Embed Size (px)
Citation preview
science.sciencemag.org/content/370/6522/1335/suppl/DC1
Supplementary Materials for
Indian monsoon derailed by a North Atlantic wavetrain
P. J. Borah, V. Venugopal, J. Sukhatme, P. Muddebihal, B. N. Goswami
*Corresponding author. Email: [email protected] or [email protected]
Published 11 December 2020, Science 370, 1335 (2020)
DOI: 10.1126/science.aay6043
This PDF file includes:
Data and Methods
Supplementary Text
Figs. S1 to S7
Tables S1 and S2
References
Data and Methods
Data
• Sea Surface Temperature (SST): 1901-2015, monthly, 1◦ gridded Had-SST from the Hadley
Centre, UK Met Office (41). This dataset can be obtained from
https://www.metoffice.gov.uk/hadobs/hadsst3/data/download.html.
• Rainfall: 1901-2015, daily, 1◦ gridded rainfall dataset from the India Meteorological Depart-
ment (IMD). This product can be obtained from www.imd.gov.in. The homogeneous
Indian monthly rainfall dataset is maintained by Indian Institute of Tropical Meteorology
(IITM), Pune and can be obtained from www.tropmet.res.in (40).
• ERA 20C: 1901-2010, daily, 1◦ gridded zonal and meridional winds and vorticity. This
product (39) can be obtained from
https://www.ecmwf.int/en/forecasts/datasets/reanalysis-datasets/
era-20c.
Methods
• Identifying El Nino Years: El Nino years were identified by an inspection of anomalies
of detrended SST during JJAS. Post-1950 events were confirmed based on estimates from
the Climate Prediction Center, NOAA, and can be accessed via “El Nino / La Nina −→
Historical Information” at https://www.cpc.ncep.noaa.gov/.
• Detrended SST Anomalies: A linear trend in JJAS mean SST is removed at each location for
the period 1901-2015, and fluctuations around the trend composited for the years of interest.
• Rainfall, Wind and Vorticity Anomalies: If F (x, y, d, yr) represents rain, wind or vorticity at
a location (x, y) on day d during JJAS, in year yr, then the daily anomalies at each location
are estimated as: Fanom(x, y, d, yr) = F (x, y, d, yr) − Fclim(x, y, d) where Fclim(x, y, d) =
1N
[yr=N∑yr=1
F (x, y, d, yr)
]. The anomalies of area-averaged rainfall (e.g., Fig. 2a) are es-
timated by calculating daily area-averaged rainfall and then subtracting the corresponding
1
daily climatology. Seasonal anomalies of F (e.g., Fig. 4) can be estimated either by sub-
tracting JJAS means from the corresponding seasonal climatology, or by averaging the daily
anomalies Fanom(x, y, d, yr) over JJAS.
• Composites: The ERA20C data is available up to 2010 and hence 2014 was not included
in the wind and vorticity composites. Furthermore, the composites presented have been
constructed based on the anomalies of 7 of the 9 NEN + Dr years for which the reanalysis
data is available. This is because the lower level (850 mb) circulation features over the Indian
region are not consistent with the observed IMD rainfall during the break periods for two of
the early era years (1901 and 1904). This could either be due to the quality of early era IMD
observations or the fidelity of reanalysis data (the years being close to spin-up period (39)).
2
Supplementary Text
Intra-category Heterogeneity. The rainfall evolution for each of the droughts suggests consider-
able intra-category heterogeneity. The spread in normal years (gray dashed lines in Figs. S3b and
S3d) is substantially lower than in either of the drought categories. Moreover, Fig. S3d indicates
that NEN + Dr can be near-normal until as late as end of July/early August, before the sudden
transition to an eventual drought.
Statistical Association between North Atlantic and NEN Droughts. The cold anomaly co-
located with the observed tropospheric vorticity anomaly is related to the fact that most NEN
droughts occur during the negative phase of the decadal variation in North Atlantic SSTs (blue
filled circles in Fig. S6a). This localized cold anomaly (Fig. 1b) is similar to a mode of variability
of the North Atlantic SSTs, commonly referred to as the East Atlantic / West Russia (EA/WR)
pattern (42). This prompts an examination of the state of the atmosphere (i.e., vorticity) in years
when the North Atlantic SST is anomalously “warm” and “cold” (Fig. S6a). A comparison of
daily total columnar vorticity in these years (Fig. S6b) suggests that there is a higher likelihood of
positive (negative) vorticity anomalies during a “cold” (“warm”) phase (18, 43). Moreover, in this
predominantly cyclonic environment characterizing the “cold” years, vorticity anomalies build up
in an episodic manner, with a tendency to occur in June and August and persist for 2 to 3 weeks
(Fig. S6b). During NEN drought years, vorticity anomalies (blue curve in Fig. S6b) are larger
in June and amplified significantly in August (Table S2), and their footprint (Fig. 4) can be seen
in the rainfall evolution during these droughts (Fig. 2 and Figs. 3j, k). These episodes are strong
enough to make the seasonal response (Fig. S5) seem like a deep tropospheric cyclonic circulation.
The statistical significance of JJAS daily total columnar vorticity anomalies during NEN + Dr
years is established via a suite of three tests (Table S2). The first test shows that these anomalies
are significantly different from the long-term mean (zero by definition) in a seasonal sense with
a p-value ≈ 10−7. The second and third tests are more stringent and involve a comparison with
years when the North Atlantic is “cold” (SST anomaly < −0.25◦C; Fig. S6a) and there is a
preference for a cyclonic environment (Fig. S6b; (19)). Here too, the columnar vorticity anomalies
3
are significantly different both in a seasonal as well as an episodic sense, with p-values of 4.6%
and 0.3%, respectively.
4
SUPPLEMENTARY FIGURES
−10 −5 0 5 10 15 200
0.1
0.2
Daily Rainfall Anomaly (mm/d)
No
rmalised
His
tog
ram
(P
DF
)
−10 −5 0 5 10 15 200
0.5
1
Cu
mu
lati
ve D
istr
ibu
tio
n F
un
cti
on
(C
DF
)
−10 −5 0 5 10 15 200
0.5
1
−10 −5 0 5 10 15 200
0.5
1EN+Dr
NEN+Dr
Normal
Fig. S1: Normalized histograms (shown as line curves) and cumulative distribution functions(CDF) of anomalies of daily area-averaged rainfall over central India (16.5N-26.5N, 74.5E-86.5E;land only; also, see also Methods section) for EN + Dr (red), NEN + Dr (blue) and normal (gray).The light-green shading highlights the difference between the two kinds of droughts as well as theirdifference between them and a normal year. The CDFs of EN + Dr (red) and NEN + Dr (blue)are statistically indistinguishable at 5% significance level (p-value of 0.09), but are both signifi-cantly different from a normal year (gray “S” curve; p-value < 10−5). The empirical CDFs havebeen compared using a two-sample Kolmogorov-Smirnov test with a null hypothesis that they areindistinguishable (38).
5
Fig. S2: Time series of anomalies of daily area-averaged rainfall (mm/d) over central India (16.5N-26.5N, 74.5E-86.5E; land only; see also Methods section), during JJAS, shown with a color-coding,for each of the drought years during (A) an El Nino year (EN + Dr) and (B) a non El Nino year(NEN + Dr). The red and blue ellipses are used to highlight the tendency of long breaks to clusterin different times of the season, in the two kinds of droughts. Colorbar is in mm/d. The average ofthese time series is shown as thin red and blue curves, respectively, in Fig. 2a.
6
1 31 62 93
−200
−150
−100
−50
0
50
Cu
mu
lati
ve o
f D
aily R
ain
fall A
no
maly
1905
1911
1918
1941
1951
19651972
1982
1987
2002
2004
2009
2015
A
EN + Dr
1 31 62 93
−200
−150
−100
−50
0
50
B
Jun Jul Aug Sep
−200
−150
−100
−50
0
50
Cu
mu
lati
ve o
f D
aily R
ain
fall A
no
maly
1901
1904
1920
1966
1968
1974
1979
1985
1986
2014
C
NEN + Dr
Jun Jul Aug Sep
−200
−150
−100
−50
0
50
D
Fig. S3: Cumulative curves of daily anomalies of area-averaged rainfall over central India (16.5N-26.5N, 74.5E-86.5E; land only; see also Methods section) for each of the drought years belongingto the two categories, along with the respective average cumulative curves: (A, B) EN + Dr and(C, D) NEN + Dr. The corresponding spread (±1 standard deviation) in each of the categories isshown as a shading in panels (B) and (D). The standard deviation is estimated based on 13 EN +Dr and 10 NEN + Dr years. The dashed gray lines in panels (B) and (D) correspond to the spread(±1 standard deviation) of 18 normal years. The solid red and blue curves in (B, D) are the sameas the ones shown in Fig. 2b.
7
1 31 62 93−4
−3
−2
−1
0
1
2
3D
aily R
ain
fall A
no
maly
(m
m/d
)
A
EN+Dr
NEN+Dr
Normal
Jun Jul Aug Sep−150
−125
−100
−75
−50
−25
0
Cu
mu
lati
ve o
f D
aily R
ain
fall A
no
maly
(m
m)
Days
B
Fig. S4: Same as in Fig. 2, but for area-averaged all-India rainfall.
8
Fig. S5: Composites of anomalies of JJAS mean wind (vectors) and vorticity (shading) in thefree troposphere, based on NEN + Dr years; the shading in the bottom-most panel represents SSTanomaly. The colorbar applies to both vorticity (× 2 × 105) and SST. The wind and vorticityanomalies are based on ERA 20th Century Reanalysis data, and SST anomalies are based on the1◦, monthly SST product from the Hadley Centre. See Methods.
9
1900 1920 1940 1960 1980 2000 −1.5
−1
−0.5
0
0.5 1
1.5S
ST
An
om
aly
A
Jun Jul Aug Sep−3
−2
−1
0
1
2
3
4
Days
To
tal
Co
lum
na
r V
ort
icit
y
B
SST < − 0.25
SST > 0.25
NEN + Dr
Fig. S6: (A) Long-term (1900-2009) time series of the North Atlantic SST anomalies (averagedover 35N-55N; 310E-340E with NEN droughts marked as blue filled circles. The red and greenbars represent “warm” and “cold” years when the SST anomaly is > 0.25◦C and < −0.25◦C (≈half the standard deviation of long-term SST data), respectively. Each of these categories compriseof approximately 30 years. (B) Time series of anomalies of daily total columnar (200, 300, 500and 700 mb) vorticity (× 105) averaged over this area during JJAS, composited over the “warm”and “cold” years. The thick red and green curves represent the respective first three harmonics.Time series of anomalies of daily total-columnar vorticity over the North Atlantic box for NEN +Dr years is shown in blue.
10
Fig. S7: Composites of anomalies of wind (vectors) and vorticity (×105; shading) at (A) 700 mband (B) 850 mb, during the 20-day period prior to the long break (see blue ellipse in Fig. S2B) inNEN+Dr years. Legend for arrows is shown in panel (B). Based on ERA 20th Century Reanalysisdata (39). See Methods section for the construction of composites.
11
SUPPLEMENTARY TABLES
Table S1: The list of Indian monsoon years where the seasonal (June through September; JJAS)rainfall anomaly is less than -10% of the long-term (1901-2015) mean, and is classified as adrought. The anomalies shown are based on the IITM homogeneous Indian monthly rainfalldataset (40) for the period 1901-2015. The droughts that occurred during an El Nino (no El Nino)year are marked in red (blue) (see also Fig. 1c).
Year Seasonal Rainfall Anomaly(% of long-term mean)
1901 -151904 -121905 -161911 -141918 -241920 -161941 -151951 -131965 -171966 -131968 -111972 -231974 -121979 -171982 -141985 -111986 -131987 -182002 -222004 -132009 -222014 -142015 -14
12
Table S2: Statistical significance of anomalies of daily total columnar vorticity during 9 NEN + Dr years (Fig. S6b; see also “Supplementary Text”).µ represents “population” mean, VN and SVN
represent the sample mean and standard deviation of N years of daily observations for the period ofinterest (in bold in column 1) during JJAS. “Cold” years represent those years when the North Atlantic SST anomaly < −0.25◦C (see Fig. S6a).The second and third tests are based on a two-sample t-test for comparing means (38).
Description Hypothesis Sample Statistics Test Statistic p-valueValue
Test1: Comparison with H0: µω = 0 V9 = 6.3× 10−6
long-term JJAS mean H1: µω = 0 SV9 = 4× 10−5 5.2 2P (|t1097| > 5.2)(2-sided) N = 9×122 = 1098 ≈ 0
Test2: Comparison with JJAS V9 = 6.3× 10−6
mean during “cold” H0: µNEN+Drω = µCold
ω V30 = 3.7× 10−6
North Atlantic years H1: µNEN+Drω < µCold
ω SV9 = 4× 10−5 1.7 P (tN9+N30−2 > 1.7)(1-sided) SV30 = 4.5× 10−5
N9 = 9 × 122 = 1098 ≈ 4.6%N30 = 30 × 122 = 3660
Test3: Comparison with August V9 = 1.9× 10−5
episode mean (days 70-90) during H0: µNEN+Drω = µCold
ω V30 = 0.97× 10−5
“cold” North Atlantic years H1: µNEN+Drω < µCold
ω SV9 = 3.4× 10−5 2.8 P (tN9+N30−2 > 2.8)(1-sided) SV30 = 4× 10−5
N9 = 9 × 21 = 189 ≈ 0.3%N30 = 30 × 21 = 630
13
References and Notes
1. B. Parthasarathy, D. A. Mooley, Some features of a long homogeneous series of Indian
summer monsoon rainfall. Mon. Weather Rev. 106, 771–781 (1978). doi:10.1175/1520-
0493(1978)106<0771:SFOALH>2.0.CO;2
2. V. Krishnamurthy, J. Shukla, Intraseasonal and interannual variability of rainfall over India. J.
Clim. 13, 4366–4377 (2000). doi:10.1175/1520-0442(2000)013<0001:IAIVOR>2.0.CO;2
3. S. Gadgil, The Indian monsoon and its variability. Annu. Rev. Earth Planet. Sci. 31, 429–467
(2003). doi:10.1146/annurev.earth.31.100901.141251
4. C. D. Hoyos, P. J. Webster, The role of intraseasonal variability in the nature of Asian
monsoon precipitation. J. Clim. 20, 4402–4424 (2007). doi:10.1175/JCLI4252.1
5. D. Sikka, “Monsoon drought in India,” Tech. Rep. 2, COLA/CARE, Maryland, USA (1999).
6. B. N. Goswami, “South Asian monsoon” in Intraseasonal Variability in the Atmosphere-
Ocean Climate System, W. K.-M. Lau, D. E. Waliser, Eds. (Springer Praxis Books,
Springer, ed. 2, 2012), pp. 21–72.
7. S. Gadgil, S. Gadgil, The Indian monsoon, GDP and agriculture. Econ. Polit. Wkly. 41, 4887–
4895(2006).
8. R. S. Nanjundiah, P. A. Francis, M. Ved, S. Gadgil, Predicting the extremes of Indian summer
monsoon rainfall with coupled ocean–atmosphere models. Curr. Sci. 104, 1380–1393
(2013).
9. B. Wang, B. Xiang, J. Li, P. J. Webster, M. N. Rajeevan, J. Liu, K.-J. Ha, Rethinking Indian
monsoon rainfall prediction in the context of recent global warming. Nat. Commun. 6,
7154 (2015). doi:10.1038/ncomms8154 Medline
10. J. Li, B. Wang, How predictable is the anomaly pattern of the Indian summer rainfall? Clim.
Dyn. 46, 2847–2861 (2016). doi:10.1007/s00382-015-2735-6
11. E. M. Rasmusson, T. H. Carpenter, The relationship between eastern equatorial Pacific sea
surface temperatures and rainfall over India and Sri Lanka. Mon. Weather Rev. 111, 517–
528 (1983). doi:10.1175/1520-0493(1983)111<0517:TRBEEP>2.0.CO;2
12. P. J. Webster, V. O. Magaña, T. N. Palmer, J. Shukla, R. A. Tomas, M. Yanai, T. Yasunari,
Monsoons: Processes, predictability, and the prospects for prediction. J. Geophys. Res.
Oceans 103, 14451–14510 (1998). doi:10.1029/97JC02719
13. K. K. Kumar, B. Rajagopalan, M. Hoerling, G. Bates, M. A. Cane, Unraveling the mystery of
Indian monsoon failure during El Niño. Science 314, 115–119 (2006).
doi:10.1126/science.1131152 Medline
14. A. G. Turner, H. Annamalai, Climate change and the South Asian summer monsoon. Nat.
Clim. Chang. 2, 587–595 (2012). doi:10.1038/nclimate1495
15. F. Fan, X. Dong, X. Fang, F. Xue, F. Zheng, J. Zhu, Revisiting the relationship between the
South Asian summer monsoon drought and El Niño warming pattern. Atmos. Sci. Lett.
18, 175–182 (2017). doi:10.1002/asl.740
16. H. Varikoden, J. V. Revadekar, Y. Choudhary, B. Preethi, Droughts of Indian summer
monsoon associated with El Niño and Non-El Niño years. Int. J. Climatol. 35, 1916–1925
(2014). doi:10.1002/joc.4097
17. X. Li, M. Ting, Recent and future changes in the Asian monsoon-ENSO relationship: Natural
or forced? Geophys. Res. Lett. 42, 3502–3512 (2015). doi:10.1002/2015GL063557
18. T. Palmer, S. Zhaobo, A modelling and observational study of the relationship between sea
surface temperature in the North-West Atlantic and the atmospheric general circulation.
Q. J. R. Meteorol. Soc. 111, 947–975 (1985). doi:10.1002/qj.49711147003
19. C. Wang, Three-ocean interactions and climate variability: A review and perspective. Clim.
Dyn. 53, 5119–5136 (2019). doi:10.1007/s00382-019-04930-x
20. B. J. Hoskins, G.-Y. Yang, The equatorial response to higher-latitude forcing. J. Atmos. Sci.
57, 1197–1213 (2000). doi:10.1175/1520-0469(2000)057<1197:TERTHL>2.0.CO;2
21. G. Branstator, J. Teng, Tropospheric waveguide teleconnections and their seasonality. J.
Atmos. Sci. 74, 1513–1532 (2017). doi:10.1175/JAS-D-16-0305.1
22. S. Bordoni, T. Schneider, Monsoons as eddy-mediated regime transitions of the tropical
overturning circulation. Nat. Geosci. 1, 515–519 (2008). doi:10.1038/ngeo248
23. R. Krishnan, V. Kumar, M. Sugi, J. Yoshimura, Internal feedbacks from monsoon–
midlatitude interactions during droughts in the Indian summer monsoon. J. Atmos. Sci.
66, 553–578 (2009). doi:10.1175/2008JAS2723.1
24. R. K. Yadav, Role of equatorial central Pacific and northwest of North Atlantic 2-metre
surface temperatures in modulating Indian summer monsoon variability. Clim. Dyn. 32,
549–563 (2009). doi:10.1007/s00382-008-0410-x
25. S. Narsey, M. J. Reeder, D. Ackerley, C. Jakob, A midlatitude influence on Australian
monsoon bursts. J. Clim. 30, 5377–5393 (2017). doi:10.1175/JCLI-D-16-0686.1
26. N. Boers, B. Goswami, A. Rheinwalt, B. Bookhagen, B. Hoskins, J. Kurths, Complex
networks reveal global pattern of extreme-rainfall teleconnections. Nature 566, 373–377
(2019). doi:10.1038/s41586-018-0872-x Medline
27. Y.-K. Lim, The East Atlantic/West Russia (EA/WR) teleconnection in the North Atlantic:
Climate impact and relation to Rossby wave propagation. Clim. Dyn. 44, 3211–3222
(2015). doi:10.1007/s00382-014-2381-4
28. L. O’Brien, M. J. Reeder, Southern Hemisphere summertime Rossby waves and weather in
the Australian region. Q. J. R. Meteorol. Soc. 143, 2374–2388 (2017).
doi:10.1002/qj.3090
29. Y. Kushnir, I. M. Held, Equilibrium atmospheric response to North Atlantic SST anomalies.
J. Clim. 9, 1208–1220 (1996). doi:10.1175/1520-
0442(1996)009<1208:EARTNA>2.0.CO;2
30. Y. Kushnir, W. A. Robinson, I. Bladé, N. M. J. Hall, S. Peng, R. Sutton, Atmospheric GCM
response to extratropical SST anomalies: Synthesis and evaluation. J. Clim. 15, 2233–
2256 (2002). doi:10.1175/1520-0442(2002)015<2233:AGRTES>2.0.CO;2
31. B. N. Goswami, M. S. Madhusoodanan, C. P. Neema, D. Sengupta, A physical mechanism
for North Atlantic SST influence on the Indian summer monsoon. Geophys. Res. Lett. 33,
L02706 (2006). doi:10.1029/2005GL024803
32. R. Lu, B. Dong, H. Ding, Impact of the Atlantic Multidecadal Oscillation on the Asian
summer monsoon. Geophys. Res. Lett. 33, L24701 (2006). doi:10.1029/2006GL027655
33. M. Rajeevan, L. Sridhar, Inter-annual relationship between Atlantic sea surface temperature
anomalies and Indian summer monsoon. Geophys. Res. Lett. 35, L21704 (2008).
doi:10.1029/2008GL036025
34. C. Wang, F. Kucharski, R. Barimalala, A. Bracco, Teleconnections of the tropical Atlantic to
the tropical Indian and Pacific Oceans: A review of recent findings. Meteorol. Z. (Berl.)
18, 445–454 (2009). doi:10.1127/0941-2948/2009/0394
35. L. Krishnamurthy, V. Krishnamurthy, Teleconnections of Indian monsoon rainfall with AMO
and Atlantic tripole. Clim. Dyn. 46, 2269–2285 (2016). doi:10.1007/s00382-015-2701-3
36. C. T. Sabeerali, R. S. Ajayamohan, H. K. Bangalath, N. Chen, Atlantic Zonal Mode: An
emerging source of Indian summer monsoon variability in a warming world. Geophys.
Res. Lett. 46, 4460–4467 (2019). doi:10.1029/2019GL082379
37. C. Stan, D. M. Straus, J. S. Frederiksen, H. Lin, E. D. Maloney, C. Schumacher, Review of
tropical-extratropical teleconnections on intraseasonal time scales. Rev. Geophys. 55,
902–937 (2017). doi:10.1002/2016RG000538
38. S. Ross, Introduction to Probability and Statistics for Scientists and Engineers (Elsevier
Academic Press, ed. 4, 2009).
39. P. Poli, H. Hersbach, D. P. Dee, P. Berrisford, A. J. Simmons, F. Vitart, P. Laloyaux, D. G.
H. Tan, C. Peubey, J.-N. Thépaut, Y. Trémolet, E. V. Hólm, M. Bonavita, L. Isaksen, M.
Fisher, ERA-20C: An atmospheric reanalysis of the twentieth century. J. Clim. 29, 4083–
4097 (2016). doi:10.1175/JCLI-D-15-0556.1
40. B. Parthasarathy, A. A. Munot, D. R. Kothawale, “Monthly and seasonal rainfall time series
for all India, homogeneous divisions and meteorological subdivisions, 1871-1994,” Tech.
Rep. RR065, Indian Institute of Tropical Meteorology, Pune, India (1996).
41. N. A. Rayner, D. E. Parker, E. B. Horton, C. K. Folland, L. V. Alexander, D. P. Rowell, E. C.
Kent, A. Kaplan, Global analyses of sea surface temperature, sea ice, and night marine air
temperature since the late nineteenth century. J. Geophys. Res. 108, 4407 (2003).
doi:10.1029/2002JD002670
42. A. G. Barnston, R. E. Livezey, Classification, seasonality and persistence of low-frequency
atmospheric circulation patterns. Mon. Weather Rev. 115, 1083–1126 (1987).
doi:10.1175/1520-0493(1987)115<1083:CSAPOL>2.0.CO;2
43. M. J. Rodwell, C. K. Folland, Atlantic air-sea interaction and model validation. Ann.
Geophys. 46, 47–56 (2003).
