Channel Morphology – UPSC

• Channel morphology (river channel morphology) is the study of the form of a river channel and the processes that shape and change it.
  • It covers channel geometry (length, width, depth, wetted perimeter, slope, bends), channel fluid dynamics (discharge, velocity), hydraulic geometry, channel types, bed topography and channel patterns.
• A channel is shaped by two flows, of water (discharge) and of sediment (amount and calibre), both governed by the channel gradient.
• Emory W. Lane (1955) put this balance as Lane’s relationship: sediment load × bed-material size is proportional to water discharge × channel slope.
  • If one side changes (a dam traps sediment, floods raise discharge), the channel aggrades or degrades until balance returns.

Channel Form and Its Controls

Channel

• A channel is the path of a river or stream outlined by its bed and banks, confined within the valley walls.
  • It is a three-dimensional form defined by its slope, cross-section and pattern.
  • It may carry perennial, intermittent (seasonal) or ephemeral (rain-fed, short-lived) flow.
• At bankfull stage the channel is just full; above it the river spills onto the floodplain as a flood.
• Effluent channels of humid regions gain groundwater downstream; influent channels of arid regions lose water to the ground and to evaporation and may dry up before reaching the sea.
  • The Ganga is effluent; the Luni, dying out in the Rann of Kachchh, behaves as an influent channel.

Channel Structure and Geometry

• A channel has three structural parts: the bed, the two banks, and the thalweg, the line joining the deepest points along the channel.

Element	What it measures	Note
Channel width (W)	Water-surface width across the channel	Maximum at bankfull stage
Channel depth (d)	Water surface to bed	Mean depth averages several gauge levels
Wetted perimeter (P)	Bed and banks in contact with water	Controls frictional resistance
Width/depth ratio	W ÷ d	High for bedload rivers, low for suspended-load rivers
Channel length	Distance along the channel mid-line	Source to mouth
Sinuosity	Channel length ÷ valley length	1.0 = perfectly straight
Crossover	Straight reach between two bends	Site of riffles
Radius of curvature (rm)	Radius of the circle through a bend	Usually 2–3 channel widths

• Cross-sectional shape decides how width responds to discharge: a rectangular channel deepens without widening, while an irregular or asymmetric one widens quickly and deepens slowly.

Thalweg Principle

• The thalweg principle is a legal rule: where a boundary between two states is said to follow a river without further description, it follows the thalweg, in practice the centre of the principal navigable channel.
  • Where there are several navigable channels, the one mainly used for downstream traffic is taken; as the thalweg migrates, such boundaries can shift.

Long Profile

• The long profile of a perennial river is usually concave upward, reflecting tectonic history, base-level change and climate, and over time it evolves towards a graded profile of equilibrium.
  • Concavity increases with relief and where discharge rises rapidly downstream; profiles approach a straight line where discharge grows slowly, as in permeable catchments.

River Processes

• Channels are continually reshaped by erosion, transportation and deposition, the same processes that build fluvial landforms.

Erosion process	Mechanism	Effect on the channel
Hydraulic action	Force of water and compressed air in cracks	Loosens bank and bed material
Corrasion (abrasion)	Load grinds against bed and banks	Potholes, deepening, widening
Corrosion (solution)	Water dissolves soluble rock	Important on limestone beds
Attrition	Particles collide and break	Rounds and fines the load

Transport mode	How the load moves	Load type
Traction	Large particles roll along the bed	Bedload
Saltation	Particles bounce along the bed	Bedload
Suspension	Fine particles carried within the flow	Suspended load
Solution	Dissolved matter	Dissolved load
Flotation	Material carried on the surface	Floating load

• Competence is the largest particle a stream can move; capacity is the total load it can carry.
• Deposition begins when the river lacks energy to carry its load, the coarsest particles settling first.

Factors Controlling Channel Morphology

• Channel form results from the interaction of four sets of variables: fluid dynamics (velocity, discharge, roughness, shear stress), channel configuration (width, depth, slope, shape, pattern), sediment load and bed and bank materials.
• Independent factors are imposed on the watershed.
  • Geology sets the bedrock, surficial materials and topography (David R. Montgomery, 1999).
  • Climate controls rainfall, runoff and the discharge regime.
  • Human action alters the watershed through anthropogenic land use, dams, embankments and sand mining.
  • Time is a further independent variable.
• Dependent factors respond to them: sediment supply (frequency, volume, calibre), discharge (frequency, magnitude, duration) and riparian vegetation, which controls bank erodibility and supplies large woody debris.
• Classical models treated channel form mainly as a function of discharge and sediment transport, with transport equal to supply at equilibrium (Stanley A. Schumm, 1971); they underplayed vegetation and boundary conditions.

Boundary condition	Effect on the channel
Valley slope	Sets the maximum channel gradient
Bank composition	Sets bank erodibility
Bedrock, compact till, coarse colluvium	Limit lateral and vertical migration; fix alignment
Stored valley-floor sediment	Floodplains, fans and terraces supply erodible material
Human structures	Bridges, embankments and flood works confine the channel

• The present channel is thus a product of both present-day and historical processes.

Adjustment Mechanisms

• The variables adjust through erosion, deposition and the movement of bedforms.
  • Bed scour lowers the gradient, deepens the channel and slows the flow; bank scour widens it and reduces depth and velocity.
  • Deposition raises the bed, making the channel shallower but locally steeper, which speeds flow downstream.
• Himalayan rivers, with heavy loads and a monsoon regime, adjust rapidly; Peninsular rivers on old resistant rocks with smaller loads are far more stable.

Hydraulic Geometry of Channels

• Hydraulic geometry analyses the relationships among discharge, velocity, channel shape, sediment load and slope.
• It rests on records from gauging stations of water level, channel shape, velocity and sediment.
• Luna B. Leopold and Thomas Maddock Jr. (1953) analysed thousands of such records and named the approach.

The Power Functions

• Discharge is the product of width, mean depth and mean velocity: Q = w × d × v.
• Each varies as a power function of discharge: w = aQ^b, d = cQ^f, v = kQ^m; suspended load rises as L = pQ^j.
  • Since Q = wdv, a × c × k = 1 and b + f + m = 1.
  • The constants only convert units; the exponents show how width, depth and velocity share a change in discharge.
  • On logarithmic paper each relation plots as a straight line.

• The relations are studied at a station (as discharge changes at one cross-section) and downstream (at mean discharge between successive stations).

Exponent	Variable	At a station	Downstream
b	Width	0.26	0.5
f	Depth	0.40	0.4
m	Velocity	0.34	0.1

At-a-Station Hydraulic Geometry

• The values are averages for 20 gauging stations in the central and south-western United States.
  • As discharge rises after heavy rain, width, depth and velocity all increase: width roughly as the fourth root, velocity roughly as the cube root, depth fastest.
• Local conditions distort these averages.
  • Riffles and bars add roughness and slow the rise of velocity.
  • Cohesive beds and banks give a lower b, so b tends to be lower in humid than in arid and semi-arid areas.
• Pattern leaves a signature: meandering and braided reaches have a higher b and lower m than straight reaches (A. David Knighton, 1975), and braided reaches have b/f above 1.
• Keith S. Richards (1973) found both f and m high over riffles, while over pools f is lower and m very high.

Downstream Hydraulic Geometry

• As mean discharge grows downstream, width, depth and mean velocity all increase; width fastest (roughly as the square root of discharge), then depth, with velocity rising only slightly.
• The velocity result was a surprise, since gradient falls downstream.
  • Increasing depth makes flow more efficient and offsets the falling slope.
  • Much of a mountain torrent’s current is lost in eddies with as much backward as forward motion.
• Charles W. Carlston (1969) found a wide range of m values: in about half of 39 streams, velocity fell or stayed constant downstream. The exponents reflect local environments, not a universal law.

Bed and Bank Materials

• Cohesive banks restrict widening and armoured beds restrict deepening; where both resist, m is larger and b and f small.
  • Non-cohesive, erodible banks give a high b; erodible banks over a cohesive bed make b greater than f.
• Erodibility depends on particle size and sorting, packing, consolidation, cementation, moisture, vegetation and, in clays, electrochemical bonding.

Sediment Load and Channel Form

• Sediment load is the second most important control after discharge.
  • Suspended load depends on turbulence and velocity; bed-material load on shear stress at the bed.
• At a station, suspended load rises faster than discharge; downstream it rises more slowly, so its concentration falls slightly.
• Suspended load moves best in narrow, deep channels; coarse bedload needs wide, shallow ones where turbulence, lift and bed shear are greater, giving bedload channels a high width/depth ratio.
• The larger and coarser the load, the steeper the gradient needed to carry it.

Evaluation and Indian Link

• Hydraulic geometry is empirical: exponents vary with channel type, materials and flow range, and they describe adjustment rather than explain it.
• Its engineering parallel is regime theory, developed on the alluvial canals of the Punjab.
  • R. G. Kennedy (1895) related non-silting, non-scouring velocity to depth on the Upper Bari Doab Canal.
  • Gerald Lacey (1930) gave the regime relation P = 4.75 √Q, the same 0.5 exponent Leopold and Maddock found for downstream width.
• It now underpins river restoration, bridge and embankment design, and the sizing of stable canals.

Channel Bed Topography

• Bed topography is the configuration of the bed in positive and negative features: riffles and pools, bars, shoals, sand islands and sand waves, formed as flow interacts with the transport of sediment.

Bedforms of Sand-Bed Channels

• Sand beds shape themselves into bedforms that change with flow strength (Daryl B. Simons and Everett V. Richardson, 1966).

Flow regime	Bedform	Flow and water surface
Lower (Froude number < 1)	Ripples, then dunes	Tranquil flow; surface out of phase with the bed
Transition	Washed-out dunes, plane bed	Resistance drops sharply
Upper (Froude number ≥ 1)	Plane bed, antidunes, chutes and pools	Rapid flow; surface in phase with the bed

Pools and Riffles

• A long profile alternates between pools, topographic lows with gentle water-surface slope and fine bed material, and riffles, topographic highs of lobate coarse sediment with steep water-surface slope.

Feature	Pool	Riffle
Position	Outer (concave) bank of bends	Crossovers between bends
Cross-section	Asymmetric, narrow, deep	Symmetric, wide, shallow
Bed material	Fine	Coarse
Low-flow velocity	Low, flat water surface	High, steep water surface
At high flow	Scoured	Receives deposition

• Spacing is regular in both straight and meandering reaches, about 5–7 channel widths from pool to pool.
  • A meander usually has one pool-riffle sequence per bend; complex bends carry extra ones. Spacing holds even as riffle material moves downstream.
• The channel-forming flow is the frequent bankfull discharge (recurrence interval about 1.5 years); low flows are too weak and extreme floods too rare to set channel dimensions.
  • At low flow the bed controls flow geometry; at bankfull flow the flow shapes the bed, scouring pools and depositing on riffles, and the depth difference between them narrows as stage rises.
• Pools favour bends and generate sinuosity; riffles straighten the channel.
• Richards (1976): riffles are wider in curved reaches and pools wider in straight ones; mean velocity is higher in curved reaches, and tight bends produce deep, asymmetric pools of finer sediment.

Keller’s Five-Stage Model

• Edward A. Keller (1972) traced how a straight alluvial channel develops pools, riffles and bends.

Stage	Bed features	Spacing (channel widths)	Planform
1	No pools or riffles; asymmetric shoals	–	Straight banks, winding flow
2	Small, closely spaced pools and riffles	3–5	Still straight; shoals persist
3	Fully formed pools and riffles; point bars	5–7	Slightly sinuous; pools ~1.5× riffle length
4	Pools enlarge further; point bars	5–7, more regular	Sinuous
5	New pools and riffles at crossovers	Mean 5–7, mode 3–7	Meandering

• Keller stressed that not every alluvial river passes through all five stages.

Reach Types in Mountain Channels

• David R. Montgomery and John M. Buffington (1997) classified mountain reaches into seven types: colluvial, bedrock, cascade, step-pool, plane-bed, pool-riffle and dune-ripple.
  • The sequence runs from steep, supply-limited headwaters to gentle, transport-limited lowland reaches, as Himalayan step-pool streams give way to pool-riffle and dune-ripple reaches on the Ganga plains.

Channel Types and Patterns

Bases of Classification

• Channels are classified by the material of bed and banks and by planform pattern.
• Schumm (1985) recognised bedrock, semi-controlled and alluvial channels.
• Rolf Kellerhals, Michael Church and Dale I. Bray (1976) used the threshold of erodibility of bed and bank materials: non-erodible, semi-erodible and erodible.
  • This avoids a contradiction in “alluvial” versus “non-alluvial”: some non-alluvial materials (marine, glaciofluvial) erode easily, while armoured alluvial beds resist movement.
• J. D. Mollard (1973) identified 17 planform types related to physiographic setting, flow, sediment supply and bank materials; Church (1992) classified patterns by the calibre and volume of sediment supply.

Bedrock Channels

• Bedrock (erosional) channels are cut into consolidated rock where potential removal exceeds sediment supply: steep mountains, glaciated hard-rock regions and zones of active uplift.
• Structure and lithology dominate their form, and many follow faults and joints.
• They erode by corrasion (boulders acting as drilling tools cut potholes), corrosion and hydraulic action.
• They have low sinuosity but deep, narrow valleys, gorges and canyons, as in the Colorado’s Grand Canyon.
• Their long profiles are stepped by knickpoints, mostly structural rather than due to rejuvenation, with rapids, waterfalls and plunge pools.
• They are stable in form but transmit disturbances downstream efficiently.
• The upper reaches of the Indus, Satluj, Ganga, Yamuna, Gandak, Kali, Tista and Brahmaputra are bedrock channels that turn alluvial on the plains; the Narmada gorge at Bhedaghat is a Peninsular example.

Alluvial Channels

• Alluvial channels flow over their own deposits, as on the Ganga-Yamuna plain and broad coastal plains.
  • They alternate between degradation and aggradation, reworking deposits each wet season, and carry sinuous, meandering or braided patterns with pools, riffles, point bars, shoals and sand islands.
• Being erodible, they change with time. Hsieh Wen Shen and Schumm (1981) identified six temporal changes:

Change	What happens
Transverse bar shift	Bars migrate across a straight channel
Alternate bar shift	Bars and thalweg shift downstream within the channel
Neck cutoff	Meander loops meet and the neck is breached
Chute cutoff	Flow cuts across a point bar
Meander shift	Bends migrate across and down the valley
Avulsion	Rapid diversion to a new course

• Cutoffs shorten the channel and steepen its gradient, causing scour upstream and deposition downstream.
• Edward J. Hickin and Gerald C. Nanson (1975) related meander migration to discharge, water-surface slope, bank material, height of the concave bank, vegetation, the curvature-to-width ratio and sediment supply.
• By degree of change, alluvial channels are stable (usually straight) or unstable (typically braided).
• By load, five alluvial types are recognised:

Type	Load	Character
Straight	Suspended	Uniform depth
Straight	Mixed	Sinuous thalweg, small coarse sediment
Highly sinuous	Suspended	Uniform width, stable banks
Meander-braided transition	Large, sandy to gravelly	Wide, shallow, chute cutoffs, shifting thalweg
Bar-braided	Coarse bedload	Unstable, erodible banks, migrating bars

Feature	Bedrock channel	Alluvial channel
Bed and banks	Consolidated rock	Own sediment
Sediment supply	Below capacity	Near or above capacity
Pattern	Straight, structurally guided	Sinuous, meandering, braided
Change over time	Slow	Rapid
Indian example	Upper Satluj, Bhedaghat	Ganga, Kosi, Brahmaputra plains

Channel Patterns

• Channel pattern is the plan view of a reach as seen from the air, and is mainly a property of alluvial channels.
• Patterns have been classified by sinuosity, stability, number of channels, slope-discharge relations and load type (Luna B. Leopold and M. Gordon Wolman, 1957; Schumm; Andrew D. Miall, 1978; Brian R. Rust, 1978).

Sinuosity and Load-Based Schemes

• Leopold and Wolman (1957) used the sinuosity index (channel length ÷ valley length): straight below 1.05, sinuous 1.05–1.5, meandering above 1.5.
  • James C. Brice (1964) measured sinuosity against the length of the meander-belt axis instead.
• Schumm (1963) classified alluvial channels by load type:

Channel	Width/depth	Sinuosity	Gradient	When aggrading	When eroding
Suspended-load	< 7	> 2.1	Gentle	Bank deposition narrows it	Bed erosion dominant
Mixed-load	7–25	1.5–2.1	Moderate	Bank then bed deposition	Bed erosion, then widening
Bedload	> 25	< 1.5	Moderate to steep	Bed deposition, islands	Widening dominant

• A consolidated scheme combining Schumm and Miall recognises five patterns:

Pattern	Morphology	Sinuosity	Load	Width/depth
Straight	Single channel, pools and riffles, winding thalweg	< 1.05	Suspended, mixed or bed	< 40
Sinuous	Single channel, pools and riffles	1.05–1.5	Mixed	< 40
Meandering	Single channel, point bars	> 1.5	Suspended or mixed	< 40
Braided	Two or more channels, unstable bars and islands	> 1.3	Bedload	> 40
Anastomosing	Two or more channels, large stable islands	> 2.0	Suspended	< 10

• Leopold and Wolman gave no measure of channel multiplicity, so their scheme could not separate meandering from braiding; Rust (1978) added a braiding parameter (braids per meander wavelength).

Sinuosity	Single channel (braiding parameter < 1)	Multiple channels (braiding parameter > 1)
Low (< 1.5)	Straight	Braided
High (> 1.5)	Meandering	Anastomosing

• Rob I. Ferguson and Alan Werritty (1983) described transitional wandering gravel rivers: low to moderate sinuosity, wide shallow channels and bare gravel bars, but less division than true braiding.
• Patterns form a continuum: experiments show that the sequence straight → meandering → braided follows stream power, which reflects discharge and sediment load.
  • For a given discharge, braided rivers occupy steeper slopes than meandering ones, a threshold first shown by Leopold and Wolman.
• As sediment supply rises relative to transport capacity, channels pass from straight to sinuous to meandering; division round stable vegetated islands gives wandering or anastomosing channels, and division round unstable bars gives braiding.

Straight Channels

• Truly straight channels are rare except where a headwater stream follows a fault (George H. Dury, 1969); natural straight reaches seldom exceed about ten channel widths.
• Even there the thalweg winds between alternate bars, with riffles at crossovers.

Meandering Channels

• Meandering (sinuosity above 1.5) is the commonest alluvial pattern, especially where silt and clay dominate the load and gradients are moderate.
• Meander geometry includes wavelength (L), amplitude, meander-belt width, radius of curvature and the concave and convex banks.
  • The cross-section is asymmetric: the outer concave bank is eroded into a steep cut bank with deep water, while the inner convex bank receives sand and gravel as a point bar on the slip-off slope.
• Richards (1982) described three stages: a straight reach (sinuosity about 1.0) with alternating fast and slow flow at bankfull; a sinuous reach (about 1.1) as pools and riffles form at 5–7 widths; and a meandering reach (about 1.5) with a regular meander belt.
• Wavelength correlates strongly with channel width (L ≈ 10–14 W) and moderately with discharge.
  • Schumm (1967) related it to bankfull discharge and the percentage of silt-clay in the banks (M): L = 618 Qb^0.43 M^-0.74, so sand and gravel rivers have longer wavelengths than silt-clay rivers.
• Regular bends reflect hydrodynamic control; irregular ones reflect topographic, sedimentary or artificial disturbance (Ferguson, 1976).
• Meandering is now seen as the normal, self-maintaining form of alluvial flow rather than an accident of obstacles; once established it persists, though its ultimate cause is still debated.
• The Ganga and Ghaghara meander across the plains; entrenched meanders of rejuvenated rivers belong to rejuvenation.

Helical Flow

• Helical (secondary) flow is the corkscrew circulation of water round a bend, superimposed on the downstream flow.
• James Thomson (1876) first explained it; Albert Einstein (1926) later used the same principle to explain meandering and bank erosion.
• Mechanism
  • Centrifugal force piles water against the outer bank, causing superelevation of the surface there.
  • Surface water moves towards the outer bank and plunges down it.
  • Near the bed, where friction slows the flow, the pressure gradient drives water back towards the inner bank.
• Effects on the meander
  • The downward current scours the outer bank, deepening the pool and undercutting the cut bank.
  • The bottom return flow sweeps bedload to the inner bank, building the point bar.
  • The thalweg hugs the outer bank, and bends migrate laterally and down-valley until neck and chute cutoffs leave oxbow lakes.
• Whether helical flow causes meandering or is caused by it remains open; it certainly maintains the form.

Braided Channels

• A braided channel divides into many shifting channels separated by unstable bars and islands within highly erodible banks.
  • Bars change position, size and shape daily, monthly and seasonally; the section is wide and shallow, and the divided channels are together narrower than the single channel above.
• Brice (1964) proposed a braiding index: 2 × (sum of lengths of bars or islands) ÷ reach length; it is stage-dependent, since bar length depends on the water level.
• Longitudinal bars are diamond-shaped and gravelly; transverse bars are tabular, flat-topped and sandy, often dune fields exposed as flow falls.
• Causes include abundant coarse bedload, highly erodible banks, steep gradients and strongly fluctuating discharge, which also strips bank vegetation.
• Leopold and Wolman (1957) showed that braiding starts with a mid-channel bar: it splits the flow towards the banks, which erode, adding load that builds further bars.
  • Braiding reflects lack of competence, not of capacity, since aggradation can occur without it; Schumm similarly held that suspended load travels in sinuous channels and bedload in straight or braided ones.
  • The wide section develops several secondary-circulation cells, bars forming where near-bed flow converges.
• Braiding is typical of alluvial fans, deltas, glacial outwash plains and strongly seasonal regimes.
• The Brahmaputra is the classic case (James M. Coleman, 1969).
  • In flood it carries about 13 million tonnes of suspended sediment a day; sand waves reach about 15 m high and migrate up to about 600 m a day.
  • Channel shifts of several hundred metres a year are common, and Majuli, one of the world’s largest river islands, has lost much of its area to bank erosion.

Anastomosing and Anabranching Channels

• Anastomosing channels are multi-thread like braided ones, but their channels and vegetated islands are stable and keep their identity with changing discharge and time; channels are narrow, deep and of low gradient.
  • Leopold and Wolman treated anastomosing as a synonym of braided; Schumm (1963) separated them on stability.
• Anabranches are offshoots that rejoin the trunk, or a neighbouring trunk, sometimes after tens of kilometres (Dury, 1969); Schumm (1977) distinguished them from anastomosing channels, and individual branches may be straight, meandering or braided.
• Nanson and Knighton (1996) now use anabranching as the umbrella term for multiple channels separated by vegetated, semi-permanent islands, with six types (1–3 lower energy, 4–6 higher energy); anastomosing rivers are its organic, cohesive, low-energy type.
• The Brahmaputra and lower Ganga show anabranching reaches with large, inhabited chars alongside braided sections.
• Classical accounts reduce all this to three basic patterns, straight, meandering and braided, with the multi-thread forms as special cases.

Channel Change and Stability

Channel Stability and Movement

• Channel stability is a channel’s propensity for vertical or lateral movement (Church, 2006), and it changes as sediment is deposited or removed.
• Changes in pattern signal changes in the watershed, making channels an indicator of watershed health; each channel type responds differently to changes in supply or discharge.

Vertical Shifts

• Aggradation: when supply exceeds transport capacity, sediment accumulates, the width/depth ratio rises and stability falls.
• Degradation: when supply is below capacity, the bed is scoured and lowered.
  • Below dams and at intensive sand-mining sites on Indian rivers, sediment starvation causes bed incision that lowers water tables and undermines bridges.

Lateral Shifts

• Progressive bank erosion follows in-channel aggradation or ordinary meander migration.
• Avulsion is a sudden shift of the channel: first-order to a new part of the floodplain, second-order by re-occupying an old channel, and third-order by switching among channels of a braided reach.
• The limit of lateral movement is set by boundary conditions: in a wide valley of erodible material a river can sweep its whole floodplain, while confinement, bridges and dykes restrict it, and riparian forest can hold a stable form that would not otherwise exist.
• Changes in in-channel storage in bars and islands give early warning of instability; braided channels are the most active, with rapid lateral migration and net aggradation.
• The Kosi is India’s clearest case.
  • Crossing one of the world’s largest megafans with a huge load, it has swung more than 120 km westward over the last two to two-and-a-half centuries, oscillating rather than moving steadily.
  • On 18 August 2008 its embankment breached at Kusaha in Nepal and the river re-occupied a channel abandoned for over a century, a second-order avulsion that displaced nearly three million people in Bihar.

Applications

• In applied geomorphology, channel form and change guide bridge siting, embankment and flood-zone planning, navigation, settlement location and river-restoration schemes.
  • Meander-belt width and migration rates fix safe setback lines for roads and villages, while wide braided reaches need long bridges with guide bunds and flood plans that expect sudden avulsion.