• Open Access

Relationship between ESI tendency and Indian monsoon rainfall: a possible mechanism

Authors

  • S. B. Kakade,

    Corresponding author
    1. Forecasting Research Division, Indian Institute of Tropical Meteorology, Pune-8, Maharashtra 411009, India
    • Forecasting Research Division, Indian Institute of Tropical Meteorology, Dr Homi Bhabha Road, Pashan, Pune, Maharashtra 411009, India.
    Search for more papers by this author
  • Ashwini Kulkarni

    1. Forecasting Research Division, Indian Institute of Tropical Meteorology, Pune-8, Maharashtra 411009, India
    Search for more papers by this author

Abstract

In this article, we have studied spatial and temporal variabilities of temperature and zonal wind field over 90°–0°N, 90°W–90°E during contrasting phases of Effective Strength Index (ESI) tendency and its relationship with Indian summer monsoon rainfall (ISMR). During positive (negative) ESI tendency, there is winter-time anomalous cooling (warming) over western Eurasia surrounding Moscow but magnitude of warming is stronger than cooling. The correlation analysis also suggests that negative ESI tendency is significantly associated with warming over western Eurasia during winter and spring seasons. The study further reveals that positive (negative) ESI tendency is indicating winter-time anomalous trough (ridge) over the Russian region surrounding Moscow. Copyright © 2011 Royal Meteorological Society

1. Introduction

In spite of the recent technological advancement, Indian economy is largely affected by the summer monsoon (June–September) rainfall activity over India. This is mainly due to the variability of Indian summer monsoon rainfall (ISMR) and its governing meteorological factors (parameters) on different time and space scales. The quantum of the ISMR received has strong socio-economic consequences. Long-range forecast of ISMR has a long history and some of the techniques used for ISMR forecasting for over past century have been reviewed by Thapliyal (1986) and Krishna Kumar et al. (1995). The first official seasonal monsoon forecast was issued by Sir Henry Blanford in 1884 using Himalayan snowfall as a predictor (Blanford, 1884). Sir Gilbert Walker was the first meteorologist who systematically examined the relationship between ISMR and global circulation parameters by statistical techniques and pointed out 28 predictors based on regression equation. Some parameters out of these 28 have lost their significance in due course of time. Banerjee et al. (1978); Kung and Sharif (1982), Bhalme et al. (1986), Parthasarathy et al. (1988, 1991), Gowariker et al. (1989, 1991), Dugam et al. (1990, 1993, 1997, 2009), Krishna Kumar et al. (1995, 1997), Munot and Pant (1998), Dugam and Kakade (1999, 2004), Kakade and Dugam (2000, 2006), Munot and Krishna Kumar (2007), etc. have also invented some new parameters for the ISMR prediction. All these studies explain how a parameter or set of parameters are physically linked with the summer monsoon rainfall activity over India. Kakade and Dugam (2000) have defined an index called Effective Strength Index (ESI) on the basis of monthly indices of North Atlantic Oscillation (NAO) and Southern Oscillation (SO). They have also shown that during excess monsoon years the ESI decreases from January to April, whereas during deficient monsoon years there is a rising tendency of ESI from January to April. Further study by Kakade and Dugam (2006) has shown that the tendency of ESI from January to April (ESI tendency) can be used as a precursor for the prediction of summer monsoon rainfall over different homogeneous regions of India like northwest India, west/central India, peninsular India and India as a whole. Because each parameter has its own temporal variability, its impact on ISMR may change drastically during two contrasting phases (positive and negative) of the parameter. In this study, we discuss possible mechanisms of the ESI tendency and monsoon relationship. It is observed that variability in temperature and zonal wind field over Eurasia mainly provides the pathway to this relationship.

2. Data used

For this study, the following data for the period 1951–2007 have been used.

  • 1.The summer monsoon rainfall (June–September) data for different homogeneous regions of India and India as a whole have been taken from the website www.tropmet.res.in. The percentage departure from long-term mean is calculated and these indices are used for further analysis.
  • 2.The 1000-hPa temperature (in degrees Kelvin) over quarter-sphere 90°–0°N, 90°W–90°E have been obtained from the website http://www.cru.uea.ac.uk/cru/data/ncep/qs_eurasia
  • 3.The 1000-hPa zonal wind data (in m s−1) over quarter-sphere 90°–0°N, 90°W–90°E have been obtained from the website http://www.cru.uea.ac.uk/cru/data/ncep/qs_eurasia
  • 4.NAO and SO data have been taken from www.cpc.ncep.noaa.gov.
  • 5.ESI is defined as the algebraic difference between monthly indices of NAO and SO. The anomalies from the annual mean have been calculated for each month and these anomaly series are then divided by the standard deviation. These series are called as ESI series of respective months. ESI tendency is the difference between April and January ESI values.

Seasonal means for all the above parameters are computed by averaging over the seasons winter (December–January–February), spring (March–April–May), summer (June–July–August) and autumn (September–October–November).

3. Results

It is a well-known fact that the monsoon activity over India is related to the strength of both the oscillations (NAO and SO) rather than the single oscillation. Therefore, it is reasonable to consider the simultaneous impact of these two oscillations on the ISMR. ESI measures the resultant signal from NAO and SO. Positive ESI indicates that the strength of the NAO is relatively more than that of the SO, which weakens the net air-flow from the southern hemisphere to the northern hemisphere. The monthly mean ESI for excess (deficient) summer monsoon years suggests positive (negative) ESI from April through September (Kakade and Dugam, 2000).

ESI tendency indicates how the relative strength of the two oscillations is evolved from winter to spring. The positive (negative) phase of ESI tendency means positive (negative) ESI from April through September (Kakade and Dugam, 2006). During 1951–2007, there are 29 episodes of positive ESI tendency and 28 episodes when ESI tendency is negative. Figure 1 shows the composite seasonal mean of NAO and SO during contrasting phases of ESI tendency. It suggests that during positive (negative) ESI tendency there is rising (falling) NAO tendency and falling (rising) SO tendency from winter to spring. Thus, ESI tendency represents the simultaneous evolution of NAO and SO from winter to spring.

Figure 1.

Composite seasonal mean NAO (blue line) and SO (pink dashed line) during (a) positive and (b) negative ESI tendency for 1951–2006

Figure 2 shows the time series of ESI tendency and percentage departure from long-term mean of ISMR for 1951–2007. The correlation coefficient (CC) between them (CC = − 0.4), is statistically significant at 5% level. The CCs of ESI tendency with ISMR during contrasting phases (positive and negative) of ESI tendency are 0.00 and − 0.39, respectively. The statistically significant CC is observed during negative ESI tendency only. Thus, there is a clear line of demarcation between the relationship of ESI tendency and ISMR as statistically insignificant (significant) during positive (negative) phase of ESI tendency.

Figure 2.

Time series of ESI tendency (dashed line) and Indian summer monsoon rainfall expressed as percentage departure from long-term mean (straight line) for 1951–2007

3.1. Temperature variability during contrasting phases of ESI tendency

Figure 3 depicts the composite mean of 1000-hPa temperature anomaly during positive (29 years) phase of ESI tendency for winter (pewint) and spring (pesprt) seasons and negative (28 years) phase of ESI tendency for winter (newint) and spring (nesprt) seasons. During positive (negative) ESI tendency, there is a negative (positive) temperature anomaly over the region (20°–90°E, 40°–60°N) and positive (negative) temperature anomalies over the region (60°–20°W, 60°–80°N) in preceding winter season whose spatial extension is reduced due to splitting into small areas in spring and shifted towards the west.

Figure 3.

Composite mean of 1000-hPa temperature anomaly during positive (29 years) phase of ESI tendency for winter (pewint) and spring (pesprt) seasons and negative (28 years) phase of ESI tendency for winter (newint) and spring (nesprt) seasons

Anomalous warm temperature over 20°–90°E, 40°–60°N (western Eurasia surrounding Moscow) in winter season may melt the Eurasian snow which creates more snow-free surface area over Eurasia and hence reducing the reflection of solar radiation back into the atmosphere due to the albedo effect. It involves relatively more solar energy in the land–atmosphere system which is again utilized for snow melting. This type of natural feedback process is not possible during anomalous cooling over the same region. Lydolph (1976) reports that the distribution of maximum snow depth is unlike that of snowfall because of the temperature influences on melting. Over the Eurasian plain, the region of maximum snowfall in the west experiences enough thawing during winter to keep the maximum depth of snow cover at not more than 20 cm except in the mountains. Therefore, just anomalous cold temperatures over 20°–90°E, 40°–60°N, during positive ESI tendency, will not allow accumulation of snow over this region, but winter-time anomalous warming over this region, during negative ESI tendency, may reduce the snow depth due to melting.

In order to understand the impact of ESI tendency on the winter-time thermal field over the region (20°–90°E, 40°–60°N) in relation to ISMR through the snow depth variability over west Eurasia surrounding Moscow, composite means of winter temperature anomaly over each 2.5° × 2.5° grid of Eurasian quarter-sphere (90°–0°N, 90°W–90°E) during above-normal and below-normal ISMR in either phases of ESI tendency are computed. Out of 29 positive ESI tendency years, 13 years have above-normal rainfall and 16 years have below-normal rainfall. Similarly, out of 28 negative ESI tendency years, 18 years show above-normal rainfall and 10 years show below-normal rainfall. The probability of above-normal rainfall during negative ESI tendency is more than the probability of below-normal rainfall during positive ESI tendency. But, it is also observed that 13 years (10 years) of positive (negative) ESI tendency show above-normal (below-normal) ISMR. Therefore, it is necessary to understand the composite mean of winter temperature field over Eurasia during contrasting phases of ESI tendency and ISMR. Figure 4 depicts winter-time composite mean temperature anomaly during (a) 16 years of positive ESI tendency and negative ISMR (peniwin), (b) 10 years of negative ESI tendency and negative ISMR (neniwin), (c) 13 years of positive ESI tendency and positive ISMR (pepiwin) and (d) 18 years of negative ESI tendency and positive ISMR (nepiwin). The figure suggests that during positive (negative) ESI tendency most of the area in the region (20°–90°E, 40°–60°N) show anomalous cooling (warming). When this thermal field condition is diluted by the opposite thermal field over the region near to the region under consideration (20°–90°E, 40°–60°N), then we get reverse rainfall activity over India (above-normal rainfall during positive ESI tendency and below-normal rainfall during negative ESI tendency) during either phases of ESI tendency. During negative ESI tendency when anomalous warming over 20°–90°E, 40°–60°N is diluted by anomalous cooling over northwest side of the region under consideration then the snow depth over western Eurasia surrounding Moscow is not reduced considerably and hence the rainfall activity over India is reduced (neniwin). Similarly, during positive ESI tendency when anomalous cooling over 20°–90°E, 40°–60°N is diluted by anomalous warming over most of the land mass over 20°–80°E, 20°–40°N then instead of below-normal rainfall activity we get above-normal rainfall activity (pepiwin).

Figure 4.

Winter-time composite mean temperature anomaly during (a) 16 years of positive ESI tendency and negative ISMR (peniwin), (b) 10 years of negative ESI tendency and negative ISMR (neniwin), (c) 13 years of positive ESI tendency and positive ISMR (pepiwin) and (d) 18 years of negative ESI tendency and positive ISMR (nepiwin)

Rajeevan et al. (2005) pointed out that land surface air temperature anomaly over northwest Europe (average of five stations: Orland, Oslo/Gendermon, Ostursund/Froson, Karlstad and De Bilt) is directly associated with ISMR. During negative phase of ESI tendency, when winter-time anomalous warming over the region (20°–90°E, 40°–60°N) is accompanied by anomalous cooling (warming) over the region surrounded by these five stations then ISMR is below-normal (above-normal). During positive phase of ESI tendency, no such contrasting anomalous temperature field is observed over the region surrounded by five stations. From the same figure, it is also observed that the direct association between land surface air temperature anomaly over northwest Europe and ISMR, as suggested by Rajeevan et al. (2005), becomes more stronger during negative phase of ESI tendency. This may be the probable mechanism for significant relationship between negative ESI tendency and ISMR.

3.2. Relationship between ESI tendency and 1000-hPa temperature anomaly

The CC between ESI tendency and 1000-hPa temperature anomaly during positive (29 years) phase of ESI tendency for winter (pesiccwint) and spring (pesiccsprt) seasons and negative (28 years) phase of ESI tendency for winter (nesiccwint) and spring (nesiccsprt) seasons are shown in Figure 5. It reveals that positive ESI tendency is inversely associated with 1000-hPa temperature anomaly over 10°–70°E, 60°–80°N region in winter season and over 0°–20°E, 35°–45°N region in spring season. Negative ESI tendency is inversely associated with temperature anomaly over the regions 10°–80°E, 50°–60°N and 30°–60°E, 40°–60°N in winter and spring seasons, respectively. In addition to this, negative ESI tendency is directly associated with temperature anomaly over the regions 70°–40°W, 60°–80°N and (30°–10°W, 60°–65°N in winter and spring seasons, respectively.

Figure 5.

Relationship between ESI tendency and 1000-hPa temperature anomaly during positive (29 years) phase of ESI tendency for winter (pesiccwint) and spring (pesiccsprt) seasons and negative (28 years) phase of ESI tendency for winter (nesiccwint) and spring (nesiccsprt) seasons

If we combine these results with the composite analysis of winter-time 1000-hPa temperature anomaly then it is observed that the region 20°–90°E, 40°–60°N experiences strong anomalous warming during negative ESI but not strong anomalous cooling during positive ESI tendency, because positive ESI tendency is showing anomalous cooling over the region 10°–70°E, 60°–80°N which is to the north of region 20°–90°E, 40°–60°N. Thus, only negative ESI tendency is affecting the temperature field over 20°–90°E, 40°–60°N.

3.3. A 1000-hPa U-wind variability during contrasting phases of ESI tendency

Meehl (1994) showed that during winter prior to strong monsoon over northern hemisphere there is anomalous ridging over Asia, which helps in keeping South Asian surface relatively warmer. Meehl (1997) further showed the role of tropical convective heating anomalies in forcing the described anomalous mid-latitude circulation via a remote Rossby wave response. Since ESI tendency is defined with mid-latitude oscillation NAO, we must understand the relationship between ESI tendency and surface zonal wind patterns over quarter-sphere 90°–0°N, 90°W–90°E.

The composite means of 1000-hPa winter and spring U-component anomaly during positive (29 years) phase of ESI tendency for winter (pewinu) and spring (pespru) seasons and negative (28 years) phase of ESI tendency for winter (newinu) and spring (nespru) seasons are calculated and depicted in Figure 6. During positive (negative) ESI tendency, there is anomalous easterly (westerly) field over the region 20°W–20°E, 50°–60°N and anomalous westerly (easterly) field over the region 40°–10°W, 20°–40°N in winter season. In spring season these anomalous westerly and easterly fields are shifted eastward. In winter season, there is a trough (ridge) over Russian region surrounding Moscow showing anomalous westerly (easterly) field to the south and anomalous easterly (westerly) field to the north during positive (negative) phase of ESI tendency. The similar U-wind field is persisted in spring during negative phase of ESI tendency only.

Figure 6.

Composite mean of 1000-hPa winter and spring U-component anomaly during positive (29 years) phase of ESI tendency for winter (pewinu) and spring (pespru) seasons and negative (28 years) phase of ESI tendency for winter (newinu) and spring (nespru) seasons

4. Conclusions

  • 1.An inverse relationship between ESI tendency and ISMR depends upon the phase of ESI tendency. The relationship is significant during negative ESI tendency only.
  • 2.Positive (negative) ESI tendency indicates winter-time anomalous cooling (warming) over western Eurasia surrounding Moscow. Only negative winter-time ESI tendency is responsible for anomalous surface warming over western Eurasia surrounding Moscow which may reduce the snow depth considerably and hence conducive for good monsoon over India.
  • 3.Positive (negative) ESI tendency is indicating winter-time anomalous trough (ridge) over Russian region surrounding Moscow. The anomalous ridge during negative phase of ESI tendency persists in spring season also.
  • 4.The direct association between winter temperature anomaly over northwest Europe and ISMR seems to be more prominent during the negative phase of the ESI tendency.

Acknowledgements

The authors thank Dr B N Goswami, Director, Indian Institute of Tropical Meteorology and Dr P N Mahajan, Head, Forecasting Research Division for all the facilities provided. We are grateful to anonymous reviewers whose suggestions have helped a lot to improve the manuscript.

Ancillary