A 33-yr climatology of shear lines occurring over the Yangtze–Huai River basin (YHSLs) of eastern China during the mei-yu season (i.e., June and July) of 1981–2013 is examined using the daily ERA-Interim reanalysis data and daily rain gauge observations. Results show that (i) nearly 75% of the heavy-rainfall days (i.e., >50 mm day−1) are accompanied by YHSLs, (ii) about 66% of YHSLs can produce heavy rainfall over the Yangtze–Huai River basin, and (iii) YHSL-related heavy rainfall occurs frequently in the south-central basin. The statistical properties of YHSLs are investigated by classifying them into warm, cold, quasi-stationary, and vortex types based on their distinct flow and thermal patterns as well as orientations and movements. Although the warm-type rainfall intensity is the weakest among the four, it has the highest number of heavy-rainfall days, making it the largest contributor (33%) to the total mei-yu rainfall amounts associated with YHSLs. By comparison, the quasi-stationary type has the smallest number of heavy-rainfall days, contributing about 19% to the total rainfall, whereas the vortex type is the more frequent extreme-rain producer (i.e., >100 mm day−1). The four types of YHSLs are closely related to various synoptic-scale low-to-midtropospheric disturbances—such as the southwest vortex, low-level jets, and midlatitude traveling perturbations that interact with mei-yu fronts over the basin and a subtropical high to the south—that provide favorable lifting and the needed moisture supply for heavy-rainfall production. The results have important implications for the operational rainfall forecasts associated with YHSLs through analog pattern recognition.
Every year from mid-June to mid-July, a steady and quasi-stationary rain belt usually is maintained over the middle and lower reaches of the Yangtze River basin in eastern China, the so-called mei-yu front (Chen and Chang 1980; Tao 1980; Ding 1992). The mei-yu front often produces heavy rainfall and floods, which are sometimes the cause for severe environmental disasters and economic losses around the Yangtze–Huai River basin (YHRB; 28°–34°N, 110°–122°E; see Fig. 1). Thus, considerable attention has been paid to studying multiscale processes involved in heavy-rain-producing mei-yu frontal systems during the past decades (Tao and Ni 2001; Ding et al. 2007; Zheng et al. 2007; Liu et al. 2008; Sampe and Xie 2010; Gao et al. 2013). However, our understanding of these processes is far from complete, especially for the roles of shear lines in producing heavy rainfall in the mei-yu frontal systems, which are typically distributed over the YHRB during the mei-yu season (i.e., June and July), also called the YHR shear lines (the YHSLs; Zhu et al. 2007).
According to the Glossary of Meteorology (American Meteorological Society 2012), a shear line is “a line or narrow zone across which is an abrupt change in the horizontal wind component parallel to this line; a line of maximum horizontal wind shear.” While this definition is similar to that used in China, shear lines during the mei-yu season over the YHRB tend to occur in the lower troposphere, that is, they are more pronounced at 850 hPa (Zhu et al. 2007). Because of varying intensities between cold and warm airflows across a shear line, the YHSLs have been previously classified into three types, that is, warm, cold, and quasi-stationary types, in the study of their associated rainfall characteristics (Zhu et al. 2007).
Previous studies have shown that shear-line systems (Wang et al. 2000), low pressure system or mesoscale vortices (Sun and Zhang 2012), low-level jets (LLJs; Chen et al. 2005), mei-yu fronts (Luo et al. 2014), and midlevel troughs (Liu et al. 2008) could provide different favorable conditions, such as quasigeostrophic ascent, ample moisture supply, and thermal advection, for heavy-rainfall production in the mei-yu system (Gao et al. 2003; Chen and Yu 1988). Specifically, given the relatively lower pressure along a shear line, mass and moisture convergence in the planetary boundary layer (PBL) and associated upward motion would facilitate the lifting of unstable air parcels in the PBL leading to convective initiation and then the development of heavy rainfall in the presence of ample moisture supply (Zhu et al. 2007; Hu et al. 2008; Li and Zhang 2014). In some extreme cases, they may lead to 5–7 days of persistent heavy rainfall (Tao 1980). Thus, the YHSLs are closely related to the generation of heavy-rainfall events during the mei-yu season. They are not only one of the major mesoscale weather systems in China, but they are also one of the important members of the East Asian summer monsoon (EASM) system (Chen et al. 1991; He et al. 2008).
Based on the statistics of the Anhui Meteorological Observatory (1976), about 66% of the shear lines in Anhui Province (see Fig. 1 for its location) during June–July from 1961 to 1972 produced heavy rainfall. On the other hand, Zhu et al. (2007) indicated that about 90% of the YHSLs could produce heavy rainfall in the YHRB. Some limitations with the two different studies should be mentioned. First, the spatiotemporal span of the samples by the first study is limited to Anhui Province, which is a portion of the YHRB, with a too-short period of 1961–72. Second, the statistics from the above two studies are out of date as a result of recent climate changes over the YHRB, especially after the 1980s (Hu et al. 2008; Li et al. 2010). In fact, after analyzing the spatiotemporal characteristics of the YHSLs and the associated rainfall under the global warming scenarios during the months of June and July in 1981–2013, Ma and Yao (2015) found that (i) nearly two-thirds of the YHSLs produced heavy rainfall, and (ii) nearly three-quarters of the heavy-rainfall events were caused by the YHSLs. Because the YHSL-associated heavy-rainfall events occur mainly from the fourth pentad of June to the second pentad of July, these events coincide with the most active period of heavy mei-yu rainfall. This implies that most of the mei-yu season over the YHRB is accompanied with shear-line activity.
Previous studies have shown the general relationship between shear lines and heavy rainfall. However, few studies have examined how heavy rainfall occurs under the different types of YHSLs. Moreover, most of them are limited to case studies (e.g., Wang 1963; Kato 1987; Larson 1976; Hu and Peng 1996; Hu 1997; X. Li et al. 2014), which lack statistical understanding of the YHSLs and the associated heavy-rainfall production. Thus, the purposes of this study are to (i) examine the occurrence frequency of different types of the YHSLs and the associated rainfall coverage and intensities during the mei-yu season using the 33-yr (i.e., 1981–2013) ERA-Interim reanalysis data and rain gauge observations, (ii) explore the quantitative relationship between different types of the YHSLs and heavy rainfall and their statistical characteristics, and (iii) investigate the characteristics of larger-scale circulations, in which the four types of YHSLs are embedded.
The next section describes the data and method used for this study. Section 3 shows the classification of YHSLs into four different types, based on their distinct flow and thermal patterns, and then describes a 33-yr climatology of the four types of YHSLs in relation to heavy-rainfall production over the YHRB. Section 4 shows the synoptic conditions in the lower -to-midtroposphere, in which the four types of YHSLs are embedded. Some common and different large-scale features among them are discussed. A summary and concluding remarks are given in the final section.
2. Data and method
The data used in this study include (i) the ERA-Interim reanalysis from the European Centre for Medium-Range Weather Forecasts (ECMWF) during the months of June and July from 1981 to 2013, with a 6-hourly temporal resolution and a 1° × 1° horizontal spatial resolution; and (ii) the daily rain gauge observations with 414 surface stations distributed over the YHRB. To be consistent with the ERA-Interim analysis, the daily rainfall amounts at individual stations are interpolated with the Cressman objective analysis technique to the YHRB domain at 1° × 1° horizontal spatial resolution.
Based on the American Meteorological Society (2012) shear-line definition and previous studies of the YHSLs (i.e., Qiao and Tan 1984), two objective criteria with the following variables, that is, the vertical relative vorticity (vor) consisting mostly of the meridional shear of zonal wind u, and null zonal wind speed along the shear line, are used to objectively identify the YHSLs (Ma and Yao 2015), that is,
where y is the north–south coordinate. The x component of the horizontal winds is used here because the YHSLs, along which the cyclonic shear vorticity is maximized, tend to be laid much more zonally than longitudinally. In addition to the above two criteria, it is required that the YHSLs have at least a 330 km zonal span within the study domain (see Fig. 1). Note that in some cases, a YHSL may be regarded as the dynamical representation of a mei-yu frontal system in the lower troposphere. However, there are some differences between the two systems. The former is a discontinuous line in horizontal winds with pronounced cyclonic shear, as defined above, whereas the latter emphasizes more its “frontal” characteristics with discontinuity in humidity. In addition, the former is typically much smaller in scale than the latter. See the next two sections for a more detailed description of the YHSL characteristics. After objectively identifying all the YHSLs during 1981–2013, they are subjectively classified into several relevant types, as will be described in detail in section 3b.
A heavy-rainfall day over the YHRB is defined as daily precipitation exceeding 50 mm at five or more surface stations in accordance with the standard of the China Meteorological Administration. Similarly, an “extreme heavy” rainfall day is derived when daily precipitation exceeds 100 mm. These definitions are similar to those used in the United States, that is, daily 2 in. (50.8 mm) and 4 in. (101.6 mm) for heavy and “extreme heavy” rainfall days, respectively (Karl et al. 1996). A YHSL day is found when a YHSL is present at least once in the four times a day (i.e., 0000, 0600, 1200, 1800 UTC) analyses. Then, a YHSL heavy-rainfall day is determined when the above two conditions are both met, plus the occurrence of heavy or extreme heavy rainfall around the YHSL. Figure 2 illustrates examples of the above definitions for a heavy-rainfall day but without a YHSL (nearby) on 23 June 1994 (Fig. 2a), a YHSL day without heavy rainfall (i.e., only <50 mm day−1) on 11 June 1994 (Fig. 2b), and a YHSL day with heavy rainfall (i.e., >50 mm day−1) on 9 June 1994 (Fig. 2c).
3. Statistical relationship between YHSLs and heavy rainfall
a. Statistical relationship between YHSLs and heavy-rainfall days
Table 1 shows the statistics of YHSL days, YHR heavy-rainfall days, and YHSL heavy-rainfall days during June and July of 1981–2013, which are on average 30.2, 33.2, and 22.0 days (out of 61 days) each year, accounting for 54%, 49.5%, and 36.1% of the total (2 month) days, respectively. This implies that the number of the YHSL heavy-rainfall days accounts for 72.8% of the heavy-rainfall days and 66.3% of the YHSL days. This result indicates the pronounced correlation between the YHSLs and heavy rainfall and provides significant evidence for studying the relationship between them. In particular, since Zhu et al. (2007) have classified the YHSLs into three different types, that is, warm, cold, and quasi-stationary, it is of great interest to examine their associated characteristics in section 3b and then explore to what extent each type of the YHSLs accounts for the generation of heavy rainfall during the mei-yu season over the YHRB in section 3c.
b. Classification of YHSLs and their spatial characteristics
In addition to the abovementioned three types of YHSLs, another type of YHSL was later found that has the characteristics of both warm and cold types of shear lines. It is usually accompanied by a mesoscale vortex with a closed cyclonic circulation, and so it is defined herein as a vortex type. This type of vortices is generated, intensified, and moved together with the associated shear lines (Long and Cheng 2004). After including such a vortex type, we will examine herein the four types of YHSLs, whose flow characteristics (mainly at 850 hPa) are described as follows. Their associated larger-scale flow patterns are presented in section 4.
A warm-type YHSL appears between a southwesterly wind to the south and a southeasterly wind to the north, with a warm tongue centered roughly along the YHSL (Fig. 3a). It is mostly oriented in the near-west–east direction, sometimes in the west-southwest–east-northeast direction, as will be seen in section 4. The warm-type YHSL is so defined because it tends to move northward into a relatively colder air mass, like a warm front. More importantly, warm advection prevails in both the southwesterly and the southeasterly flows. Heavy and extreme rainfall occurs usually in the warm and moist southwesterly flows with the orientations being nearly parallel to the YHSL. In fact, Fig. 3a shows clearly a demarcation of rainfall and rainfall-free areas across the YHSL, except along the western boundary of the YHRB where light (<50 mm day−1) rainfall takes place likely as a result of topographical lifting by the sloping surface of the Qinling mountain range (see its location in Fig. 1). Its topographical effects can also be reflected by the associated large temperature anomalies near the northwestern border of the YHRB (Figs. 3a–d).
In contrast, a cold-type YHSL develops as a warm and moist southwesterly wind to the south meets a relatively colder and drier northeasterly wind to the north, with a cold tongue centered roughly along the YHSL, and it is mostly oriented in the northeast–southwest direction (Fig. 3b). The cold type is so defined because it tends to move more south- and southeastward into a warmer air mass, like a cold front. Indeed, this movement is determined by pronounced cold advection in the northeasterly flow. Note the three major differences between the cold and the warm type of YHSLs: significant cold advection in a northeasterly wind with a south- to southeastward displacement in the former versus significant warm advection in a southeasterly wind with a northward displacement in the latter. Heavy and extreme rainfall in the cold type occurs mostly in the warm and moist southwesterly flow, like that in the warm type.
A quasi-stationary YHSL exists between easterly winds with a weak cold tongue to the north and westerly winds with a pronounced warm tongue to the south, which appears often in the near-west–east direction (Fig. 3c). Its flow and rainfall structures are nearly steady and quasi-stationary, and so named as the quasi-stationary type. In this type, heavy and extreme rainfall is typically distributed on the southern side of the YHSL. The “frontal” type of lifting in the southern half portion of the cold tongue appears to account for the generation of lighter rainfall on the northern side of the YHSL but over the eastern two-thirds portion of the YHRB, while the orographic lifting is responsible for lighter rainfall in the western third of the YHRB.
A vortex-type YHSL is seen as a closed circulation that is more evident at 850 hPa in its western (or southwestern) and eastern (or northeastern) semicircle resembling the cold and warm type of shear-line flow, respectively (Fig. 3d). Such a closed circulation is shallow and typically could not be seen above 700 hPa, as will be shown in section 4. Heavy and extreme rainfall occurs mostly in the warm and moist southwesterly flow in the southeastern semicircle, namely, with a comma-shaped pattern. This type of YHSL moves eastward with the vortex along the mei-yu front.
c. Frequencies of the YHSL-associated heavy-rainfall days
Given the above classification criteria, the frequencies (i.e., numbers) of the four types of YHSL heavy-rainfall days during June and July of 1981–2013 can be analyzed. Figure 4 shows that the warm type is of the highest frequency with 276 days (out of 1403 days in 1981–2013), accounting for 38.1% of the total YHSL heavy-rainfall days. It is followed by the cold type and the vortex type, accounting for 21.8% and 21.3%, respectively, with a YHSL-heavy-rainfall frequency of 158 and 154 days. The quasi-stationary type has the lowest frequency of 136 days, accounting for 18.8% of the total YHSL-heavy-rainfall days. On average, we may consider about 40% of the YHSL-heavy-rainfall days to be warm-type YHSLs, with the remaining three types contributing 20% each. The highest frequency of the warm type is more associated with the dominant EASM flows during June and July, in which the persistent warm and moist southwesterly airflow prevails.
d. Spatial distribution of the YHSL-associated rainfall
Figure 5 shows the spatial distribution of the annual-averaged rainfall amounts associated with the YHSLs during June and July of 1981–2013. We see that the maximum annual rainfall amounts exceeding 350 mm take place at the borders of Anhui, Jiangxi, and Hubei Provinces, exhibiting the pattern of ample rainfall in the south-central region but scant rainfall in the northwest region of the YHRB. This rainfall pattern is attributable to the fact that the YHSL-associated rainfall occurs mainly on the south of the shear line, that is, in the southwesterly–westerly flows.
The spatial distributions of the annual-averaged rainfall amounts associated with the four types of YHSLs are given in Fig. 6, showing different patterns and intensities among them. Specifically, the warm-type rainfall pattern is southeast–northwest oriented, with ample amounts in the south-central region of the YHRB (i.e., peaked in northern Jiangxi Province) but scant amounts in the northeastern region (Fig. 6a). In contrast, the cold-type rainfall pattern is southwest–northeast oriented, and concentrated mainly in the eastern half region, but with scant amounts in the northwestern region (Fig. 6b). The quasi-stationary-type rainfall pattern is west-southwest–east-northeast-oriented and mainly concentrated in the southern region (i.e., peaked in southern Jiangxi Province), but with scant amounts in the northern region of the YHRB (Fig. 6c). The vortex-type rainfall pattern is more circular and mainly concentrated in the central region, with scant amounts in the northwestern region (Fig. 6d).
It is evident by comparing Figs. 5 and 6 that the spatial distribution of the annual total rainfall of all the YHSLs resembles more that of the vortex type of YHSLs. This is due likely to the eastward propagation of the associated vortices along the mei-yu fronts that pass through the central to south-central region.
e. Rainfall intensity and relative contributions of individual types of YHSLs
To gain insight into the relative rainfall intensity of the four types of YHSLs, Fig. 7 compares the daily averaged numbers of stations with heavy-rainfall rates, that is, ranging between 50 and 100 mm day−1, and extreme heavy-rainfall rates, that is, exceeding 100 mm day−1, associated with the four types of YHSLs. We see first that the numbers of stations with the heavy-rainfall rates are much more than those with the extreme heavy-rainfall rates. Among the four types of YHSLs, the vortex type has the largest number of stations recording heavy-rainfall amount (32.2) and extreme heavy-rainfall amount (7.8). The vortex type tends to be an extreme heavy-rain producer because of the more organized convergence toward the lower pressure center in the PBL. This result explains to a certain extent why the annual rainfall pattern of the vortex type resembles closely that of all the YHSLs (cf. Figs. 6d and 5), as mentioned before. It is followed by the cold type, with daily averaged heavy and extreme heavy-rainfall stations of 23.6 and 5.8, respectively, and then the quasi-stationary type with 22.2 and 5.4 rain gauge stations, respectively. The least is the warm type, with 18.3 stations and 4.6 stations, respectively.
The above analysis indicates that the warm type of heavy rainfall occurs at the least number of rain gauge stations, but with more days and more widespread extreme heavy rainfall than all the other three types (cf. Figs. 7, 6a, and 4). By comparison, the vortex type occurs less frequently (Fig. 4), but with the largest number of stations recording heavy and extreme heavy rainfall (Fig. 7). Our calculations show the respective contributions of 33%, 25%, 17%, and 25% by the warm, cold, quasi-stationary, and vortex type of YHSLs to the total rainfall of all the YHSLs. Apparently, these relative contributions incorporate both the YHSL heavy-rainfall days and the station numbers recording heavy and extreme heavy rainfall. In particular, the two extreme contributions of the warm type and the quasi-stationary type could be clearly seen from Figs. 6a and 6c, showing that the warm type of YHSLs contributes more ample rainfall to the total rainfall amounts over the YHRB, while the quasi-stationary type contributes the least in terms of both the YHRB-averaged and the maximum amounts. This difference can be attributed partly to the different heavy-rainfall days (Fig. 4), and partly to the different numbers of heavy-rainfall recording stations between the two types of YHSLs (Fig. 7). We may hypothesize that the four different types of YHSLs and their distinct flow patterns and relative rainfall contributions should be more or less determined by larger-scale flows/forcing and moisture supply, as will be discussed in the next section.
Note the more widespread heavy rainfall with the largest contribution to the total (heavy) rainfall amount by the warm type of YHSLs and the more frequent occurrences of extreme heavy rainfall (i.e., 100 mm day−1) by the vortex type. These two results have important implications for operational weather forecasters on the coverage and amount of rainfall that should be expected after seeing the predicted pattern of a YHSL, especially for the heavy-rain-producing YHSL of the vortex type.
4. Synoptic circulation characteristics of the YHSLs
To verify our previous hypothesis that each type of YHSL with distinct flow patterns should be closely related to its associated larger-scale circulations, in the next section, we examine the composite synoptic circulation characteristics associated with the four types of heavy-rain-producing YHSLs. For each type, the composite analysis is performed by selecting 10 analogous heavy-rain-producing YHSL cases, in which at least 10 heavy-rainfall recorded stations should take place in the vicinity of each YHSL. In addition, it requires that the distribution of the 10 analogous YHSLs be as close as possible to each other to minimize any undesirable effects due to the averaging of mesoscale perturbations. Table 2 lists 40 YHSL cases that are used to plot the synoptic-scale geopotential height, horizontal winds, and precipitable water fields associated with the four types of YHSLs at 850 and 500 hPa, which are given in Figs. 8 and 9, respectively. The upper-level maps are not shown herein, since the YHSLs are basically low-tropospheric phenomena, with some distinct features in the midtroposphere, and since little differences in their large-scale flows among the four types could be seen in the upper troposphere.
a. The warm type of synoptic circulation
At 850 hPa (Fig. 8a), the warm-type YHSL exhibits as a west–east-oriented trough axis extending from a southwest vortex (Tao 1980; Ding 1992; J. Li et al. 2014) east-northeastward across the YHRB. In particular, this type of YHSL is dominated by a warm and moist southwesterly LLJ of >10 m s−1 associated with the western Pacific subtropical high (WPSH) to the south and a weak southeasterly flow to the north that is sandwiched between the southwest vortex on the east side of the Tibet Plateau and a weak height ridge downstream. Clearly, the cyclonic vorticity intensity and displacement of this type YHSL are determined more by the southwest vortex and the WPSH. In addition, precipitable water of more than 50 mm is distributed over the YHRB, with much less moisture content on its north, which is one of the important characteristics during the mei-yu season, as can also be seen from Figs. 8b–8d. Given the low-tropospheric north–south temperature gradients, albeit small during the months of June and July, mesoscale (i.e., >1000 km) isentropic lifting of the southwesterly warm and moist air associated with the EASM (Raymond and Jiang 1990; Zhang and Zhang 2012) provides the needed lifting and moisture supply for the generation of heavy rainfall in the southern portion than that in the northern portion of the YHRB.
At 500 hPa (Fig. 9a), we see the presence of a weak trough on the west side of the YHRB that is closely associated with the southwest vortex shown at 850 hPa. Ahead of the trough axis are more intense southwesterly flows in the southern portion and relatively weaker westerly flows in the northern portion. The former are consistent with the more westward extension of the WPSH, as often defined by the 588-dagpm height contour, rather than the climate mean. This deep layer of cyclonic flows on the upstream with the more intense southwesterly flows of the WPSH helps maintain the warm-type YHSL in the YHRB.
b. The cold type of synoptic circulation
The cold-type YHSL at 850 hPa is essentially the northeastward-tilted trough axis associated with a midlatitude traveling disturbance, with the warm and moist advection by a southwesterly LLJ of 10–12 m s−1 ahead of and north-to-northeasterly cold and dry advection behind the axis (Fig. 8b). Unlike a common midlatitude trough, this type YHSL could span a long distance, namely, extending from a vortex in the Yellow Sea—that is, centered at 35°N, 125°E—to the foothills of the Tibet Plateau. Because of the presence of cold advection to the northwest of the YHSL, a surface cold front may be expected, which acts to move the YHSL southeastward or southward, again the so-called cold type. Heavy rainfall tends to take place near the surface front, in contrast to that which occurs more in the warm sector (i.e., ahead of a surface front) in the warm-type YHSL (cf. Figs. 8a,b).
The cold-type YHSL is no longer evident at 500 hPa (Fig. 9b), as the curvature of the synoptic trough decreases rapidly with height. Although the YHRB is located at the trough bottom, the associated quasigeostrophic forcing could still play some roles in lifting the warm and moist air ahead of the surface front. Moreover, the more pronounced northwesterly flow behind the trough axis and the more eastward-distributed WPSH would allow the cold-type YHSL to be well maintained as it moves eastward across the YHRB.
c. The quasi-stationary type of synoptic circulation
The quasi-stationary type YHSL at 850 hPa is oriented along the west–east direction between warm and moist southwesterly to westerly flows at the northern edge of the WPSH and relatively cold and dry easterly flows at the bottom of a cold high pressure system at higher latitudes (Fig. 8c). Like the cold-type YHSL, the typical mei-yu frontal characteristics, that is, with large contrasts in the precipitable water, are evident for this quasi-stationary-type YHSL. A zonally distributed lower (i.e., than 144 dagpm) height (pressure) zone must coincide with the YHSL, which would facilitate cross-isobaric mass convergence in the PBL. Such a lower-pressure zone tends to be much weaker than the other three types of YHSLs. On the downstream of this high is a deep-layer near-upright and symmetric trough system. This YHSL could become quasi-stationary because of the slow variation of the high/trough system to north and the stationary nature of the WPSH. Because of the reduced moisture supply at the northern edge of monsoonal airflow and the weaker lower pressure zone (i.e., implying weaker convergence in the PBL) along the YHSL, less heavy rainfall than that in the other types of YHSLs could be generated over the YHRB. Moreover, little rainfall could be produced in the anticyclonic cold flows to the north of the YHSL. In contrast, extreme heavy rainfall may take place to the south of the region, as indicated by the presence of a small portion of the heavy-rainfall area in the southern YHRB (see Figs. 6c and 8c), where ample moisture supply by a southwesterly LLJ is available.
Because of the rapid weakening of the cold high pressure system with height, at 500 hPa the YHRB is dominated by westerly to southwesterly flows in the southern portion and northwesterly flows in the northern portion (Fig. 9c). The more westerly flows on the south of the YHRB could be attributed to the more eastward distribution of the WPSH than that associated with the warm and cold types of YHSLs (cf. Figs. 9a–c), as indicated by the 588-dagpm contour near 135°E that is more than 20° of longitude to the east of that in the warm type (Figs. 9a,c). This deep, near-steady westerly flow facilitates the setup of a quasi-stationary-type YHSL when a high pressure system approaches from higher latitudes.
d. The vortex type of synoptic circulation
For the vortex type, the YHRB is centered with a cyclonic vortex at 850 hPa (Fig. 8d), with a warm- and a cold-type YHSL extending eastward and southwestward from the center, respectively. In some cases, this vortex originates from a southwest vortex as it propagates eastward along the mei-yu front into the YHRB. The vortex-type YHSL appears to be longer lived than the other three types due partly to the conservation of cyclonic vorticity above the PBL in the presence of weak vertical shear and horizontal deformation and partly to the continuous latent heating induced convergence in the lower troposphere and the resulting amplification of cyclonic vorticity (Zhang and Fritsch 1987, 1988; Bartels and Maddox 1991). In general, mesovortices of this kind tend to produce heavy rainfall, given enough moisture supply (Zhang and Fritsch 1987, 1988). In fact, a southwesterly LLJ, albeit the weakest among the four types of YHSLs (cf. Figs. 8a–d), is always present to the south of the vortex. Despite its weakness, the LLJ plays an important role in supplying the needed moisture content for the generation of heavy rainfall in the southwesterly flow regions, that is, ahead of the cold type of YHSL and behind the warm type of YHSL, being more localized near the vortex center.
The lower-level closed cyclonic circulation becomes a trough at 500 hPa, as the larger-scale westerly mean flow increases with height but still has pronounced cyclonic shear vorticity over the YHRB (Fig. 9d). As compared with the other three types of synoptic circulation, the WPSH has the least influence on the development of the vortex-type YHSL, as indicated by the far eastward location of its 588-dagpm contour, that is, much farther eastward from 140°E. Otherwise, the vortex and rainfall intensities could be further strengthened with increased moisture supply by the southwesterly LLJ.
e. Common and different features of the synoptic circulations among the four types of YHSLs
It is evident from the above analysis that the four types of YHSLs have some common and different configurations of synoptic circulation. Some common features include a lower height (pressure) zone along the YHSLs favoring mass and moisture convergence in the PBL and a southwesterly LLJ accounting for moisture supply in the lower troposphere, the presence of the WPSH to the south from the low-to-midtroposphere that influences the intensity of the southwesterly LLJ, and midlevel westerly traveling troughs providing lower pressure for the maintenance of YHSLs and favorable quasigeostrophic forcing for preconditioning a heavy-rain-producing environment. Major differences among the synoptic circulations of the four types of YHSLs occur mainly in the lower troposphere (e.g., 850 hPa), such as the presence of a southwest vortex or just a mesoscale vortex, a high pressure system to the north, and a southwestward-tilted tough. Some unique features of some types of YHSLs are a closed cyclonic circulation of the vortex type that is often visible up to 500 hPa over the YHRB; a pronounced southwestward-extended trough associated with the cold type, with an intense northerly component over the YHRB; and a slow-moving high pressure system providing an easterly flow in the north of the YHRB for maintaining the quasi-stationary type. The dominant warm type of YHSLs, that is, 38.1% (Fig. 4), could be attributed to the important influences of the warm EASM with the southward invasion of fewer higher-latitude disturbances during the mei-yu season. Our analysis also indicates that some YHSLs could be converted from one type to another, for example, from a warm type to a vortex type when the southwest vortex moves into the YHRB (cf. Figs. 8a,d), or from a cold type to a vortex type when the bottom trough amplifies to a vortex (cf. Figs. 8b,d).
In addition, weak thermal advection could be seen across the YHSLs (Fig. 3), although warm advection of various intensities tends to prevail in the southwesterly flow. This indicates that the YHSL-related heavy rainfall is much less dynamically driven (i.e., by vertical differential vorticity advection and thermal advection), but it is driven more by convergence associated with the directional shear of the horizontal winds (i.e., lower pressure) in the lower troposphere with the needed moisture supply from the southwesterly monsoonal LLJ.
5. Summary and concluding remarks
In this study, a 33-yr climatology of the YHSLs and heavy-rainfall days over the YHRB during the mei-yu season of 1981–2013 is examined using the ERA-Interim reanalysis and daily rain gauge observations. To help understand their statistical characteristics and processes involved in heavy-rainfall production, the YHSLs are classified into warm, cold, quasi-stationary, and vortex type, and their mesoscale and synoptic circulation characteristics are analyzed. Major results may be summarized as follows:
During the mei-yu season when the East Asian summer monsoon is active, the YHSL is one of the major heavy-rain-producing weather systems. Results show that about 66% of the YHSLs can produce heavy rainfall, and nearly 75% of the heavy-rainfall days are accompanied by the YHSLs. The YHSL-related heavy rainfall presents a pattern of “plentiful amount in the south-central basin and scant amounts in the northwestern basin,” with the maximum rainfall occurring near the borders of Anhui, Jiangxi, and Hubei Provinces.
The four types of YHSLs presented do show distinct flow and thermal (e.g., warm vs cold tongues) patterns among them, with some differences in orientation and displacement. The associated thermal advection and the common moisture source from the westerly to southwesterly flows play an important role in determining the distribution and amplitude of rainfall over the YHRB.
The warm-type-YHSL-related rainfall intensity is the weakest among the four types, but it is the largest contributor to the total mei-yu rainfall amounts associated with the YHSLs due to its large frequency of occurrences. The frequency of quasi-stationary type is the lowest, contributing the smallest to the total rainfall, whereas the vortex type is the more frequent extreme heavy-rain producer.
A composite analysis of the large-scale flows also shows some common and distinct synoptic circulation associated with the four types of YHSLs. The common larger-scale features include a lower height (pressure) zone along the YHSLs, a southwesterly LLJ, the WPSH to the south from the low to midtroposphere, and midlevel westerly traveling disturbances. Major differences occur mainly in the lower troposphere, such as the presence of a mesoscale vortex, a high pressure system to the north, and a southwestward-tilted tough.
Note that the classification of the four types of YHSLs and the associated rainfall distribution and amounts presented herein provide a basis for the operational forecast use of analogs with the assumption that two similar flow patterns will likely produce similar rainfall distributions and amounts (Van Den Dool 1994; Roebber and Bosart 1998; Ren et al. 2020). In this regard, the above results have important implications for operational daily rainfall forecasts on what coverage and how heavy an amount of rainfall should be expected after seeing the predicted pattern of a YHSL, especially for the heavy-rain-producing YHSL of the vortex type. We mention also that this study is limited to the mature stage of the YHSLs. Thus, in a forthcoming journal article, we will examine the evolutionary processes associated with the four types of heavy-rain-producing YHSLs.
This study was supported by the National Natural Science Foundation of China (91937301, 41775048, 41775050, and 91637105), the Second Tibetan Plateau Scientific Expedition and Research (STEP) program (2019QZKK0105), and the National Key R&D Program of China (2018YFC1507804).