NO.5

615

BUEH Cholaw, LI Yan, LIN Dawei, et al.

Interannual Variability of Summer Rainfall over the Northern Part of China and the Related Circulation Features




BUEH Cholaw1∗ (

), LI Yan2 (

 ), and LIAN Yi ( ü)

), LIN Dawei1,3 (

4,5

1 International Center for Climate and Environment Sciences, Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing 100029 2 College of Atmosphere Science, Chengdu University of Information Technology, Chengdu 610225 3 University of Chinese Academy of Sciences, Beijing 100049 4 Laboratory of Research for Middle–High Latitude Circulation and East Asian Monsoon, Changchun 130062 5 Institute of Meteorological Sciences of Jilin Province, Changchun 130062 (Received December 22, 2015; in final form May 20, 2016)

ABSTRACT In this study, interannual variability of summer rainfall over the northern part of China (NPC) and associated circulation patterns were investigated by using long-term (1961–2013) observational and reanalysis data. Two important NPC rainfall modes were identified by empirical orthogonal function analysis: the first is characterized by an almost uniformly distributed rainfall anomaly over most parts of the NPC, while the second shows rainfall variability in Northeast China (NEC) and its out-of-phase relationship with that in North China (NC) and the northern part of Northwest China. The results also suggest that the NPC summer rainfall anomalies are also closely associated with those in some other parts of China. It is revealed that the circumglobal teleconnection pattern associated with the anomalous Indian summer monsoon (ISM) and the Polar/Eurasia (PEA) pattern work in concert to constitute the typical circulation pattern of the first rainfall mode. The cooperative engagement of the anomalous ISM circulation and the PEA pattern is fundamental in transporting water vapor to the NPC. The study emphasizes that the PEA pattern is essential for the water vapor transport to the NPC through the anomalous midlatitude westerly. In the second NPC rainfall mode, the typical circulation pattern is characterized by the anomalous surface Okhotsk high and the attendant lower tropospheric circulation anomaly over NEC. The circulation anomaly over NEC leads to a redistribution of water vapor fluxes over the NPC and constitutes an out-of-phase relationship between the rainfall anomalies over NEC and NC. Key words: summer rainfall, northern part of China, circumglobal teleconnection pattern, Polar/Eurasia pattern, water vapor transport Citation: Bueh Cholaw, Li Yan, Lin Dawei, et al., 2016: Interannual variability of summer rainfall over the northern part of China and the related circulation features. J. Meteor. Res., 30(5), 615– 630, doi: 10.1007/s13351-016-5111-5.

1. Introduction The northern part of China (NPC) is an interlaced zone of agriculture, animal husbandry, and forestry, covering Northeast China (NEC), North China (NC), and Northwest China (NWC). Affected by the arid, semi-arid, and semi-humid climates from west to east, the ecological zone in the NPC is particularly sensitive to the interannual variability of summer

rainfall. As is well known, summer rainfall over the NPC is primarily controlled by the midlatitude circulation system, and it is also influenced in part by the East Asian summer monsoon (EASM) circulation due to its geographical location at the northern fringe of the EASM region. Undoubtedly, understanding the interannual variability of summer rainfall in the NPC is essential for improving its prediction. A number of studies have revealed that the inter-

Supported by the National Natural Science Foundation of China (41375064 and 41630424) and National Science and Technology Support Program of China (2015BAC03B03). ∗ Corresponding author: [email protected]. ©The Chinese Meteorological Society and Springer-Verlag Berlin Heidelberg 2016

616

JOURNAL OF METEOROLOGICAL RESEARCH

annual variability of summer rainfall over the eastern part of China is closely associated with a meridionally elongated circulation pattern, i.e., the East Asia/Pacific (EAP) teleconnection pattern (Huang and Li, 1987; Nitta, 1987; Huang and Sun, 1992; Lu, 2004; Bueh et al., 2008). This pattern primarily reflects variations of three key circulation systems: the western Pacific subtropical high (WPSH), the Okhotsk/Yakutsk high over Northeast Asia, and the trough between them (the Meiyu trough). If these three systems are stably strengthened (weakened) simultaneously, above-normal (below-normal) precipitation occurs over the middle and lower reaches of the Yangtze River and below-normal (above-normal) precipitation occurs over North and South China. The summer rainfall over North China is associated with not only the EAP pattern but also a zonally oriented midlatitude wave train pattern over Eurasia (Zhang and Tao, 1998; Liang et al., 2011). Liang et al. (2011) found that increases (decreases) in rainfall over NC are associated with a Rossby wave train pattern along the midlatitude westerly over Eurasia, in which a cyclonic (anticyclonic) anomaly center is located over the southern side of Lake Baikal and an anticyclonic (cyclonic) anomaly anchored around the Korean Peninsula. Among the several anomaly centers, the anticyclonic (cyclonic) anomaly center around the Sea of Japan favors (disfavors) a westward and northward extension of the WPSH, corresponding to increased (decreased) rainfall over NC. Researchers have also revealed that there is a significant positive correlation between the summer rainfall anomalies over NC and India (Guo, 1992; Kripalani and Singh, 1993; Kripalani et al., 1997; Zhang, 1999; Kripalani and Kulkarni, 2001; Ding and Wang, 2005; Liu and Ding, 2008; Lin et al., 2016). It has been realized that the summertime circumglobal teleconnection (CGT) pattern along the Asian–African westerly jet acts as a bridge to link the Indian summer monsoon (ISM) and NC rainfall (Lu et al., 2002; Ding and Wang, 2005; Liu and Ding, 2008; Lin et al., 2016). The variation of summer rainfall over NEC and the associated circulation features have been analyzed by several previous studies (e.g., Lian and An, 1998;

VOL.30

Sun et al., 2002). Sun et al. (2002) pointed out that the mid-tropospheric circulation in wet years over NEC is characterized by a negative height anomaly over NEC and a positive height anomaly over central and western Siberia; whereas in drought years, these two height anomalies almost reverse in sign. They also emphasized the association of NEC summer rainfall with the EASM and ISM. By comparing the summer circulation patterns in two flooding years, i.e., 1954 and 1998, of the Yangtze River valley, Yao and Dong (2000) noted a close link between NEC rainfall and blocking-type circulation over the Okhotsk/Yakutsk region. In the summer of 1998, the blocking-type circulation was situated over the Okhotsk/Yakutsk region and supplied persistent and moderate cold air to NEC. As a consequence, the interaction of the cold air brought by the blocking-type circulation and the warm and wet air transported by the EASM caused serious flooding over NEC. In contrast, in the summer of 1954, the blocking-type circulation was located to the south of Okhotsk/Yakutsk, covering part of NEC, and caused drought conditions over NEC, despite the EASM flow being similar to that in the summer of 1998. Liu et al. (2010) also revealed similar circulation features associated with NEC summer rainfall anomalies and particularly emphasized the semi-annual signals involved in the variation of NEC summer rainfall. The interannual variability of NWC summer rainfall has also been investigated in a number of studies (e.g., Li et al., 1997; Shi et al., 2003; Yang and Zhang, 2007; Chen and Dai, 2009; Yang et al., 2009). There are two dominant summer rainfall patterns in NWC: a uniformly distributed rainfall anomaly pattern, and a see-saw rainfall pattern with an east–west orientation (Li et al., 1997; Chen and Dai, 2009). Based on an empirical orthogonal function (EOF) analysis, Chen and Dai (2009) found that the leading EOF pattern of NWC summer rainfall is characterized by a see–saw anomaly distribution, with the corresponding mid-tropospheric circulation anomalies showing a north–south dipole anomaly pattern over central Asia and western Siberia. Specifically, in years of more (less) rainfall in western NWC and less (more) rainfall in eastern NWC, the mid-tropospheric circulation upstream of NWC features a negative (positive)

NO.5

BUEH Cholaw, LI Yan, LIN Dawei, et al.

height anomaly center over central Asia and a positive (negative) height anomaly center over western Siberia. Compared to that of the leading EOF pattern, the time series of the second EOF pattern (uniformly distributed) also shows a strong interannual variability (Li et al., 1997; Chen and Dai, 2009). Additionally, Yang and Zhang (2007) studied the variation of western NWC (Xinjiang Region) rainfall and its associated circulation features. They revealed that increased (decreased) rainfall in western NWC is locally caused by an anomalous cyclonic (anticyclonic) circulation over central Asia that corresponds to an active (inactive) central Asian cut-off vortex circulation, which is consistent with the results of Chen and Dai (2009). Yang et al. (2009) also reported that western NWC summer rainfall is negatively correlated with ISM rainfall. These findings suggest that the recent weakening of the ISM (Wu, 2005) might be linked with increased (decreased) rainfall in western (eastern) NWC (Chen and Dai, 2009). Though the variation of summer rainfall over the NPC has been extensively studied, thus far, most of these studies have been conducted separately over different sections of the NPC. However, due to the two following facts, it is worthwhile taking into account the summer rainfall variability over the NPC as a whole: (1) the circulation anomalies associated with summer rainfall anomalies over NC, NEC, and NWC are often closely linked with one another; thus, they may be considered as a coherent variability of the Northern Hemispheric circulation; and (2) as presented previously, the summer rainfall anomalies in these three sub-regions (NC, NEC, and NWC) bear a close relationship with the variation of the ISM. Therefore, in the present study, we examine the summer rainfall variability over the NPC as a whole, with the aim to reveal the typical circulation anomalies across a larger horizontal scale. Following this introduction, a brief description of the datasets and methods is given in Section 2; the different rainfall modes for the NPC identified by EOF analysis are described in Section 3; the typical circulation features and related water vapor conditions for the rainfall modes of the NPC are presented in Section 4; and a summary and discussion are provided in Section 5.

617

2. Data and methods We used a suite of high-quality gridded monthly precipitation data over China (1961–2013) archived by the China Meteorological Administration. The data have a spatial resolution of 0.5◦ × 0.5◦ and are based on the interpolation of observations from 2474 stations over the whole China (Zhao et al., 2014). We also utilized NCEP–NCAR reanalysis data (Kalnay et al., 1996), including geopotential height, horizontal wind, and specific humidity. These data are available on horizontal grids with a resolution of 2.5◦ × 2.5◦ and 17 vertical pressure levels, spanning from 1961 to 2013. We also used sea level pressure (SLP) and surface pressure fields with the same horizontal resolution and time period. The vertically integrated (from surface to 300 hPa) water vapor content (WVC; precipitable water) and water vapor flux (Q) were calculated as in Su and Feng (2014):  ps 1 qdp, gρ  ps 300 1 Q= qV dp, gρ 300

WVC =

(1) (2)

where p and ps are pressure and surface pressure, q is specific humidity, ρ is liquid water density, g is gravity acceleration, and V is horizontal wind. A standardized ISM index was defined as the summer (June–July–August) mean rainfall over all of India (Parthasarathy et al., 1994) to represent the ISM’s intensity, as in Lin et al. (2016). An index of the Polar/Eurasia (PEA) pattern (Barnston and Livezey, 1987) in summer was defined by using the monthly index of the PEA pattern archived by the Climate Prediction Center, NCEP, NOAA, USA (http://www.cpc.ncep.noaa.gov/data/teledoc/telecontents.shtml). To focus on year-to-year variations, throughout the paper, the interannual components of variables (i.e., precipitation, geopotential height, wind, water vapor flux, and WVC) and indices (i.e., the ISM and PEA indices) are used. The long-term trends and decadal variations with periods longer than 9 yr were removed by using a 9-point quadratic polynomial fit

618

JOURNAL OF METEOROLOGICAL RESEARCH

method (Gorry, 1990), in which “9-point” corresponds to “9 yr” and all weights were obtained based on least square approximation. An EOF analysis (Jolliffe, 1986) was performed based on the covariance matrix of summer precipitation data in the NPC domain. Finally, we used composite analysis to derive the rainfall and circulation anomaly fields. An “anomaly” was defined relative to the corresponding multiyear mean (1961–2013) quantity. We employed the Student’s t-test (Wilks, 1995) to assess the statistical significance of the results in the composite and linear correlation (or regression) analyses. 3. Rainfall modes The summer rainfall modes for the NPC were obtained with an EOF analysis based on the summer mean precipitation data (1961–2013) over the NPC domain. The NPC domain was chosen as the part of China to the north of 36◦ N, but the northwestern portion of the Tibetan Plateau was excluded (see Fig. 1a). To identify the rainfall modes on the interannual timescale, we only retained the interannual component of the precipitation data at each gridpoint of the NPC for the EOF analysis, as introduced in Section 2. The two leading EOF modes account for 16.3% and 10.5% of the total variance, respectively, and they are statistically distinct from each other, and from the remaining EOF modes, according to the criterion of North et al. (1982). Figure 1 shows the first and second EOF patterns, the time series of the corresponding principal components (PCs; standardized), and the correlation coefficients between the two PCs and the rainfall over all of China. The first EOF pattern reflects an almost uniform distribution of rainfall anomalies over most parts of the NPC, though opposite anomalies are observed over the southwestern part and eastern ends of the NPC (Fig. 1a). Specifically, the NC rainfall anomaly shows a uniformly distributed pattern, while rainfall anomalies with opposite sign are observed in southwestern NWC and the marginal areas of NEC. In this mode, the rainfall anomaly pattern over the NWC bears a resemblance to the leading mode of the NWC

VOL.30

summer rainfall identified by Chen and Dai (2009). The correlation field between the first PC time series (Fig. 1b) and the rainfall over China resembles the first EOF pattern over the NPC (Figs. 1a and 1c). Meanwhile, it is noteworthy that the first PC has a significant positive correlation with the summer rainfall over southwestern China and the middle reaches of the Yangtze River, as it is also negatively correlated with the rainfall over southeastern China and the Huaihe River valley (Fig. 1c). In the second EOF pattern (Fig. 1d), a prominent rainfall anomaly center is situated over the whole of NEC, and is out-of-phase with the rainfall anomalies over NC and North Xinjiang. Besides, the rainfall anomaly over NEC also shows an in-phase relationship with that of southern Xinjiang. These features are even more clearly indicated in Fig. 1f. The NEC rainfall anomaly pattern illustrated in Fig. 1f is similar to the leading rotated EOF pattern of the NEC summer rainfall identified by Sun et al. (2002; Fig. 3), reminiscent of the influences of the cold vortex circulation system. It is also apparent from Fig. 1f that the second PC has a significant positive correlation with the summer rainfall over southern and southwestern China. To examine whether the two rainfall modes for the NPC show symmetry between positive and negative polarities, we performed composite analyses. The years in the composite analysis were selected according to the two following criteria: (1) the amplitudes (absolute values) of PCs (standardized) were greater than 1; and (2) in the same year, the difference between amplitudes of the first and second PCs was greater than 0.5. The second criterion was adopted to exclude some cases with projection of both EOF patterns. For brevity, hereafter, we use the term “positive anomaly year” (“negative anomaly year”) to represent years with PCs larger than +1 (smaller than –1). The positive and negative anomaly years are listed in Table 1, and they are also indicated by solid and open dots, respectively, in Figs. 1b and 1e. The composite analyses described throughout the paper were performed in this manner. Figure 2 shows the composite rainfall anomalies over China as percentages in the positive and negative anomaly years for the two NPC rainfall

NO.5

BUEH Cholaw, LI Yan, LIN Dawei, et al.

modes. As shown in Figs. 2a and 2b, the NPC rainfall anomalies in the positive and negative anomaly years are nearly a mirror image of one another, reflecting symmetry of the first rainfall mode in its positive and negative polarities. Moreover, the rainfall anomalies over the Huaihe River valley, southern Tibet, and Guangxi Region in Fig. 2a are also shown in Fig. 2b with reversed signs. For the second NPC rainfall mode, the rainfall anomalies over NEC,

619

NC, and NWC in the positive phase (Fig. 2c) could be considered as mirror images of those in the negative phase (Fig. 2d). Meanwhile, the rainfall anomalies over southern China also show symmetry between the positive and negative phases. As mentioned in Section 1, when identifying the summer rainfall patterns over the NPC, most previous studies performed these determinations regionally and separately over NC, NWC, and NEC (Li et al., 1997;

Fig. 1. (a) The first EOF pattern (arbitrary units) of summer rainfall over the NPC, (b) the standardized time series of the first EOF, and (c) the correlation coefficients between the time series of the first EOF and the rainfall over China. (d–f) As in (a–c), but for the second EOF pattern. Solid and open dots in (b, e) indicate the positive and negative anomaly years in the composite analysis. Isolines in (c, f) are drawn at intervals of 0.15. Positive (negative) values are indicated by green solid (yellow dashed) lines and zero lines are drawn in gray. Heavy (light) shadings in (c, f) mark the regions where the correlation exceeds the 95% (90%) confidence level.

620

JOURNAL OF METEOROLOGICAL RESEARCH

VOL.30

Fig. 2. Composite rainfall anomalies (%) over China in (a, c) positive and (b, d) negative anomaly years of the (a, b) first and (c, d) second NPC rainfall modes. Intervals are drawn every 7%. Positive (negative) values are indicated by solid (dashed) lines and zero lines are drawn in gray. Heavy (light) shadings mark the regions where the composite anomalies are significant at the 95% (90%) confidence level.

Sun et al., 2002; Liu and Ding, 2008; Chen and Dai, 2009; Lin et al., 2016). In fact, as inferred from Figs. 1 and 2 of the present study, the rainfall variabilities over NC, NWC, and NEC could be taken into account in the framework of the rainfall variability for the whole of the NPC. Furthermore, a benefit of this approach is that we can unify the related circulation patterns as a whole from a large-scale perspective, which will be discussed in the next section. In addition, the information indicated in Figs. 1c, 1f, and 2 suggests that the NPC rainfall anomalies are closely associated with those in some other parts of China, demonstrating the fact that they may serve as a reference. 4. Typical circulation patterns 4.1 First NPC rainfall mode Figure 3 displays the composite fields of the 300and 500-hPa geopotential height (Z300 and Z500)

anomalies in the positive and negative anomaly years of the first NPC rainfall mode. The height anomaly patterns in the mid and upper troposphere basically show a quasi-geostrophic structure. It is interesting to note from Fig. 3a that the Z300 anomalies over the subtropical and midlatitude portions of the Northern Hemisphere resemble the CGT pattern defined by Ding and Wang (2005). In particular, over the Asian Continent, the two positive anomaly centers over West/Central Asia and over the Sea of Japan are almost identical to those of the CGT pattern (Ding and Wang, 2005; Fig. 1). They revealed that the boreal summer CGT is closely associated with the strength of the ISM, and further suggested that the ISM heat source might be instrumental in maintaining the CGT through interacting with the global wave train. To clarify the association of these two positive anomaly centers with the ISM, we plotted the regressed anomalies of Z300 in summer against the sum-

NO.5

621

BUEH Cholaw, LI Yan, LIN Dawei, et al.

Table 1. Positive and negative anomaly years for the two NPC rainfall modes Positive anomaly years Negative anomaly years

First EOF 1964, 1973, 1984, 1990, 2008, and 2012 1965, 1968, 1972, 1980, 1989, 1997, and 2010

Second EOF 1977, 1981, 1991, 2002, 2006, and 2009 1970, 1992, 1995, 2000, 2004, and 2007

Fig. 3. Composite (a, c) Z300 and (b, d) Z500 anomalies (gpm) in (a, b) positive and (c, d) negative anomaly years of the first NPC rainfall mode. Contours are drawn at intervals of 6 gpm, and shadings are as in Fig. 2. Positive (negative) values are indicated by solid (dashed) lines and zero lines are eliminated. The lowest point in each panel is drawn at 20◦ N, 90◦ E.

mer ISM index at its unit standard deviation (+1σ) (Fig. 4a). It is seen that these two positive anomaly centers are indeed a reflection of the intensity of the ISM. Specifically, Figs. 3a and 3b illustrate a strong ISM, while Figs. 3c and 3d indicate a weak ISM. However, by comparing Figs. 3a and 4a, it is seen that the circulation features associated with the first

NPC rainfall mode cannot be completely explained by the ISM alone, as the significant anomalies over the mid- and high-latitude portions of the Eurasian continent shown in Figs. 3a and 3b are absent in Fig. 4a. In Fig. 3a, a significant positive anomaly center is situated to the northwest of Lake Baikal, straddling the West Siberian Plain and the central Siberian Plateau,

622

JOURNAL OF METEOROLOGICAL RESEARCH

and two negative anomaly centers are located over Mongolia/northern China and eastern Europe, to the southeast and southwest of the above-mentioned positive anomaly center. These three anomaly centers are reminiscent of the PEA pattern defined by Barnston and Livezey (1987). They found that the PEA pattern primarily represents the fluctuations of the circumpolar vortex circulation, with the positive (negative) phase reflecting an enhanced (weakened) circumpolar vortex, and the associated circulation anomalies over midlatitude Eurasia. To further examine the association of the PEA pattern with the circulation anomalies corresponding to the first NPC rainfall mode, the regressed anomalies of Z300 in summer against the summer PEA index at its +1σ are shown in Fig. 4b. It is clear that the three anomaly centers over the midand high-latitude portions of the Eurasian continent in Figs. 3a and 3b can be explained by the PEA pattern in its negative phase (PEA− ; Fig. 4b). More specifically, in the positive anomaly years of the first NPC rainfall mode, the polar vortex is weakened over the Barents/Kara/Laptev seas and the neighboring subarctic regions, and concurs with the negative height anomaly centers over Mongolia/northern China and Europe. In the negative anomaly years of the first NPC rainfall mode, these circulation anomalies are reversed in sign (Figs. 3c, 3d, and 4b).

VOL.30

Therefore, these findings suggest that a strong (weak) ISM and a PEA− (PEA+ ) pattern work in concert to constitute the typical circulation anomalies of the positive (negative) phase of the first NPC rainfall mode. The relationship between the ISM and the NC rainfall anomaly, as indicated in Figs. 2 and 3, is consistent with the findings of Liu and Ding (2008) and Lin et al. (2016). The circulation anomalies over Central/West Asia and eastern Europe (Fig. 3), which are associated with the see–saw rainfall anomalies over NWC (Fig. 2), also resemble those in Chen and Dai (2009). Again, these findings suggest that, on the interannual timescale, the boreal NWC rainfall pattern also has a connection with the ISM. However, thus far, the importance of the PEA pattern on rainfall over the NPC has not previously been reported in literatures. It seems that the finding about the impact of the PEA pattern on the NPC rainfall anomalies in the present study can be ascribed to the consideration of the rainfall variation over the NPC as a whole. As is well known, summer rainfall over the NPC is closely associated with the local WVC in the tropospheric column, which is equivalent to precipitable water. Importantly, the WVC over the NPC is influenced not only by the water vapor transport of the midlatitude westerly flow, but also the subtropical or tropical monsoonal circulation in summer. Therefore,

Fig. 4. Regressed Z300 anomalies against the standardized (a) ISM and (b) PEA indices at unit standard deviation. Contours are drawn at intervals of 5 gpm. Positive (negative) values are indicated by solid (dashed) lines and zero lines are eliminated. Shadings are as in Fig. 1c. The lowest point in each panel is drawn at 20◦ N, 90◦ E.

NO.5

BUEH Cholaw, LI Yan, LIN Dawei, et al.

it is necessary to examine how the first NPC rainfall mode is associated with the water vapor transport of the midlatitude circulation system and the summer monsoonal circulation. Figure 5 shows the vertically integrated water vapor flux anomalies and WVC anomalies for positive and negative anomaly years of the first NPC rainfall mode. In positive anomaly years of the first NPC rainfall mode, the water vapor transport to the NPC, as indicated by the vertically integrated water vapor flux anomalies, is apparently strengthened by both the EASM circulation and the midlatitude westerly flow (Fig. 5a), resulting in an extensive rainfall increase (Fig. 2a). The EASM circulation is enhanced, as indicated by the northward displacement of the WPSH, and the anomalous southerly at the western flank of the WPSH is conducive to an above-normal water vapor condition over NC and NEC. The strengthened westerly over the subtropical northern Indian Ocean

623

also indicates a strong ISM circulation. These findings are consistent with the synchronous variations of the ISM and EASM and their relationship to NC summer rainfall variation (Zhang, 1999; Liu and Ding, 2008). In fact, the impact of the Asian summer monsoon circulation on NC and NEC rainfall has been extensively investigated (e.g., Guo, 1992; Kripalani and Singh, 1993; Kripalani et al., 1997; Zhang, 1999; Kripalani and Kulkarni, 2001; Sun et al., 2003; Ding and Wang, 2005; Liu and Ding, 2008; Lin et al., 2016). However, as seen in Figs. 5a and 5c, if one considers the NPC rainfall anomalies as a whole, the midlatitude zonal wind anomaly to the south of Lake Baikal is also a very important contributor to the water vapor transport in the NPC. Specifically, as shown in Fig. 5a, the enhanced westerly in the southern flank of the cyclonic circulation to the southeastern side of Lake Baikal corresponds to strengthened water vapor transport to the NPC. Notably, it is just the cyclonic

Fig. 5. Vertically integrated (a, c) water vapor flux (arrows; kg s−1 m−1 ) and the anomalous divergence of water vapor flux (color shading; 10−4 g s−1 cm−2 ) and (b, d) WVC (mm) anomalies in (a, b) positive and (c, d) negative anomaly years of the first NPC rainfall mode. In (b, d), intervals are 0.08 mm, positive (negative) values are indicated by red solid (blue dashed) lines, and zero lines are drawn in black. Shadings are as in Fig. 2.

624

JOURNAL OF METEOROLOGICAL RESEARCH

circulation to the southeastern side of Lake Baikal that acts as an anomaly center of the PEA− pattern (Fig. 4b). Though the Z300 anomalies associated with a strong ISM (Fig. 4a) also have a negative anomaly center between the two primary positive anomaly centers over the Asian continent, this center is weak and situated to the southwest of Lake Baikal, largely unconnected with the cyclonic circulation shown in Fig. 5a. These findings suggest that the PEA− pattern is of great importance in transporting water vapor to the NPC. Corresponding to the enhancement of water vapor transport to the NPC, the WVC over most parts of the NPC increases (Fig. 5b), particularly over NEC and eastern NC. As indicated by the color shading in Fig. 5a, convergence of water vapor flux anomalies over the NEC, eastern NC, and northern Xinjiang, reflects the fact that precipitation exceeds evaporation locally (Su and Feng, 2014). The WVC decreases over southern NWC and western NC (Fig. 5b) and matches well with the local rainfall decrease (Fig. 2a) and local divergence of water vapor flux anomalies (Fig. 5a). In negative anomaly years of the first NPC rainfall mode, the situation is almost reversed compared to that in positive anomaly years. The water vapor transport to the NPC is markedly reduced due to the weakening of both the midlatitude westerly flow and the EASM circulation (Fig. 5c), resulting in decreased rainfall there. In addition, the PEA+ pattern is closely associated with reduced water vapor transport to the NPC through its anticyclonic circulation anomaly to the southeastern side of Lake Baikal. The WVC over the NPC decreases (Fig. 5d), particularly over NEC and eastern NC, consistent with the local precipitation deficit (Fig. 2b) and divergence of water vapor fluxes (color shading in Fig. 5c). The divergence of water vapor fluxes reflects the observation that evaporation exceeds precipitation locally. To provide further evidence regarding the important role played by the PEA pattern in the NPC summer rainfall anomalies, we plotted the regressed rainfall anomalies against the PEA index (reversed sign) (Fig. 6). It is seen that the NPC rainfall anomalies in Fig. 6 resemble those in Figs. 1a, 1c, and 2a. It can be inferred that the NPC rainfall anomalies in the

VOL.30

positive phase of the PEA pattern would be similar to those in Fig. 2b. At its unit standard deviation, the PEA can explain up to 13% of the rainfall anomaly over NEC and up to 9% over NC and northern Xinjiang. This suggests that the PEA pattern does play a crucial role in the occurrence of extensive rainfall anomalies over the NPC. As for the rainfall over the NPC, in comparison, the influence of the CGT pattern associated with the anomalous ISM circulation is mainly limited to NC, rather than the NPC as a whole (Guo, 1992; Kripalani and Singh, 1993; Liu and Ding, 2008; Lin et al., 2016). 4.2 Second NPC rainfall mode Figure 7 shows the composite Z500 and SLP anomalies in the positive and negative anomaly years for the second NPC rainfall mode. The Z300 anomalies are similar to the Z500 anomalies in structure (figure omitted). In positive anomaly years (Figs. 6a and 6b), the most pronounced positive anomaly center is located around the Sea of Okhotsk, which is close to NEC. In fact, the close relationship between NEC summer rainfall and the surface Okhotsk high (OKSH) has been reported in a number of previous studies (e.g., Wang, 1992; Lian and An, 1998; Yao and Dong, 2000; Liu et al., 2010; Xie and Bueh, 2016). It has been recognized that the stronger the OKSH, the cooler and

Fig. 6. Regressed summer rainfall anomalies (%) against the standardized PEA index at unit standard deviation. For comparison, signs of regression anomalies are reversed. Intervals are drawn every 2.5%. Positive (negative) values are indicated by solid (dashed) line and zero lines are drawn in gray. Shadings are as in Fig. 1c.

NO.5

BUEH Cholaw, LI Yan, LIN Dawei, et al.

625

Fig. 7. Composite (a, c) Z500 anomalies (gpm) and (b, d) SLP anomalies (hPa) in (a, b) positive and (c, d) negative anomaly years of the second NPC rainfall mode. Isolines are drawn at intervals of 5 gpm in (a, c), and 0.4 hPa in (b, d). Positive (negative) values are indicated by solid (dashed) lines and zero lines are eliminated. Shadings are as in Fig. 2. The black boxes in (b, d) indicate the domain of the Sea of Okhotsk, which are used in calculating the OKSH index. The lowest point in each panel is drawn at 20◦ N, 90◦ E.

wetter the NEC region, and vice versa (e.g., Lian and An, 1998; Xie and Bueh, 2016). As seen from Fig. 7b, a strong OKSH is accompanied by a deepened low pressure system over NEC and Japan, consistent with the local rainfall increase (Fig. 2c). On the other hand, a weak OKSH is related to a higher SLP over NEC and Japan (Fig. 7d), giving rise to reduced rainfall locally (Fig. 2d). It is also indicated in Figs. 7c and 7d that the anomaly center around the Sea of Okhotsk has a baroclinic structure vertically. Against

the background of strong land–sea thermal contrast in summer, such a baroclinic structure is necessary, in that the mid-tropospheric anomaly can play an effective role in the formation of an anomalous OKSH (Nakamura and Fukamachi, 2004). In comparison, as shown in Figs. 7a and 7b, the strong OKSH center and the corresponding Z500 anomaly center are almost in the same location. Despite this, the OKSH center is oriented in the southwest–northeast direction, but the Z500 anomaly center is oriented in the northwest–sout-

626

JOURNAL OF METEOROLOGICAL RESEARCH

heast direction, still reflecting a baroclinic structure. As demonstrated in Fig. 7a, the blocking-type circulation around the Sea of Okhotsk is also accompanied by a positive anomaly center over eastern Europe. These two anomaly centers straddle the Eurasian continent, showing a “double blocking” structure, but it is slightly different from the “double blocking” centers during the Meiyu season. As compared to those during the Meiyu season, the two anomaly centers in Fig. 7a are even separated in the east–west direction. Recently, Xie and Bueh (2016) investigated the blocking circulation over the Sea of Okhotsk and its association with the NEC cold vortex on the intraseasonal timescale. They found that a long-lived blocking event over the Sea of Okhotsk is always accompanied by a positive anomaly center over eastern Europe, whereas the latter is absent for short-lived blocking events over the Sea of Okhotsk. This suggests that a positive anomaly center over eastern Europe may also be a contributing factor for the formation of a strong OKSH. Though the vertical structures indicated in the Z500 and SLP anomalies are different over the Sea of Okhotsk in positive and negative anomaly years of the second rainfall mode, the surface high anomalies are observed to be in a consistent position (black boxes in Figs. 7b and 7d). This suggests that the typical circulation associated with the second NPC rainfall mode is primarily characterized by an anomalous OKSH system. Figure 8 presents the vertically integrated water vapor flux anomalies and WVC anomalies for positive and negative anomaly years of the second NPC rainfall mode. In positive anomaly years, a cyclonic circulation anomaly is anchored over NEC (Fig. 8a), which is consistent with decreased SLP (Fig. 7b) and the negative height anomaly center at 500 hPa (Fig. 7a) over the region. As a result, the vertically integrated water vapor fluxes converge over NEC (Fig. 8a), in accordance with the rainfall increase there (Fig. 2c). Most parts of NC are affected by the northerly anomalies within the western flank of the cyclonic circulation anomaly, corresponding to the local rainfall decrease there (Fig. 2c). Additionally, an anticyclonic

VOL.30

circulation anomaly resides to the west of the abovementioned cyclonic circulation anomaly, which is also discernable in Fig. 7a, corresponding also to the rainfall decrease over NC and northern NWC (Fig. 2c). The increased WVC over NC and northern NWC (Fig. 8b) implies that the evaporation exceeds precipitation locally. In negative anomaly years of the second NPC rainfall mode (Figs. 8c and 8d), the situation is almost reversed compared to that in Figs. 8a and 8b. An extensive anticyclonic circulation anomaly controls NEC, eastern NC, and Mongolia, which is consistent with the rainfall decrease over NEC (Fig. 2d). However, the anomalous easterlies in the southern flank of the anticyclonic circulation anomaly transport extra water vapor to NC and NWC, corresponding to the rainfall increase there (Fig. 2d). It is also interesting to note that the vertically integrated water vapor flux anomalies over NEC and southern China show a similar feature (cyclonic or anticyclonic), both in positive and negative anomaly years (Figs. 8a and 8c). This fact is also consistent with the in-phase relationship between the rainfall anomalies over these two regions for the second NPC rainfall mode (Figs. 1f, 2c, and 2d). As presented previously, the circulation anomalies, and thus the vertically integrated water vapor flux anomalies, for the second NPC rainfall mode are most closely related to the variation of the OKSH system. To provide further evidence to this view, we constructed a summer OKSH index in terms of the area-averaged SLP over the domain (52◦ –65◦ N, 135◦ – 170◦ E)—black boxes in Figs. 7b and 7d—and examined the related rainfall anomalies over China. Figure 9 shows the regressed summer rainfall anomalies over China against the OKSH index. Clearly, the NPC rainfall anomaly pattern, as indicated by Figs. 1f, 2c, and 2d, are almost reproduced in Fig. 9; even the inphase relationship between the rainfall anomalies over NEC and southern China is apparent. These findings again suggest that the variation of the OKSH system is closely related to the circulation pattern, the water vapor condition, and the rainfall anomaly pattern over the NPC, acting as a primary contributor in the sec-

NO.5

BUEH Cholaw, LI Yan, LIN Dawei, et al.

627

Fig. 8. As in Fig. 5, but for the second NPC rainfall mode.

ond NPC rainfall mode. 5. Summary and discussion 5.1 Summary The present study used long-term (1961–2013) observational and reanalysis data to document the summer rainfall modes and the associated typical circulation patterns for the NPC on interannual timescale. Two important NPC rainfall modes were identified in the EOF analysis. The first EOF pattern is characterized by a uniformly distributed rainfall anomaly over most parts of the NPC, with opposite anomalies over southwestern NWC and the northern and southeastern tips of NEC. The second EOF pattern primarily shows rainfall variability in NEC and its out-of-phase relationship with that in NC and northern NWC. It is revealed that the CGT pattern associated with a strong (weak) ISM and the PEA− (PEA+ ) pattern work in concert to constitute the typical circulation pattern of the first NPC rainfall mode in its

positive (negative) phase. In the second NPC rainfall mode, the typical circulation pattern is characterized by an anomalous OKSH and the accompanying surface pressure anomaly over NEC and Japan. Specifically, in the positive (negative) phase of the second rainfall

Fig. 9. Regressed summer rainfall anomalies (%) against the standardized OKSH index at unit standard deviation. Positive (negative) values are indicated by solid (dashed) lines and zero lines are drawn in gray. Intervals are 2.5% and shadings are as in Fig. 1c.

628

JOURNAL OF METEOROLOGICAL RESEARCH

mode, a strong (weak) OKSH concurs with decreased (increased) SLP over NEC. In terms of water vapor transport, it was found that the cooperative engagement of a strong (weak) ISM circulation and PEA− (PEA+ ) pattern is fundamental in the anomalous water vapor transport to the NPC for the positive (negative) phase of the first NPC rainfall mode. This study emphasizes that the PEA pattern is essential for the water vapor transport by the midlatitude westerly flow, causing extensive rainfall anomalies over the NPC. In the positive phase of the second NPC rainfall mode, a cyclonic anomaly center of water vapor fluxes around NEC, which is associated with a strong OKSH, facilitates a favorable water vapor condition and rainfall in NEC. In the negative phase of the second rainfall mode, the situation is almost reversed. It is also revealed that the cyclonic or anticyclonic anomaly feature in the water vapor flux anomalies over NEC is opposite to that over NC and northern NWC, but similar to that over southern China. This is consistent with the rainfall anomalies illustrated by the second NPC rainfall mode. 5.2 Discussion The present study identifies several circulation patterns typical of rainfall anomalies over the NPC. Though the CGT pattern, representing typical circulation features of the ISM and EASM, has been explored in conjunction with NC rainfall anomalies, the crucial roles of the PEA pattern in the rainfall anomalies over the NPC, which are heavily emphasized in this study, have not been previously reported in literatures. The OKSH system and its relation to summer rainfall over NEC and the Yangtze–Huaihe River valley have been studied extensively. As presented in this study, the year-to-year variation of the OKSK is also closely associated with the rainfall anomalies in NC and southern China, as reflected in the second NPC rainfall mode. It seems that considering the NPC as a whole, rather than separating it into several parts (such as NC, NWC, and NEC), is a good option for improving our understanding of rainfall variability and the related typical circulation patterns in the NPC. It is also

VOL.30

suggested that the rainfall variations over the NPC are closely associated with those in some other parts of China, acting as an important reference. Though this study has identified two important modes of summer rainfall variability over the NPC, and explored their typical circulation patterns, comprehensive analyses regarding the underlying physical mechanisms are lacking. Wu et al. (2009) and Wu et al. (2012) found that a tripole pattern of sea surface temperature anomalies (SSTAs) over North Atlantic in summer, which is closely associated with the spring North Atlantic Oscillation, plays an important role in the interannual variation of EASM rainfall, through triggering a teleconnection pattern over northern Eurasia. The PEA pattern in the present study resembles the above-mentioned teleconnection pattern over northern Eurasia, suggesting a possible relationship between the PEA pattern and the North Atlantic tripole SSTA pattern. On the other hand, more recently, Wu et al. (2015) revealed that Tibetan Plateau snow cover in summer can influence the interannual variations of Eurasian heat-wave frequency through a wave train pattern over Eurasia. Interestingly, the wave train pattern is also similar to the PEA pattern. These findings suggest that the typical circulation patterns associated with summer rainfall variability over the NPC may possibly be connected with the North Atlantic SSTA and the anomalous snow cover over Eurasia. These issues remain to be investigated in future work. The prediction of summer rainfall in the NPC has long been a challenging task. The strongly irregular features of the interannual signals involved in the summer rainfall in different regions of the NPC lead to low predictability. At present, the capacity to predict summer rainfall in this region by analyzing signals from the preceding winter, including external forcing signals and internal atmospheric signals, is poor (Liu et al., 2010). However, in an attempt to address this issue, effort has been made to explore the oceanic and atmospheric signals from the preceding spring. Sun et al. (2003) noted that above-normal (below-normal) summer precipitation in NEC is often preceded by a negative (positive) 500-hPa height anomaly over the

NO.5

BUEH Cholaw, LI Yan, LIN Dawei, et al.

region from Lake Baikal to Yakutsk from April to May. Li et al. (2005) found that summer rainfall anomalies over NC and NWC are significantly correlated with the strength and phase of the Pacific Decadal Oscillation in the preceding spring. Upon analyzing the summer rainfall variability in Inner Mongolia, Li et al. (2016) found that it is closely associated with the timing of the spring-to-summer seasonal circulation transition over mid–high latitude Asia. All of these studies on summer rainfall over the NPC hint that there may be some precursory atmospheric and oceanic signals in the preceding spring. This issue deserves further investigation in future studies. Acknowledgments. The figures in this study were plotted by using NCARG Command Language (UCAR/NCAR/CISL/VETS, 2012).

REFERENCES Barnston, A. G., and R. E. Livezey, 1987: Classification, seasonality and persistence of low-frequency atmospheric circulation patterns. Mon. Wea. Rev., 115, 1083–1126. Bueh Cholaw, Shi Ning, Ji Liren, et al., 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. Chen Dongdong and Dai Yongjiu, 2009: Characteristics and analysis of typical anomalous summer rainfall patterns in Northwest China over the last 50 years. Chinese J. Atmos. Sci., 33, 1247–1258. (in Chinese) Ding, Q. H., and B. Wang, 2005: Circumglobal teleconnection in the Northern Hemisphere summer. J. Climate, 18, 3483–3505. Gorry, P. A., 1990: General least-squares smoothing and differentiation by the convolution (Savitzky–Golay) method. Analy. Chem., 62, 570–573. Guo

Qiyun, 1992: Teleconnection between the floods/droughts in North China and Indian summer monsoon rainfall. Acta Geographica Sinica, 47, 394–402. (in Chinese)

Huang Ronghui and Li Weijing, 1987: Influence of the heat source anomaly over the western tropical Pacific on the subtropical high over East Asia. Proceedings of the International Conference on the General Cir-

629

culation of East Asia. Chengdu, China, 40–45. (in Chinese) Huang, R. H., and F. Y. Sun, 1992: Impact of the tropical western Pacific on the East Asian summer monsoon. J. Meteor. Soc. Japan, 70, 243–256. Jolliffe, I. T., 1986: Principal Component Analysis. Springer-Verlag, New York, USA, 290 pp. Kalnay, E., M. Kanamitsu, R. Kistler, et al., 1996: The NCEP/NCAR 40-year reanalysis project. Bull. Amer. Meteor. Soc., 77, 437–471. Kripalani, R. H., and S. V. Singh, 1993: Large-scale aspects of India-China summer monsoon rainfall. Adv. Atmos. Sci., 10, 71–84. Kripalani, R. H., and A. Kulkarni, 2001: Monsoon rainfall variations and teleconnections over South and East Asia. Int. J. Climatol., 21, 603–616. Kripalani, R. H., A. Kulkarni, and S. V. Singh, 1997: Association of the Indian summer monsoon with the Northern Hemisphere mid-latitude circulation. Int. J. Climatol., 17, 1055–1067. Li Dongliang, Xie Jinnan, and Wang Wen, 1997: A study of summer precipitation features and anomaly in Northwest China. Scientia Atmospherica Sinica, 21, 331–340. (in Chinese) Li, Q., S. Yang, V. E. Kousky, et al., 2005: Features of cross-Pacific climate shown in the variability of China and US precipitation. Int. J. Climatol., 25, 1675–1696. Li Yan, Bueh Cholaw, Lin Dawei, et al., 2016: The dominant modes of summer precipitation over Inner Mongolia and its typical circulation characteristics. Chinese J. Atmos. Sci., doi: 10.3878/j.issn.10069895.1509.15187. (in Chinese) Lian Yi and An Gang, 1998: The relationship among East Asian summer monsoon, El Ni˜ no and low temperature in Songliao Plains of Northeast China. Acta Meteor. Sinica, 56, 724–735. (in Chinese) Liang Feng, Tao Shiyan, Wei Jie, et al., 2011: Variation in summer rainfall in North China during the period 1956–2007 and links with atmospheric circulation. Adv. Atmos. Sci., 28, 363–374. Lin Dawei, Bueh Cholaw, and Xie Zuowei, 2016: Relationship between summer rainfall over North China and India and its genesis analysis. Chinese J. Atmos. Sci., 40, 201–214, doi: 10.3878/j.issn.10069895.1503.14339. (in Chinese) Liu Yunyun and Ding Yihui, 2008: Analysis and numerical simulation of the teleconnection between Indian

630

JOURNAL OF METEOROLOGICAL RESEARCH

summer monsoon and precipitation in North China. Acta Meteor. Sinica, 66, 789–799. (in Chinese) Liu, S., S. Yang, Y. Lian, et al., 2010: Time-frequency characteristics of regional climate over Northeast China and their relationships with atmospheric circulation patterns. J. Climate, 23, 4956–4972. Lu, R. Y., 2004: Associations among the components of the East Asian summer monsoon system in the meridional direction. J. Meteor. Soc. Japan, 82, 155–165. Lu, R. Y., J. H. Oh, and B. J. Kim, 2002: A teleconnection pattern in upper-level meridional wind over the North African and Eurasian continent in summer. Tellus A, 54, 44–55. Nakamura, H., and T. Fukamachi, 2004: Evolution and dynamics of summertime blocking over the Far East and the associated surface Okhotsk high. Quart. J. Roy. Meteor. Soc., 130, 1213–1233. Nitta, T., 1987: Convective activities in the tropical western Pacific and their impact on the Northern Hemisphere summer circulation. J. Meteor. Soc. Japan, 65, 373–390. North, G. R., T. L. Bell, R. F. Cahalan, et al., 1982: Sampling errors in the estimation of empirical orthogonal functions. Mon. Wea. Rev., 110, 699–706. Parthasarathy, B., A. A. Munot, and D. R. Kothawale, 1994: All-India monthly and seasonal rainfall series: 1871–1993. Theor. Appl. Climatol., 49, 217–224. Shi Yafeng, Shen Yongping, Li Dongliang, et al., 2003: Discussion on the present climate change from warmdry to warm-wet in Northwest China. Quaternary Sciences, 23, 152–164. (in Chinese) Su Tao and Feng Guolin, 2014: The characteristics of the summer atmospheric water cycle over China and comparison of ERA-Interim and MERRA reanalysis. Acta Physica Sinica, 63, 249201. (in Chinese) Sun Li, An Gang, and Dong Li, 2002: The characteristics of summer drought and flood in northeast area of China. Scientia Geographica Sinica, 22, 311–316. (in Chinese) Sun Li, An Gang, and Tang Xiao1ing, 2003: Relationship between the Northeast Asian summer south wind anomaly and the precipitation in Northeast China. Chinese J. Atmos. Sci., 27, 425–434. (in Chinese) Wang, Y. F., 1992: Effects of blocking anticyclones in Eurasia in the rainy season (Meiyu/Baiu season). J. Meteor. Soc. Japan, 70, 929–951.

VOL.30

Wilks, D. S., 1995: Statistical Methods in the Atmospheric Sciences. Academic Press, San Diego, CA, 121 pp. Wu Bingyi, 2005: Weakening of Indian summer monsoon in recent decades. Adv. Atmos. Sci., 22, 21–29. Wu, Z. W., B. Wang, J. P. Li, et al., 2009: An empirical seasonal prediction model of the East Asian summer monsoon using ENSO and NAO. J. Geophys. Res., 114, D18120, doi: 10.1029/2009JD011733. Wu, Z. W., J. P. Li, Z. H. Jiang, et al., 2012: Possible effects of the North Atlantic Oscillation on the strengthening relationship between the East Asian summer monsoon and ENSO. Int. J. Climatol., 32, 794–800. Wu, Z. W., P. Zhang, H. Chen, et al., 2015: Can the Tibetan Plateau snow cover influence the interannual variations of Eurasian heat wave frequency? Climate Dyn., doi: 10.1007/s00382-015-2775-y. Xie, Z. W., and C. Bueh, 2016: Cold vortex events over Northeast China associated with the Yakutsk– Okhotsk blocking. Int. J. Climatol., doi: 10.1002/ joc.4711. Yang Lianmei and Zhang Qingyun, 2007: Circulation characteristics of interannual and interdecadal anomalies of summer rainfall in North Xinjiang. Chinese J. Geophys., 50, 412–419. (in Chinese) Yang Lianmei, Xiaokaiti Duolaite, and Zhang Qingyun, 2009: Relationships between rainfall anomalies in Xinjiang summer and Indian rainfall. Plateau Meteor., 28, 564–572. (in Chinese) Yao Xiuping and Dong Min, 2000: Research on the features of summer rainfall in Northeast China. Quart. J. Appl. Meteor., 11, 297–303. (in Chinese) Zhang Qingyun and Tao Shiyan, 1998: Influence of Asian mid–high latitude circulation on East Asian summer rainfall. Acta Meteor. Sinica, 56, 199–211. (in Chinese) Zhang Renhe, 1999: The role of Indian summer monsoon water vapor transportation on the summer rainfall anomalies in the northern part of China during the El Ni˜ no mature phase. Plateau Meteor., 18, 567– 574. (in Chinese) Zhao Yufei, Zhu Jiang, and Xu Yan, 2014: Establishment and assessment of the grid precipitation datasets in China for recent 50 years. J. Meteor. Sci., 34, 414– 420. (in Chinese)