The Asian monsoon can be divided into the East Asian and the South Asian or Indian monsoon systems (Ding et al., 2004). Based on a summer monsoon index derived from MSLP gradients between land and ocean in the East Asian region, Guo et al. (2003) found a systematic reduction in the East Asian summer monsoon during 1951 to 2000, with a stronger monsoon dominant in the first half of the period and a weaker monsoon prevailing in the second half (Figure 3.35). This long-term change in the East Asian monsoon index is consistent with a tendency for a southward shift of the summer rain belt over eastern China (Zhai et al., 2004). However, Figure 3.35, based on the newly developed Hadley Centre MSLP data set version 2 (HadSLP2; Allan and Ansell, 2006), suggests that although there exists a weakening trend starting in the 1920s, it is not reflected in the longer record extending back to the 1850s, which shows marked decadal-scale variability before the 1940s.
There is other evidence that changes in the Asian monsoon occurred about the time of the 1976–1977 climate shift (Wang, 2001) along with changes in ENSO (Huang et al., 2003; Qian et al., 2003), and declines in land precipitation are evident in southern Asia and, to some extent, in Southeast Asia (see Figure 3.14). Gong and Ho (2002) suggested that the change in summer rainfall over the Yangtze River valley was due to a southward rainfall shift and Ho et al. (2003) noted a sudden change in Korea. These occurred about the same time as a change in the 500 hPa geopotential height and typhoon tracks in summer over the northern Pacific (Gong et al., 2002; see Section 3.6.3) related to the enlargement, intensification and southwestward extension of the northwest Pacific subtropical high. When the equatorial central and eastern Pacific is in a decadal warm period, summer monsoon rainfall is stronger in the Yangtze River valley but weaker in North China. A strong tropospheric cooling trend is found in East Asia during July and August. Accompanying this summer cooling, the upper-level westerly jet stream over East Asia shifts southward and the East Asian summer monsoon weakens, which results in the tendency towards increased droughts in northern China and floods in the Yangtze River valley (Yu et al., 2004b).
Figure 3.35. Annual values of the East Asia summer monsoon index derived from MSLP gradients between land and ocean in the East Asia region. The definition of the index is based on Guo et al. (2003) but was recalculated based on the HadSLP2 (Allan and Ansell, 2006) data set. The smooth black curve shows decadal variations (see Appendix 3.A).
Rainfall during the Indian monsoon season, which runs from June to September and accounts for about 70% of annual rainfall, exhibits decadal variability. Observational studies have shown that the impact of El Niño is more severe during the below-normal epochs, while the impact of La Niña is more severe during the above-normal epochs (Kripalani and Kulkarni, 1997a; Kripalani et al., 2001, 2003). Such modulation of ENSO impacts by the decadal monsoon variability was also observed in the rainfall regimes over Southeast Asia (Kripalani and Kulkarni, 1997b). Links between monsoon-related events (rainfall over South Asia, rainfall over East Asia, NH circulation, tropical Pacific circulation) weakened between 1890 and 1930 but strengthened during 1930 to 1970 (Kripalani and Kulkarni, 2001). The strong inverse relationship between El Niño events and Indian monsoon rainfalls that prevailed for more than a century prior to about 1976 has weakened substantially since then (Kumar et al., 1999; Krishnamurthy and Goswami, 2000; Sarkar et al., 2004), involving large-scale changes in atmospheric circulation. Shifts in the Walker Circulation and enhanced land-sea contrasts appear to be countering effects of increased El Niño activity. Ashok et al. (2001) also found that the IOD (see Section 188.8.131.52) plays an important role as a modulator of Indian rainfall. The El Niño-Southern Oscillation is also related to atmospheric fluctuations both in the Indian sector and in northeastern China (Kinter et al., 2002).