Impact of dam on inundation regime of flood plain wetland of punarbhaba river basin of barind tract of Indo-Bangladesh

International Soil and Water Conservation Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎

HOSTED BY

Contents lists available at ScienceDirect

International Soil and Water Conservation Research journal homepage: www.elsevier.com/locate/iswcr

Original Research Article

Impact of dam on inundation regime of flood plain wetland of punarbhaba river basin of barind tract of Indo-Bangladesh Swapan Talukdar a,n, Swades Pal b a b

Department of Geography, University of Gour Banga, Malda, India Department of Geography, University of Gour Banga, Malda, India

art ic l e i nf o

a b s t r a c t

Article history: Received 31 January 2017 Accepted 4 May 2017

Present study raises a serious issue of wetland loss and transformation due to damming and water diversion. At present study, it is noticed that overall rainfall trend (
0.006) of the study period (1978– 2015) remains unchanged but riparian wetland area is attenuated after damming both pre monsoon (March to May) and post monsoon season (October to December). Total wetland area in pre- and postmonsoon seasons is respectively reduced from 42.2 km2 to 27.87 km2. and from 277.86 sq.km. to 220.90 sq.km. in post dam period. Transformation of frequently inundated wetland area into sparsely inundated wetland is mainly triggered by flow modification due to installation of Komardanga dam and Barrage over Punarbhaba and its major tributary Tangon river. Sparsely inundated seasonal wetland area is rapidly reclaimed for agricultural practice. This extreme issue will invite instability in socio-ecological setup of the neighbouring region. & 2017 International Research and Training Center on Erosion and Sedimentation and China Water and Power Press. Production and Hosting by Elsevier B.V. This is an open access article under the CC BY-NCND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Keywords: Delimitation of wetland Damming on river Seasonal regime Flow modification Flood plain modification Wetland deterioration

1. Introduction Impact of dam on hydrological alteration of river flow modification is well documented in varied spatial scale and unit by a host of scholars (Leopold, Wolman, & Miller 1964; Petts, 1984; Williams & Wolman, 1984; Graf, 1999, 2006; Brandt, 2000; Thoms & Sheldon, 2000; Gain & Giupponi, 2014; Pal, 2015; Pal, 2016a,, 2016b). Most of the works established the fact that in dam after condition flow volume is reduced and sometimes altered flow is sub critical in reference to environmental flow (Adel, 2001; Richter, Baumgartner, Wigington, & Braun, 1997; Olden & Poff, 2002; & Gain & Giupponi, 2015; Pal, 2016b). They also documented the socio-ecological impact of dam in its downstream reach, growing water scarcity and ecological modification and re-adjustment. Water development, mostly related to dams and diversions, contributed to the declines of more threatened and endangered species than any other resource-related activity (Losos, Hayes, Phillips, Wilcove, & Alkire, 1995). Adel (2001) has carried out a detailed investigation on Farakka water diversion-induced social-ecological consequences. Apart from hydro-geomorphological modifications n Correspondence to: University of Gour Banga, Mokdumpur, Englishbazar, Malda 732101, India. E-mail addresses: [email protected] (S. Talukdar), [email protected] (S. Pal). Peer review under responsibility of International Research and Training Center on Erosion and Sedimentation and China Water and Power Press.

of channel due to dam installation, its impact on flood plain, flood plain connectivity, nutrient exchange between river and flood plain etc. have also documented by the scholars (Guy, 1981; Ward and Stanford, 1995; Finger, Schmid, & Wüest, 2006, 2007; McCartney, 2009). First and foremost impact of water level lowering in flood plain region is decreasing flood plain command area as well as quality habitat for flood plain ecology (Light, Vincent, Darst, & Price, 2006; Ligon, Dietrich, & Trush, 1995). Most explicitly, it ultimately hampers the renewal of soil fertility and restricts free fish flow to the flood plain, diversity of fish production (Grift, Buijse, Van Densen, & Klein Breteler, 2001) on which primary producers of the human ecological tropics are depended on. Along with dam and reservoir induced hydrological alteration, climate change effects on flow change in Lower part of Brhmaputra river basin is also investigated by Gain and Giupponi (2015). However, the study is concerned addressing different drivers of flow change; dam is thought to be the most dominant driver exerts most explicit and well detected impacts. Study related to connectivity of main river with its flood plain reveal that loss of connectivity compel dearth of water scarcity in the flood plain. Flood plain wetlands are mostly depending on such timing, frequency and magnitude of inundation. Due fast change of flood plain character in last 50 years, more than 70% flood plain wetland is either lost or converted to other forms (Constanza et al., 1997). However, wetland is considered as a heart of the environment, lungs of the hydrological system because it helps to purify natural water and biodiversity supermarket because of it is a good abode

http://dx.doi.org/10.1016/j.iswcr.2017.05.003 2095-6339/& 2017 International Research and Training Center on Erosion and Sedimentation and China Water and Power Press. Production and Hosting by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Please cite this article as: Talukdar, S., & Pal, S. International Soil and Water Conservation Research (2017), http://dx.doi.org/10.1016/j. iswcr.2017.05.003i

S. Talukdar, S. Pal / International Soil and Water Conservation Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎

2

of diversified species specially rare and endangered species (Cowardin, Carter, Golet, & LaRoe, 1979a). Most important contribution of wetland is its irreplaceable services like water purification, nutrient recycling, recharging of water, carbon sequestration etc. (Curie, Gaillard, Ducharne, & Bendjoudi, 2007). But in most part of the world there is no definite boundary of wetlands as it is a transitional part of terrestrial and aquatic land. In US, Australia, New Zealand some relevant works addressing wetland delimitation and conversion are carried out by Cowardin, Carter, Golet, and LaRoe (1979b); Tiner, (1990, 1993); Adam (1992); Johnston and Barson (1993); Clarkson (2013); Johnson and Gerbeaux (2004). In Indian sub continent, estimate of wetland and wetland loss is found at large scale from the report of NRSA. But small scale work is very rare and these are confined in some regional pockets of India where large size recognized wetlands are found, more specifically, these are based on some wetlands of national and international importance. For example, East Kolkata wetland is well reported by Kundu, Pal, and Saha (2008); National park of Assam is studied by Parihar, Panigrahy, and Parihar (1986), Mathur et al. (2005). In most of the large-scale studies reveals total wetland area of India shows diversified results, as NRSA (1998) shows total wetland area is 5.31 3 m ha, MoEE (1990) reports total wetland area is 15.3 m ha. Seasonal fluctuation of wetland area and high rainfall dependency on expansion and squeezing of wetland withstand against accurate estimation of stable wetland area and so wetland characters is getting changed with fast rate and it is difficult to provide a viable wetland map (Borro, Morandeira, Salvia, Minotti, Perna & Kandus, 2014). Therefore, studies is also carried out by Ji, Zhang, and Wylie (2009); Liu, Song, Peng, and Ye (2012); Wu, Lane, and Liu (2014) in aim to provide more stable base of wetland area delimitation and estimate in highly dynamic wetland area extraction. Most of them adopted wetland area extraction indices like Normalized Difference Water Index (NDWI) (Gao, 1996; McFeeters, 1996), Normalized Difference Moisture Index (NDMI) (Wilson & Sader, 2002), Modified Normalized Difference Water Index (MNDWI) (Xu, 2006), Water Ratio Index (WRI) (Shen & Li, 2010), Normalized Difference Vegetation Index (NDVI) (Rouse, Haas, Schell, & Deering, 1973) and Automated Water Extraction Index (AWEI) (Feyisaa, Meilbya, Fensholtb, & Proudb, 2014) etc. for separating water bodies from non water bodies and executed frequency approach for providing more reliable wetland area. Enormous amounts of data at different spatial, spectral, and temporal resolutions are provided by different remote sensing satellites. These data have become primary sources and used extensively for extracting and detecting surface water and its changes throughout decades (Li et al., 2013; Tang, Ou, Dai, & Xin, 2013; Xu, 2006; Zhou, Hong, & Huang, 2011). In the present work, thrust is mainly given on delimiting seasonal pattern of wetland and investigating impact of rainfall and dam or reservoir induced river flow modification on spatial dynamics of flood plain wetlands. Detected change is also tested for establishing exact role of dam on changing spatial extent of wetland in seasonal scale.

pre monsoon (March-May) rainfall is 0.00. Monsoon (June–September) is
0.046, and Post-monsoon (October- December) is 0.025. So it can be said that there is no significant changes of rainfall (1978–2014). Therefore, catchment agricultural area depends on either river water or subsurface water for irrigation purpose. Alluvial and laterite soil dominates the entire basin area. Inundation process renews soil make it suitable for intensive agriculture. For mitigating seasonal water scarcity for irrigation, Komardanga dam has been constructed over Punarbhaba river in 1992.

3. Materials and methods 3.1. Data used Landsat-TM images represent valuable and continuous records of the earth's surface during the last 3 decades (USGS, 2014). Moreover, the entire Landsat archive is now available free-ofcharge to the scientific public, which represents a wealth of information for identifying and monitoring changes in manmade and physical environments (Chander, Markham, & Helder, 2009; El Bastawesy, 2014). LANDSAT TM and LANDSAT 8 OLI have been obtained from the US Geological Survey (USGS) Global Visualization Viewer. The obtained Landsat data (Level 1 Terrain Corrected (L1T) product) were pregeoreferenced to UTM zone 45 North projection using WGS-84 datum. Detail specification of the satellite imageries are mentioned in Table 1. The other necessary corrections have been carried out in this study. The basin area is delineated using Google earth. Arc GIS 10.1 and ERDAS IMAGINE 9.2 are used for the entire study. Base map of this present study has been prepared from Google earth imageries, 2014. Three hours interval water level data since 1978–2014 have been collected from Haripur Gauge station (lat. 24°53'24"N and long. 88°19'16"E) at Malda (under Irrigation & Waterways Deptt. Govt. of West Bengal) over Punarbhaba river for analyzing seasonal hydrological divide in reference to the installation of dam. 3.2. Method for detecting seasonal flow regime Water level data for pre monsoon (March to May) and post monsoon (October to February) periods have been considered. Flow regime of pre and post reservoir is being plotted on line graphs to illustrate season wise differences of flow following Richard and Julien (2003); Batalla, Gomez, and Kondolf (2004); Rödel and Hoffmann (2005); Graf (2006); Pal (2016b). Seasonal discharge gap between pre and post reservoir conditions have been calculated for both the seasons. Least square regression models for pre and post reservoir conditions have been generated and concerned coefficient of determination (R2) have been calculated. Seasonal flow instability (IX) is also being calculated following Cuddy and Della Valle (1978) with the Eq. (1).

IX = CV x 1 − R2

(1)

2

2. Study area Punarbhaba River (length: 160 km.) basin, covering an area 5265.93 sq. km, is a sub basin of Mahananda River majorly in the Barind tract of India and Bangladesh. Elevation of this basin ranges from 89 m (at the source region) to 12 m (at the confluence) (Fig. 1). The average pre-monsoon (March-May) rainfall is 14.46% of annual rainfall, monsoon (June-September) is 70.16%, and postmonsoon (October-December) is 12.24%, with annual average rainfall ranges from 257.42 cm to 508.05 cm. The trend of average

Where, R is coefficient of determination and CV is Coefficient of variation of selected time series discharge. Less IX value indicates less instability and vice versa. 3.3. Methods for wetland detection Numerous methods are available for detecting water body and wetland from Landsat images e.g. NDVI (Townshend & Justice, 1986), NDWI (McFeeters, 1996), MNDWI (Xu, 2006), WI etc. Not all these methods are equally applicable for all spatio-temporal scale (Ji et al., 2009; Xu, 2006). The differentiation of water surface area

Please cite this article as: Talukdar, S., & Pal, S. International Soil and Water Conservation Research (2017), http://dx.doi.org/10.1016/j. iswcr.2017.05.003i

S. Talukdar, S. Pal / International Soil and Water Conservation Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎

3

Fig. 1. Shows Punarbhaba river basin with drainage networks, river gauge station and location of dam. Table 1 Presents the specifications of Landsat TM, ETM þ , OLI images. Satellite

Sensor

Path/Row

Year

Resolution (m)

Wavelength (μm)

Landsat-5

TM

139/42,43

1980–2010

30

Band Band Band Band Band Band

Landsat- 8

(Thematic

139/42,43

2014

30

Band 1: 0.435–0.451

(For band 8 resolution is 15 m)

Band 2: 0.452–0.512 Band 3: 0.533–0.590

Mapper) OLI

Band Band Band Band Band Band Band Band

(Operational Land Imager) and TIRS (Thermal Infrared Sensor)

from other terrestrial feature is difficult task using a common threshold value because of its dynamic nature that changes depending on the sub pixel land-cover component (Ji et al., 2009), specially in the plain riparian region. Therefore, an optimal threshold value between the two significantly different features (water and non-water) is determined from a histogram by analyzing the frequency distribution of grey levels in the image (Liu et al., 2012; Wu et al., 2014). Das and Pal (2016) identified NDWI is the best-fitted methods for water body extraction from Landsat imageries and defined optimum threshold for separating water body from non-water body. Frequency approach is adopted for classifying wetland into different vulnerability classes. Frequency approach was first time adopted by Borro et al. (2014) and thereon a good number of scholars namely Sun, Markthyer, Renard, and Lang (2014); Nguyen, Gaume, and Payrastre (2013); Zamman et al.

1: 0.45–0.52 2: 0.52–0.60 3: 0.63–0.69 4: 0.76–0.90 5: 1.55–1.75 7: 2.08–2.35

4: 0.636–0.673 5: 0.851–0.879 6: 1.566–1.651 7: 2.107–2.294 8: 0.50–0.68 9: 1.363–1.384 10: 10.60–11.19 11: 11.50–12.51

(2012) have applied this approach for different purposes. Kayastha, Thomas, Galbraith, and Banskota (2012) applied this technique for detecting dependable classification of wetland in reference to water availability over time, Brown and Matlock (2011) used it for vulnerability and water scarcity classification. Li et al. (2015) applied this technique for seasonal wetland cover area. 3.4. Detection of seasonal water presence area We estimated the water presence frequency of each pixel in seasonally-flooded wetland map using frequency approach (Eq. (2)). N

WPFsj =

∑i =s 1 BFpij Ns

(2)

Please cite this article as: Talukdar, S., & Pal, S. International Soil and Water Conservation Research (2017), http://dx.doi.org/10.1016/j. iswcr.2017.05.003i

S. Talukdar, S. Pal / International Soil and Water Conservation Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎

4

Where, WPFsj is the water presence frequency value of the pixel j for the sets and represents the proportion of images 1 with WPF equal to 1 in the pixel j (WPFij) in relation to the total number of images in the analyzed set of images (Ns). The pixel value of each WPF map was reclassified into three equal frequency class zones: low frequency class (40.33), moderate frequency class (0.33–0.66) and high frequency class (40.66). It is difficult to delineate wetland area in the flood plain region, as its character is highly dynamic. For relatively stable delimitation of wetland area, consideration and aggregation of a number of year representing wetland area is required and it is done here. To identify seasonal variation of wetland area aggregation of satellite imageries for pre and post monsoon seasons for different years separately have done. In this case as maximum number of satellite imageries available for a particular period, its result will be more creditable. However, data-sparse environments those are common in developing countries (Sanyal, Densmore, & Carbonneau, 2014). For identifying the impact of dam, we have subdivided entire temporal spectrum of dataset for each season into pre and post dam temporal dataset. For viable finding, seasonal variation of wetland area in pre and post dam conditions has been shown from composite WPF maps. Table 2 showing the frequency structure of the dataset used for different seasons both during pre and post dam condition. 3.5. Assessing temporal shift in distribution pattern of water presence area Chi square test, (Pearson, 1900) as a non-parametric testing method, is applied here for assessing the pattern of water presence area distribution in wetland area (Eq. (3)). From this distribution pattern, it is tried to find out whether installation of dam imparts any effect on biasing the pattern of water presence area distribution. Both pre and post monsoon data is used individually for testing the parameter. Another contingency table depicting number of year before and after dam during pre and post monsoon period were above and below mean water presence area and same test is conducted for capturing the significance of the detected change. Phi coefficient (Eq. (4)) and Coefficient of contingency (Eq. (5)) are respectively used for above mentioned cases for showing the magnitude of association between the parameters. Mukhopadhyay and Pal (2011) used this test for flood frequency analysis in post Massanjore dam condition in Mayurakshi river.

χ2 =

∑

(Oij − Eij )2 Eij

(3)

Where, X2 represent chi square, Oij and Eij are the observed and estimated frequencies respectively of the ith class. 3.5.1. Phi coefficient (Φ)

φ=

χ2 N

(4) 2

Where, X represent chi square and N means the number of observation. 3.5.2. Coefficient of contingency (C)

C=

χ2 χ +N 2

(5)

Where, χ represent chi square and N means the number of observation. 2

Table 2 Landsat image dataset used for different seasons both during pre and post dam conditions. Season

Pre dam period (1980– 1991)

Post dam period (1992–2015)

Pre monsoon (March-May)

1980, 1982, 1983, 1985, 1986, 1987, 1988, 1989, 1990, 1991 Total frequency ¼ 10 1980, 1983, 1984, 1985, 1986, 1987, 1988, 1989, 1990, 1991 Total frequency ¼ 10

1993, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2004, 2005, 2006, 2008, 2009, 2010, 2014 Total frequency¼ 17 1993, 1994, 1996, 1999, 2000, 2003, 2004, 2006, 2008, 2009, 2010, 2012, 2014, 2015 Total frequency¼ 18

Post Monsoon (October-December)

3.6. Assessing role of dam on wetland area variability Rainfall is the major variable for determining water presence area. In the riparian wetland, more rainfall actually inflates wetland area and vice-versa. But, if such fact is not found, other determining factors of water presence area should be considered as dominant factor. For reaching this aim, simple linear regression, range of variability approach (RVA), anomaly estimation approach have been adopted. 1) Simple linear regressions between rainfall and water presence area; water level of river and water presence area has separately been calculated. It is considered that if rainfall strongly control water presence area, not water level of the river, then role of the dam on controlling wetland area will be nullified and vice versa. 2) The wetland hydrological (rainfall-flow-water presence area trinity) variability and its associated characteristics (timing, frequency, duration and rates of change) play decisive role in sustaining aquatic ecosystem (Poff & Ward, 1989; Walker, Sheldon, & Puckridge, 1995; Richter et al., 1996 for detail). Natural phenomena like rainfall, river flow and dependent wetland area are not entirely stable. At the same time, variation of the intensity of these variables does not indicate instability. Where the factors are random, range of variability approach (RVA) (Richter et al., 1997) can be applied for detecting stable unstable divide or upper and lower threshold of the respective variable. Such approach was applied by Smakhtin, Shilpakar, and Hughes (2006); Monk, Wood, Hannah, and Wilson (2007); Gain and Giupponi (2014) etc. for delimiting threshold discharge estimation and determination of environmental flow. Most of the cases they considered mean plus 7SD (standard deviation) was considered as flow threshold. Same approach can also be applied for estimating threshold rainfall, and water presence area, which could be tolerable. In this approach RVA is calculated following (mean
SD) r parameter r (mean þ SD). After doing this for all the variables, simple year wise comparison is done. It is accounted whether the year having rainfall, water level below lower threshold limit, water presence area is very low, or it is also below lower threshold. More synchronized frequency indicates greater similarity as well as strong linkages or vice versa. 3) In anomaly estimation approach, negative and positive anomaly (actual wetland area-mean wetland area) has been calculated for different years and similar comparison is carried out alike RVA to see the degree of similarity or dissimilarity. For this it is calculated that how many years attain RVA range or beyond the upper or lower limits in case of rainfall, flow level and wetland area since 1988–2014. Similarity of such incidents (in terms of year) between rainfall and wetland area; water level and wetland area is counted. Fig. 2 shows flow diagram that addressing methodology of the workflow.

Please cite this article as: Talukdar, S., & Pal, S. International Soil and Water Conservation Research (2017), http://dx.doi.org/10.1016/j. iswcr.2017.05.003i

S. Talukdar, S. Pal / International Soil and Water Conservation Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎

5

Fig. 2. Flow diagram addressing methodology of the workflow.

4. Results and analysis 4.1. Alteration of flow regime in river Reservoirs often convert highly variable natural flows into more stable flows, with reduced monthly and seasonal variations (Batalla et al., 2004) but situation may also be reversed. Irrigation storage may produce short-term variability in flows during peak demands, with otherwise constant flows (Walker, Sheldon, & Puckridge, 1995). The altered seasonal pattern can influence oxygen levels, temperature, suspended solids, drift of organisms, and cycling of organic matter and other nutrients, as well as having

direct impacts on biota (Ward & Stanford, 1995). In the present study area seasonal change of rainfall and water requirement regulate water flow regime of the river. Tables 3–5 respectively show the average, maximum and minimum water level gap of different seasons in pre and post dam periods, respective variation and instability in flow character. From these tables, it is evident that there is significant gap of water level in all seasons between pre and post dam conditions. Maximum gap (average W.L.: 52.24%; Max. W.L.: 44.59%; Min. W.L.: 55.75%) is recorded during pre monsoon season. In a word, water level in all forms has reduced in dam after periods. Williams and Wolman (1984), Collier, Webb, and Schmidt (1996), Batalla et al. (2004), Guo et al. (2012) have established that dam

Table 3 Seasonal water level gap before and after dam in reference to average, maximum and minimum water level. Pre and post monsoon seasons are being considered. Water Level type (water Season level-zero value)

Water level (m.) Before dam condition

Gap (m.) % of gap CV in % After dam condition

Instability index

Before dam condition

After dam condition

Before dam condition

After dam condition

Average water level

Pre monsoon 6.96 Post 8.17 monsoon

3.33 5.53

3.64 2.64

52.24 32.34

4.95 3.84

10.30 8.41

2.38 3.84

4.15 7.08

Maximum water level

Pre monsoon 7.51 Post 8.89 monsoon

4.16 6.51

3.35 2.38

44.59 26.74

3.15 3.07

7.91 8.88

2.43 2.93

3.66 7.1

Minimum water level

Pre monsoon 6.97 Post 7.45 monsoon

3.08 5.07

3.89 2.38

55.75 31.96

3.86 4.65

9.62 7.90

2.72 4.5

3.4 7.11

Please cite this article as: Talukdar, S., & Pal, S. International Soil and Water Conservation Research (2017), http://dx.doi.org/10.1016/j. iswcr.2017.05.003i

S. Talukdar, S. Pal / International Soil and Water Conservation Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎

6

Table 4 Wetland area and associated pixel frequency for pre and post monsoon seasons both during pre and post dam conditions under different water presence frequency (WPF) or inundation frequency (IF) classes. Before dam construction

After dam construction

year

IF class (number of pixel)

Total pixel

High

Area (sq.km) Low

Moderate

Total (sq.km)

Seasonal wetland (sq.km.)

165.37

Low

Moderate

High

Apr-88

17996

6383

1284

25663

16.20

5.74

1.16

23.10

Nov-88

132768

40078

36565

209411

119.49

36.07

32.91

188.47

Apr-93 Oct-93 Mar-99 Nov-99 Mar-04 Nov-04 Apr-10 Nov-10 Mar-14 Nov-14

19054 87328 16305 126436 13013 13013 1639 48749 10746 33792

6066 48156 12968 29330 10261 10261 1347 37822 4190 22282

4484 65953 7739 13251 3845 3845 1277 28676 5651 15399

29604 201437 37012 169017 27119 27119 4263 115247 20587 71473

17.15 78.59 14.74 113.79 11.71 11.71 1.48 43.87 9.67 30.41

5.46 43.34 11.72 26.40 9.23 9.23 1.21 34.04 3.77 20.05

4.04 59.36 6.99 11.93 3.46 3.46 1.15 25.81 5.09 13.86

26.64 181.29 33.45 152.11 24.41 118.05 3.84 103.72 18.53 64.33

154.65 118.66 93.64 99.88 45.8

Table 5 Area under different inundation or water presence frequency zones in pre monsoon season. Pre dam period

Post dam period

Frequency (%)

Area (sq.km)

Frequency

Area (sq.km)

0–15 15–34 34–80 Total

22.14 13.41 6.65 42.2

0–18 18–40 40–74 Total

13.58 8.32 5.97 27.87

WPF status

Vulnerability in terms of water scarcity

Suitability for agriculture

Ecosystem stability

Low Moderate High

Highly vulnerable Moderate vulnerable Low vulnerable

Suitable area for agriculture Moderate Poor

Low Moderate High

Fig. 3. Yearly average water level change of Punarbhaba river in pre and post dam periods in (a) pre monsoon season (b) post monsoon season.

Fig. 4. Yearly maximum water level change Punarbhaba river in pre and post dam periods in (a) pre monsoon season (b) post monsoon season.

reduces the peakiness of flow in their respective works and it is also true in the present study area. Peakiness of flow is reduced by 44.59% and certainly it will have strong impact on flood frequency and inundated area in the river basin. Figs. 3(a-b), 4(a-b) and 5(a-b) respectively depict yearly average, maximum and minimum water level conditions in (a) pre monsoon and (b) post monsoon seasons. All these figures clearly show that in dam after periods water levels have attenuated at significant scale although the degree is not identical for all the periods.

The mentioned tables also measured that keeping parallelism with water level attenuation, but variability of flow is increased to some extent. For instance, average and minimum water level variability of flow are inflated from 4.95% to 10.30% and from 3.86% to 9.62% during pre monsoon those are maximum over the year (Table 3), maximum water level variability is increased from 3.07% to 8.88% during post monsoon season. Inflation of flow variability is not usual situation in post dam period as recorded by the scholars associated in this field (Guo et al., 2012) but it is found

Please cite this article as: Talukdar, S., & Pal, S. International Soil and Water Conservation Research (2017), http://dx.doi.org/10.1016/j. iswcr.2017.05.003i

S. Talukdar, S. Pal / International Soil and Water Conservation Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎

7

Fig. 5. Yearly minimum water level change Punarbhaba river in pre and post dam periods in (a) pre monsoon season (b) post monsoon season.

here perhaps due to water regulation rules from dam. Instability index measured for average, maximum and minimum water level in different seasons also show same trend of instability rise in dam after condition and seasonal variation of instability is also quite similar to the result of coefficient of variation as stated. 4.2. Seasonal change of wetland area Except the river and the riparian wetland fed by ice melt water or recharged by underground perennial spring, most of the wetland in the continental parts is seasonal in Indian sub continent. The degree of seasonal water presence area is strongly controlled by seasonal variation of rainfall as well as river flow regime. In the present study, it is considered that the water presence area sustains even up to pre monsoon is treated as permanent. Therefore to identify the seasonal part of wetland, water presence area for pre monsoon season is overlain on water presence area for post monsoon season. Extra part beyond perennial water presence area i.e. in the peripheral part of the same is treated as seasonal

wetland. Fig. 6(a-h) depicts seasonal and permanent wetland part in 1988, 1989,1991,1993,1999, 2008,2010 and 2016. This is also extracted for other years also in this purpose. All these maps shows very irregular pattern of seasonal wetland area but geographical location of the wetlands is quite identical. Core of the major wetlands are same over time. Most of the perennial segment of wetlands is found in very contiguous part of the confluence reach of Punarbhaba river. At the confluence segment of its major tributary Tangon, most part of the wetlands is seasonal in nature. Presence of low land and meeting of two major tributary as well as cumulative assemblages of water have yield perennial wetland in the confluence segment of river Punarbha. Position of water level very nearer to surface (2–3 m. below ground level) also supports surface water stagnation in the wetland. In the middle segment of the catchment, few man-made large water bodies represent as perennial wetland. Table 4 quantifies the area of water presence during monsoon and pre monsoon as well as seasonal wetland extent in selected years. From this table it is found that perennial and seasonal wetland area is declining in trend. In 1988,

Fig. 6. Spatial pattern of wetland in different seasons (a) 1988 (b) 1989 (c) 1991 (d) 1993 (e) 1999 (f) 2008 (g) 2010 (h) 2016.

Please cite this article as: Talukdar, S., & Pal, S. International Soil and Water Conservation Research (2017), http://dx.doi.org/10.1016/j. iswcr.2017.05.003i

8

S. Talukdar, S. Pal / International Soil and Water Conservation Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎

of agriculture association; high, moderate and low in terms of stability and continuity of ecosystem. 4.4. Impact of dam on seasonal inundation characters

Fig. 7. Seasonal water presence area during pre monsoon and post monsoon seasons in different years.

total seasonal wetland area was 165.37 sq.km. which is 45.8 sq.km. in 2010. From the gross loss of wetland area since 1988 to present seasonal wetland loss is more vulnerable than perennial. The calculated rate of wetland loss is extremely alarming. Fig. 7 shows comparative pattern of wetland area between post and pre monsoon in different years. From this graph, it is clear to state that pre monsoon wetland area is quite random but clear cut trend of wetland wasting is noticed in post monsoon season. The gap between post and pre monsoon wetland area is very high all through the temporal resolution. Pre monsoon water presence area is effectively determined by pre monsoon rainfall, temperature intensity and water use for agriculture purposes. 4.3. Composite pattern of wetland association In methodology section, it is already mentioned that for providing relative creditable wetland area delineation, compositing of wetland for different years can be done. In this line of thinking such composite models of wetland area have been prepared both for pre and post monsoon periods. Figs. 8a-b and 9a-b represent composite water presence area during pre and post dam periods of both pre and post monsoon seasons. Each map has been categorized into three distinct inundation or water presence frequency zones in relative terms based on the calculated highest and lowest value and area under each zone during pre and post monsoon seasons are tabulated in Tables 5, 6. These high, moderate and low inundation frequency wetland classes have been further qualitatively defined as low, moderate and high vulnerable zones in terms of water scarcity; poor, moderate and suitable area in terms

4.4.1. Pre monsoon status before and after dam Similarly, season specific compositing of water presence area is extracted using the maps of pre and post dam conditions. Figs. 8a and 8b respectively depict the water presence frequency and wetland area for pre monsoon season during pre and post dam periods. From these figures visually, it is noticed that both wetland area and area under high inundation frequency are declined in after dam condition. Table 5 accounts area under different zones of inundation frequency. In pre dam condition, total areal extension of quite perennial wetland was 42.2 sq, km. and it becomes 27.87 sq.km. in dam after condition (Table 5). Low inundation frequency zone is highly vulnerable because of its very fast rate shrinkage. In pre monsoon period, area of water presence under low inundation frequency (0–15%) is reduced from 22 sq.km. to 13 sq. km. in dam after scenario (Fig. 10a). Area under high inundation frequency is slightly increased in post dam period (Fig. 10b) and it is because of concentration of water in the lowland core wetland from widely irrigated land for pre monsoon paddy cultivation. Relatively elevated land is captured for pre monsoon agricultural practice. 4.4.2. Post monsoon status before and after dam On the other hand, in post monsoon period, total wetland area is reduced from 277.86 sq.km. to 220.90 sq.km. in post dam period. It means in dam after condition, total lost wetland area is 56.96 sq.km. The intensity of wetland transformation of high inundation frequency to successively lower frequency zone is significantly high (27.44 sq. km.) followed by low inundation frequency zone (16.18 sq.km.) (Table 6). This is very usual fact because in after dam condition post monsoon water level has reduced (32.34%) and as a result of this relatively elevated area surrounding wetland core which was inundated in pre dam condition but inundation frequency is reduced in post dam period. Maximum water level of Punarbhaba river is reduced by 44.59% which negatively influences peripheral wetland area with relatively elevated area. Reduction of peak water level does indicate that this area will not receive water even during peak water level regime. Therefore, in post dam period, 16.18 sq.km. wetland area with low inundation frequency is reduced. Figs. 10a and 10b of pre and post dam periods of post monsoon season highlight the spatial pattern of the above-mentioned fact.

Fig. 8. Nature of inundation frequency in pre monsoon season (a) pre dam (b) post dam periods.

Please cite this article as: Talukdar, S., & Pal, S. International Soil and Water Conservation Research (2017), http://dx.doi.org/10.1016/j. iswcr.2017.05.003i

S. Talukdar, S. Pal / International Soil and Water Conservation Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎

9

Fig. 9. Nature of inundation frequency in post monsoon season (a) pre dam (b) post dam periods. Table 6 Area under different inundation or water presence frequency zones in post monsoon season. Pre dam period

Post dam period

Frequency (%)

Area (sq.km)

0–23 23–46 46–89

120.79 91.66 65.51 277.86

WPF status

Vulnerability in terms of water scarcity

Suitability for agriculture

Ecosystem stability

Low Moderate High

Highly vulnerable Moderate vulnerable Low vulnerable

Suitable area for agriculture Moderate Poor

Low Moderate High

Area (sq.km) 0–20 20–45 45–84

104.61 78.22 38.07 220.90

Fig. 10. Area under different water presence frequency classes in wetland during (a) pre monsoon time(b) post monsoon time.

4.4.3. Trend of seasonal wetland area during pre and post dam conditions Table 7 accounts the seasonal water presence area, fluctuation of areal extent, trend of wetland area and pattern of instability in pre and post dam periods. From this table it is clear that in pre and

post monsoon time 28.57% and 31.07% wetland area respectively are reduced in dam after condition. Level of fluctuation is increased significantly in both seasons after dam due to quite whimsical regulation of dam water downstream. Declining trend is accelerated during pre monsoon season in post dam period and it

Table 7 Seasonal pattern of wetland area, areal fluctuation, trend and instability in pre and post dam conditions. Season

Pre Monsoon

Average Area (Sq.km) Pre dam condition

Post dam condition

37.25

26.61 123.49

Post monsoon 179.16

Gap

% of Gap CV in %

Trend model and R2

Instability index

Pre dam condition

Post dam condition

Pre dam condition

Post dam condition

Pre dam condition

Post dam condition

10.64 28.57

61.98

74.55

62.5

23.48

53.15

y ¼
2.314x þ 46.27(0.308) y¼
12.40x þ 201.9(0.287)

55.16

55.67 31.07

y ¼ 18.79x þ 3.242 (0.489) y ¼
15.12x þ 242.6(0.112)

20.91

41.4

Please cite this article as: Talukdar, S., & Pal, S. International Soil and Water Conservation Research (2017), http://dx.doi.org/10.1016/j. iswcr.2017.05.003i

S. Talukdar, S. Pal / International Soil and Water Conservation Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎

10

Fig. 11. Wetland area fluctuation in pre and post dam condition (a) pre monsoon (b) post monsoon. Table 8 Frequency of year having more and less than mean wetland area. Number of frequency (year)

4 mean (142.78 sq. km.)

o mean (142.78 sq. km.)

Total

Before dam After dam Total

9 4 13

1 10 11

10 14 24

is indicated by reducing of Coefficient of determination (R2 reduced from 0.112 to 0.287) and it is because of greater water diversion through dam for irrigation use. But this shift is quite reverse in post monsoon period (R2 increased from 0.489 to 0.308). Wetland area in post dam period is reduced from 37.25 sq.km. to 26.61 sq.km. during pre monsoon period and from 179.16 sq.km. to 123.49 sq.km. in the post monsoon period (Figs. 11a and 11b and Table 8). Here, another fact should be noted that immediate after installation of dam in few years water presence area was high (Fig. 11b) and it is reflected in increasing R2 value in the post monsoon (Table 7). It is perhaps due to contemporary operation principles of the dam and relatively enhanced release of water from dam. If those few years were excluded from analysis, in post monsoon season also, water presence area would be in declining trend. Instability index in both seasons in dam after period is enhanced due to irregular inundation magnitudes in response to inconsistent flow regulation. It might be imaged out different result if the mentioned unstable phases were excluded. 4.5. Assessment of the water presence area distribution in temporal resolution 4.5.1. Chi square test Estimated chi square values are 371.27 and 726.79 during pre and post monsoon seasons respectively where tabulated value is 52.62 at 0.001 level of significance at one tailed test. From this fact it can be said that dam has significant role for determining water presence in

wetland cover area in both the seasons. Coefficient of contingency for pre and post monsoon periods are respectively 0.966 and 0.984 which indicate a strong force controlling the natural water presence dynamics in wetland area. The factors can be of either rainfall fluctuation or flow regime of the river water. Rainfall pattern shows there is no significant variation which is observed in water presence area fluctuation within wetland command area. It does mean flow regime of Punarbhaba River plays vital role for modifying the water presence rhythm in the wetland area. To specify the role of dam on water presence characters in wetland area, 2  2 contingency table is prepared and chi square test is being done. From this analysis it is proved that post monsoon water presence area is strongly guided by the operational principle of the dam. Calculated chi square value based on Table 8 is 11.06, which is greater than tabulated value (10.83) at 0.001 level of significance. Phi coefficient is 2.35. Linear regression analysis is carried out between rainfall and wetland area, river flow and wetland area to assesses the priority of the parameters which to be considered more dominant. From this analysis, it is found that the controlling power of rainfall on wetland area is very negligible (y¼0.256x þ 84.22 and R2 ¼0.053) (Fig. 12a) which is usually unexpected because riparian seasonal wetlands are majorly controlled by rainfall. However, river flow is significantly determine areal fluctuation of wetland (y ¼16.29x
168.5 and R2 ¼0.312) (Fig. 12b). As the river flow is guided by dam stated earlier, it can be concluded that flow regime of the river is dominant drivers than rainfall. 4.6. Ecological assessment 4.6.1. RVA analysis and similarity testing Table 9 depicts RVA of post monsoon rainfall, flow level and wetland area and respective departure. Similarity incidents are counted for each year and comparison is carried out between rainfall and wetland area and flow level and wetland area. Similarity of incidents between rainfall and wetland area in post monsoon season is 45% to total frequency considered. Out of them maximum (27.27%) frequencies come under RVA range. But in case

Fig. 12. Linear regression showing influences of rainfall and flow level on wetland extension (a) rainfall and wetland area fluctuation and (b) flow regime and wetland area fluctuation.

Please cite this article as: Talukdar, S., & Pal, S. International Soil and Water Conservation Research (2017), http://dx.doi.org/10.1016/j. iswcr.2017.05.003i

S. Talukdar, S. Pal / International Soil and Water Conservation Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎ Table 9 Range of variability of rainfall, flow level and wetland area and similarity incidents count among them in post monsoon season (- sign indicates actual value is below the lower RVA threshold limit; þ sign indicates actual value is above the upper RVA threshold limit and unsigned years mean actual value remains within the threshold limits). Year

Rainfall (mm.)

1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Year count (-) Year count ( þ) Year count (unsigned)

289.37 228.31 269.44 209.03 229.48 203.24 273.70 218.34 197.25 209.30 185.75 278.35 277.83 232.12 323.37 335.17 186.95 213.82 302 188.75 217.25

RVA (213.74– 327.70)

– –

– – –

–

–

07 (31.81%) 00 (00%) 15 (68.18%)

Flow level (m.)

RVA (20.35– 21.98)

Wetland area (sq. km.)

RVA (137.09– 221.23)

18.89 18.82 19.18 19.50 19.57 20.87 20.49 19.27 19.57 19.38 18.53 17.53 18.22 17.30 17.08 18.52 17.39 18.46 18.69 13.05 19.13 18.05

– – – – –

181.29 70.67

–

270.59

þ

152.11 240.84

þ

– – – – – – – – – – – – – – – 20 (90.90%) 00 (00%) 02 (9.10%)

123.72 118.0464

– –

92.15

–

78.77 103.72 64.33

– – –

86.54

–

61.25

– 09 (40.90%) 02 (9.10%) 11 (50%)

of similarity incidents between flow level and wetland area is 50% and out of this 41% frequency denotes that fall of water level in the river causes shrinkages of wetland area. This fact signifies the control of flow level on wetland area fluctuation. Table 10 describes the results of RVA for rainfall, flow level and wetland area in pre monsoon season. In case of pre monsoon, similarity incidents is only 18.18% to total frequencies when comparison is carried out between rainfall and wetland area. On the other hand, 81.81% dissimilar frequencies indicate very least control of rainfall for defining wetland area. Similarity incidents, in case of flow level and wetland area, are 31.81% indicating relatively greater control on wetland area. In this connection, it is to be mentioned that during pre monsoon, flow level in river is very low therefore; there has very little chance of inundation of river and feeding wetlands astride. But relatively elevated water level in river support sustaining water level nearer to surface as well as facilitates stagnation of water level within wetland. 4.6.2. Analysis of anomaly and similarity testing For proving the fact that is there any impact of rainfall and flow level on wetland area, anomaly of all the parameters have been calculated and tested similarity. In post monsoon period, in respect to total considered year (1980–2014), similarity is 38.88% between rainfall and wetland area. It does mean only in 38.88% years rainfall trend coincide with changing trend of wetland area. In after dam period, this frequency is reduced from 42.85% to 36.36% indicating the losing control on determining wetland area. Similarity incidents are 52.77% to total considered frequencies when comparison is done between flow level and wetland area. It is reduced from 64.28% to 45.45% in post dam period (Table 9). It is also to be mentioned that not only the proportion of similarity is changed but also the direction of anomaly is changed. In post dam

11

Table 10 Range of variability of rainfall, flow level and wetland area and similarity incidents count among them in pre monsoon season(- sign indicates actual value is below the lower RVA threshold limit; þ sign indicates actual value is above the upper RVA threshold limit and unsigned years mean actual value remains within the threshold limits). Year

Rainfall (mm.)

RVA (32.61– 70.24)

1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Year count (-) Year count (þ ) Year count (unsigned)

62.00533 47.78867 36.002 45.371 68.335 63.797 39.447 44.18633 59.71233 40.21 81.33333 þ 38.2 46.4 64.26667 48.9 17.6 – 99 þ 74.1 þ 67 24 – 42.33333 02 (9.10%) 03 (13.63%) 17 (77.27%)

Flow level (m.)

19.2 18.33667 17.45333 17.57333 17.91333 18.09667 16.80667 18.29333 16.88667 17.16 17.29667 16.00333 16.41333 14.28667 16.12667 15.00667 15.69 14.99 13.78 13.54667 13.58 14.73667

RVA (18.97– 20.95)

Wetland area (sq. km.)

RVA (14.16– 60.34)

26.64 – – – – – – – – – – – – – – – – – – – – 20 (90.90%) 00 (00%) 02 (9.10%)

9.89 66.9 73.26 38.54 33.45 27.61 11.31

– þ þ

–

24.41 23.36 32.14 20.95 8.7 3.84 6.21

– – –

18.53 05 (22.72%) 02 (9.10%) 17 (77.27%)

period, positive anomaly is changed into negative anomaly and certainly this is due to dramatic fall of flow level in the river. Almost identical fact is found in pre monsoon time (Table 10).

5. Conclusion From the analysis it is evident that after attenuating river flow level by 44.59–55.75% in different seasons due to control and diversion of water through dam and as consequences of that in post monsoon period, total wetland area is reduced from 277.86 sq.km. to 220.90 sq.km. and in pre monsoon period it is reduced from 42.2 sq.km. to 27.87 sq.km. High frequency wetland area is transformed into moderate frequency wetland to low frequency wetland. Time series wetland map analysis reveals very inconsistent pattern of seasonal wetland area but geographical location of the wetlands is quite identical. Core of the major wetlands are same over time. Most of the perennial segments of wetlands are found in very contiguous part of the confluence reach of Punarbhaba river. At the confluence segment of its major tributary Tangon, most part of the wetlands is seasonal in nature. From the statistical analysis it can be stated that dam is significantly responsible for altering the inundation scenario of the river and riparian wetland. Growing loss of water presence in the wetland area exposed them to the scope of alteration. These wetlands are strongly impacted either by agricultural extension and urban expansion. Always not due to agricultural invasion wetland deterioration takes place, often such seasonal wetland regime and lowering depth of wetland indirectly invites peasants to be on there for agriculture. Regulation of water flow through dam has speeded up such processes. Certainly, if such cardinality of control continues, rate of conversion of perennial wetland to seasonal or seasonal to

Please cite this article as: Talukdar, S., & Pal, S. International Soil and Water Conservation Research (2017), http://dx.doi.org/10.1016/j. iswcr.2017.05.003i

12

S. Talukdar, S. Pal / International Soil and Water Conservation Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎

ephemeral will be accelerated. In present study, it is established that dam has both decelerated the inundation frequencies but enhanced shrinking wetland area and thereby wetland started to experience new hydrological paradigm. So, policies regarding regulation of dam are vital issue for sustaining wetlands and wetland ecology. Present study also accounted the amount of water that should be released from dam for nurturing the downstream channel ecology, flood plain ecology as well as riparian flood plain wetland. It means for the sake of immediate benefits like irrigating agricultural field and other purposes harvesting water from dam, irreplaceable long-term goods and services of the wetland is sacrificed. It is disappointing enough. Rethinking for staying creatively with river and riparian wetlands compromising with some immediate return or with less return is the way-out amidst this emerging crisis.

Acknowledgements We must acknowledge Mr. Tapas Roy, Assistant Engineer, North Bengal Planning Division, India for providing day to day discharge and water level data of river Punarbhaba.

References Adam, P. (1992). Wetlands and wetland boundaries: problems, expectations, perceptions and reality. Wetlands (Australia), 11, 60–67. Adel, M. M. (2001). Effect on water resources from upstream water diversion in the Ganges basin. Journal of Environmental Quality, 30, 356–368. Batalla, R. J., Gomez, C. M., & Kondolf, G. (2004). Reservoir-induced hydrological changes in the Ebro River basin (NE Spain). Journal of Hydrology, 290, 117–136. Borro, M., Morandeira, N., Salvia, Minotti, P., Perna, P., & Kandus, P. (2014). Mapping shallow lakes in a large South American floodplain: A frequency approach on multitemporal Landsat TM/ETM data. Journal of Hydrology, 512, 39–52. Brandt, S. A. (2000). Classification of geomorphological effects downstream of dams. Catena, 40, 375–401. Brown, A., Matlock, M.D. (2011). A review of water scarcity indices and methodologies. The sustainability consortium, University of Arkansas. Chander, G., Markham, B. L., & Helder, D. L. (2009). Summary of current radiometric calibration coefficients for Landsat MSS, TM, ETM þ , and EO-1 ALI sensors. Remote Sensing of Environment, 113, 893–903. Clarkson, B. R. (2013). A vegetation tool for wetland delineation in New Zealand. Landcare Research, Gate 10 Silverdale Road. New Zealand: University of Waikato Campushttp://dx.doi.org/10.7931/J2TD9V77. Collier, M., Webb, R. H., & Schmidt, J. C. (1996). Dams and rivers: A primer on the downstream effects of dams. US Geological Survey Circular, 1126, 94. Constanza, R. R., R d′Arge, R., de Groot, S., Farber, M., Grasso, B., aeem, R. V., … van der Belt, M. (1997). The value of the world's ecosystem services and natural capital. Nature, 387, 253–260. Cowardin, L.M., Carter, V., Golet, F.C., and LaRoe, E. T. (1979). Classification of Wetlands and deepwater Habitats of the United States. U.S. Department of the Interior, Fish and Wildlife Service Biological Services Program FWS/OBS-79131. Cowardin, L.M., Carter, V., Golet, F.C., and LaRoe, E.T. (1979). Classification of wetlands and deepwater habitats of the United States. U.S. Fish and Wildlife Service Report FWS/OBS-79/31, 131 p. Cuddy, J. D., & Della Valle, P. A. (1978). Measuring the instability of time series data. Oxford Bulletin of Economics and Statistics, 40(1), 79–85. Curie, F., Gaillard, S., Ducharne, A., & Bendjoudi, H. (2007). Geomorphological methods to characterise wetlands at the scale of the Seine watershed. Elsevier Science Direct, 375, 59–68. Das, R. T., & Pal, S. (2016). Identification of water bodies from multispectral landsat imageries of Barind tract of West Bengal. International Journal of Innovative Research and Review, 4(1), 26–37. El Bastawesy, M. (2014). The geomorphological and hydrogeological evidences for a Holocene deluge in Arabia. The Arabian Journal of Geosciences. , http://dx.doi. org/10.1007/s12517-014-1396-9. Feyisaa, G. L., Meilbya, H., Fensholtb, R., & Proudb, S. R. (2014). Automated Water Extraction Index: A new technique forsurface water mapping using Landsat imagery. Remote Sensing Environment, 140, 25–35. Finger, D., Schmid, M., & Wüest, A. (2006). Effects of upstream hydropower operation on riverine particle transport and turbidity in downstream lakes. Water Resources Research, 42, W08429. http://dx.doi.org/10.1029/2005WR004751. Finger, D., Schmid, M., & Wüest, A. (2007). Comparing effects of oligotrophication and upstream hydropower dams on plankton and productivity in perialpine lakes. Water Resources Research, 43, W12404. http://dx.doi.org/10.1029/ 2007WR005868.

Gain, A. K., & Giupponi, C. (2014). Impact of the Farakka Dam on Thresholds of the Hydrologic Flow Regime in the Lower Ganges River Basin (Bangladesh). Water, 6, 2501–2518. Gain, A. K., & Giupponi, C. (2015). A dynamic assessment of water scarcity risk in the Lower Brahmaputra River Basin: An integrated approach. Ecological Indicators, 48, 120–131. Gao, B. C. (1996). NDWI -A Normalized Difference Water Index for Remote Sensing of Vegetation Liquid Water from Space, 58. Elsevier. Graf, W. L. (1999). Dam nation: A geographic census of large American dams and their hydrologic impacts. Water Resources Research, 35, 1305–1311. Graf, W. L. (2006). Downstream hydrologic and geomorphic effects of large dams on American rivers. Geomorphology, 79, 336–360. Grift, R. E., Buijse, A. D., Van Densen, W. L. T., & Klein Breteler, J. G. P. (2001). Restoration of the river-floodplain interaction: benefits for the fish community in the river Rhine. Archiv für Hydrobiologie Supplementband Large Rivers, 12, 173–185. Guo, H., Hub, Q., Zhang, Q., & Feng, S. (2012). Effects of the Three Gorges Dam on Yangtze River flow and river interaction with Poyang Lake, China: 2003–2008. Journal of Hydrology, 416(417), 19–27. Guy, P. R. (1981). River bank erosion in the mid-Zambezi Valley, downstream of Lake Kariba. Biological Conservation, 20, 199–212. Ji, L., Zhang, L., & Wylie, B. (2009). Analysis of dynamic thresholds for the normalized difference water index. Photogramm Eng Remote Sensing, 75, 1307–1317. Johnson, P. N., & Gerbeaux, P. (2004). Wetland types in New Zealand (p. 184) Wellington: Department of Conservation. Johnston, R. M., & Barson, M. M. (1993). Remote sensing of Australian wetlands: An evaluation of LANDSAT TM data for inventory and classification. Australian Journal of Marine and Freshwater Resources, 44, 235–252. Kayastha, N., Thomas, V., Galbraith, J., & Banskota, A. (2012). Monitoring wetland change using inter-annual landsat time-series data. Wetlands, 32, 1149–1162. http://dx.doi.org/10.1007/s13157-012-0345-1. Kundu, N., Pal, M., Saha, S. (2008). East Kolkata wetland: a resource recovery system through productive activities. Proceedings of Taal 2007: the 12th world lake conference, 868-881. Leopold, L. B., Wolman, M. G., & Miller, J. P. (1964). Fluvial Processes in Geomorphology. San Francisco, CA, USA: WH Freeman and Sons. Li, L., Vrieling, A., Skidmore, A., Wang, T., Muñoz, A. R., & Turak, E. (2015). Evaluation of MODIS spectral indices for monitoring hydrological dynamics of aSmall, seasonally Flooded wetland in Southern Spain. Wetland. , http://dx.doi.org/ 10.1007/s13157-015-0676-9. Li, W., Du, Z., Ling, F., Zhou, D., Wang, H., & hang, X. (2013). A comparison of land surface water mapping using the normalized difference water index from TM, ETM þ and ALI. Remote Sensing, 5, 5530–5549. Light, H. M., Vincent, K. R., Darst, M. R., & Price, F. D. (2006). Water-level decline in the Apalachicola River, Florida, from 1954 to 2004, and effects on floodplain habitats. U.S. Geological Survey Scientific Investigations Report (pp. 5173–5183), 5173–5183. Ligon, F. K., Dietrich, W. E., & Trush, W. J. (1995). Downstream ecological effects of dams. BioScience, 45, 183–192. Liu, Y., Song, P., Peng, J., & Ye, C. (2012). A physical explanation of the variation in threshold for delineating terrestrial water surfaces from multi-temporal images: Effects of radiometric correction. International Journal of Remote Sensing, 33 (18), 5862–5875. Losos, E., Hayes, J., Phillips, A., Wilcove, D., & Alkire, C. (1995). Taxpayer-subsidized resource extraction harms species: Double jeopardy. BioScience, 45, 446–455. Mathur, V., Verma, A., Dudley, N., Stolton, S., Hockings, M., & James, R. (2005). Opportunities and Challenges for Kaziranga National Park, Assam Over the Next Fifty Years. Enhancing Our Heritage Project Team, UNF-UNESCO, Paris, p. 15. McCartney, M. (2009). Living with dams: Managing the environmental impacts. Water Policy, 11, 121–139. McFeeters, S. K. (1996). The Use Of The Normalized Difference Water Index (NDWI) in the delineation of open water features. International Journal of Remote Sensing, 17, 1425–1432. Ministry of Environment and Forests (MoEF) (1990). Wetlands of India: A Directory. New Delhi: MoEF Government of India. Monk, W. A., Wood, P. J., Hannah, D. M., & Wilson, D. A. (2007). Selection of river flow indices for the assessment of hydroecological change. River Research and Applications, 23, 113–122. Mukhopadhyay, S., & Pal, S. (2011). Trend of flood at riverine Bengal Basin of Kandi block of Murshidabad district: A hydrogeomorphological overview. Indian Journal of Geography & Environment, 12, 09–18. Nguyen, C. C., Gaume, E., & Payrastre, O. (2013). Regional flood frequency analysis involving extraordinary flood evens at ungauged sites, further development and validation. Journal of Hydrology, 508, 385–396. http://dx.doi.org/10.1016/ j-jhydrol.2013.09.058. Olden, J. D., & Poff, N. L. (2002). Redundancy and choice of hydrologic indices for Characterizing streamflow regimes. River Research and Applications (p. 18), 18. Pal, S. (2015). Impact of Massanjore Dam on hydro-geomorphological modification of Mayurakshi River, Eastern India, Environment Development and Sustainability. Springer Science, 17. Pal, S. (2016aa). Impact of Tilpara barrage on backwater reach of Kushkarni River: A tributary of Mayurakshi River. Environment, Development and Sustainability. , http://dx.doi.org/10.1007/s10668-016-9833-4. Pal, S. (2016bb). Impact of water diversion on hydrological regime of Atreyee River of Indo-Bangladesh. International Journal of River Basin Management. , http://dx.

Please cite this article as: Talukdar, S., & Pal, S. International Soil and Water Conservation Research (2017), http://dx.doi.org/10.1016/j. iswcr.2017.05.003i

S. Talukdar, S. Pal / International Soil and Water Conservation Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎ doi.org/10.1080/15715124.2016.1194282. Parihar, J., Panigrahy, S., & Parihar, J. (1986). Remote sensing based habitat assessment of Kaziranga National Park In: D. Kamat, & H. Panwar (Eds.), Wildlife Habitat Evaluation Using Remote Sensing Techniques (pp. 157–164). Dehra Dun: Indian Institute of Remote Sensing / Wildlife Institute of India. Pearson, K. (1900). On the criterion that given system of deviation from the probable in the case of a correlated system of variables is such that it can be reasonable supposed to have arisen from random sampling (PDF). Philosophical Magazine Series, 550(302), 157–175. http://dx.doi.org/10.1080/ 14786440009463897. Petts, G. E. (1984). Impounded Rivers: Perspectives for Ecological Management (p. 326) Chichester: Wiley. Poff, N. L., & Ward, J. V. (1989). Implications of streamflow variability and predictability for lotic community structure: a regional analysis of streamflow patterns. Canadian Journal of Fisheries and Aquatic Sciences, 46, 1805–1818. Richard, G., & Julien, P. (2003). Dam impacts On And restoration Of An alluvial River – Rio Grande, new Mexico. International Journal of Sediment Research, 18(2), 89–96. Richter, B. D., Baumgartner, J. V., Powell, J., & Braun, D. P. (1996). A method for assessing hydrologic alteration within ecosystems. Conserv. Biol., 10, 1163–1174. Richter, B. D., Baumgartner, J. V., Wigington, R., & Braun, D. P. (1997). How much water does a river need? Freshwater Biology, 37, 231–249. Rödel, R., & Hoffmann, T. (2005). Quantifying the efficiency of river regulation. Advances in Geosciences, 5, 75–82. Rouse, J.W., Haas, R.H., Schell, J.A., & Deering, D.W. (1973). Monitoring Vegetation Systems in the Great Plains with ERTS (Earth Resources Technology Satellite). In Proceedings of the Third Earth Resources Technology Satellite Symposium, 351,309–317. Sanyal, J., Densmore, A. L., & Carbonneau, P. (2014). 2D finite element inundation modelling in anabranching channels with sparse data: Examination of uncertainties. Water Resource Management, 28(8), 2351–2366. Shen, L., & Li, C. (2010). Water Body Extraction from Landsat ETM þ Imagery Using Adaboost Algorithm. In Proceedings of the 18th International Conference on Geoinformatics, Beijing, China, pp. 1–4. Smakhtin, V. U., Shilpakar, R. L., & Hughes, D. A. (2006). Hydrology-based assessment of environmental flows: An example from Nepal. Hydrological Sciences Journal–des Sciences Hydrologiques, 51(2), 207–222.

13

Sun, X., Markthyer, Renard, B., & Lang, M. A. (2014). general regional frequency analysis framework for quantifying local scale climate effects: A case study of ENSO effect on southeast queensland rainfall. Journal of Hydrology, 512, 53–68. http://dx.doi.org/10.1016/j-jhydrol-2014.02.025. Tang, Z., Ou, W., Dai, Y., & Xin, Y. (2013). Extraction of water body based on Landsat TM5 imagery - A case study in the Yangtze River. Advances in Information and Communication Technology, 393, 416–420. Thoms, M. C., & Sheldon, F. (2000). Water resource development and hydrological change in a large dryland river. The Barwon Darling River, Australia. Journal of Hydrology, 228, 10–21. Tiner, R. W., Jr. (1990). Use a high-altitude aerial photography for inventorying forested Wetlands in the United States. Forest Ecology and Management, 33/34, 593–604. Tiner, R. W. (1993). The primary indicators method- a practical approach to wetland recognition and delineation in the United States. Wetlands, 13(l), 50. Townshend, J. R., & Justice, C. O. (1986). Analysis of the dynamics of African Vegetation Using the Normalized Difference Vegetation Index. International Journal of Remote Sensing, 7(11), 1435–1445. Walker, K. F., Sheldon, F., & Puckridge, J. T. (1995). A perspective on dryland river ecosystem. Regulated Rivers, 11, 85–104. Ward, J. V., & Stanford, J. A. (1995). Ecological connectivity in alluvial river ecosystems and its disruption by flow regulation. Regulated Rivers, 11, 105–119. Williams, G. P., & Wolman, M. G. (1984). Downstream effects of dams in alluvial rivers, US geological survey. Professional (p. 1286), 1286. Wilson, E. H., & Sader, S. A. (2002). Detection of forest harvest type using multiple dates of Landsat TM imagery. Remote Sensing of Environment, 80, 385–396. Wu, Q., Lane, C., & Liu, H. (2014). An effective method for detecting potential woodland vernal pools using high-resolution LiDAR data and aerial imagery. Remote Sensing, 6(11), 11444–11467. Xu, H. (2006). Modification of normalised difference water index (NDWI) to enhance open water features in remotely sensed imagery. International Journal of Remote Sensing, 27, 3025–3033. Zamman, A. M., Rahman, A., & Haddak, K. (2012). Regional flood frequency analysis in Arid region – A case study for Australia. Journal of Hydrology, 475, 74–83. http://dx.doi.org/10.1016/j-jhydrol.2012.08.054. Zhou, H., Hong, J., & Huang, Q. (2011). Landscape and water quality change detection in urban wetland: A post-classification comparison method with IKONOS data. Procedia Environmental Sciences, 10, 1726–1731.

Please cite this article as: Talukdar, S., & Pal, S. International Soil and Water Conservation Research (2017), http://dx.doi.org/10.1016/j. iswcr.2017.05.003i