1. Introduction
According to the Clausius–Clapeyron equation, the saturated water vapor pressure in the atmosphere will increase by about 7% for each degree of global warming (Trenberth et al. 2003), which has further aggravated the intensity and frequency of extreme precipitation events. The population exposed to extreme precipitation events have increased by about 234% since the twenty-first century (2000–19) compared to the previous 20 years (1980–99) (UN Office for Disaster Risk Reduction 2020). China is one of the countries prone to severe flooding, especially for the Yangtze–Huaihe River basin (YHRB). A growing number of persistent extreme precipitation events (PEPEs) have been observed over the YHRB since the 1990s (Chen and Zhai 2013) and have exerted great social impacts. For example, the YHRB suffered a record-breaking mei-yu in 2020, which affected more than 54.811 million people and led to a direct economic loss of more than 144.43 billion Chinese Yuan (Zhang et al. 2020). Therefore, understanding the possible mechanisms of PEPEs in the YHRB is imperative for the local communities, socioeconomic development, and government’s scientific decision-making.
The PEPEs are closely associated with the stable maintenance of abnormal atmospheric circulation systems, whose lifespan covers from days to weeks (Chen et al. 2019; Zhai et al. 2016). For instance, the precipitation extremes are found to be prolonged by opportune configuration among different atmospheric teleconnections, such as the East Asia–Pacific (EAP) teleconnection, the Silk Road teleconnection, and the Eurasia teleconnection (Chen and Zhai 2015; Chen et al. 2019; Guo and Wang 2023; Liu et al. 2024). These three teleconnection patterns actually reflect the interactions among tropical, subtropical, and the mid–high latitudes’ systems. Among them, the EAP pattern is typically characterized by the configuration of the western North Pacific subtropical high (WNPSH), the mei-yu trough, and the Okhotsk blocking high (Bueh et al. 2008). It is found that the concurrent combinations of the westward extension of the WNPSH and the mid–high latitude blocking highs (e.g., Ural blocking and Okhotsk blocking) are responsible for the occurrence of PEPEs, which result in the intersection of warm moist and cold dry air in the YHRB (Chen and Zhai 2013; Sun et al. 2020; Wang et al. 2017). The eastward extension of the South Asian high (SAH) could release condensational heat and trigger a positive geopotential height anomaly in the upper troposphere (Ren et al. 2015; Zhang et al. 2016), which is conducive to the formation of upward motion (Chen and Zhai 2016; Wei et al. 2014) and favors the development of PEPEs in the YHRB.
In addition to the aforementioned circulation systems, the Northeast China cold vortex (NEC-CV) is also an essential factor affecting the extreme precipitation (Cao et al. 2024; Chen et al. 2023; Li et al. 2024). As a well-known vigorous cutoff low system in Northeast Asia, the long-lived NEC-CVs could cause local instability and lead to various high-impact weather (Hsieh 1949; Shi et al. 2020, 2023). The mei-yu rainfall has been proven to be closely related to the NEC-CVs, with the total precipitation increased by stronger NEC-CVs (He et al. 2006; Wang et al. 2010). So far, however, the systematic analysis of individual and combined importance of the NEC-CVs for extreme precipitation is still constrained to Northeast China or the evaluation of typical short-term strong convective processes and remains unexplored for the PEPEs in the YHRB. Additionally, a NEC-CV-like weather system is clearly found in the recent studies (e.g., Chen and Zhai 2014; Cheng et al. 2021), with a deep low pressure located to the north of the YHRB during the PEPEs. Hence, it is pertinent to explore the possible mechanisms of the NEC-CVs in influencing the PEPEs over the YHRB.
The organization of this study is as follows. Section 2 briefly describes the data and methods used in the research. Section 3 qualitatively demonstrates the importance of the NEC-CVs for PEPEs and objectively identifies the PEPEs that coincide with the NEC-CVs by applying state-of-the-art technique. Section 4 presents the dynamic processes of NEC-CVs controlling the PEPEs over the YHRB and discusses the corresponding physical mechanisms. To the end, the main conclusions are provided in section 5.
2. Data and methods
a. Data
In this study, the daily gridded precipitation dataset CN05.1 (Wu and Gao 2013) is used for the identification of PEPEs during the summer season (June–August). This dataset was constructed by interpolating data over more than 2400 observing stations in China, with a horizontal resolution of 0.25° × 0.25° and ranging from 1961 to 2022. The daily ERA5 reanalysis dataset during 1961–2022 derived from the European Centre for Medium-Range Weather Forecasts (ECMWF; Hersbach et al. 2020) is also applied in this study, including total precipitation, geopotential height, air temperature, meridional/zonal wind, vertical velocity, specific humidity, and surface pressure. The reanalysis dataset has a horizontal resolution of 1° × 1° and 37 vertical levels ranging from 1000 to 1 hPa. Meanwhile, the daily gridded precipitation data obtained from the Climate Prediction Center (CPC; Xie et al. 2010) are used to validate the robustness of relevant results, which has a horizontal resolution of 0.5° × 0.5° and starts from 1979 to 2022.
b. Definition of PEPEs
At present, there is no unified standard for the definition of PEPEs. Previous studies have defined PEPEs in terms of intensity by using a fixed criterion (e.g., ≥50 mm day−1) (Chen and Zhai 2013; Lin et al. 2020) or a quantile threshold (e.g., 95% percentile calculated from the probability distribution of daily precipitation during a specific period) at individual grid points. However, these methods may have certain limitations. On the one hand, it is not appropriate to define the extreme precipitation with the criterion 50 mm for regions like North China and Northeast China, as the anomalous events with such intensity are rarer compared with those in the lower latitudes. On the other hand, it is difficult to objectively define a regional extreme precipitation event when considering its spatial extent. Previous studies have defined regional anomalous events by a certain number of stations or grid points with the occurrence of extreme events in the study area (Chen and Zhai 2013; Lin et al. 2020), which is subject to the human annotation and rather empirical.
In this study, an autoencoder network is utilized to detect PEPEs, which is an emergent unsupervised learning technique (or neural network) to objectively identify anomalous events from daily precipitation data by recent studies (Murakami et al. 2022; Huang et al. 2023). Compared to the commonly used methods before, this approach is independent from subjective human annotation or predefined arbitrary thresholds (Liu et al. 2016; Racah et al. 2017) and has been proven to be more accurate than conventional methods for anomalous events (Zhao and Du 2016). The PEPEs are detected by using a similar scheme documented in Murakami et al. (2022) and Huang et al. (2023). In short, the selected higher-dimensional daily precipitation anomaly fields (e.g., the YHRB) are first compressed to a lower-dimensional latent representation (encoder). Then, the lower-dimensional data are reconstructed to the higher-dimensional data with the same sizes as the original input fields (decoder). In general, a well-trained autoencoder can reproduce the original input fields very well when a sample is frequently represented in the training dataset and hence leads to smaller mean-square errors (MSEs). In other words, if a rare sample appears in the training dataset, a larger error will be caused in the process of decoder, resulting in a large MSE between the spatial patterns before and after autoencoding. Therefore, the anomalous precipitation events can be detected with the events having larger MSEs (here the 90th percentile of the MSE distribution). The PEPE is defined as an anomalous event lasting for at least 2 days.
c. Identification and track of NEC-CVs
The NEC-CV is recognized by the daily 500-hPa geopotential height (Z500) and air temperature in summer (June–August), following the traditional definition by Sun et al. (1994). The corresponding scheme mainly consists of three steps (Shi and Zhai 2024): 1) the detection of a closed low, 2) the detection of a cold core or trough, and 3) the track of a NEC-CV exists within Northeast China (110°–140°E, 35°–60°N) for certain life cycle (at least 3 days). A detailed description of the automated identification procedure for a NEC-CV can be referred in Shi and Zhai (2024).
d. SOM
e. Vertically integrated water vapor flux
f. Horizontal wave activity flux
3. Impacts of the NEC-CVs on the PEPEs and relative circulation classification
a. Features of precipitation related to NEC-CVs
To get insight into the effects of NEC-CVs on the summer precipitation over China, we start with a dynamic composite of the coherent precipitation during the occurrence of NEC-CVs (Fig. 1). Here, the precipitation occurring simultaneously with the NEC-CVs is considered as the cold vortex–related precipitation, regardless of whether it is directly or indirectly caused by the NEC-CVs. Spatially, it can be found that there are two maxima centers for the intensity and frequency of precipitation (Figs. 1a,b), with one mainly situated in the immediate vicinity of the NEC-CVs and the other located in the southwest quadrant. On the contrary, little impact has been observed for the northern part to the core area of NEC-CVs, which is primarily attributed to the dominant role played by cold advection and dry cold air in this area. This means that the impact of the NEC-CVs on precipitation is mainly manifested in two aspects, including the influence on regional precipitation and the effect on long-distance precipitation. Heavy precipitation (≥25 mm day−1) also exhibits a similar distribution but with higher occurrence probabilities in the southeast quadrant (Fig. 1c). However, rainstorm (≥50 mm day−1) is more likely to occur in the southern part of the NEC-CVs (Fig. 1d), implying that rainstorm is closely related to the long-distance effect of the NEC-CVs. The distribution of the PEPEs is also characterized by two local maxima, but with higher frequency observed in the immediate vicinity of the NEC-CVs (Fig. 1f). It is not incompatible with the results using fixed thresholds (Figs. 1c,d), as the anomalous precipitation with the criterion 50 mm is rarer in the high latitudes compared with that in the lower latitudes (Fig. 1e). This in turn reflects the rationality of using objective identification methods for PEPEs. Similar patterns have also been found for the precipitation derived from ERA5 reanalysis (1940–2022) at a longer time scale (results not shown).
Spatial distributions of (a) the mean precipitation intensity (mm day−1), (b) the precipitation frequency (%), (c) the precipitation frequency (%) for heavy rain (≥25 mm day−1), (d) the precipitation frequency (%) for rainstorm (≥50 mm day−1), (e) the mean precipitation intensity (mm day−1) for PEPEs, and (f) the precipitation frequency (%) for PEPEs based on the CPC daily grid precipitation dataset from 1979 to 2022. The black contours (gpm) indicate the composite geopotential height at 500 hPa. The low center of each NEC-CV is taken as point (0, 0) to perform the composite. The PEPEs in (e) and (f) are identified by using the conventional method, with the precipitation greater than or equal to the 90% percentile of daily climatology at individual grid point for at least 2 days.
Citation: Journal of Climate 38, 10; 10.1175/JCLI-D-24-0299.1
Taking each NEC-CV into account, one might question the distance in which the precipitation can be estimated as NEC-CV-related precipitation. Most previous studies have used fixed radius thresholds to estimate the precipitation related to cyclonic weather systems, for example, a value ranging from 250 to 1000 km in the studies regarding tropical cyclones (Chen et al. 2010; Dare et al. 2012; Li and Zhou 2015) and a threshold 660 km for cutoff lows (Abatzoglou 2016). Perhaps, it is inaccurate for NEC-CVs, as the horizontal scale of each NEC-CV varies greatly (250–1250 km) in summer (Chen et al. 2023; Shi and Zhai 2024). Larger or smaller thresholds can overestimate or underestimate the related properties of precipitation to some extent. In such a case, precipitation at different distances from the NEC-CV center is averaged to determine the effective radius of influence of NEC-CVs (Fig. 2). Clearly, the mean rainfall and frequency decrease exponentially with increasing distance from the NEC-CV centers. The changes in rainfall and frequency start to level off at a distance of 4 times NEC-CV radius by using CPC and ERA5. This suggests that the NEC-CV influence is actually limited to this range. Hence, this relative threshold is more appropriate to estimate the precipitation related to NEC-CVs. Daily precipitation at a grid point is then classified as NEC-CV rainfall when the grid lies within the relative threshold of each NEC-CV (4× radius).
Mean precipitation intensity (red) and frequency (blue) averaged within different distances from the NEC-CV centers. The circle and triangle indicate the results derived from ERA5 and CPC, respectively. The distance is converted to relative thresholds with the radius of each NEC-CV.
Citation: Journal of Climate 38, 10; 10.1175/JCLI-D-24-0299.1
Based on the fourfold threshold, the distributions of the relative contribution of the NEC-CVs to the precipitation in eastern China during summer are shown in Fig. 3. It turns out that the maximum center of the extreme precipitation related to NEC-CVs is located over Northeast China, including the northeastern part of Inner Mongolia and the western part of Heilongjiang Province. This is consistent with previous studies that NEC-CVs account for a large proportion of precipitation over Northeast China in summer. The relative contribution of extreme precipitation related to NEC-CVs decreases from northeast to southwest areas, with the secondary centers mainly located over North China and the YHRB (35%–40%). The extreme precipitation related to NEC-CVs can also explain a considerable proportion of total extreme precipitation over Northeast China, North China, East China, and Southeast China (Figs. 3e–h). In addition, a more pronounced impact of the extreme precipitation related to NEC-CVs can be found for more extreme precipitation (Figs. 3g,h), suggesting the nonnegligible role of NEC-CVs in controlling the extreme precipitation in eastern China.
Contributions of (left) NEC-CV rainfall to total rainfall frequency and (right) total rainfall amount for total rainfall anomalies (%) during summer over 1961–2022. (a),(b) The results of all precipitation events; (c),(d) the extreme events (defined as the precipitation greater than or equal to the 90th percentile at individual grids). (e),(f) and (g),(h) The extreme precipitation events persisting for at least 2 days (greater than or equal to 90th and 95th percentiles, respectively). The black box indicates the selected area of the YHRB (112.5°–123.5°E, 29.5°–33.5°N).
Citation: Journal of Climate 38, 10; 10.1175/JCLI-D-24-0299.1
b. Identification of PEPEs associated with NEC-CVs
Based on the autoencoder neural network technique, the abnormal days of precipitation over the YHRB can be determined (Fig. 4a), with the maximum value found in year 2020, when the record-breaking mei-yu was observed (Ding et al. 2021). During 1961–2022, 130 PEPEs are identified over the YHRB (Fig. 4b). The PEPEs mainly occur in late June and early July (Fig. 4c), which is consistent with the date of rainy season over the YHRB.
(a) Number of days (day) with anomalous precipitation, (b) calendar of the PEPEs, and (c) its corresponding distributions over the YHRB during summer over 1961–2022. The blue and orange lines indicate the total PEPEs and PEPEs related to NEC-CVs, respectively.
Citation: Journal of Climate 38, 10; 10.1175/JCLI-D-24-0299.1
Generally, the human visual judgment of whether a certain type of weather system plays a major role in an event is highly subjective, thus making it difficult to objectively pick out PEPEs that are closely associated with NEC-CVs. Here, taking the advantage of SOM, the 500-hPa geopotential height and the corresponding precipitation can hence be jointly classified (Fig. 5). A different number of SOM clusters have been tested by calculating the silhouette coefficient, with a higher score indicating better performance. The silhouette coefficient generally decreases with the increasing number of clusters. However, when classified into three clusters, we found that it was insufficient to distinguish the relationship between the NEC-CVs and the corresponding precipitation. As for the clusters ranging from 4 to 6, the distribution of samples in each category is not uniform, and there are also many samples with negative silhouette coefficients. In contrast, the silhouette coefficient is relatively higher when the cluster number is set as 7, which is second to the case of three clusters and has a relatively uniform distribution for each cluster.
Silhouette plot for SOM clusters with 500-hPa geopotential height and the coherent precipitation during the occurrence of NEC-CVs in summer. The red dashed line indicates the mean value of the silhouette coefficient for each cluster.
Citation: Journal of Climate 38, 10; 10.1175/JCLI-D-24-0299.1
Figure 6 further illustrates the specific spatial distributions of the seven clusters. It is shown that the NEC-CVs are closely related to the local precipitation and long-distance rainfall, which further confirms the robustness of the conclusion obtained in Section 3a. Obviously, the precipitation in the Northeast China is mainly influenced by the local effects of NEC-CVs, such as categories 1, 2, and 7, with the corresponding high-value areas of precipitation located in the southeastern region ahead of the NEC-CVs. However, higher precipitation is observed over the YHRB for types 1, 3, 6, and 7. Among them, significant precipitation is found for eastern China in type 7, but only accounts for about 1.97%. As for the other three SOM clusters (types 2/4/5), although they have a comparable contribution to the categories, the maximum center of the coherent precipitation is located to the north and the south of the YHRB, respectively, which suggests the weaker influence of NEC-CVs on the extreme precipitation over the YHRB. Therefore, combined with the calendar of detected NEC-CVs, it can be said that a PEPE is related to a NEC-CV when a coherent sample of daily precipitation lies in the above four SOM patterns and corresponds to a NEC-CV. In such sense, 36 events are selected as the PEPEs associated with NEC-CVs, accounting for about 27.7% of the total PEPEs.
SOM clusters of 500-hPa geopotential height (contours; gpm) and coherent precipitation (shading; mm day−1). The black box indicates the selected area of the YHRB (112.5°–123.5°E, 29.5°–33.5°N).
Citation: Journal of Climate 38, 10; 10.1175/JCLI-D-24-0299.1
c. Classification of PEPEs associated with NEC-CVs
As can be observed in Fig. 6, type 1 and type 3 are characterized by pronounced differences in the longitudinal positions of NEC-CVs, which accounts for about 28% and 18% of the total samples, respectively. Because the resulting SOM maps represent the most important features of the input space (Gibson et al. 2017; Hewitson and Crane 2002; Zhou et al. 2020), these two categories suggest the possible role of the longitudinal changes of the NEC-CVs’ positions. In addition, previous studies have also pointed out that different locations of NEC-CVs may have a different influence on the PEPEs (Chen et al. 2023; Xu and Qi 2023). Will the westward-/eastward-displaced NEC-CVs have great impacts on the PEPEs over the YHRB? Figure 7 further depicts the plot of the trajectories for 36 events. The result shows that most NEC-CVs move along a northerly or middle path, and only a few processes of NEC-CVs are located very southward. From the viewpoint of precipitation timing of each NEC-CV, it can be seen that the majority of heavy precipitation usually occurs at the early stage of NEC-CVs (around the first 4 days). Only a small number of extreme precipitation events are found to occur particularly late to the east of Northeast China, with the positions of corresponding NEC-CVs far away from the YHRB and hovering to the west of the Sea of Okhotsk for over 10 days. An obvious gap is found at around 125°E for the average locations of NEC-CVs (Fig. 7b), suggesting the possible role of longitudinal positions for NEC-CVs. Hence, to elucidate the possible impacts of NEC-CVs on anomalous precipitation events, the PEPEs related to NEC-CVs are grouped into two different groups based on the mean location of each NEC-CV track. If the mean position of a NEC-CV is smaller/greater than 125°E, the corresponding PEPE is referred to as a western-/eastern-type NEC-CV precipitation (type W/type E).
Trajectories of the NEC-CVs related to PEPEs over the YHRB. The red and blue points in (a) represent the starting and ending point of a NEC-CV track, respectively. The green points represent the centers of NEC-CVs coincident with heavy precipitation, with the corresponding numbers indicating the day of the center point within the respective cold vortex process, and gray points represent the remaining positions of the NEC-CVs. The yellow ellipse indicates the main range of the cold vortex locations with coherent precipitation. The red and blue points in (b) represent the average location of type W and type E, respectively.
Citation: Journal of Climate 38, 10; 10.1175/JCLI-D-24-0299.1
4. Dominant atmospheric circulation patterns for different types of PEPEs
a. Configuration of circulation systems
A composite analysis is conducted for the 36 events, and the corresponding distributions of the precipitation and water vapor for the two types of PEPEs are shown in Fig. 8. It is found that the rain belts are located along the middle and lower reaches of the Yangtze River, with the maximum precipitation concentrated in the YHRB. Wider coverage of precipitation anomaly is also observed for type W (Fig. 8a), with a secondary center located in the middle reaches of the Yangtze River basin. This is important for the refined forecast in the different provinces over China, especially for the border areas between Chongqing and Guizhou Provinces. Meanwhile, higher precipitation is also found over the northern part of Northeast China in type W.
The anomalies of (a),(b) mean precipitation and (c),(d) integrated water vapor flux (vectors; 100 kg m−1 s−1) from 300 to 1000 hPa along with the water vapor flux divergence (shading; 10−5 kg m−2 s−1) averaged for (left) type W and (right) type E. The stippling indicates the composite anomalies are statistically significant at the 95% confidence level (Student’s t test).
Citation: Journal of Climate 38, 10; 10.1175/JCLI-D-24-0299.1
From the perspective of moisture transport, the two types of PEPEs are linked with the anomalous water vapor transport to the YHRB along the moisture channel over the periphery of the WNPSH. Accordingly, a prominent convergence band extends from the YHRB to the southern part of the Korean Peninsula (Figs. 8c,d). In addition, it is worth noting that the sources of water vapor are quite different between type W and type E. In the case of type W, there is an enhanced water vapor flux from the western North Pacific to the South China Sea, which eventually converges with the southward cold air associated with the NEC-CVs over the YHRB (Fig. 8c). Moreover, relatively weaker water vapor is found to transport via the Bay of Bengal–Central South Peninsula in type W, while the anomalous warm and moist water vapor are mainly confined to the South China Sea for type E (Fig. 8d). On the contrary, Northeast Asia is characterized by an anomalous southwestward dry air, thereby leading to a more southward displaced convergence zone for type E.
The circulation configuration for the two types of PEPEs related to NEC-CVs is depicted in Fig. 9. It is obvious that the two types of PEPEs share some common characteristics. For example, the eastward extension of the SAH and southward shift of upper-level jet stream are found at 200 hPa, with the YHRB located on the right hand side of the upper-level jet stream, favoring the upward motion of a secondary circulation (Figs. 9a,b). At 500 hPa, the enhanced WNPSH is observed over the western North Pacific, together with a significant north–south-oriented dipole pattern over Northeast Asia, which is more conducive to the southward intrusion of cold air to the YHRB and intersects with the southwest warm and humid airflow along the periphery of WNPSH (Figs. 9c,d). The temperature and humidity gradients are further increased, which ultimately leads to the stable maintenance of fronts in the YHRB. At 850 hPa, the YHRB is dominated by a significant anomalous low pressure system, while the lower latitudes are characterized by an anomalous high pressure system (Figs. 9e,f). The YHRB lies in the area with a sharp meridional gradient due to the negative and positive anomalies at mid–high latitudes. Meanwhile, the circulation configuration shows a pattern with upper-level divergence and lower-level positive vorticity over the YHRB, which in turn further supports the strong secondary upward motion.
(a),(b) Composite wind field (vectors; m s−1), geopotential height for the SAH (blue solid contours; 12 520 gpm), and upper-level jet stream (shading; ≥25 m s−1) at 200 hPa, with the dashed lines corresponding to their climatology, respectively. The black dots indicate the statistical significance of the wind field is above the 95% confidence level. The divergence field is also shown (green contours; ≥1 × 10−5 s−1). (c),(d) Composite geopotential height (solid contours; gpm) and standardized anomalies of geopotential height with respect to the climatology at 500 hPa. The red solid and dashed lines indicate the WNPSH (5880 gpm). (e),(f) Composite anomalous wind field (vectors; m s−1), relative vorticity (shading; 10−5 s−1), and standardized anomalies of geopotential height (blue contours; solid lines for positive values and dashed lines for negative values) with respect to the climatology at 850 hPa. The black dots in (c)–(f) indicate the statistical significance of standardized anomalies of geopotential height is statistically significant at the 95% confidence level (Student’s t test). Gray shading indicates the main body of the Tibetan Plateau (≥4000 m).
Citation: Journal of Climate 38, 10; 10.1175/JCLI-D-24-0299.1
Compared with type W, the positions of the SAH and WNPSH shift more eastward for type E. At 500 hPa, the mid–high latitudes exhibit a pattern with two troughs and two ridges for type W. The upstream blocking high continuously transports cold air to the NEC-CV and further deepens the low trough, while the downstream blocking high is conducive to the stabilization of the NEC-CV over Northeast China and prevents it from moving downstream. In contrast, type E is mainly controlled by the cyclonic anomalies, indicating a higher baroclinity of the low trough and inducing the southwestward intrusion of the NEC-CVs in the mid–upper troposphere. The anomalous anticyclonic anomaly located over high latitudes is more widespread for type E. As for 850 hPa, different from the continental-scale low pressure in type W, a prominent southwest–northeast-oriented belt of low pressure from central China to the Japan Island is found for type E.
b. Features of the circulation evolution and key precursor signals
The previous section has described the key circulation configurations related to the two types of PEPEs in the YHRB and how do the key circulation systems evolve and develop in the precursor period? This section uses composite analysis to discuss the precursor circulation signals for the individual PEPEs. In the following text, day 0 represents the peak day of the PEPE and day d denotes the dth day prior to (negative) or after (positive) the peak date.
1) 200 hPa
According to the analysis in Section 4a, the occurrence and development of PEPEs are closely related to the stable divergence in the upper troposphere, in which the SAH and upper-level jet stream play important roles. During the two types of PEPEs, the upper-level jet stream is basically stable and less mobile (Fig. 10). Compared with type E, the position of the southern boundary of the upper-level jet axis is more southward for type W, while the preceding SAH shifts eastward and gradually retreats westward (Fig. 10a). Unlike the clear signal that can be detected about 10 days earlier at 500 hPa, the eastward extension of the SAH is not observed until day −2, when the WNPSH also enhances westward. For type E, the SAH and WNPSH are relatively weaker in the early stage (Fig. 10b). With the movement in opposite directions, the SAH and WNPSH eventually extend to the southern part of the YHRB and mutually configure on day 0.
Distribution of the time evolution of the WNPSH (5880 gpm) at 500 hPa, the SAH (12 520 gpm), and jet stream (u ≥ 25 m s−1) at 200 hPa for (a) type W and (b) type E. The solid lines with different colors represent the number of days from the peak date of PEPEs.
Citation: Journal of Climate 38, 10; 10.1175/JCLI-D-24-0299.1
In terms of the circulation distributions, it is found that anomalous circulation systems are intensifying in the early stage (up to 1.5 std) and reach the maximum on day −2 (1.8 std) for type W (Fig. 11). The NEC-CVs are located to the north of the upper-level jet exit during their mature stage. The air parcels decelerate while moving toward the jet exit, which produce rightward ageostrophic wind components and result in upper-level divergence to the south of the jet entrance. The meridional geopotential height gradient gradually increases with the deepening of the upper low pressure system and further prompts the acceleration of the westerly jet stream. A significant divergent area is found at 200 hPa since day −4, when the YHRB is located to the south of the jet entrance. The upper-level divergence is conducive to the development of abnormal upward motion over the YHRB. In addition, there is a “−+−+” wave train in the mid–high latitudes, facilitating the propagation of upstream Rossby wave energy to Northeast China (Hu et al. 2011; Lian et al. 2010; Nie et al. 2023; Xie and Bueh 2015). As for type E, the mid–high latitudes are controlled by broad negative anomalies (1.2 std) in the preceding stage, which also weaken the SAH. The SAH develops from day −2 onward and reaches its maximum (1.2 std) on the peak day, resulting in significant divergence over the YHRB.
Distributions of the standardized anomalies of geopotential height (contours) and wave activity flux (vectors; m2 s−2) for (first column),(second column) type W and (third column),(fourth column) type E at 200 hPa, with the green solid lines indicating the horizontal divergence (≥1 × 10−5 s−1). The shaded areas indicate the geopotential height is statistically significant at the 95% confidence level (Student’s t test).
Citation: Journal of Climate 38, 10; 10.1175/JCLI-D-24-0299.1
2) 500 hPa
The temporal evolution of the 500-hPa geopotential height is shown in Fig. 12. It is found that the precursor signal of the NEC-CV can be traced back to day −12 for type W, accompanied by a significant negative anomaly and a low trough in the eastern part of the Lake Baikal. Simultaneously, a significant positive anomaly exists in the upstream of the Baikal trough, indicating the intensification and development of the Ural blocking high (UBH) toward the polar areas. The persistence of the UBH facilitates the continuous intrusion of polar air, which is conducive to the deepening and downstream movement of the trough over the Lake Baikal. The Okhotsk blocking high (OBH) starts to develop poleward from day −9 onward and acts to slow down the upstream low trough, which also makes it difficult to move downstream. On day −6, along with the development of the OBH, a negative anomaly prevails over Northeast China, gradually forming a north–south-oriented dipole pattern and further favoring the southward transport of cold air to the YHRB. The dipole structure reaches its maximum on day −3, when the OBH has abnormally extended poleward and weakens until day −1. The existence of the blocking highs over the Ural Mountains and the Sea of Okhotsk favors the local aggregation of cold air, which provides favorable dynamic conditions for the occurrence and development of the NEC-CVs. From day −9 onward, a significant positive anomaly is detected in the western North Pacific, indicating the westward extension and intensification of the WNPSH. The ridge point of the WNPSH is the most westward on day −3 and persists to the south of the YHRB, resulting in a cyclonic wind shear in the north side of the WNPSH and favoring the maintenance of the NEC-CV. The above results confirm that the configuration of the NEC-CV, the OBH, and the WNPSH plays an important role in the occurrence and development of type W (Liang et al. 2009).
The temporal evolution of 500-hPa geopotential height fields (contours) for (first column),(second column) type W and (third column),(fourth column) type E, along with their anomalies (shading; gpm). The red solid lines denote the WNPSH (5880 gpm). The shaded areas indicate that the geopotential heights are statistically significant at the 95% confidence level (Student’s t test).
Citation: Journal of Climate 38, 10; 10.1175/JCLI-D-24-0299.1
In the preceding period, frequent activities of shortwave trough are observed for type E in the mid–low latitudes of East Asia, with cold air continuously transported eastward at the bottom of the trough. Meanwhile, the high-latitude region of Northeast Asia is controlled by a blocking situation, and the low pressure trough in the Sea of Okhotsk begins to deepen and enhance. On day −8 (results not shown), along with the arrival of the upstream shortwave trough and the collapse of northern blocking situation, a cyclonic anomaly prevails over the midlatitudes of Northeast Asia. The WNPSH is anomalously eastward, making it difficult for warm air to be transported northward. On day −3, the intensifying OBH and the upstream Lake Baikal blocking high (BBH) drive the polar cold air to the stabilized trough in the Sea of Japan, which further deepens the trough toward southwest. On peak days, the type E event is found to occur with the westward extension of the WNPSH. The configuration of the circulation pattern resembles the single-blocking mode in previous studies (Chen and Zhai 2014; Wang et al. 2017).
3) 850 hPa
The occurrence of PEPEs relies on the adequate supply of water vapor at the lower level. Accordingly, the low-level wind anomalies, total deformation, and water vapor transport are analyzed for the two types of PEPEs (Fig. 13). The total deformation (
The distribution of the 850-hPa anomalous wind field (vectors; m s−1), total deformation (red contours; >3 × 10−5 s−1), anomalies of normalized geopotential height (solid/dashed lines for positive/negative values), and water vapor flux (shading) for (first column),(second column) type W and (third column),(fourth column) type E, where wind field anomalies are only available for regions that have passed the 95% significance test (Student’s t test).
Citation: Journal of Climate 38, 10; 10.1175/JCLI-D-24-0299.1
In the early stage, it can be seen that the geopotential height field exhibits the development and adjustment of the NEC-CVs for type W, accompanied by active water vapor transport. An anomalous southwestern water vapor transport exists in the southeastern region of China, although it is not very strong. The corresponding maximum moisture convergence is located in the East China Sea or the southern part of the YHRB. On day −2, due to the anomalous southwesterly flow caused by the intensifying anticyclone (>1.4 std) over the western North Pacific, anomalous water vapor starts to be transported to the Yangtze River basin and dominates the YHRB on day 0 (>2.6 std). At this time, there is a strong cyclonic anomaly in the high latitudes of Northeast Asia (>1.8 std), which increases the meridional geopotential gradient and results in the intersection of the anomalous warm and humid airflow over the YHRB. A long wind shear line can also be found in the northern part of the anomalous moisture channel, thereby enhancing the vertical motion in the YHRB. The abnormal water vapor continues to propagate downstream to Japan on day +2. However, a strong wind shear zone can be found in the northeast of the YHRB, due to the fact that the mid–high latitudes are still under the control of the anomalous low pressure center.
In contrast to type W, there is almost no abnormal water vapor activity observed for type E in the preceding period, when the eastern part of China is dominated by a continental-scale low pressure system. From day −2 onward, with the enhancement of the WNPSH, the abnormal southwest airflow along the periphery of the WNPSH guides the warm and humid water vapor to the YHRB. On the peak day, the left boundary of the moisture channel is more westerly for type E. The anomalous moisture continues to be transported downstream and decays very quickly since day −2.
5. Conclusions
In this study, we first examine the features of precipitation occurrence related to the NEC-CVs during summer. The results show that the NEC-CVs have local and long-distance influences on the precipitation, with heavier and more frequent occurrence of rainfall found in the vicinity and the southern part of the closed lows. The contribution of NEC-CVs to related precipitation is quantified by using a relative threshold of 4× radius for each NEC-CV. The NEC-CVs are found to have a distinct geographic influence on precipitation over central-eastern China, especially for the more extreme precipitation events. For the precipitation extremes lasting for at least 2 days, the NEC-CVs contribute substantially to the total persistent extreme precipitation in the YHRB (about 35%–40%).
The PEPEs over the YHRB are detected based on the autoencoder machine learning technique. The PEPEs related to the NEC-CVs are identified by the joint SOM cluster, with 36 such events selected (account for about 27.7% of the total PEPEs). The corresponding NEC-CVs mainly travel along the middle and northerly path and the occurrence of coherent PEPEs is mainly concentrated in the first 4 days of the respective NEC-CV process.
Two distinct types of PEPEs are then classified by the mean position of each NEC-CV, namely, type W and type E. Accompanied by the anomalous moisture transporting from the western North Pacific and the South China Sea, respectively, type W and type E show different distributions of the precipitation anomalies, which highlight their importance in the refined forecast. The possible role of the NEC-CVs in the two types of PEPEs is further investigated, and a concept of the physical process is presented in Fig. 14 based on the detailed composite. Anomalous eastward extension of the SAH and southward shift of the jet axis are found at the upper levels for both types of PEPEs. The YHRB is dominated by the cyclonic anomaly due to the NEC-CVs at 500 hPa, accompanied with a large-scale low pressure at the lower level. The north–south-oriented dipole pattern over Northeast Asia is favorable for the transport of the cold dry air to the YHRB. The YHRB is located to the south of the upper-level jet axis and characterized by the lower-level convergence and the upper-level divergence, which is conducive to the enhancement of the upward motion.
Schematic diagram of the PEPEs over the YHRB under the influence of NEC-CVs for (a) type W and (b) type E, respectively. The dashed lines indicate the climatology of the SAH and WNPSH, while the solid ones indicate the observed ones.
Citation: Journal of Climate 38, 10; 10.1175/JCLI-D-24-0299.1
The diagnosis of the circulation evolution shows that the precursor signal of the NEC-CVs can be traced back to about 12/8 days before the precipitation peak for type W/type E at 500 hPa. The concurrent combinations of the preceding UBH and OBH provide favorable dynamical conditions for the maintenance of the NEC-CVs in type W, whereas the BBH is responsible for the development of NEC-CVs in type E. Compared with type E, apart from the enhanced SAH and WNPSH, there are a “−+−+” wave train and a southerly jet stream in the upper troposphere. However, the SAH in type E anomalously extends eastward on day −2, with its ridge line exceeding the farthest position of the counterpart in type W. The two types of PEPEs are both governed by the combined wind shear and intense moisture transport over the YHRB, except for a stronger, longer, and more easterly moisture channel observed in type W.
In summary, this study qualitatively demonstrates the role of NEC-CVs in influencing the PEPEs over the YHRB. These findings could enhance the comprehension of the atmospheric dynamics responsible for the PEPEs and may improve the accuracy of medium-range weather forecasts and the effectiveness of early warning mechanisms that can mitigate flood hazards and their subsequent socioeconomic repercussions. Additionally, this research might uncover novel perspectives regarding the capabilities and constraints of modeling in accurately depicting the critical synoptic-scale thermodynamic drivers of the PEPEs.
Acknowledgments.
This research was supported by the Joint Funds of the National Natural Science Foundation of China (U2142205 and U2342212), the National Key R&D Program of China (2023YFC3007700 and 2023YFC3007702), the Joint Research Project for Meteorological Capacity Improvement (24NLTSQ007), the Science and Technology Development Plan in Jilin Province of China (20230203135SF), the Special Fund for Innovative Development of China Meteorological Administration (CXFZ2022J007), the Opening Fund of Key Opening Laboratory for Northeast China Cold Vortex Research (2022SYIAEKFMS01), and the China Meteorological Administration Youth Innovation Team (CMA2024QN05).
Data availability statement.
ERA5 data can be downloaded from the Copernicus Climate Data Store (https://cds.climate.copernicus.eu) and CPC precipitation data are available from the website (https://psl.noaa.gov/data/gridded/data.cpc.globalprecip.html), last accessed May 30, 2024.
REFERENCES
Abatzoglou, J. T., 2016: Contribution of cutoff lows to precipitation across the United States. J. Appl. Meteor. Climatol., 55, 893–899, https://doi.org/10.1175/JAMC-D-15-0255.1.
Bueh, C., N. Shi, L. Ji, J. Wei, and S. Tao, 2008: Features of the EAP events on the medium-range evolution process and the mid- and high-latitude Rossby wave activities during the Meiyu period. Chin. Sci. Bull., 53, 610–623, https://doi.org/10.1007/s11434-008-0005-2.
Cao, Y., Y. Zheng, J. Sun, S. Hua, and J. Sheng, 2024: Spatiotemporal distributions and environmental characteristics of three types of regional severe convective weather processes associated with the Northeast China cold vortex (in Chinese). Acta Meteor. Sin., 82, 22–36, https://doi.org/10.11676/qxxb2024.20230029.
Chen, J.-M., T. Li, and C.-F. Shih, 2010: Tropical cyclone– and monsoon-induced rainfall variability in Taiwan. J. Climate, 23, 4107–4120, https://doi.org/10.1175/2010JCLI3355.1.
Chen, X., X. Zhuge, X. Zhang, Y. Wang, and D. Xue, 2023: Objective identification and climatic characteristics of heavy-precipitation northeastern China cold vortexes. Adv. Atmos. Sci., 40, 305–316, https://doi.org/10.1007/s00376-022-2037-y.
Chen, Y., and P. Zhai, 2013: Persistent extreme precipitation events in China during 1951–2010. Climate Res., 57, 143–155, https://doi.org/10.3354/cr01171.
Chen, Y., and P. Zhai, 2014: Two types of typical circulation pattern for persistent extreme precipitation in Central–Eastern China. Quart. J. Roy. Meteor. Soc., 140, 1467–1478, https://doi.org/10.1002/qj.2231.
Chen, Y., and P. Zhai, 2015: Synoptic-scale precursors of the East Asia/Pacific teleconnection pattern responsible for persistent extreme precipitation in the Yangtze River Valley. Quart. J. Roy. Meteor. Soc., 141, 1389–1403, https://doi.org/10.1002/qj.2448.
Chen, Y., and P. Zhai, 2016: Mechanisms for concurrent low-latitude circulation anomalies responsible for persistent extreme precipitation in the Yangtze River Valley. Climate Dyn., 47, 989–1006, https://doi.org/10.1007/s00382-015-2885-6.
Chen, Y., P. Zhai, Z. Liao, and L. Li, 2019: Persistent precipitation extremes in the Yangtze River Valley prolonged by opportune configuration among atmospheric teleconnections. Quart. J. Roy. Meteor. Soc., 145, 2603–2626, https://doi.org/10.1002/qj.3581.
Cheng, Y., L. Wang, and T. Li, 2021: Two distinct types of 10–30-day persistent heavy rainfall events over the Yangtze River Valley. J. Climate, 34, 9571–9584, https://doi.org/10.1175/JCLI-D-20-0741.1.
Dare, R. A., N. E. Davidson, and J. L. McBride, 2012: Tropical cyclone contribution to rainfall over Australia. Mon. Wea. Rev., 140, 3606–3619, https://doi.org/10.1175/MWR-D-11-00340.1.
Ding, Y., Y. Liu, and Z.-Z. Hu, 2021: The record-breaking mei-yu in 2020 and associated atmospheric circulation and tropical SST anomalies. Adv. Atmos. Sci., 38, 1980–1993, https://doi.org/10.1007/s00376-021-0361-2.
Doan, Q.-V., H. Kusaka, T. Sato, and F. Chen, 2021: S-SOM v1.0: A structural self-organizing map algorithm for weather typing. Geosci. Model Dev., 14, 2097–2111, https://doi.org/10.5194/gmd-14-2097-2021.
Gibson, P. B., S. E. Perkins-Kirkpatrick, P. Uotila, A. S. Pepler, and L. V. Alexander, 2017: On the use of self-organizing maps for studying climate extremes. J. Geophys. Res. Atmos., 122, 3891–3903, https://doi.org/10.1002/2016JD026256.
Guo, Z., and L. Wang, 2023: Impact of the “combined modality” of Silk-Road and East Asia–Pacific teleconnection patterns on the heavy precipitation in the early stage of Meiyu in the Yangtze–Huaihe River region in 2020 (in Chinese). Chin. J. Atmos. Sci., 47, 1171–1182, https://doi.org/10.3878/j.issn.1006-9895.2202.21215.
He, J., Z. Wu, Z. Jiang, C. Miao, and G. Han, 2006: “ Climate effect” of the Northeast Cold Vortex and its influences on Meiyu (in Chinese). Chin. Sci. Bull., 51, 2803–2809, https://doi.org/10.1360/csb2006-51-23-2803.
Hersbach, H., and Coauthors, 2020: The ERA5 global reanalysis. Quart. J. Roy. Meteor. Soc., 146, 1999–2049, https://doi.org/10.1002/qj.3803.
Hewitson, B. C., and R. G. Crane, 2002: Self-organizing maps: Applications to synoptic climatology. Climate Res., 22, 13–26, https://doi.org/10.3354/cr022013.
Hsieh, Y.-P., 1949: An investigation of a selected cold vortex over North America. J. Meteor. Sci., 6, 401–410, https://doi.org/10.1175/1520-0469(1949)006<0401:AIOASC>2.0.CO;2.
Hu, K., R. Lu, and D. Wang, 2011: Cold vortex over Northeast China and its climate effect (in Chinese). Chin. J. Atmos. Sci., 35, 179–191, https://doi.org/10.3878/j.issn.1006-9895.2011.01.15.
Huang, Z., X. Tan, X. Wu, X. Tan, J. Fu, and B. Liu, 2023: Long-term changes, synoptic behaviors, and future projections of large-scale anomalous precipitation events in China detected by a deep learning autoencoder. J. Climate, 36, 4133–4149, https://doi.org/10.1175/JCLI-D-22-0737.1.
Li, R. C. Y., and W. Zhou, 2015: Interdecadal changes in summertime tropical cyclone precipitation over southeast China during 1960–2009. J. Climate, 28, 1494–1509, https://doi.org/10.1175/JCLI-D-14-00246.1.
Li, Y., X. Cui, G. Li, and L. Chen, 2024: Moisture sources and contributions of source regions to the northeast cold vortex rainstorm in June 2021 (in Chinese). Chin. J. Atmos. Sci., 48, 1–15, https://doi.org/10.3878/j.issn.1006-9895.2210.22202.
Lian, Y., C. Bueh, Z. Xie, B. Shen, and S. 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., 32, 429–439, https://doi.org/10.3878/j.issn.1006-9895.2010.02.16.
Liang, H., Y. Wang, and Z. Guo, 2009: The teleconnection relationship between the northeast cold vortex and the subtropical high, the Ohotstk high in summer (in Chinese). J. Meteor. Sci., 29, 793–796, https://doi.org/10.3969/j.issn.1009-0827.2009.06.013.
Lin, A., D. Gu, D. Peng, B. Zheng, and C. Li, 2020: Definition indicators of regional persistent heavy precipitation processes with large-scale characteristics. J. Trop. Meteor., 36, 289–298, https://doi.org/10.16032/j.issn.1004-4965.2020.014.
Liu, D., L. Wang, Z. Guan, and R. Bao, 2024: Maintenance mechanism for the summertime + EAP/-SR combination pattern. Climate Dyn., 62, 6555–6578, https://doi.org/10.1007/s00382-024-07224-z.
Liu, Y., and Coauthors, 2016: Application of deep convolutional neural networks for detecting extreme weather in climate datasets. arXiv, 1605.01156v1, https://arxiv.org/abs/1605.01156.
Murakami, H., T. L. Delworth, W. F. Cooke, S. B. Kapnick, and P.-C. Hsu, 2022: Increasing frequency of anomalous precipitation events in Japan detected by a deep learning autoencoder. Earth’s Future, 10, e2021EF002481, https://doi.org/10.1029/2021EF002481.
Nie, Y., Y. Zhang, J. Zuo, M. Wang, J. Wu, and Y. Liu, 2023: Dynamical processes controlling the evolution of early-summer cut-off lows in northeast Asia. Climate Dyn., 60, 1103–1119, https://doi.org/10.1007/s00382-022-06371-5.
Racah, E., C. Beckham, T. Maharaj, S. E. Kahou, Prabhat, and C. Pal, 2017: Extreme weather: A large-scale climate dataset for semi-supervised detection, localization, and understanding of extreme weather events. Proc. 31st Int. Conf. on Neural Information Processing Systems, Long Beach, CA, ACM, 3405–3416, https://dl.acm.org/doi/10.5555/3294996.3295099.
Ren, X., D. Yang, and X.-Q. Yang, 2015: Characteristics and mechanisms of the subseasonal eastward extension of the South Asian high. J. Climate, 28, 6799–6822, https://doi.org/10.1175/JCLI-D-14-00682.1.
Shi, C., and P. Zhai, 2024: Changes in climatic features of Northeast China Cold Vortex as reflected by ERA5 and CRA-40. Atmos. Res., 300, 107233, https://doi.org/10.1016/j.atmosres.2024.107233.
Shi, C., Y. Lian, X. Yang, D. Fu, B. Shen, S. Li, and G. Liu, 2020: The relationship between winter cut-off cold vortexes in northeast Asia and Northern Hemisphere as well as their connections with extreme low temperature events in China (in Chinese). Acta. Meteor. Sin., 78, 778–795, https://doi.org/10.11676/qxxb2020.049.
Shi, C., P. M. Zhai, and Y. Lian, 2023: Advances in research of upper level cut-off cold vortex (in Chinese). Meteor. Mon., 49, 513–524, https://doi.org/10.7519/j.issn.1000-0526.2023.021501.
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, X., R. Jin, T. Xiao, N. Yang, and W. Wei, 2020: Statistical characteristics of Asian blocking activity during the Yangtze-Huaihe Meiyu season (in Chinese). Acta Meteor. Sin., 78, 580–592, https://doi.org/10.11676/qxxb2020.043.
Takaya, K., and H. Nakamura, 2001: A formulation of a phase-independent wave-activity flux for stationary and migratory quasigeostrophic eddies on a zonally varying basic flow. J. Atmos. Sci., 58, 608–627, https://doi.org/10.1175/1520-0469(2001)058<0608:AFOAPI>2.0.CO;2.
Trenberth, K. E., A. Dai, R. M. Rasmussen, and D. B. Parsons, 2003: The changing character of precipitation. Bull. Amer. Meteor. Soc., 84, 1205–1218, https://doi.org/10.1175/BAMS-84-9-1205.
UN Office for Disaster Risk Reduction, 2020: The human cost of disasters—An overview of the last 20 years (2000-2019). Accessed 27 May 2024, https://reliefweb.int/report/world/human-cost-disasters-overview-last-20-years-2000-2019.
Wang, L.-j., J. H. He, D. Si, M. Wen, and S. Zhong, 2010: Analysis of impacts of northeast Cold Vortex processes on Meiyu rainfall period over Yangtze-Huaihe River Basin (in Chinese). Trans. Atmos. Sci., 33, 89–97, https://doi.org/10.3969/j.issn.1674-7097.2010.01.012.
Wang, X., Y. Ding, and Q. Zhang, 2017: Circulation pattern and moisture transport for summertime persistent heavy precipitation in eastern China (in Chinese). Climatic Environ. Res., 22, 221–230, https://doi.org/10.3878/j.issn.1006-9585.2016.16056.
Wei, W., R. Zhang, M. Wen, X. Rong, and T. Li, 2014: Impact of Indian summer monsoon on the South Asian High and its influence on summer rainfall over China. Climate Dyn., 43, 1257–1269, https://doi.org/10.1007/s00382-013-1938-y.
Wu, J., and X.-J. Gao, 2013: A gridded daily observation dataset over China region and comparison with the other datasets (in Chinese). Chin. J. Geophys., 56, 1102–1111, https://doi.org/10.6038/cjg20130406.
Xie, P., M. Chen, and W. Shi, 2010: CPC unified gauge-based analysis of global daily precipitation. 24th Conf. on Hydrology, Atlanta, GA, Amer. Meteor. Soc., 2.3A, https://ams.confex.com/ams/90annual/webprogram/Paper163676.html.
Xie, Z., and C. Bueh, 2015: Different types of cold vortex circulations over northeast China and their weather impacts. Mon. Wea. Rev., 143, 845–863, https://doi.org/10.1175/MWR-D-14-00192.1.
Xu, S., and L. Qi, 2023: Critical influence of the Northeast Cold vortex in different positions on precipitation. Climate Dyn., 60, 867–881, https://doi.org/10.1007/s00382-022-06365-3.
Zhai, P. M., L. Li, B. Zhou, and Y. Chen, 2016: Progress on mechanism and prediction methods for persistent extreme precipitation in the Yangtze-Huai River Valley (in Chinese). J. Appl. Meteor. Sci., 27, 631–640, https://doi.org/10.11898/1001-7313.20160511.
Zhang, F., T. Chen, F. Zhang, X. Shen, and Y. Lan, 2020: Extreme features of severe precipitation in Meiyu period over the middle and lower reaches of Yangtze River Basin in June-July 2020 (in Chinese). Meteor. Mon., 11, 1405–1414, https://doi.org/10.7519/j.issn.1000-0526.2020.11.002.
Zhang, P., Y. Liu, and B. He, 2016: Impact of East Asian summer monsoon heating on the interannual variation of the South Asian high. J. Climate, 29, 159–173, https://doi.org/10.1175/JCLI-D-15-0118.1.
Zhao, W., and S. Du, 2016: Learning multiscale and deep representations for classifying remotely sensed imagery. ISPRS J. Photogramm. Remote Sens., 113, 155–165, https://doi.org/10.1016/j.isprsjprs.2016.01.004.
Zhou, B., P. Zhai, and Y. Chen, 2020: Contribution of changes in synoptic‐scale circulation patterns to the past summer precipitation regime shift in eastern China. Geophys. Res. Lett., 47, e2020GL087728, https://doi.org/10.1029/2020GL087728.
