Theor Appl Climatol (2012) 109:447–459 DOI 10.1007/s00704-012-0591-4

ORIGINAL PAPER

Weakening of Indian summer monsoon rainfall in warming environment Ashwini Kulkarni

Received: 27 May 2011 / Accepted: 18 January 2012 / Published online: 2 February 2012 # Springer-Verlag 2012

Abstract Though over a century long period (1871–2010) the Indian summer monsoon rainfall (ISMR) series is stable, it does depict the decreasing tendency during the last three decades of the 20th century. Around mid-1970s, there was a major climate shift over the globe. The average all-India surface air temperature also shows consistent rise after 1975. This unequivocal warming may have some impact on the weakening of ISMR. The reduction in seasonal rainfall is mainly contributed by the deficit rainfall over core monsoon zone which happens to be the major contributor to seasonal rainfall amount. During the period 1976–2004, the deficit (excess) monsoons have become more (less) frequent. The monsoon circulation is observed to be weakened. The midtropospheric gradient responsible for the maintenance of monsoon circulation has been observed to be weakened significantly as compared to 1901–1975. The warming over western equatorial Indian Ocean as well as equatorial Pacific is more pronounced after mid-70s and the co-occurrence of positive Indian Ocean Dipole Mode events and El Nino events might have reinforced the large deficit anomalies of Indian summer monsoon rainfall during 1976–2004. All these factors may contribute to the weakening of ISMR.

1 Introduction For the agrarian country like India whose economy largely swings with the whims of the summer (June through September) monsoon rainfall, year-to-year variability of monsoon is a very vital issue. Though it has been well discussed A. Kulkarni (*) Indian Institute of Tropical Meteorology, NCL PO, Pashan, Pune 411008, India e-mail: [email protected]

that on the century scale the Indian monsoon rainfall series has remained stable, it has shown striking changes in its relationship with important climate controls like El Nino– Southern Oscillation (Kripalani and Kulkarni 1997; Krishna Kumar et al. 2006) and Eurasian and Himalayan snow cover in the last three decades of the 20th century (Kripalani and Kulkarni 1999; Kripalani et al. 2003; Kripalani et al. 2001) which makes the seasonal prediction of Indian summer monsoon more challengeable. Also seasonal mean monsoon rainfall shows epochal variability of approximately 30 years since 1870s. The seasonal rainfall series has been characterized by below normal epochs (1901–1930 and 1961–1990) and the above normal epochs (1871– 1900 and 1931–1960) of approximately 30 years each (Kripalani and Kulkarni 1997). If this epochal variability is assumed to be continued, Indian monsoon rainfall should have entered into the above normal epoch around 1990, which did not happen. After 1990, the epochal variability of Indian summer monsoon rainfall (ISMR) has been disturbed. Kothawale et al. (2008) have shown that there is significant warming over Arabian sea, Bay of Bengal as well as equatorial south Indian ocean over last hundred years and the trend has been accelerated in the period 1971–2002. It is well documented that the SSTs of the Indian Ocean have had an impact on the Indian monsoon (Shukla 1975; Shukla and Mooley 1987). Clark et al. (2000) found a significant positive correlation between winter Arabian Sea SSTs and the following monsoon season All-India rainfall. Also, the Indian landmass has experienced unequivocal rise in temperature since after mid-70s. The land temperatures over India have shown around 0.2°C rise per decade in the period 1971–2007 (Kothawale et al. 2010). Of the last 15 years, 13 happened to be the warmest years on record of more than one and half century data which goes hand in hand with the

448

rainfall are described in section 6. The conclusions of this study are given in section 7.

2 Data The data sets used in this study are (1) High resolution (1°×1° lat/long) daily gridded rainfall data set for the Indian region for 1901–2004 prepared by India Meteorological Department (Rajeevan et al. 2008). For this analysis, the rainfall data over 1,384 stations with minimum 70% data availability were considered. The interpolation method proposed by Shepard (1968) had been used for interpolating station rainfall data into regular grids. The quality of this data set was examined by comparing with the similar kind of data set developed for the period 1951–2004 (Rajeevan et al. 2006). The data are available for 1 January to 31 December for every year; here, we are using the data for Indian summer monsoon season, 1 June to 30 September, i.e., 122 days for each of these 104 years (1901–2004). Since this 3

(a)

2 1 0 -1 Standardised Rainfall

global temperatures. Our aim of this study is to examine whether there is any change in the behavior of Indian summer monsoon rainfall and the monsoon circulation associated with this consistent warming since mid-70s. Goswami et al. (2006) have shown that the Indian summer monsoon rainfall series over the past century has been remained stable in spite of the steady rise in global mean temperatures. They have shown that over core monsoon zone, the contribution from increasing heavy rain events is offset by decreasing moderate events and hence on the long term the series is trendless. Using the century long daily high resolution gridded data, Rajeevan et al. (2008) have shown that the extreme rainfall events are increasing over central India, and this increasing trend could be associated with the increasing trend of sea surface temperatures and surface latent heat flux over tropical Indian ocean. Though the extremes have been extensively studied, not much attention is paid to the typical features of Indian monsoon variability in the warmer period at the end of the 20th century. In view of the consistent rise in temperatures after 1970s, it has been observed that the Indian monsoon rainfall has decreased at the rate of 1.5 mm/year over the period 1971–2002 (Kothawale et al. 2008). The year-to-year variability of ISMR is partly due to the external surface boundary forcings and partly due to internal dynamics. The variability associated with internal dynamics is inherently unpredictable. Several recent modeling studies have shown that a significant fraction of the inter-annual variability (IAV) of the seasonal mean Indian summer monsoon is governed by internal chaotic dynamics (e.g., Goswami 1998). The intra-seasonal variability dominant during the monsoon season constitutes this internal dynamics. Thus, the year-to-year changes in the intra-seasonal variability (ISV) are an important potential source of interannual fluctuations of ISMR strength (Ferranti et al. 1997; Goswami 1998; Sperber et al. 2000; Goswami and Mohan 2001; Kripalani et al. 2004). However, Kulkarni et al. (2009) have shown that the contributions of intra-seasonal variability have also been weakened after 1970s. Hence, in general, it has been observed that there is a drastic change in the surface temperature as well as monsoon characteristics around mid-1970s. In this study, we examine the changes in rainfall on seasonal as well as subseasonal scale and make an attempt to discuss the role of various factors in the weakening quantum of Indian summer monsoon rainfall in the period 1976–2004 as compared to the earlier period 1901–1975. The data sets used in this study are described in section 2. Section 3 deals with analysis of regional changes. Section 4 gives contribution of excess/deficit monsoons to the weakening monsoon rainfall after mid-1970s. The characteristics features of rainfall on sub-seasonal scale are discussed in section 5. The circulation features associated with the weakening

A. Kulkarni

-2 -3 3

1920

1940

1960

1980

2000

1920

1940

1960 Years

1980

2000

(b)

2 1 0 -1 -2 -3

Fig. 1 a Inter-annual variability of ISMR for the period 1901–2004 (top panel bars) and the 13-point filtered series (smooth curve). Green (red) line depicts +1 (−1) standard deviation. b Time series of standardized ISMR with linear trend line for 1901–1975(violate) and 1976–2004 (black)

Weakening of Indian summer monsoon rainfall

gridded data set is based on fixed rainfall network, it can be used for examining long-term rainfall trends. (2) All India seasonal summer monsoon rainfall (June–September) for the period 1871–2010 has been downloaded

449

from the website of Indian Institute of Tropical Meteorology www.tropmet.res.in. (3) The monthly mean global Sea Surface temperature data set HadISST at the resolution of 1°×1° (lat/long) for

Fig. 2 Composite ISMR anomalies (millimeter) for two sub-period 1901–1975 and 1976–2004 (top panels) and the t value for difference (1901– 1975)–(1976–2004) (bottom panel)

450

the period 1871–2010 has been downloaded from the website of http://badc.nerc.ac.uk. (4) The NCEP-NCAR reanalysis data set (Kalnay et al. 1996) for the 850 hPa vector winds for the period 1948–2004 is used to examine the dynamic circulation features. Also, NCEP reanalyzed 500 hPa geopotential height data have been used.

3 Inter-annual variability Seasonal variation of rainfall is the most distinguishing feature of the monsoonal regions of the world. About 80% of the annual rainfall over a large part of India occurs during the summer monsoon period (June to September). The variation in the all-India average seasonal rainfall has been widely studied and may be considered as a measure of the intensity of the planetary-scale monsoon over the Indian region. 3.1 All-India summer monsoon rainfall Since high resolution gridded daily rainfall data is available for 1901–2004, we analyze the Indian summer monsoon rainfall (ISMR) series for the same period. Figure 1 depicts the inter-annual variability of ISMR for the period 1901– 2004. To remove the sub-decadal fluctuations, 13-point

A. Kulkarni

filter (IPCC 2007) has been applied. The smooth orange curve shows 13-point filtered series. This series clearly shows the epochal variability with epochs approximately of 30 years; however, it is clearly seen that after 1990s, the series did not enter into the above normal epoch, actually the below normal epoch of 1961–1990 continued after 1990. In fact after 1960, the rainfall remained near/below normal. Even after removing the multi-decadal variability, the rainfall does show increasing (decreasing) trend before (after) mid-1970s. The century long series is highly random and free from long-term trend (Mann–Kendall trend value −0.25, not statistically significant); however (Fig. 1, bottom panel), it clearly shows increasing trend in the period 1901– 1975 (trend value +1.8, significant at 10% level) and decreasing trend in 1976–2004 (trend value −1.28, significant at 20% level). The mean and coefficient of variation of ISMR for the period 1901–2004 are 843.8 mm and 9.5%. For the period 1901–1975 (1976–2004), the mean and CV are 850.0% and 9.6% (827.6% and 9.1%), respectively. Though the difference in mean rainfall in these two time periods is statistically significant at a low level of significance of 20% (t value01.28), the ISMR is decreasing at the rate of 1.5 mm/year in the recent period (Kothawale et al. 4

SW

2 0 -2 -4 1901 4

1921

1941

1961

1981

2001

1921

1941

1961

1981

2001

1921

1941

1961

1981

2001

1921

1941

1961

1981

2001

NW

2

Standardized Rainfall

0 -2 -4 1901 4

CEN

2 0 -2 -4 1901 4

SE

2 0 -2 -4 1901

Years

Fig. 3 Map of 1° latitude/longitude grid over India for which the daily rainfall data (1901–2004) is used in the study. Four regions are shaded (1) south-west, (2) north-west, (3) central, and (4) south-east region

Fig. 4 Time series (bars) for 1901–2004 of four regions in Fig 3. Red line depicts 13-point filtered series

Weakening of Indian summer monsoon rainfall

2008). The seasonal mean rainfall is reduced by ∼3% in 1976–2004 as compared to 1901–1975. To examine the exact region where the rainfall has been reduced in 1976–2004, the composite anomaly seasonal rainfall patterns are prepared. Figure 2 (top panels) depicts seasonal rainfall anomalies for the two periods 1901–1975 and 1976–2004 and the t value for the difference (bottom panel). The composite anomalies based on the period 1901– 1975 are positive over the entire country; however, the rainfall over west coast and central India which have major contributions to seasonal quantum of monsoon rainfall is substantially reduced in the period 1976–2004, and the reduction over core monsoon zone and southern parts of west coast is statistically significant as can be seen from the bottom panel. The seasonal rainfall over North-eastern parts has been increased significantly in the recent period. The positive rainfall anomalies over these regions in 1901–1975 have become large deficits in 1976–2004 which has contributed a lot in reducing the average all-India seasonal rainfall. 3.2 Regional rainfall analysis To study the regional aspects of the seasonal monsoon rainfall, the time series of seasonal rainfall over four specific regions: (1) south-west (SW), (2) north-west (NW), (3) central India (CEN), and (4) south-east (SE) region as shown in Fig. 3 have been constructed by averaging the

451

rainfall over appropriate grids over these regions. The regions are exactly as considered by Kulkarni et al. (2009). The seasonal rainfall time series for these four regions is given in Fig. 4. The smooth curve represents the 13-point filtered time series. The rainfall over central India remains continuously in the below normal phase since 1970, while that over South-east region, it is consistently in the above normal phase after 1970s. In north-west and southwest region, it shows the oscillatory nature; however, it remains much closer to long-term mean value over southwest region. Central India depicts significant decreasing long-term trend over the period 1901–2004 (Mann–Kendall trend value0−2.02). Over south-west region, significant increasing trend (trend value 2.35) is observed over 1901– 1975, while it shows decreasing tendency over 1976–2004. 3.3 Spatial trends To get the exact region of significant contribution in reduced rainfall after mid-1970s, the seasonal rainfall time series at every grid has been applied by a nonparametric Mann– Kendall test for randomness against trend. Figure 5 depicts the Mann–Kendall trend values at every grid for three periods 1901–1975 (first panel), 1976–2004 (middle panel), and 1901–2004 (last panel). Over the long period of 1901–2004, significant decreasing trend has been observed over large portion of central India, i.e., over the passage of low pressure systems from bay of Bengal to Indian landmass as well

Fig. 5 Spatial distribution of Mann–Kendall trend statistic in seasonal rainfall for three time epochs 1901–1975, 1976–2004, and 1901–2004. Negative/decreasing (positive/increasing) trends are given in red (green)

452

as northernmost tip of the country. The west coast and south-eastern coast shows increasing tendency. These trends tally very well with the trends given by Guhathakurta and Rajeevan (2008). However, during 1901–1975, increasing trend is seen over large part of the landmass, west of 80°E, slightly significant decreasing rainfall is seen near foothills of Himalayas and north-eastern parts. Hence, the region over which the seasonal rainfall had shown substantial increasing trend in 1901–1975 has become very weak in the later period 1976–2004 which might have reduced the strength of ISMR in this period.

4 Contribution of extreme Indian monsoons The frequency and intensity of extreme monsoons like deficit and excess monsoons also contribute to the overall behavior of ISMR. The standardized series of ISMR has been computed by subtracting long-term mean and dividing by its standard

A. Kulkarni

deviation. The years when standardized rainfall anomaly is less than (greater than) −1 (+1) are designated as deficit (excess) monsoon years. It has been well discussed that during years of extreme deficits/excess, the rainfall anomaly tends to be homogeneous over the entire country (Shukla 1987; Mooley and Shukla 1987) while during normal monsoon years, the rainfall anomaly tends to be inhomogeneous over the country (Xavier and Goswami 2007) with a reasonably large region of excess/ deficit appearing during such years. We examined this aspect for deficit/excess monsoons after 1970s. Figure 6 depicts seasonal rainfall anomalies for three deficit monsoons 1979, 1987, and 1985 (top panels) and three excess monsoons 1983, 1988, and 1994 (bottom panels). All the three deficit monsoons do show large regions of positive anomalies. Similarly during all-India excess years of 1988 and 1994, there are the regions of negative rainfall departures. In 1988, the most important region—core monsoon zone shows negative departures. Also, in 1994, the entire south-east peninsula and north-east region is under positive rainfall departures,

Fig. 6 Seasonal rainfall anomalies (millimeter) for typical all-India deficits 1979, 1987, and 1985 (top) and excess 1983, 1988, and 1994 (bottom)

Weakening of Indian summer monsoon rainfall

hence the statement “during excess/deficit monsoons the rainfall anomalies tend to be homogeneous over the entire

453

country” may not hold after 1975. The spatial variability has been increased after 1975.

Fig. 7 Composite seasonal rainfall anomalies (millimeter) in all-India deficits (left) and excess (right) in three time epochs

454

4.1 Frequency and intensity of extreme monsoons During the period 1901–2004, there are 20 deficit monsoons out of which 13 are in 1901–1975 and 7 are in 1976–2004. Thus, during 1901–1975 probability of getting deficit monsoon is 17% which has increased to 24% during the epoch 1976–2004. The average standardized rainfall associated with deficit monsoons in the first period is −1.53 while that in the second period is −1.47. During 1901–1975 (1976–2004), there are 12 (3) excess monsoons with average standardized rainfall 1.48 (1.34). The probability of excess monsoon is reduced from 16% in earlier period to 10% in recent period. Though the difference in mean rainfall is not statistically significant, the deficit monsoons have increased in frequency during the epoch 1976–2004. Also, less number of excess monsoons are observed in this period which are weaker as compared to those in 1901–1975. This might have contributed to the decreasing mean all India rainfall in the period 1976–2004. Interestingly, during the period 1901–1975, average standardized rainfall associated with deficit (excess) monsoons is −1.53 (1.48)

A. Kulkarni

which might have compensated while for 1976–2004, it is −1.47 (1.34). Thus, deficit monsoons are more severe during this later period as compared to excess.

4.2 Spatial rainfall distribution during extreme monsoons To examine the spatial patterns of summer monsoon rainfall associated with excess/deficit monsoons during these two subperiods, the composites of rainfall anomalies during extreme monsoons have been prepared. Figure 7 depicts the composite summer monsoon rainfall anomalies associated with all-India deficits and excess monsoons during the three time epochs 1901–2004, 1901–1975, and 1976–2004. There are 20 deficit monsoons during 1901–2004 (1901, 1904, 1905, 1911, 1918, 1920, 1941, 1951, 1965, 1966, 1968, 1972, 1974, 1979, 1982, 1985, 1986, 1987, 2002, and 2004) and 15 excess monsoons (1910, 1916, 1917, 1933, 1942, 1947, 1955, 1956, 1959, 1961, 1970, 1975, 1983, 1988, 1994). Consequently, the composites for 1901–1975 are based on 13 deficits and 12

Fig. 8 Composite monthly rainfall difference (1901–1975)-(1976–2004) in millimeter during four monsoon months: June, July, August, and September

Weakening of Indian summer monsoon rainfall

excess monsoons, while for 1976–2004, they are based on 7 deficits and 3 excess monsoons. The spatial distribution of seasonal rainfall anomalies associated with all-India deficit monsoons during period 1901–1975 do show some positive rainfall anomalies over north-eastern parts as well as southern parts of the country. While during 1976–2004, except a small strip on the lee side of western Ghats and north-eastern parts, the entire country, especially the core monsoon region is under negative departures. Also, in the recent three decades, central India and west coast, which happen to be major contributors to the strength of the all-India rainfall, experience more severe negative rainfall anomalies as compared to previous period. In the composites associated with all-India excess monsoons, during 1901–1975, except a small tip over eastern parts, entire country has positive rainfall departures. During 1976–2004, the region under positive departures has shrunk and also the magnitude of rainfall has been reduced over southern parts of west coast as well as central India. In summary, in the period 1976–2004 during all-India deficits (excesses), the area under negative (positive) departures has increased (reduced) and the rainfall has reduced (not much enhanced) as compared to that during 1901– 1975, which contributes to total reduction in area averaged all-India seasonal rainfall during the period 1976–2004.

5 Sub-seasonal variability To examine the rainfall variability over smaller time scales, we study characteristic features of monthly rainfall during the two

455

sub-periods. Generally, it has been observed that July and August are the major contributors to summer monsoon rainfall over India. During severe deficit monsoons, July rainfall happens to be the tremendous failure. Also, it has been shown that when the rainfall gets recharged in September, generally, it results in good seasonal Indian monsoon (Kulkarni et al. 2006). Here, we examine the rainfall variability on monthly scale during these two periods 1901–1975 and 1976–2004.

5.1 Inter-annual variability The rainfall time series on monthly scale over entire land mass has been prepared for 1901–2004 for the four monsoon months, June to September. The century long time series are highly random in all the 4 months and are free from any trend (figure not given). However, during the period 1901–1975, June rainfall showed decreasing tendency (trend value −0.68), though not statistically significant. July rainfall depicts statistically significant increasing trend (2.01), significant at 5% level. August (1.22) and September (1.64) do show increasing tendency but not significant. In the recent three decades, June rainfall depicts increasing tendency (0.29). July (−0.64), August (−2.03), and September (−0.24) are showing decreasing tendencies. August rainfall is significantly decreasing over this period 1976–2004. Since August rainfall is a major contributor to seasonal rainfall, the decrease in all-India summer monsoon rainfall during 1976–2004 may be due to substantial reduction in August rainfall and decreasing rainfall in July and September also contribute.

Fig. 9 Composite June–September SST anomalies for a 1948–1975, b1976-2004, c difference, and d t value

456

5.2 Sub-seasonal spatial distribution To bring out the regions of enhancement/weakening of monthly rainfall, we prepare the composites for difference of seasonal rainfall (1901–1975)–(1976–2004) for the 4 months. Figure 8 shows these composites for 4 months, June to September. In the period 1976–2004, the June

A. Kulkarni

rainfall has enhanced over west coast, north-west and northern parts of the country, while it is slightly reduced over east coast and some parts of north-east India. The rainfall over core monsoon zone and the west coast are substantially reduced in the month of July. These two regions are of very much vital importance to all-India summer monsoon rainfall. In the month of August, some regions south of 20°S do

Fig. 10 Composite June–September 850 hPa vector wind anomalies (meter per second) for the time epochs 1948–1975 and 1976–2004 (top panels) and the t value for the difference (1976–2004)–(1948–1975)

Weakening of Indian summer monsoon rainfall

6 Dynamics associated with weakening rainfall The behavior of Indian monsoon rainfall has changed in the last three decades of the 20th century. The decade after 1990 did not follow epochal behavior of Indian monsoon rainfall which was there for more than a century. Also, its teleconnections with important climatic modes like ENSO, Eurasian as well as Himalayan snow cover in these decades have been weakened (Kripalani and Kulkarni 1997, 1999; Kripalani et al. 2003). We try to examine the dynamics associated with the changing pattern of Indian monsoon in the period 1976–2004. 6.1 Sea surface temperatures The changes in global SSTs are the major driving force of the earth's year-to-year climate variability. The variations in SSTs over equatorial Pacific affects the climate over more than 50% area of the globe. Also, after 1950s, another dominant mode of SST variability, Indian ocean dipole (IOD) mode, has been shown to have significant association with the changes in circulation patterns (Saji et al. 1999). Gadgil et al. (2004) have shown that as large anomalies of ISMR are linked to El Nino Southern Oscillation (ENSO), they are also linked to the equatorial Indian Ocean Oscillation, the atmospheric component of the coupled IOD mode. They have shown that the large excess/deficit anomalies of ISMR are associated with the composite index based on both IOD and ENSO. The largest anomalies of ISMR occur when these two modes reinforce each other. Since these two are a measure of two modes of the coupled system, the impact of two modes on ISMR is additive. Since NCEP reanalysis parameters are available from 1948, for comparison the composites for temperature are also based on 1948–1975 and 1976–2004. During 1948– 1975, no ENSO event co-occurred with positive IOD event, while during 1976–2004, there were four events when they co-occurred (1982, 1994, 1997, and 2006). For 1982, ISMR happened to be severe deficit, while in 1994, it was excess. Figure 9 shows the composite SST anomalies during two sub-periods, 1948–1975 and 1976–2004 (top panels), their difference and the t value for testing the difference (bottom panels). It is clearly seen that during 1948–1975, the western block of IOD (10°S–10°N, 50°–70°E) is cooler than the

eastern block (10°S–Eq, 90°–110°E) with very small temperature anomalies, hence the conditions over Indian Ocean are neutral. Comparatively, NINO3 region is warmer hence ENSO mode was dominant during this period, also it is observed that central Pacific is warmer than the eastern Pacific, i.e., during 1948–1975 there are more frequent Modoki events. In the period 1976–2004, both the oceans show pronounced warming, equatorial Indian Ocean is warmer than the Pacific. Also, western equatorial Indian Ocean SST anomalies are 0.3–0.5°C more than the southeast equatorial Indian Ocean anomalies which suggests that in this period there are more IOD events. The difference between temperature anomalies is significant (panel (d)) even at 1% level over Indian ocean, hence IOD mode is more prominent in the recent period. As shown by Gadgil et al. (2004), the large rainfall anomalies do occur in 1976– 2004 due to additive effect of the two modes, ENSO and IOD. 6.2 Circulation patterns The Asian summer monsoon circulation, a component of the atmospheric general circulation, is mainly forced by the asymmetric pattern of atmospheric heating and cooling over major land masses and oceans. To examine whether there is any difference in the mean monsoon circulation in the two periods, we use NCEP/NCAR reanalyzed vector winds for 850 hPa for the summer monsoon season. Since the reanalysis data is available from 1948, we prepare the composites of vector wind anomalies based on two sub-periods: 1948–1975 and 1976–2004. Figure 10 shows these composites. During the period 1948–1975, the strong south-westerly winds are observed crossing the equator and bringing moisture from Arabian sea to Indian landmass. Also, strong easterly anomalies are observed over equatorial Pacific, which is a typical character of ENSO events. This can 100 Mid-tropospheric Gradient

show increase in rainfall; however, central and north India experience reduction in August rainfall. In September, the entire central India and north India exhibit reduced rainfall. Hence, major reduction in rainfall in July, August, and September on the vital zones of west coast and core monsoon zone might be responsible for the reduced quantum of monsoon rainfall in the period 1976–2004.

457

Y=764.4-0.35*X

80 60 40 20 0

1950

1960

1970

1980 Years

1990

2000

Fig. 11 Time series of mid-tropospheric gradient for 1948–2004. Red line depicts the significant weakening trend

458

be observed from Fig. 9 that during the period 1948– 1975, ENSO was the dominat mode. During recent period 1976–2004, the monsoon circulation is weak. Also, strong northerly extra-tropical winds are prominent over Indian longitudes which may bring cold weather from northern latitudes and may weaken the monsoon circulation over India. During recent period, the trade winds are very weak over Bay of Bengal which may hinder the passage of low pressure rain bearing systems over Bay of Bengal resulting into weak monsoon rains. Strong westerly anomalies along the equator suggest the transport of entire moisture to the eastern Pacific and shift of Walker circulation to the east and hence comparatively drier conditions over the Indian landmass. 6.3 Mid-tropospheric circulation It has been well studied that the mid-tropospheric circulation is mainly responsible for maintaining the monsoon over the period of 4 months, once it sets in. The gradient between geopotential heights at 500 hPa between two centers (50°–100°E, 20°–40°N) and (60°–100°E, 10°S– 10°N), defined as mid-tropospheric gradient is known to maintain the monsoon circulation throughout the season. Figure 11 shows the time series of this gradient for the period 1948–2004 which shows strong statistically significant decreasing trend, significant at 1% level which implies that the factor responsible for maintaining the monsoon is weakening over this period. Hence, not only the circulation is weakened, but the force required for maintaining the monsoon has also been weakened significantly in the recent period. The average gradient for the period 1948–1975 (−10.6) is significantly stronger than that for the period 1976–2004 (−5.4). The difference is statistically significant even at 1% level (t value0−3.9).

7 Conclusions It is well known fact that the globe as well as the Indian landmass has been warming consistently since mid-1970s. In this paper, an attempt has been made to examine whether associated with this unequivocal warming the nature of Indian summer monsoon rainfall has also changed. The features of monsoon are compared in two sub-periods 1901–1975 and 1976–2004. The main conclusions of this study are: (1) Though on a century long period the Indian summer monsoon is stable, free from any long term trend, it does depict decreasing tendency during the period 1976–2004. The average rainfall during the period

A. Kulkarni

(2) (3)

(4)

(5) (6)

1976–2004 is approximately 3% less as compared to average for 1901–1975 The deficit (excess) monsoons have become more (less) frequent during the recent period 1976–2004. The analysis over sub-seasonal and over regional scale shows that the July and August rainfall have decreased in the period 1976–2004 over the south-west and core monsoon region. In the recent period, 1976–2004, the IOD events happen to be more frequent, also co-occurrence of ENSO and IOD has been increased leading to the largest anomalies of ISMR. The low level monsoon circulation has weakened after mid-1970s. Mid-tropospheric gradient has been substantially weakened resulting in the weakening of the maintenance of monsoon circulation throughout the season.

Acknowledgements Author wish to thank Dr B N Goswami, Director, Indian Institute of Tropical Meteorology (IITM), Dr R Krishnan, Executive Director, Center for Climate Change Research, IITM for all the facilities provided. The valuable suggestions by the anonymous reviewers are duly acknowledged. Author is grateful to India Meteorological Department for providing the century long high resolution daily rainfall data on the Indian domain.

References Clark CO, Cole JE, Webster PJ (2000) Indian Ocean SST and Indian summer rainfall: predictive relationships and their decadal variability. J Climate 13:2503–2519 Ferranti L, Slingo JM, Palmer TN, Hoskins BJ (1997) Relations between interannual and intraseasonal monsoon variability as diagnosed from AMIP integrations. Q J R Meteorol Soc 123:1323–1357 Gadgil S, Vinaychandran PN, Francis PA, Gadgil S (2004) Extremes of the Indian summer monsoon rainfall, ENSO and equatorial Indian Ocean oscillation. Geophys Res Lett 31:L12213 Goswami BN (1998) Inter-annual variations of Indian monsoon in a GCM: external conditions versus internal feedback. J Climate 11:501–522 Goswami BN, Mohan RSA (2001) Intraseasonal oscillations and interannual variability of the Indian summer monsoon. J Climate 14:1180–1198 Goswami BN, Venugopal V, Sengupta D, Madhusoodanan MS, Xavier PK (2006) Increasing trend of Extreme Rain Events over India in a Warming Environment. Science 314:1442–1445 Guhathakurta P, Rajeevan M (2008) Trends in the rainfall pattern over India. Int J Climatol 28:1453–1469 IPCC (2007) Summary for policymakers. In: Solomon S, Qin D, Manning M, Chen Z, Marquis M, Averyt KB, Tignor M, Miller HL (eds) Climate change 2007: the physical science basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge Kalnay E et al (1996) The NCEP/NCAR 40-year reanalysis project. Bull Am Meteorol Soc 77:437–471

Weakening of Indian summer monsoon rainfall Kothawale DR, Munot AA, Borgaonkar HP (2008) Temperature variability over the Indian ocean and its relationship with Indian summer monsoon rainfall. Theor Appl Climatol 92:31–45 Kothawale DR, Munota A, Krishna Kumar K (2010) Recent trends in pre-monsoon daily temperature extremes over India. J Earth Syst Sci 119:51–65 Kripalani RH, Kulkarni A (1997) Climatic impact of El Nino/La Nina on the Indian monsoon: a new perspective. Weather 52:39–46 Kripalani RH, Kulkarni A (1999) Climatology and variability of Historical Soviet Snow Depth Data: some new perspectives in SnowIndian monsoon teleconnections. Clim Dyn 15:475–489 Kripalani RH, Kulakrni A, Sabade SS (2001) El Nino Southern Oscillation, Eurasian snow cover and Indian monsoon rainfall. Proc Indian Natl Sci Acad 67(A):361–368 Kripalani RH, Kulkarni A, Sabade SS (2003) Western Himalayan snow cover and Indian monsoon rainfall: a re-examination with INSAT and NCEP/NCAR. Theor Appl Climatol 72:1–18 Kripalani RH, Kulkarni A, Sabade SS, Revadekar JV, Patwardhan SK, Kulkarni JR (2004) Intra-seaonal oscillations during monsoon 2002 and 2003. Curr Sci 87:325–334 Krishna Kumar K, Rajgopalan B, Hoerling M, Bates G, Cane M (2006) Unraveling the mystery of Indian monsoon failure during El Nino. Science 314:115–119 Kulkarni A, Sabade SS, Kripalani RH (2006) Intra-seasonal vagaries of Indian summer monsoon rainfall. Research Report of Indian Institute of Tropical Meteorology, India, No RR-114, pp 51 Kulkarni A, Sabade SS, Kripalani RH (2009) Spatial variability of intra-seasonal oscillations during extreme Indian monsoons. Int J Climatol 29:1945–1955

459 Mooley DA, Shukla J (1987) Variability and forecasting of the summer monsoon rainfall over India. In: Chang CP, Krishnamurti TN (eds) Monsoon Meteorology (Oxford Monograph on Geology and Geophysics, No. 7, p. 544), Oxford University Press, New York, pp. 26–59 Rajeevan M, Bhate J, Kale JD, Lal B (2006) High resolution daily gridded rainfall data for the Indian region: analysis of break and active monsoon spells. Curr Sci 91:296–306 Rajeevan M, Bhate J, Jaswal AK (2008) Analysis of variability and trends of extreme rainfall events over India using 104 years of gridded daily rainfall data. Geophys Res Lett 35:L18707 Saji NH, Goswami BN, Vinayachandran PN, Yamagata T (1999) A dipole mode in the tropical Indian Ocean. Nature 401:360–363 Shepard D (1968) A two-dimensional interpolation function for irregularly-spaced data. Proceedings of the 1968 23rd ACM National Conference. Assoc. Comput. Mach, New York, pp 517–524 Shukla J (1975) Effect of Arabian Sea surface temperature anomaly on Indian summer monsoon: a numerical experiment with the GFDL Model. J Atmos Sci 32:503–511 Shukla J (1987) Interannual variability of monsoons. In: Monsoons. Jay S.Fein and Pamela L. Stephens (eds), John Wiley & Sons, Inc., pp. 399-464 Shukla J, Mooley DA (1987) Empirical prediction of the summer monsoon rainfall over India. Mon Weather Rev 115:695–703 Sperber KR, Slingo JM, Annamalai H (2000) Predictability and the relationship between subseasonal and interannual variability during the Asian summer monsoon. Q J R Meteorol Soc 126:2545–2574 Xavier PK, Goswami BN (2007) Promising alternative to prediction of seasonal mean all India rainfall. Curr Sci 93:195–202