You are on page 1of 18

JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 112, F03025, doi:10.

1029/2006JF000734, 2007

Predicting downstream hydraulic geometry:


A test of rational regime theory
Brett C. Eaton1 and Michael Church1
Received 5 December 2006; revised 14 April 2007; accepted 7 June 2007; published 21 September 2007.

[1] The classical equations of hydraulic geometry are purely empirical, but the
widespread similarity of the scaling (downstream) form of them suggests that they express
some important underlying regularities in the morphology of stream channels through
the drainage network. A successful physical theory of river regime must be able to
reproduce and explain this regularity. In this paper we test the regime theory of Eaton et al.
(2004) using selected data of hydraulic geometry. We first use data from environments in
which bank strength presumably does not vary greatly, such as in anabranched channel
systems and deltas. Regime models parameterized by assuming uniform relative bank
strength plausibly describe the observed bankfull channel geometries in these systems. We
then test a modified bank strength formulation for vegetated gravel bed rivers against
downstream hydraulic geometry data sets in which relative bank strength is supposed to
vary with channel scale. Assuming a uniform effective cohesion due to riparian
vegetation, the regime model is again able to reproduce details of the channel geometry.
Both analyses show that the classical hydraulic geometry represents only an approximation
of the variation of channel form. If we have confidence in the theory, we may infer
information about bank strength and bed material transport. The pattern of variation in
these quantities, as well as discharge, through the drainage system lends approximate
regularity to stream channel scaling that is summarized in the empirical relations.
Citation: Eaton, B. C., and M. Church (2007), Predicting downstream hydraulic geometry: A test of rational regime theory,
J. Geophys. Res., 112, F03025, doi:10.1029/2006JF000734.

1. Introduction theory’’ [see, e.g., Lacey, 1930; Blench, 1969]). These


investigators recognized that a critical control over river
[2] Downstream hydraulic geometry [after Leopold and
channel form is exercised by the materials that form the bed
Maddock, 1953] has become a standard tool to describe the
and banks of the channel, but they were able to quantify the
variation with flow of stream channel geometry through the
effect only by an index ‘‘silt factor’’ [Lacey, 1930; Blench,
channel network. The equations are empirical, but signifi-
1969] or by stratifying regime relations according to the
cant attempts have been made to develop a theoretical basis
character of bed and bank materials [Lane, 1957; Simons
for them [e.g., Langbein and Leopold, 1966; Kirkby, 1977;
and Albertson, 1963]. This factor has mostly been ignored
Parker, 1978, 1979; Chang, 1980; Yang et al., 1981; White
in considering the hydraulic geometry of rivers, although
et al., 1982; Davies, 1987; Cao, 1996; Huang and Nanson,
thoughtful investigators have introduced the related factor
2000]. By their very construction they are incomplete
of vegetation-mediated bank strength [Andrews, 1984; Hey
statements with no strictly physical interpretation, but they
and Thorne, 1986; Huang and Nanson, 1998], again by
can properly be interpreted as scaling relations between the
index methods, while others have restricted investigations to
‘‘effective channel-forming flow’’ and the accommodation
a particular type of bounding sediment [Bray, 1973; Neill,
of that flow by the covariation of stream width, depth and
1973; Charlton et al., 1978; Andrews, 1984; Hey and
velocity. The usual power law representation of them is, to
Thorne, 1986].
this extent, appropriate.
[4] The purpose of this paper is to explore the hydraulic
[3] The hydraulic geometry arose by analogy with
geometry of selected river systems in terms of a rational
‘‘regime relations’’ established early in the 20th century
regime theory, both as a test of the theory and in order to
for the design of unlined irrigation canals. The earliest
attempt to learn more about the covariation of terrain and
relations were deliberately empirical [Kennedy, 1895; Lindley,
hydrological conditions through a drainage system that
1919], but they were succeeded by attempts to clothe the
leads to the familiar empirical equations for the variation
relations in the semblance of a theory (hence ‘‘regime
of width, depth and velocity.
1
Department of Geography, University of British Columbia, Vancouver,
British Columbia, Canada. 2. Statement of the Regime Approach
[5] A rational regime theory specifies a set of physical
Copyright 2007 by the American Geophysical Union.
0148-0227/07/2006JF000734
relations that apply the relevant governing conditions to

F03025 1 of 18
F03025 EATON AND CHURCH: HYDRAULIC GEOMETRY F03025

enumerated above). It turns out that this condition is


mathematically equivalent to the condition that the imposed
sediment load be passed with the smallest expenditure of
energy (in a particular sense, ‘‘with maximum efficiency’’
[Millar, 2005] or, more obscurely, ‘‘at capacity’’). Accord-
ingly, the preferred gradient– if it be accessible to the river –
is the minimum channel gradient [Eaton et al., 2004]. One
important way in which the maximum resistance solution
differs from other proposals is that it is constrained by the
relative erodibility of the bed and the banks. A reduction in
slope for a constant transport capacity is accomplished by
narrowing the channel, thereby increasing the channel depth
and mean boundary shear stress. If the channel becomes too
narrow and deep, the shear stress acting on the banks will
exceed the critical threshold for erosion and the banks will
fail, therefore bank strength places a limit on the degree to
Figure 1. Relation between apparent friction angle and the which a channel can narrow and deepen, and thus places a
relative strength of the channel banks compared to the limit on the degree to which the channel gradient can be
channel bed. reduced.
[8] The regime models used in the following analyses are
based on two separate approaches to the bank strength
determine the hydraulic geometry of a self-formed channel. problem. The first, more general model represents bank
The complete hydraulic geometry includes, in addition to strength as a function of the bed erodibility via an apparent
width, depth and mean flow velocity, channel gradient and friction angle, f0 [after Millar and Quick, 1993], which
planform geometry. The principal governing conditions are describes the ratio between bank strength and bed strength
the imposed flow and sediment load (which includes both [Millar, 2005], as shown in Figure 1. Because of the simple
caliber and quantity of bed material sediment), the valley nature of this model, it is not necessary to make any
gradient down which the flow and sediment must be passed, assumptions about the bank stratigraphy or the physical
and the character of the deposits in which the channel is origin of the bank strength, which is expressed merely as
formed. relative to the bed strength. While this is an intuitive way to
[6] We know that, if the discharge (Q), bed material think about bank strength, the parameter f0 is difficult to
sediment transport rate (Qbm) and sediment caliber (D) are estimate.
specified, then a channel geometry comprising gradient (S), [9] The model assumes a trapezoidal channel geometry,
hydraulic radius (R) and wetted perimeter (P) can be which facilitates the use of empirical functions to estimate
calculated. Applying a bank stability criterion does not the shear stress distribution along the channel boundary on
isolate a unique solution since a range of channel geome- the basis of the channel shape and the average boundary
tries satisfies the bank stability constraint. In order to isolate shear stress [Millar and Quick, 1993]. The threshold for
a unique solution one must apply some sort of optimality bank erosion depends on the specified critical dimensionless
criterion, such as minimum variance of hydraulic quantities shear stress (or Shields number) for entrainment of the
[Langbein and Leopold, 1966] (a purely statistical concept), representative grain size used to characterize the banks
minimum unit stream power [Yang, 1976], maximum sed- (t c*), the relative strength of the banks (f0) and the side
iment transport efficiency [Kirkby, 1977], minimum power slope angle of the trapezoid. The critical shear stress for
expenditure [Chang, 1979], maximum sediment transport- entrainment of the bank material is selected according to the
ing capacity [White et al., 1982], least action [Huang and character of the bed, and the reference grain size used in the
Nanson, 2000], or minimum energy expenditure [Huang et analysis. For gravel bed rivers in which the surface median
al., 2004]. Since these optimality criteria all arrive at nearly grain size, D50, is used in the bank stability analysis, t c* is
identical solutions (but from different assumptions), the set at 0.03; when the surface D84 is used, t c* is set at 0.02,
differences between them lie not in the predictions that they and if the D90 is used t c* is set at 0.015. Since the typical
make so much as in the physical implication of the ratio of D90:D50 in gravel bed streams is approximately two,
assumption. setting t c*(D90) = 0.015 is tantamount to applying an
[7] We have recently proposed what we believe to be a entrainment threshold of t c*(D50)  0.03, which is cited
plausible physically based solution [Eaton et al., 2004]. The as an appropriate threshold for gravel bed streams [Neill,
stable state is the one in which resistance to flow is 1968; Andrews, 1982]. For sand bed rivers, a set of linear
maximized subject to the condition that the imposed water functions that estimate the critical Shields number as a
and sediment load – that is, the load delivered from function of the grain Reynolds number (Re*) is used in
upstream – be passed on the adopted channel gradient. the bank stability analysis [Van Rijn, 1984].
This condition still permits exchange of sediment between [10] Bed material sediment transport capacity (Qbm) is
the load and the bed, hence lateral movement of the estimated in the model using two different equations. For
channel, but no net change occurs in the average channel gravel bed streams, the Parker [1990] equation is used.
form. Flow resistance is maximized with respect to the When the bed material is primarily sand, we use the Van
fluvial system, which refers to the valley gradient, Sv, as the Rijn [1984] equation. In both cases, the calculated sediment
controlling system gradient (cf. the governing conditions transport capacity is interpreted as merely an index of the

2 of 18
F03025 EATON AND CHURCH: HYDRAULIC GEOMETRY F03025

actual transport capacity since bed load transport equations [15] We face a serious problem in attempting to test these
are not very accurate. two models against empirical data. We know of no field
[11] A regime solution is found using as input parameters data set that presents sufficient information to make a
values of slope (S), formative discharge (Q), a flow resis- rigorous test of a complete regime theory, primarily because
tance parameter (in this analysis, Manning’s n) and surface there exists no practical measurement procedure to deter-
grain size (often D50). Solution geometries are found that mine the shear strength of channel banks in the field, and no
satisfy the constraint that the banks are stable for the range field report of such data is known to us. The general
of possible trapezoid bed widths, and the predicted bed objective of this paper, then, is to obtain an indirect test
material sediment transport capacity (Qbm) is plotted against of our regime models by applying the appropriate formula-
the bed width. The chosen solution is the one associated tion to available data sets of the hydraulic geometry of
with the highest sediment transport capacity (i.e., the most alluvial channels. Specifically, we (1) apply the regime
efficient channel), since this is equivalent to the minimum model to data sets in which the relative bank strength (f0)
slope for a constant value of Qbm [White et al., 1982] which can plausibly be assumed to be uniform, (2) apply the
in turn is equivalent to the maximum flow resistance for the regime model to data sets in which the vegetation-related
fluvial system [Eaton et al., 2004]. bank strength (C0) can be assumed to be uniform, and
[12] In the original presentation of the regime model (3) invert the regime model to fit the observed widths by
[Eaton et al., 2004], flow resistance was indexed using varying the relative bank strength (f0) for the data sets
the Darcy-Weisbach coefficient. Manning’s n is adopted referred to in (2). By applying physically based models, we
here since it is related to powers of the hydraulic variables, are able to investigate the physical processes underlying
which is convenient for scaling arguments. The regime downstream hydraulic geometry relations.
models have been coded in MATLABTM for numerical
solution. 3. Systems With Nominally Uniform Relative
[13] For gravel bed streams with vegetated floodplains, Bank Strength
we use an alternative bank strength model proposed by
Eaton [2006]. This model parameterizes bank strength for 3.1. Methods
vegetated banks by recognizing a root-laced, semicohesive [16] We found five cases in which it seems reasonable to
upper bank section of height H overlying a lower bank with assume that relative bank strength is approximately uni-
properties similar to those of the streambed, to which a form. For each data set, the regime formulation using f0 was
frictional strength proportional to the true friction angle, f, adopted. The value of f0 used in the model was varied to
is applied. The parameter H is thought to be roughly bring the predicted and the observed bankfull widths and
equivalent to the rooting depth for the dominant riparian depths as close to agreement as possible. Each data set
vegetation, and can be interpreted as an effective cohesion presented values of Q, S, surface grain size and sometimes n
(C0) due to riparian vegetation. Since two parameters are differently and, presumably, with various degrees of accu-
required (C0 and f), this model is slightly more complicated racy. Therefore we were forced to make a range of assump-
than the f0 model. The primary advantage of it is that it tions about the input parameters.
explicitly incorporates the scale-dependent nature of bank [17] We investigate the model performance by fitting
strength due to riparian vegetation: while the relative bank power laws to the modeled channel dimensions and com-
strength (f0) for a forested stream declines downstream paring these equations to the empirical power laws. For the
(because the channel steadily grows deeper than the riparian sake of consistency, we have calculated the power law
rooting depth), becoming similar to streams with grassy relations for both the empirical data sets and the modeled
banks for bankfull flows greater than about 400 m3/s [Eaton channel dimensions by applying linear regression to the log-
and Millar, 2004], the absolute vegetation-related bank transformed variates: this allows us to directly compare the
strength for a forest (C0) remains constant, even though scale relations associated with the regime model with the
the importance of the vegetation in determining the channel scale relations associated with the empirical data without
shape declines [Eaton, 2006]. encountering the potentially biasing effects of different
[14] This model imposes a composite channel cross curve-fitting procedures. We also calculate RMS error and
section, comprising vertical upper banks atop a trapezoidal mean bias statistics, both expressed as a percent of the
lower channel. The stability analysis is applied to the sides estimate, for the regime model predictions, and RMS errors
of the lower trapezoid, which represents the cohesionless for the empirical equations. These statistics help us to assess
gravel toe of the channel. Deriving a regime solution, one the performance of the regime model and to compare the
proceeds in exactly the same way as above, with one further relative ability of both approaches to explain the observed
complication: once the channel depth predicted by the variance in the data set.
model becomes less than H, there is no cohesionless gravel 3.1.1. Colorado River
toe upon which to perform the bank stability analysis. This [18] The first case is a data set on the single-thread,
affects a very small number of streams in the data sets that meandering Colorado River published by Pitlick and Cress
we analyze, since it applies only in situations where the [2002]. We consider only those reaches that are identified as
relative bank strength is very high (i.e., small channels in alluvial or quasi-alluvial which, the authors suggest, are
forested streams where root systems dominate the channel generally free to adjust their channel width. All of the
banks). In this circumstance, we bypass the bank stability reaches are relatively wide and deep (>2.4 m; relative
analysis and simply maximize the sediment transport roughness, D50/d, <0.025) with minimal cohesive bank
capacity [after White et al., 1982]. This procedure selects sediment, so we have made the assumption that the relative
the narrowest and deepest possible channel. bank strength is fairly uniform and that it is relatively low.

3 of 18
F03025 EATON AND CHURCH: HYDRAULIC GEOMETRY F03025

Figure 2. Data from the Colorado River [Pitlick and Cress, 2002] used to estimate slope, Manning’s n
and D50 as functions of Q. The functions shown in the plots were used to specify the regime model (see
Figure 3); the normalized standard errors for the estimates of S, n, and D50 are 8.9%, 1.4%, and 6.5%,
respectively.

The authors estimated the bankfull flows using the cross from measurements of average velocity and hydraulic radius
sectional areas and estimates of Manning’s n from fitting a reported by Ellis and Church [2005], using the Strickler
1-D flow model to the reach of interest [Pitlick and Cress, equation to estimate the grain Manning’s n for the measured
2000]. We use their (unpublished) estimates for bankfull surface D84. Since total flow resistance is typically under-
flow and Manning’s n in conjunction with the published estimated by the Strickler equation, the predicted slopes are
channel slopes and surface textures of Pitlick and Cress lower than they ought to be. The average slope for the
[2002]. We developed empirical equations for interpolating system is tolerably well known, and when the estimates of
S, n and D50 as functions of Q (Figure 2). We then used the slopes for each study reach are scaled upward by 50%,
these functions to generate input parameters for the regime the average of the estimates agrees with the known average
model over the observed range of Q, and compared the slope for the system. We use these scaled slope values to
predicted channel geometry with the observed geometry. parameterize the regime model. Since D84 was used to
Since we have parameterized the model by choosing a value estimate S, we use D50 for the bank stability analysis. Since
of Q, then estimating S, n, and D50 as a function of Q, the we have already used the measurements of velocity and
regime model produces smooth trends. In order to determine depth to estimate S, it would be inappropriate to use them
the effect of using these functions to smooth the model again to estimate n. Therefore we have held n constant (at a
output, we also ran the model using the raw input data for value of 0.03) to generate predictions of width and depth: n
each study reach. values back calculated from our adjusted slopes range from
[19] One of the reaches (Professor Valley) [Pitlick and 0.024 to 0.036 with a mean value of 0.029.
Cress, 2002] was associated with a bankfull discharge about 3.1.3. Columbia River
twice that of the reaches immediately upstream and down- [22] Tabata and Hickin [2003] reported a data set on sand
stream, and was excluded from the analysis on the grounds bed anabranch channels of the upper Columbia River. The
that the identified channel geometry most likely represents banks of these channels are almost entirely composed of silt
the dimensions of a low terrace rather than a bankfull and clay, which we believe is consistent with a uniformly
channel. high relative bank strength. Data reported by Tabata and
3.1.2. Fraser River Hickin [2003] were used to estimate average values of S
[20] A data set from gravel bedded anabranches of the (0.001 m/m), n (0.028), and D50 (0.5 mm). The formative
Fraser River was published by Ellis and Church [2005]. The discharge was estimated from field measurements in each
bankfull channel dimensions and bankfull discharge were anabranch that was surveyed. The regime model applied to
carefully estimated by interpolation amongst measurements the Columbia held S, n, D50 and f0 constant and varied only
to higher than bankfull stage for a number of anabranches Q, generating predictions of width and depth that vary
and for the main channel of the Fraser River. Since all of the smoothly with Q.
anabranch channels are relatively deep (>2.5 m; D50/d, 3.1.4. Danube Delta
<0.029), and are formed in materials having similar sedi- [23] The fourth case is a data set on the sand bed delta
mentological properties and riparian vegetation, we can distributary channels of the Danube delta [Mikhailov, 1970].
assume that the relative bank strength is uniform (and These channels are sedimentologically similar to the anab-
relatively low). ranches in the upper Columbia River, and we have assumed
[21] The regime model applied to this system uses the that they are associated with a uniform, high relative bank
measured values of Q and D50 for each study reach. The strength. We used the estimated formative discharges in
channel gradient is small enough to be difficult to measure conjunction with the relations reported by Mikhailov that
accurately over the length (250 m) of a single study reach. express S and n as functions of Q to parameterize the model.
Therefore values of the slope for each reach were estimated

4 of 18
F03025 EATON AND CHURCH: HYDRAULIC GEOMETRY F03025

Figure 3. Plots of the Colorado River regime analysis [Pitlick and Cress, 2002]. (left) Bed material
transport concentration (Qbm/Q) estimated by the model plotted against Q. Qbm/Q is a proxy for the bed
material yield (Qbm/area) at each point, since drainage area and formative discharge are strongly
correlated. (middle) Modeled channel shape (as described by the W/d ratio) plotted against the predicted
dimensionless shear stress. (right) Predicted channel widths and depths (W, d) plotted against Q. The
available data from the Colorado River are plotted for comparison.

The data reported support the selection of only an average, geometry is the variation of grain size and slope down-
representative bed grain size (0.15 mm). stream, possibly driven by a declining supply of bed
3.1.5. Laitaure Delta material along the system.
[24] The last case comprises data from the sand bed [27] The regime model fits the data from the Colorado
channels of the Laitaure delta in northern Sweden, reported River when the relative bank strength is set to about 1.25
by Andrén [1994]. The energy gradient for these channels (f0 = 36°). Figure 3 plots both the regime model predictions
was held constant (0.0002 m/m, consistent with values and the observed channel geometry against the estimated Q.
provided by Andrén [1994, p. 84]), as was Manning’s n While the relation between predicted W and Q is fairly well
(0.025). On the basis of the empirical data reported by described by a power law, the relation predicted by the
Andrén, grain size was varied as a function of discharge: regime model between d and Q is not well described by a
power function. The empirical data are at least as consistent
D ¼ 0:14Q0:18 R2 ¼ 0:54 ð1Þ with the regime-based depth relation, insofar as they suggest
no clear preference for a power law fit. The RMS error for
[25] The regime model was parameterized using the estimating width and depth using the regime model is
reported formative discharge values, the constant values of almost identical to the error associated with the empirical
S and n, and the function relating the characteristic grain power fits (Table 1). Table 1 also presents the RMSE
size to Q. estimates for a regime model based on the raw field data
(not shown in Figure 3) rather than the fitted functions of
3.2. Results Figure 2: the values are nearly identical to the results
3.2.1. Colorado River discussed above.
[26] Pitlick and Cress [2002] express their empirically [28] While the regime model results are not entirely
based hydraulic geometry equations using the dimension- consistent with power law fits, they can be used to estimate
less parameters of Andrews [1984], W* = W/D50, d* = d/ dimensionless hydraulic geometry relations of the same
Q form as equations (2) and (3). The equations are
D50, and Q* = pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi , where s is the specific gravity
ðs1ÞgD50 D250
of the sediment grains. The equations based on their data are
W * ¼ 34:3Q*0:34 ; R2 ¼ 0:99; Std: Err: ðbÞ ¼ 0:0067 ð4Þ
W * ¼ 32:4Q*0:34 ; R2 ¼ 0:99; Std: Err: ðbÞ ¼ 0:0075 ð2Þ
d* ¼ 0:051Q*0:55 ; R2 ¼ 0:99; Std: Err: ð f Þ ¼ 0:0089 ð5Þ
0:55 2
d* ¼ 0:055Q* ; R ¼ 0:99; Std: Err: ð f Þ ¼ 0:0119 ð3Þ
While these equations are not identical to equations (2) and
(3) they are sufficiently similar that we can conclude that the
These equations differ from the results typically reported regime model has reasonably reproduced the hydraulic
from studies of downstream hydraulic geometry (see geometry equations for this reach.
Knighton [1998, p. 173] for a summary), in which width [29] In order to assess the downstream changes in the
/ Q0.5, and depth / Q0.4, approximately, such that the transport capacity, the Parker [1990] equation was used to
width/depth ratio typically increases downstream. Pitlick calculate the bed material sediment transport concentration,
and Cress’s [2002] data set has the peculiar property that the Qbm/Q. Bed material concentration declines systematically
width/depth ratio decreases downstream. The authors downstream at a rate that is faster than if Qbm were constant
speculate that the origin of this feature of the hydraulic (Figure 3). This implies that the absolute yield of bed

5 of 18
F03025 EATON AND CHURCH: HYDRAULIC GEOMETRY F03025

Table 1. RMSE Values for Width and Depths Estimated Using Regime Models
Width Depth Empirical Power Law
RMSE RMSE
Data Set Bank Strength RMSE, % BIAS, % RMSE, % BIAS, % Width, % Depth, %
Colorado River f0 = 36° 15.9 (16.2)a 8.8 (9.0) 8.1 (5.5) 0.5 (0.9) 15.2 8.8
Fraser River anabranches f0 = 30° 19.3 (14.6)b 6.3 (1.2) 31.3 (19.9) 29.1 (17.2) 28.8 16.5
Columbia River anabranches f0 = 78° 20.3 20.7)c 5.9 (5.8) 35.5 (18.4) 30.1 (9.7) 19.7 19.6
Danube Delta distributaries f0 = 83° 36.0 (33.5)c 1.0 (18.3) 56.7 (26.9) 49.0 (15.4) 31.8 23.4
Laitaure Delta f0 = 83° 36.2 33.0)c 0.7 (17.7) 37.9 (18.9) 21.3 (4.8) 36.9 24.9
Salmon River Area
group 1 C0 = 1.9 kPa 22.3 0.9 23.4 8.0 15.8 17.3
group 2 C0 = 3.0 kPa 12.9 1. 11.2 0.8 13.8 8.7
Mtn Str. Coloradod
Thin vegetation C0 = 1.7 kPa 15.7 6.3 10.8 0.3 12.5 11.5
Thick vegetation C0 = 1.9 kPa 14.8 6.4 11.4 0.0 8.2 11.4
a
Values in brackets for the Colorado River are based on using the raw field data for S, n, and D50.
b
Values in parentheses for the Fraser River are based on a data set excluding the two smallest channels.
c
Values in parentheses are based on data sets excluding the smallest channels (for Columbia and Laitaure, Q < 20 m3/s; for Danube, Q < 200 m3/s).
d
Reported statistics are based on the 5-year floods.

material sediment declines downstream, which is consistent The regime model reproduces the observed width scaling
with the speculation by Pitlick and Cress [2002]. Since the fairly well (Figure 4) when the relative bank strength is set
transport capacity predicted by the model depends on the equal to unity (f0 = 30°), but does a poorer job of predicting
channel width, depth and surface texture, which the regime the channel depths, particularly for the smaller anabranches
model faithfully reproduces, the spatial variation in bed in the system. The model predictions are adequately
material transport capacity does not depend on the model described by power functions. The power laws fitted to
assumptions at all, but only on the quality of the initial data the regime predictions are
and the performance of the Parker formula as an index of
bed material transport rates.
W ¼ 2:60Q0:57 ; R2 ¼ 0:93; Std: Err: ðbÞ ¼ 0:0054 ð8Þ
3.2.2. Fraser River Anabranches
[30] The empirical hydraulic geometry equations for these
anabranches are not typical either, with an increase in d ¼ 1:45Q0:18 ; R2 ¼ 0:66; Std: Err: ð f Þ ¼ 0:0044 ð9Þ
bankfull discharge being accommodated by a dispropor-
tionately large increase in width. The equations reported by
Ellis and Church [2005] are which are reasonably close to the empirical relations of Ellis
and Church [2005]. The RMS error for width and depth
W ¼ 2:60Q0:60 ; R2 ¼ 0:81; Std: Err: ðbÞ ¼ 0:106 ð6Þ based on the regime model are similar to those associated
with the empirical fits (Table 1): the regime model appears
to do a better job of predicting the channel widths, while the
d ¼ 0:76Q0:25 ; R2 ¼ 0:76; Std: Err: ð f Þ ¼ 0:058 ð7Þ empirical equations are superior for predicting the observed
depths.

Figure 4. Plots of the regime analysis of Ellis and Church [2005]. (left) Bed material sediment
concentration plotted against discharge. (middle) Widths and depths predicted by the model plotted
against the widths and depths measured in the field. (right) Predicted (circles) and observed (pluses)
channel widths and depths plotted against Q. Since D50 and S covary and are not strongly related to Q, it
is necessary to use the raw field data, which precludes the definition of smoothly varying regime
predictions.

6 of 18
F03025 EATON AND CHURCH: HYDRAULIC GEOMETRY F03025

Figure 5. Plots of the regime analysis of Tabata and Hickin [2003]. (left) Bed material sediment
concentration plotted against discharge. (right) Observed channel widths and depths (circles), as well as
the widths and depths predicted by the regime model (solid lines), plotted against Q. The shaded areas
represent alternate states associated with weaker banks, the dashed line indicating the limit value f0 = 68°.

[31] The relation between bed material transport concen- [34] Power law relations fit the modeled widths quite
tration and Q for the Fraser River anabranches is the reverse well, but not the modeled depths (see Figure 5). The
of the pattern evident in the Colorado River (Figure 4). equations are
Since the anabranch channels do not reflect downstream
trends, the increase in bed material concentration with
W ¼ 2:13Q0:74 ; R2 ¼ 0:99; Std: Err: ðbÞ ¼ 0:0071 ð12Þ
increasing discharge implies that the smaller channels are
less efficient at transporting bed material than are the larger
ones.
d ¼ 1:89Q0:076 ; R2 ¼ 0:85; Std: Err: ð f Þ ¼ 0:015 ð13Þ
3.2.3. Columbia River Anabranches
[32] The anabranches of the upper Columbia River ex-
hibit an empirical hydraulic geometry rather similar to that These power fits are systematically different from the
of the gravel bed anabranches in the Fraser River. The empirical fits presented in equations (10) and (11). The
empirical equations are differences are primarily attributable to the poor model
performance for the smaller channels.
3.2.4. Danube Delta
W ¼ 3:24Q0:64 ; R2 ¼ 0:84; Std: Err: ðbÞ ¼ 0:075 ð10Þ
[35] The empirical hydraulic geometry equations for the
delta channels investigated by Mikhailov [1970] are similar
to the equations typically observed for downstream hydrau-
d ¼ 1:04Q0:19 ; R2 ¼ 0:30; Std: Err: ð f Þ ¼ 0:077 ð11Þ lic geometry:
They are fairly well described by our regime model when
the banks are nearly 10 times stronger than the bed (f0 = W ¼ 5:96Q0:48 ; R2 ¼ 0:89; Std: Err: ðbÞ ¼ 0:029 ð14Þ
78°, see Figure 5). The observed widths and depths are
predicted about as well as for the Fraser River (see Table 1):
again, the empirical equations for depth outperform the d ¼ 0:56Q0:35 ; R2 ¼ 0:86; Std: Err: ð f Þ ¼ 0:024 ð15Þ
regime model. The smallest anabranches seem to be
consistently wider and shallower than predicted, just as The modeling results are shown on Figure 6. The regime
they were in the Fraser River system. model assumes relatively strong banks (f0 = 83°), and
[33] The relation between bed material transport concen- produces results similar to the analysis of the Columbia
tration (Qbm/Q) and Q is also similar to that of the Fraser River anabranches. The bed material transport concentration
River (Figure 5), with progressive division of the channel declines as channels are progressively subdivided, and most
into smaller and smaller threads being associated with a (but not all) of the smallest channels are, again, wider and
reduction of bed material transport concentration, albeit less shallower than predicted by the regime model. While the
rapidly than for the Fraser River. Since both locations are regime model predicts widths about as well as the empirical
aggradational settings, the positive correlation between power fits, it predicts depths less well than the empirical
transport efficiency and channel size suggests that the equations (Table 1). The RMS errors associated with the
processes producing the channel divisions are closely linked empirical fits are also rather high compared to the previous
to the occurrence of bed material deposition in these examples.
environments.

7 of 18
F03025 EATON AND CHURCH: HYDRAULIC GEOMETRY F03025

Figure 6. Plots of the regime analyses of Mikhailov [1970] and Andrén [1994] using the same
conventions as in Figure 5. The limit alternate states indicated by the dashed lines are associated with f0 = 63°.

[36] The hydraulic geometry equations fit to the channel [38] The data for the Laitaure Delta were modeled using
dimensions predicted by the regime model for the Danube the same bank strength value as was used for the Danube
Delta are delta channels. The power law equations based on the
modeled widths and depths are
W ¼ 1:64Q0:65 ; R2 ¼ 0:99; Std: Err: ðbÞ ¼ 0:0031 ð16Þ
W ¼ 3:26Q0:66 ; R2 ¼ 0:99; Std: Err: ðbÞ ¼ 0:0034 ð20Þ

d ¼ 2:23Q0:19 ; R2 ¼ 0:98; Std: Err: ð f Þ ¼ 0:0041 ð17Þ


d ¼ 0:91Q0:16 ; R2 ¼ 0:99; Std: Err: ð f Þ ¼ 0:0068 ð21Þ
Like the Columbia River results, the equations associated
with the regime model are quite different from those based Once again, the regime equations based on the predicted
on the empirical fit to the data, primarily because of poor geometry are not similar to the empirical equations based on
model performance for at least some of the smaller the reported data, primarily because of deviations from the
distributary channels. model at the lower discharges, resulting in large RMS errors
3.2.5. Laitaure Delta for the modeled depths (Table 1).
[37] The empirical equations for the Laitaure Delta, based
on Andrén’s [1994] data are 4. Systems With Spatially Varying Relative Bank
Strength
W ¼ 10:0Q0:39 ; R2 ¼ 0:53; Std: Err: ðbÞ ¼ 0:13 ð18Þ
4.1. Methods
[39] The above analyses targeted environments in which
d ¼ 0:36Q0:385 ; R2 ¼ 0:69; Std: Err: ð f Þ ¼ 0:090 ð19Þ
relative bank strength appears to be at least approximately
constant. In most drainage basins, this is clearly not the
case. However, by applying the second formulation of the
While these equations are different from the expected
regime model, in which bank strength is divided into a
empirical hydraulic geometry, they are also quite poorly
component due to the influence of vegetation (C0) and one
known, having the largest uncertainties for the value of the
due to the true frictional strength (f), we can make some
exponents, and the largest RMS errors (Table 1).

8 of 18
F03025 EATON AND CHURCH: HYDRAULIC GEOMETRY F03025

http://earth.google.com/) reveals that the floodplains have


a nearly continuous cover of grass, shrubs, and occasional
deciduous trees. The nature and extent of the riparian
vegetative cover combined with the range of channel sizes
suggest that, in the Salmon River area, the effects of stream
bank vegetation dominate bank strength in the upper rea-
ches, but have a much reduced effect along the larger stream
channels.
[41] Emmett’s [1975] data were used to investigate the
variation in relative bank strength in the downstream
direction that is required to produce the observed empirical
hydraulic geometry using a rational regime model. Since
bank strength is one of the key variables affecting the
channel dimensions predicted by our regime model, we
selected only those stream channels that, according to the
map of geology and surficial deposits provided by Emmett
[1975], formed in modern alluvium or unconsolidated
glaciofluvial deposits. Another important factor affecting
Figure 7. Empirical hydraulic geometry for Salmon River the regime model predictions is the coarse sediment sup-
data [Emmett, 1975] and for two subsets of the data, defined plied to the stream channel. No data exist to allow us to
on the basis of boundary conditions and lithology of the predict which areas have a high bed material sediment yield
contributing area. and which have low sediment yield, so the selected data
were further classified on the basis of the dominant litho-
logies in their drainage basins. The result is to divide
progress. This approach is limited to gravel bed streams
Emmett’s data set into two groups (Figure 7). The first
where it is reasonable to assume that the banks comprise a
group contains 13 stations along the main stem of Salmon
lower, cohesionless gravel toe overlain by a vegetation-
River, which is dominated by granitic rock from the Idaho
reinforced vertical upper bank, characterized by the effec-
Batholith, as well as rock units mapped as undifferentiated
tive cohesion due to vegetation, C0. We have analyzed in
Paleozoic rock [Emmett, 1975]. The second group contains
some detail two data sets on mountain streams in Idaho and
7 stations in the drainage basin of the East Fork Salmon
in Colorado that exhibit fairly typical hydraulic geometry
River, which is dominated by rocks from the Challis
relations.
Volcanics.
4.1.1. Salmon River
[42] In order to apply the regime model to the two data
[40] The first data set, from the Salmon River area of
sets, we used the reported values of Q, S, D50 and D84 for
Idaho (Figure 7), was published by Emmett [1975]. Even
each group to generate functions relating S, n, D50 and D84 to
though the Salmon River is located in a semiarid environ-
Q, as described above for Colorado River (see Figure 8 and
ment, an inspection of satellite imagery (Google Earth,
Table 2). These functions were then used in the regime

Figure 8. Trends in S, n, D50, and D84 with the reported formative discharge for the two groups of
streams from the Salmon River area data set [Emmett, 1975]. The equations of the lines are provided in
Table 2. These fitted lines were used to parameterize the regime model (see Figure 10).

9 of 18
F03025 EATON AND CHURCH: HYDRAULIC GEOMETRY F03025

Table 2. Data-Based Trends and Correlation Coefficients Used to reported by Andrews were adjusted upward to reflect the
Reconstruct the Hydraulic Geometry for Two Groups of Streams, fact that we are invoking larger formative discharges (see
Salmon River Areaa Appendix A for details). Since neither the geometry nor,
Variable Group 1 Group 2 consequently, the mean velocity for the 5-year or 20-year
Slope, m/m 0.054 Q0.66, R2 = 0.69 0.071 Q0.55, R2 = 0.83 events are accurately known, we have applied a constant
Manning’s n 0.04 (assumed constant) 0.12 Q0.37, 0.79 Manning’s n of 0.03 to all of the channels.
D50, mm 19 Q0.19, 0.19 16 Q0.31, 0.67 4.1.3. Inversion of the Relative Bank Strength Model
D84, mm 56 Q0.081, 0.09 43 Q0.25, 0.88
a
[48] For both of the foregoing data sets, we have esti-
Data are from Emmett [1975]. mated the relative bank strength, f0, associated with the
observed channel geometry by inverting the regime model.
To do this, we applied the regime model using the reported
model to estimate S, n, D50 and D84 for various specified data for each study reach. In the computer program, we
values of Q. varied the parameter f0 for each study reach until the
[43] The estimates of the surface D50 were used in the predicted channel width matched the observed width. The
Parker [1990] equation to calculate the bed material sedi- purpose of this analysis is to link the patterns in sediment
ment transport capacity, Qbm, and the estimates of the transport concentration evident in the C0 regime model to
surface D84 were used in the bank stability analysis on the the more easily interpretable changes in relative bank
grounds that bank stability is most likely to be controlled by strength, f0.
the larger grains found in the bed and therefore in the lower
4.2. Results
part of the channel bank.
[44] According to the analysis presented by Eaton [2006] 4.2.1. Salmon River
the riparian vegetation type in the Salmon River area is [49] Salmon River exhibits fairly typical empirical hydrau-
consistent with C0 between 1.5 and 3.8 kPa, which corre- lic geometry (Figure 7). Power law fits for group 1 and
sponds to Hey and Thorne’s [1986] type II and III channels. group 2, respectively, are
We adjusted C0 for each group to produce the best agree-
ment between the modeled channel geometry and the W ¼ 3:30Q0:53 ; R2 ¼ 0:87; Std: Err: ðbÞ ¼ 0:0612 ð22Þ
observed channel dimensions.
4.1.2. Mountain Streams in Colorado
[45] The second data set (Figure 9), on gravel bed rivers d ¼ 0:20Q0:42 ; R2 ¼ 0:78; Std: Err: ð f Þ ¼ 0:066 ð23Þ
in Colorado, was published by Andrews [1984]. He classi-
fied channels into ones with ‘‘thin’’ and ones with ‘‘thick’’
riparian vegetation, which should roughly correlate with the W ¼ 4:36Q0:32 ; R2 ¼ 0:77; Std: Err: ðbÞ ¼ 0:077 ð24Þ
relative erodibility of the stream banks. In this case, the data
come from streams in different environments with different
bank vegetation, so we cannot generate reasonable empir- d ¼ 0:25Q0:36 ; R2 ¼ 0:92; Std: Err: ð f Þ ¼ 0:048 ð25Þ
ical relations between discharge and the other required input
data (S, n and grain size). We therefore applied the regime The exponents of the width equations in the two groups of
model to each study reach and evaluated model perfor- data (equation (22) and (24)) are significantly different at
mance using 1:1 plots of the expected and observed channel
dimensions. We used the D90 reported by Andrews for the
bank stability analysis and the reported values of D50 to
calculate an index of the sediment transport capacity via the
Parker [1990] equation.
[46] The initial analysis using the reported bankfull flows
suggested that, for many of Andrews’s stream channels, the
bed was actually more resistant to erosion than the banks,
which is at odds with previous findings [Eaton and Millar,
2004; Eaton, 2006]. A comparison of the Andrews hydrau-
lic geometry equations against empirical data collected from
similarly vegetated gravel bed channels suggests that the
formative discharge is probably higher than the reported
bankfull flows (see Appendix A). This conclusion is sup-
ported by the fact that, for the majority of the streams in
the data set, the reported bankfull discharge is less than the
mean annual instantaneous peak flow reported by the
USGS, on average about 25% less. Some researchers have
claimed that, in semiarid regions like Colorado, the forma-
tive discharge may occur less frequently than in humid
environments; in truly arid regions, possibly as rarely as Figure 9. Hydraulic geometry data from Andrews [1984].
once in 200 years [Baker, 1977; Friedman et al., 1996]. The data are stratified by riparian vegetation classification.
[47] In our analysis, we use the 5-year and 20-year floods Emmett’s [1975] data from Salmon River are shown for
to parameterize the regime model: the bankfull depths comparison.

10 of 18
F03025 EATON AND CHURCH: HYDRAULIC GEOMETRY F03025

Figure 10. Results of regime modeling for groups 1 and 2, Salmon River, using the effective cohesion
bank stability analysis described by Eaton [2006]: (left) bed material concentration plotted against
discharge and (right) observed widths and depths (plotted as circles), overlain by the widths and depths
predicted by the regime model (dashed lines).

a = 0.002 (i.e., the 99.8% confidence level), and the 4.2.2. Mountain Streams in Colorado
coefficients are significantly different at a = 0.17. The [51] For Andrews’ [1984] streams with thin riparian
exponents in the depth equations (equations (23) and (25)) vegetation, we obtained the following empirical equations
are significantly different at a = 0.14, while the coefficients for width and depth as functions of the originally reported
are different at a = 0.20. While the confidence levels for bankfull flows:
these differences are generally modest at best, there appear
to be systematic differences between the two groups. These W ¼ 4:17Q0:50 ; R2 ¼ 0:95; Std: Err: ðbÞ ¼ 0:032 ð26Þ
differences may be due to differences in the yield of bed
material sediment from the drainage basin, to differences in
the typical bed material grain shape and surface arrange- d ¼ 0:26Q0:36 ; R2 ¼ 0:89; Std: Err: ð f Þ ¼ 0:037 ð27Þ
ment or, possibly to systematic differences in the bank
conditions. For thick vegetation, the empirical equations are
[50] The results of the regime modeling are shown in
Figure 10. The effective cohesion value that produced the W ¼ 3:38Q0:54 ; R2 ¼ 0:96; Std: Err: ðbÞ ¼ 0:041 ð28Þ
best fit to the data for group 1 was 1.9 kPa, and for group 2
it was 3.0 kPa. Both values are consistent with the range
exhibited in the Hey and Thorne [1986] data set when
d ¼ 0:25Q0:33 ; R2 ¼ 0:85; Std: Err: ð f Þ ¼ 0:050 ð29Þ
analyzed by Eaton [2006]. As in the Colorado River data set
analyzed above, the bed material transport concentration
declines downstream in the Salmon River. The predicted The results of the regime model analysis assuming a
channel dimensions are not related to the formative dis- formative discharge with a return period of between 5 years
charge by a simple power function. The analysis produced and 20 years are shown in Figure 11. The downstream
trends that are clearly not power laws, but the modeled variations in width and depth are successfully reproduced
trends are about as consistent with the data as are empirical by the model assuming, for the 5-year flood, that C0 equals
power fits, particularly for group 2 (Table 1). 1.7 kPa and 1.9 kPa for the ‘‘thin vegetation’’ and ‘‘thick
vegetation’’ channels, respectively. Substantially similar

11 of 18
F03025 EATON AND CHURCH: HYDRAULIC GEOMETRY F03025

Figure 11. Results of regime modeling for Andrews [1984] channels with ‘‘thin’’ and ‘‘thick’’ bank
vegetation using the effective cohesion bank stability analysis described by Eaton [2006]: (top) results
using the 5-year flood data and (bottom) results using the 20-year flood. (left) Bed material sediment
concentration plotted against discharge (with a dashed line indicating constant Qbm) and (middle and
right) one-to-one plots of the predicted versus observed widths and depths, respectively.

results were obtained for the 20-year flood by assuming C0 cannot be weaker since this would imply that they would
values of 2.2 kPa and 2.6 kPa for the thin and thick data be unstable at the identified bankfull discharge.
sets. The pattern of declining bed material concentration is
also evident in both of these analyses. The RMS errors for 5. Discussion
the fitted regime models using the 5-year flood are similar
to (albeit slightly greater than) the RMS errors for the [55] Physically based regime models reasonably fit down-
empirical equations using the originally reported bankfull stream hydraulic geometry data in a range of environments,
flows (Table 1). reproducing the typical relations in which widths vary
approximately with Q0.5 and depths vary with Q0.4 [Emmett,
4.3. Inversion of the Relative Bank Strength Model 1975; Andrews, 1984] as well as relations in which there is
[52] Figure 12 displays the results of the analysis of an unusually rapid increase in width [Tabata and Hickin,
Emmett’s [1975] data set using the regime model with 2003; Ellis and Church, 2005] or depth [Pitlick and Cress,
relative bank strength (f0). The observed channel widths 2002] with discharge. The goodness of fit of the predictions
at each observation point were fit by varying f0. The is measured using the root mean squared error (RMSE) for
estimated relative bank strength declines systematically the calculated widths and depths, as well as a mean bias (for
downstream for both groups. The uppermost reaches have the regime model only). Table 1 summarizes the model
relatively erosion-resistant banks that are about 5 to 7 times performance for all of the data sets examined. Generally, the
stronger than the channel bed, while the largest streams RMSE for width estimates using the regime model are about
have banks that are between 1.5 and 2 times stronger than the same as (and sometimes smaller than) the RMSE for the
the bed (see Figure 1). empirical equations, and the bias in the width estimates is
[53] Figure 13 presents the same analysis applied to the low (i.e., <10%). The same is true for the RMSE for
Andrews [1984] data set: the choice of formative discharge modeled depths, except in the anabranch and delta channels.
has little effect on the general trends predicted by the model. For those data sets, the empirical equations perform sub-
The bed material transport concentration again declines as stantially better and the modeled depths exhibit a significant
formative discharge increases, as does the relative bank mean bias that is about the same magnitude as the RMSE:
strength. The range of relative bank strengths is slightly this is attributable to poor model performance for the
lower than in Figure 12, with the thick vegetation channels smallest anabranch channels (see Figures 4, 5, and 6).
being associated with higher relative bank strengths than the When the regime model is tested against only the larger
thin vegetation channels. channels, the RMS errors for the modeled depths are similar
[54] It is worth noting that the regime model selects the to those for the empirical equations, and the mean bias
narrowest possible stream channel that would be stable values are significantly reduced (Table 1).
given the governing conditions of Q, S, and D84 so that, [56] Overall, the regime models summarized in Table 1
while it is possible that the banks are stronger than indicated perform about as well as empirical equations (particularly
by the apparent friction angle shown in Figure 1, they for the single-thread gravel bed channels), since the models
generally explain about as much variance as do the empir-

12 of 18
F03025 EATON AND CHURCH: HYDRAULIC GEOMETRY F03025

Figure 12. Results of regime modeling for groups 1 and 2, Salmon River [Emmett, 1975]. The circles
on all of the plots represent the regime model fit to reproduce the observed channel width at each study
reach. (left) Bed material transport concentration (Qbm/Q); dashed lines represent a line of constant Qbm.
(right) Relative bank strength (f0) required in the regime model to match the observed bankfull width for
each station; the line at which bed and bank erodibility becomes equal is indicated by dashed lines.

ical equations. However, since it might be expected that the a wider range of flows and bypassing storage as bed
empirical equations would outperform the physically based material. Continuity of mass in a nonaggrading system
theory, because they are based on finding a relation that implies that a downstream decline in bed material concen-
minimizes the RMSE of the observations while regime tration must be accompanied by a downstream increase in
models yield predictive relations based on the physical either the suspended sediment concentration or the dis-
processes supposed to underlie the observed patterns, the solved material concentration. Sklar et al. [2006] suggest
regime model performance is judged to be quite good. Since that, beyond some critical length scale set by the relative
the regime models are physically based, they are transfer- resistance of the basin lithology, bed load transport rates
able from one basin to another, whereas the empirical fits cease to scale with drainage area and become effectively
are not. Furthermore, the invocation of a physically based constant, a behavior that is reproduced by our model. In
model leads to further insights about the structure and contrast, this sort of downstream change in Qbm/Q is not
function of stream channel networks, and is a suitable tool evident in our data sets from anabranching and delta
for predicting their response to environmental change. systems, which are not, in fact, instances of downstream
[57] The indicated decline in bed material sediment hydraulic geometry at all.
concentration with formative discharge in the single-thread [58] Anabranch and delta systems exhibit scale relations
channel systems examined herein (all of our cases of true in the true sense: they represent systematic variation in
downstream hydraulic geometry) suggests a common scale channel geometry with changing channel size, while gradi-
effect. Headwaters, in which Qbm/Q predicted by our model ent and channel boundary materials remain relatively sim-
is high, appear to transport a relatively large proportion of ilar within each data set. The regime model consistently
the total sediment supplied to the channel as bed load (we predicts that bed material concentration increases with
draw a rough parallel between bed material and bed load discharge: the relation between Qbm/Q and Q is interpreted
transport for the purposes of this discussion). As sediment is to be evidence that, when single thread channels divide into
transported farther from its original source, it weathers and multiple threads, they become less efficient and can trans-
wears into smaller and smaller sizes. In addition to direct port less sediment. This observation is consistent with a
abrasion during transport [Pizzuto, 1995; Sklar et al., 2006] well-known principle of hydraulic engineering that, to
floodplains, where sediment is stored in a wet environment, remain stable, small distributary channels must exclude
may be particularly effective at weathering bed material bed material transport; that is, they must carry a much
sediment into finer materials [Jones and Humphrey, 1997]. reduced sediment load. Amongst the four data sets, there is
If the grain size is reduced sufficiently, the material is a remarkable inverse correlation between the exponents for
transported in suspension, moving longer distances and in the equations of width and depth (R2 = 0.93) which reveals

13 of 18
F03025 EATON AND CHURCH: HYDRAULIC GEOMETRY F03025

Figure 13. Plots of the regime analysis of the Andrews [1984] data: (top) analysis using the 5-year flood
events and (bottom) 20-year analysis. (left) Bed material transport concentration (Qbm/Q). Dashed lines
represent constant Qbm. (right) Relative bank strength (f0) required in the regime model to match the
observed bankfull width for each station; the line at which bed and bank erodibility becomes equal is
indicated using dashed lines.

that, together, width and depth accounts for a about 82% of Makaske et al. [2002]. They reported that channels on the
the variation in channel capacity in every case. The covari- Columbia River floodplain initially form on top of crevasse
ation is presumably introduced by the degree to which splay deposits (stage 1) and are bounded primarily by sand.
channels flowing in similar materials must adjust their As a result, they tend to be fairly wide and shallow, and they
sediment transporting capacity to remain in equilibrium. would be poorly described by a regime model that assumes
[59] This result is contrary to the speculation made by the existence of strong, cohesive channel banks. As the
other researchers that channel division increases the bed channels abstract more water and mature, they deepen by
material sediment concentration and that division in ana- eroding the crevasse splay sands and by depositing cohesive
branch systems occurs in order to accommodate an increase sediments along their banks. Once they attain depths on the
the sediment supply [Nanson and Knighton, 1996; Jansen order of 2 m, they typically become capable of eroding
and Nanson, 2004]. Obviously, our choice of bed material laterally and maintaining a dynamic equilibrium (stage 2):
transport law has influenced the model predictions so that these channels are likely in regime, and may be well
the validity of the underlying transport laws, and the way in described over a range of formative discharges by a regime
which we have used them in the model, must be considered model assuming a constant, high relative bank strength. At
when commenting on larger geomorphic issues such as some point, often because of the development of log jams,
channel pattern and sediment transport efficiency. the flow into the anabranch is reduced (stage 3), and the
[60] The regime model performs poorly for the smallest channels undergo either rapid vertical aggradation, produc-
anabranch and distributary channels across all four data sets. ing wide, shallow channels similar to stage 1, or a more
It overpredicts the depths of the smallest channels, and gradual combination of vertical and lateral aggradation,
therefore overpredicts the transport concentration, indicat- producing narrow, deeper channels that may or may not
ing that the effect of channel division on the transport be in regime with the fluid and sediment fluxes supplied to
capacity in the smallest anabranches is in fact greater than them.
predicted by the model. That this result is common to all [62] Therefore it appears that the larger channels (in the
four data sets suggests that either the assumptions about the Columbia system at least) may be in regime with respect to
parameters used to specify the model are systematically the fully developed cohesive banks (represented by f0 =
wrong, or that the fundamental premise of the model (i.e., 78°), while the smaller channels may be in regime with
that the channels are in regime) is inaccurate. respect to a range of lower relative bank strengths. These
[61] A reasonable explanation for poor model perfor- channels are covered in Figure 5 by a shaded region
mance is suggested by a sedimentological reconstruction bounded by an upper (f0 = 78°) and lower (f0 = 68°) bank
of the Columbia River anabranch dynamics published by strength value. It is possible that a similar sequence of

14 of 18
F03025 EATON AND CHURCH: HYDRAULIC GEOMETRY F03025

channel initiation and development is responsible for the sediment supply [Dietrich et al., 1989]. However, they also
range of channel shapes characteristic of the smallest claim that t* remains essentially constant, which would
distributary channels on both the Danube and Laitaure seem to be at odds with declining transport capacity. The
deltas as well (see Figure 6). Other small channels may regime model supports the conjecture that bed material
be rapidly shrinking as they aggrade, ultimately forming transport rates are declining, since Qbm/Q declines by
part of the floodplain: these channels are not likely to be at almost two orders of magnitude as one moves down the
equilibrium with respect to the supply of sediment, and system, and also suggests that, as the ratio W/d decreases
therefore should not be well described by a regime model. downstream, t* also decreases. The decrease is relatively
Nonequilibrium behavior is likely at least part of the small (from just 0.046 to 0.039), and remains close to the t*
explanation of the poor model performance, and seems value of 0.049 reported by Pitlick and Cress [2002]. It
to be the most likely explanation for the Fraser River should be noted that the scatter in estimates of t* of Pitlick
anabranches. and Cress is much greater than the range of variation in the
[63] While the first five cases analyzed in the paper model, implying that physically reasonable relations
represent carefully chosen situations for which we could between the parameters W/d and t* may be obscured by
make the assumption that the relative banks strength is limited measurement resolution.
approximately uniform, one cannot apply such assumptions
to most downstream hydraulic geometry sets. In order to 6. Conclusions
fully explain the variations in channel geometry expressed
in the data sets presented by Emmett [1975] and Andrews [67] We draw the following conclusions from our analy-
[1984], we would require data that are seldom collected. ses of downstream hydraulic geometry using rational (phys-
However, we can reasonably speculate on the physical ically based) regime models.
processes that underlie these variations. [68] 1. Fully elaborated regime models of alluvial channel
[64] The ability of the modified regime model described form are not simple scaling relations, as has been widely
by Eaton [2006] to replicate the general downstream supposed, because of the variable influence of several
hydraulic geometry of both the Emmett [1975] data set governing factors. However, power law scaling relations
(see Figure 10) and the Andrews [1984] data set (see furnish very good approximations of the underlying rela-
Figure 11) by adopting reasonable values of vegetation- tions in many cases when one or more of the governing
related bank strength, C0, suggests that the declines in relations remains essentially constant.
relative bank strength and bed material concentration with [69] 2. Systematic variations in relative bank strength and
increasing Q are driven primarily by vegetation-related bank in bed material transport capacity are subsumed within the
strength. The effect of vegetation is limited to the zone typically observed empirical hydraulic geometry. By fitting
influenced by the tensile strength of the root systems. For a physically based model to the data of hydraulic geometry,
small streams, the rooting depth may be similar to the we approach the underlying physical structure of the stream
average channel depth, and the banks may be many times network organization (with respect to grain size variation,
more erosion-resistant than the bed. However, as the chan- transport capacity, boundary resistance and channel gradi-
nel depth increases (holding the rooting depth constant), the ent), as well as the scaling of channel dimensions with
relative importance of the vegetation declines and the formative discharge. That is, we obtain a glimpse of how
relative erodibility of the banks approaches that of the bed stream networks function.
[see Eaton and Millar, 2004]. This scale effect is likely to [70] 3. In stream networks where riparian vegetation
be ubiquitous in channels where vegetation dominates the influences bank strength, relative bank strength declines
bank stability, and the results of our analyses are consistent downstream with increasing discharge. This scale effect is
with these plausible conjectures (see Figures 12 and 13). consistent with the notion that the contribution of riparian
Variations in vegetation type and riparian land use along vegetation to bank strength declines as the channel depth
stream systems, as appears to be the case for the streams in becomes large relative to the rooting depth of the riparian
Colorado [Andrews, 1984], are likely to further modify the species. The hydraulic geometry of stream networks can be
scale effect. modeled by assuming an appropriate (and constant) value
[65] The analysis of the Andrews [1984] data set also for the absolute bank strength, represented as an effective
indicates that the channels with the highest potential bed cohesion, C0.
material transport concentration tend to be those with thick [71] 4. The bed material sediment concentration in stream
riparian vegetation. These channels, probably laterally sta- networks generally appears to decline with discharge in
ble, are deeper than channels with less dense riparian conjunction with relative bank strength which, in nonaggrad-
vegetation. The channels with less dense vegetation and ing systems, is presumably counterbalanced by an increase in
weaker channel banks are more likely to be laterally the suspended sediment (and/or dissolved material) concen-
unstable, but they are less efficient in conveying bed tration in order to maintain continuity of mass.
material downstream. [72] 5. For large stream channels, like the Fraser River
[66] Our regime model also provides insight into the and the Colorado River, the channel geometry can be
downstream variations in the Shields number (t*) along reasonably modeled by assuming a uniform relative bank
the reach of the Colorado River studied by Pitlick and Cress strength, f0. The atypical hydraulic geometry relations
[2002]. The authors explain the unusual hydraulic geometry reported by Pitlick and Cress [2002] and Ellis and Church
along this reach as the result of a decrease in Qbm which, [2005] are reproduced using such models.
they point out, may have produced a downstream increase [73] 6. The scaling relations for anabranched channels
in the armor ratio, since the armor ratio is related to and delta distributary channels, where true friction and

15 of 18
F03025 EATON AND CHURCH: HYDRAULIC GEOMETRY F03025

Figure A1. Application of the Andrews [1984] hydraulic geometry equations to the data set of Hey and
Thorne [1986]. (left) Results using the surface grain size and bankfull discharge reported by Hey and
Thorne, applying the equation for thin vegetation to Hey and Thorne’s type I and type II channels and the
equations for thick vegetation to types III and IV. (right) Results when the discharge values are all scaled
back, as indicated.

cohesion dominate bank strength, are reasonably repro- requirements appear to be reasonable and necessary for
duced using regime models that assume a uniform relative the construction of a physically based theory to explain
bank strength, f0. For these channels, bed material concen- the observed hydraulic geometry of rivers. In addition to
tration increases with formative discharge, indicating that providing insight into hydraulic geometry relations, we
division of a channel into multiple threads is associated believe that regime models can usefully be incorporated
with reduced sediment transport efficiency and likely with into numerical models of drainage network dynamics and
aggradation, as well. even landscape development. They represent a computa-
[74] 7. The smallest channels examined in the sand bed tionally effective method for relating channel morphology
anabranch and distributary channels in this study conform to to governing conditions and, insofar as they escape unqual-
a regime model that assumes a lower bank strength. In the ified empiricism, they are superior to the empirical hydrau-
case of the Columbia River, there is a sedimentological lic geometry equations commonly used in drainage network
basis for believing that this variation is due to real changes models.
in the bank strength. The same argument may apply to the
Laitaure and Danube deltas, though there are no published Appendix A
reports to support this assertion. Some of the small channels
are probably shrinking as they aggrade: they would not [77] The initial analysis of the Andrews [1984] data set
therefore be in equilibrium with their governing conditions using a relative bank strength formulation indicated that
and should not be expected to agree with the model many of the channels identified as having thin riparian
predictions. vegetation and one channel with thick vegetation have
[75] Physically based models relating channel geometry banks that are more erodible than the bed. This is surprising,
to the governing conditions of discharge, sediment supply, since previous analyses using the regime model [Eaton and
grain size and slope can be constructed in a wide variety of Millar, 2004; Eaton, 2006] indicate that channel banks with
environments using appropriately configured regime models, riparian vegetation similar to that of the Colorado streams
but only if the relative erodibility of the channel boundary is still remain about 50% stronger than the channel bed and,
adequately represented. A key consideration is the choice of on average, exhibit apparent friction angles of 40°. Only a
bank stability analysis method and the choice of bank handful of the smaller channels with thick vegetation and a
stability parameter values. Direct knowledge of the bed single channel with thin vegetation exhibit widths consistent
material discharge would seem to be of secondary impor- with even the modest bank strength index of 40°. This
tance, since the reach-averaged slope and characteristic unexpected result led us to question whether or not the
grain size are set in response to the long-term average bed reported bankfull discharges actually correspond to the
material discharge, and since channels have the capacity to formative flow responsible for the observed widths and
respond to small year-to-year variations in sediment dis- depths. (A previous analysis of these data by Millar [2005]
charge without modifying the gross channel morphology did not consider this possibility.)
simply by varying the bed surface texture and structure. [78] Andrews estimated bankfull flows by projecting
[76] While application of a regime model requires knowl- surveyed indicators of the bankfull elevation along the
edge of more than just the formative discharge, the addi- stream channel parallel to the water surface slope to a
tional parameters are unarguably important in determining nearby gauge section, where the recorded stage-discharge
the channel morphology, so the increased information relation was used to estimate the bankfull Q. Since bankfull

16 of 18
F03025 EATON AND CHURCH: HYDRAULIC GEOMETRY F03025

Table A1. Discharge Data for 5-Year and 20-Year Floods in the likely formed by flows less frequent than those identified as
Andrews [1984] Gravel Bed Channels bankfull flows in the original data set.
Discharge Data, m3/s Channel Geometry, m [81] On the basis of the existing data for each USGS
USGS Mean gauging station, we estimated the peak flows with return
Station Qbf Annual 5 Year 20 Year W d(Qbf) d(Q5) d(Q20) periods of 5 and 20 years (Table A1) and used these values
6620000 85.2 86.8 119 166 47.2 0.97 1.22 1.63 in the regime model. Streams for which the period of record
7083000 7.08 7.85 9.8 15.5 9.1 0.52 0.55 0.75 was not sufficient to accurately estimate the 20-year peak
7091000 9.8 18.7 26.4 37.6 11.5 0.46 0.75 0.87 flow have been excluded from the analysis. The 5-year
9022000 2.69 3.89 5.04 7.83 7.0 0.29 0.38 0.58
9035900 8.36 8.49 10.8 13.4 9.2 0.43 0.47 0.49
flood frequency best corresponds to the empirical result
9036000 22.6 26.2 33.8 42.0 17.5 0.73 0.93 0.95 from the test using the Hey and Thorne [1986] data set:
9074800 4.45 10.2 12.6 15.9 8.4 0.49 0.64 0.74 these data were used in regime models. Since there is a large
9078100 3.17 6.59 8.51 11.7 6.3 0.39 0.73 0.77 degree of uncertainty in this post hoc analysis, we have also
9078200 2.52 4.12 5.52 8.03 5.6 0.41 0.53 0.62
9081600 49.0 64.3 81.8 103 34.1 0.84 1.02 1.30
run the regime analysis using the 20-year floods. In order to
9112500 37.5 67.1 85.9 119 26.0 0.91 1.62 1.65 estimate the water depth associated with the 5-year and
9115500 7.08 11.1 16.0 22.1 11.6 0.52 0.80 0.96 20-year floods, we used recorded water stages reported by
9124500 42.0 47.8 62.6 70.2 24.9 0.88 1.08 1.15 the USGS for the historic peak flows at each gauging
9242500 101 115 139 155 36.6 1.45 1.65 1.72 station. We compared the 5-year and 20-year water stage
9244410 167 224 282 326 53.3 1.63 2.22 2.52
9251000 255 293 362 413 83.8 1.85 2.38 2.71 values to the water stage for a flow close to the reported
9253000 72.2 63.7 82.7 91.5 30.5 1.13 1.22 1.54 bankfull stage. In order to avoid the effects of net aggrada-
9257000 114 135 173 203 36.6 1.65 2.03 2.28 tion or degradation at the gauge section, water stage values
Ratio to Qbf /Qi 0.77 0.59 0.50 were chosen as close together in time as possible. The
estimated difference in the stage between the 5-year and
20-year flood and the flood magnitude reported by Andrews
depth is difficult to measure in the field, and since the stage [1984] was added to the bankfull depths that Andrews
relations at gauge cross sections are not necessarily the originally reported. The result was an average increase in
same as the stage relations elsewhere in a river reach, it is at water depth of about 33% for the 5-year event and 55% for
least possible that formative discharges were systematically the 20-year event (Table A1).
underestimated.
[79] It is also possible that the observed channel widths [82] Acknowledgments. Rudy Slingerland generously made the data
are associated with flows that are larger than the bankfull of H. Andrén available to us, and John Pitlick kindly provided unpublished
flows. Since, in semiarid regions, the rate at which vegeta- data on the Colorado River. We also extend our thanks to Alex Densmore,
Rob Ferguson, Tom Lisle, and an anonymous reviewer for their extensive
tion can colonize an eroded surface is substantially slower and very helpful comments on the paper; all of their comments have
than it is in a humid environment, it is likely that stream resulted in a much improved paper.
channels in arid environments are influenced to a greater
degree by less frequent flood events [Baker, 1977; Friedman References
et al., 1996]. Andrén, H. (1994), Development of the Laitaure Delta, Swedish Lappland,
188 pp., Inst. of Earth Sci., Ph.D. dissertation, Phys. Geogr., Uppsala
[80] In order to assess the possibility that the widths Univ., Uppsala, Sweden.
reported by Andrews [1984] were formed during flows Andrews, E. (1982), Bank stability and channel width adjustment, East
higher than the reported bankfull value, we first used Fork River, Wyoming, Water Resour. Res., 18, 1184 – 1192.
Andrews, E. D. (1984), Bed material entrainment and the hydraulic geo-
Andrews’s dimensionless hydraulic geometry equations to metry of gravel-bed rivers in Colorado, Geol. Soc. Am. Bull., 95, 371 –
predict the widths of stable gravel bed streams with similar 378.
bank vegetation information [Hey and Thorne, 1986]. Hey Baker, V. R. (1977), Stream-channel response to floods, with examples
and Thorne [1986] classified bank vegetation into four from central Texas, Geol. Soc. Am. Bull., 88, 1057 – 1071.
Blench, T. (1969), Mobile-Bed Fluviology, 168 pp., Univ. of Alberta Press,
classes. We have assumed that the dimensionless equations Edmonton, Canada.
for thin vegetation reported by Andrews [1984] should Bray, D. I. (1973), Regime relations for Alberta gravel-bed rivers, in Fluvial
describe Hey and Thorne’s types I and II channels, which Processes and Sedimentation: Proceedings of Hydrology Symposium No.
9, pp. 440 – 452, Natl. Res. Counc. of Can., Ottawa.
are the least densely vegetated, and that Andrews’s equa- Cao, S. (1996), Fluvial Hydraulic Geometry, 148 pp., Chengdu Univ. of
tions for thick vegetation should correspond to Hey and Sci. and Technol. Press, Chengdu, China.
Thorne’s types III and IV channels. When Hey and Thorne’s Chang, H. H. (1979), Minimum stream power and river channel patterns,
J. Hydrol., 41, 303 – 327.
discharges are entered into the Andrews equations, they Chang, H. H. (1980), Geometry of gravel streams, J. Hydraul. Div. Am.
consistently overpredict the observed channel widths in Hey Soc. Civ. Eng., 106, 1443 – 1456.
and Thorne’s data set. We reduced the discharge used in the Charlton, F. G., P. M. Brown, and R. W. Benson (1978), The hydraulic
equations until the observed and predicted widths agreed, geometry of some gravel rivers in Britain, Hydraul. Res. Station,
Wallingford, U. K.
on average (Figure A1). The Andrews equations predict Davies, T. R. H. (1987), Channel boundary shape—Evolution and equili-
widths that are, on average, consistent with the channel brium, in Rivers—Environment, Process and Form, IBG Spec. Publ. 18,
widths observed by Hey and Thorne when their Q estimates edited by K. Richards, pp. 228 – 248, Inst. of Br. Geogr., London.
Dietrich, W. E., J. W. Kirchner, H. Ikeda, and F. Iseya (1989), Sediment
are reduced to approximately 66% of their original values. supply and the development of the coarse surface layer in gravel-bedded
This, combined with the initial results using the regime rivers, Nature, 340, 215 – 217.
model, suggests that the widths observed by Andrews Eaton, B. C. (2006), Bank stability analysis for regime models of vegetated
[1984] were probably formed by flows consistently greater gravel bed rivers, Earth Surf. Processes Landforms, 31, 1438 – 1444,
doi:10.1002/esp.1364.
than the reported bankfull flows. We therefore surmise that Eaton, B. C., and R. G. Millar (2004), Optimal alluvial channel width under
the bankfull widths observed in the field by Andrews were a bank stability constraint, Geomorphology, 62, 35 – 45.

17 of 18
F03025 EATON AND CHURCH: HYDRAULIC GEOMETRY F03025

Eaton, B. C., M. Church, and R. G. Millar (2004), Rational regime model of Mikhailov, V. N. (1970), Hydrologic-morphometric characteristics of delta
alluvial channel morphology and response, Earth Surf. Processes Land- branches, in Hydrology of Deltas: Proceedings of the Bucharest Sympo-
forms, 29, 511 – 529. sium, 6 – 14 May, 1969, Stud. Rep. Hydrol., vol. 9, pp. 146 – 157, Int.
Ellis, E. R., and M. Church (2005), Hydraulic geometry of secondary Assoc. of Sci. Hydrol., Louvain, Belgium.
channels of lower Fraser River, British Columbia, from acoustic Doppler Millar, R. G. (2005), Theoretical regime equations for mobile gravel-bed
profiling, Water Resour. Res., 41, W08421, doi:10.1029/2004WR003777. rivers with stable banks, Geomorphology, 64, 207 – 220.
Emmett, W. W. (1975), The channels and waters of the Upper Salmon River Millar, R. G., and M. C. Quick (1993), Effect of bank stability on geometry
Area, Idaho, U.S. Geol. Surv. Prof. Pap., 870-A. of gravel rivers, J. Hydraul. Div., Am. Soc. Civ. Eng., 119, 1343 – 1363.
Friedman, J. M., W. R. Osterkamp, and W. M. J. Lewis (1996), Channel Nanson, G. C., and A. D. Knighton (1996), Anabranching rivers: Their
narrowing and vegetation development following a Great Plains flood, cause, character and classification, Earth Surf. Processes Landforms,
Ecology, 77, 2167 – 2181. 21, 217 – 239.
Hey, R. D., and C. R. Thorne (1986), Stable channels with mobile gravel Neill, C. R. (1968), Note on initial movement of coarse uniform bed-material,
beds, J. Hydraul. Div. Am. Soc. Civ. Eng., 112, 671 – 689. J. Hydraul. Res., 6, 173 – 176.
Huang, H. Q., and G. C. Nanson (1998), The influence of bank strength on Neill, C. R. (1973), Hydraulic geometry of sand rivers in Alberta, in Fluvial
channel geometry: An integrated analysis of some observations, Earth Processes and Sedimentation: Proceedings of Hydrology Symposium No.
Surf. Processes Landforms, 23, 865 – 876. 9, pp. 453 – 461, Natl. Res. Counc. of Can., Ottawa.
Huang, H. Q., and G. C. Nanson (2000), Hydraulic geometry and maximum Parker, G. (1978), Self-formed rivers with equilibrium banks and mobile
flow efficiency as products of the principle of least action, Earth Surf. bed: Part 2. The gravel river, J. Fluid Mech., 89, 127 – 156.
Processes Landforms, 25, 1 – 16. Parker, G. (1979), Hydraulic geometry of active gravel rivers, J. Hydraul.
Huang, H. Q., H. H. Chang, and G. C. Nanson (2004), Minimum energy as Div. Am. Soc. Civ. Eng., 105, 1185 – 1201.
the general form of critical flow and maximum flow efficiency and for Parker, G. (1990), Surface-based bedload transport relation for gravel riv-
explaining variations in river channel pattern, Water Resour. Res., 40, ers, J. Hydraul. Res., 28, 417 – 436.
W04502, doi:10.1029/2003WR002539. Pitlick, J., and R. Cress (2000), Longitudinal trends in channel character-
Jansen, J. D., and G. C. Nanson (2004), Anabranching and maximum flow istics of the Colorado River and implications for food-web dynamics,
efficiency in Magela Creek, northern Australia, Water Resour. Res., 40, 52 pp., U. S. Fish and Wildlife Serv., Grand Junction, Colo.
W04503, doi:10.1029/2003WR002408. Pitlick, J., and R. Cress (2002), Downstream changes in the channel geo-
Jones, L. S., and N. F. Humphrey (1997), Weathering-controlled abrasion in metry of a large gravel bed river, Water Resour. Res., 38(10), 1216,
a coarse-grained meandering reach of the Rio Grande: Implications for doi:10.1029/2001WR000898.
the rock record, Geol. Soc. Am. Bull., 109, 1080 – 1088. Pizzuto, J. E. (1995), Downstream lining in a network of gravel-bedded
Kennedy, R. G. (1895), The prevention of silting in irrigation canals, Proc. rivers, Water Resour. Res., 31, 753 – 759.
Inst. Civ. Eng., 119, 281 – 290. Simons, D. B., and M. L. Albertson (1963), Uniform water conveyance
Kirkby, M. J. (1977), Maximum sediment efficiency as a criterion for channels in alluvial material, Am. Soc. Civ. Eng. Trans., 128, 65 – 107.
alluvial channels, in River Channel Changes, edited by K. J. Gregory, Sklar, L. S., W. E. Dietrich, E. Foufoula-Georgiou, B. Lashermes, and
pp. 429 – 442, John Wiley, Chichester, U. K. D. Bellugi (2006), Do gravel bed river size distributions record channel
Knighton, D. (1998), Fluvial Forms and Processes, 383 pp., John Wiley, network structure?, Water Resour. Res., 42, W06D18, doi:10.1029/
New York. 2006WR005035.
Lacey, G. (1930), Stable channels in alluvium, Proc. Inst. Civ. Eng., 229, Tabata, K. K., and E. J. Hickin (2003), Interchannel hydraulic geometry and
259 – 384. hydraulic efficiency of the anastomosing Columbia River, southeastern
Lane, E. W. (1957), A study of the shape of channels formed by natural British Columbia, Canada, Earth Surf. Processes Landforms, 28, 837 –
streams flowing in erodible material, 106 pp., U. S. Army Eng. Div., Mo. 852.
River, Omaha, Nebr. Van Rijn, L. C. (1984), Sediment transport, part I: Bed load transport,
Langbein, W. B., and L. B. Leopold (1966), River meanders—Theory of J. Hydraul. Eng., 110, 1431 – 1456.
minimum variance, U. S. Geol. Surv. Prof. Pap., 422-H. White, W. R., R. Bettess, and E. Paris (1982), Analytical approach to river
Leopold, L. B., and T. Maddock (1953), The hydraulic geometry of stream regime, J. Hydraul. Div. Am. Soc. Civ. Eng., 108, 1179 – 1193.
channels and some physiographic implications, U.S. Geol. Surv. Prof. Yang, C. T. (1976), Minimum unit stream power and fluvial hydraulics,
Pap., 252. J. Hydraul. Div., Am. Soc. Civ. Eng., 102, 769 – 784.
Lindley, E. S. (1919), Regime canals, paper presented at Punjab Engineer- Yang, C. T., C. C. S. Song, and M. J. Woldenberg (1981), Hydraulic
ing Congress, Inst. of Br. Geogr., Lahore, Pakistan. geometry and minimum rate of energy dissipation, Water Resour. Res.,
Makaske, B., D. G. Smith, and H. J. A. Berendsen (2002), Avulsions, 17, 1014 – 1018.
channel evolution and floodplain sedimentation rates of the anastomosing
upper Columbia River, British Columbia, Canada, Sedimentology, 49, 
1049 – 1071. M. Church and B. C. Eaton, Department of Geography, University of
British Columbia, Vancouver, BC, Canada V6T 1Z2. (brett.eaton@ubc.ca)

18 of 18

You might also like