1
2
Severe Precipitation in Northern India in June 2013: Causes, Historical
Context, and Changes in Probability
3
4
5
6
7
8
Deepti Singh, Daniel E. Horton, Michael Tsiang, Matz Haugen, Moetasim Ashfaq, Rui
Mei, Deeksha Rastogi, Nathaniel C. Johnson, Allison Charland, Bala Rajaratnam and
Noah S . Diffenbaugh*
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Affiliations: Singh, Charland, Diffenbaugh, Horton - Department of
Environmental Earth System Science and Woods Institute for the Environment,
Stanford University, California, USA; Tsiang, Haugen, Rajaratnam - Department
of Environmental Earth System Science, Department of Statistics, and Woods
Institute for the Environment, Stanford University, California, USA; Ashfaq, Mei,
Rastogi – Climate Change Science Institute, Oak Ridge National Laboratory, Oak
Ridge, Tennessee, USA; Johnson - International Pacific Research Center,
University of Hawaii at Manoa, Honolulu, Hawaii and Scripps Institution of
Oceanography, University of California-San Diego, La Jolla, California.
23
24
25
26
27
28
29
corresponding author: Stephanie C. Herring, Stephanie.Herring@NOAA.govSummary
30
Statement
1
1
31
We find that June 2013 precipitation in northern India was a century-scale event
32
in the observational period, with evidence of increased likelihood in the present climate
33
compared to the pre-industrial.
34
35
The Event: June 2013 Flooding in Northern India
36
Parts of mountainous northern India – including Himachal Pradesh, Uttarakhand,
37
and Uttar Pradesh – experienced extremely heavy precipitation during 14-17 June 2013
38
(Fig. 1a,b). Landslides, debris flows, and extensive flooding caused catastrophic damage
39
to housing and infrastructure, impacted >100,000 people, and resulted in >5,800 deaths
40
(Dobhal et al. 2013; Dube et al. 2013; Dubey et al. 2013; Joseph et al. 2013; Mishra;
41
Srinivasan 2013). Subsequent heavy rains on 24-25 June hampered rescue efforts,
42
ultimately leaving thousands without food or shelter for >10 days (Prakash 2013).
43
Causes of the mid-June precipitation and associated flooding have been analyzed
44
in detail (Dobhal et al. 2013; Dube et al. 2014; Mishra; Srinivasan 2013; Prakash 2013).
45
Anomalously early arrival of monsoon-like atmospheric circulation over India (Fig. 1c,d)
46
brought heavy rains to the mountainous regions where snow-cover typically melts prior
47
to monsoon onset (Dube et al. 2014; Joseph et al. 2013). Snow-cover in local river basins
48
was ~30% above normal in early-June 2013 (Durga Rao et al. 2014). Heavy precipitation
49
led to rapid snowmelt, overwhelming the regional hydrologic system, causing glacial lake
50
outburst floods, and triggering catastrophic mass wastage events (Andermann et al. 2012;
51
Dubey et al. 2013; Durga Rao et al. 2014; Prakash 2013; Siderius et al. 2013).
2
2
52
The upper- and lower-level synoptic conditions in early- and mid-June supported
53
the anomalously early monsoon-like circulation (Fig.S1) and excessive precipitation in
54
northern India (Fig. 1a,b). In the upper atmosphere (200 mb), a persistent anticyclonic
55
anomaly formed over Central Asia (Fig. 1e). This upper-level blocking pattern guided
56
mid-to-high-latitude troughs southward, thereby facilitating the advection of relatively
57
cold, dry, high-potential-vorticity air to the upper levels of the atmosphere over northern
58
India (Joseph et al. 2013). In the lower atmosphere (850 mb), low-pressure systems
59
formed over both the northern Bay of Bengal and the northern Arabian Sea (Joseph et al.
60
2013), with the Bay of Bengal system moving inland over central India for the duration
61
of the event (Fig. 1h). Low-level convergence associated with these systems and a
62
stronger-than-normal Somali Jet facilitated anomalous moisture advection to the Indian
63
subcontinent (Fig. 1c). These co-occurring upper- and lower-level dynamics are
64
consistent with a convectively unstable atmosphere (Hong et al. 2011; Ullah; Shouting
65
2013; Wang et al. 2011), which, when combined with orographic forcing from the
66
surrounding northwestern Himalayan terrain, create an environment ripe for intense
67
mesoscale convection (Houze et al. 2011).
68
69
In this study, we analyze the dynamics of this event within the context of the
historical and pre-industrial climates.
70
71
Historical context
3
3
72
We contextualize June-2013 precipitation using the Indian Meteorological
73
Department (IMD) 1951-2013 1×1 gridded dataset, with the caveat that the rain-gauge
74
network in the region could have changed over this period (Rajeevan et al. 2010).
75
Cumulative June precipitation exceeded the 80th percentile over much of central and
76
northern India, and the maximum quantile over a majority of the flood region (Fig. 1a).
77
This domain (29-33N, 77.5-80E) received unprecedented 4-day total precipitation from
78
14-17 June, with the heaviest day (16 June) exceeding the previous one-day June
79
maximum by 105% (Fig S2). As a consequence, the flood region recorded the highest
80
total accumulated June precipitation in the 1951-2013 record, with the previous
81
maximum June total equaled by the 17th of June, and exceeded by 31% by the end of the
82
month (Fig. 1b).
83
Monsoon dynamics and thermodynamics were also unusual relative to June
84
climatological norms. The monsoon onset date is closely associated with the reversal of
85
the zonally averaged (52-85E) meridional tropospheric (500-200mb) ocean-to-continent
86
(5˚N-30˚N) temperature gradient (Ashfaq et al. 2009; Webster et al. 1998), and with the
87
vertical easterly zonal wind shear between 850mb and 200mb averaged over 0-30N and
88
50-90E (Li; Yanai 1996; Webster et al. 1998; Wu et al. 2012; Xavier et al. 2007). The
89
2013 MTG reversal dates were among the earliest on record (Figs. 1d). The early MTG
90
reversal resulted from anomalously high land temperatures (~2) (Fig. S1c,d), which co-
91
occurred with record-low Eurasian snow-cover (NOAA 2013). In addition, as a result of
92
the early monsoon-like circulation, low-level atmospheric humidity exceeded 2 above
93
the climatological 14-17 June mean (Fig. 1c).
4
4
94
Synoptic conditions were likewise extremely rare for mid-June. We categorize the
95
occurrence of upper- and lower-level daily June atmospheric patterns in the NCEP R1
96
reanalysis using self-organizing map (SOM) cluster analysis (Borah et al. 2013;
97
Chattopadhyay et al. 2008; Hewitson; Crane 2002; Johnson 2013; Kohonen 2001) (see
98
Supplemental Methods). SOM analyses reveals persistent upper-level blocking patterns
99
from 10-17 June and lower-level troughing patterns from 11-17 June (Fig. S2).
100
Additionally, the upper- and lower-level patterns (Fig. 1f,i) that persisted during the core
101
of the event (14-17 June) are each historically associated with heavy precipitation over
102
northern India (Fig. 1g,j). Although occurrence of the core-event upper-level pattern is
103
not rare for June (median frequency of occurrence), the 850mb pattern is much less
104
common (<6 percentile frequency of occurrence). Further, mid-June 2013 was the only
105
instance that the core-event upper- and lower-level patterns co-occurred in June during
106
the 1951-2013 period. The atmospheric configuration associated with the unprecedented
107
mid-June extreme precipitation therefore appears to also have been unprecedented.
108
(We note that this configuration is not necessarily unprecedented later in the
109
monsoon season. For example, the co-occurrence of upper-level blocking with tropical
110
moisture advection is similar to the conditions identified during the July 2010 Pakistan
111
floods, and during heavy precipitation events that occur during the core monsoon season
112
(Hong et al. 2011; Houze et al. 2011; Lau; Kim 2011; Ullah; Shouting 2013; Webster et
113
al. 2011).)
114
5
5
115
Quantifying the likelihood of a 2013-magnitude event
116
We first quantify the likelihood of the June-2013 total precipitation in the
117
observed climate. Given the rarity of the event in the observed record (Fig. 2a), we fit a
118
Pareto (heavy-tailed) distribution to the 1951-2012 observations (Fig. 2a,S3a). From the
119
Pareto distribution, we estimate the sample-quantile (Qo) and return period (Ro) of the
120
June 2013 total precipitation in the present climate (see Supplemental Methods). We find
121
that the 2013 event exceeds the 99th percentile in the observed distribution (Qo= 99.1th
122
quantile), yielding a return period of 111 years (Fig. 2a). Because the Pareto is a heavy-
123
tailed distribution, extreme events are less likely to be found anomalous, and thus the
124
corresponding return period can be considered a lower bound.
125
Next, we assess the influence of anthropogenic forcings on the likelihood of
126
extreme June precipitation using the historical (“20C”) and preindustrial (“PI”)
127
simulations from the CMIP5 climate model archive (Taylor et al. 2012). We use the
128
Kolmogorov-Smirnov (“K-S”) goodness-of-fit test to identify the models that most
129
closely simulate the observed distribution of June total precipitation over the impacted
130
region (Fig. 1a). (To control for the mean-bias in the models, we first re-center each
131
model’s distribution so that the model mean matches the observed mean.) Because the
132
simulated change in likelihood of extremes can be heavily influenced by biases in the
133
simulated distribution, we restrict our analyses to 11 models whose K-S value exceeds
134
0.2 (Fig. S3), ensuring a comparatively good fit of the overall distribution, including in
135
the tails. Among these 11 models that pass this goodness-of-fit criterion, 4 show greater
136
mean and variability of June precipitation in the 20C simulations (Fig. 2b). However, 7 of
6
6
137
the 11 show increased exceedance of the PI 99th percentile value (Fig. 2c), suggesting
138
increased probability of extremely high June precipitation in the current climate. This
139
result is consistent
140
primarily from increased atmospheric-moisture availability
141
Soden 2008; O'Gorman; Schneider 2009).
with studies that indicate an increase in extremes
(Allan;
142
Next, we use Pareto distributions to estimate the return period of the June-2013
143
total precipitation in the 20C and PI simulations. To control for the variability-bias in the
144
models, we first determine the magnitude of the 111-year event (Qo= 99.1th quantile) in
145
the fitted 20C distribution (PrH), and then determine the quantile (QPI) corresponding to
146
PrH in the fitted PI distribution (see Supplemental Methods; Fig. S3). Further, we
147
quantify the uncertainty in these likelihood estimates (Qo/QPI) using the bootstrap (Fig.
148
2d). We find that 5 of the 11 models show >50% likelihood that the extreme June total
149
precipitation has higher probability in the 20C climate. In addition, of the three models
150
that have high K-S > 0.8 and similar sample sizes in the 20C and PI populations (Fig. 2d),
151
two suggest >50% likelihood that the extreme June total precipitation has higher
152
probability in the 20C climate, and the third model suggests ~50% likelihood. Further,
153
the model with the largest 20C ensemble (CNRM-CM5) demonstrates a ~50% likelihood
154
that the probability of the extreme June total precipitation has at least doubled in the 20C
155
climate. CNRM-CM5 also has the highest skill in simulating the summer monsoon
156
precipitation and lower-level wind climatology (Sperber et al., 2013).
157
7
7
158
Conclusions
159
Our statistical analysis, combined with our diagnosis of the atmospheric
160
environment, demonstrates that the extreme June-2013 total precipitation in northern
161
India was at least a century-scale event. Precise quantification of the likelihood of the
162
event in the current and pre-industrial climates is limited by the relatively short
163
observational record, and by the resolution and ensemble size of the small subset of
164
models that credibly simulate the seasonal rainfall distribution over northern India.
165
Indeed, an attempt to quantify the probability of the unprecedented 4-day precipitation
166
total would present even greater analytical challenges. However, despite these limitations,
167
our analyses of the observed and simulated June precipitation provide evidence that
168
anthropogenic forcing of the climate system has increased the likelihood of such an event,
169
a result in agreement with previous studies of trends in rainfall extremes in India
170
(Goswami et al. 2006, Krishnamurthy et al. 2009, Ghosh et al. 2012, Singh et al. 2014).
171
172
8
8
173
174
References
175
176
177
Allan, R. P., and B. J. Soden, 2008: Atmospheric Warming and the Amplification of
Precipitation Extremes. Science, 321, 1484.
178
179
180
Andermann, C., L. Longuevergne, S. Bonnet, A. Crave, P. Davy, and R. Gloaguen, 2012:
Impact of transient groundwater storage on the discharge of Himalayan rivers.
Nature Geosci., 5, 127-132.
181
182
183
Ashfaq, M., Y. Shi, W.-W. Tung, R. J. Trapp, X. Gao, J. S. Pal, and N. S. Diffenbaugh,
2009: Suppression of south Asian summer monsoon precipitation in the 21st
century. Geophys. Res. Lett., 36, L01704.
184
185
186
187
Borah, N., A. K. Sahai, R. Chattopadhyay, S. Joseph, S. Abhilash, and B. N. Goswami,
2013: A self-organizing map–based ensemble forecast system for extended range
prediction of active/break cycles of Indian summer monsoon. Journal of
Geophysical Research: Atmospheres, 118, 9022-9034.
188
189
190
Chattopadhyay, R., A. K. Sahai, and B. N. Goswami, 2008: Objective Identification of
Nonlinear Convectively Coupled Phases of Monsoon Intraseasonal Oscillation:
Implications for Prediction. Journal of the Atmospheric Sciences, 65, 1549-1569.
191
192
Dobhal, D. P., A. K. Gupta, M. Mehta, and D. D. Khandelwal, 2013: Kedarnath disaster:
facts and plausible causes. Current Science, 105.
193
194
195
196
197
Dube, A., R. Ashrit, A. Ashish, K. Sharma, G. R. Iyengar, E. N. Rajagopal, and B. S.;,
2013: Performance of NCMRWF Forecast Models in Predicting The Uttarakhand
Heavy Rainfall Event During 17-18 June 2013. [India] National Centre for Medium
Range Weather Forecasting Research Report NMRF/RR/08/2013, 35 pp. . N. C. f.
M. R. W. Forecasting, Ed.
198
199
200
Dube, A., R. Ashrit, A. Ashish, K. Sharma, G. R. Iyengar, E. N. Rajagopal, and Basu S.,
2014: Forecasting the heavy rainfall during Himalayan flooding - June 2013.
Weather and Climate Extremes.
201
202
Dubey, C. S., D. P. Shukla, A. S. Ningreichon, and A. L. Usham, 2013: Orographic
Control of the Kedarnath disaster. Current Science, 105, 1474-1476.
203
204
205
Durga Rao, K. H. V., V. Venkateshwar Rao, V. K. Dadhwal, and P. G. Diwakar, 2014:
Kedarnath flash floods: a hydrological and hydraulic simulation study, Current
Science, 106.
9
9
206
207
Goswami, B. N., V. Venugopal, et al. (2006). "Increasing Trend of Extreme Rain Events
Over India in a Warming Environment." Science 314(5804): 1442-1445.
208
209
Ghosh, S., D. Das, et al. (2012). "Lack of uniform trends but increasing spatial variability
in observed Indian rainfall extremes." Nature Clim. Change 2(2): 86-91.
210
211
Hewitson, B. C., and R. G. Crane, 2002: Self-organizing maps: Applications to synoptic
climatology. Climate Research, 22, 13-26.
212
213
214
Hong, C.-C., H.-H. Hsu, N.-H. Lin, and H. Chiu, 2011: Roles of European blocking and
tropical-extratropical interaction in the 2010 Pakistan flooding. Geophysical
Research Letters, 38, L13806.
215
216
217
Houze, R. A., K. L. Rasmussen, S. Medina, S. R. Brodzik, and U. Romatschke, 2011:
Anomalous atmospheric events Leading to the summer 2010 floods in Pakistan.
Bulletin of the American Meteorological Society, 92, 291-298.
218
219
Johnson, N. C., 2013: How Many ENSO Flavors Can We Distinguish?*. Journal of
Climate, 26, 4816-4827.
220
221
Joseph, S., and Coauthors, 2013: Extended Range Prediction of Uttarakhand Heavy
Rainfall Event by an Ensemble Prediction System based on CFSv2. .
222
Kohonen, T., 2001: Self-Organizing Maps. 3rd ed. Vol. 30, Springer
223
224
225
Krishnamurthy, C. K. B., U. Lall, et al. (2009). "Changing Frequency and Intensity of
Rainfall Extremes over India from 1951 to 2003." Journal of Climate 22(18): 47374746.
226
227
228
Lau, W. K. M., and K.-M. Kim, 2011: The 2010 Pakistan Flood and Russian Heat Wave:
Teleconnection of Hydrometeorological Extremes. Journal of Hydrometeorology,
13, American Meteorological Society--403.
229
230
231
Li, C., and M. Yanai, 1996: The Onset and Interannual Variability of the Asian Summer
Monsoon in Relation to Land–Sea Thermal Contrast. Journal of Climate, 9, 358375.
232
233
Mishra, A., and J. Srinivasan, 2013: Did a cloud burst occur in Kedarnath during 16 and
17 June 2013. Current Science, 105.
234
NOAA, 2013: State of the Climate: Global Analysis for May 2013.
235
236
237
O'Gorman, P. A., and T. Schneider, 2009: The physical basis for increases in
precipitation extremes in simulations of 21st-century climate change. Proc Natl
Acad Sci U S A, 106, 7.
10
10
238
239
Prakash, S., 2013: Brief Report on visit to Alaknanda Valley, Uttarakhand Himalaya
during 22-24 June 2013.
240
241
Rajeevan, M., S. Gadgil, and J. Bhate, 2010: Active and break spells of the Indian
summer monsoon. Journal of Earth System Science, 119, 229-247.
242
243
Siderius, C., and Coauthors, 2013: Snowmelt contributions to discharge of the Ganges.
Science of The Total Environment, 468–469, Supplement, S93-S101.
244
245
246
Singh, D., M. Tsiang, et al. (2014). "Observed changes in extreme wet and dry spells
during the South Asian summer monsoon season." Nature Clim. Change 4(6): 456461.
247
248
249
Sperber, K. R., H. Annamalai, et al. (2013). "The Asian summer monsoon: an
intercomparison of CMIP5 vs. CMIP3 simulations of the late 20th century."
Climate Dynamics 41(9-10): 2711-2744.
250
251
Taylor, K. E., R. J. Stouffer, and G. A. Meehl, 2012: An Overview of CMIP5 and the
Experiment Design. Bulletin of the American Meteorological Society, 93, 485-498.
252
253
254
Ullah, K., and G. Shouting, 2013: A diagnostic study of convective environment leading
to heavy rainfall during the summer monsoon 2010 over Pakistan. Atmospheric
Research, 120–121, 226-239.
255
256
257
Wang, S.-Y., R. E. Davies, W.-R. Huang, and R. R. Gillies, 2011: Pakistan's two-stage
monsoon and links with the recent climate change. Journal of Geophysical
Research: Atmospheres, 116, D16114.
258
259
Webster, P. J., V. E. Toma, and H. M. Kim, 2011: Were the 2010 Pakistan floods
predictable? Geophysical Research Letters, 38, L04806.
260
261
262
Webster, P. J., V. O. Magana, T. N. Palmer, J. Shukla, R. A. Tomas, M. Yanai, and T.
Yasunari, 1998: Monsoons: Processes, predictability, and the prospects for
prediction. J. Geophys. Res., 103, 14451-14510.
263
264
Wu, G., Y. Liu, B. He, Q. Bao, A. Duan, and F.-F. Jin, 2012: Thermal Controls on the
Asian Summer Monsoon. Sci. Rep., 2.
265
266
267
Xavier, P. K., C. Marzin, and B. N. Goswami, 2007: An objective definition of the Indian
summer monsoon season and a new perspective on the ENSO–monsoon
relationship. Quarterly Journal of the Royal Meteorological Society, 133, 749-764.
268
269
270
11
11
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
Figure 1 | Precipitation characteristics and synoptic environment. (a) June-2013 gridcell cumulative precipitation percentiles relative to June climatology (1951-2012). White
box highlights the severe flooding domain (29-33˚N, 77.5-80˚E). (b) Daily cumulative
precipitation distribution over the flood domain. (c) 14-17 June 2013 composite lowerlevel wind and specific humidity anomalies relative to 14-17 June climatology. (d)
Climatological and 2013 meridional temperature gradient (MTG), defined as the zonallyaveraged (52-85˚E) pentad mean tropospheric (200-500mb) temperature difference
between 30˚N and 5˚N. (e,f) 14-17 June 2013 composite upper- and lower-level wind and
geopotential height anomalies relative to the 14-17 June climatology. (g,h) Upper- and
lower-atmosphere Self-Organizing Map (SOM) patterns that correspond to 14-17 June
2013. Pattern matches are autonomously selected from 35 SOM nodes, generated from an
analysis of all 1951-2013 June days. (i,j) Composite precipitation for all June days during
the 1951-2013 period that were associated with the upper- and lower-level SOM patterns
shown in (g) and (h).
288
289
290
12
12
291
292
293
294
295
296
297
298
299
300
301
302
303
304
Figure 2 | Extreme precipitation statistics in the current and pre-industrial climates.
(a) Probability density function of the Pareto fitted observed cumulative-June
precipitation distribution (black line; 1951-2012), and probability of occurrence of the
June-2013 cumulative precipitation magnitude in this distribution (red). The return period
of the June-2013 magnitude in the observed distribution is indicated on the plot. (b)
Change in mean and standard deviation of precipitation between the CMIP5 historical
("20C") and pre-industrial ("PI") simulations. Grey dots represent all available CMIP5
models and colored symbols represent “A1” models that meet the Kolmogorov-Smirnov
(“K-S”) goodness-of-fit test criteria (p-value > 0.2). (c) Percent of years in the 20C
simulations of A1 models that exceed the respective PI quantiles of the A1 models. The
numbers on the plot indicate the fraction of A1 models that exceed the PI quantiles in the
13
13
305
306
307
308
309
310
311
312
20C simulations. (d) Box-plot representing the distribution of ratios of the return period
of June-2013 magnitude event in the PI and 20C simulations, calculated using the
bootstrap. The lines in the boxes represent the median of the distribution for each model;
the bounds of the boxes represent the 25th and 75th percentiles; the whiskers extend to
the edges of 1.5*interquartile range; and points outside of those bounds are shown
individually. The number of years indicated for the “20C Yrs” and “PI Yrs” columns are
the total years available from all realizations within each scenario. The color bar
corresponding to the box-plot indicates p-values from the Kolmogorov-Smirnov test.
313
314
14
14