Thunderstorms with Extreme Lightning Activity in China: Climatology, Synoptic Patterns, and Convective Parameters

Abstract

Intense convection is often accompanied by high-frequency lightning and is highly prone to producing heavy rainfall, strong winds, hail, and tornadoes, frequently resulting in significant damage and loss of life. It is necessary to understand the mechanisms and meteorological conditions of intense convection. This study utilizes the Thunderstorm Feature Dataset from 2010–2018 to analyze the characteristics of thunderstorms with extreme lightning activity (TELAs), defined as thunderstorms whose lightning frequency ranks in the top 1%. Four regions with relatively high thunderstorm activity were selected for analysis: Northeast China (NEC), North China (NC), South China (SC), and the Tibetan Plateau (TP). In NEC, TELAs primarily occur just west of upper-level westerly troughs (UWT), including cold vortices. In NC, TELAs are mainly associated with UWT and subtropical highs (STH). In SC, TELAs are related to frontal systems, easterly waves, tropical cyclones, and STH. In TP, TELAs are generated by TP vortices. Before the TELA process, vertically integrated moisture divergence (VIMD) and convective available potential energy (CAPE) show the most notable anomalies. Except for the TP, TELAs are typically located between centers of anomalies with positive and negative geopotential height (500 hPa) and near centers of anomalies with positive CAPE and negative VIMD, accompanied by notable increases in surface temperature and wind speed. These findings offer a valuable reference for the early warning and forecasting of intense convection.

1. Introduction

According to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC), continued global warming is projected to increase the frequency and intensity of extreme weather events in more regions in the future [1]. In most studies, intense convection is defined based on extreme weather events such as heavy rainfall, strong winds, hail, and tornadoes and their intensity [2]. Consequently, the characteristics and physical mechanisms of intense convection have received increasing attention from researchers in recent years [3,4,5,6,7,8,9,10,11].

In this study, we define intense convection by its lightning frequency, referring to them as Thunderstorms with Extreme Lightning Activity (TELA). Lightning can be an important indicator of storm intensity due to its close relationship with the convective intensity that determines the disaster risk of thunderstorms. Generally, thunderstorms with stronger convection can lift more water vapor and precipitation particles to higher altitudes, increase the number of ice particles within the cloud, and improve collision rebound and charge-transfer efficiencies between the ice particles [12], which leads to stronger electrification and high-frequency lightning activity. Several studies have shown that lightning frequency is a good proxy for the convective intensity of thunderstorms [13,14,15,16,17]. For instance, Barthe et al. [14] found a strong correlation between the maximum updraft velocity and the flash rate of severe storms. Liu et al. [16] discovered that in tropical and subtropical regions, the correlation coefficients between flash rates and the volumes and areas with radar reflectivity greater than 30, 35, and 40 dBZ in the mixed-phase region (−5 °C to −35 °C) exceed 0.8. Strong convection is a main cause not only of intense lightning activity but also of other extreme weather events. Thunderstorms with more intense lightning activity, especially total lightning (intracloud and cloud-to-ground lightning), are more likely to be accompanied by extreme weather events [18,19,20,21]. Thus, defining intense convection based on extreme lightning activity has a solid theoretical foundation, especially considering that lightning itself is one of the significant extreme weather events [22,23,24,25,26]. Moreover, the development of lightning detection technology and the continuous expansion of lightning observation networks have provided a robust observational basis for identifying extreme thunderstorms through lightning location data.

The formation, evolution, internal structure, intensity, movement, and organization of convective systems are closely related to environmental conditions [6,27,28,29,30,31]. Mora et al. [32] classified extreme thunderstorm days (>2000 cloud-to-ground flashes per day) into four types based on average upper-level synoptic patterns—shortwave troughs, lows, cyclonic vortex, and ridge—conducted a comprehensive analysis of the thermodynamic conditions and synoptic patterns for each type and found that thunderstorms associated with shortwave troughs and lows were more likely to produce extreme weather events such as hail and heavy rainfall. Liu et al. [8] used ERA-Interim reanalysis data to analyze synoptic patterns in seven regions where high-flash-rate thunderstorms are most active (south-central United States, northwest Mexico, Argentina, the Sahel, Congo, Colombia, and the Himalayas). They comprehensively discussed the lifting mechanisms, convective instability, and moisture conditions, finding that thermodynamic features associated with high-flash-rate thunderstorms include CAPE values, small to moderate CIN values, and abundant moisture convergence with low-level warm-air advection. Thunderstorms with high flash rates tend to have larger CAPE values and wind shear compared to those with low flash rates. Favorable environmental conditions for high-flash-rate thunderstorms vary by region. For example, in tropical regions, high-flash-rate thunderstorms are characterized by higher CAPE, lower CIN, and weaker wind shear, while in subtropical regions, moderate CAPE, CIN, and stronger low-to-mid-level wind shear are more conducive to high-flash-rate thunderstorms.

Previous research mainly focused on case studies of individual extreme thunderstorms [33,34,35,36,37,38,39], while there are still many deficiencies in statistical studies on the climatological characteristics of TELAs and the associated synoptic patterns. This study utilizes the Thunderstorm Feature Dataset (TFD) developed from the Black Body Temperature (TBB) and Cloud Classification products of the Fengyun-2E (FY-2E) satellite, combined with lightning location data from WWLLN, and analyzes the spatial and temporal distribution characteristics of TELAs in China. In addition, the synoptic patterns and atmospheric parameters were examined using ERA5 reanalysis data, revealing the environmental conditions favorable to TELAs. This research will deepen our understanding of the mechanisms of TELAs and provide references for monitoring and forecasting intense convection.

2. Materials and Methods

2.1. Thunderstorm Feature Dataset (TFD)

The TFD developed by Ma and Zheng [40] was established based on the TBB product from the FY-2E geostationary satellite and WWLLN lightning location data. In the TFD, thunderstorm clouds are defined as cloud areas with TBB $\le -$32 °C and where WWLLN lightning occurs within an hour. Figure 1 illustrates the identification process of a thunderstorm cloud at 06:00 UTC on 31 May 2015. The black solid line encloses the cloud area with TBB $\le -$32 °C of this thunderstorm sample. The red solid line outlines the strong convective core (SCC) within the thunderstorm, defined as areas with TBB $\le -$52 °C. The blue solid line represents an ellipse fitted using the method of minimum second-order moments. The yellow “+” indicates WWLLN lightning strokes detected between 20 min before and 40 min after the current time. The green solid line outlines clusters of WWLLN lightning strokes, which were grouped using the DBSCAN algorithm [41], and contain more than two lightning strokes within a radius of 12 km. The solid deep green line marks the cluster with the highest frequency of WWLLN lightning strokes, which is defined in this study as the Lightning Active Area (LAA) of the thunderstorm. The center of LAA is defined as the Lightning Active Center (LAC), as indicated by the deep green “+” symbol in Figure 1.

The TFD ranges from 2010 to 2018, with some data gaps in January 2010 and June 2015. It covers the entire full-disk observation area of the FY-2E satellite, and its spatial resolution matches that of the FY-2E TBB data. The spatial resolution at the satellite’s sub-point (located at 104.5 °E on the equator before July 2015 and 86.5 °E on the equator thereafter) is 5 km. Since the time resolution of TFD is 1 h, the thunderstorm samples in TFD are defined as Thunderstorm Hour Samples (THSs). The TFD records the time and location information for each THS, and TBB feature, as well as parameters that characterize the form of thunderstorm clouds, including the major and minor axes’ lengths, the rotation angles of the fitted ellipse, the area of the thunderstorm cloud (the area with TBB $\le -$32 °C), and the average, minimum, and median values of TBB within the cloud. Additionally, it stores the number and location of SCCs and lightning clusters within the cloud. TFD has been used in previous studies, providing valuable insights into thunderstorm activity and structure, and the data quality has been validated [40,42,43].

2.2. Study Areas

Based on the TFD, the spatial distribution of thunderstorm activity in China and its surrounding areas is shown in Figure 2. Note that the THS count reflects not only the number of thunderstorm samples but also the duration of thunderstorm processes. For example, a single thunderstorm sample lasting three hours is considered equal to three individual thunderstorm samples within one hour. Thunderstorms are most active in southern China, southwestern China, and the central and eastern Tibetan Plateau. Northern China exhibits weak thunderstorm activity, although the North China Plain and Northeast China are regions with relatively higher thunderstorm activity. The weakest thunderstorm activity is observed in western China, including Xinjiang, northern Qinghai, Gansu, and western Inner Mongolia. Because of the complex and diverse climatic environments in China, the thunderstorm activity and the characteristics of thunderstorms vary significantly across different climate zones [42,44]. Based on the spatial distribution of thunderstorms, this study selects four representative regions with relatively high thunderstorm activity: Northeast China (NEC), North China (NC), South China (SC), and the Tibetan Plateau (TP), as indicated by the black rectangles in Figure 2.

THSs with LACs located within each study area were selected for analysis. According to the quality control method used by Ma and Zheng [40], approximately 10% of anomalous THSs (characterized by big cloud areas but very few WWLLN lightning strokes) were removed from each region. The final number of THSs from highest to lowest across the four areas were as follows: 31460 in SC, 12898 over the TP, 8274 in NC, and 5758 in NEC.

2.3. The Definition of Thunderstorm with Extreme Lightning Activity (TELA)

Since the WWLLN lightning within thunderstorm clouds often presents multiple lightning clusters, as shown in Figure 1, we used the lightning frequency within the LAA (the region enclosed by the dark green line in Figure 1) to identify thunderstorms with intense convective activity, which we refer to as TELAs. The probability and cumulative probability distributions of the WWLLN lightning frequency within the LAA for the THSs in each study area are shown in Figure 3. The lightning frequency generally follows a log-normal distribution, with more THSs distributed in intervals with low lightning frequency within the LAA. The lightning frequency within the LAA rankings from highest to lowest are SC, NEC, NC, and TP, with mean values of 63, 52, 46, and 10, respectively. The 99th-percentile values are approximately 744, 639, 517, and 56, respectively, as indicated by the blue vertical lines in Figure 3. Overall, the four study areas exhibit significant differences in thunderstorm activity, lightning frequency within the LAA, and its distribution. These differences can be attributed to varying climatic backgrounds, indicating differences in the atmospheric conditions that trigger thunderstorms. Considering these regional differences, as well as the potential impacts of the uneven spatial distribution of WWLLN detection efficiency [45,46], we do not use a fixed value of WWLLN lightning frequencies to define TELAs. Instead, TELAs are defined as the THS with the top 1% of lightning frequency within the LAA in each study area. The number of TELAs is 315 for SC, 136 for the TP, 84 for NC, and 58 for NEC.

Using hourly precipitation data provided by national weather stations of the China Meteorological Administration, TELAs were matched with the weather stations located within the TELA cloud area and its fitted ellipse. The maximum hourly precipitation recorded by these stations within a 7-h period (3 h prior to the TELA + 1 h of the TELA + 3 h after the TELA) is presented in Figure 4. The results show that TELAs in SC are associated with the highest average maximum hourly precipitation, reaching 71 mm/h. NC and NEC exhibit lower mean values of 57 mm/h and 40 mm/h, respectively. TELAs over the TP demonstrate the lowest average maximum hourly precipitation, with only 7 mm/h. Additionally, 81%, 91%, and 100% of TELAs in NEC, NC, and SC are associated with maximum precipitation exceeding 20 mm/h, meeting the criterion for short-duration heavy rainfall [47]. However, only about 3% of the TELAs over the TP meet the 20 mm/h threshold. It is important to note that precipitation over the TP is significantly lower than in eastern China [48], primarily due to the small scale and short lifespans of convective systems over the TP [49]. Li [50] used the 95th percentile as the criterion for extreme precipitation over the TP, finding that it ranged from 1.96 to 5.30 mm/h, with a mean value of just 3.54 mm/h. Applying this criterion, 73% of the TELAs in this study could be considered as extreme precipitation. The correlation between TELAs and extreme precipitation further supports the validity of using lightning activity as an indicator for identifying extreme thunderstorms.

2.4. Reanalysis Data and Selection of Convective Parameters

ERA5 reanalysis data provided by the European Centre for Medium-Range Weather Forecasts (ECMWF) are utilized to investigate the atmospheric environment conducive to TELAs. ERA5 offers a comprehensive dataset of climate variables for the atmosphere, land, and oceans, with a spatial resolution of 0.25° × 0.25°. In this study, parameters corresponding to 6 h before each TELA from two datasets (ERA5 hourly data on pressure levels and single levels) were mainly used. The parameters are listed in Table 1, and certain convective parameters, such as 0–6 km wind shear, mid-level average relative humidity, and the K-index, were calculated from the ERA5 reanalysis data.

The analysis of the atmospheric background of strong convection is often based on mesoscale synoptic analysis [8,29,32,51]. In this study, the synoptic pattern analysis and classification of TELAs were used to duplicate samples belonging to the same weather process. The numbers of TELA processes in each study area are as follows: 33 in NEC, 51 in NC, 165 in SC, and 104 over the TP. The specific analysis steps were as follows: (1) ERA5 reanalysis data, including wind, temperature, pressure, and humidity parameters at pressure levels, were selected to plot the synoptic pattern for each TELA process. TELA processes and their synoptic pattern in each area were analyzed and classified. (2) For each synoptic type, composite analysis was performed over a 30° × 30° region centered on the LAC position of each TELA process, and the conceptual diagrams were generated for supplementary analysis.

To quantify the deviation of meteorological parameters from their standard climatological normal (SCN) before TELA processes, anomalies and standardized anomalies (N) of each parameter were calculated. Referring to Hart and Grumm (2001) [52], we used the data of a 21-day window centered on the TELA process (including the event day itself and 10 days before and after), spanning the 30-year period from 1991 to 2020, which can ensure the representative of SCN and standard anomalies. For calculating standardized anomalies we used Equations (1) and (2),

(1)$N={\displaystyle \frac{X-\mu }{\sigma }}$

(2)$\sigma =\sqrt{{\displaystyle \frac{1}{n-1}}\sum _{i=1}^n{(X_i-\mu )}^2}$

where X represents the meteorological parameter, while $\mu$ denotes its SCN. The anomaly is calculated as X−$\mu$, and $\sigma$ represents the standard deviation of the parameter. The sample size (n) for each TELA process is determined to be 630 samples (30 years × 21 days). The standardized anomalies (N) can measure the abnormality of factors in extreme events. A larger absolute value of standardized anomalies indicates a greater the anomalies of the factor from its SCN [53]. According to the standard normal distribution, the probability of standardized anomalies exceeding an absolute value of 1 is roughly 30%, while the probability of standardized anomalies exceeding an absolute value of 2 is about 5%, signifying a notable anomaly from the SCN.

3. Results

3.1. Climatological Characteristics of TELAs

Figure 5a illustrates the distribution of the percentage of monthly THS counts, showing that thunderstorm activity in all regions is more active during the warm season (May to September). In NEC, thunderstorm activity peaks earliest, reaching its maximum in June. Subsequently, thunderstorm activity peaks in July, August, and September in NC, SC, and the TP, respectively. Figure 5b shows the distribution of the percentage of monthly TELA counts, indicating that TELA activity is also active during the warm season. The peak of monthly TELA activity occurs in July and August for NEC and NC, respectively, one month later than the peak of total thunderstorm activity. In SC, the distribution of monthly TELA activity exhibits a double-peak pattern, with the highest peak in May and a secondary peak in August, which contrasts with the single peak in total thunderstorm activity in August. Over the TP, the monthly variation in TELA activity is similar to total thunderstorm activity.

In terms of diurnal variation (Figure 6), the peak times of total thunderstorm activity in all four study areas occur in the afternoon (Figure 6a), indicating the function of surface heating in triggering convection. Figure 6b shows that TELA activity in NEC and NC is generally active from 05:00 to 15:00 Local Solar Time (LST). In SC, TELA activity is present during the daytime (06:00–22:00, LST). These patterns contrast with the diurnal variation of total thunderstorm activity, potentially suggesting that weather systems play a more prominent role in these cases. Over the TP, the diurnal variation of TELA activity still exhibits a clear single-peak pattern, but the peak occurs between 18:00 and 19:00 (LST), which is significantly later than the peak time of total thunderstorm activity, which occurs between 15:00 and 16:00 (LST).

Figure 7 presents boxplots of feature parameters for the total THSs (in blue) and TELA (in orange) across the study areas, with their mean values also indicated in the figure. Figure 7a shows that the cloud area of TELA (defined by the area where TBB $\le -$32 °C) is significantly (5–12 times) larger than that of the total THSs. Among the four regions, TELA in SC exhibits the largest average cloud area at approximately 2.02 × 10⁵ km², while TELA in NC has the smallest average cloud area at about 1.17 × 10⁵ km². Figure 7b demonstrates that the average minimum TBB within TELA clouds is 12–22 °C lower than that of the total THSs in each region, indicating that TELAs have significantly higher cloud tops and stronger convection. In particular, the mean value of minimum TBB for TELA in SC reaches −76.3 °C, while in NEC, it is −58.3 °C, representing the lowest cloud tops. These variations in TELA cloud-top height align with the general trend of decreasing tropopause altitude with increasing latitude.

Figure 7c–e are the boxplots of features related to the SCCs (areas with TBB $\le -$52 °C) within thunderstorm clouds. The results indicate that the counts (Figure 7c), area (Figure 7d), and area proportion (Figure 7e) of SCCs in TELA clouds are significantly larger than those in the total THSs. The counts of SCCs in TELA clouds are higher than those of the total THSs by 2–8. The area of SCCs in TELA clouds is 7–14 times larger than the area for total THSs, while the proportion of the cloud area occupied by SCCs is approximately 5%–13% higher in TELA clouds. The TELA in TP shows the highest count of SCCs, with a mean value of 10. In SC, TELA has the largest SCC area, averaging 9.7 × 10⁴ km² and exhibits the highest proportion of cloud area occupied by SCCs, with a mean value of 53%.

3.2. Main Synoptic Weather Types and Patterns Corresponding to TELAs

The large-scale atmospheric circulation typically provides a favorable background for the thunderstorms, and the diurnal variation characteristics of TELAs (Figure 6b) further indicate that their occurrence should be closely related to the synoptic situation. The synoptic patterns associated with strong convective activity in China are diverse [51,54,55]. To ensure that the composite analysis of environmental meteorological conditions for TELA is representative, it is necessary to first classify the synoptic weather types associated with each TELA process. This section reveals several main synoptic weather types of TELA processes in each study area and conducts a composite analysis of each synoptic pattern. The monthly variation in TELA for each synoptic weather type is shown in Figure 8.

Among the 33 TELA processes in NEC, 31 (94%) were associated with upper-level westerly troughs (UWT), while the remaining two showed no significant weather systems. TELA in NEC typically occurred in the region east of the 500 hPa westerly trough or on the eastern side of a cut-off low-pressure system (cold vortex in NEC) formed after the trough had deepened. TELAs were most frequent from June to September, corresponding to the active period of UWTs and cold vortices [56,57].

In NC, 20 TELA processes (39%) were also associated with the UWT, while another 26 processes (51%) were influenced by the subtropical high (STH), and the remaining 5 TELA processes showed no significant weather systems. From late July to early August, the STH moved northward to 30–35 °N, and the convective activities along its outer edges often triggered intense convection, including thunderstorms, strong winds, and hail in NC. TELAs near the western edge of the STH were most concentrated in July and August. As the STH retreated southward in September, the UWT synoptic pattern once again dominated the TELA processes in NC.

The synoptic patterns associated with TELA processes in SC can be classified into four types: Frontal type (96, 58%), Easterly Wave (EW) type (6, 4%), Tropical Cyclone (TC) type (45, 27%), and STH-edge type (18, 11%). During the pre-flood season, Frontal-type TELAs dominated in SC, triggered by the convergence of cold and warm air flows. After late May, the South China Sea summer monsoon brought more abundant southwest moisture into East Asia, leading to an increase in Frontal-type TELAs. At the same time, SC was located on the western edge of the STH, where favorable thermal conditions caused a rise in STH-edge-type TELAs in June. After July, as the STH moved northward, the region came under the influence of easterly winds, which often led to inverted troughs and a concentration of EW-type TELAs in August. TC-type TELAs primarily occurred from June to September, peaking in August. After September, as the STH gradually retreated southward, STH-edge-type TELAs reappeared.

Previous studies have identified shear lines, low vortices, and low troughs as the primary weather systems influencing convective activity over the TP [58,59]. However, since shear lines and low vortices are often closely associated and frequently occur together, both as components of TP low-pressure systems [60], this study defines all 104 TELA processes over the TP as TP-type TELA, which predominantly occur from June to September, peaking in July.

3.2.1. Synoptic Patterns for TELAs in NEC

Figure 9 presents a conceptual diagram derived from the composite analysis of UWT-type TELAs in NEC (see Figure A1). Overall, the favorable thermodynamic and kinematic environments for UWT-type TELAs are characterized by strong thermal instability caused by intense low-level warm and moist air, combined with dynamic disturbances from the UWT, which lead to strong vertical upward motion.

At the 500 hPa level, the TELA location is predominantly controlled by westerly, high-speed winds. There is a pronounced trough to the west of the TELA, where air convergence and upward motion are prominent. The strong southwestern winds at the 850 hPa level combine with the strong westerly wind at the 500 hPa level, creating significant vertical wind shear, which provides favorable dynamic conditions for the development of strong convection. At the 850 hPa level, there is a low-pressure center to the west of the TELA, with a pronounced warm ridge to the southwest. This enhances the vertical temperature lapse rate, further increasing the convective instability east of the trough and promoting the occurrence of TELAs. The TELA is located precisely within a significant moisture zone, where low-level humidity is high. The presence of abundant warm, moist air from the southwest ensures moisture supply, supporting the development of strong convective activity.

3.2.2. Synoptic Patterns for TELAs in NC

Figure 10a illustrates the conceptual synoptic pattern for UWT-type TELAs in NC. Overall, the weather system configuration of UWT-type TELAs in NC is similar to that in NEC, with the main difference being slightly lower humidity in NC. The high-humidity zone in NC is at the southwest of the TELA.

Figure 10b is the conceptual diagram for STH-edge-type TELAs in NC. At the 500 hPa level, the TELA is at the northwestern side of the 5880 gpm (geopotential meter) contour, with its northern side controlled by the westerlies, which can reach wind speeds of more than 10 m/s. At the 850 hPa level, a weak low-pressure trough exists to the west of the TELA. This is because the westerly flow on the northwestern side of the STH is often accompanied by a westerly trough, which further enhances the occurrence of strong convection. Among the 26 STH-edge-type TELAs in NC, 7 were influenced by both the westerly trough and the STH, providing more favorable dynamic conditions for TELA processes. Under the control of the STH, the surface temperature is relatively high, and a warm ridge is observed at the 850 hPa level, which helps to establish thermal instability. To the southwest of the TELA, there is significant transport of warm, moist air, which ascends along the northern edge of the STH. This moist air interacts with the cold, dry air from the north, further enhancing thermal instability and providing favorable conditions for the development of strong convection.

3.2.3. Synoptic Patterns for TELAs in SC

Figure 11a shows the conceptual synoptic pattern of Frontal-type TELAs in SC. At the 500 hPa level, the TELA location is primarily under the control of westerly flow, with relatively strong wind speeds averaging around 8 m/s. A shallow trough is located to the west of the TELA, and the TELA occurs east of this trough. As the low-pressure trough moves eastward with the westerly flow, it triggers southward cold air at the lower levels. The upward motion is east of the trough. At the 850 hPa level, relative humidity is extremely high near the TELA. Distinct temperature ridges and troughs are observed to the southwest and northeast, respectively, where cold air from the north converges with warm, moist air from the south. This convergence creates a distinct frontal zone and shear line, providing strong dynamic lifting mechanisms. Additionally, a low-pressure center exists to the west of the TELA, further enhancing the upward motion.

Figure 11b shows the conceptual diagram of the synoptic pattern for EW-type TELA, TELA located to south of the STH. At the 500 hPa level, the easterly flow dominates, with wind speeds around 6 m/s. To the east of the trough line, the airflow is southeasterly, while to the west of the trough line, it is northeasterly. The TELA process is situated to the west of the easterly wave trough line, which is conducive to convergence and upward air motion. At the 850 hPa level, a warm center is located to the northwest of the TELA.

The TC-type TELA occurs to the southwest of the TC’s low-pressure center, where the wind speed and direction converge (Figure 11c). At lower levels, humidity is high, representing abundant moisture supply. To the west of the TELA, a warm zone is developing, with raised temperatures that promote the formation of TELAs.

Figure 11d presents the conceptual diagram for STH-edge-type TELA in SC. At the 500 hPa level, the TELA is located to the west of the 5880 gpm geopotential height contour. At the 850 hPa level, a warm temperature zone is present to the west of the TELA. Similar to the STH-edge-type TELA in NC, the warm, moist air transported by the southwesterly flow along the northwestern edge of the STH provides favorable moisture and thermal conditions for establishing thermal instability. However, unlike the NC, where cold air from the north interacts with the warm, moist air from the southwest, there is no significant low-pressure trough to the west of the TELA in SC. Consequently, this type of TELA in SC is primarily a localized thermodynamic thunderstorm triggered by the warm and moist conditions.

3.2.4. Synoptic Patterns for TELAs over the TP

The conceptual diagram of the synoptic patterns for TELAs over the TP is shown in Figure 12. At the 200 hPa level, the TELAs are located to the northeast of the high-pressure region. The eastward extension of the South Asian High is conducive to the weakening of the westerlies over the TP, which favors the development of low vortices over the TP. At the 500 hPa level, TELA is positioned in a region with high temperatures, with warm and moist air flows originating from the south of the Himalayas. The temperature and humidity conditions are favorable for convection. To the north of the TELA location, the confluence of airflow from the north and south enhances the development of low vortices and shear lines over the TP, promoting the occurrence of TELA.

3.3. The Characteristics of Convective Parameters Before TELA

To further investigate the anomalies of meteorological parameters before TELA from the SCN, the study calculated and ranked the standardized anomalies (N) for various synoptic weather types across the four study areas (Figure 13). Overall, the ranking results of the standardized anomalies of meteorological parameters before TELA show significant similarity across the different study areas. Except for the lower anomalies of dynamical parameters before the EW type TELA in SC, the parameters of other types exhibit significant anomalies compared to the SCN, including thermodynamic parameters (e.g., CAPE, T2m), humidity parameters (e.g., VIMD, MF), and dynamical parameters (e.g., 10WS, Shear06). Among these, VIMD and CAPE show the highest anomalies, with standardized anomalies generally exceeding a value of 5. This suggests that the marked anomalies in these parameters could serve as potential indicators for predicting the occurrence of TELA. Additionally, the standardized anomalies of 10 WS and T2m are also relatively high, indicating these parameters increased before the TELA process.

The spatial distribution of meteorological parameter anomalies for different weather types of TELAs varies as well. Based on the previous results, we selected environmental parameters with significant anomalies to analyze the spatial distribution of their anomalies before the TELA process.

Figure 14 presents the distribution of anomalies of CAPE, moisture flux, and VIMD before the TELA process. Overall, TELAs tend to occur near the centers of positive CAPE anomalies and negative VIMD anomalies. UWT-type TELA in NEC (Figure 14a) occurs to the northeast of the CAPE anomaly center, where CAPE values increase by over 800 J/kg compared to SCN. Moisture is supplied from the southwest, and there is a negative anomaly in VIMD where the TELA is located, indicating moisture convergence. In contrast, the TELA in NC (Figure 14b) is located to the northwest of the CAPE anomaly center, with maximum positive CAPE anomalies around 600 J/kg, which is smaller than in NEC. Moisture is transported from the west and south, with negative VIMD anomalies present at the TELA location. However, the moisture flux anomalies are weaker compared to the UWT-type TELAs in NEC. For the STH-edge-type TELA in NC (Figure 14c), the positive anomaly of CAPE exceeds 1000 J/kg, and moisture primarily comes from the warm, moist air transported northward along the edge of the STH, showing significant anomalies in southwesterly moisture flux. Both moisture and thermal conditions are superior to those of the UWT-type TELA in NC.

In SC, the second-highest CAPE anomaly was shown before the Frontal-type TELA process (Figure 14d), with a maximum value exceeding 1000 J/kg. Moisture is mainly supplied from the west, with significant moisture flux and a pronounced region of negative VIMD, indicating strong moisture convergence. The EW-type TELA (Figure 14e) exhibits the highest positive CAPE anomaly (in SC), reaching up to 1200 J/kg, accompanied by large amounts of moisture transported from the ocean by the easterly flow. The TC-type TELA (Figure 14f) is located to the west of the cyclone center, where the tropical cyclone supplies abundant moisture. The STH-edge-type TELA (Figure 14g) is located to the northwest of the CAPE anomaly center, with southwesterly flow transporting moisture northward along the edge of the STH and converging at the TELA location.

For the TP-type TELA (Figure 14h), CAPE increases by over 200 J/kg compared to the SCN. Moisture is transported eastward along the westerlies to the south of the TELA location and then turns southeastward, flowing toward the TP. However, due to the high elevation of the TP acting as a barrier, moisture flux near the TELA is not substantial.

Figure 15 illustrates the distribution of anomalies of surface wind, temperature, and 500 hPa geopotential height before the TELA process. Except for the TP (Figure 15h), TELA primarily occurs between the centers of positive and negative pressure anomalies, where the pressure anomaly isolines are most densely packed, indicating the highest horizontal pressure anomaly gradient. Except for the EW-type (Figure 15e) and TC-type TELA (Figure 15f), the center of negative pressure anomalies is located to the northwest of the TELA, while the center of positive pressure anomalies is to the southeast. For the EW-type TELA (Figure 15e), the center of positive-pressure anomalies is positioned to the northwest of the TELA, and the center of negative-pressure anomalies lies to the southeast. The TC-type TELA (Figure 15f) is located on the western side of a low-pressure anomaly center, while the TP-type TELA (Figure 15h) is located directly at the center of a negative pressure anomaly.

Before the TELA process, surface temperatures significantly increase, and wind speeds at the surface notably strengthen. The EW-type TELA (Figure 15e) and TP-type TELA (Figure 15h) exhibit the most pronounced temperature increases, with values exceeding 5 K. In particular, before the TP-type TELA process, the entire TP area experiences significant warming, with intense surface heating providing favorable thermal conditions for the development of convective activities over the TP [61].

4. Discussion

The findings of this study reveal that the synoptic patterns and meteorological conditions associated with TELAs share certain similarities with those of other intense convection [29,30,51,62,63,64]. For example, in NEC, most TELAs occur to the east of the UWT, consistent with studies that demonstrate the close link between intense convection and the UWT (including the Northeast China Cold Vortex) [64,65]. In NC, TELAs are typically found either east of the UWT or along the western edge of the STH. This result aligns with the findings of Liu et al. [63], who indicated that extremely persistent heavy rainfall in NC is closely related to the STH and local vortices. In SC, TELAs are associated with frontal systems, easterly waves, tropical cyclones, and the STH, which are included in the synoptic types for intense convection in China proposed by Xu et al. [51]. Over the TP, TELAs are generally linked to low-pressure systems, including troughs, shear lines, and low vortices [58,59,60], identified as the primary synoptic patterns driving extreme precipitation events on the TP [66].

From the perspective of meteorological parameters, we found that the conditions required for TELAs are consistent with established knowledge, which identifies conditional instability, moisture, and a source of lift as critical factors for intense convection [67,68]. In this study, the parameters related to thermal, dynamic, and humidity conditions exhibited significant anomalies before the onset of TELAs. Notably, the main parameters with the most prominent standardized anomalies, such as CAPE, VIMD, and 10-m wind speed, were consistent across all four study areas (Figure 13). This consistency indicates their broad applicability in forecasting intense convection. Additionally, standardized anomalies offer an effective method for quantifying and comparing deviations among meteorological variables of differing scales, holding considerable potential for forecasting intense convection. Moreover, in terms of spatial distribution, anomalies of parameters (Figure 14 and Figure 15) reveal more notable characteristics than raw parameter values (Figure A1), suggesting their importance in predicting the occurrence and location of intense convection.

5. Conclusions

This study is based on the TFD from 2010 to 2018 and focuses on four areas in China with relatively frequent thunderstorm activity: NEC, NC, SC, and the TP. Thunderstorms with lightning frequency within the LAA ranking in the top 1% are defined as TELA. The temporal distribution and structural characteristics of TELA were analyzed, and using ERA5 reanalysis data, the synoptic patterns and environmental meteorological conditions favorable to TELA were investigated. The main conclusions are as follows:

1. TELA activity is most active during the warm season (May to September). In NEC, the peak occurs in July, and in NC, it peaks in August, both lagging behind the peak in total thunderstorm activity by one month. TELA activity in SC shows a first peak in May and a secondary peak in August, differing from the single August peak of total thunderstorms. TELA activity is above the TP peaks in September. Except for the TELA being over TP, TELAs are more active from morning to early afternoon (05:00–15:00, LST), in contrast to total thunderstorm activity, which tends to peak in the afternoon. Over the TP, TELAs present a single peak around 18:00–19:00 (LST), which is later than the peak of total thunderstorm activity (15:00–16:00, LST).

2. Compared to total thunderstorms, TELAs exhibit cloud areas 5 to 12 times larger, mean values of maximum TBB that are 12–22 °C colder and 2 to 8 times more SCCs, SCC areas that are 7 to 14 times larger, and proportions of the SCC area that are approximately 5%–13% higher. The maximum average precipitation within 3 h of a TELA is 40 mm/h in NEC, 57 mm/h in NC, 71 mm/h in SC, and 7 mm/h on the TP, indicating intense rainfall in each study area.

3. TELAs are often associated with significant weather systems. In NEC, most TELAs (94%) typically occur to the east of the UWT (including cold vortex systems) where strong low-level warm and moist air lead to thermal instability, combined with dynamic disturbances from UWT that trigger intense vertical motion. In NC, in addition to TELAs associated with the UWT (39% of all TELAs in NC), warm moist air from the southwest interacts with cold, dry air from the north, leading to TELA at the western edge of the STH (STH-edge-type TELA, 51% of all TELAs in NC). Most TELAs (58%) in SC are Frontal-type TELAs, caused by the convergence and lifting from strong interactions between northern cold air and southern warm air. Additionally, EW-type, TC-type, and STH-edge-type TELAs account for 4%, 27%, and 11% of all TELAs in SC, respectively (96, 58%). Over the TP, TELA occurs to the north of the 200 hPa high-pressure center, where the eastward extension of the South Asian High weakens the westerlies and fosters the development of TP low vortices. Meanwhile, at the 500 hPa level, the convergence of northern and southern air flows further supports the occurrence of TELA. Conceptual models of these synoptic patterns are provided in the study, offering insights for early warnings of extreme weather events.

4. Before the TELA process, significant anomalies in thermal, moisture, and dynamic conditions are observed across all study areas. VIMD and CAPE show the greatest anomalies, with standardized anomalies exceeding 5. TELAs are typically located near the centers of positive CAPE anomalies and negative VIMD anomalies. The sources of water vapor for TELA vary by region and synoptic pattern, including westerly moisture transport (for UWT-type, Frontal-type in SC, and TP-type TELA), southwestern moisture transport along the western edge of the STH (for STH-edge-type TELA in NC and SC), and easterly moisture transport (for EW-type and TC-type TELA in SC). At the 500 hPa level, except over the TP, TELA occurs between the centers of positive and negative pressure anomalies, where the pressure anomaly isolines are most densely packed, indicating the highest horizontal pressure anomaly gradient. Before the TELA process, surface temperature and wind speed increased significantly.

This study suggests that lightning data can serve as an effective indicator to identify intense convection, which may be accompanied by other extreme weather events such as heavy precipitation, hail, and tornadoes. Furthermore, insights into the anomalous characteristics of atmospheric circulation and meteorological parameters preceding TELAs provide valuable references for the early warning and forecasting of intense convection.

Author Contributions

Conceptualization, R.M. and D.Z.; methodology, R.M. and D.Z.; software, R.M. and D.Z.; validation, D.Z., Y.Z., W.Y., W.Z. and B.Z.; formal analysis, R.M. and D.Z.; investigation, R.M., D.Z., W.Y. and W.Z; resources, R.M., D.Z., W.Y., W.Z. and B.Z.; data curation, R.M.; writing—original draft preparation, R.M. and D.Z.; writing—review and editing, R.M. and D.Z.; visualization, R.M. and D.Z.; supervision, D.Z., Y.Z., W.Y., W.Z. and B.Z.; project administration, D.Z. and B.Z.; funding acquisition, D.Z. and B.Z. All authors have read and agreed to the published version of the manuscript.

Data Availability Statement

The thunderstorm feature dataset (TFD) associated with this paper can be accessed from http://doi.org/10.57760/sciencedb.02003 or from the corresponding author.

Conflicts of Interest

The authors declare no conflicts of interest.

Abbreviations

The following abbreviations are used in this manuscript.

TELA	Thunderstorms with Extreme Lightning Activity
WWLLN	World Wide Lightning Location Network
CAPE	Convective Available Potential Energy
TFD	Thunderstorm Feature Dataset
TBB	The Black Body Temperature
SCC	Strong Convective Core
LAA	Lightning Active Area
LAC	Lightning Active Center
THS	Thunderstorm Hour Sample
NEC	Northeast China
NC	North China
SC	South China
TP	Tibetan Plateau
SCN	Standard Climatological Normal
LST	Local Solar Time
UWT	Upper-level Westerly Troughs
STH	Subtropical High
EW	Easterly Wave
TC	Tropical Cyclone
gpm	geopotential metre
RH	Relative Humidity
T2m	2 m Temperature
VT	Vertical Totals
VIMD	Vertically Integrated Moisture Divergence
viwve/viwvn	Vertical integral of eastward/northward water vapour flux
MF	Moisture Flux
D2m	2 m Dewpoint temperature
PW	Precipitable Water
MRH	Mid-level Relative Humidity
u10/v10	10 m u/v Component of wind
10 WS	10 m Wind Speed
Shear06	0–6 km Wind Shear
SP	Surface Pressure
Kindex	K-index
CT	Cross Totals
TT	Total Totals

Footnotes

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Figures and Table

Figure 1. Schematic diagram of thunderstorm cloud area identification. The black solid line encloses the cloud area with TBB [Forumla omitted. See PDF.]32 °C. The red solid line outlines areas with TBB [Forumla omitted. See PDF.]52 °C. The blue solid line represents an ellipse fitting the cloud area with TBB [Forumla omitted. See PDF.]32 °C. The yellow “+” indicates WWLLN lightning strokes. The green solid line outlines clusters of WWLLN lightning strokes.

Figure 2. Spatial distribution of the density of Thunderstorm Hour Samples (THSs) in China during 2010–2018, 1° × 1° grid. The black boxes show the four areas of study (NEC, NC, SC and TP).

Figure 3. Probability distributions of WWLLN flash counts within the LAA of THSs in NEC (a), NC (b), SC (c), and the TP (d).

Figure 4. Distributions of maximum hourly precipitation corresponding to TELA in each study area (NEC, NC, SC and TP). The central solid black line on each box indicates the median, and the bottom and top edges of the box indicate the 25th and 75th percentiles, respectively. The whiskers extend to the most extreme data points not considered outliers. The mean and median values are indicated in blue and black text, respectively.

Figure 5. Percentage of monthly counts of THSs (a) and TELAs (b) in each study area.

Figure 6. Percentage of hourly counts of THSs (a) and TELAs (b) in each study area.

Figure 7. Distributions of (a) cloud area, (b) minimum TBB, (c) number, (d) area, and (e) area proportion of SCCs for THSs (blue columns) and TELA (orange columns) in four study areas. The components on the boxplot are consistent with those in Figure 4.

Figure 8. Monthly variation in the counts of TELA processes corresponding to different synoptic weather types. (a) UWT type in NEC, (b) UWT type and (c) STH-edge type in NC, (d) Frontal type, (e) EW type, (f) TC type (g) STH-edge type in SC, (h) TP type TELAs. UWT: Upper-level Westerly Through, STH: Subtropical High, EW: Easterly Wave, TC: Tropical Cyclone.

Figure 9. Conceptual diagram of the synoptic pattern of UWT-type TELA in NEC (light blue area: 850 hPa RH ≥ 70%; dark blue solid line: 850 hPa geopotential height contours; light blue solid line: 500 hPa geopotential height contours; orange dashed line: 850 hPa temperature contours; gray arrow: 850 hPa wind; black arrow: 500 hPa wind).

Figure 10. Conceptual diagram of the synoptic pattern of (a) UWT-type and (b) STH-edge-type TELA in NC (light blue area: 850 hPa RH ≥ 70%; dark blue solid line: 850 hPa geopotential height contours; light blue solid line: 500 hPa geopotential height contours; orange dashed line: 850 hPa temperature contours; gray arrow: 850 hPa wind; black arrow: 500 hPa wind).

Figure 11. Conceptual diagram of the synoptic pattern of (a) Frontal-type, (b) EW-type, (c) TC-type, (d) STH-edge-type TELA in SC (light blue area: 850 hPa RH ≥ 80%; dark blue solid line: 850 hPa geopotential height contours; light blue solid line: 500 hPa (a: 700 hPa) geopotential height contours; orange dashed line: 850 hPa temperature contours; gray arrow: 850 hPa wind; black arrow: 500 hPa (a: 700 hPa) wind).

Figure 12. Conceptual diagram of the synoptic pattern of TP-type TELA over the TP (light blue area: 850 hPa RH ≥ 80%; dark blue solid line: 500 hPa geopotential height contours; light blue solid line: 200 hPa geopotential height contours; orange dashed line: 500 hPa temperature contours; gray arrow: 500 hPa wind; pink arrow: 10m wind).

Figure 13. Distributions and ranking results of the standardized anomalies (N) of environmental meteorological parameters before TELA processes for different synoptic weather types. (a) UWT-type in NEC, (b) UWT-type and (c) STH-edge-type in NC, (d) Frontal-type, (e) EW-type, (f) TC-type (g) STH-edge-type in SC, (h) TP-type TELAs. The red “o” symbol indicates the mean value and the outliers on the boxplot are plotted individually using the “+” symbol. The other components on the boxplot are consistent with those in Figure 4.

Figure 14. Spatial distribution of the anomalies of VIMD (fill color), moisture flux (vector arrows), and CAPE (unit: J/kg, contours) before TELA processes for different synoptic weather types. (a) UWT-type in NEC, (b) UWT-type and (c) STH-edge-type in NC, (d) Frontal-type, (e) EW-type, (f) TC-type (g) STH-edge-type in SC, (h) TP-type TELAs.

Figure 15. Same as Figure 14, but the fill color, vector arrows, and the contours indicate the anomalies of 2m temperature (K), 10m wind, and 500 hPa geopotential height (unit: 10 gpm), respectively.

Table 1

Environmental meteorological parameters used in this study.

Name	Abbreviation	Unit	Resource	Notes
Convective available potential energy	CAPE	J·kg−1	ERA5 on single levels	Thermodynamic parameter
2 m Temperature	T2m	K	ERA5 on single levels
Vertical Totals	VT	K	Calculate based on ERA5; $VT=T_{850}-T_{500};V{T}_{TP}=T_{500}-T_{300}$
Vertically integrated moisture divergence	VIMD	kg·m−2	ERA5 on single levels	Moisture parameter
Vertical integral of eastward/northward water vapour flux	viwve/viwvn	kg·m−1·s−1	ERA5 on single levels
Moisture flux	MF	kg·m−1·s−1	Calculate based on viwve and viwvn
2 m Dewpoint temperature	D2m	K	ERA5 on single levels
Precipitable water	PW	mm	Calculate based on ERA5
400–700 hPa Mid-level relative humidity	MRH	%	Calculate based on ERA5; 200–400 hPa in TP
10 m u/v Component of wind	u10/v10	m·m−1	ERA5 on single levels	Dynamical parameter
10 m wind speed	10 WS	m·s−1	Calculate based on u10 and v10
0–6 km Wind shear	Shear06	m·s−1	Calculate based on ERA5
Surface pressure	SP	hPa	ERA5 on single levels
K-Index	Kindex	K	Calculate based on ERA5; $Kindex=T_{850}-T_{500}+T{d}_{850}-(T_{700}-T{d}_{700});Kinde{x}_{TP}=T_{500}-T_{300}+T{d}_{500}-(T_{400}-T{d}_{400})$	Mixed parameter
Cross totals	CT	K	Calculate based on ERA5; $CT=T{d}_{850}-T_{500};C{T}_{TP}=T{d}_{500}-T_{300}$
Total totals	TT	K	$TT=VT+CT$

Appendix A

Figure A1. Spatial distribution of the relative humidity at 850 hPa level (fill color), 850 hPa (h: 500 hPa) geopotential height contours (unit: 10 gpm, dark blue contours), 500 hPa (d: 700 hPa, h: 200 hPa) geopotential height contours (unit: 10 gpm, light blue contours), 850 hPa (h: 500 hPa) temperature contours (orange), 850 hPa wind (black arrow), 500 hPa wind (gray arrow), and 10m wind (pink arrow) before TELA processes for different synoptic weather types. (a) UWT-type in NEC, (b) UWT-type and (c) STH-edge-type in NC, (d) Frontal-type, (e) EW-type, (f) TC-type (g) STH-edge-type in SC, (h) TP-type TELAs.

References

1. IPCC2021. Climate Change 2021: The Physical Science Basis; Technical Report, Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change Cambridge University Press: Cambridge, UK, 2021.

2. Johns, R.H.; Doswell III, C.A. Severe local storms forecasting. Weather. Forecast.; 1992; 7, pp. 588-612. [DOI: https://dx.doi.org/10.1175/1520-0434(1992)007<0588:SLSF>2.0.CO;2]

3. Brooks, H.E. Severe thunderstorms and climate change. China Sci. Data; 2013; 123, pp. 129-138. [DOI: https://dx.doi.org/10.1016/j.atmosres.2012.04.002]

4. Allen, J.T.; Karoly, D.J. A climatology of Australian severe thunderstorm environments 1979–2011: Inter-annual variability and ENSO influence. Int. J. Climatol.; 2014; 34, pp. 81-97. [DOI: https://dx.doi.org/10.1002/joc.3667]

5. Taszarek, M.; Allen, J.T.; Groenemeijer, P.; Edwards, R.; Brooks, H.E.; Chmielewski, V.; Enno, S.E. Severe convective storms across Europe and the United States. Part I: Climatology of lightning, large hail, severe wind, and tornadoes. J. Clim.; 2020; 33, pp. 10239-10261. [DOI: https://dx.doi.org/10.1175/JCLI-D-20-0345.1]

6. Taszarek, M.; Allen, J.T.; Púčik, T.; Hoogewind, K.A.; Brooks, H.E. Severe convective storms across Europe and the United States. Part II: ERA5 environments associated with lightning, large hail, severe wind, and tornadoes. J. Clim.; 2020; 33, pp. 10263-10286. [DOI: https://dx.doi.org/10.1175/JCLI-D-20-0346.1]

7. Liu, N.; Liu, C.; Tissot, P.E. A bayesian-like approach to describe the regional variation of high-flash rate thunderstorms from thermodynamic and kinematic environment variables. J. Geophys. Res. Atmos.; 2019; 124, pp. 12507-12522. [DOI: https://dx.doi.org/10.1029/2019JD031254]

8. Liu, N.; Liu, C.; Chen, B.; Zipser, E.J. What are the favorable large-scale environments for the highest-flash-rate thunderstorms on earth?. J. Atmos. Sci.; 2020; 77, pp. 1583-1612. [DOI: https://dx.doi.org/10.1175/JAS-D-19-0235.1]

9. Raupach, T.H.; Martius, O.; Allen, J.T.; Kunz, M.; Lasher-Trapp, S.; Mohr, S.; Rasmussen, K.L.; Trapp, R.J.; Zhang, Q. The effects of climate change on hailstorms. Nat. Rev. Earth Environ.; 2021; 2, pp. 213-226. [DOI: https://dx.doi.org/10.1038/s43017-020-00133-9]

10. Koch, E.; Koh, J.; Davison, A.C.; Lepore, C.; Tippett, M.K. Trends in the Extremes of Environments Associated with Severe U.S. Thunderstorms. J. Clim.; 2021; 34, pp. 1259-1272. [DOI: https://dx.doi.org/10.1175/JCLI-D-19-0826.1]

11. Liu, N.; Liu, C.; Tissot, P.E. Relative Importance of Large-Scale Environmental Variables to the World-Wide Variability of Thunderstorms. J. Geophys. Res. Atmos.; 2022; 127, e2021JD036065. [DOI: https://dx.doi.org/10.1029/2021JD036065]

12. Takahashi, T. Riming Electrification as a Charge Generation Mechanism in Thunderstorms. J. Atmos. Sci.; 1978; 35, pp. 1536-1548. [DOI: https://dx.doi.org/10.1175/1520-0469(1978)035<1536:REAACG>2.0.CO;2]

13. Zipser, E.J.; Lutz, K.R. The Vertical Profile of Radar Reflectivity of Convective Cells: A Strong Indicator of Storm Intensity and Lightning Probability?. Mon. Weather. Rev.; 1994; 122, pp. 1751-1759. [DOI: https://dx.doi.org/10.1175/1520-0493(1994)122<1751:TVPORR>2.0.CO;2]

14. Barthe, C.; Deierling, W.; Barth, M.C. Estimation of total lightning from various storm parameters: A cloud-resolving model study. J. Geophys. Res. Atmos.; 2010; 115, D24202. [DOI: https://dx.doi.org/10.1029/2010JD014405]

15. Dahl, J.M.L.; Höller, H.; Schumann, U. Modeling the Flash Rate of Thunderstorms. Part I: Framework. Mon. Weather. Rev.; 2011; 139, pp. 3093-3111. [DOI: https://dx.doi.org/10.1175/MWR-D-10-05031.1]

16. Liu, C.; Cecil, D.J.; Zipser, E.J.; Kronfeld, K.; Robertson, R. Relationships between lightning flash rates and radar reflectivity vertical structures in thunderstorms over the tropics and subtropics. J. Geophys. Res. Atmos.; 2012; 117, D06212. [DOI: https://dx.doi.org/10.1029/2011JD017123]

17. Basarab, B.M.; Rutledge, S.A.; Fuchs, B.R. An improved lightning flash rate parameterization developed from Colorado DC3 thunderstorm data for use in cloud-resolving chemical transport models. J. Geophys. Res. Atmos.; 2015; 120, pp. 9481-9499. [DOI: https://dx.doi.org/10.1002/2015JD023470]

18. Xu, W. Thunderstorm Climatologies and Their Relationships to Total and Extreme Precipitation in China. J. Geophys. Res. Atmos.; 2020; 125, e2020JD033152. [DOI: https://dx.doi.org/10.1029/2020JD033152]

19. Poręba, S.; Taszarek, M.; Ustrnul, Z. Diurnal and Seasonal Variability of ERA5 Convective Parameters in Relation to Lightning Flash Rates in Poland. Weather. Forecast.; 2022; 37, pp. 1447-1470. [DOI: https://dx.doi.org/10.1175/WAF-D-21-0099.1]

20. Pineda, N.; Bech, J.; Rigo, T.; Montanyà, J. A Mediterranean nocturnal heavy rainfall and tornadic event. Part II: Total lightning analysis. Atmos. Res.; 2011; 100, pp. 638-648. [DOI: https://dx.doi.org/10.1016/j.atmosres.2010.10.027]

21. Schultz, C.J.; Petersen, W.A.; Carey, L.D. Lightning and severe weather: A comparison between total and cloud-to-ground lightning trends. Weather. Forecast.; 2011; 26, pp. 744-755. [DOI: https://dx.doi.org/10.1175/WAF-D-10-05026.1]

22. Xu, X. Lightning Induced Disasters and Its Detecting and Forecasting. Meteorol. Mon.; 2004; 30, pp. 17-21. (In Chinese) [DOI: https://dx.doi.org/10.3969/j.issn.1000-0526.2004.12.004]

23. Ma, M.; Lu, W.; Zhang, Y.; Meng, Q. Analysis of Lightning Disasters in China and Their Correlative Factors. Adv. Earth Sci.; 2008; 23, pp. 856-865. (In Chinese) [DOI: https://dx.doi.org/10.3321/j.issn:1001-8166.2008.08.009]

24. Zheng, L.; Sun, J.; Wei, J. Thunder events in China: 1980–2008. Atmos. Ocean. Sci. Lett.; 2010; 3, pp. 181-188. [DOI: https://dx.doi.org/10.1080/16742834.2010.11446866]

25. Zhang, W.; Meng, Q.; Ma, M.; Zhang, Y. Lightning casualties and damages in China from 1997 to 2009. Nat. Hazards; 2011; 57, pp. 465-476. [DOI: https://dx.doi.org/10.1007/s11069-010-9628-0]

26. Qie, X.; Zhang, Y.; Yuan, T.; Zhang, Q.; Zhang, T.; Zhu, B.; Lu, W.; Ma, M.; Yang, J.; Zhou, Y. et al. A review of atmospheric electricity research in China. Adv. Atmos. Sci.; 2015; 32, pp. 169-191. [DOI: https://dx.doi.org/10.1007/s00376-014-0003-z]

27. Kolendowicz, L. Synoptic patterns associated with thunderstorms in Poland. Meteorol. Z.; 2012; 21, 145. [DOI: https://dx.doi.org/10.1127/0941-2948/2012/0272]

28. Wapler, K.; James, P. Thunderstorm occurrence and characteristics in Central Europe under different synoptic conditions. Atmos. Res.; 2015; 158, pp. 231-244. [DOI: https://dx.doi.org/10.1016/j.atmosres.2014.07.011]

29. Liu, N.; Liu, C. Synoptic environments and characteristics of convection reaching the tropopause over northeast China. Mon. Weather. Rev.; 2018; 146, pp. 745-759. [DOI: https://dx.doi.org/10.1175/MWR-D-17-0245.1]

30. Ma, R.; Sun, J.; Yang, X. An eight-year climatology of the warm-season severe thunderstorm environments over North China. Atmos. Res.; 2021; 254, 105519. [DOI: https://dx.doi.org/10.1016/j.atmosres.2021.105519]

31. Zhou, Z.; Zhang, Q.; Allen, J.T.; Ni, X.; Ng, C.P. How Many Types of Severe Hailstorm Environments Are There Globally?. Geophys. Res. Lett.; 2021; 48, e2021GL095485. [DOI: https://dx.doi.org/10.1029/2021GL095485]

32. Mora, M.; Riesco, J.; Dávila, F.d.P.; Sorianoc, L.R. Atmospheric background associated with severe lightning thunderstorms in Central Spain. Int. J. Climatol.; 2015; 35, pp. 558-569. [DOI: https://dx.doi.org/10.1002/joc.4002]

33. Aran, M.; Amaro, J.; Arús, J.; Bech, J.; Figuerola, F.; Gayà, M.; Vilaclara, E. Synoptic and mesoscale diagnosis of a tornado event in Castellcir, Catalonia, on 18th October 2006. Atmos. Res.; 2009; 93, pp. 147-160. [DOI: https://dx.doi.org/10.1016/j.atmosres.2008.09.031]

34. Zheng, D.; MacGorman, D.R. Characteristics of flash initiations in a supercell cluster with tornadoes. Atmos. Res.; 2016; 167, pp. 249-264. [DOI: https://dx.doi.org/10.1016/j.atmosres.2015.08.015]

35. Xu, S.; Zheng, D.; Wang, Y.; Hu, P. Characteristics of the two active stages of lightning activity in two hailstorms. J. Meteorol. Res.; 2016; 30, pp. 265-281. [DOI: https://dx.doi.org/10.1007/s13351-016-5074-6]

36. Zhang, Z.; Zheng, D.; Zhang, Y.; Lu, G. Spatial–temporal characteristics of lightning flash size in a supercell storm. Atmos. Res.; 2017; 197, pp. 201-210. [DOI: https://dx.doi.org/10.1016/j.atmosres.2017.06.029]

37. Qie, X.; Yuan, S.; Chen, Z.; Wang, D.; Liu, D.; Sun, M.; Sun, Z.; Srivastava, A.; Zhang, H.; Lu, J. et al. Understanding the dynamical-microphysical-electrical processes associated with severe thunderstorms over the Beijing metropolitan region. Sci. China Earth Sci.; 2021; 64, pp. 10-26. [DOI: https://dx.doi.org/10.1007/s11430-020-9656-8]

38. Rigo, T.; Rodríguez, O.; Bech, J.; Farnell, C. An observational analysis of two companion supercell storms over complex terrain. Atmos. Res.; 2022; 272, 106149. [DOI: https://dx.doi.org/10.1016/j.atmosres.2022.106149]

39. Madhulatha, A.; Mohapatra, M.; Devi, K.S.; Pattanaik, D.R.; Rajeevan, M.; Das, A.K.; Jenamani, R.K.; Kumar, N. Climatology of thermodynamic indices and background synoptic conditions responsible for severe convection during pre- to post-monsoon seasons over Indian region. Int. J. Climatol.; 2024; 44, pp. 2607-2635. [DOI: https://dx.doi.org/10.1002/joc.8471]

40. Ma, R.; Zheng, D. A dataset of thunderstorm features of the FY-2E full-disc observation area from 2010 to 2018. China Sci. Data; 2023; 8, pp. 1-14. (In Chinese) [DOI: https://dx.doi.org/10.11922/11-6035.csd.2022.0068.zh]

41. Hutchins, M.L.; Holzworth, R.H.; Brundell, J.B. Diurnal variation of the global electric circuit from clustered thunderstorms. J. Geophys. Res. Space Phys.; 2014; 119, pp. 620-629. [DOI: https://dx.doi.org/10.1002/2013JA019593]

42. Ma, R.; Zheng, D.; Zhang, Y.; Yao, W.; Zhang, W.; Cuomu, D. Spatiotemporal lightning activity detected by wwlln over the tibetan plateau and its comparison with lis lightning. J. Atmos. Ocean. Technol.; 2021; 38, pp. 511-523. [DOI: https://dx.doi.org/10.1175/JTECH-D-20-0080.1]

43. Du, Y.; Zheng, D.; Ma, R.; Zhang, Y.; Lyu, W.; Yao, W. Thunderstorm activity over the Qinghai–Tibet Plateau indicated by the combined data of the FY-2E geostationary satellite and WWLLN. Remote. Sens.; 2022; 14, 2855. [DOI: https://dx.doi.org/10.3390/rs14122855]

44. Zhang, H.; Zheng, Y. Thunderstorm climatology over mainland China based on hourly observations during 1971—2010. Acta Meteorol. Sin.; 2022; 80, pp. 54-66. (In Chinese) [DOI: https://dx.doi.org/10.11676/qxxb2022.004]

45. Rodger, C.J.; Werner, S.; Brundell, J.B.; Lay, E.H.; Thomson, N.R.; Holzworth, R.H.; Dowden, R.L. Detection efficiency of the VLF World-Wide Lightning Location Network (WWLLN): Initial case study. Ann. Geophys.; 2006; 24, pp. 3197-3214. [DOI: https://dx.doi.org/10.5194/angeo-24-3197-2006]

46. Hutchins, M.L.; Holzworth, R.H.; Brundell, J.B.; Rodger, C.J. Relative detection efficiency of the World Wide Lightning Location Network. Radio Sci.; 2012; 47, pp. 1-9. [DOI: https://dx.doi.org/10.1029/2012RS005049]

47. Xiaodin, Y. Nowcasting thinking and method of flash heavy rain. Torrential Rain Disasters; 2013; 32, pp. 202-209. (In Chinese) [DOI: https://dx.doi.org/10.3969/j.issn.1004-9045.2013.03.002]

48. Lu, Y.; Jiang, S.; Ren, L.; Zhang, L.; Wang, M.; Liu, R.; Wei, L. Spatial and temporal variability in precipitation concentration over mainland China, 1961–2017. Water; 2019; 11, 881. [DOI: https://dx.doi.org/10.3390/w11050881]

49. Qie, X.; Wu, X.; Yuan, T.; Bian, J.; Lu, D. Comprehensive Pattern of Deep Convective Systems over the Tibetan Plateau–South Asian Monsoon Region Based on TRMM Data. J. Clim.; 2014; 27, pp. 6612-6626. [DOI: https://dx.doi.org/10.1175/JCLI-D-14-00076.1]

50. Li, J. Hourly station-based precipitation characteristics over the Tibetan Plateau. Int. J. Climatol.; 2018; 38, pp. 1560-1570. [DOI: https://dx.doi.org/10.1002/joc.5281]

51. Xu, A.; Sun, J.; Dongbei, X.; Wan, X.; Guo, Y. Basic Synoptic Situation Classification and Element Character of Severe Convection in China. Meteorol. Mon.; 2014; 40, pp. 400-411. (In Chinese) [DOI: https://dx.doi.org/10.7519/j.issn.1000-0526.2014.04.002]

52. Hart, R.E.; Grumm, R.H. Using Normalized Climatological Anomalies to Rank Synoptic-Scale Events Objectively. Mon. Weather. Rev.; 2001; 129, pp. 2426-2442. [DOI: https://dx.doi.org/10.1175/1520-0493(2001)129<2426:UNCATR>2.0.CO;2]

53. Grumm, R.H.; Hart, R. Standardized Anomalies Applied to Significant Cold Season Weather Events: Preliminary Findings. Weather. Forecast.; 2001; 16, pp. 736-754. [DOI: https://dx.doi.org/10.1175/1520-0434(2001)016<0736:SAATSC>2.0.CO;2]

54. Wang, H.; Luo, Y.; Zhang, R. Synoptic pattern and severe weather associated with the wide convection over Southeast China during the summer monsoon period. J. Meteorol. Res.; 2015; 29, pp. 41-58. [DOI: https://dx.doi.org/10.1007/s13351-014-4069-4]

55. Chen, Y.; Luo, Y.; Liu, B. General features and synoptic-scale environments of mesoscale convective systems over South China during the 2013–2017 pre-summer rainy seasons. Atmos. Res.; 2022; 266, 105954. [DOI: https://dx.doi.org/10.1016/j.atmosres.2021.105954]

56. Yang, B.; Wang, L.; Guan, Y. Characteristics and Development Mechanisms of Northeast Cold Vortices. Adv. Meteorol.; 2021; 2021, 6636192. [DOI: https://dx.doi.org/10.1155/2021/6636192]

57. Hu, K.; Lu, R.; Wang, D. Seasonal climatology of cut-off lows and associated precipitation patterns over Northeast China. Meteorol. Atmos. Phys.; 2010; 106, pp. 37-48. [DOI: https://dx.doi.org/10.1007/s00703-009-0049-0]

58. Lin, Z. Analysis of Tibetan Plateau vortex activities using ERA-Interim data for the period 1979–2013. J. Meteorol. Res.; 2015; 29, pp. 720-734. [DOI: https://dx.doi.org/10.1007/s13351-015-4273-x]

59. Lin, Z.; Yao, X.; Guo, W.; Du, J. Vertical structure of Tibetan Plateau Vortex in boreal summer. Theor. Appl. Climatol.; 2021; 145, pp. 427-440. [DOI: https://dx.doi.org/10.1007/s00704-021-03640-x]

60. Liu, Z.; Li, G.; Zhang, B. Statistical Characteristics of the Concurrence between Qinghai-Tibetan Plateau Vortex and Qinghai-Tibetan Plateau Shear Line. Plateau Meteorol.; 2018; 37, pp. 1233-1240. (In Chinese) [DOI: https://dx.doi.org/10.7522/j.issn.1000-0534.2018.00028]

61. Xu, X.; Chen, L. Advances of the study on Tibetan Plateau experiment of atmospheric sciences. J. Appl. Meteorol. Sci.; 2006; 17, pp. 756-772. (In Chinese) [DOI: https://dx.doi.org/10.3969/j.issn.1001-7313.2006.06.013]

62. Zhao, Y.; Li, J. Dominant Synoptic Systems for Summer Precipitation over the Complex Terrain of Southwestern China. Mon. Weather. Rev.; 2024; 152, pp. 2443-2460. [DOI: https://dx.doi.org/10.1175/MWR-D-24-0009.1]

63. Liu, H.; Huang, X.; Fei, J.; Zhang, C.; Cheng, X. Spatiotemporal features and associated synoptic patterns of extremely persistent heavy rainfall over China. J. Geophys. Res. Atmos.; 2022; 127, e2022JD036604. [DOI: https://dx.doi.org/10.1029/2022JD036604]

64. Gao, J.; Gao, H. Influence of the northeast cold vortex on flooding in Northeast China in summer 2013. J. Meteorol. Res.; 2018; 32, pp. 172-180. [DOI: https://dx.doi.org/10.1007/s13351-018-7056-3]

65. Lian, Y.; Shen, B.; Li, S.; Liu, G.; Yang, X. Mechanisms for the formation of Northeast China cold vortex and its activities and impacts: An overview. J. Meteorol. Res.; 2016; 30, pp. 881-896. [DOI: https://dx.doi.org/10.1007/s13351-016-6003-4]

66. Sun, J.; Yao, X.; Deng, G.; Liu, Y. Characteristics and synoptic patterns of regional extreme rainfall over the Central and Eastern Tibetan Plateau in boreal summer. Atmosphere; 2021; 12, 379. [DOI: https://dx.doi.org/10.3390/atmos12030379]

67. Doswell, C.A., III. The distinction between large-scale and mesoscale contribution to severe convection: A case study example. Weather. Forecast.; 1987; 2, pp. 3-16. [DOI: https://dx.doi.org/10.1175/1520-0434(1987)002<0003:TDBLSA>2.0.CO;2]

68. Doswell, C.A., III; Schultz, D.M. On the use of indices and parameters in forecasting severe storms. E-J. Sev. Storms Meteorol.; 2006; 1, pp. 1-22. [DOI: https://dx.doi.org/10.55599/ejssm.v1i3.4]