River morphology describes a river's cross-sectional shape, sedimentation, and erosion. The meandering parameters, oxbow formation and decadal land usage land cover (LULC) fluctuations of Barak River were investigated using 21 meandering spans to measure river morphological changes. The decadal meandering parameters were calculated reach-wise and section-wise to characterize river morphological changes. It was observed from the paired t-test that the river width significantly changed during the study period (1990–2020). Strong inter-relationships between the meandering parameters are shown from the regression analysis. The morphological investigation found a reduction in the centerline distance due to variations in the radius of curvature caused by the internal arc's reduction. As a result, the average sinuosity has decreased over time. The current work used SVM and ML techniques for LULC classification, and a comparison of ML and SVM techniques was also done. The SVM technique performs better. The decadal LULC analysis suggests that between 1990 and 2020, the areas of water bodies, forests, and bare land types declined. Whereas, agricultural and settlement areas increased. River morphology is substantially impacted by agriculture and urbanization, particularly in areas where oxbows occur simultaneously, since this work may apply to other similar meandering river management along the alluvial flood plain.

  • River morphology is affected by meandering parameters, land resource changes, and oxbow development.

  • A minor change in meandering parameters like sinuosity, radius of curvature, etc., can affect river width, migration rate, and morphology. A statistical analysis can estimate the river's meandering parameters.

  • Oxbow formation influence morphology. Future oxbows can be identified through morphological study, GIS technique.

The flow and sediment regimes combine with the landscape's physiographic features and vegetation cover to give alluvial floodplain rivers their dynamic nature (Ward & Stanford 1995). In the alluvial stretch, the rivers meander frequently and take different courses. Subsequently, the river morphology changes on a spatial and temporal scale. Several variables influence river morphology (Lancaster & Bras 2002). These phenomena represent the hydro-geomorphologic forms caused by the river's lateral movement (Yousefi et al. 2016). The river morphology mainly describes the cross-sectional shape of the river channel and the process under which the river section changes due to sedimentation and erosion from banks and beds (Manjusree et al. 2013). Meanders are significant features that can alter the morphology of floodplains (Lagasse et al. 2004; Guneralp et al. 2012). The reason and process of the changing nature of morphological traits have been investigated by various researchers over the last few decades (Kotoky et al. 2005; Sarkar et al. 2012; Sinha & Ghosh 2012; Gogoi & Goswami 2014; Verma et al. 2021). Inner banks erode, and outer banks accumulate to keep a river balanced (Debnath et al. 2017) to avoid lateral and vertical instabilities in the river (Resmi et al. 2018). Bank erosion and channel migration can occur on different timescales, depending on the environment (days, years, decades).

Riverbank erosion has long-term impacts on natural disasters (Das et al. 2014). Erosion along riverbanks affects several nations, negatively affecting the landscape and having an effect on the ecosystem and society. Over two million people in Bangladesh lost their homes between 1970 and 2000 due to bank erosion in the Padma and Jamuna rivers (Islam & Rashid 2011). The Himalayan foothill region's society benefits from an integrated investigation of erosion, its causes, and a calculated move toward modifying mitigation procedures in fluvial dynamicity (Prokop et al. 2020), and bank erosion would have an impact on the River Mahananda in Sub-Himalayan North Bengal (Chakraborty & Saha 2022). Space Application Centre (SAC), Ahmedabad, India, and the Brahmaputra Board, India (1996) studied river erosion on Majuli Island to identify and delineate island portions that had changed along the bank line due to the river's dynamic behavior.

Many studies have been conducted on meander parameters and the variables that affect them (Hooke 2013; Nabegu 2014). The sinuosity index is one of the most crucial factors for determining and measuring river morphological changes’ spatial and temporal direction (Ozturk & Sesli 2015; Sapkale et al. 2016). The sinuosity index and riverbank erosion have an opposite relationship (Nath & Ghosh 2022). Flood risk increases as stream sinuosity decreases, and riverbanks fail when driving forces exceed resisting forces (Mohammed-Ali et al. 2021). In addition to having an irregular meander belt with variable width and wavelength, they also have distorted geometric patterns. The channel planform characteristics of a meandering system, such as its width, wavelength, and radius of curvature, exhibit clear correlations (Leopold 1994). In contrast, meandering streams in alluvial zones frequently form oxbows. One of the primary causes of oxbow lakes’ formation is soil erosion. A new flood plain was generated by an oxbow (Ijafiya & Yonnana 2018), resulting in significant local or regional effects (Hickin & Nanson 1984). Hence, river morphology studies and geomorphic risks are required for sustainable river basin management (Arnaud-Fassetta et al. 2009; Papini et al. 2011). Oxbow lakes and their subsequent sediment filling are significant for channel self-confinement and meander belt evolution dynamics. In this regard, several river morphological factors such as radius of curvature, centerline distance, sinuosity, axis length, river width, and the impact of anthropogenic activities or changes in land use/land cover (LULC) and affect due to oxbow were investigated to identify morphological changes over the river.

Meandering behavior has been linked to severe issues for human life, such as the destruction of residential and agricultural lands in floodplain zones. Several researchers have investigated the consequences of urbanization on the hydrological, geomorphologic, and biological processes of the fluvial system (Clark & Wilcock 2000; Kondolf et al. 2007; McCann 2013; Das et al. 2014). Assessing bank erosion hazards for the detailed study of fluvial erosion is the most crucial aspect of correlating decline and land use changes (Deng et al. 2019; Twisa & Buchroithner 2019; Neupane et al. 2020). Human intervention alters river courses by focusing on artificial barriers that regulate flow rates (Gregory 2006). Aerial images can be used to compare and evaluate channel geometry changes, particularly river width and urbanization's impact on the land because significant land use changes occur over short enough time scales (Galster et al. 2008). Urbanization and channel morphology are closely linked in many aspects. For instance, Leopold (1970) showed that a river's channel width in the United States was reduced by 20 years due to urban areas. Urbanization has many effects, such as making channels wider and more homogeneous (O'Driscoll et al. 2009; O'Driscoll et al. 2010). This rise cut down on the length of the flow and made the drainage less intense. Urbanization affects the stability of channel banks, the flow of streams, the transport and deposition of sediment, and the widening of channels (McCann 2013). Stream flow and storm runoff in Baltimore, Maryland, have increased due to urbanization in some basin regions (Nelson et al. 2006). Minor land use changes dramatically influenced regional soil erosion rates and sediment transfer to China's Lushi river basin (Wang et al. 2012). Hazarika et al. (2015) found that river dynamics in the upper Brahmaputra plains affected land use and negatively impacted the livelihoods of those living in floodplains. Assessing the impact of agricultural land conversion on soil erosion (Papini et al. 2011; Zhu et al. 2012) is essential. The growing human population is one of many factors that affect how agricultural land is used over time. The soils along riverbanks are vulnerable and erode quickly due to changes in farming practices and vegetation clearing for agricultural use. As a result of harvesting, crops frequently have lower soil organic matter (SOM) than forests (Zhu et al. 2012). Land use and management processes influence soil's chemical and physical composition (Bhattacharya et al. 2016). These land use changes reduce precipitation infiltration, forest cover interception, and drainage efficiency. Flooding and river meandering are impacted by deforestation (Barasa & Pereram 2018; Adnan et al. 2019). Pesticides and herbicides lose effectiveness with heavy use and intensive farming. In response, parasites become more effective and frequent. Intensive farming destroys soil by interfering with natural processes. Pesticides kill soil microorganisms that control composting and organic matter assimilation, reducing soil stability and erodibility. Tree clearance, slash-and-burn techniques, and clearing of forest areas for agriculture have caused deforestation and soil erosion (Sial et al. 2022). Aside from human intervention, changes in the meander characteristics can also be linked to local management operations such as urbanization expansion and the construction of levees, riprap, a dam, and a road (Nelson et al. 2013).

Field data and numerical models estimated meandering and its morphological changes (Xiao et al. 2019). Unfortunately, most alluvial river lengths are unreachable by field technique. Remote sensing is often used to map surface phenomena such as sinuosity index, land use change, and river meandering parameters (Rawat et al. 2013). Very few literatures have documented the effects of LULC transitions, particularly concerning the morphology of rivers and how much of an impact they have. Given the importance of LULC change in channel alterations, the current study attempts to link those changes to understand the phenomenon.

Study area

River Barak enters Assam at 24° N latitude and 93° E longitude (Figure 1), flowing from the Naga Hills. The river Barak has multiple abandoned meandering loops and demonstrates decadal shifting (Choudhury et al. 2014). Nearly 900 km of the Barak River are in Bangladesh, and the remaining 532 km are in India. The Barak valley in Assam was followed by 129 km of the 532 km of India. The Barak valley, which is in the southwest monsoon zone, is around 25–30 km wide (Choudhury et al. 2014). Katakhal, Jiri, Chiri, Modhura, Longan, Sonai, Rukni, and Singla are the principal tributaries of the river Barak. Surma and Kushiyara split the river Barak as it flows toward Bangladesh. The flood plain buffer zone on both sides of the river, covering about 3 km, has been used as the study area (shown in Figure 1). The landscape comprises hills and plain land terrain, the plain region is surrounded on three sides by hills, except in the west (Singh & Ghosh 2022). Erosion has caused the Barak river channel to move northward towards Silchar (Das 2012). Because the river flow followed a severe meandering pattern, many erosion events resulted in river migration during most of the research period. Figure 2 shows images from field visits to several erosion-vulnerable zones within the study area. Meanders flow through flood plains when alluvial rivers erode and deposit silt along their banks, causing course shifts. The cut-off formation is always a potential in the research region due to the substantial sinuosity and erosion. The study has many sharp curves (Figures 1, 4(a), 6). Even though most of the Silchar town boundary is a meander's loop, only a few bends are protected by protection structures.
Figure 1

Study Area map.

Figure 2

Field visit conducting vurnarable zone within study area.

Figure 2

Field visit conducting vurnarable zone within study area.

Close modal

The River Barak, which flows through the Cachar district, is one of North East India's most meandering rivers. After the Brahmaputra and Chambal rivers, it is the third-largest river in NE India. In several places, the valley is influenced by the river's high meandering and shifting behavior (Das et al. 2007; Das 2012). Agricultural fields, tea garden complexes, rural and urban inhabitants, and abandoned and uncultivated areas compose the floodplain areas, mixing multiple LULCs (Das & Das 2014). Increased urbanization in recent decades has also been a concern for the valley. Human development and its major effects on LULC changes are seen in the Barak River (Annayat et al. 2022). Several environmental problems have arisen as a result of the rapid conversion of open space and vegetation to built-up a region, which is impeding Silchar's ability to grow sustainably (Devi et al. 2019). Land usage, land cover, and river morphology are all linked, according to literature reviews. LULC investigations, on the other hand, are critical for land and water resource management, particularly in flood-prone, highly meandering regions such as Barak. The Barak River's channel behavior was analyzed using satellite imagery and some field data to find river segments that remained steady between 1910 and 1988 (Bardhan 1993). To anticipate channel migration over Barak, an attempt is made to explain and assess the empirical approach and time sequence extrapolation method. It is discovered that Nanson and Hickin technique predicts meander migration better than other empirical methods (Annayat & Sil 2020a). Between 1918 and 2003, a quantitative investigation of the Barak River found an increasing trend in erosion and deposition (Laskar & Phukon 2012). Furthermore, considerable bank shifting in the Barak River was investigated, with the study predicting that it will soon have severe consequences for the economy and people's livelihood (Annayat & Sil 2020a). As a result of the study area's significant sinuosity variation, which causes erosion, the formation has a continual tendency to cut off (Nath & Ghosh 2022).

The study is especially important for this form of highly meandering river management because it indicates how high meandering river morphology has changed over the period and identified some major causes of the changes. This study deals with how the Barak River's meander parameters are interrelated to each other and how it influences river morphology. To the literature, it has been demonstrated that the River Barak has had a significant impact over time and causes regional morphological vulnerability. The links between land use land cover and river morphology are complex. But LULC studies are vital for both land and water resource management. Although there was some research on river morphology (Das 2012; Laskar & Phukon 2012; Annayat & Sil 2020a) and some research was conducted on land use changes over the basin (Das 2012; Devi et al. 2019, Annayat & Sil 2020b). But, these studies were carried out LULC analysis using the ML approach (Das 2012; Devi et al. 2019, Annayat & Sil 2020b; Annayat et al. 2022). However, a literature review revealed that using the SVM technique for LULC classification offered more accuracy. As a result, SVM and ML techniques were utilized in the current work for LULC classification, and a comparison between ML and SVM techniques was also made. Which are provided more preciously quantify the LULC changes in this region, which is essential for related meander river morphological studies.

For statistical analysis over meander parameters, there was some study conducted using the GIS platform that analyze the significance of meander parameters, but very few studies quantify the inter-statistical relationship between each parameter concerning others. In this study, a regression analysis is performed to identify the inter-relationship among meandering parameters. Future trends in river morphology and meander evolution can be predicted by investigating the meander parameters. Better river management and damage reduction may be possible by having an understanding of the evolving meander parameters. One of the key elements responsible for regional and global morphological alterations has been recognized as oxbow formation. It is important to identify future oxbow-prone zone to limit the morphological changes. According to the literature, river bank erosion and its impacts are one of the disasters that have a major influence on some of the world's main flood plains. Barak is a highly meandering river that has undergone significant morphological changes over time. This study provided a better understanding of the dynamics and actual meandering characteristics of alluvial rivers along with morphological causes of vulnerability in the context of the significance of this river and its catchment area.

Data collection

The multispectral remote sensing and Landsat data used in this study were collected from the United States Geological Survey (USGS). Landsat images were taken every ten years from 1990 to 2020 (Table 1).

Table 1

Satellite data used in the study

Spacecraft IDLandsat 5Landsat 7Landsat 7Landsat 8
Date 16/01/1990 29/02/2000 08/02/2010 04/02/2020 
Sensor ID TM ETM + ETM + ETM + 
WRS Path 146 146 146 146 
WRS Row 43 43 43 43 
Resolution(m) 30 × 30 30 × 30 30 × 30 30 × 30 
Spacecraft IDLandsat 5Landsat 7Landsat 7Landsat 8
Date 16/01/1990 29/02/2000 08/02/2010 04/02/2020 
Sensor ID TM ETM + ETM + ETM + 
WRS Path 146 146 146 146 
WRS Row 43 43 43 43 
Resolution(m) 30 × 30 30 × 30 30 × 30 30 × 30 

Technique

The meandering parameter, LULC changes, and river migration assessment of the study were all determined using layer-stacking Landsat images. Landsat data and digitized river pathways were used to track the shifting of river bank lines along the river. In morphological assessments based on digitized river paths, oxbows’ formation and effects are useful. LULC mapping with a supervised classification technique was utilized to identify land use trends across the research area. This study focused on morphological changes in the river, as well as changes in LULC characteristics and the identification of potential oxbow formation zones.

Data used

The methods used to clarify morphological changes utilizing remotely sensed data for assessing bend characteristics and amounts of LULC change throughout the channel length in the chosen study reach are shown in Figure 3. Different Landsat pictures were used in this study (Table 1). No atmospheric correction was added to the images because there were no clouds present at the observation time. The spatial resolution (30 m) of images utilized for land use maps is one of the most crucial topics in this study; these restrictions are a source of uncertainty in this investigation.
Figure 3

Flowchart of methodology.

Figure 3

Flowchart of methodology.

Close modal
Figure 4

(a) selection of location for morphological analysis. (b) Meandering parameters of study meanders: axis length (A), meander neck length (L), river width (W), radius curvature (R), Centerline distance (S). (c) Identification of oxbow formation.

Figure 4

(a) selection of location for morphological analysis. (b) Meandering parameters of study meanders: axis length (A), meander neck length (L), river width (W), radius curvature (R), Centerline distance (S). (c) Identification of oxbow formation.

Close modal

River path digitization

For every satellite image between 1990 and 2020, the river bank has been located and defined. The river bank line may be easily identified using the NDWI image. The computation of river channel migration, changes in river channel width owing to erosion and accretion, and identification of erosion and accretion sites is followed by the preparation of the classified images, vector profile, and reference points. The distance between the reference point and the river banks on either side was digitized as a line feature to compute river channel migration. Following that, the attribute table's computed geometry options were used to determine the lengths of these lines. As the distance between the vector line and channel bank increases, positive values show the channel shifting toward the inner side of the river, which caused accretion. The channel shifted toward the outer banks, causing erosion as the distance between the vector line and the channel bank decreased, as indicated by the negative values. Through area calculation using GIS software tools for polygon areas with shifting banklines over the study period, the erosion and deposit area have been estimated.

Meander parameters

The width, wavelength, and radius of curvature of the channel planform of a meandering system all exhibit clear correlations with one another, to prior research. The study river reach's channel centerlines and bank lines were digitized and grouped into 21 meander loops (Figure 4(a)) to consider meander parameters. Hooke's meander change models have been used to identify meander loop changes. The meander modifications in this model were identified based on visual and spatial changes. Axis length (A), meander neck length (L), river width (W), the radius of curvature (R), centerline distance (S), and sinuosity (C) are the specific meander parameters (Figure 4(b)). The longest distance between the internal arc and the meander neck is the axis length. The shortest distance between two meander loops is known as the meander neck. The distance between two meander apexes is measured in water flow length. River width (W) is an average of five cross-sections (the shortest distances between river banks) in a meander curve, and the radius of curvature is the radius of the largest inner circle in a meander loop (Hooke 2013). The sinuosity index is used to identify and measure changes in river morphology (Ozturk & Sesli 2015). The following Equation (1) was used to calculate sinuosity (C) (Schumm's Model 1963):
(1)
where, C = Sinuosity, OL = Observed path, EL = Expected straight path.

Measurement tools in ArcGIS 10.2 and AUTOCAD 2020 were used to measure morph metric parameters for all meander loops between 1990 and 2020.

River migration and oxbow formation

One of the most influential factors in changing river morphology is river migration and oxbow formation (Johnson & Paynter 1967; Sharma et al. 2010; Tal & Paola 2010; Micheli & Larsen 2011). Unfortunately, most alluvial river lengths are unreachable by field technique.

Multi-temporal high-resolution data allows us to track river configuration changes, new channels or oxbow lakes, and riverbank erosion/deposition (Rawat et al. 2013). The shifting of river bank lines along the Barak River was traced using Landsat data from 1990 to 2020. River migration and oxbow formation were studied using Landsat images to digitize river paths. For referring river migration rate, the position of river banks migration was checked. A U-shaped meander in a river is occasionally cut off from the actual river channel that caused it known as an oxbow. When the gap between two river sections is considerably reduced, it is referred to as a high oxbow-prone zone (Figure 4(c)). To further understand how oxbows form, Figure 4(c) provides an example where, in 1990, the gap between two meander necks of a segment was marked by X, and in 2010, the gap had constantly decreased to become Z. The segment's necks coming together between 2010 and 2020 caused the oxbow to form (X > Y > Z > Oxbow formation). So, future oxbows can be introduced by continuously reducing the space between sharp curve necks.

Land use mapping between 1990 and 2020 and their change detection

Changes in land use have influenced hydrology and sediment transport, which have a considerable impact on soil erosion (Brierley & Fryirs 2005; Wang et al. 2012; Mahapatra et al. 2014; Hegazy & Kaloop 2015; Hazarika et al. 2015). Hydrology, climate, land use, and human involvement affect riverbank erosion and deposition (Arohunsoro et al. 2014; Barman 2016; Barasa & Pereram 2018, Adnan et al. 2019; Raj & Singh 2020).

The flowchart shows the sequential land use mapping processes (Figure 2). There are two subtypes of machine learning techniques: supervised and unsupervised techniques (Wu et al. 2019). Supervised categorization uses user-developed spectral signatures of established categories. The maximum likelihood (ML) and SVM algorithms were used in supervised classification. An overview of the two algorithms is given in the subsections that follow (Taati et al. 2015).

The ML classifier (supervised classification) has been widely employed due to its user-friendly nature. The ML classifier uses the likelihood that a pixel corresponds to a particular class. Landsat images were categorized using ML. The GIS software used the reference data samples to build region of interest polygons. Pre-processing using Landsat images for Land use mapping requires subsetting, layer stacking, and complete composition. The LULC mapping of the study area aims to determine changes in land use patterns from 1990 to 2020. The probability that a pixel belongs to a particular class is used by the ML classifier. The Bayesian equation shown in Equation (2) is used to determine the weighted distance, or likelihood D, of an unknown measurement vector X. (Hall 1999).
(2)
where

D = weighted distance (likelihood)

C = particular class

X = measurement vector of the candidate pixel

Mc = mean vector of the sample of class c

ac = the likelihood that any candidate pixel belongs to class c as a percentage

Covc = the pixels in the sample of the class c covariance matrix

= Covc's determinant (matrix algebra)

Covc − 1 = inverse of Covc

T = transposition function (matrix algebra).

More recent techniques, such as artificial neural networks (ANN), SVM, RF, decision trees, and other models, have drawn a lot of interest in remote sensing-based applications, such as LULC classification, over the past decade. As a result, many machine learning techniques have been used in multiple studies on LULC modeling (Teluguntla et al. 2018; Abdullah et al. 2019; Zhang et al. 2019; Talukdar et al. 2020). Additionally, a few types of research have been conducted to determine which machine learning classifier is the most appropriate and accurate for LULC mapping (Camargo et al. 2019; Jamali 2019). It has been discovered that ANN, SVM, and RF generally offer more accuracy compared to other conventional classifier algorithms, while SVM and RF are the best machine-learning techniques for LULC classification (Ma et al. 2017). The most extensively utilized algorithms in the research community for multispectral and hyperspectral image classification tasks are RF and SVM (Belgiu & Draguţ 2016; Georganos et al. 2018; Zhang et al. 2019; Talukdar et al. 2020).

SVM is a non-parametric supervised machine learning technique (Vapnik & Izmailov 2017) that was first made to solve binary classification problems (Maxwell et al. 2018). It is based on the concept of structural risk minimization (SRM), which maximizes and separates the hyper-plane and data points closest to the spectral angle mapper (SAM) of the hyper-plane. With the SVM algorithm, a hyperplane is first made based on the sample sets used for training. Then, the segmented objects are put into one of the known land cover classes. Support vectors are training samples or bordering samples that define the SVM's hyper-plane or margin (Shi et al. 2019). The best performance is demonstrated by the genetically optimized SVM using support vectors for both simulated and actual hyperspectral satellite data (Mathur & Foody 2008). The process's training quality will determine how effectively the SVM works. Utilizing linearly separable classes is the simplest method for training the SVM. If there is a vector W perpendicular to the linear hyper-plane (which determines the direction of the discriminating plane) and a scalar b showing the offset of the discriminating hyperplane from the origin, then these classes are considered to be linearly separable. This is the case if the training data with k number of samples is represented as ‘Xi, Yi,’ where I = 1, 2… k. If there is a vector W perpendicular to the linear hyperplane (which determines the direction of the discriminating plane) and a scalar b showing the offset of the discriminating hyperplane from the origin, then these classes are considered to be linearly separable. These classes are regarded as linearly separable if the training data with k number of samples is represented as (Xi, Yi), I = 1, 2… k where X € Rn is an n-dimensional space and y € (−1, +1) is a class label. To distinguish the data points in the two classes, class 1 is represented as −1 and class 2 is represented as +1, two hyper-planes can be utilized shown in Equations (3) and (4).
(3)
(4)
The five categories of the LULC study were bare soil, settlements, agricultural land, forests, and water bodies. A post-classification accuracy assessment was carried out to assess the correctness of the categorized maps. The accuracy of LULC maps has been investigated using methodology such as the Kappa coefficient (Rahman et al. 2019). The kappa coefficient was calculated using the formula below in Equation (5) (Mourya et al. 2020).
(5)
where m is the number of rows in the matrix, denotes the number of observations in row i and column i, Ti denotes the total number of categorized pixels in class i and Ai denotes the total number of actual data pixels in class i. In this study, the accuracy of LULC maps with randomly selected points was assessed using the Kappa coefficient technique. The points were picked to reflect each of the LULC courses fairly evenly and came from throughout the study area. The ground observations were acquired from the Google Earth Pro domain because field data for 1990 and 2005 was unavailable. In addition, ground observations for 2020 were derived partly via field visits and the Google Earth engine. The best method for correlating morphological assessment was chosen based on an accuracy test between the two methods.

In this aspect, several factors such as river meandering parameters, oxbow lake formation, and decadal LULC changes are analyzed to quantify river morphological changes (Hickin & Nanson 1984; Hooke 2013; Gogoi & Goswami 2014; Yousefi et al. 2016; Nath & Ghosh 2022).

The findings of the studies are presented, including changes in LULC in the study area over time, particular changes in meander characteristics through time, river migration analysis, as well as the effect of oxbow formation on river morphology.

LULC analysis

Using the ML and SVM classifier technique, training samples for each type of LULC (bare soil, settlement, agricultural land, forestry, and water body) were created for the study area. According to the accuracy assessment process, the overall accuracy level for 1990, 2005, and 2020 are suggested to be 88.70%, 90.32%, and 91.93%, respectively, with Kappa statistics for ML classifiers of 0.85, 0.87, and 0.89. The overall accuracy level for 1990, 2005, and 2020 is 90.03%, 91.2%, and 92.7%, respectively, with Kappa values for SVM classifiers of 0.87, 0.88, and 0.9. When compared to the ML technique, the SVM technique performs better according to kappa statistics. Table 2 and Figure 5(a)–5(d) are summarizes the results of the LULC study performed using the SVM algorithm.
Table 2

LULC analysis

LULC classes19902005Rise/loss of area 1990–2005 (sq. km)2020Rise/loss of area 2005–2020 (sq. km)Area covered by classes (1990)Area covered by classes (2020)
Water 59.19 57.56 −1.6 53.22 −1.34 10.76% 9.68% 
Forest 151.08 127.89 −23.19 114.83 −13.06 27.48% 20.88% 
Agriculture 48.33 64.29 +15.96 74.39 +10.1 8.79% 13.53% 
Bareland 270.01 259.61 −13.4 242.8 −15.81 49.11% 44.16% 
Settlement 21.16 40.42 +19.26 64.53 +24.11 3.84% 11.73% 
LULC classes19902005Rise/loss of area 1990–2005 (sq. km)2020Rise/loss of area 2005–2020 (sq. km)Area covered by classes (1990)Area covered by classes (2020)
Water 59.19 57.56 −1.6 53.22 −1.34 10.76% 9.68% 
Forest 151.08 127.89 −23.19 114.83 −13.06 27.48% 20.88% 
Agriculture 48.33 64.29 +15.96 74.39 +10.1 8.79% 13.53% 
Bareland 270.01 259.61 −13.4 242.8 −15.81 49.11% 44.16% 
Settlement 21.16 40.42 +19.26 64.53 +24.11 3.84% 11.73% 
Figure 5

(a) LULC mapping 1990. (b) LULC mapping 2005, (c) LULC mapping 2020 (d) LULC Analysis 1990–2020.

Figure 5

(a) LULC mapping 1990. (b) LULC mapping 2005, (c) LULC mapping 2020 (d) LULC Analysis 1990–2020.

Close modal
Figure 6

Changes in meander parameter.

Figure 6

Changes in meander parameter.

Close modal

Settlement increase over time was one of the most influential parameters in the study area. In 1990, settlement land covered 21.16 sq. km, 40.42 sq. km in 2005, and 64.53 sq. km in 2020. Other land use parameters are also affected by settlement or urbanization, such as agricultural land, which covered 48.33 sq. km of the study area in 1990, 64.29 sq. km in 2005, and 74.39 sq. km in 2020. Increased urbanization was also one of the key influencing factors for deforestation. In 1990, forestry covered over 151.08 sq. km, which decreased to 127.89 sq. km in 2005 and 114.83 sq. km in 2020, according to the LULC mapping. The area matrices of the various LULC types indicate how the Bareland has been converted into other land classes, particularly settlement, and how this has influenced the gradual increase in agricultural land area. Agricultural land and built-up area have shown a significant upward trend, whereas forestation and barren areas have shown a declining trend. As a result of the disturbance caused by settlement, agricultural land changes, and a significant decline in forestation, some of the region's important influencing factors were founded.

The findings are comparable to those of Gebrehiwot et al. (2014) in the Blue Nile basin's Birr and Upper Didesa basins, as well as Gashaw et al. (2017) in northeastern Ethiopia's Dera area, where urbanization and agricultural land have increased while forest land has decreased.

Changes in meander parameters

The study reach's channel centerlines were digitized, and 21 meander loops were chosen for the time interval between the study period and the end of the study period (Figure 6). From 1990 to 2020, morphological parameters for all meander loops were measured, and the morphological change of meander loops was estimated (Table 3).

Table 3

Changes in meander parameter

Section202020102000199020202010200019902020201020001990
 Radius of curvature Center line distance Meander neck length 
P1 864 1,019 881 927 5,873 6,015 6,128 5,961 3,684 3,741 3,762 3,658 
P2 1,906 1,833 1,525 1,802 5,100 5,697 5,506 5,373 3,677 3,984 4,183 4,044 
P3 3,838 3,501 2,993 3,707 9,200 8,836 9,228 9,184 6,870 6,859 6,830 6,915 
P4 3,265 3,015 2,982 3,903 9,059 8,497 8,500 8,443 7,413 7,530 7,541 7,599 
P5 2,579 2,573 2,324 1,444 7,032 6,862 6,976 6,860 4,831 4,731 4,728 4,689 
P6 895 1,267 755 1,223 9,224 10,119 10,050 19,485 4,416 4,231 4,324 4,334 
P7 637 1,447 1,241 1,114 3,754 3,418 3,201 3,203 1,947 1,877 1,871 2,154 
P8 592 563 351 384 3,332 2,899 3,026 4,412 1,891 1,762 1,884 1,800 
P9 1,105 1,089 795 612 4,100 4,146 4,018 3,983 2,199 2,221 2,266 2,228 
P10 876 499 355 296 4,686 4,702 4,919 4,820 1,677 1,538 1,681 1,640 
P11 849 910 775 848 4,963 4,843 4,830 4,867 1,441 1,443 1,451 1,500 
P12 573 578 536 570 4,882 4,928 4,885 4,902 1,103 1,116 1,107 1,049 
P13 907 711 846 790 3,996 3,772 3,760 3,792 1,861 1,770 1,780 1,769 
P14 327 262 297 469 2,996 3,275 3,155 3,225 932 1,096 1,053 972 
P15 962 717 889 759 3,976 3,909 3,941 3,984 1,779 1,759 1,715 1,789 
P16 843 651 805 693 7,886 8,036 7,835 7,787 1,348 1,379 1,278 1,365 
P17 1,287 794 1,208 671 6,679 6,738 5,700 6,662 3,156 3,094 3,063 3,046 
P18 420 475 509 775 5,873 5,789 5,695 5,558 1,321 1,285 1,268 1,280 
P19 886 917 833 712 4,548 4,296 4,279 4,202 1,875 1,848 1,890 1,843 
P20 1,798 1,435 1,335 790 3,724 3,475 2,944 2,563 2,793 2,784 2,640 2,425 
P21 630 691 1,097 614 3,120 3,223 3,093 3,020 2,316 2,265 2,437 2,498 
 Axis length Sinuosity River width 
P1 1,789 1,820 1,894 1,843 1.59 1.60 1.62 1.63 161 152 156 160 
P2 1,447 1,507 1,552 1,451 1.38 1.42 1.31 1.32 163 153 152 158 
P3 2,844 2,689 2,694 2,651 1.33 1.29 1.35 1.32 169 154 141 140 
P4 1,311 1,371 1,408 1,501 1.22 1.13 1.12 1.11 152 117 114 125 
P5 2,207 2,145 2,213 2,201 1.46 1.45 1.47 1.46 133 113 108 106 
P6 3,989 3,978 3,875 4,400 2.08 2.39 2.32 4.5 140 122 118 115 
P7 1,254 1,295 1,268 921 1.92 1.82 1.71 1.48 148 119 115 109 
P8 1,309 1,142 1,181 1,797 1.76 1.64 1.60 2.45 145 110 114 121 
P9 1,586 1,698 1,641 1,564 1.86 1.87 1.77 1.78 179 117 115 125 
P10 1,427 1,687 1,623 1,455 2.79 3.05 2.92 2.93 143 133 119 133 
P11 1,727 1,901 1,744 1,871 3.44 3.55 3.32 3.24 131 113 121 118 
P12 2,031 2,093 2,140 2,063 4.42 4.41 4.41 4.67 148 110 115 113 
P13 1,539 1,495 1,452 1,374 2.15 2.13 2.11 2.14 135 107 108 111 
P14 1,362 1,523 1,496 1,497 3.21 2.98 2.99 3.31 169 158 150 147 
P15 1,478 1,554 1,467 1,516 2.23 2.22 2.29 2.23 146 127 131 124 
P16 2,844 2,906 2,973 2,741 5.85 5.82 6.13 5.70 131 109 121 115 
P17 2,210 2,254 2,234 2,166 2.11 2.17 1.86 2.18 135 121 110 113 
P18 2,670 2,580 2,590 2,532 4.44 4.50 4.49 4.34 129 115 117 119 
P19 1,891 1,877 1,720 1,543 2.42 2.32 2.26 2.27 153 140 135 125 
P20 877 818 463 217 1.33 1.24 1.11 1.05 135 127 130 127 
P21 656 806 538 278 1.34 1.42 1.26 1.20 143 135 125 123 
Section202020102000199020202010200019902020201020001990
 Radius of curvature Center line distance Meander neck length 
P1 864 1,019 881 927 5,873 6,015 6,128 5,961 3,684 3,741 3,762 3,658 
P2 1,906 1,833 1,525 1,802 5,100 5,697 5,506 5,373 3,677 3,984 4,183 4,044 
P3 3,838 3,501 2,993 3,707 9,200 8,836 9,228 9,184 6,870 6,859 6,830 6,915 
P4 3,265 3,015 2,982 3,903 9,059 8,497 8,500 8,443 7,413 7,530 7,541 7,599 
P5 2,579 2,573 2,324 1,444 7,032 6,862 6,976 6,860 4,831 4,731 4,728 4,689 
P6 895 1,267 755 1,223 9,224 10,119 10,050 19,485 4,416 4,231 4,324 4,334 
P7 637 1,447 1,241 1,114 3,754 3,418 3,201 3,203 1,947 1,877 1,871 2,154 
P8 592 563 351 384 3,332 2,899 3,026 4,412 1,891 1,762 1,884 1,800 
P9 1,105 1,089 795 612 4,100 4,146 4,018 3,983 2,199 2,221 2,266 2,228 
P10 876 499 355 296 4,686 4,702 4,919 4,820 1,677 1,538 1,681 1,640 
P11 849 910 775 848 4,963 4,843 4,830 4,867 1,441 1,443 1,451 1,500 
P12 573 578 536 570 4,882 4,928 4,885 4,902 1,103 1,116 1,107 1,049 
P13 907 711 846 790 3,996 3,772 3,760 3,792 1,861 1,770 1,780 1,769 
P14 327 262 297 469 2,996 3,275 3,155 3,225 932 1,096 1,053 972 
P15 962 717 889 759 3,976 3,909 3,941 3,984 1,779 1,759 1,715 1,789 
P16 843 651 805 693 7,886 8,036 7,835 7,787 1,348 1,379 1,278 1,365 
P17 1,287 794 1,208 671 6,679 6,738 5,700 6,662 3,156 3,094 3,063 3,046 
P18 420 475 509 775 5,873 5,789 5,695 5,558 1,321 1,285 1,268 1,280 
P19 886 917 833 712 4,548 4,296 4,279 4,202 1,875 1,848 1,890 1,843 
P20 1,798 1,435 1,335 790 3,724 3,475 2,944 2,563 2,793 2,784 2,640 2,425 
P21 630 691 1,097 614 3,120 3,223 3,093 3,020 2,316 2,265 2,437 2,498 
 Axis length Sinuosity River width 
P1 1,789 1,820 1,894 1,843 1.59 1.60 1.62 1.63 161 152 156 160 
P2 1,447 1,507 1,552 1,451 1.38 1.42 1.31 1.32 163 153 152 158 
P3 2,844 2,689 2,694 2,651 1.33 1.29 1.35 1.32 169 154 141 140 
P4 1,311 1,371 1,408 1,501 1.22 1.13 1.12 1.11 152 117 114 125 
P5 2,207 2,145 2,213 2,201 1.46 1.45 1.47 1.46 133 113 108 106 
P6 3,989 3,978 3,875 4,400 2.08 2.39 2.32 4.5 140 122 118 115 
P7 1,254 1,295 1,268 921 1.92 1.82 1.71 1.48 148 119 115 109 
P8 1,309 1,142 1,181 1,797 1.76 1.64 1.60 2.45 145 110 114 121 
P9 1,586 1,698 1,641 1,564 1.86 1.87 1.77 1.78 179 117 115 125 
P10 1,427 1,687 1,623 1,455 2.79 3.05 2.92 2.93 143 133 119 133 
P11 1,727 1,901 1,744 1,871 3.44 3.55 3.32 3.24 131 113 121 118 
P12 2,031 2,093 2,140 2,063 4.42 4.41 4.41 4.67 148 110 115 113 
P13 1,539 1,495 1,452 1,374 2.15 2.13 2.11 2.14 135 107 108 111 
P14 1,362 1,523 1,496 1,497 3.21 2.98 2.99 3.31 169 158 150 147 
P15 1,478 1,554 1,467 1,516 2.23 2.22 2.29 2.23 146 127 131 124 
P16 2,844 2,906 2,973 2,741 5.85 5.82 6.13 5.70 131 109 121 115 
P17 2,210 2,254 2,234 2,166 2.11 2.17 1.86 2.18 135 121 110 113 
P18 2,670 2,580 2,590 2,532 4.44 4.50 4.49 4.34 129 115 117 119 
P19 1,891 1,877 1,720 1,543 2.42 2.32 2.26 2.27 153 140 135 125 
P20 877 818 463 217 1.33 1.24 1.11 1.05 135 127 130 127 
P21 656 806 538 278 1.34 1.42 1.26 1.20 143 135 125 123 

Between 1990 and 2020, mean river width estimates reflect an increasing trend in river width. The average river width in 1990 was 125.09 meters, but by 2020 it had increased to 147.07 meters. The average sinuosity of the sections decreased as the centerline distance between time spans decreased. According to the study, the average centerline distance between sections was 5,823.14 meters in 1990 and 5,428.71 meters in 2020. Therefore, during the two decades, the average sinuosity (C) declined. According to sinuosity analysis, P6 indicates the highest content in sinuosity across the research period. According to the research, the sinuosity of the sector was 4.5 in 1990 and 2.08 in 2020. This is because, between 1990 and 2000, an oxbow formed near Sonaimukh (Kaptanpur), cutting off a substantial stretch of the river route from the main Barak. As a result, in this section of the river, a significant decrease in centerline distance (10.26 km) was found between study periods. The highest radius of curvature was discovered in section P3, and the lowest radius of curvature was discovered in section P18. Within the study period, the most significant variation in radius of curvature was observed in P5, P6, P7, P10, and P17. Paired t-tests were used between 1990 and 2020 to determine the significance of variance in the river meandering parameters.

Only the meander width showed a significant change within a 5% significant level between the research periods, according to the samples paired t-test (Table 4). (W). The radius of curvature (R), centerline distance (S), sinuosity (C), meander neck length (L), and meander axis length (A) do not change significantly between periods. Where R-squared (R2) is a statistical measure that measures the amount of variation explained by other independent variables in a regression model for a dependent variable. The degree of marginal significance in a statistical hypothesis test is referred to as the p-value. Acceptance of the null hypothesis is indicated by the ‘p’ value or probability value where strong significance variation is indicated by a p-value less than 5%. The outcomes of variable correlation are summarized in Table 5.

Table 4

Performance of paired t-test

Meander parameterAverage difference between variableStandard deviationT valueDegree of freedomSig. (* significance)
Radius of curvature 139.8 941.54 −1.45 20 0.08 
Center line distance −394.42 2,898.09 0.79 20 0.22 
Meander neck length −3.19 1,794.12 0.09 20 0.46 
Sinuosity −0.09 1.27 0.75 20 0.22 
Axis length 41.23 817.70 −0.72 20 0.23 
River width 21.95 18.22 −8.25 20 3.56 × 10−8
Meander parameterAverage difference between variableStandard deviationT valueDegree of freedomSig. (* significance)
Radius of curvature 139.8 941.54 −1.45 20 0.08 
Center line distance −394.42 2,898.09 0.79 20 0.22 
Meander neck length −3.19 1,794.12 0.09 20 0.46 
Sinuosity −0.09 1.27 0.75 20 0.22 
Axis length 41.23 817.70 −0.72 20 0.23 
River width 21.95 18.22 −8.25 20 3.56 × 10−8
Table 5

Variable correlation and regression analysis

Variable correlation R2 R(p) S(p) L(p) A(p) W(p) Intercept 
0.92 0.08 5.63 × 10−6 1.14 × 10−7 0.02 0.16 −0.10 
Variable correlation R2 C(p) S(p) L(p) A (p) W(p) Intercept 
0.85 0.07 0.21 0.008 0.80 0.60 −5.38 
Variable correlation R2 C(p) R(p) L(p) A (p) W(p) Intercept 
0.97 5.63 × 10−6 0.21 2.14 × 10−7 2.03 × 10−5 0.04 901.20 
Variable correlation R2 C(p) R(p) S(p) A (p) W(P p) Intercept 
0.98 1.14 × 10−7 0.008 2.14 × 10−7 0.005 0.06 −469.61 
Variable correlation R2 C(p) R(p) S(p) L (p) W(p) Intercept 
0.86 0.02 0.80 2.03 × 10−5 0.005 0.12 −692.6 
Variable correlation R2 C(p) R(p) S(p) L (p) A(p) Intercept 
0.34 0.15 0.60 0.04 0.06 0.12 135.13 
Regression equation C = −0.10 + 0.00,043R + 0.0001S − 0.001L + 0.0007A + 0.01W 
   R = −5.38 + 453.5 C −0.38S +0.96L + 0.08 A −4.13W 
   S = 901.20 + 737.1C − 0.26R + 1.18L + 0.88 A − 12.52W 
   L = −469.61–606.73 C +0.39 R + 0.71S − 0.52A + 8.99W 
   A = −692.6 − 449.19C + 0.05R + 0.80S − 0.79L + 9.35W 
   W = 135.13 + 11.98C − 0.004R − 0.02S + 0.02L + 0.01 A 
Variable correlation R2 R(p) S(p) L(p) A(p) W(p) Intercept 
0.92 0.08 5.63 × 10−6 1.14 × 10−7 0.02 0.16 −0.10 
Variable correlation R2 C(p) S(p) L(p) A (p) W(p) Intercept 
0.85 0.07 0.21 0.008 0.80 0.60 −5.38 
Variable correlation R2 C(p) R(p) L(p) A (p) W(p) Intercept 
0.97 5.63 × 10−6 0.21 2.14 × 10−7 2.03 × 10−5 0.04 901.20 
Variable correlation R2 C(p) R(p) S(p) A (p) W(P p) Intercept 
0.98 1.14 × 10−7 0.008 2.14 × 10−7 0.005 0.06 −469.61 
Variable correlation R2 C(p) R(p) S(p) L (p) W(p) Intercept 
0.86 0.02 0.80 2.03 × 10−5 0.005 0.12 −692.6 
Variable correlation R2 C(p) R(p) S(p) L (p) A(p) Intercept 
0.34 0.15 0.60 0.04 0.06 0.12 135.13 
Regression equation C = −0.10 + 0.00,043R + 0.0001S − 0.001L + 0.0007A + 0.01W 
   R = −5.38 + 453.5 C −0.38S +0.96L + 0.08 A −4.13W 
   S = 901.20 + 737.1C − 0.26R + 1.18L + 0.88 A − 12.52W 
   L = −469.61–606.73 C +0.39 R + 0.71S − 0.52A + 8.99W 
   A = −692.6 − 449.19C + 0.05R + 0.80S − 0.79L + 9.35W 
   W = 135.13 + 11.98C − 0.004R − 0.02S + 0.02L + 0.01 A 

For each variable, a regression equation was developed using a regression analysis between the variable. Except for river width (W), all dependent variables had R2 statistics that were satisfactory.

Effect of river morphology due to river migration and oxbow development

To better understand river migration in the study area, a section-by-section analysis was performed. Section-by-section river migration and rate of variations are summarized in Table 6. Section-wise total migration of the river from 1990 to 2020 is shown in Figure 7(a).
Table 6

River migration analysis over the study period

SecTotal migration 1990–2020Migration rate (1990–2000) (m/year)
Avg. m/yearMigration rate (1990–2010) (m/year)
Avg. m/yearMigration rate (1990–2020) (m/year)
Avg. m/year
RightLeftRightLeftRightLeft
P1 30.69 1.01 0.63 0.82 0.58 0.55 0.56 0.47 0.51 0.49 
P2 27.18 0.89 0.63 0.76 0.72 0.64 0.68 0.16 0.36 0.26 
P3 31.26 2.24 1.06 1.65 0.68 0.36 0.52 0.21 0.35 0.28 
P4 67.3 3.12 1.18 2.15 3.86 1.83 2.84 2.92 1.12 2.02 
P5 106.35 5.87 2.11 3.99 5.3 2.77 4.03 4.38 2.54 3.46 
P6 199.48 8.17 2.92 5.54 5.02 16.32 10.67 8.16 5.26 6.71 
P7 393.75 13.73 7.88 10.8 12.89 10.82 11.85 6.01 9.99 9.0 
P8 105.16 3.37 2.64 3.0 3.4 1.31 2.35 1.58 1.5 1.54 
P9 77.73 4.9 3.21 4.05 2.42 1.63 2.02 1.22 1.98 1.6 
P10 34.94 1.47 2.14 1.8 1.01 1.35 1.18 0.6 0.62 0.61 
P11 53.92 3.28 2.79 3.03 1.68 2.07 1.87 1.44 1.02 1.23 
P12 50.91 4.75 3.55 4.15 3.65 1.94 2.79 1.53 1.33 1.43 
P13 37.36 1.3 1.17 1.23 1.33 1.07 1.2 0.55 0,42 0.48 
P14 22.77 1.18 1.07 1.12 0.41 0.62 0.51 0.35 0.40 0.37 
P15 80.62 5.76 2.74 4.25 2.61 2.11 2.36 2.63 2.27 2.45 
P16 55.04 1.84 1.97 1.9 2.25 2.17 2.21 1.21 1.15 1.18 
P17 78.1 2.38 2.89 2.63 1.91 3.02 2.46 0.94 1.57 1.25 
P18 52.62 2.84 2.21 2.52 1.58 2.22 1.9 0.73 1.32 1.02 
P19 83.39 9.92 5.18 7.55 5.72 4.33 5.02 4.22 2.94 3.58 
P20 602.86 5.99 15.99 10.99 13.84 13.22 13.33 15.04 11.29 13.16 
P21 350.68 4.15 4.63 4.41 8.54 11.65 10.09 5.73 4.83 5.28 
SecTotal migration 1990–2020Migration rate (1990–2000) (m/year)
Avg. m/yearMigration rate (1990–2010) (m/year)
Avg. m/yearMigration rate (1990–2020) (m/year)
Avg. m/year
RightLeftRightLeftRightLeft
P1 30.69 1.01 0.63 0.82 0.58 0.55 0.56 0.47 0.51 0.49 
P2 27.18 0.89 0.63 0.76 0.72 0.64 0.68 0.16 0.36 0.26 
P3 31.26 2.24 1.06 1.65 0.68 0.36 0.52 0.21 0.35 0.28 
P4 67.3 3.12 1.18 2.15 3.86 1.83 2.84 2.92 1.12 2.02 
P5 106.35 5.87 2.11 3.99 5.3 2.77 4.03 4.38 2.54 3.46 
P6 199.48 8.17 2.92 5.54 5.02 16.32 10.67 8.16 5.26 6.71 
P7 393.75 13.73 7.88 10.8 12.89 10.82 11.85 6.01 9.99 9.0 
P8 105.16 3.37 2.64 3.0 3.4 1.31 2.35 1.58 1.5 1.54 
P9 77.73 4.9 3.21 4.05 2.42 1.63 2.02 1.22 1.98 1.6 
P10 34.94 1.47 2.14 1.8 1.01 1.35 1.18 0.6 0.62 0.61 
P11 53.92 3.28 2.79 3.03 1.68 2.07 1.87 1.44 1.02 1.23 
P12 50.91 4.75 3.55 4.15 3.65 1.94 2.79 1.53 1.33 1.43 
P13 37.36 1.3 1.17 1.23 1.33 1.07 1.2 0.55 0,42 0.48 
P14 22.77 1.18 1.07 1.12 0.41 0.62 0.51 0.35 0.40 0.37 
P15 80.62 5.76 2.74 4.25 2.61 2.11 2.36 2.63 2.27 2.45 
P16 55.04 1.84 1.97 1.9 2.25 2.17 2.21 1.21 1.15 1.18 
P17 78.1 2.38 2.89 2.63 1.91 3.02 2.46 0.94 1.57 1.25 
P18 52.62 2.84 2.21 2.52 1.58 2.22 1.9 0.73 1.32 1.02 
P19 83.39 9.92 5.18 7.55 5.72 4.33 5.02 4.22 2.94 3.58 
P20 602.86 5.99 15.99 10.99 13.84 13.22 13.33 15.04 11.29 13.16 
P21 350.68 4.15 4.63 4.41 8.54 11.65 10.09 5.73 4.83 5.28 
Figure 7

(a) Section-wise river migration analysis. (b) River Migration near Sonabarighat (section P6). (c) Identification of Future oxbow formation location.

Figure 7

(a) Section-wise river migration analysis. (b) River Migration near Sonabarighat (section P6). (c) Identification of Future oxbow formation location.

Close modal

Total migration was calculated from the river centerline position from 1990 to 2020. According to river migration analysis, the areas P7, P20, and P21 were the most severely impacted. P20 section was the most vulnerable of all the sections, as it was found near 602.86 meters of land that had been impacted by river migration throughout the study period. P1, P2, P3, and P14 were among the most stable sections. In such sections, annual shifting of fewer than 2 meters has been recorded. Whereas for P20, the rate of migration was also considerably high when compared to other sections of the study area, where near 13.16 meters of moving was still causing land degradation every year because to NH53, this portion of the river is now more vital to Pachgram (Section P20). The only highway connecting Silchar to the rest of the country runs through it.

During the study period, a large section of the river path was cut off from the main Barak near Sonaimukh (Kaptanpur), and an oxbow formed over section P6. One of the main explanations for the largest sinuosity fluctuation across the decade was oxbow formation (Figure 3(c)).

As a result, overall river migration and migration rates were much higher in P6, P7, and P8 than in other sections. As Sonabarighat Pt II (Section P7) has been subjected to significant migration due to river migration. Piers were built here between 2012 and 2013 (Figure 7(b)), which were used to connect the two ends of the river in preparation for constructing a bridge. As per collective information, due to some issues, construction activities were suspended in 2013. Over a decade, the river has moved an enormous amount of land, reaching the last piers of the bridge. Hare, over 150-meter migration was observed within the last decade. During the field visit, it was observed that a roadway near Govindapur (Section P6) had entirely deteriorated in the recent decade due to massive migration activity. Within this zone, there is still around 6.71-meter shifting per year. The upstream river section is still suffering from river migration, and oxbow development is one factor that makes the area vulnerable to river migration. As a result, determining the future oxbow tendency ability zone is crucial for river management, especially in this type of river where the township is close to the flood plain. An oxbow is formed when two meander necks get near together (Figure 3(c)). Following the river's path, some future oxbow tendency-able zones inside the study area were notified (Figure 7(c)).

As a result, the riskiest zone between sections P6 and P7 is Dhamalia, due to future oxbow tendencies. This section of the river (Dhamalia) is more prone to oxbow formation. If the river is not protected immediately, oxbows will form, cutting Kanakpur, Gangapur, and Neairgram off the main Barak (between sections P6 and P7). The distance between the river's two meander neck ends was over 750 meters in 1990 and is now 490 meters. According to migration analysis, the gap is constantly decreasing. Baghadhar (section P10) and Manikpur (section P16) are similar, with the observed gap between the two meander necks decreasing significantly during the study period. If an oxbow formation takes place, it will be affected similarly to the oxbow formed in Sonaimukh (section P6), resulting in downstream instability.

River system morphological changes are linked like a chain (Lagasse et al. 2004). Centerline distance (S) decreased due to recorded changes in the radius of curvature (R) induced by decreasing the internal arc. Axis length (A) also increased during this process. The sinuosity of a meander is also affected by changes in other meander parameters. As a result, the average sinuosity (C) has decreased over time. Changes in river width (W) are inversely proportional to changes in centerline distance (S) (Yousefi et al. 2016). If the length of the meander neck (L) remains constant over time, the radius of curvature (R) will decrease, and meander sinuosity (C) will increase (Crosato 2008; Heo et al. 2009). As a result, the radius of curvature and sinuosity were inversely proportional. Here, the average radius of curvature over the sections shows increasing trends and the meander neck length (L) remains almost constant over the study period; therefore, average sinuosity shows a decreasing trend. In this study, the average radius of curvature over the sections has been growing, and the average sinuosity has decreased. In general, the results show a direct link between changes in the radius of curvature and changes in sinuosity, the meander neck length, centerline distance, and meander axis length. The regression analysis is performed to identify the inter-relationship among meandering parameters. The performance of regression analysis is checked by R2 value (Table 5). It shows in Table 5 that the inter-relationship between each parameter with other meandering parameters is fairly good as R2 statistics varying in between 0.85 and 0.98. The regression equation for L concerning other meandering parameters is a good fit. On the other hand, the regression equation for W concerning other meandering parameters is a poor fit as R2 is equal to 0.34.

The study area's water body, forest, and bareland type areas decreased by 6.92 sq. km, 34.36 sq. km, and 22.27 sq. km, respectively, between 1990 and 2020. On the other hand, agricultural and settlement areas increased by 15.4 and 48.15 sq. km, respectively. Land use parameters, such as the rate of settlement and the rate of agriculture practiced, fluctuated significantly over a decade in the current study, with a constant loss of forest land observed throughout the study area. Integrated river basin management strategies, regulating rapid settlement growth and agriculture practices, and reforesting significantly mitigate the rapid rates of LULC conversions at a watershed. This study shows that the upstream section (sections P1 to P5) has the most hilly-dense forestation and has less urbanization than the downstream section of the study area. As a result, river migration in those sections was significantly lower than in other sections. According to LULC mapping, agriculture operations increased at their highest between (sections P6 to P10), with P5 to P8 being some of the most vulnerable sections identified by the migration study. The most vulnerable sections of the research area were sections P20 and P21, in which agriculture has a slightly greater influence near the river bank. One of the primary contributing factors of section P6 for considerable fluctuation in sinuosity is oxbow development. As a result of the river's morphological changes, it lost over 10.26 kilometers of length from the main Barak. Hence, this part and the downstream segment suffered substantially due to river migration. On the other hand, the most probable future oxbow-prone zones identified in this analysis are Dhamalia, Baghadhar, and Manikpur. Deforestation and urbanization are two major contributors to river basin LULC changes, which are linked to river morphology (Woldemichael et al. 2012; Behera et al. 2018; Li et al. 2020). This finding supports the hypothesis that LULC changes affect river morphology. River morphological changes often take decades to become noticeable, although a human intervention can boost channel alterations (Rinaldi et al. 2005). Therefore, the rate of morphological changes, as well as future oxbow formation, will accelerate more if land use changes as significantly as those previously seen. This study suggests that more research to be done to detect prospective changes in terms of LULC changes, as well as a long-term predicted model for predicting future morphological changes in rivers.

Understanding geomorphic responses to global land use change requires a historical perspective. Prior conditions, such as previous disturbances to which systems are responding must be considered while making a change. The current study shows that utilizing RS and GIS, multi-temporal satellite data (Landsat) may be used to monitor river morphological activity with LULC and meandering parameter variations. The study's main purpose was to use remote sensing data to analyze and detect morphological changes in a section of the Barak River from 1990 to 2020. The average centerline distance (S) in this analysis decreased over time, which influenced the change in sinuosity of that section. Before 2000, a cutoff was created between sections P6 and P7, reducing the main centerline distance (S) from the river to around 10.26 kilometers. Oxbow formation was one of the key factors for the highest sinuosity fluctuation over the decade. Therefore, identifying the formation zone for future oxbow formation was essential to achieving the aims of this research. Here, the most dangerous zones for future oxbow development were Dhamalia (P6 to P7), Baghadhar (P10), and Manikpur (P16). If the land is not protected immediately, a large portion of the main river will be cut off, making the area more vulnerable downstream. All dependent variables had acceptable R-Square statistics, except river width (W). Where river width (W) was also affected due to changes in river centerline distance(S), therefore average width of the river shows the incremental trend. Then a regression equation was established for each meandering variable.

Observations of geomorphic responses to land use may only be accurately interpreted if a thorough understanding of prior land use changes and patterns was obtained. According to the investigation critically from coarse resolution imageries, water, forest, and bareland type areas decreased in the study area, whereas agricultural and settlement areas increased during the study period. Agricultural activities are influenced by urbanization or settlement rise. Intensive agricultural land was largely afflicted by river migration. Changes in river migration rates substantially impact other river meandering parameters and morphology. LULC changes and meandering river parameters are all linked to river morphology. The parameters of meanders can be used to forecast future trends in river morphology and meander evolution. Understanding how meander parameters are changing could lead to better river management and mitigation of the damage caused by these changes.

In this investigation, images with a 30 m pixel reconstruction are employed. This is the consequence of colour misunderstanding between Landsat image stacks. In contrast, the colour associated with urbanization resembles bare land, and the colour associated with agriculture resembles forestation. As a result, selecting the correct class using only Google Maps is very much challenging. As a result, the necessary site-specific knowledge should be required for this type of assessment using coarse-resolution images.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.

Abdullah
M. D. Y. A.
,
Masrur
A.
,
Adnan
M. S. G.
,
Baky
M. A. A.
,
Hassan
K.
&
Dewan
A.
2019
Spatio-temporal patterns of land use/land cover change in the heterogeneous coastal region of Bangladesh between 1990 and 2017
.
Remote Sensing
11
,
790
.
doi:10.3390/rs11070790
.
Annayat
W.
&
Sil
B. S.
2020b
Assessing channel morphology and prediction of centerline channel migration of the Barak River using geospatial techniques
.
Bulletin of Engineering Geology and the Environment
79
,
5161
5183
.
https://doi.org/10.1007/s10064-020-01894-9
.
Annayat
W.
,
Ashwini
K.
&
Sil
B. S.
2022
Channel migration detection and landuse/landcover changes for Barak River in Northeast India
.
Journal of the Geological Society of India
98
(
7
),
981
990
.
Arnaud-Fassetta
G.
,
Astrade
L.
,
Bardou
E.
,
Corbonnois
J.
,
Delahaye
D.
,
Fort
F.
,
Gautier
E.
,
Jacob
J.
,
Peiry
J. L.
,
Piégay
H.
&
Penven
M. J.
2009
Fluvial geomorphology and flood-risk management. géomorphologiefluvialeetgestion des risquesfluviaux
.
Géomorphologie. Relief, Processus, Environnement
2
,
109
128
.
Arohunsoro
S. J.
,
Owolabi
J. T.
&
Omotoba
N. I.
2014
Watershed management and ecological hazards in an urban environment: the case of river Ajilosun in Ado Ekiti, Nigeria
.
European Journal of Academic Essays
1
(
2
),
17
23
.
Barasa
B. N.
&
Pereram
E. D. P.
2018
Analysis of land-use change impacts on flash flood occurrences in the Sosiani River basin Kenya
.
International Journal of River Basin Management
16
(
2
),
179
188
.
ISSN: 1571-5124
.
Bardhan
M.
1993
Channel stability of Barak river and its tributaries between Manipur- Assam and Assam- Bangladesh borders as seen from satellite imagery
. In:
Proc. Nat. Syrup. on Remote Sensing Applications for Resource Management with Special Emphasis on N.E. Region
,
Nov. 25–27
,
held in Guwahati
, pp.
481
485
.
Barman
T.
2016
Relation of riverbank erosion with agricultural landuse practices on floodplain of River Jaldhaka (Mansai) at Tikiner Char, Bhowerthana Village, Coochbehar, West Bengal, India
.
International Journal of Research in Geography, (IJRG)
.
ISSN 2454-8685. http://dx.doi.org/10.20431/2454-8685.0202001
.
Bhattacharya, S. S., Kim, K. H., Das, S., Uchimiya, M., Jeon, B. H., Kwon, E. & Szulejko, J. E. 2016 A review on the role of organic inputs in maintaining the soil carbon pool of the terrestrial ecosystem. Journal of environmental management 167, 214–227.
Behera
M. D.
,
Tripathi
P.
,
Das
P.
,
Srivastava
S. K.
,
Roy
P. S.
,
Joshi
C.
,
Behera
P. R.
,
Deka
J.
,
Kumar
P.
,
Khan
M. L.
,
Tripathi
O. P.
,
Dash
T.
&
Krishnamurthy
Y. V. N.
2018
Remote sensing based deforestation analysis in Mahanadi and Brahmaputra river basin in India since 1985
.
Journal of Environmental Management
206
,
1192
1203
.
https://doi.org/10.1016/j.jenvman.2017.10.015
.
Belgiu
M.
&
Draguţ
L.
2016
Random forest in remote sensing: a review of applications and future directions
.
ISPRS International Journal of Geo-Information
114
,
24
31
.
Brierley
G. J.
&
Fryirs
K.
2005
Geomorphology and River Management: Applications of the River Styles Framework
.
Blackwell
,
Oxford
, p.
398
.
Choudhury
P.
,
Roy
P.
,
Nongthombam
J.
,
Ullah
N.
,
Devi
A.
&
Debbarman
S.
2014
Flood Damage Mitigation Measures for Barak Valley in South Assam Including Effects of Climate Change. Assam State Disaster Management Final Report
.
Clark
J. J.
&
Wilcock
P. R.
2000
Effects of land-use change on channel morphology in northeastern Puerto Rico
.
Geological Society of America Bulletin
112
(
12
),
1763
1777
.
Crosato
A.
2008
Analysis and Modelling of River Meandering
.
IOS Press
,
Amsterdam
, p.
251
.
Das
P.
2012
Study of Barak river meander and associated hazard around Silchar Town, Assam, using remote sensing and GIS
.
Earth Science India
5
(
II
),
51
59
.
EISSN: 0974–8350
.
Das
T.
&
Das
A. K.
2014
Mapping and identification of homegardens as a component of the trees outside forests using remote sensing and geographic information system
.
Journal of the Indian Society of Remote Sensing
42
(
1
),
233
242
.
https://doi.org/ 10.1007/s12524-013-0310-3
.
Das
J. D.
,
Dutta
T.
&
Saraf
A. K.
2007
Remote sensing and GIS application in change detection of the Barak river channel, NE India
.
Journal of the Indian Society of Remote Sensing
35
(
4
),
301
312
.
Das
K. T.
,
Halder
H. K.
&
Gupta
I. D.
2014
River bank erosion induced human displacement and its consequences
.
Living Reviews in Landscape Research
.
https://doi.org/10.12942lrlr-2014-3
.
Debnath
J.
,
Das
N.
,
Ahmed
I.
&
Bhowmik
M.
2017
Channel migration and its impact on land-use/land cover using RS andGIS: a study on Khowai River of Tripura, North-East India
.
The Egyptian Journal of Remote Sensing and Space Science
20
(
2017
),
197
210
.
Deng
L.
,
Kim
D. G.
,
Li
M.
,
Huang
C.
,
Liu
Q.
,
Cheng
M.
,
Shangguan
Z.
&
Peng
C.
2019
Land-use changes driven by ‘Grain for Green'program reduced carbon loss induced by soil erosion on the Loess Plateau of China
.
Global and Planetary Change
177
,
101
115
.
Galster
J. C.
,
Pazzaglia
F. J.
&
Germanoski
D.
2008
Measuring the impact of urbanization on channel widths using historic aerial photographs and modern surveys 1
.
JAWRA Journal of the American Water Resources Association
44
(
4
),
948
960
.
Gashaw
T.
,
Tulu
T.
,
Argaw
M.
&
Worqlul
A. W.
2017
Evaluation and prediction of land use/ land cover changes in the Andassa watershed, Blue Nile Basin, Ethiopia
. Environmental Systems Research
6
,
17
.
doi:10.1186/s40068-017-0094-5
.
Gebrehiwot
S. G.
,
Bewket
W.
,
Gärdenäs
A. I.
&
Bishop
K.
2014
Forest cover change over four decades in the Blue Nile Basin, Ethiopia: comparison of three watersheds
.
Regional Environmental Change
14
(
1
).
doi:10.1007/s10113-013-0483-x
.
Georganos
S.
,
Grippa
T.
,
Vanhuysse
S.
,
Lennert
M.
,
Shimoni
M.
&
Wolff
E.
2018
Very high resolution object-based land use–land cover urban classification using extreme gradient boosting
.
IEEE Geoscience and Remote Sensing Letters
15
(
4
),
607
611
.
Gogoi
C.
&
Goswami
C.
2014
A study on channel migration of the Subansiririver in Assam using remote sensing and GIS technology
.
Current Science
106
(
8
),
1113
1120
.
Gregory
K. J.
2006
The human role in changing river channels
.
Geomorphology
79
(
3–4
),
172
191
.
Guneralp
I.
,
Abad
J. D.
,
Zolezzi
G.
&
Hooke
J.
2012
Advances and challenges in meandering channels research
.
Geomorphology
163–164
,
1
9
.
Hall
M. A.
1999
Correlation-based Feature Selection for Machine Learning
.
Doctoral Dissertation
,
The University of Waikato
.
Hazarika
N.
,
Das
A. K.
&
Borah
S. B.
2015
Assessing land-use changes driven by river dynamics in chronically flood affected Upper Brahmaputra plains, India, using RS-GIS techniques
.
The Egyptian Journal of Remote Sensing and Space Science
.
https://doi.org/10.1016/j.ejrs.2015.02.001
.
Hegazy
I. R.
&
Kaloop
M. R.
2015
Monitoring urban growth and land-use change detection with GIS and remote sensing techniques in Daqahlia governorate, Egypt
.
International Journal of Sustainable Built Environment
4
(
1
),
117
124
.
Hickin
E. J.
&
Nanson
G. C.
1984
Lateral migration rates of river bends
.
Journal of Hydraulic Engineering
110
(
11
),
1557
1567
.
Hooke
J. M.
2013
River Meandering
. In:
Treatise on Geomorphology
(
Wohl
E.
&
Schroder
J.
, eds),
Academic Press
,
San Diego, CA
, Vol.
9
, pp.
260
288
.
Ijafiya
D. J.
&
Yonnana
E.
2018
Development of Chute cutoff in the lower course of River Mayo-Inne, Yola South, Nigeria
.
Asian Journal of Geographical Research
1
(
1
),
1
9
.
Johnson
R. H.
&
Paynter
J.
1967
Development of a cutoff on the River Irk at Chadderton, Lancashire
.
Geography
52(1), 41–49.
Kondolf
G. M.
,
Anderson
S.
,
Lave
R.
,
Pagano
L.
,
Merenlender
A.
&
Bernhardt
E. S.
2007
Two decades of river restoration in California: what can we learn?
Restoration Ecology
15
(
3
),
516
523
.
Kotoky
P.
,
Bezbaruah
D.
,
Baruah
J.
&
Sarma
J. N.
2005
Nature of bank erosion along the Brahmaputra river channel, Assam, India
.
Current Science
88(4),
634
640
.
Lagasse
P. F.
,
Zevenbergen
L. W.
,
Spitz
W. J.
,
Thorne
C. R.
,
Associates
A.
&
Collins
F.
2004
Methodology for Predicting Channel Migration, Rivers. NCHRP Web-Only Document 67 (Project 24-16)
.
Lancaster
S. T.
&
Bras
R. L.
2002
A simple model of river meandering and its comparison to natural channels
.
Hydrological Processes
16
,
1
26
.
doi:10.1002/hyp.273
.
Laskar
A. A.
&
Phukon
P.
2012
Erosional vulnerability and spatio-temporal variability of the Barak River, NE India
.
Current Science Association
103
(
1
),
80
86
.
Leopold
L. B.
1994
River morphology as an analog to Darwin's theory of natural selection
.
Proceedings of the American Philosophical Society
138
(
1
),
31
47
.
Leopold, L. B. 1970 An improved method for size distribution of stream bed gravel. Water Resources Research 6(5), 1357–1366.
Ma
L.
,
Li
M.
,
Ma
X.
,
Cheng
L.
,
Du
P.
&
Liu
Y.
2017
A review of supervised object-based land-cover image classification
.
ISPRS Journal of Photogrammetry and Remote Sensing
130
,
277
293
.
Mahapatra
M.
,
Ramakrishnan
R.
&
Rajawat
A. S.
2014
Monitoring long-term morphological changes of Narmada estuary using remote sensing and GIS techniques
.
Journal of Geomatics
8(1), 125–129.
Manjusree
P.
,
Satyanarayana
P.
,
Bhatt
C. M.
,
Sharma
S. V. S. P.
&
Srinivasa
R. G.
2013
Remote Sensing and GIS for River Morphology Studies
.
NNRMS(B)-18, ISRO, Hyderabad
.
Maxwell
A. E.
,
Warner
T. A.
&
Fang
F.
2018
Implementation of machine-learning classification in remote sensing: an applied review
.
International Journal of Remote Sensing
39
(
9
),
2784
2817
.
McCann
C. J.
2013
Urbanization and its Effects on Channel Morphology
.
Doctoral Dissertation
.
Micheli
E. R.
&
Larsen
E. W.
2011
River channel cutoff dynamics, Sacramento River, California, USA
.
River Research and Applications
27
(
3
),
328
344
.
Mohammed-Ali
W.
,
Mendoza
C.
&
Holmes
R. R.
2021
Riverbank stability assessment during hydro-peak flow events: the lower Osage River case (Missouri, USA)
.
International Journal of River Basin Management
19(3), 335–343.
ISSN: 1571-5124 1814-2060
.
Mourya, M., Kumari, B., Tayyab, M., Paarcha, A. & Rahman, A. 2021 Indices based assessment of built-up density and urban expansion of fast growing Surat city using multi-temporal Landsat data sets. GeoJournal 86(4), 1607–1623.
Nabegu
A. B.
2014
Impact of urbanization on channel morphology: some comments
.
IOSR Journal of Environmental Science, Toxicology and Food Technology (IOSR-JESTFT)
8
,
40
45
.
Nath
A.
&
Ghosh
S.
2022
Meandering rivers’ morphological changes analysis and prediction–a case study of Barak river, Assam
.
H2Open Journal
5(2), 289–306.
Nelson
P. A.
,
Smith
J. A.
&
Miller
A. J.
2006
Evolution of channel morphology and hydrologic response in an urbanizing drainage basin
.
Earth Surface Processes and Landforms: The Journal of the British Geomorphological Research Group
31
(
9
),
1063
1079
.
Neupane
B.
,
Singh
B. K.
,
Poudel
P.
,
Panthi
S.
&
Khatri
N. D.
2020
Habitat occupancy and threat assessment of gharial (Gavialis gangeticus) in the Rapti River, Nepal
.
Global Ecology and Conservation
24
,
e01270
.
O’Driscoll, M., Soban, J. & Lecce, S. 2009 Stream channel enlargement response to urban land cover in small Coastal Plain watersheds, North Carolina. Phys. Geogr. 30, 528–555.
O'Driscoll
M.
,
Clinton
S.
,
Jefferson
A.
,
Manda
A.
&
McMillan
S.
2010
Urbanization effects on watershed hydrology and in-stream processes in the southern United States
.
Water
2
(
3
),
605
648
.
Ozturk
D.
&
Sesli
F. A.
2015
Determination of temporal changes in the sinuosity and braiding characteristics of the Kizilirmak River, Turkey
.
Polish Journal of Environmental Studies
24
(
5
),
2095
2112
.
Papini
R.
,
Valboa
G.
,
Favilli
F.
&
L'Abate
G.
2011
Influence of land-use on organic carbon pool and chemical prop­erties of verticCambisols in central and southern Italy
.
Agriculture, Ecosystems & Environment
140
,
68
79
.
Prokop
P.
,
Wiejaczka
Ł.
,
Sarkar
S.
,
Bryndal
T.
,
Bucała-Hrabia
A.
,
Kroczak
R.
,
Soja
R.
&
Płaczkowska
E.
2020
Morphological and sedimentological responses of small stream channels to extreme rainfall and land use in the Darjeeling Himalayas
.
Catena
188
,
104444
.
Rahman
M. S.
,
Mohiuddin
H.
,
Al Kafy
A.
,
Sheel
P. K.
&
Di
L.
2019
Classification of cities in Bangladesh based on remote sensing derived spatial characteristics
.
Journal of Urban Management
8
(
2
),
206
224
.
Raj
C.
&
Singh
V.
2020
Assessment of planform changes of the Ganga River from Bhagalpur to Farakka during 1973–2019 using satellite imagery
.
ISH Journal of Hydraulic Engineering
.
doi:10.1080/09715010.2020.1812123
.
Rawat
J. S.
,
Biswas
V.
&
Kumar
M.
2013
Changes in land-use/cover using geospatial techniques: a case study of Ramnagartown area, district Nainital, Uttarakhand, India
.
The Egyptian Journal of Remote Sensing and Space Sciences
16
,
111
117
.
Resmi
S. R.
,
Patel
P. L.
&
Timbadiya
P. V.
2018
Morphological study of Upper Tapi river using remote sensing and GIS techniques
.
ISH Journal of Hydraulic Engineering
.
doi:10.1080/09715010.2017.1409089
.
Rinaldi
M.
,
Wyżga
B.
&
Surian
N.
2005
Sediment mining in alluvial channels: physical effects and management perspectives
.
River Research and Applications
21
(
7
),
805
828
.
SAC and Brahmaputra Board
1996
Report on Bank Erosion on Majuli Island, Assam: A Study Based on Multi Temporal Satellite Data
.
Space Application Centre, Ahmedabad and Brahmaputra Board
,
Guwahati
.
Sapkale
J. B.
,
Kadam
Y. U.
,
Jadhav
I. A.
&
Kamble
S. S.
2016
River in planform and variation in sinuosity index: a study of Dhamni river, Kolhapur (Maharashtra), India
.
International Journal of Scientific and Engineering Research
7
,
863
867
.
Sarkar
A.
,
Garg
R. D.
&
Sharma
N.
2012
RS-GIS based assessment of river dynamics of Brahmaputra River in India
.
Journal of Water Resource and Protection
4
,
63
72
.
doi:10.4236/jwarp.2012.42008. Published Online February 2012. Available from: http://www.SciRP.org/journal/jwarp
Schumm
S. A.
1963
A tentative classification of alluvial river channels
.
Geological Survey Circular
477
.
Sharma
N.
,
Amoako
J. F.
,
Hutton
C. W.
&
Clark
M.
2010
Hazard vulnerability and risk on the Brahmaputra Basin: a case study of river bank erosion
.
The Open Hydrology Journal
4
,
211
226
.
Shi
Z.
,
Hao
H.
,
Zhao
M.
,
Feng
Y.
,
He
L.
,
Wang
Y.
&
Suzuki
K.
2019
A deep CNN based transfer learning method for false positive reduction
.
Multimedia Tools and Applications
78
(
1
),
1017
1033
.
Sial
A. K.
,
Shankar
T.
,
Maitra
S.
,
Sairam
M.
,
Sagar
L.
,
Kumar
P. P.
&
Rout
R. K.
2022
Effect of nutrient optimization on growth, yield and quality of sunflower under southern Odisha conditions
.
Crop Research
57
(
3
),
195
201
.
Singh
M. K.
&
Ghosh
S.
2022
Groundwater flow modeling for Cachar, India using MODFLOW: a case study
.
ISH Journal of Hydraulic Engineering
.
doi:10.1080/09715010.2020.1868357
.
Taati
A.
,
Sarmadian
F.
,
MousavI
A.
,
Pour
C. T. H.
&
Shahir
A. H. E.
2015
Land use classification using support vector machine and maximum likelihood algorithms by Landsat 5 TM images
.
Walailak Journal of Science and Technology (WJST)
12
(
8
),
681
687
.
Tal
M.
&
Paola
C.
2010
Effects of vegetation on channel morphodynamics: results and insights from laboratory experiments
.
Earth Surface Processes and Landforms
35
(
9
), 1014–1028.
Talukdar
S.
,
Singha
P.
,
Mahato
S.
,
Pal
S.
,
Liou
Y. A.
&
Rahman
A.
2020
Land-use land-cover classification by machine learning classifiers for satellite observations – a review
.
Remote Sensing
12
(
7
),
1135
.
doi:10.3390/rs12071135
.
Teluguntla
P.
,
Thenkabail
P. S.
,
Oliphant
A.
,
Xiong
J.
,
Gumma
M. K.
,
Congalton
R. G.
,
Yadav
K.
&
Huete
A.
2018
A 30-m landsat-derived cropland extent product of Australia and China using random forest machine learning algorithm on Google Earth Engine cloud computing platform
.
ISPRS Journal of Photogrammetry and Remote Sensing
144
,
325
340
.
Vapnik
V.
&
Izmailov
R.
2017
Knowledge transfer in SVM and neural networks
.
Annals of Mathematics and Artificial Intelligence
81
(
1
),
3
19
.
Ward
J. V.
&
Stanford
J. A.
1995
Ecological connectivity in alluvial river ecosystems and its disruption by flow regulation
.
Regulated Rivers: Research & Management
11
(
1
),
105
119
.
doi:10.1002/rrr.3450110109
.
Woldemichael
A. T.
,
Hossain
F.
,
Pielke
R.
Sr.
&
Beltran-Przekurat
A.
2012
Understanding the impact of dam-triggered land use/land cover change on the modification of extreme precipitation
.
Water Resources Research
48
(
9
).
https://doi.org/10.1029/2011WR011684
.
Wu
L.
,
Zhu
X.
,
Lawes
R.
,
Dunkerley
D.
&
Zhang
H.
2019
Comparison of machine learning algorithms for classification of LiDAR points for characterization of canola canopy structure
.
International Journal of Remote Sensing
40
,
5973
5991
.
Yousefi
S.
,
Pourghasemi
H. R.
,
Hooke
J.
,
Navartil
O.
&
Kidová
A.
2016
Changes in morphometric meander parameters identified on the Karoon River, Iran, using remote sensing data
.
Geomorphology
.
doi:10.1016/j.geomorph.2016.07.034
.
Zhang
S.
,
Yao
L.
,
Sun
A.
&
Tay
Y.
2019
Deep learning based recommender system: a survey and new perspectives
.
ACM Computing Surveys (CSUR)
52
(
1
),
1
38
.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/).