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

net/publication/342255915

Assessing the Status of Changing Channel Regimes of Balason and

Mahananda River in the Sub-Himalayan West Bengal, India

Article in Earth Systems and Environment · June 2020

DOI: 10.1007/s41748-020-00160-y

CITATIONS READS

25 1,433

3 authors:

Suman Mitra Asit Kumar Roy

University of Calcutta Jadavpur University
10 PUBLICATIONS 46 CITATIONS 15 PUBLICATIONS 125 CITATIONS

SEE PROFILE SEE PROFILE

Lakpa Tamang
University of Calcutta
22 PUBLICATIONS 102 CITATIONS

SEE PROFILE

All content following this page was uploaded by Lakpa Tamang on 07 October 2021.

The user has requested enhancement of the downloaded file.

Earth Systems and Environment
https://doi.org/10.1007/s41748-020-00160-y

ORIGINAL ARTICLE

Assessing the Status of Changing Channel Regimes of Balason

and Mahananda River in the Sub‑Himalayan West Bengal, India
Suman Mitra1 · Asit Kumar Roy2 · Lakpa Tamang1

Received: 8 March 2020 / Accepted: 7 June 2020

© King Abdulaziz University and Springer Nature Switzerland AG 2020

Abstract
Rapid infrastructural development, triggered by the economic policies implemented in 1991, subsequently modified the river
channel planform of the Terai region of northern West Bengal, India, through perceptible channel fragmentation, excessive
in-channel sediment mining and rampant landuse alterations of the floodplains. In this context, this study has attempted to
assess the impacts of anthropogenic interventions on fluvial regime of the lower reaches of Balason and Mahananda River
of the sub-Himalayan region over the last 30 years by adopting a combined methodological framework of remote sensing
and field survey. Results showed that initially the naturally vegetated areas had converted into crop land and grass land, tea
plantation or bare land, and afterwards, the majority of it had been converted into built-ups. Besides, a noticeable amount
of channel narrowing was directly proportionate to unrestricted sediment mining and embanking of the river channels, evi-
denced between 1987 and 2017. Comparison of measured multi-temporal channel width along with cross-profiles showed
remarkable channel narrowing (18.8 m/year) as well as significant bed lowering (3.15 m). Studies reveal that the rivers have
started showing signs of losing their existing equilibrium condition and if that happens then that will lead to the sinking of
the ground water table, decreasing flood occurrence interval, destabilization of existing infrastructures associated with riv-
ers, and destruction of river ecology. In order to restore the state of socio-hydrological as well as eco-hydrological amenities
of these rivers, certain recommendations have also been made towards promoting the wise use of riverine resources by the
local communities and policy makers.

Keywords Anthropogenic intervention · Bed lowering · Channel narrowing · Landuse alteration · Sediment mining

1 Introduction the water and sediment flux of river systems worldwide.

Along with that anthropogenic activities such as landuse
River planform or river topography is the function of several transformation, damming, embanking, in-channel mining
natural and anthropogenic factors which primarily deals in and construction of overpasses on the river channels, can
varying sediment as well as water flux (Strahler 1956; Greg- also amplify channel modification (Strahler 1956; Surian
ory 2006). Among several natural factors; climate change, and Rinaldi 2003; Paul and Meyer 2001; Gregory 2006;
tectonic movement, large floods and volcanic eruptions Khaleghi and Surian 2019; Birerley and Fryirs 2005). At this
are such elements, which play a crucial role in regulating time of growing anthropogenic pressure on environment,
it cannot be possible to discuss channel adjustments with-
* Lakpa Tamang out assessing the nature and degree of human intervention
[email protected] experienced by it. Genetically, the river itself has the inher-
Suman Mitra ent feedback mechanism which helps it regain its previous
[email protected] equilibrium after large scale catastrophic events. However,
Asit Kumar Roy since anthropogenic activities have become integrated into
[email protected] the fluvial system; consequently, the river cannot attain its
previous equilibrium as earlier (Strahler 1956; Paul and
1
Department of Geography, University of Calcutta, Meyer 2001). Thus, human-induced channel adjustments
Kolkata 700019, India
in several geo-environmental setups have been drawing the
2
Department of Geography, Jadavpur University, attention of the researchers for the last 70 years (Scorpio and
Kolkata 700032, India

13
Vol.:(0123456789)
 S. Mitra et al.

Rosskopf 2016; Khaleghi and Surian 2019; Radecki-Pawlik 2002; Knighton 1989). Since rivers flowing in the sub-Him-
et al. 2014; Chin 2006). alayan foothills are excellent suppliers of riverine sediments
Cumulative effects of anthropogenic activities can cause of various sizes from boulder and pebble to sand which are
irreplaceable alterations to the river’s physiography and chiefly used in infrastructural developmental works, hence it
among all these, urbanisation has become one of the most is heavily mined to fulfil the need of the construction aggre-
critical ones (Abass et al. 2018; Ramachandra et al. 2012; gates (Tamang and Mandal 2015).
Pradhan 2004; Hollis and Luckett 1976). It has been revealed Degree of channel planform dynamics is significantly
from several literatures that urbanisation has substantial dependent on the pattern of landuse transformation of the
effects on the sediment regime as well as on effective dis- riparian zone of a river. Riparian zones act as a buffer to the
charge of a channel. Furthermore, with increasing impervi- fluvial system since it controls river’s lateral migration, and
ous area, infiltration rate decreases, and in connection with sediment discharge to the river; it influences overland flow
that peak discharge also increases alternatively; which may and the fluvial ecosystem as well (Gordon and Meentemeyer
result in-channel widening through abrasion of loose bank 2006; Picco et al. 2017). Specifically, urbanization acts as
materials (Gregory 2006; Paul and Meyer 2001). Damming one of the most influential altering forces of landuses, which
and embanking act as the major constraining factors to riv- rapidly impacts the channel planform with the increasing
er’s longitudinal and lateral connectivity (Gregory 2006). pressure of settlers at the river banks. If the pace of planform
Concretized large dams and embankments can directly alter dynamics is high enough then through reshaping and recon-
sediment and water flux by disconnecting the rivers from figuring the channel, the river will quickly attain another
its headwaters as well as floodplains and thus river plan- stage of equilibrium. However, in the slower pace of change,
form will adjust to an irreparable extent. Due to damming, the river will maintain a quasi-equilibrium stage, which may
water gets stored in the reservoirs and consequently become last for several years to decades (Leopold 1968; Morisawa
sediment free. This water, often termed as hungry water, and La Flure 1979; Grable and Harden 2006).
possesses higher potential energy and after being released Channel adjustment instigated by human intervention has
from dams it can instantaneously increase the stream power, become a worldwide phenomenon, especially in the fast-
which may lead to vigorous down cutting. Similarly, as the growing countries like India (Gregory 2006; Douglas 1996;
river channel gets embanked, its lateral movements halt, thus Bandyopadhyay and De 2018; Birerley and Fryirs 2005).
at the time of storm water flux loaded with bedload and sus- Being a river-dominated landmass and having prolonged
pended sediment load, cannot get spilled over the floodplain. history of river valley civilizations; India has an intricate
Now to reallocate the storm water flux, the channel starts river-human interaction over the ages. Since the industrial
scouring its bed and thus the river starts incising its course infrastructural development gained rapidity after the eco-
(Gregory 2006; Kondolf et al. 2002). Deep incised valley nomic reform of 1991, the anthropogenic pressure on natural
floor may cause disconnection to any interactions with the resources has increased many folds. These developmental
groundwater table and also may expose the bridge pillars activities have not been restricted only to the major cities or
caused due to vigorous down cutting of channel (Gregory towns but also expanded to the areas which were resource-
2006; Kondolf et al. 2002). Past evidences showed that these ful and characterised by lesser accessibility. Ecologically
kinds of channel modification eventually stimulate channel diverse and topographically distinct sub-Himalayan foothills
incision and channel narrowing, which is capable of causing or Terai region is an appropriate example of such area which
higher magnitude of flood vulnerability at the downstream is going through significant landuse alterations along with
during sudden storm water flux (Strahler 1956; Gregory degradation of natural landscape or natural features (Deb-
2006; Paul and Meyer 2001; Rinaldi et al. 2005; Kondolf nath et al. 2017; Ramachandra et al. 2012; Rao and Kadam
et al. 2002). Heavily encroached river banks, bars can face 2016; Panagariya 2005). In this region nearly 30% area
severe damages due to this as narrowed channels can’t hold under forest cover had decreased, mostly in the last century;
the sudden storm water flux into it. and simultaneously 35% and 82% area under tea plantation
Among all the other anthropogenic actions, in-channel and built-up area respectively were increased (Prokop 2018).
sediment mining acts as a direct modifier to river topography Eventually, the fluvial system of Terai region, which is char-
as it directly affects the sediment flux and thereafter mini- acterised by a unique transitional spatial extent between
mises the carrying capacity of streams and its energy. Due lesser Himalaya and floodplains of Himalayan foothills, is
to heavy sediment withdrawal, the river channel was almost at present under serious anthropogenic pressure and can’t
freed from its bedload and thus dominant sediment size be evaluated without measuring the dimension of anthropo-
would reduce to smaller particles which resulted in increas- genic intervention.
ing stream flow energy; and in connection with that valley A significant research gap had been found regarding lan-
deepening and channel pattern alteration occurred (Gregory duse alteration and other anthropogenic disturbances and its
2006; Paul and Meyer 2001; Kondolf 1997; Kondolf et al. impact on river planform especially in the Terai region, since

13 Published in partnership with CECCR at King Abdulaziz University

Assessing the Status of Changing Channel Regimes of Balason and Mahananda River in the…

most of the previous works were only focused on the impacts (De 1998; Sarkar 1989). Other than the sedimentary and
of anthropogenic activities such as damming, degrading hydrological environment, the climatic condition also dis-
quality of river water due to contamination etc. (Bandyo- played an unchanged scenario, since no significant temporal
padhyay and De 2017, 2018; Wiejaczka et al. 2014). In this change has been observed in the precipitation pattern (R2
context, very few investigations were conducted on the sub- is 0.001) of Siliguri over time (Fig. 2). Since the flow of
Himalayan Terai rivers with special reference to human- Balason and Mahananda are partially spared from any longi-
induced landuse alterations on the floodplain landscape and tudinal barriers, and major channel forming factors (i.e. geo-
its ultimate effect on channel planform modification. Keep- logical setting, sediment and precipitation) have remained
ing in mind this research gap, the present study aims to give constant; then it can be assumed that the discharge of these
an insight on how the landuse landcover (LULC) pattern rivers has not varied significantly. However, the deteriorated
of Siliguri and its surrounding areas has evolved over the conditions of these rivers indicate some other factors, which
last 30 years and along with this, assessment of the nature act as a major stressor on the fragile fluvial systems. It can
and degree of other anthropogenic activities on the channel be hypothesised that the anthropogenic actions have a nota-
of Balason and Mahananda River. This study also tries to ble amount of share in-channel modification.
establish the relationship between the various responses of Siliguri, the second largest urban agglomeration of West
the fluvial system to human intervention, through the assess- Bengal, being situated on the southern part of Darjeeling
ment of channel physiography and quantification of anthro- district, is getting tremendously urbanised at the expense of
pogenic activities using geospatial technologies. all its natural properties (Debnath et al. 2017). Since it is a
gateway of North–Eastern India and acts as a major nodal
point for international trades with China, Nepal and Bhu-
2 Methodology tan; the immense economic potentiality of Siliguri and its
surroundings attract a mass of the population from nearby
2.1 Delineation of the Study Area states as well as countries as settlers. This Himalayan fore-
deep consists of Quaternary alluvial sediment, experiencing
A well-defined area of interest (AOI) has been selected from a large scale transformation in landuses over the last three
the lower part of the catchment with a spatial coverage of decades. This area has average elevation varying from 100
393.18 km2. The area lies between the latitudinal exten- to 200 m, also characterised by large tea plantation, mono-
sion of 26°40′ N to 26°49′ N and longitudinal extension of specific forest lands, braided streams, largely urbanised
88°14′ E to 88°28′ E with a mean elevation of 100 m com- patches and interspersed hamlets with dozens of settlers.
prising of the floodplains of two studied rivers i.e. Balason The southern extent of the study region is mainly charac-
and Mahananda (Fig. 1). The upper stretches of the Balason terised by dense settlement patches with scattered fallow
and Mahananda floodplains are situated at a topographical land on each side of the wide river beds. On the contrary,
transition between the Siwalik Himalaya and its expansive the northern part has higher elevation (200–300 m), embel-
piedmont plain also known as Terai. Hence, this topographic lished with tea plantation, forested areas and small patches
integration between two landscape units covers a complete of rural settlements.
gamut of environmental variability, which helps to develop
a unique fluvial system in terms of hydrological, morpho- 2.2 Data Acquisition and Data Preparation
logical and ecological aspects. These two main rivers of this
study area i.e. Balason and Mahananda; originated from the Three consecutive multispectral LANDSAT datasets with
lower Siwalik and started depositing its bedload just below 30 m spatial resolution, acquired by Thematic Mapper (TM)
their debouching points at Gulma and Dudhia respectively. sensor and Operational Land Imager (OLI)/Thermal Infrared
The entire region is mostly dominated by the quaternary Sensor (TIRS) sensor, as well as a single-band panchromatic
deposits (~ 50,000 years BP), containing sedimentary and CORONA image of the year of 1968 with 2.5 m spatial reso-
metamorphosed rocks such as Darjeeling Gneisses, Dail- lution, downloaded from United State Geological Services
ing Schist, Sandstones and Quartzite (Wadia 1975). The (USGS) Earth Explorer (https​://earth​explo​rer.usgs.gov),
entire Terai region of northern West Bengal has uniformity had been used for geospatial data exploration (Table 1).
in terms of stratigraphic setting which can be distinguished The post-monsoon season was considered at the time of
as a separate geomorphic unit (Gansser 1964; Wadia 1975). data selection to obtain cloud-free images. A long 50 years’
These rivers are characterised by high fluctuating discharge, data had been assessed, after selecting the satellite datasets
in case of Balason it ranged between 5.814 cumec in pre of the year 1968, 1987, 2002, and 2017. Along with these,
monsoon of 1989 to 61.579 cumec in post-monsoon, 1989, the Census data of 1981, 1991, 2001 and 2011 of Darjil-
and in case of Mahananda it ranged between 1.804 cumec ing and Jalpaiguri district were also taken into account for
in pre monsoon of 1987 to 40.173 in post-monsoon of 1987 having a comprehensive outline regarding the demographic

Published in partnership with CECCR at King Abdulaziz University 13

 S. Mitra et al.

Fig. 1  Geographical location and spatial distribution of the specific showing Balason Mahananda River system overlaid in digital eleva-
study sites; a Map indicating West Bengal along with border shar- tion model with division of studied reaches, sections from where river
ing neighbouring countries and states, the study area has been high- LULC dynamicity and river widths are measured respectively, gaug-
lighted; b District wise division of northern part of West Bengal indi- ing stations where cross-profiles are measured, boundary of Siliguri
cating the Balason Mahananda River system in the study area; c Map Municipal Corporation (SMC) and various other places of interest

strength of Siliguri Municipal Corporation (SMC) and its Image processing was primarily conducted using Erdas
surroundings. Imagine 2014 Version 14.01 software. The radiometric
correction was performed on the three satellite datasets

13 Published in partnership with CECCR at King Abdulaziz University

Assessing the Status of Changing Channel Regimes of Balason and Mahananda River in the…

Fig. 2  Status of rainfall variability in Siliguri Municipality and its surroundings from 1993 to 2017 Source: Irrigation and Waterways Depart-
ment, Siliguri Division; Government of West Bengal

Table 1  Geospatial datasets used in the study permissible range. CORONA image of 1968 was not used
Date of acquisi- Satellite Sensor Path/row Resolution (m)
for supervised LULC classification due to its single-band
tion availability; so except this, the images of 1987, 2002 and
2017 were classified based on the existing LULC’s. The data
22/12/1968 CORONA – – 2.5 derived from these classifications were further processed
31/12/1987 Landsat 5 TM 139/041 30 using the platform of Microsoft Excel. Data related to the
16/12/2002 Landsat 5 TM 139/041 30
channel width was obtained from repetitive field investiga-
01/12/2017 Landsat 8 OLI-TIRS 139/041 30 tions and were also processed using Microsoft Excel.
Intensive field investigation had been carried out to col-
lect ground control points (GCPs) using handheld Geo-
to convert the digital number (DN) values into top of graphic Positioning System (GPS) of Garmin E-Trex 20
atmosphere (TOA) radiance values and then it was again for rectifying these images geometrically. During a pilot
transformed into bottom of atmosphere (BOA) reflectance survey, various landuses were identified from 48 selected
product. As the selected study region is mostly occupied sites within the study region and geocoded by 170 col-
with the urban built-ups, hence it can be assumed that the lected GCPs. The results of that survey were further used
atmosphere of this region is filled with particulate matters. for accuracy assessment of the classified raster images for
So, the dark matter subtraction method had been used to the successive timeframes. The channel cross-sectional
perform the necessary atmospheric correction (Campbell data of Balason and Mahananda River were procured from
1993; Chavez 1988). Usually, the greater amount of aerosol the previous research work of 1987 and 1989 respectively.
content in the air, higher will be the atmospheric scatter- Whereas the recent cross-sectional data were obtained
ing effect or irradiance recorded in the image. To get rid of through field-based total station survey using Leica TC-805
this difficult atmospheric scattering, atmospheric correction by the authors. It was measured during the non-monsoon
using dark object subtraction was done (Campbell 1993; Liu period of 2018 (April 2018).
et al. 2005; Chavez 1988). To reduce the geometric distor-
tions, the process of Geometric correction was also per- 2.3 LULC Analysis
formed on the geospatial data sets using the collected GCPs
from the field survey. Firstly, the CORONA image of 1968 Having an accurate and detailed geospatial information
was rectified using the GCPs; then the other three images regarding the changing scenario of landuse/landcovers
were co-registered using that pre-rectified image. Between (LULC’s) throughout the study region, maximum likeli-
two successive images, the root mean square error (RMSE) hood classifier was chosen, which was found to be a most
was kept as less than 0.5 pixel, which was considered as the suitable and well established classification algorithm among

Published in partnership with CECCR at King Abdulaziz University 13

 S. Mitra et al.

contemporary researchers (Shalaby and Tateishi 2007; 1994). Widths of the river channel (in between banks)
Lillesand et al. 2008; Roy and Datta 2018). After adopt- were also measured sequentially from the selected satellite
ing necessary pre-processing steps, all three selected raster images used in this study i.e. 1968, 1987, 2002 and 2017.
images were classified into seven profound LULC classes In the last 50 years, as the most destructive flood event
(i.e. river, river deposit, natural vegetation, tea plantation, of northern West Bengal was recorded in 1968; it can be
crop land and grass land, fallow land and built-up area) stated that a new equilibrium was attained at that time
based on a priori knowledge of investigators regarding the by the channels, which has been quantified by measuring
study region and the classification scheme proposed by the the channel width (Leopold and Wolman 1957). Floods,
National Remote Sensing Centre (Arveti et al. 2016). In playing an important role in-channel forming; the highest
these distinctively different LULC classes, the pixels under magnitude (magnitude 5) flood in the recent history of
river class represented the perennial area within the river northern West Bengal, the flood of 1968; subsequently
channel. Whereas; the wide river beds comprised of fluvial lead to the planform attributes of 1968 to be taken as a
deposits, were represented by the class river deposit. Fal- benchmark for this study (Feio et al. 2014; Wang et al.
low lands mainly comprised of agricultural fields without 2008; Roy 2011).
any crops and bare arable lands, while, the built-up class Both of the rivers Balason and Mahananda were divided
included all types of impervious concretized features, such into 2 km long reaches and river widths were measured
as buildings, concretized pavements, roads and other urban across the middle of every reach. They were divided into
infrastructures. Only the crop land and grass land were con- 10 and 11 reaches and were termed as A1, A2 to A10 and
sidered as a single class as they can’t be distinguishable for B1, B2 to B11; respectively. Image-based measurements
their homogenous spectral reflectance. Most importantly, the of channel widths from recent times were further validated
accuracy of this image classification was further assessed by during field-based observations. But neither people’s per-
the error matrices using the 170 pre-collected GCPs. These ception nor the channel relicts were found to be useful to
GCPs were allotted in a stratified random manner to the dif- validate the cross-sectional widths of 1968 for both the
ferent LULC classes for checking the classification accuracy rivers. It happened chiefly due to the very low number
without any biases (Shalaby and Tateishi 2007). of settlements located at the study sites then. A radical
transformation of the adjoining floodplain has happened
2.4 Change Detection Analysis over the last 50 years (1968–2107), but due to its much
higher resolution of 2.5 m, the bank lines were very easily
Post-classification change detection had been performed to detectable.
analyse the LULC dynamics of the study area. Primarily Like the channel planform, river cross-profiles being the
the spatial extent of each class was calculated into square function of its water and sediment flux, are one of the essen-
kilometre ­(km2); thereafter those values were converted into tial parameters to be assessed for detection of the effect of
percentage for standardization. The net areal change (NAC) the channel forming process on river bed topography (Leo-
for each LULC classes was also derived by comparing the pold and Wolman 1957; Harrelson et al. 1994; Tamang and
areal coverage for two successive time frames (Coppin et al. Mandal 2015). Since both the rivers have been experienc-
2004; Disperati and Virdis 2015). Through generation of ing vigorous sediment mining activity for several decades,
transformation matrices, the pixel-to-pixel transformations three-channel cross-profiles from each river were compared
of LULC’s were detected over time within a space and quan- in this study to detect the fluctuations of the river bed topog-
tified into k­ m2 for assessing the anthropogenic stress on raphy during 1987–2017. On the stretch of Mahananda the
nature. Recognition of the LULC conversion of every single profiles were situated at 250 m upstream of Central Water
pixel between two periods (i.e. 1987–2002 and 2002–2017) Commission (CWC) gauging station at Champasari Bridge
was the ultimate objective of these produced conversion (Mahananda Upper Gauge or U.G.) on NH-10/AH-2 (pre-
matrices (Serra et al. 2003; Islam et al. 2016; Roy and Datta viously NH-31)and 50 m upstream of Champasari Bridge
2018; Sinha et al. 2016). 1987–2002 was termed as first half (Mahananda Central Gauge or C.G.) and 200 m downstream
and 2002–2017 as the second half. of Champasari Bridge (Mahananda Down Gauge or D.G.).
On the other hand, for River Balason, the profiles were situ-
2.5 Measurements of the Channel Planform ated 425 m upstream (Balason U.G.) of CWC gauging sta-
Attributes and River Profiling tion at Matigara Balason Bridge on NH-10/AH-2 (previously
NH-31) and 25 m upstream (Balason C.G.) of Matigara Bal-
In this study, bank to bank width of the entire channel ason Bridge and 375 m downstream (Balason D.G.) of Mati-
was measured instead of channel passage or waterways; gara Balason Bridge. The measured river cross-profiles were
as those can be reshaped in quick successions, especially compared with previous cross-profiles measured in 1987 and
in the braided rivers (Bertoldi et al. 2009; Harrelson et al. 1989 by superimposing them on a constant axis.

13 Published in partnership with CECCR at King Abdulaziz University

Assessing the Status of Changing Channel Regimes of Balason and Mahananda River in the…

2.6 Assessment of the Relationship Between from 1987 to 2017 was experienced by the built-up
Channel Planform and LULC Dynamics coverage. It had increased from 28.78 km 2 in 1987 to
43.36 km 2 in 2002 and 104.95 km 2 in 2017 covering
After measuring the channel widths of three successive larger than one-fourth (26.69%) of the total area. How-
years; i.e. 1987, 2002 and 2017 they were further compared ever, the major expansion was observed in between 2002
with the dynamic areal coverage of the adjoining LULC and 2017 which was 61.59 km 2 (142.06%) as a conse-
classes within the buffer of 1 km. The buffer of 1 km is taken quence of major infrastructural development which took
as the five times of average channel width. It was observed place during this period. Noticeably, this built-up expan-
during the field survey that the area of influence of these sion mainly occurred along the rivers and major roads
reaches comes under this 1 km buffer area. These temporal of NH-10/AH-2 (previously NH-31). It has facilitated
areal data (­ km2) of different LULC’s were further converted the increase of the number of census towns (C.T.) at the
into a percentage (%) for standardising it into a more under- adjoining rural–urban continuum and as a result of that 10
standable than the absolute values. Channel widths of 1968 new census towns emerged within the study area during
were also taken into account to portray the gradual change 2001–2011. Tea plantation also portrayed an increasing
from the least disturbed scenario. trend throughout the study period, as in 1987 the land
occupancy of this class was 41.56 km2, which increased
to 50.48 and 59.49 km 2 in 2001 and 2017 respectively.
3 Results It created an ambiguity, as in spite of losing profitabil-
ity the area under tea plantation was increasing stead-
3.1 Spatio‑Temporal Pattern of LULC ily. As per the local respondents, the reason of such sur-
Transformation prising growth in tea plantation was reopening of some
tea gardens through re-plantation at a larger extent and
As it could be seen from Fig. 3 and Table 2, an abrupt maintaining the gardens by wiping out threatened exotic
change with an overall increase of 76.17 km2 (264.63%) shrubs and bushes. On another hand, after a rapid decline

Fig. 3  Landuse and Landcover (LULC) classification map of the study area for the time frame of a 1987, b 2002, and c 2017

Published in partnership with CECCR at King Abdulaziz University 13

 S. Mitra et al.

Table 2  LULC change statistics of the study are from 1987 to 2017
LULC classes Year 1987 Year 2002 Year 2017 % 1987–2002 % 2002–2017
Area ­(km2) Area (%) Area ­(km2) Area (%) Area ­(km2) Area (%) Area ­(km2) Area (%) Area ­(km2) Area (%)

River 8.93 2.27 1.41 0.36 2.78 0.71 − 7.52 − 84.21 1.37 97.15
River deposit 22.45 5.71 23.68 6.02 16.65 4.24 1.23 5.50 − 7.03 − 29.67
Natural vegetation 117.51 29.89 76.18 19.37 77.96 19.83 − 41.33 − 35.17 1.78 2.33
Tea plantation 41.56 10.57 50.49 12.84 59.49 15.13 8.93 21.47 9.01 17.84
Crop land and grass land 76.45 19.44 83.18 21.15 57.32 14.58 6.73 8.80 − 25.86 − 31.09
Fallow land 97.50 24.80 114.89 29.22 74.03 18.83 17.38 17.83 − 40.86 − 35.56
Built-up area 28.78 7.32 43.36 11.03 104.96 26.69 14.58 50.64 61.60 142.06
Total 393.1849 100 393.1849 100 393.1849 100

of 35.17% during 1987–2002, natural vegetation had one. The landuse transformation matrices indicated that
arrested its declining trend during 2002–2017 with spa- between 1987 and 2002; area coverage of 9.04, 6.95, and
tial coverage of 77.96 km2. Noticeably, except the built- 26.60 km 2 of natural vegetation, river deposit and crop
up area and tea plantation, all other landuse classes have land and grass land had transformed into fallow land;
never followed a unidirectional trend towards increasing as opposed to the second temporal span of 2002–2017.
or decreasing. Large portions of crop land and grass land with an area
This overall landuse transformation pattern observed of 18.24 km2 areal coverage had transformed into fallow
from Table 3 portrays a pronounced landuse transfor- land during the second half. Though, earlier a noticeable
mation pattern of a growing urban centre, from natural amount of natural vegetation (14.54 km 2) was altered
to quasi-natural or pseudo-natural to purely artificial into crop land and grass land, this transformation might

Table 3  Landuse transformation matrix of the study area

(a) Landuse transformation matrix from 1987 to 2002
1987 2002
River River deposit Natural Tea plantation Crop land and Fallow land Built-up area
vegetation grass land

River 0.57 4.29 0.03 0.03 0.47 2.72 0.83

River deposit 0.58 11.98 0.25 0.07 1.61 6.95 1.01
Natural vegetation 0.01 1.09 67.80 18.75 14.54 9.04 6.27
Tea Plantation 0.00 0.24 5.15 22.51 10.70 2.47 0.49
Crop land and grass land 0.02 1.23 1.80 6.57 34.53 26.60 5.69
Fallow land 0.11 3.79 0.62 2.16 19.30 60.38 11.14
Built-up area 0.12 1.07 0.53 0.39 2.03 6.73 17.92
2002 2017
(b) Landuse transformation matrix from 2002 to 2017
1987 2002
River River deposit Natural Tea plantation Crop land and Fallow land Built-up area
vegetation grass land

River 0.51 0.45 0.00 0.01 0.09 0.09 0.26

River Deposit 1.33 10.85 0.33 0.22 1.38 5.99 3.58
Natural Vegetation 0.04 0.25 59.53 8.84 2.03 1.44 4.04
Tea Plantation 0.02 0.11 10.41 28.16 4.87 3.75 3.16
Crop land and grass land 0.17 0.85 4.47 16.05 24.26 18.24 19.13
Fallow land 0.48 3.34 2.24 5.39 23.03 43.10 37.30
Built-up area 0.24 0.81 0.96 0.82 1.64 1.40 37.4868

13 Published in partnership with CECCR at King Abdulaziz University

Assessing the Status of Changing Channel Regimes of Balason and Mahananda River in the…

be happened due to the introduction of seasonal crops or sprayed rapidly in the adjoining rural areas at the expense
change in cropping pattern. It was evidenced that the built- of all above-mentioned LULC classes and as a result of
up area had been supplemented chiefly by fallow land with that several C.T.’s, and large villages had also emerged.
the transformation of 11.14 and 37.29 km2 during first Accuracy had been assessed for all three images using
half and second half respectively in a process of altering the 170 GCP’s collected from the field during the survey
the quasi-natural setting to a complete artificial concre- procedure of 2017. Information gathered from residents
tized landcover. Along with this, a remarkable increase regarding past landuses were also taken into account dur-
has been seen in the transformation of crop land and ing the classification of multi-temporal satellite images.
grass land into the built-up area covering 5.69 km2 during The overall classification accuracy was 82.35%, 83.53%,
1987–2002 to 19.13 km 2 in 2002–2017. Some portions and 85.29% for the years of 1987, 2002 and 2017 respec-
of tea plantation had also transformed directly into the tively, whereas the derived Kappa statistics for all three
built-up area (3.16 km2) between 2002 and 17. As a result classified images were 0.78, 0.80 and 0.82.
of reclamation of aggradational landforms of both Bala-
son and Mahananda River, some of the areas under river 3.2 Characteristics of River Profiles
deposit had also transformed into crop land and grass land
along with fallow land. Due to encroachments on point A detailed comparison of the six river cross-profiles meas-
bars and mid-channel bars, 3.58 km2 areas were directly ured at Balason and Mahananda portrayed a considerable
transformed into the built-up area from the area covered change in terms of the width of the cross section as well
with river deposit during the second half. Since the SMC as the depth of the river channel. Except for the profile
was almost saturated with settlements, built-up areas had situated at 25 m upstream of Matigara Balason Bridge i.e.

Fig. 4  Comparison of river cross-profiles; a1 Balason upper gauge files are compared between 1989 and 2018 and for Mahananda it is
(U.G.) situated 425 m upstream of Balason River gauging station at 1987 and 2018 Data source: Cross-section data of 2018 were meas-
Matigara; a2 Balason central gauge (C.G.) situated 25 m upstream of ured at field by the authors, cross-section data for Balason of 1989
gauging station and (a3). Balason down gauge (D.G.) situated 375 m was procured from unpublished PhD thesis of Prof. Sunil Kumar De;
downstream of gauging station. b1 Mahananda upper gauge (U.G.) Department of Geography, North-Eastern Hill University, Shilong
situated 250 m upstream of Mahananda gauging station at Cham- and cross-section data for Mahananda of 1987 was procured from
pasari; b2 Mahananda central gauge (C.G.) situated 50 m upstream unpublished PhD thesis of Prof. Subir Sarkar; Department of Geog-
of gauging station and b3 Mahananda down gauge (D.G.) situated raphy and Applied Geography, University of North Bengal, Siliguri
200 m downstream of gauging station. For Balason River cross-pro-

Published in partnership with CECCR at King Abdulaziz University 13

 S. Mitra et al.

Balason C.G. (a2 in Fig. 4) all other profiles surveyed at 2.45 m and 1.43 m respectively. It was a bit on the slower
both rivers showed a remarkable decrease in the width of side than that of Balason.
the river channel. Balason U.G. (a1 in Fig. 4) and Balason
D.G. (a3 in Fig. 4) lost its width by 80 m and 100 m in just 3.3 Relationship Between LULC Dynamics
29 years. In case of Mahananda River, a similar trend was and Channel Planform Attributes
observed as Mahananda U.G, C.G. and D.G. (b1, b2 and
b3 in Fig. 4) were narrowed by 70, 45 and 50 m respec- The whole channel regime of Balason River, except reaches
tively in just 31 years. Concerning decreasing channel like a3 and a4, showed tendencies of channel narrowing
width, all these six profiles have also portrayed substantial from 1968 to 2017. Only these two reaches, a3 and a4 had
down cutting at an average rate of > 2 m in less than thirty widened to some extent at an average rate of 4.84 m and
years. In the case of Balason U.G, the vertical difference 2.31 m per year. Being a braided river, Balason had shown
between the past and present thalweg point was 3.15 m enough dynamicity from a1 to a4, but the majority of the
and for Balason C.G. and D.G. it was 2.58 m and 1.9 m rest of the channel was under vigorous negative change
respectively. On the contrary in the case of Mahananda from section a5 to a7. The highest registered narrowing was
U.G., C.G. and D.G., rates of down cutting were 2.06 m, observed at section a5, where the channel narrowed at an
average of 18.8 m per year. On the contrary a8, a9 and a10

Fig. 5  Relationship between LULC dynamics and channel width for dynamics for 2002 and river channel width for 1987 and 2002 are
the rivers Balason (a) and Mahananda (b); have portrayed here. ai shown; aiii and biii portray the situation prevailed in 2017 in terms of
and bi are showing LULC dynamics of 1987 and comparison of river LULC and comparison of channel width between 2002 and 2017
channel width between 1968 and 1987. In case of aii and bii; LULC

13 Published in partnership with CECCR at King Abdulaziz University

Assessing the Status of Changing Channel Regimes of Balason and Mahananda River in the…

experienced very lesser narrowing rate compared to a5, a6 migrants came in large numbers into this region (Datta 1998,
and a7 because at these reaches the river was laterally dis- 2003). People from the neighbouring states such as Bihar,
connected by embankments for many years so the stretch Jharkhand, Assam and Sikkim also participated in the event
had adjusted anyway. From the data provided by Irrigation of population growth throughout this exclusive region of
and Waterways Department, Siliguri Division, Government Himalayan foothills (Chatterjee 2001; Debnath et al. 2017;
of West Bengal and from the field survey the past condition Ghosh 2018; Datta 2003). Eventually, a stage-wise pattern
of the embankments were revealed. The dominant LULC of LULC alteration was identified from the analysis of geo-
associated with the sections from a5 to a10 were character- spatial data, which was the outcome of the above-mentioned
ised by rapidly increasing built-ups with encroached point multi-sequential migration streams. In the early nineteenth
bars and continuous stretch of concrete embankments. The century, through massive deforestation, forested lands were
upper reaches from a1 to a4 had still maintained their natural converted into farmland; and tea plantation was begun under
channel regime; with larger areas covered by natural vegeta- the supervision of former British rulers. However, when this
tion and tea plantation and few small patches of built-ups. tea industry and agricultural practices were proven to be
At these upper reaches i.e. a1 to a4; no such features that unprofitable for the stakeholders, they were pushed to sell
disconnect the rivers with its floodplains or hillslopes were these lands or left it as fallow and got engaged in more prof-
observed (Fig. 5). itable economic practices mostly in the sediment mining
Among the eleven reaches on Mahananda, only the from river bed (Prokop 2018; Khawas 2002; Tirkey 2005;
uppermost reach, i.e. b1, experienced continuous channel Tirkey and Gardner 2006). Thus, within just a few decades,
widening at an average rate of 1.33 m per year, whereas all the morphology of extensive fluvial landscape had changed
other reaches, from b2 to b11 was characterised by the ten- dramatically and reached the verge of problematic urbani-
dencies of intensified channel narrowing. The lower reaches sation practices as one-fourth of the study area was trans-
of Mahananda showed continuous narrowing trend with the formed into the impervious surface.
highest narrowing of 334 m for section b9 during 1968–2017 LULC transformation of an area is primarily regulated by
at an average rate of 6.82 m per year. The lower reaches, the aspiration for better living and developmental approach
i.e. b10 and b11, has not been fully assessed due to less towards the economic prosperity of an area (Senanayake
areal coverage of the CORONA image of 1968. Interest- et al. 2013, Pradhan 2004). Being the main urban centre
ingly, at b10, channel width had increased at an average rate of northern West Bengal, Siliguri has immense economic
of 11.5 m per year from 1987 to 2002 but largely decreased potential as it acts as a major node of connectivity with
at an average rate of 17.5 m per year from 2002 to 2017. north-eastern states of India and other neighbouring coun-
The 11 sections along the Mahananda River can be divided tries such as Nepal, Bhutan, Bangladesh and China. In terms
into three broader groups concerning the pattern of landuses of locational importance, Siliguri remains the connection
observed at the associated areas. Areas under the reaches of with eight North–East Indian states of high strategic impor-
b5, b6, b7, b8 and b9 were featured by a continuous progres- tance with the other parts of the country. Siliguri Municipal
sion of built-up areas with the almost destruction of natural Corporation (SMC) or former Siliguri Municipality, estab-
vegetation. Reaches such as b3, b4 and b9, b10 was charac- lished in 1967, had extensively sprawled out after the 1990s.
terised by dense built-up areas with a complete absence of This areal expansion of urban built-ups was just an after
natural LULC features; i.e. crop land and grass land, fallow effect of population growth in SMC with a decadal growth of
land, and natural vegetation etc. In contrast with all the other 40.53%, 76.70% and 33.88% for 1981–1991, 1991–2001 and
cross sections, areas associated with uppermost sections, b1 2001–2011 respectively. The areal extent of SMC had also
and b2 still had large patches of natural vegetation cover increased accordingly from 15.59 km2 in 1981, 26.37 km2
with almost no impervious surface (Fig. 5). in 1994 (after getting the status of Municipal Corporation in
1994) and 41.9 km2 in 2001 which remains constant in 2011
also. (Bureau of Applied Economics and Statistics 1986,
4 Discussion 1996–97, Census of India 2006: Darjiling district; Census
of India 2014: Darjiling district; Census of India 2006: Jal-
4.1 Understanding the Drivers of LULC Change paiguri district; Census of India 2014: Jalpaiguri district).
The result of LULC transformation depicted that the
Since the time of British rule in India, Siliguri had experi- built-up area had stretched towards Dabgram, situated in
enced an immense burden of immigration of large numbers the north-eastern part of the SMC. This newly grown area
of tribal people, who were forced to settle in Siliguri and its had also been absorbed within the administrative setup of
surrounding hilly areas to run the tea plantation sector. Ini- SMC in 2001. During 2001–2011, in just one decade, 10
tially the majority of immigrants came from Nepal but after new C.T.’s had emerged in addition to 2 previous Notified
the geopolitical setback of 1947 and 1971 and Bangladeshi Municipalities in the adjoining rural areas of Matigara and

Published in partnership with CECCR at King Abdulaziz University 13

 S. Mitra et al.

Naxalbari C.D. blocks of Darjeeling district. It was observed b10, b11 (Fig. 5); were marked transforming very slowly
that a huge number of people, immigrated in Siliguri mainly into built-ups, which was directly reflected into the rela-
to have a better livelihood, were settled in adjacent rural tively wider channel width in this segment for both of the
areas of SMC as the living cost was relatively low there river. Hence it could be stated that urbanised landcover was
comparing to the main city and consequently, a rapid LULC inversely proportionate to channel width as with increasing
transformation took place on that area. An interesting demo- urbanisation in the adjoining floodplain the river channel
graphic lag was observed as, in spite of having a greater might lose its width largely (Yousefi et al. 2017; Vanacker
population growth rate in the first half, (1987–2002) built-up et al. 2005; Hollis and Luckett 1976).
area was not much expanded lower comparing to the sec- Large concrete embankments and bed stabilizing struc-
ond half (2002–2017). The second half was characterised tures had been introduced in Mahananda a long time back to
by much lower population growth than that of the first half. protect people from bank failures and floods. Areas adjoin-
From 1991 to 2001 the population growth was 76.7% within ing the major bridges like Champasari bridge or Hill cart
the SMC area and the growth in the built-up area from 1987 road bridge were embanked at the time of earliest estab-
to 2002 was 50.6%. Whereas from 2001 to 2011 the popu- lishment and as Siliguri grew up by its side Mahananda
lation growth was only 33.9% (SMC area) and the growth got embanked. Concrete embankments were introduced to
of the built-up area was nearly 142% from 2002 to 2017. It Balason recently as Siliguri started absorbing its contigu-
was affirmed from the community response that this trend of ous rural areas. At present, almost half of the Balason River
built-up expansion could be considered as the general trend had embanked with large concrete embankments and bed
for this region. stabilization structures and the process was ongoing at a
rapid rate (Table 4).
4.2 Site‑specific Effects of LULC Transformation Compared to Mahananda, Balason has relatively large
on Channel Planform and Topography catchment area (size of the Balason catchment is 255.27
sq. km compared to Mahananda which is only 78.42 sq.
According to the results of multi-temporal classified maps, km) and is characterised by a relatively higher volume
the entire study area can be divided into three LULC seg- of water and sediment flux in the flood season of June to
ments considering the stages of LULC transformation. The October. Due to that high sediment yield Balason is get-
northernmost part was characterised by dense natural veg- ting heavily mined compare to Mahananda River which
etation patches with lowest built-up cover hence low anthro- is always narrower compared to Balason, due to lesser
pogenic activities. While the middle portion of the study amount of water and sediment flux. Mahananda also was
area was becoming the most urbanised, with almost two- heavily mined but the mining pits were only constrained
thirds of this region under built-up cover; the southernmost to some reaches; on the other hand, in case of Balason, the
part was distinguished by a quasi-natural landuse setting entire stretch of Balason flowing through the study area
with patches of built-ups along with agricultural land and was heavily mined (Tamang and Mandal 2015) (Fig. 6).
barren lands. This area was reclaimed a long time before Though the responses of these two fluvial systems i.e.
but mainly used for agricultural practices. However, pre- narrowing and deepening and change in channel pattern
sent trends portrayed that this area also was getting covered (Fig. 7); were same, but the nature of anthropogenic pressure
by impervious surface. In case of Mahananda, the adjoin- was not similar in every case. While Balason was deteriorat-
ing floodplain and point bars were reclaimed a long time ing from heavy sediment mining, Mahananda was experi-
before, mostly during the time of massive urbanization in encing the adverse impacts of getting laterally disconnected
SMC during the 1990s, whereas, the floodplain of Balason, due to embankments and increased imperviousness in its
comprised of rural areas, got in touch with such urbanisa- floodplain. The effect of sediment mining was on the lower
tion process in very recent time. The expansive river bed side for Mahananda and, on the contrary, floodplain of Bal-
of Balason dissected by large point bars, mid-channel bars ason was less urbanised than that of Mahananda. Hence,
and vast floodplains; were reclaimed in a slipshod manner down cutting and narrowing of both these channels could
as these adjacent rural areas were starting to transform into not be linked with just any single causative factor. Changes
built-up spaces rapidly. While river reaches such as a1, a2,
a3 and b1, b2, b3 (Fig. 5); lying in the northernmost part,
Table 4  Increase in percentage Year Balason Mahananda
were distinguished by wide and expansive river channel; of river banks protected
middle part of the study area from section a7 to a9 and b5 by concrete embankments 1968 1 2
to b9 had recorded a major channel narrowing; where the (1968–2018) Source:
1987 5 7
Satellite image, Irrigation and
landuse adjacent to the channel was rapidly transforming or 2002 36 42
Waterways Department, Siliguri
already transformed into impervious urbanised areas. How- Division, Government of West 2018 46 77
ever, reaches in the southernmost part i.e. reach a10 and Bengal and Field Survey

13 Published in partnership with CECCR at King Abdulaziz University

Assessing the Status of Changing Channel Regimes of Balason and Mahananda River in the…

Fig. 6  Changes in planform of Balason due to extensive gravel mining in the channel bed; a is at Khaprail bazar and b is at Naukaghat, conflu-
ence of Balason with Mahananda

Fig. 7  Gradual change in channel width of Balason and Mahananda River from 1968 to 2017

in the channel pattern also had come as an inevitable effect intense anthropogenic activities such as urbanisation of the
of these factors (Fig. 7). In order to justify this kind of chan- adjoining floodplains, sediment mining, encroachment on
nel narrowing and down cutting it could be said that since river planform etc.
there was no longitudinal disconnectivity found in both of
the rivers, the rainfall did not show any significant change 4.3 Possible Future Patterns of Channel Planform
during the studied years. Moreover, since the entire river and Topography
stretch is lying on similar kinds of Stratigraphic settings,
then it could be perceived that the rapid down cutting and Channel narrowing, channel incision and changes in chan-
narrowing only happened due to the cumulative action of nel pattern were found to be highly interlinked in the study

Published in partnership with CECCR at King Abdulaziz University 13

 S. Mitra et al.

area. With respect to the study area, when the natural chan- bridges will collapse (Kondolf 1997; Kondolf et al. 2002;
nel forming variables remained constant, and development Khaleghi and Surian 2019; Wishart et al. 2008). Further-
of concrete embankments, sediment mining and urbanisa- more, as a consequence of channel incision, groundwater can
tion proved to be the main reasons of channel narrowing, seep into the river channel and the piezometric level will go
and then reconfiguration of the river topography becomes down, which may create a crisis of potable water for the local
inescapable. If channel narrowing process continues in this inhabitants. Usually, groundwater gets recharged by river
way, then as an immediate response to that river channels water during flood season. Whereas in the leaning months,
of Balason and Mahananda would be deeply incised and the river channel maintains its flow by getting recharged
consequently, several structures such as railway and roadway from aquifers. In this situation, if the channel incised deeper

Fig. 8  Evidences of anthropogenic stresses and its impact on Bala- embanked Mahananda River bank with bed stabilizing structures; e
son and Mahananda River; a Encroachments of a point bar found and f Exposed piers of a railway bridge at Matigara and unstable pil-
on Mahananda River bank at Champasari; b Large amount of sedi- lars of a roadway bridge at Champasai on National Highway 10 over
ment is getting withdrawn from Balason River bed at Matigara; c Balason and Mahananda respectively as a consequence of bed lower-
Ground water is getting discharged from exposed water carrying layer ing. Photographs are taken by authors
as a result of bed lowering on Balason River at Tarabari; d Heavily

13 Published in partnership with CECCR at King Abdulaziz University

Assessing the Status of Changing Channel Regimes of Balason and Mahananda River in the…

than the level of acquirers, then the local groundwater table the eco-hydrological perspective, it would simply destruct
cannot maintain its hydrological budget with fluvial system, the natural habitat condition which may lead to mass extinc-
which may prove to be undesirable for the locals, who reside tion of living inherent biota. So, further studies on this area
at the adjacent riparian zones and floodplains, and directly can be directed towards these problems and on replicating
depend on groundwater for their daily water requirements this methodology to find out the nature, degree, impact of
(Khaleghi and Surian 2019; Rinaldi et al. 2005). During anthropogenic intervention on other major rivers of Terai
the field survey, the above-mentioned phenomena were evi- region.
denced very prominently in the form of displaced bridge Restoration of the channel form or habitat condition
pillars (Fig. 8) on both the river and severe water crisis for of these channels is a timely need. Most importantly the
6–8 months of a year, specifically manifested at the Tarabari restoration programme must include the scheme of giving
village situated on the right bank of Balason river (Biswas better job opportunities to the locals instead of sediment
2016). Eventually, the groundwater level of their wells went mining. On the other hand, for the sake of infrastructural
down, immediately after the offset of monsoon rainfall of developmental activities, construction aggregates sourced
October and it cannot regain its former level until the onset by sediment mining are highly needed. So to have this
of monsoonal rainfall in June. In a broader dimension, the supply and demand evenly balanced, there should be a
nexus between uncontrolled in-channel gravel mining and two-way approach of channel restoration projects; firstly,
channel narrowing can damage the natural cycle of the entire estimate the sediment budget and upon that try to regulate
fluvial system, natural sediment sorting, river bed morphol- demand and supply rates and secondly, implementation of
ogy as well as the inherent ecological properties, which were a plan for sustainable withdrawal of sediments as well as
subsequently capable of altering the channel pattern of the of aquatic resources leaving the channels as a continuous
river. Some of the above-mentioned problems have already source of resources. Riverbanks and bars must be free from
started to condense and being considered as an intimidation any encroachments and no further concrete embankments
for the stakeholders to be alert and responsive to the changes shall be built in the upstreams. There shall be regulation
taking place in the fluvial system; otherwise those may get imposed by local authorities on future urbanisation pro-
intensified in the eminent years, and can be detrimental for cesses, which shall give dividends to the prevailing hydro-
the river system (Gregory 2006; Paul and Meyer 2001). logical systems of this region. New settlements or urban
amenities should have enough open unpaved surfaces for
groundwater recharge and along with that; the riparian zone
5 Conclusion must be free from any such concretization and should be
left in its natural stage. Hence future urbanization as well
This study can be conceived as the pioneering work on as restoration management policy should have the inherent
analysing the effects of anthropogenic activities on chan- resilience to combat further difficulties to come.
nel morphology in the lower Darjiling Himalaya using a
combined methodology of geospatial data analysis, field Acknowledgements This study received financial support (Fellow-
ship Reference No. 3128/NET-JUNE 2014) extended by the University
investigations, along with community responses. Commu- Grants Commission, India to the first author. Authors are thankful to
nity responses helped immensely to understand not only the Prof. Subir Sarkar, Professor, Department of Geography and Applied
past landuse scenario but also the past geomorphic setting. Geography, University of North Bengal; Prof. Sunil Kumar De, Pro-
This study has shown that rapid landuse transformation took fessor, Department of Geography, North-Eastern Hill University; for
permitting us to use data from their unpublished PhD thesis. Authors
place adjacent to the river channels. Along with this, intense are also grateful to Dr. Debajit Datta, Assistant Professor, Department
anthropogenic activities are getting operated in-channel; of Geography, Jadavpur University; for his valuable support during the
which is significantly altering the river planform as well preparation of the manuscript. We also express our gratitude towards
as river bed topography causing severe eco-hydrological those who provided their logistic support during fieldwork.
as well as socio-hydrological hazards. Amplification of the
Funding This study had received financial support (Fellowship Refer-
magnitude of storm water flux, disturbed sediment budget, ence No. 3128/NET-JUNE 2014) extended by the University Grants
is fuelling the incision process which can be destructive to Commission, India to the first author.
bridge pillars, bed stabilization structures and groundwa-
ter interactions. Increased down cutting fuelled by sedi- Compliance with Ethical Standards
ment mining, lateral disconnection of the river channel and
increase in impervious area can result in disconnection of Conflicts of interest The authors declare that they have no conflict of
river-ground water interchange, and initial tendencies are interest.
directing towards this. It may lead to ground water scarcity Availability of Data and Material The data or materials used in this
which can be detrimental for a huge population living on the article are part of the ongoing Ph.D. research work of the first author,
floodplains and prove to be an enormous social hazard. On hence not to be shared until its completion. Satellite images used here

Published in partnership with CECCR at King Abdulaziz University 13

 S. Mitra et al.

are freely available, however, the data extracted from these, and used Datta P (1998) Migration in India with special reference to Nepali
in this article, can’t be shared. Migration. Unpublished PhD Thesis. Department of Economics.
University of Calcutta, Kolkata, West Bengal. http://hdl.handl​
Code Availability Not applicable for this article. e.net/10603​/15886​9
Datta A (2003) Human migration: a social phenomenon. Mittal Pub-
lications, New Delhi
De SK (1998) A Study of the Fluvial Dynamics of the River Balasan.
Unpublished PhD Thesis. Department of Geography. University
of Calcutta, Kolkata, West Bengal. http://hdl.handl​e.net/10603​
References /15991​0
Debnath M, Ray S, Islam N, Sar N (2017) Migration patterns and
Abass K, Adanu SK, Agyemang S (2018) Peri-urbanisation and loss of urban growth in north–east India: a study in Siliguri city. Quest
arable land in Kumasi Metropolis in three decades: evidence from J UGC-HRDC 11:118–123. https​: //doi.org/10.5958/2249-
remote sensing image analysis. Land Use Policy 72:470–479. 0035.2017.00016​.X
https​://doi.org/10.1016/j.landu​sepol​.2018.01.013 Disperati L, Virdis SGP (2015) Assessment of land-use and land-
Arveti N, Etikala B, Dash P (2016) Land use/land cover analysis based cover changes from 1965 to 2014 in Tam Giang-CauHai
on various comprehensive geospatial data sets: a case study from Lagoon, Central Vietnam. Appl Geogr 58:48–64. https​://doi.
Tirupati area, south India. Adv Remote Sens 5:73–82. https​://doi. org/10.1016/j.apgeo​g.2014.12.012
org/10.4236/ars.2016.52006​ Douglas I (1996) The impact of land-use changes, especially logging,
Bandyopadhyay S, De SK (2017) Human Interference on River Health. shifting cultivation, mining and urbanization on sediment yields
Springer, Switzerland in humid tropical Southeast Asia: a review with special refer-
Bandyopadhyay S, De SK (2018) Anthropogenic impacts on the mor- ence to Borneo. In: IAHS publications-series of proceedings
phology of Haora River, Tripura, India. Geomorphologie: relief, and reports-International Association of Hydrological Sciences,
processus, environment 24:151–166. https​://journ​als.opene​ditio​ vol 236, pp 463–472
n.org/geomo​r phol​ogie/12019​ Feio MJ et al (2014) Least disturbed condition for European Medi-
Bertoldi W, Zanoni L, Tubino M (2009) Planform dynamics of braided terranean rivers. Sci Total Environ 476:745–756. https​://doi.
streams. Earth Surf Process Landf 34:547–557. https​: //doi. org/10.1016/j.scito​tenv.2013.05.056
org/10.1002/esp.1755 Gansser A (1964) Geology of the Himalayas. Inter-science Publish-
Birerley GJ, Fryirs KA (2005) Geomorphology and River management: ers, London
applications of the River Styles Framework. John, New York Ghosh A (2018) The importance of being Siliguri: border effect and
Biswas P (2016) West Bengal: in this village of 70 wells, 65 dry up the ‘Untimely’ city in North Bengal. In: Neilson B et al (eds)
in four months. The Indian Express. https​://india​nexpr​ess.com/ Logistical Asia. Springer, Singapore, pp 135–154
artic​le/elect​ions-2016/citie​s/kolka​ta/west-benga​l-in-this-villa​ge- Gordon E, Meentemeyer RK (2006) Effects of dam operation and
of-70-wells​-65-dry-up-in-four-month​s-27541​17 land use on stream channel morphology and riparian vegetation.
Bureau of Applied Economics and Statistics (1986) District statistical Geomorphology 82:412–429. https​://doi.org/10.1016/j.geomo​
handbook 1981: Darjeeling District. Government of West Bengal, rph.2006.06.001
West Bengal Grable JL, Harden CP (2006) Geomorphic response of an Appalachian
Bureau of Applied Economics and Statistics (1996–97) District Sta- valley and ridge stream to urbanization. Earth Surf Process Landf
tistical Handbook 1996–97, Darjeeling District, West Bengal, 31:1707–1720. https​://doi.org/10.1002/esp.1433
Government of West Bengal Gregory KJ (2006) The human role in changing river channels.
Campbell JB (1993) Evaluation of the dark-object subtraction tech- Geomorphology 79:172–191. https​://doi.org/10.1016/j.geomo​
nique for adjustment of multispectral remote-sensing data. In: rph.2006.06.018
Carlotto MJ (ed) SPIE Proceedings of the digital image process- Harrelson CC, Rawlins CL, Potyondy JP (1994) Stream channel refer-
ing and visual communications technologies in the earth and ence sites: an illustrated guide to field technique. In: Gen. Tech.
atmospheric sciences II, 1819, pp 176–188 Rep. RM-245. Fort Collins, CO, US Department of Agriculture,
Census of India (2006) District Census Handbook 2001: Darjiling Forest Service, Rocky Mountain Forest and Range Experiment
District, West Bengal Station, vol 61, pp 245
Census of India (2006) District Census Handbook 2001: Jalpaiguri Hollis GE, Luckett JK (1976) The response of natural river channels to
District, West Bengal urbanization: two case studies from southeast England. J Hydrol
Census of India (2014) District Census Handbook 2011: Darjiling 30:351–363. https​://doi.org/10.1016/0022-1694(76)90118​-9
District, West Bengal Islam MR, Miah MG, Inoue Y (2016) Analysis of land use and land
Census of India (2014) District Census Handbook 2011: Jalpaiguri cover changes in the coastal area of Bangladesh using Landsat
District, West Bengal imagery. Land Degrad Dev 27:899–909. https​://doi.org/10.1002/
Chatterjee P (2001) A time for tea: women, labour, and post/colonial ldr.2339
politics on an Indian plantation. Duke University Press, Durham Khaleghi S, Surian N (2019) Channel adjustments in Iranian Rivers: a
Chavez PS (1988) An improved dark-object subtraction technique for review. Water 11:672. https​://doi.org/10.3390/w1104​0672
atmospheric scattering correction of multispectral data. Remote Khawas V (2002) Environment and rural development in Darjeeling
Sens Environ 24:459–479. https ​ : //doi.org/10.1016/0034- Himalaya: issues and concerns. Centre for Environmental Plan-
4257(88)90019​-3 ning and Technology, Ahmadabad
Chin A (2006) Urban transformation of river landscapes in a global Knighton AD (1989) River adjustment to changes in sediment load:
context. Geomorphology 79:460–487. https​://doi.org/10.1016/j. the effects of tin mining on the Ringarooma River, Tasmania,
geomo​r ph.2006.06.033 1875–1984. Earth Surf Process Landf 14:333–359. https​://doi.
Coppin P, Jonckheere I, Nackaerts K, Muys B, Lambin E (2004) org/10.1002/esp.32901​40408​
Digital change detection methods in ecosystem monitor- Kondolf GM (1997) Hungry water: effects of dams and gravel min-
ing: a review. Int J Remote Sens 25:565–1596. https​: //doi. ing on river channels. Environ Manage 21:533–551. https​://doi.
org/10.1142/97898​12777​249_0001 org/10.1007/s0026​79900​048

13 Published in partnership with CECCR at King Abdulaziz University

Assessing the Status of Changing Channel Regimes of Balason and Mahananda River in the…

Kondolf GM, Piégay H, Landon N (2002) Channel response to Geography. University of North Bengal, Darjeeling, West Bengal.
increased and decreased bedload supply from land use change: http://hdl.handl​e.net/10603​/18233​2
contrasts between two catchments. Geomorphology 45:35–51. Scorpio V, Rosskopf CM (2016) Channel adjustments in a Mediter-
https​://doi.org/10.1016/S0169​-555X(01)00188​-X ranean river over the last 150 years in the context of anthropic
Leopold LB (1968) Hydrology for urban land planning: a guidebook and natural controls. Geomorphology 275:90–104. https​://doi.
on the hydrologic effects of urban land use. US Department of the org/10.1016/j.geomo​rph.2016.09.017
Interior, Geological Survey Senanayake IP, Welivitiya WDDP, Nadeeka PM (2013) Remote sens-
Leopold LB, Wolman MG (1957) River channel patterns: braided, ing based analysis of urban heat islands with vegetation cover
meandering, and straight. US Government Printing Office in Colombo city, Sri Lanka using Landsat-7 ETM + data. Urban
Lillesand TM, Kiefer RW, Chipman JW (2008) Remote sensing and Clim 5:19–35. https​://doi.org/10.1016/j.uclim​.2013.07.004
image interpretation. Wiley, New York Serra P, Pons X, Sauri D (2003) Post-classification change detection
Liu XP, Deng RR, Peng XJ (2005) A fast atmospheric correction with data from different sensors: some accuracy considerations.
method based on TM imagery. Scientia Geographica Sinica 1 Int J Remote Sens 24:3311–3340. https​://doi.org/10.1080/01431​
Morisawa M, La Flure E (1979) Hydraulic Geometry, Stream Equaliza- 16021​00002​1189
tion and Urbanization. In: Rhodes DD, Williams GP (eds) Pro- Shalaby A, Tateishi R (2007) Remote sensing and GIS for mapping
ceedings of the Tenth annual geomorphology symposia series: and monitoring land cover and land-use changes in the north–
adjustments of the fluvial system. Binghamton, New York, pp western coastal zone of Egypt. Appl Geogr 27:28–41. https​://doi.
333–350 org/10.1016/j.apgeo​g.2006.09.004
Panagariya A (2005) India in the 1980s and the 1990s: A Triumph of Sinha P, Verma NK, Ayele E (2016) Urban built-up area extraction
Reforms. India’s and China’s Recent Experience with Reform and and change detection of Adama municipal area using time-series
Growth. Palgrave Macmillan, London Landsat images. Int J Adv Remote Sens GIS 5:1886–1895. https​
Paul MJ, Meyer JL (2001) Streams in the urban landscape. Annu Rev ://doi.org/10.23953​/cloud​.ijars​g.67
Ecol Syst 32:333–365. https​://doi.org/10.1146/annur​ev.ecols​ Strahler AN (1956) The nature of induced erosion and aggradation. In:
ys.32.08150​1.11404​0 William LT Jr (ed) Man’s role in changing the face of the earth.
Picco L, Comiti F, Mao L, Tonon A, Lenzi MA (2017) Medium and Chicago, Univ. Chicago Press, Chicago, pp 621–638
short term riparian vegetation, island and channel evolution in Surian N, Rinaldi M (2003) Morphological response to river engineer-
response to human pressure in a regulated gravel bed river (Piave ing and management in alluvial channels in Italy. Geomorphology
River, Italy). CATENA 149:760–769. https​://doi.org/10.1016/j. 50:307–326. https​://doi.org/10.1016/S0169​-555X(02)00219​-2
caten​a.2016.04.005 Tamang L, Mandal DK (2015) Bed material extraction and its effects
Pradhan PK (2004) Population growth, migration and urbanisation, on the forms and processes of the lower Balason River in the
Environmental consequences in Kathmandu valley, Nepal. In: Darjeeling Himalayas, India. Geogr Pol 88:393–405. https​://doi.
Unruh JD et al (eds) Environmental change and its implications org/10.7163/GPol.2015.3
for population migration. Springer, Dordrecht, pp 177–199 Tirkey LP (2005) Tea plantations in the Darjeeling district, India: Geo-
Prokop P (2018) Tea plantations as a driving force of long-term land ecological and socio-economic impacts in post-independence
use and population changes in the Eastern Himalayan piedmont. period. Natural Resources Institute, University of Manitoba,
Land Use Policy 77:51–62. https​://doi.org/10.1016/j.landu​sepol​ Manitoba
.2018.05.035 Tirkey LP, Gardner JS (2006) The geo-ecological and socio-economic
Radecki-Pawlik A, Bucała A, Plesiński K, Oglęcki P (2014) Ecohy- implications of tea estate closures in Darjeeling, Indian Himalaya.
drological conditions in two catchments in the Gorce Mountains: In: International symposium towards sustainable livelihoods and
Jaszcze and Jamne streams–Western Polish Carpathians. Eco- ecosystems in mountainous regions, Chiang Mai, Thailand
hydrol Hydrobiol 14:229–242. https​://doi.org/10.1016/j.ecohy​ Vanacker V, Molina A, Govers G, Poesen J, Dercon G, Deckers S
d.2014.06.003 (2005) River channel response to short-term human-induced
Ramachandra TV, Aithal BH, Sanna DD (2012) Insights to urban change in landscape connectivity in Andean ecosystems. Geo-
dynamics through landscape spatial pattern analysis. Int J morphology 72:340–353. https ​ : //doi.org/10.1016/j.geomo​
Appl Earth Obs Geoinf 18:329–343. https​://doi.org/10.1016/j. rph.2005.05.013
jag.2012.03.005 Wadia DN (1975) Geology of India. Allahabad, Tata McGraw Hill
Rao A, Kadam K (2016) 25 years of liberalisation: a glimpse of India’s Publishing Co. Ltd
growth in 14 charts. Firstpost. https​://www.first​post.com/busin​ Wang L, Brenden T, Seelbach P, Cooper A, Allan D, Clark R, Wiley
ess/25-years​-of-liber​alisa​tion-a-glimp​se-of-india​s-growt​h-in-14- M (2008) Landscape based identification of human disturbance
chart​s-28776​54.html gradients and reference conditions for Michigan streams. Envi-
Rinaldi M, Wyżga B, Surian N (2005) Sediment mining in alluvial ron Monit Assess 141:1–17. https ​ : //doi.org/10.1007/s1066​
channels: physical effects and management perspectives. River 1-006-9510-4
Res Appl 21:805–828. https​://doi.org/10.1002/rra.884 Wiejaczka Ł, Bucała A, Sarkar S (2014) Human role in shaping the
Roy S (2011) Flood Hazards in Jalpaiguri District and its Management. hydromorphology of Himalayan rivers: study of the Tista River
Unpublished PhD Thesis. Department of Geography & Applied in Darjeeling Himalaya. Curr Sci 106:717–724
Geography. University of North Bengal, Darjeeling, West Bengal. Wishart D, Warburton J, Bracken L (2008) Gravel extraction and plan-
http://hdl.handl​e.net/10603​/14993​1 form change in a wandering gravel-bed river: the River Wear,
Roy AK, Datta D (2018) Analyzing the effects of afforestation on estu- Northern England. Geomorphology 94:131–152. https​://doi.
arine environment of river Subarnarekha, India using geospatial org/10.1016/j.geomo​rph.2007.05.003
technologies and participatory appraisals. Environ Monit Assess Yousefi S, Moradi HR, Keesstra S, Pourghasemi HR, Navratil O,
190:645. https​://doi.org/10.1007/s1066​1-018-7030-7 Hooke J (2017) Effects of urbanization on river morphology of the
Sarkar S (1989) A geo- environmental appraisal of the Upper Talar River, Mazandarn Province, Iran. Geocarto Int 34:276–292.
Mahananda Basin of the Darjeeling Himalaya West Bengal. https​://doi.org/10.1080/10106​049.2017.13867​22
Unpublished PhD Thesis. Department of Geography & Applied

Published in partnership with CECCR at King Abdulaziz University 13

View publication stats