Page 1
A case independent approach on the impact of climate
change effects on combined sewer system performance
M. Kleidorfer, M. Moderl, R. Sitzenfrei, C. Urich and W. Rauch
ABSTRACT
M. Kleidorfer
M. Moderl
R. Sitzenfrei
C. Urich
W. Rauch
Unit of Environmental Engineering,
Faculty of Civil Engineering,
University of Innsbruck,
Technikerstrasse 13,
A6020 Innsbruck,
Austria
E-mail: [email protected] ;
[email protected] ;
[email protected] ;
[email protected] ;
[email protected]
Design and construction of urban drainage systems has to be done in a predictive way, as
the average lifespan of such investments is several decades. The design engineer has to
predict many influencing factors and scenarios for future development of a system (e.g.
change in land use, population, water consumption and infiltration measures). Furthermore,
climate change can cause increased rain intensities which leads to an additional impact on
drainage systems. In this paper we compare the behaviour of different performance indicators
of combined sewer systems when taking into account long-term environmental change effects
(change in rainfall characteristics, change in impervious area and change in dry weather flow).
By using 250 virtual case studies this approach is—in principle—a Monte Carlo Simulation in
which not only parameter values are varied but the entire system structure and layout is
changed in each run. Hence, results are more general and case-independent. For example
the consideration of an increase of rainfall intensities by 20% has the same effect as an
increase of impervious area of + 40%. Such an increase of rainfall intensities could be
compensated by infiltration measures in current systems which lead to a reduction of
impervious area by 30%.
Key words | climate change, combined sewer system, land use, modelling, performance
indicator, rainfall intensities, uncertainties, urban drainage
INTRODUCTION
The use of computer models is a state of the art instrument
in different fields of water resources and environmental
engineering (e.g. Chau 2006). Beside a sufficient model
calibration (e.g. Cheng et al. 2002; Kleidorfer et al. 2009) the
consideration of possible future development scenarios
becomes more and more important. Design and construc-
tion of urban drainage systems has to be done in a
prospective way, as the average lifespan of such investments
is several decades. The design engineer has to assess many
prospective influencing factors e.g. change in land use,
population, water consumption and infiltration. Further-
more climate change can cause increased rainfall intensities
which leads to an additional impact on drainage systems
(Rowell 2005; Arnbjerg-Nielsen & Fleischer 2009). Although
climate change models contain substantial uncertainties, the
consideration of climate change effects has become an
important issue to estimate the possible impact on existing
drainage systems (respectively costs of possible climate
change adaption measures) (Ashley et al. 2005). For example
Butler et al. (2007) evaluated the impact of climate change
on sewer storage tank performance using one case study.
Semadeni-Davies et al. (2008) took into account the impacts
of climate change, as well as the land use change due to
urbanisation.Mark et al. (2008) describes the climate change
adaption strategies of the Scandinavian countries and
presents three case studies, where climate change adaption
plans have been arranged. In one case study even future city
development due to urbanisation was taken into account.
doi: 10.2166/wst.2009.520
1555 Q IWA Publishing 2009 Water Science & Technology—WST | 60.6 | 2009
Page 2
To evaluate climate change impact on combined sewer
systems, performance indicators have to be defined for the
drainage system, e.g. prediction of flooding or prediction of
discharge to receiving water bodies. Impact of climate
change can be analysed under various aspects (e.g. regard-
ing impact on urban flooding or impact on pollutant
discharge). Therefore a multitude of different performance
indicators can be evaluated (Berggren 2008). These per-
formance indicators respond differently on a variation of
boundary conditions, consequently they hardly can be
compared. Hence, when possible future rainfall conditions
are considered for evaluating specific system behaviour (e.g.
urban flooding) the impact on other performance indicators
(e.g. pollutant discharge) is not clear, but can only be
analysed for a specific case study.
In this work we compare the behaviour of different
performance indicators of combined sewer systems when
taking into account long-term environmental change effects
(climate change, land use and population change). Applying
the case study generator presented by Moderl et al. (2009)
250 virtual case studies are generated and subsequently
simulated. While the approach is—in principle—a Monte
Carlo Simulation, not only some parameter values are varied
but the entire system structure and layout is changed in each
run. Further, for each of these generated case studies
parameters are varied to predict environmental change
scenarios. Thus not only one specific case study is assessed
in terms of environmental change effects, but a whole range
of systems. The simulation results are evaluated statistically.
As different performance indicators and their interaction are
taken into account this study helps to understand the
correlation between possible future scenarios and combined
sewer system performance in a case-independent way.
Hence, this study focuses on two questions:
† Is there a correlation between different performance
indicators for evaluating the impact of climate change on
system behaviour? How do adaption strategies for
preventing urban flooding influence combined sewer
overflow (CSO) discharge?
† What is the impact of climate change—induced by
increase of rainfall intensities—as compared to the
impact of future development (urbanisation, land use-
change) and data uncertainties?
METHODS
Model
This work is based on hydrodynamic computer simulation
of combined sewer systems. The model used in this study is
the well known hydrodynamic model SWMM 5.0 (Rossman
2008). Hence the impact on flooding as well as on combined
sewer overflow (CSO) discharge to the receiving water
can be analysed based on a detailed assessment of the
hydraulics in the system.
Case studies
For a case independent conclusion 250 virtual case-studies
(VCSs) with varying system properties were generated using
the case study generator described by Moderl et al. (2009).
The virtual case studies were generated based on a
stochastic approach and designed according to state-of-
the art design guidelines and can be downloaded from the
Institute’s homepage. Of course the VCSs are simplifica-
tions of reality neglecting aspects of real world systems. For
example the VCSs are generated using a branching process
and therefore do not reflect loops in the system. Negative
slopes of conduits were not considered in the used set of
VCSs. Furthermore they are not associated with a natural
urban development which would lead to areas of different
degree of safety for flood protection (i.e. design of the sewer
system on higher return periods in areas of special interest
(e.g. industrial areas) than in rural areas). CSO structures
and detention volumes are located randomly distributed in
the system not only close to the receiving water and the
sewer layout is not interacting with other infrastructure
systems (e.g. water supply systems, roads). Nevertheless
Moderl et al. (2009) showed that the VCSs sufficiently
perform similar as real world systems for prediction of
surface flooding and CSOs (Moderl 2009). Additionally a
real world case study (RWCS) is used for comparison. Some
system characteristics of the real world case study are
presented in Table 1 and Figure 1 shows the system layout
of the real world case study compared to one of the 250
virtual case studies.
Figure 2 shows the 5%, the 50% (median) and the 95%
percentile of empirical cumulative distribution functions
1556 M. Kleidorfer et al. | Climate change effects on combined sewer system performance Water Science & Technology—WST | 60.6 | 2009
Page 3
(CDFs) of the 250 VCSs (grey) and the CDF of the RWCS
(black) for impervious area per node (Aimp/node), for the
conduit diameters and for dry weather flow (DWF/node)
per junction. Hence each CDF illustrates the variance of
these parameters within the case studies. For example in the
RWCS about 50% of the nodes are connected to a conduit
diameter of about 1m or less (Figure 2, middle, black line).
A more detailed analysis and description of system
characteristics of those 250 VCSs is available in (Moderl
2009). This comparison of VCSs and the RWCS shows that
the VCSs are designed in similar way as the RWCS. Only
the distribution of DWF per junction is different as in the
RWCS there is no DWF at each node. Nevertheless, the
impact of this difference is expected to be insignificant.
Rainfall input for the VCSs as well as for the RWCS are
design storm events (Euler II) with different return periods
(RP) (RP ¼ 0.5; 1; 2; 3; 5 and 10 yr) evaluated from real
rainfall data. RP ¼ 10 yr. was chosen as upper limit
according to guiding rules (OWAV-RB 11 2009) as this is
maximum return period required for the proof of sufficient
hydraulic sewer capacity. The rain gauge used is located in
Innsbruck, Austria and the evaluation is described in the
description of the Austrian rainfall database (Rauch &
Kinzel 2007). Innsbruck is a city located in Tyrol, Austria.
The climate is alpine, so the region is characterised by cold
winters and summers with intense rainfall. The duration of
the design storm events is assumed with 120 minutes and
the time step used is 5 minutes. Figure 3 shows diagrams
of the design storm events exemplified for RP ¼ 1 yr and
RP ¼ 10 yr. The Euler II design storms are as well as the
temporal resolution and the length of the rainfall event
chosen according to guiding rules of Austria (OWAV-RB 11
2009) and Germany (DWA-A 118E 2006). Furthermore Lei
(1996) and Rauch et al. (1998) show that 5 minute timesteps
are sufficient for simulation of urban drainage systems.
Applications and a further description of the Euler II design
storm is also available from De Toffol (2009).
Performance indicators (PI)
Usually sewer system performance is expressed by means
of performance indicators PI which should provide key
information needed to assess the efficiency of the system.
A PI is always a qualitative index of a particular aspect.
(De Toffol 2009). In this study we evaluated 7 different
performance indicators according to Moderl (2009) for
combined sewer system behaviour to analyse different
aspects of system performance. Four of them are evaluated
in detail and presented below. Performance indicators PI1,
PI2 and PI3 are emission based and represent CSO
discharge to the receiving water. PI4 represents flooding
in the system during extreme events. All performance
indicators used range between 0 (total system failure) and
1 (perfect system behaviour). While legal and societal
requirements with respect to the performance indicators
have changed in Europe during the last years for PI1, PI2
and PI3 with introduction of the EU Water Framework
Table 1 | System characteristics of real world case study
Parameter Real world case study
Subcatchments 200
Junction nodes 246
Outfall nodes 34
Storage volume 5,100m3
Total area 2,500ha
Impervous area 774ha
Average fraction imperviousness 0.31
Inflow to wastewater treatment plant 2.2m3/s
Figure 1 | Virtual case study (left) and real world case study (right) represented by the Software SWMM 5.0.
1557 M. Kleidorfer et al. | Climate change effects on combined sewer system performance Water Science & Technology—WST | 60.6 | 2009
Page 4
Directive (2000/60/EC 2000), requirements for PI4 expres-
sing a degree of safety regarding flood protection (CEN
1996, 1997) have been almost constant over time. Further
discussion regarding that topic is available for example
from De Toffol (2009) or Engelhard (2008).
Hydraulic CSO efficiency: PI1
The CSO efficiency is used in Austrian guidelines to
evaluate a combined sewer system’s performance over a
simulation period of at least 10 years with recorded rainfall
data. The indicator represents the percentage of surface
runoff which reaches the waste water treatment plant as an
average over the simulation period. Here Euler II design
storms (single events) are simulated to reduce computing
time and so PI1 does not concur with the methodology from
the guidelines. However it can be used to compare
behaviour of systems due to different boundary conditions.
Nevertheless for future studies a comparison with sewer
system performance using long-time rainfall series is
recommended. PI1 is calculated from total overflow volume
of the entire system (VCSO) and total surface runoff
generated (VR) after
PI1 ¼ 12VCSO
VR½0j1�:
Pollutant CSO efficiency: PI2 and PI3
The CSO efficiency for pollutants is a modification of PI1.
Therein the CSO discharge of the pollutant mass (MCSO) is
calculated and normalized by the sum of the pollutant mass
generated by surface runoff (MR) and by dry weather flow
(MDWF). PI2 is calculated for total suspended solids (TSS)
and PI3 for total nitrogen (TN) after
PI2;3 ¼ 12MCSO
MR þMDWF½0j1�:
The two water quality parameters, total suspended
solids (TSS) and total nitrogen (TN) were selected for
Figure 2 | CDFs of VCSs compared to RWCS.
Figure 3 | Design storm event Euler II for RP ¼ 1 yr (left) and RP ¼ 10 yr (right).
1558 M. Kleidorfer et al. | Climate change effects on combined sewer system performance Water Science & Technology—WST | 60.6 | 2009
Page 5
the analysis since they exhibit very different behaviour; TSS
is particulate, while TN is mainly dissolved (Taylor et al.
2005). Although estimation of pollutant concentrations
contains substantial uncertainties (see e.g. Dotto et al.
2008 or Freni et al. 2008 or Kleidorfer et al. 2009 (in press))
the performance indicators can be used to compare
performances of different systems.
Flooding efficiency: PI4
When estimating urban floods due to sewer system over-
loads usually a highly spatial distributed evaluation is
required to identify possible weak points in the system.
Using 1D models often the flooded volume per manhole is
assessed. Here we use a system-wide performance indicator
which also follows the idea of calculating a normalized
efficiency in order to compare system behaviour. Hence,
PI4 is calculated from total ponded volume (VP) and surface
runoff (VR) after
PI4 ¼ 12VP
VR½0j1�:
Environmental change scenarios/data uncertainties
Possible scenarios for future developments of urban
drainage systems cannot be generalised due to completely
different conditions in different regions. For this study the
impact on three main parameters in urban drainage
modelling is considered: dry weather flow (DWF), paved
area (AP) and rainfall intensities (I). The parameter values
for the predictions are taken from literature. As a bandwidth
of possible future scenarios is analysed, a realistic set
of values rather than a perfect prediction is sufficient.
The parameter variation can also be interpreted as
uncertainties in data collection. Table 2 shows the impact
of possible future scenarios and data uncertainties on model
parameters.
Climate change
The change in precipitation is not calculated by means of
climate change models but rather straightforward as given
change of the rain intensity (in variable proportions).
Arnbjerg-Nielsen (2008) calculated climate factors for
consideration of climate change in design of urban drainage
systems for Denmark. Therein three different approaches
are used. He estimates the increase in design intensities as
climate change factor (CCF) by 10–50% depending on
duration, return period and anticipated technical lifetime
of sewer systems. Although this CCF regionally varies
(De Toffol et al. 2006) a bandwidth of increase in design
intensities due to climate change of 10–50% is assumed.
Consequently for each 5min timestep j the rainfall intensity
under consideration of climate change conditions ICCF,j is
calculated from original rainfall intensity Ij and the climate
chance factor CCF after
ICCF;j ¼ Ij·CCF:
Land use change and population change
Land use change due to urbanisation and expansion of
cities has an impact on pavement of urban areas which can
lead to an increase of the fraction imperviousness. But
on the other hand measures for on-site infiltration of
stormwater reduce surface runoff. Dry weather runoff is
influenced by a possible change in population and by a
Table 2 | Environmental change effects and data uncertainties in urban drainage models
Parameter Dry weather flow Paved area Rainfall intensity
Environmental change Change in population Urbanisation/change in land use Climate-change
Change in water consumption
Data uncertainties Measurement uncertainties Determination of fraction of area Spatial distribution
imperviousness Resolution
Measurement uncertainties
1559 M. Kleidorfer et al. | Climate change effects on combined sewer system performance Water Science & Technology—WST | 60.6 | 2009
Page 6
change in water consumption. Hence both parameters—
effective impervious fraction (EIF) and dry weather flow
(DWF)—can increase or decrease in future development. In
this study a bandwidth from 2 60% to þ 60% for both
parameters is assumed. Consequently the parameter for
possible future conditions (effective impervious area EIFfA
or respectively dry weather flow DWFfDWF) is calculated
from current conditions and an area factor fA or a dry
weather flow factor fDWF after
EIFfA ¼ EIF·fA and DWFfDWF¼ DWF·fDWF:
For example a þ 60% increase of impervious area for
the RWCS leads to an increase of the effective impervious
fraction from 0.31 to 0.5.
Simulation and analysis
One real world case study and 250 virtual case studies
are simulated with design storm events (Euler II) of
different return periods. Deviation of impervious area
and dry weather flow is considered in steps
of 2 60%, 2 30%, ^ 0%, þ 30% and þ 60%. Additionally
climate change factors of 1.1, 1.2, 1.3, 1.4 and 1.5 are
analysed.
All together more than 30,000 simulation runs over a
24 hour period are evaluated.
RESULTS AND DISCUSSION
Correlation of performance indicators
Figure 4 shows distributions and correlations of the
performance indicators analysed for the return periods
(RP) 1 (year) and 10 (years) (RP ¼ 1 yr and RP ¼ 10 yr).
Here the dot plots illustrate the correlation between the 4
performance indicators analysed. For example in the plot in
the first column and the fourth row the correlation between
PI1 and PI4 is presented, whereas PI1 is plotted at the
abscissa and PI4 is plotted at the ordinate of the subplot.
The diagonal plots show the distributions of the perform-
ance indicators as histograms. The performance indicators
in all scatter plots range from 0 to 1.
The emission based performance indicators PI1, PI2
and PI3 are clearly correlated, for example high values
for PI1 come along with high values for PI2. On the
other side PI4— which represents flooding—decreases
with increasing emission based performance indicators.
This corresponds with findings by Butler et al. (2008) who
analysed the relationship between flood volume and
receiving water quality in an integrated urban wastewater
system. Increased conduit diameters improve the system
capacity leading to a reduction of PI4. On the other hand
more combined flow can be conveyed downstream
resulting in an increase of CSO discharge (i.e. a decrease
of PI1, PI2 and PI3).
Figure 4 | Distribution of performance indicators for return period 1 yr (left) and 10 yr (right).
1560 M. Kleidorfer et al. | Climate change effects on combined sewer system performance Water Science & Technology—WST | 60.6 | 2009
Page 7
The comparison of different return periods in Figure 4
shows that an increase of the rainfall intensities leads to a
shift of the parameter distributions from high values to
lower values. This effect is highest for PI4. One can clearly
see that for RP ¼ 1 yr PI4 is 1 (or close to 1) for most virtual
case studies, which indicates that no flooding occurs. For
RP ¼ 10 yr the distribution of PI4 changes significantly
indicating that more flooding occurs.
Due to the good correlation between PI1 and both
pollutant performance indicators PI2 and PI3 we assume
that that emission based performance indicators behave
equally. Consequently for a clear arrangement of the
following analysis only PI1 and PI4 are presented.
Figure 5 presents box plots of PI1 and PI4 in order to
illustrate the variance of PI1 and PI4 of the different case
studies and to demonstrate how performance indicators
change with return periods. On each box, the central mark
is the median, the edges of the box are the 25th and 75th
percentiles and outliers of that range are plotted
individually.
Here again one can clearly see how PI1 and PI4 as
expected decrease with increasing rainfall intensities.
Additionally one can see that the evaluation of the
performance indicators of the VCSs cover a wide range of
possible simulation results.
Impact of environmental change scenarios
For a better estimation of impact of different environmental
change scenarios the effects are evaluated pairwise. Per-
formance indicators are (nonlinear) functions of model
structure Q (including all aspects of the model with respect
to the case study analysed as e.g. system layout, DWF,
catchment area …), rainfall event R, CCF, fA and fDWF. For
a specific case study i and a specific rainfall event n PI can
be written in generalised way as a function of CCF, fA and
fDWF for given Q and R
PI ¼ fðCCF; fA; fDEFjQi;RnÞ:
Consequently by equalising sewer system performance
respectively with two factors kept constant, one factor can
be expressed as function of another one. Hence impact of
specific environmental change effects can easily be com-
pared. For example by equalising
fðfA;CCF ¼ 1; fDWF ¼ 1Þ ¼ fðCCF; fA ¼ 1; fDWF ¼ 1Þ
the area factor fA can be expressed as a function of CCF
fA ¼ fðCCFÞ
and finally it is possible to evaluate which factors lead to the
same deviation of performance indicators. This evaluation
is done by means of linear interpolation between calculated
points.
Figure 6 presents the impact of CCF compared to the
impact of fA for the VCSs (median and boxplots) and for the
RWCS for PI1 on the left hand side and for PI4 on the right
hand side. Here is important to note that the figures don’t
present absolute values of PI, but PI is only used to
compare the impact of CCF with respect to the impact of fA.
The figure shows the evaluation for RP ¼ 5 yr but other
return periods behave very similar. Results show for
example that a CCF of 1.2 has the same effect on system
performance as an increase in impervious area of
about þ 40% when regarding PI1. This increase of rainfall
intensities could be compensated by a reduction of
impervious area (e.g. by infiltration measures) by about
30%. We expect that the effect that a reduction of
impervious area can have higher impact on system
performance than an increase of impervious area is true
due to nonlinear relations between system characteristics
and performance indicators (e.g. mobilisation of storage
volume due to backwater effects).
VCSs and the RWCS behave similar. Only for high
fAs . 1.4 the impact of an increase of paved area is
higher for the RWCS compared to VCSs when regarding
PI1. This indicates that impact of possible environmentalFigure 5 | Box plot of PI1 (left) and PI4 (right) for different return periods.
1561 M. Kleidorfer et al. | Climate change effects on combined sewer system performance Water Science & Technology—WST | 60.6 | 2009
Page 8
change scenarios (i.e. change of performance indicators
compared to current conditions) is rather similar for
different systems although absolute value of PIs vary as
shown in Figures 4 and 5.
Impact of DWF is not presented here as analysis of
simulation results showed that a change of DWF has
marginal effects on PI1 and PI4. Furthermore a CCF
cannot be compensated by a reduction of DWF in a
reasonable range.
CONCLUSION
In this study the possible impact of climate change induced
increase of rainfall intensities on combined sewer system
performance is analysed in a case independent way by using
simulation results of 250 virtual case studies and one real
world case study. In addition the impact of possible change
in DWF (e.g. increase by population growth or decrease
water saving measures) and possible change in impervious
area (e.g. increase by further pavement in urban areas or
decrease by a boost of infiltration measures) is taken into
account.
The different performance indicators analysed represent
different system behaviours (e.g. PI1 = CSO discharge;
PI4 = flooding) but show a very similar pattern. The impact
of an increase of rainfall intensities has the highest impact
on system performance followed by the impact of a
variation in impervious area. Variation of DWF marginally
changes system behaviour especially when only minor
changes in future development are realistic.
VCSs and the RWCS behave very similar. This indicates
that impact of possible environmental change scenarios
(i.e. change of performance indicators compared to current
conditions) is rather similar for different systems although
absolute value of PIs vary as shown in Figures 4 and 5.
For example the consideration of a climate change
factor of 1.2 has the same effect as an increase of impervious
area of þ 40%. Such an increase of rainfall intensities by
1.2 could be compensated by infiltration measures in
current systems which lead to a reduction of impervious
area by 30%. Certainly this different behaviour of area
reduction and area increase needs further examination in
upcoming studies.
Concluding, in this paper some general coherences of
sewer system behaviour under future development scen-
arios and climate change scenarios are presented. This is a
contribution to enhanced system understanding taking into
account long-term environmental change effects.
As this study is limited by some (inevitable) short-
comings of the VCSs further research should focus on more
real world case studies with different regional character-
istics and on more realistic VCS. For example Sitzenfrei
et al. (2009) and Urich et al. (2009) develop a agent based
Figure 6 | Comparison of climate change factor and area factor.
1562 M. Kleidorfer et al. | Climate change effects on combined sewer system performance Water Science & Technology—WST | 60.6 | 2009
Page 9
software for generating virtual case studies based on cellular
automata which is expected to represent interactions of
different infrastructure systems in a more realistic way.
Additionally using real rainfall time series instead of design
storm events may lead to an improved analysis especially for
estimating impact on performance indicators related to
CSO discharge.
REFERENCES
2000/60/EC 2000 Directive of the European Parliament and of the
Council establishing a framework for Community action in the
field of water policy.
Arnbjerg-Nielsen 2008 Quantification of climate change impacts on
extreme precipitation used for design of sewer systems. In:
Proceedings of the 11th International Conference on Urban
Drainage, Edinburgh, Scotland, UK, 2008.
Arnbjerg-Nielsen, K. & Fleischer, H. S. 2009 Feasible adaptation
strategies for increased risk of flooding in cities due to climate
change. Water Sci. Technol. 60(2), 273–281.
Ashley, R. M., Balmforth, A. J., Saul, A. J. & Blanskby, J. D.
2005 Flooding in the future—predicting climate change, risks
and responses in urban areas. Water Sci. Technol. 52(5),
265–273.
Berggren, K. 2008 Indicators for urban drainage system—assessment
of climate change impacts. In: Proceedings of the 11th
International Conference on Urban Drainage, Edinburgh,
Scotland, UK.
Butler, D., McEntee, B., Onof, C. & Hagger, A. 2007 Impact of
climate change on sewer storage tank performance. Water Sci.
Technol. 56(12), 29–35.
Butler, D., Fu, G. & Khu, S.-T. 2008 The relationship between
sewer flood volume and receiving water quality in an
integrated urban wastewater system. In: Proceedings of the
BHS 2008, Exeter, UK.
CEN 1996 EN 752—Drain and sewer systems outside buildings—
Part 2: Performance Requirements. European Commitee for
Standardization (CEN), Brussels, Belgium.
CEN 1997 EN 752—Drain and sewer systems outside buildings—
Part 4: Hydraulic design and environmental considerations.
European Commitee for Standardization (CEN), Brussels,
Belgium.
Chau, K. W. 2006 Particle swarm optimization training algorithm
for ANNs in stage prediction of Shing Mun River. J. Hydrol.
329(3–4), 363–367.
Cheng, C. T., Ou, C. P. & Chau, K. W. 2002 Combining a fuzzy
optimal model with a genetic algorithm to solve multi-
objective rainfall–runoff model calibration. J. Hydrol. 268,
72–86.
De Toffol, S. 2009 Sewer system performance assessment—an
indicator based methodology (PhD Thesis). Innsbruck
University Press, Innsbruck.
De Toffol, S., Burstrom, Y. D. S. & Rauch, W. 2006 On the
effect of spatial variances in historical rainfall time series to
CSO performance evaluation. Water Sci. Technol. 54(6–7),
25–31.
Dotto, C., Deletic, A. & Fletcher, T. D. 2008 Analysis of uncertainty
in flow and water quality from a stormwater model. In:
Proceedings of the 11th International Conference on Urban
Drainage, Edinburgh, Scotland, 2008.
DWA-A 118E 2006 Hydraulic Dimensioning and Verification of
Drain and Sewer Systems. Deutsche Vereinigung fur
Wasserwirtschaft, Abwasser und Abfall e.V. (DWA).
Engelhard, C. 2008 Assessing the Impact of Urban Draining
Measures with Regard to the Water Framework Directive (PhD
Thesis). Innsbruck University Press, Innsbruck.
Freni, G., Mannina, G. & Viviani, G. 2008 Uncertainty in urban
stormwater quality modelling: The effect of acceptability
threshold in the GLUE methodology. Water Res. 42(8–9),
2061–2072.
Kleidorfer, M., Moderl, M., Fach, S. & Rauch, W. 2009
Optimization of measurement campaigns for calibration of a
conceptual sewer model. Water Sci. Technol. 59(8),
1523–1530.
Kleidorfer, M., Deletic, A., Fletcher, T. D. & Rauch, W. 2009
Impact of input data uncertainties on urban stormwater model
parameters. Water Sci. Technol. 60(6), 1545–1554.
Lei, J. H. 1996 Uncertainty Analysis of Urban Rainfall–Runoff
Modelling. Norwegian University of Science and Technology,
Department of Hydraulic and Environmental Engineering,
Dissertation.
Mark, O., Svensson, G., Konig, A. & Linde, J. J. 2008 Analyses and
Adaptation of Climate Change Impacts on Urban Drainage
Systems. In: Proceedings of the 11th International Conference
on Urban Drainage, Edinburgh, Scotland, UK, 2008.
Moderl, M. 2009 Modelltechnische Analyse von Netzwerksystemen
der Siedlungswasserwirtschaft. University of Innsbruck, Unit of
Environmental Engineering, Dissertation.
Moderl, M., Butler, D. & Rauch, W. 2009 A stochastic approach for
automatic generation of urban drainage systems. Water Sci.
Technol. 59(6), 1137–1143.
OWAV-RB 11 2009 Richtlinie fur die abwassertechnische
Berechnung und Dimensionierung von Abwasserkanalen; 2.
vollstandig uberarbeitete Auflage. Osterreichisches
Normungsinstitut, Wien.
Rauch, W. & Kinzel, H. 2007 OKOSTRA—Eine Datenbank zur
Ausgabe von hoch aufgelasten Niederschlagsserien aus
Osterreich. Wiener Mitteilungen, pp. I1–I14.
Rauch, W., Thurner, N. & Harremoes, P. 1998 Required accuracy of
rainfall data for integrated urban drainage modeling. Water
Sci. Technol. 37(11), 81–89.
Rossman, L. A. 2008 Storm Water Management Model—User’s
Manual Version 5.0. National Risk Management Research
Laboratory—U.S. Environmental Protection Agency.
Rowell, D. P. 2005 A scenario of European climate change for the
late 21st century: seasonal means and interannual variability.
Clim. Dyn. 25, 837–843.
1563 M. Kleidorfer et al. | Climate change effects on combined sewer system performance Water Science & Technology—WST | 60.6 | 2009
Page 10
Semadeni-Davies, A., Hernebring, C., Svensson, G. & Gustafsson,
L.-G. 2008 The impact of climate change and urbanisation on
drainage in Helsingborg, Sweden: combined sewer system.
J. Hydrol. 350(1–2), 100–113.
Sitzenfrei, R., Fach, S., Kinzel, H. & Rauch, W. 2009 A multi-layer
cellular automata approach for algorithmic generation of
virtual urban water systems—VIBe. In: Proceedings of the 8th
International Conference on Urban Drainage Modelling, 7–11
Sept. 2009.
Taylor, G. D., Fletcher, T. D., Wong, T. H. F. & Breen, P. F. 2005
Nitrogen composition in urban runoff—implications
for stormwater management. Water Res. 39(10),
1982–1989.
Urich, C., Sitzenfrei, R., Moderl, M. & Rauch, W. 2009 An agent
based approach for generating virtual sewer systems in the
software VIBe. In: Proceedings of the 8th International
Conference on Urban Drainage Modelling, 7–11
Sept. 2009.
1564 M. Kleidorfer et al. | Climate change effects on combined sewer system performance Water Science & Technology—WST | 60.6 | 2009