i
Environmental Sustainability
Indicators
for
Urban River Management
LIFE02 ENV/UK/000144
March 2004
Environmental Sustainability Indicators for Urban River Management
i
The SMURF project, Sustainable Management of Urban Rivers and Floodplains, is a partnership
between Birmingham City Council, Environment Agency, Severn Trent Water, HR Wallingford,
Staatliches Umweltamt Herten, University of Birmingham and King’s College London.
Environmental Sustainability
Indicators
For Urban River Management
Produced by
Angela Boitsidis and Angela Gurnell
Department of Geography
King’s College London
March 2004
Environmental Sustainability Indicators for Urban River Management
ii
This report is a contribution to research generally and it would be imprudent for third parties to
rely on it in specific applications without first checking its suitability.
Various sections of this report rely on data supplied by or drawn from third party sources. King’s
College London accepts no liability for loss or damage suffered by the client or third parties as a
result of errors or inaccuracies in such third party data.
King’s College London will only accept responsibility for the use of its material in specific projects
where it has been engaged to advise upon a specific commission and given the opportunity to
express a view on the reliability of the material for the particular application.
Environmental Sustainability Indicators for Urban River Management
iii
Acknowledgements
The authors would particularly like to thank Professor G.E. Petts, University of Birmingham, and
Dr P Armitage, Centre for Ecology and Hydrology, Dorset, for their valuable discussions whilst we
were developing the present research and writing this report. We should also like to thank Ms M.
Lee for her assistance with the field survey of the River Tame. The methodology described in this
report has its foundation in an approach that was developed under funding from the Natural
Environment Research Council’s URGENT Thematic Programme.
Environmental Sustainability Indicators for Urban River Management
iv
Environmental Sustainability Indicators for Urban River Management
v
Contents
Title page i
Acknowledgements...........................................................................................................................iii
Contents .......................................................................................................................................... v
1. Introduction ........................................................................................................................ 1
2. Engineered stretches within a spatial hierarchical framework............................................ 2
3. A methodology for surveying urban river stretches............................................................ 5
3.1 The River Habitat Survey..................................................................................... 5
3.2 The Urban River Survey (URS) ........................................................................... 6
3.3 Background Measurements .................................................................................. 7
3.4 Spot-check Measurements.................................................................................... 7
3.5 Once-Only Measurements.................................................................................... 9
3.6 Cumulative Measurements ................................................................................. 11
4. Aggregate indices from URS data (primary environmental indicators)............................ 15
4.1 Indices describing Materials............................................................................... 15
4.2 Indices describing Physical Habitat Features ..................................................... 16
4.3 Indices describing Vegetation Structure and Biomass ....................................... 17
5. Classification of urban river stretches (secondary environmental indicators) .................. 20
5.1 Data Analysis ..................................................................................................... 20
5.2 Materials Attributes............................................................................................ 20
5.3 Vegetation Attributes.......................................................................................... 24
5.4 Environmental Indicators ................................................................................... 25
6. Sector scale indices (tertiary environmental indicators) ................................................... 29
6.1 Flow-related indicators....................................................................................... 29
6.2 Sediment-related indicators................................................................................ 29
6.3 Water quality indicators ..................................................................................... 30
6.4 Biotic indicators ................................................................................................. 31
6.5 Land Use and Land Availability......................................................................... 32
7. Combining indicators to address scenarios of change ...................................................... 33
7.1 Stretch Scores..................................................................................................... 33
7.2 Management Recommendations......................................................................... 35
7.3 A scoring system to underpin scenario modelling of engineering and
management change ........................................................................................... 37
7.4 Detailed Assessment of Stretch Characteristics and Constraints........................ 38
8. References ........................................................................................................................ 40
Tables
Table 1 Subdivisions of river channel planform character, cross section character and bed and
bank reinforcement that can be combined to define the engineering type for a reach of
urban channel....................................................................................................................4
Table 2 Groups of variables that must be considered for incorporation into an URS....................6
Table 3 Variables included in the URS Background Information and their compatibility with
those of the RHS methodology .........................................................................................7
Table 4 Comparison of spot-check variables between RHS and URS methodologies ..................8
Table 5 Land use types used to describe land use at the catchment scale (Level 1) and finer
spatial scales (Level 2)....................................................................................................10
Table 6 URS channel vegetation types and example species (Adapted from EA, 1997).............11
Environmental Sustainability Indicators for Urban River Management
vi
Contents continued
Table 7 Comparison of attributes assessed in the cumulative measurements within the RHS
and URS surveys.............................................................................................................12
Table 8 Types of pollution recorded in the URS..........................................................................13
Table 9 Categories of habitat features recorded in the URS. .......................................................14
Table 10 Synthetic Indices derived from the Urban River Survey relating to three different sets
of characteristics of urban river stretches: ‘Materials’, ‘Physical Habitat’ and
‘Vegetation’. ...................................................................................................................15
Table 11 Bank Protection Types....................................................................................................16
Table 12 Index Values for the dominant flow type within the stretch. ..........................................17
Table 13 Descriptions of the characteristics of stretches attributed to different Materials
clusters ............................................................................................................................22
Table 14 Descriptions of the characteristics of stretches attributed to different Physical Habitat
clusters. ...........................................................................................................................23
Table 15 Descriptions of the characteristics of stretches attributed to different vegetation
clusters. ...........................................................................................................................25
Table 16 Water Quality Criteria for defining River Ecosystem Classes (from NRA, 1994) .........31
Table 17 The GQA classification of rivers according to their biological quality incorporating
RIVPACS derived EQI indices (EA, Pers. Comm). .......................................................32
Table 18 Scoring system for defining the quality and potential of urban river channels...............34
Table 19 Overall scores, classes and management recommendations associated with categories
of urban river stretch.......................................................................................................36
Figures
Figure 1 A hierarchy of six spatial scales at which urban river data may be collected, stored
and analysed, with examples of the data types that might be collected at each scale .......3
Figure 2 Catchment and local controls on the geomorphology of river stretches ...........................3
Figure 3 Macrophyte types grouped according to their perceived attenuation of flow in urban
channels. .........................................................................................................................19
Figure 4 7 clusters of urban river stretches defined by their materials characteristics ..................21
Figure 5 6 clusters of urban river stretches defined by their physical habitat characteristics........22
Figure 6 8 clusters of urban river stretches defined by their Vegetation characteristics ...............24
Figure 7 Flow chart for allocating urban river stretches to the relevant Materials Class ..............26
Figure 8 Flow chart for allocating urban river stretches to the relevant Physical Class................27
Figure 9 Flow chart for allocating urban river stretches to the relevant vegetation class..............28
Figure 10 Frequency distribution of overall scores for 106 stretches of the River Tame................35
Environmental Sustainability Indicators for Urban River Management
1
1. Introduction
This report describes a method for deriving Environmental Indicators for urban river assessment.
The method is based upon the following stages:
• Conceptualising the functioning of urban rivers within a spatially hierarchical framework
(Section 2).
• Recognising that urban river stretches affected by different levels of engineering
intervention form the key spatial scale for urban river assessment, but that processes and
forms at other spatial scales are also of importance (Section 2)
• Proposing an Urban River Survey which provides data to characterise the environmental
characteristics of engineered stretches whilst maintaining compatibility with the widelyused
River Habitat Survey (Section 3)
• Evaluating summary indices from URS data to provide quantitative information on
environmental properties of engineered stretches. The values of these summary indices are
Primary Environmental Indicators since they provide a quantitative description of a range
of specific properties of engineered stretches (e.g. bank and bed material calibre, in-channel
and bank vegetation biomass, extent of bank or bed reinforcement of different types) and
thus provide a detailed description of each stretch (Section 4)
• Classifying engineered stretches into a few (6 to 8) classes according to groups of primary
environmental indicators describing channel and bank ‘Materials’; the number and types of
‘Physical Habitats’; the distribution, biomass and type of ‘Vegetation’. ‘Materials’,
‘Physical Habitat’ and ‘Vegetation’ are Secondary Environmental Indicators. The classes
ascribed to each of these environmental indicators can be arranged along a gradient of
diversity, abundance or modification where each class represents a typical combination of
the primary environmental indicators. In some cases the class to which a stretch is allocated
could be directly altered by engineering intervention (e.g. the introduction of increased
reinforcement materials could change the ‘Materials’ class), whereas in other cases stretch
classes may be associated with engineering intervention as a result of more indirect links
between engineering and stretch properties (e.g. the biomass and character of bank
vegetation, the number and types of in-channel physical habitats). If association with
engineering is apparent, then the classifications can be used to consider the likely
consequences of change in channel engineering for the class to which a stretch might be
allocated (Section 5). If there is no direct link with engineering, the class of a stretch can
still be used to assess scenarios of natural recovery or imposed change.
• Incorporating Tertiary Environmental Indicators operating at the sector scale (e.g.
properties of the flow regime) or observed within the same sector as the engineered stretch
(e.g. water quality or biotic indices and scores), and assessing their relative importance in
constraining recovery potential (Section 6). Such indicators allow an assessment of the
potential of a stretch for modification, enhancement or rehabilitation as a result of just
changes in the way it is engineered (e.g. flow energy, water quality, availability of
propagules, flood plain extent and land use).
Environmental Sustainability Indicators for Urban River Management
2
2. Engineered stretches within a spatial
hierarchical framework
Frissell et al.(1986) proposed a hierarchical framework for stream habitat assessment and
classification, based on the assumption that river ecosystems are largely controlled by physical
patterns and processes which interact at a range of spatial scales. The framework had a spatially
nested, hierarchical structure, with five spatial scales of river unit in the hierarchy: stream
network, segment or sector, stretch or reach, pool-riffle, and microhabitat. Small objects, such as
patches of river-bed sediment are set within a framework of intermediate scale units (e.g. poolriffles)
and larger scale units (e.g. sectors of river between tributary confluences). This
hierarchical framework has been adopted in parts of the United States and South Africa as a
basis for river assessment (Beechie and Sibley, 1990; Wadeson and Rowntree, 1994). It is a
robust starting point for designing the spatial structure and sampling regime of new monitoring
programmes, as well as providing a conceptual framework for integrating data from different
sources and for devising river classification schemes.
A spatially hierarchical framework is particularly useful for storing, analysing and classifying
information on urban rivers because the character of urban rivers at all spatial scales is heavily
constrained by the range of engineering works undertaken at different times for a variety of
purposes. An hierarchical framework can be constructed around the engineering modifications
that have been made to urban rivers. Figure 1 illustrates that there are five spatial scales within
the framework that we have devised for urban rivers: the catchment (entire stream network),
sector (major tributaries and unbranched sections of river between tributary junctions), stretch
(river reach exhibiting a single engineering ‘type’), habitat (individual pool, riffle, bar etc.) and
patch. Engineered stretches that reflect differences in the nature and degree of engineering
intervention are the key spatial units to which units at other spatial scales can be linked. Thus
engineered stretches may be aggregated into river sectors and catchment networks, or
subdivided into habitats and patches.
Engineered stretches have a single ‘type’ of engineering intervention based on a combination of
(i) the river planform; (ii) the channel cross section, and (iii) the amount of bank and bed
reinforcement. Tables 1 lists the various subdivisions of these three properties that can be
combined to identify 144 potential engineering ‘types’.
Figure 2 summarises the catchment and local controls on the geomorphology of river stretches
and provides the rationale for defining the scale of the engineering stretch as the key to
understanding and classifying urban river channels. Under natural conditions, the form of a river
channel is determined by interactions between the river flow and sediment transport regimes and
the boundary materials within which the channel is developed. The flow and sediment regimes
of urban rivers are heavily modified by catchment-scale, sector-scale and stretch-scale
influences on hydrological and hydraulic processes, and on the availability of sediment. The
channel margin materials are frequently modified at the stretch-scale as a result of engineering
intervention.
Environmental Sustainability Indicators for Urban River Management
3
Figure 1 A hierarchy of six spatial scales at which urban river data may be collected, stored and
analysed, with examples of the data types that might be collected at each scale
Figure 2 Catchment and local controls on the geomorphology of river stretches (the bold text in
italics refers to factors which can adjust in many rural river channels but which are
frequently fixed by engineering works in urban river channels)
CATCHMENT
Catchment
characteristics
SECTOR
Main tributary
or section of
the main river
between
tributary
junctions
River corridor land use
Water quality data
Hydrological data Channel
morphology
Vegetation
Reinforcement
Materials
Flow Types
Hydraulics
HABITAT
Physical
habitat feature
(riffle; bar etc.)
Biological data
Sediments
Local hydraulic data
STRETCH
Up to 500m in
length of any
single
engineering type
PATCH
Patch of
vegetation or
sediment
Biological data
Sediments
Local hydraulic
data
GEOMORPHOLOGY
Channel Size
Bed and Bank Form
Bed and Bank Sediment Size
and Heterogeneity
Vegetation Pattern and Structure
CATCHMENT PROCESSES
Flow Regime
Sediment Transport Regime
Water Quality Regime
Water and wind-transported
Propagules
SITE CHARACTERISTICS
Slope
Bed and Bank Materials
(including vegetation impact on
hydraulics and material strength)
Locally-derived propagules
PHYSICAL HABITAT
Character
Diversity
Dynamics / Stability
Environmental Sustainability Indicators for Urban River Management
4
This stretch-scale modification places severe constraints on the degree to which the river form
can adjust to variations in river flow and sediment transport, and it suggests that the fundamental
scale for differentiating the nature and diversity of physical habitat within urban rivers is the
river stretch, distinguished by engineering ‘type’.
Table 1 Subdivisions of river channel planform character, cross section character and bed and
bank reinforcement that can be combined to define the engineering type for a reach of
urban channel.
(i) Alterations to the river’s
planform
(ii) Re-engineering of the
channel cross section
(iii) Re-inforcement of the
channel bed and banks
Semi-Natural Semi-Natural No re-inforcement
Straight Restored Bed only
Meandering Cleaned 1 bank only
Recovered Enlarged Bed and 1 bank only
Two-stage Both banks only
Resectioned Full
Because the engineered stretch is the key spatial scale for defining the characteristics of urban
rivers within the hierarchical framework of Figure 1, this report concentrates on a methodology
for defining environmental indicators for urban river stretches. Nevertheless, the position of a
stretch within a sector or the entire river network influences the flow and water quality regime to
which it is subject and also the availability of plant and faunal propagules to colonise the stretch
(Figure 2) and so some sector-scale summary process indicators are also relevant to the
assessment of urban river stretches. It is, therefore, the interaction of stretch (i.e. primary and
secondary) and sector scale (i.e. tertiary) indicators that provide an assessment of the present
state of the stretch, its likely response to changes in management, and its potential for recovery.
Environmental Sustainability Indicators for Urban River Management
5
3. A methodology for surveying urban river
stretches
The majority of river geomorphological and habitat surveys have been devised for the appraisal
of rural rivers (Newson, 2002). Whilst such surveys can be applied to urban rivers, they may not
be sufficiently sensitive to urban river characteristics to produce discrimination between
different urban river reaches. Whereas the dimensions of rural channels closely reflect the
magnitude and frequency of the fluvial processes that they transmit (Wharton, 1995) and the
frequency of geomorphological features, such as pools and riffles are also scaled to the channel
dimensions (e.g. Leopold and Wolman, 1957), this is not often the case for urban channels. Not
only is the channel size frequently a product of channel engineering, but urban channels often
contain artificial structures and materials which have significant hydraulic impacts that control
sediment dynamics and the creation of particular habitat types, such as bars and pools. Thus,
urban channels may not display the number or pattern of physical habitats that are encountered
in less heavily impacted channels. As a result, any physical assessment of urban rivers must
place heavy emphasis on channel engineering. Channels subject to single types of engineering
can range in length from a few metres to several hundred metres, but commonly fall in a range
of 200m to 1km. Since a 500m reach of river is now widely adopted as the basis for many
surveys of rural rivers in the UK, especially for the River Habitat Survey (RHS), surveys of a
standard 500m reach length of a single engineering type are adopted here for surveys of urban
rivers.
3.1 THE RIVER HABITAT SURVEY
The current operational habitat survey technique in the UK is the Environment Agency’s River
Habitat Survey (RHS). Its provides a rapid assessment of physical and hydraulic habitat, riparian
structure, and ecological potential of 500m stretches of river, that can be completed by nonspecialists
following a short training course (Raven et al., 1997). The spatial scale and
widespread usage of the RHS method provides a good foundation for the development of a
survey methodology specifically focussed on urban rivers. Moreover, the usefulness of both
surveys is enhanced if compatibility between them can be maintained.
In brief, the RHS is comprised of 4 basic components: (i) Background Measurements; (ii) Spotcheck
Measurements; (iii) Once-Only Measurements; and (iv) Cumulative Measurements.
Background measurements include the date, time of the survey, grid reference, and general
conditions for the assessment (adverse weather, and channel bed visibility). Properties that relate
the stretch to its catchment, provide a context for the survey and can be derived mainly from
secondary sources (e.g. altitude, geology, distance from source, slope) or a brief assessment in
the field (e.g. valley form) are also recorded.
Spot-check measurements are recorded within 1m wide transects across the channel located
every 50m along the stretch (10 spot checks per 500m stretch), with a ‘catch-all’ column for the
final 50m in the stretch. The attributes associated with each spot-check are assessed by eye from
either the bank or from within the channel, and include the physical attributes of the channel
(channel substrate, bank materials, in-stream features such as bars, flow types, and forms and
modifications of the channel and banks), in-channel macrophytes, the bank vegetation in terms
of its complexity, and immediate land use (5m from the bank top),
Once only measurements are assessed once within the stretch. They include bank and channel
width, water depth, bank top and bank full height, and embanked height.
Cumulative measurements comprise all of the measurements contained within the RHS ‘sweepup’
section. A continuous assessment is made along the 500m stretch and a single recording
made at the end of the survey. These attributes include the presence of trees and their associated
features, bank profile types, land use, channel features, artificial features, special features and
Environmental Sustainability Indicators for Urban River Management
6
management attributes. A total of 12 separate categories are evaluated, comprising some 80
different measurements.
3.2 THE URBAN RIVER SURVEY (URS)
An important feature of the RHS is its success in harmonising the wide variety of data to be
collected into a simple rapid survey. The basic structure of the survey and the definitions of the
variables are maintained within the Urban River Survey (URS). The variables contained within
each section of the URS may differ from the RHS, however, reflecting the differences between
required for a survey focussing specifically upon urban channels. Much emphasis in the RHS is
placed on the frequency of habitat features which are assessed using the spot-check
measurements. Emphasis is also placed on artificial features found within the stretch, and the
associated land use. In an urban context a different level of importance needs to be placed on
these categories. For example, in urban rivers, the channel has often been heavily modified for
flood defence, thus importance needs to be placed on characterising the habitat features
associated with such modifications and with processes of recovery. Geomorphological features
are often infrequent or lacking in urban channels, and their presence and frequency are,
therefore, better represented as counts along the entire stretch (cumulative measurements) rather
than by regularly spaced spot-checks. Artificial features also require a more detailed
characterisation both in terms of their frequency and extent, especially measurements associated
with the reinforcement of the channel bed and banks. Furthermore, water quality requires a more
detailed assessment in urban channels to cover the range of water quality problems that might
occur. These problems can be recorded using easily identifiable indicators such water colour,
turbidity, algal growth or smell. Building on the current RHS, the types of variables within each
of the four specific groups that need to be assessed within the URS are listed in Table 2, and are
developed in greater detail below.
The URS methodology is designed to be compatible with the RHS, and is therefore equally
simple to complete. Complex stretches of urban channel such as those located within
rehabilitation schemes, and even stretches that remain unmodified can be surveyed within the
average one hour time span that the RHS methodology advocates, while heavily engineered or
extremely uniform stretches may take as little as 20 minutes to complete. However, the survey
has been designed to be sensitive enough to recognise even small changes in channel form,
which may be important in the hostile conditions that exist within urban channels. Each
component of the URS is described and justified below, referring to equivalent measures in the
RHS where necessary.
Table 2 Groups of variables that must be considered for incorporation into an URS
Geomorphological
Variables
Hydraulic Variables Ecological Variables Other Variables
Channel Substrate
Materials
Flow Type Bank face Structure Land use (5m and 50m from
the bank)
Bank Materials Channel Dimensions Bank top structure Water and Sediment odours
Channel
Dimensions
Habitat features Macrophyte type and
amount
Oils
Bank profiles Special features Trees and associated
features
Surface scum
Bank Protection
type/amount
Species present Gross pollution
Habitat Features Nuisance species
present
Clarity
Special features Alders/Diseased alders
present
Number of input pipes
Evidence of Recent
Management
Number of leach points
Evidence of Recent
Management
Environmental Sustainability Indicators for Urban River Management
7
3.3 BACKGROUND MEASUREMENTS
Background measurements incorporate the variables that (i) relate the stretch to its catchment
and to the river sector within which it is located; (ii) are survey-specific, placing it in a temporal
context, and (iii) define the general character of the stretch against which the detailed attributes
of the river channel and riparian corridor can be placed (Table 3). Several of the variables enable
the surveyed stretch to be placed in a hierarchical catchment framework, as advocated by Frissell
et al, (1986). The Hydrocatchment Identification Number (HC ID) relates the stretch to the
catchment and the sector code relates the stretch to the network sector in which it lies. These are
complemented by the river name and central grid reference.
Variables that define survey specific details (the date and time of the survey, the surveyor name
and RHS accreditation number, and the location from which the survey is made, i.e. bank or
channel), conditions at the time of survey (adverse conditions and bed visibility), and give a
pictorial reference for the stretch (photographs taken) are retained from the RHS. The remaining
variables describe the general character of the stretch including the engineering type, and indices
of river quality that are generally applied by the Environment Agency. The latter include the
General Quality Assessment (GQA) chemical quality and biological quality (Nixon et al., 1996).
The latter uses the RIVPACS programme (Wright et al., 1993) to predict target values of the
faunal parameters number of taxa, and BMWP score (Biological Monitoring Working Party) and
ASPT (Average Score Per Taxon) that can be compared with observed values. BMWP scores
are no longer used by the Environment Agency to calculate the GQA grading of rivers, but have
proved a useful descriptor of biological quality (Hawkes, 1997).
Table 3 Variables included in the URS Background Information and their compatibility with
those of the RHS methodology
RHS COMPATIBLE VARIABLES URS SPECIFIC VARIABLES
River Name Hydrocatchment ID Number (from EA)
Central Grid Reference Sector Code
Surveyor Name Stretch ID Code
Accreditation Number (from RHS course) Stretch Name
Date of survey Stretch Engineering Type
Time of survey Observed BMWP Score
Adverse Conditions Affecting Results Predicted BMWP Score
Bed Visible Observed ASPT Score
Photographs Taken Predicted ASPT Score
Site Surveyed From (Bank/channel) RHS Data Available?
Distance from Source (km) GQA Biological Quality Class
GQA Water Quality Class
Solid Geology Code
Drift Geology Code
3.4 SPOT-CHECK MEASUREMENTS
Spot-check measurements when combined with a final 50m sweep up category represent the
frequency and pattern of the features found within the river channel. Table 4 compares the
properties recorded within the RHS and URS. The key changes to the RHS methodology are
found in detailing the physical characteristics of the channel at each spot-check. Bank protection
in urban rivers is a fundamental component of the channel structure. The frequency of different
types of protection, and the mosaic of types found along each bank greatly influence flow
hydraulics and the type of habitats found within the stretch. Furthermore, the composition of the
bank material influences the durability of each type of protection. For example, gabions placed
in a predominantly sand bank material may be washed out through erosion at a faster rate than
gabions placed in a more cohesive bank. The URS therefore, records the underlying ‘natural’
bank materials in a separate category using the classes of sediment calibre adopted in the RHS
(e.g. cobble, gravel/sand, clay etc.), while the bank protection is recorded using descriptors
derived specifically for the URS (e.g. gabions, rip-rap, sheet piling etc.). The measurements of
Environmental Sustainability Indicators for Urban River Management
8
bank and channel modifications and features recorded in the RHS, have been omitted in the URS
spot-check measurements, and included in the cumulative measurements.
Both bank modifications and some channel modifications are implicit in the definition of the
urban stretch and, therefore, inclusion of these attributes within the spot-check section of the
survey are unnecessary. Other features recorded by the RHS, such as dams and fords, are
generally absent from urban channels, where the engineered function of the river is to transport
large amounts of water away from the urban area as quickly as possible. Nevertheless, it is
important to include all types of modification within the survey and for this reason these
attributes are recorded in the cumulative measurements. ‘Natural’ bank and channel features
such as bars and eroding cliffs, while important features of urban channels, are relatively
infrequent, and the collection of data on these attributes is best served by an overall assessment
of the stretch rather than by a regularly spaced assessment of the channel. Thus, a combination
of the stretch definition, which summarises its engineering type, together with a series of
cumulative measures rather than spot-checks provide an appropriate summary of bank and
channel features and modifications within the URS.
Table 4 Comparison of spot-check variables between RHS and URS methodologies
RHS SPOT-CHECK PARAMETERS URS SPOT-CHECK PARAMETERS
Bank Materials Bank Materials
Bank Modifications Bank Protection
Bank Features
Channel Substrate Channel Substrate
Flow Type Flow Type
Channel Modifications
Channel Features
Bank Top structure Bank Top Structure
Bank Face Structure Bank Face Structure
Bank Top Land Use (5m) Bank Top Land Use (5m)
Channel Vegetation Channel Vegetation
Channel substrate is an important component of urban rivers, especially where artificial
substrates have been placed within the channel. The measurement of this attribute is similar to
the RHS methodology, with the categories of channel substrate being retained within the URS.
Where artificial substrates occur, however, the presence of mobile substrates overlaying
artificial materials is also recorded. This is important for evaluating the channel’s capacity for
forming features such as riffles and bars despite the rigid bed reinforcement, which ultimately
may affect the ecological diversity of the channel. Measurements of flow type, bank face and
bank top structure remain unchanged within the URS methodology.
The dominant land use on the bank top is also recorded in the URS. The RHS methodology
categorises the typical land use types found across a range of river environments and classifies
urban and suburban development as a single homogenous category. However, the urban
environment is not a single expanse of land development but a complex mix of fragmented
‘natural’ land cover types set within and between different types of urban development.
Moreover, different land use types, even in urban areas, affect stretches of rivers differently (i.e.
industrial, residential, parkland). Any measurements of land use must therefore reflect this
heterogeneity. The URS, therefore adopts a two-tiered classification of land-use proposed by
Anderson et al. (1976) and modified by Meador et al. (1993) (Table 5). Level 1 uses entirely
remotely sensed data to categorise the land into 6 broad categories: Urban, Agricultural,
Rangeland, Forest land, Wetland, and Barren land, and is typically applied at a catchment scale.
These broad categories are then subdivided into 21 Level 2 land use types (Table 5) that can be
assessed either by using aerial photographs or during the field survey of river stretches. The
URS also records the available floodplain width, which is the width of any open land use
adjacent to the channel.
The final attribute measured in this section of the URS survey is that of channel vegetation.
Channel vegetation in the urban environment is important for three reasons. Firstly, the diversity
Environmental Sustainability Indicators for Urban River Management
9
of the aquatic macrophytes is important for the ecological integrity of the system. Secondly,
channel vegetation affects flow patterns, and excessive growth of some types of macrophyte,
such as submerged fine leaved species (e.g. Potamogeton pectinatis) can affect channel
conveyance, creating problems for flood management. Thirdly, excessive growth of macrophyte
types such as filamentous algae induce marked diurnal fluctuations of dissolved oxygen within
the water column (Pitcairn and Hawkes, 1973; Kirk, 1994), affecting the ecology of the system.
The RHS method for measuring channel vegetation has been retained within the URS, but with
one important change, the extent of each vegetation type is recorded, reflecting the importance
of excessive macrophyte and algal growth (Table 6). An additional category that is critical to the
ecological interpretation of the urban river is the explicit recording of non-visible channels in
comparison with those that possess no vegetation cover. Modified channels frequently possess
high concentrations of suspended material, especially after rainfall, which reduce the visibility of
the channel bed, whilst low levels of shading combined with increased nutrient inputs from
sewage effluents increase the extent of aquatic macrophytes. It is important, therefore, to make
the distinction between no vegetation cover, and vegetation that is not visible. It is also
important to note that although the macrophyte type ‘Filamentous Algae’ is not strictly a
macrophyte, algal species such as Cladophora are a characteristic indicator of organic pollution
and may grow to lengths of up to 10m (Pitcairn and Hawkes, 1973). Therefore, they are included
in the appraisal of channel vegetation.
3.5 ONCE-ONLY MEASUREMENTS
The channel dimensions assessed in the RHS have been retained within the URS (i.e. bankfull
width, water width, water depth, banktop height, embanked height, trashline height, and location
of measurement), and an additional measurement of the amount of macrophyte cover at this
point within the stretch is included for completeness. Channel dimensions are important
properties of urban rivers. Although these dimensions may not always be able to adjust freely in
response to fluvial processes in urban rivers, they nevertheless impact on the geomorphological
features that are found within the channel. For example, depositional berms and marginal bars
might be expected in overwidened channels, whilst reinforced straightened, or narrow,
overdeepened channels, might produce fewer geomorphological features of a coarser sediment
calibre than natural channels with a similar flow and sediment transport regime.
Environmental Sustainability Indicators for Urban River Management
10
Table 5 Land use types used to describe land use at the catchment scale (Level 1) and finer spatial
scales (Level 2)
Level 1 Land Use Codes Level 2 Land Use Codes Description
UR (Urban) Re Residential
Cm Commercial
In Industrial
Ic Industrial/Commercial
Tr Transport
Sw Sewage Treatment Works
Ld Landfill/Refuse Deposits
Dr Derelict Land
Cn Contaminated land
AG (Agricultural) Cr Cropland
Pa Pasture
Or Orchard
Fe Close Feeding (Battery Farms etc.)
FO (Forested) Co Coniferous
Dd Deciduous
Ow Open Woodland
PA (Pasture) He Heathland
Sc Scrub
Op Open Parkland (Community Grass
etc.)
Rc Recreational Land (Playing Fields)
Ce Cemeteries/Crematoria
Es Estate Lands (Inc. MOD)
OW (Open Water) La Lake
Rv Reservoir
Ca Canal
Rq Reclaimed Quarry
Tb Tributary
WE (Wetland) Fo Forested
Nf Non-forested
BA (Bare) Sm Strip mines/Open Cast
Ex Exposed Rock
Tn Transitional
Environmental Sustainability Indicators for Urban River Management
11
Table 6 URS channel vegetation types and example species (Adapted from EA, 1997).
RHS Channel Vegetation Type Macrophyte Code Example Species
None NON No vegetation present
Non Visible Channel NVC Channel bed not visible
Mosses/liverworts/ lichens LML Exposed or submerged
Emergent broad leaved herbs EBH Apium spp; Rorippa spp
Emergent reeds/sedges/rushes RSR Sparganium erectum; Glyceria maxima;
Schoenoplectus; Typha; Phragmites; Juncus
spp; Carex spp
Floating leaved (rooted) RFL Nuphar lutea; Potamogeton natans;
Sparganium emersum
Free-floating FFL Lemna spp; Hydrocharis; Ceratophyllum;
Stratiotes.
Amphibious AMP Polygonum amphibium; Agrostis
stolonifera; Glyceria fluitans; Alopecurus
geniculatus; Myosotis scorpiodes.
Submerged Broad-leaved SBL Nuphar spp; Elodea spp; Callitriche spp.
Submerged linear-leaved SLL Sparganium erectum; Butomus umbellatus;
Typha spp; Sagittaria sagittifolia
Submerged fine-leaved SFL Ranunculus spp; Myriophyllum spp;
Ceratophyllum spp.
Filamentous algae FAL Cladophora; Enteromorpha
3.6 CUMULATIVE MEASUREMENTS
The attributes included in this section of the survey are intended to provide an overall impression
of the quality of the stretch, and how well the channel may be recovering from past
modifications (Table 7). Within the urban context, quality can be determined by the diversity of
the channel morphology, the vegetational structure of the riparian zone, the level of water
pollution, and the recovery potential of the stretch.
Pollution exerts influence over river ecology through direct means such as toxic chemicals and
leachates, or indirectly by degradation of potential habitats for biota. The RHS methodology
limits the identification of pollution to a simple presence or absence within the stretch. Within
the urban river, however, pollution is an important consideration where sewage effluent is often
a primary component of the river’s base flow, and industrial effluents and runoff from roads are
also frequently major water quality impacts. To address the increased potential for pollution in
urban river channels, eight pollution characteristics are recorded (Table 8). The first five of these
measures are assessed on an Absent/Present/Extensive (APE) scale, where extensive relates to
more than 33% of the stretch being affected by a particular pollution type. Clarity of the water is
assessed as being good (water is clear and channel substrate is clearly visible), poor (the channel
substrate is not visible due to high turbidity) or average (where the clarity of the water falls
between these two extremes). This measurement is partly dependant upon discharge. However,
the survey should be carried out under dry conditions when water levels are ‘normal’, thereby
reducing adverse effects on the clarity of the water through increased water flows. The final two
pollution measures (number of input pipes and number of leach points) are assessed as a total
count of each within the stretch. The number of input pipes within a stretch serves to identify
likely points of pollution pulses characteristic of urban rivers, while the leach points identify
more diffuse pollution that may be important in the general water quality of the river.
Environmental Sustainability Indicators for Urban River Management
12
Table 7 Comparison of attributes assessed in the cumulative measurements within the RHS and
URS surveys.
RHS CUMULATIVE MEASUREMENTS URS CUMULATIVE MEASUREMENTS
Land use (within 50m of bank top) Land use (within 50m of bank top)
Bank Profiles Bank Profiles
Trees and associated features Trees and Associated Features
Channel Features Habitat Features
Recent Management Recent Management
Features of Special Interest Features of Special Interest
Choked Channel Choked Channel
Nuisance Plant Species Nuisance Plant Species
Alders Alders
Overall Characteristics
Number of Riffles, Pools and Point bars
Artificial Features
Wildlife Species Present
Extent of Pollution
Bank Protection
Other Information
Other measures of quality relate to the riparian structure of the channel. Riparian structure is
particularly important in the urban environment where rivers may act as wildlife corridors
(Goode, 1989) and so the RHS assessment of trees within the stretch has been retained within
the URS. Measurements describing the structure of the bank top and face, overlying vegetation
(such as trees) and the recent management of the riparian zone give an integrated impression of
the riparian structure of the stretch being surveyed, that is fundamental to its ecological
potential. Trees are often not present along modified channels, where appropriate substrates for
tree growth are frequently limited. Furthermore, trees may be removed from stretches where
they are perceived to be a significant contributing factor to flooding. Their presence, however, is
important where shading reduces macrophyte growth within the channel, and also for providing
cover for both aquatic and riparian species.
Nuisance plant species are another major problem in urban environments where frequent
disturbance of the banks and surrounding corridors provide ideal habitats for species such as
Himalayan balsam and Japanese knotweed, allowing them to out-compete native vegetation and
degrade the riparian zone. The RHS method of recording nuisance species using a
presence/absence measure is expanded to incorporate a simple cover scale within the URS, to
reflect the increased extent and potential importance of these species: Absent; Single Individual
(a single plant within the stretch); Isolated Clumps (a few small clusters of plants within the
stretch); Frequent (present in 25-33%of the stretch); Extensive (>33% of the stretch).
Environmental Sustainability Indicators for Urban River Management
13
Table 8 Types of pollution recorded in the URS.
Pollution Type Description
Water Odours Typically refers to the classic sewage effluent odours, but may also
include industrial chemical aromas such as ammonia. Especially
important where the pollutant is colourless.
Sediment Odours Describes the characteristic odour emitted by anoxic sediments, and
can easily be tested by inserting a ranging pole through the surface
of the sediments.
Oils Can be extensive in urban channels where surface runoff from roads
is a major source of pollutants, and is characteristically seen floating
on the water surface, or released from toxic sediments during testing
for sediment odours.
Surface scum Consists of foams caused by the presence of phosphate detergents
during surface mixing. It is usually seen by sewage outfalls, but may
also refer to floating mats of small particles of debris and thin foams
forming in slow flowing waters.
Gross pollution A characteristic of urban channels and incorporates larger items of
urban trash including shopping trolleys, mechanical parts, and litter
Clarity Primarily assessing the level of suspended materials, but may also
include the discharge of coloured effluents
Number of input pipes These include outfalls, land drainage pipes and small industrial
outfalls.
Number of leach points Characteristic of drainage from contaminated land. The leachate may
contain ferric matter which can be readily identified by its orange
colour.
Other measures of quality include recent management, wildlife species present, alders/diseased
alders present (required for the national assessment of the incidence of Phytophthora root
disease – Environment Agency, 1997), choked channel and other information (i.e. presence of
weirs etc). These are recorded as presence/absence measurements in an identical manner to the
RHS methodology. Land use 50m from the bank top is also recorded in the URS, but using the
Level 2 land use types described in Table 5. Furthermore, land use for each bank is also
recorded as a percentage cover to provide greater resolution in the assessment of the riparian
structure and quality in the urban channel.
The presence of habitat features (equivalent to the RHS channel features) can be useful in the
assessment of both urban channel quality, and urban channel recovery. Habitat features can be
grouped into two distinct types: flow habitats or flow types (e.g. riffles, runs), and physical
features such as bars (Table 9). Flow types are the surface expression of three-dimensional flow
structures, and are associated with characteristic circulation patterns, ranges of flow velocities
and bed forms within the river. The RHS methodology recognises ten categories of flow type,
ranging from free fall to no flow (dry bed). It is important to assess their extent in urban
channels, where even the smallest amount of variation in flow type may provide enough refugia
for fauna to successfully inhabit a relatively hostile environment. Physical habitat features, such
as riffles, pools and bars, are both an influence on and a result of hydraulic factors and are,
therefore, important in both rural and urban habitat surveys as their varied morphological,
sedimentological and hydraulic characteristics define the mosaic of physical habitats seen at the
stretch scale. The RHS methodology assesses the presence of habitat features on an APE scale.
However, some types of engineered stretch may produce a relatively homogenous channel in
terms of its habitats, and the presence of even small amounts of variation may be sufficient to
increase the ecological quality of the channel. Therefore, a more accurate assessment of the
frequency of these habitats is required for the URS. To this end, the flow habitats are measured
as a percentage of the stretch, whereas other habitat features are recorded as a total count of each
type present within the stretch. The presence of special features such as open waters, and
Environmental Sustainability Indicators for Urban River Management
14
adjacent wetland types such as bogs and fens are recorded in the URS in an identical manner to
the RHS
Table 9 Categories of habitat features recorded in the URS.
FLOW HABITATS
RECORDED AS % OF
STRETCH
OTHER HABITATS RECORDED AS
TOTAL NUMBERS IN STRETCH
Cascade Exposed bedrock
Rapid Rock/boulder
Riffle Waterfall
Run Backwater
Boil Sand/Silt deposits
Glide Mature Island
Pool Unvegetated mid-channel bar
Ponded Reach Vegetated mid-channel bar
Marginal Deadwater Unvegetated point bar
Stagnant water Vegetated point bar
Unvegetated side bar
Vegetated side bar
Woody debris
Other measures of channel heterogeneity and recovery are also included in the cumulative
measurements. The amount of each bank protection type is recorded as a percentage of the
stretch. This expands upon the URS spot-check measurements which can be used to describe the
mosaic of protection types along the stretch. When the two measures (spot-check and
cumulative) are combined they can be used to assess which types of protection are more
important in the urban channel. Bank profiles are of fundamental importance for the assessment
of channel recovery through erosion and sediment deposition. Two different types of bank
profile can be present in a stretch, namely natural and artificial profiles. Artificial bank profiles
are particularly significant in urban channels since they provide particular riparian habitats and
offer characteristic controls on flow hydraulics. However, ‘natural’ features of recovery, which
are incorporated in the RHS, such as eroding banks are also important in the urban channel, as
are natural components of the bank profile such as undercutting of the bank toe, which provide
refugia during spate flows, and may also be indicative of the onset of recovery processes in
highly modified stretches. Recovery processes allow natural profiles to become superimposed
upon the artificial profiles. Artificial and natural bank profiles are, therefore, grouped separately
within the URS, and each bank profile type within these two groups is recorded as a percentage
of the stretch, rather than the APE scale used in the RHS. This allows even small amounts of
recovery, for example through undercutting of the bank toe, to be incorporated into the survey,
whilst still maintaining a reliable assessment of the actual level of modification.
Environmental Sustainability Indicators for Urban River Management
15
4. Aggregate indices from URS data (primary
environmental indicators)
Many aggregate indices can be developed using the URS data and can mainly be attributed to
one of three groups, which describe ‘Materials’, ‘Physical Habitat’ and ‘Vegetation’ features
(Table 10). The derivation of each index is described in this section, where each index is
constrained to have a similar numerical range (typically 0 to 10 or, –9 to +9).
Table 10 Synthetic Indices derived from the Urban River Survey relating to three different sets of
characteristics of urban river stretches: ‘Materials’, ‘Physical Habitat’ and ‘Vegetation’.
MATERIALS PHYSICAL HABITAT VEGETATION
Proportion Immobile Substrate Number of Flow Types No. Channel Vegetation Types
SEDCAL Dominant Flow Type Channel Vegetation Cover
Proportion Immobile Left Bank
Materials
Number Natural Bank Profiles Dominant Channel Vegetation
Type
BANKCAL (left bank) Proportion Natural Bank Profiles Total Tree Score
Proportion Immobile Right Bank
Materials
Number Artificial Bank Profiles Total Tree Feature Score
BANKCAL (right bank) Proportion Artificial Bank Profiles BANKVEG (left top)
BANKPROT (left bank) Number of Habitat Types BANKVEG (left face)
BANKPROT (right bank) BANKVEG (right top)
Proportion No Bank Protection
(NONE)
BANKVEG (right face)
Proportion Biodegradable
Protection (BIO)
Proportion Open Matrix
Protection (OMP)
Proportion Solid Protection
(SOL)
4.1 INDICES DESCRIBING MATERIALS
Channel Substrate: The channel substrate was separated into two components: immobile
materials (concrete, brick, and bedrock) and mobile materials (silt, sand, gravel etc). The URS
records the predominant mobile substrate at each spot check (10 cross sections along a 500m
stretch) according to categories compatible with the Wentworth particle size scale. The
SEDCAL index converts these spot-check measurements into an approximate average particle
size for the stretch in phi units.
Environmental Sustainability Indicators for Urban River Management
16
SEDCAL = (-8*BO) + (-7*CO) + (-3.5*GP) + (1.5*SA) +(1.5*SI) + (9*CL)
(BO + CO + GP + SA + SI + CL)
(BO=boulder; CO=cobble; GP=gravel-pebble; SA=sand; SI=silt; CL=clay)
An index of the proportion of the stretch with bed reinforcement is:
Proportion Immobile Substrate = number of spot-checks with immobile materials x 10
number of spot-checks
Bank Materials: Since data are gathered separately for the two river banks in the URS, the
synthetic indices were also estimated for both banks, although they could also be combined to
give a stretch summary.
The URS records similar measurements for mobile bank materials as for the channel substrate
based upon the Wentworth scale. The BANKCAL index converts these spot-check
measurements into an approximate average particle size for the stretch banks in phi units
BANKCAL = (-9*BO) + (-8*CO) + (-2.5*GS) + (4*EA) + (9*CL)
(BO+CO+GS+EA+CL)
(EA=earth)
The proportion of Immobile Bank Materials (concrete, concrete and brick, laid stone, sheet
piling, and bedrock) is calculated in the same way as for the Immobile Substrate.
Bank Protection: The various types of protection used in urban channels can be placed into
different categories according to their attributes, and then be ascribed a numerical value relating
to their durability and permeability (Table 11).
Table 11 Bank Protection Types
CATEGORY BANK PROTECTION TYPES NUMERICAL
VALUE
None None; Washed Out 0
Biodegradable Reeds; Wood piling; Willow spiling 1
Open Matrix Rip-rap; Gabions; Builders waste 2
Solid Concrete; Concrete and Brick; Brick/Laid stone; Sheet piling 3
The numerical value given to each category can then be used to calculate the level of protection
for each bank.
BANKPROT = (0*NONE) + (1*BIO) + (2*OMP) + (3*SOL) x 3
(NONE+BIO+OPM+SOL)
The overall type and level of protection for the stretch can be taken from the URS cumulative
measurements. The types of protection are grouped into the four categories of none,
biodegradable, open matrix and solid, and the proportion of the stretch that these types represent
is then calculated. From this the Proportion of No Bank Protection (NONE), Biodegradable
Protection (BIO), Open Matrix Protection (OMP) and Solid Protection (SOL) can be
estimated.
4.2 INDICES DESCRIBING PHYSICAL HABITAT FEATURES
Flow Types: The water surface pattern of flow types reflects the three-dimensional flow patterns
induced by the form and roughness of the channel, and is therefore an important indicator of
flow hydraulics and channel bed morphology.
Environmental Sustainability Indicators for Urban River Management
17
Two indices help to characterise this hydraulic and morphological diversity. Firstly, the
dominant flow type gives an indication of the general character of the stretch. This can be
easily determined from the spot-check measurements, by selecting the flow type which is
recorded the most times. Where two categories occur with equal frequency, the flow habitats
recorded in the cumulative measurements can be used to determine the dominant flow type
within the stretch. The flow types in part reflect the flow velocity, and so the dominant flow type
can be arranged along a flow velocity gradient (Table 12) from faster flow types (index value 1)
to slower flow types (index value 10). Secondly, the number of flow types within a stretch is
important for looking at hydraulic and bed form variability, and can be ascertained by a count of
the number of different flow types recorded in the spot-checks. Together these indices give an
indication of the heterogeneity of the stretch in terms of its hydraulics, and associated channel
morphology.
Table 12 Index Values for the dominant flow type within the stretch.
Flow Type Index Value
Free Fall (FF) 1
Chute Flow (CH) 2
Chaotic Flow (CF) 3
Broken Standing Waves (BW) 4
Unbroken Standing Waves (UW) 5
Rippled (RP) 6
Smooth (SM) 7
Upwelling (UP) 8
No Perceptible Flow (NP) 9
Dry Channel (DR) 10
Habitat features: While the nature and extent of habitats within a stretch can be drawn from the
raw URS data, a simple count of the number of different habitat types observed within the
stretch (not the total number of habitats) provides a simple, integrative index of the diversity of
habitats that are present. This integrative index represents a count of in-channel habitat types,
including both the morphological (e.g. bars, islands, riffles, pools etc.), and hydraulic (flow type)
habitats that are present.
Bank Profiles: The URS recognises two different categories of bank profiles, artificial and
natural, reflecting the historical management practises and the level of bank profile recovery
from past modification. Where the urban channel shows evidence of recovery processes through
erosion, natural profile components become superimposed on artificial profiles, giving a total of
observed profiles of over 100%. Similarly, where an urban channel displays two different types
of modification (e.g. two stage channel and reinforced banks), the total proportion of artificial
profiles can exceed 100%. It is important to distinguish channels that show evidence of
recovery, in order to explore the effects that different types of engineering may have on the
urban channel. To this end, separate indices are developed for natural and artificial bank
profiles. The number of natural profiles and the number of artificial profiles comprise two
of the indices from this group of measurements, which can be ascertained from the cumulative
measurements. This gives an impression of the heterogeneity of the channel in terms of its bank
characteristics, and provides an indication of the processes involved in the recovery of the
channel. The proportion of artificial profiles, and the proportion of natural profiles comprise
the final two indices to come from this group of measurements. The quantitative nature of these
measurements allows a simple index to be derived by dividing the recorded percentage by 10.
4.3 INDICES DESCRIBING VEGETATION STRUCTURE AND BIOMASS
Bank Face and Top Structure: One of the characteristics of urban channels is the uniformity of
the bank in terms of its vegetation. To reduce the roughness of the channel, tall vegetation tends
to be removed, thus increasing the capacity of the channel. The URS records the vegetation
Environmental Sustainability Indicators for Urban River Management
18
structure in the same way as the RHS (B = bare, U = uniform, S = simple and C = complex). The
spot check measurements can be combined into a simple index of the bank vegetation, where
higher values represent increased vegetation complexity. The engineered channel may display
different complexities of vegetation between the banks, and between the top and face of the
bank. Therefore a calculation is made for each bank top and face structure using the following
calculation:
BANKVEG = (0*B) + (1*U) + (2*S) + (3*C) x3
(B+U+S+C)
The presence of trees is calculated on a scale of absent to continuous for the entire length of the
stretch river banks, and the right and left bank can be added together to give a representative
index of cover or total tree score for the whole stretch (None = 0, Isolated/Scattered = 1,
Regularly spaced = 2, Occasional Clumps = 3, Semi-continuous = 4, Continuous = 5).
Tree features (shading of channel, overhanging boughs, exposed bankside roots, underwater tree
roots, fallen trees, coarse woody debris) represent the degree to which marginal trees directly
influence the river channel environment. They are measured on the APE scale, where Absent,
Present, and Extensive score 0,1 and 2 respectively. The scores for each of the 6 features are
added together to give an index of tree influence along the stretch called the total tree feature
score. Combined, these two scores represent the extent and level of impact of the different tree
associated features.
Channel vegetation: To fully assess the nature of the macrophytes within the channel, the
measurements taken in the spot-checks can be separated into 3 important components.
The number of channel vegetation types can be used to indicate the macrophyte diversity,
which in turn can help to indicated water quality. The number of vegetation types consists of a
simple count of the number of different macrophyte groups over the stretch. The number of
types could therefore be as high as 10 in stretches that possess high species numbers.
The dominant channel vegetation type can be easily determined by a simple addition of the
percentage cover each species has at each spot-check. However, in order to represent this
numerically for future analysis, the vegetation types must be ranked according to the properties
they possess, or the effect they have on urban channels. In terms of management, each
macrophyte type will have a greater or lesser effect on the attenuation of flow in the channel.
Thus the macrophyte types can be arranged on a linear scale, from low to high, according to
their potential effects on the hydraulic regime of the stretch (Figure 3). Where non-visible
channels are recorded two possible options exist to explore the data fully. Where a complete
spot-check is recorded as being non-visible the measurement is excluded from the result and the
number of spot-checks used is reduced. Where partial measurements of channel vegetation have
been recorded and the rest of the channel is not-visible at any individual spot-check, the
vegetation measurements are extrapolated to represent the entire channel at that spot-check.
The average cover of the channel combined with the dominant vegetation type, can be used to
assess diversity, but may also help to highlight areas where the management of macrophytes is
important. The calculation of the average channel vegetation cover is taken for the entire
stretch. Again where non visible channels are present, the same principles apply for assessing the
entire stretch as for calculating the dominant macrophyte species. Once this has been achieved, a
simple average is taken of all the macrophyte types to give the % cover for the stretch. To
maintain the same range as other indices, the percentage result is divided by a factor of 10. It is
reasonable to hypothesise that different engineering types will promote different levels of
channel cover, and diversity in terms of the macrophyte growth, although this might be
confounded by the amount of shading present at different stretches.
Environmental Sustainability Indicators for Urban River Management
19
Low attenuation of flow 0 = NON none
1 = LML Liverworts / mosses / lichens
2 = FFL Free-floating
3 = AMP Amphibious
4 = EBH Emergent broadleaved herbs
5 = FAL Filamentous algae
6 = RFL Floating leaved (rooted)
7 = SLL Submerged linear-leaved
8 = SBL Submerged broadleaved
9 = SFL Submerged fineleaved
High attenuation of flow 10 = RSR Emergent reeds/sedges/rushes
Figure 3 Macrophyte types grouped according to their perceived attenuation of flow in urban
channels.
The pollution measurements can be divided into three different indices. The total pollution
score can be calculate from the first 5 variables listed in Table 8, which are measured on the
APE scale, and the clarity which can be assigned a similar score where good, average and poor
score 0,1 and 2 respectively: the higher the score, the higher the extent of pollution within the
stretch.
The number of leach points and the number input pipes comprise the other 2 indices for
pollution. The total number of each that is measured in the URS (Table 8) is converted into a
score as follows:
Number of Score
pipes
0 = 0
1 = 1
2 = 2
3 = 3
4 = 4
5 = 5
6-9 = 6
10-14 = 7
15-20 = 8
20-30 = 9
>30 = 10
Environmental Sustainability Indicators for Urban River Management
20
5. Classification of urban river stretches
(secondary environmental indicators)
A set of secondary environmental indicators have been devised by analysing URS surveys of
106 stretches of the river Tame catchment. The surveys were undertaken during summer 2003,
and the secondary environmental indicators are derived by classifying groups of the primary
indicators described in section 4. The classifications and resultant secondary environmental
indicators presented in this section represent the refined versions of classifications that were
developed previously by Davenport et al (in press) from a smaller dataset.
5.1 DATA ANALYSIS
Cluster analysis was applied to the derived aggregate indices (primary environmental indicators)
that are described in section 4 and listed in Table 10. Because of the similar numerical range in
these aggregate indices, cluster analysis was applied to the untransformed data. Davenport et al
(in press) tested various clustering algorithms (within-group average linkage, between-group
average linkage, centroid, and Ward’s algorithm) using a more restricted data set based on
applying the URS to stretches of the River Tame. Ward’s clustering algorithm was finally
selected because it produced distinct, compact clusters of similar size conforming to the view
that the algorithm generates ‘the most appealing overall results in terms of cluster size, shape
(compactness), density and internal homogeneity’ (Griffith and Amrhein 1997, p220).
Therefore, Ward’s algorithm was used in the present analyses.
Once each cluster analysis was complete, and a dendrogram describing the hierarchical
agglomeration of the objects produced, the identification of the number of clusters or classes that
best described the data was, inevitably, somewhat subjective. The dendrogram was inspected to
identify agglomerations to between 3 to 8 clusters and the number of clusters selected within this
range was based on the generation of the most clearly defined groups within the dendrogram and
the degree to which the clusters had an interpretable meaning.
The validity and meaning of the clusters was assessed by (i) applying non-parametric (Kruskal
Wallace) analysis of variance (ANOVA) to identify which of the individual attributes provided a
statistically significant (P<0.05) discrimination between the clusters; (ii) inspecting box and
whisker plots for each of the discriminatory attributes to identify which clusters were
discriminated by each attribute and the strength of the discrimination; and (iii) identifying
whether the clusters were comprised of any distinct engineering types, which might suggest a
causal impact of engineering on cluster characteristics.
Using the indices listed in Table 10, cluster analysis was applied separately to stretch scores on
the ‘Materials’, ‘Physical Habitat’ and ‘Vegetation’ aggregate indices. The aim was to identify
groupings within the data that could be arranged along a gradient reflecting increasing
complexity or naturalness in relation to these three secondary environmental indicators, so that
the defined classes could represent ‘intensities’ of the ‘Materials’, ‘Physical Habitat’ and
‘Vegetation’ Environmental Indicators. The clusters or classes to which each surveyed stretch
was allocated for each of these three environmental indicators was also compared with the type
of engineering that had been applied to assess whether there were links that would support the
direct modelling of scenarios of changed engineering modification on stretch characteristics or
whether the defined clusters were independent of engineering type.
5.2 MATERIALS ATTRIBUTES
The indices used within the Materials cluster analysis (Table 10), reflect the character of the
natural bed and bank materials and of the artificial materials used to reinforce the channel banks
and/or bed. Therefore, the attributes that underpin the cluster analysis reflect the potential
susceptibility of the river bed and banks to modification through fluvial processes. 7 classes of
stretches were identified (Figure 4). The proportion of Immobile Substrate and the proportion of
Biodegradable Protection were not significant in discriminating between the clusters, although
Environmental Sustainability Indicators for Urban River Management
21
the former provides a useful index for discriminating the Heavily Engineered Class of stretches.
In contrast, the calibre of the sediment and bank materials (SEDCAL and BANKCAL
respectively) and the type and amounts of bank protection (BANKPROT, Proportion No Bank
Protection, Proportion Open Matrix Protection, Proportion Solid Protection) were found to be
important discriminatory attributes. Although the engineering type was not included in the
analysis, this is clearly reflected in the level of reinforcement and so, not surprisingly, each
cluster is comprised of stretches that possessed distinct types of engineering modification
(Table 13). The combination of engineering and natural materials is reflected in the names given
to the clusters.
100.00
-273.69
-647.38
-1021.07
Similarity
SNF
SNCSNMLEHEENME
MATERIALSCLASS
Figure 4 7 clusters of urban river stretches defined by their materials characteristics
Environmental Sustainability Indicators for Urban River Management
22
Table 13 Descriptions of the characteristics of stretches attributed to different Materials clusters
Group Name: abbreviation Description of discriminating
Primary (Materials) Indicators
Description of broad Engineering
Characteristics
SEMI-NATURAL (COARSE):
SNC
Low proportions of bank protection.
Coarser substrates (SEDCAL) and
bank materials (average BANKCAL).
More natural planforms and cross sections
(developed through natural processes,
recovery or restoration), typically with
some sinuosity
SEMI-NATURAL (MIXED):
SNM
Low proportions of bank protection,
with mixed substrates typically
corresponding to silt/sand with some
gravels (SEDCAL).
Artificial (mainly straight) planforms, and
cross-sections but with limited
reinforcement
SEMI-NATURAL (FINE):
SNF
Low proportions of bank protection.
Finer (typically clay) substrates
(SEDCAL) and bank materials
(BANKCAL).
Natural sinuous planforms and crosssections
with limited reinforcement
LIGHTLY ENGINEERED:
LE
Coarser bed and bank materials
(SEDCAL, BANKCAL). Moderate
proportions (ca. 30-85%) of open
matrix protection (gabions, rip rap etc).
Artificial (usually sinuous) planforms, and
cross-sections with significant
reinforcement
ENGINEERED:
EN
High (ca. 90-100%) proportions of
open matrix bank protection and
moderate proportions (ca. 20-50%) of
solid bank protection. Low proportions
of immobile substrate.
Artificial (mainly straight) planforms and
cross-sections with extensive
reinforcement
MODERATELY ENGINEERED:
ME
High proportions (50-90%) of solid
bank protection (concrete, laid stone
etc.) but low proportions of immobile
substrate (i.e. bed reinforcement).
Artificial (mixed sinuosity) planforms and
cross-sections with extensive
reinforcement
HEAVILY ENGINEERED: HE High proportions (ca. 100%) of solid
bank protection (concrete, laid stone
etc.) and immobile substrate.
Heavily engineered, straight planforms and
high levels of reinforcement on the banks
and the bed.
-1551.40
-1000.93
-450.47
100.00
Similarity
Physical Class
SNS SNA RC US UM UA
Figure 5 6 clusters of urban river stretches defined by their physical habitat characteristics
23
EnvironmentalSustainabilityIndicatorsforUrbanRiverManagement
Table 14 Descriptions of the characteristics of stretches attributed to different Physical Habitat clusters.
Group Name:
abbreviation
Description of discriminating Primary
(Physical) Indicators
Typical Physical Habitat Characteristics Description of Broad Engineering Characteristics
SEMI-NATURAL
(ACTIVE):
SNA
Very low Proportions of Artificial Bank
Profiles and very high Proportions of
Natural Bank Profiles. >7 different
Habitat Types.
Mixed flow regime (20-30% riffle), some evidence
of pool formation. Typically contain 7-8
vegetated/ unvegetated bars.
Predominantly natural sinuous planforms.
SEMI-NATURAL
(STABLE):
SNS
Very low Proportions of Artificial Bank
Profiles and very high Proportions of
Natural Bank Profiles. <7 different
Habitat Types.
Flow dominated by glides (50-90%) with no
evidence of pool formation. Typically contain 0-4
vegetated/unvegetated bars.
Predominantly natural sinuous planforms.
RECOVERING:
RC
Moderate Proportions of Artificial Bank
Profiles (40-100%) and high Proportions
of Natural Bank Profiles (50-100%). < 7
different habitat types.
High levels of active bank recovery from
engineering intervention. Some evidence of mixed
flow regime (5-20% riffle, 60% glide) with some
pool formation. Typically contain 1-4 vegetated/
unvegetated bars.
Predominantly artificial sinuous planforms.
UNIFORM
ACTIVE;
UA
High Proportions of Artificial Bank
Profiles (ca. 100%), and moderate to high
Proportions of Natural Bank Profiles (ca.
0-50%). 5-9 different habitat types.
Some evidence of active channel recovery through
bank erosion. Some evidence of mixed flow
regime (10-25% riffle, 40-60% glide) but no
distinct pool formation. Typically contain 2-6
vegetated/unvegetated bars.
Predominantly artificial planforms most with some
sinuosity.
UNIFORM
MODERATELY
ACTIVE:
UM
High Proportions of Artificial Bank
Profiles (ca. 100%) and low Proportions
of Natural Bank Profiles (ca. 0-30 %).
Lower numbers (3-4) of Habitat Types.
Little evidence of channel recovery through bank
erosion. Channel dominated by glides (70-90%).
Typically contain 1-2 vegetated/unvegetated bars.
Predominantly artificial straight planforms
UNIFORM
STABLE:
US
High Proportions of Artificial Bank
Profiles (ca. 100%). Very Low
Proportions of Natural Bank Profiles (ca.
0%). Very low Numbers (1-2) of Habitat
Types.
No evidence of active channel recovery through
bank erosion. Flow almost entirely glides (90-
100%) Typically contain 0 bars.
Predominantly artificial straight planforms.
Environmental Sustainability Indicators for Urban River Management
24
5.3 VEGETATION ATTRIBUTES
The final cluster analysis is primarily concerned with the characteristics of the in-channel and
bank vegetation. 8 clusters were identified and all of the variables included in the cluster analysis
showed significant discrimination between clusters. The dominant channel vegetation type
(unvegetated channels, algal dominated channels, and vegetated channels) discriminated two
large clusters: an unvegetated cluster and a cluster that was predominantly vegetated with the
algal – dominated channels forming a distinct subgroup (Figure 6). The total tree score and
average bank face and top complexity provide additional discriminatory factors. Again, the
clusters were associated with different types of engineering, especially with the sinuosity of the
channel (Table 15).
100.00
-437.52
-975.04
-1512.55
VEGETATION CLASS
Similarity
VMC1 VMC2 VLC VHT ALG ULCUHC UMC
Figure 6 8 clusters of urban river stretches defined by their Vegetation characteristics
Environmental Sustainability Indicators for Urban River Management
25
Table 15 Descriptions of the characteristics of stretches attributed to different vegetation clusters.
Group Name:
abbreviation
Description of discriminating Primary (Vegetation)
Indicators
Description of broad
Engineering Characteristics
Vegetated Moderate
Complexity:
VMC1
Vegetated channels with low Total Tree Scores (< 6)
representing isolated scattered to occasional clumps, and
a higher mean bank top than bank face vegetation
complexity (average top BANKVEG>4 and average face
BANKVEG <5).
Mainly sinuous channels,
either natural or artificial.
Vegetated Moderate
Complexity:
VMC2
Vegetated channels with low Total Tree Scores (< 6), and
a higher bank face than bank top complexity (average top
BANKVEG <4 and average face BANKVEG >5 resp.).
Mainly artificial straight
channels.
Vegetated Low
Complexity:
VLC
Vegetated channels with low Total Tree Scores (< 6), and
low bank face BANKVEG and top BANKVEG indices.
Mainly artificial straight
planforms with some natural
planforms (typically running
through farmland).
Vegetated High Trees:
VHT
Vegetated channels with high Total Tree Scores (>6)
equivalent to semi-continuous – continuous tree cover.
Artificial planforms usually
straight but some display
sinuosity.
Algal Channels: ALG Channels dominated by algae (Dominant Vegetation
Type = Algae)
Mainly artificial straight
planforms.
Unvegetated Low
Complexity:
ULC
An aggregate group comprised of unvegetated channels
with either relatively low levels of Total Tree Scores (<6)
or with a higher tree cover combined with low bank face
and top complexity (average face BANKVEG <6.5 and
average top BANKVEG <6).
Mainly artificial planforms
usually straight but some
display sinuosity.
Unvegetated Moderate
Complexity:
UMC
Unvegetated stretches with higher Total Tree Scores
(>6), combined with low average top BANKVEG (<6.5)
and high average bank face BANKVEG (>6).
Mainly artificial planforms
usually straight but some
display sinuosity.
Unvegetated High
Complexity:
UHC
Unvegetated stretches with high Total Tree Scores (>6),
combined with high bank top vegetation complexity
(average top BANKVEG >6.5).
Mainly artificial planforms
with some sinuosity.
5.4 ENVIRONMENTAL INDICATORS
The classifications presented in sections 5.2 to 5.4 illustrate that there are three broad secondary
environmental indicators (Materials, Physical Habitat, Vegetation) which can be used to allocate
engineered stretches to 7, 6 and 8 different classes, respectively. These secondary environmental
indicators all appear to be related to the level and type of engineering to some degree, although
the strongest associations are with Materials and the weakest are with Vegetation. In order to
identify the class to which any new stretch should be allocated without re-running the entire
cluster analysis this section proposes three decisions trees (Figures 7 to 9) for this purpose.
These decision trees enable a newly-surveyed stretch to be allocated to an appropriate
‘Materials’, ‘Physical habitat’ and ‘Vegetation’ class. The decision trees do not incorporate all of
the variables (Primary Environmental Indicators) that were used to define the classes because
many of these are highly correlated. Thus, the decision trees incorporate the minimum number of
key variables that are needed to allocate a newly-surveyed stretch to the various classes.
Environmental Sustainability Indicators for Urban River Management
26
Figure 7 Flow chart for allocating urban river stretches to the relevant Materials Class
< 80% ≥ 80%
≤ 10% > 10%
≤ 70%
> 70% ≥ 90%
≥-1.5 <-1.5
OPEN
MATRIX
SOLID
SNF
SNC
SNM
LE
EN ME HE
≥ 2.5 < 2.5
Proportion
Immobile
Substrate
Amount Bank
Protection
(Inverse of Proportion
No Bank Protection)
SEDCAL
BANKCAL
Dominant
Protection
Type
(Proportion OMP,
Proportion SOL)
CLASS
Environmental Sustainability Indicators for Urban River Management
27
Figure 8 Flow chart for allocating urban river stretches to the relevant Physical Class
≤50% > 50%
≥90% < 90% > 50% ≤ 50%
≤7 >7 1 – 2 3 - 4 5+
SNS SNA RC US UM UA
Proportion
Artificial
Bank Profiles
Proportion
Natural
Bank
Profiles
Number of
Habitat
Types
CLASS
282828
EnvironmentalSustainabilityIndicatorsforUrbanRiverManagement
Figure 9 Flow chart for allocating urban river stretches to the relevant vegetation class.
OTHER ALGAE NONE
≤ 6 >6 ≤ 6 >6
>5 ≤5 ≤6.5 >6.5
≤ 4 > 4 <6 ≥6
UHCULC UMCALGVHT
≤ 4
VLCVMC2VMC1
Dominant
Vegetation
Type
Total
Tree Score
Average
BANKVEG
(Face)
Average
BANKVEG
(TOP)
CLASS
Environmental Sustainability Indicators for Urban River Management
29
6. Sector scale indices (tertiary environmental
indicators)
Whilst the secondary environmental indicators can be used to assess the likely outcomes of
applying different management and engineering options to a stretch, the success of such actions
may be constrained by other, largely sector-scale, characteristics. For example:
• If water quality or transported sediment quality are low, changes in physical habitat are
unlikely to yield any ecological benefit.
• If sediment supply from upstream is low, then changes in physical habitat dependent upon
geomorphological adjustment within a stretch will not be sustained.
• Certain changes in the engineering of reaches may be inherently unstable because of the
energy of sector-scale river flows, or may not meet flood-defence requirements.
• Even if water quality, sediment supply and flow regimes do not present constraints on the
outcomes of changes in the secondary environmental indicators, land use and land
availability may restrict the space available for such changes.
Thus the tertiary environmental indicators provide simple indicators of such constraints on the
potential to achieve particular stretch – scale objectives.
(N.B. Where surveys are available within a stretch, many of the indicators described in this
section are recorded in the URS using existing data sources (see section 3). However, as a result
of the limited availability of such surveys, they may only be available for locations elsewhere
within the sector and so they are presented here as sector-scale indices).
6.1 FLOW-RELATED INDICATORS
The river flow regime is a constraint in terms of achieving flood defence targets (high flow
magnitude), ensuring that any stretch–scale modifications do not result in major channel
instabilities (high flow energy), and ensuring that there is sufficient aquatic habitat to support
species during low flows (low flow magnitude and depth).
To reflect the above sensitivities, flow indices including the mean annual flood magnitude (and
its unit stream power within stretches of differing width), the maximum recorded flow and the 5,
50 and 95 percentiles of the flow duration curve are required for channel network sectors.
6.2 SEDIMENT-RELATED INDICATORS
The concentration, calibre and quality of sediment transported by the river are constraints in
terms of the range of geomorphological adjustment that can occur and the biota that can be
supported. Whilst the flow regime provides the energy to move sediments and construct and
erode landforms, the availability of sediment for transport is essential for landform construction,
and the calibre of the sediment constrains the degree of adjustment and the types of landform
that can be supported. The concentration of transported fine sediment is sometimes determined
as a component of analyses of water quality (see below), but coarser sediments are also required
for landform construction and their transport is not routinely monitored. Thus, the availability of
sediment for driving morphological change can only be determined by evidence of the presence
of active sediment sources (e.g. eroding banks, sedimentary bedforms) upstream of any
particular stretch. Such information can be derived from URS surveys of upstream stretches as
can the likely calibre of sediment delivered from those sources. Sediment quality or
contamination is also an important constraint on the success of river channel modifications. Both
water quality and sediment quality can severely limit the ecological benefits of channel
modifications.
Environmental Sustainability Indicators for Urban River Management
30
6.3 WATER QUALITY INDICATORS
Water quality modifications found in urban rivers arise from three key sources: domestic and
sewage effluents, industrial effluents, and road runoff. Each of these effluents include different
and characteristic types of pollutant. Domestic and sewage effluents primarily contribute to the
nutrient enrichment of the river system through inputs of nitrogenous and phosphate compounds.
The exact nature of industrial effluents largely depends upon the industrial processes being used.
However, heavy industrial processes such as metal working may create an increase in metal
concentrations within the river (copper, lead, aluminium, iron cadmium etc.). Surface runoff
from roads increases levels of electrical conductivity, especially in winter where salting of the
roads is employed for de-icing. Lead and petrol derived hydrocarbons are also components of
road surface runoff.
Water quality indices have been developed across Europe in response to legislation controlling
the safety limits of sanitary determinants for bathing and drinking waters (e.g. EEC, 1975; EC
1991; 1994) and the perceived impacts that poor water quality has upon the ecology of river,
lake and marine systems. Typically, these individual indices have been combined to provide a
basic classification of water quality for any given sector of river. The water quality indicators
can, therefore, based upon the requirements of the water quality classification developed for use
in each member country.
Whilst many indicators could be used to represent sector-scale water quality, the SMURF study
of the River Tame catchment will use the River Ecosystem Classification (RE), which is
currently used within the UK and combines 8 key parameters to assign a river to one of 5 classes
(Table 16). The listed indices are derived from either continuously monitored data, or from spotcheck
samples taken monthly. The former is clearly preferable as the indices are based on a
larger sample of determinations. However, continuous monitoring in urban rivers is typically
limited to only one or two sites within a catchment and therefore, the RE classification is often
based upon a spot sampling regime. In order to ensure that results are not biased by
uncharacteristic local events (such as a single pollution episode), the RE rating is based upon a
minimum of 12 samples, with a recommended monthly sampling regime.
Environmental Sustainability Indicators for Urban River Management
31
Table 16 Water Quality Criteria for defining River Ecosystem Classes (from NRA, 1994)
Class Do
(%)
BOD
Mg/L
Total
Ammonia
Un-Ionised
Ammonia
pH Hardness Dissolved
Copper
Total
Zinc
RE1 80 2.5 0.25 0.021 6-9 <10
>10 and <50
>50 and <100
>100
5
22
40
112
30
200
300
500
RE2 70 4.0 0.6 0.021 6-9 <10
>10 and <50
>50 and <100
>100
5
22
40
112
30
200
300
500
RE3 60 6.0 1.3 0.021 6-9 <10
>10 and <50
>50 and <100
>100
5
22
40
112
300
700
1000
2000
RE4 50 8.0 2.5 - 6-9 <10
>10 and <50
>50 and <100
>100
5
22
40
112
300
700
1000
2000
RE5 20 15.0 9.0 - - - - -
6.4 BIOTIC INDICATORS
The combined effects of degraded channel morphology and water quality within urban rivers is
often an equally poor ecological community. Contemporary bioassessment classification systems
are typically based upon the natural distribution of species within river systems and their
tolerance to environmental conditions and pollution, and can be used to assess levels of sublethal
stress within the system. Most European countries have developed, or are developing,
biotic classification systems similar to those developed for water quality, based upon the
response of the benthic macroinvertebrate community to poor water quality. Therefore, the
indicators for biotic integrity of the river can be based upon the classification developed by each
member country.
The present study will adopt the UK approach to biological classification of rivers. This has been
based the Biological Monitoring Working Party (BMWP) scoring system, developed between
1976 to 1978, and revised in 1981 (Hawkes, 1997). It assigns a score from 1 to 10 to families of
taxa according to their perceived tolerance to pollution (mainly organic pollution). Each family
scores once within the sample, and the scores for each taxa are added together to give a total
score for a site. This score can then be used within a classification index such as the General
Quality Assessment to assign each site to a class according to its quality of benthic invertebrate
community (Nixon et al., 1996).
The need for a more predictive approach to river management instigated the development of the
River InVertebrate Prediction And Classification System (RIVPACS), a computer-based tool
that predicts the list of taxa that should be found at any given site according to the physical and
chemical conditions within the channel (Wright et al., 1993). This type of prediction can be used
as a reference condition for assessing the level of degradation at a site due to water quality
problems. In order to ensure that the system can be fully integrated into river management, the
taxa list can be used to produce predicted BMWP and ASPT scores. The predictive capability of
RIVPACS and its ability to produce a reference condition for polluted rivers, has allowed its
incorporation into the General Quality Assessment (GQA) employed by river managers in the
UK (Nixon et al., 1996).
The GQA scheme grades rivers into 6 classes ranging from A to F where A represents high
quality and F represents poor quality rivers (Table 17). The observed and predicted BMWP and
Environmental Sustainability Indicators for Urban River Management
32
ASPT scores are used to produce an ecological quality index (EQI) which is used to classify
each site, where the observed value is divided by the predicted value.
Table 17 The GQA classification of rivers according to their biological quality incorporating
RIVPACS derived EQI indices (EA, Pers. Comm).
GQA GRADE INFERRED
QUALITY
EQI TAXA
(BMWP)
EQI ASPT BMWP SCORE
A Very good > 0.85 1.00 >95
B Good 0.70-0.84 0.90-0.99 68-95
C Fairly good 0.55-0.69 0.77-0.89 51-67
D Fair 0.45-0.69 0.65-0.76 35-50
E Poor 0.30-0.44 0.50-0.64 13-34
F Bad <0.30 <0.50 0-12
6.5 LAND USE AND LAND AVAILABILITY
The URS directly records land use at both stretch and sector scale (Table 5). These data form a
final class of tertiary environmental indicator since they represent land use, land quality or open
land availability constraints on the engineering options that may be considered for a particular
stretch.
Environmental Sustainability Indicators for Urban River Management
33
7. Combining indicators to address scenarios of
change
The preceding sections have described the derivation of primary and secondary environmental
indicators at the stretch scale and tertiary indicators at the sector scale. However, in order to
enable river managers to consider management scenarios these different levels of indicators need
to be combined.
The secondary environmental indicators allocate a stretch to different classes according to its
Materials, Physical Habitat and Vegetation characteristics. The work has shown that the type of
engineering applied to a stretch has a significant influence on the class to which a stretch is
allocated, with the strongest associations being apparent in the Materials classes and the weakest
in the Vegetation classes. As a result, the consequences of changed engineering can be explored
in relation to these three classifications, and the influence of scenarios of vegetation management
can be additionally explored in relation the Vegetation classification. Thus the secondary
environmental indicators provide a simple means of characterising the physical properties of a
stretch and their dependence upon engineering and to some extent vegetation management. In
the case of algal channels, pollution management is also significant. These secondary
environmental indicators also permit consideration of the consequences of changes primarily in
engineering but also in vegetation management.
7.1 STRETCH SCORES
In order to combine the three different classifications to produce a single index of the overall
quality of a stretch, scores have been assigned to each group within the Materials, Physical and
Vegetation classifications (Table 18)
The scores assigned to the materials classes reflect the change from semi-natural (score = 1) to
heavily engineered stretches (score = 5). The semi-natural coarse (SNC) and fine (SNF) classes
essentially reflect the different alluvial sediments bounding the river channel and are, therefore,
assigned the same score of 1. The semi-natural mixed class (SNM), with its coarser banks and
finer substrates is actually comprised of straightened channels with resectioned banks. The
lightly engineered (LE) class also displays engineering modification but with more sinuous than
SNM with some reinforcement usually applied to eroding banks on the outside of the bends. As
a result of the similar (low) degree of engineering modification both of these groups have been
assigned a score of 2. The engineered (EN) and moderately engineered (ME) classes are
distinguished from each other by the difference in dominant protection type (open matrix and
solid, respectively), and the ME class contains less overall reinforcement than the EN class. Both
of the classes also show a marked increase in the level of engineering from the SNC, SNF, SNM
and LE classes and have, therefore, each been assigned a score of 4. The final heavily
engineered (HE) class comprises stretches that have both reinforced bed and banks and is
assigned a score of 5
Scores assigned to the physical classes reflect the degree to which the channel has been modified
and the degree to which the channel is recovering either some or all of its physical habitat
features. The semi-natural active class is assigned a score of 1 reflecting the fact that this class
possesses a high degree of naturalness with many different habitat types. The recovering class
(RE) is characterised by some engineering intervention but also high levels of recovery, creating
a greater variety in physical habitat features and flow types than the semi-natural stable (SNS)
class. Therefore, scores of 2 and 3 have been assigned to the RE and SNS classes, respectively.
The remaining, more heavily engineered classes display less habitat types and lower levels of
recovery than any of the previously-discussed classes. The uniform active (UA), uniform
moderately active (UM) and uniform stable classes (US) display decreasing levels of sinuosity,
natural bank profiles and habitats with increasing channel stability and so these classes have
been assigned scores of 4, 5, and 6 respectively.
Environmental Sustainability Indicators for Urban River Management
34
Scores assigned to the vegetation classes reflect the level and type of in-channel vegetation, and
the complexity of the riparian vegetation. In general a vegetated channel is more desirable than
an unvegetated channel (with the exception of Algal channels), a complex riparian habitat is
more desirable than a uniform one, and a moderate to high tree cover is preferable to either no
trees or a channel completely shaded by trees. Therefore, the two vegetated moderate complexity
channels (VMC1 and VMC2) are given the best score of 1. The vegetated high tree (VHT) class
scores higher than the unvegetated high complexity (UHC) class (2 and 3 respectively). Both of
these classes possess high levels of trees, however, the fact that one class is vegetated suggests
that less shading is present allowing vegetation to grow in the channel. As stated, a moderate
riparian complexity is preferable to a low complexity, and therefore the unvegetated moderate
complexity (UMC) class is assigned a score of 4. The vegetated low complexity (VLC) class is
assigned a score of 5, whereas the unvegetated low complexity (ULC) class is assigned a score
of 6. Finally the algal dominated channels (ALG) are assigned a score of 7 reflecting the fact that
these filamentous algae are an indicator of poor water quality, and therefore represent the least
desirable class of all. Nevertheless, it should be stressed that, with the exception of the ALG
class, a mixture of these vegetation classes is required at the catchment or sector level, to
provide variation along the river.
Table 18 Scoring system for defining the quality and potential of urban river channels.
MATERIALS PHYSICAL HABITAT VEGETATION
Class Score Class Score Class Score
SNF
(semi-natural fine)
1 SNA
(semi-natural active)
1 VMC1(vegetated moderate
complexity)
1
SNC
(semi-natural coarse)
1 RC
(recovering)
2 VMC2 (vegetated moderate
complexity)
1
SNM
(semi-natural mixed)
2 SNS
(semi-natural stable)
3 VHT
(vegetated high trees)
2
LE
(lightly engineered)
2 UA
(uniform active)
4 UHC (unvegetated high
complexity)
3
EN
(engineered)
4 UM (uniform
moderately active)
5 UMC (unvegetated
moderate complexity)
4
ME (moderately
engineered)
4 US
(uniform stable)
6 VLC (vegetated low
complexity)
5
HE
(highly engineered)
5 ULC (unvegetated low
complexity)
6
ALG
(algal low complexity)
7
Each stretch can therefore be assigned a score according to its Materials, Physical and
Vegetation class, and these scores can then be added to give an overall score for the stretch,
where a low score represents a good quality channel.
The decision trees and scoring system were applied to each of the 106 stretches of the River
Tame surveyed during summer 2003. The frequency distribution of these overall scores (Figure
10) shows a good range of scores, although none of the stretches display a ‘perfect
environmental score’ of 3 or a ‘worst environmental condition’ score of 18.
Environmental Sustainability Indicators for Urban River Management
35
Figure 10 Frequency distribution of overall scores for 106 stretches of the River Tame.
7.2 MANAGEMENT RECOMMENDATIONS
This report has described how urban rivers may be comprehensively assessed to provide a
detailed overall score indicative of the current condition of any individual stretch of river.
However, management of urban rivers requires an understanding of the potential condition of a
stretch if modifications for flood defence or rehabilitation were to be undertaken. This section
presents a basic model of how the potential condition of a stretch may be defined, and how to
model likely scenarios for rehabilitation of urban rivers.
Section 7.1 described how each stretch can be given individual scores for each of the Materials,
Physical and Vegetation classes (Table 18). These scores can be added to give an overall score
for each stretch. Table 19 defines 6 categories of urban river stretches, assigns the overall scores
associated with each category, defines the type of material, physical, and vegetation class that
comprises these scores, and describes the type of management that might be undertaken to
rehabilitate a stretch falling in each category.
Number of stretches for each overall score
0
2
4
6
8
10
12
14
16
18
20
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Ov e r a l l S c or e
EnvironmentalSustainabilityIndicatorsforUrbanRiverManagement
36
Table 19 Overall scores, classes and management recommendations associated with categories of urban river stretch.
CATEGORY OVERALL
SCORES
CLASSES MOST
ASSOCIATED WITH
SCORES (M=Materials;
P=Physical; V=Vegetation)
MANAGEMENT RECOMMENDATIONS
Very Good 3-5 M: SNC/SNF
P: SNA/RC/SNS
V: VMC1 & 2/ UHC/VHT
Predominantly semi-natural and recovering stretches, with good vegetation and tree cover. The
recommendation is to leave these stretches free of management and to protect them from development.
Good 6-8 M: SNC/LE
P: RC/SNS/UA/UM
V: VMC1 & 2/ UHC/VHT
Semi-natural, recovering and a few uniform channels displaying some activity, with good vegetation
complexity and tree cover. The recommendation is to remove any remaining reinforcement to allow the
channel to recover more freely. These stretches should also be protected from further development.
Average 9-11 M: SNC/SNM/LE/ME
P: RC/UA/US/SNS
V: ULC/UHC/VMC2
Stretches with varying levels of engineering, but displaying some level of either recovery or activity, with
little vegetation complexity or too much tree cover. The recommendation is, where possible, to reduce the
levels of immobile substrates and bank materials and increase sinuosity. Tree cover and bank top and face
vegetation should be managed to provide increased variety and complexity. These channels show moderate
to high levels of activity and should be targeted for rehabilitation where opportunities arise.
Below Average 12-13 M: SNC/LE/HE
P: UA/UM
V: ULC/UMC/ALG
Stretches with varying levels of modification but showing high levels of activity, combined with low bank
vegetation complexity or algal dominated channels with few trees. These channels show moderate to high
levels of activity and should be targeted for rehabilitation where opportunities arise. The recommendation is
to reduce or alter the level and/or type of reinforcement and increase channel sinuosity where possible.
Increased tree cover through planting, and management of the bank face and top vegetation to improve
complexity should be undertaken. Algal dominated channels should also be assessed for improvements to
water quality.
Poor 14-16 M: HE/ME/EN
P: UM/US/UA
V: ULC/UMC/ VLC/ALG
Moderate to heavily engineered channels with low to moderate levels of activity, low complexity of bank
vegetation, few trees and often algal dominated channels. The recommendation is to assess the water quality
for improvement of in-channel vegetation diversity, and undertake a detailed assessment of the level of
rehabilitation required to improve the physical condition of the channel. Where possible, a reduction of
reinforcement level and/or type and an increase in sinuosity of the channel is desirable. Tree planting should
be introduced to improve riparian complexity.
Very Poor 17-18 M: HE
P: US
V: ULC/ALG
Heavily engineered, algal-dominated, stable channels with little vegetation complexity. Significant
improvements to water quality should be initiated, followed by a detailed assessment of rehabilitation
needs. Aesthetic rehabilitation may be the best option in the short term. Wherever possible this should be
followed by some reduction in the level of reinforcement and an increase in channel sinuosity.
Environmental Sustainability Indicators for Urban River Management
37
7.3 A SCORING SYSTEM TO UNDERPIN SCENARIO MODELLING OF
ENGINEERING AND MANAGEMENT CHANGE
To scenario-model the impact of changed engineering, the decision trees that are used to allocate
a reach to a particular class (Figures 7, 8 and 9) can, to some extent, be re-used. However, direct
human modification can only be applied to certain components of the decision trees, since others
are not directly physically manipulable:
• All components of the Materials decision tree apart from BANKCAL and SEDCAL (which
discriminate between the SNF, SNM and SNC classes) can be directly manipulated and so
Figure 7 can be used to assess a new Materials score based on a scenario of changed
engineering within 5 classes.
• For the Physical Habitat decision tree (Figure 8), the proportion of artificial bank profiles
can be manipulated but the proportion of natural bank profiles (which distinguishes SNS &
SNA from RC and from US, UM & UA) cannot be manipulated. However, one important
reason why there are more natural / active bank profiles in the SNS & SNA classes than in
the US, UM & UA is that reaches in the US, UM & UA class tend to be relatively straight,
whereas those in the SNS & SNA tend to be more sinuous, with those in the RC class
having a mixed, intermediate sinuosity. Therefore a highly sinuous, intermediate or straight
channel planforms have been introduced to discriminate between these three groups of
classes for scenario modelling.
• For vegetation, the algal channels reflect relatively poor water quality and so cannot be
influenced by physical modification of reaches. For the other classes, presence or absence of
in-channel vegetation cover cannot easily be manipulated, although shading of the channel
will reduce the in-channel vegetation. However, tree cover (as reflected in the Total Tree
Score) and bank top and face vegetation complexity (BANKVEG) are manipulable factors
that discriminate between the classes. Thus the presence or absence of in-channel vegetation
cover provides a context around which manipulation of vegetation on the banks can be
undertaken.
Thus, Materials, Physical habitat and Vegetation have manipulable properties that can be
scored (Table 20) to closely match the scores allocated in Table 18. This revised and
simplified scoring system can be used to compare the current state of a stretch with a range
of scenarios based upon the physical manipulation of materials, level of bank protection,
sinuosity and bank vegetation retention, planting and management. Because the properties
in Table 20 are manipulable, stretches can be given a score for their current state or for any
manipulated state, so that the level of ‘improvement’ or ‘degradation’ resulted from specific
management changes can be broadly assessed.
Environmental Sustainability Indicators for Urban River Management
38
Table 20 Stretch scoring system based on manipulable stretch properties to support scenario
modelling.
Class Threshold values for discriminating indicators Score
MATERIALS Proportion
Immobile
Substrate
Proportion Bank
Protection
Predominantly
Open Matrix
Protection
Predominantly
Solid
protection
SN (F, M & C) <80% <10% 1
LE <80% >10%, <70% 2
EN <80% >70% Yes 4
ME <80% >70% Yes 4
HE >80% >90% 5
PHYSICAL Proportion
Artificial
Bank Profiles
Sinuosity
SNS, SNA <50% Natural, relatively
high, sinuosity
1
RC Artificial moderate
sinuosity
2
UA >50% Artificial low
sinuosity
4
US, UM >50% Artificial Straight 5
VEGETATION Dominant
Vegetation
Type
Total Tree score Average
BANKVEG
(face)
Average
BANKVEG
(top)
VMC1 Vegetated-
Other
<6 >5 < 4 1
VMC2 Vegetated-
Other
<6 > 4 1
VHT Vegetated-
Other
>6 2
UHC Unvegetated >6 > 6.5 3
UMC Unvegetated >6 < 6.5 4
VLC Vegetated-
Other
<6 < 5 < 4 5
ULC Unvegetated <6
OR
>6 <6
6
ALG Vegetated-
Algae
7
N.B. Note that scenario modelling of the vegetation should not involve manipulation of the
dominant in-channel vegetation. This represents a controlled condition around which scenarios of
bank vegetation change can be explored.
7.4 DETAILED ASSESSMENT OF STRETCH CHARACTERISTICS AND
CONSTRAINTS
Whilst the previous sections link broad management recommendations to the overall score for a
stretch it is important to realise that such an approach is extremely generalised. The overall
score is a very simple index of stretch character and quality, which provides a simple basis for
selecting groups of stretches for a more detailed assessment. Such a detailed assessment should
incorporate the other indicators and supporting information included in this report.
The individual secondary environmental indicators based on the Materials, Physical and
Vegetation characteristics of the stretch provide a far more detailed assessment of its character,
and inspection of the individual primary environmental indicators can provide even grater
detail, supported where necessary by information from the raw URS records.
Environmental Sustainability Indicators for Urban River Management
39
However, in considering scenarios of engineering change the tertiary environmental
indicators provide initial information on constraints which may limit the potential success of
any particular option. Thus water and sediment quality indicators can support an assessment of
whether any genuine in-channel ecological benefit can be gained. In essence, if water and/or
sediment quality are poor, then no improvement in physical habitat will yield an improvement in
the aquatic ecology of the stretch. Under such circumstances, changes in engineering may yield
aesthetic benefits and improvements in riparian ecology, but water quality improvement will be
essential before the in-channel ecosystem can benefit.
In addition, flow-related indicators can provide an initial assessment of the likely stability of a
change in engineering by considering (e.g. the unit stream power at bank-full stage), and
sediment calibre and supply indicators may be relevant to stability and also to the degree to
which morphological change may be achieved. Such indicators may also be of ecological
significance in indicating whether low flows will be sufficient to support an enhanced aquatic
ecosystem. Moreover, a combination of flow and water quality indices, may allow consideration
of the consequences of different flow regulation scenarios for water quantity and quality within a
stretch.
Finally, some simple sector-scale tertiary indicators relating to floodplain land use and flood
plain width indicate whether there are constraints in land availability or land quality that may
prevent certain engineering options. For example, a restricted flood plain or presence of
contaminated land could place severe constraints on engineering options that include a change
from a straight to sinuous river planform.
Environmental Sustainability Indicators for Urban River Management
40
8. References
Anderson, J.R., Hardy, E.E., Roach, J.T. and Witmer, R.E. (1976). A land use and land cover
classification system for use with remote sensor data. US Geological Survey Professional Paper 964
Beechie, T.J. and Sibley, T.H. (1990) Evaluation of the TFW Stream Classification
System on South Fork Stillaguamish Streams. Water Research Centre Project
No A-164-Wash.USA.
Davenport, A.J., Gurnell, A.M. and Armitage P.D. (in press). Habitat survey and classification of urban
rivers. River Research and Applications.
EEC (1975). Directives concerning the quality required of surface water intended for the abstraction of
drinking water in the member states. 75/440.EEC. Official Journal L194/39 25 July 1975.
EC (1991). Council Directive of 21st
May 1991 concerning urban Waste Water Treatment. EC Council
Directive 91/271/EEC
EC (1994). Proposal for a council directive on the ecological quality of water. COM (93) C90-final.
Brussels, 15/6/94. 94/0152 (syn).
Environment Agency, 1997. River Habitat Survey: 1997 Field Survey Guidance Manual, Incorporating
SERCON. Environment Agency, Bristol, UK.
Fox, P.J.A., Naura, M. and Scarlett, P. 1998. ‘An Account of the Derivation and Testing of a Standard
Field Method, River Habitat Survey’. Aquatic Conservation: Marine and Freshwater Ecosystems, 8, 455-
475.
Frissell, C.A., Liss, W.J., Warren, C.E. and Hurley, M.D. (1986). A Hierarchical Framework for Stream
Habitat Classification: Viewing Streams in a Watershed Context. Environmental Management, 10(2),
199-214
Goode, D.A. 1989. ‘Urban Nature Conservation in Britain’. Journal of Applied Ecology 26, 859-873.
Griffith, D.A. and Amrhein, C.G. 1997. Multivariate Statistical Analysis for Geographers. Prentice Hall:
USA.
Hawkes, H.A. 1997. ‘Origin and Development of the Biological Monitoring Working Party Score
System’. Water Research, 32, 964-968.
Holmes, N.T.H., Newman, J.R., Dawson, F.H., Chadd, S., and Rouen, K.J. (1997). Mean Trophic Rank:
A User’s Manual. Environment Agency R&D Draft Technical Report 694/NW/05.
Kirk, J.T.O. 1994. Light and photosynthesis in aquatic ecosystems. Cambridge University Press,
Cambridge, UK.
Leopold, L.B. and Wolman, M.G. 1957. River channel patterns: braided, meandering and straight.
USGS Professional Paper No. 282B
Meador M.R., Hupp, C.R., Cuffney, T.F. and Gurtz, M.E. 1993. Methods for characterizing stream
habitat as part of the national water-quality assessment programme. US Geological Survey Open File
Report 93-408.
Newson, M.D. 2002. ‘Geomorphological concepts and tools for sustainable river ecosystem
management’. Aquatic Conservation: Marine and Freshwater Ecosystems, 12, 365-379.
Environmental Sustainability Indicators for Urban River Management
41
Nixon, S.C., Clarke, S.J., Dobbs, A.J. and Everard, M. 1996. Development and Testing of General
Quality Assessment Schemes. NRA R&D Report 27, HMSO, London, UK.
NRA (1994). Water Quality Objectives: Procedures used by the National Rivers Authority for the
purpose of the surface waters (River Ecosystem Classification) regulations. NRA Bristol.
Pitcairn, C.E.R. and Hawkes, H.A. 1973. ‘The role of phosphorus in the growth of Cladophora’. Water
Research, 7, 159-171.
Raven, P.J., Fox, P., Everard, M., Holmes, N.T.H. and Dawson, F.H. (1997). River Habitat Survey: a
new system for classifying rivers according to habitat quality.In: Freshwater Quality: Defining the
Indefinable? Boon, P.J. and Howell, D.L. (Eds.) The Stationary Office, Edinburgh, pp 215-234.
Wadeson, R.A. and Rowntree, K.M. (1994). A Hierarchical Geomorphological Model for the
Classification of South African River Systems. Proceedings of a joint South African and Australian
workshop. Cape Town February 7-11th
, South Africa.
Wharton, G. 1995. ‘Information from channel geometry-discharge relationships’. In: Gurnell, A.M. and
Petts, G.E. (eds.), Changing River Channels, 325-346, Wiley, Chichester, UK.
Wright, J.F., Furse, M.T. and Armitage, P.D. 1993. ‘RIVPACS – a technique for evaluating the
biological quality of rivers in the UK’. European Water Pollution Control, 3, 15-25.
Environmental Sustainability Indicators for Urban River Management
42
SMURF, Sustainable Management of Urban Rivers
and Floodplains, is supported by the EC LIFEEnvironment
programme (Ref LIFE02 ENV/UK/000144).
The project is a partnership between Birmingham City
Council, Environment Agency, Severn Trent Water,
HR Wallingford, Staatliches Umweltamt Herten,
University of Birmingham and King’s College London.
i
Environmental Sustainability Indicators
for
Urban River Management
Appendix 1
Application of the Assessment and Scoring
System to Rivers in Other EU Member States
LIFE02 ENV/UK/000144
April 2004
Environmental Sustainability Indicators for Urban River Management
3
The SMURF project, Sustainable Management of Urban Rivers and Floodplains, is a partnership
between Birmingham City Council, Environment Agency, Severn Trent Water, HR Wallingford,
Staatliches Umweltamt Herten, University of Birmingham and King’s College London.
Environmental Sustainability
Indicators
For Urban River Management
Appendix 1
Application of the Assessment and
Scoring System to Rivers in Other
EU Member States
Produced by
Angela Boitsidis and Angela Gurnell
Department of Geography
King’s College London
Environmental Sustainability Indicators for Urban River Management
4
This report is a contribution to research generally and it would be imprudent for third parties to rely on
it in specific applications without first checking its suitability.
Various sections of this report rely on data supplied by or drawn from third party sources. King’s
College London accepts no liability for loss or damage suffered by the client or third parties as a result
of errors or inaccuracies in such third party data.
King’s College London will only accept responsibility for the use of its material in specific projects
where it has been engaged to advise upon a specific commission and given the opportunity to express a
view on the reliability of the material for the particular application.
Environmental Sustainability Indicators for Urban River Management
5
Acknowledgements
The authors would particularly like to thank Professor G.E. Petts, University of Birmingham, and Dr P
Armitage, Centre for Ecology and Hydrology, Dorset, for their valuable discussions whilst we were
developing the present research and writing this report. We should also like to thank Dr Harald Rahm,
Staatliches Umwetamt Herten, and Dr Dana Kominkova, Czech Technical Institute, for their assistance in
organising the fieldwork within Europe, and Ms M. Lee, King’s College London, for her assistance with
the field survey of the River Tame. The methodology described in this report has its foundation in an
approach that was developed under funding from the Natural Environment Research Council’s URGENT
Thematic Programme.
Environmental Sustainability Indicators for Urban River Management
6
Environmental Sustainability Indicators for Urban River Management
7
Contents
Title page i
Acknowledgements............................................................................................................................ 5
Contents .......................................................................................................................................... 7
1 Site Descriptions …………………………………………………………………………..8
2 Results: The Botic River, Prague, Czech Republic……………………………..………....9
3 Results: The River Emscher, Duisberg, Germany ………………………………………..11
4 Discussion………………………………………………………………………………….13
5 Suggestions for Further Development……………………………………………………..13
Environmental Sustainability Indicators for Urban River Management
8
1. Site Descriptions
In order to ensure the methodology is applicable to urban rivers in other EU countries,
the Urban River Survey (URS) was applied to 19 stretches along the Botic River,
Prague during May 2004, and 18 stretches along the River Emscher, Germany during
July 2004.
The Botic River is a short urban river (34.5km) draining an area of 135km2. It rises in
the east of Prague in a rural environment passing through a reservoir before flowing
through the city and into the Vlatva river. The Botic has been highly engineered with
many weirs to improve oxygenation. Industrial effluents have historically been
discharged into the river, and contamination with metals, especially copper and nickel,
is high.
The River Emscher drains an area of approximately 865km2
, within the highly
developed area of North-Rhine Westphalia, Germany, running through the major towns
of Dortmund, and Duisburg before discharging into the river Rhine (Figure 1). The river
has, historically, been shortened, straightened, and deepened to allow floodplain
development and mining and facilitate flood defence. The subsidence in the area caused
by mining prevented the construction of sewer networks, resulting in the river itself
being used to transport raw effluents away from the developed areas. Consequently, a
treatment plant located near the confluence with the river Rhine was constructed to treat
the entire river.
Figure 1 – River Emscher Catchment Map
The river has been subjected to very high levels of engineering, with 50% of the main
channel possessing embankments up to 8m in height, and most possessing concrete
trapezoidal channels, and highly maintained riparian vegetation. The water quality of
the Emscher is very poor due to the levels of raw effluent being discharged to it.
However, the entire river is being systematically rehabilitated, and a sewer network
constructed.
Environmental Sustainability Indicators for Urban River Management
9
2. Results: Botic River, Prague, Czech Republic.
A range stretches of the Botic river subject to different types of engineering were
selected for assessment. These ranged from semi-natural stretches in the upstream
section, to fully reinforced concrete stretches downstream. Table 1 provides the
outcomes of various classifications applied to the surveyed stretches.
SITE MATERIALS
CLASS
PHYSICAL
CLASS
VEGETATION
CLASS
TOTAL
SCORE
OVERALL CLASS
1CHP SNC SNA UHC 5 VERY GOOD
1MAT SNM SNA UHC 6 GOOD
1STY SNC SNA UHC 5 VERY GOOD
1UKR LE RC UHC 7 GOOD
2UBD LE SNA UHC 6 GOOD
2UBU SNC SNA UHC 5 VERY GOOD
3DKS LE SNS UHC 8 GOOD
3ISH SNC SNA UHC 5 VERY GOOD
3MAB LE SNS UHC 8 GOOD
3PRA LE RC UHC 7 GOOD
3ZAB ME UA ULC 14 POOR
4AMV HE UM UHC 13 BELOW AVERAGE
4HLS HE UM UHC 13 BELOW AVERAGE
4KBT HE US UHC 14 POOR
4KRM HE UM ULC 16 POOR
4NFM HE UM UMC 14 POOR
4PHS HE US ULC 17 VERY POOR
4SEK HE UM ULC 16 POOR
4UMS HE UM UHC 13 BELOW AVERAGE
Table 1: Results of the application of the classification system to the Botic river,
Prague.
Table 1 shows that only four of the surveyed stretches were allocated to physical classes
indicative of high stability (SNS, US), with the majority showing at least some signs of
activity. This stability is especially notable in the Heavily Engineered (HE) materials
class where 75% of the stretches were allocated to physical classes indicative of some
level of active change. Also notable are the relatively high complexity of the riparian
zone of the channel indicated by the vegetation classes, with only 4 stretches displaying
low complexity (ULC), and the fact that all stretches exhibit no aquatic vegetation (i.e.
have codes starting with U).
The distribution of the stretches according to their overall class (Figure 2), shows a
bimodal distribution, with no stretches falling into the average class. This is, in part, due
to the high level of complexity displayed by the riparian areas, reflecting low
maintenance, that even extends to completely reinforced stretches with concrete
channels (e.g. Figure 3).
Environmental Sustainability Indicators for Urban River Management
10
0.00
5.00
10.00
15.00
20.00
25.00
30.00
35.00
VERY GOOD GOOD AVERAGE BELOW AVERAGE POOR VERY POOR
CLASS
%AGEOFSTRETCHES
Figure 2: Distribution of stretches of River Botic according to overall class
Figure 3: Riparian complexity along the river Botic, Prague.
Environmental Sustainability Indicators for Urban River Management
11
3. Results: River Emscher, Duisberg, Germany.
Stretches exhibiting a range of engineering types were selected for assessment in the
Emscher catchment, although the selection was limited as a result of safety concerns;
the width of the river at the upstream end; and the availability of engineering types.
Table 2 lists the outcomes of the various classifications applied to the surveyed
stretches.
SITE MATERIALS
CLASS
PHYSICAL
CLASS
VEGETATION
CLASS
TOTAL
SCORE
OVERALL CLASS
2ESD HE US ALG 18 VERY POOR
2ESU HE UM ALG 17 VERY POOR
2ESS LE UA VHT 8 GOOD
2HCS SNM US VMC1 9 AVERAGE
2LBS HE US VHT 13 BELOW AVERAGE
3ALS HE US ALG 18 VERY POOR
3DLS SNF UM VHT 8 GOOD
3NHS SNF UA ULC 11 AVERAGE
3OST SNF UA UHC 8 GOOD
3SWW SNM US VMC2 9 AVERAGE
9GWP HE UM ULC 16 POOR
9HGS SNM UM UHC 10 AVERAGE
9INS HE UM UHC 13 BELOW AVERAGE
9SML EN UA ULC 14 POOR
9WHS HE UM UHC 13 BELOW AVERAGE
12DLD HE UM ULC 16 POOR
12DLU HE US ULC 17 VERY POOR
12RAU HE UM VMC2 11 AVERAGE
Table 2: Results of the application of the classification system to the River Emscher,
Germany.
Although a range of materials classes was found within the catchment, it is clear from
the allocation of stretches to physical classes that this is a highly modified river system,
with no stretches exhibiting a physical class indicative of semi-natural characteristics.
However, the river does show signs of active adjustment with 60% of the Heavily
Engineered (HE) stretches displaying moderate activity (physical class UM), and only
six stretches being attributed to a stable class (US). The vegetation classes illustrate a
mix of in-channel and riparian characteristics and it was apparent during the survey that
the vegetation was mainly managed for flood defence purposes.
The distribution of stretches according to their overall score (Figure 4) shows that 31%
of the stretches fall in the average or good category, despite being so heavily modified
and having very poor water quality (Figure 5).
Environmental Sustainability Indicators for Urban River Management
12
0.00
5.00
10.00
15.00
20.00
25.00
30.00
VERY GOOD GOOD AVERAGE BELOW AVERAGE POOR VERY POOR
OVERALL CLASS
%AGEOFSTRETCHES
Figure 4: Distribution of surveyed stretches of River Emscher according to the overall
class
Figure 5: Example of a partially rehabilitated stretch on the River Emscher, Germany.
Environmental Sustainability Indicators for Urban River Management
13
4. Discussion
The application of the Urban River Survey to two rivers in other EU member states has
shown that the method is robust in identifying stretches of river that display higher
quality features such as morphological habitats and complex vegetation. The survey has
proved to be very simple to apply, and the classification system developed for UK
rivers has proved to be applicable to these other urban river systems. The results show
that even in heavily modified stretches of river some variation exists which may be
important for long term sustainability of urban rivers flowing through large
conurbation.
The scoring system developed for the SMURF project enables managers to prioritise
rehabilitation schemes effectively and details priorities for effective rehabilitation.
However, the application of this methodology to these two additional rivers has
highlighted some problems. Primarily, the scoring system that has been developed is
applied unweighted to stretches of the river. However, this leads to the vegetation
classification exerting greater influence on the overall score than physical changes to
the channel. Thus, the manipulation of the riparian vegetation will result in a greater
movement in the overall classification than manipulation of the banks or channel
materials.
A further problem is highlighted by the application to the River Emscher. In this
catchment, the toxic sediments and water quality that exists within much of the channel
prevents the growth of any in-channel vegetation, including filamentous algae. This
results in these channels being given a higher score that channels where water/sediment
quality is sufficiently good to support the growth of algae. Thus, it is vitally important
that the proposed scoring system is fully integrated with a water quality classification,
and, where possible, to a sediment quality classification, in order to identify whether the
underlying problem is one of engineering design or water quality.
5. Suggestions for further development
This work suggests that the methodology is simple, effective and robust in its
application to a range of urban rivers. Further work needs to be undertaken to enable the
survey to be fully integrated with the current nationally accepted survey methodologies
(e.g. River Habitat Survey in the UK), and also to ensure that variation between
individual surveyors is minimised.
Further assessment of the scoring system needs to be undertaken to determine an
acceptable weighting system for the three components: Materials, Physical habitat and
Vegetation, to ensure that each component is given equal importance within the overall
system.
SMURF, Sustainable Management of Urban Rivers
and Floodplains, is supported by the EC LIFEEnvironment
programme (Ref LIFE02 ENV/UK/000144).
The project is a partnership between Birmingham City
Council, Environment Agency, Severn Trent Water,
HR Wallingford, Staatliches Umweltamt Herten,
University of Birmingham and King’s College London.