1 Introduction

Alluvial fans possess useful archival interest since they record interactions between tectonic deformations, climatic perturbations and changes in eustatic cycles in the Late Quaternary period (Blair and McPherson 1994; Chakraborty and Ghosh 2010; Pickering et al. 2018; Ventra and Clarke 2018). Such depositional landforms develop along the margins of a highland and a subsiding continental basin over a broad spectrum of climatic settings, ranging from hyper-arid to humid temperate to monsoonal tropical (Bull 1977; Blair and McPherson 1994; Ventra and Clarke 2018). The stacked intercalations of stream flow and sediment gravity flow deposits in alluvial fans attest to the dynamicity of the mountain front domain for tectonic reworking and climate change. The southern foot-hill piedmont surface of the Himalayan mountain belt (figure 1) represents one of the most seismically and neotectonically active areas in the Indian subcontinent (Valdiya 1986; Nakata 1989; Starkel et al. 2015; Mugnier et al. 2022). The near-continuous piedmont surface in this belt developed due to the coalescence of alluvial fans (Valdiya 1986; Nakata 1989). The uninterrupted siliciclastic sedimentary succession of these alluvial fans has immense importance in constraining the boundary conditions for their evolution during the Quaternary Period (Guha et al. 2007; Chakraborty et al. 2010; Chakraborty and Ghosh 2010; Goswami et al. 2013; Mandal and Sarkar 2016; Mugnier et al. 2022). Changes in channel morphometric parameters over a considerable period in the recent past and well-exposed sedimentary successions along channel walls and bars offer ample opportunities to unravel the neotectonic activities and the influence of climatic changes in the foot-hill region and adjacent mountain belt.

Figure 1
figure 1

(A) The Himalayan foot-hills. (B) The study area with locales of the earthquakes that occurred from 1930–2020. The sources of data from the National Earthquake Information Centre of USGS; 10/07/2022 https://www.usgs.gov/programs/earthquake-hazards/national-earthquake-informationcenterneic#:~:text=The%20National%20Earthquake%20Information%20Center%20(NEIC)%2C%20a%20part%20of,10%20miles%20west%20of%20Denver. (C) Studied segment of the Nagrakata piedmont surface (Nagrakata section of Nakata 1972) in the Jalpaiguri district of West Bengal, India. AB, CD, and EF showing longitudinal and transverse profile locations are shown in figure 5.

Full size image

A segment of the piedmont surface (26°41′03″–27°00′21″N; 88°51′07″–89°03′21″E) in the Jalpaiguri district of West Bengal, India (figure 1), with minimum anthropogenic activities, has been selected for the present study. Recent earthquakes (figure 1B) and heavy monsoonal downpours with marked variations (figure 2) indicate neotectonic activities and climatic fluctuations in the selected study area. The Jaldhaka and Daina rivers occupy reincised valleys on this piedmont surface. This paper aims to explore the short-term intricacies in the variability of channel morphometric characteristics in response to neotectonic activities and climatic changes in this area over the last ninety years, spanning 1930 and 2020.

Figure 2
figure 2

(A) Month-wise rainfall data of the studied region spanning 1930–2020. (B) The month-wise plot of rainfall data of the studied region spanning 1930–2020. Monthly precipitation gridded time series data collection with a timeframe of 1901–2021 and a precision of 0.5° × 0.5° resolution is acquired from CRU TS v4.06 (Climatic Research Unit, University of East Anglia in Norwich, UK; Harris et al. 2020).

Full size image

2 Geotectonic setting of the study area

The study area (figure 1) represents a low dipping (~5°) piedmont surface that hosts the river valleys of the Jaldhaka in the west and the Daina in the east with their tributaries. The slope is variously incised by other smaller tributaries of the Jaldhaka and Daina rivers (figure 1). Lineament mapping using Surfer 13 and Arc GIS 10.3.1, followed by intensive ground checks (183 ground control points), reveals the development of north–south and east–west trending faults (figure 3). The two most prominent east–west running ridges in this area (figure 3) are the eastward extensions of the regional thrusts, namely Matiali and Chalsa thrusts (Nakata 1989). Another east–west trending thrust, Bharadighi Thrust (Nakata 1989), also developed further south of the Chalsa thrust (figure 3); however, its surficial expression is subdued. These blind thrusts are buried under fault propagation folds (Kar et al. 2014). These thrusts postdate the Himalayan Frontal Thrust (HFT) of Ganssar (1964) since they have affected the Quaternary fan deposits of the area (Mugnier et al. 2022). However, in the study area, the HFT is not well exposed (Guha et al. 2007). Another east–west trending elevated ridge (figure 3), with a northerly dipping scarp face, represents another contemporary thrust (eastward continuity of the Matiali thrust?) and is known as the Thaljhora fault (Goswami et al. 2013). Presently the Thaljhora channel flows along this fault. These thrusts are the youngest manifestation of the continuing convergence between Indian and Eurasian plates. Furthermore, there is an opinion that these thrusts represent younger splays of the HFT (Valdiya 1986; Hodges 2000; Kar et al. 2014; Mandal and Sarkar 2016). Transverse faults in the foot-hill region are common (Som et al. 2012; Goswami et al. 2013; Patra and Saha 2019). The River Jaldhaka follows a nearly north–south trending fault, sub-parallel to the Teesta fault (figure 3). A 30–40 m steep fault-scarp along the eastern bank of the Jaldhaka River signifies the development of a north–south trending fault (figure 4). The plain lying west of this fault, now occupied by the Chapramari Forest, represents the down-thrown block. The lateral shifts along the course of the Jaldhaka River mark the off-sets on the Jaldhaka-fault across the younger thrusts (figure 3). The lateral shifts of the Jaldhaka fault also indicate that the younger east–west running thrusts and faults have a strong strike-slip component (also see Mukul and Matin 2005). All the tributaries of the Jaldhaka river flow in a northeast–southwest direction, and most of them coincide with the minor faults in this area (figure 3).

Figure 3
figure 3

Neo-tectonic lineaments in the Himalayan foothill piedmont region of North Bengal show locations of the longitudinal Matiali, Chalsa, and Bharadighi thrusts and the transverse Teesta and Jaldhaka faults.

Full size image
Figure 4
figure 4

Field photograph of the fault scarps along the eastern bank of the Jaldhaka River.

Full size image

The longitudinal section of the fan surface is relatively rugged in the northern upper part and becomes near to flat in the southern lower part (figure 5A). The transverse section across the proximal part of the fan surface reveals similar ruggedness (figure 5B), which becomes flat near the distal part (figure 5C). The rugged topography in the upper part of the fan surface results from the intersection of longitudinal and transverse thrusts and faults in the area.

Figure 5
figure 5

(A) AB longitudinal section showing the rugged surface in the proximal part and low dipping smooth surface in the distal part of the studied fan. (B) CD transverse section showing a sculptured surface of the middle part of the fan region. (C) EF transverse section showing near the flat surface of the outer fan region. The data for the profile sections (smoothed for better expression) are derived from the SRTM DEM.

Full size image

The deep fault scarps and terrace walls along channels exposed good sections of the Quaternary deposits (figure 6). A close examination of the different sections of scarps and terrace walls reveals the presence of different sedimentary facies types, namely matrix-supported gravel facies, clast-supported gravel facies, normally graded gravelly sand facies alternating gravelly facies with imbricated clasts, and trough cross-stratified gravelly sand facies (figure 7). These different sediment types represent cohesive debris flow, grain flow, turbulent high-density sediment gravity flow, and fluviatile gravelly sand deposition (Chakraborty and Ghosh 2010; Mandal and Sarkar 2016; Roy 2016). The architecture of these alternating gravelly and sandy deposits represents a prograded alluvial fan deposit accumulated in a foredeep trough created after the emplacement of the post-Siwalik thrust (Main Frontal Thrust, Gansser 1981). A gradual downslope change in grain size from the fan apex to the southern end of the fan attests to a single and continuous phase of sediment accumulation with seasonal variation to form a prograded and multi-lobed alluvial fan deposit (c.f. Chakraborty and Ghosh 2010; Kar et al. 2014; Mandal and Sarkar 2016). Bouldery gravel bars dominate the gravelly bars and channel floor deposits of the reincised channels in the upper reaches (figure 8A). Gravelly bar and sandy inter-bar channel deposits cover the lower reaches and constitute the lower (southward) fluvial plain of the study area (figure 8B).

Figure 6
figure 6

Field photograph showing stacked Quaternary deposits along the fault scarps.

Full size image
Figure 7
figure 7

A composite stratigraphic column of the alluvial fan shows the sedimentary facies' architecture.

Full size image
Figure 8
figure 8

(A) Field photograph of the bouldery gravels on the channel floor of the upper part of the alluvial fan. (B) Field photograph of the gravelly bar and sandy inter-bar channel deposits in the outer fan region.

Full size image

Signatures of glacial reworking of the fan deposits during the last glacial maxima are not yet reported, and at the same time, the effects of eustatic changes on the fan evolution are yet to be worked out (see Chakraborty et al. 2010; Chakraborty and Ghosh 2010; Mandal and Sarkar 2016; Pickering et al. 2018).

3 Methodology

Field visits and RS-GIS applications are adopted to study the channel characteristics in the study area. To carry out a time series analysis of the channel evolution pattern for the period between 1930 and 2020, available topographical maps and satellite images are used (table 1). The main synthesis of the present study depends mainly on the database generated on Jaldhaka and Daina rivers. However, their tributaries, Jiti, Thaljhora, Ghatia, and Kuji Daina, are also considered. The Survey of India topographical map published in 1930 was utilized in the absence of satellite images. All topographical maps and cloud-free satellite images were geometrically rectified in the same coordinate system (Universal Transverse Mercator: Zone North 45; Datum WGS1984). The SRTM Digital Elevation Model (DEM) has been used to analyze topographic and geomorphic indices. Riverbank lines have been digitized for every year superimposed on each other, and changes were measured using the spatial analysis tool of ArcGIS (10.3.1). Kumai Forest to Ranirhat More stretch of the Jaldhaka River and the Daina River from Chengmari Tea Garden to the confluence with the Jaldhaka stretch (figure 9A) are detailed in this study. Contours with 30-m intervals for the study area are generated using DEM (figure 9A). DEM was utilized to avoid the use of restricted Survey of India topographic maps for the study area which is close to the international border. Usually, DEM offers high accuracy in studying the Himalayan foot-hill region (Mukul et al. 2017). Moreover, 183 GCPs were utilized to analyze the errors in preparing the DEM. In this study, the Jaldhaka and Diana rivers are segmented into seven and six representative reaches, respectively, based on the 30-m elevation change (figure 9A). The changing positions of the confluence point of the Jaldhaka and Daina rivers across time are also considered to address the dynamicity of the study area.

Table 1 Details of the topographical maps, satellite images and DEM are used in the study.
Full size table
Figure 9
figure 9

(A) Contour map constructed from DEM shows the studied reaches of the Jaldhaka and Daina rivers. (B) and (C) showing convex SL Index profiles for the studied stretches of the rivers Jaldhaka and Daina. The data for the profile sections (smoothed for better expression) are derived from the SRTM DEM.

Full size image

4 Present study

Stream length gradient index (SL), sinuosity index (SI), braiding index (BI), channel (Thalweg) migration, and the shifting confluence point of the Jaldhaka and Daina rivers are estimated for a period last 90 years to evaluate the change in channel characteristics under the influence of neotectonic movements and climatic variations.

4.1 Stream length gradient index (SL)

The stream length gradient index (SL) for a reach in a channel stretch reflects the role of tectonic and lithological changes in the surrounding region, if there are any (Hack 1973; Lifton and Chase 1992; Keller and Pinter 1996, 2002; Burbank and Anderson 2001). The SL index values rise when streams flow across a tectonically active area or a zone with higher lithological resistance (Keller and Pinter 2002; Mood et al. 2016). Applications of SL index analysis for channels in unconsolidated sediments have also yielded good results (Goswami 2017; Ayaz et al. 2018). The SL index is measured for a particular reach of a defined stretch of a river using the formula (after Hack 1973): SL = (ΔH/ΔL) × L, where ΔH represents the difference of elevation between two endpoints of the reach, ΔL is the length of the reach, and L is the stream length of the channel from the uppermost point of the stretch to the midpoint of a reach under consideration. The computed SL values of the reaches of the studied stretches of the Jaldhaka and Diana rivers depict interesting results (table 2 and figure 9). These high values coincide with the zones where the Jaldhaka river valley crosses the Thaljhora (Matiali?) and Chalsa thrusts, and the Daina River crosses the Chalsa thrust (figure 9B and C). Thus, the convex profile of the SL values across the Thaljhora (Matiali?) and Chalsa thrusts indicate the role of neotectonism in the region during the period under consideration.

Table 2 Values of stream length gradient index of the different reaches of the Jaldhaka and Daina rivers.
Full size table

4.2 Channel morphology

The sinuosity index (SI) and braiding index (BI) of the river Jaldhaka and Daina have been calculated following Leopold and Wolman (1957) and Brice (1964) to address the channel morphology.

4.2.1 Sinuosity index (SI)

The sinuosity index (SI) of a channel, according to Leopold and Wolman (1957), is SI = OL/EL, where OL represents the channel thalweg length, and EL represents the valley length. Morisawa (1985) used the SI value to classify the channel type into five types, namely, straight channel (SI < 1.05), sinuous channel (SI > 1.05), meandering channel (SI > 1.5), braided channel (SI > 1.3), anastomosing channel (SI > 2.0). A braided pattern represents multiple channels where the many divided channel ways always shift (Schumm 1963). However, all these channel patterns manifest the variation in energy conditions and sediment load in a continuous fluvial channel system with local constraints (Schumm 1963; Petts and Foster 1985).

SI is better reflected in smaller segments of a fluvial channel, for example, in one meander length (Ebisemiju 1994). This study measured SI for different reach lengths along the Jaldhaka and Daina rivers for the last 90 years, spanning 1930 and 2020 (table 3). The study reveals interesting results. The Jaldhaka River remained straight to slightly sinuous throughout the study period in the upper two reaches (J1 and J2). However, in reach J3, the SI values were initially high and became low with time, indicating that the high sinuous channel pattern became straight. The SI values for the lower two reaches (J6 and J7) were always high and remained sinuous throughout the study period (figure 10A).

Table 3 Reach-wise sinuosity index values of the Jaldhaka and Daina rivers for the period from 1930 to 2020.
Full size table
Figure 10
figure 10

(A) Plots showing the reach-wise variation of sinuosity index values of the rivers Jaldhaka and Daina for the period from 1930 to 2020 and plots of average values, (B) Plots showing the reach-wise variation of braiding index values of the rivers Jaldhaka and Daina for the period from 1930 to 2020 and plots of average values.

Full size image

Similarly, most of the upper reaches of the Diana River (D1, D2, D3, and D4) show low SI values; however, the lower two reaches (D5 and D6) show comparatively higher values. Moreover, there are few high SI values around 2011 for D2 and D3. Like the Jaldhaka River, the SI values of all the reaches of the Daina river show a gradual fall (figure 10A).

An important point from the present study reveals that the Jaldhaka and Daina rivers are becoming straight within the last 90 years. With time, a drop in SI values suggests the gradual imposition of the control of the regional N–S trending Jaldhaka fault and NE–SW trending Daina fault on the channel morphology. The high SI values in J3, J6, J7, D5, and D6 stand for the higher rate of sediment deposition in the lower reaches in response to prolong monsoonal downpours (figure 2). Intermittent rise of SI values also points to higher precipitation rates related to the climatic variations across years.

4.2.2 Braiding index (BI)

Braiding index (BI) for the Jaldhaka and Daina rivers has been measured after Brice (1964) to evaluate the degree of braiding; Braiding Index = 2(sum of the length of islands or bars)/length of the reach. The computed BI values for the two studied rivers show high variability (table 4). The BI values for the Jaldhaka River range between 0.76 and 4.88; for the Daina River, it varies from 0.50 to 4.56. The average BI values of the two studied rivers show increasing trends during the latter part of the studied period; however, a significant rise in BI index took place after the year 2011 (figure 10B). The BI values for the years 2017 and 2020 also remain high in comparison to the initial phases of the study period. Such an increase in the BI values in the lower reaches after 2011 agrees with the higher degree of sediment accumulation in the lower part of the fan with a reduced valley slope (Chakraborty et al. 2010). The convex profile of the SL index of the studied reaches also supports this observation.

Table 4 Reach-wise braiding index values of the Jaldhaka and Daina rivers for the period from 1930 to 2020.
Full size table

There is a common agreement that BI values frequently change with time due to changes in factors like clogging of channels, the volume of the sediment load carried by a river, the average size of the sediment, competency of the river, and the presence of erodible banks (Church 1972; Morisawa 1985). All these factors directly relate to the rate of precipitation and availability of the sediments in the source region and along the banks. The high degree of variability of the BI values for the studied rivers approves fluctuations in the precipitation rate in the upper reaches, availability of the easily erodible sediments of variable sizes from boulder to mud from the unconsolidated Quaternary alluvial fan deposits, and frequent changes in stream capacity and competency. An increase in channel multiplicity with the development of longitudinal bars within channels giving rise to high BI values and widening the river beds are the common features in the lower reaches of the Jaldhaka and Daina rivers. Such a higher rate of sediment shedding in the lower reaches of the studied rivers, particularly from 2011 onward (figure 10B), might have a causal relationship with the ‘2011 Sikkim Earthquake’ with a magnitude of Mw 6.9 (figure 1B) along with higher monsoonal downpour (figure 2). The earthquake might have triggered landslides in the Himalayan region and enhanced the availability of loose sediment (see also Martha et al. 2015).

4.3 Channel migration

The channel migration rates for the Jaldhaka and Daina rivers have been measured for 1930–1977, 1977–1997, 1997–2011, 2011–2017, and 2017–2020. The computed lateral river shifting with migration rate and direction of shift for both the rivers within the studied stretches are shown in table 5. The digitized thalweg lines of the main channels of the two rivers for the years 1930, 1977, 1997, 2011, 2017, and 2020 are superimposed one above the other, to estimate the avulsion pattern over the 90 years (figure 11A). The channel migration is limited in the upper reaches of the Jaldhaka River, north of Chalsa thrust (figure 11B). However, the lower reaches show a higher degree of channel migration, increasing in the low southern gradient fluvial plane (figure 11B). Irrespective of the position of the reaches, the Daina River shows a higher degree of channel migration in the studied stretch (figure 11C). Moreover, the lower reaches of both the rivers show maximum lateral channel migration values (table 5).

Table 5 Reach-wise variation of channel migration across the transects of the rivers Jaldhaka and Daina for the period from 1930 to 2020.
Full size table
Figure 11
figure 11

(A) Reach-wise variation of channel migration along the transects of the rivers Jaldhaka and Daina. (B) Reach-wise plot of migration rate of the river Jaldhaka. (C) Reach-wise plot of migration rate of river Daina.

Full size image

Restricted lateral channel migration in the upper reaches of the Jaldhaka river points to a strong fault control of the Jaldhaka fault (tectonic control). Dumping the sediment load (gravelly sand) in the lower reaches leads to channel migration. A higher rate of sediment shedding also suggests high monsoonal discharge with flood-like situations and bank erosion. The formation of bars under this situation might have influenced this channel migration in the lower reaches of the Jaldhaka River. Because of the low gradient, the Daina River registered a higher migration during the studied period under the same climatic condition of higher monsoonal discharge, bank erosion, and a high sedimentation rate.

4.4 Jaldhaka–Daina River confluence dynamics

The confluence points of the Jaldhaka and Daina rivers remain extremely dynamic during the study period (also see Chakraborty and Mukhopadhyay 2014). The confluence point (CP 1) was situated ∼5.25 km downstream from the current location in 1930 and moved 2.88 km upstream (CP 2) from 1930 to 1977 (figure 12). A huge upstream migration of the confluence point (CP 3) for 4.10 km took place in the next 20 years, from 1977 to 1997 (figure 12). From 1997 to 2011, the confluence point moved 1.88 km downstream (CP 4) from its earlier position (figure 12). Further, the confluence point shifted 0.37 km upstream (CP 5) from 2011 to 2017 (figure 12). A downstream movement of 0.36 km of the confluence point (CP 6) took place between 2017 and 2020 (figure 12). This study on the rate of migration of the confluence point of the rivers Jaldhaka and Daina depicts an initial faster upstream migration, followed by a slower rate of alternating downstream and upstream migrations.

Figure 12
figure 12

Showing the dynamic behaviour of the confluence point of the rivers Jaldhaka and Daina from 1930 to 2020.

Full size image

The initial upstream migration between 1930 and 1997 for a distance of 6.98 km (figure 12) attests to valley-floor aggradation in response to a prolonged period of monsoonal discharge with a higher rate of sedimentation and channel aggradation. After 1997, there were fluctuations in monsoonal discharge with variable sediment load for deposition, as evidenced by the alternating downstream and upstream movements of the confluence point. Furthermore, the present study also reveals another alternative, the confluence point shifted upstream when the channel shifted towards each other. However, the channel avulsion has a definite linkage with monsoonal discharge and a higher sedimentation rate (figure 12; also, see Ghosh and Mukhopadhyay 2021). Therefore, the dynamicity of the confluence point of the Jaldhaka and Daina rivers over 90 years reveals the aggradation of the river valleys.

5 Discussion

Channel evolution in foredeep alluvial fans responds rapidly to changes in climatic conditions and tectonic deformations (Burbank 1992; Dalrymple et al. 1994; Blum and Tornqvist 2000). However, eustatic rise and fall impose an overall control on the deepening of valley floors and sediment dispersal patterns (Pickering et al. 2019). In the frame of the foreland basin model, the studied fan system might have time equivalent correlatives in the lower deltaic part (Sarkar et al. 2009; Pickering et al. 2018, 2019) and also in the deep basinal fans in the Bay of Bengal. Since the present study was conducted for 90 years of the immediate past (1930–2020), the role of eustatic rise or fall will be minimum in the morphometric changes of the channels flowing through the studied piedmont. As a result, controlling neotectonics and climatic changes in channel evolution during the studied period will gain much importance, particularly where anthropogenic activity is at a minimum level. There is a conflict regarding the control of neotectonics (Gupta 1997; Goswami et al. 2011; Mandal and Sarkar 2016) or heavy monsoonal discharge (Burbank 1992; Kar et al. 2014) or both factors (Nakata 1972; Valdiya et al. 1992; Sinha et al. 2005; Singh et al. 2016) on the evolution of the channel patterns in the piedmont of the Himalayan foot-hill. Development of the channels along regional and local thrust and faults in the fans suggests undoubted control of the Himalayan neotectonic movements on the evolution of the channel system (Gansser 1981; Goswami et al. 2011, 2013; Bhattacharya et al. 2022). The present study area is no exception to it. Development of the unpaired terraces (at least three) along the rivers Jaldhaka and Daina also supports episodic exhumation of the piedmont containing the Quaternary siliciclastic succession (see Guha et al. 2007; Chakrabarti et al. 2019; Mugnier et al. 2022). Such deformational movements might have imposed controls on channel characteristics like a convex profile of the SL values across the Matiali and Chalsa thrusts (figure 9). However, the lower reaches of the Jaldhaka and Daina rivers recorded a comparatively higher sedimentation rate, particularly from 2011 onward, as evidenced by the general increase in Braiding index (BI), a higher rate of the channel migration, and dynamicity of the confluence point. This increased rate of sediment shedding in the lower reaches with a low gradient might have occurred under the influence of higher monsoonal discharge (figure 2) and neotectonism during the study period (2011 Sikkim earthquake and other earthquakes shown in figure 1B).

It is important to note that different reaches of the studied stretches of the two rivers over the last 90 years (1930–2020) show a high degree of variation in SL, SI, BI, and channel migration rate. These results point to distinctly different responses of the studied reaches of the rivers to the tectonic and climatic causes. The present study demonstrates that neotectonics guide the evolution of channel patterns in the Himalayan foothill terrain, and monsoonal discharges play an important role.

6 Conclusion

The piedmont surface of the Nagrakata area developed through the coalescence of the Quaternary alluvial fan deposits and is subjected to Himalayan neotectonic movements. Active thrusts and faults control the drainage pattern on the piedmont surface. Jaldhaka and Daina rivers and their tributaries developed along lineaments representing dislocation planes. Stream length gradient index, sinuosity index, braiding index, channel migration, and change of position of the confluence of the Jaldhaka and Daina rivers, studied in successive reaches, provide information regarding the evolution of the channel pattern over 90 years (1930–2020). The studied parameters indicate that neotectonics and climate were equally important in the evolution of the channel pattern of the two rivers. However, control of these factors was variable in different segments of the rivers. Moreover, aggradations and channel migration in the lower reaches of the rivers are strongly influenced by the rate of neotectonic activities and monsoonal discharge.