Flood Routing - Theory & Concepts

Flood Routing

$Learning Objectives$

Define flood routing and differentiate between attenuation and lag.
Compare and contrast lumped (hydrologic) routing with distributed (hydraulic) routing.
Understand the principles and applications of Reservoir Routing (Level Pool Routing).
Apply the Muskingum Method for channel routing and calculate routing coefficients.
Explain the significance of the Saint-Venant Equations in hydraulic routing.

Techniques to predict the flood hydrograph at a downstream section, including Reservoir Routing and Channel Routing methods like Muskingum, crucial for flood forecasting and control structure design.

What is Flood Routing?

$Flood Routing$

The technique of determining the flood hydrograph at a downstream section of a river or reservoir by utilizing the data of flood flow at one or more upstream sections.

It is used to predict two main properties of a flood wave:

$Magnitude (Attenuation)$

The reduction of the peak flow downstream as the flood wave spreads out.

$Timing (Lag or Translation)$

The delay in time of the peak flow from the upstream section to the downstream section.

$Lumped vs. Distributed Routing$

Hydrologic Routing (Lumped): Uses the continuity equation and storage-discharge relationships (e.g., Muskingum, Modified Puls). It treats the reach as a "black box," calculating flow only at the ends of the reach.
Hydraulic Routing (Distributed): Solves the continuity and momentum equations (St. Venant Equations) simultaneously. It describes flow as a function of both space and time at every point along the channel.

Hydraulic Routing and the St. Venant Equations

While hydrologic routing methods (Muskingum, Level Pool) are lumped models based only on continuity, Hydraulic Routing is a distributed method based on the 1D shallow water equations, known as the St. Venant equations.

$Saint-Venant Equations$

A pair of partial differential equations that describe unsteady open channel flow:

Continuity Equation: Conservation of Mass.
Momentum Equation: Conservation of Momentum, balancing forces of gravity, friction, pressure gradient, and convective/local acceleration.

Saint-Venant Continuity Equation

The conservation of mass equation for unsteady open channel flow.

\frac{\partial A}{\partial t} + \frac{\partial Q}{\partial x} = q

Variables

Symbol	Description	Unit
$A$	Cross-sectional area of flow	$m^2$
$t$	Time	s
$Q$	Discharge	$m^3/s$
$x$	Longitudinal distance along the channel	m
$q$	Lateral inflow per unit length	$m^2/s$

$Simplified Hydraulic Models$

Depending on the channel slope and flow conditions, some terms in the Momentum equation can be neglected to simplify calculations:

Kinematic Wave: Assumes friction and gravity forces balance exactly (ignoring pressure and acceleration). It translates the wave but does not attenuate it.
Diffusion Wave: Includes pressure gradients in addition to friction and gravity, allowing the wave to both translate and attenuate (flatten) over time.
Dynamic Wave: Uses the full Saint-Venant equations, accounting for backwater effects, tidal influence, and rapid flow changes that other models cannot handle.

Reservoir Routing (Level Pool Routing)

Used for routing floods through reservoirs or lakes where the water surface is horizontal. The inflow hydrograph ( $I$ ) is modified by the reservoir's storage ( $S$ ) to produce an outflow hydrograph ( $O$ ).

$Reservoir Routing Principle$

Reservoir Routing assumes a level water surface (Level Pool Routing). Because the pool is level, the storage within the reservoir is a function solely of the outflow discharge or water surface elevation ( $S = f(O)$ ).

Continuity Equation

The fundamental principle of conservation of mass applied to a reservoir.

I - O = \frac{dS}{dt}

Variables

Symbol	Description	Unit
$I$	Inflow rate	$m^3/s$
$O$	Outflow rate	$m^3/s$
$dS/dt$	Rate of change of storage	$m^3/s$

Discretized for a time interval $\Delta t$ :

Discretized Continuity Equation

The discrete form of the continuity equation used for step-by-step numerical routing.

\frac{I_1 + I_2}{2} \cdot \Delta t - \frac{O_1 + O_2}{2} \cdot \Delta t = S_2 - S_1

Variables

Symbol	Description	Unit
$I_1, I_2$	Inflow at beginning and end of time step	$m^3/s$
$O_1, O_2$	Outflow at beginning and end of time step	$m^3/s$
$S_1, S_2$	Storage at beginning and end of time step	$m^3$
$\Delta t$	Time step duration	s

This equation is often solved using the Storage-Indication Method or Modified Puls Method by constructing a curve of $(2S/\Delta t + O)$ vs $O$ to find the unknown outflow $O_2$ at each time step.

Level-Pool Routing by the Modified Puls Method

Use the storage-indication relation and discretized continuity to compute reservoir outflow through successive time intervals.

The storage-indication table relates O to 2S/Δt + O using the reservoir storage-elevation and outflow relations. For each interval, compute the right-hand side from the known inflows, prior storage and prior outflow; interpolate the table for the new outflow, then update storage and carry the new state into the next interval. Keep discharge, storage and time units consistent with the chosen Δt.

Muskingum Method (Channel Routing)

Used for routing floods in river channels. Unlike reservoirs, the water surface in a channel slopes downstream. Therefore, storage in a channel must account for both prism storage (the volume below a line parallel to the streambed) and wedge storage (the volume between the actual water surface profile and the prism level during flood passage). Thus, channel storage is a function of both inflow ( $I$ ) and outflow ( $O$ ).

Muskingum Storage Equation

Relates channel storage to a weighted combination of inflow and outflow.

S = K [xI + (1-x)O]

Variables

Symbol	Description	Unit
$S$	Total channel storage	$m^3$
$K$	Storage time constant, approximating the travel time of the flood wave through the reach.	hours or seconds
$x$	Weighting factor representing the relative importance of inflow vs. outflow on storage (0 to 0.5).	-
$I$	Inflow rate	$m^3/s$
$O$	Outflow rate	$m^3/s$

$Interpreting the Weighting Factor (x)$

$x=0$ : Wedge storage is zero. Storage depends only on outflow, equivalent to Reservoir-type storage (Linear Reservoir).
$x=0.5$ : Pure translation without attenuation (wedge storage equals prism storage).
Typical natural streams have $x \approx 0.2$ .

Muskingum Routing Equation

By substituting the Muskingum Storage Equation into the discretized continuity equation, we derive an explicit equation to calculate the outflow at the next time step ( $O_2$ ) from known values ( $I_1, I_2, O_1$ ).

Muskingum Routing Equation

Calculates the outflow at the end of a time step based on current and past inflows and past outflow.

O_2 = C_0 I_2 + C_1 I_1 + C_2 O_1

Variables

Symbol	Description	Unit
$O_2$	Outflow at the end of the time step	$m^3/s$
$I_2$	Inflow at the end of the time step	$m^3/s$
$I_1$	Inflow at the beginning of the time step	$m^3/s$
$O_1$	Outflow at the beginning of the time step	$m^3/s$
$C_0, C_1, C_2$	Muskingum routing coefficients	-

Where the coefficients are derived as:

Routing Coefficient C0

Equation for calculating the Muskingum coefficient C0.

C_0 = \frac{-Kx + 0.5\Delta t}{K(1-x) + 0.5\Delta t}

Variables

Symbol	Description	Unit
$C_0$	Muskingum routing coefficient C0	-
$K$	Storage time constant	hours
$x$	Weighting factor	-
$\Delta t$	Time step duration	hours

Routing Coefficient C1

Equation for calculating the Muskingum coefficient C1.

C_1 = \frac{Kx + 0.5\Delta t}{K(1-x) + 0.5\Delta t}

Variables

Symbol	Description	Unit
$C_1$	Muskingum routing coefficient C1	-
$K$	Storage time constant	hours
$x$	Weighting factor	-
$\Delta t$	Time step duration	hours

Routing Coefficient C2

Equation for calculating the Muskingum coefficient C2.

C_2 = \frac{K(1-x) - 0.5\Delta t}{K(1-x) + 0.5\Delta t}

Variables

Symbol	Description	Unit
$C_2$	Muskingum routing coefficient C2	-
$K$	Storage time constant	hours
$x$	Weighting factor	-
$\Delta t$	Time step duration	hours

$Coefficient Verification$

For the routing to conserve mass correctly, the sum of the coefficients must always equal exactly 1: $C_0 + C_1 + C_2 = 1$ .

$Numerical Stability$

For numerical stability and to prevent negative outflows, the routing interval $\Delta t$ should be chosen such that $2Kx \le \Delta t \le K$ .

Muskingum-Cunge Method

The original Muskingum method requires historical flood data to calibrate the parameters $K$ and $x$ . The Muskingum-Cunge method overcomes this limitation by linking the routing parameters directly to the physical hydraulic characteristics of the channel (geometry, slope, roughness).

$Physical Basis of Muskingum-Cunge$

By matching the Muskingum finite difference scheme with a simplified form of the Saint-Venant momentum equation (diffusion wave), Cunge demonstrated that $K$ and $x$ can be calculated directly.

$K$ is related to the travel time through the reach length $\Delta x$ at celerity $c$ ( $K = \Delta x / c$ ).
$x$ is derived from the channel width ( $B$ ), slope ( $S_0$ ), and discharge ( $Q$ ), acting as a diffusion term. This makes Muskingum-Cunge a pseudo-hydraulic method, highly powerful for ungauged rivers.

$Interactive Simulation: Muskingum Routing$

Use the simulation below to explore how altering the travel time ( $K$ ) and weighting factor ( $x$ ) affects the attenuation and lag of a flood wave.

Muskingum Channel Routing Simulation

Adjust the Storage Time Constant (K) and Weighting Factor (x) to see how the flood wave is attenuated and delayed.

K (Storage Constant): 12 hours

x (Weighting Factor): 0.2

Δt (Time Step): 6 hours

C0: 0.0476

C1: 0.4286

C2: 0.5238

ΣC: 1

Loading chart...

Flood-Routing Method Selection and Verification

Select a routing model from the physical system and required fidelity, then verify parameters, time step, continuity, and the routed hydrograph before use.

Define routing objective → Collect inflow hydrograph, geometry, storage and boundary data; Collect inflow hydrograph, geometry, storage and boundary data → Level-pool reservoir or lake?; Level-pool reservoir or lake? — Yes → Build storage–elevation and outflow relation; Build storage–elevation and outflow relation → Route with level-pool / Modified Puls continuity; Route with level-pool / Modified Puls continuity → Continuity, attenuation, lag and behavior credible?; Level-pool reservoir or lake? — No → Need distributed backwater or rapid-wave detail?; Need distributed backwater or rapid-wave detail? — Yes → Select Saint-Venant dynamic routing; Select Saint-Venant dynamic routing → Set model boundaries and time step; Set model boundaries and time step → Route the hydrograph step by step; Need distributed backwater or rapid-wave detail? — No → Estimate K and x for the reach; Estimate K and x for the reach → Routing OK?; Routing OK? — No → Adjust parameters or routing interval; Adjust parameters or routing interval — Recheck → Estimate K and x for the reach; Routing OK? — Yes → Route the hydrograph step by step; Route the hydrograph step by step → Continuity, attenuation, lag and behavior credible?; Continuity, attenuation, lag and behavior credible? — No → Revise time step, parameters or model choice; Revise time step, parameters or model choice — Reassess → Define routing objective; Continuity, attenuation, lag and behavior credible? — Yes → Document verified routed hydrograph and assumptions; Document verified routed hydrograph and assumptions → Use routed result

Define routing objective: terminator
Collect inflow hydrograph, geometry, storage and boundary data: input output
Level-pool reservoir or lake?: decision
Build storage–elevation and outflow relation: process
Route with level-pool / Modified Puls continuity: process
Need distributed backwater or rapid-wave detail?: decision
Select Saint-Venant dynamic routing: process. Use distributed continuity and momentum equations when backwater, spatial variation or rapid-wave behavior requires hydraulic detail.
Set model boundaries and time step: process
Estimate K and x for the reach: process. Calibrate the original Muskingum method from flood data. For ungauged reaches, estimate parameters from channel geometry, slope, roughness and discharge with Muskingum-Cunge.
Routing OK?: decision. Check C0 + C1 + C2 = 1 and, for the lesson's Muskingum formulation, 2Kx ≤ Δt ≤ K with consistent time units.
Revise time step, parameters or model choice: process
Adjust parameters or routing interval: process
Route the hydrograph step by step: process
Continuity, attenuation, lag and behavior credible?: decision
Document verified routed hydrograph and assumptions: document
Use routed result: terminator

Key Takeaways

Flood Routing predicts the changes in a flood wave (magnitude and timing) as it moves downstream.
Attenuation is the reduction of the peak discharge. Lag (Translation) is the delay in time of the peak flow.
Hydrologic routing (lumped) uses simplified continuity, treating the reach as a single control volume.
Hydraulic routing (distributed) solves the 1D Saint-Venant equations (continuity and momentum) for highly accurate distributed flow modeling, accounting for complex channel hydraulics.
Reservoir Routing assumes a horizontal water surface (Level Pool Routing), where storage is a function solely of outflow ( $S = f(O)$ ).
The Modified Puls Method is the standard numerical technique for solving level pool routing.
Channel Routing must account for both prism and wedge storage, meaning channel storage depends on both inflow and outflow ( $S = f(I, O)$ ).
The Muskingum Method is the standard hydrologic channel routing technique. The parameter $K$ represents travel time, and $x$ weights the relative effects of inflow vs. outflow on storage.
The sum of Muskingum routing coefficients ( $C_0, C_1, C_2$ ) is always unity.
The Muskingum-Cunge Method physically bases $K$ and $x$ on channel geometry, making it suitable for ungauged streams.

Previous TopicHydrographs - Examples & Applications

Quiz Me

Next TopicFlood Routing - Examples & Applications

Prev Next

Quiz Me