See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/275340859

Trend analysis of rainfall and temperature data for India SK Jain, V Kumar
Current Science(Bangalore) 102 (1), 37-49

Article in Current science · January 2012

CITATIONS READS

236 5,018

2 authors, including:

Sharad K. Jain
National Institute of Hydrology
159 PUBLICATIONS 3,261 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Observation and Modeling of varous hydrological processes in a Small Watershed in Upper Ganga Basin View project

Impact of climate and landuse change on hydrological response (surface and sub-subsurface) of Beas basin (up to Pandoh dam) (Researcher project - JOINTINDNOR)
View project

All content following this page was uploaded by Sharad K. Jain on 06 January 2016.

The user has requested enhancement of the downloaded file.

 REVIEW ARTICLE

Trend analysis of rainfall and temperature

data for India
Sharad K. Jain1,* and Vijay Kumar2
1
Department of Water Resources Development and Management, Indian Institute of Technology, Roorkee 247 667, India
2
National Institute of Hydrology, Roorkee 247 667, India (Now at Ministry of Earth Sciences, New Delhi)

water, along with the danger of increasing occurrences of

This article aims to review studies pertaining to trends
in rainfall, rainy days and temperature over India. droughts and floods. The southwest (SW) monsoon,
Sen’s non-parametric estimator of slope has been which brings about 80% of the total precipitation over the
frequently used to estimate the magnitude of trend, country, is critical for the availability of freshwater for
whose statistical significance was assessed by the drinking and irrigation. Changes in climate over the
Mann–Kendall test. Spatial units for trend analysis Indian region, particularly the SW monsoon, would have
vary from station data to sub-division to sub-basin/ a significant impact on agricultural production, water
river basins. There are differences in the results of the resources management and overall economy of the country.
various studies, and a clear and consistent picture of The heavy concentration of rainfall in the monsoon
rainfall trend has not emerged. Although the different months (June–September) results in scarcity of water in
units (sub-basins or sub-divisions) may have a non- many parts of the country during the non-monsoon periods.
zero slope value, few values are statistically signifi-
In view of the above, a number of studies have attemp-
cant. In a study on basin-wise trend analysis, 15 basins
had decreasing trend in annual rainfall; only one basin ted to investigate the trend of climatic variables for
showed significant decreasing trend at 95% confidence the country. These studies have looked at the trends on the
level. Among six basins showing increasing trend, one country scale, regional scales and at the individual sta-
basin showed significant positive trend. Most of the tions. This article gives an exhaustive coverage of the
basins had the same direction of trend in rainfall and reported studies dealing with two variables which are
rainy days at the annual and seasonal scale. critical in hydrologic studies: rainfall and temperature.
Regarding trends in temperature, the mean maxi- Temperature and its changes impact a number of hydrologi-
mum temperature series showed a rising trend at most cal processes including rainfall and these processes, in turn,
of the stations; it showed a falling trend at some sta- also impact temperature (e.g., cooling due to rain/snow).
tions. The mean minimum temperature showed a ris- Due to the uneven distribution of rainfall and the mis-
ing as well as a falling trend. At most of the stations in
match between water availability and demand, large stor-
the south, central and western parts of India a rising
trend was found. Some stations located in the north age reservoirs are required to redistribute the natural flow
and northeastern India showed a falling trend in in accordance with the requirements of specific regions.
annual mean temperature. Most of the data used The design of hydro-infrastructure is generally based on
in trend analysis pertained to the stations located in the assumption that climate is stationary1. Changes in
urban areas and these areas are sort of heat islands. rainfall due to global warming will influence the hydro-
This article also highlights the need of a network of logical cycle and the pattern of stream-flows and de-
baseline stations for climatic studies. mands (particularly agricultural), requiring a review of
hydrologic design and management practices. Changes in
Keywords: Climate change, rainfall, trend, river basin, run-off and its distribution will depend on likely future
seasonal analysis, temperature data. climate scenarios2. The trend analysis of rainfall, tem-
perature and other climatic variables on different spatial
AGRICULTURE and related sectors, food security, and scales will help in the construction of future climate sce-
energy security of India are crucially dependent on the narios. Using these water availability in different basins
timely availability of adequate amount of water and a can be assessed in the context of future requirements.
conducive climate. The rainfall received in an area is an
important factor in determining the amount of water River basins of India and their key features
available to meet various demands, such as agricultural,
industrial, domestic water supply and for hydroelectric In India, river basins having catchment areas exceeding
power generation. Global climate changes may influence 20,000 sq. km are classified as major basins. The country
long-term rainfall patterns impacting the availability of has 12 such basins whose total catchment area
is 2.528 million sq. km. Among the major rivers, the
*For correspondence. (e-mail: [email protected]) Ganga–Brahmaputra–Meghna (Barak) system is the

CURRENT SCIENCE, VOL. 102, NO. 1, 10 JANUARY 2012 37

REVIEW ARTICLE

largest. Three major basins, viz. Indus, Ganga and Bra- Least square lines are also drawn in Figure 2 to illustrate
hmaputra are snow-fed, the remaining are purely rainfall- a possible linear trend in the rainfall and rainy-day data.
dependent. Seven other medium-sized basins of the As seen from Figure 2, rainfall in the Brahmaputra and
Sabarmati, Mahi, Narmada and Tapi rivers flowing west Godavari basins shows a decreasing trend, whereas in the
and the Subarnarekha, Brahmani, Baitarani and Cauvery Ganga and Krishna basins, no trend is seen. Similarly,
rivers flowing east together cover 15% of the total drainage rainy days indicate decreasing trend in the Brahmaputra
area of India. The catchment area of medium rivers is and Godavari basins and increasing trend in three other
about 0.25 million sq. km; Subarnarekha with catchment river basins. Figure 2 also shows varying amounts of
area of 19,000 sq. km is the largest among the medium trend in rainfall and rainy days over different river basins.
rivers in the country. Major river basins of India are
shown in Figure 1 and key features of the river basins are
Trend analysis – general methodology
provided in Table 1.
An understanding of rainfall variability for the country
Trend analysis of a time series consists of the magnitude
is necessary to appreciate the impacts of climate change;
of trend and its statistical significance. Obviously, differ-
it is also important for water management. The key rain-
ent workers have used different methodologies for trend
fall statistics of different river basins is given in Table 1.
detection. Kundzewicz3 has discussed the change detec-
The variability of rainfall and rainy days during winter,
tion methodologies for hydrologic data.
pre-monsoon and post-monsoon seasons is high. WFR2
In general, the magnitude of trend in a time series is
basin receives the maximum mean annual rainfall,
determined either using regression analysis (parametric
whereas Luni+ and Sabarmati basin receive the minimum
test) or using Sen’s estimator method (non-parametric
mean annual rainfall. Regarding seasonal patterns, the
method)4. Both these methods assume a linear trend in
maximum rainfall in the pre-monsoon season is received
the time series. Regression analysis is conducted with
over Brahmaputra and Barak+ basins; in the monsoon
time as the independent variable and rainfall/temperature
season over WFR1; in the post-monsoon season over
as the dependent variable. The regression analysis can be
WFR2 and in winter season over Indus basin. There are
carried out directly on the time series or on the anomalies
large temporal variations in rainfall; expectedly, the
(i.e. deviation from mean). A linear equation, y = mt + c,
coefficient of variation is the largest in arid regions and
defined by c (the intercept) and trend m (the slope), can
the smallest for rivers with high water yield.
be fitted by regression. The linear trend value represented
Figure 2 shows the temporal variation of annual rain-
by the slope of the simple least-square regression line
fall and annual rainy days for a few major river basins.
provided the rate of rise/fall in the variable.
Sen’s estimator has been widely used for determining
the magnitude of trend in hydro-meteorological time
series5–7. In this method, the slopes (Ti) of all data pairs
are first calculated by

x j − xk
Ti = for i = 1, 2,…, N, (1)
j−k

where xj and xk are data values at time j and k (j > k)

respectively. The median of these N values of Ti is Sen’s
estimator of slope which is calculated as

⎧TN +1 N is odd,
⎪ 2
β = ⎨1
( )
(2)
⎪ TN + TN + 2 N is even.
⎩2 2 2

A positive value of β indicates an upward (increasing)

trend and a negative value indicates a downward
(decreasing) trend in the time series.

Significance of trend

Figure 1. River basin map of India. EFR, East flowing rivers; WFR, To ascertain the presence of statistically significant trend
West flowing rivers. in hydrologic climatic variables such as temperature and

38 CURRENT SCIENCE, VOL. 102, NO. 1, 10 JANUARY 2012

 REVIEW ARTICLE

CURRENT SCIENCE, VOL. 102, NO. 1, 10 JANUARY 2012 39

REVIEW ARTICLE

Figure 2. Temporal variation of annual rainfall and rainy days for four major river basins of India. Filled circles represent rainfall and empty cir-
cles represent rainy days. Thick and thin black lines are the linear trend lines in rainfall and rainy days respectively. Note that the y-axis scale for
Brahmaputra River has a different scale.

precipitation with reference to climate change, non-para- n

metric Mann–Kendall (MK) test has been employed8–13. N ( N − 1)(2 N + 5) − ∑ t (t
k =1
k k − 1)(2tk + 5)
The MK test checks the null hypothesis of no trend var( S ) = , (6)
versus the alternative hypothesis of the existence of 18
increasing or decreasing trend.
The statistics (S) is defined as14 where n is the number of tied (zero difference between
compared values) groups and tk the number of data points
N −1 N in the kth tied group. The standard normal deviate
S= ∑ ∑ sgn( x
i =1 j =i +1
j − xi ), (3) (Z-statistics) is then computed as16

⎧ S −1
⎪ if S > 0
where N is the number of data points. Assuming ⎪ var( S )
(xj – xi) = θ, the value of sgn(θ) is computed as follows: ⎪
Z =⎨ 0 if S = 0 (7)
⎪ S +1
⎪ if S < 0.
⎧ 1 if θ > 0 ⎪⎩ var( S )
⎪
sgn(θ ) = ⎨ 0 if θ > 0 (4)
⎪−1 if θ > 0,
⎩ If the computed value of |Z| > zα/2, the null hypothesis
(H0) is rejected at α level of significance in a two-sided
This statistics represents the number of positive differ- test. In this analysis, the null hypothesis was tested at
ences minus the number of negative differences for all 95% confidence level.
the differences considered. For large samples (N > 10),
the test is conducted using a normal distribution15, with Studies of rainfall trends in India
the mean and the variance as follows:
Many studies have attempted to determine the trend in
E[S] = 0, (5) rainfall on both country and regional scales. Most of

40 CURRENT SCIENCE, VOL. 102, NO. 1, 10 JANUARY 2012

 REVIEW ARTICLE

these deal with the analysis of annual and seasonal series monsoon and post-monsoon months25. Pattanaik35 found
of rainfall for some individual stations or groups of sta- decreasing trend in monsoon rainfall over northwest and
tions. Some past studies related to changes in rainfall central India during 1941–2002.
over India have concluded that there is no clear trend in Trend analysis of rainfall data of 135 years (1871–
average annual rainfall over the country17–22. Though the 2005) indicated no significant trend for annual, seasonal
monsoon rainfall in India exhibited no significant trend and monthly rainfall on an all-India basis22. Annual and
over a long period of time, particularly in the all-India monsoon rainfall decreased, and pre-monsoon, post-
scale, pockets of significant long-term rainfall changes monsoon and winter rainfall increased over the years,
were identified in some studies22–26. with maximum increase in the pre-monsoon season.
Parthasarathy and Dhar27 found that the annual rainfall Monsoon months of June, July and September witnessed
for the period 1901–1960 had a positive trend over Cen- decreasing rainfall, whereas August showed increasing
tral India and the adjoining parts of the peninsula, and a trend on an all-India basis. Further analysis on sub-
decreasing trend over some parts of eastern India. A sta- divisional basis (30 sub-divisions) indicated that half of
ble northeast monsoon rainfall was found over Tamil them have increasing trend in annual rainfall, but for only
Nadu by Dhar et al.28. Pant and Hingane29 found increasing three sub-divisions, namely Haryana, Punjab and coastal
trend in mean annual and SW monsoon rainfall over Karnataka, this trend was statistically significant. Only
meteorological sub-divisions of Punjab, Haryana, west the Chhattisgarh sub-division had significant decreasing
Rajasthan, east Rajasthan and west Madhya Pradesh during trend out of the 15 sub-divisions showing decreasing
the period 1901–1982. Rupa Kumar et al.30 reported that trend in annual rainfall. All the five regions showed non-
northeast peninsula, northeast India and northwest penin- significant trend in annual, seasonal and monthly rainfall
sula experienced a decreasing trend (ranged between –6% for most of the months.
and –8% of the normal per 100 years) in the monsoon Analysis of rainfall data at five stations (Srinagar, Kul-
rainfall. But the west coast, central peninsula and north- gam, Handwara, Qazigund and Kukarnag) in the Kashmir
west India experienced increasing trend (about 10–12% Valley for the period 1903–1982 showed that the first
of the normal per 100 years) in monsoon rainfall. three stations experienced a decreasing trend in annual
Trends in monsoon rainfall in different sub-divisions rainfall; the largest decrease was for Kulgam (–20.16% of
and the whole of India were examined for the period mean/100 yrs) and the smallest for Srinagar26. The
1871–1988 by Subbaramayya and Naidu31. Decreasing decreasing trend in winter rainfall was statistically
trends in the central and western Indian sub-divisions significant (95% confidence level) at Kulgam and Hand-
during the late 19th century and again in the 1960s were wara, whereas none of the increasing trends in the
observed. The later trend was reversed in the early 1970s. pre-monsoon and post-monsoon season was significant.
Kothyari and Singh32 found decreasing trend in monsoon Srinagar and Handwara witnessed a decreasing (non-
rainfall and rainy days in the Ganga basin, and monsoon significant) trend in annual rainy days, whereas Kulgam
and annual rainfall over India starting around the second experienced the opposite trend. All the stations experi-
half of the 1960s. Singh and Sontakke33 analysed rainfall enced a decreasing trend in monsoon and winter rainy
data for the period 1829–1999 over the Indo-Gangetic days. Qazigund and Kukarnag experienced decreasing
Plains (IGP) and found significant increasing trend annual rainfall, whereas Srinagar showed increasing
(170 mm/100 yrs) from 1900 in the summer monsoon annual rainfall during 1962–2002. Annual, pre-monsoon,
rainfall over western IGP; non-significant decreasing post-monsoon and winter rainfall increased (non-
trend (5 mm/100 yrs) from 1939 over central IGP and significant), whereas monsoon rainfall decreased (non-
non-significant decreasing trend (50 mm/100 yrs) during significant) at Srinagar during the last century.
1900–1984 and non-significant increasing trend (480 mm/ Studies of temporal variation in monthly, seasonal and
100 yrs) during 1984–1999 over eastern IGP. annual rainfall over Kerala for 1871–2005 by Krishna-
Trend detection of rainfall for the period 1901–1984 at kumar et al.36 revealed significant decrease in SW mon-
11 stations in Himachal Pradesh indicated increasing soon rainfall, and an increase in post-monsoon season.
trend in annual rainfall at 8 stations24. Further, eight other Rainfall during winter and summer seasons showed
stations showed a decreasing trend in monsoon rainfall. insignificant increasing trend. Basistha et al.37 analysed
The decreasing trend of annual and monsoon rainfall in 80 years’ (1901–1980) rainfall data from 30 rain gauge
Shimla was statistically significant at 95% confidence stations in the Indian Himalayas and found an increasing
level. Ramesh and Goswami34 analysed daily gridded ob- trend between 1901 and 1964, and a decreasing trend
served rainfall data for the period 1951–2003 and found between 1965 and 1980. Pal and Al-Tabbaa38 found
decreasing trends in both early and late monsoon rainfall decreasing trends in the spring and monsoon rainfall,
and number of rainy days over India. Analysis of rainfall and increasing trends in the autumn and winter rainfall
amount during different seasons indicated decreasing over India during the period 1954–2003.
tendency in the summer monsoon rainfall over the Indian Rajeevan et al.39 and Guhathakurta and Rajeevan40
land mass and increasing trend in the rainfall during pre- analysed a rainfall series created using a network of 1476
CURRENT SCIENCE, VOL. 102, NO. 1, 10 JANUARY 2012 41
REVIEW ARTICLE
Table 2. Sen’s estimator of slope for rainfall and rainy days*

Annual Pre-monsoon Monsoon Post-monsoon Winter

Basin Rainfall Rainy days Rainfall Rainy days Rainfall Rainy days Rainfall Rainy days Rainfall Rainy days
(mm/yr) (days/yr) (mm/yr) (days/yr) (mm/yr) (days/yr) (mm/yr) (days/yr) (mm/yr) (days/yr)

Ganga 0.000 0.111 0.341 0.048 –0.302 0.000 –0.016 0.000 0.081 0.000
Indus –1.825 –0.182 0.880 0.000 –3.160 –0.167 –0.261 –0.045 0.864 0.000
Brahmaputra –0.442 –0.064 –1.935 –0.025 0.570 0.095 –0.057 –0.061 0.146 0.000
Barak+ 2.662 –0.200 1.344 0.000 1.012 –0.111 0.191 –0.053 0.455 0.042
Luni+ –0.597 –0.104 0.109 0.000 –1.039 –0.125 0.028 0.000 0.057 0.000
Sabarmati –0.990 –0.200 –0.006 0.000 –1.242 –0.231 0.013 0.000 –0.014 0.000
Mahi –2.183 –0.138 0.003 0.000 –2.153 –0.139 0.000 0.000 0.000 0.000
Narmada –2.409 –0.114 –0.030 0.000 –2.394 –0.067 0.097 0.000 0.030 0.000
Tapi –1.387 –0.167 –0.073 0.000 –1.367 –0.139 0.379 0.030 0.000 0.000
WFR1 –10.164 –0.176 –0.318 –0.020 –8.472 –0.079 –0.515 –0.111 0.002 0.000
WFR2 –7.011 –0.375 –2.482 –0.152 –5.612 –0.176 0.926 –0.032 0.050 0.000
BBS 2.222 0.000 0.728 0.045 1.026 –0.034 –0.583 –0.079 0.056 0.000
Mahanadi –2.905 0.053 0.172 0.029 –3.130 0.000 –0.091 0.000 0.127 0.000
Godavari –2.727 –0.276 –0.400 0.000 –2.635 –0.200 0.209 –0.031 0.166 0.000
Krishna –0.271 0.024 –0.569 0.000 –0.827 0.026 0.247 0.000 0.155 0.000
Cauvery 0.879 0.000 –0.563 0.000 0.075 0.028 1.748 0.050 0.024 0.000
Pennar 1.257 0.000 –0.398 –0.027 1.040 0.000 0.114 0.047 0.023 0.000
EFR1 –1.231 –0.190 0.418 0.053 –1.156 –0.167 –1.391 –0.091 0.386 0.034
EFR2 –0.442 –0.044 0.525 0.000 –1.722 –0.125 –0.136 0.000 0.255 0.000
EFR3 4.928 0.111 –0.630 –0.031 2.575 0.070 1.398 0.038 0.227 0.000
EFR4 0.445 –0.032 –0.345 –0.032 –0.214 –0.047 0.659 0.000 0.197 0.000
EFR5 –0.950 –0.333 –0.800 –0.143 –0.500 –0.125 0.491 0.000 –0.246 –0.032

*Bold values indicate statistical significance at 95% confidence level according to the Mann–Kendall test.

rain gauge stations for the period 1901–2003. It showed river basins of India. Using the data of 316 rain gauges,
significant decreasing trend in the monsoon season for Singh et al.45 found that annual rainfall over major basins
three sub-divisions (Jharkhand, Chhattisgarh and Kerala) in Central India (Sabarmati, Mahi, Narmada, Tapi, Goda-
and significant increasing trend over eight sub-divisions vari and Mahanadi) showed decreasing trend since the
(Gangetic West Bengal, Western Uttar Pradesh, Jammu 1960s, whereas Indus, Ganga, Brahmaputra, Krishna and
and Kashmir, Konkan and Goa, Madhya Maharashtra, Cauvery basins showed an increasing trend.
Rayalseema, coastal Andhra Pradesh and North Interior Singh et al.13 studied the changes in rainfall over the
Karnataka). Monthly, seasonal and annual precipitation last century in nine river basins of northwest and Central
of five meteorological sub-divisions of Central North India. The rate of change of rainfall in each of 43 stations
East India showed a significant decreasing trend of was estimated by trend line slope and these point values
4.6 mm/yr for Jharkhand and 3.2 mm/day for Central were interpolated to get the spatial distribution of rainfall
North East India during the period 1889–2008 (ref. 41). change over the study area. They found increasing trends
During December, all the meteorological sub-divisions, in annual rainfall over eight river basins in the 2–19%
except Jharkhand have shown a significant decreasing range of the mean per 100 years.
trend of rainfall in the recent past. Kumar and Jain46 carried out a detailed analysis to
Some studies have investigated the trend and magni- determine the trends in rainfall amount and the number of
tude of variations in rainfall on the basin-scale. Mirza et rainy days in Indian river basins using daily gridded rain-
al.42 studied the changes over Ganges, Brahmaputra and fall data at 1° × 1° resolution provided by the India Mete-
Meghna basins and found that precipitation in the Ganges orological Department (IMD). The magnitude of trend in
basin was by and large stable. One sub-division of the annual and seasonal rainfall and rainy days as determined
Brahmaputra basin showed a decreasing trend, whereas by the Sen’s estimator is given in Table 2. Six river ba-
another showed an increasing trend. For Meghna basin, sins had increasing trend in annual rainfall and 15 river
one sub-division showed decreasing trend and another basins had the opposite trend, whereas the Ganga basin
showed increasing trend. Rao43 did not find any signifi- had no trend. The increase in annual rainfall varied
cant trend in monsoon and annual rainfall over Mahanadi between 0.45 mm/yr (for EFR4 basin) and 4.93 mm/yr
basin during the period 1901–1980. Ranade et al.44 found (for EFR3 basin); the decrease was maximum for WFR1
no trend in the starting or ending date, duration and total basin (–10.16 mm/yr) and minimum for Krishna basin
rainfall of the hydrological wet season over different (–0.27 mm/yr). Annual rainfall of only two basins showed

42 CURRENT SCIENCE, VOL. 102, NO. 1, 10 JANUARY 2012

 REVIEW ARTICLE

Figure 3. Trends and magnitude pertaining to rainfall changes for different river basins in India. Significant trends are displayed in
boxes. a, Annual rainfall (% of mean/100 yrs); b, annual rainy days (% of mean/100 yrs).

significant trend and the trends for majority of the basins monsoon, 10 in the post-monsoon and 19 in winter showed
were non-significant. In the case of annual rainy days, 4 no change in the rainy days. In the monsoon season, the
river basins showed increasing trend, 15 river basins maximum increase in the number of rainy days was in the
showed decreasing trend and 3 showed no change. The Brahmaputra basin (0.095 days/yr), whereas the maximum
maximum decrease was shown by WFR2. The increase in decrease was in the Sabarmati basin (–0.231 days/yr).
rainy days varied from 0.024 to 0.111 days/yr, with Trends and magnitude of changes in annual rainfall and
maximum increase in the Ganga and EFR3. The decreas- rainy days in terms of percentage of mean per 100 years
ing trend in annual rainy days in the Godavari, WFR2 and for basins are shown in Figure 3 (ref. 46). Direction of
EFR5 river basins was statistically significant at 95% trend in rainfall and rainy days was found to be similar
confidence level. for most river basins. For annual data, Mahanadi and
Analysis of seasonal trends showed that pre-monsoon Krishna experienced decreasing trend in annual rainfall
rainfall increased over 9 river basins and decreased over and increasing trend in rainy days, which means that
13; monsoon rainfall increased over 6 basins and decreased droughts may become more recurrent in Krishna (a water-
over 16; post-monsoon rainfall increased over 13 basins stressed basin)46. Barak+ and EFR4 have experienced
and decreased over 8; and winter rainfall increased over increasing rainfall and decreasing rainy days, which
18 basins and decreased over 2 of them. In the monsoon implies that floods may become more intense. Similarly,
season, the maximum increase was of the order of in the monsoon season, Barak+ and BBS (positive trend
2.58 mm/yr for EFR3 and maximum decrease was for in rainfall and negative in rainy days) and Krishna (nega-
WFR1 (–8.47 mm/yr). The Ganga basin, with no change tive trend in rainfall and positive trend in rainy days)
in annual rainfall, experienced a small decreasing trend in experienced the opposite trend in rainfall and rainy days.
the monsoon and post-monsoon rainfall. There is some Similar to annual data, these trends are non-significant.
notion of weakening summer (SW) monsoon and strength-
ening in winter (NE) monsoon. Monsoon rainfall indi- Studies of extreme rainfall trends in India
cated negative significant trend in the Indus and WFR1
river basins. In the seasonal rainy days, 4 river basins Studies by various authors25,47–53 show that, in general,
each in the pre-monsoon, monsoon and post-monsoon the frequency of more intense rainfall events in many
season and 2 in the season experienced an increasing parts of Asia has increased whereas the number of rainy
trend. Eleven river basins in the pre-monsoon, 3 in the days and total annual precipitation has decreased. An

CURRENT SCIENCE, VOL. 102, NO. 1, 10 JANUARY 2012 43

REVIEW ARTICLE

increase in intense rainfall events leads to more severe also studied the changes in frequency and magnitudes of
floods and landslides. extreme monsoon rainfall deficiency and excess from
Decreasing trends, particularly for stations in the hilly 1871 to 2005 over 5 regions in India. The extreme mon-
terrain, were found in annual extreme rainfall over soon seasonal precipitation exhibited a negative tendency
Kerala54. Rakhecha and Soman55 while analysing extreme leading to increasing frequency and magnitude of mon-
events from 1 to 3 days duration for 316 stations across soon rainfall deficit and decreasing frequency and magni-
India for the period 1901–1980 found that increasing tude of monsoon rainfall excess.
trends in these events were not statistically significant at The long-term trend in monsoon season extreme rain-
most stations. They reported that ‘the extreme rainfall fall events for 1951–2005 was analysed by Pattanaik and
series at stations over the west coast north of 12-degrees- Rajeevan62. The average frequency of extreme rainfall
N and at some stations to the east of the Western Ghats events along with the contribution of extreme rainfall
over the central parts of the peninsula showed a signifi- events to the seasonal rainfall showed a significant
cant increasing trend at 95% level of confidence. Stations increasing trend (above the 98% confidence level) over
over the southern peninsula and over the lower Ganga India during monsoon season and also during June and
valley have been found to exhibit a decreasing trend at July. It was also found that the increasing trend of contri-
the same level of significance. bution from extreme rainfall events is balanced by a
Sen Roy and Balling56 analysed daily rainfall data for decreasing trend in low rainfall events.
1910–2000 at 129 stations and found generally increasing Kothawale et al.63 also studied the association between
trend in a contiguous region extending from the north- El Niño Southern Oscillation (ENSO) and monsoon rain-
western Himalayas in Kashmir through most of the Deccan fall over India and reported a strong association between
Plateau in the south and decreasing values in the eastern El Niño events and deficient monsoon rainfall. Nearly
part of the Gangetic Plains and parts of Uttarakhand. 60% of major droughts over India have occurred in asso-
Trends in extreme rainfall indices for the period 1901– ciation with El Niño events. Strong association between
2000 were examined for 100 stations over India by Joshi monsoon droughts and El Niño events was noted by the
and Rajeevan57. Most of the extreme rainfall indices for authors. On the other hand, La Niña events were associ-
the SW monsoon season and annual period showed ated with more rainfall during monsoon and cooling.
significant positive trends over the west coast and Earlier, Krishna Kumar et al.64 had shown that the rela-
northwestern parts of the peninsula. However, two hill tion between the Indian monsoon and ENSO weakened in
stations (Shimla and Mahabaleshwar) showed decreasing recent decades.
trend in some of the extreme rainfall indices.
Using high resolution daily gridded rainfall data for the
period 1951–2003, Goswami et al.53 showed that there Studies of temperature trends in India
were significant rising trends in the frequency and magni-
tude of extreme rain events over central India during the The surface temperatures over a given region vary sea-
monsoon season. They also found significant decreasing sonally and annually depending upon latitude, altitude
trend in the frequency of moderate events during the and location with respect to geographical features such as
same period, thus leading to no significant trend in the a water body (river, lake or sea), mountains, etc. Probably
mean rainfall. The results of this study were contradicted one of the most widely quoted aspects of climatic change
for some places when analysis was performed at a finer and the one that will have important ramifications on a
resolution (1° lat. × 1° long.) by Ghosh et al.58. range of sectors, including water is the significant
Variability and long-term trends of extreme rainfall increase in the global mean air temperature during the
events over central India were examined by Rajeevan et past century. Since the hydrologic cycle is a thermally
al.59 using 104 years (1901–2004) of high-resolution driven system, rise in global temperature is likely to
daily gridded rainfall data. They found statistically signi- accelerate this cycle. Identification of the temperature
ficant long-term trend of 6% per decade in the frequency trends and their projection has been the subject matter of
of extreme rainfall events. According to them, the in- a large number of studies.
creasing trend of extreme rainfall events in the last five Most of the temperature trend studies in India focus on
decades could be associated with the increasing trend of the analysis of annual and seasonal temperature data for a
sea surface temperatures and surface latent heat flux over single station or a group of stations. Such studies date
the tropical Indian Ocean. back to at least 50 years. In one of the first studies, trends
Pal and Al-Tabbaa60 studied the trends in seasonal rain- in the annual mean, maximum and minimum tempera-
fall extremes in Kerala, using gridded daily data for tures over the whole country were studied by Pramanik
1954–2003. They found winter and post-monsoon extreme and Jagannathan65, who did not find any general tendency
rainfall having an increasing tendency with statistically for an increase or decrease in these temperatures. The
significant changes in some regions and decreasing trends study by Srivastava et al.66 on decadal trends in climate
in spring seasonal extreme rainfall. Pal and Al-Tabbaa61 over India gave the first indications that the diurnal
44 CURRENT SCIENCE, VOL. 102, NO. 1, 10 JANUARY 2012
 REVIEW ARTICLE

asymmetry of temperature trends over India is quite rising trend and some stations in the north and northeast-
different from that over many other parts of the globe. ern parts showed a falling trend in annual mean tempera-
Pant and Kumar67 analysed seasonal and annual air tem- ture. In the winter season, a rising trend in mean
perature series for 1881–1997 and showed that there is a temperature was observed at 39 stations and a falling
significant warming trend of 0.57°C per 100 years. The trend at 19 stations. Most of the stations in the coastal
magnitude of warming was higher in the post-monsoon and southern areas showed an increasing trend, whereas a
and winter seasons. The monsoon temperature did falling trend was exhibited by some stations in the north,
not show a significant trend in any major part of the central and eastern India. The mean pre-monsoon tem-
country, except for a significant negative trend over north- perature showed a rising trend at 35 stations and a falling
west India. trend at 23 stations. Most of the stations in the eastern
Mean annual temperature was found to be increasing region and in the foothills of the Himalayas showed a
over the west coast, interior peninsula, north central and falling trend. For the mean monsoon temperature series,
northeastern regions of India along with the whole of 27 stations showed a rising trend and 18 stations a falling
India during the period 1901–1982 by Hingane et al.68. trend. In the mean post-monsoon temperature series, 59
Pant and Hingane29 found decreasing trend in mean stations showed a rising trend and six a falling trend. A
annual surface air temperature for 1901–1982 over the rising trend dominates all over India in this season.
northwest Indian region consisting of the meteorological Arora et al.71 also found that the percentage of signifi-
sub-divisions of Punjab, Haryana, west Rajasthan, east cant trends was high and there is a rising trend in most
Rajasthan and west Madhya Pradesh. Data from seven cases, except for temperatures for mean pre-monsoon,
stations for 1901–1980 showed highly significant warm- mean monsoon, pre-monsoon mean minimum and mon-
ing trend in the mean maximum, mean minimum and soon mean minimum. There was an increase of 0.42°C
average mean temperatures of the Mahanadi river basin43. (100 year)–1 in the annual mean temperature, 0.92°C (100
Trend analyses of maximum and minimum temperature year)–1 in the mean maximum temperature and 0.09°C
data at 121 stations in India for 1901–1987 by Rupa (100 year)–1 in the mean minimum temperature. Averaged
Kumar et al.69 showed increasing maximum temperature for all seasons, there was a rise of 1.1°C (100 year)–1 in
and trendless minimum temperature, resulting in rise in mean winter temperature, 0.94°C (100 year)–1 in mean
mean and diurnal range of temperature. post-monsoon temperature, and a fall of –0.40°C
Kothyari and Singh32 found increasing annual maxi- (100 year)–1 in mean pre-monsoon temperature.
mum temperature over the Ganga basin and increasing Atmospheric surface temperature in India has increased
average annual temperature for India starting around the in the last century by about 1°C and 1.1°C during winter
second half of 1960s. The annual surface air temperature and post-monsoon months respectively. Also, decrease in
in IGP showed significant rising trend (0.53°C/100-yrs) the minimum temperature during summer monsoon and
during 1875–1958 and significant decreasing trend its increase during post-monsoon months have created a
(–0.93°C/100-yrs) during 1958–1997 (ref. 33). Sinha Ray large difference of about 0.8°C in the seasonal tempera-
and De21 summarized information on trends in the occur- ture anomalies. Opposite phases of increase and decrease
rence of extreme events over India. An increasing trend in the minimum temperatures in the southern and northern
of 0.35°C over the last 100 years was observed in the regions, respectively, have been noticed in the inter-
temperature records. Extreme maximum and minimum annual variability. In north India, the minimum tempera-
temperatures showed an increasing trend in the south and ture showed a sharp decrease between 1955 and 1972,
a decreasing trend in the north. Temporal variation in the and then a sharp increase. But in south India, the mini-
temperature over Pune city during the period 1901–2000 mum temperature showed a steady increase25.
revealed significant decrease in mean annual and mean Bhutiyani et al.72 found increasing trend in maximum,
maximum temperature70. Winter season experienced minimum, mean and diurnal temperature range over the
decrease in temperature, whereas monsoon season tem- northwestern Himalayan region during the 20th century.
perature was found to be increasing. Frequency of occurrence of hot days and hot nights
Arora et al.71 analysed a series of annual and seasonal showed widespread increasing trend, whereas that of cold
mean temperature, annual mean maximum temperature days and cold nights showed widespread decreasing trend
and annual mean minimum temperature using the data during the period 1970–2005 over India as a whole and
from 125 stations distributed throughout India. For annual seven homogeneous regions63. The frequency of occurrence
series, mean maximum temperature showed a rising trend of hot days was found to have significantly increased
at 63 stations and a falling trend at 8 stations. The mean over the east coast, west coast and interior peninsula,
minimum temperature showed a rising trend at 33 sta- whereas that of cold days showed significant decreasing
tions and a falling trend at 31 stations. In the mean tem- trend over Western Himalaya and the west coast. The
perature series, a rising trend was observed at 53 stations three regions, east coast, west coast and northwest,
and a falling trend at 17 stations. Most of the stations in showed significant increasing trend in the frequency of
the south, central and western parts of India showed a hot nights.
CURRENT SCIENCE, VOL. 102, NO. 1, 10 JANUARY 2012 45
REVIEW ARTICLE

Trends in annual and seasonal temperatures in the four increased by 0.20°C per decade due to significant increases
most populated cities of India – Delhi, Kolkata, Mumbai in both maximum and minimum temperatures. Also, in-
and Chennai showed positive change with different rates in crease in minimum temperature was much steeper than
different seasons73. The maximum temperature at Mumbai the maximum temperature. On the whole, winter and
during winter and monsoon was significantly increasing, summer monsoon temperatures showed a significant
whereas minimum temperature showed significant de- increasing trend over almost the entire country, whereas
crease. The remaining cities recorded significant increase post-monsoon temperatures significantly increased over
in minimum temperature during winter. Analysis of relatively smaller number of regions.
maximum, minimum and mean temperatures; diurnal (ii) On larger spatially aggregated scales, the tempera-
temperature range; and sunshine duration in eight sites in ture trends in India were found to be quite consistent and
North East India by Jhajharia and Singh74 showed de- in agreement with global and hemispheric trends. How-
creasing trends in diurnal temperature range correspond- ever, on smaller regional scales and for different sub-
ing to annual, seasonal (pre-monsoon and monsoon) and periods, trends were not always consistent with Indian
monthly (September) timescales at three sites. Diurnal aggregated temperatures. Further, the trends were also in-
temperature range increases were also observed at three fluenced by the variability of rainfall in the monsoon and
other sites in the monsoon and post-monsoon seasons as post-monsoon seasons. It was opined that the recent tem-
well as in June, October and December. Temperature perature changes in some parts of the country may also be
remained practically trendless in winter and pre-monsoon due to the relative influence of greenhouse gases and
season and increased in the monsoon and post-monsoon aerosols.
seasons over NE India. Decreasing trends in sunshine (iii) Accelerated warming was observed during 1971–
duration were observed mainly on annual, seasonal 2007 and this was attributed mainly to intense warming
(winter and pre-monsoon) and monthly (January–March) in the decade 1998–2007. During 1998–2007, maximum
timescales. temperature was significantly higher compared to the
Maximum temperature showed significant rising trend long-term (1901–2007) mean throughout India, with a
of 0.008°C/yr during monsoon season; 0.014°C/yr during stagnated trend during this period, whereas minimum
post-monsoon season and 0.008°C/yr in the annual maxi- temperature showed an increasing trend, almost equal
mum temperature during the period 1914–2003 for Central to that observed during 1971–2007. It is pertinent to
Northeast India41. Minimum temperature showed signifi- note that recently, the year 2010 has been reported to
cant rising trend of 0.012°C/yr during post-monsoon sea- rank in the top three warmest years since the beginning of
son and significant falling trend of 0.002°C/yr during instrumental climate records in 1850, according to data
monsoon season. Studies of the long-term trends and sources compiled by the World Meteorological Organiza-
variations of the monthly maximum and minimum tem- tion (WMO).
peratures in various climatological regions in India by Pal (iv) On the seasonal scale, maximum temperature has
and Al-Tabbaa75 showed increasing monthly maximum significantly increased in all seasons during the period
temperature, though unevenly, over the last century. 1901–2007. On this scale, pronounced warming trends in
Minimum temperature changes were found more variable mean temperature were observed in two seasons – winter
than maximum temperature changes, both temporally and and monsoon. However, for the recent period only, winter
spatially, with results of lesser significance. and post-monsoon temperatures showed significant
The monthly maximum and minimum temperature data warming trends and the other seasons showed a warming
from 121 stations well distributed over the country for the tendency (trend not significant). In contrast, minimum
period 1901–2007 were used by Kothawale et al.76 and temperature showed a significant warming trend in most
from these data, they computed seasonal and annual seasons during 1971–2007.
trends in surface air temperature over the country and (v) The global mean annual and seasonal (DJF, MAM,
seven homogeneous regions (western Himalaya, north- JJAS and ON) temperatures have significantly increased
west, north-central, northeast, east coast, west coast and by 0.82°C, 0.89°C, 0.94°C, 0.72°C and 0.76°C per 100
interior peninsula) during three periods: 1901–2007, years, respectively, during the period 1901–2007. Indian
1971–2007 and 1998–2007. Key findings of this compre- mean annual and seasonal temperatures also showed a
hensive study are described here. significant warming trend in all seasons. The magnitude
(i) Annual mean (average of maximum and minimum), of the warming trend of winter and post-monsoon seasons
maximum and minimum temperatures showed significant was almost the same for these two areas, whereas
warming trends of 0.51, 0.72 and 0.27°C (100 year)–1, pre-monsoon and monsoon temperature trends for India
respectively, during 1901–2007. The temperature has in- were half that of the global trend (Figure 4).
creased gradually and continuously over this period. This Kothawale and Rupa Kumar77 had shown a strong
warming was mainly due to increasing temperatures in negative simultaneous correlation between Indian mon-
the winter and post-monsoon seasons. During the three soon mean surface temperature and monsoon rainfall.
decades from 1971 to 2007, annual mean temperature Kothawale et al.63 concluded that on interannual time
46 CURRENT SCIENCE, VOL. 102, NO. 1, 10 JANUARY 2012
 REVIEW ARTICLE

show a large variability. In a recent study, 6 river basins

have shown increasing trend in annual rainfall in the
0.27–10.16 mm/yr range whereas 15 river basins have
shown decreasing annual rainfall in the 0.45–4.93 mm/yr
range. Further, 4 river basins have shown increasing trend
in rainy days and 15 river basins have shown the opposite
trend. Seasonal analysis shows that in the pre-monsoon
season, rainfall increased over nine river basins and rainy
days increased in four river basins; in the monsoon sea-
son, rainfall increased over six river basins and rainy
days in four river basins; in the post-monsoon season,
rainfall increased over 13 river basins and rainy days on
four river basins and in the winter season, rainfall in-
creased over 18 river basins and rainy days increased
over 2 river basins. Rainfall decreased in 13 river basins
in the pre-monsoon season, in 16 basins in the monsoon
and in 8 basins in the post-monsoon. Some of these are
major and water-deficit basins. Majority of river basins in
the pre-monsoon, post-monsoon and winter season have
shown no change in rainy days. Further, majority of the
basins show neither increasing nor decreasing significant
trend in seasonal rainfall and rainy days. Absence of sig-
nificant change in the number of rainy days shows that at
present there is no clear-cut indication of any major
change in rainfall intensities. Temperature data, however,
show nearly monotonous rising trends although the
behaviour of the maximum and the minimum temperature
at different locations is different.
In trend analysis studies, the results significantly depend
Figure 4. Global (red) and Indian (blue) annual and seasonal mean
temperature variations during 1901–2007. *p < 0.05; **p < 0.01. DJF, upon the period of data and the stations whose data are
Previous year December–February; MAM, pre-monsoon March–May; used. There is another observation that most of the data
JJAS, monsoon June–September; ON, post-monsoon October– used in trend detection pertain to the stations that are
November (Source: Kothawale et al.63).
located in urban areas and these areas are a sort of heat
islands. Thus, the study of trends using this data may not
be the correct depiction of the reality and this aspect
scales, Indian mean temperatures are strongly correlated
needs to be addressed. These concerns, in fact, highlight
with SST in the eastern Pacific and the equatorial Indian
the importance of identifying a network of baseline stations
Ocean. A strong interannual link between Indian tem-
for change detection studies. Countries, such as the USA,
peratures and Indian Ocean SST was found. ENSO was
have identified such networks. Existing stations with long
found to be impacting Indian temperatures significantly.
series of good quality will be the obvious candidates for
The composite maximum temperature anomalies of El
the baseline network and it will be necessary to add new
Niño years were statistically significant and positive dur-
stations so that the whole country is properly represented.
ing monsoon and post-monsoon seasons over large areas
More studies to detect teleconnections between Indian
of the country and the composite anomalies of La Niña
rainfall and temperatures and ENSO/SST are also needed.
years were almost opposite to El Niño years. Interest-
In a major exercise, an assessment of impact of climate
ingly, the year 2010 was special for climate scientists
change in the 2030s on four key sectors of the Indian
since El Niño conditions prevailed in the tropical Pacific
economy, namely agriculture, water, natural ecosystems
for the first four months but it quickly changed and a La
and biodiversity, and health in four climate sensitive re-
Niña pattern had emerged by June.
gions of India, namely the Himalayan region, the Western
Ghats, the coastal areas and the northeastern region has
Concluding remarks been carried out by the Ministry of Environment and
Forests, GoI78. There is a need of an integrated nation-
An understanding of the spatial and temporal distribution wide program of change detection, impact assessment,
and changing patterns in rainfall is a basic and important and adaptation and mitigation.
requirement for the planning and management of water Finally, while interpreting the results of trend analysis,
resources. Expectedly, the river-basin rainfall trends the observations of Cohn and Lins79 are worth repeating:
CURRENT SCIENCE, VOL. 102, NO. 1, 10 JANUARY 2012 47
REVIEW ARTICLE

‘that reported trends are real yet insignificant indicates a 24. Kumar, V., Singh, P. and Jain, S. K., Rainfall trends over
worrisome possibility: natural climatic excursions may be Himachal Pradesh, Western Himalaya, India. Conference on
Development of Hydro Power Projects – A Prospective Challenge,
much larger than we imagine. So large, perhaps, that they Shimla, 20–22 April 2005.
render insignificant the changes, human-induced or other- 25. Dash, S. K., Jenamani, R. K., Kalsi, S. R. and Panda, S. K., Some
wise, observed during the past century’. evidence of climate change in twentieth-century India. Climatic
Change, 2007, 85, 299–321.
26. Kumar, V. and Jain, S. K., Trends in seasonal and annual rainfall
1. Mehrotra, R., Sensitivity of runoff, soil moisture and reservoir and rainy days in Kashmir valley in the last century. Quaty Intl.,
design to climate change in central Indian river basins. Climatic 2010, 212, 64–69.
Change, 1999, 42, 725–757. 27. Parthasarathy, B. and Dhar, O. N., Secular variations of regional
2. IPCC, Summary for policymakers. In Climate Change 2007: The rainfall over India. Q. J. R. Meteorol. Soc., 1974, 100, 245–257.
Physical Science Basis (eds Solomon, S. D. et al.), Cambridge 28. Dhar, O. N., Rakhecha, P. R. and Kulkarni, A. K., Fluctuations in
University Press, Cambridge, UK, 2007. northeast monsoon rainfall of Tamil Nadu. Int. J. Climatol., 1982,
3. Kundzewicz, Z. W., Change detection in hydrological records – a 2, 339–345.
review of the methodology. Hydrol. Sci., J., 2004, 49(1), 7–19. 29. Pant, G. B. and Hingane, L. S., Climatic changes in and around
4. Sen, P. K., Estimates of the regression coefficient based on Kend- the Rajasthan desert during the 20th century. Int. J. Climatol.,
all’s tau. J. Am. Stat. Assoc., 1968, 63, 1379–1389. 1988, 8, 391–401.
5. Lettenmaier, D. P., Wood, E. F. and Wallis, J. R., Hydro- 30. Rupa Kumar, K., Pant, G. B., Parthasarathy, B. and Sontakke,
climatological trends in the continental United States, 1948–88. N. A., Spatial and sub-seasonal patterns of the long term trends of
J. Climate, 1994, 7, 586–607. Indian summer monsoon rainfall. Int. J. Climatol., 1992, 12, 257–
6. Yue, S. and Hashino, M., Temperature trends in Japan: 1900– 268.
1990. Theor. Appl. Climatol., 2003, 75, 15–27. 31. Subbaramayya, I. and Naidu, C. V., Spatial variations and
7. Partal, T. and Kahya, E., Trend analysis in Turkish precipitation trends in the Indian monsoon rainfall. Int. J. Climatol., 1992, 12,
data. Hydrol. Process, 2006, 20, 2011–2026. 597–609.
8. Yu, Y. S., Zou, S. and Whittemore, D., Non-parametric trend 32. Kothyari, U. C. and Singh, V. P., Rainfall and temperature trends
analysis of water quality data of rivers in Kansas. J. Hydrol., in India. Hydrol. Process, 1996, 10, 357–372.
1993, 150, 61–80. 33. Singh, N. and Sontakke, N. A., On climatic fluctuations and envi-
9. Douglas, E. M., Vogel, R. M. and Knoll, C. N., Trends in flood ronmental changes of the Indo-Gangetic Plains, India. Climatic
and low flows in the United States: impact of spatial correlation. Change, 2002, 52(3), 287–313.
J. Hydrol., 2000, 240, 90–105. 34. Ramesh, K. V. and Goswami, P., The shrinking Indian summer
10. Yue, S., Pilon, P. and Phinney, B., Canadian streamflow trend monsoon. Research Report RR CM 0709, CSIR Centre for
detection: impacts of serial and cross-correlation. Hydrol. Sci. Mathematical Modelling and Computer Simulation, Bangalore,
J., 2003, 48, 51–63. 2007.
11. Burn, D. H., Cunderlik, J. M. and Pietroniro, A., Hydrological 35. Pattanaik, D. R., Analysis of rainfall over different homogeneous
trends and variability in the Liard river basin. Hydrol. Sci. regions of India in relation to variability in westward movement
J., 2004, 49, 53–67. frequency of monsoon depressions. Nat. Hazards, 2007, 40(3),
12. Singh, P., Kumar, V., Thomas, T. and Arora, M., Changes in rain- 635–646.
fall and relative humidity in different river basins in the northwest 36. Krishnakumar, K. N., Rao, G. S. L. H. V. P. and Gopakumar,
and central India. Hydrol. Process., 2008, 22, 2982–2992. C. S., Rainfall trends in twentieth century over Kerala, India.
13. Singh, P., Kumar, V., Thomas, T. and Arora, M., Basin-wide Atmos. Environ., 2009, 43, 1940–1944.
assessment of temperature trends in the north-west and central 37. Basistha, A., Arya, D. S. and Goel, N. K., Analysis of historical
India. Hydrol. Sci. J., 2008, 53, 421–433. changes in rainfall in the Indian Himalayas. Int. J. Climatol.,
14. Salas, J. D., Analysis and modeling of hydrologic time series. In 2009, 29, 555–572.
Handbook of Hydrology (ed. Maidment, D. R.), McGraw-Hill, 38. Pal, I. and Al-Tabbaa, A., Assessing seasonal precipitation trends
New York, 1993, pp. 19.1–19.72. in India using parametric and non-parametric statistical techniques.
15. Helsel, D. R. and Hirsch, R. M., Statistical Methods in Water Theor. Appl. Climatol., 2010, DOI: 10.1007/s00704-010-0277-8.
Resources, Elsevier, New York, 1992. 39. Rajeevan, M., Bhate, J., Kale, J. D. and Lal, B., High resolution
16. Hirsch, R. M., Helsel, D. R., Cohn, T. A. and Gilroy, E. J., Statistical daily gridded rainfall data for the Indian region: Analysis of break
treatment of hydrologic data. In Handbook of Hydrology (ed. and active monsoon spells. Curr. Sci., 2006, 91(3), 296–306.
Maidment, D. R.), McGraw-Hill, New York, 1993, pp. 17.1–17.52. 40. Guhathakurta, P. and Rajeevan, M., Trends in the rainfall pattern
17. Mooley, D. A. and Parthasarthy, B., Fluctuations of all India over India. Int. J. Climatol., 2008, 28, 1453–1469.
summer monsoon rainfall during 1871–1978. Climatic Change, 41. Subash, N., Sikka, A. K. and Ram Mohan, H. S., An investigation
1984, 6, 287–301. into observational characteristics of rainfall and temperature in
18. Sarker, R. P. and Thapliyal, V., Climate change and variability. Central Northeast India – a historical perspective 1889–2008.
Mausam, 1988, 39, 127–138. Theor. Appl. Climatol., 2010, DOI: 10.1007/s00704-010-0299-2.
19. Thapliyal, V. and Kulshreshtha, S. M., Climate changes and trends 42. Mirza, M. Q., Warrick, R. A., Ericksen, N. J. and Kenny, G. J.,
over India. Mausam, 1991, 42, 333–338. Trends and persistence in precipitation in the Ganges, Brahmapu-
20. Lal, M., Climatic change – implications for India’s water re- tra and Meghna river basins. Hydrol. Sci. J., 1998, 43, 845–858.
sources. J. Indian Water Resour. Soc., 2001, 21, 101–119. 43. Rao, P. G., Climatic changes and trends over a major river basin in
21. Sinha Ray, K. C. and De, U. S., Climate change in India as evi- India. Climate Res., 1993, 2, 215–223.
denced from instrumental records. WMO Bull., 2003, 52, 53–59. 44. Ranade, A., Singh, N., Singh, H. N. and Sontakke, N. A., On vari-
22. Kumar, V., Jain, S. K. and Singh, Y., Analysis of long-term rain- ability of hydrological wet season, seasonal rainfall and rainwater
fall trends in India. Hydrol. Sci. J., 2010, 55, 484–496. potential of the river basins of India (1813–2006). J. Hydrol. Res.
23. Srivastava, H. N., Sinha Ray, K. C., Dikshit, S. K. and Muk- Dev., 2008, 23, 79–108.
hopadhaya, R. K., Trends in rainfall and radiation over India. 45. Singh, N., Sontakke, N. A., Singh, H. N. and Pandey, A. K.,
Vayu Mandal, 1998, 41–45. Recent trend in spatiotemporal variation of rainfall over Indian

48 CURRENT SCIENCE, VOL. 102, NO. 1, 10 JANUARY 2012

 REVIEW ARTICLE
investigation into basin-scale rainfall fluctuations. IAHS Publica- 63. Kothawale, D. R., Munot, A. A. and Krishna Kumar, K., Surface
tion No. 296, 2005, pp. 273–282. air temperature variability over India during 1901–2007, and its
46. Kumar, V. and Jain, S. K., Trends in rainfall amount and number association with ENSO. Climate Res., 2010, 42, 89–104.
of rainy days in river basins of India (1951–2004). Hydrol. Res., 64. Krishna Kumar, K., Rajagopalan, B. and Cane, M. K., On the
2010, 42(4), 290–306. weakening relationship between the Indian monsoon and ENSO.
47. Sinha Ray, K. C. and Srivastava, A. K., Is there any change in Science, 1999, 284, 2156–2159.
extreme events like droughts and heavy rainfall? INTROPMET- 65. Pramanik, S. K. and Jagannathan, P., Climatic changes in India
97, 2–5 December 1997, IIT New Delhi, 1999. rainfall. Indian J. Meteorol. Geophys., 1954, 4, 291–309.
48. Khan, T. M. A., Singh, O. P. and Sazedur Rahman, M. D., Recent 66. Srivastava, H. N., Dewan, B. N., Dikshit, S. K., Rao, P. G. S.,
sea level and sea surface temperature trends along the Bangladesh Singh, S. S. and Rao, K. R., Decadal trends in climate over India.
coast in relation to the frequency of intense cyclones. Mar. Geo- Mausam, 1992, 43, 7–20.
desy, 2000, 23, 103–116. 67. Pant, G. B. and Kumar, K. R., Climates of South Asia, John Wiley,
49. Shrestha, A. B., Wake, C. P., Dibb, J. E. and Mayewski, P. A., Chichester, UK, 1997.
Precipitation fluctuations in the Nepal Himalaya and its vicinity 68. Hingane, L. S., Rupa Kumar, K. and Murty, V. R., Long-term
and relationship with some large scale climatological parameters. trends of surface air temperature in India. Int. J. Climatol., 1985,
Int. J. Climatol., 2000, 20, 317–327. 5, 521–528.
50. Mirza, M. Q., Global warming and changes in the probability of 69. Rupa Kumar, K., Krishankumar and Pant, G. B., Diurnal asymme-
occurrence of floods in Bangladesh and implications. Global Envi- try of surface temperature trends over India. Geophys. Res. Lett.,
ron. Change, 2002, 12, 127–138. 1994, 21, 677–680.
51. Lal, M., Global climate change: India’s monsoon and its variabi- 70. Gadgil, A. and Dhorde, A., Temperature trends in twentieth
lity. J. Environ. Stud. Policy, 2003, 6, 1–34. century at Pune, India. Atmos. Environ., 2005, 39, 6550–6556.
52. Min, S. K., Kwon, W. T., Park, E. H. and Choi, Y., Spatial and 71. Arora, M., Goel, N. K. and Pratap Singh, Evaluation of tempera-
temporal comparisons of droughts over Korea with East Asia. Int. ture trends over India. Hydrol. Sci. J., 2005, 50, 81–93.
J. Climatol., 2003, 23, 223–233. 72. Bhutiyani, M. R., Kale, V. S. and Pawar, N. J., Long-term trends
53. Goswami, B. N., Venugopal, V., Sengupta, D., Madhusoodanam, in maximum, minimum and mean annual air temperatures across
M. S. and Xavier, P. K., Increasing trends of extreme rain events the Northwestern Himalaya during the twentieth century. Climatic
over India in a warming environment. Science, 2006, 314, 1442– Change, 2007, 85, 159–177.
1445. 73. Dhorde, A. and Gadgil, A. S., Long-term temperature trends at
54. Soman, M. K., Krishna Kumar, K. and Singh, N., Decreasing four largest cities of India during the twentieth century. J. Indian
trend in the rainfall of Kerala. Curr. Sci., 1988, 57, 7–12. Geophys. Union, 2009, 13, 85–97.
55. Rakhecha, P. R. and Soman, M. K., Trends in the annual extreme 74. Jhajharia, D. and Singh, V. P., Trends in temperature, diurnal
rainfall events of 1 to 3 days duration over India. Theor. Appl. temperature range and sunshine duration in Northeast India. Int. J.
Climatol., 1994, 48, 227–237. Climatol., 2010, DOI:10.1002/joc.2164.
56. Sen Roy, S. and Balling, R. C., Trends in extreme daily precipita- 75. Pal, I. and Al-Tabbaa, A., Long-term changes and variability of
tion indices in India. Int. J. Climatol., 2004, 24, 457–466. monthly extreme temperatures in India. Theor. Appl. Climatol.,
57. Joshi, U. R. and Rajeevan, M., Trends in precipitation extremes 2010, 100, 45–56.
over India. Research Report No: 3/2006, National Climate Centre, 76. Kothawale, D. R., Revadekar, J. V. and Rupa Kumar, K., Recent
India Meteorological Department, Pune, 2006. trends in pre-monsoon daily temperature extremes over India.
58. Ghosh, S., Luniya, V. and Gupta, A., Trend analysis of Indian J. Earth Syst. Sci., 2010, 119, 51–65.
summer monsoon rainfall at different spatial scales. Atmos. Sci. 77. Kothawale, D. R. and Rupa Kumar, K., Tropospheric temperature
Lett., 2009, 10, 285–290. variation over India and links with the Indian summer monsoon:
59. Rajeevan, M., Bhate, J. and Jaswal, A. K., Analysis of variability 1971–2000. Mausam, 2002, 53, 289–308.
and trends of extreme rainfall events over India using 104 years of 78. MOEF, Climate change and India: A ‘4 × 4’ assessment. Ministry
gridded daily rainfall data. Geophys. Res. Lett., 2008, 35, L18707. of Environment and Forests, New Delhi, 2010.
60. Pal, I. and Al-Tabbaa, A., Trends in seasonal precipitation 79. Cohn, T. A. and Lins, H. F., Nature’s style: naturally trendy. Geo-
extremes – an indicator of climate change in Kerala, India. J. phys. Res. Lett., 2005, 32, L23402, doi: 10.1029/2005GL024476.
Hydrol., 2009, 367, 62–69. 80. Jain, S. K., Agarwal, P. K. and Singh, V. P., Hydrology and Water
61. Pal, I. and Al-Tabbaa, A., Regional changes in extreme monsoon Resources of India, Springer, The Netherlands, 2007.
rainfall deficit and excess in India. Dyn. Atmos. Oceans, 2009,
doi:10.1016/j.dynatmoce.2009.07.001
62. Pattanaik, D. R. and Rajeevan, M., Variability of extreme rainfall
events over India during southwest monsoon season. Meteorol.
Appl., 2010, 17, 88–104. Received 26 September 2011; accepted 4 October 2011

CURRENT SCIENCE, VOL. 102, NO. 1, 10 JANUARY 2012 49

View publication stats