as linear systems) become relatively less common, whereas the small and short-lived convective storms (such as cellular storms) occur more and more frequently.

Table 2

Monthly variation of the frequency and average lifetime (h) of convective storm cases.

Table 2

The average lifetime of all convective storm cases is 9.8 h (Fig. 5a). A total of 48.9% of convective storms have a lifetime of 6–12 h. A peak appears around 6–9 h, accounting for 31.4%. The shortest and longest lifetimes of convective storms are 2.8 and 38.3 h, respectively. Although the lifetime of convective storms in spring and summer in southern China is comparable to that in the southeastern United States (Geerts 1998), apparent differences in seasonal distribution still exist between the two regions. In spring, the lifetime of convective storms in southern China (11.4 h) is about 4.2 h longer than that in the southeastern United States (7.2 h), whereas the longer-lived convective storms in the southeastern United States (10–12 h) are mainly occurred in autumn and winter. In this survey, only 39.9% of convective storms have a lifetime of more than 9 h, while 73.4% in central East China (Zheng et al. 2013). On average, in China, the lifetime of convective storms in low latitudes is shorter than that in middle and high latitudes (11.4 h over North China and 14 h over central East China). Because of the thermodynamical differences between the north and the south, the shorter-lived cellular systems are peculiarly prone to being triggered in the south. For instance, the percentage of convective storms with a lifetime of less than 6 h in southern China (28.8%) is 2.4 times more than that in central East China [12.2%, see Zheng et al. (2013)] and more than 4 times as much as that in North China [7%, see Ma et al. (2021)].

Fig. 5.
Fig. 5.

Frequency distributions as functions of lifetime for (a) all convective storm cases and duration for three broader categories of convective storms in spring and summer over southern China: (b) cellular storms, (c) nonlinear systems, and (d) linear systems. The number above the histogram denotes the percentage (%) among each sample.

Citation: Monthly Weather Review 150, 11; 10.1175/MWR-D-22-0061.1

Figures 5b–d present in detail the frequency distributions as a function of duration for three typical categories of convective storms over southern China: cellular storms, linear systems, and nonlinear systems. In spring and summer, both cellular storms and nonlinear systems exhibit a clearly left-skewed distribution, with an average duration of 4.2 and 3.7 h, respectively. In contrast, linear systems persist relatively longer, ranging from 1 to 33 h, with an average of 6.2 h, and peak at 3–6 h, accounting for about one-third. In addition, a considerable number of linear convective storms (11.7%) live longer than the half-period of the diurnal cycle. Approximately two-thirds of the nonlinear systems (Fig. 5c) last for less than 3 h, twice as much as that of linear systems (Fig. 5d). As may be expected, the average duration of any type of cellular modes is shorter than that of nonlinear and linear modes (Table 3). Nearly 95% of IC, CC, and BL modes survive for 1–6 h, whereas only a small fraction of NL (14%), BE (18%), EL (15%), and LS (10%) modes and a quarter of TS modes can persist for longer than 6 h. Among the six linear modes, the duration of TS (5.2 h) and BE (5.0 h) modes is longer than that of EL (4.0 h), LS (3.5 h), PS (3.4 h), and NS (3.2 h) modes. By contrast, the average duration of linear convective storms over the central United States (Parker and Johnson 2000) is almost twice as long as that over southern China.

Table 3

Duration (h) of each convective mode.

Table 3

Furthermore, the frequency for nonlinear systems presents the least difference between spring and summer (Fig. 5c). On the contrary, the frequency for cellular storms with a lifetime of less than 6 h is significantly more in summer than in spring (Fig. 5b), while that for linear systems with a lifetime of more than 6 h is considerably more in spring than in summer (Fig. 5d). It is mainly because in spring, the troposphere is essentially baroclinic with more frontal convective rainfall, whereas in summer (except for June), baroclinic forcing becomes very weak or even sometimes absent, thus triggering more small-scale and short-lived cellular storms. It is worth further exploring whether the linear convective storms with long lifetime in spring occur in an environment where the wind speed and thermal stability profile are suitable for such type of squall lines.

b. Interannual and monthly variations

As shown in Fig. 6a, the annual average occurrences of convective modes are approximately 2513, with a maximum and minimum of 3644 and 1510, respectively. The peak of convective storms (3644) appears in 2019, followed by 2693 in 2016. With respect to the monthly variation (Fig. 6b), the numbers of convective storms range from 1035 to 2776, with an average of 2094. The frequency of convective storms peaks between May and June, then gradually decreases in July, August, and April, and reaches a valley in March. Besides, each methodology exhibits apparent differences in terms of interannual and monthly variations. For example, the number of cellular storms and nonlinear systems peaks in 2019, while that of linear systems peaks in 2016 (Fig. 6a). Cellular storms occur frequently between May and August, while nonlinear and linear systems are most common between April and June (Fig. 6b). However, TS and EL modes share a commonality that both of them account for large proportions on both the interannual and monthly scales. Both cellular and nonlinear systems present a single-peak pattern with a maximum in June, whereas linear systems exhibit a clear double-peak pattern, with the major peak in May and the secondary peak in August. Furthermore, the seasonal variation of cellular and linear systems demonstrates that May and June are the transient months of spring and summer seasons, as discussed in the previous section. The frequently observed linear systems during the early summer (May–June) are probably associated with the high wind shear and intense interactions between the northern cold, dry airflow and southern warm, wet airflow over southern China (Liu and Guo 2012). Previous studies have noted that the major contribution of linear systems in August comes from the pre-TC (preceding landfalling tropical cyclones) squall lines (Meng and Zhang 2012; Meng et al. 2013).

Fig. 6.
Fig. 6.

Number of convective storms associated with each morphology by (a) year and (b) month. The blue, green, and red labels represent the numbers of cellular storms, nonlinear systems, and linear systems, respectively.

Citation: Monthly Weather Review 150, 11; 10.1175/MWR-D-22-0061.1

c. Diurnal variations

There is a marked diurnal pattern of convective storm activity. Figure 7 shows the diurnal variations of convective storms in three essential stages (initiation, maturity and dissipation) during their life cycles in spring and summer. The initiation (dissipation/decaying) time of a storm is subjectively determined by the time when the first (last) radar image meets the criteria defined in section 2. The maturity time is defined as the time when the radar echo of composite reflectivity factor reaches its maximum value during the lifetime, rather than the central time of an MCS life cycle adopted in previous studies (Geerts 1998; Parker and Johnson 2000). According to the frequency of convective storms during the mature stage (the red curve in Fig. 7), the primary peak and several secondary peaks of occurrences are identified.

Fig. 7.
Fig. 7.

Frequencies of convective storm cases in the stages of initiation, maturity, and dissipation during their life cycles in (a) spring and (b) summer. Times are LST.

Citation: Monthly Weather Review 150, 11; 10.1175/MWR-D-22-0061.1

As shown in Fig. 7, convective storms can be triggered and then develop at any time of the day. However, apparent similarities and differences exist between spring and summer. The common feature is that the primary peaks (Peak_1 and Peak_4) possessed by convective storms in spring and summer shared the same time period, which appear during early afternoon to evening (1400–2000 local standard time, LST = UTC + 8 h.) and persist for about 6 h. This primary peak herein are similar to those in the southeastern United States (Geerts 1998), but apparently different from the late evening peak (between 1800 and 0200 LST) in the central United States (Wallace 1975; Maddox 1980; Parker and Johnson 2000). Besides, the frequency of primary peaks of convective storms in summer is more than twice as much as that in spring, indicating that convective storms during afternoon to evening in summer are more active than those in spring.

Moreover, two secondary peaks of convective storms occurrences appear in spring (Fig. 7a, one forms during 2200–2300 LST, matures during 0100–0300 LST, and dissipates during 0500–0700 LST, as denoted by Peak_2; the other forms at 0800 LST, matures at 1000 LST, and dissipates at 1200 LST, as denoted by Peak_3), while only one secondary peak exists in summer (Fig. 7b, forms at 0200 LST, matures during 0300–0600 LST, and dissipates at 1000 LST, as denoted by Peak_5). These results are consistent with the double-peak pattern in diurnal cycles of thunderstorms documented in this region (Zheng et al. 2011). Due to the number of samples, the amplitudes of these nocturnal secondary peaks are not as great as those of the primary peaks. However, these nocturnal storms in China have also been well documented in previous studies (Meng et al. 2013; Zheng et al. 2013; Bai et al. 2020). It should be pointed out that the duration of convective storms between the late night and early morning (Peak_2 and Peak_5) is relatively longer than that in daytime (Peak_1, Peak_3, and Peak_4), which is most likely resulted from the influence of environmental conditions. Previous studies (Blackadar 1957; Hoecker 1963) have determined that the diurnal variations of the low-level jets are critical for the development and enhancement of nocturnal convective storms. Similar secondary maxima in MCSs were well documented in the central (Parker and Johnson 2000) and southeastern (Geerts 1998) United States. It should be attributed to utilizing the regional radar data that a secondary maximum has been revealed, which is unapparent in satellite imagery.

As shown in Fig. 8, most modes exhibit a clear diurnal preference. Cellular storms (Figs. 8a–c) have a similar primary peak (between afternoon and evening) and a valley (in the morning). More than two-thirds (69%) of IC modes (Fig. 8a) occur between afternoon and evening (1300–2100 LST) and peak at 1600 LST. During this period, the frequency of IC modes in summer is approximately 3 times more than that in spring. Nevertheless, a secondary peak appears at late night (0000–0400 LST) for both the CC and BL modes, which is not evident for IC mode.

Fig. 8.
Fig. 8.

Diurnal variation of the frequency of occurrence among each convective mode in spring and summer. Times are LST.

Citation: Monthly Weather Review 150, 11; 10.1175/MWR-D-22-0061.1

NL mode (Fig. 8d) presents two peaks at 0900 and 1900 LST, as well as two valleys at 1500 and 0400 LST, respectively. There is little seasonal difference between spring and summer in the diurnal distribution of NL mode. However, compared with other modes, NL mode contributes the most to the peak of convective storms between 0800 and 1200 LST in spring (Fig. 7a, as denoted by Peak_3). Similar to NL mode, EL mode also presents a double-peak pattern, especially in summer.

The diurnal cycle of frequency for TS mode (Fig. 8f) shows a primary peak during the early evening (1700–1900 LST), which is about 2 h later than the peak time for most initial convective modes (e.g., IC, CC, BL, and NS, see later) during the late afternoon (1500–1700 LST). Furthermore, a secondary peak exists during the late night to the early morning. While this secondary maximum is relatively small for TS mode with respect to the amplitude of line pattern, it appears to be one of the major peaks for both LS and EL modes.

Most NS modes occur between 1400 and 1900 LST. PS modes with the highest frequency occur between 1200 and 1700 LST, apparently different from the observed maxima for PS modes during the evening to midnight (1700–0000 LST) in the central United States (Parker and Johnson 2000). In addition, most BE modes are observed in spring, with only a few occurring during the afternoon to the early night in summer.

d. Geographical distributions of initial convective modes

The geographical distribution of the frequency of convective storm initiation is shown in Fig. 9. The location of a storm is the centroid point of radar echo greater than 30 dBZ or the geometric center if the radar echo is scattered. It is shown that there are striking differences among each initial convective mode in spring and summer. In spring, cellular storms occur frequently in the region south of the Yangtze River (Fig. 9a), among which the largest frequency of their initiation is evident around 23°N, 112°E, near the boundaries between Guangxi and Guangdong Provinces (the names of the provinces are given in Fig. 1a). In summer, cellular storms are significantly more active (Fig. 9b), with their occurrences being approximately 1.6 times more than those in spring, and concentrated along the coastal, as well as in the Pearl River basin and high mountain regions, such as Nanling and Wuyi Mountains (the terrain is given in Fig. 1a). Approximately 65.4% of convective storms are initiated in these regions, which are likely induced by the differential heating between land and sea or mountain ranges and low-lying plains, indicating that both land-sea distribution and topography are important factors for triggering convection.

Fig. 9.
Fig. 9.

Geographical distribution of the frequency (blue shading) among the three typical types of initial convective modes: (a),(b) cellular storms (CS); (c),(d) NL; and (e),(f) NS modes at their formation times in (left) spring and (right) summer over southern China. Darker shading denotes higher frequency. The numbers represent the number of times that the centroids of initial convection occur within the 2° × 2° latitude–longitude grid squares. The red hollow circle denotes the centroid location of each initial convective storm. The green hollow circle denotes the average centroid of these storms. The arrow denotes the velocity of each initial convection (m s−1; the scale is in the bottom-left corner). The yellow solid box represents relatively high-frequency area.

Citation: Monthly Weather Review 150, 11; 10.1175/MWR-D-22-0061.1

In spring, NL systems are mainly initiated in the northwest of Hunan and Jiangxi Provinces, as well as northern Guangdong Province (Fig. 9c). These areas are located on the windward slopes of the mountains, demonstrating that the effects of the local terrain produce an important impact on triggering convection. In summer, the genesis of NL systems is concentrated near the junction of Guangxi and Guangdong Provinces, as well as the eastern coastal areas of Guangdong Province (Fig. 9d). It should be noted that the average latitude of NL mode is higher than that of the other two initial modes.

In spring, NS systems (Fig. 9e) forced by synoptic-scale systems are triggered over northwestern Guangxi Province (eastern Yunnan–Guizhou Plateau), and then develop and move to the southeast, affecting Guangxi and Guangdong Provinces. In contrast, in summer, NS systems are concentrated in Guangdong Province with frequent formation and evolution (Fig. 9f). It should be pointed out that the initial convection over northwestern Guangxi Province is dominated by NS mode, which could be a consequence of the quasi-stationary frontal system that triggered and strengthened by the southward movement of the weak cold air from a shortwave trough over the Tibetan Plateau through the Yunnan–Guizhou Plateau (Lin et al. 2014), rather than the uplifting effect of the terrain.

In general, the mean tracks of initial convection in spring and summer (the green arrows in Fig. 9) clearly reveal the dominant eastward movements of convective storms. Approximately 79.7% of convective storms in spring and summer have an eastward component of movement, and exhibit considerable seasonal differences. The eastward movements of convective storms frequently occur in spring, accounting for 97.2%, while only 67.4% in summer, which are inseparable from the synoptic-scale weather systems dominated in different seasons. In spring, the eastward-propagating shortwave troughs, the midlatitude dry and cold air outbreaks, and the low-level jets in the prevailing southwesterly monsoon (Wu et al. 2020) are important factors in controlling the eastward movements of convective storms (Figs. 9a,c,e). Conversely, in summer, due to the influence of the western Pacific subtropical high and tropical cyclones (TCs), relatively more westward movements are observed in lower latitudes (Figs. 9b,d,f). Obviously, the average moving speed of all initial convection in spring is faster than that in summer. In recent years, similar behaviors of eastward movements of MCSs have been well documented over South China in the study by Chen et al. (2022).

e. Transitions of convective modes

As shown in Fig. 10, the dominant modes of convective storms frequently evolve into the other modes in their initial, mature and decaying stages. As may be expected, among the three initial modes, convective storms triggered by the cellular storms (CS) (1054 cases) are most prevalent, accounting for 82% of all cases, followed by NL mode (140 cases) and NS mode (90 cases). By comparison, the initial modes in central East China are dominated by NL and NS systems (Zheng et al. 2013), accounting for 77% of all cases. It is demonstrated that CS systems are most prevalent during the initial stage of convective storms, resulting in the highest frequency among all modes (36.2%, as shown in Fig. 4). Therefore, the environmental conditions for CS systems deserve preferential investigation in the future study. Furthermore, given the total convective storms distribution as detailed above (Fig. 4), it is not surprising that a majority of convective storms are eventually characteristic of NL (521 cases, accounting for 40%) or CS (423 cases, accounting for 33%) modes. This proportion is even higher in those convective storm cases that dominated by TS/EL. More specifically, about 76% (82%) of convective storms whose dominant mode is TS (EL) eventually evolve into NL systems. About 38.9% (38.6%) of convective storm cases that begin as CS (NL) maintain this state from initiation until dissipation (Figs. 10a,b). Conversely, this phenomenon only accounted for 7.8% of convective storm cases that begin as NS, whereas approximately 90% of these cases evolve into other linear systems (Fig. 10c).

Fig. 10.
Fig. 10.

Schematic diagrams of evolutionary pathways for convective storms initiated as (a) cellular storms (CS, 1054 cases), (b) NL (140 cases), and (c) NS modes (90 cases) in this study. Labels along each pathway denote the dominant modes in the initial, mature, and decaying stages during the life cycles of convective storms. The number of convective storm cases following each step is also indicated. The thick blue arrows denote the frequent transitions between convective modes by the events occurrence. Only 3 cases are unclassifiable among the 1287 cases, accounting for 0.2%.

Citation: Monthly Weather Review 150, 11; 10.1175/MWR-D-22-0061.1

It is worth noting that TS systems are most frequently developed in convective storms that begin as NS (43%, Fig. 10c), followed by those begin as NL (25%, Fig. 10b). BE systems present similar characteristics. Comparatively, EL systems are prone to developing in convective storms that begin as NL (23%, Fig. 10b), followed by those begin as NS (19%, Fig. 10c). It is likely that convective lines start to accelerate forward over time, thus favoring the formation of convective storms with TS/BE structure during their later stages (Parker and Johnson 2000). Actually, with the development of convection, the linear echoes embedded in stratiform precipitation (EL) are more frequently identified from those modes dominated by large-scale stratiform precipitation (e.g., NL) in radar imagery, leading to little difference in the number of cases with EL and TS as their dominant modes in the mature stages (Fig. 10b). Additionally, convective storms dominated by LS and PS modes are relatively rare, accounting for less than 6%, likely because of the unique juxtaposition of low-level convergence boundaries and upper-level winds within the dynamic structures of LS and PS modes (Parker and Johnson 2000) determining that such convective modes are not generally produced during the evolution of convective storms.

4. Severe weather related to convective morphologies

This study further investigates the severe weather observations produced by each of the 10 convective structures. As shown in Fig. 11, linear systems produce the most severe weather observations (accounting for 60.2%), followed by nonlinear systems (20.0%) and cellular storms (19.8%). These percentages can be compared to a 7-yr dataset over North China: 44.4% for linear systems, 41.5% for nonlinear systems, and 10.7% for cellular storms. In addition, among the linear systems, TS systems (27.9%) are most conducive to producing severe weather observations, followed by EL systems (17.6%). NS (7.1%) and BE (3.9%) systems are ranked the third and fourth most frequent, respectively, with the least proportion for PS (2.1%) and LS (1.6%) systems.

Fig. 11.
Fig. 11.

Percentage of severe weather observations produced by each of the 10 convective morphologies. A total of 14 036 observations are examined in spring and summer (March–August) of 2015–19 over southern China. Light gray denotes cellular storms, medium gray denotes nonlinear systems, and dark gray denotes linear systems.

Citation: Monthly Weather Review 150, 11; 10.1175/MWR-D-22-0061.1

According to the definition of severe weather events in section 2, a total of 12 557 severe weather events are identified. We further examine the possible correspondence between each mode and the following specific weather events: rain events (rainfall ≥ 20 mm h−1) only, wind events (maximum wind gust speed ≥ 17 m s−1) only, rain and wind events, rain or wind events, and weak events (neither rain nor wind events). As shown in Fig. 12, rain events are most common for all modes, with a peak of 46% and 58% in spring and summer, respectively. PS systems tend to produce wind events in spring. However, in summer, wind events seem to be independent of PS systems, indicating that wind events generated by PS systems are generally accompanied by rain events. In spring and summer, LS systems present the highest proportion of rain events among all linear systems, but their proportion of rain and wind events is the lowest. A plausible explanation is that LS systems experience weak rear-to-front storm-relative winds in the middle- and upper-troposphere and thus move slowly (Parker and Johnson 2000), resulting in becoming the leading producers of rain events. It should be noted that LS systems also posed a distinct flood threat in the United States (Pettet and Johnson 2003).

Fig. 12.
Fig. 12.

The (top) percentage of severe weather events, (middle) average severe weather observations per event, and (bottom) number of severe weather events generated by each mode in (a) spring and (b) summer. Types of severe weather events are abbreviated as follows: rain events (rainfall ≥ 20 mm h−1), wind events (maximum wind gust speed ≥ 17 m s−1), rain only, wind only, rain and wind, rain or wind, and weak events (neither rain nor wind events).

Citation: Monthly Weather Review 150, 11; 10.1175/MWR-D-22-0061.1

Linear systems produce the highest frequency of severe weather events, among which BE and TS systems are most significant. Clearly, the average numbers of severe weather observations for each event (the red and blue dashed lines in Fig. 12) generated by BE and TS systems in spring are ranked the first and second highest, respectively. Additionally, rain or wind events are most common for BE systems, accounting for 91% in spring and 84% in summer, followed by PS systems in spring (73%) and EL systems in summer (72%). This is in contrast to the previous results of Gallus et al. (2008) and Duda and Gallus (2010), in which it was found that PS systems produced the largest number of flooding per case over the central United States. In terms of the average number of severe weather observations per event, BL systems are most prolific among the cellular systems. As reported in Gallus et al. (2008) and Duda and Gallus (2010), BL systems were most favorable for producing severe hail and tornado events in the central United States. Given that hail events are always accompanied by either heavy rainfall or severe wind events, or even both, the present results largely resemble those discovered by previous authors.

To further investigate the influence of each convective mode on different types of severe weather events, the short-duration heavy rainfall and severe convective wind events are each divided into two subsets, namely, intense rainfall (20–50 mm h−1), extremely intense rainfall (≥50 mm h−1), severe wind (17–24 m s−1), and extremely severe wind (≥24 m s−1). As shown in Fig. 13, all modes are found to pose some risks of severe weather, but the types and percentages of severe weather events vary substantially from mode to mode. BL systems are most conducive for the above four severe weather types among the cellular storms, followed by IC systems. Meanwhile, NL systems are the principal producers of short-duration heavy rainfall events, whereas the associated severe convective wind events are least common among all systems. In linear systems, LS systems are most efficient in generating extremely intense rainfall in summer, followed by EL systems (Fig. 13d). Unsurprisingly, LS systems hardly produce extremely severe wind in spring and summer. It is worth noting that BE systems seldom appear in spring and summer as mentioned above, however, their probabilities of extremely intense rainfall and extremely severe wind events are the highest among all systems. In spring, PS systems are the abundant producers of severe convective wind events, ranking the second highest probability, which are likely related to the strong middle-tropospheric vertical wind shear exhibited in PS systems (Parker and Johnson 2000). Given the largest number of extremely intense rainfall and extremely severe wind events generated by TS systems in spring, it is not necessarily advisable for local forecasters to wait until seeing a TS echo pattern from radar images before issuing warnings for severe weather. Additionally, as shown in Fig. 13b, only approximately 13% of severe wind events are associated with EL systems, a lower percentage than other linear systems in summer. However, they produce the largest proportion of intense rainfall events (60%) except BE systems, suggesting that EL is another dominant structure for summer heavy rainfall that forecasters should focus on.

Fig. 13.
Fig. 13.

Percentage of severe weather events associated with each convective mode in (a),(c) spring and (b),(d) summer. The short-duration heavy rainfall and severe convective wind events are each divided into two subsets, namely, intense rainfall (20–50 mm h−1), extremely intense rainfall (≥50 mm h−1), severe wind (17–24 m s−1), and extremely severe wind (≥24 m s−1). Numbers of severe weather events of each type are given in parentheses beneath the morphology codes.

Citation: Monthly Weather Review 150, 11; 10.1175/MWR-D-22-0061.1

As shown in Fig. 14, in general, linear systems are much more likely to produce severe weather than either cellular storms or nonlinear systems, exhibiting more significantly in spring than in summer. As expected, the number of linear systems and associated severe convective wind events are more prevalent in spring than in summer. On the contrary, the convective activity of cellular storms in summer is apparently much more common than that in spring. Moreover, in summer, the number of short-duration heavy rainfall events associated with nonlinear systems is about 1.5 times as much as that in spring, which is likely related to the abundant water vapor supply after the onset of the South China Sea monsoon.

Fig. 14.
Fig. 14.

The proportion of severe weather events contributed by cellular storms (CS), nonlinear systems, and linear systems in (a) spring and (b) summer. Four weather types: rain events (rainfall ≥ 20 mm h−1), wind events (maximum wind gust speed ≥ 17 m s−1), rain and wind events, and weak events (neither rain nor wind events). Also given in parentheses after the morphology codes are the total numbers of events of each type.

Citation: Monthly Weather Review 150, 11; 10.1175/MWR-D-22-0061.1

5. Summary and discussion

In spring and summer, convective storms occur frequently over southern China under either strong or weak synoptic forcings. Using radar mosaics of composite reflectivity and national surface observations, a total of 12 564 morphology samples from 1287 discrete convective cases are identified to be responsible for 14 036 severe weather observations during a 5-yr period. The convective storms identified from radar images are classified into 10 categories, including three types of cellular storms, one nonlinear system and six types of linear systems. Then, through examining each mode in terms of frequencies, durations, temporal and spatial variations, geographical distributions, as well as transitions, we attempt to reveal the distinct features of convective organizational modes over southern China. Finally, the severe weather events associated with each organizational mode are further investigated. The major findings are summarized as follows:

  1. Cellular storms account for 36.2% of all samples (19.1% CC, 15.8% IC, and 1.3% BL). Linear systems account for 34.2% (14.6% TS, 11.2% EL, 5.0% NS, 1.4% PS, 1.1% LS, and 0.9% BE). Nonlinear systems (NL) compose the remaining 29.6%. Convective storms are quite common in spring, mainly in summer, and peak in June. In spring, convective storm cases are longer lived (11.4 h on average), while in summer, convective storm cases tend to be shorter lived (8.6 h on average). The duration of TS/BE modes is relatively longer than that of other modes. The activities of convective storm cases exhibit a primary peak during 1400–2000 LST, with two (one) secondary peaks (peak) in spring (summer). Meanwhile, most modes exhibit a clear diurnal preference.

  2. About 79.7% of convective storms in spring and summer exhibit the dominant eastward movements, and present considerable seasonal differences. In spring, the strong synoptic forcing and local orographic effect (lifting/blocking) play a key role in triggering convection. While in summer, both land–sea distribution and terrain are important factors for triggering convection. It is noted that convective storms frequently evolve among convective modes during their life spans. Convective storms initiated by cellular storms are most prevalent (82% of all cases), followed by NL (11%) and NS (7%). TS mode is most common in the mature stages of convective storms, followed by EL mode. Most convective storms eventually evolve into NL (40% of all cases) or CS (33%). In addition, convective storms dominated by LS and PS modes occasionally occur over southern China.

  3. The severe weather observations produced by linear systems are most prevalent with a proportion of 60.2%, followed by nonlinear systems (20.0%) and cellular storms (19.8%). Moreover, TS systems are the most prolific (27.9%) among the linear systems, followed by EL systems (17.6%). BE systems possess the highest probability of short-duration heavy rainfall and severe convective wind events among all systems, particularly for extremely intense rainfall (≥50 mm h−1) and extremely severe wind (≥24 m s−1) events, although they seldom occur in spring and summer. In summer, EL systems produce the second largest proportion (60%) of intense rainfall events. Given the slow movement, not surprisingly, LS systems are most effective in generating extremely intense rainfall in summer and thus pose a distinct flood threat, while they hardly produce extremely severe wind in spring and summer due to weak rear-to-front storm-relative winds. It is worth mentioning that PS systems are the abundant producers of severe convective wind events in spring, ranking the second highest probability, which are likely related to the strong middle-tropospheric vertical wind shear. Consistent with the results from Gallus et al. (2008) and Duda and Gallus (2010), BL is considered as one of the most conducive morphologies because it produces the largest proportion of severe weather events among the cellular storms. Besides, NL systems have the weakest capability to produce severe weather events among all systems.

The results presented in this paper advance our understanding of the particular features of organizational modes and associated severe convective weather over southern China, where the prediction of warm-sector convective storms currently remains challenging in operational forecasting and research communities. However, it should be noted that due to the limitations on the spatial and temporal resolution of radar data and surface observations, this work provides a basic description of convective morphologies during a period of only 5 years. It is therefore necessary to employ additional years of radar data together with high-density regional automatic observations to perform climatological studies of convective organizational modes and further expand the current conclusions in the future. Further investigations in combination with proximity sounding properties should also be conducted to explore the convective lifting mechanisms and thermodynamic conditions associated with severe weather and their spatial and temporal variability between convective structures (i.e., cellular, nonlinear, linear). These efforts are crucial to help forecasters toward better understanding the pre-convective environments in which such severe weather occur and thus leading to better predictions and warnings for severe local thunderstorms.

Acknowledgments.

This research was jointly supported by the National Key Research and Development Program of China (Grant 2019YFC1510400), the National Natural Science Foundation of China (Grants 41975054, 41930967, and 42075014), and the Natural Science Foundation of Guangdong Province, China (Grant 2021A1515011539). We acknowledge the High Performance Computing Center of Nanjing University of Information Science and Technology for their support of this work. We are also grateful to the anonymous reviewers for their insightful and detailed comments toward improving the original version of this paper.

Data availability statement.

The surface observational data, cloud-to-ground lightning data, and composite Doppler radar reflectivity data were provided by the National Meteorological Information Center of the China Meteorological Administration via http://data.cma.cn/en.

REFERENCES

  • Anderson, C. J., and R. W. Arritt, 1998: Mesoscale convective complexes and persistent elongated convective systems over the United States during 1992 and 1993. Mon. Wea. Rev., 126, 578599, https://doi.org/10.1175/1520-0493(1998)126<0578:MCCAPE>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Anderson, G. D., 2010: The first weather satellite picture. Weather, 65, 8787, https://doi.org/10.1002/wea.550.

  • Bai, L., G. Chen, and L. Huang, 2020: Image processing of radar mosaics for the climatology of convection initiation in South China. J. Appl. Meteor. Climatol., 59, 6581, https://doi.org/10.1175/JAMC-D-19-0081.1.

    • Search Google Scholar
    • Export Citation
  • Blackadar, A. K., 1957: Boundary layer wind maxima and their significance for the growth of nocturnal inversions. Bull. Amer. Meteor. Soc., 38, 283290, https://doi.org/10.1175/1520-0477-38.5.283.

    • Search Google Scholar
    • Export Citation
  • Bluestein, H. B., and M. H. Jain, 1985: Formation of mesoscale lines of precipitation: Severe squall lines in Oklahoma during the spring. J. Atmos. Sci., 42, 17111732, https://doi.org/10.1175/1520-0469(1985)042<1711:FOMLOP>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Carlberg, B. R., W. A. Gallus Jr., and K. J. Franz, 2018: A preliminary examination of WRF ensemble prediction of convective mode evolution. Wea. Forecasting, 33, 783798, https://doi.org/10.1175/WAF-D-17-0149.1.

    • Search Google Scholar
    • Export Citation
  • Chen, J., Y. Zheng, X. Zhang, and P. Zhu, 2013: Distribution and diurnal variation of warm-season short-duration heavy rainfall in relation to the MCSs in China. Acta Meteor. Sin., 27, 868888, https://doi.org/10.1007/s13351-013-0605-x.

    • Search Google Scholar
    • Export Citation
  • Chen, Y., Y. Luo, and B. Liu, 2022: General features and synoptic-scale environments of mesoscale convective systems over South China during the 2013–2017 pre-summer rainy seasons. Atmos. Res., 266, 105954, https://doi.org/10.1016/j.atmosres.2021.105954.

    • Search Google Scholar
    • Export Citation
  • Cotton, W. R., R. L. George, P. J. Wetzel, and R. L. McAnelly, 1983: A long-lived mesoscale convective complex. Part I: The mountain–generated component. Mon. Wea. Rev., 111, 18931918, https://doi.org/10.1175/1520-0493(1983)111<1893:ALLMCC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Du, Y., and G. Chen, 2018: Heavy rainfall associated with double low-level jets over southern China. Part I: Ensemble-based analysis. Mon. Wea. Rev., 146, 38273844, https://doi.org/10.1175/MWR-D-18-0101.1.

    • Search Google Scholar
    • Export Citation
  • Du, Y., and G. Chen, 2019a: Heavy rainfall associated with double low-level jets over southern China. Part II: Convection initiation. Mon. Wea. Rev., 147, 543565, https://doi.org/10.1175/MWR-D-18-0102.1.

    • Search Google Scholar
    • Export Citation
  • Du, Y., and G. Chen, 2019b: Climatology of low-level jets and their impact on rainfall over southern China during the early-summer rainy season. J. Climate, 32, 88138833, https://doi.org/10.1175/JCLI-D-19-0306.1.

    • Search Google Scholar
    • Export Citation
  • Duda, J. D., and W. A. Gallus Jr., 2010: Spring and summer Midwestern severe weather reports in supercells compared to other morphologies. Wea. Forecasting, 25, 190206, https://doi.org/10.1175/2009WAF2222338.1.

    • Search Google Scholar
    • Export Citation
  • Feng, S., Q. Hu, and W. Qian, 2004: Quality control of daily meteorological data in China, 1951–2000: A new dataset. Int. J. Climatol., 24, 853870, https://doi.org/10.1002/joc.1047.

    • Search Google Scholar
    • Export Citation
  • Fujita, T. T., 1978: Manual of downburst identification for Project NIMROD. NOAA/NASA Tech. Rep. NRC-04-74-239, 118 pp.

  • Gallus, W. A., Jr., N. A. Snook, and E. V. Johnson, 2008: Spring and summer severe weather reports over the Midwest as a function of convective mode: A preliminary study. Wea. Forecasting, 23, 101113, https://doi.org/10.1175/2007WAF2006120.1.

    • Search Google Scholar
    • Export Citation
  • Geerts, B., 1998: Mesoscale convective systems in the southeast United States during 1994–95: A survey. Wea. Forecasting, 13, 860869, https://doi.org/10.1175/1520-0434(1998)013<0860:MCSITS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Hoecker, W. H., Jr., 1963: Three southerly low-level jet systems delineated by the weather bureau special pibal network of 1961. Mon. Wea. Rev., 91, 573582, https://doi.org/10.1175/1520-0493(1963)091<0573:TSLJSD>2.3.CO;2.

    • Search Google Scholar
    • Export Citation
  • Houze, R. A., Jr., 1989: Observed structure of mesoscale convective systems and implications for large-scale heating. Quart. J. Roy. Meteor. Soc., 115, 425461, https://doi.org/10.1002/qj.49711548702.

    • Search Google Scholar
    • Export Citation
  • Houze, R. A., Jr., B. F. Smull, and P. Dodge, 1990: Mesoscale organization of springtime rainstorms in Oklahoma. Mon. Wea. Rev., 118, 613654, https://doi.org/10.1175/1520-0493(1990)118<0613:MOOSRI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Jirak, I. L., W. R. Cotton, and R. L. McAnelly, 2003: Satellite and radar survey of mesoscale convective system development. Mon. Wea. Rev., 131, 24282449, https://doi.org/10.1175/1520-0493(2003)131<2428:SARSOM>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Li, J., B. Wang, and D. Wang, 2012: The characteristics of mesoscale convective systems (MCSs) over East Asia in warm seasons. Atmos. Oceanic Sci. Lett., 5, 102107, https://doi.org/10.1080/16742834.2012.11446973.

    • Search Google Scholar
    • Export Citation
  • Lin, Z., M. Lin, K. Lin, and H. Luo, 2014: A convective process of quasi-stationary front triggered by southward-moving weak cold air from Tibetan Plateau (in Chinese). J. Trop. Meteor., 30, 111118.

    • Search Google Scholar
    • Export Citation
  • Liu, L., Q. Xu, P. Zhang, and S. Liu, 2008: Automated detection of contaminated radar image pixels in mountain areas. Adv. Atmos. Sci., 25, 778790, https://doi.org/10.1007/s00376-008-0778-x.

    • Search Google Scholar
    • Export Citation
  • Liu, X., and X. Guo, 2012: Analysis and numerical simulation research on severe surface wind formation mechanism and structural characteristics of a squall line case (in Chinese). Chin. J. Atmos. Sci., 36, 11501164, https://doi.org/10.3878/j.issn.1006-9895.2012.11212.

    • Search Google Scholar
    • Export Citation
  • Liu, Y., X. Yao, J. Fei, X. Yang, and J. Sun, 2021: Characteristics of mesoscale convective systems during the warm season over the Tibetan Plateau based on FY-2 satellite datasets. Int. J. Climatol., 41, 23012315, https://doi.org/10.1002/joc.6959.

    • Search Google Scholar
    • Export Citation
  • Lombardo, K. A., and B. A. Colle, 2010: The spatial and temporal distribution of organized convective structures over the northeast and their ambient conditions. Mon. Wea. Rev., 138, 44564474, https://doi.org/10.1175/2010MWR3463.1.

    • Search Google Scholar
    • Export Citation
  • Luo, L., M. Xue, and K. Zhu, 2020 : The initiation and organization of a severe hail-producing mesoscale convective system in East China: A numerical study. J. Geophys. Res. Atmos., 125, e2020JD032606, https://doi.org/10.1029/2020JD032606.

  • Luo, Y., Y. Gong, and D.-L. Zhang, 2014: Initiation and organizational modes of an extreme-rain-producing mesoscale convective system along a mei-yu front in East China. Mon. Wea. Rev., 142, 203221, https://doi.org/10.1175/MWR-D-13-00111.1.

    • Search Google Scholar
    • Export Citation
  • Luo, Y., and Coauthors, 2020: Science and prediction of heavy rainfall over China: Research progress since the reform and opening-up of new China. J. Meteor. Res., 34, 427459, https://doi.org/10.1007/s13351-020-0006-x.

    • Search Google Scholar
    • Export Citation
  • Ma, R., J. Sun, and X. Yang, 2021: A 7-yr climatology of the initiation, decay, and morphology of severe convective storms during the warm season over North China. Mon. Wea. Rev., 149, 25992612, https://doi.org/10.1175/MWR-D-20-0087.1.

    • Search Google Scholar
    • Export Citation
  • Maddox, R. A., 1980: Mesoscale convective complexes. Bull. Amer. Meteor. Soc., 61, 13741400, https://doi.org/10.1175/1520-0477(1980)061<1374:MCC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Maddox, R. A., 1981: Satellite depiction of the life cycle of a mesoscale convective complex. Mon. Wea. Rev., 109, 15831586, https://doi.org/10.1175/1520-0493(1981)109<1583:SDOTLC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Maddox, R. A., 1983: Large-scale meteorological conditions associated with midlatitude, mesoscale convective complexes. Mon. Wea. Rev., 111, 14751493, https://doi.org/10.1175/1520-0493(1983)111<1475:LSMCAW>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Maddox, R. A., C. F. Chappell, and L. R. Hoxit, 1979: Synoptic and meso-α scale aspects of flash flood events. Bull. Amer. Meteor. Soc., 60, 115123, https://doi.org/10.1175/1520-0477-60.2.115.

    • Search Google Scholar
    • Export Citation
  • Meng, Z., and Y. Zhang, 2012: On the squall lines preceding landfalling tropical cyclones in China. Mon. Wea. Rev., 140, 445470, https://doi.org/10.1175/MWR-D-10-05080.1.

    • Search Google Scholar
    • Export Citation
  • Meng, Z., D. Yan, and Y. Zhang, 2013: General features of squall lines in East China. Mon. Wea. Rev., 141, 16291647, https://doi.org/10.1175/MWR-D-12-00208.1.

    • Search Google Scholar
    • Export Citation
  • Orlanski, I., 1975: A rational subdivision of scales for atmospheric processes. Bull. Amer. Meteor. Soc., 56, 527530, https://doi.org/10.1175/1520-0477-56.5.527.

    • Search Google Scholar
    • Export Citation
  • Pacey, G. P., D. M. Schultz, and L. Garcia-Carreras, 2021: Severe convective windstorms in Europe: Climatology, preconvective environments, and convective mode. Wea. Forecasting, 36, 237252, https://doi.org/10.1175/WAF-D-20-0075.1.

    • Search Google Scholar
    • Export Citation
  • Parker, M. D., and R. H. Johnson, 2000: Organizational modes of midlatitude mesoscale convective systems. Mon. Wea. Rev., 128, 34133436, https://doi.org/10.1175/1520-0493(2001)129<3413:OMOMMC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Pettet, C. R., and R. H. Johnson, 2003: Airflow and precipitation structure of two leading stratiform mesoscale convective systems determined from operational datasets. Wea. Forecasting, 18, 685699, https://doi.org/10.1175/1520-0434(2003)018<0685:AAPSOT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Ren, Z., and Coauthors, 2015: Development of three step quality control system of real time observation data from AWS in China (in Chinese). Meteor. Mon., 41, 12681277, https://doi.org/10.7519/j.issn.1000-0526.2015.10.010.

    • Search Google Scholar
    • Export Citation
  • Schiesser, H. H., R. A. Houze Jr., and H. Huntrieser, 1995: The mesoscale structure of severe precipitation systems in Switzerland. Mon. Wea. Rev., 123, 20702097, https://doi.org/10.1175/1520-0493(1995)123<2070:TMSOSP>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Schoen, J. M., and W. S. Ashley, 2011: A climatology of fatal convective wind events by storm type. Wea. Forecasting, 26, 109121, https://doi.org/10.1175/2010WAF2222428.1.

    • Search Google Scholar
    • Export Citation
  • Schumacher, R. S., and K. L. Rasmussen, 2020: The formation, character and changing nature of mesoscale convective systems. Nat. Rev. Earth Environ., 1, 300314, https://doi.org/10.1038/s43017-020-0057-7.

    • Search Google Scholar
    • Export Citation
  • Smith, B. T., R. L. Thompson, J. S. Grams, C. Broyles, and H. E. Brooks, 2012: Convective modes for significant severe thunderstorms in the contiguous United States. Part I: Storm classification and climatology. Wea. Forecasting, 27, 11141135, https://doi.org/10.1175/WAF-D-11-00115.1.

    • Search Google Scholar
    • Export Citation
  • Snively, D. V., and W. A. Gallus Jr., 2014: Prediction of convective morphology in near-cloud-permitting WRF Model simulations. Wea. Forecasting, 29, 130149, https://doi.org/10.1175/WAF-D-13-00047.1.

    • Search Google Scholar
    • Export Citation
  • Tao, Z., H. Wang, X. Wang, and Y. Ma, 1998: A survey of meso-α-scale convective systems over China during 1995 (in Chinese). Acta Meteor. Sin., 56, 166177, https://doi.org/10.11676/qxxb1998.016.

    • Search Google Scholar
    • Export Citation
  • Trapp, R. J., S. A. Tessendorf, E. S. Godfrey, and H. E. Brooks, 2005: Tornadoes from squall lines and bow echoes. Part I: Climatological distribution. Wea. Forecasting, 20, 2334, https://doi.org/10.1175/WAF-835.1.

    • Search Google Scholar
    • Export Citation
  • Wallace, J. M., 1975: Diurnal variations in precipitation and thunderstorm frequency over the conterminous United States. Mon. Wea. Rev., 103, 406419, https://doi.org/10.1175/1520-0493(1975)103<0406:DVIPAT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Wang, H., Y. Luo, and B. J.-D. Jou, 2014: Initiation, maintenance, and properties of convection in an extreme rainfall event during SCMREX: Observational analysis. J. Geophys. Res. Atmos, 119, 13 206–13 232, https://doi.org/10.1002/2014JD022339.

    • Search Google Scholar
    • Export Citation
  • Wang, X., and C. Cui, 2011: A number advances of the research on the heavy rain mesoscale convective systems (in Chinese). Torrential Rain Disasters, 30, 97106.

    • Search Google Scholar
    • Export Citation
  • Wu, N., X. Ding, Z. Wen, G. Chen, Z. Meng, L. Lin, and J. Min, 2020: Contrasting frontal and warm-sector heavy rainfalls over South China during the early-summer rainy season. Atmos. Res., 235, 104693, https://doi.org/10.1016/j.atmosres.2019.104693.

    • Search Google Scholar
    • Export Citation
  • Yang, X., and J. Sun, 2018: Organizational modes of severe wind-producing convective systems over North China. Adv. Atmos. Sci., 35, 540549, https://doi.org/10.1007/s00376-017-7114-2.

    • Search Google Scholar
    • Export Citation
  • Yang, X., J. Sun, and Y. Zheng, 2017: A 5-yr climatology of severe convective wind events over China. Wea. Forecasting, 32, 12891299, https://doi.org/10.1175/WAF-D-16-0101.1.

    • Search Google Scholar
    • Export Citation
  • Yu, X., and Y. Zheng, 2020: Advances in severe convection research and operation in China. J. Meteor. Res., 34, 189217, https://doi.org/10.1007/s13351-020-9875-2.

    • Search Google Scholar
    • Export Citation
  • Zhang, S., S. Tao, Q. Zhang, and J. Wei, 2002: Large and meso-α scale characteristics of intense rainfall in the mid- and lower reaches of the Yangtze River. Chin. Sci. Bull., 47, 779786, https://doi.org/10.1360/02tb9176.

    • Search Google Scholar
    • Export Citation
  • Zheng, L., J. Sun, and J. Wei, 2011: The diurnal variation of thunder events in China (in Chinese). Torrential Rain Disasters, 30, 137144.

    • Search Google Scholar
    • Export Citation
  • Zheng, L., J. Sun, X. Zhang, and C. Liu, 2013: Organizational modes of mesoscale convective systems over central East China. Wea. Forecasting, 28, 10811098, https://doi.org/10.1175/WAF-D-12-00088.1.

    • Search Google Scholar
    • Export Citation
Save
  • Anderson, C. J., and R. W. Arritt, 1998: Mesoscale convective complexes and persistent elongated convective systems over the United States during 1992 and 1993. Mon. Wea. Rev., 126, 578599, https://doi.org/10.1175/1520-0493(1998)126<0578:MCCAPE>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Anderson, G. D., 2010: The first weather satellite picture. Weather, 65, 8787, https://doi.org/10.1002/wea.550.

  • Bai, L., G. Chen, and L. Huang, 2020: Image processing of radar mosaics for the climatology of convection initiation in South China. J. Appl. Meteor. Climatol., 59, 6581, https://doi.org/10.1175/JAMC-D-19-0081.1.

    • Search Google Scholar
    • Export Citation
  • Blackadar, A. K., 1957: Boundary layer wind maxima and their significance for the growth of nocturnal inversions. Bull. Amer. Meteor. Soc., 38, 283290, https://doi.org/10.1175/1520-0477-38.5.283.

    • Search Google Scholar
    • Export Citation
  • Bluestein, H. B., and M. H. Jain, 1985: Formation of mesoscale lines of precipitation: Severe squall lines in Oklahoma during the spring. J. Atmos. Sci., 42, 17111732, https://doi.org/10.1175/1520-0469(1985)042<1711:FOMLOP>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Carlberg, B. R., W. A. Gallus Jr., and K. J. Franz, 2018: A preliminary examination of WRF ensemble prediction of convective mode evolution. Wea. Forecasting, 33, 783798, https://doi.org/10.1175/WAF-D-17-0149.1.

    • Search Google Scholar
    • Export Citation
  • Chen, J., Y. Zheng, X. Zhang, and P. Zhu, 2013: Distribution and diurnal variation of warm-season short-duration heavy rainfall in relation to the MCSs in China. Acta Meteor. Sin., 27, 868888, https://doi.org/10.1007/s13351-013-0605-x.

    • Search Google Scholar
    • Export Citation
  • Chen, Y., Y. Luo, and B. Liu, 2022: General features and synoptic-scale environments of mesoscale convective systems over South China during the 2013–2017 pre-summer rainy seasons. Atmos. Res., 266, 105954, https://doi.org/10.1016/j.atmosres.2021.105954.

    • Search Google Scholar
    • Export Citation
  • Cotton, W. R., R. L. George, P. J. Wetzel, and R. L. McAnelly, 1983: A long-lived mesoscale convective complex. Part I: The mountain–generated component. Mon. Wea. Rev., 111, 18931918, https://doi.org/10.1175/1520-0493(1983)111<1893:ALLMCC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Du, Y., and G. Chen, 2018: Heavy rainfall associated with double low-level jets over southern China. Part I: Ensemble-based analysis. Mon. Wea. Rev., 146, 38273844, https://doi.org/10.1175/MWR-D-18-0101.1.

    • Search Google Scholar
    • Export Citation
  • Du, Y., and G. Chen, 2019a: Heavy rainfall associated with double low-level jets over southern China. Part II: Convection initiation. Mon. Wea. Rev., 147, 543565, https://doi.org/10.1175/MWR-D-18-0102.1.

    • Search Google Scholar
    • Export Citation
  • Du, Y., and G. Chen, 2019b: Climatology of low-level jets and their impact on rainfall over southern China during the early-summer rainy season. J. Climate, 32, 88138833, https://doi.org/10.1175/JCLI-D-19-0306.1.

    • Search Google Scholar
    • Export Citation
  • Duda, J. D., and W. A. Gallus Jr., 2010: Spring and summer Midwestern severe weather reports in supercells compared to other morphologies. Wea. Forecasting, 25, 190206, https://doi.org/10.1175/2009WAF2222338.1.

    • Search Google Scholar
    • Export Citation
  • Feng, S., Q. Hu, and W. Qian, 2004: Quality control of daily meteorological data in China, 1951–2000: A new dataset. Int. J. Climatol., 24, 853870, https://doi.org/10.1002/joc.1047.

    • Search Google Scholar
    • Export Citation
  • Fujita, T. T., 1978: Manual of downburst identification for Project NIMROD. NOAA/NASA Tech. Rep. NRC-04-74-239, 118 pp.

  • Gallus, W. A., Jr., N. A. Snook, and E. V. Johnson, 2008: Spring and summer severe weather reports over the Midwest as a function of convective mode: A preliminary study. Wea. Forecasting, 23, 101113, https://doi.org/10.1175/2007WAF2006120.1.

    • Search Google Scholar
    • Export Citation
  • Geerts, B., 1998: Mesoscale convective systems in the southeast United States during 1994–95: A survey. Wea. Forecasting, 13, 860869, https://doi.org/10.1175/1520-0434(1998)013<0860:MCSITS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Hoecker, W. H., Jr., 1963: Three southerly low-level jet systems delineated by the weather bureau special pibal network of 1961. Mon. Wea. Rev., 91, 573582, https://doi.org/10.1175/1520-0493(1963)091<0573:TSLJSD>2.3.CO;2.

    • Search Google Scholar
    • Export Citation
  • Houze, R. A., Jr., 1989: Observed structure of mesoscale convective systems and implications for large-scale heating. Quart. J. Roy. Meteor. Soc., 115, 425461, https://doi.org/10.1002/qj.49711548702.

    • Search Google Scholar
    • Export Citation
  • Houze, R. A., Jr., B. F. Smull, and P. Dodge, 1990: Mesoscale organization of springtime rainstorms in Oklahoma. Mon. Wea. Rev., 118, 613654, https://doi.org/10.1175/1520-0493(1990)118<0613:MOOSRI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Jirak, I. L., W. R. Cotton, and R. L. McAnelly, 2003: Satellite and radar survey of mesoscale convective system development. Mon. Wea. Rev., 131, 24282449, https://doi.org/10.1175/1520-0493(2003)131<2428:SARSOM>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Li, J., B. Wang, and D. Wang, 2012: The characteristics of mesoscale convective systems (MCSs) over East Asia in warm seasons. Atmos. Oceanic Sci. Lett., 5, 102107, https://doi.org/10.1080/16742834.2012.11446973.

    • Search Google Scholar
    • Export Citation
  • Lin, Z., M. Lin, K. Lin, and H. Luo, 2014: A convective process of quasi-stationary front triggered by southward-moving weak cold air from Tibetan Plateau (in Chinese). J. Trop. Meteor., 30, 111118.

    • Search Google Scholar
    • Export Citation
  • Liu, L., Q. Xu, P. Zhang, and S. Liu, 2008: Automated detection of contaminated radar image pixels in mountain areas. Adv. Atmos. Sci., 25, 778790, https://doi.org/10.1007/s00376-008-0778-x.

    • Search Google Scholar
    • Export Citation
  • Liu, X., and X. Guo, 2012: Analysis and numerical simulation research on severe surface wind formation mechanism and structural characteristics of a squall line case (in Chinese). Chin. J. Atmos. Sci., 36, 11501164, https://doi.org/10.3878/j.issn.1006-9895.2012.11212.

    • Search Google Scholar
    • Export Citation
  • Liu, Y., X. Yao, J. Fei, X. Yang, and J. Sun, 2021: Characteristics of mesoscale convective systems during the warm season over the Tibetan Plateau based on FY-2 satellite datasets. Int. J. Climatol., 41, 23012315, https://doi.org/10.1002/joc.6959.

    • Search Google Scholar
    • Export Citation
  • Lombardo, K. A., and B. A. Colle, 2010: The spatial and temporal distribution of organized convective structures over the northeast and their ambient conditions. Mon. Wea. Rev., 138, 44564474, https://doi.org/10.1175/2010MWR3463.1.

    • Search Google Scholar
    • Export Citation
  • Luo, L., M. Xue, and K. Zhu, 2020 : The initiation and organization of a severe hail-producing mesoscale convective system in East China: A numerical study. J. Geophys. Res. Atmos., 125, e2020JD032606, https://doi.org/10.1029/2020JD032606.

  • Luo, Y., Y. Gong, and D.-L. Zhang, 2014: Initiation and organizational modes of an extreme-rain-producing mesoscale convective system along a mei-yu front in East China. Mon. Wea. Rev., 142, 203221, https://doi.org/10.1175/MWR-D-13-00111.1.

    • Search Google Scholar
    • Export Citation
  • Luo, Y., and Coauthors, 2020: Science and prediction of heavy rainfall over China: Research progress since the reform and opening-up of new China. J. Meteor. Res., 34, 427459, https://doi.org/10.1007/s13351-020-0006-x.

    • Search Google Scholar
    • Export Citation
  • Ma, R., J. Sun, and X. Yang, 2021: A 7-yr climatology of the initiation, decay, and morphology of severe convective storms during the warm season over North China. Mon. Wea. Rev., 149, 25992612, https://doi.org/10.1175/MWR-D-20-0087.1.

    • Search Google Scholar
    • Export Citation
  • Maddox, R. A., 1980: Mesoscale convective complexes. Bull. Amer. Meteor. Soc., 61, 13741400, https://doi.org/10.1175/1520-0477(1980)061<1374:MCC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Maddox, R. A., 1981: Satellite depiction of the life cycle of a mesoscale convective complex. Mon. Wea. Rev., 109, 15831586, https://doi.org/10.1175/1520-0493(1981)109<1583:SDOTLC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Maddox, R. A., 1983: Large-scale meteorological conditions associated with midlatitude, mesoscale convective complexes. Mon. Wea. Rev., 111, 14751493, https://doi.org/10.1175/1520-0493(1983)111<1475:LSMCAW>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Maddox, R. A., C. F. Chappell, and L. R. Hoxit, 1979: Synoptic and meso-α scale aspects of flash flood events. Bull. Amer. Meteor. Soc., 60, 115123, https://doi.org/10.1175/1520-0477-60.2.115.

    • Search Google Scholar
    • Export Citation
  • Meng, Z., and Y. Zhang, 2012: On the squall lines preceding landfalling tropical cyclones in China. Mon. Wea. Rev., 140, 445470, https://doi.org/10.1175/MWR-D-10-05080.1.

    • Search Google Scholar
    • Export Citation
  • Meng, Z., D. Yan, and Y. Zhang, 2013: General features of squall lines in East China. Mon. Wea. Rev., 141, 16291647, https://doi.org/10.1175/MWR-D-12-00208.1.

    • Search Google Scholar
    • Export Citation
  • Orlanski, I., 1975: A rational subdivision of scales for atmospheric processes. Bull. Amer. Meteor. Soc., 56, 527530, https://doi.org/10.1175/1520-0477-56.5.527.

    • Search Google Scholar
    • Export Citation
  • Pacey, G. P., D. M. Schultz, and L. Garcia-Carreras, 2021: Severe convective windstorms in Europe: Climatology, preconvective environments, and convective mode. Wea. Forecasting, 36, 237252, https://doi.org/10.1175/WAF-D-20-0075.1.

    • Search Google Scholar
    • Export Citation
  • Parker, M. D., and R. H. Johnson, 2000: Organizational modes of midlatitude mesoscale convective systems. Mon. Wea. Rev., 128, 34133436, https://doi.org/10.1175/1520-0493(2001)129<3413:OMOMMC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Pettet, C. R., and R. H. Johnson, 2003: Airflow and precipitation structure of two leading stratiform mesoscale convective systems determined from operational datasets. Wea. Forecasting, 18, 685699, https://doi.org/10.1175/1520-0434(2003)018<0685:AAPSOT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Ren, Z., and Coauthors, 2015: Development of three step quality control system of real time observation data from AWS in China (in Chinese). Meteor. Mon., 41, 12681277, https://doi.org/10.7519/j.issn.1000-0526.2015.10.010.

    • Search Google Scholar
    • Export Citation
  • Schiesser, H. H., R. A. Houze Jr., and H. Huntrieser, 1995: The mesoscale structure of severe precipitation systems in Switzerland. Mon. Wea. Rev., 123, 20702097, https://doi.org/10.1175/1520-0493(1995)123<2070:TMSOSP>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Schoen, J. M., and W. S. Ashley, 2011: A climatology of fatal convective wind events by storm type. Wea. Forecasting, 26, 109121, https://doi.org/10.1175/2010WAF2222428.1.

    • Search Google Scholar
    • Export Citation
  • Schumacher, R. S., and K. L. Rasmussen, 2020: The formation, character and changing nature of mesoscale convective systems. Nat. Rev. Earth Environ., 1, 300314, https://doi.org/10.1038/s43017-020-0057-7.

    • Search Google Scholar
    • Export Citation
  • Smith, B. T., R. L. Thompson, J. S. Grams, C. Broyles, and H. E. Brooks, 2012: Convective modes for significant severe thunderstorms in the contiguous United States. Part I: Storm classification and climatology. Wea. Forecasting, 27, 11141135, https://doi.org/10.1175/WAF-D-11-00115.1.

    • Search Google Scholar
    • Export Citation
  • Snively, D. V., and W. A. Gallus Jr., 2014: Prediction of convective morphology in near-cloud-permitting WRF Model simulations. Wea. Forecasting,