Has modulation of Indian Summer Monsoon Rainfall by

Has modulation of Indian Summer Monsoon Rainfall by Sea Surface Temperature of the equatorial Pacific Ocean, weakened in recent years? SRIVASTAVA et al (2015) Clim Dyn DOI 10. 1007/s 00382 -015 -2470 -z

SSTs of Equatorial east pacific Ocean have significant influence on the ISMR Results of many studies suggest that the relationship between the deficient monsoon rainfall with El-Nino events, was stronger than the relationship of excess rainfall with the La-Nina events.

Points of inflection Post mid seventies, there were frequent El Nino events, still the ISMR was in normal range on most of the occasions. There was prolonged warming in the equatorial Pacific Oceans, during 1991 -1994. ISMR was almost normal during the period and it was above normal in 1994. During the year 1997, which witnessed one of the most severe EL-Nino events, ISMR was slightly above normal. Similarly, abnormal deficiency in the ISMR during 2002 could not be satisfactorily explained with the help of the moderate El-Nino event.

21 year running correlation coefficients between the ISMR and monsoon season (June-September) Nino 3 index series for the period 1951 -2010, plotted at the middle year of the 21 year window.

Premise of this study To identify a meteorological field which has imprints of the “Walker circulation/ Reverse Walker circulation” to be considered as proxy of the “Walker circulation/ Reverse Walker circulation”. And To examine whether the “Walker circulation/ Reverse Walker circulation” is still solely driven by SSTs of the equatorial east Pacific Ocean.

Composite velocity potential anomaly diagram of 0. 21 sigma level of the monsoon season for El Nino (1951, 1953, 1957, 1965, 1969, 1972, 1976, 1982, 1987, 1991, 1997, 2002, 2004 and 2009) Years.

Composite velocity potential anomaly diagram of 0. 21 sigma level of the monsoon season for La Nina (1954, 1955, 1956, 1964, 1970, 1971, 1973, 1975, 1988, 1999 and 2007) years.

Data used in the Study For this study, we have used daily and monthly data of velocity potential, geopotential height, at different pressure levels for the period 1951 -2010. ISMR data from the IMD archives and monthly Nino 3 index series and Sea Surface Temperature data from NCEP/NCAR (ER SST data) were used in the study. We divided the data into two parts 1951 -1980 and 1981 -2010 and did analysis separately for both the periods. This was done due to reported climate shift year around 1978 / 1979 and major shift in evolution of El –Nino events a major forcing influencing the ISMR (Wang 1995) around the same time.

Methodology v We did Principal Component Analysis (PCA) of monsoon months (June to September) Velocity potential of the area bound between latitudes 300 N to 300 S longitudes 600 E to 1050 W (domain of the study) for 1951 -1980 & 19812010 periods, separately. v Subsequently, we identified principal components correlated with SSTs of the Nino 3 region and the ISMR and carried further analysis.

Correlation Matrix for the period 1951 -1980 Rainfall % Depature(Jun – Sept. ) All India Rainfall % (Jun. Sep) Departure Velocity Potential at 0. 21 sigma level for the area bound between 29. 50 S-29. 50 N, 60. 0 0 E 105. 00 E Velocity Potential at 0. 21 sigma level for the area bound between 29. 50 S - 29. 50 N, 60. 0 0 E - 105. 00 E PC 1 PC 2 PC 3 PC 4 PC 5 PC 6 1. 00 PC 1 -0. 42 1. 00 PC 2 -0. 28 0. 00 1. 00 PC 3 -0. 01 0. 00 1. 00 PC 4 0. 07 0. 0 1. 00 PC 5 -0. 13 0. 0 1. 00 PC 6 -0. 19 0. 0 0. 0 1. 00 -0. 37 0. 68 0. 28 -0. 20 0. 13 -0. 18 0. 11 NINO 3 Index JUN– SEP NINO 3 Index Jun – Sept 1. 00

Correlation Matrix for the period 1981 -2010 All India Rainfall % Dep. (Jun – Sept) All India Rainfall % Departure (Jun- Sep) Velocity Potential at 0. 21 sigma level for the area bound between 29. 50 S-29. 50 N, 60. 0 0 E -105. 00 E Velocity Potential at 0. 21 sigma level for the area bound between 29. 50 S - 29. 50 N, 60. 0 0 E - 105. 00 E PC 1 PC 2 PC 3 PC 4 PC 5 PC 6 1. 00 PC 1 -0. 46 1. 00 PC 2 -0. 15 0. 0 1. 00 PC 3 -0. 21 0. 0 1. 00 PC 4 0. 20 0. 0 1. 00 PC 5 -0. 07 0. 0 1. 00 PC 6 0. 08 0. 0 0. 0 1. 00 -0. 16 0. 47 0. 13 -0. 14 -0. 10 -0. 06 0. 25 NINO 3 Index JUN– SEP NINO 3 Index Jun – Sept 1. 00

SPATIAL PATTERN OF EIGEN VECTORS OF THE PC 1 AND CORRESPONDING SPATIAL PATTERN OF THE SSTs

SPATIAL PATTERN OF EIGEN VECTORS OF THE PC 1 AND CORRESPONDING SST ANOMALIES FOR 1951 -1980 PERIOD (51, 57, 61, 63, 65 and 72)

SPATIAL PATTERN OF EIGEN VECTORS OF THE PC 1 AND CORRESPONDING SST ANOMALIES FOR 1981 -2010 PERIOD (82, 86, 87, 91, 94, 97, 02, 04 and 09)

Composite SST anomaly pattern for the months during which spatial pattern in the tropical velocity potential anomaly was just opposite (opposite polarity) to that of the Eigen vectors of the PC 1 for the periods 1951 -1980. and 1981 -2010. (54, 55, 64, 70, 73 and 75) (81, 88, 95, 96, 98, 07 and 10)

SPATIAL PATTERN OF EIGEN VECTOR OF SPATIAL PATTERN OF EIGEN THE PC 2 AND CORRESPONDING SPATIAL VECTORS OF THE PC 2 PATTERN OF THE SSTs

Spatial pattern of Eigen vectors of PC 2 of velocity potential data at 0. 21 sigma level (June- September) for 1951 – 1980 period and corresponding SST anomalies (61, 76, 79 and 80)

Spatial pattern of Eigen vectors of PC 2 of velocity potential data at 0. 21 sigma level (June- September) for 1981 – 2010 period and corresponding SST anomalies

Extra tropical circulation anomalies and the ‘Walker circulation’ It is well known that transient waves in the mid and upper troposphere active in the extra tropics. At times these waves move much south and interact with the tropical monsoon circulation systems. There are studies (Chang et al. 2001) which indicate that extra tropical circulation anomalies in recent years have (up to certain extent) countered adverse impact of the frequent occurring El Niño events over the ISMR in recent years. Subsequently, Ding and Wang (2005, 2007) and Srivastava et al. (2007, 2014) also highlighted growing role of extra tropics in modulating the ISMR.

Therefore, it would not be incorrect to hypothesize that the extra tropical circulation anomalies may be influencing tropical velocity potential field at 0. 21 sigma level, probably more in recent years. In view of above and keeping in the mind that transient waves in the extra-tropics move in middle and upper troposphere, we did PCA of vertically averaged geopotential height anomaly (for 600 to 300 h. Pa levels) for region bound between latitude 300 N – 600 N, longitude 600 E – 1050 W (same longitudinal domain but for the extra tropics) of the monsoon months for the two periods separately. We further calculated correlation between its PCs and the PC 1 of the velocity potential field of the 0. 21 sigma level and probable physical mechanism / reasons for their apparent association were looked into.

Correlation Matrix for the period 1951 -1980 Rainfall % Depature(Jun – Sept. ) Velocity Potential at 0. 21 sigma level for the area bound between 29. 50 S-29. 50 N, 60. 0 0 E -105. 00 E PC 1 All India Rainfall % (Jun. Sep) Departure PC 2 Average geopotential height anomaly for the area bound between 30 N 0 -600 N, 60. 0 0 E 105. 00 E PC 1 PC 2 NINO 3 Index Jun – Sept 1. 00 Velocity Potential at 0. 21 sigma level for the area bound between 29. 50 S-29. 50 N, 60. 0 0 E- 105. 00 E PC 1 -0. 42 1. 00 PC 2 -0. 28 0. 00 1. 00 Average geopotential height anomaly for the area bound between 30 N 0600 N, 60. 00 E 105. 00 E PC 1 -0. 15 0. 11 -0. 24 1. 00 0. 00 -0. 30 PC 2 -0. 01 0. 16 0. 00 1. 00 0. 04 -0. 37 0. 68 0. 26 -0. 30 0. 04 1. 00 NINO 3 Index JUN– SEP

Correlation Matrix for the period 1981 -2010 All India Rainfall % Dep. (Jun – Sept) Velocity Potential at 0. 21 sigma level for the area bound between 29. 50 S-29. 50 N, 60. 0 0 E- 105. 00 E PC 1 All India Rainfall % Departure (Jun- Sep) PC 2 Average geopotential height anomaly for the area bound between 30 N 0 -600 N, 60. 0 0 E 105. 00 E PC 1 PC 2 NINO 3 Index Jun – Sept 1. 00 Velocity Potential at 0. 21 sigma level for the area bound between 29. 50 S-29. 50 N, 60. 0 0 E- 105. 00 E PC 1 -0. 46 1. 00 PC 2 -0. 15 0. 0 1. 00 Average geopotential height anomaly for the area bound between 30 N 0 -600 N, 60. 00 E -- 105. 00 E PC 1 -0. 30 0. 46 0. 19 1. 00 0. 00 -0. 04 PC 2 0. 27 -0. 02 -0. 06 0. 00 1. 00 0. 25 NINO 3 Index JUN– SEP -0. 16 0. 47 0. 13 -0. 04 0. 25 1. 00

SPATIAL PATTERN OF EIGEN VECTORS OF SPATIAL PATTERN OF EIGEN THE PCs AND CORRESPONDING SPATIAL VECTORS OF THE PC 2 PATTERN OF THE GPM HTS

Spatial pattern of Eigen vectors of PC 2 of average geopotential height (600 h Pa to 300 h Pa levels) of monsoon months (June- September) for a) 1951 – 1980 and b) 1981 -2010 periods. (a) (b)

Spatial pattern of Eigen vectors of PC 1 of average geopotential height (600 h Pa to 300 h Pa levels) of monsoon months (June- September) for a) 1951 – 1980 and b) 1981 -2010 periods. (a) (b)

Composite average geopotential height anomaly pattern for the days during which spatial pattern in average geopotential height anomaly was similar to that of the Eigen vectors of the PC 1 for the period 1951 -1980.

Composite average geopotential height anomaly pattern for the days during which spatial pattern in average geopotential height anomaly was similar to that of the Eigen vectors of the PC 1 for the period 1981 -2010.

Trend in the PC 1 series of monsoon months (June. September) of tropical velocity potential at 0. 21 sigma level for the period 1951 -2010.

Results from the Analysis 1951 -1980 Period SSTs V P FIELD ISMR 1981 -2010 Period ET C A SSTs V P FIELD ISMR

Answer to some questions WHETHER MONSOON RAINFALL LIKE 1997, HAS BECOME INDEPENDENT OF EL-NINO EVENTS ( IMPACT OF EASTWARD SHIFT OF WALKER CIRCULATION OVER CIRCULATION PATTERN CAUSING MONSOON RAINFALL OVER INDIA) WHY “CONTRACTION OF MONSOON SEASON CONCEPT” WAS NOT NECESSARY TO EXPLAIN ISMR-ENSO RELATION BEFORE CLIMATE SHIFT EVEN THOUGH YASUNARI T (1987) USING PAST EL-NINO EVENTS, HIGHLIGHTED THAT EL-NINO EVENTS TEND TO LOWER EURASIA TEMPERATURE ( ENSO-MONSOON RELATION HAS REMAINED STEADY AND IF A PHYSICALLY BASED DEFINITION OF MONSOON SEASON IS CONSIDERED, ISMR-ENSO RELATION IS INTACT)

CONCLUSIONS We find that during 1951 -1980 period, two different patterns of the velocity potential field, forced by probably two distinct modes of El Nino episodes, were associated with the ISMR. These two El Nino episodes, respectively, correspond to the strong El Nino events, where in warming was extended up to the date line (primarily zonal) and the moderate El Nino events in which, warming having north south extension, was limited to the eastern Pacific Ocean only.

CONCLUSIONS However, in recent years, only the first pattern of the velocity potential field, induced by the strong El Nino events (warming extending up to the date line), was correlated with the ISMR. Further, in the 1981 -2010 period, velocity potential field at 0. 21 sigma level over the tropical Pacific and Indian Oceans, which appeared to be primarily driven by SST anomalies of the equatorial Pacific Ocean in the 1951=1980 period, was found to be significantly correlated with the extra tropical circulation anomalies also. Therefore, modulation of the ISMR through velocity potential field over the tropical Indian and Pacific Oceans, in the later period, may have additional significant impact of the extra-tropical circulation anomalies. This might have to led weak correlation between the ISMR and SSTs of the Nino 3 region, which is actually being observed in recent years.

“ We do not know one millionth of one percent about anything” -Thomas Alva Edison

Thank you 10/2/2020 34