1. Introduction
The cutoff low (CL) is an important midlatitude weather system that has drawn the attention of researchers since the 1940s. The CL manifests as a closed cyclonic circulation in the middle and upper troposphere, detached completely from the main westerly stream (Palmén 1949; Palmén and Nagler 1949). Having a polar origin, it is characterized by a local cold core and, thus, a low tropopause, and is also known as a cold vortex (Hsieh 1949). Hoskins et al. (1985) presented a definition of the CL as a closed region of high potential vorticity (PV) on an isentropic surface. Regionally, CLs not only cause low temperatures, but sometimes also cause large amounts of rainfall (Tao 1980; Nieto et al. 2007). The CLs play an important role in the stratosphere–troposphere exchange of chemical substances, such as ozone and hydrocarbons (Bamber et al. 1984; Liu et al. 2013).
Nieto et al. (2005) described three distinct preferred occurrence areas of CLs in the Northern Hemisphere: southern Europe and the eastern Atlantic, the northeast Pacific, and the northern China–Siberia region. Climatologically, CLs occur over the three distinct preferred areas around the major troughs of the large-scale circumpolar flow, which are distinguished from the synoptically occurred CLs in any other region. The northeast China cold vortex (NCCV) prevails from May to August (Sun et al. 1994). Since it occurs most frequently in strongest amplitude during May to mid-June, this period was regarded as the active period of the NCCV (Xie and Bueh 2012). The NCCV can bring persistent cool weather, even leading to a considerable reduction of local crop production during frequent NCCV years (Zheng et al. 1992).
Since NCCVs may cause destructive local weather conditions, a number of studies have investigated their typical circulation features. NCCVs are closely associated with the East Asian jet (EAJ; Bell and Bosart 1989). Specifically, recurrent or persistent NCCVs tend to occur when the EAJ exhibits strong bifurcation or split features (Sun 1997; Sun et al. 2000). The blocking-type circulation is also characterized by a split jet stream (Rex 1950), suggesting that it may play an important role in maintaining an NCCV. Zheng et al. (1992) showed that the formation of the NCCV circulation is associated with a ridge over Lake Baikal or a block over the Yakutsk area. Sun et al. (1994) further noted that approximately 77% of NCCVs are associated with a blocking-type circulation over northeast Asia. In particular, in the years of highly recurrent NCCVs, anomalous circulation centers with a “plus–minus–plus” distribution in the midtroposphere were observed over the region of Lake Balkhash to Lake Baikal, northeast China, and the Yakutsk–Okhotsk region, respectively. Wang and Lupo (2009) proposed that the block over the Sea of Okhotsk induces an intensification of NCCV circulation through Rossby wave propagation in early summer. Moreover, Wang et al. (2011) noted that the maintenance of the block over the Sea of Okhotsk that is in phase with a Rossby wave excited by Tibetan Plateau heating favors the maintenance of NCCV circulation. Lian et al. (2010) found that the persisting blocking-type circulation over the Ural Mountains is also favorable for persistent NCCVs. Wu et al. (2009) investigated the recurrent and persistent NCCVs during May 2008, during which NCCVs prevailed for 28 of 31 days with 10 sequences. They concluded that a persistent ridge anchored between the Ural Mountains and Lake Baikal contributed to the recurrent NCCV circulation. Another extreme case of NCCVs occurred in June 2009, during which the NCCV circulation recurrently appeared for 27 of the 30 days of the month. Ren et al. (2010) and Zhao et al. (2010) studied this case and suggested that the double blocking circulation over the Ural Mountains and the Sea of Okhotsk is the main contributor to the NCCV activity. As for the downstream background circulation, Xie and Bueh (2012) suggested that, from May to August, the monthly west Pacific teleconnection pattern (Barnston and Livezey 1987) in its negative phase (hereafter WP− for brevity) is a favorable background circulation background of the NCCVs. Although these studies are based on monthly mean circulations, they indicate that the formation of NCCV circulation is closely associated with various types of ridges or blocking-type circulations. Therefore, it is necessary to objectively classify NCCV events according to their relationship with different types of ridges or blocking-type circulations.
As previously mentioned, NCCVs exert great influence on surface air temperature (SAT) distributions and rainfall patterns. These can be expected to be diverse because of the various types of NCCV configurations. As for the weather impacts of the NCCV circulation, researchers have paid a great deal of attention to the locations of NCCV-induced rainfall. Sun et al. (1995) found that the precipitation tends to concentrate in the southeastern quadrant of the NCCV circulation, where abundant water vapor transport and strong ascending motion are induced by the anomalous southerly or southwesterly wind associated with the NCCV. Sun et al. (2002) also found that heavy rainfalls caused by NCCVs mainly occur to the east of the NCCV center. By investigating a typical case in July 2007, Wang et al. (2012) showed that heavy rainfall also takes place in the northeast of the NCCV center. By comparing three cases of NCCV circulations and their rainfall patterns in the summer of 1998, Zhao and Sun (2007) revealed that the rainfall patterns associated with the NCCVs vary with different types of NCCV-related blocks. As for the different impacts of the different types of NCCVs on the local temperature, it has not yet been studied extensively and the weather impacts caused by different types of NCCVs have not been well documented.
Some studies have attempted to categorize NCCVs into several groups according to their locations. For example, Sun et al. (1994) classified NCCVs into three groups according to their geographic locations. However, since the location of a NCCV is closely related to the EAJ, the classification of NCCVs with geographic location is considerably affected by the seasonal variation of the EAJ. When Zhang (2008) investigated NCCVs’ impact on the mesoscale convective system over northeast China, he grouped NCCVs into zonal and meridional types based only on their orientations. In addition, Zhong (2011) categorized NCCVs according to their lifetimes, dividing them into persistent and transient types. He made the point that an NCCV inclines to be persistent if a trough lies to its west, with the latter feeding cold air and positive vorticity to the former. On the other hand, a ridge located on the upstream side is unfavorable. However, Sun et al. (1994) noted that the duration of a NCCV associated with the blocking-type circulation over the Sea of Okhotsk tends to be two days longer than the NCCV average lifetime (4 days). Nonetheless, a complete picture of NCCV signatures and evolutions has not been provided in the various classification schemes mentioned above. In particular, a deeper insight is required to describe the essential differences between short-lived and long-lived NCCVs.
In the present study, we classified NCCVs into four groups based on rotated empirical orthogonal function (REOF) analysis, for the active period of NCCVs (1 May–15 June). As shown later, such a classification captures not only the distinct features of the different types of NCCVs themselves, but also illustrates their close associations with the different kinds of ridges or blocking-type circulations. In the following section, we describe the data and the method of classification. Section 3 investigates the signatures of different types of NCCV circulations. In section 4, we present the life cycles of different types of NCCVs on the intraseasonal time scale. Based on their life cycles, the distinction between the short-lived and long-lived NCCVs and the association with the blocking activities are also discussed. In section 5, the weather impacts caused by different types of NCCVs are presented. A summary and discussion are provided in the final section.
2. Data and methodology
The daily observational data used in this study came from the NCEP–NCAR 40-Year Reanalysis Project (Kalnay et al. 1996), including the geopotential height, horizontal wind, and air temperature. These data were on a grid with a horizontal resolution of 2.5° × 2.5° and 17 vertical pressure levels. We adopted a suite of high-quality gridded daily precipitation and SAT data over China archived by the China Meteorological Administration. It was based on the interpolation of observations from 2400 stations (Zhao et al. 2014). The data were in spatial resolution of 0.5° × 0.5°, spanning from 1961 to the present. The analysis period was the active period of NCCVs (1 May–15 June) over the years 1965–2011. A low-pass filter with an 8-day cutoff period was applied to depict the intraseasonal evolution of large-scale anomalies (Nakamura 1994).
NCCV events were defined if the following three conditions were satisfied in the region of 35°–55°N, 115°–140°E: first, if a minimum center (or vortex center) could be identified on the low-pass-filtered 500-hPa geopotential height (Z500) field. Specifically, a grid (i, j) was regarded as a minimum center if its daily mean Z500 (
Schematic representation of the method used to detect/identify the (a) vortex, (b) cold trough, and (c) cold core at 500 hPa at a grid point (i, j). The circles in (a) and (c) denote the closed contours and the curve in (b) denotes the cold trough.
Citation: Monthly Weather Review 143, 3; 10.1175/MWR-D-14-00192.1
Daily blocking events were identified using a modified blocking index of Schwierz et al. (2004) based on the vertically averaged PV anomalies from 500 to 150 hPa (Small et al. 2014). The index considers blocking events as large-scale quasi-three-dimensional patterns with lifetime exceeding 5 days. The vertically averaged PV anomaly (mAPV) is designed to identify the most physically relevant characteristic of a blocking anticyclone: a negative PV underneath a raised tropopause. The anomaly is defined relative to the zonal mean PV on each level, which ensures that the selected features are near the tropopause and not in the stratosphere. We identified every closed mAPV anomaly exceeding −1.0 PV unit (PVU; 1 PVU = 10−6 K kg−1 m2 s−1) across the Northern Hemisphere between 40° and 75°N and drew its bounding rectangle. Any rectangle smaller than 15° × 15° and 106 km2 was discarded. Then a blocking event is identified if the rectangles overlap in space by at least 15° × 15° for 5 consecutive days. This index identifies not only blocks with reversals of the PV gradient but also blocking anticyclones from highly amplified ridges. In addition, this index is more effective at detecting blocking events during summer as the jet stream tends to be confined to the upper troposphere.
We applied the Student’s t test (Wilks 1995) to assess the statistical significance of the linear regression coefficient and composite anomalies of Z500, PV, and SAT. Since precipitation is not normally distributed, the bootstrap approach (Wilks 1995), a nonparametric test, was adopted to assess the statistical significance of composite precipitation anomalies. In the following section, we classified the identified NCCVs based on the REOF analysis, which has been widely used to study low-frequency variability (Wallace and Gutzler 1981; Horel 1984; Barnston and Livezey 1987).
3. Classification of NCCV event
a. Classification
Taking into account the close relationship between NCCVs and the circulation over Eurasia, we adopted the REOF analysis to determine the major modes of the NCCV. The data for the REOF analysis were constructed from the Z500 anomalies on the peak days of the 103 NCCV events, over the region covering 35°–75°N, 40°–160°E. The anomalies were area weighted by the square root of the cosine of the latitude. A traditional EOF analysis was performed, based on the covariance matrix of daily Z500 anomalies. Since the physical modes were not necessarily orthogonal (Horel 1981; Simmons et al. 1983), we rotated the leading EOF patterns to avoid difficulties in interpreting the obtained EOF spatial patterns. However, when the number of EOFs to be rotated increases, the leading patterns of REOF lose structure and become localized (Hannachi et al. 2006). To obtain suitable spatial patterns, we rotated different numbers of leading EOFs using Kaiser row normalization and the varimax criterion (Kaiser 1958). A total of 10 leading EOFs, accounting for 78.7% of total variance, were found to be appropriate for the rotation. The obtained REOF patterns were ordered according to the variances of the corresponding rotated principal components (RPCs).
Figure 2 shows daily Z500 anomalies at the peak days of NCCV events regressed against four leading normalized RPCs. As expected, the leading four modes show four distinct configurations. The first pattern is characterized by an apparent positive anomaly center over the Ural Mountains, and centers with the opposite sign over Lake Baikal and the Atlantic/west European sector (Fig. 2a). Similarly, the “minus–plus–minus” height anomalies for the REOF2 pattern appear slightly eastward with respect to that of REOF1 pattern, with a conspicuous positive anomaly center over the Yenisei River valley (Fig. 2b). In sharp contrast to the two leading patterns, the third REOF shows a pronounced positive anomaly center over northeast Asia and a moderate positive center over the Ural Mountains (Fig. 2c). In this pattern, two oppositely signed centers are situated over Lake Baikal and the northwest Pacific to the east of Japan, respectively. In the REOF4 spatial pattern (Fig. 2d), wave train–like height anomalies are distributed from the North Atlantic to Lake Baikal, with the primary positive anomaly located over Lake Baikal. The remaining six REOFs, to a certain degree, share similar features with one of the four leading REOFs. For example, the REOF5, REOF8, and REOF9 patterns resemble the REOF1 pattern, although they show new secondary anomaly centers over the most upstream side or the most downstream side of the wave train–like height anomalies (not shown), compared to that of REOF1. We emphasize that the four leading REOFs not only represent the four largest fractions of variances, but also show a localized feature. Based on the fact that the neighboring ridges or blocks at the specific locations directly influence the NCCV activities, as discussed in section 1, we categorized the NCCVs in association with the REOF1, REOF2, REOF3, and REOF4 into four types: Ural type (UR), Yenisei type (YNS), Yakutsk–Okhotsk type (YO), and Baikal type (BKL) NCCVs. Specifically, NCCV events were categorized into four types in terms of their normalized indices, obtained through projecting the height anomalies at the peak days of NCCV events onto each spatial pattern of the four leading REOFs. Then, a category of NCCV events was identified if the highest projection index on a specific REOF pattern was greater than 0.5 and the difference between two leading projection indices was greater than 0.5. The latter criterion ensured that the NCCV category was mainly associated with a particular ridge or block. If these two criteria were not satisfied simultaneously, the corresponding NCCVs were categorized as “others.” Each NCCV category was farther subdivided into two groups according to their duration. Simply, short-lived (SL) NCCV events had a 3–5-day lifetime, while those persisting for longer were classed as long lived (LL).
Maps of Z500 anomalies (m) at the peak day of NCCV events regressed against the four leading RPCs of the NCCV-related circulations over the Eurasian continent: (a) REOF1, (b) REOF2, (c) REOF3, and (d) REOF4. The contour interval is 10 m. Shading marks the regions above the 95% confidence level. The polygon designates the region in northeast China and the lowest point in each panel is drawn at 20°N, 90°E.
Citation: Monthly Weather Review 143, 3; 10.1175/MWR-D-14-00192.1
As displayed in Table 1, 72 NCCV events (i.e., about 70%) conform to one of the four identified types. The remaining 31 NCCV events belong to the others category, and show a mix of two or more of the four leading REOF patterns. Most UR- and YNS-type NCCV events are SL, while YO- and BKL-type NCCV events tend to be LL, particularly the YO-type NCCV events. This indicates that blocking-type circulations over the Yakutsk–Okhotsk region may contribute to the persistence of NCCV events.
The number of each of the four NCCV types, and subdivisions of these types into short- and long-lived categories.
b. Circulation features
Since the REOF patterns, as shown in Fig. 2, are obtained with the height anomalies at the peak times of the NCCV events, they are not capable of representing the NCCV circulation features relative to climatological annual cycle. To depict the circulation features of NCCVs, composite Z500 charts for the four types of NCCVs and the corresponding anomalies relative to the climatological annual cycle are shown in Fig. 3. Figure 3 shows that a closed contour of low height (at 5570, 5500, 5540, and 5500 m, respectively) appears over northeast China, indicative of the NCCV. However, the composite closed low of the UR-type NCCVs is not so apparent in Fig. 3c, because of the dispersed locations of their NCCV centers. Analysis of the individual NCCVs (not shown) reveals that each of NCCVs does have a well-defined closed low. Correspondingly, there is a negative anomaly center over northeast China (Fig. 3), slightly southward displaced compared to the closed low. Meanwhile, for each type of NCCV event, the ridge or blocking-type circulation is evidently observed in the neighboring region of the NCCV, consistent with the four leading REOF patterns (Fig. 2). For BKL- and YNS-type NCCV events (Figs. 3a,b), the ridge resides over Lake Baikal and the Yenisei River valley, upstream of the NCCV. A striking zonal wave train–like pattern is observed from north Europe to northeast China.
Composite Z500 (contours; m) and associated anomalies (shading; see scale at bottom) at the peak days of (a) 21 BKL-, (b) 18 YNS-, (c) 19 UR-, and (d) 14 YO-NCCV events. The contour interval is 40 m, and the thick solid contours (at 5570, 5500, 5540, and 5500 m, respectively) denote NCCVs. Dotting marks the regions above the 90% confidence level. The polygon designates the region in northeast China and the lowest point in each panel is drawn at 20°N, 90°E.
Citation: Monthly Weather Review 143, 3; 10.1175/MWR-D-14-00192.1
The composite chart of UR-type NCCVs is characterized by a pronounced ridge over the Ural Mountains and a broad trough over Lake Baikal, with a closed low over northeast China (Fig. 3c). A northeast–southwest-oriented positive height anomaly extends from the Ural Mountains to the Yakutsk region and an elongated center of opposite sign extends from Lake Baikal to northeast China (Fig. 3c). For the YO-type NCCV (Fig. 3d), a prominent block anchors over the Yakutsk–Okhotsk region, with the cold vortex along its southwestern flank residing over northeast China. Correspondingly, a pronounced positive center prevails over the Yakutsk–Okhotsk region, and an oppositely signed center extends from Novaya Zemlya to northeast China. The anomalous configurations of UR- and YO-type NCCVs over East Asia show a meridionally dipole distributed signature, in sharp contrast to the wave train pattern of the BKL- and YNS-types of NCCVs.
Figure 4 shows longitude–height cross sections of geopotential height anomalies for the four types of NCCVs along the latitude of their centers. The NCCV is a deep weather system, spanning all of the troposphere up to the stratosphere with the strongest amplitude at the tropopause level (about 250 hPa). Notably, the negative height anomalies of BKL-type NCCVs extend to 30 hPa, whereas those of YO-type NCCVs only reach 150 hPa (Figs. 4a,d). It is shown that the negative height anomalies for the BKL- and YNS-type NCCVs exhibit an apparent westward tilt with height (Figs. 4a,b), reflecting a baroclinic conversion of available potential energy to kinetic energy associated with the circulation anomalies. In contrast, the negative height anomalies of UR- and YO-type NCCVs generally show the equivalent barotropic structure.
Longitude–height cross sections of geopotential height anomalies of (a) BKL- and (b) YNS-type NCCV events on the peak day along the latitude of 42.5°N, and (c) UR- and (d) YO-type NCCV events on the peak day along the latitude of 47.5°N. Isolines are drawn every 20 m, and the zero lines are omitted. Dark (light) shading marks the regions above the 95% (90%) confidence level.
Citation: Monthly Weather Review 143, 3; 10.1175/MWR-D-14-00192.1
4. Intraseasonal evolution of the NCCV event
As shown in section 3, there are two major configurations of circulation anomalies associated with NCCVs: a wave train–like pattern and meridional dipole pattern (Fig. 3). Since the circulation anomalies associated with the BKL- and YO-type NCCVs well represent the two major patterns, we choose these two types of NCCVs to address their typical evolution processes on the intraseasonal time scale in this section. Moreover, the primary ridge associated with the former type is situated on the upstream side of the NCCV (Figs. 3a–c), while the blocking-type circulation for the latter type is located on the downstream side of the NCCV (Fig. 3d). Differences between the LL and SL events for each of the two NCCV types are also discussed in terms of their formation and maintenance features.
a. BKL-type NCCV event
Figure 5 shows the Hovmöller diagram of Z500 anomalies along the line connecting the four anomaly centers of the circulation anomalies associated with the BKL-type NCCV (Fig. 3a). Hereafter, for brevity, day N (−N) refers to the day that is N days after (before) the peak day of the NCCV event. The four anomaly centers are strengthened progressively from the west to east, indicative of the propagation of Rossby wave packets. Anomalous centers of LL-NCCVs appear to be quasi-stationary, and the NCCV anomaly center remains over northeast China from days −2 to 3 (Fig. 5a). By contrast, anomaly centers of the SL-NCCV moves eastward with time, particularly after the peak day (Fig. 4b). After day 2, the NCCV center is displaced eastward of 135°E, away from northeast China.
(a) Composite Z500 anomalies (m) of BKL-type NCCV events at the peak day, as in Fig. 3a. A base line is drawn as connecting the four anomaly centers, along which the (b),(c) Hovmöller diagrams are generated. The polygon designates the region in northeast China. (b) The Hovmöller diagram in the time–longitude cross section for composite 500-hPa height anomalies of the LL-NCCV event. The contours interval is 20 m, and the thick solid lines indicate zero lines. Dark (light) shading marks the regions above the 95% (90%) confidence level. (c) As in (b), but for the SL-NCCV event.
Citation: Monthly Weather Review 143, 3; 10.1175/MWR-D-14-00192.1
To compare the circulation features between the LL- and SL-NCCV events, the corresponding composite daily Z500 anomalies are displayed in Fig. 6. We first discuss the situation of the LL-NCCV event. On day −4 (Fig. 6a), a pronounced positive height anomaly center resides around the Scandinavian Peninsula, and a weak negative anomaly center appears over the Ural Mountains. From day −2 to the peak day (Figs. 6b,c), the wave train–like anomalies intensify progressively eastward to northeast China. As a result, the negative height anomaly center develops on day −2 and deepens dramatically at the peak day over northeast China, indicating the formation and maturation of the NCCV circulation. Additionally, the primary positive height anomaly center for the BKL-type LL-NCCV event is mainly situated to the northeast of Lake Baikal, while that in Fig. 3a is just over Lake Baikal. The dipole pattern over northeast Asia/northwestern Pacific resembles the WP− pattern. On day 2 (Fig. 6d), despite the weakening amplitudes of the anomaly centers over the Ural Mountains, Lake Baikal, and northeast China, their positions are stationary.
Composite Z500 anomalies (m) in the intraseasonal evolution process for the LL-NCCV BKL-type event on (a) day −4, (b) day −2, (c) the peak day, and (d) day 2. The contours interval is 20 m, and the zero lines are omitted. Dark (light) shading marks the regions above the 95% (90%) confidence level. The polygon designates the region in northeast China and the lowest point in each panel is drawn at 20°N, 90°E. (e)–(h) As in (a)–(d), but for the SL-NCCV event.
Citation: Monthly Weather Review 143, 3; 10.1175/MWR-D-14-00192.1
The evolution of the circulation anomalies associated with the BKL-type SL-NCCV event is very different from that of the LL-NCCV event. From days −4 to −2 (Figs. 6e,f), prior to the SL-NCCV, a meridional pattern structure with a positive height anomaly band over the subarctic coast of Eurasia and a negative anomaly center over the Ural Mountains is formed and intensified. A weak positive height anomaly center also forms to the southwestern side of Lake Baikal. At the peak day (Fig. 6g), the positive height anomaly center amplifies and dominates over Lake Baikal, while the negative anomaly center of NCCV deepens dramatically and resides in the eastward side of the corresponding cold vortex center of the LL event (Fig. 6c). By day 2 (Fig. 6h), the NCCV moves eastward out of northeast China. Unlike the LL-NCCV event, no evidently positive anomaly is observed over the Yakutsk–Okhotsk region throughout the lifetime of the SL event, which signifies a distinction between the typical circulations of the LL and SL events.
Figure 7 shows the daily evolutions of the composite Ertel PV and horizontal wind on the 320-K isentropic surface during the BKL-type NCCV event. Both the PV and wind fields are unfiltered and the 320-K isentropic surface approximately corresponds to 300–500-hPa layers in the mid- and high latitudes in late spring and early summer. The Ertel PV was defined as (Hoskins et al. 1985), 
Composite daily evolution of unfiltered PV (contours and shading; see scale bar at bottom) and horizontal wind (arrows; see scale at bottom) on the 320-K isentropic surface for the BKL-type LL-NCCV event on (a) day −4, (b) day −2, (c) the peak day, (d) day 2, and (e) day 4. The contour interval is 1 PVU, and the zero lines are omitted. Dotting marks the regions above the 90% confidence level. The polygon designates the region in northeast China and the lowest point in each panel is drawn at 20°N, 90°E. (f)–(j) As in (a)–(e), but for the SL-NCCV event.
Citation: Monthly Weather Review 143, 3; 10.1175/MWR-D-14-00192.1
For the LL event, on day −4 (Fig. 7a), a high-PV air to the northwestern side of Lake Baikal intrudes southeastward. Then, on day −2 (Fig. 7b), an isolated high-PV center forms to the west of northeast China, and a low-PV center forms to its northern side, constituting a meridional dipole structure. It can be inferred from the cross-PV-contour wind that the negative PV advection contributes to the formation of the low-PV center, though the PV advection itself is not drawn. From days 0 to 2 (Figs. 7c,d), the low-PV air extends northeastward and merges with a low-PV center over the Sea of Okhotsk, showing a zonal low-PV band. The NCCV circulation is indicated as the isolated high-PV center. During this time, the PV field over northeast Asia and the northwestern Pacific exhibits an apparent anticyclonic Rossby wave breaking (Thorncroft et al. 1993). Meanwhile, the PV advection associated with the westerly diffluence over northeast Asia favors the wave breaking structure of the PV field. Nakamura (1994) concluded that the positive (negative) correlation between PV and meridional wind corresponds to a divergence (convergence) of wave activity flux. As seen in Figs. 7c and 7d, a negative correlation between PV and meridional wind is apparent over northeast Asia and it corresponds to a wave activity flux convergence, reflecting the wave obstruction over northeast Asia. It is consistent with the LL feature of the NCCV. At day 4 (Fig. 7e), the high-PV feature over northeast China weakens with the disappearance of the wave breaking structure, and the total PV field is recovered to the wave structure.
As for the SL event, from days −4 to −2 (Figs. 7f,g), there is also an intrusion of high-PV air from Lake Baikal into northeast China. From days 0 to 2 (Figs. 7h,i), unlike the LL event, the PV field keeps a wave train structure from the Scandinavian Peninsula to northeast China and the westerly diffluence over northeast Asia is considerably weak. Consequently, the high-PV band near Lake Baikal extends southeastward to northeast China. Moreover, a moderate low-PV air is amplified over the Kamchatka Peninsula (Fig. 7i), indicative of the downward wave propagation. At day 4 (Fig. 7j), the NCCV circulation disappears and a strong westerly dominates over northeast Asia. It suggests that the PV fields during the LL event experience an anticyclonic wave breaking event, thus showing a quasi-stationary feature, whereas those in the SL event signify the wave propagation feature.
b. YO-type NCCV event
Figure 8 shows the time–latitude diagram of the composite Z500 anomalies along the line connecting the two anomaly centers of circulation anomalies associated with the YO-type NCCV event on the peak day (Figs. 8a,b). The remarkable feature for both the LL and SL events is that the positive anomaly center over northeast Asia amplifies quickly and expands southward prior to the peak day. Correspondingly, the NCCV deepens over northeast China (Figs. 8c,d), particularly for the LL event. The blocking-type circulation reaches its maximum amplitude in the LL event on day 2 and then gradually weakens and slowly moves southward, while the NCCV gradually weakens and shifts southward after day 2 (Fig. 8c). However, two such height anomalies of the SL-NCCV event move southward more quickly than those for the LL event (Fig. 8d). As a result, the cold vortex moves out of northeast China after day 3.
Composite Z500 anomalies of (a) the LL and (b) SL event for YO-type NCCV at the peak day. A base line is drawn connecting the two anomaly centers, along which the (c),(d) time–latitude plots are generated. The polygon designates the region in northeast China. (c) The time–latitude cross section for the composite 500-hPa height anomalies of the LL-NCCV event. The contour interval is 20 m, and the thick solid line indicates the zero line. Dark (light) shading marks the regions above the 95% (90%) confidence level. (d) As in (c), but for the SL-NCCV event.
Citation: Monthly Weather Review 143, 3; 10.1175/MWR-D-14-00192.1
Figure 9 displays the daily evolution of the composite Z500 anomalies for the YO-type LL- and SL-NCCV events in detail. For the LL event, on day −4 (Fig. 9a), a dipole pattern with a broad positive height anomaly extending from Alaska to northeast Asia and a negative height anomaly center over the North Pacific is very similar to the WP− pattern. Upstream, a moderate positive anomaly center and a negative anomaly center are located around the Ural Mountains and Lake Baikal, respectively, indicative of wave train circulation anomalies. From day −2 to the peak day (Figs. 9b,c), the primary positive anomaly center over northeast Asia strongly amplifies and displaces westward, while the negative height anomaly center over Lake Baikal deepens and moves eastward to northeast China. On day 2 (Fig. 9d), the negative height anomaly center still anchors over northeast China, albeit with a slightly weakened amplitude. The daily evolution of circulation anomalies associated with the SL event shares some features of the LL event, particularly the westward expansion of the positive height anomaly center over northeast Asia (Figs. 9e–h). A broad and strong positive height anomaly center mainly dominates over the Bering Strait, which corresponds to North Pacific block, and then strengthens with progressively westward displacement. The negative height anomaly with a northeast–southwest axis deepens abruptly over northeast China on the peak day. In the decaying stage of the blocking circulation (day 2), the cold vortex over northeast China shifts eastward to the Sea of Japan.
As in Fig. 6, but for the YO-type NCCV event.
Citation: Monthly Weather Review 143, 3; 10.1175/MWR-D-14-00192.1
Composites of daily PV and horizontal wind fields on the 320-K isentropic surface are shown in Fig. 10. For the LL event, from days −4 to −2 (Figs. 10a,b), a prominent high-PV tongue over the Yenisei River valley intrudes southeastward into the western side of northeast China with a strong northwesterly wind. To the east of northeast China, a negative PV advection by a southwesterly wind induces a moderate ridge on day −2. On the peak day (Fig. 10c), due to the prominent positive PV advection, evidently closed high-PV air indicating an NCCV forms over northeast China. Then, on day +2 (Fig. 10d), the closed high-PV air moves anticlockwise to the east of northeast China. Concurrently, a low-PV tongue extends from the Sea of Okhotsk to northeast Asia, due to the negative PV advection by the southeasterly wind associated with the NCCV. During this time, the PV field over northeast Asia and the northwestern Pacific exhibits striking cyclonic Rossby wave breaking (Thorncroft et al. 1993). On day 4 (Fig. 10e), the high-PV feature over northeast China weakens with the waning of the wave breaking structure.
As in Fig. 7, but for the YO-type NCCV event.
Citation: Monthly Weather Review 143, 3; 10.1175/MWR-D-14-00192.1
As for the SL event, from days −4 to −2 (Figs. 10f,g), there is also an evident intrusion of high-PV tongue into northeast China. From days 0 to 2 (Figs. 10h,i), a moderate closed high-PV air is completely detached from the high-PV tongue over Lake Baikal and shifted toward northeast China. In sharp contrast to the LL event, the PV field maintains a wave train structure from the Scandinavian Peninsula to northeast China and a negative PV advection by the southwesterly wind is observed from Lake Baikal to northwestern northeast China. On day 4 (Fig. 10j), a strong low-PV tongue dominates over northeast China. Consequently, the PV fields during the LL event experience a cyclonic wave breaking event, thus presenting a quasi-stationary feature, while that in the SL event is characterized by the wave propagation.
The formation of YNS-type event bears a strong resemblance to that of the BKL-type event, and is characterized by wave train circulation anomalies emanating from western Atlantic to northeast China (not shown). As for the UR-type NCCV event, the cold vortex circulation over northeast China forms with the eastward displacement of a broad trough over middle Asia–northwest China, which is driven by an amplified positive height anomaly band over the Ural Mountains (not shown). A positive height anomaly over the Sea of Okhotsk, observed for the LL event for YNS- and UR-type NCCV, favors persistence of the NCCV.
c. Blocking activity associated with the NCCV events
As discussed above, the NCCV events are closely related to the blocking-type circulations. However, these results were mainly based on the composite anomalies, which may not adequately represent blocking activity. To address this concern, we used the blocking index defined by Small et al. (2014) to investigate the blocking activity for the four types of NCCVs. The blocking frequency of an NCCV event was defined as the percentage of the number of blocking days in the total number of days in its lifetime. Figure 11 shows the climatology of blocking frequency and those for the four types of NCCV events. As seen in Fig. 11a, climatologically, blocks frequently occur over a vast area covering North America, the North Atlantic, Europe, and the Ural Mountains during the NCCV active period. The locations of the maximum blocking frequencies for the four types of NCCV events, as shown in Figs. 11b–e, are consistent with the corresponding ridge or block in Fig. 3. Specifically, the center of blocking frequency extends from Lake Baikal to the Yakutsk–Okhotsk region for the BKL-type NCCV event, while the center resides over the Yenisei River basin for the YNS-type NCCV event (Figs. 11b,c). As for the UR-type NCCV events, blocks occur over Europe with a high frequency of 45% and over the Ural Mountains with a frequency of 30%. Moreover, a weak blocking frequency (up to 10%) is also observed over the Yakutsk–Okhotsk region. For the YO-type NCCV events, blocks dominate over northeast Asia with a frequency of 25% and simultaneously appear over Europe.
(a) Climatology of blocking frequency (%) during 1 May–15 Jun, and composite blocking frequency during NCCV events of the (b) BKL, (c) YNS, (d) UR, and (e) YO type. The first contour is at 5% and the interval is 5%. Dark (light) shading marks the regions above the 95% (90%) confidence level. The polygon designates the region in northeast China.
Citation: Monthly Weather Review 143, 3; 10.1175/MWR-D-14-00192.1
Figure 12 shows the blocking frequencies during the four types of NCCV events. For BKL- and YNS-type NCCV events, the block over Lake Baikal and the Yenisei River valley occurs more frequently in the LL event than in the SL event (Figs. 12a–d). Therefore, the block plays a much more important role for the persistence of the LL event in these two types of NCCVs. In contrast, for the UR-type NCCV, the block is more active over the Ural Mountains in the SL event than in the LL event (Figs. 12e,f). This is also true for the YO-type NCCV event, for which block mainly occurs over northeast Asia (Figs. 12g,h). In addition, for the YO-type LL-NCCV event, the block is also observed over the Ural Mountains, reflecting the contribution of the block over the Ural Mountains to the persistence of the YO-type NCCV event. Consequently, the frequent occurrence of the block over the key region is closely associated with the UR- and YO-type SL-NCCV events, rather than the corresponding LL events.
Composite blocking frequency (%) for the LL-NCCV event of (a) BKL, (b) YNS, (c) UR, and (d) YO type. The first contour is 10% and the interval is 10%. Dark (light) shading marks the regions above the 95% (90%) confidence level. The polygon designates the region in northeast China. (e)–(h) As in (a)–(d), but for the SL-NCCV event.
Citation: Monthly Weather Review 143, 3; 10.1175/MWR-D-14-00192.1
5. Weather impacts
Since the circulations for the four types of NCCV events have diverse configurations, their resultant SAT and precipitation anomaly distributions might be expected to differ. Figure 13 shows the composite SAT anomalies and horizontal wind at 850 hPa induced by the four types of NCCV events on the peak day. All types of NCCV events induce the northerly or northwesterly wind over northeast China, causing cold SAT anomalies there. However, the distribution of cold SAT anomalies varies for the four NCCV types. Specifically, the cold SAT anomalies for the BKL-type NCCV event are distributed over northeast China to the south to 50°N, with a minimum of −2.5°C (Fig. 13a). Compared to the situation of the BKL-type NCCV, the YNS-type NCCV influences a more extensive region and the resultant cold SAT anomaly is much stronger (Fig. 13b). The cold SAT anomalies prevail over northeast China and most parts of north China, with the cold anomaly core being lower than −3.5°C. In addition, the cold SAT anomalies associated with the YNS-type NCCV are also observed over the southern part of China. Such extensive cold SAT anomalies induced by the cold vortex are consistent with the cold air advection with a broader extent (Fig. 13b). In contrast, the cold SAT anomalies caused by the UR-type NCCV only cover northeast China and the center amplitude is only −1.5°C (Fig. 13c). Similarly, the influence of the YO-type NCCV is also limited to northeast China, though the strongest cold SAT anomaly is lower than −3°C (Fig. 13d). Both the strength and extent of the cold SAT anomalies induced by the UR- and YO-type NCCVs can be interpreted by the strength and extent of the northerly flow (Figs. 13c,d).
Composite SAT anomalies (contours; °C) and horizontal wind at 850 hPa (arrows, see scale at bottom) induced by (a) BKL-, (b) YNS-, (c) UR-, and (d) YO-type NCCV events on the peak day. The contour interval is 0.5°C, and zero lines are omitted. Dark (light) shading marks the regions above the 95% (90%) confidence level.
Citation: Monthly Weather Review 143, 3; 10.1175/MWR-D-14-00192.1
Previous studies have shown that the rainfall associated with an NCCV event tends to occur while the cold vortex circulation develops (Sun et al. 1995; Chen et al. 2008), and this was found to be the case for all four NCCV types in this study. Taking this fact into account, we averaged the precipitation anomalies during days −1 and 0 of each NCCV event, as displayed in Fig. 14. For all four types of NCCVs, above-normal precipitation mainly occurs at the eastern (or northeastern) portion of the vortex circulation, consistent with previous studies (Sun et al. 2002; Wang et al. 2012). The BKL- and YNS-type NCCVs lead to above-normal precipitation (1–3 mm day−1) over northeast China, but with a limited regional extent (Figs. 14a,b). The prevalence of the cold vortex circulation also causes below-normal precipitation anomalies on the northern side of the Yangtze River. In addition, these two types of NCCVs are also associated with above-normal precipitation over southwestern China for an unknown reason. In contrast, for the UR- and YO-type NCCVs, the relevant positive rainfall anomalies (up to 3 mm day−1) cover the middle and eastern parts of northeast China to the east of 120°E (Figs. 14c,d), showing a broader extent than those of the BKL- and YNS-type NCCVs. In addition, the YO-type NCCV also causes abundant rainfall anomalies over southern China and theYangtze River valley. Figure 3 shows that the zonal extents of the BKL- and YNS-type NCCVs are smaller than those of the UR- and YO-type NCCVs. This suggests that the zonal extent of the vortex circulation corresponds well to the extent of the rainfall anomaly distribution.
Composite precipitation anomalies (mm day−1) averaged over days −1 and 0 of (a) BKL-, (b) YNS-, (c) UR-, and (d) YO-type NCCV events. The first contour of positive (negative) anomaly is 1 (−1) mm day−1 and the interval is 2 mm day−1. Shading marks the regions above the 95% confidence level of surface air temperature anomalies.
Citation: Monthly Weather Review 143, 3; 10.1175/MWR-D-14-00192.1
6. Summary and discussion
Based on the long-term data in the active period of NCCVs (1 May–15 June) during 1965–2011, this study has used rotated EOF analysis to classify NCCV events into four typical types: BKL-, YNS-, UR-, and YO-type NCCV events. The four types of NCCV events were characterized by the distinct locations of ridges or blocking-type circulations situated close to the NCCVs. For the BKL-, YNS-, UR-, and YO-type NCCV events, ridges were observed over Lake Baikal, the Yenisei River valley, the Ural Mountains, and Yakutsk–Okhotsk region, respectively. Horizontally, the BKL- and YNS-type NCCV events were characterized by wave train circulation anomalies over mid- and high-latitude Eurasia, whereas the UR- and YO-type events featured a meridional height anomaly pattern. The NCCV was a deep weather system, spanning from the troposphere to the stratosphere, with its center at the tropopause. Vertically, the BKL- and YNS-type NCCVs exhibited a westward-tilted baroclinic structure, while the UR- and YO-type NCCVs showed an equivalent barotropic feature.
The formation and maintenance processes of the four types of NCCVs were diverse. The BKL-type NCCV formed with the wave train height anomalies originating from the upstream. An anticyclonic circulation anomaly was present over the Yakutsk–Okhotsk region prior to the LL-NCCV formation, whereas it was absent for the formation of the SL-NCCV. This suggests that such an anticyclonic anomaly maintained the accompanying NCCV circulation more persistently. Moreover, the NCCVs were associated with the intrusion of high-PV air over Lake Baikal. For the LL event, the wave structure of the PV field tended to break down anticyclonically over northeast China around the peak day of the NCCV, causing a temporary obstruction of the wave activity, whereas this phenomenon is absent for the SL event, signifying the distinction between the LL and SL event. The evolution of the YNS-type NCCV and its LL and SL characteristics resembled those of the BKL-type NCCV. In contrast, the formation of the UR-type NCCVs was initiated by an eastward displacement of a negative height anomaly center from Lake Baikal, driven by the southeastward expansion of a broad blocking-type circulation over the Ural Mountains. The YO-type NCCV formation was characterized by a westward expansion of a predominant blocking circulation over the Bearing Strait to northeast Asia and an eastward displacement of the negative height anomaly center over the southeastern side of Lake Baikal to northeast China. Unlike the BKL-type event, the high-PV center for the LL YO-type event deepened counterclockwise and broke down cyclonically, causing a temporary wave activity obstruction. The geographical locations of the maximum blocking frequencies during the four types of NCCV events agreed well with their corresponding predominant positive height anomalies close to the NCCV circulation. The blocking-type circulation contributed to the maintenance of LL BKL- and YNS-type NCCVs, whereas it favored SL UR- and YO-type NCCV events.
Because the four NCCVs types exhibited diverse configurations, they had different effects on the weather. Though they shared the similar northeast–southwest-elongated structure, the YNS- and BKL-type NCCV circulations caused different cold SAT anomaly distributions. The former caused cold air temperatures not only over northeast China, but also over central and south China, whereas the latter led only to regional cold SAT anomalies over northeast China. These two types of cold vortex circulation brought increased precipitation to northeast China, but the rainfall extents were regionally limited. Because of the zonally elongated structure of UR- and YO-type NCCVs, the resultant cold SAT anomalies were primarily restricted to northeast China. Compared to those of the YNS- and BKL-type NCCV circulations, the increase in rainfall was broader for the UR and YO-type NCCVs, and extended to cover the middle and eastern parts of northeast China.
This study mainly focused on the circulation features of four typical types of NCCVs, but the related dynamics involved in the evolution process remains unanswered. Because of the diversity of the four types of NCCV circulation, the processes of low-frequency Rossby wave energy dispersion and transient eddy feedback forcing might play different roles for the different types of NCCV events. As seen from Figs. 6 and 9, prior to the formation of the NCCV circulation, the WP− pattern already exists on the downstream side of the cold vortex circulation. However, it is still unknown how the WP− pattern contributes to the formation and maintenance of the NCCV event. Besides, NCCV circulation is also closely associated with the subtropical circulation system, such as the western Pacific subtropical high (Sun et al. 1994). Therefore, how the subtropical circulation system affects the formation process of the NCCV circulation deserves future investigation. We hope to address these issues in future work.
Acknowledgments
The authors thank Professor Ji Li-Ren for comments on an earlier version of the paper. Thanks also go to the writers of the NCARG Command Language (UCAR/NCAR/CISL/VETS, 2012), which was employed to plot the figures. This work was jointly supported by the National Natural Science Foundation of China (Grants 41305047 and 41375064), and the National Key Technologies R&D Program of China (Grants 2015BAC03B03 and 2009BAC51B02).
REFERENCES
Bamber, D. J., P. G. W. Healey, B. M. R. Jones, S. A. Penkett, A. F. Tuck, and G. Vaughan, 1984: Vertical profiles of tropospheric gases: Chemical consequences of stratospheric intrusions. Atmos. Environ., 18, 1759–1766, doi:10.1016/0004-6981(84)90351-2.
Barnston, A. G., and B. E. Livezey, 1987: Classification, seasonality, and persistence of low-frequency atmospheric circulation patterns. Mon. Wea. Rev., 115, 1083–1126, doi:10.1175/1520-0493(1987)115<1083:CSAPOL>2.0.CO;2.
Bell, G. D., and L. F. Bosart, 1989: A 15-year climatology of Northern Hemisphere 500 mb closed cyclone and anticyclone centers. Mon. Wea. Rev., 117, 2142–2164, doi:10.1175/1520-0493(1989)117<2142:AYCONH>2.0.CO;2.
Chen, L.-Q., L.-Z. Zhang, and X.-S. Zhou, 2008: Characteristic of unstable energy distribution in cold vortex over northeastern China and its relation to precipitation area (in Chinese). Plateau Meteor., 27, 339–348.
Hannachi, A., I. T. Jolliffe, D. B. Stephenson, and N. Trendafilov, 2006: In search of simple structures in climate: Simplifying EOFs. Int. J. Climatol., 26, 7–28, doi:10.1002/joc.1243.
Horel, J. D., 1981: A rotated principal component analysis of the interannual variability of the Northern Hemisphere 500 mb height field. Mon. Wea. Rev., 109, 2080–2092, doi:10.1175/1520-0493(1981)109<2080:ARPCAO>2.0.CO;2.
Horel, J. D., 1984: Complex principal component analysis: Theory and examples. J. Climate Appl. Meteor., 23, 1660–1673, doi:10.1175/1520-0450(1984)023<1660:CPCATA>2.0.CO;2.
Hoskins, B. J., M. E. McIntyre, and A. W. Robertson, 1985: On the use and significance of isentropic potential vorticity maps. Quart. J. Roy. Meteor. Soc., 111, 877–946, doi:10.1002/qj.49711147002.
Hsieh, Y.-P., 1949: An investigation of a selected cold vortex over North America. J. Meteor., 6, 401–410, doi:10.1175/1520-0469(1949)006<0401:AIOASC>2.0.CO;2.
Hu, K. X., R. Y. Lu, and D. H. Wang, 2010: Seasonal climatology of cut-off lows and associated precipitation patterns over Northeast China. Meteor. Atmos. Phys., 106, 37–48, doi:10.1007/s00703-009-0049-0.
Kaiser, H. F., 1958: The varimax criterion for analytic rotation in factor analysis. Psychometrika, 23, 187–200, doi:10.1007/BF02289233.
Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77, 437–471, doi:10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO;2.
Lian, Y., C. Bueh, Z. W. Xie, B. Z. Shen, and S. F. Li, 2010: The anomalous cold vortex activity in Northeast China during the early summer and the low-frequency variability of the northern hemispheric atmosphere circulation (in Chinese). Chin. J. Atmos. Sci., 34, 429–439.
Liu, C. X., Y. Liu, X. Liu, and K. Chance, 2013: Dynamical and chemical features of a cutoff low over northeast China in July 2007: Results from satellite measurements and reanalysis. Adv. Atmos. Sci., 30, 525–540, doi:10.1007/s00376-012-2086-8.
Nakamura, H., 1994: Rotational evolution of potential vorticity associated with a strong blocking flow configuration over Europe. Geophys. Res. Lett., 21, 2003–2006, doi:10.1029/94GL01614.
Nieto, R., and Coauthors, 2005: Climatological features of cutoff low systems in the Northern Hemisphere. J. Climate, 18, 3085–3103, doi:10.1175/JCLI3386.1.
Nieto, R., and Coauthors, 2007: Analysis of the precipitation and cloudiness associated with COLs occurrence in the Iberian Peninsula. Meteor. Atmos. Phys., 96, 103–119, doi:10.1007/s00703-006-0223-6.
Palmén, E., 1949: Origin and structure of high-level cyclones south of the maximum westerlies. Tellus, 1, 22–31, doi:10.1111/j.2153-3490.1949.tb01925.x.
Palmén, E., and K. M. Nagler, 1949: The formation and structure of a large-scale disturbance in the westerlies. J. Meteor., 6, 228–242, doi:10.1175/1520-0469(1949)006<0228:TFASOA>2.0.CO;2.
Ren, L., C. W. Wang, L. Jiao, and L. Jin, 2010: Preliminary analysis on the causes of rainy and cold weather over Heilongjiang province in June 2009. J. Anhui Agric. Sci., 38, 6775–6777.
Rex, D. F., 1950: Blocking action in the middle troposphere and its effect upon regional climate. Tellus, 2, 196–211, doi:10.1111/j.2153-3490.1950.tb00331.x.
Schwierz, C., M. Croci-Maspoli, and H. C. Davies, 2004: Perspicacious indicators of atmospheric blocking. Geophys. Res. Lett., 31, L06125, doi:10.1029/2003GL019341.
Simmons, A. J., J. M. Wallace, and G. W. Branstator, 1983: Barotropic wave propagation and instability, and atmospheric teleconnection patterns. J. Atmos. Sci., 40, 1363–1392, doi:10.1175/1520-0469(1983)040<1363:BWPAIA>2.0.CO;2.
Small, D., E. Atallah, and J. R. Gyakum, 2014: An objectively determined blocking index and its Northern Hemisphere climatology. J. Climate, 27, 2948–2970, doi:10.1175/JCLI-D-13-00374.1.
Sun, L., 1997: A study of the persistence activity of northeast cold vortex in China. Scientia Atmos. Sin., 21, 297–307.
Sun, L., X. Y. Zheng, and Q. Wang, 1994: The climatological characteristics of northeast cold vortex in China (in Chinese). Quart. J. Appl. Meteor., 5, 297–303.
Sun, L., Q. Wang, and X. Tang, 1995: A composite diagnostic analysis of cold vortex of storm-rainfall and non-storm rainfall types (in Chinese). Meteor. Monogr., 21, 7–10.
Sun, L., G. An, Y. Lian, B. Z. Shen, and X. L. Tang, 2000: A study of the persistent activity of northeast cold vortex in summer and its general circulation anomaly characteristics (in Chinese). Acta Meteor. Sin., 55, 704–714.
Sun, L., G. An, Z. T. Gao, X. L. Tang, L. Ding, and B. Z. Shen, 2002: A composite diagnostic study of heavy rain caused by the northeast cold vortex over Songhuajiang-Nenjiang River basin in summer of 1998 (in Chinese). J. Appl. Meteor. Sci., 13, 156–162.
Tao, S., 1980: Heavy Rainfall in China. Science Press, 224 pp.
Thorncroft, C. D., B. J. Hoskins, and M. E. McIntyre, 1993: Two paradigms of baroclinic-wave life-cycle behaviour. Quart. J. Roy. Meteor. Soc., 119, 17–55, doi:10.1002/qj.49711950903.
Wallace, J. M., and D. S. Gutzler, 1981: Teleconnections in the geopotential height field during the Northern Hemisphere winter. Mon. Wea. Rev., 109, 784–812, doi:10.1175/1520-0493(1981)109<0784:TITGHF>2.0.CO;2.
Wang, P., X. Y. Shen, and S. T. Shou, 2012: A numerical study and rainfall analysis of a cold vortex process over Northeast China (in Chinese). Chin. J. Atmos. Sci., 36, 130–144.
Wang, Y., and A. R. Lupo, 2009: An extratropical air–sea interaction over the North Pacific in association with a preceding El Niño episode in early summer. Mon. Wea. Rev., 137, 3771–3785, doi:10.1175/2009MWR2949.1.
Wang, Y., X. Xu, A. R. Lupo, P. Li, and Z. Yin, 2011: The remote effect of the Tibetan Plateau on downstream flow in early summer. J. Geophys. Res., 116, D19108, doi:10.1029/2011JD015979.
Wilks, D., 1995: Statistical Methods in the Atmospheric Sciences. Academic Press, 121 pp.
Wu, Y.-Q., Q. Yan, X.-Y. Kang, L.-Y. Jin, and G.-X. Chang, 2009: Anomaly characteristics of persistent activity northeast cold vortex in May of 2008 (in Chinese). J. Meteor. Environ., 25, 13–17.
Xie, Z., and C. Bueh, 2012: Low frequency features of northeastern China cold vortex and its background circulation pattern (in Chinese). Acta Meteor. Sin., 70, 704–716.
Zhang, L. X., 2008: The study on MCS in cold vortex over northeastern. Ph.D. thesis, Nanjing University of Information Science and Technology, 11 pp. [Available from Nanjing University of Information Science and Technology, 219 N. Ningliu Rd., Nanjing 210044, China.]
Zhao, S., and J. Sun, 2007: Study on cut-off low-pressure systems with floods over Northeast Asia. Meteor. Atmos. Phys., 96, 159–180, doi:10.1007/s00703-006-0226-3.
Zhao, Y. F., J. Zhu, and Y. Xu, 2014: Establishment and assessment of the grid precipitation datasets in China for recent 50 years (in Chinese). J. Meteor. Sci., 34, 414–420.
Zhao, Y. J., M. Wang, K. Tian, and P. He, 2010: Rare low temperature rain origin analysis at high latitude area in June. J. Anhui Agric. Sci., 38, 6765–6767.
Zheng, X. Y., T. Z. Zhang, and H. R. Bai, 1992: Heavy Rainfall in Northeastern China. China Meteorological Press, 219 pp.
Zhong, S. X., 2011: Structural features of cold vortex and its formation mechanism of heavy rainfall over Northeast China. Ph.D. thesis, Chinese Academy of Sciences, 16 pp. [Available from Institute of Atmospheric Physics, Chinese Academy of Sciences, 40 N. Huayanli Rd., Beijing 100029, China.]
