1. Introduction
In China, mesoscale vortices (with a horizontal scale ranging from 2 to 2000 km; Orlanski 1975; Fu et al. 2020) frequently cause huge economic losses due to their associated heavy rainfall and strong winds (Tao 1980; Zhao et al. 2004). According to previous studies, there are at least three mesoscale-vortex sources in China (Fig. 1): the Tibetan Plateau (Feng et al. 2014; Curio et al. 2019), the Sichuan Basin and its surrounding regions (Fu et al. 2015; Ni et al. 2017; Feng et al. 2020), and the middle reaches of the Yangtze River basin (i.e., the Dabie Mountain and its surrounding regions; Zhang et al. 2015; Fu et al. 2017). Mesoscale vortices that form within these sources are named as the Tibetan Plateau vortex (TPV), the southwest vortex (SWV), and the Dabie vortex (DBV), respectively. Previous studies (Feng et al. 2014; Fu et al. 2015; Curio et al. 2019) find that most TPVs cannot vacate the Tibetan Plateau, and most SWVs cannot move out from the Sichuan Basin. In contrast, the DBVs can often move out from their source region (SR) (Fu et al. 2016b) and affect the downstream areas. In history, North, East, and central China (i.e., downstream areas of the DBVs) frequently experience floods due to the torrential rainfall induced by the DBVs (Zhao et al. 2004). For instance, the catastrophic floods over the middle and lower reaches of the Yangtze River in 1998 and 2020 (Luo et al. 2020; Ding et al. 2021) are mainly due to the DBVs (Fu et al. 2022a).
The topography features of the DBV source (the white dashed box; 26°–34°N, 111°–118°E) and surrounding regions (shading; m). The small figure at the lower-left corner represents the classification of DBVs’ moving directions (EE = eastward moving; NE = northeastward moving; NN = northward moving; NW = northwestward moving; WW = westward moving; SW = southwestward moving; SS = southward moving; SE = southeastward moving).
Citation: Journal of Climate 38, 7; 10.1175/JCLI-D-24-0304.1
Due to its strong catastrophability, DBVs have been a research focus since they are discovered. A series of studies are conducted on the DBVs, to clarify their synoptic environments (Gao and Xu 2001; Yang et al. 2010), three-dimensional structures (Zhang et al. 2015; Chen and Dell’osso 1984; Sun et al. 2010), formation/evolution mechanisms (Hu and Pan 1996; Dong and Zhao 2004; Fu et al. 2013), and precipitation features (Ninomiya 2000; Zhou and Bai 2010; Shen et al. 2013; Fu et al. 2016a). However, there is a striking flaw for all existing research studies on the DBVs: thus far, no studies investigate the long-term climatological features of this type of vortices, and all the basic features about the DBVs are obtained from short-term (≤14-yr) statistic and/or case studies, which have notable limitations to represent the DBVs’ general features. To this end, the primary purpose of this study is to investigate the long-term (42-yr) climatological features of the DBVs, to update the existing statistical features of the DBVs based on a long-term statistic, and to explore key mechanisms that govern DBVs’ multiscale temporal variations. This is not only useful to fill the knowledge gap about the DBVs’ long-term climatological features but also helpful to reach a more comprehensive understanding of this type of mesoscale vortices. The remaining part of this article is structured as follows: the data and methods used are presented in section 2, DBVs’ multiscale temporal variations are investigated in section 3, DBVs’ horizontal distribution and tracks are discussed in section 4, other key features are shown in section 5, and finally, a conclusion and discussion is reached in section 6.
2. Data and methods
The fifth major global reanalysis produced by the European Centre for Medium-Range Weather Forecasts (ERA5) data (Hersbach et al. 2020) during 42 (1979–2020) warm seasons (May–September), which has a temporal resolution of hourly and a spatial resolution of 0.25° × 0.25°, are used in this study. The ERA5 data are used for detecting and tracking the DBVs and also for the composite analyses. The final run data of the integrated multisatellite retrievals for Global Precipitation Measurement (GPM; Hou et al. 2014), which has a temporal resolution of 30 min and a horizontal resolution of 0.1° × 0.1°, are utilized to investigate the DBV-associated precipitation (accumulated rainfall is calculated by summing the 30-min precipitation). Evaluations on the GPM precipitation data by previous studies (Du et al. 2022; Wei et al. 2023) show that, although this dataset may underestimate the extreme precipitation intensity, the GPM can capture the key characteristics of the precipitation in China credibly.
a. Detecting and tracking method for the Dabie vortices
In this study, we use the method developed by Zhang et al. (2023) to detect and track the DBVs. The method contains three main steps: i) determining the vortex structure (i.e., a mesoscale closed cyclonic circulation) at the DBVs’ central level (850 hPa; Fu et al. 2016b); ii) tracking the vortex in the horizontal direction (if the distance of two vortex structures at two adjacent hours is ≤300 km, and the correlation coefficient between the wind fields of the two vortex structures is ≥0.9, these two vortex structures are regarded as belonged to a same DBV); and iii) identifying the vortex in the vertical direction [if the distance of two vortex structures at two adjacent vertical levels (we interpolate the ERA5 data into the equidistant vertical levels with an interval of 25 hPa) is ≤100 km, and the correlation coefficient between their wind fields is ≥0.9, these two vortex structures are regarded as belonged to a same DBV]. As the evaluation from Zhang et al. (2023) shows, this method shows a hit rate of ∼95% in detecting mesoscale vortices, which guarantees its usefulness for identifying the DBVs. Moreover, all detected DBVs have undergone strictly manual verifications, which further remove the errors during the vortex identification.
b. Vortex parameters
In this study, the location of a DBV is represented by its center, which is the centroid of the vortex’s closed cyclonic circulation. The first/last time when a DBV can be detected is defined as its formation/dissipation time, and the period between the vortex’s formation and dissipation is defined as its lifespan. The vertical extent of a DBV is defined as all continuous vertical levels for the appearance of notable vortex structures that belong to the same vortex (Fu et al. 2022b). The vertical level with the largest/smallest pressure is defined as the bottom/top level of the DBV, and the thickness of the vortex is defined as the difference between the bottom and top levels.
If a DBV can i) move out from the source region (26°–34°N, 111°–118°E) and ii) the distance between its formation and dissipation locations is ≥3°, then the DBV is classified into the mobile type; otherwise, the DBV is classified into the quasi-stationary (QS) type. For the mobile type of DBVs, its moving direction is defined by using the angle between its displacement vector [from the mean location at a DBV’s formation time (FT) and 1 h later to the mean location at its dissipation time and 1 h earlier] and the due east direction (lower-left corner of Fig. 1).
For each hour, the rainfall associated with a DBV is defined as the hourly precipitation ≥0.1 mm, which is located within a radius of 3° from the vortex’s center (Fu et al. 2022a). This circle (with a radius of 3°) is also defined as the DBV’s impact region (IR). We calculate the mean rainfall intensity of DBVs as follows: i) at every hour, calculate its horizontal averaged hourly rainfall (≥0.1 mm) within its impact region, and ii) calculate the temporal mean of the horizontal averaged hourly rainfall during the DBVs’ whole lifespan. We define the maximum rainfall intensity of DBVs as the maximum hourly rainfall within its impact region during its whole lifespan.
The DBVs’ occurrence frequency during a period is defined as the total number of DBVs within this period. Although the occurrence frequency can indicate the total number of DBVs during a period, it cannot show the total number of hours when DBVs appear. To make up for this deficiency, we raise another definition: the appearance frequency. The DBVs’ appearance frequency is defined as the total number of times when the DBVs can be detected by using the hourly ERA5 data during a period. For example, if a DBV is detected at 0000, 0100, and 0200 UTC during its whole lifetime, its lifespan is 2 h, its occurrence frequency is 1, and its appearance frequency is 3. There is a relationship between the lifespan of a DBV’s lifespan and its appearance frequency: (appearance frequency − 1) × 1 h (ERA5’s temporal resolution) = DBV’s lifespan. Therefore, for long-lived DBVs, their appearance frequencies ≈ their lifespans. In this study, a significant trend means it exceeds the confidence level of 95% (Mann–Kendall test). A significant correlation means the correlation coefficient (CC) between two factors passes the significance of 95% (the Student’s t test).
c. Vorticity budget
3. Multiscale temporal variations
a. Annual variations
During the 42 warm seasons, DBVs’ occurrence frequency is ∼3721, with ∼89 DBVs occurred in each season on average (Fig. 2a). The DBVs’ appearance frequency is ∼36 357 (Fig. 2b), implying that, on average, ∼23.6% of the time of a warm season (3672 h) is affected by DBVs. All these means the DBVs are an important type of mesoscale systems that frequently affect central and eastern China. Both the occurrence and appearance frequencies of DBVs are characterized by significant annual variations, with a correlation coefficient of ∼0.71 between them (exceeding the confidence level of 95%). The maxima of the occurrence and appearance frequencies appear in 2014 (∼114 DBVs) and 1998 (∼1190 times), respectively, and the minima of them appear in 1995 (∼57 DBVs) and 2000 (∼495 times), respectively. It should be noted that, in 1998 and 2000, when the DBVs’ appearance frequencies reach maximum and minimum, China experiences catastrophic flood and severe drought in central and eastern China (Luo et al. 2020; Fu et al. 2022a; Peng et al. 2023; Chyi and He 2023), respectively. Overall, during the 42 warm seasons, no significant linear trends are observed in the DBVs’ occurrence and appearance frequencies.
(a) The DBVs’ occurrence frequency in the warm seasons of different years, where the black dashed line is the mean value. (b) The DBVs’ appearance frequency in the warm seasons of different years, where the black dashed line is the mean value. (c) The DBVs’ occurrence frequency in each warm season (light blue dashed line) and their 9-yr running mean (thick blue solid line), where the black dashed line is the trend of the running mean during the whole period and the green/red dashed lines show the decreasing/increasing trends of the 9-yr running mean in different periods.
Citation: Journal of Climate 38, 7; 10.1175/JCLI-D-24-0304.1
b. Interdecadal variations
To analyze the features of the DBVs’ occurrence frequency on an interdecadal scale, we conduct a 9-yr running mean to the annual occurrence frequency. As Fig. 2c illustrates, the interdecadal variation of the DBVs’ occurrence frequency is notable, which can be divided into four stages: from 1983 to 1992, from 1992 to 2001, from 2001 to 2009, and from 2009 to 2016, during which significant trends are around −1.3, +1.5, −0.5, and +1.3 DBVs per year, respectively. Overall, from 1983 to 2016, the DBVs’ occurrence frequency shows a significant increasing trend of ∼0.3 DBV a−1. To understand the DBVs’ increasing trend, we calculate the 9-yr running means of the intensity and coverage indices (https://cmdp.ncc-ma.net/download/Monitoring/Index/Index_definition.docx) of the western Pacific subtropical high (WPSH), the summer monsoon index (SMI) (Wang and Fan 1999), and the source-region averaged 2-m temperature (TMP) and 850-hPa divergence (DIV), as these factors all can affect DBVs’ occurrence. It is found that, from 1983 to 2016, on an interdecadal scale, the WPSH (32.0 and 1.2 days−1), summer monsoon (0.23 day−1), and the source-region averaged divergence (1.3 × 10−7 s−1 day−1) and temperature (0.3°C day−1) all show significant increasing trends (Table 1; Fig. 3), implying their impacts tend to strengthen. All these factors show significant correlations to the DBVs’ occurrence frequency (Table 2): stronger WPSH (CC = ∼0.72), stronger summer monsoon (CC = ∼0.34/−0.48), stronger source-region averaged divergence (CC = ∼0.47), and higher source-region averaged 2-m temperature (CC = ∼0.53) are all corresponding to more DBVs. However, not all of these factors are favorable for DBVs’ formation. According to Fu et al. (2016b, 2017), four terms of the vorticity budget equation (including horizontal and vertical advection of vorticity, vertical stretching/shrinking, and tilting) can effectively reflect the mechanisms governing the DBVs’ formation. Of these, an enhancing divergence would reduce the cyclonic-vorticity production through vertical shrinking, which is detrimental for DBVs’ formation. In addition, the enhancing divergence may be caused by the WPSH’s intensifying (CC between WPSH and divergence is ∼0.49; Table 2): as the WPSH enhances, convection would be suppressed and descending motions would be strengthened; stronger descending motions contribute to stronger lower-level divergence (fluid continuity) and a warmer surface (via subsidence warming; from Table 2, CC between WPSH and 2-m temperature is ∼0.71). Stronger descending motions would reduce upward transport of cyclonic vorticity, which decelerate the DBVs’ formation (Fu et al. 2016b). In contrast, a stronger summer monsoon would result in stronger horizontal transport of vorticity (Fig. 3c), which would contribute to the convergence of cyclonic vorticity within the source region. This is favorable for the DBVs’ formation. We further calculate the vorticity budget within the source region by using Eq. (1) and show the results in Figs. 3c–f. It can be seen that terms HAV and TIL show a significantly increasing trend and a significantly positive correlation with DBVs’ occurrence frequency, implying they may be crucial for the DBVs’ increasing trend. In contrast, terms STR and VAV act conversely, as they show a significantly decreasing trend and a significantly negative correlation.
Trends (from 1983 to 2016) of the 9-yr running means of DBVs’ warm-season occurrence frequency (DOF), the intensity, and coverage indices of the WPSH, the SMI, and the SR averaged 2-m TMP and 850-hPa DIV. Bold means the trends exceed the 95% confidence level according to the Mann–Kendall test.
(a) The 9-yr running means of DBVs’ occurrence frequency (blue line), the intensity, and coverage indices of the WPSH (orange lines) and the SR averaged 2-m TMP (yellow line; °C), where the black dashed line is the trend of TMP. (b) As in (a), but for the SR averaged 850-hPa DIV (yellow line; 10−7 s−1) and the summer monsoon index. (c)–(f) The 9-yr running means of DBVs’ occurrence frequency (blue line) and the SR averaged vorticity budget terms (HAV, VAV, STR, and TIL) at 850 hPa (10−10 s−2). The black dashed lines are the trends of different budget terms (positive/negative trends that exceed the 95% confidence level are shown in red/blue values), and the CCs between DBVs’ occurrence frequency and different vorticity budget terms are shown in red/blue values (exceed the 95% confidence level) at the top of the figures.
Citation: Journal of Climate 38, 7; 10.1175/JCLI-D-24-0304.1
Pearson CCs between the 9-yr running means of different variables: DOF, the intensity and coverage indices of the WPSH (IIW, CIW), the SMI [the first/second is the index developed by Wang and Fan (1999)/Li and Zeng (2002)], and the SR averaged 2-m TMP and 850-hPa DIV. Bold means the correlations exceed the 95% confidence level according to the t test.
c. Monthly variations
The DBVs’ occurrence and appearance frequencies are characterized by notable monthly variations (Figs. 4a,b), with the maxima appeared in June (on average, DBVs can be detected in ∼33.4% of the time). For the occurrence frequency, its minimum appears in August, whereas the minimum of appearance frequency appears in September (on average, DBVs can be detected in ∼17.6% of the time). The difference is mainly because the lifespans of the DBVs in August are longer than those of September (Fig. 4c). Overall, from May to June, DBVs’ monthly mean occurrence and appearance frequencies are larger than those from July to September.
(a) Monthly variation of the DBVs’ occurrence frequency, where the numbers are the maximum and minimum values, respectively. (b) As in (a), but for the DBVs’ appearance frequency, where the percentages show the contributions of DBVs to the total duration of the month. (c) As in (a), but for the mean lifespans of the DBVs in different months.
Citation: Journal of Climate 38, 7; 10.1175/JCLI-D-24-0304.1
The DBVs’ monthly variations are closely related to the monthly variation of the background circulation (BC). As Fig. 5 shows, the wind within the source region and its southern/northern section increases from May to July and then decreases sharply in August and September. This is consistent with the monthly variation of the summer monsoon (Wang and Fan 1999). From May to July, the wind within southern section is larger than that within northern section (Fig. 5f), which results in strong convergence within the source region; after that, the difference between the wind within southern and northern sections decreases notably, with the decrease in convergence within the source region. The maximum difference between the wind within the southern and northern sections appears in June, resulting in the strongest convergence in the source region (Fig. 5f). This provides the most favorable condition for DBVs’ formation (by producing cyclonic vorticity through convergence-related vertical stretching). In contrast, in August and September, divergence dominates the source region, which means conditions become less favorable for DBVs’ formation. This explains why DBVs’ occurrence/appearance frequencies decrease sharply in these two months (Figs. 4a,b). Moreover, in June, the WPSH shows the largest coverage and intensity indices, whereas, in August and September, these two indices become much smaller than those of the other months (Figs. 5a–c). This is also consistent with the DBVs’ monthly variation features, as the WPSH can modulate the summer monsoon’s intensity and range (Wang 2002; Huang et al. 2012; Zhang et al. 2015).
(a)–(e) The 30-yr (from 1991 to 2020) monthly averaged 500-hPa geopotential height (black contour; gpm), 850-hPa wind (a full bar represents 4 m s−1), and 850-hPa vertical velocity (shading; Pa s−1), where the blue box is the SR, and the blue dashed line divides it into the northern and southern sections. (f) The northern section, southern section, and SR averaged wind speed and DIV.
Citation: Journal of Climate 38, 7; 10.1175/JCLI-D-24-0304.1
d. Diurnal variations
The DBVs’ diurnal variation is characterized by two peaks (Fig. 6a), one is at 2200 UTC (accounting ∼9.8% for the total occurrence frequency of DBVs) and the other is at 1000 UTC (∼6.8%), both of which are much larger than the mean percentage of the remaining part (∼3.8%). These two DBV peaks are consistent with the two rainfall peaks (i.e., from 1000 to 1200 UTC and from 2200 to 0000 UTC) within the source region of DBVs (not shown). The minimum occurrence frequency appears at 0400 UTC, which accounts ∼2.7% for the total occurrence frequency of DBVs. For the DBVs with a lifespan of ≥6 and ≥12 h, the two DBV peaks (2200 and 1000 UTC) also appear, but of a smaller percentage (not shown). The DBV forms at the peaks of 1000 UTC (1800 local standard time) and 2200 UTC (0600 local standard time) are defined as the afternoon (AN) and early morning (EM) types of DBVs, respectively. From Fig. 6b, the ANs and EMs both show notable monthly variations, with their maxima appeared in July (∼7.6%) and August (∼10.8%), respectively, which is not consistent with the monthly variations of the DBVs’ occurrence and appearance frequencies. The EM peaks are notable for the whole warm season, whereas the AN peak in August is not notable. We calculate the maximum difference among different EM peaks (1.8%) and divide it by the mean of the EM peaks (i.e., 9.8%). This value is defined as the change amplitude for the EM (∼18.4%). According to this definition, the change amplitude for the AN is ∼38.2%, and that for EM+AN is ∼9.6%. It is clear that EM has a much smaller change amplitude than that of the AN. Therefore, it can be concluded that the EM peak is more stable than the AN peak.
(a) The diurnal variation of DBVs’ occurrence frequency (%), where the light blue shows the minimum occurrence frequency, and the orange and blue shading show the occurrence frequencies of AN and EM types of DBVs, respectively. (b) The monthly variations of the contributions of the AN_DBV and EM_DBV, where the thick black percentages are the sum of the both types’ percentages.
Citation: Journal of Climate 38, 7; 10.1175/JCLI-D-24-0304.1
To understand the DBVs’ diurnal variation, we calculate the dynamical (divergence, vertical velocity, horizontal wind) and thermodynamical factors (2-m temperature, sensible heat flux, K index) averaged within the source region (Fig. 7). As the yellow line in Fig. 7c (only the DBVs’ formation time is used in calculation) shows, ∼1 h before the AN_ and EM_DBV peaks (i.e., 0900 and 2100 UTC), a peak of convergence appears, which is conducive to DBVs’ formation through the production of cyclonic vorticity by vertical stretching (Fu et al. 2017). This is the direct reason for why these two peaks appear. For the divergence of the background circulation (i.e., all the time during the 42 warm seasons is used in calculation), it shows a similar variation to that of the formation time (the correlation between them is ∼0.82; p < 0.05), implying the background circulation acts as a base for the formation of the AN_ and EM_DBV peaks. However, the difference between the divergence of the background circulation and the formation time is notable ∼1 h before the AN_ and EM_DBV peaks. Therefore, these two DBV peaks cannot be produced only by the local diurnal variation of background circulation, or in other words, nonlocal forcings are crucial for them. This can also be confirmed by the diurnal variation of 850-hPa wind averaged within the source region (Fig. 7d), as the difference between horizontal wind of the background circulation and the formation time is remarkable ∼1 h before the AN_ and EM_DBV peaks.
(a) The diurnal variation of the DBVs’ occurrence frequency (gray shading), the SR averaged 2-m TMP (blue line; K), and sensible heat flux (orange line; W m−2) during the 42 warm seasons. Green shading highlights 0900 and 2100 UTC. (b) As in (a), but for the K index (yellow line; K) and vertical velocity (red lines; Pa s−1). (c) As in (a), but for the 850-hPa DIV (10−6 s−1), where BC means all the time during the 42 warm seasons is used in calculation, and FT means only the DBVs’ FT is used in calculation. (d) The 42-warm season mean diurnal variation of the 850-hPa wind (a full bar is 4 m s−1; wind speed < 1 m s−1 is marked by numbers) within the SR, where the thin dashed lines mark the AN_DBV and EM_DBV peaks, and yellow shading marks the time when the BC’s wind reaches its maximum.
Citation: Journal of Climate 38, 7; 10.1175/JCLI-D-24-0304.1
In terms of background circulation features, for the AN peak, at 0900 UTC, sensible heat flux keeps positive (Fig. 7a), but it begins to change from positive to negative, implying the change from the convective to stable boundary layer (Markowski and Richardson 2010). Correspondingly, the vertical velocity ω at 850 hPa begins to change from ascending to descending (thick red line in Fig. 7b), and ω950 (ascending motions) weakens with time. There are three striking features for the diurnal variation of ω: i) the decreasing rate of ω850 from 0800 to 0900 UTC is smaller than that from 0600 UTC (when the 2-m temperature reaches its maximum) to 0800 UTC, ii) ω950 weakens at a quasi-steady decreasing rate from 0600 to 0900 UTC, which makes iii) the difference between the vertical velocity at 850 and 950 hPa (i.e., ω850 − ω950) reduce. According to the continuity equation (∂u/∂x + ∂υ/∂y) = −(∂ω/∂p) (where u and υ are the zonal and meridional winds, respectively; p is the pressure), a decrease in −(∂ω/∂p) corresponds to a decrease in the horizontal divergence (∂u/∂x + ∂υ/∂y). This explains why a weakening in divergence that contributes to the AN_DBV peak appears in 0900 UTC (Fig. 7c). For the EM peak, at 2100 UTC, sensible heat flux is negative (Fig. 7a; stable boundary layer), 2-m temperature reaches its minimum, and K index (larger means stronger convective instability) is below its diurnal mean value (Fig. 7b). All these mean that the local thermodynamical conditions are not conducive to maintain ascending motions. Correspondingly, from 1700 to 2100 UTC, ω850 (ascending motions) decreases in intensity, and ω950 changes from ascending to descending motions, with the difference between them (ω850 − ω950) increasing in value. According to (∂u/∂x + ∂υ/∂y) = −(∂ω/∂p), the convergence within the source region enhances. This explains why a strong convergence that contributes to the EM_DBV peak appears in 2100 UTC (Fig. 7c). As discussed above, the variation of vertical motions is closely related to the convergence’s two peaks (i.e., 0900 and 2100 UTC), which directly contribute to the formation of the AN and EM peaks. Overall, compared with the EM_DBVs, the difference between the divergence/wind of the background circulation and the formation time is much larger for the AN_DBVs, implying the nonlocal forcings play a more important role in the AN_DBVs’ formation. This explains why the EM peak is more stable than the AN peak.
The diurnal variations of HAV, VAV, STR, and TIL are shown in Fig. 8. It can be seen that, of the four terms, only STR and VAV show positive values (Figs. 8b,c), implying that they are conducive to DBVs’ formation. In terms of diurnal cycle features, VAV has only one peak which appears from 1000 to 1100 UTC (Fig. 8b), whereas STR has two peaks which appears from 0900 to 1200 and 2000 UTC, respectively (Fig. 8c). The first peak of STR and the single peak of VAV are closely related to the AN_DBVs (VAV is larger than twice STR, which means it is more important), and the second peak of STR is crucial for the EM_DBVs. Therefore, the AN_DBVs and EM_DBVs are generated through different mechanisms. The three peaks of STR and VAV indicate that dynamical conditions are favorable for the formation of AN_DBVs and EM_DBVs. In terms of correlation, the 2-m temperature only shows significant correlations with HAV and STR (Figs. 8a,c), implying that a warmer surface tends to associate with a smaller negative HAV and a smaller positive STR. The former contributes to DBVs’ formation, whereas the latter acts conversely. As Figs. 8f and 8h show, the meridional wind only shows significant correlations with VAV and TIL. This indicates that a larger southerly wind tends to associate with a smaller positive VAV (Fig. 8f) and a smaller negative TIL. The latter contributes to DBVs’ formation, whereas the former acts conversely.
(a) The diurnal variation of the SR averaged 2-m TMP (blue line; °C) and HAV (black line; 10−10 s−2), where the CC between them is shown in value (exceeding the 95% confidence level is highlighted in red; otherwise, it cannot pass the significance test) at the top of the figure, and the two thick red lines mark 1000 and 2200 UTC. (b),(d) As in (a), but for VAV, STR, and TIL, respectively. The thin dashed line is the zero line. (e) The diurnal variation of the SR averaged υ-component wind (red line; m s−1), where the CC between them is shown in value (exceeding the 95% confidence level is highlighted in red; otherwise, it cannot pass the significance test) at the top of the figure. (f)–(h) As in (e), but for VAV, STR, and TIL, respectively.
Citation: Journal of Climate 38, 7; 10.1175/JCLI-D-24-0304.1
4. Horizontal distribution and tracks
a. Distribution of DBVs’ occurrence and impact
As Fig. 9a shows, although DBVs can form at almost any location within the source region, the distribution of their occurrence frequency is characterized by notable unevenness. Of these, the centers with an occurrence frequency of ≥10 DBVs are mainly located within the northern section of Hubei, the junction zone of Hubei and Hunan, and the northern section of Jiangxi. In contrast, the southwestern, northeastern, and southeastern sections of the source region show a relatively low occurrence frequency. To clarify the relationship between DBVs and terrain, we calculate the correlation coefficient between the DBVs’ occurrence frequency and terrain height. The correlation coefficient is −0.31 (exceeding the confidence level of 99.9%), implying that more DBVs tend to form over the terrain with a lower altitude. For the DBVs with a lifespan of ≥6 h (Fig. 9b), we defined them as the longer-lived DBVs. Overall, longer-lived DBVs show a different horizontal distribution from that of all DBVs (Figs. 9a,b), and there are three striking differences: i) the occurrence frequency of longer-lived DBVs is much smaller than those of all DBVs, ii) around 8% of the source region does not experience the formation of longer-lived DBVs, and iii) centers with an occurrence frequency of ≥4 longer-lived DBVs are mainly located in the middle-western and eastern sections of the source region (Fig. 9b).
(a) The DBVs’ occurrence frequency (shading dots; times), i.e., the total counts of all DBVs at different locations (reflected by DBVs’ centers), where the red dashed box shows the DBVs’ SR. (b) As in (a), but for the occurrence frequency of the DBVs with a lifespan of ≥6 h. (c) The appearance frequency of DBVs’ impacts (shading; times), where the thick red contour outlines the frequency of 500 times, and the orange dashed box outlines the SR. JS = Jiangsu; ZJ = Zhejiang; SD = Shandong.
Citation: Journal of Climate 38, 7; 10.1175/JCLI-D-24-0304.1
As Fig. 9c shows, DBVs exert notable impacts on the regions in central, North, and South China, as well as the Korean Peninsula and Japan. The highest appearance frequency of DBVs’ impacts appears within the center of the source region, whereas the southwestern and southeastern sections of the source region show the lowest frequency within the source region. The region with an impact frequency of ≥500 times (∼12 times during a warm season on average) is roughly orientated in a southwest–northeast direction, which is corresponding to DBVs’ tracks (section 4b). Within this region, the impact frequency mainly decreases northeastwards, with Jiangsu showing the largest frequency outside the source region.
b. DBVs’ tracks
According to section 2b, we classify the DBVs into mobile and QS types. As most of the DBVs with a lifespan < 6 h (shorter-lived DBVs) cannot move out from the source region, we mainly focus on the longer-lived DBVs. According to statistic, during 42 warm seasons, there are a total of 1491 longer-lived DBVs. Of these, 347 DBVs belonged to the mobile type, which accounts for ∼9.3% of all DBVs, implying most DBVs (including 1144 longer-lived DBVs and 2230 shorter-lived DBVs) are characterized by a quasi-stationary behavior. The tracks of the longer-lived DBVs are shown in Fig. 10a, of which ∼60.8% belong to the EE type (Fig. 10b). These DBVs can move across the East China Sea and Yellow Sea and exert influences on middle and southern Japan. Around 30% of the longer-lived DBVs belong to the NE type, which exerts notable effects on the Korean Peninsula (Figs. 10a,b). The sum contribution of the EE and NE types is ∼90.8%, since the 500-hPa westerly and southwesterly winds dominate the source region during warm seasons (Figs. 5a–e) and act as steering flow. About 5.5% of the longer-lived DBVs are classified into the SE type (ranks third), which can impact Guangdong and Taiwan. In addition, there are no WW/NW type of DBVs, since easterly and southeasterly steering flow (DBVs’ steering flow is mainly located in the layer from 950 to 500 hPa) rarely dominates the source region.
(a) The tracks of different types of DBVs (EE = eastward moving; NE = northeastward moving; NN = northward moving; NW = northwestward moving; WW = westward moving; SW = southwestward moving; SS = southward moving; SE = southeastward moving; QS = quasi stationary), which the colored dots show the centers of the DBVs along their tracks, and the gray shading outlines the impact ranges of the DBVs. (b) The relative contributions of different moving directions to the mobile type of DBVs. (c) The monthly variation of the QS type, mobile type, and long-lived DBVs (it equals the sum of the former two types), where the percentages are the ratios of the mobile type to the long-lived DBVs in each month.
Citation: Journal of Climate 38, 7; 10.1175/JCLI-D-24-0304.1
The monthly variations of the QS and mobile types (Fig. 10c) are similar to the DBVs’ monthly variation of occurrence frequency (Fig. 4a): their maxima all appear in June, and the values in May and June are larger than those of July, August, and September. The largest percentage (∼28.9%) of the mobile type of DBVs appears in May (Fig. 10c), when strong westerly wind dominates 500 hPa of the source region (Fig. 5a) and acts as steering flow. From May to August, the ratios of the mobile type of DBVs decrease from ∼28.9% to ∼20.3%, corresponding to the weakening of the westerly wind at 500 hPa. The minimum ratio (∼15.5%) appears in September (Fig. 10c), when easterly wind appears at 850 hPa (Fig. 5e), which is detrimental for the eastward displacement of the DBVs.
5. Other key features
a. Lifespan features
Because most reanalysis data have a temporal resolution of 6 h (e.g., ERA-I, FNL, CFSR, JRA55), the lower limit of the DBVs’ lifespans that they can determine is 6 h. As we use hourly ERA5 reanalysis data in detecting/tracking DBVs, vortices with a lifespan of ≥ 1 h can be determined. From Fig. 11a, among the 3721 DBVs, 1113 have a lifespan of 1–2 h (accounting for ∼29.9%) and 1119 have a lifespan of 3–5 h (30.1%). These shorter-lived DBVs (with a lifespan of <6 h) account for a sum contribution of ∼60%; however, they cannot be detected by the reanalysis data with a temporal resolution of 6 h or coarser. The DBVs with a lifespan of 6–23 h occupy a proportion of ∼32.1%, and those DBVs that last for no less than 1 day (≥24 h) account for ∼7.9%. Overall, the proportions of the DBVs decrease in their proportions as their lifespans increase from 1 to 35 h, whereas the contribution of 36–41 h is higher than that of 30–35 h. It should be noted that the DBVs with a lifespan of ≥48 h contribute ∼2.5% to the DBVs’ total occurrence frequency, which is higher than the averaged percentage of the four groups from 24 to 47 h (∼1.4%).
(a) The occurrence frequencies of the DBVs with different lifespans, where the black numbers are the occurrence frequencies of the DBVs with different lifespans and the red numbers are the percentage. (b) The annual occurrence frequency of the DBVs with a lifespan of ≥6 and <6 h, where the blue and green dashed lines mark the 42-yr mean of the occurrence frequencies of the DBVs with a lifespan of <6 and ≥6 h, respectively; the black line with small boxes show the proportion of the DBVs with a lifespan of ≥6 h; and the black dashed line marks the 42-yr mean of the proportion of the DBVs with a lifespan of ≥6 h.
Citation: Journal of Climate 38, 7; 10.1175/JCLI-D-24-0304.1
As Fig. 11b shows, the longer- and shorter-lived DBVs both show notable annual variations. During the 42 warm season, the mean occurrence frequencies of longer- and shorter-lived DBVs are ∼53.1 and ∼35.5, respectively. The maximum occurrence frequencies of the longer- and shorter-lived DBVs appear in 2015 and 2014, respectively, and the minimum occurrence frequencies of them appear in 2000 and 1995, respectively. In addition, longer-lived DBVs’ proportion also shows a notable annual variation, with a mean proportion of ∼40%, a minimum proportion of ∼30.2% (1985), and a maximum proportion of 50.9% (1988). In the study period, no significant trends are detected in the occurrence frequencies of longer- and shorter-lived DBVs and the proportion of longer-lived DBVs.
b. Vertical extents
Following the method described in section 2b, we investigate the vertical extents of the longer-lived DBVs. As Fig. 12b shows, although some DBVs’ top levels are located around 300 hPa or even higher, over 75% of the 1491 longer-lived DBVs’ highest top levels are located below 500 hPa. This implies that the DBVs are mainly situated in the mid- and lower troposphere. In addition, over 75% of the longer-lived DBVs’ lowest bottom levels are located in the layer of 975–900 hPa (mean is ∼938 hPa). On average, the difference between the longer-lived DBVs’ highest top levels and lowest bottom levels is ∼274 hPa (i.e., 938–664 hPa), which is thicker than the mean maximum extents of the longer-lived DBVs (as DBVs’ highest top levels and lowest bottom levels sometimes do not appear together). This indicates that DBVs mainly belonged to a type of shallow mesoscale vortices. In terms of the temporal mean calculated within the respective lifespans of the longer-lived DBVs (Fig. 12a), ∼50% of the longer-lived DBVs’ top levels are located in the layer of 825–725 hPa (the mean value is ∼753 hPa), and ∼50% of their bottom levels are located in the layer of 925–875 hPa (the mean value is ∼903 hPa). Therefore, on average, the mean vertical extent of the longer-lived DBVs is ∼150 hPa (i.e., 903–753 hPa), with the central level located at ∼828 hPa (i.e., the mean of 903 and 753 hPa). Also, these confirm the longer-lived DBVs are a type of shallow mesoscale vortices that are mainly located in the mid- and lower troposphere, and its central level is around 825 hPa (for the reanalysis data that have a vertical interval of 50, 850 hPa can be used as DBVs’ central level).
(a) The box-and-whisker plot of DBVs’ temporal averaged (within their respective lifespans) vertical extents (only longer-lived DBVs are considered), where the left column shows the vortices’ bottom levels (hPa) and the right column illustrates the top levels (hPa). (b) As in (a), but for the (left) lowest bottom levels and (right) highest top levels during DBVs’ lifespans. The box range represents the 25th–75th percentile, the red lines in the box represent the median values, and the green triangles mark the mean values.
Citation: Journal of Climate 38, 7; 10.1175/JCLI-D-24-0304.1
c. Precipitation features
As documented in section 2b, we use the half-hourly GPM data from 2001 to 2020 for calculating the precipitation associated with DBVs. Overall, the accumulated DBV-associated precipitation shows a similar distribution to that of the appearance frequency of DBVs’ impacts (cf. Figs. 9c and 13a), with the strongest rainfall center (annual mean is above 240 mm) appearing within the joint region of Hubei, Jiangxi, and Hunan. Outside the source region, there is a secondary rainfall center (annual mean is above 50 mm), which is located over the sea east of Shandong and Jiangsu. Outside the region with DBVs’ appearance frequency ≥ 500 (A500; the thick red line in Fig. 13a), the DBV-associated precipitation is much weaker. For the heaviest DBV-associated hourly precipitation during 20 warm seasons, its strongest center (≥100 mm h−1) appears in the offshore regions east of Shandong and Jiangsu (Fig. 13b), which is much stronger than those within the source region. In addition, some regions outside A500 (mainly over sea) also experience strong DBV-associated precipitation. Overall, inside the DBVs’ impact region (i.e., the shading regions in Fig. 8c), ∼2.7% of the strongest local hourly precipitation from 2001 to 2020 is directly related to the DBVs (Fig. 13c); for the region inside A500, the proportion is ∼7.2%; and for the region inside the source region, the proportion is much higher (∼15.2%). In addition, outside the source region, ∼1.4% of the strongest local hourly precipitation is directly related to the DBVs (Fig. 13c), and even outside A500, the DBVs is directly related to the strongest local hourly precipitation in some regions.
(a) The accumulated (in the warm seasons from 2001 to 2020) DBV-associated precipitation (shading; mm), where the thick red contour outlines the appearance frequency of 500 times, and the white dashed box outlines the SR. (b) As in (a), but for the heaviest DBV-associated hourly precipitation during the warm seasons from 2001 to 2020. (c) The proportion (%) of grids (0.1° × 0.1°) where the heaviest hourly precipitation is related to the DBVs, relative to the DBVs’ IR, the region inside/outside the SR, and the region inside/outside the area with an DBVs’ appearance frequency ≥ 500 (A500; the thick red line in Fig. 12a). HB = Hubei; HN = Hunan; JX = Jiangxi; BHS = Bohai Sea.
Citation: Journal of Climate 38, 7; 10.1175/JCLI-D-24-0304.1
On average, within the source region, the accumulated DBV-associated precipitation accounts ∼14.5% for the total accumulated precipitation during the 20 warm seasons from 2001 to 2020 (Fig. 14a; Table 3). There is a center of >20%, which is located within the center of the source region. This implies that DBVs play an important role in producing rainfall within the source region. For the heavy rainfall (≥15 mm h−1) and short-duration heavy rainfall (≥20 mm h−1), they show similar distributions (Figs. 14b,c). On average, within the source region, the accumulated DBV-associated heavy rainfall accounts ∼18.6% for the total accumulated heavy rainfall during the 20 warm seasons (Table 3), and that for the accumulated DBV-associated short-duration heavy rainfall is ∼19.8% (Table 3). For the heaviest DBV-associated hourly precipitation during 20 warm seasons, its ratio to the strongest local hourly precipitation from 2001 to 2020 shows many strong centers (>50%) within the source region (Fig. 14d), with a source-region averaged value of ∼72.5% (Table 3). This means that as the rainfall intensity grows, the relative importance of DBVs’ contribution increases.
(a) The proportion of the accumulated DBV-associated precipitation (shading; %) relative to the total accumulated precipitation during the warm seasons from 2001 to 2020, where the black contour highlights the contour of 20%, the thick red contour outlines the appearance frequency of 500 times, and the black dashed box outlines the SR. (b) As in (a), but for the heavy rainfall (≥15 mm h−1) and the contour of 30%. (c) As in (a), but for the short-duration heavy rainfall (≥20 mm h−1) and the contour of 50%. (d) The ratio of the heaviest DBV-associated hourly precipitation during 20 warm seasons to the strongest local hourly precipitation from 2001 to 2020 (shading; %), where the black contour highlights the contour of 50%.
Citation: Journal of Climate 38, 7; 10.1175/JCLI-D-24-0304.1
Proportion (%) of the DBV-related accumulated total precipitation (ATP) relative to the warm-season ATP during the 20 warm seasons (from 2001 to 2020); proportion of the DBV-related accumulated heavy precipitation (AHP) relative to the warm-season AHP during the 20 warm seasons; proportion of the DBV-related accumulated short-duration heavy precipitation (SDP) relative to the warm-season accumulated SDP during the 20 warm seasons; proportion of the DBV-related heaviest torrential precipitation (HTP) relative to the HTP during the 20 warm seasons. AWIR = average within the whole impact region of DBVs (i.e., the shading area in Fig. 9); A500 = average within the region with an appearance frequency ≥ 500 (i.e., the thick red line in Fig. 9); ADSR = average within the DBVs’ SR.
In terms of mean rainfall intensity, the QS type has a mean value of ∼1.4 mm h−1 (Fig. 15a), which is smaller than that of the mobile type (∼1.9 mm h−1). For the maximum rainfall intensity (Fig. 15b), the mean value of the mobile type (∼41.9 mm h−1) is almost 2 times of that of the QS type (∼23.8 mm h−1). In the mobile type, the EE, NE, and SE account for a sum proportion of ∼96.3%. Their mean and maximum rainfall intensities differ from each other notably: As Fig. 15c shows, the mean value of the EEs’ mean rainfall intensity (1.9 mm h−1) is the largest among these three types, whereas those of the NE and SE are similar to each other. For the maximum rainfall intensity (Fig. 15d), the SEs’ mean value is the largest (50.4 mm h−1), the EE ranks second (42.8 mm h−1), and the NE ranks third (38.5 mm h−1). This is consistent with the horizontal distribution of specific humidity during warm seasons, which mainly decreases from south to north (Fig. 15e).
(a) The box-and-whisker plot of the mean rainfall intensity for the QS and mobile types of DBVs. (b) The box-and-whisker plot of the maximum rainfall intensity for the QS and mobile types of DBVs. (c) As in (a), but for the EE, NE, and SE types of DBVs. (d) As in (b), but for the EE, NE, and SE types of DBVs. The black triangles highlight the stronger/strongest type of DBV. (e) The 20-warm-season (from 2001 to 2020) averaged 825-hPa specific humidity (shading; g kg−1), where the black dashed box is the SR, and gray shading outlines terrain higher than 1750 m.
Citation: Journal of Climate 38, 7; 10.1175/JCLI-D-24-0304.1
6. Conclusions and discussion
As one of the most important types of mesoscale vortices that generate in China, the DBVs have formed a research focus since they are discovered. However, for all existing researches on the DBVs, there is a striking flaw, since no studies have ever investigated the long-term climatological features of this type of vortices, and all the basic features about the DBVs are obtained from short-term statistics and/or case studies. To this end, we conduct a 42-yr statistical study on the DBVs by using the hourly ERA5 reanalysis data, which partly fill these knowledge gaps.
Statistics indicate DBVs are an important type of mesoscale systems that frequently affect central and eastern China, because on average, DBVs’ occurrence frequency is ∼17.4 month−1 (that documented in Fu et al. (2016b) is ∼12.2 month−1) and ∼23.6% of the time of a warm season is affected by DBVs. Around 60% of the DBVs last for <6 h; those with a lifespan of ≥12 h occupy a proportion of ∼20.0%, which is similar to that (19.5%) found by Fu et al. (2016b); and those last for ≥24 h only account for ∼7.9%. Overall, longer-lived DBVs are a type of shallow (mean thickness is ∼150 hPa) mesoscale vortices that are mainly located in the mid- and lower troposphere (mean central level is ∼825 hPa). These are similar to the findings documented in Fu et al. (2016b).
In terms of spatial features, DBVs can form at almost any location of the source region, with high-frequency centers mainly located in northern Hubei, northern Jiangxi, and the Junction of Hubei and Hunan. DBVs play a crucial role in producing rainfall within the source region. As the rainfall intensity grows, the relative importance of DBVs’ contribution increases from ∼13.6% (accumulated precipitation) to ∼71.5% (heaviest precipitation). Outside the source region, DBVs can induce heavy rainfall and short-duration heavy rainfall along their tracks, particularly for the offshore regions east of Shandong and Jiangsu. Inside/outside the source region, ∼15.2%/∼1.4% of the strongest local hourly precipitation is directly related to the DBVs. Most DBVs belong to the QS type, and only ∼9.3% of the DBVs (347 DBVs) can vacate their source region. On average, the mobile type shows stronger mean and maximum rainfall intensities than those of the QS type. Of the mobile type, ∼60.8% move eastward, ∼30% move northeastward, and ∼5.5% move southeastward. These three types of DBVs can induce rainfall over North and South China, as well as the Korean Peninsula and Japan. Overall, the SE type has the largest maximum rainfall intensity, and the EE has the strongest mean rainfall intensity.
In terms of temporal features, occurrence and appearance frequencies of DBV both feature significant annual variations, but there are no significant linear trends. In contrast, on an interdecadal scale, DBVs’ occurrence frequency shows a significant increasing trend of ∼0.3 DBV a−1, which may correspond to the enhancing horizontal transport of vorticity by the intensifying summer monsoon (Table 1). DBVs’ occurrence and appearance frequencies are characterized by notable monthly variations, with the maxima appeared in June. This is consistent with the monthly variations of summer monsoon and its associated divergence. DBVs’ diurnal variation shows an EM peak at 2200 UTC (∼9.8% in proportion), and an AN peak at 1000 UTC (∼6.8%), with the former more stable than the latter. The reason for these two peaks is that, corresponding to the variation of vertical motions, two peaks of convergence appear within the source region ∼1 h before the AN_ and EM_DBV peaks, respectively. The AN and EM peaks cannot be produced only by the local diurnal variation of background circulation; instead, the nonlocal forcings are crucial for them, particularly for the AN peak.
Compared with previous DBV-related studies, new findings obtained in this study mainly include i) interdecadal and diurnal variation features of DBVs and their possible mechanisms, ii) mechanisms governing DBVs’ monthly variation, iii) features of DBVs’ impact region, and iv) features of DBV-associated short-duration heavy rainfall and heaviest rainfall. These are helpful to reach a more comprehensive understanding of DBVs. However, as GPM rainfall underestimates heavy rainfall notably, DBVs’ contribution in precipitation may show inevitable uncertainty. Therefore, it is better to update the results by using station observations. In addition, this paper does not discuss composite background circulations under which DBVs are generated, particularly for the EM and AN types of DBVs. Solving these scientific problems will provide more information about DBVs’ formation.
As an alternative for real observational data, reanalysis datasets show inevitable errors, which may influence the credibility of the research results based on them. The present study is based on the ERA5 data, which can capture well the observed diurnal variation features in surface pressure, air temperature, and relative humidity (Dai 2023), but shows two gaps in the diurnal cycle of 925-hPa wind due to the 12-h data assimilation window (Chen et al. 2021). Known and unknown advantages and disadvantages of the ERA5 data can affect the DBVs’ features obtained in this study. To reduce the uncertainties caused by using different types of reanalysis data for investigating features of the DBVs, in addition to the ERA5 data, we recommend to conduct similar analyses on the DBVs (as this study) by using more reanalysis datasets such as the Japanese 55-year Reanalysis (JRA-55; Kobayashi et al. 2015), Climate Forecast System Reanalysis (CFSR; Saha et al. 2010), and so on. After obtaining the detailed features of the DBVs based on each type of reanalysis data, a further study on discovering commonalities among different types of datasets is crucial to reach more reliable results about the DBVs. As this study has shown a detailed report about the DBVs based on the ERA5 data, it can be used as a reference for the comparisons for future studies. The ERA5 reanalysis data use a 12-h data assimilation window (Hersbach et al. 2020; Chen et al. 2021), which may affect the features of DBVs’ diurnal cycle. As the two peaks of DBVs’ formation (i.e., 1000 and 2200 UTC; Fig. 6a) are one of the most striking features for DBVs’ diurnal cycle, we compare the DBVs derived from ERA5 at 1200/0000 UTC (2 h after 1000/2200 UTC) with the DBVs derived from JRA55 at 1200/0000 UTC (its temporal resolution and data assimilation window are 6 h) by using 200 randomly selected DBVs (including 100 DBVs with a lifespan of <6 h, 50 DBVs with a lifespan of 6 h ≤ lifespan < 12 h, and 50 DBVs with a lifespan of ≥12 h), to check whether the two datasets are consistent. If the distance between the centers of the vortices derived from two types of datasets is ≤3° (the mean radius of the DBVs), we regard the two vortices as a same vortex (i.e., the ERA5 and JRA55 are consistent); otherwise, the two datasets are not consistent. We found that, for the group of DBVs with a lifespan of <6 h, ∼66% of the DBVs are consistent; for the group of DBVs with a lifespan of 6 h ≤ lifespan < 12 h, ∼74% of the DBVs are consistent; and for the group of DBVs with a lifespan of ≥12 h, ∼88% of the DBVs are consistent. Overall, ∼73.5% of the 200 DBVs are consistent for the ERA5 and JRA-55. This ensures the reliability of this study.
Acknowledgments.
The authors thank ECMWF for providing the ERA5 data. This research was supported by the National Natural Science Foundation of China (42475008) and Strategy Priority Research Program of Chinese Academy of Sciences (XDB0760400).
Data availability statement.
The ERA5 data used in this study (Hersbach et al. 2020) are freely available on Copernicus Data Store at https://cds.climate.copernicus.eu/cdsapp#!/dataset/reanalysis-era5-pressure-levels?tab=overview. The GPM IMERG data were downloaded from the Goddard Earth Sciences Data and Information Services Center (https://disc.gsfc.nasa.gov/datasets?keywords=GPM&page=1).
REFERENCES
Chen, G., Y. Du, and Z. Wen, 2021: Seasonal, interannual, and interdecadal variations of the East Asian summer monsoon: A diurnal-cycle perspective. J. Climate, 34, 4403–4421, https://doi.org/10.1175/JCLI-D-20-0882.1.
Chen, S.-J., and L. Dell’osso, 1984: Numerical prediction of the heavy rainfall vortex over Eastern Asia monsoon region. J. Meteor. Soc. Japan, 62, 730–747, https://doi.org/10.2151/jmsj1965.62.5_730.
Chyi, D., and L. He, 2023: Stage characteristics and mechanisms of extreme high temperature in China in summer of 2022. J. Appl. Meteor. Sci., 34, 385–399, https://doi.org/10.11898/1001-7313.20230401.
Curio, J., R. Schiemann, K. I. Hodges, and A. G. Turner, 2019: Climatology of Tibetan Plateau vortices in reanalysis data and a high-resolution global climate model. J. Climate, 32, 1933–1950, https://doi.org/10.1175/JCLI-D-18-0021.1.
Dai, A., 2023: The diurnal cycle from observations and ERA5 in surface pressure, temperature, humidity, and winds. Climate Dyn., 61, 2965–2990, https://doi.org/10.1007/s00382-023-06721-x.
Ding, Y., Y. Liu, and Z.-Z. Hu, 2021: The record-breaking mei-yu in 2020 and associated atmospheric circulation and tropical SST anomalies. Adv. Atmos. Sci., 38, 1980–1993, https://doi.org/10.1007/s00376-021-0361-2.
Dong, P., and S. Zhao, 2004: A diagnostic study of mesoscale lows (disturbances) on meiyu front and associated heavy rainfall (in Chinese). Chin. J. Atmos. Sci., 28, 876–891, https://doi.org/10.3878/j.issn.1006-9895.2004.06.07.
Du, Y., D. Wang, J. Zhu, Z. Lin, and Y. Zhong, 2022: Intercomparison of multiple high-resolution precipitation products over China: Climatology and extremes. Atmos. Res., 278, 106342, https://doi.org/10.1016/j.atmosres.2022.106342.
Feng, S.-L., S.-L. Jin, S.-M. Fu, J.-H. Sun, and Y.-C. Zhang, 2020: Formation of a kind of heavy-precipitation-producing mesoscale vortex around the Sichuan Basin: An along-track vorticity budget analysis. Atmos. Sci. Lett., 21, e949, https://doi.org/10.1002/asl2.949.
Feng, X., C. Liu, R. Rasmussen, and G. Fan, 2014: A 10-yr climatology of Tibetan Plateau vortices with NCEP climate forecast system reanalysis. J. Appl. Meteor. Sci., 53, 34–46, https://doi.org/10.1175/JAMC-D-13-014.1.
Fu, S., F. Yu, D. Wang, and R. Xia, 2013: A comparison of two kinds of eastward-moving mesoscale vortices during the mei-yu period of 2010. Sci. China Earth Sci., 56, 282–300, https://doi.org/10.1007/s11430-012-4420-5.
Fu, S., W. Li, J. Sun, J. Zhang, and Y. Zhang, 2015: Universal evolution mechanisms and energy conversion characteristics of long-lived mesoscale vortices over the Sichuan Basin. Atmos. Sci. Lett., 16, 127–134, https://doi.org/10.1002/asl2.533.
Fu, S.-M., J.-H. Sun, J. Ling, H.-J. Wang, and Y.-C. Zhang, 2016a: Scale interactions in sustaining persistent torrential rainfall events during the Mei-yu season. J. Geophys. Res. Atmos., 121, 12 856–12 876, https://doi.org/10.1002/2016JD025446.
Fu, S.-M., J.-P. Zhang, J.-H. Sun, and T.-B. Zhao, 2016b: Composite analysis of long-lived mesoscale vortices over the middle reaches of the Yangtze River valley: Octant features and evolution mechanisms. J. Climate, 29, 761–781, https://doi.org/10.1175/JCLI-D-15-0175.1.
Fu, S.-M., J.-H. Sun, Y.-L. Luo, and Y.-C. Zhang, 2017: Formation of long-lived summertime mesoscale vortices over Central East China: Semi-idealized simulations based on a 14-year vortex statistic. J. Atmos. Sci., 74, 3955–3979, https://doi.org/10.1175/JAS-D-16-0328.1.
Fu, S.-M., J.-P. Zhang, H. Tang, L.-Z. Jiang, and J.-H. Sun, 2020: A new mesoscale-vortex identification metric: Restricted vorticity and its application. Environ. Res. Lett., 15, 124053, https://doi.org/10.1088/1748-9326/abcac6.
Fu, S.-M., H. Tang, J.-H. Sun, T.-B. Zhao, and W.-L. Li, 2022a: Historical rankings and vortices’ activities of the extreme mei-yu seasons: Contrast 2020 to previous Mei-yu seasons. Geophys. Res. Lett., 49, e2021GL096590, https://doi.org/10.1029/2021GL096590.
Fu, S.-M., Y.-C. Zhang, H.-J. Wang, H. Tang, W.-L. Li, and J.-H. Sun, 2022b: On the evolution of a long-lived mesoscale convective vortex that acted as a crucial condition for the extremely strong hourly precipitation in Zhengzhou. J. Geophys. Res. Atmos., 127, e2021JD036233, https://doi.org/10.1029/2021JD036233.
Gao, K., and Y.-M. Xu, 2001: A simulation study of structure of mesovortexes along meiyu front during 22∼30 June 1999 (in Chinese). Chin. J. Atmos. Sci., 25, 740–756, https://doi.org/10.3878/j.issn.1006-9895.2001.06.02.
Hersbach, H., and Coauthors, 2020: The ERA5 global reanalysis. Quart. J. Roy. Meteor. Soc., 146, 1999–2049, https://doi.org/10.1002/qj.3803.
Hou, A. Y., and Coauthors, 2014: The Global Precipitation Measurement mission. Bull. Amer. Meteor. Soc., 95, 701–722, https://doi.org/10.1175/BAMS-D-13-00164.1.
Hu, B.-W., and E.-F. Pan, 1996: Two kinds of cyclonic disturbances and their accompanied heavy rain in the Yangtze River Valley during the mei-yu reriod (in Chinese). J. Appl. Meteor. Sci., 7, 138–144.
Huang, R., J. Chen, L. Wang, and Z. Lin, 2012: Characteristics, processes, and causes of the spatio-temporal variabilities of the East Asian monsoon system. Adv. Atmos. Sci., 29, 910–942, https://doi.org/10.1007/s00376-012-2015-x.
Kirk, J. R., 2003: Comparing the dynamical development of two mesoscale convective vortices. Mon. Wea. Rev., 131, 862–890, https://doi.org/10.1175/1520-0493(2003)131<0862:CTDDOT>2.0.CO;2.
Kobayashi, S., and Coauthors, 2015: The JRA-55 reanalysis: General specifications and basic characteristics. J. Meteor. Soc. Japan, 93, 5–48, https://doi.org/10.2151/jmsj.2015-001.
Li, J., and Q. Zeng, 2002: A unified monsoon index. Geophys. Res. Lett., 29, 1274, https://doi.org/10.1029/2001GL013874.
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, 427–459, https://doi.org/10.1007/s13351-020-0006-x.
Markowski, P., and Y. Richardson, 2010: Mesoscale Meteorology in Midlatitudes. John Wiley and Sons, 414 pp.
Ni, C., G. Li, and X. Xiong, 2017: Analysis of a vortex precipitation event over southwest China using AIRS and in situ measurements. Adv. Atmos. Sci., 34, 559–570, https://doi.org/10.1007/s00376-016-5262-4.
Ninomiya, K., 2000: Large- and meso-α-scale characteristics of Meiyu/Baiu front associated with intense rainfalls in 1–10 July 1991. J. Meteor. Soc. Japan, 78, 141–157, https://doi.org/10.2151/jmsj1965.78.2_141.
Orlanski, I., 1975: A rational subdivision of scales for atmospheric processes. Bull. Amer. Meteor. Soc., 56, 527–530.
Peng, J., S. Sun, and D. Liu, 2023: The extreme hot event along the Yangtze Basins in August 2022. J. Appl. Meteor. Sci., 34, 527–539, https://doi.org/10.11898/1001-7313.20230502.
Saha, S., and Coauthors, 2010: The NCEP Climate Forecast System Reanalysis. Bull. Amer. Meteor. Soc., 91, 1015–1058, https://doi.org/10.1175/2010BAMS3001.1.
Shen, H., G. Zhai, J. Yin, W. Zhang, and B. Zha, 2013: Feature analysis of mesoscale vortex over lower reaches of Yangtze River during meiyu period (in Chinese). Chin. J. Atmos. Sci., 37, 923–932, https://doi.org/10.3878/j.issn.1006-9895.2012.12075.
Sun, J., S. Zhao, G. Xu, and Q. Meng, 2010: Study on a mesoscale convective vortex causing heavy rainfall during the mei-yu season in 2003. Adv. Atmos. Sci., 27, 1193–1209, https://doi.org/10.1007/s00376-009-9156-6.
Tao, S.-Y., 1980: Rainstorms in China. Science Press, 225 pp.
Wang, B., and Z. Fan, 1999: Choice of South Asian summer monsoon indices. Bull. Amer. Meteor. Soc., 80, 629–638, https://doi.org/10.1175/1520-0477(1999)080<0629:COSASM>2.0.CO;2.
Wang, H., 2002: The instability of the East Asian summer monsoon–ENSO relations. Adv. Atmos. Sci., 19 (1), 1–11, https://doi.org/10.1007/s00376-002-0029-5.
Wei, X., Y. Yu, B. Li, and Z. Liu, 2023: Representativeness of two global gridded precipitation data sets in the intensity of surface short-term precipitation over China. Remote Sens., 15, 1856, https://doi.org/10.3390/rs15071856.
Yang, Y., W. Gu, R. Zhao, and J. Liu, 2010: The statistical analysis of low vortex during meiyu season in the lower reaches of the Yangtze (in Chinese). J. Appl. Meteor. Sci., 21, 11–18.
Zhang, J., S. Fu, J. Sun, X. Shen, and Y. Zhang, 2015: A statistical and compositional study on the two types of mesoscale vortices over the Yangtze River Basin (in Chinese). Climatic Environ. Res., 20, 319–336, https://doi.org/10.3878/j.issn.1006-9585.2015.14164.
Zhang, J., S. Jin, S. Feng, H. Han, and S. Fu, 2023: A new objective identification method for mesoscale vortices: Three-dimensional tracking and quantitative evaluation (in Chinese). Chin. J. Atmos. Sci., 47, 1434–1450, https://doi.org/10.3878/j.issn.1006-9895.2111.21178.
Zhang, R., 2015: Changes in East Asian summer monsoon and summer rainfall over eastern China during recent decades. Sci. Bull., 60, 1222–1224, https://doi.org/10.1007/s11434-015-0824-x.
Zhao, S.-X., Z.-Y. Tao, J.-H. Sun, and N.-F. Bei, 2004: Study on Mechanism of Formation and Development of Heavy Rainfalls on Meiyu front in Yangtze River. China Meteorological Press, 282 pp.
Zhou, Y.-S., and L. Bai, 2010: Structural analyses of vortex causing torrential rain over the Changjiang-Huaihe River Basin during 8 and 9 July 2003 (in Chinese). Chin. J. Atmos. Sci., 34, 629–639.
