Subseasonal Characteristics of Diurnal Variation in Summer Monsoon Rainfall over Central Eastern China

6684 J O U R N A L O F C L I M A T E VOLUME 23 Subseasonal Characteristics of Diurnal Variation in Summer Monsoon Rainfall over Central Eastern China WEIHUA YUAN LASG, Institute of Atmospheric Physics, Chinese Academy of Sciences, and Graduate School of Chinese Academy of Sciences, Beijing, China RUCONG YU Beijing Climate Center, and LaSW, Chinese Academy of Meteorological Sciences, China Meteorological Administration, Beijing, China HAOMING CHEN AND JIAN LI LaSW, Chinese Academy of Meteorological Sciences, China Meteorological Administration, Beijing, China MINGHUA ZHANG Institute for Terrestrial and Planetary Atmospheres, School of Marine and Atmospheric Sciences, Stony Brook University, Stony Brook, New York (Manuscript received 22 April 2010, in final form 5 August 2010) ABSTRACT Subseasonal characteristics of the diurnal variation of the summer monsoon rainfall over central eastern China (258 408N, 1108 1208E) are analyzed using hourly station rain gauge data. Results show that the rainfall in the monsoon rain belt is dominated by the long-duration rainfall events ($7 h) with early-morning peaks. The long-duration rainfall events and early-morning diurnal peaks experience subseasonal movement that is similar to that of the monsoon rain belt. When the monsoon rainfall is separated into the active and break periods, the long-duration early-morning precipitation dominates the active period, which is in sharp contrast to the short-duration (#6 h) rainfall with leading late-afternoon diurnal peaks during the break period. The combination of different diurnal features of monsoon rainfall in the active and break monsoon periods also explains the less coherent diurnal phases of summer mean rainfall over central eastern China. The cause of the early-morning peak of rainfall during the active monsoon period is discussed. 1. Introduction Diurnal variability of rainfall is an important aspect of regional climate. When precipitation occurs regularly during particular periods of the day, the atmospheric system is usually characterized by certain dynamical and thermal conditions (Sorooshian et al. 2002). Understanding these conditions is not only important to predict regional precipitation events but also necessary to validate and improve climate models. Corresponding author address: Dr. Rucong Yu, China Meteorological Administration, No. 46, Zhongguancun South Street, Beijing 100081, China. E-mail: yrc@cma.gov.cn Summer rainfall over contiguous China has pronounced diurnal variations with considerable regional features (Yu et al. 2007b). Using rain gauge records, Yu et al. (2007b) showed a full description of diurnal variations for summer rainfall over contiguous China except for most northwestern China. They found coherent rainfall diurnal variation except over central eastern China: over southern and northeastern China, summer rainfall in most stations peaks in the late afternoon (LA); over southwestern China, hourly rainfall peaks around midnight. However, they showed that there are no uniform diurnal peaks of the summer rainfall among different stations in central eastern China, and similar results are reported in He and Zhang (2010). But the regionally averaged diurnal variation of summer precipitation is DOI: 10.1175/2010JCLI3805.1 Ó 2010 American Meteorological Society

15 DECEMBER 2010 Y U A N E T A L. 6685 characterized by two comparable peaks: one in the early morning (EM) and the other in LA (Yu et al. 2007b). This rainfall diurnal characteristic in central eastern China can be partly attributed to the distinct diurnal features of rainfall with different duration time. Yu et al. (2007a) revealed that the long-duration rainfall event usually has the maximum around EM and the shortduration event gets the maximum around LA in this area. The LA peak of short-duration rainfall events may be explained by the diurnal variation of low-level atmospheric stability caused by solar heating, but the physical mechanisms of the prevailing EM long-duration rainfall events have not been clearly understood up to now. The diurnal and seasonal cycles are the two most fundamental modes of the variability of global climate systems, which are associated with large and well-defined variations in the solar forcing (Yang and Slingo 2001). Besides the distinct diurnal variations, the rainfall of the East Asian summer monsoon (EASM) also exhibits pronounced subseasonal transitions, which have been documented in numerous literatures (e.g., Tao and Chen 1987). The monsoon rainfall commences from the South China Sea during mid- and late May and extends abruptly to the Yangtze River Valley in about early to mid-june. Then the monsoon rain belt penetrates to north China. The passage of the monsoon rain belt is followed by a break period that also propagates northward, and the rainy season in some areas revives after the break (Tao and Chen 1987; Ding 1992; Qian et al. 2002; Chen et al. 2004; Ding and Chan 2005). Simultaneous with progressions of the monsoon rain belt, the monsoon airflow demonstrates similar movements. These stepwise jumps and retreats are closely related to seasonal changes in the general circulation in East Asia, especially the seasonal evolution of the planetary frontal zone, subtropical high over the West Pacific, and the southwesterly low-level jet that is a crucial factor to maintaining the rainfall in the monsoon rain belt (Ramage 1952; Matsumoto 1985; Chen and Li 1995; Wang et al. 2000; Ding et al. 2001; Chen et al. 2004). Meanwhile, the rainfall and reflectivity in mei-yu present significant diurnal variation, which usually enhance from midnight to EM (Luo et al. 2003; Geng and Yamada 2007; Chen et al. 2010). Previous studies based on limited station observations or Tropical Rainfall Measuring Mission (TRMM) products presented indications that the nighttime or morning rainfall over eastern China may undergo seasonal advance and retreat, which is possibly related to the movement of EASM (Chen et al. 2009; Yin et al. 2009). In these studies, only the monthly characteristics of rainfall diurnal variations were analyzed. However, the movement of EASM rainfall always occurs abruptly in a few days (Ding and Chan 2005), so process understanding on a shorter time scale is needed. In addition, the sparse density of stations and the limited latitudinal coverage of the TRMM observations restrict the analysis of the subseasonal movements of the diurnal rainfall. Moreover, these previous studies only focus on the total rainfall, while the rainfall diurnal characteristic of the monsoon rain belt resembles that of the long-duration rainfall that responded to large-scale circulations (Chen 1983; Yu et al. 2007a). The contributions of the rainfall with different duration time to the monsoon rain belt have never been discussed. The possible linkages of the summer monsoon rain belt to both the long-duration rainfall events and EM rainfall motivate us to investigate the subseasonal variations of rainfall diurnal features (including the rainfall amount, frequency, and intensity) and durational characteristics according to the subseasonal movement of the summer monsoon rain belt based on a finer time interval and a denser surface station network. Considering 1108E as the west rim of EASM (Wang and LinHo 2002; Ding and Chan 2005), this study focuses on the subseasonal characteristics of rainfall diurnal features over central eastern China (258 408N, 1108 1208E) from June (onset of mei-yu) to September (end of EASM). Results show that the subseasonal movement of the monsoon rain belt is mainly contributed by the long-duration rainfall with EM diurnal peaks. The monsoon rainfall presents alterative EM and LA diurnal peaks corresponding to the active and break monsoon periods. The rest of paper is organized as follows. Section 2 describes the datasets and analysis methods. The linkage between the long-duration rainfall events and monsoon rain belt is presented in section 3. Section 4 examines the subseasonal characteristics of EM and LA diurnal phases in summer monsoon rainfall. Section 5 further analyzes the rainfall diurnal features in active and break monsoon periods. The summary and discussion are given in section 6. 2. Datasets and analysis methods The hourly rain gauge records at 575 stations covering mainland China in 1961 2006 have been used in this study. The data are merged by two independent datasets: one consists of 575 stations during 1961 2000 (D1) and the other contains records from 706 stations during 1991 2006 (D2, records during 2001 06 are used in this study). D2 and the hourly data used in Yu et al. (2007b) are from the same data source. The two datasets were obtained from the national climatic reference network and the national weather surface network of China and were collected and quality controlled by the National

6686 J O U R N A L O F C L I M A T E VOLUME 23 FIG. 1. Locations of the 288 rain gauge stations used in this study, including six mountain stations marked by triangles. Meteorological Information Center (NMIC) of the China Meteorological Administration (CMA). The records of both D1 and D2 were made by siphon or tippingbucket rain gauges. The digital records of D1 were collected manually, while D2 were collected either manually or automatically by computer. The overlapped records between D1 and D2 (during 1991 2000) have been compared and no essential difference was found. Only the stations (288 in total) that have more than 85% of summer days (June September) without missing values during 1961 2006 over eastern China are used in the analysis. Among the 288 stations used in the final analysis, 247 stations have data without missing values in every summer for at least 25 years during 1961 90. For the rest of the stations, only 9 stations have fewer than 20 years of nonmissing observation during this period. The locations of the 288 stations are shown in Fig. 1. The hourly rainfall event is defined as the one with more than or equal to 0.1 mm precipitation accumulated during an hour. The hourly mean rainfall amount (the accumulated rainfall amount), frequency (the percentage of observational hours having measurable precipitation), and intensity (the mean rate in precipitation hours) are calculated at each hour averaged in 1961 2006. The time of diurnal peak is represented by the hour when the maximum of the mean precipitation amount, frequency, or intensity occurs, and the amplitude is calculated following Dai et al. (1999). The hourly rainfall event is further classified according to its continuous durations without any intermittence or at most a 1-h intermittence. Therefore rainfall after an intermittence is considered to belong to a new rainfall event FIG. 2. The normalized diurnal variation of rainfall amount (normalized by the daily mean) for rainfall events with different duration time averaged over central eastern China (258 408N, 1108 1208E). The x axis represents the LST, and the y axis denotes the duration time. The contour interval is 0.1. Dashes represent negative contours. if the intermittence lasts for 2 h or longer (Yu et al. 2007a). The number of hours between the start and the end of every event is defined as the duration time. The normalized diurnal variations of rainfall amount for events with different duration hours averaged in summer (June September) over central eastern China are shown in Fig. 2. The hourly rainfall amount is normalized with the daily mean for a better comparison among rainfall events with different duration times. As illustrated in Fig. 2, the rainfall occurring in LA [1400 2000 local solar time (LST)] and EM (0200 0800 LST) dominates the total rainfall. The accumulated LA and EM rainfall amount contributes to more than 71% of the total. The LA diurnal peaks are mainly contributed by the rainfall lasting less than or equal to 6 h, and the EM peaks are primarily contributed by the rainfall lasting more than 6 h. In the subsequent discussions of this study, the long-duration rainfall is defined as rainfall lasting more than 6 h and the short-duration rainfall as rainfall lasting 1 6 h. The Japanese 25-yr reanalysis (JRA-25; Onogi et al. 2007) for the period from 1979 to 2006 is used in this study to describe the large-scale circulation. Hourly wind records at six mountain stations over China (marked by triangles in Fig. 1) during 1991 2006 are also applied for the validation of the diurnal wind evolution revealed by the reanalysis. The records at mountain stations have

15 DECEMBER 2010 Y U A N E T A L. 6687 FIG. 3. The time latitude cross section of 10-day mean rainfall amount (mm day 21 )of (a) long-duration, (b) short-duration, and (c) total rainfall averaged in 1108 1208E. The five black solid lines in (a) represent the four periods discussed in this study. been demonstrated to be useful indicators of tropospheric low-level winds (Yu et al. 2009; Chen et al. 2010). To analyze the subseasonal features of rainfall diurnal variations, the 10-day mean data are obtained by averaging the original datasets.asthenumbersofdays among months are different, the days from the 21st day to the last day are regarded as the last 10 days for each month. 3. Linkage between long-duration rainfall and monsoon rain belt According to Yu et al. (2007a), long-duration rainfall events dominate the summer rainfall over central eastern China. To investigate its subseasonal variation and contribution to the total rainfall, in Fig. 3 we show the time latitude cross sections of 10-day mean rainfall amount for long-duration, short-duration, and total rainfall averaged between 1108 and 1208E. The dominant long-duration rainfall (Fig. 3a) experiences similar subseasonal advance and retreat as that of total rainfall (Fig. 3c), and the latter has been discussed extensively before (e.g., Tao and Chen 1987). In Figs. 3a and 3c, the major rain belt commences in mid-june along the Yangtze River Valley (about 278 328N) and lasts to about July. In early to mid-july, rainfall in the Huaihe River Valley (about 328 378N) is at its annual maximum. At the end of July, the rain belt jumps to north China (about 378 408N), at the northernmost position of summer monsoon. The rainy season over north China persists from about late July to mid-august. From the end of August, the monsoon rain belt moves back to the Huaihe and Yangtze River Valleys. After mid- September, the rainy season comes to the end for most central eastern China. Contrary to the subseasonal variation of long-duration and total rainfall, the increase of short-duration rain mainly occurs in July and August with few regional differences except for south China (Fig. 3b). Short-duration rainfall in south China increases even earlier than May (figure not shown). The large rainfall amount center there in mid-july to mid- August corresponds to the second rainy season over south China, which is mainly caused by typhoons, the

6688 J O U R N A L O F C L I M A T E VOLUME 23 FIG. 4. The spatial distributions of (a) long-duration, (b) short-duration, and (c) total rainfall amount averaged in mid- and late June (mm day 21 ). The Yangtze and Yellow Rivers are drawn as black lines. movement of the ITCZ, and other tropical disturbances (Ramage 1952; Chen 1994; Chen et al. 2004). Therefore, to further emphasize the subseasonal migration of longduration and total rainfall amounts shown in Figs. 3a and 3c, four periods (marked by five black lines in Fig. 3a) have been selected to represent the major standing stages of the monsoon rain belt for different regions. Besides the subseasonal movement, the long-duration rainfall also presents similar spatial distribution as that of total rainfall. Averaged in mid- and late June, rainfall centers extending along the middle and lower reaches of the Yangtze River Valley can be found in the distributions of both long-duration (Fig. 4a) and total (Fig. 4c) rainfall, with the spatial correlation coefficient reaching 0.99 over central eastern China. Conversely, the shortduration rainfall mainly prevails in south China (Fig. 4b). During the migration of the monsoon rain belt from July to September, the long-duration rainfall always resembles the total rainfall in the spatial distribution (figures omitted). The results above indicate that the long-duration rainfall is closely linked to the rainfall in the monsoon rain belt, which has demonstrated to be highly related to the principal frontal zone (Reed 1960; Matsumoto 1985; Chen et al. 2004). The departures of temperature and specific humidity between the long-duration rainfall days and climatology derived from the reanalysis averaged in mid- and late June during 1979 2006 are shown in Fig. 5. The long-duration rainfall days here are defined as the days when the long-duration rainfall occurs at more than 20% of the stations in the middle and lower reaches of the Yangtze River Valley (278 328N, 1108 1208E). The vertical distribution of temperature around 308N (dashed contours in Fig. 5a) is characterized by the positive anomalies at the higher level and negative anomalies at the lower level. It suggests that the cold air mass intrudes

15 DECEMBER 2010 Y U A N E T A L. 6689 FIG. 5. (a) The temperature departures (black dashed contours, K) and specific humidity (gray solid contours, g kg 21 ) averaged in 1108 1208E (the mean in the long-duration rainfall days relative to the climatology) derived from JRA-25 data in 1979 2006 during mid- and late June. Shaded are regions significant at the 5% level according to the Student s t test for temperature departures. The contours of specific humidity are all above the 95% confidence level, and the interval of contours is 0.1 g kg 21. (b) Wind anomalies at 850 hpa (the mean in the longduration rainfall days relative to the climatology) derived from JRA-25 data in 1979 2006 during mid- and late June. from the north in the lower troposphere, forcing the warm air to rise, which is also reflected in the low-level wind field. Figure 5b presents the 850-hPa wind anomalies between the mean of long-duration rainfall days and the climatology. As illustrated in Fig. 5b, the area north of the Yangtze River Valley is dominated by the northeasterly anomalies, while southwesterly ones prevail over southeastern China. Accompanied by the convergence zone lying along the Yangtze River Valley, this area is also characterized by the positive moisture anomalies (solid gray contours in Fig. 5a). These coupled temperature, humidity, and wind fields in the long-duration rainfall days experience subseasonal movement that follows the monsoon rain belt (figures not shown). It is consistent with previous studies, which point out that the frontal zone corresponding to the monsoon rain belt has a migration in concert with the monsoon s evolution (Matsumoto 1985; Chen et al. 2004). 4. Subseasonal characteristics of EM and LA diurnal phases in summer monsoon rainfall As illustrated in Fig. 2, the summer monsoon rainfall in central eastern China mainly occurs in EM and LA. The subseasonal migration of EM and LA rainfall is analyzed in this section. The percentages of the EM rainfall amount (Fig. 6a) and frequency (Fig. 6b) to total rainfall averaged in 1108 1208E show obvious subseasonal movements as those of total and long-duration rainfall shown in Fig. 3. Compared with the persistent large percentages around 308N throughout June for the rainfall amount, the migration of EM frequency in the areas between the Yangtze and Yellow River Valleys is more obvious during this period. Figure 6c shows the subseasonal variability of EM rainfall intensity. As indicated in Fig. 6, both the increased frequency and enhanced intensity over most central eastern China contribute to the increase of EM rainfall amount, which is also mentioned in Zhou et al. (2008). While the intensity does not experiences notable enhancement in July in the area between 358 and 388N, the increase of frequency is the major contributor to the augment of the rainfall amount. The LA rainfall amount (Fig. 7) presents a distinct transition from that of the monsoon rain belt and EM rainfall. During June and early July, the LA rainfall dominates north and south China, while the rain belt extends along the Yangtze and Huaihe River Valleys. When the rain belt moves northward, the percentage of LA rainfall amount in the Yangtze and Huaihe River Valleys increases. Along with the revival of monsoon rainfall in this area, the dominant LA rainfall is substituted by EM rainfall. The LA rainfall frequency and intensity exhibit similar movement as that of rainfall amount (figures omitted). To further analyze the relation of subseasonal movement between the monsoon rain belt and rainfall diurnal phases, we show the spatial distributions of the diurnal peaks and amplitudes of rainfall frequency averaged in the four periods (mentioned in section 3) in Fig. 8. As

6690 J O U R N A L O F C L I M A T E VOLUME 23 FIG. 6. The time latitude cross section of 10-day mean percentages of EM (a) rainfall amount and (b) frequency vs total rainfall averaged in 1108 1208E (%). (c) The time latitude cross section of 10-day mean EM rainfall intensity averaged in 1108 1208E (mm h 21 ). indicated by the colors of vectors, the rainfall in central eastern China mainly peaks in LA (red vectors) and EM (green vectors). The EM diurnal peaks exhibit bandlike spatial distributions. During mid- and late June (Fig. 8a), the EM-peak band mainly extends along the area around 308N, where the rain belt is located. When the monsoon rainfall in the Huaihe River Valley gets the annual maximum (early and mid-july, Fig. 8b), the less coherent diurnal peaks in Fig. 8a over this area shift uniformly to EM. The region south of the EM-peak band, where the rainfall gets the maximum in LA, also undergoes northward enlargement with prominent amplitude. During late July to mid-august (Fig. 8c), the EM rainfall peaks shift northward to north China, coinciding with the migration of the monsoon rain belt. The EM diurnal peaks along the Yangtze and Huaihe River Valleys are substituted by LA peaks. In late August and early September (Fig. 8d), EM rainfall peaks withdraw to the south and again dominate the Huaihe River Valley. To be noted, the band of EM diurnal peaks presents a north tilt with latitude in four periods, and it is tilted more toward the north south direction when the rain belt locates at higher latitude. It can be also found in the spatial distributions of the long-duration and total rainfall amounts (figures not shown). 5. Rainfall diurnal features in active and break monsoon periods Analyses in section 3 have demonstrated the fundamental roles of long-duration rain in determining the subseasonal movement of the summer monsoon rain belt over central eastern China. Results in section 4 showed the correlation between locations of the EM diurnal peaks and the rain belt. Therefore, the long-duration rainfall with EM diurnal peak is closely related to the summer monsoon rain belt. Following Chen et al. (2010), the EM long-duration rainfall days are selected to represent the days when the rain belt locates in a region. As the monsoon rain belt undergoes stepwise movement, four target regions (labeled in Fig. 8) have been selected

15 DECEMBER 2010 Y U A N E T A L. 6691 with much reduced amplitude. During the break period, the total, short-duration, and long-duration rainfalls all get the hourly maximum in LA, while the long-duration rainfall also has a secondary EM diurnal peak. These results suggest that monsoon rainfall at different latitudes is dominated by the long-duration rainfall events with EM diurnal peak during the active period, which is quite opposite to the leading LA diurnal peak of short-duration rainfall being during the break period. The different diurnal features of monsoon rainfall between the active and break monsoon periods can also explain the two comparable diurnal peaks in summer rainfall averaged over central eastern China. FIG. 7. The time latitude cross section of 10-day mean percentages of LA rainfall amount to that of total rainfall averaged in 1108 1208E (%). to represent locations of the monsoon rain belt for the four periods, according to the spatial distributions of EM peaks (shown in Fig. 8) and accumulated rainfall amount (figure not shown). The number of stations in each region is listed in Table 1. The days in mid- and late June when long-duration events with EM peak occur at 8% or more stations in region 1 (labeled in Fig. 8a) are defined as the active monsoon period for this region and the remaining days of summer as the break period. The active and break monsoon periods of other regions are similarly selected. The number of days in active monsoon period for each region is listed in Table 1. To emphasize the diurnal features of the monsoon rain belt, the diurnal variations during three time regimes(thewholesummer,andtheactiveandbreak monsoon periods) are compared in the four regions. As the diurnal features of different regions during the same time regime are similar, we only show the situation for region 1. Averaged in the whole summer (Fig. 9a), two comparable diurnal peaks of total rainfall exist in EM and LA. The EM diurnal peak is mainly attributed to the long-duration rainfall, and the LA peak with more pronounced amplitude is attributed to the short-duration rainfall. In the active period (Fig. 9b), the diurnal phase of total rainfall shifts to EM and the amplitudes of diurnal variations for both long-duration and total precipitation are large. The short-duration rainfall has high values from LA to midnight and reaches the minimum around noon, 6. Summary and discussions a. Summary Using the hourly station rain gauge data in 1961 2006, the subseasonal characteristics of rainfall diurnal variations over central eastern China are analyzed. The major conclusions are summarized below. 1) The long-duration rainfall, the major contributor of summer monsoon rainfall, experiences similar subseasonal movement to that of summer monsoon rain belt. Conversely, the short-duration events mainly occur in July and August over most of central eastern China. The vertical distributions of temperature and humidity anomalies in long-duration rainfall days resemble those of the principal frontal zone, indicating the close linkage between the long-duration rainfall and summer monsoon rain belt. 2) The EM rainfall show subseasonal movement coinciding with that of the monsoon rain belt. The augment of EM rainfall amount in the monsoon rain belt is the result of both the increased frequency and enhanced intensity over most of central eastern China. However, the increase of frequency is the major contributor to the augment of the rainfall amount in the area between 358 and 388N, and the intensity does not experiences notable enhancement. 3) Rainfall at different latitudes of central eastern China exhibits similar diurnal features when partitioned into active and break periods. During the active period, monsoon rainfall is dominated by the long-duration events with EM diurnal peak. In contrast, during the break period, monsoon rainfall is dominated by shorter-duration events with LA diurnal peak. The different diurnal variation of monsoon rainfall between the active and break monsoon periods can explain the two comparable diurnal peaks averaged in summer central eastern China.

6692 J O U R N A L O F C L I M A T E VOLUME 23 FIG. 8. Spatial distributions of the diurnal peaks (LST) and amplitude (% of daily mean) of rainfall diurnal cycle averaged in (a) mid- and late June, (b) early and mid-july, (c) late July to mid-august, (d) late August and early September. The shading represents the amplitude, and unit vectors denote the LST of the maximum precipitation (see phase clock and colors). The green (blue) vectors represent the diurnal peaks occurring between 0200 and 0800 (between 2000 and 0200) LST, and the red (black) vectors for the diurnal peaks occurring between 1400 and 2000 (between 0800 and 1400) LST. The Yangtze and Yellow Rivers are drawn as black lines. Four distinct regions are also labeled.

15 DECEMBER 2010 Y U A N E T A L. 6693 TABLE 1. The number of stations (NS) and days (ND) in active monsoon period accumulated in 1961 2006 for the four regions marked in Fig. 8. Region 1 Region 2 Region 3 Region 4 NS 87 48 57 41 ND 305 239 274 168 b. Discussion The major finding in this study is that rainfall in the monsoon rain belt is dominated by the long-duration rainfall with EM diurnal peaks. The diurnal features of long-duration rainfall have been demonstrated to be closely related to the diurnal evolution of the low-level southwesterlies (Chen et al. 2010). To analyze possible mechanisms responsible to rainfall diurnal features in the active monsoon period, the composite diurnal variations of low-level wind field are calculated. Figure 10 shows spatial distributions of composite 850-hPa wind anomalies (relative to the daily mean, black vectors) at 0800 BJT (Beijing time) of the reanalysis averaged in the corresponding four periods for the active monsoon days. As there is only four time slices in the reanalysis, the counterpart of records of six mountain stations at 0800 BJT (red vectors) is also presented for the validation. Consistent with Chen et al. (2010), wind anomalies in the reanalysis generally exhibit a close resemblance to those at mountain stations, although the records at mountain stations only cover the period of 1991 2006. Compared with the climatology, the low-level wind fields in the active monsoon period during mid- and late June present southwesterly anomalies over southern China and northeasterly ones over the north of Yangtze River Valley, which is quite similar to those illustrated in Fig. 5. As shown in Fig. 10a, both the southwesterly and northeasterly anomalies are enhanced at 0800 BJT, which form a cyclonic anomaly elongating the monsoon rain belt (marked by blue lines). The combination of the cyclonic vorticity and enhanced moisture transported by the southwesterly winds provides a favorable environment for the nocturnal rainfall systems to develop and maximize at EM (Tao and Chen 1987; Zhou and Yu 2005; Yu et al. 2009). At other time slices, easterly (2000 BJT), northerly (1400 BJT), and southeasterly anomalies (0200 BJT) dominate the area of the monsoon rain belt (figures not shown). As a result, the convergence of wind and moisture transportation of the rainfall zone is weakened and unfavorable for the rainfall maximum. Accompanied with the subseasonal movements of the monsoon rain belt, the northeasterly and southwesterly flow anomalies at 0800 BJT, along with the shear line FIG. 9. The diurnal variations of rainfall amount (normalized with daily mean) averaged in (a) summer, (b) the active monsoon period, and (c) the break monsoon period for region 1. The solid lines represent the total rainfall. The dashed (dotted) lines represent the long-duration (short duration) rainfall events. The x axis denotes the LST. formed by them that favors the EM diurnal peak, undergo similar movements (Figs. 10b d). The location of the shear line always corresponds to that of the monsoon rain belt (marked by the blue lines). From June to August (Figs. 10a c), the northward component of the southwesterlies over the southern boundary of the shear line increases gradually and the shear line is tilted more toward the north south direction. The changes in the tilted spatial distributions of EM diurnal phases presented in Fig. 8 and those of total and long-duration rainfall amounts (figure not shown) could be partly attributed to the variations of the tilted shear line and low-level largescale circulations. Acknowledgments. This work was jointly supported by the Major National Basic Research Program of China (973 Program) on Global Change under Grant 2010CB951900 and the National Natural Science Foundation of China under Grants 40921003, 40625014, and

6694 J O U R N A L O F C L I M A T E VOLUME 23 FIG. 10. The composite wind anomalies (m s 21 ) at 0800 BJT in the active monsoon period (relative to the daily mean) derived from JRA-25 data at 850 hpa for 1979 2006 (black vectors) and records at six mountain stations at 0800 BJT (red vectors) for 1991 2006, during (a) mid- and late June, (b) early and mid-july, (c) late July to mid- August, and (d) late August and early September. The blue lines show the four regions representing the locations of the monsoon rain belt. The red dots mark the locations of the six mountain stations. 40705025. Additional support is provided by the U.S. Department of Energy to Stony Brook University. REFERENCES Chen, G., W. Sha, and T. Iwasaki, 2009: Diurnal variation of precipitation over southeastern China: Spatial distribution and its seasonality. J. Geophys. Res., 114, D13103, doi:10.1029/ 2008JD011103. Chen, G. T.-J., 1983: Observational aspects of the mei-yu phenomenon in sub-tropical China. J. Meteor. Soc. Japan, 61, 306 312., 1994: Large-scale circulations associated with the East Asian summer monsoon and the mei-yu over south China and Taiwan. J. Meteor. Soc. Japan, 72, 959 983. Chen, H., R. Yu, J. Li, W. Yuan, and T. Zhou, 2010: Why nocturnal long-duration rainfall presents an eastward-delayed diurnal phase of rainfall down the Yangtze River Valley. J. Climate, 23, 905 917. Chen, T. C., S. Y. Wang, W. R. Huang, and M. C. Yen, 2004: Variation of the East Asian summer monsoon rainfall. J. Climate, 17, 744 762. Chen, Y. L., and J. Li, 1995: Large-scale conditions favorable for the development of heavy rainfall during TAMEX IOP-3. Mon. Wea. Rev., 123, 2978 3002.

15 DECEMBER 2010 Y U A N E T A L. 6695 Dai, A. G., F. Giorgi, and K. E. Trenberth, 1999: Observed and model-simulated diurnal cycles of precipitation over the contiguous United States. J. Geophys. Res., 104, 6377 6402. Ding, Y. H., 1992: Summer monsoon rainfalls in China. J. Meteor. Soc. Japan, 70, 373 396., and J. C. L. Chan, 2005: The East Asian summer monsoon: An overview. Meteor. Atmos. Phys., 89, 117 142., Y. Zhang, Q. Ma, and G. Q. Hu, 2001: Analysis of the largescale circulation features and synoptic systems in East Asia during the intensive observation period of GAME/HUBEX. J. Meteor. Soc. Japan, 79, 277 300. Geng, B., and H. Yamada, 2007: Diurnal variations of the meiyu/ baiu rain belt. SOLA, 3, 61 64. He, H., and F. Zhang, 2010: Diurnal variations of warm-season precipitation over northern China. Mon. Wea. Rev., 138, 1017 1025. Luo, J., Z. Wang, and J. Zhou, 2003: Geographic causes of the daily variation of rain storm in the mei-yu frontal jet. J. Nanjing Inst. Meteor., 26, 371 377. Matsumoto, J., 1985: Precipitation distribution and frontal zones over East Asia in the summer of 1979. Bull. Dep. Geogr. Univ. Tokyo, 17, 45 61. Onogi, K., and Coauthors, 2007: The JRA-25 reanalysis. J. Meteor. Soc. Japan, 85, 369 432. Qian, W., H. S. Kang, and D. K. Lee, 2002: Distribution of seasonal rainfall in the East Asian monsoon region. Theor. Appl. Climatol., 73, 151 168. Ramage, C. S., 1952: Variation of rainfall over south China through the wet season. Bull. Amer. Meteor. Soc., 33, 308 311. Reed, R. J., 1960: Principal frontal zones of the Northern Hemisphere in winter and summer. Bull. Amer. Meteor. Soc., 41, 591 598. Sorooshian, S., X. Gao, K. Hsu, R. A. Maddox, Y. Hong, H. V. Gupta, and B. Imam, 2002: Diurnal variability of tropical rainfall retrieved from combined GOES and TRMM satellite information. J. Climate, 15, 983 1001. Tao, S., and L. Chen, 1987: A review of recent research on the East Asian summer monsoon in China. Monsoon Meteorology, Monogr. Geol. Geophys., Vol. 7, Oxford University Press, 60 92. Wang, B., and LinHo, 2002: Rainy season of the Asian Pacific summer monsoon. J. Climate, 15, 386 398. Wang, W.-C., W. Gong, and H. Wei, 2000: A regional model simulation of the 1991 severe precipitation event over the Yangtze Huai River Valley. Part I: Precipitation and circulation statistics. J. Climate, 13, 74 92. Yang, G. Y., and J. Slingo, 2001: The diurnal cycle in the tropics. Mon. Wea. Rev., 129, 784 801. Yin, S. Q., D. L. Chen, and Y. Xie, 2009: Diurnal variations of precipitation during the warm season over China. Int. J. Climatol., 29, 1154 1170. Yu, R., Y. Xu, T. Zhou, and J. Li, 2007a: Relation between rainfall duration and diurnal variation in the warm season precipitation over central eastern China. Geophys. Res. Lett., 34, L13703, doi:10.1029/2007gl030315., T. Zhou, A. Xiong, Y. Zhu, and J. Li, 2007b: Diurnal variations of summer precipitation over contiguous China. Geophys. Res. Lett., 34, L01704, doi:10.1029/2006gl028129., J. Li, and H. Chen, 2009: Diurnal variation of surface wind over central eastern China. Climate Dyn., 33, 1089 1097. Zhou, T., and R. Yu, 2005: Atmospheric water vapor transport associated with typical anomalous summer rainfall patterns in China. J. Geophys. Res., 110, D08104, doi:10.1029/2004jd005413.,, H. Chen, A. Dai, and Y. Pan, 2008: Summer precipitation frequency, intensity, and diurnal cycle over China: A comparison of satellite data with rain gauge observations. J. Climate, 21, 3997 4010.