1. Introduction
The Pacific–Japan (PJ) teleconnection pattern (Nitta 1987) is a dominant driver of East Asian summer monsoon (EASM) circulation variability (e.g., Kosaka et al. 2011), leading to a meridional dipole of rainfall anomalies over the western North Pacific and East Asia. After the late 1990s or the early 2000s, however, the PJ pattern has weakened or is rather absent (Huang et al. 2018; Xu et al. 2019; Li and Lu 2020; Sun et al. 2021), weakening EASM precipitation variability during the past two decades, as the El Niño–Southern Oscillation (ENSO) influence on the PJ pattern (Xie et al. 2010) has weakened over recent decades (e.g., Kubota et al. 2016). As a result, the PJ pattern is no longer a dominant driver of EASM precipitation variability, and East Asian summer climate displays a marked change after the late 1990s.
Despite the weakening of the PJ pattern, western Japan has frequently experienced heavy rainfall events, especially during mei-yu–baiu period over the past several years, when severe natural disasters frequently occurred with devastating socioeconomic impacts. The major causes of each individual heavy rainfall event are diversely suggested, such as oceanic evaporation (Sekizawa et al. 2019) and upper-tropospheric trough (Yokoyama et al. 2020) in early summer 2018, tropical Indian Ocean (Takaya et al. 2020) and subtropical moistening (Zhao et al. 2021) in early summer 2020. Climatologically, the mei-yu–baiu rainband, which is a major part of EASM system, is accompanied by moisture transport from the subtropics (e.g., Ninomiya and Murakami 1987). However, the weakening of the PJ pattern in the positive phase contributes to suppress moisture transport from the subtropics to the mei-yu–baiu rainband through the influence of the western Pacific subtropical high (WPSH) (Huang et al. 2018; Xu et al. 2019; Li and Lu 2020). Thus, despite a short time scale, the frequent occurrence of heavy rainfall events in western Japan is difficult to explain by the conventional framework of the PJ pattern.
The EASM precipitation variability is also affected by the Silk Road pattern (Enomoto et al. 2003) through the meandering of the Asian jet (Kosaka et al. 2011, 2012; Wu et al. 2016; Chowdary et al. 2019). The relationship between the Asian Jet and EASM precipitation is investigated by using climate model outputs (Horinouchi et al. 2019). The Silk Road pattern can be regarded as the Eurasian sector of the circumglobal teleconnection pattern (Ding and Wang 2005) and is suggested to link to heavy rainfall event over western Japan in early summer 2020 (Horinouchi et al. 2021). Although the Silk Road pattern also has a decadal shift in the late 1990s (Wang et al. 2017; Stephan et al. 2019), oceanic variability does not directly force the Silk Road pattern (Kosaka et al. 2012; Stephan et al. 2019). Instead, recent Arctic warming is suggested to lead to amplification of circumglobal wave trains through the weakening of the westerly jet (Coumou et al. 2018). Indeed, midlatitude summer circulation has weakened in conjunction with a reduction in the poleward temperature gradient (Coumou et al. 2015). Thus, climate change needs to be considered for better understanding of the Silk Road pattern or circumglobal teleconnection pattern.
As noted above, recent EASM precipitation variability has been difficult to understand within the conventional framework of the PJ and Silk Road patterns. Besides, most previous studies focus on each individual heavy rainfall event. Therefore, why heavy rainfall events occur frequently in western Japan remains unclear, except for an increase of water vapor due to global warming (Imada et al. 2020). To address this question, investigating EASM or mei-yu–baiu precipitation variability itself should be essential, such as Kosaka et al. (2011) who described interannual variability in mei-yu–baiu precipitation from 1979 to 2007. The present study updates the mei-yu–baiu precipitation variability in a view of interdecadal changes or a modulation of interannual variability under climate change and tries to understand the frequent occurrence of heavy rainfall events. We reveal that cumulus convection has been more dominant for the characteristic of EASM precipitation variability including mei-yu–baiu rainfall than frontal structure over the subtropical seas since the mid-2000s, which may favor the frequent occurrence of heavy rainfall events in western Japan.
The rest of the paper is organized as follows: section 2 describes the data and analysis method. Section 3 documents EASM precipitation variability based on both a linear trend and empirical orthogonal function (EOF) analysis. Sections 4 and 5 investigate interdecadal changes in atmospheric circulation associated with EASM precipitation and a decadal shift in EASM precipitation variability including mei-yu–baiu rainfall, respectively. In section 6 we present our conclusions and discuss the frequent occurrence of heavy rainfall events.
2. Data and method
Atmospheric data are from the Japanese 55-year Reanalysis (JRA-55; Kobayashi et al. 2015). We use two global precipitation products based on a rain gauge over land from Climate Prediction Center (CPC; Chen et al. 2008) with a 0.5° × 0.5° horizontal grids and an integration of satellite data and a rain gauge from the Global Precipitation Climatology Project (GPCP; Huffman et al. 1997; Adler et al. 2003) with a 2.5° × 2.5° horizontal grids. Sea surface temperature (SST) data are obtained from the Hadley Centre Global Sea Ice and Sea Surface Temperature (HadISST; Rayner et al. 2003).
This study mainly focuses on June–August (JJA) including mei-yu–baiu period (June–July) for 42 years from 1979 to 2020. To determine the major modes of precipitation variability, we performed an EOF analysis of JJA GPCP precipitation in EASM region (20°–50°N, 100°–160°E). Such EOF analysis based on precipitation yields almost the same pattern as covariability of precipitation and atmospheric circulation (Kosaka et al. 2011). To examine interdecadal changes, we performed running correlation analysis after linearly detrending. The significance test used in this study is a standard two-tailed t test with degrees of freedom based on the number of years. To further examine decadal changes in the relationship between atmospheric circulation and precipitation, we defined GPCP precipitation averaged over around the Southwest Islands (25°–30°N, 125°–135°E) as SWI precipitation and Yangtze River basin (28°–34°N, 110°–122°E) as YRB precipitation, respectively, normalized by each standard deviation. Vertical p velocity variance is based on 6-hourly JRA-55 data, diabatic heating is derived as a residual of thermodynamic equation, and wave activity flux is formulated by Takaya and Nakamura (2001).
3. Changes in EASM precipitation
a. Linear trends
First, we explore how EASM precipitation has increased over the period of 1979–2020. Figures 1a and 1b show CPC and GPCP precipitation trends, respectively. CPC precipitation significantly increases over southeastern China and western Japan, whereas GPCP precipitation increases over southeastern China and around the Southwest Islands in southern Japan. No significantly increased GPCP precipitation is identified over western Japan. In particular, the GPCP precipitation has the largest increase over the southern East China Sea and around the Southwest Islands, which is also captured by recent satellite data during mei-yu–baiu period for 1998–2019 (Takahashi and Fujinami 2021). As heavy rainfall events generally occur in a short time scale, such as a week or so at the longest, daily precipitation variance can be regarded as an indicator of heavy rainfall events. Compared with precipitation trends (Fig. 1a), western Japan has much larger precipitation variance than southeastern China (Fig. 1c). In fact, the number of precipitation days over 50 mm, which has a close relationship with daily precipitation variance (Fig. 1 in the online supplemental material), has markedly increased over western Japan (not shown). Figure 1d compares time series of precipitation variance over western Japan with those over YRB (surrounded by black rectangle in Fig. 1c) as the major mei-yu–baiu rainfall regions. Daily precipitation variance over western Japan had been close to that over YRB by the early 1990s, but then has rapidly increased with large interannual variability, especially since the early 2000s, consistent with recent frequent occurrence of heavy rainfall events. By contrast, YRB precipitation variance remains relatively constant, although YRB rainfall in 2020 was the heaviest in the past half century (e.g., Zhou et al. 2021). The difference in both regions implies that the characteristic of precipitation variability over western Japan has changed over the recent decades. Upward vertical velocity at 300 hPa from the JRA-55 dataset also has the largest trend over around the Southwest Islands in southern Japan (Fig. 2a), well capturing GPCP precipitation trends (Fig. 1b). Stronger vertical velocity variance based on 6-hourly data further extends into western Japan (Fig. 2b), suggesting the contribution of short-term-scale rainfall to the increased precipitation.
The spatial pattern of linear trends (decade−1) in (a) CPC and (b) GPCP precipitation (mm day−1) for 1979–2020. (c) As in (a), but for daily precipitation variance. Hatching denotes the 95% confidence level. (d) Time series of daily precipitation variance averaged over western Japan [larger black rectangle in (a) and (c); 30°–35°N, 129°–137°E] and the Yangtze River basin [black rectangle in (c); 28°–34°N, 110°–122°E]. Thick black lines indicate 11-yr running means.
Citation: Journal of Climate 36, 8; 10.1175/JCLI-D-22-0292.1
The spatial pattern of linear trends (decade−1) in (a) vertical p velocity (Pa s−1) and (b) its variance (10−3 Pa2 s−2) at 300 hPa. Dotted contours indicate the 95% confidence level.
Citation: Journal of Climate 36, 8; 10.1175/JCLI-D-22-0292.1
b. EOF analysis
The largest increased precipitation over around the Southwest Islands is also supported by an EOF analysis. Figures 3a and 3b show the first and second EOF modes (EOF-1 and EOF-2, respectively) of GPCP precipitation over 20°–50°N, 100°–160°E (explaining 16.4% and 14.5% of the total variance, respectively). EOF-1 represents the PJ pattern, similar result to previous studies (Kosaka et al. 2011; Matsumura et al. 2016), and the first principal component (PC) has a slightly downward trend with a maximum in 2020 (Fig. 3c). EOF-2 with a distinct upward trend (Fig. 3d) well captures not only the GPCP precipitation trends (Fig. 1b) but also the CPC precipitation trends over western Japan (Fig. 1a). To examine whether both EOF modes accounts for precipitation variability over YRB and western Japan, we show 11-yr running correlations of CPC precipitation with the PCs in Fig. 4. Correlation of YRB precipitation with PC-1 had mostly reached at the 95% significance level prior to 2000, but rapidly fell in the early 2000s, and has recovered over the recent years. The rapidly decreased correlation in the early 2000s is consistent with the weakening of the PJ pattern (Huang et al. 2018; Xu et al. 2019; Li and Lu 2020; Sun et al. 2021), although the recent recovery might reflect the minimum (2018) and maximum (2020) in PC-1. Similarly, correlation of precipitation over western Japan with PC-1 had also reached at the significance level during the 1990s, with a higher correlation than that with PC-2 by the late 2000s. Since then, however, the correlation with PC-2 has exceeded that with PC-1 and held robust since 2010. This result supports that the PJ pattern is no longer a leading driver of precipitation variability over western Japan.
(a) First and (b) second EOF modes of precipitation (mm day−1) and (c),(d) corresponding PCs for the period 1979–2020. Thick black lines indicate a linear trend.
Citation: Journal of Climate 36, 8; 10.1175/JCLI-D-22-0292.1
(a) The 11-yr running correlations of CPC precipitation over the Yangtze River basin with PC-1 (blue line) and PC-2 (red line). (b) As in (a), but for western Japan. Dashed line denotes the 95% confidence level.
Citation: Journal of Climate 36, 8; 10.1175/JCLI-D-22-0292.1
4. Changes in atmospheric circulation
This section investigates what causes the precipitation EOF-2. While the precipitation EOF-1 (i.e., the PJ pattern) is associated with ENSO in the preceding winter (r = 0.41; above the 99% confidence level) (e.g., Xie et al. 2010), there is no EOF-2 correlation with the ENSO (r = −0.04), as in Kosaka et al. (2011). Figures 5a and 5b show trends in zonal mean temperature and its regressed anomalies onto PC-2, respectively. Interestingly, the EOF-2 has nothing to do with tropical warming but is closely related to Arctic warming, in contrast to the EOF-1. The Asian jet weakening coincides with Arctic warming in both linear trends (Fig. 5c) and EOF-2 (Fig. 5d), likely through a reduction in the poleward temperature gradient (Coumou et al. 2018). As the Silk Road pattern is linked to the Asian jet, we show trends in streamfunction at 300 hPa and its regressed anomalies onto PC-2 in Fig. 6. No wave train trend is identified over the Pacific Ocean. Whereas in the regression onto EOF-2 (Fig. 6b), wave train from East Asia near western Japan toward North America is quite evident, and the Silk Road pattern over the Eurasian continent is weaker. In particular, the Rossby waves propagate northeastward from 120° to 130°E near the jet entrance and then reach the eastern Pacific Ocean along the westerly jet. The substantial difference between linear trends and the EOF-2 in upper-tropospheric circulation means that we cannot further discuss changes in EASM precipitation and atmospheric circulation using a simple linear trend. To better understand EASM precipitation variability, the rest of the paper examines a modulation of interannual variability by comparing the first half and second half of our analysis periods.
(a) Zonal mean temperature trends (K decade−1) and (b) regressed anomalies of zonal mean temperature (K) onto PC-2. (c),(d) As in (a) and (b), but for zonal wind (m s−1) averaged over 0°–150°E. Dotted contours indicate the 95% confidence level and solid black contours in (c) and (d) indicate climatological zonal wind (10 and 20 m s−1).
Citation: Journal of Climate 36, 8; 10.1175/JCLI-D-22-0292.1
(a) The spatial pattern of linear trends (decade−1) and (b) regressed anomalies onto PC-2 in zonally asymmetric streamfunction (106 m2 s−1) at 300 hPa. Dotted contours indicate the 95% confidence level and solid black contours indicate climatological zonal wind at 15, 18, and 21 m s−1. Vectors denote wave activity flux calculated by using the regressed anomalies.
Citation: Journal of Climate 36, 8; 10.1175/JCLI-D-22-0292.1
Figures 7a and 7b show regressed anomalies of detrended 300-hPa streamfunction onto SWI precipitation averaged over a rectangular region for 1979–99 and 2000–20. The Silk Road–like pattern over Eurasia is slightly visible in the earlier and recent decades but is not as clear as the component of trends (Fig. 6). Similar pattern to the EOF-2, wave train is generated from East Asia near western Japan toward North America in the recent decades, although no significant wave train is identified in the earlier decades. The wave activity propagates upward from the midtroposphere over East Asia around 120°–130°E, rather than the upper troposphere, which intensifies anticyclonic anomaly in the upper troposphere over the east of Japan around 150°E (supplemental Fig. 2). The Asian jet has a zonally uniform structure in the earlier decades, whereas in the recent decades East Asia is located near the jet entrance, which help excite the wave propagation toward North America along the jet, as intensified cyclonic circulation over East Asia acts as a wave source near the jet entrance, like the Silk Road pattern (Enomoto et al. 2003). This result suggests that recent EASM precipitation variability enables to act as a forcing of the wave train over the Pacific Ocean.
As in Fig. 6b, but for regression onto SWI precipitation (black rectangle; 25°–30°N, 125°–135°E) for (a) 1979–99 and (b) 2000–20.
Citation: Journal of Climate 36, 8; 10.1175/JCLI-D-22-0292.1
This atmospheric circulation difference in the first half and second half of the periods might be related to the weakening of the PJ pattern. To clarify the difference with the PJ pattern, we show 850-hPa geopotential height anomalies regressed onto YRB precipitation. YRB precipitation variability is associated with the PJ pattern through the southwesterly along the western flank of the WPSH in the earlier decades (Fig. 8a), together with a weaker cyclonic anomaly over Japan (e.g., Kosaka et al. 2012). However, the meridional dipole of circulation anomalies over the western North Pacific and Japan is absent in the recent decades (Fig. 8b), which is in good agreement with previous studies (Huang et al. 2018; Xu et al. 2019; Li and Lu 2020; Sun et al. 2021). Specifically, recent northeastward-intensified WPSH (Matsumura et al. 2015; Matsumura and Horinouchi 2016), which resembles the anomaly of the summer 2020 (e.g., Zhou et al. 2021), is likely to contribute to the absence of the cyclonic anomaly over Japan. Lower-tropospheric circulation associated with SWI precipitation also shows pronounced difference between the earlier and recent decades (Figs. 8c,d). In the earlier decades, there is little cyclonic anomaly over SWI area. In the recent decades, however, wave train appears to be generated northeastward from SWI area, tilting northward with height (Fig. 7b), which reflects a tilt of the jet axis during mei-yu–baiu period. These results indicate the robust relationship between SWI precipitation variability and wave propagation from East Asia. This wave train can help supply moist air toward western Japan through southerlies and locally intensify the WPSH through air–sea interaction in the northwest Pacific (Matsumura et al. 2016).
As in Fig. 7, but for 850-hPa geopotential height (m) and horizontal wind (vectors; m s−1) onto (a),(b) YRB precipitation (black rectangle; 28°–34°N, 110°–122°E) and (c),(d) SWI precipitation.
Citation: Journal of Climate 36, 8; 10.1175/JCLI-D-22-0292.1
5. Decadal shift in EASM precipitation variability
The difference between the first half and second half of the periods in interannual variability is also supported by precipitation EOF analysis. When the EOF analysis is applied to 2000–20, EOF-1 (19.3%) is replaced with increased SWI precipitation pattern and EOF-2 (16.8%) represents the PJ pattern. Although this decadal shift in EASM precipitation variability may be partly explained by the weakening of the PJ pattern, why SWI precipitation markedly increases and the associated wave train is generated from East Asia remains unclear. This section explores the characteristic of EASM precipitation variability.
a. Vertical structures
Figure 9 shows a latitude–vertical section of regressed anomalies of specific humidity, diabatic heating, and equivalent potential temperature (θe) at 130°E onto SWI precipitation. The typical mei-yu–baiu rainband is accompanied by characteristic frontal structures with sharp gradients in specific humidity and θe at the lower to midtroposphere (e.g., Ninomiya and Murakami 1987). Although we focus on JJA including mei-yu–baiu period, such mei-yu–baiu frontal structures on interannual variability are evident in the earlier decades; sharp north–south gradients in specific humidity (Fig. 9a) and θe (Fig. 9c), and narrow heating band with strong convergence by lower-level southerlies (Fig. 9b). When SWI precipitation (25°–30°N) increases, both specific humidity and θe anomalies are small in the lower level (1000–800 hPa), which are likely to be unfavorable for vertically deep convection that are prevalent in the tropics. Specifically, negative θe anomalies, which are the largest in the lower level and tilt northward with height, well correspond to climatologically strong poleward θe gradient (dashed contours). Thus, precipitation variability indicates frontal structures that precipitation increases when the meridional θe gradient is strengthened. In addition, the lower-level southerlies are pronounced, suggesting the contribution of moisture transport from the subtropics and convergence along the frontal rainband.
Latitude–vertical section of regressed anomalies of (a) specific humidity (g kg−1), (b) diabatic heating (K day−1) and meridional circulation (vectors; m s−1 and Pa s−1; vertical component is multiplied by 20), and (c) equivalent potential temperature θe (K) at 130°E onto SWI precipitation for 1979–99. (d)–(f) As in (a)–(c), but for 2000–20. Dotted contours indicate the 95% confidence level. Dashed contours in (c) and (f) denote climatological poleward θe gradient −0.8 K per degree latitude.
Citation: Journal of Climate 36, 8; 10.1175/JCLI-D-22-0292.1
By contrast, the recent decades have little mei-yu–baiu frontal structures on interannual variability. Specific humidity anomalies positively increase the strongest at the lower troposphere and decrease at the north of the moist region (Fig. 9d). Coincident with the lower-tropospheric moistening, lower-level convergence, upward motion, and upper-tropospheric divergence are intensified by enhancing diabatic heating with a wider heating band (Fig. 9e), raising θe from the lower to upper troposphere (Fig. 9f). These results represent a characteristic of cumulus convection, which neutralizes moist instability by supplying moisture from lower to upper troposphere. Unlike the earlier decades, specific humidity and θe anomalies markedly increase in the lower level (900–700 hPa), favorable for deep convection. In addition, negative θe anomalies along climatologically strong poleward θe gradient are significantly weakened especially at the midtroposphere. These results indicate that in the earlier decades frontal structures are essential for precipitation variability, while in the recent decades, deep convection is more dominant for precipitation variability than frontal features. The lower-level cyclonic circulation associated with SWI precipitation (Fig. 8d) can be explained by a response to the enhanced convective heating (Sampe and Xie 2010).
To examine when the SWI precipitation variability changes the characteristic from frontal to cumulus convective structures, Fig. 10a shows vertical distribution of 21-yr running regressions (contours) and trends (shadings) in specific humidity. As regression is based on our defined SWI precipitation, we also show linear trend as a more objective method, although there is the difference between interannual variability and linear trend. The midtroposphere is anomalously the moistest in both the regressions and trends by the mid-2000s. Since the mid-2000s, however, lower-tropospheric moistening has been most pronounced, especially in trends with surface moistening. Furthermore, midtroposphere has gradually moistened with surface moistening. We obtain the similar results to JJA when focusing on June–July during mei-yu–baiu period (Fig. 10b). Thus, considering the distinct mei-yu–baiu frontal structures in the earlier decades (Figs. 9a–c), these results indicate that cumulus convection has been more dominant for the characteristic of precipitation variability including mei-yu–baiu rainfall than frontal structure since the mid-2000s.
Vertical distribution of 21-yr running regressions (contours, with thick zero line) of specific humidity (g kg−1) onto SWI precipitation and 21-yr running trends (shaded; decade−1) in specific humidity at 25°N, 130°E during (a) June–August and (b) June–July.
Citation: Journal of Climate 36, 8; 10.1175/JCLI-D-22-0292.1
b. Convective instability
The atmospheric vertical structures demonstrate the decadal shift in EASM precipitation variability. However, this is based on regression onto our defined SWI precipitation, although specific humidity trends capture the features. Has the characteristic of rainfall changed only over SWI area? Figs. 11a and 11b show correlation of interannual variability between precipitable water and total precipitation. In the earlier decades, local precipitable water does not contributes to local precipitation in the mei-yu–baiu rainband including over SWI (r = 0.18 based on GPCP precipitation). By contrast, in the recent decades when the weakening of the positive PJ pattern acts to suppress moisture transport by reducing horizontal advection of lower-level θe (e.g., Li and Lu 2020), local precipitable water has a close relationship with local precipitation along the mei-yu–baiu rainband, especially over the southern East China Sea and off the Pacific coast of Japan including over SWI (r = 0.65 based on GPCP precipitation), supporting the decadal shift in the characteristic of precipitation variability. This relation is relatively strengthened over China, possibly due to the suppression of moisture transport associated with the weakening of the PJ pattern.
Correlation of interannual variability between precipitable water and total precipitation based on JRA-55 data for (a) 1979–99 and (b) 2000–20. (c),(d) As in (a) and (b), but for surface air temperature (SAT) and the ratio of convective precipitation to total precipitation. (e),(f) As in (c) and (d), but for convective instability defined by the vertical gradient of equivalent potential temperature between 850 and 500 hPa. Contours denote mean total precipitation rate (mm day−1) in (a) and (b) and mean convective precipitation rate (mm day−1) in (c) and (d).
Citation: Journal of Climate 36, 8; 10.1175/JCLI-D-22-0292.1
To further better identify the decadal shift in EASM precipitation variability, we discuss convective instability. Corresponding to Sampe and Xie (2010) based on mei-yu–baiu period, correlation of the ratio of convective to total precipitation (from JRA-55) with surface air temperature (SAT) tends to be high along the mei-yu–baiu rainband in the earlier decades (Fig. 11c). We note that cumulus convective parameterization for reanalysis data [an economical version of the Arakawa–Schubert scheme is used in the JRA-55 (Kobayashi et al. 2015)] diagnoses the occurrence of deep convection, thus being qualitatively meaningful for precipitation variability. This linkage has been strengthened over the East China Sea, around SWI and the far east of Japan in the recent decades (Fig. 11d), indicating that local SAT warming plays a greater role in local convection than the earlier decades. Note that EASM precipitation variability including mei-yu–baiu rainfall is primarily dependent on larger-scale atmospheric forcings, especially on the WPSH and Asian jet, as noted in the introduction. As convective precipitation depends on cumulus convective parameterization, we further show SAT correlation with convective instability defined by the vertical gradient of θe between 850 and 500 hPa in Figs. 11e and 11f. The vertical gradient of θe between 850 and 500 hPa is regarded as a representative of vertical stability in the mei-yu–baiu front (e.g., Ninomiya et al. 2002). Higher correlations along the mei-yu–baiu rainband roughly correspond to the SAT correlation with convective precipitation in the earlier decades (Fig. 11c), confirming the linkage between local convective precipitation and convective instability. In the recent decades, however, SAT correlation with convective instability has markedly fallen along the mei-yu–baiu rainband, especially over the East China Sea and off the Pacific coast of Japan including SWI area, inconsistent with the enhancement of warm surface for convection (Fig. 11d). The correlation stays relatively high over China and rises over northern Japan and the far east of Japan.
To understand why the linkage between SAT and convective instability disappears in the recent decades, we show scatterplot of SST and convective instability over SWI area (Figs. 12a,b), as deep convection depends largely on local SST (e.g., Graham and Barnett 1987; Fu et al. 1994). Corresponding to SAT correlation with convective instability, warmer SST enhances convective instability in the earlier decades (r = 0.58; above the 99% confidence level), whereas in the recent decades there is no linear relation between SST and convective instability. Mean convective instability is almost the same in 1979–99 and 2000–20 (0.8 and 0.83 K 100 hPa−1) but is not a reliable indicator for convection (Sampe and Xie 2010). The difference between the two periods is that SST is mostly over 28°C in the latter periods, which exceeds the SST threshold for convection (e.g., Graham and Barnett 1987; Fu et al. 1994). In the earlier decades, local precipitable water does not contribute to local precipitation in the mei-yu–baiu rainband (Fig. 11a), suggesting that atmospheric forcing accompanied with moisture transport (i.e., the PJ pattern or WPSH) is more dominant for precipitation variability rather than local convection. Under such atmospheric forcing, it may be difficult to apply the SST threshold to midlatitudes like the tropics. By contrast, in the recent decades, local convection is more essential for precipitation variability (Fig. 11b). The linear relation between SST and convective instability abruptly broke in the mid-2000s (Fig. 12c), when the decadal shift in specific humidity also occurred (Fig. 10), and the correlation is close to zero afterward. Similar change to the correlation with SST, interannual variability of convective instability has gradually weakened, reached a minimum in the mid-2000s, and it remains weak afterward (Fig. 12d). Although vertical gradient of θe might not be an adequate indicator for tropical convection, these results suggest that since the mid-2000s, the bottoming out of both the SST correlation and convective instability in interannual variability reflects the release of unstable stratification due to cumulus convection (i.e., moist neutral stratification) over the warmer SST.
Scatterplot of SST (°C) and convective instability (K 100 hPa−1) averaged over SWI area for (a) 1979–99 and (b) 2000–20. (c) The 11-yr running correlations between SST and convective instability and (d) the 11-yr running standard deviations of convective instability averaged over SWI area. The dashed line in (c) denotes the 95% confidence level.
Citation: Journal of Climate 36, 8; 10.1175/JCLI-D-22-0292.1
Making use of the weakened interannual variability of convective instability, we can objectively evaluate the region of enhanced cumulus convection. Figure 13a shows the ratio of standard deviation in 2000–20 to 1979–99 for convective instability. Recent weakening of interannual variability of convective instability is evident over the southern East China Sea and off the Pacific coast of Japan, with a minimum over SWI area, supporting the regression analysis (Fig. 9) and the precipitation EOF-2 pattern (Fig. 3b). It appears that the enhanced cumulus convection is located along the warmer Kuroshio current, where the SST contour that corresponds to the threshold for convection has been moving northward (Fig. 13b). Indeed, warm SST in the Kuroshio helps enhance the local precipitation in early summer for 2003–08 (Sasaki et al. 2012). While EASM precipitation variability is primarily dependent on larger-scale atmospheric forcing, recent local SST warming may enhance the contribution to the local convection. Further research needs to clarify the impact of local SST warming on cumulus convection. The weakened interannual variability of convective instability is more pronounced in June–July than JJA, further extending into western Japan (Fig. 13c). The consistency between vertical structures and convective instability indicates that cumulus convection is more dominant for the characteristic of EASM precipitation variability including mei-yu–baiu rainfall than frontal structure over the southern East China Sea and off the Pacific coast of Japan. The pronounced weakening over China (Figs. 13a,c) is likely to be associated with a weakening of convective activity (Figs. 11c,d), possibly resulting from the weakening of the PJ pattern.
The ratio (%) of standard deviation in 2000–20 to 1979–99 for convective instability during (a) JJA and (c) JJ. Contours denote standard deviation for 1979–99. (b) JJA SST distribution (27°, 28°, and 29°C) for 1979–99 (dashed contours) and 2000–20 (solid contours).
Citation: Journal of Climate 36, 8; 10.1175/JCLI-D-22-0292.1
6. Conclusions and discussion
We have investigated EASM precipitation variability in views of interdecadal changes and a modulation of interannual variability from 1979 to 2020. Our results reveal that cumulus convection has been more dominant for the characteristic of EASM precipitation variability including mei-yu–baiu rainfall than frontal structure over the subtropical seas since the mid-2000s. Both linear trend and precipitation EOF-2 show that recent SWI precipitation including western Japan has markedly increased, which is closely related to Arctic warming. The EOF-2 has taken the place of the PJ pattern (EOF-1) as a dominant driver of precipitation variability over western Japan in the past decade. Regression analysis indicates the decadal shift in the characteristic of SWI precipitation variability from variations of frontal intensity to those of cumulus convective activity in the mid-2000s. Cumulus convection is enhanced through neutralizing moist instability over the southern East China Sea and off the Pacific coast of Japan along the Kuroshio, where the SST contour that corresponds to the threshold for convection has been moving northward. The enhanced cumulus convection enables to act as a forcing for Rossby wave train from East Asia near western Japan toward North America along the westerly jet or for the eastward extension of the Silk Road pattern.
While SWI precipitation has increased in both linear trends (Fig. 1b) and EOF analysis (Fig. 3b), there is the considerable difference between both methods in atmospheric circulation variability (Fig. 6). EOF analysis better captures the atmospheric circulation variability than linear trends, indicating the importance of interannual variability. However, EOF analysis also varies the result when the period is different. For example, there is a possibility that the precipitation EOF-2 pattern is more pronounced when the PJ pattern weakens. In fact, the EOF-1 and -2 are not statistically separated, based on a rule of thumb by North et al. (1982), although the EOF-1 is associated with ENSO, while the EOF-2 has a close relationship with Arctic warming, suggesting different modes. We need to consider a modulation of interannual variability under climate change, especially when discussing future projection of EASM precipitation.
Here, we attempt to understand why heavy rainfall events occur frequently in western Japan by focusing on the decadal shift in EASM precipitation variability. Figure 14 shows regressed anomalies of vertical p velocity and its variance (based on 6-hourly data) at 300 hPa onto YRB and SWI precipitation. Ascending motions associated with YRB precipitation intensify at the east toward Japan for 1979–99 (Fig. 14a) but turn northeast toward the Korean Peninsula for 2000–20 (Fig. 14c). Stronger vertical velocity variance further extends to northeastward (Fig. 14d), accounting for increased daily precipitation variance over the Korean Peninsula (Fig. 1c). Overall, the intensified ascent appears to reflect the changes in the PJ pattern (Figs. 8a,b). Similarly, ascending motions associated with SWI precipitation intensify at the northeast along off the Pacific coast of Japan for 1979–99 (Fig. 14e) but turn north toward western Japan for 2000–20 (Fig. 14g), and stronger vertical velocity variance further extends to northward over the entire western Japan (Fig. 14h). This decadal shift in ascent variability is in good qualitative agreement with the EOF-2 pattern (Fig. 3b) and increased daily precipitation variance over western Japan (Fig. 1c). These results are consistent with Sampe and Xie (2010) who clarified that using a linear baroclinic model, mei-yu–baiu convection is enhanced eastward under June climatology and northward under August climatology through warm advection due to the westerly jet. As the August jet is weaker than the June jet, recent Asian jet weakening (Figs. 5c,d) is suggested to contribute to the frequent occurrence of heavy rainfall events in western Japan, as a result of downstream response to an upstream latent heating or a diabatic Rossby wave (Shibuya et al. 2021).
Regressed anomalies of (a),(c) vertical p velocity (Pa s−1) and (b),(d) its variance (10−3 Pa2 s−2) at 300 hPa onto YRB precipitation for (a),(b) 1979–99 and (c),(d) 2000–20. (e)–(h) As in (a)–(d), but for SWI precipitation. Dotted contours indicate the 95% confidence level.
Citation: Journal of Climate 36, 8; 10.1175/JCLI-D-22-0292.1
How moisture is accumulated and transported to western Japan? Moisture transport accompanied with lower-level southerly is the largest at the east side of SWI area in both the heavy rainfall event of early summer 2018 (Sekizawa et al. 2019) and long-term changes (Takahashi and Fujinami 2021). In the early summer 2018, upstream convective activity is observed over the southern East China Sea and around SWI area, possibly helping intensify the lower-level southerlies (Takemura et al. 2019) (e.g., Fig. 8d) for moisture transport to downstream western Japan through favoring evaporation from the Kuroshio (Sekizawa et al. 2019). In the heavy rainfall event of July 2020, Zhao et al. (2021) also clarify that the moisture from subtropical regions contributed the most to the moisture accumulation and suggest the role of recent SST warming in the subtropical western Pacific for the local evaporation and convection. These heavy rainfall events also suggest the characteristic of cumulus convection over the subtropical seas. In conclusion, the frequent occurrence of heavy rainfall events in western Japan is likely to be enhanced by the combination between the decadal shift in EASM precipitation variability and the Asian jet weakening, under the weakening of the PJ pattern, although how the Asian jet affects EASM precipitation variability including heavy rainfall event needs to be further investigated.
We further reveal that enhanced cumulus convection enables to act as a driver for Rossby wave train from East Asia near western Japan toward North America. While our findings are derived from the datasets for 1979–2020, the wavelike train from East Asia across the Pacific Ocean (Figs. 7b and 8d) with increased SWI precipitation was observed in summer of 2021 (not shown), possibly favoring heatwaves in northern Japan through locally intensifying the WPSH (Fig. 8d) and the Pacific Northwest areas of the United States and Canada by a blocking high (e.g., Philip et al. 2022). Although enhanced cumulus convection can help propagate eastward the wave train, it might also be linked to atmospheric forcings, such as the Silk Road pattern (Fig. 6b) or the circumglobal teleconnection pattern. Further research needs to clarify the influence of the Silk Road pattern on the eastward wave propagation and mei-yu–baiu precipitation variability including heavy rainfall event.
Acknowledgments.
We wish to thank three anonymous reviewers for their constructive comments. This work was supported by the Environment Research and Technology Development Fund (JPMEERF20192004 and JPMEERF20222002) of the Environmental Restoration and Conservation Agency, Japan, and the Japan Society for the Promotion of Science (JSPS) KAKENHI grants (Grant 19H05697).
Data availability statement.
All data used in this study are publicly available and can be downloaded from the corresponding websites (JRA55: https://jra.kishou.go.jp/JRA-55/index_en.html; GPCP: https://www.esrl.noaa.gov/psd/data/gridded/data.gpcp.html; HadISST: https://www.metoffice.gov.uk/hadobs/hadisst/; and CPC: https://psl.noaa.gov/data/gridded/data.cpc.globalprecip.html).
REFERENCES
Adler, R. F., and Coauthors, 2003: The version-2 Global Precipitation Climatology Project (GPCP) Monthly Precipitation Analysis (1979–present). J. Hydrometeor., 4, 1147–1167, https://doi.org/10.1175/1525-7541(2003)004<1147:TVGPCP>2.0.CO;2.
Chen, M., W. Shi, P. Xie, V. B. S. Silva, V. E. Kousky, R. Wayne Higgins, and J. E. Janowiak, 2008: Assessing objective techniques for gauge-based analyses of global daily precipitation. J. Geophys. Res., 113, D04110, https://doi.org/10.1029/2007JD009132.
Chowdary, J. S., K. Hu, G. Srinivas, Y. Kosaka, L. Wang, and K. Koteswara Rao, 2019: The Eurasian jet streams as conduits for East Asian Monsoon variability. Curr. Climate Change Rep., 5, 233–244, https://doi.org/10.1007/s40641-019-00134-x.
Coumou, D., J. Lehmann, and J. Beckmann, 2015: The weakening summer circulation in the Northern Hemisphere mid-latitudes. Science, 348, 324–327, https://doi.org/10.1126/science.1261768.
Coumou, D., G. Di Capua, S. Vavrus, L. Wang, and S. Wang, 2018: The influence of Arctic amplification on mid-latitude summer circulation. Nat. Commun., 9, 2959, https://doi.org/10.1038/s41467-018-05256-8.
Ding, Q. H., and B. Wang, 2005: Circumglobal teleconnection in the Northern Hemisphere summer. J. Climate, 18, 3483–3505, https://doi.org/10.1175/JCLI3473.1.
Enomoto, T., B. J. Hoskins, and Y. Matsuda, 2003: The formation mechanism of the Bonin high in August. Quart. J. Roy. Meteor. Soc., 129, 157–178, https://doi.org/10.1256/qj.01.211.
Fu, R., A. D. Del Genio, and W. B. Rossow, 1994: Influence of ocean surface conditions on atmospheric vertical thermodynamic structure and deep convection. J. Climate, 7, 1092–1108, https://doi.org/10.1175/1520-0442(1994)007<1092:IOOSCO>2.0.CO;2.
Graham, N. E., and T. P. Barnett, 1987: Sea surface temperature, surface wind divergence, and convection over tropical oceans. Science, 238, 657–659, https://doi.org/10.1126/science.238.4827.657.
Horinouchi, T., S. Matsumura, T. Ose, and Y. N. Takayabu, 2019: Jet-precipitation relation and future change of mei-yu/baiu rainband and subtropical jet in CMIP5 coupled GCM simulations. J. Climate, 32, 2247–2259, https://doi.org/10.1175/JCLI-D-18-0426.1.
Horinouchi, T., Y. Kosaka, H. Nakamigawa, H. Nakamura, N. Fujikawa, and Y. N. Takayabu, 2021: Moisture supply, jet, and silk-road wave train associated with the prolonged heavy rainfall in Kyushu, Japan in early July 2020. SOLA, 17B (Special_Edition), 1–8, https://doi.org/10.2151/sola.2021-019.
Huang, Y., B. Wang, X. Li, and H. Wang, 2018: Changes in the influence of the western Pacific subtropical high on Asian summer monsoon rainfall in the late 1990s. Climate Dyn., 51, 443–455, https://doi.org/10.1007/s00382-017-3933-1.
Huffman, G. J., and Coauthors, 1997: The Global Precipitation Climatology Project (GPCP) combined precipitation dataset. Bull. Amer. Meteor. Soc., 78, 5–20, https://doi.org/10.1175/1520-0477(1997)078<0005:TGPCPG>2.0.CO;2.
Imada, Y., H. Kawase, M. Watanabe, M. Arai, H. Shiogama, and I. Takayabu, 2020: Advanced risk-based event attribution for heavy regional rainfall events. npj Climate Atmos. Sci., 3, 37, https://doi.org/10.1038/s41612-020-00141-y.
Kobayashi, S., and Coauthors, 2015: The JRA-55 reanalysis: General specifications and basic characteristics. J. Meteor. Soc. Japan, 93, 5–48, https://doi.org/10.2151/jmsj.2015-001.
Kosaka, Y., S.-P. Xie, and H. Nakamura, 2011: Dynamics of interannual variability in summer precipitation over East Asia. J. Climate, 24, 5435–5453, https://doi.org/10.1175/2011JCLI4099.1.
Kosaka, Y., J. S. Chowdary, S. Xie, Y. Min, and J. Lee, 2012: Limitations of seasonal predictability for summer climate over East Asia and the northwestern Pacific. J. Climate, 25, 7574–7589, https://doi.org/10.1175/JCLI-D-12-00009.1.
Kubota, H., Y. Kosaka, and S.-P. Xie, 2016: A 117-year long index of the Pacific–Japan pattern with application to interdecadal variability. Int. J. Climatol., 36, 1575–1589, https://doi.org/10.1002/joc.4441.
Li, X., and R. Lu, 2020: Breakdown of the summertime meridional teleconnection pattern over the western North Pacific and East Asia since the early 2000s. J. Climate, 33, 8487–8505, https://doi.org/10.1175/JCLI-D-19-0746.1.
Matsumura, S., and T. Horinouchi, 2016: Pacific Ocean decadal forcing of long-term changes in the western Pacific subtropical high. Sci. Rep., 6, 37765, https://doi.org/10.1038/srep37765.
Matsumura, S., S. Sugimoto, and T. Sato, 2015: Recent intensification of the western Pacific subtropical high associated with the East Asian summer monsoon. J. Climate, 28, 2873–2883, https://doi.org/10.1175/JCLI-D-14-00569.1.
Matsumura, S., T. Horinouchi, S. Sugimoto, and T. Sato, 2016: Response of the Baiu rainband to northwest Pacific SST anomalies and its impact on atmospheric circulation. J. Climate, 29, 3075–3093, https://doi.org/10.1175/JCLI-D-15-0691.1.
Ninomiya, K., and T. Murakami, 1987: The early summer rainy season (Baiu) over Japan. Monsoon Meteorology, C.-P. Chang and T. N. Krishnamurti, Eds., Oxford University Press, 93–121.
Ninomiya, K., T. Nishimura, W. Ohfuchi, T. Suzuki, and S. Matsumura, 2002: Features of the Baiu front simulated in an AGCM (T42L52). J. Meteor. Soc. Japan, 80, 697–716, https://doi.org/10.2151/jmsj.80.697.
Nitta, T., 1987: Convective activities in the tropical western Pacific and their impact on the Northern Hemisphere summer circulation. J. Meteor. Soc. Japan, 65, 373–390, https://doi.org/10.2151/jmsj1965.65.3_373.
North, G. R., T. L. Bell, R. F. Cahalan, and F. J. Moeng, 1982: Sampling errors in the estimation of empirical orthogonal functions. Mon. Wea. Rev., 110, 699–706, https://doi.org/10.1175/1520-0493(1982)110<0699:SEITEO>2.0.CO;2.
Philip, S. Y., and Coauthors, 2022: Rapid attribution analysis of the extraordinary heat wave on the Pacific Coast of the U.S. and Canada in June 2021. Earth Syst. Dyn., 13, 1689–1713, https://doi.org/10.5194/esd-13-1689-2022.
Rayner, N. A., D. E. Parker, E. B. Horton, C. K. Folland, L. V. Alexander, D. P. Rowell, E. C. Kent, and A. Kaplan, 2003: Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century. J. Geophys. Res., 108, 4407, https://doi.org/10.1029/2002JD002670.
Sampe, T., and S.-P. Xie, 2010: Large-scale dynamics of the meiyu-baiu rainband: Environmental forcing by the westerly jet. J. Climate, 23, 113–134, https://doi.org/10.1175/2009JCLI3128.1.
Sasaki, Y. N., S. Minobe, T. Asai, and M. Inatsu, 2012: Influence of the Kuroshio in the East China Sea on the early summer (baiu) rain. J. Climate, 25, 6627–6645, https://doi.org/10.1175/JCLI-D-11-00727.1.
Sekizawa, S., T. Miyasaka, H. Nakamura, A. Shimpo, K. Takemura, and S. Maeda, 2019: Anomalous moisture transport and oceanic evaporation during a torrential rainfall event over western Japan in early July 2018. SOLA, 15A, 25–30, https://doi.org/10.2151/sola.15A-005.
Shibuya, R., Y. N. Takayabu, and H. Kamahori, 2021: Dynamics of widespread extreme precipitation events and the associated large-scale environment using AMeDAS and JRA-55 data. J. Climate, 34, 8955–8970, https://doi.org/10.1175/JCLI-D-21-0064.1.
Stephan, C. C., N. P. Klingaman, and A. G. Turner, 2019: A mechanism for the recently increased interdecadal variability of the Silk Road pattern. J. Climate, 32, 717–736, https://doi.org/10.1175/JCLI-D-18-0405.1.
Sun, L., X.-Q. Yang, L. Tao, J. Fang, and X. Sun, 2021: Changing impact of ENSO events on the following summer rainfall in eastern China since the 1950s. J. Climate, 34, 8105–8123, https://doi.org/10.1175/JCLI-D-21-0018.1.
Takahashi, H. G., and H. Fujinami, 2021: Recent decadal enhancement of Meiyu-Baiu heavy rainfall over East Asia. Sci. Rep., 11, 13665, https://doi.org/10.1038/s41598-021-93006-0.
Takaya, K., and H. Nakamura, 2001: A formulation of a phase-independent wave-activity flux for stationary and migratory quasigeostrophic eddies on a zonally varying basic flow. J. Atmos. Sci., 58, 608–627, https://doi.org/10.1175/1520-0469(2001)058<0608:AFOAPI>2.0.CO;2.
Takaya, Y., I. Ishikawa, C. Kobayashi, H. Endo, and T. Ose, 2020: Enhanced Meiyu-Baiu rainfall in early summer 2020: Aftermath of the 2019 super IOD event. Geophys. Res. Lett., 47, e2020GL090671, https://doi.org/10.1029/2020GL090671.
Takemura, K., S. Wakamatsu, H. Togawa, A. Shimpo, C. Kobayashi, S. Maeda, and H. Nakamura, 2019: Extreme moisture flux convergence over western Japan during the heavy rain event of July 2018. SOLA, 15A, 49–54, https://doi.org/10.2151/sola.15A-009.
Wang, L., P. Xu, W. Chen, and Y. Liu, 2017: Interdecadal variations of the Silk Road pattern. J. Climate, 30, 9915–9932, https://doi.org/10.1175/JCLI-D-17-0340.1.
Wu, B., T. Zhou, and T. Li, 2016: Impacts of the Pacific–Japan and circumglobal teleconnection patterns on the interdecadal variability of the East Asian summer monsoon. J. Climate, 29, 3253–3271, https://doi.org/10.1175/JCLI-D-15-0105.1.
Xie, S.-P., Y. Du, G. Huang, X.-T. Zheng, H. Tokinaga, K. Hu, and Q. Liu, 2010: Decadal shift in El Niño influences on Indo–western Pacific and East Asian climate in the 1970s. J. Climate, 23, 3352–3368, https://doi.org/10.1175/2010JCLI3429.1.
Xu, P., L. Wang, W. Chen, J. Feng, and Y. Liu, 2019: Structural changes in the Pacific–Japan pattern in the late 1990s. J. Climate, 32, 607–621, https://doi.org/10.1175/JCLI-D-18-0123.1.
Yokoyama, C., H. Tsuji, and Y. N. Takayabu, 2020: The effects of an upper-tropospheric trough on the heavy rainfall event in July 2018 over Japan. J. Meteor. Soc. Japan, 98, 235–255, https://doi.org/10.2151/jmsj.2020-013.
Zhao, N., A. Manda, X. Guo, K. Kikuchi, T. Nasuno, M. Nakano, Y. Zhang, and B. Wang, 2021: A Lagrangian view of moisture transport related to the heavy rainfall of July 2020 in Japan: Importance of the moistening over the subtropical regions. Geophys. Res. Lett., 48, e2020GL091441, https://doi.org/10.1029/2020GL091441.
Zhou, Z.-Q., S.-P. Xie, and R. Zhang, 2021: Historic Yangtze flooding of 2020 tied to extreme Indian Ocean conditions. Proc. Natl. Acad. Sci. USA, 118, e2022255118, https://doi.org/10.1073/pnas.2022255118.