• Alexander, L. V., and Coauthors, 2006: Global observed changes in daily climate extremes of temperature and precipitation. J. Geophys. Res., 111 , D05109. doi:10.1029/2005JD006290.

    • Search Google Scholar
    • Export Citation
  • Chang, C. P., Ed. 2004: East Asian Monsoon. World Scientific Series on Meteorology of East Asia, Vol. 2, World Scientific, 564 pp.

  • Chen, C. S., , and J. M. Huang, 1999: A numerical study of precipitation characteristics over Taiwan Island during the winter season. Meteor. Atmos. Phys., 70 , 167183.

    • Search Google Scholar
    • Export Citation
  • Chen, C. S., , and Y. L. Chen, 2003: The rainfall characteristics of Taiwan. Mon. Wea. Rev., 131 , 13231341.

  • Chen, T. C., , W. R. Huang, , and W. A. Gallus Jr., 2002: An East Asian cold surge: Case study. Mon. Wea. Rev., 130 , 22712290.

  • Groisman, P. Ya, , R. W. Knight, , D. R. Easterling, , T. R. Karl, , G. C. Hegerl, , and V. N. Razuvaev, 2005: Trends in intense precipitation in the climate record. J. Climate, 18 , 13261350.

    • Search Google Scholar
    • Export Citation
  • Hung, C-w, 2007: The History of Meteorological Observatories in Taiwan. Taiwan Interminds Publishing Inc., 293 pp.

  • Hung, C-w, 2009: Temperature discontinuity caused by relocation of meteorological stations in Taiwan. Terr. Atmos. Ocean. Sci., 20 , 607617. doi:10.3319/TAO.2008.07.11.01(A).

    • Search Google Scholar
    • Export Citation
  • Hung, C-w, , H-H. Hsu, , and M-M. Lu, 2004: Decadal oscillation of spring rain in northern Taiwan. Geophys. Res. Lett., 31 , L22206. doi:10.1029/2004GL021344.

    • Search Google Scholar
    • Export Citation
  • Jhun, J. G., , and E. J. Lee, 2004: A new East Asian winter monsoon index and associated characteristics of the winter monsoon. J. Climate, 17 , 711726.

    • Search Google Scholar
    • Export Citation
  • Kinter, J. L., , K. Miyakoda, , and S. Yang, 2002: Recent changes in the connection from the Asian monsoon to ENSO. J. Climate, 15 , 12031215.

    • Search Google Scholar
    • Export Citation
  • Liu, Q., , X. Jiang, , S-P. Xie, , and W. T. Liu, 2004: A gap in the Indo-Pacific warm pool over the South China Sea in boreal winter: Seasonal development and interannual variability. J. Geophys. Res., 109 , C07012. doi:10.1029/2003JC002179.

    • Search Google Scholar
    • Export Citation
  • Mitchell, T. D., , and P. D. Jones, 2005: An improved method of constructing a database of monthly climate observations and associated high resolution grids. Int. J. Climatol., 25 , 693712.

    • Search Google Scholar
    • Export Citation
  • Nakamura, H., , and T. Izumi, 2002: Interannual and decadal modulations recently observed in the Pacific storm-track activity and East Asian winter monsoon. J. Climate, 15 , 18551874.

    • Search Google Scholar
    • Export Citation
  • Okuma, K., 1930: Note on NE monsoon at Boko observatory. J. Meteor. Soc. Formosa, 2 , 3839.

  • Rayner, N. A., , P. Brohan, , D. E. Parker, , C. K. Folland, , J. J. Kennedy, , M. Vanicek, , T. Ansell, , and S. F. B. Tett, 2006: Improved analyses of changes and uncertainties in sea surface temperature measured in situ since the mid-nineteenth century: The HadSST2 dataset. J. Climate, 19 , 446469.

    • Search Google Scholar
    • Export Citation
  • Uppala, S. M., and Coauthors, 2005: The ERA-40 Re-Analysis. Quart. J. Roy. Meteor. Soc., 131 , 29613012.

  • Webster, P. J., , V. O. Magana, , T. N. Palmer, , J. Shukla, , R. A. Tomas, , M. Yanai, , and T. Yasunari, 1998: Monsoons: Processes, predictability, and the prospects for prediction. J. Geophys. Res., 103 , (C7). 1445114510.

    • Search Google Scholar
    • Export Citation
  • Wu, M. J., 1998: Climate variation in Taiwan—Temperature and precipitation. Amer. Sci., 20 , 295318.

  • Zhai, P., , X. Zhang, , H. Wan, , and X. Pan, 2005: Trends in total precipitation and frequency of daily precipitation extremes over China. J. Climate, 18 , 10961108.

    • Search Google Scholar
    • Export Citation
  • Zhang, Y. K., , R. Sperber, , and J. S. Boyle, 1997: Climatology and interannual variation of the East Asian winter monsoon: Results from 1979–95 NCEP/NCAR reanalysis. Mon. Wea. Rev., 125 , 26052619.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    (a) The topography map in East Asia. The white circles are the locations of the weather stations within the large rainfall amount area over southern China. The data from these stations were used in this study. (b) The climatology of 1962–2001 mean DJF circulation at 850 hPa from ERA-40 and seasonal rainfall (gray shading; mm) from the CRU.

  • View in gallery

    DJF rainfall observed in Taipei. (a) The interannual variation of rainfall; (b) the number of days with daily rainfall ≥30 mm; (c) the number of days with rain ≥0.1 mm in each winter season. The interannual variation (thin gray bars; see the right-hand side labels) along with the decadal accumulated values (black bar; see the left-hand side labels) are plotted together in (b) and (c).

  • View in gallery

    The composite daily rainfall pattern during DJF from the high-spatial-resolution ARMTS and conventional weather stations in 1997–2005 for days with negative meridional wind observed at Agincourt Island. These cases are divided into (a) without and (b) with frontal passage cases based on the JMA weather maps. The wind rose analysis (m s−1) and the total number of cases used for composites are also provided.

  • View in gallery

    The topography and locations of the conventional weather stations whose data were used in this study. The stations with open white circles (the station numbers are shaded in the boxes) and filled black circles (the station numbers are not shaded in the boxes) refer to stations sensitive and insensitive to the northeasterly monsoonal wind, respectively. The line legends shown here are for the time series in Figs. 5 and 6.

  • View in gallery

    The interannual variation of DJF rainfall from the (a) four plains stations and (b) three windward slope stations over northern Taiwan; and (c) the average rainfall over southern China. The locations of the rainfall stations used for (a) and (b) are displayed in Fig. 4. The stations from southern China used in (c) are displayed in Fig. 1a.

  • View in gallery

    As in Fig. 5, but for the adjusted interannual variations of DJF rainfall. The rainfall amount for each DJF season is divided by the 1961–90 mean for each station or region. The gray shadings are used to emphasized the early period (1962–81; light gray shading) and latter period (1982–2001; dark gray shading) discussed in the text. The letters E and L refer to the El Niño and La Niña events.

  • View in gallery

    The upper two time series (lines a and b) are the averages of the wind speed at Agincourt Island. The line a (black) is the mean wind speed only for the cases whose meridional components are <0. The line b (gray) is the mean wind for all cases without considering the wind directions. The line c is similar to the line a, but for the average wind speed observed at the 4 island stations in adjacent to Taiwan. The line d is similar to the line b, but for the average of total wind speed at 850 hPa along the coast of East Asia (20°–35°N, 120°–135°E). The thin and thick lines indicate the original interannual variations and the 11-yr running means, respectively.

  • View in gallery

    The DJF rainfall differences (1982–2001 minus 1962–81) over the land region and the SST. The positive values of rainfall differences are stippled from light to dark with an interval of 20 mm, while the gray shading is for the SST with an interval of 0.3°C.

  • View in gallery

    (a) As in Fig. 8, but for the DJF circulation and specific humidity (gray, interval 0.3 g kg−1) at 850 hPa. (b) Moisture divergence at 850 hPa displayed as streamlines of moisture flux and potential (gray shading for positive values).

  • View in gallery

    The time series a is the average of DJF SST over the SCS (15°–22°N, 110°–120°E). The other three time series (b, c, and d) are the average of the DJF specific humidity over the same region in line a at 1000, 925, and 850 hPa, respectively. The gray line plotted at 1981–82 is used to indicate the abrupt change point discussed in the text. The thin and thick lines indicate the original interannual variations and the 11-yr running means, respectively.

  • View in gallery

    A conceptual model to explain rainfall brought by eastward-moving fronts that produce more rainfall to southern China and the plains of northern Taiwan. The shading in the SCS indicates that the increase of SST and the low-level moisture (specific humidity q). Although the northerly wind along the eastern China coast has weakened during the recent decades, the anomalous southerly wind over the SCS can transport more moisture to the southern China area.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 83 83 6
PDF Downloads 72 72 6

Weakening of the Winter Monsoon and Abrupt Increase of Winter Rainfalls over Northern Taiwan and Southern China in the Early 1980s

View More View Less
  • 1 Department of Geography, National Taiwan Normal University, Taipei, Taiwan
  • | 2 Department of Atmospheric Sciences, National Taiwan University, Taipei, Taiwan
© Get Permissions
Full access

Abstract

The rainfall characteristic of the East Asian winter monsoon (EAWM) is less emphasized in previous works. This study reveals that the circulation of the EAWM weakened in recent decades, which results in a decrease of winter rainfall over several windward coastal areas over East Asia including the hills in northern Taiwan. In contrast, there is an abrupt increase of rainfall in southern China and the plains of northern Taiwan during the early 1980s. This is due to the increase in sea surface temperature and lower-troposphere moisture over the South China Sea and the anomalous northward flow that enhances the moisture transport to southern China. Because more moisture is provided for the frontal system that moves eastward, the fronts frequently come with abundant moisture and a well-developed rainband in winter. Therefore, the plains of northern Taiwan receive more rainfall after the 1980s.

Corresponding author address: Dr. Chih-wen Hung, Department of Geography, National Taiwan Normal University 162, Sec. 1, Hoping East Rd. 106 Taipei, Taiwan. Email: cwhung@ntnu.edu.tw

Abstract

The rainfall characteristic of the East Asian winter monsoon (EAWM) is less emphasized in previous works. This study reveals that the circulation of the EAWM weakened in recent decades, which results in a decrease of winter rainfall over several windward coastal areas over East Asia including the hills in northern Taiwan. In contrast, there is an abrupt increase of rainfall in southern China and the plains of northern Taiwan during the early 1980s. This is due to the increase in sea surface temperature and lower-troposphere moisture over the South China Sea and the anomalous northward flow that enhances the moisture transport to southern China. Because more moisture is provided for the frontal system that moves eastward, the fronts frequently come with abundant moisture and a well-developed rainband in winter. Therefore, the plains of northern Taiwan receive more rainfall after the 1980s.

Corresponding author address: Dr. Chih-wen Hung, Department of Geography, National Taiwan Normal University 162, Sec. 1, Hoping East Rd. 106 Taipei, Taiwan. Email: cwhung@ntnu.edu.tw

1. Introduction

The East Asian region is influenced by the large low-level cyclonic and anticyclonic Asian monsoon circulation in summer and winter, respectively. For most of the Asian monsoon regions such as South Asia, the summer monsoon season is referred to as the “rainy season,” while the winter is usually a dry season (Chang 2004; Webster et al. 1998). Over East Asia, there is a similar wet (dry) and warm (cold) monsoonal characteristic for the summer (winter) season. Several previous studies discussed the variabilities of the East Asian winter monsoon (EAWM) from the interannual to the decadal time scale (i.e., Chang 2004; Nakamura and Izumi 2002; Zhang et al. 1997). Most of them used the maximum wind event associated with the cold surges to describe the characteristics of the EAWM. The meridional wind and the pressure gradient along the edge of East Asia related to the strength of the Siberian high and the Aleutian low are usually examined to reveal the variabilities of the EAWM.

Because the cold surge is an important feature of the EAWM, the cold or warm seasonal mean winter temperature is then emphasized to represent the strong or weak winter monsoon over Korea, Japan, and eastern China (e.g., Jhun and Lee 2004). For this reason, the rainfall has drawn less attention in these studies. However, because of the orographic effect over the windward area (the topography is shown in Fig. 1a), the low-level EAWM circulation results in rainfall over the coast of the Japan Sea in Japan, northern Taiwan, the eastern Philippines, and eastern Indo-China peninsula (Fig. 1b; the climatology of 1962–2001 mean December–February circulation at 850 hPa and the rainfall pattern. See more details in later sections). In northern Taiwan, the northerly and northeasterly flow of the EAWM can produce the orographic rain along the coastal regions during winter (Chen and Chen 2003; Chen and Huang 1999; Chen et al. 2002). Therefore, the strength of the northeasterly monsoonal wind has been emphasized since very early winter monsoon study in Taiwan. In the preliminary work by Okuma (1930), the July–August temperature over mainland China is used to predict the strength of the northeasterly winter monsoonal wind in Taiwan. Although the result by Okuma (1930) did not include any physical explanation, the attempt on predicting the winter monsoon represented by the wind speed in Taiwan is recognized. In addition to this orographic effect associated with the monsoonal wind, the frontal activity is another source of rainfall in the East Asian region during winter, especially over southern China and northern Taiwan. Previously, the decadal trend of the EAWM represented by the storm activity was studied (Nakamura and Izumi 2002). However, most EAWM research is less focused on the rainfall variations on the interannual and decadal time scale.

Under the global warming trend, several previous reports show that increasing rainfall intensity has occurred in many regions around the world (e.g., Alexander et al. 2006; Groisman et al. 2005). The positive trend of the rainfall over southern China during winter is pointed out by Zhai et al. 2005. However, the decadal change of the rainfall characteristics associated with the EAWM has not been seriously studied. Taiwan is an island located east of China and strongly influenced by the East Asian monsoon in both summer and winter seasons. Traditional wisdom from the local weather forecasters in Taiwan believed that the stronger northeasterly winter monsoonal wind would produce more rainfall (Wu 1998). However, in the present study, a sudden abrupt increase of winter rainfall across the plains of northern Taiwan is found during the early 1980s, while the weakening of the winter monsoonal wind is observed adjacent to Taiwan.

This abrupt increase of winter rainfall cannot be simply explained by the decadal change of the EAWM wind near northern Taiwan, because the mean winter wind speed along the coast of East Asia weakens in recent decades (shown in later sections). In addition, the rainfall over southern China has a similar abrupt increase as the plains of northern Taiwan experienced. The main purpose of this work is to determine the mechanism that results in this sudden increase of winter rainfall, which is in addition to the orographic lifting. The rest of this study is organized as follows. The data used in this work are provided in section 2. The abrupt increase of winter rainfall in northern Taiwan after 1980s and the associated weakening of the EAWM are discussed in sections 3 and 4, respectively. Finally, the conclusions are presented in section 5.

2. Data

The daily rainfall data (some can be traced back into 1897) from 8 conventional weather stations (Tamsui, Anbu, Taipei, Chutzi Lake, Keelung, Taichung, and Hsinchu) were obtained from the central weather bureau in Taiwan (Hung 2007, 2009), and similar data observed from 377 automatic rainfall stations [the high-spatial-resolution Automatic Rainfall and Meteorological Telemetry System (ARMTS)] since 1997 were used. Daily surface wind data from Agincourt, Penghu, Tungchitao, and Lanyu were also acquired. Agincourt is an island at a distance of 56 km away from northern Taiwan with less topographic influences. Therefore, it is a good location to detect the winter wind variations in adjacent to Taiwan. Several further analyses of wind variation in this study are then based on the observation in this island.

The surface weather map at 0000, 0600, 1200, and 1800 coordinated universal time (UTC) provided by the Japan Meteorological Agency (JMA) in December 1996–February 2005 was also used to define whether there were fronts passing through Taiwan. In addition, several variables from the 40-yr European Centre for Medium-Range Weather Forecasts (ECMWF) Re-Analysis (ERA-40) (Uppala et al. 2005) in 1958–2002 are obtained for further analysis. The monthly Second Hadley Centre Sea Surface Temperature dataset (HadSST2) (Rayner et al. 2006) from 1870 to 2007 provided by the Met Office Hadley center, and the monthly precipitation from the Climatic Research Unit time series (Mitchell and Jones 2005) from 1901–2002 were also used. In the present study, the month January is used to represent the winter each study year. For example, the 1951 December–February (DJF) stands for the season from December 1950 to February 1951. The definition of El Niño and La Niña episodes are acquired from the Climate Prediction Center of the National Oceanographic and Atmospheric Administration (more information is available online at http://www.cpc.noaa.gov) based on the Niño-3.4 index.

3. The abrupt increase of winter rainfalls in northern Taiwan

The increasing trend of extreme rainfall events has been mentioned under the global warming climate (Groisman et al. 2005). Taipei, the largest city located in northern Taiwan, has well-collected daily rainfall data, which makes the long-term trend study available to represent the rainfall variability. The winter DJF rainfall in Taipei for each year are shown in Fig. 2a. An abrupt increase of rainfall can be seen after the 1980s. For further analysis of this abrupt change, the number of days whose daily rainfall is larger than or equal to 30 mm and the total rain days in each winter season (0.1 mm is used as a threshold to define the rain day) for Taipei are shown in Figs. 2b and 2c, respectively. Although the number 30 mm in Fig. 2b is arbitrarily selected, we choose this value partly because of the heavy rainfall definition, 15 mm h−1, from the Central Weather Bureau (CWB). We double this number to make it a threshold value for the daily amount (30 mm) and display it in Fig. 2b. The interannual variations (thin gray bars) along with the decadal accumulated values (wide black bars) since 1900 are plotted together in Figs. 2b and 2c. The decadal accumulation is calculated every 10 years from 1900. For example, the first decadal accumulated bar is the sum of total number from 1900 to 1909. Apparently, the winter rain day is significantly decreased in the second half of the twentieth century, and this tendency is continued to the early twenty-first century, while the total winter rainfall and the heavy rainfall events suddenly increased in the early 1980s. Since the heavier winter rainfall in Taiwan are usually related to the rainband associated with the frontal passage, this implies that the further investigation of rainfall types is required to understand the major reason that causes the abrupt increase of winter rainfall.

Previously, Hung et al. (2004) found that the spring rain in Taipei along with several adjacent stations has a significant decadal oscillation that is highly correlated with the Pacific decadal oscillation (PDO) over the North Pacific Ocean. The oscillated ocean signal of the PDO during the late 1970s coincides with the 1976/77 shift, which is a very large climatic jump in the Pacific. As described in Kinter et al. 2002, the 1976/77 climatic jump is related to the atmospheric circulation over the entire North Pacific Ocean. The correlation between the Asian monsoon and the ENSO is then dramatically changed after 1976/77. Kinter et al. 2002 discussed this change mainly from the Asian monsoon rainfall in June–August (JJA), which is not the same as the spring rain over northern Taiwan (Hung et al. 2004) or the winter monsoon rainfall described here. The winter rainfall in Taipei observed in Fig. 2 has no significant decadal oscillation. Instead, it show a sudden abrupt increase in the early 1980s.

Although traditional wisdom believed that the stronger northeasterly winter monsoonal wind results in a heavier rainfall (Wu 1998), the station data from northern Taiwan seem not to support this view for all regions that receive winter rain in Taiwan. Before further analysis, a background check for these conventional weather stations should first be done. There are some stations located on windward slope regions sensitive to the northeasterly monsoonal wind, but some others are located on the plains areas without the same characteristic. The composite analysis for the daily rainfall pattern from high spatial automatic rainfall stations and conventional weather stations in 1997–2005 is shown in Fig. 3. For all the northerly wind days (with negative meridional wind υ at Agincourt), the cases without and with fronts passing through are shown in Figs. 3a and 3b, respectively. Note that the JMA surface weather maps were acquired here to define the days with or without fronts in the Taiwan region. The wind rose analysis is also provided for each composite group. The Fig. 3a shows that the rainfall region is confined in the northern and the northeastern coast of Taiwan where the dynamic lifting due to the topographic obstruction is an important mechanism. Since there are no fronts passing through in these cases, the major large-scale circulation pattern is the northeasterly monsoonal wind as the wind rose analysis from Agincourt shows. However, the rainfall can be seen island-wide in Fig. 3b. This implies that the front along with a rainband results in the rainfall for a large area of Taiwan. The wind from Agincourt for these cases in Fig. 3b is northerly as the wind rose plot shows. For further analysis in the following, the topography and the locations of the conventional weather stations whose data will be used in this study are plotted in Fig. 4. These weather stations in Fig. 4 are provided with symbols and lines that will be used in Figs. 5 and 6.

All seven stations on the island of Taiwan are divided into two groups with open white and filled black circles in Fig. 4. It can be seen that the three stations with open white circles are all in the mountains or windward slope regions and in the northern coast of Taiwan, while others with filled black circles are located on the plains, which requires a frontal rainband to bring rainfall. The total DJF rainfall amounts shown for the four plains stations (Fig. 5a) are much smaller than the three windward slope stations (Fig. 5b). We calculated the correlation coefficients (CC) of the rainfall and the wind speed in Agincourt and show them in Table 1. The high correlation (CC = 0.57 ∼ 0.62) between rainfall from the windward stations and the wind speed at Agincourt agrees with the traditional view from the local weather forecasters in Taiwan that the increasing rainfall over northern Taiwan are mainly due to the increase of the northeasterly monsoonal wind speed. However, Table 1 also indicates that the plains stations do not have good correlations between rainfall and the monsoonal wind (CC = 0.04 ∼ 0.38). It is concluded that the costal windward stations (with open circles, the station numbers are shaded) and plains stations (with filled black circles, the station numbers are not shaded) in Fig. 4 are stations sensitive and insensitive to the northeasterly monsoonal wind, respectively.

To compare the rainfall characteristics between northern Taiwan and southern China, the rainfall average over southern China (the area indicated in Fig. 1b) is shown in Fig. 5c. The abrupt rainfall increase in the early 1980s in the plains of northern Taiwan is similar to the southern China variations. The increase of the winter rainfall over southern China in recent decades is indicated by Zhai et al. 2005. The correlation coefficients between the rainfall over southern China and the stations in Taiwan were calculated and shown in Table 1. The stations over the plains regions in Taiwan have a higher correlation with rainfall over southern China (CC = 0.65 ∼ 0.74) than do the windward slope stations (CC = 0.31 ∼ 0.34). This implies that the sensitive windward stations in northern Taiwan have different rainfall characteristics from the plains stations, which, on the other hand, have similar rainfall variations to the southern China region.

To clearly show the sudden rainfall increase in the early 1980s, we present a further analysis of rainfall data from the weather stations. Because the ranges of rainfall amount from each station are not the same (as shown in Fig. 5), a simple statistic analysis is applied here to adjust the values. The rainfall amount for each DJF season divided by the 1961–90 means for each station [the period 1961–90 is chosen here based on the discussions in the Intergovernmental Panel on Climate Change Fourth Assessment Report (IPCC AR4)]. The 1961–90 DJF means for Taipei, Tamsui, Hsinchu, Taichung, Chutzi Lake, Anbu, and Keelung and the mean for the southern China region are 305, 398, 265, 122, 879, 1021, 1106, and 153 mm, respectively. The interannual variations of rainfall divided by the above values from four plains stations, three windward slope stations, and the average rainfall over southern China are shown in Figs. 6a, 6b, and 6c, respectively. The result shows that all four plains stations and southern China had sudden rainfall increases in the early 1980s (the same as the abrupt increase in Taipei shown in Fig. 2a), while the three windward slope stations did not have such an abrupt change.

To answer the reason for this sudden increase of rainfall, one may suspect that El Niño and La Niña have impacts on this winter rainfall variations in Figs. 6a and 6c because some studies suggested the relationship between the El Niño and the EAWM (e.g., Zhang et al. 1997). For this concern, the El Niño and La Niña years after 1980 were denoted as E and L in Fig. 6a. The interannual variations of the winter rainfall for these plains stations do not have a one-to-one relationship with El Niño and La Niña. Although large El Niño events did occur before 1980, a similar winter signal was not observed in Taiwan. This indicates that El Niño and La Niña events were not the main factors for the abrupt rainfall increase in the early 1980s on the decadal time scale. On the other hand, the question of why the abrupt increase of the winter rainfall over the plains regions in Taiwan occurred in the early 1980s and not over the mountains and coastal windward slope area becomes important.

4. The weakening of EAWM and the increase of the moisture supply from the South China Sea

The climatology of 1962–2001 mean DJF circulation at 850 hPa from ERA-40 and the rainfall pattern from the Climatic Research Unit (CRU) (Fig. 1b) clearly show that the EAWM circulation results in rainfall over the coast of the Japan Sea in Japan, northern Taiwan, the eastern Philippines, and eastern Indo-China peninsula because of the orographic lifting effects. In the anticyclonic circulation of the EAWM, the mean wind along the coast of eastern China is mainly the northerly and northeasterly. Using the data from Agincourt Island, it is found that the mean total wind speed (without considering the wind directions) and the mean wind speed with northerly component (meridional component υ is negative) associated with the EAWM during DJF significantly weakens in recent decades, especially after the late 1970s (Fig. 7). In addition, we have acquired another three island stations adjacent to Taiwan to compare with the one single station result. The three stations are Penghu, Tungchitao, and Lanyu (locations shown in Fig. 4). The averaged wind speed from all 4 island stations (Agincourt, Penghu, Tungchitao, and Lanyu) has a similar decreasing trend as one single station, Agincourt. The ERA-40 850-hPa wind speed near northern Taiwan (25°–30°N, 120°–125°E, not shown) or along the coast of East Asia (20°–35N, 120°–135°E) also have a similar tendency. All these results indicate the weakening of the winter monsoon strength, which is represented by the wind speeds.

From the previous analysis, we can conclude that rainfall from the plains region in northern Taiwan and southern China show a sudden increase in the early 1980s (Figs. 6a and 6c). However, the monsoonal wind speed already started to decrease after the late 1970s and continued this decreasing trend in the early 1980s. These findings indicate that the early 1980s is an important abrupt change point for the rainfall characteristics associated with the EAWM in northern Taiwan and southern China and need further investigations.

For this reason, using 1981–82 as a point to divide the second half of the twentieth century into the early period (1962–81; light gray shading in Fig. 6) and the latter period (1982–2001; dark gray shading in Fig. 6), the difference (latter-minus-early period) can be examined to present the contrast of this abrupt change. In Fig. 8, the differences of rainfall over the land region and the SST for two periods are shown. Similar differences of the 850-hPa circulation for two periods and the specific humidity at 850 hPa are provided in Fig. 9a. These differences in two periods show a strong southerly wind anomaly, which indicates that the original anticyclonic circulation of the EAWM gets weaker after 1980s (same as the wind observed in Fig. 7). In addition, it also indicates that the weakening of the EWAM circulation resulted in a decrease of rainfall over the coast of the Japan Sea in Japan, the eastern Philippines, and the eastern Indo-China peninsula. However, there is a tendency for increasing rainfall in southern China and northern Taiwan.

The decrease of wind speed over East Asia along the coast of the southern China and Taiwan area would not cause the increase of rainfall in windward slopes of northern Taiwan. The observation in Fig. 6b shows that the rainfall over the windward slopes of northern Taiwan has been slightly decreasing over the past half century. This characteristic is similar to those windward regions such as the eastern Philippines and eastern Indo-China peninsula. On the other hand, the rainfall over southern China increased after the 1980s (Fig. 8; also Fig. 6c), which is the same as the plains in northern Taiwan (Fig. 6a). To answer the question of why the plains stations in northern Taiwan and southern China receive more rainfall in recent decades, the moisture source over the South China Sea (SCS) for the eastward-moving fronts should be examined.

In recent decades, because of global warming, the total moisture increased around the world, but the regional trend of it may not be the same. Figure 8 shows that the SST over the SCS has increased as most oceanic areas in the tropics. As the mean DJF SST over the SCS is getting warmer, the low-level (850 hPa) specific humidity (q; from ERA-40) over this region is increasing as expected (Fig. 9a; same as 925 and 1000 hPa, not shown). In addition, the northward anomalous wind transports more moisture (Fig. 9a) from the SCS to the southern China area where the eastward-moving front received abundant moisture to develop the rainband. The moisture convergence at 850 hPa also covered Taiwan and southern China (Fig. 9b), which suggests that the moisture supply increased after the early 1980s.

To show the time evolution for these related variables in Figs. 8 and 9, the time series for the SST and low-level specific humidity over the SCS (15°–22°N, 110°–120°E) are displayed in Fig. 10. The DJF SST (HadSST2) over the SCS remains at about 23.5°C from the early 1960s to the early 1980s, but it suddenly increases about 1°C within the most recent 20 years. This increasing tendency is similar to the global warming trend observed in the IPCC AR4 report, but the 1°C magnitude in about 20 years is larger than the global warming trend (about 0.17°C decade−1 in the most recent 25 yr). Liu et al. 2004 suggested that the weakened winter monsoon is associated with a spindown of the SCS circulation and a deceleration of its western boundary current, which can result in a warming over the SCS.

Because of the abrupt increase of the SCS SST in the early 1980s, the 1000 ∼ 850 hPa specific humidity (ERA-40) over the same region increases steadily (Fig. 10). Along with the weakening of the EAWM circulation, these evidences show that the moisture supply from the SCS to the southern China and Taiwan regions has increased in recent decades. A conceptual model provided in Fig. 11 displays the increase of the SST and low-level moisture over the SCS with the weakening of the EAWM and the enhancing of the moisture transport from the SCS to southern China. These explain why the plains of northern Taiwan and the southern China area have the abrupt increase of rainfall after the early 1980s.

5. Conclusions

From the long-term station records, the climatic change of the winter rain in northern Taiwan can be summarized as follows. In the winter season, northern Taiwan is under the influence of the EAWM. The mechanisms to produce rainfall are different for the windward slopes in the northern coast of Taiwan and the plains area. For the windward coastal areas of northern Taiwan, the orographic lifting is very important. As long as the wind direction is favorable for this mechanism with a moisture supply from the adjacent ocean, it is easy to have rain. High correlation between the northerly wind speed and the rainfall in the windward coastal area agrees with this traditional wisdom from the local weather forecasters. However, for the plains regions, the major contributor to rainfall is frontal systems. The fronts come with a well-organized rainband that has abundant moisture and can bring a widely distributed rainfall to northern Taiwan. The long-term climate data records show that the windward slopes in northern Taiwan do not have a significant change of winter rainfall over the past 100 years, but the plains receive a rainfall abrupt increase after the 1980s.

A conceptual model is provided in this study to explain how the rainfall brought by the eastward-moving fronts can produce more rainfall to southern China and the plains over northern Taiwan (Fig. 11). In recent decades, because of the increase of SST over the SCS, the moisture in this region increased significantly after the early 1980s, which is similar to the global warming trend but with higher magnitudes. The decrease of the EAWM over East Asia in recent decades and the increase of moisture over the SCS lead to more moisture transport to southern China. Under this circumstance, more moisture is received by the frontal system that moves eastward from southern China to Taiwan. When the fronts frequently come with abundant moisture and well-developed rainband, they bring a higher probability of large rainfall occurrence across the plains of northern Taiwan.

This conceptual model is contradictory to the traditional view from the local weather forecasters in Taiwan. It was believed that the increase of rainfall over northern Taiwan is mainly due to the increase of the northeasterly monsoonal wind speed. In this study, we confirm that this traditional view is still valid for the windward coastal regions of northern Taiwan, which is sensitive to the monsoonal wind. However, the plains regions have different characteristics. Our result shows that the strength of the EAWM weakened after early 1980s. The weakening of the EAWM circulation has resulted in the decrease of rainfall over several windward coastal regions such as the eastern Philippines and the eastern Indo-China peninsula but not southern China and the plains areas of northern Taiwan to which fronts bring large rainfall. The main factor for the abrupt increase of rainfall over the plains of northern Taiwan is the increase of anomalous moisture supply from the SCS to the southern China area from which the fronts are moving. Although the present study provided an answer to why the rainfall abruptly increased over the plains of northern Taiwan and southern China, further studies are required to understand what causes the northerly anomalous flow after the 1980s to transport the increased moisture under the global warming climate, which may relate to the sudden warming of the SCS in 1980s. This anomalous flow is possibly influenced by the PDO in the northern Pacific and the pressure gradient between the Siberian high and the Aleutian low. However, this is an intriguing but unsolved question.

Acknowledgments

The authors thank reviewers for their useful comments on the manuscript. The suggestions from H.-H. Hsu and J.-Y. Yu are especially appreciated. Special thanks are extended to Y.-M. Li for his help in data preparation. This work was supported by the National Science Council under Grants NSC 97-2111-M-003-004-MY2 and NSC 98-2625-M-492-011.

REFERENCES

  • Alexander, L. V., and Coauthors, 2006: Global observed changes in daily climate extremes of temperature and precipitation. J. Geophys. Res., 111 , D05109. doi:10.1029/2005JD006290.

    • Search Google Scholar
    • Export Citation
  • Chang, C. P., Ed. 2004: East Asian Monsoon. World Scientific Series on Meteorology of East Asia, Vol. 2, World Scientific, 564 pp.

  • Chen, C. S., , and J. M. Huang, 1999: A numerical study of precipitation characteristics over Taiwan Island during the winter season. Meteor. Atmos. Phys., 70 , 167183.

    • Search Google Scholar
    • Export Citation
  • Chen, C. S., , and Y. L. Chen, 2003: The rainfall characteristics of Taiwan. Mon. Wea. Rev., 131 , 13231341.

  • Chen, T. C., , W. R. Huang, , and W. A. Gallus Jr., 2002: An East Asian cold surge: Case study. Mon. Wea. Rev., 130 , 22712290.

  • Groisman, P. Ya, , R. W. Knight, , D. R. Easterling, , T. R. Karl, , G. C. Hegerl, , and V. N. Razuvaev, 2005: Trends in intense precipitation in the climate record. J. Climate, 18 , 13261350.

    • Search Google Scholar
    • Export Citation
  • Hung, C-w, 2007: The History of Meteorological Observatories in Taiwan. Taiwan Interminds Publishing Inc., 293 pp.

  • Hung, C-w, 2009: Temperature discontinuity caused by relocation of meteorological stations in Taiwan. Terr. Atmos. Ocean. Sci., 20 , 607617. doi:10.3319/TAO.2008.07.11.01(A).

    • Search Google Scholar
    • Export Citation
  • Hung, C-w, , H-H. Hsu, , and M-M. Lu, 2004: Decadal oscillation of spring rain in northern Taiwan. Geophys. Res. Lett., 31 , L22206. doi:10.1029/2004GL021344.

    • Search Google Scholar
    • Export Citation
  • Jhun, J. G., , and E. J. Lee, 2004: A new East Asian winter monsoon index and associated characteristics of the winter monsoon. J. Climate, 17 , 711726.

    • Search Google Scholar
    • Export Citation
  • Kinter, J. L., , K. Miyakoda, , and S. Yang, 2002: Recent changes in the connection from the Asian monsoon to ENSO. J. Climate, 15 , 12031215.

    • Search Google Scholar
    • Export Citation
  • Liu, Q., , X. Jiang, , S-P. Xie, , and W. T. Liu, 2004: A gap in the Indo-Pacific warm pool over the South China Sea in boreal winter: Seasonal development and interannual variability. J. Geophys. Res., 109 , C07012. doi:10.1029/2003JC002179.

    • Search Google Scholar
    • Export Citation
  • Mitchell, T. D., , and P. D. Jones, 2005: An improved method of constructing a database of monthly climate observations and associated high resolution grids. Int. J. Climatol., 25 , 693712.

    • Search Google Scholar
    • Export Citation
  • Nakamura, H., , and T. Izumi, 2002: Interannual and decadal modulations recently observed in the Pacific storm-track activity and East Asian winter monsoon. J. Climate, 15 , 18551874.

    • Search Google Scholar
    • Export Citation
  • Okuma, K., 1930: Note on NE monsoon at Boko observatory. J. Meteor. Soc. Formosa, 2 , 3839.

  • Rayner, N. A., , P. Brohan, , D. E. Parker, , C. K. Folland, , J. J. Kennedy, , M. Vanicek, , T. Ansell, , and S. F. B. Tett, 2006: Improved analyses of changes and uncertainties in sea surface temperature measured in situ since the mid-nineteenth century: The HadSST2 dataset. J. Climate, 19 , 446469.

    • Search Google Scholar
    • Export Citation
  • Uppala, S. M., and Coauthors, 2005: The ERA-40 Re-Analysis. Quart. J. Roy. Meteor. Soc., 131 , 29613012.

  • Webster, P. J., , V. O. Magana, , T. N. Palmer, , J. Shukla, , R. A. Tomas, , M. Yanai, , and T. Yasunari, 1998: Monsoons: Processes, predictability, and the prospects for prediction. J. Geophys. Res., 103 , (C7). 1445114510.

    • Search Google Scholar
    • Export Citation
  • Wu, M. J., 1998: Climate variation in Taiwan—Temperature and precipitation. Amer. Sci., 20 , 295318.

  • Zhai, P., , X. Zhang, , H. Wan, , and X. Pan, 2005: Trends in total precipitation and frequency of daily precipitation extremes over China. J. Climate, 18 , 10961108.

    • Search Google Scholar
    • Export Citation
  • Zhang, Y. K., , R. Sperber, , and J. S. Boyle, 1997: Climatology and interannual variation of the East Asian winter monsoon: Results from 1979–95 NCEP/NCAR reanalysis. Mon. Wea. Rev., 125 , 26052619.

    • Search Google Scholar
    • Export Citation

Fig. 1.
Fig. 1.

(a) The topography map in East Asia. The white circles are the locations of the weather stations within the large rainfall amount area over southern China. The data from these stations were used in this study. (b) The climatology of 1962–2001 mean DJF circulation at 850 hPa from ERA-40 and seasonal rainfall (gray shading; mm) from the CRU.

Citation: Journal of Climate 23, 9; 10.1175/2009JCLI3182.1

Fig. 2.
Fig. 2.

DJF rainfall observed in Taipei. (a) The interannual variation of rainfall; (b) the number of days with daily rainfall ≥30 mm; (c) the number of days with rain ≥0.1 mm in each winter season. The interannual variation (thin gray bars; see the right-hand side labels) along with the decadal accumulated values (black bar; see the left-hand side labels) are plotted together in (b) and (c).

Citation: Journal of Climate 23, 9; 10.1175/2009JCLI3182.1

Fig. 3.
Fig. 3.

The composite daily rainfall pattern during DJF from the high-spatial-resolution ARMTS and conventional weather stations in 1997–2005 for days with negative meridional wind observed at Agincourt Island. These cases are divided into (a) without and (b) with frontal passage cases based on the JMA weather maps. The wind rose analysis (m s−1) and the total number of cases used for composites are also provided.

Citation: Journal of Climate 23, 9; 10.1175/2009JCLI3182.1

Fig. 4.
Fig. 4.

The topography and locations of the conventional weather stations whose data were used in this study. The stations with open white circles (the station numbers are shaded in the boxes) and filled black circles (the station numbers are not shaded in the boxes) refer to stations sensitive and insensitive to the northeasterly monsoonal wind, respectively. The line legends shown here are for the time series in Figs. 5 and 6.

Citation: Journal of Climate 23, 9; 10.1175/2009JCLI3182.1

Fig. 5.
Fig. 5.

The interannual variation of DJF rainfall from the (a) four plains stations and (b) three windward slope stations over northern Taiwan; and (c) the average rainfall over southern China. The locations of the rainfall stations used for (a) and (b) are displayed in Fig. 4. The stations from southern China used in (c) are displayed in Fig. 1a.

Citation: Journal of Climate 23, 9; 10.1175/2009JCLI3182.1

Fig. 6.
Fig. 6.

As in Fig. 5, but for the adjusted interannual variations of DJF rainfall. The rainfall amount for each DJF season is divided by the 1961–90 mean for each station or region. The gray shadings are used to emphasized the early period (1962–81; light gray shading) and latter period (1982–2001; dark gray shading) discussed in the text. The letters E and L refer to the El Niño and La Niña events.

Citation: Journal of Climate 23, 9; 10.1175/2009JCLI3182.1

Fig. 7.
Fig. 7.

The upper two time series (lines a and b) are the averages of the wind speed at Agincourt Island. The line a (black) is the mean wind speed only for the cases whose meridional components are <0. The line b (gray) is the mean wind for all cases without considering the wind directions. The line c is similar to the line a, but for the average wind speed observed at the 4 island stations in adjacent to Taiwan. The line d is similar to the line b, but for the average of total wind speed at 850 hPa along the coast of East Asia (20°–35°N, 120°–135°E). The thin and thick lines indicate the original interannual variations and the 11-yr running means, respectively.

Citation: Journal of Climate 23, 9; 10.1175/2009JCLI3182.1

Fig. 8.
Fig. 8.

The DJF rainfall differences (1982–2001 minus 1962–81) over the land region and the SST. The positive values of rainfall differences are stippled from light to dark with an interval of 20 mm, while the gray shading is for the SST with an interval of 0.3°C.

Citation: Journal of Climate 23, 9; 10.1175/2009JCLI3182.1

Fig. 9.
Fig. 9.

(a) As in Fig. 8, but for the DJF circulation and specific humidity (gray, interval 0.3 g kg−1) at 850 hPa. (b) Moisture divergence at 850 hPa displayed as streamlines of moisture flux and potential (gray shading for positive values).

Citation: Journal of Climate 23, 9; 10.1175/2009JCLI3182.1

Fig. 10.
Fig. 10.

The time series a is the average of DJF SST over the SCS (15°–22°N, 110°–120°E). The other three time series (b, c, and d) are the average of the DJF specific humidity over the same region in line a at 1000, 925, and 850 hPa, respectively. The gray line plotted at 1981–82 is used to indicate the abrupt change point discussed in the text. The thin and thick lines indicate the original interannual variations and the 11-yr running means, respectively.

Citation: Journal of Climate 23, 9; 10.1175/2009JCLI3182.1

Fig. 11.
Fig. 11.

A conceptual model to explain rainfall brought by eastward-moving fronts that produce more rainfall to southern China and the plains of northern Taiwan. The shading in the SCS indicates that the increase of SST and the low-level moisture (specific humidity q). Although the northerly wind along the eastern China coast has weakened during the recent decades, the anomalous southerly wind over the SCS can transport more moisture to the southern China area.

Citation: Journal of Climate 23, 9; 10.1175/2009JCLI3182.1

Table 1.

The correlation coefficients between the monthly DJF rainfall data in the stations (Tamsui, Taipei, Hsinchu, Taichung, Anbu, Chutzi Lake, and Keelung) and the mean DJF wind speed at Agincourt. Similar analysis is also applied to the DJF rainfall over southern China.

Table 1.
Save