Secular and seasonal variations of winter monsoon weather patterns in Japan since the early 20th century

Authors

  • Junpei Hirano,

    Corresponding author
    1. Department of Geography, Tokyo Metropolitan University, Minami-ohsawa 1-1, Hachioji, Tokyo 192-0397, Japan
    • Department of Geography, Tokyo Metropolitan University, Minami-Ohsawa 1-1, Hachioji-Shi, Tokyo 192-0397, Japan.
    Search for more papers by this author
  • Jun Matsumoto

    1. Department of Geography, Tokyo Metropolitan University, Minami-ohsawa 1-1, Hachioji, Tokyo 192-0397, Japan
    Search for more papers by this author

Abstract

Winter weather conditions in Japan are characterized by a clear contrast between the regions on the Japan Sea side and those on the Pacific Ocean side of the Japanese Islands. Weather conditions in the Japan Sea side are characterized by snowfall or rainfall under winter monsoon outbreak conditions. However, the Pacific Ocean side has fine weather because the central ridges of mountains block snow clouds coming from the Japan Sea. This particular weather pattern can easily be recognized from the daily precipitation distributions. In this study, we used the daily precipitation data recorded by the Japan Meteorological Agency since 1901 to define winter monsoon weather pattern based on the contrast in the spatial distribution of daily precipitation between the regions on the Japan Sea side and those on the Pacific Ocean side. On the basis of seasonal changes in the occurrence frequencies of winter monsoon weather pattern, we delimited the beginning and ending pentads of the winter monsoon season for each year. We aimed to clarify the characteristics of long-term variations in the lengths of the winter monsoon season and in the seasonal march of winter weather conditions in Japan since the early 20th century. The results revealed a significant decreasing trend in the lengths of the winter monsoon season since the early 20th century. We also found that the occurrence frequencies of winter monsoon weather pattern have been decreasing significantly since the early 20th century. In addition to the long-term decreasing trend, we detected an abrupt decrease in the occurrence frequencies of winter monsoon weather pattern in the mid-1980s, which suggests a weakening of the winter monsoon after the mid-1980s. Copyright © 2010 Royal Meteorological Society

1. Introduction

The East Asian winter monsoon is one of the most prominent monsoon circulation systems existing over the Northern Hemisphere. It strongly influences local weather conditions over the East Asian region. For example, it brings abundant snow to the Japan Sea side of the Japanese Islands. Active cold surges influence not only the weather conditions over East Asia but also the occurrence of heavy rainfall in the Indochina Peninsula (Yokoi and Matsumoto, 2008). Many studies have been undertaken to identify secular variations in the East Asian winter monsoon (Chen et al., 2000; Gong et al., 2001; Jhun and Lee, 2004; Wang et al., 2009). However, there have been very few studies on the long-term variations in the seasonal march of winter weather conditions and the length of the winter monsoon season in East Asia. In East Asia, information on both secular and seasonal variations in weather conditions is necessary to understand the features of climatic variations in detail, because weather conditions often change within a small timescale of less than 1 month or one season (Inoue and Matsumoto, 2007). Consequently, investigating long-term variations in the seasonal march of winter weather conditions is important for gaining a better understanding of winter climate variations over East Asia.

In Japan, several attempts have been made to clarify long-term variations in the lengths of the winter monsoon season and in the seasonal march of winter weather conditions in the latter half of the 20th century. Yamakawa (1988) attempted to detect the typical winter monsoon-type pressure pattern around Japan based on the inspection of daily weather charts. On the basis of seasonal transitions in the occurrence frequencies of winter monsoon-type pressure pattern, he investigated variations in the dates of the beginning and the end of the winter monsoon season from 1941 to 1985. He pointed out that the lengths of the winter monsoon season and the occurrence frequencies of winter monsoon-type pressure pattern were increasing during the period 1941–1985. However, his process for detecting daily pressure pattern types involves some subjectivity. Inoue and Matsumoto (2003) investigated secular and seasonal variations in sunshine rates in Japan for the period 1951–2000. They defined the dates of the beginning and the end of the winter monsoon season based on the seasonal transitions in the differences between sunshine rates in the regions on the Japan Sea side and those on the Pacific Ocean side. As a result, they discovered that the lengths of the winter monsoon season have been increasing during the late 20th century, despite the warming trend in winter temperatures.

However, no studies have investigated secular and seasonal variations in winter weather conditions in Japan from the beginning of the 20th century. This is because of the lack of daily meteorological data in Japan in the first half of the 20th century. Monthly or seasonal meteorological data are inappropriate for studying secular and seasonal variations in weather conditions in detail. Daily or sub-monthly data are necessary. Kimura et al. (2009) proposed an objective method for detecting typical winter monsoon-type pressure pattern by using the daily mean sea level pressure (SLP) data obtained from the Japanese 25-year Reanalysis Project (JRA-25) data. They indicated that typical winter monsoon-type pressure pattern could be detected with high accuracy by using the JRA-25 reanalysis data. However, we cannot apply their methods to the early 20th century because JRA-25 reanalysis data are not available for this period. Consequently, we cannot also delimit the dates of the beginning and the end of the winter monsoon season since the early 20th century based on daily pressure data.

Recently, the Japan Meteorological Agency (JMA) digitized daily precipitation data recorded since 1901 at 51 meteorological observatories in Japan. The quality of these precipitation data was checked by the observation department of the JMA. These precipitation data are homogeneous since 1901 (Fujibe et al., 2006). Typical daily weather patterns under the winter monsoon flow are characterized by a clear contrast between the weather in the regions on the Japan Sea side and those on the Pacific Ocean side, and this can easily be recognized from the daily precipitation distribution (Suzuki 1962). Therefore, the homogeneous daily precipitation data observed by the JMA are valuable in investigating secular and seasonal variations in winter weather conditions in detail.

Knowledge of past climatic conditions is important in understanding the present climate and predicting the future climate. It is therefore necessary to identify the characteristics of long-term variations in the length of the winter monsoon season and in the seasonal march of winter weather conditions in Japan since the early 20th century.

In this study, by using the JMA daily precipitation data since 1901, we defined the winter monsoon weather pattern based on the contrast in the spatial distribution of daily precipitation between the regions on the Japan Sea side and those on the Pacific Ocean side. On the basis of the seasonal changes in the occurrence frequencies of winter monsoon weather pattern, we delimited the beginning and ending pentads of the winter monsoon season for each year. We aimed to clarify the characteristics of the long-term variations in the seasonal march of winter weather conditions and in the lengths of the winter monsoon season in Japan from the early 20th century.

In Section 2, we describe the data and analysis procedures used in this study. In Section 3, long-term variations in the occurrence number of winter monsoon weather pattern since 1901 are investigated. In Section 4, secular and seasonal variations in the occurrence frequencies of winter monsoon weather pattern are investigated. In Section 5, we attempt to delimit the beginning and ending pentads of the winter monsoon season for each year since the early 20th century on the basis of seasonal variations in the occurrence frequencies of winter monsoon weather pattern. Secular variations are also discussed in Section 5. Finally, conclusions and discussion are presented in Section 6.

2. Data and analysis procedures

In this study, we attempted to detect typical winter monsoon weather pattern on the basis of the climatic division by Suzuki (1962). Suzuki (1961) reported the typical daily precipitation distributions over the Japanese Islands under winter monsoon outbreak conditions. On the basis of the spatial distribution of daily precipitation patterns, Suzuki (1962) classified Japan into three different climatic regions. The detailed characteristics of the climatic regions identified by Suzuki (1962) are as follows: (1) The Japan Sea side region: when the winter monsoon is in progress, precipitation is commonly observed over this region; (2) Transient region: when the winter monsoon is strong (weak), precipitation is (not) observed in this region and (3) The Pacific Ocean side region: no precipitation is observed under the winter monsoon flow because the high central ridges of backbone mountain block snow clouds coming from the Japan Sea side. Usually, fine dry weather prevails over this region. Because both Suzuki (1962) and this study use daily precipitation data, the climatic divisions identified by Suzuki (1962) can be applied to this study for recognizing particular daily weather pattern under winter monsoon outbreak condition.

Daily precipitation data for the period 1901–2009 were obtained from the CD-ROMs released by the JMA. Among the 48 JMA stations located in the main islands of Japan, we selected six stations for the Japan Sea side region on the basis of Suzuki (1962). For the Pacific Ocean side region, we selected nine stations that are located in the Pacific Ocean side region of Suzuki's (1962) divisions. We did not use the other 33 JMA stations because most of them are located in the transient region between the Japan Sea side and the Pacific Ocean side regions. In this transient region, precipitation is observed only during the day, when the winter monsoon is strong (Suzuki, 1962). Consequently, precipitation data from these 33 stations are thought to be inappropriate for recognizing the typical winter monsoon-type weather pattern. As a result, 15 JMA stations were used in this study (Figure 1).

Figure 1.

Locations of daily precipitation data used in this study. Filled squares indicate stations located in the Japan Sea side region. Filled triangles indicate stations located in the Pacific Ocean side region

To recognize particular daily weather pattern under the winter monsoon flow, we focused on the contrast in daily weather conditions between the six stations located in the region of the Japan Sea side and the nine stations located in the region on the Pacific Ocean side. Under the winter monsoon flow, the spatial variability in the amount of daily precipitation is large, even in the Japan Sea side region. In particular, heavy precipitation events in the Japan Sea side region are very localized (Kodama et al., 1995). Consequently, a small threshold value of daily precipitation is considered appropriate for detecting synoptic-scale weather patterns under the winter monsoon flow. In this study, we selected 1 mm/day as the threshold value for daily precipitation. Then, we defined ‘winter monsoon weather pattern’ on the basis of the following criteria: (1) precipitation (≥1 mm/day) is observed for at least four neighbouring stations in the region on the Japan Sea side and (2) precipitation (≥1 mm/day) is not observed in any of the nine stations located in the region on the Pacific Ocean side.

To investigate the relationship between daily weather patterns and daily synoptic pressure patterns around Japan, we used daily mean SLP of the NCEP/NCAR reanalysis dataset (Kalnay et al., 1996). Although this data set starts from 1948, we used daily mean SLP data from 1980/1981 to 2008/2009 because remarkable discrepancy between SLP of the NCEP/NCAR data and that of observed pressure data have been reported for the Asian region prior to the late 1970s (Wu et al., 2005). Yang et al. (2002) also pointed out that the quality of NCEP/NCAR reanalysis data over Asian region may be low prior to 1968. In order to obtain a general SLP pattern for the days with the winter monsoon weather pattern, we selected all winter monsoon weather pattern days from 1980/1981 to 2008/2009. Then, we constructed a composite map of the daily mean SLP pattern for the winter monsoon pattern days during this period. Figure 2 shows a composite map of the SLP field for winter monsoon weather pattern days during the period from 1980/1981 to 2008/2009. In this figure, a low SLP area exists in the northern Pacific Ocean and a high SLP area exists over the Eurasian Continent. This indicates that winter monsoon weather pattern corresponds to the typical winter monsoon-type pressure pattern (known as ‘west-high, east-low-type pressure pattern’).

Figure 2.

Composite map of daily mean SLP fields for winter monsoon weather pattern day. Unit is hPa. Contours show the deviation from the spatial mean. The contour interval is 2 hPa. Solid contours indicate positive deviations from the spatial mean. Broken contours indicate negative deviations from the spatial mean

We investigated the features of secular and seasonal variations in the occurrence frequencies of winter monsoon weather pattern since the early 20th century. On the basis of seasonal variations in occurrence frequencies, we attempted to delimit the beginning and ending pentads of the winter monsoon season for each year since 1901.

3. Long-term variations in the occurrence number of winter monsoon weather pattern

In this section, we report the occurrence numbers of winter monsoon weather pattern during the cold season (1 October to 30 April) for each year since 1901. Figure 3(a) shows the long-term variations in the occurrence number of the winter monsoon weather pattern.

Figure 3.

(a) Long-term variations in the occurrence numbers of winter monsoon weather pattern. The grey line indicates the 5-year running mean. The broken line indicates the linear regression lines of the occurrence numbers. (b) Time series of the Lepage statistic (HK) when the sampling numbers of the two groups are 12 years. The 95% confidence level of this test is indicated by a broken line

To check the reliability of our analysis, we compared variations in the occurrence number of the winter monsoon weather pattern with those of the winter monsoon-type pressure pattern identified from daily synoptic weather charts (Yamakawa and Yoshino, 2002). Figure 4 represents this relationship for the period 1981/1982 to 1999/2000. In this figure, numbers of the ‘weather pattern’ are generally lower than those of ‘pressure pattern’. Yamakawa and Yoshino (2002) did not consider the strength of the winter monsoon when classifying the daily pressure pattern types. Consequently, the ‘winter monsoon-type pressure pattern’ by Yamakawa and Yoshino (2002) is thought to include a weak winter monsoon-type pressure pattern, which usually prevails over northern Japan. When such a pattern prevails over northern Japan, precipitation is usually observed only in the northern part of the Japan Sea side region. In this study, we did not detect such precipitation patterns because there are few JMA stations in the northern part of the Japan Sea side region (Figure 1). Consequently, undetected weak winter monsoon-type pressure pattern in our study is thought to be the main reason why the number of days of ‘weather’ pattern is generally smaller than those of the ‘pressure’ pattern. The purpose of this study is to detect the typical synoptic-scale winter monsoon weather pattern, which prevails over the entire area of the Japanese islands. Therefore, we do not consider the difference in the number of days showing the ‘weather pattern’ and ‘pressure pattern’ to be a serious problem in our analysis. Although these differences exist, variations in the occurrence number of the winter monsoon weather pattern and in winter monsoon-type pressure pattern agree and show a high positive correlation (r = 0.86). In the late 1960s, JMA changed the precipitation measurement instruments from cylinder-type rain gauge instruments to tipping-bucket rain gauge instruments. In order to check whether this change affected our results, we run the standard normal homogeneity test (Alexandersson, 1986), the Buishand range test (Buishand, 1982) and the Pettitt test (Pettit, 1979) over the time series of the occurrence numbers of the winter monsoon weather pattern, shown in Figure 3(a). In the test results, we could not detect any homogeneity breaks around the late 1960s. Consequently, we confirmed that the change in the type of rain gauge instruments in the late 1960s did not affect the results of our study. These facts indicate that investigating the occurrence number of the winter monsoon weather pattern is effective in understanding the long-term variations of winter synoptic weather conditions. We then examined the features of the long-term trends and secular variations in the occurrence number of the winter monsoon weather pattern based on Figure 3(a). To identify the characteristics of the long-term trends, we applied the Mann–Kendall rank statistic (Kendall, 1938) to this time series. We found a significant (p < 0.05) decreasing trend in the occurrence number of the winter monsoon weather pattern since 1901. This implies that the winter monsoon has become weaker during the past 109 years. In order to detect abrupt increases or decreases in the occurrence numbers, we applied the Lepage test (Lepage, 1971) to the time series of the occurrence number. The Lepage test is a nonparametric test used to investigate significant differences between two samples. The Lepage test has often been used to detect discontinuous climate changes (Yonetani, 1992, 1993; Inoue and Matsumoto, 2007). The results of the Lepage test (sample number n1 = n2 = 12) are shown in Figure 3(b). The Lepage statistic (HK), which indicates a degree of discontinuity, has a peak in 1986/1987, which is significant at a 95% confidence level. We also checked the results of the Lepage test when the sample numbers (n1 and n2) changed from 11 to 15 (figures not shown), and we confirmed that the peak in the mid-1980s unchanged. Therefore, discontinuity in the mid-1980s is thought to be statistically robust. The occurrence number of the winter monsoon weather pattern abruptly decreased after this discontinuity in the mid-1980s (Figure 3(a)). This suggests sudden weakening of the winter monsoon after the mid-1980s. Except for this abrupt change, there were no significant changes in the occurrence number of the winter monsoon weather pattern. In the early 20th century, high occurrence numbers were observed in the early 1910s, the late 1920s and the early 1940s. Low occurrence numbers were observed in the late 1940s and the early 1950s. After the early 1950s, occurrence numbers increased until the early 1980s. This implies that the winter monsoon became stronger from the early 1950s to the early 1980s. Yamakawa (1988) indicated that occurrence frequencies of winter monsoon-type pressure pattern were increasing in the period from the 1940s to the early 1980s, which roughly corresponds to the increase in the occurrence number of the winter monsoon weather pattern observed in this study.

Figure 4.

Relationship between the occurrence numbers of winter monsoon weather pattern and the occurrence numbers of winter monsoon-type pressure pattern for the period 1981/1982 to 1999/2000. Data for the occurrence numbers of the winter monsoon-type pressure pattern were obtained from Yamakawa and Yoshino (2000)

After the abrupt decrease in the mid-1980s, the occurrence number of the winter monsoon weather pattern was about 10 days less than those before the mid-1980s. It should be noted that this abrupt decrease almost coincides with the abrupt winter climatic changes over the Northern Hemisphere in the mid-1980s (Watanabe and Nitta, 1999; Yasunaka and Hanawa, 2002). Yasunaka and Hanawa (2008) indicated that winter temperatures in Japan abruptly increased after the mid-1980s. Jhun and Lee (2004) and Wang et al. (2009) indicated that East Asian winter monsoon weakened after the mid-1980s, corresponding to the abrupt decrease in the occurrence number of the winter monsoon weather pattern observed in this study. In the late 1990s, the occurrence numbers increased again slightly and were about 8 days higher than those in the early 1990s.

4. Secular and seasonal variations in the occurrence frequencies of winter monsoon weather pattern

In this section, we investigated secular and seasonal variations in the occurrence frequencies of winter monsoon weather pattern since the early 20th century. First, the occurrence frequencies of winter monsoon weather pattern for each pentad from pentad 56 (3–7 October) to pentad 24 (26–30 April) were calculated for each year of the study period. For leap years, we omitted the data for 29 February. Occurrence frequencies were then smoothed by the five-pentad running mean and the 5-year running mean in order to represent secular and seasonal variations in the occurrence frequencies. The smoothed data are shown as an isopleth diagram (Figure 5).

Figure 5.

Isopleth diagram representing seasonal and secular variations in occurrence frequencies of winter monsoon weather pattern. Unit is %. Contour interval is 10%

Before the 1920s, the occurrence frequency peaks were seen in early winter (December to early January). In mid-winter (mid-January to early February), the occurrence frequencies were lower than those in early winter. It is notable that these characteristics were not observed in any other years in the study period. These features suggest that an unusual seasonal march of winter monsoon occurred before the 1920s. In the 1930s and the 1940s, the occurrence frequency peaks were observed in mid-January. In the early 1950s, occurrence frequencies over the whole winter season became lower than those in the 1940s. After the late 1950s, the occurrence frequencies over the whole winter season increased again. In particular, two periods—the late 1950s to the late 1960s and the late 1970s to the early 1980s—were characterized by high occurrence frequencies over the whole winter season. After the mid-1980s, occurrence frequencies over the whole winter season decreased, and the peak of the occurrence frequencies became less prominent than that of the early 1980s.

5. Long-term variations in the lengths of winter monsoon season since the early 20th century

In this section, we attempted to delimit the beginning and ending pentads of the winter monsoon season based on the secular and seasonal variations in the occurrence frequencies of the winter monsoon weather pattern shown in Figure 5. Several studies have attempted to delimit the boundaries of the natural seasons in Japan, including the dates of the beginning and the end of the winter monsoon season (Maejima, 1967; Yoshino and Kai, 1977; Yamakawa, 1988; Inoue and Matsumoto, 2003). As was mentioned in Section 1, Yamakawa (1988) delimited the dates of the beginning and the end of the winter monsoon season from 1941 to 1985 based on seasonal changes in the occurrence frequencies of winter monsoon-type pressure pattern. Inoue and Matsumoto (2003) delimited the dates of the beginning and the end of the winter monsoon season from 1951 to 2000 based on the seasonal changes in sunshine rates. However, their study period starts after the mid-20th century, which seems too short a timescale to distinguish decadal scale variations from the long-term trends. Consequently, it is necessary to delimit the beginning and ending pentads of the winter monsoon season for each year since the early 20th century based on the seasonal variations in the occurrence frequencies of winter monsoon weather pattern defined in this study. The characteristics of the long-term variations in the beginning and ending pentads, and in the lengths of the winter monsoon season since the early 20th century, were therefore investigated in this section. The situations after the mid-20th century were compared with those reported in the previous studies (Yamakawa, 1988; Inoue and Matsumoto, 2003).

To delimit the beginning and ending pentads of the winter monsoon season, occurrence frequencies of winter monsoon weather pattern, smoothed by five-pentad and 5-year running mean (Figure 5) were used. Because Inoue and Matsumoto (2003) used pentad-based data of sunshine rates, smoothed by five-pentad and 5-year running mean, to delimit the seasonal boundaries, smoothed data shown in Figure 5 are useful to compare our results with those by Inoue and Matsumoto (2003). On the basis of the isopleth diagram shown in Figure 5, beginning (ending) pentad in each year was defined as follows: the first (last) pentad when the occurrence frequency exceeds the threshold value (20%) for more than three consecutive pentads.

Figure 6 shows the long-term variations in the beginning and ending pentads of the winter monsoon season. Early starts to the winter monsoon season were observed in the early 1910s, the 1920s and the late 1930s. Late starts were observed in the early 1930s and the early 1990s. The period of the ending pentads was more stable than that of the beginning pentads. Standard deviations of the beginning and ending pentads were 1.75 and 1.71 pentads, respectively.

Figure 6.

Long-term variations in the beginning and ending pentads of winter monsoon season. Broken lines indicate the linear regression lines of the beginning and ending pentads

Figure 7 shows the long-term variations in the lengths of the winter monsoon season during the study period. We observe long winter monsoon season in the early 1910s, the late 1920s, the 1940s and the early 1980s. Short winter monsoon season is observed in the early 1900s, the early 1930s, the early 1950s and the early 1990s. In order to identify the characteristics of the long-term trend, we applied the Mann–Kendall rank statistic (Kendall, 1938) to the time series of the beginning and ending pentads and to the lengths of the winter monsoon season. This revealed that the lengths of the winter monsoon season have been decreasing significantly (p < 0.01) during the study period, although no significant trends were observed in the time series of the beginning and ending pentads.

Figure 7.

Long-term variations in the length of winter monsoon season. Broken line indicates the linear regression lines of length

Yamakawa (1988) pointed out that the lengths of the winter monsoon season increased after the 1970s. In this study, winter monsoon season in the mid-1970s and the early 1980s was longer than those in the 1950s and the early 1960s (Figure 7), which roughly corresponds to Yamakawa's (1988) results. Yamakawa (1988) did not analyse variations in the lengths of the winter monsoon season before 1941 and after 1985. He did not therefore identify the long-term decreasing trends in the lengths of the winter monsoon season since the early 20th century observed in this study.

On the basis of the contrast between the seasonal changes in the sunshine levels observed in the regions on the Japan Sea side and the Pacific Ocean side, Inoue and Matsumoto (2003) delimited the dates of the beginning and the end of the winter monsoon season for each year since 1951. These data indicated that the lengths of the winter monsoon season had increased since the 1980s, which is contrary to the decreasing trend in the lengths found by this study. Moreover, their results seem to contradict the weakening of the winter monsoon after the mid-1980s pointed out by Jhun and Lee (2004) and Wang et al. (2009). Therefore, more detailed studies of the relationship between variations in sunshine rates and the winter monsoon are needed.

6. Discussion and conclusions

In this study, by using daily precipitation data recorded in Japan since 1901, we defined the winter monsoon weather pattern based on the contrast in spatial distribution of daily precipitation between the regions on the Japan Sea side and those on the Pacific Ocean side. First, we investigated secular and seasonal variations in the occurrence frequencies of winter monsoon weather pattern since the early 20th century. We then delimited the beginning and ending pentads of the winter monsoon season for each year on the basis of seasonal variations in the occurrence frequencies of winter monsoon weather pattern. We discussed the features of the long-term variations in winter weather conditions and in the lengths of the winter monsoon season since the early 20th century. The main results of this study are as follows:

  • (1)The occurrence numbers of the winter monsoon weather pattern during the cold season (October to April) have been decreasing significantly since the early 20th century.
  • (2)An abrupt decrease in the occurrence frequencies of winter monsoon weather pattern was observed in the mid-1980s, which suggests sudden weakening of the winter monsoon after the mid-1980s.
  • (3)Before the 1920s, the occurrence frequencies of the winter monsoon weather pattern in mid-winter (mid-January to early February) were lower than those in early winter (December to early January). These features suggest that an unusual seasonal march of the winter monsoon occurred before the 1920s.
  • (4)The lengths of the winter monsoon season have been decreasing significantly since the early 20th century.

As mentioned in Section 5, Inoue and Matsumoto (2003) indicated that the lengths of the winter monsoon season have been increasing in the latter half of the 20th century, which is contrary to the decreasing trend in the lengths of the winter monsoon season found by this study. They delimited the dates of the beginning and the end of the winter monsoon season based on the seasonal changes in the difference in sunshine levels between the regions on the Japan Sea side and the Pacific Ocean side. However, weather conditions in these two regions are influenced not only by the variations of the winter monsoon but also by extratropical cyclones, which pass along the southern coast of Japan and over the Japan Sea (Tasaka, 1988). Consequently, it is not clear whether increases in the differences in sunshine levels between the two regions can be explained by variations in the winter monsoon alone. In this study, however, winter monsoon weather pattern corresponds to the typical winter monsoon-type pressure pattern (Figure 2). Thus, differences in the data and in the methods are thought to be the main reasons for the different results obtained in this study and those obtained by Inoue and Matsumoto (2003).

On the physical mechanisms for the secular and seasonal variations in winter weather conditions, many recent studies have proposed possible physical mechanisms of the winter monsoon variability. Nakamura et al. (2002) found out enhancement of storm track activity in mid-winter over north western pacific after the late 1980s with weakening of the Asian winter monsoon, which is in good agreement with a decrease in occurrence frequency of the winter monsoon weather pattern after the mid-1980s found by us. Panagiotopoulos et al. (2005) found out a steep weakening trend of Siberian high after 1978, which started earlier than the abrupt decrease in the occurrence number of the winter monsoon weather pattern in the mid-1980s (Figure3(a)). According to Jhun and Lee (2004), variability in the intensity of the East Asian winter monsoon is influenced by both the Siberian high and the Aleutian low. Consequently, the winter-time weather pattern in Japan is thought to be affected not only by Siberian high strength but also by the Aleutian low. Further studies are necessary to reveal the relationship between changes of winter-time weather condition in Japan and long-term variability of the Aleutian low.

A number of studies simulated the change of Asian winter monsoon using the multi-general circulation models with global warming issues. Hu et al. (2000) investigated the Asian winter monsoon response to global warming through a long-term integration of transient greenhouse warming using a global-coupled atmosphere/ocean/sea ice/land surface climate model. They indicated that the intensities of the Asian winter monsoon are clearly reduced in the global warming scenario. Their results indicated a pronounced linear decrease in Asian winter monsoon strengths after the 1990s. Kimoto (2005) simulated changes in East Asian circulation patterns by using coupled climate models. His results also indicated weakening of the East Asian winter monsoon under the global warming scenario. Hori and Ueda (2006) investigated the impact of global warming on the East Asian winter monsoon by using coupled atmosphere–ocean global climate models. They indicated weakening of East Asian winter monsoon and northern shift of Aleutian low. In this study, we found that occurrence frequencies of winter monsoon weather pattern and lengths of the winter monsoon season have been decreased in a centennial timescale. It is therefore necessary to investigate whether these changes relate with global warming.

In the early 20th century (before the 1920s), the occurrence frequencies of the winter monsoon weather pattern in mid-winter were lower than those in early winter. Further work is needed to identify the seasonal changes in atmospheric circulation patterns that can explain this unusual seasonal march of weather conditions.

Our results indicate that we can clarify detailed features of winter climatic variations based on daily weather distribution patterns. This method can be applied to the reconstruction of winter synoptic weather conditions in Japan before the early 19th century, periods for which modern meteorological data are not available. Because historical daily weather documents for the 18th and 19th centuries can be obtained from various sources in Japan (Yoshimura, 1993; Mizukoshi, 1993; Hirano and Mikami, 2008), it is possible to identify days with typical winter monsoon weather pattern in historical periods based on daily weather distribution pattern reconstructed from historical documents. Further studies are therefore necessary to clarify these features of climatic variations not only for the modern instrumental period but also for historical periods.

Acknowledgements

We would like to thank all the members of the Climatology Laboratory of Tokyo Metropolitan University for their useful comments and suggestions. This study is supported by the Global Environment Research Fund B092 and A092 of the Japanese Ministry of the Environment and by the grant-in-aid for the scientific research from the Japanese Ministry of Education, Culture, Sports, Science and Technology (No. 20240075).

Ancillary