Journal of the Geological Society of India
Drought Analysis and Its Implication in Sustainable Water Resource Management in
Barind Area, Bangladesh
--Manuscript Draft--
Manuscript Number:
JGSI-D-15-00136R2
Full Title:
Drought Analysis and Its Implication in Sustainable Water Resource Management in
Barind Area, Bangladesh
Article Type:
Research Article
Corresponding Author:
Chowdhury Sarwar Jahan, Ph.D.
University of Rajshahi
BANGLADESH
Corresponding Author Secondary
Information:
Corresponding Author's Institution:
University of Rajshahi
Corresponding Author's Secondary
Institution:
First Author:
ATM Sakiur Rahman, M.Sc.
First Author Secondary Information:
Order of Authors:
ATM Sakiur Rahman, M.Sc.
Chowdhury Sarwar Jahan, Ph.D.
Quamrul Hasan Mazumder, Ph.D.
Md. Kamruzzaman, Ph.D.
Takahiro Hosono, Ph.D.
Order of Authors Secondary Information:
Funding Information:
Abstract:
Rajshahi University Research Grant
(Fiscal Year 2015-16)
Rajshahi University Research Grant
(Fiscal Year 2015-2016)
Prof. Chowdhury Sarwar Jahan
Prof. Quamrul Hasan Mazumder
The study analyzes drought using Standardized Precipitation Index (SPI) and MannKendall (MK) Trend Test in the context of the impacts of drought on groundwater table
(GWT) in the Barind area, Bangladesh during the period 1971-2011. The area
experienced twelve moderate to extreme agricultural droughts in the years 1972, 1975,
1979, 1982, 1986, 1989, 1992, 1994, 2003, 2005, 2009 and 2010. Some of them
coincide with El Niño events. Hydrological drought also occurred almost in the same
years. However, relationship between all after drought events and El Niño is not clear.
Southern and central parts of the area frequently suffer from hydrological drought,
northern part is affected by agricultural drought. Trends in SPI values indicate that the
area has an insignificant trend towards drought, and numbers of mild and moderate
drought are increasing. GWT depth shows strong correlation with rainy season SPI
values such that GWT regaining corresponds with rising SPI values and vice versa.
However, 2000 onwards, GWT depth is continuously increasing even with positive SPI
values. This is due to over-exploitation of groundwater and changes in cropping
patterns. Agricultural practice in Barind area based on groundwater irrigation is
vulnerable to drought. Hence, adaptation measures to minimize effects of drought on
groundwater ought to be taken.
Powered by Edit orial Manager® and ProduXion Manager® from Aries Syst em s Corporat ion
Copyright Transfer Form (signed & scanned)
Click here to access/download
Copyright Transfer Form (signed & scanned)
Signed JGSI Copyright.pdf
Authors' Response to Reviewers' Comments
Dear Reviewer,
Thank you very much for reviewing the paper. We have incorporated all the corrections as pointed
out by you.
Title Page
Drought Analysis and Its Implication in Sustainable
Water Resource Management in Barind Area,
Bangladesh
1. ATM Sakiur Rahman1
Research Fellow
Institute of Environmental Science
University of Rajshahi
Rajshahi-6205, Bangladesh
Email: shakigeo@gmail.com
2. Chowdhury Sarwar Jahan2
Professor
Department of Geology and Mining
and
Pro-Vice Chancellor
University of Rajshahi
Rajshahi-6205, Bangladesh
Email: sarwar_geology@yahoo.com
3. Quamrul Hasan Mazumder2
Professor
Department of Geology and Mining
University of Rajshahi
Rajshahi-6205, Bangladesh
Email: qhm27@yahoo.com
4. Md. Kamruzzaman3
Associate Professor
Institute of Bangladesh Studies
University of Rajshahi
Rajshahi-6205, Bangladesh
Email: mkzaman.ru@gmail.com
5. Takahiro Hosono4
Associate Professor
Department of Earth Sciences
Priority Organization for Innovation and Excellence,
Kumamoto University, Kurokami 2-29-1
Kumamoto 860-8555, Japan
Email: hosono@kumamoto-u.ac.jp
Blinded Manuscript
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
Click here to view linked References
Drought Analysis and Its Implication in Sustainable
Water Resource Management in Barind Area,
Bangladesh
Abstract: The study analyzes drought using Standardized Precipitation Index (SPI) and
Mann-Kendall (MK) Trend Test in the context of the impacts of drought on groundwater
table (GWT) in the Barind area, Bangladesh during the period 1971-2011. The area
experienced twelve moderate to extreme agricultural droughts in the years 1972, 1975, 1979,
1982, 1986, 1989, 1992, 1994, 2003, 2005, 2009 and 2010. Some of them coincide with El
Niño events. Hydrological drought also occurred almost in the same years. However,
relationship between all after drought events and El Niño is not clear. Southern and central
parts of the area frequently suffer from hydrological drought, northern part is affected by
agricultural drought. Trends in SPI values indicate that the area has an insignificant trend
towards drought, and numbers of mild and moderate drought are increasing. GWT depth
shows strong correlation with rainy season SPI values such that GWT regaining corresponds
with rising SPI values and vice versa. However, 2000 onwards, GWT depth is continuously
increasing even with positive SPI values. This is due to over-exploitation of groundwater and
changes in cropping patterns. Agricultural practice in Barind area based on groundwater
irrigation is vulnerable to drought. Hence, adaptation measures to minimize effects of drought
on groundwater ought to be taken.
Keywords: Bangladesh, Barind Area, Drought Analysis, Impact of Drought on Groundwater,
Mann-Kendal Test, SPI
INTRODUCTION
Drought is a kind of extreme climatic natural disaster that makes the environment,
agriculture and social economy collectively suffer from severe damages affecting more
people than any other form of natural disaster (Wilhite, 1993 and 2000). It can affect drinking
water supply. However, agriculture is the most sensitive sector (Benitez and Domecq, 2014).
Thus, drought characterization has received much attention, a number of studies have been
carried out for the analysis of droughts. Sonmez et al. (2005) studied spatial and temporal
dimensions of drought vulnerability in Turkey using the SPI. Their results showed drought
vulnerability portraying a very diverse but consistent picture with varying time steps. The
increase in drought hazard is the result of increased frequency and severity of meteorological
drought, which then may lead to increased societal vulnerability to drought. Labedzki (2007)
studied local drought frequency in Central Poland using SPI. The results of analysis indicated
SPI greatly helps to identify and characterize local droughts. Comparing field observation
and measurements of agricultural and hydrological drought, the study showed that shorter
time step better reflect agricultural drought development than longer time step. Khan et al.
(2008) tracked and assess the impact of rainfall on water table in irrigation areas by SPI
method in Australia. The study found that SPI correlates well with GWT and can also capture
major drought patterns, and SPI values can be attributed to improvement in irrigation
management practices. Zhang et al. (2009) studied the changes in drought and wetness in the
Pearl River basin, Chain. The study detected the trends in dry and wet episodes decided by
SPI using Mann-Kendall Test. The study found that the Pearl River basin tends to be dryer in
the rainy season. Mishra et al. (2009) characterized drought based on the average rainfall time
1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
series of Kansabati basin of West Bengal in India using probabilistic approach. The study
reveals that interval time of drought increases from lower to higher step SPI series. Huang et
al., (2013) studied the spatio-temporal changes and frequency analysis of drought in the Wei
River Basin, China. The study results showed that the trends of SPI value with modified
Mann-Kendall Test indicated that the western basin has a significantly wet trend, whilst the
eastern basin has a trend towards drought. Benitez and Domecq (2014) analyzed
meteorological drought episodes in Paraguay using SPI. They found that the southern most
parts of the country are affected by severe droughts producing damage to soybean and corn
crop during the rainy season. The study also revealed that the impact assessment of a specific
drought requires involving its causes and the spatial and temporal distribution of the
precipitation anomalies.
Bangladesh suffered from nine droughts of major magnitude since its independence in
1971 and drought is a recurrent phenomenon in the northwest (Paul, 1998). Most of the
studies related to drought impact assessment mainly depend on crop production, yield loss
and no standard drought index method has been used for the assessment of droughts in
Bangladesh (Shahid and Behrawan, 2008). Shahid and Behrawan (2008) investigated the
extent and impact of droughts in the western part of Bangladesh using SPI method. The study
has been given emphasis on socio-economic and physical indicators of drought vulnerability
for Bangladesh. Some recent studies in the study area of NW Bangladesh (Fig. 1) which is
popularly known as Barind area have given emphasis on inland salinity (Mazumder et al.,
2014), long-term trend analysis of groundwater table and sustainable water resources
management (Rahman et al., 2015), stress on groundwater resources (Jahan et al., 2015) and
adapting cropping systems (Kamruzzaman et al., 2015). A number of studies have been
carried out on hydrogeology (Ahmed and Burgess, 1995; Islam and Kanumgoe, 2005),
groundwater occurrence potential (Haque et al., 2000; Azad and Bashar, 2000) and flow of
groundwater (Jahan and Ahmed, 1997) of the study area. Changes in drought or wetness in
developing countries such as Bangladesh did not receive enough concern and no
comprehensive study has been carried out on the driving forces of these events and their
impact on groundwater table in response to climate change. The objectives of this study are:
(1) To characterize the drought using SPI method and their relation with El Niño-Southern
Oscillation (ENSO) (2) To detect the trends in drought/wetness using MK Test and (3) To
find out the impact of drought on groundwater resources and its implications for sustainable
water resources management in the Barind area, NW Bangladesh.
STUDY AREA
The study area, former Greater Rajshahi district that includes newly formed ChapaiNawabganj, Naogaon and Rajshahi districts comprising 25 Upazilas (sub-district) in the NW
Bangladesh (Fig. 1). It covers an area of 7587 km2. Geographically, the area extends from
24o08/N to 25o13/N latitudes and from 88o01/ to 89o10/E longitudes. This area enjoys a
subtropical monsoon climate characterized by three seasons: cool and dry winter (Nov-Feb)
with almost no rainfall, hot and dry summer (Mar-May) and rainy season (Jun-Oct)
characterized by heavy rainfall. In the study area, the annual rainfall is 1600 mm which is less
than the national average of 2550 mm (Jahan et al., 2010). The study area, NW Bangladesh,
is characterized by two distinct landforms (1) the Barind Tract which is dissected and
undulating and (2) the floodplains (Fig. 2). The area lies in the catchment of the River Ganges
(Padma) with drainage system predominantly of the Atrai, Mahananda, Purnabhaba rivers
along with other minor seasonal streams. Except the Padma, all others are seasonal in nature.
Physiography map of the study area is shown in Fig. 2. The accretion to groundwater from
2
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
rainwater and floodwater during the monsoon results in the rise of groundwater level. After
monsoon, part of the water recharged into groundwater body gets discharged into the rivers,
streams and low-lying areas (Jahan et al., 2010).
Fig.1. Study area with meteorological, raingauge and permanent hydrograph stations
Geologically, the area is underlain by stream and inter-stream Recent and Pleistocene
sediments. Neogene sediments directly overlie the Gondwana sediments. The Barind Tract
flanked by actively subsiding regions has formed into horst block at the close of the
Pleistocene (Morgan and McIntire, 1959). Faulting is still active with vertical movement at
the rate of 0.4-1.1 mm/yr (Hoque, 1982). Aquifer system of NW Bangladesh is characterized
by single to multiple layers (two to four) of Plio-Pleistocene age (thickness 5.0-42.5 m). In
the Barind area there occurs semi-impervious clay-silt aquitard of Recent-Pleistocene period
(thickness 3.0-47.5 m) (Jahan et al., 2007). The aquifer characteristics of the area reveal (1)
lower values of transmissivity (<500 m2/day) in the central part which is suitable for domestic
water supply, (2) medium (500-1000 m2/day) and (3) higher (>1000 m2/day) transmissivity
values in the remaining part suitable for irrigation and domestic needs (Pitman, 1981).
DATA AND METHODOLOGY
There is only one meteorological station at Rajshahi in the study area. Data of this station
for the period of 1971-2011 have been collected from Bangladesh Meteorological
Department (BMD), Dhaka. There are fourteen raingauge stations in the study area. Data of
these stations for the period of 1971-2011 have also been collected from Bangladesh Water
Development Board (BWDB), Dhaka. The collected data for each station contain several
missing values. The highest missing value (4.7%) is encountered in Porsha raingauge station.
To fill missing link of rainfall, multiple imputation method has been applied. This method is
preferred to single imputation and regression imputation methods (Recha et al., 2012).
3
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
Multiple imputations do not suffer from the problem of underestimating the sampling error
because it appropriately adjusts the standard error for missing data (Enders, 2010). Moreover,
it yields complete data set for analysis. Rahman et al. (2016) prepared a groundwater table
database collected from BWDB for the period of 1991-2011. Complied groundwater table
data for 15 monitoring wells which are close to the raingauge stations have been used in the
present study. The locations of meteorological station, raingauge stations and groundwater
monitoring wells are shown in Fig. 1.
Fig. 2. Physiographic map of the study area (Brammer, 1996 and Alam, 1998)
In recent years numerous indices have been developed to detect and monitor droughts, but
the Palmer drought severity index (PDSI) and the standardized precipitation index (SPI) are
the more commonly used indices (Mishra, 2009). In this paper the SPI (McKee et al., 1993
and 1995) is employed to track drought, and assess the impact on water table in the Barind
area, NW Bangladesh. In the present study, SPI for 3, 6 and 12 months’ time steps are
computed as SPI-3 and SPI-6 normally focused on agricultural impacts and SPI-12 related to
hydrological impacts (Benitez and Domecq, 2014).
To compute SPI, historic rainfall data of each station are fitted to a gamma probability
distribution function:
� − −�/ for X > … … … … … … …
�
Here x> 0 it is the amount of precipitation, α > 0 is a shape parameter, β > 0 is a scale
parameter, and τ (α) defines the gamma function. The maximum likelihood solutions are
applied to optimally estimate the gamma distribution parameters, α and β for each station and
for each (3, 6 and 12 months) time steps:
� � =
=
Where,
=
�̅
�
+√ +
�
and … … … … … … … … … … . .
……………………………………………………….
4
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
∑ �� �̅
……………………………………….
�
Here, n is the number of precipitation observations.
This allows the rainfall distribution at the station to be effectively represented by a
mathematical cumulative probability function � :
� = ln �̅ −
�
� = ∫� �
�=
�
∫�
��
−
−�/
�……………
As gamma function is undefined for � = and = � =
> where � =
is the
probability of zero precipitation, the cumulative probability becomes:
� = + −
� ……………………………………..
Here, q is the probability of a zero. The cumulative probability � is then transformed
to the standard normal distribution to yield the SPI (McKee et al., 1993). The details
regarding the SPI calculation procedures can also be found in McKee et al. (1993 and 1995),
Guttaman (1999) and Hughes and Saunders (2002). Categories of drought based on SPI
values are given in Table 1. “R” statistical language by R Development Core Team has been
used to calculate SPI values from rainfall data sets. Drought occurrences have been calculated
by taking ratio of drought occurrences in each time step to the total drought occurrences in
the same time step and drought category (Sonmez et al., 2005).
Table 1. Drought categories based on SPI values
SPI values
Class
0 to -0.99
Mild drought
-1 to -1.49
Moderate drought
-1.50 to -1.99 Severe drought
< -2
Extreme drought
To find out the relation between El Niño-Southern Oscillation (ENSO) and SPI events,
and the National Oceanic and Atmospheric Administration (NOAA) index the Oceanic Niño
Index (ONI) have been used for identifying El Niño (warm) and La Niña (cool) events.
Details regarding the data can be found in the website (http://ggweather.com/enso/oni.htm) of
Golden Gate Weather Services, Canada. These data sets have also been used for study of
rainfall variability in Bangladesh by Ahasan et al. (2010).
The non-parametric rank based Mann-Kendall (MK) method (Mann, 1945; Kendall,
1975) is commonly used to assess the significance of monotonic trends in hydro-climatic
time series data. This test has the advantage of not assuming any distribution form for time
series data but it is powerful as its parametric competitors. The trends in SPI time series data
have also been investigated by MK Test and yielded good results (e.g., Zhang et al., 2009;
Huang et al., 2014). The test is based on the test statistic S, which is given as:
n−
n
S = ∑ ∑ sign x − x
=
= +
+
Where, sign(x − x ) = sign(R J − R ) = {
…………………
if (x − x ) >
if (x − x ) =
− if (x − x ) <
……………
The value of S indicates the trend direction as a negative value indicates decreasing trend
and positive value rising trend.
5
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
Mann-Kendall has documented that when n≥ , the test statistics S is approximately
normally distributed with mean and variance are as follows:
� S = ………………………………………
[n n−
n+ ⟧−∑m t t −
t+
i
i=1 i i
………………
Var S =
8
Where n is the number data of points, t is the number of ties for the i the value, and m is
the number of tied values. Using equations 7 and 10 the test statistic Z is computed from the
following formula:
S−
, for S >
√Var S
Z=
, for S =
………………………
S+
, for S <
{√Var S
The null hypothesis, Ho, that there is no trend in the records is either accepted or rejected
depending on whether the computed value of is less than or more than the critical value of
obtained from the area under normal curve table. At the 5% level of significance, the null
hypothesis of no trend is rejected if | |>1.965.
The groundwater table data have been used for a comparison with the SPI values. Khan et
al in 2008 performed regression and correlation analysis to find out the relation between SPI
values and groundwater table data of different irrigation areas in Australia. The study
compared SPI value of September and GWT data of this month. However, in the present
study multiple linear regression between average GWT data and rainy season’s monthly SPI
values has been performed to find out the impact of drought on GWT table as GWT attains
the highest level at the end of the rainy season in Bangladesh (UNDP, 1982; WAPRO, 2000;
BGS and DPHE, 2001; Harvey et al., 2006) which means that the rainy season rainfall plays
major role to recoup the GWT position. Thus, decrease in rainfall/ drought in any month of
rainy season or decrease of rainfall in consecutive months during rainy season may affect the
average GWT depth and hence annual minimum GWT depth will be affected. Multiple linear
regressions have been carried out by the following formula:
=
+
� +
� +
� +
� +
� ………………….
Where, is dependent variable,
is intercept,
………
are regression coefficient
and � … … … . � are independent variables.
GIS has been used as a management and decision tool by many researchers (e.g., Fischer
et al., 1996; Chappell et al., 2003) for the hydro-meteorological research. In this study
mapping is performed using the Geostatistical Analyst tool integrated into ArcGIS 9.3
software.
RESULTS AND DISCUSSION
Drought in the study area and ENSO Phases (El Niño and La Niña)
To characterize the regional droughts, monthly time series of rainfall data for the period
of 1971-2011 have been generated by average of all stations data. The standardized
precipitation indexes (SPIs) for different time steps based on the average rainfall over the
study area are shown in Fig. 3 (a-c).
6
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
In this paper, the SPI values are used as indicators of drought of different typesagricultural (SPI-3), meteorological (SPI-6) and hydrological (SPI-12). Table 2 displays SPI3 of rainy months (June to October) during the El Niño or La Niña events recorded from
1971 to 2011. For agricultural drought (SPI-3 ≤ -1), moderate to extreme droughts occurred
in the years 1972, 1975, 1979, 1982, 1986, 1989, 1992, 1994, 2003, 2005, 2009 and 2010.The
comparative analysis of the SPI-3 with El Niño or La Niña shows that in general El Niño
events are more frequently associated with moderate to extreme drought while wet events
(SPI-3 ≥ 1) are more frequently associated with La Niña events. There is a strong coincidence
between moderate to extreme drought occurrence and El Niño events in the years of 1972,
1982, 1986, 1992, 1994, and 2009, but positive SPI-3 values in the years of 1987 and 1997
show some disagreement. Seasonal average rainfall in Bangladesh shows a negative general
tendency during strong El Niño years with few discrepancies and the relation between rainfall
variability and ENSO index is very high in the Ganges basin (Chowdhury, 2003). However,
La Niña years that are consistent with positive SPI-3 are 1971, 1973, 1988, 1995, 1998, 1999,
2000, 2007 and 2011, but with negative SPI-3 values in 1975 and 2010. Strong La Niña
years, except 1975, received excess rainfall in Bangladesh (Chowdhury, 2003). Droughts also
occurred in 1979, 2003 and 2005 with no prominent El Niño or La Niña events. Thus, the
frequency and intensity of droughts is not explained only by El Niño variability in the study
area.
For meteorological drought (SPI-6 ≤ -1), moderate to extreme droughts were encountered
in 1972, 1975, 1979, 1982, 1986, 1992, 1994, 1995, 1996, 2005, 2009 and 2010. For
hydrological drought (SPI-12 ≤ -1), moderate to extreme droughts were found in the years of
1972, 1975, 1976, 1979, 1982, 1983, 1986, 1992, 1994, 1995, 2009 and 2010. The
comparative analysis of the SPI-6 and SPI-12 with El Niño or La Niña shows almost similar
results of SPI-3.
Fig.3. SPI series based on average rainfall in the study area a) SPI-3, b) SPI-6 and c) SPI-12
The Figure 4 shows that the spatial distribution of SPI-3, SPI-6 and SPI-12 for the rainy
season’s months (June to October) of different years. The map generally shows that the
7
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
southern and central parts of the area suffer from severe and extreme droughts, whereas the
north-eastern portion suffers from mild to moderate droughts. Spatial distribution of the SPI3 of June 1972 indicates that the north-eastern part suffered from severe to extreme droughts
with SPI-3 values ranging from -2.99 to -1.49 and southern part experiences mild drought,
whereas in July 1994 northern part of the area suffered severe to extreme droughts. Spatial
distribution of SPI-6 of October 1982 reveals that the southern part of the area experienced
severe to extreme droughts with SPI-6 values ranging from -2.87 to -1.49 and at the same
time northern part suffered mild while moderate droughts. At the same time step in August in
2009, the central part of the area experienced moderate to severe droughts, whereas
southeastern part experienced mild drought. SPI-12 of October 1992 indicates that the
southern part experienced severe to extreme droughts, whereas northern part experienced
mild drought. Southern and central parts suffered from moderate to extreme droughts
according to the SPI-12 of September 2010, whereas northern part experienced mild drought.
Thus southern and central parts of the area more frequently suffered from severe to extreme
hydrological droughts as severity of drought increases with increasing time steps, whereas
northern part suffered from agricultural drought as shorter time step (SPI-3) values were high.
Table 2. Correspondence between El Niño and La Niña events and SPI-3 dry and wet events
SPI-3
Running 3-Month Mean ONI values
ENSO
Jun
Jul
Aug
Sep
Oct
AMJ
MJJ
JJA
JAS
ASO
1971 1.05
0.47
0.83
0.30
0.40
-0.7
-0.7
-0.7
-0.7
-0.7
ML
1972 -1.62 -2.71 -1.17 -0.65 -0.22
0.6
0.8
1.1
1.4
1.6
SE
1973 1.96
1.78
0.54
0.20
0.79
-0.5
-0.8
-1.0
-1.2
-1.3
SL
1975 -1.48 -0.19 -0.57 -0.11 -1.16
-0.8
-1
-1.1
-1.2
-1.4
SL
1977 2.03
2.60
1.34 -0.71 -1.30
0.3
0.4
0.4
0.4
0.5
WE
1982 -0.53 -1.18 -0.38 -1.57 -1.99
0.5
0.7
0.7
1.0
1.5
SE
1986 -0.38 -0.97 -1.54 0.04
1.39
-0.1
0
0.3
0.5
0.7
WE
1987 -0.85 0.81
3.05
3.15
2.32
1.0
1.2
1.4
1.6
1.6
SE
1988 1.15
1.14
1.80
0.39
0.47
-0.8
-1.2
-1.3
-1.2
-1.3
SL
1989 -0.31 0.50 -1.15 -0.02 -0.86
-0.6
-0.4
-0.3
-0.3
-0.3
WL
1991 0.89
1.41
0.50
0.94
0.72
0.5
0.7
0.8
0.7
0.7
WE
1992 -1.69 -1.02 -0.59 -0.24 -0.77
1.0
0.7
0.3
0
-0.2
ME
1994 -0.23 -1.61 -1.36 -1.94 -0.98
0.4
0.4
0.4
0.4
0.5
WE
1995 -0.91 -0.82 0.63
2.00
2.41
0.2
0
-0.2
-0.4
-0.7
WL
1997 0.82
0.65
1.35
0.43 -0.15
0.7
1.2
1.5
1.8
2.1
SE
1998 -0.43 0.44
0.80
1.37
1.26
0.4
-0.2
-0.7
-1.0
-1.2
ML
1999 0.16
0.75
0.93
1.12
1.12
-0.9
-1.0
-1.0
-1.1
-1.1
ML
2000 1.53 -0.24 -0.75 0.62
1.31
-0.8
-0.7
-0.6
-0.5
-0.6
WL
2007 0.56
1.44
1.08
0.55 -0.34
-0.3
-0.3
-0.4
-0.6
-0.8
ML
2009 -1.23 -1.59 -1.41 -0.51 -0.21
0.2
0.4
0.5
0.6
0.8
ME
2010 -0.03 -1.27 -1.71 -2.02 -1.35
0.1
-0.4
-0.9
-1.2
-1.4
ML
2011 1.08
0.15
0.43 -0.32 -0.28
-0.3
-0.2
-0.2
-0.4
-0.6
WL
N.B: WE=Weak El Niño, ME=Moderate El Niño, SE=Strong El Niño, WL=Weak La Niña,
ML=Moderate La Niña, SL=Strong La Niña, AMJ= April, May and June; MJJ=May, June and
July; JJA= June, July and August; JAS=July, August and September; ASO=August, September and
October
Annual drought occurrences in the study area have been analyzed based on frequency of the
events for each drought category at multiple-time steps (SPI-3, SPI-6 and SPI-12) to clear the
understanding of droughts phenomena. The spatial distribution maps of annual drought
occurrences for SPI-3, SPI-6 and SPI-12 are shown in Fig. 5.
8
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
The spatial analysis of mild drought occurrences (%) indicates that they most frequently tend
to occur in the north-eastern and south-western areas at 3, 6 and 12 month time steps, while
the north-western and southern parts are characterized by the lowest frequencies at the same
time steps (Fig. 5). Moderate droughts occur more frequently in the eastern part of the study
area at 3 month time steps, and the moderate droughts exhibit some variable behavior when
the time step increases to 6 and 12-month as they tend to occur in the central part contiguous
with the western part. Annual occurrences (%) of moderate drought of SPI-3, SPI-6 and SPI12 range from 6.5 to 11.6, 7.4 to 12.9 and 6.7 to 12.5 respectively.
Fig.4 Spatial distribution of SPI values of rainy season’s months (June-October) for the
selected years
Annual occurrences (%) of severe droughts range from 2.5 to 4.7, 1.6 to 5.9 and 1.7 to 6.4
at 3, 6 and 12 months’ time steps respectively. Severe droughts also show variability with
increasing time steps. At the 3-month time step, they most frequently tend to occur in the
south-eastern parts that cover High Barind Tract, while this tendency becomes low towards
the north and very low in the north-eastern Atrai Flood Plain area. At the 6-month time steps,
it shows maximum frequency in the south, south-western and north-western parts and at 12month time steps maximum frequency occurs in the north-western High Barind area. Annual
occurrence (%) of the extreme droughts at 3, 6 and 12 months’ time steps vary from 0.6 to
2.2, 0.4 to 3.1 and 0.2 to 4.8 respectively in the study area. North western part of the High
Barind that extends up to the Central High Barind Tract exhibits maximum frequencies at
shorter time step (Fig. 6). The spatial distribution of occurrences (%) of extreme droughts
shows maximum frequency in the central part of the area that covers the whole High Barind
Tract at 6-month time step and at the 12-month time steps. As the severity of drought
increases like severe and extreme droughts for the same time step, major changes are
9
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
observed at maximum frequencies, there is a shift in the high drought occurrences toward
south and central parts of the area. In other words, majority of the historical droughts that
occurred in the southern and central part pars of Barind area are severe to extreme. These
findings also reveal that the southern and central parts of the area are more likely to be
affected from hydrological drought, with consequent very high depletion of groundwater
resources.
Fig.5 Drought occurrences (%) at 3, 6 and 12-month time steps
Temporal Drought Characterization
Results of the MK Test
The significant results of the MK Trend Test of SPI-3, SPI-6 and SPI-12 concerning
every month in the Barind area are given in Table 3. In terms of the Z value of MK Test,
when its absolute value is greater than 1.96, the trend is significant at the 5 % significance
level. The positive value of Z represents a wet trend and vice versa. The results of analysis in
Table 3 are arranged according to the climatic seasons for identifying the seasonal
characteristics. Both positive and negative trends are identified by the MK test in the SPI time
series. However, most of the trends are insignificant at the 5% significance levels. As shown,
the majority of the trends in the winter months’ time series at SPI-3, SPI-6 and SPI-12 time
steps are negative, accounting for about 65 ( total winter month time series 60 (4×15) and
negative trends found in 39 time series), 63 and 53% of the time series respectively. The
significant negative trends in different months of winter season are found at two stations
(Rajshahi, and Mahadebpur) in SPI-3 and SPI-12 series, at two stations (Rajshahi and
Rohonpur) in SPI-6 series, while significant positive trends in all winter months are found at
Nawabganj Sadar in SPI-12 series.
Similar to the winter month’s series, the summer month’s SPI series show decreasing
trends in 53 (total summer month time series 45 (3×15) and negative trends found in 24 time
series), 38 and 57% of the time series at SPI-3, SPI-6 and SPI-12 time steps, but no
significant positive and negative trends are detected in the SPI-3 and SPI-6 series by the trend
tests. However, significant positive trends are found in all months in SPI-12 series at
Nawabganj Sadar station as in the winter season and negative trends in all months in SPI-12
at Rajshahi station.
Similar to the other seasonal series, most of the trends in SPI values in rainy season’s
months are decreasing, accounting for about 61 (total rainy season month time series 75
10
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
(5×15) and negative trends found in 41 time series), 59 and 50% of the months in rainy
season in SPI-3, SPI-6 and SPI-12 time series respectively. More significant negative trends
are found in rainy season months than in the other seasonal series. Five significant negative
trends are observed at Mahadebpur, Naogaon Sadar, Sapahar and Rohonpur stations in the
SPI-3 time series, significant negative trends are also found in SPI-6 and SPI-12 series.
Again, significant positive trends in all SPI series are observed at Nawabganj Sadar station in
different rainy season’s (June-October) months. Hence, the whole study area except
southwestern corner has a tendency towards recurrence of drought. Jahan et al. (2015)
pointed out that most of the stations in Barind area are characterized by insignificant negative
annual rainfall trends and their distribution reveal that the declining trends occurred over the
area except the southwestern corner. Therefore, findings of the present study are also
consistent with other study.
Table 3. Mann-Kendall Z statistics of SPI-3, SPI-6 and SPI-12 time series
Stations
Winter Season
Summer Season
Rainy Season
Z statistics of SPI-3 time series
Nov
Rajshahi
Dec
Jan
Feb
Mar
Apr
-0.66 -0.83 -1.96* -1.67 -0.84 -0.51
May
Jun
Jul
Aug
Sep
Oct
0.00 -0.13 -0.86 -1.66 -1.52 -1.00
Mahadebpur -0.30 -0.04 -2.20* -2.25* -1.38 -1.21 -1.20 -0.64 -2.45* -1.93 -2.01 -0.64
Naogaon
-0.47 -0.54 -0.79 -1.21 -0.92 -0.03 -0.06 0.25 -1.29 -1.40 -2.01* -0.83
Sapahar
-0.47 -0.54 -0.79 -1.21 -0.92 -0.03 -0.06 0.25 -1.29 -1.40 -2.01* -0.83
Nawabganj
1.01
0.24
0.07
0.63
1.26
1.95 2.39* 2.21* 1.44
Rohonpur
-0.86
0.08
-0.46 -0.91 -0.51 -0.07
0.60
1.83 -0.26 -1.18 -2.53* -1.30
0.33
1.04
0.57
Z statistics of SPI-6 time series
Rajshahi
-2.16* -1.52 -1.48 -0.70 -0.38 -1.33 -0.69 -0.27 -0.92 -1.22
Mahadebpur -1.56 -1.76 -0.73 -0.43 0.28 -0.89 -1.31 -0.92 -2.91* -2.17*
Nawabganj 2.57* 1.56 0.78 1.48 1.04 1.45 1.89 1.80 2.58* 2.10*
Rohonpur
-1.09 -2.40* -1.54 -0.94 0.20 0.73 1.12 1.73 -0.20 -0.47
-1.47 -2.04*
-2.10* -1.63
2.38* 2.27*
-0.71 -1.09
Z statistics of SPI-12 time series
Rajshahi
-1.49 -1.83 -1.88 -1.99* -2.11* -2.47* -2.66* -1.75 -1.41 -0.86 -1.25 -1.55
Manda
1.36
0.91
0.92
1.06
0.87
1.12
1.15
1.50 1.99* 1.43
1.50
1.67
Mahadebpur -1.92* -2.03* -1.82 -1.95 -1.91 -1.78 -1.74 -1.85 -2.56* -2.60* -1.93 -1.99*
Nawabganj
2.94* 2.49* 2.35* 2.38* 2.55* 2.69* 2.41* 2.21* 2.70* 3.19* 2.92* 2.97*
Analysis of Decadal Drought Events
The numbers of drought events during different months of all the three seasons at SPI-3,
SPI-6 and SPI-12 time steps are summed up by decades to identify trends in drought events
by categories (Fig. 6). The total number of dry episodes found from SPI values is the highest
in the recent decades (2001-2010) and the number of mild and moderate drought events at
SPI-3, SPI-6 and SPI-12 time steps are increasing rapidly in rainy and summer seasons with
steady rising trends in winter seasons, though the numbers of severe and extreme drought
events have shown fluctuation. As is known, rainy season is the season of rain feed paddy
cultivation which needs a lot of water, the frequent mild and moderate droughts may cause
damage to the crop production. To minimize the loss of crop production from dry spells,
farmers need to frequently irrigate their land even during the rainy season and they are
incurring extra expenditure for this (Kamruzzaman et al., 2015). Decreasing trend of rainfall
in the Barind area has already created stress on the groundwater resources (Jahan et al.,
11
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
2015). It will inevitably intensify in the future as the area frequently suffers from mild and
moderate droughts and shows drying trends. As the area is the food granary in Bangladesh,
the rainy season is the main rice production season. It will hamper the food safety of the
country leading to the price hike of rice.
Fig.6. Cumulative decadal drought events in study area during (a) rainy season, (b) summer
and (c) winter
Impact of Drought on Groundwater Tables
Rahman et al. (2016) found that GWT reaches maximum depth in April to May and
regains its original position after rainy season (September to October) in the study area. Thus,
rainfall of rainy season plays a vital role to regains its original position. This means annual
minimum depth of groundwater table is influenced by rainy season rainfall and annual
average depth too is affected. To find out the relation between rainfall variability and depth of
GWT table, multiple correlations and multiple linear regressions between annual minimum
and annual average depth of GWT and SPI time series have been carried out for the period of
1991-2011.
12
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
Figure 7a shows depth of annual average and annual minimum GWT with rainy season’s
monthly SPI-3 values (July to October) during the period of 1991 to 2011 of Rajshahi area. In
general, the annual average and annual minimum depths of GWT are influenced by rainy
season’s monthly SPI-3 values. The depth to GWT of a particular year is deeper than before
and after this particular year with negative SPI-3 values like 1992, 1994, 1996, 2009 and
2010 and is lower with positive SPI-3 values in 1991, 1993, 1997,1998 and 1999. The results
show that the changes in SPI-3 values coincided with the changes in the depth of GWT. The
average GWT get deeper with a negative SPI value as less water is available to recharge its
(Khan et al., 2008). However, after the year 2000 depth to GWT table is continuously
increasing even with the positive SPI values during the rainy season’s months. Therefore,
there are other influencing factors on GWT depth and these are very much influencing on
GWT depth along with dry episodes after the year 2000 in Rajshahi area. Figure 7b shows the
annual average and annual minimum depth to GWT with rainy season’s monthly SPI-6
values for the 1991 to 2011 period in Naogaon area. The results indicate that similar to the
Rajshahi area changes in SPI value are related to corresponding changes in GWT depth and
post 2007 depth to GWT is continuously increasing. Figure 7c shows the annual average and
minimum depth to GWT with rainy season’s monthly SPI-12 values for the 1991 to 2011
period in Nawabganj area. In general, the results indicate that negative SPI-12 values
coincide with higher GWT depth and vice-versa. However, after the year 2001 annual
average depth to GWT table is continuously increasing even with the positive SPI values
during the rainy season as in Rajshahi. Hence, there are other influencing factors on GWT
depth along with dry episodes after the year 2001 in Nawabganj area. Though there were few
wet year after 1996 onwards, the minimum groundwater levels have not increased
proportionately. Other factors come into play, suggesting that groundwater table is also
influenced by water management practices and changes in cropping patterns. Rahman et al.
(2016) found that irrigated areas by shallow tube wells (STWs), deep tube wells (DTWs) and
PPs (Power pumps) in the area increased rapidly.
The multiple linear regression and multiple correlation analysis of SPI time series on
annual minimum depth and annual average depth of GWT have been performed and are
given in Table 4. Multiple correlation coefficient (r) in the study area ranges from 0.65 to
0.38 (Avg. = 0.50) for annual minimum depth of GWT, 0.66 to 0.31 (Avg. = 0.47) for annual
average depth of GWT in case of SPI-3, 0.77 to 0.15 (Avg. = 0.51) for annual minimum
depth of GWT, 0.81 to 0.17 (Avg. = 0.47) for annual average depth of GWT in case SPI-6,
and also 0.73 to 0.24 (Avg. = 0.50) for annual minimum depth of GWT and 0.74 to 0.23
(Avg. = 0.49) for annual average depth of GWT in case SPI-12. Multiple linear regression
coefficient (r2) in the study area ranges from 0.42 to 0.14 (Avg. = 0.26) for annual minimum
depth of GWT and 0.44 to 0.10 (Avg. = 0.23) for annual average depth of GWT in case of
SPI-3, 0.59 to 0.02 (Avg. = 0.28) for annual minimum depth of GWT and 0.66 to 0.03 (Avg.
= 0.25) for annual average depth of GWT in case SPI-6, 0.53 to 0.06 (Avg. = 0.27) for annual
minimum depth of GWT and 0.54 to 0.05 (Avg. = 0.26) for annual average depth of GWT in
case SPI-12. Multiple linear regression and multiple correlation analyses show that depth of
GWT are influenced by SPI values and the two series are positively correlated such that
regaining of GWT corresponds with the rising SPI values and vice versa. The correlation and
regression coefficients for flood plain areas are higher than the Barind Tract area (Table 4)
such as correlation and regression coefficients between groundwater observation well (Well
no-RJ084) located in the Ganges Flood Plain area and rainy season’s monthly SPI values of
Rajshahi meteorological station are 0.77 and 0.59 for annual minimum depth of GWT and
SPI-6 values. However, these values are 0.15 and 0.02 for the same cases although the
groundwater observation well (Well no RJ-052) and raingauge stations are located in the
High Barind Tract. These spatial variations of correlation and regression coefficients are
13
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
related to the depth of GWT and possible due to local groundwater dynamics, groundwater
recharge and discharge zones (Khan et al., 2008). As it is known that depth of GWT is higher
in the Barind Tract than the Flood plain areas (Rahman et al., 2016) and hydraulic
conductivity is low (Jahan and Ahmed, 1997), the correlation and regression coefficients are
lower in the Barind Tract area, illustrating that the climatic impacts are greater in flood plain
areas where groundwater table is shallow. Thus overall results suggest that depth of GWT is
influenced by rainfall, and hence by drought, which have implications for sustainable
groundwater management schemes.
Fig.7. Comparison of annual average changes in GWT depth and rainy season’s months (JunOct) SPI values (a) SPI-3 and GWT of Rajshahi area, (b) SPI-6 and GWT of Naogaon
area, and (c) SPI-12 and GWT of Nawabganj area
CONCLUSIONS
This study has analyzed droughts and impacts of drought on groundwater in the Barind area
using the SPI, MK trend test and multiple linear regression methods. The results of the
analysis indicate that the study area suffered from twelve droughts since independence of
14
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
Bangladesh in 1971. The relationship between drought and El Niño shows that in general El
Niño events are more frequently associated with drought. Droughts also occurred with no
prominent El Niño or La Niña events, suggesting the necessity of new research, focusing on
new drivers to explain the cause of the droughts in the study area.
Table 4. Multiple correlation and multiple linear regression coefficients of rainy season’s SPI
values and annual minimum and annual average depth of GWT for the period of 1991-2011
Stations
Rajshahi
Godagari
Tanore
Atrai
Badalgachi
Manda
Mahadebpur
Porsha
Naogaon
Shapahar
Bholahat
Nachole
Nawabgonj
Gomostapur
Shibganj
Average
SPI-3 &
Min GWT
r
r2
0.65
0.42
0.49
0.24
0.38
0.14
0.47
0.22
0.55
0.30
0.53
0.28
0.44
0.19
0.41
0.17
0.55
0.30
0.53
0.28
0.47
0.22
0.41
0.17
0.45
0.20
0.57
0.32
0.57
0.33
0.50
0.26
SPI-3 &
Avg. GWT
r
r2
0.66 0.44
0.46 0.21
0.36 0.13
0.50 0.25
0.48 0.23
0.36 0.13
0.34 0.11
0.38 0.15
0.52 0.27
0.56 0.31
0.59 0.35
0.31 0.10
0.33 0.11
0.54 0.29
0.63 0.40
0.47 0.23
SPI-6 &
Min GWT
r
r2
0.77 0.59
0.40 0.16
0.44 0.19
0.50 0.25
0.55 0.30
0.52 0.27
0.45 0.20
0.15 0.02
0.49 0.24
0.73 0.54
0.67 0.45
0.31 0.10
0.44 0.19
0.53 0.28
0.61 0.38
0.51 0.28
SPI-6 &
Avg. GWT
r
r2
0.72 0.52
0.31 0.10
0.44 0.19
0.55 0.31
0.55 0.30
0.34 0.12
0.39 0.15
0.21 0.04
0.55 0.31
0.69 0.48
0.81 0.66
0.17 0.03
0.25 0.06
0.52 0.27
0.52 0.27
0.47 0.25
SPI-12 &
Min GWT
r
r2
0.71 0.50
0.68 0.46
0.27 0.07
0.55 0.30
0.60 0.36
0.33 0.11
0.57 0.33
0.34 0.12
0.57 0.33
0.24 0.06
0.54 0.29
0.32 0.10
0.39 0.15
0.55 0.30
0.73 0.53
0.50 0.27
SPI-12 &
Avg. GWT
r
r2
0.54 0.29
0.63 0.40
0.30 0.09
0.74 0.54
0.64 0.41
0.30 0.09
0.62 0.38
0.31 0.10
0.58 0.33
0.25 0.06
0.58 0.33
0.23 0.05
0.33 0.11
0.65 0.42
0.53 0.28
0.49 0.26
The spatial distribution of SPI values of different drought years suggested that southern
and central parts of the area suffer from severe and extreme droughts, whereas north-eastern
portion suffers from mild to moderate droughts. Annual occurrences (%) of drought also
indicate that mild and moderate droughts most frequently tend to occur in the north-eastern
and south-western parts, while the southern and central parts of the area are affected by
severe to extreme droughts with increasing time steps. These findings confirm that at longer
time steps hydrologic drought is likely to occur in the southern and central parts, while the
north-eastern part suffers from agricultural drought.
The MK trend analysis of SPI time series indicates that the study area except the
southwestern portion has an insignificantly dry trend, accounting for 61, 59 and 50% of the
time series of rainy season months at SPI-3, SPI-6 and SPI-12 time steps respectively. Mild
and moderate droughts are increasing rapidly in the rainy and summer seasons with steady
raising trends in the winter seasons, though the numbers of severe and extreme drought
events show fluctuation. It is obvious that drought phenomenon will create more vulnerable
environment for the agricultural sector in the food granary of Bangladesh and water resources
in the Barind area.
For further understanding the impact of drought on GWT, the multiple linear regression
coefficient and correlation between annual average and minimum have been calculated for
the period of 1991-2011. The results show that groundwater table are influenced by SPI vales
and the two series were positively correlated such that regaining of groundwater table
correspond with the rising SPI values and vice versa. However, 2000 or onward, depth to
GWT table is continuously increasing even with the positive SPI values during the rainy
15
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
season’s months. This is due to the over-exploitation of groundwater in the area by deep tubewells and changes in cropping pattern.
The overall results suggest that droughts have impacts on groundwater levels, and have
implications for sustainable agriculture in terms of cropping patterns and drought assistance
schemes. Therefore, the water resources managers need to pay high attention to the potential
impact of drought and explore in advance scientific and effective drought adaptation
measures to minimize loss.
References
AHASAN, M.N., CHOWDHARY, M.A.M and QUADIR, D.A. (2010) Variability and trends of
summer monsoon rainfall over Bangladesh. Jour. Hydro. Meteo., v.7, no.1, pp.1-17.
AHMED, K. and BURGESS, W. (1995) Bils and the Barind aquifer, Bangladesh. In: Brown,
A.G. (Ed.) Geomorphology and groundwater. Wiley, New York
ALAM, M.S. (1998) Paleoclimatic impact on the flood basin accretion and paleosol
development in northwestern Bangladesh. Jour. Nepal. Geo. Soc., v.18, pp.227-238.
AZAD, M.A.S. and BASHAR, K. (2000) Groundwater zonation of Nawabganj Sadar Thana and
its relation to groundwater chemistry. Bangladesh Jour. Geol., v.19, pp.57-71.
BENITEZ, J.B. and DOMECQ, R.M. (2014) Analysis of meteorological drought episodes in
Paraguay. Clim. Change, v.127, pp.15-25. DOI 10.1007/s10584-014-1260-7
BGS and DPHE (BRITISH GEOLOGICAL SURVEY and DEPARTMENT OF PUBLIC HEALTH
ENGINEERING), (2001) Arsenic contamination of groundwater in Bangladesh, Keyworth,
UK.
BRAMMER, H. (1996) The Geography of the Soils of Bangladesh, 1st Ed, The University
Press Limited.
CHAPPELL, A., HERITAGE, G.L., FULLER, I.C., LARGE, A.R.G. and MILAN, D.J. (2003)
Geostatistical analysis of ground-survey elevation data to elucidate spatial and temporal
river channel change. Earth Surf. Proc. Land, v.28, pp.349-370.
CHOWDHURY, A.R. (2003) The El Niño-Southern Oscillation (ENSO) and seasonal floodingBangladesh. Theor. App. Climatol., v.76, no.1-2, pp.105-124. DOI:10.1007/s00704-0030001-z
ENDERS, C.K. (2010) Applied missing data analysis, The Guilford Press, New York, ISBN
978-1 60623-639-0, retrieved from http://books.google.co.ke/on 30 Jul 2011.
FISCHER, M.M., SCHOLTEN, H.J. and UNWIN, D.J. (1996) Spatial analytical perspectives on
GIS. Taylor & Francis Ltd, London.
GUTTMAN, N.B. (1999) Accepting the standardized precipitation. Jour. Amer. Water Resour.
Assoc., v.35, no.2, pp.311-322.
HAQUE, M.N., KERAMAT, M. and RAHMAN, A.M.A. (2000) Delineation of groundwater
potential zones in the western Barind Tract of Bangladesh. J. Bangladesh Natl. Geogr.
Assoc., v.21-26, pp.13-20.
HARVEY, C.F., ASHFAQUE, K.N., YU, W., BADRUZZAMAN, A.B.M., ALI, M.A., OATES, P.M.,
MICHAEL, H.A., NEUMANN, R.B., BECKIE, R., ISLAM, S. and AHMED, M.F. (2006)
Groundwater dynamics and arsenic contamination in Bangladesh. Chem. Geol., v.228,
pp.112-136.
HOQUE, M. (1982) Tectonic set up of Bangladesh and its relation to hydrocarbon
accumulation, Phase-I: Center for Policy Research, Dhaka University and Universities
Field Staff International (UFSI), USA publication, 177p.
16
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
HUANG, S., CHANG, J., HUANG, Q., and CHEN, Y. (2014) Spatio-temporal changes and
frequency analysis of drought in the Wei River Basin, China. Water Resour. Manag.,
v.28, no.10, pp.3095-3110. DOI 10.1007/s11269-014-0657-4
HUGHES, B.L. and SAUNDERS, M.A. (2002) A drought climatology for Europe. Int. J. Climat.,
v.22, no.13, pp.1571-1592.
ISLAM, M.M. and KANUMGOE P. (2005) Natural recharge to sustainable yield from the Barind
aquifer: a tool in preparing effective management plan of groundwater resources. Water
Sci. Technol., v.52, no.12, pp.251-258.
JAHAN, C.S. and AHMED, M. (1997) Flow of groundwater in the Barind area, Bangladesh:
implication of structural framework. J. Geol. Soc. India, v.50, pp.743-752.
JAHAN, C.S., ISLAM, M.A., MAZUMDER, Q.H., ASADUZZAMAN, M., ISLAM, M.M., ISLAM, M.O.
and SULTANA, A. (2007) Evaluation of depositional environment and aquifer condition in
the Barind Area, Bangladesh, using Gamma Ray Well Log data. J. Geol. Soc. India, v.70,
pp.1070-1076.
JAHAN, C.S., MAZUMDER, Q.H., GHOSE, S.K. and ASADUZZAMAN, M. (1994) Specific yield
evaluation: Barind area, Bangladesh. J. Geol. Soc. India, v.44, pp.283-290.
JAHAN, C.S., MAZUMDER, Q.H., ISLAM, A.T.M.M. and ADHAM, M.I. (2010) Impact of
irrigation in Barind Area, NW Bangladesh - An evaluation based on the meteorological
parameters and fluctuation trend in groundwater table. J. Geol. Soc. India, v.76, pp.134142.
JAHAN, C.S., MAZUMDER, Q.H., KAMRUZZAMAN, M. and RAHMAN, A.T.M.S. (2015) Stress on
Groundwater Resource in Drought Prone Barind Area, Bangladesh: Study of Climate
Change and Irrigation Effect. In: M. Thangarajan, C. Mayilswami, P.S. Kulkarni and V.P.
Singh (Eds.) Groundwater Prospecting, Evaluation and Management of Aquifers,
Cambridge Scholars Publishing, UK. Accepted on 18 January, 2015.
KAMRUZZAMAN, M., RAHMAN, A.T.M.S. and JAHAN, C.S. (2015) Adapting Cropping
Systems under Changing Climate in NW Bangladesh. Lambert Academic Publishing,
Germany, ISBN- 978-3-659-69174-4.
KENDALL, M.G. (1975) Rank Correlation Methods. Griffin, London.
KHAN, S., GABRIEL, H.F. and RANA, T. (2008) Standard precipitation index to track drought
and assess impact of rainfall on watertables in irrigation areas. Irrig. Drainage Sys., v.22,
no.2, pp.159-177. DOI 10.1007/s10795-008-9049-3.
LABEDZKI, L. (2007) Estimation of local drought frequency in central Poland using the
standardized precipitation index SPI. Irrig. Drainage, v.56, no.1, pp.67-77. DOI:
10.1002/ird.285
MANN, H.B. (1945) Nonparametric tests against trend, Econometrica, v.13, pp.245-259.
MAZUMDER, Q.H., JAHAN, C.S., MAZUMDER F., ISLAM M.A., JAMAN S., ALI M.N., RAHMAN,
A.T.M.S., AREFIN M.R. and AHASAN A. (2014) Geoelectric model and hydrochemistry of
salinity affected lower Atrai Floodplain aquifer, NW Bangladesh: An approach for
irrigation management. J. Geol. Soc. India, v.84, pp.431-441.
MCKEE, T.B., DOESKEN, N.J. and KLEIST, J. (1993) The relationship of drought frequency and
duration to time scales. In: Proceedings of eighth conference on applied climatology,
American Meteorological Society, Jan 17-23, 1993, Anaheim CA.
MCKEE, T.B., DOESKEN, N.J. and KLEIST, J. (1995) Drought monitoring with multiple time
scales. In: Proceedings of 9th conference on applied climatology, Dallas, TX.
MISHRA, A.K., SINGH, V.P. and DESAI, V.R. (2009) Drought characterization: a probabilistic
approach. Stoch. Environ. Res. Risk Assess., v.23, no.1, pp.41-55. DOI 10.1007/s00477007-0194 2.
MORGAN, J.P. and MCINTIRE, W.G. (1959) Quaternary geology of Bengal Basin, East
Pakistan and India. Geological Society of America Bulletin, v.70, pp.319-342.
17
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
PAUL, B.K. (1998) Coping mechanisms practiced by drought victims (1994/5) in North
Bengal, Bangladesh. Appl. Geogr., v.18, no.4, pp.355-373.
PITMAN, G.T.K. (1981) Aquifer and Recharges Evaluation in Bangladesh, Technical Note
No.8 UNDP/UNDTCD Groundwater Survey, BGD/74/009, BWDB, GWC.
RAHMAN, A.T.M.S., KAMRUZZAMAN, M., JAHAN, C.S. and MAZUMDER, Q.H. (2016) LongTerm trend analysis of water table using 'MAKESENS' model and sustainability of
groundwater resources in drought prone Barind Area, NW Bangladesh. J. Geol. Soc.
India., v.87, no.2, pp.179-193.
RECHA, C.W., MAKOKHA, G.L., TRAORE, P.S., SHISANYA, C., LODOUN, T. and SAKO, A.
(2012) Determination of seasonal rainfall variability, onset and cessation in semi-arid
Tharaka district, Kenya, Theor. App. Climatol., v.108, no.3-4, pp.479-494. doi:
10.1007/s00704-011-0544-3.
SHAHID, S. and BEHRAWAN, H. (2008) Drought risk assessment in the western part of
Bangladesh. Nat Hazards, v.46, no.3, pp.391-413.
SONMEZ, F.K., KOMUSCU, A.U., ERKAN, A. and TURGU, E. (2005) An Analysis of Spatial and
Temporal Dimension of Drought Vulnerability in Turkey Using the Standardized
Precipitation Index. Nat. Hazards, V.35, PP-243-264. DOI 10.1007/S11069-004-5704-7.
UNDP (UNITED NATIONS DEVELOPMENT PROGRAMME), (1982) Groundwater Survey: the
Hydrogeological Conditions of Bangladesh, United Nations Development Programme
(UNDP), Technical Report DP/UN/BGD-74-009/1, New York, pp.113.
WARPO (WATER RESOURCES PLANNING ORGANIZATION), (2000) National Water
Management Plan Project, Draft Development Strategy, Main final, Vol. 2, Water
Resources Planning Organization (WARPO), Dhaka.
WILHITE, D.A. (1993) Drought assessment, management and planning: theory and case
studies. Kluwer Academic Publishers, USA, pp.293.
WILHITE, D.A. (2000) Drought as a natural hazard: Concepts and definitions, Chapter 1, In:
D.A. Wilhite (Ed.), Drought: A Global Assessment, Natural Hazards and Disasters Series,
Routledge Publishers, UK.
ZHANG, Q., XU, C-Y. and ZHANG, Z. (2009) Observed changes of drought/wetness episodes
in the Pearl River basin, China, using the standardized precipitation index and aridity
index. Theor. App. Climatol., v.98, no.1-2, pp.89-99.
18