Resource Directed Measures for Protection of Water Resources: River Ecosystems
R17: THE PROCEDURE FOR GENERATING HYDRAULIC DATA FOR THE INTERMEDIATE AND
COMPREHENSIVE ECOLOGICAL RESERVE: QUANTITY
Senior Author:
A.L. Birkhead, Streamflow Solutions
Contributing authors:
W.S. Rowlston, Department of Water Affairs and Forestry
A. Jordanova, Department of Water Affairs and Forestry
Editor:
Heather MacKay, Department of Water Affairs and Forestry
Lizette Guest, Guest Environmental Management
Version:
1.0
Date:
24 September 1999
M:\f_rdm_october\rivers\version 1.0\riv_appR17_version1.0.doc
Department of Water Affairs and Forestry, South Africa
Version 1.0: 24 September 1999
R17/1
Resource Directed Measures for Protection of Water Resources: River Ecosystem
Appendix R17:
The Procedure for Generating Hydraulic Information and
Intermediate and Comprehensive Ecological Reserves (Quantity)1
This document is an adaptation of the hydraulics chapter for the Building Block Methodology (BBM) manual
(Rowlston et al., in prep), which describes the role of hydraulics in determining the water quality component of
the Comprehensive Ecological Reserve (CER) using the BBM. Much of the material presented therein is
equally applicable to the Intermediate RDM determination, and the chapter is therefore adapted here to avoid
continual cross-referencing and unnecessary duplication.
R17.1 The role of hydraulics in the ecological functioning of a
river
The flow of water in a river channel and its physical structure are intimately related in a cycle of cause and
effect, both spatially and temporally. Depending on the susceptibility of the channel to flow related change,
channel morphology is determined by the geology as well as by the sediment and flow regimes, whilst local
hydraulic conditions are determined by the geometry and flow resistance of the channel. Local hydraulics and
channel morphology are the primary determinants of the availability of physical habitat which, in turn, controls
ecosystem functioning. A quantitative understanding of the flow regime of a river, its physical structure, and its
discharge-depth regime, derived jointly and severally from hydrological, geomorphological and hydraulic
analyses, is therefore a prerequisite for deriving quantitative information about its ecological functioning.
R17.2 The role of hydraulics in the Building Block Methodology
Holistic methods for IFR determination (refer to Tharme et al., 1996) quantify ecological flow requirements for
the various biotic components of rivers in terms of parameters such as flow depth, flow velocity, wetted
perimeter and water surface width, adding time as a parameter by referring to the frequency of occurrence (or
level of assurance) of a particular flow rate, or the duration of inundation resulting from a particular flooding
event.
Hydrologists, water engineers and water resource managers, on the other hand, are more comfortable when
dealing with the water needs of humankind, and habitually express these needs in terms of volume and time.
This quantification can range from an instantaneous flow rate in cubic metres per second (m3/sec), to long-term
requirements in millions of cubic metres per annum (Mm3/a).
Although both approaches are completely valid in their own context, the application of the BBM requires an
interface between them: this interface is found in the hydraulic analysis of flow in natural open channels. The
results of hydraulic analyses and modelling therefore form the essential link between the way in which the
hydrologist, engineer and water resource manager express the flow of water in the river, and the ways in which
river ecologists express the water requirements of the river ecosystem itself.
The primary product of hydraulics work comprises a series of relationships between flow rate and, among
others, flow depth, flow velocity, wetted perimeter and water surface width. Fluvial geomorphologists and river
ecologists (benthic macroinverterbrate, fish and riparian vegetation biologists) use this information to quantify
flow requirements in ways which are described in other parts of the documentation detailing the Reserve
determination. The invertebrate specialist, for example, considers the availability of hydraulic habitat for the
invertebrate community which is characteristic of the IFR sites and/or river. Fish biologists take into account
the requirement for a critical depth or velocity for fish passage for particular purposes in various seasons, for the
inundation of particular habitats, particularly for breeding, and for habitat refuges in low flow conditions.
1
Editor’s note: Use of terminology IFR/BBM/Reserve is inconsistent and must be reviewed for version 1.1
Department of Water Affairs and Forestry, South Africa
Version 1.0: 24 September 1999
R17/2
Resource Directed Measures for Protection of Water Resources: River Ecosystem
Riparian vegetation specialists draw on information about the flooding requirements for recruitment of key
riparian trees and marginal vegetation, and the effects of morphological changes on the extent, for example, of
reed beds. The geomorphologist requires an estimate of the hydraulic shear stress to determine the flow at which
various sediment sizes may be entrained (mobilised), transported and deposited, and the water levels (stages)
necessary to inundate particular morphological units (terraces and benches for instance) for channel
maintenance.
It is important to note that there is, of necessity, great emphasis on the hydraulic characterisation of low flows in
the BBM. The difficulties attendant on low flow hydraulics work - when compared with the analyses of high
flows and floods which are more familiar to engineering hydraulicians - are not to be underestimated, and are
repeatedly revisited throughout this document.
Except for the special case of sediment as a component of water quality, and its transport or deposition in a river
channel, hydraulics as discussed in this document does not specifically deal with water quality considerations.
Water quality modelling is, however, dependent on the hydraulic characterisation of flow through a river
system.
R17.3 The role of hydraulics in the Intermediate and
Comprehensive Ecological Reserves
Resource constraints, both financial and temporal, dictate that the Intermediate RDM determination is a scaleddown version of the CER. The scale reduction of the IFR 2 applies to all aspects and has important ramifications
for the hydraulics component. Holistic methods for IFR determination are dependent on an acceptable degree of
accuracy in the characterisation of river hydraulics at the IFR sites. The means of achieving reasonable
confidence in the hydraulics over a range of flows, and particularly at low f lows, may be assessed by
considering the influence of data requirements and site complexity on overall hydraulic confidence for
Comprehensive and Intermediate type assessments.
Figure 1 is a plot of site complexity against available observed data (from a hydraulics perspective). The
confidence that may be achieved in the hydraulic characterisation is represented by the shaded area between
axes, with increased shading density representing increased confidence. For a given degree of site complexity,
increased observed data results in improved confidence. For a Comprehensive determination, 4 to 5
observations of discharge and stage over a range of f lows are generally undertaken, and site complexity ranges
from low (preferable) to medium (often necessitated by geomorphological and biotic considerations). The
hydraulic confidence for lFR3 sites therefore generally falls within the rectangular box, i.e. medium overall
confidence, with higher confidence achieved where site complexity is low and the observed rating data set
covers a wide range of flows.
By necessity, the available observed hydraulic data for an Intermediate Ecological Reserve assessment is
considerably lower then that for a Comprehensive determination, as indicated by the arrow on the ‘observed
data’ axis. Consequently, in order to achieve a reasonable degree of accuracy, the site complexity must
necessarily be low.
R17.4 The sequence of necessary activities
R17.4.1 Site selection
The selection of sites for consideration in an IFR exercise is dealt with in detail in supporting documentation,
from which it is clear that a wide range of factors must be taken into account (See Appendix R22).
2
Editor’s note: Here IFR is used to indicate “the water quantity component of the ecological Reserve”
3
Editor’s note: See note 1 above.
Department of Water Affairs and Forestry, South Africa
Version 1.0: 24 September 1999
R17/3
Resource Directed Measures for Protection of Water Resources: River Ecosystem
Figure 1: Illustration of the overall confidence that may be achieved in the hydraulic characterisation at
IFR sites as a function of site complexity (hydraulic) and availability of observed data.
As the principal purpose of an IFR exercise is to determine the flow regime which will maintain an acceptable
level of ecological functioning in the river, biotic considerations will dominate the selection of appropriate sites.
Resource constraints will almost always dictate that the reach of river under investigation has to be characterised
by a relatively small number of sites, and this in turn dictates that the limited number of sites used should
illustrate as high a degree of habitat - and therefore biotic - diversity as possible. Consequently, thus far in the
relatively brief history of IFR determination in South Africa, sites with riffles have been widely used. Such
geomorphological features are hydraulically complex, especially at the low flows which receive considerable
attention in IFR determinations. Under these conditions depths of flow are usually the same order of size as the
roughness elements (gravels, cobbles and boulders) which constitute the river bed, and which result in wide
variations and non-uniformity of flow velocities. These factors complicate the hydraulic analysis.
Whilst it is important for the hydraulics specialist not to expect that hydraulic considerations will enjoy absolute
pre-eminence in site selection, it is equally important for the hydraulician to influence the selection process to
the extent that the sites chosen are not of such hydraulic complexity that reliable hydraulic analysis becomes
impractical within the limits of available resources. Under these circumstances a site which is difficult to
analyse will almost certainly produce hydraulic information which is of low confidence, with consequent
negative implications for the IFR assessment process. An example is a site characterised by multiple distributary
channels flowing at different water levels.
The hydraulic complexity of the sites selected for an IFR exercise has a profound influence on the ways in
which hydraulic data are analysed, particularly in respect of the proportions of observed and modelled data
required for the production of reliable relationships between flow rate and, for instance, depth and velocity. As a
general rule: the more hydraulically complex the site, the greater the reliance on observed data for reliable
results from the hydraulic analysis. Conversely, the hydraulic characterisation of a simpler site may be achieved
by using relatively sparse observed data, followed by the use of appropriate hydraulic modelling techniques.
Department of Water Affairs and Forestry, South Africa
Version 1.0: 24 September 1999
R17/4
Resource Directed Measures for Protection of Water Resources: River Ecosystem
This can not be overemphasized for an IER, where the ability to provide hydraulic information of reasonable
accuracy, based on minimal observed data, is required (refer to Fig. 1). Sites selected for an IER should
therefore ideally be characterised by prismatic, single active channels; uniform energy (water surface) gradients;
conditions where flow resistance is not strongly influenced by stage or discharge; and the ability to accurately
assess the discharge through the site.
Discharge through a rapid or riffle is relatively difficult to measure directly with any confidence and, in
selecting such a site for direct, manual gauging, it is necessary to consider whether the discharge may be more
reliably measured at a nearby site where the cross-section is prismatic, the flow is materially uniform (i.e. does
not change with distance along the river), and considerably (say, ten times) deeper than the roughness elements
constituting the bed. This surrogate discharge site should be sufficiently close to the IFR site that any losses or
inflows between the two sites are minor, and can safely be ignored. If the need arises for such a surrogate
discharge site, its position must be identified, and arrangements made for it to be surveyed.
Location of cross-section/s
(a)
Comprehensive Ecological Reserve
Once the IFR site has been selected, it is important that adequate time and effort is assigned to the selection of
cross-sections, and that all specialists are involved. The hydraulic characterisation of the site - and therefore the
characterisation of its physical habitat - is primarily confined to the cross-sections, and therefore the success of
the process is largely dependent on their appropriate location.
Although an IFR site is three-dimensional, spatially linked two-dimensional cross-sections are used to describe
the river geometry, and the relationships between discharge and the hydraulic determinants mentioned
previously. Methods are currently being investigated for extending the hydraulic characterisation to provide a
more representative spatial description of the IFR sites, without the need for full three-dimensional
topographical surveys and hydraulic modelling (see 12 following) 4.
The hydraulics specialist must determine the number and location of channel cross-sections required to
characterise the site hydraulically, but it is difficult to predefine the number of cross-sections required, since it is
a function of both the local biotic and abiotic characteristics. Experience has shown that the following approach
is appropriate for a potentially difficult and time-consuming task:

Firstly, each specialist present at the site selection should identify and justify the location of ‘non-hydraulic’
cross-sections;

Secondly, the positions and importance of all cross-sections should be assessed with a view to combining
cross-sections without loss of essential information; and

Finally, additional cross-sections for hydraulic purposes (positioned at changes in the water slope and
channel geometry) should be located by the hydraulician, and the purpose of their inclusion explained to the
specialists on the team.
When selecting the additional hydraulic sections, the hydraulician must bear in mind that hydraulic controls (i.e.
determinants of the relationship between discharge and flow depth) are a function of discharge. These must
therefore be selected at a discharge appropriate to IFR-type application; that is, with greater emphasis on low
flows.
(b) Intermediate Ecological Reserve
Site selection for an IER is undertaken by the hydraulician and river ecologist who jointly represent the site
specific needs of the various IFR specialists, and therefore must be adept with the morphological and biological
characteristics required of IFR sites. Resource constraints usually dictate that one cross-section per site is
surveyed, and it is consequently essential that the profile is correctly positioned from both a hydraulic and biotic
perspective. Due the minimal time available between site selection and the on-site specialist meeting, it is
necessary and cost-effective to undertake the survey immediately following site selection. This implies that the
hydraulician should have sufficient expertise, and the necessary equipment, to carry out the survey.
4
Editor’s note: Author to check and correct cross-references.
Department of Water Affairs and Forestry, South Africa
Version 1.0: 24 September 1999
R17/5
Resource Directed Measures for Protection of Water Resources: River Ecosystem
R17.4.2 Site survey
The purpose of the survey from the hydraulics viewpoint is to define the cross-sectional profile of the river
channel in sufficient detail to enable hydraulic measurements and analysis to be undertaken at sufficient levels
of resolution, and to identify features of interest to river scientists.
Whilst the site survey is essential to the hydraulics work, it should not be regarded as the sole province or
responsibility of the hydraulician for a CER, where the site dependent IFR components are represented by
specialists at the site selection and on-site data collection activities. In addition to the hydraulics-related details
of the channel, the survey should also be used to acquire data and information in respect of, for instance, the
location and type of instream, marginal and riparian vegetation, which may then be portrayed on the site plan
and plots of the channel cross-sections. Requirements in this respect should be obtained from the other
specialists for the CER, but the responsibility lies solely with the hydraulician and ‘representative’ river
ecologist for an IER. Consequently, the IER site selection team is required to be familiar with site
characteristics appropriate for determining flow requirements of the various IFR components. Furthermore,
assuming that the hydraulician and representative river ecologist undertake the site survey immediately
following site selection, the survey should include pertinent geomorphological features (such as for example
terraces and bankfull levels positioned off the surveyed cross-section) and key riparian trees that are tagged for
future identification by the riparian vegetation specialist at the on-site specialist meeting.
The survey should extend from bank to bank of the macro-channel, and should incorporate all significant
changes in slope along the profile. It must be appreciated that the roughness elements along the profile that are
frequently transported (i.e. annually) constitute the overall resistance of the river channel and therefore need not
be surveyed in minute detail. Larger sedimentary obstructions that are moved infrequently, however, must be
included in the cross-sectional survey, since these features reduce the channel area for all but the highest floods.
At least two permanent bench marks must be placed at each of the cross-sections and clearly marked for future
identification. The bench marks form the local datum for linking the plan orientation and elevation of the crosssections, longitudinal river bed and water surface profiles, and it is therefore essential that the bench marks are
related to each other in elevation to an acceptable accuracy (+1cm), particularly for sites characterised by mild
water surface slopes.
It is preferable to survey the defined channel cross-sections at the outset of the IFR study, and preferably during
low-flow conditions. It may, however, be necessary to select sites when higher flow conditions prevail, due to
untimely climatic conditions, or because the site selection has been scheduled during the wet season. Under such
circumstances, stage level data (and discharge data, discussed in 4.35) may be collected along the river profile
and reconciled with the positioning of cross-sections at a later date. If, during the course of the IFR exercise, a
high flow event occurs which causes the cross-section to change (through scour of bed or banks, or deposition of
sediments), it will be necessary to re-survey the channel. Significant changes in the cross-section will necessitate
a re-evaluation of work undertaken previously, and may invalidate some of the results. Although it is difficult to
directly include provision for such eventualities in the work programme, implications in terms of additional
resources and changes to the scheduling of the overall study must be addressed in the work outline.
In order to develop stage-discharge relationships for the cross-sections, at every site visit when the flow rate is
measured, water levels relative to the local bench marks need to be surveyed at the banks of each active channel
along the cross-sectional profiles. At sites where access into the river is safe, longitudinal river bed and water
surface profiles should also be surveyed at the lowest bed level, extending approximately ten channel widths
beyond the downstream and upstream cross-sections. These longitudinal water surface profile data are
particularly necessary for cases where a single cross-section is used to characterise a site - resource constraints
often dictate that this is so (refer to 4.1.2.2 for an IER).
Whilst recording bed and water levels along the longitudinal profile the opportunity should also be taken to
record depth-averaged velocities at the lowest bed level. The use of this data in the overall process is, as yet, not
fully defined, but in the few cases where such data has been gathered it has proved useful to fish and
macroinvertebrate biologists in developing a more three-dimensional picture of the way in which velocities vary
along the river.
5
Editor’s note: Author to check all cross-references
Department of Water Affairs and Forestry, South Africa
Version 1.0: 24 September 1999
R17/6
Resource Directed Measures for Protection of Water Resources: River Ecosystem
The equipment best suited for undertaking the survey is a total station linked to a data logger, with the data
recorded in an unreduced (raw) format (i.e. horizontal and vertical angles, and sloping distance) and not reduced
co-ordinates. The survey data may then be readily reduced using trigonometric principles to produce crosssectional profiles, plan orientation of sections, stage levels and longitudinal profiles.
Some hydraulicians may have sufficient expertise, and the necessary equipment, to carry out the survey (This
the most pragmatic (and cost-effective) approach for an IER, given the time constraints). Alternatively, expert
surveyors may be employed for the task. In this latter case experience has shown that useful results are obtained
only if the surveyors know and understand the purpose of the work, and the reasons for the details they are
asked to record. Irrespective of who actually carries out the survey work, the hydraulician is responsible for
defining the level of detail required in the survey.
It appears to be self-evident that survey work must be carried out at the actual sites. However, in undertaking
hydraulic analysis in respect of flood flows, with which engineering hydraulicians are generally more familiar, it
is often possible to derive all the necessary physical details of the river channel from contour plans,
orthophotographs, and aerial and terrestrial photographs to an acceptable degree of accuracy. This is definitely
not the case when dealing with the low flows which are so important in IFR determinations, where the necessary
detail can only be acquired through observations at the sites themselves.
The importance of visual information on the sites cannot be overemphasised, and every opportunity should be
taken to take photographs at as wide a range of flow rates as possible. Surprisingly accurate quantitative
information about incremental changes in width and depth can often be derived from photographs by relating
them to known dimensions on the cross-sectional survey, and by observing the degree of inundation of, for
instance, a prominent feature such as a large boulder, or the extent of inundation of marginal vegetation.
At each visit to each site, at least three photographs of each cross-section should be taken, each from a
subsequently identifiable and repeatable position, across the channel along the surveyed cross-section, and
facing upstream and downstream with the surveyed cross-section in the foreground. Photographs must be
related to a known flow rate, and must be dated.
Visits to site for hydraulic data collection are often more frequent than for the other specialist disciplines. At the
outset the hydraulician should obtain from the other specialists their requirements, if any, for regular
photography for their particular purposes, and build these needs into the programme of site activities at each
visit.
R17.4.3 Measurement of discharge
It is essential that, to be of any use in the IFR determination, parameters such as flow depth and velocity,
specified by the other specialists in the process, are related to a known discharge (flow rate).
Various methods exist for the measurement of discharge, including the use of existing rated sites (natural river
sections and structural gauges) and manual techniques including the velocity-area and dilution methods.
A gauging weir or rated cross-section located in close vicinity to the IFR site provides a useful means of
obtaining discharge data. The integrity of data derived from hydrological stations must not however be taken
for granted, and should be checked with the authority responsible for its operation. The gauge should be
sufficiently close to the site so that intervening inflows and losses may be neglected. Furthermore, care must be
exercised during unsteady flow conditions, to account for the travel time and attenuation of flow between the
local IFR site and the remote gauging station. A method for synthesising rating relationships based on the
measurement of an unsteady flow event is provided by Birkhead and James (1998).
The velocity-area method is undoubtedly the most commonly applied manual technique for determining the
flow rate in natural, medium to large watercourses, whereas dilution techniques are better suited to turbulent
rivers where other methods are difficult to apply (e.g. a rock-strewn river of high bed slope). Details and
standards for the application of these manual gauging techniques are given in the British Standard for the
Measurement of Liquid Flow in Open Channels (BS 3680), and must be consulted to ensure correct application
of the methods. See also Gordon et al. (1992).
Point velocities recorded during manual discharge measurement may be used to determine the distribution of
velocity across the channel and, if depth requires more than one depth-averaged velocity measurement,
Department of Water Affairs and Forestry, South Africa
Version 1.0: 24 September 1999
R17/7
Resource Directed Measures for Protection of Water Resources: River Ecosystem
vertically in the water column. The hydraulician should ascertain if this velocity data will be of use to any of the
other specialists, notably the fish biologist and the fluvial geomorphologist, in their deliberations. Reference is
made in 12 to further developments and application of velocity data in IFR determination.
In order to observe as wide a range of discharges as possible, it is highly desirable to undertake discharge and
related measurements, by whatever method, over at least one hydrological season. This, however, does not
necessarily guarantee that a suitable range of flows will be encountered due to the possibility of unfavourable
climatic conditions (e.g. a failed wet season or unseasonally high flows in the dry period). Under such
exceptional circumstances the IFR specialist meeting may be postponed or, if this is not possible, additional data
collection and refinement of the hydraulics and IFR recommendations will be required following the specialist
meeting. For an IER, time constraints will often dictate that two observed rating points do not cover a wide
range of discharges, and this should be taken into account when assessing the confidence in the extrapolated
hydraulic relationships.
Measuring discharge by direct methods during high flows in most rivers, where depth and velocity of flow
militate against safely entering the water, requires the use of boats or other techniques, and demands high
standards of safety to avoid accidents. Dangers from the natural inhabitants of the river such as hippopotami and
crocodiles, and the risk of contracting river-related diseases such as giardia and bilharzia, must also be taken
into account.
When high flows make entry into the river, by whatever means, impossible or inadvisable, stage levels at the
banks should be recorded at the cross-sections, as well as upstream and downstream of the sections as discussed
in 4.2. Where possible, floats should be used to measure surface velocities. The use of surface velocities to
obtain an estimate of discharge is described in BS 3680.
R17.4.4 Data analysis and modelling
This component of the hydraulics task is concerned with converting observed cross-sectional and flow data, the
latter often for a limited range of flow rates, into relationships between flow rate and the more biologically
useful parameters for the entire range of flows of interest in the IFR exercise.
When sufficient observed rating data exists, it is possible to establish rating functions based entirely on field
measurements by fitting relationships of the form given by Birkhead and James (1998) to the observed data.
(Refer to Fig. 2). Care must be exercised with the extrapolation of the modelled rating relationship above and
below the highest and lowest recorded flow, respectively. The validity of the extrapolation can be assessed by
computing inferred resistance coefficients and average flow velocities beyond the range of observed data, and
comparing these with reasonable values based on experience, and resistance coefficients given in the literature
(e.g. Barnes, 1967; Hicks and Mason, 1991 and Ven Te Chow, 1959).
Department of Water Affairs and Forestry, South Africa
Version 1.0: 24 September 1999
R17/8
Resource Directed Measures for Protection of Water Resources: River Ecosystem
Figure 2: Example plots of a rating relationship on log-normal scale showing the observed data, and a
cross-section profile.
Synthesis of a rating relationship essentially involves interpolation between sparse data points and extrapolation
beyond the limit of recorded data. As discussed previously with reference to Manning’s resistance coefficient,
flow resistance in natural watercourses is generally a function of stage, particularly at low stages. The rating
relationship may therefore be synthesised between observed data by interpolating the flow resistance coefficient.
The selection of a suitable resistance relationship and corresponding coefficient (e.g. Darcy-Weisbach, Chézy or
Manning) is considered arbitrary in the existing application and should be based on pragmatic considerations
such as for example the relevant resistance equation applied in available software (if this is to be used in the
analysis), experience and familiarity. This is because, although certain relationships are theoretically more
rigorous than others, it is illogical to apply the most rigorous modelling approach in a situation where the
resistance coefficient is essentially a ‘composite’ calibration factor based on field data. This factor and the
energy losses cannot be derived solely from consideration of the measurable physical dimensions of the
resistance components (e.g. size of the roughness elements, vegetation type and density, channel plan form, etc).
Extrapolated rating relationships may be developed by extrapolating the resistance coefficient based on
calibrated data (i.e. fitting a relationship to the resistance coefficient versus discharge (or stage) data and
extrapolating). Calibrated resistance coefficients are usually determined based on non-uniform flow profile
computations (i.e. backwaters), or by assuming linear variation of the energy slope for a site represented by a
single cross-section. Once again, care must be taken by judicially assessing the values of the extrapolated
coefficient and comparing these with values based on experience, and published in the literature.
A major difficulty with low-flow hydraulic analysis at many IFR sites (pools, for instance), is the estimation of
the stage of zero discharge; that is, the water level at which the flow ceases. The most appropriate method for
estimating the stage of zero discharge (in the absence of observed flow data) is to survey the longitudinal profile
downstream of the cross-section within the deepest portion of the active channel to ascertain the level of the
downstream bed which causes the upstream backup. Alternatively, extrapolation of the observed rating data to
zero discharge may also provide a useful, albeit approximate, estimate of the stage of zero discharge.
Department of Water Affairs and Forestry, South Africa
Version 1.0: 24 September 1999
R17/9
Resource Directed Measures for Protection of Water Resources: River Ecosystem
Once the rating relationship for a river section has been developed, the relationships between discharge and
other hydraulic determinants (e.g. average flow velocity, wetted perimeter and average flow depth - refer to Fig.
3) may readily be computed using the cross-sectional geometry.
Figure 3: Example plots of discharge against flow depth (maximum and average), average velocity and
wetted perimeter on normal scale axes.
Hydraulic analysis and modelling must only be carried out by skilled practitioners who are familiar with low
flow techniques and problems, as the errors inherent in the application of the more traditional approaches of
analysis, more suited to high flows, resonate throughout the entire process. For instance, as discussed earlier, the
values of the resistance parameter Manning’s n which must be applied to low flows in a riffle are considerably
higher than the range of values used in high flow analyses. Application of inappropriately low n values results in
significant underestimation of flow depths, and concomitant overestimation of velocities, for specific flow rates.
This in turn prompts overestimation of flow rates to achieve particular flow depths and velocities for, for
instance, fish passage, and thereby inflates the instream flow requirement.
Although the hydraulic information supplied for use within IFR assessments should be ‘best estimates’, it is
considered advisable to err on the side of conservatism when uncertainties are present due to limited observed
data. This is particularly applicable for an Intermediate determination where only two rating observations are
allowed for, and procured within a short time period. Conservatism in this respect implies overestimating the
discharge required to provide specific flow depths - a consequence of underestimating the resistance coefficient
as discussed above. If a CER is undertaken at a later stage and recommends significantly higher flows, then it
may be very difficult to provide for such recommendations if the balance of the resource has been allocated to
other users.
Department of Water Affairs and Forestry, South Africa
Version 1.0: 24 September 1999
R17/10
Resource Directed Measures for Protection of Water Resources: River Ecosystem
R17.5 Minimum, acceptable and ideal data sets
R17.5.1 Site selection
It is absolutely essential for the hydraulics specialist to visit, inspect, photograph and carry out (or oversee) the
site survey for both an Intermediate and Comprehensive Ecological Reserve. As emphasized previously, it is
completely impractical, when dealing with low flows, to work from contour maps or contoured site plans alone.
R17.5.2 Site Survey
The accuracy of the survey (cross-sectional profile and water levels) must be comparable to the level of
accuracy with which results are presented in the starter document and at the specialist meeting. For example, it
is fallacious to predict changes in discharge due to 1cm variations in river stage when water levels are accurate
to, say, only +5cm.
R17.5.3 Observed flow data - stage and discharge
In general, the more observed stage versus discharge data available, the higher the confidence is in the hydraulic
relationships derived. Every possible opportunity should be taken to visit the site and gather such data.
Sites characterised by prismatic river cross-sections and uniform flow conditions require relatively fewer
observed data points than more complex sites containing rapids and riffles, since the hydraulic relationships for
the former are relatively easier to synthesize analytically - and consequently more appropriate for an
Intermediate level of determination (refer to 4.1).
Ideally, observations should be made at significantly different flow rates. The flow range 0-5 m3/sec is better
characterised by flow rates of 0.2, 0.5, 1,2 and 5 m3/sec, than by 0.1, 0.12, 0.15 and 5 m3/sec. This requires
careful planning of site visits to maximise the likelihood of procuring a range of discharges. The range of flows
is river dependent, and a function of how much each flow increment influences the depth of flow. Some floodrelated information (which can sometimes be derived from recording the level of flood-borne debris on the river
banks or in trees, or from information obtained from residents living close to the river, related to known flows
from the hydrological record) is useful in fixing a high point of the curve in extrapolating the observed low flow
data.
Intermediate Ecological Reserve

Absolute minimum: one data point at an appropriate low flow, plus stage of zero discharge.

Acceptable: absolute minimum condition, plus one data point within the intermediate to high flow range.
Comprehensive Ecological Reserve

Absolute minimum: one data point at an appropriate low flow, plus stage of zero discharge

Acceptable: three points distributed over the low flow range of interest, plus stage of zero discharge, plus
some flood-related data if possible.

Ideal: six data points, good distribution, plus stage of zero discharge, plus some flood-related data.
R17.6 Summary of what to include in starter documents, and how
this is used by specialists
The following is a list of what is required in the starter document:

Brief summary of the method used to determine the rating relationships.
Department of Water Affairs and Forestry, South Africa
Version 1.0: 24 September 1999
R17/11
Resource Directed Measures for Protection of Water Resources: River Ecosystem






Record of techniques used to collect discharge data.
Tabulated observed discharge versus flow depth and resistance coefficient data.
Plots of the rating relationships on log-normal scale showing the observed data, and the cross-sections with
annotated salient features (refer to Fig. 2) for different flow ranges (i.e. low, intermediate and high).
Plots of discharge against flow depth (maximum and average), average velocity and wetted perimeter (refer
to Fig. 3) on normal scale axes.
Explicit estimate of the accuracy in the rating relationships over the range of observed data. This may be
assessed by calculating the average absolute difference in flow depth (or discharge) between the observed
and modelled data.
An assessment of the confidence in the extrapolated rating relationships.
R17.7 Roles and responsibilities at the specialist meeting




Explanation of data and information presented. Interpretation of data and assistance to other specialists in
deriving useful and appropriate information from the hydraulic data.
Assessment of levels of confidence in data.
Assessment of the limitations of data.
Transfer of technology.
R17.8 Roles and responsibilities after the specialist meeting
At this stage in the development of the IFR process its credibility with water resource managers depends to a
large extent on the quality and reliability of the hydraulic information. The fundamental importance of reliable
hydraulic information to the work of the other specialists involved in the process demands that results of the
highest possible quality, in terms both of completeness and confidence, should be presented at the specialist
meeting. Shortcomings in the hydraulic information will almost always militate against a successful outcome,
and it should never deliberately be assumed that hydraulic inadequacies can be made good at some later stage.
Post-specialist meeting hydraulic work in connection with monitoring the effectiveness of the specified IFR to
achieve the objectives of the modified flow regime is discussed in 13 following.
R17.9 Example of terms of reference for hydraulics specialists
The terms of reference for the hydraulics specialist have been discussed previously, in the descriptions of the
tasks necessary to carry out the study. The terms of reference should, however, in addition, provide an explicit
breakdown of the allocation of resources (time and rate of remuneration) for the essential tasks, including:








Site selection (CER - 5 days/4 sites; IER - 3 days/2 sites, including survey)
River channel cross-sectional and longitudinal profile surveys (CER - 1 day/site)
Field trips for collection of hydraulic data (CER - 3 days/4 sites, IER - 1.5 days/2 sites )
Reduction of survey and hydraulic data (CER - 0.5 day/site visit/4 sites, IER - 0.5 day/2 sites)
Hydraulic analysis and modelling (CER - 2 days/site, IER - 2 days/2 sites)
Reporting (CER - 2 days/4 sites, IER - 0.5 day/2 sites)
Specialist meeting (refer to supporting documentation for specialist meetings)
Post-specialist meeting activities (e.g. additional data collection and refinement, monitoring and scenario
meetings).
Notes:
 The time allocations for tasks involving field work (site selection and data collection) are dependant on the
location of the river system and site accessibility, and the allocations provided above are therefore
approximate.
 Allowance must be made for an assistant in undertaking the survey and hydraulic data collection. The
minimum time required for collection of hydraulics data is half a day per site, with a reasonable allocation
of one day per site to complete the all of the tasks described in this document.
Department of Water Affairs and Forestry, South Africa
Version 1.0: 24 September 1999
R17/12
Resource Directed Measures for Protection of Water Resources: River Ecosystem
R17.10
Minimum and optimal specialist training
If the results of the hydraulics task are to successfully support the work of the other specialists, all of its aspects
- site selection, site survey, discharge measurement, and the analysis of hydraulic data - require a high degree of
specialist knowledge, expertise and experience, particularly in the requirements of low flow work. Furthermore,
the limited hydraulics data used for the generation of hydraulic information in the IER (2 rating observations) is
likely to place greater reliance on specialist expertise and previous experience than for a CER where allowance
is made for 4 rating observations over a longer time frame thus improving the likelihood of recording flow
related data over a wider range of discharges.
It is however possible that local non-specialist assistance can be enlisted in gathering water level and flow rate
data, and a photographic record. This can provide additional observed data points whilst reducing the number of
site visits by the hydraulician. Extreme care must however be taken with the selection and training of such
assistance, since experience has shown that the integrity of data cannot be assured, with potential dire
consequences for the IFR study.
To illustrate, one such attempt was made in which the staff of a nature reserve area, through which the river in
question ran, were asked to observe and record relevant data. The intention was to supplement a limited number
of on-site observations made by specialists, to obtain a record for the whole annual flow cycle. Although
considerable efforts were made to define the necessary sequence of tasks (verbally on site, and in writing using
illustrations, as simply as possible, the resulting data proved to be of virtually no use in augmenting data
gathered by hydraulics specialists. This was by no means due to lack of intelligence on the part of the people
involved, who were all well-educated environmentalists, nor from a lack of understanding of what the objectives
of the exercise were, but arose from an understandable unfamiliarity with the hydraulic concepts involved. Such
attempts should not be made lightly, nor should great reliance be placed on the results.
R17.11
Potential pitfalls
The potential pitfalls have been discussed in the relevant sections of this document, with the following concerns
deserving re-emphasis:







Inappropriate location of cross-sections (biotic and hydraulic).
Surveys undertaken in insufficient detail and not linked to a common datum.
Inability to relocate fixed bench marks.
Inaccurate stage and discharge measurements.
Insufficient data over a limited flow range.
Inappropriate hydraulic analyses, particularly for low flows.
Sites too hydraulically complex to achieve a reasonable level of confidence for an IER (refer to Fig. 1).
R17.12
Further developments
As discussed previously, spatially linked two-dimensional cross-sections are used to describe the river geometry
and relationships between discharge and various hydraulic determinants. Consequently, the determination of the
IFR is strongly focussed on the geometry of two-dimensional river sections. In order to provide a less spatially
fixed description of the site, recently (Birkhead, 1998) flow depth data have been analysed as frequency
distributions of available depth across river sections as a function of discharge.
Similar analyses for depth-averaged velocity would provide useful information for invertebrate and fish
ecologists, but is only advisable using measured velocity data, due the complexity of velocity distribution across
non-prismatic channel sections. In order to further develop the role of hydraulics in IFR determination using the
BBM, is it therefore necessary to invest the time required to collect these data.
The application of full three-dimensional modelling techniques in order to obtain an improved characterisation
of the hydraulic conditions across the sites is prohibitive, primarily in terms of the data collection requirements
for an IFR type analysis. There is potential, however, to extend the existing hydraulic modelling to provide data
on habitat availability in quasi three-dimensions by integrating the hydraulic characterisation with habitat
Department of Water Affairs and Forestry, South Africa
Version 1.0: 24 September 1999
R17/13
Resource Directed Measures for Protection of Water Resources: River Ecosystem
mapping between cross-sections. Broadly, habitat mapping provides a plan description of the habitat conditions
at the sites, and the hydraulic analysis a continuous longitudinal water surface profile between cross-sections.
Consequently, it is possible to model (based on a common or water surface profile datum) the changes in flow
depth and velocity at the mapped sites.
R17.13
Monitoring
The overall objective of monitoring is to assess the effectiveness of the specified Instream Flow Requirements
in maintaining the existing condition, or to achieve the desired condition of the river. Two phases are envisaged:
establishment of baseline conditions, followed by monitoring to detect changes in the baseline conditions.
Ideally the baseline conditions, defining the hydraulic characteristics of all the sites for a full range of flows,
should have been determined prior to the specialist meeting. If this was not possible for any reason, baseline
monitoring should include collection of additional data to complete the hydraulic characterisation. This could
relate to supplementary observations to augment an incomplete flow range, or repeat observations to raise
confidence in a data set in which doubt arose at the specialist meeting. Additional data collection at IER IFR
sites following the specialist meeting may not only be used to refine the flow recommendations, but also serves
to provide information for a CER that may be undertaken at a later stage. If very high flows occur subsequent to
the specialist meeting and are believed to have altered the morphology of the channel, and consequently the
site’s hydraulic characteristics, resurvey followed by full hydraulic analysis will be required to establish a new
baseline condition.
Long-term monitoring involves repeat visits (regular, and also after high flow events) to the sites to resurvey the
cross-sections to detect morphological changes and collect flow-related data, followed by re-evaluation of the
hydraulic relationships.
Hydraulic monitoring relates closely to hydrological monitoring, the aim of which is to determine the degree of
concurrence between the flow regime specified by the IFR and that which actually eventuates.
R17.14
Conclusions
Experience with the determination of the Instream Flow Requirements of South African rivers has repeatedly
shown that the availability of reliable hydraulic information is of quintessential importance to the success of the
process as a whole. If biologically relevant data in terms of depths and velocities of flow cannot be related to
rates of flow, the instream flow requirements cannot be quantified. Although the work required to gather and
analyse hydraulic data can be costly and time-consuming, especially in rivers in remote areas, or where access is
difficult, the investment pays rich dividends in confidence. It is infinitely preferable to undertake this work prior
to the IFR specialist meeting, where it can be confidently used by the other disciplines, rather than to enter the
meeting with inadequate or unreasonably low confidence information, thereby having to subsequently revisit the
entire procedure with improved hydraulic information.
References & bibliography
Barnes, H.H. (1967). Roughness Characteristics of Natural Channels. U.S. Geological Survey Water Supply
Paper 1849. U.S. Geological Survey, Washington D.C., 1-9.
Birkhead, A.L. (1998). Mkomazi River Instream Flow Hydraulics. In: Site Visit Document for the 1998
Workshop to Identify the Instream Flow Requirements of the Mkomazi River, Scottburgh, Kwa-Zulu Natal, 2427 March 1998.
Birkhead, A.L. and James, C.S. (1998). Synthesis of rating curves from local stage and remote discharge
monitoring using non-linear Muskingum routing. Jour. Hydrology, 205, 52-65.
BS 3680, British Standard for the Measurement of Liquid Flow in Open Channels. Part 2: Dilution Methods.
Part 3A: Velocity-area Methods.
Department of Water Affairs and Forestry, South Africa
Version 1.0: 24 September 1999
R17/14
Resource Directed Measures for Protection of Water Resources: River Ecosystem
Gordon, N.D., McMahon, T.A. and Finlayson, B.L. (1992). Stream Hydrology: An Introduction for Ecologists.
John Wiley and Sons. 420 pp.
Henderson, F.M. (1966). Open Channel Flow. Macmillan. N.Y. 522 pp.
Hicks, D.M. and Mason, P.D. (1991). Roughness Characteristics of New Zealand Rivers. Water Resources
Survey Paper, DSIR, 1-13.
Rowlston, B., Jordanova, A. and Birkhead, A.L., in prep. Hydraulics chapter for the Building Block
Methodology used in Instream Flow Reqirement assessment. Water Research Commission Report.
Tharme, R. (1996). Review of international methodologies for the quantification of the Instream Flow
Requirements of rivers. Draft report.
Ven Te Chow (1959). Open Channel Hydraulics. McGraw-Hill, N.Y. 680 pp.
Glossary of terms
Flow/flow rate/discharge - Volumetric flux per unit time past a fixed point along the river (m 3/s).
Flow velocity - Speed at which water moves per unit time past a fixed point in a given direction (m/s).
Flow depth - Vertical distance between the water surface and bed at a given point in the river channel, or the
maximum depth along a cross-section (m).
Resistance - Overall resistance to flow imposed by the river channel, including all resistance components, e.g.
bed roughness, vegetation, channel plan form, etc. (Manning’s n - s/m1/3; Chézy’s C - m1/2/s; Darcy-Weisbach’s
f - dimensionless).
Roughness - Resistance imposed by the fluvial sediments.
Stage/water level - Elevation of the water surface relative to a datum (m).
Surface width - Horizontal extent of the water surface, measured along the channel cross-section (m).
Wetted perimeter - Amount of channel in contact with flow, measured along the cross-section (m).
Department of Water Affairs and Forestry, South Africa
Version 1.0: 24 September 1999
R17/15