Assessment of Urban Flood Resilience in Barcelona For Current and Future Scenarios. The RESCCUE Project

sustainability
Article
Assessment of Urban Flood Resilience in
Barcelona for Current and Future Scenarios.
The RESCCUE Project
Beniamino Russo 1,2,3, * , Marc Velasco 1 , Luca Locatelli 1 , David Sunyer 1 , Daniel Yubero 1 ,
Robert Monjo 4 , Eduardo Martínez-Gomariz 3,5 , Edwar Forero-Ortiz 5 ,
Daniel Sánchez-Muñoz 6 , Barry Evans 7,8 and Andoni Gonzalez Gómez 9
1 AQUATEC (SUEZ Advanced Solutions), Paseo de la Zona Franca, 46-48, 08038 Barcelona, Spain;
marc.velasco@suez.com (M.V.); luca.locatelli@aquatec.es (L.L.); dsunyer@aquatec.es (D.S.);
dyuberop@aquatec.es (D.Y.)
2 Grupo de Ingeniería Hidráulica y Ambiental (GIHA) (Group of Hydraulic and Environmental Engineering),
Escuela Politécnica de La Almunia (EUPLA, Universidad de Zaragoza) (Technical College of La Almunia,
University of Zaragoza), Calle Mayor, 5, 50100 Zaragoza, Spain
3 Flumen Research Institute, Universitat Politècnica de Catalunya, Jordi Girona 1-3, 08034 Barcelona, Spain;
eduardo.martinez-gomariz@upc.edu or eduardo.martinez@cetaqua.com
4 Fundación de Investigación del Clima (FIC) (Climate Research Fundation), Calle Gran Vía, 22, 28019 Madrid,
Spain; rma@fic.es
5 Cetaqua, Water Technology Centre, Carretera d’Esplugues, 75, 08940 Barcelona, Spain; eaforero@cetaqua.com
6 IREC, Power Systems department, Jardins de les Dones de Negre, 1, 2a pl., 08930 Barcelona, Spain;
dsanchezm@irec.cat
7 Centre for Water Systems, University of Exeter, Exeter EX4 4QF, UK; b.evans@exeter.ac.uk
8 School of Built Environment, College of Sciences, Massey University, Auckland 0745, New Zealand
9 Ajuntament de Barcelona (Barcelona Municipality), Carrer de Torrent de l’Olla 218, 08012 Barcelona, Spain;
agonzalezgom@bcn.cat
* Correspondence: brusso@unizar.es; Tel.: +34-932-479-869

Received: 26 June 2020; Accepted: 6 July 2020; Published: 13 July 2020
Abstract: The results of recent climate projections for the city of Barcelona show a relevant increment
of the maximum rainfall intensities for the period 2071–2100. Considering the city as a system
of systems, urban resilience is strictly linked to the proper functioning of urban services and the
knowledge of the cascading effects that may occur in the case of the failure of one or more critical
infrastructures of a particular strategic sector. In this context, the aim of this paper is to assess
urban resilience through the analysis of the behavior of the main urban services in case of pluvial
floods for current and future rainfall conditions due to climate change. A comprehensive flood risk
assessment including direct, indirect, tangible and intangible impacts has been performed using
cutting edge sectorial and integrated models to analyze the resilience of different urban services
(urban drainage, traffic, electric and waste sectors) and their cascade effects. In addition, the paper
shows how the information generated by these models can be employed to feed a more holistic
analysis to provide a general overview of the city’s resilience in the case of extreme rainfall events.
According to the obtained results, Barcelona could suffer a significant increase of socio-economic
impacts due to climate change if adaptation measures are not adopted. In several cases, these impacts
have been geographically distributed showing the specific situation of each district of the city for
current and future scenarios. This information is essential for the justification and prioritization of
the implementation of adaptation measures.
Keywords: urban resilience; cascading effects; climate change; pluvial floods; 1D/2D coupled models
Sustainability 2020, 12, 5638; doi:10.3390/su12145638 www.mdpi.com/journal/sustainability

Sustainability 2020, 12, x FOR PEER REVIEW 2 of 26
Sustainability 2020, 12, 5638 2 of 25
1. Introduction
1. Introduction
Urban resilience refers
Urban resilience refers toto the
the ability
ability of of an
an urban
urban system—and
system—and all all its constituent socio-ecological
its constituent socio-ecological
and socio-technical networks across temporal and spatial scales—to
and socio-technical networks across temporal and spatial scales—to maintain or rapidly return maintain or rapidly return to to
desired functions in the face of a disturbance, to adapt to change and to quickly
desired functions in the face of a disturbance, to adapt to change and to quickly transform systems transform systems that
limit current
that limit or future
current adaptive
or future capacity
adaptive [1]. [1].
capacity
In this context, a city can be considered
In this context, a city can be considered as as aa system
system of of systems,
systems, and
and itsits urban
urban resilience
resilience is is strictly
strictly
linked
linked toto the
theproper
properfunctioning
functioningofofurban urban services
services and thethe
and knowledge
knowledge of the cascading
of the cascadingeffects that may
effects that
occur in the case of the failure of one or more critical infrastructures of a particular
may occur in the case of the failure of one or more critical infrastructures of a particular strategic strategic sector [2].
Moreover, urban areas are complex systems that cannot be understood
sector [2]. Moreover, urban areas are complex systems that cannot be understood by sectorial and by sectorial and disciplinary
approaches
disciplinary alone [3,4], and
approaches alonethe[3,4],
focusand of smart cities
the focus ofmodels on strengthening
smart cities different sectors
models on strengthening with
different
technological advancement could contribute to building upon a city’s resilience
sectors with technological advancement could contribute to building upon a city’s resilience in terms in terms of dealing
with natural
of dealing hazards
with natural [5].hazards [5].
This
This paper shows how
paper shows howpluvial
pluvialflood
floodurban
urbanresilience
resiliencecancanbebeassessed
assessed byby analyzing
analyzing thethe
behavior
behavior of
critical urban services and the related cascading effects in the case of failures
of critical urban services and the related cascading effects in the case of failures due to heavy storm due to heavy storm events.
With
events.this aim,
With sectorial
this and integrated
aim, sectorial models
and integrated have been
models developed
have been and calibrated
developed and calibrated to analyze
to analyze the
resilience of several urban services in Barcelona for current (baseline scenario)
the resilience of several urban services in Barcelona for current (baseline scenario) and future and future (business as
usual scenario) rainfall conditions up to the horizon of 2100 [6]. In addition,
(business as usual scenario) rainfall conditions up to the horizon of 2100 [6]. In addition, the the information generated
by these models,
information together
generated by with
thesethe historical
models, information
together with theavailable
historicalfor each urbanavailable
information service, has for been
each
used to feed a more holistic model which covers all the urban services
urban service, has been used to feed a more holistic model which covers all the urban services of the city. This twofold
of the
approach, including
city. This twofold risk treatment
approach, (implementation
including risk treatment of adaptation strategies)
(implementation in a comprehensive
of adaptation strategies)flood
in a
risk management process, is presented in Figure
comprehensive flood risk management process, is presented in Figure 1.1.
Strategic urban services

modelling
Climate change &
extreme Events scenarios
Hazard impacts &

cascading effects
Resilience & adaptation

strategies for the market
uptake
Holistic resilience Initial holistic resilience

assessment & assessment to diagnose
management the problems
Simulation of the effects
of climate change
scenarios
Inclusion of Resilience
and adaptation
strategies
Figure 1. Twofold
Figure approach
1. Twofold to achieve
approach an urban
to achieve resilience
an urban assessment
resilience for current
assessment and future
for current scenarios.
and future
scenarios.
A flood-resilient city can be defined as a city which is able to resist, absorb, accommodate
and recover from the city
A flood-resilient effects
canofbea defined
flood hazard in awhich
as a city timelyisand
ableefficient
to resist,manner,
absorb, including
accommodate through
and
the preservation
recover from the and restoration
effects of a floodofhazard
its essential basic structures
in a timely andmanner,
and efficient functions [7]. In this
including context,
through the
flood resilience assessment has been performed in Barcelona through a 1D/2D urban drainage
preservation and restoration of its essential basic structures and functions [7]. In this context, flood model
linked to other
resilience urbanhas
assessment services models to in
been performed evaluate thethrough
Barcelona cascade aeffects
1D/2Dproduced by urban
urban drainage floods
model on
linked
traffic, electric and waste collection systems. The employment of a coupled 1D/2D
to other urban services models to evaluate the cascade effects produced by urban floods on traffic,urban drainage
model
electricproviding
and wasteflow variables
collection (flow depths,
systems. flow velocity
The employment of aand flood1D/2D
coupled extension)
urban ondrainage
urban surfaces
model
during
providingpluvial
flowflood events
variables is essential
(flow depths,toflow
perform tangible
velocity and intangible
and flood extension)riskonassessments.
urban surfacesMoreover,
during
pluvial flood events is essential to perform tangible and intangible risk assessments. Moreover, the
results of these
the results integrated
of these models
integrated havehave
models beenbeen
usedused
to feed a holistic
to feed tool to
a holistic assess
tool the resilience
to assess of theof
the resilience
city
the as a whole.
city as a whole.
The
Thepaper
paperproposes
proposesspecific
specificand
andholistic
holisticapproaches
approachestotoassess
assesspluvial
pluvialflood
floodresilience
resiliencein
inurban
urban
areas.
areas.The
Theapproaches
approachesare arecomplementary
complementaryand andinterconnected
interconnectedand andcan
canbebeused
usedtotounderstand
understandthe the
interrelations
interrelationsbetween
betweenurban
urbanservices
servicesand
andinfrastructures,
infrastructures,asaswell
wellasasrepresenting
representingaavaluable
valuabletool
toolfor
for
decision making.
decision making.
2.2.Materials
Materialsand
andMethods
Methods
2.1.The
2.1. TheEffects
EffectsofofClimate
ClimateChange
Changeon
onMaximum
MaximumRainfall
RainfallIntensity
IntensityininBarcelona
Barcelona
Recently,the
Recently, theClimate
ClimateResearch
ResearchFoundation
Foundation(Fundación
(Fundaciónde deInvestigación
Investigacióndel
delClima;
Clima;hereafter,
hereafter,
FICfrom
FIC fromthe
theacronym
acronymininSpanish)
Spanish)provided
providedclimate
climateprojections
projectionsand
andpredictions
predictionsfor
fordifferent
differentclimate
climate
variables in Barcelona, the results of which are summarized in Figure 2. These results
variables in Barcelona, the results of which are summarized in Figure 2. These results confirm the confirm the
same trends as other previous studies developed for the city and are in line with the data
same trends as other previous studies developed for the city and are in line with the data from the from the
lastClimate
last ClimatePlan
Planpublished
publishedby byBarcelona
BarcelonaCity
CityCouncil
Council[8].
[8].According
Accordingtotothethedata
dataprovided
providedby byFIC,
FIC,
phenomena such as extreme rainfall, heatwaves and droughts could experience significant
phenomena such as extreme rainfall, heatwaves and droughts could experience significant increases increases
duetotoan
due anacceleration
accelerationofofthe
thehydrological
hydrologicalcycle
cycle[6,9].
[6,9].
Extremescompass
Figure2.2.Extremes
Figure compassrose
rosefor
forBarcelona:
Barcelona:maximum
maximumpoint pointchange
changeininextreme
extremeclimate
climateevents
events
over the century, taking into account return periods between 2 and 100 years. The center
over the century, taking into account return periods between 2 and 100 years. The center represents represents
nochanges,
no changes,and
andthetheedge
edgecorresponds
correspondstotoan anincrease
increaseofof100%
100%forforevery
everyvariable
variableexcept
exceptfor
forheat
heatwave
wave
days(the
days (theborder +1000%)and
borderisis+1000%) andextreme
extremetemperature
temperature(the
(theborder +10°C).
borderisis+10 ◦ C).Thick
Thicklines
linesrepresent
representthe
the
median scenario, and the shaded area is the uncertainty region
median scenario, and the shaded area is the uncertainty region (5–95%). (5–95%).
Particularly, in the case of maximum rainfall intensities and the horizon of 2071–2100 for the city
Particularly, in the case of maximum rainfall intensities and the horizon of 2071–2100 for the city
of Barcelona, the value of the coefficient of climate change (defined as the ratio between future and
of Barcelona, the value of the coefficient of climate change (defined as the ratio between future and
current maximum intensities, for certain return periods and time intervals) [10,11] was found to be
current maximum intensities, for certain return periods and time intervals) [10,11] was found to be in
in a range between 1.07 and 1.26 depending on the frequency and duration of each maximum rainfall
a range between 1.07 and 1.26 depending on the frequency and duration of each maximum rainfall
intensity (Figure 3 and Table 1).
intensity (Figure 3).
1.30
Climate change
1.25
coefficients
1.20
1.15
1.10
1.05
1.00
5 10 15 20 25 30 35 40 45 50 55 60
Duration (in minutes)
T1 T2 T5 T10 T25 T50 T100 T500
Figure 3. Fiftieth percentiles of the climate change coefficients obtained for different rainfall durations
Figure 3. Fiftieth percentiles of the climate change coefficients obtained for different rainfall durations
and return periods for the horizon 2071–2100 for the city of Barcelona [6].
and return periods for the horizon 2071–2100 for the city of Barcelona [6].
These results were obtained using statistical spatial and temporal downscaling techniques on
Table 1. Loosely coupled models developed and used for the flood resilience analysis in Barcelona.
20 future pluviometric series provided by 10 general atmospheric circulation models, forced by
Representative
Loosely Coupled Concentration Pathway (RCP) 4.5 and 8.5 scenarios and previously validated for a
Involved Sectors Main Purposes
historicalModel
control period (1976–2005) [6]. The climate change coefficients in Figure 3 represent the 50th
percentile of the results obtained. Flood hazard assessment and socio-economic
Once these
1D/2D climate change
coupled coefficients
Urban were obtained,flood
drainage they risk
wereassessment
applied to for
synthetic
peoplestorms
and with
different return periods (T1, T10, T50, T100 and T500) used for the last drainage propertiesmaster plan of the city
of Barcelona [11] obtained through
Flooding—traffic the alternating
Urban drainage and blocks method.ofFigure
Assessment flood 4hazard
showsandthe flood
urbanimpacts
drainage
Barcelona project
model design storm for a return
Surface traffic period of 10 years with a duration of
on traffic systemapproximately 2h
and 30 min
Sustainability after
2020,
Flooding—electric the
12, x application
FOR PEER of
REVIEW climate
Urban drainage andchange coefficients for each different rainfall duration.
Assessment of flood hazard and flood impacts26 5 of
model electric system on electric system

200
Flooding—waste Urban drainage and Assessment of flood hazard on waste
180
collecting
160 model waste collecting model collecting system
Intensity (mm/h)
140
These120results were obtained using statistical spatial and temporal downscaling techniques on 20
100
future pluviometric series provided by 10 general atmospheric circulation models, forced by
80
Representative Concentration Pathway (RCP) 4.5 and 8.5 scenarios and previously validated for a
60
historical control
40
period (1976–2005) [6]. The climate change coefficients in Figure 3 represent the 50th
percentile of20 the results obtained.
Once these
0 climate change coefficients were obtained, they were applied to synthetic storms
0:05
0:10
0:15
0:20
0:25
0:30
0:35
0:40
0:45
0:50
0:55
1:00
1:05
1:10
1:15
1:20
1:25
1:30
1:35
1:40
1:45
1:50
1:55
2:00
2:05
2:10
2:15
2:20
2:25
2:30
2:35
2:40
with different return periods (T1, T10, T50, T100 and T500) used for the last drainage master plan of
the city of Barcelona [11] obtained through the alternating blocks method. Figure 4 shows the urban
Elapsed time
drainage Barcelona project design storm for a return period of 10 years with a duration of
Figure 4.
approximately 4. Urban
Urban drainage
drainage
2 h and Barcelona
30 min project
Barcelona
after the design
project storm
design
application withwith
storm
of a return
climate periodperiod
a return
change of 10 years and
of 10
coefficients a duration
foryears
eachand a
different
of 2 h and
duration 35 min [6].
of 2 h and 35 min [6].
rainfall duration.
2.2. 1D/2D Coupled Approaches for Urban Pluvial Modelling
2.2. 1D/2D Coupled Approaches for Urban Pluvial Modelling
Urban areas have a complex topography and contain small-scale elements such as streets and
Urban areas have a complex topography and contain small-scale elements such as streets and
buildings that are usually not taken into account in standard river floodplain studies [12]. Therefore,
buildings that are usually not taken into account in standard river floodplain studies [12]. Therefore,
a higher resolution is required to represent features at the city scale, although this may lead to larger
a higher resolution is required to represent features at the city scale, although this may lead to larger
computational time, notwithstanding the fact that urban model areas are generally smaller than a river
computational time, notwithstanding the fact that urban model areas are generally smaller than a
floodplain. For all of these reasons, urban, pluvial flooding requires a different modeling approach
river floodplain. For all of these reasons, urban, pluvial flooding requires a different modeling
than the one used for fluvial flooding [12].
approach than the one used for fluvial flooding [12].
During the last two decades, several authors have published papers about the need to develop
and use urban stormwater models (USMs) based on coupled approaches (the modeling of the surface
and sewer flows at the same time by 1D/1D or 1D/2D models) to represent adequately urban flood
caused by surcharged sewers [13–16] and carry out realistic flood risk assessments [17].
Although the choice between using a 1D or a 2D surface overland flow model (to be coupled to
During the last two decades, several authors have published papers about the need to develop
and use urban stormwater models (USMs) based on coupled approaches (the modeling of the surface
and sewer flows at the same time by 1D/1D or 1D/2D models) to represent adequately urban flood
caused by surcharged sewers [13–16] and carry out realistic flood risk assessments [17].
Although the choice between using a 1D or a 2D surface overland flow model (to be coupled to a
1D sewer model) determines the accuracy of results and the computational time required to obtain
them, when the flow overtops street curbs and does not remain within the street profile, using a 2D
model is crucial [12,18].
In 1D/2D USMs, the underground sewer network is represented by a 1D sewer model while the
surface flow is computed using a 2D model. The 2D model reproduces the urban surface topography
and is essential to achieve a more realistic simulation of the flow spreading across complex urban
surfaces, with results such as flow depths and velocities anywhere in the urban model area [12].
USM can be semi-distributed (SD) or fully distributed (FD). SD models, commonly applied in
urban stormwater modeling, are based on subcatchment units where rainfall is applied, while runoff
is estimated and routed according to specific hydrological losses and rainfall–runoff transformation
methods. FD models, which are generally more detailed and theoretically more realistic, are based on
the two-dimensional (2D) discretization of the overland surface, where runoff volumes are estimated
and directly routed by the 2D overland flow module [19]. Both kinds of approaches can be followed to
create 1D/2D coupled models that are able to simulate, at the same time, the behavior of the sewer
system and the urban surfaces and their mutual interaction in case of pluvial flooding events (Figure 5).
Finally, hybrid models (H) can account for runoff produced by rainfall which is directly applied from
subcatchment units formed by building areas (roofs, terraces and courtyards) and directly conveyed
into the sewer systems; for the other impervious (streets, sidewalks, squares, etc.) and pervious
(parks and natural areas) urban surfaces, the 2D overland flow model computes and routes the runoff
produced by the rainfall directly applied to these surfaces [20]. These approaches are represented in
Figure 5.
(a) SD (b) FD (c) H
Figure
Figure5. 5.Scheme
Schemeofofsemi-distributed (SD) (a),
semi-distributed (SD) (a), fully-distributed
fully-distributed(FD)
(FD)(b)(b)and
and hybrid
hybrid (H)(H)
(c)(c) 1D/2D
1D/2D
coupled
coupledurban
urbanstormwater
stormwatermodel
model (USM)
(USM) approaches (adaptedfrom
approaches (adapted from[19]).
[19]).InInbrown,
brown, subcatchment
subcatchment
units are
units arerepresented,
represented,while
whileblue
bluelines
lines and
and arrows indicatethe
arrows indicate thepathway
pathwayofofthe the runoff
runoff from
from thethe source
source
(subcatchment units or discretized 2D surface)
(subcatchment units or discretized 2D surface) to the sewer system.
sewer system.
TheTheamount
amountofofrunoff
runoffentering
entering the
the underground sewernetwork
underground sewer networkisislimited
limitedbybythethe hydraulic
hydraulic
efficiency of surface drainage structures (inlets, transversal grates, etc.) [21–23] and their
efficiency of surface drainage structures (inlets, transversal grates, etc.) [21–23] and their state state
of of
maintenance and clogging [23,24], although these aspects are often neglected in
maintenance and clogging [23,24], although these aspects are often neglected in urban drainage urban drainage
models
models [19].
[19].Generally,
Generally,SD
SDmodels
models apply the runoff
apply all the runoff estimated
estimatedinina agiven
givensubcatchment
subcatchment directly
directly
into the
into theselected
selectedcomputational
computational node
node ofof the sewer system, without
sewer system, withoutaccounting
accountingforforthethe hydraulic
hydraulic
capacity
capacityofofsurface
surfacedrainage
drainagecapacity.
capacity. With this
this assumption,
assumption,this
thiskind
kindofofmodel
model only
only considers
considers thethe
flooding
flooding thatoccurs
that occurswhen
whenthe
thesewer
sewer system
system surcharges
surchargesandandneglects
neglectsurban
urbanfloods produced
floods produced by by
thethe
poor capacity of inlet systems [19]. On the contrary, FD modelling packages, such as Infoworks ICM
(Integrated Catchment Modeling) software [25], can take into account the hydraulic performance of
surface drainage systems connecting network nodes with the 2D overland surface mesh by weirs,
orifices and other experimental equations [20,21].
poor capacity of inlet systems [19]. On the contrary, FD modelling packages, such as Infoworks ICM
(Integrated Catchment Modeling) software [25], can take into account the hydraulic performance of
surface drainage systems connecting network nodes with the 2D overland surface mesh by weirs,
orifices and other experimental equations [20,21].
2.3. Barcelona Semi-Distributed 1D/2D USM

After the two first investigations concerning the development and calibration of a detailed 1D/2D
USM in Barcelona, covering approximately half of the administrative land of the city (more than
50 km2 ) [20,26], a new SD model has been developed and calibrated in the framework of this work [2]
and the new drainage master plan of the city (PDISBA, from the acronym in Spanish) [11,27].
The large amount of effort related to the analysis of the deficit of surface drainage systems through
the last two drainage master plans of the city and the consequent implementation of thousands of
inlets in all the urban catchments allow the assumption that stormwater could be quickly introduced
into the sewer system by avoiding uncontrolled runoff circulation and aiming to develop an SD 1D/2D
USM (referred to as 1D/2D USM in the following).
Moreover, the new 1D/2D USM presents two relevant improvements with respect to the previous
ones: the model includes the main and secondary sewer network, reaching a total length of 1650 km of
pipes, and covers the whole hydrological area of the city (administrative land and upstream surfaces),
exceeding 120 km2 of model domain.
The model, with more than 85,000 nodes and a discretized 2D domain in an unstructured mesh of
more than 1,360,000 cells, was developed through the Infoworks ICM (Integrated Catchment Modeling)
software (www.innovyze.com) [25] and was calibrated and validated using the historical data recorded
by more than 100 flow depth gauges located in the city’s sewage network and more than 20 rain gauges
distributed in the analyzed domain using Thiessen polygons [11,27]. The average size of the 2D cells
for overland flow modeling is in the range of 25–100 m2 .
The model required high-quality topographic information (physical data from the network,
digital terrain model 2 × 2 m with a resolution in height of approximately 15 cm) and phenomenological
information (rainfall data and flow level for the calibration phase), in addition to an adequate hardware
configuration to reduce computation time during numerical simulations [20].
The new 1D/2D USM allowed the estimation of flow variables (flow depth, flow velocity and flood
extension) on the surface prone areas by several numerical simulations of historic events and synthetic
storm hyetographs for current and future scenarios. These values were used for the flood hazard and
intangible risk assessment (concerning pedestrian and vehicular circulation) and the evaluation of
tangible direct and indirect impacts. The same outputs were also used to feed other integrated models
of critical urban services to assess the cascading effects of floods in these sectors.
2.4. Modeling of the Effects of Pluvial Floods on Several Urban Services

Projections of rainfall and sea level rise were used to feed the 1D/2D coupled USM to analyze the
hydraulic behavior of the underground sewer system and the overland flood-prone areas in the case of
pluvial floods for current and future rainfall conditions.
Concerning the assessment of multiple hazards and risks, the proposed methodology is based on
the development of coupled models and tools (“loosely or integrated models”); thus, the outputs of
certain models are used as inputs in others according to the scheme presented in Figure 6 [27].
Projections of rainfall and sea level rise were used to feed the 1D/2D coupled USM to analyze
the hydraulic behavior of the underground sewer system and the overland flood-prone areas in the
case of pluvial floods for current and future rainfall conditions.
Concerning the assessment of multiple hazards and risks, the proposed methodology is based
on the development
Sustainability 2020, 12, 5638of coupled models and tools (“loosely or integrated models”); thus, the outputs
7 of 25
of certain models are used as inputs in others according to the scheme presented in Figure 6 [27].
Figure 6. Scheme of loosely coupled model approach used for the multi-hazards
multi-hazards assessment
assessment [27].
[27].
The aim
aim ofof the
the developed
developed loosely
loosely coupled
coupled models
models was
was the
the assessment
assessment of of multi-hazards
multi-hazards and
and
multi-risks (including direct and indirect impacts) produced by urban pluvial floods
multi-risks (including direct and indirect impacts) produced by urban pluvial floods and the and the cascading
effects on other
cascading effectsurban services
on other urban(electrical system, waste
services (electrical collection
system, system and
waste collection surface
system andtraffic system).
surface traffic
Table 1 summarizes
system). the analyzed
Table 1 summarizes the services
analyzed affected byaffected
services pluvial floods in Barcelona,
by pluvial floods inthe behaviorthe
Barcelona, of
which was based on the developed loosely coupled models [27]. Figure 7
behavior of which was based on the developed loosely coupled models [27]. Figure 7 shows the shows the analyzed
interrelationships to assess the
analyzed interrelationships resilience
to assess of some main
the resilience city main
of some services
cityinservices
the case inof urban
the case flooding
of urban
episodes [28].
flooding episodes [28].
Table 1. Loosely coupled models developed and used for the flood resilience analysis in Barcelona.
Loosely Coupled Model Involved Sectors Main Purposes

Flood hazard assessment and socio-economic
1D/2D coupled model Urban drainage
flood risk assessment for people and properties
Assessment of flood hazard and flood impacts on
Flooding—traffic model Urban drainage and surface traffic
traffic system
Assessment of flood hazard and flood impacts on
Flooding—electric model Urban drainage and electric system
electric system
Flooding—waste collecting Urban drainage and waste collecting Assessment of flood hazard on waste collecting
model model system
Figure 7.
Figure 7. Diagram
Diagramofofthe
theimpact analyses
impact carried
analyses out out
carried within this paper
within and the
this paper andpotential cascading
the potential effects.
cascading
effects.
2.5. Social Flood Impacts Model
Pluvial
2.5. Social flood
Flood impacts
Impacts can be classified into tangible and intangible impacts and direct and indirect
Model
impacts [29]. In this study, socio-economic impacts produced by pluvial flooding have been assessed
Pluvial flood impacts can be classified into tangible and intangible impacts and direct and
indirect impacts [29]. In this study, socio-economic impacts produced by pluvial flooding have been
assessed according to comprehensive and detailed methodologies carried out and implemented in
previous investigations in several urban areas [29,30].
In the social field, for the assessment of the intangible impacts, human risk focuses on the safety
effects.
2.5. Social Flood Impacts Model

Pluvial flood impacts can be classified into tangible and intangible impacts and direct and
indirect impacts
Sustainability [29].
2020, 12, 5638In this study, socio-economic impacts produced by pluvial flooding have 8been of 25
assessed according to comprehensive and detailed methodologies carried out and implemented in
previous investigations in several urban areas [29,30].
according
In the to comprehensive
social field, for the and detailedofmethodologies
assessment carried out
the intangible impacts, and implemented
human risk focuses onintheprevious
safety
investigations in several urban areas [29,30].
of pedestrians and vehicles exposed to pluvial flooding events. Risk is defined as the combination of
hazardIn and
the social field, foraccording
vulnerability the assessment of the intangible
to the approach proposedimpacts, human
by Turner risk
et al. focuses
[31] on the safety
and implemented
of pedestrians and vehicles exposed to pluvial flooding events. Risk is defined
in previous studies [29,30]. According to this, hazard assessment is based on the severity and as the combination of
hazard and vulnerability according to the approach proposed by Turner et
frequency of the hydrodynamic variables and is classified based on specific flood hazard al. [31] and implemented in
previous studies
experimental [29,30].
criteria According
regarding to this, and
pedestrian hazard assessment
vehicular is based
stability on the
in urban severity
flooded and (Figure
areas frequency8)
of the hydrodynamic
[27,32–34]. variables and is classified based on specific flood hazard experimental criteria
regarding pedestrian and vehicular stability in urban flooded areas (Figure 8) [27,32–34].
(a) (b)
Figure 8.
Figure Experimentalflood
8. Experimental floodhazard
hazard criteria
criteria for
for (a)
(a) pedestrians
pedestrians and
and (b)
(b) vehicles.
vehicles.
Regarding flood
Regarding flood vulnerability
vulnerability for
for pedestrians,
pedestrians,itit is
is considered
considered toto be
be aa function
function of
of exposure
exposure and
and
sensitivity, taking into account several indicators such as demographic density, the
sensitivity, taking into account several indicators such as demographic density, the percentage of percentage of
people of a critical age and of foreign inhabitants and the number of critical infrastructures. By
people of a critical age and of foreign inhabitants and the number of critical infrastructures. By setting setting
thresholds for
thresholds for the
the proposed
proposed indicators,
indicators, the
the vulnerability
vulnerability of of each
each census
census district
district can
can be
be qualitatively
qualitatively
scored and classified as low, medium and high. On the other hand, in order to assess the
scored and classified as low, medium and high. On the other hand, in order to assess the vulnerabilityvulnerability
for vehicular circulation, the exposure for each urban street, expressed in terms of vehicular daily
intensity, is considered. Based on this value, flood vulnerability regarding vehicular circulation
is qualitatively scored and classified as low, medium and high, in a similar manner to pedestrian
vulnerability [28–30].
Methods for risk determination can be qualitative or quantitative, with both having limitations.
If we define risk as the probability or threat of a hazard occurring in a vulnerable area, flood risk can
be assessed through a flood risk map related to a determined scenario and return period by combining
hazard and vulnerability maps [29,30]. Pedestrians and vehicles are expected to be the most potentially
affected by floods in Barcelona. Their risk is related to their stability, and in Barcelona, this is assessed
for the present (baseline) and also for the future (business as usual (BAU)) scenarios according to the
rainfall variable projections for different return periods.
Qualitative risk assessment defines hazards, vulnerability and risk levels by significance levels
such as “high”, “medium” and “low” and evaluates the resultant level of risk against qualitative criteria.
In this case, hazard and vulnerability maps are generally elaborated through specific criteria and
indexes, and so risk maps will be created by multiplying the vulnerability index (1, 2 or 3, corresponding
to low, medium and high vulnerability) by the hazard index (1, 2 or 3, corresponding to low, medium
and high hazard). Finally, the total risk varies from 1 to 9, where higher levels indicate higher risk
according to the following risk matrix (Figure 9) previously employed in other works [29,30].
criteria. In this case, hazard and vulnerability maps are generally elaborated through specific criteria
and indexes, and so risk maps will be created by multiplying the vulnerability index (1, 2 or 3,
corresponding to low, medium and high vulnerability) by the hazard index (1, 2 or 3, corresponding
to low, medium and high hazard). Finally, the total risk varies from 1 to 9, where higher levels
indicate higher
Sustainability risk
2020, 12, according to the following risk matrix (Figure 9) previously employed in other
5638 9 of 25
works [29,30].
Hazard
Low Medium High
Low Low Low Medium
Vulnerability Medium Low Medium High
High Medium High High
Figure 9. Proposed
Figure 9. Proposed flood
flood risk
risk matrix
matrix for
for pedestrians and vehicles
pedestrians and vehicles [28,30].
[28,30].
2.6. Economic Flood

2.6. Economic Flood Impacts
Impacts Models
Models
Regarding
Regarding economic flood risk
economic flood risk assessment
assessment in in Barcelona,
Barcelona, tangible
tangible direct
direct and
and indirect
indirect damage
damage
were
were considered
considered[28].[28]. Specifically,
Specifically,tailored
tailoredflood
flooddepth–damage
depth–damagecurvescurves were
were developed
developed forfor
thethecase of
case
Barcelona [35,36] and used to feed a detailed damage model regarding properties and
of Barcelona [35,36] and used to feed a detailed damage model regarding properties and vehicles (the vehicles (the two
most affected
two most assetsassets
affected by pluvial floodsfloods
by pluvial in the in
city).
the The
city).model was already
The model successfully
was already appliedapplied
successfully for the
city of Badalona [30,37] and validated using insurance claims according to the data
for the city of Badalona [30,37] and validated using insurance claims according to the data received received from the
Spanish public insurance company “Consorcio de Compensación de Seguros
from the Spanish public insurance company “Consorcio de Compensación de Seguros (CCS)” [28]. (CCS)” [28]. This public
entity covers
This public all the
entity damage
covers produced
all the damageby extraordinary
produced events, such
by extraordinary as damage
events, related
such as damage to natural
related
hazards (e.g., pluvial floods). On the other hand, indirect damages were assessed
to natural hazards (e.g., pluvial floods). On the other hand, indirect damages were assessed by an by an econometric
model, achieving
econometric model, a constant
achieving relation between
a constant direct
relation and indirect
between tangible
direct and damage.
indirect This
tangible model This
damage. was
also
modelvalidated
was alsousing field data
validated using[28].
field data [28].
2.7. Integrated Flooding–Surface

2.7. Integrated Flooding–Surface Traffic
Traffic Model
Model
The
The simulations
simulations related
related to
to the
the integrated
integrated flooding–surface traffic model
flooding–surface traffic model in in Barcelona
Barcelona were
were
carried out using a mesoscale model and the TransCAD Transportation
carried out using a mesoscale model and the TransCAD Transportation Planning Planning software (https:
software
//www.caliper.com/tcovu.htm)
(https://www.caliper.com/tcovu.htm) adopted by city council’s
adopted by mobility department
city council’s (https://www.caliper.
mobility department
com/tcovu.htm) [38]. The mesoscale model simulated the vehicular flow
(https://www.caliper.com/tcovu.htm) [38]. The mesoscale model simulated the vehicular flow in each link of theinstreet
each
network; each link contained detailed information regarding the volume of traffic,
link of the street network; each link contained detailed information regarding the volume of traffic, its typology (for
example:
its typology (for example: number of cars, trucks, bicycles, etc.), travel time, the residual capacityetc.
number of cars, trucks, bicycles, etc.), travel time, the residual capacity of the section, of
Flood maps produced
the section, etc. Floodthrough the city’s 1D/2D
maps produced through USMthe were
city’sused as inputs
1D/2D USM forwere a dynamic
used as traffic
inputsmodel
for a
to estimate
dynamic the effects
traffic modelwithin the city’s
to estimate thesurface
effectstransportation network
within the city’s produced
surface by pluvial
transportation floods.
network
A recent study
produced developed
by pluvial floods.byAPyatkova
recent study et al. [39] analyzed
developed how theetflow
by Pyatkova depth
al. [39] information
analyzed how the can
flowbe
used as criteria to approximate the reduction of vehicular free-flow speeds to 20 kmh −1 along streets
depth information can be used as criteria to approximate the reduction of vehicular free-flow speeds
that
to 20have
kmhstanding water, that
−1 along streets withhave
a reduction
standing 0 kmh−1
to water, where
with the watertodepths
a reduction 0 kmhexceed
−1 where a threshold
the water
value (Table 2). As previously mentioned, the mesoscale traffic model contained information on a wide
number of parameters relating to traffic flows for each road section, including the maximum speeds
allowed on each section. To simulate the effect of flooding within the traffic model, we needed to adjust
the maximum allowable speed parameters based on food model outputs. This approach involved
using geospatial analysis as a precursor to modify the input data of a traffic model, as outlined in
Evans et al. [38], where the outputs from the 1D/2D USM were used to spatially define vehicular speed
restrictions along the road network. The results from the traffic model run under flooded conditions
were compared to the benchmark traffic model (the traffic model run under dry weather conditions)
and the impacts in terms of relative disruption to traffic flows were analyzed.
Table 2. Flood hazard effects on traffic flow (from [38,39]).
Flood Depth Range (m) Hazard Classification Maximum Vehicle Speed (km/h)
Flow depth < 0.1 Low Road speed limit
0.1 < Flow depth < 0.3 Medium 20
Flow depth > 0.3 m High 0 (Road closed)
Table 2. Flood hazard effects on traffic flow (from [38,39]).
Flood Depth Range (m) Hazard Classification Maximum Vehicle Speed (km/h)
Flow depth < 0.1 Low Road speed limit
0.1 < Flow depth
Sustainability 2020, 12, 5638
< 0.3 Medium 20 10 of 25
Flow depth > 0.3 m High 0 (Road closed)
Here, flood hazard analysis

analysis was performed
performed by aa GIS
GIS (Geographic
(Geographic Information
Information System)
System) spatial
spatial
analysis of the flooded road links; the rules applied in relation to traffic speed reductions are outlined
in Table
Table 2,
2, and the flow depths were provided
provided by the 1D/2D coupled
coupled USM.
USM. The results of this analysis
were
were used
usedtotoselect
selectand
andmodify thethe
modify maximum
maximum allowable speed
allowable limits
speed for flooded
limits roadsroads
for flooded withinwithin
the traffic
the
model based on the flood hazard. Using these new input parameters, the traffic model
traffic model based on the flood hazard. Using these new input parameters, the traffic model was run was run and
the
andresults compared
the results to the normal
compared to the n(dry weather)
normal (drytraffic modelling
weather) trafficconditions.
modelling Detailed information
conditions. Detailed
about the approach can be found in the work published by Evans et al. [38].
information about the approach can be found in the work published by Evans et al. [38].
2.8. Integrated Flooding–Electric

2.8. Integrated Flooding–Electric System
System Model
Model
This model analyzed
This model analyzed thethepotential
potentialeffects
effectsofofpluvial
pluvialfloods
floodsononthe
theelectrical
electrical system
system of of
thethe city
city of
of Barcelona, with special emphasis on critical infrastructures such as high and
Barcelona, with special emphasis on critical infrastructures such as high and medium-voltage medium-voltage
substations,
substations, as
as well
well as
as distribution
distribution centers,
centers, taking
taking into
into account
account the
the possible
possible effects
effects of
of climate
climate change.
change.
The model was designed for the hazard, risk and cost assessment of the electrical
The model was designed for the hazard, risk and cost assessment of the electrical assets, assets, as shownas
in Figure
shown in 10.
Figure 10.
Energy Non Supplied

Cost (ENSC)
Substation Failure
Probability (PF)
Electrical Percentage of Auxiliary Generation
Defining
Total cost
locations Sample of area afected Cost (AGC)
substations
water
Potentially
coupled 1D/2D depth Fragility Curve Effective Damage
affected
USW model probability Cost (EDC)
Bussines Cost (BC)
MAIN INPUTS HAZARD ASSESSMENT RISK ASSESSMENT COST ASSESSMENT
Figure 10.
Figure 10. Integration
Integration of
of GIS
GIS spatial
spatial analysis
analysis for flood assessment
for flood assessment on
on the
the electrical
electrical model.
model.
In particular,
particular, the
the flooding
flooding hazard
hazard level
level of each electrical infrastructure
infrastructure waswas assessed on the basis
on flood influence
influence areas
areas of
of55m,m,2525mmandand3030mmininradius with
radius withrespect
respectto to
their location
their depending
location depending on
thethe
on asset type
asset (distribution
type (distribution center
center(DC),
(DC),medium-voltage
medium-voltage(MV) (MV)andand high-voltage
high-voltage (HV) substation
respectively),as
respectively), aswell
wellasasconsidering
consideringthe theflow
flow depths
depths values
values every
every 2 m2 in
morder
in order to avoid
to avoid locallocal errors
errors and
and potential
potential uncertainties
uncertainties of theofelectrical
the electrical asset location
asset location and ofand
the of the source
source data provided
data provided by the
by the 1D/2D
1D/2Dmodel.
USW USW Inmodel. In addition,
addition, a 10 cm threshold
a 10 cm threshold was used to was used to
consider consider local
significant significant local
flooding. flooding.
Using these
parameters, the flood affections were classified as complete, partial or null, quantifying the percentage
of flooded surface in each area of influence of each electric infrastructure according to the methodology
proposed by Sánchez et al. [28,40].
One of the most important uncertainties of this model was the lack of knowledge about the
specific location of critical electrical infrastructures (sometimes located on surfaces and at other times
underground or with self-protection elements which were not always known).
For the impacts analysis, a vulnerability curve (known as a fragility curve in the energy sector)
of the electrical infrastructure proposed by the Federal Emergency Management Agency [41] was
used. The curve relates the probability of failure of an electrical infrastructure to the flood depth.
Furthermore, this curve was partially modified to carry out a sensitivity analysis of the final results
regarding this input [40]. The results obtained from the analysis of the percentage of flooding surface in
each area and from the fragility curve were computed to obtain a probability of failure, later categorized
into four different risk categories as shown in Table 3.
Table 3. Failure probabilities for electrical assets.
Probability Range Categorical Description

PF < 0.01 Low Failure Probability (LFP)
0.01 < PF ≤ 0.1 Moderate Failure Probability (MFP)
0.1 < PF ≤ 0.5 High Failure Probability (HFP)
PF > 0.5 Non-Acceptable Failure Probability (NAFP)
The cost assessment was based on estimations based on GIS computing; furthermore,
we established the supply area of each electrical location using Thiessen polygons and obtained
the power supplied through an estimation of the consumers per area based on the census of the city.
Based on these estimations, it was possible to extract the number of consumers affected and the
time needed to repair the substation as well as the cost of the energy not supplied, the cost incurred by
businesses, auxiliary generation and the damage received by the location [40].
2.9. Integrated Flooding–Waste Collection System Model

In the case of pluvial floods, waste containers can lose their stability due to buoyancy, dragging or
overtopping, thereby allowing debris and leachate to escape from the containers and contaminate the
floodwater and the environment [42]. On the other hand, the containers displaced by the flow can
obstruct superficial drainage pathways or obstruct narrow streets, exacerbating the effects of the flood.
Consequently, waste containers’ stability in the case of pluvial flooding is definitely an environmental,
safety and health concern which needs to be addressed in a context of urban flood resilience assessment.
In order to analyze the significance of this problem in Barcelona, an integrated flooding–waste
collection system model was developed and validated. In the city, there are more than 27,000 containers,
which can be classified according to the type of waste they contain (waste, organic, paper and cardboard,
plastic and packaging and glass), their volume in liters (3200, 3000, 2400, 2200 and 1800 L) or the manner
in which they can be loaded (lateral, bilateral, rear or underground) (Figure 11) [27,42]. In order to
study the stability of these containers, stability curves depending on the type of container, their filling
degree and the overland flow parameters (flow depth and velocity) were created [42].
(b)
(a) (c)
Figure11.
Figure 11.(a)
(a) Container
Container distribution
distributionin inBarcelona
Barcelonaclassified according
classified according to to
fraction type
fraction andand
type types of of
types
containersas
containers as they
they are
are loaded
loaded by by the
the bin
binlorry,
lorry,(b)
(b)lateral
lateralload
loadand
and(c)(c)bilateral load.
bilateral Adapted
load. from
Adapted from
from Martínez-Gomariz et
Martínez-Gomariz et al. [42]. al. [42].
3.10. Holistic Model of Urban Resilience

The information and the results provided by the 1D/2D USM were used as inputs to the
HAZUR® holistic tool for the evaluation of the potential cascading effects produced by pluvial floods
on several main urban services (as well as others not contemplated by the integrated flood models
previously described) and to estimate their recovery time for current and future scenarios. This
Finally, on the basis on the location of the containers (Figure 11) and the flow parameters provided
by the 1/D/2D USM model, flood (a)
hazard maps showing the unstable waste (c)
containers were created
for an historical storm event to validate the model and several synthetic project storms of 1, 10 and
Figure 11. (a) Container distribution in Barcelona classified according to fraction type and types of
50 years [27,42].
containers as they are loaded by the bin lorry, (b) lateral load and (c) bilateral load. Adapted from
from Martínez-Gomariz et al. [42].
2.10. Holistic Model of Urban Resilience
3.10.information
The Holistic Modelandof Urban provided by the 1D/2D USM were used as inputs to the HAZUR®
Resilience
the results
holistic tool
Theforinformation
the evaluation and of
thethe potential
results cascading
provided by theeffects
1D/2Dproduced
USM were byused
pluvial floods on
as inputs several
to the
main HAZUR
urban services
® holistic(as
toolwell as evaluation
for the others notofcontemplated by the integrated
the potential cascading floodby
effects produced models
pluvialpreviously
floods
described) and to
on several estimate
main urban their recovery
services (as welltime for current
as others and future by
not contemplated scenarios. This analysis
the integrated involved
flood models
previously
34 urban services described)
groupedand intotonine
estimate
sectors their
withrecovery time for
563 critical current and future
infrastructures. scenarios.
The main urbanThis
sectors
analysis involved 34 urban services grouped into nine sectors with 563 critical
and services analyzed included the water cycle, energy, telecommunications, transport, emergencies, infrastructures. The
publicmain urban
health, sectors andand
environment services
greenanalyzed includedwaste
infrastructures, the water cycle, energy,
and citizens. telecommunications,
The HAZUR ® tool is capable
transport, emergencies, public health, environment and green infrastructures, waste and citizens. The
of analyzing cascading effects generated from certain impacts (Figure 12). In the case of pluvial floods,
HAZUR® tool is capable of analyzing cascading effects generated from certain impacts (Figure 12).
impacts and cascade effects on electric and transport sectors were assessed for several synthetic project
In the case of pluvial floods, impacts and cascade effects on electric and transport sectors were
stormsassesseddifferent
with return
for several periods
synthetic (T1,storms
project T10, T150, T100 and
with different T500)
return for current
periods (baseline)
(T1, T10, and
T150, T100 andfuture
(BAU)T500)
scenarios.
for current (baseline) and future (BAU) scenarios.
FigureFigure 12. Cascading

12. Cascading effects
effects simulated
simulated byby HAZUR® for
HAZUR for aapluvial
®
pluvialflood with
flood a 10a year
with return
10 year period
return period
in Barcelona. in Barcelona.
Taking into account the down-times included in the HAZUR® tool (obtained from the sectorial
models or by expert assessment) and considering the interdependencies which exist between several
services and infrastructures, the cascading effects can be simulated [43]. Figure 12 presents an example
of cascading effects generated by a 10 year return period flood event in Barcelona. In this case, as can be
seen in Figure 12, the tool allows us to see that the lack of capacity of the drainage system in the upper
part of the city generates a flood on the high-speed ring “Ronda de Dalt”, stopping the circulation of
vehicles. This affects many other services (such as medical emergency services, the local police or the
citizens), and it also causes the failure of the other high-speed ring “Ronda Litoral”, which in turn
would affect the same services as before (in another area of the city) as well as the port of the city due
to the connection of this highway with that infrastructure.
3. Results
The 1D/2D USM and the derived loosely coupled (integrated) models described in the previous
section were developed and validated using field data provided by sensors and historic collected
information to estimate the potential effects of climate change on the urban drainage sectors and the
cascading effects on other main urban services [27]. The use of this modeling approach allowed us
to achieve valuable results in terms of flood hazards, as well as in terms of socio-economic risk and
impacts on other sectors of the city. In this section, the specific results directly related to the urban
drainage sector and the other analyzed urban services (surface transport, electric system and waste
collection) are presented.
3.1. Assessment of Social Impacts Produced by Pluvial Floods
3.1.1. Flood Risk for Pedestrians

The hydraulic behavior of the urban drainage system of the city (considering both the hydraulic
response of the sewer network and the overland flow on the urban surfaces) was simulated using
the 1D/2D USM for a large set of synthetic projects storms with return periods T of 1, 10, 50, 100 and
500 years for current and future scenarios. As our first results, following the specific flood hazard
criteria for pedestrians presented in Section 2.5, detailed flood hazard maps were created.
The results also allowed the estimation of flood hazards for pedestrians for each district of
the city (with a total area of approximately 102 km2 ) and their evolution in the case of a climate
change scenario. The results concerning the current scenario show that areas classified as having
high flood hazard conditions have reduced risk for low return periods (null for T = 1 and less than
5% for T = 10, with this last one being the designed return period for the sewer system of the city),
which progressively increases for higher return periods. This notwithstanding, the results show that
climate change scenarios could produce an increase in high flood hazard areas of between 20% and
50%. Additionally, for the simulation corresponding to return period T1 and BAU scenario, the high
flood hazard area is null for each district [28].
To assess the flood risk for pedestrians, flood hazard was combined with the human vulnerability,
which was achieved according to the indicators mentioned in Section 2.5 (details can be found in [28]).
Vulnerability was qualitatively assessed for each census district in low, medium and high levels.
These classification ranges resulted in the vulnerability map presented in Figure 13, which was also
considered for 2020,
Sustainability BAU12,due toPEER
x FOR the REVIEW
mainly consolidated urbanistic characteristics of the city [28].
14 of 26
Figure
Figure 13.13.Vulnerability
Vulnerability maps
maps for
forpedestrians.
pedestrians.
As stated
As stated above,
above, thethe flood
flood riskresults
risk results were
were presented
presented ininterms of of
terms flood riskrisk
flood maps for all
maps forthe
all the
considered return periods and scenarios (baseline and BAU). Figure 14 shows the flood risk maps
related to a rainfall storm event with a return period of 10 years for both scenarios.
related to a rainfall storm event with a return period of 10 years for both scenarios.
Figure 13. Vulnerability maps for pedestrians.
As stated above, the flood risk results were presented in terms of flood risk maps for all the
Sustainability to a12,
related2020, 5638 storm event with a return period of 10 years for both scenarios.
rainfall 14 of 25
(a) (b)
Figure
Figure 14. Example
14. Example of flood
of flood risk
risk mapsfor
maps forpedestrians
pedestrians for
foraasynthetic
synthetic1010
year return
year period
return projected
period projected
storms
storms related
related to baseline
to (a) (a) baseline
andand
(b)(b) businessas
business asusual
usual (BAU)
(BAU)scenarios.
scenarios.
Furthermore,
Furthermore, the the high-risk
high-risk area
area (in(in percentage)for
percentage) forpedestrians
pedestrians was
was broken
brokendowndowninto
intodistricts
districts in
in order to observe the riskiest districts in terms of pedestrians’ stability. Moreover, in order to
order to observe the riskiest districts in terms of pedestrians’ stability. Moreover, in order to highlight
highlight the effect of climate change in terms of the increase of high-risk areas in Barcelona, we also
the effect of climate change in terms of the increase of high-risk areas in Barcelona, we also present the
present the variation of high flood risk areas for pedestrians according to the 10 districts into which
variation of high
Barcelona flood risk areasdivided
is administratively for pedestrians
(Figure 15).according to to
It is possible theobserve
10 districts intoincreases
the major which Barcelona
of high is
administratively
flood risk areas divided
(around(Figure
30% for15).
the It is possible
whole district to observe the
of Barcelona) major
with increases
respect of highchange
to the climate flood risk
areascoefficients
(around 30% for the whole district of Barcelona) with
(from 12% to 16%) for the same return periods [28]. respect to the climate change coefficients
Sustainability
(from 12% to 2020,
16%)12, xfor
FOR PEER
the REVIEW
same return periods [28]. 15 of 26
Climate Change
60% 20%
high risk area
Increase of
50%
coefficients
40%
30% 15%
20%
10%
0% 10%
T1 T10 T50 T100 T500
Floods related to return periods
Ciutat Vella Eixample
Sants-Montjuic Les Corts
Sarrià-St. Gervasi Gràcia
Horta-Guinardó Nou Barris
St. Andreu St. Martí
Barcelona Climate Change Factors
Figure 15. Expected increase of high-risk areas according to the future conditions.
Figure 15. Expected increase of high-risk areas according to the future conditions.
3.1.2. Flood Risk for Vehicles
The flow variables (flow depths and flow velocity) provided by the 1D/2D USM were also used to
The flood
generate flow variables (flowfor
hazard maps depths and for
vehicles flow velocity)
each provided
district by and
of the city the 1D/2D USM were
their evolution inalso
the used
case
to agenerate
of climate flood
change hazard mapsThe
scenario. forresults
vehicles for each district
concerning of thescenario
the current city and show
their evolution
that areasin the case
classified
of a high
with climate
floodchange
hazardscenario. Thehave
conditions results concerning
reduced the current
risk compared scenario
to the case ofshow that areas
pedestrians; classified
in particular,
with hazard
high high flood
is null for T =conditions
hazard have
1 and is less reduced
than T = compared
5% forrisk 10, with this to last
the one
casebeing
of pedestrians;
the designed in
particular,
return periodhigh
for hazard is null
the sewer for of
system T the
= 1 city,
and which
is less progressively
than 5% for Tincreases
= 10, with
forthis
higherlastreturn
one being the
periods.
designed
This return period
notwithstanding, thefor the show
results sewerthatsystem
climateof change
the city,scenarios
which progressively
could produceincreases
an average forincrease
higher
return
of 30% periods. This notwithstanding,
for the whole city with a peak of the50%
results show that
for specific climateAdditionally,
districts. change scenarios could
for the produce
simulation
an average increase of 30% for the whole city with a peak of 50% for specific districts. Additionally,
for the simulation corresponding to return period T1 and the BAU scenario, the high flood hazard
area is null for each district [28].
In order to assess the vehicles’ vulnerability, three levels were also proposed based on a unique
indicator: the vehicular flow intensity (VFI) expressed in veh/day. Depending on this value and
defined thresholds, the vulnerability of each urban road was classified into three levels (low, medium
The flow variables (flow depths and flow velocity) provided by the 1D/2D USM were also used
to generate flood hazard maps for vehicles for each district of the city and their evolution in the case
Sustainability
of a climate 12, 5638 scenario. The results concerning the current scenario show that areas classified
2020,change 15 of 25
with high flood hazard conditions have reduced risk compared to the case of pedestrians; in
particular, high hazard is null for T = 1 and is less than 5% for T = 10, with this last one being the
corresponding to return period T1 and the BAU scenario, the high flood hazard area is null for each
designed return period for the sewer system of the city, which progressively increases for higher
district [28].
return periods. This notwithstanding, the results show that climate change scenarios could produce
In order to assess the vehicles’ vulnerability, three levels were also proposed based on a unique
an average increase of 30% for the whole city with a peak of 50% for specific districts. Additionally,
indicator: The vehicular flow intensity (VFI) expressed in veh/day. Depending on this value and
for the simulation corresponding to return period T1 and the BAU scenario, the high flood hazard
defined
area isthresholds,
null for each thedistrict
vulnerability
[28]. of each urban road was classified into three levels (low, medium
and high) [28]. The
In order finalthe
to assess vulnerability map is shown
vehicles’ vulnerability, in Figure
three 16. also proposed based on a unique
levels were
Furthermore, for vehicles, flood risk was assessed through
indicator: the vehicular flow intensity (VFI) expressed in veh/day. the elaboration
Dependingofonflood
thisrisk maps
value andfor
alldefined
the considered
thresholds, return periods andofscenarios
the vulnerability (baseline
each urban road was and BAU).into
classified Figure 17levels
three shows themedium
(low, flood risk
maps
andrelated
high) [28].to aTherainfall
finalstorm event with
vulnerability mapaisreturn
shownperiod of 10
in Figure 16.years for both scenarios.
Figure 16. Vulnerability map for vehicles. Green, orange and red colors indicate low vulnerability
(vehicular flow intensity (VFI) < 100), medium (100 < VFI < 1000) and high (VFI > 1000), respectively.
Furthermore, for vehicles, flood risk was assessed through the elaboration of flood risk maps for
all the considered return periods and scenarios (baseline and BAU). Figure 17 shows the flood risk
maps related
Figure to a rainfall storm
16. Vulnerability event
map for with aGreen,
vehicles. returnorange
periodand
of 10 years
red forindicate
colors both scenarios.
low vulnerability
(vehicular flow intensity (VFI) < 100), medium (100 < VFI < 1000) and high (VFI > 1000), respectively.
(a) (b)
Figure
Figure 17.17. Exampleofofflood
Example floodrisk
riskmaps
mapsfor
for vehicles
vehicles for synthetic
synthetic10
10year
yearreturn
returnperiod
periodprojected
projectedstorms
storms
related
related to to
(a)(a) baseline
baseline andand (b)BAU
(b) BAUscenarios.
scenarios.
In this case, the assessment has also been broken down into districts in order to observe the
riskiest districts in terms of vehicles’ stability. Moreover, in order to highlight the effect of climate
change in terms of the increase of high-risk areas in Barcelona, we present the variation of high-flood
risk areas for vehicles in all of the districts (Figure 18). In this case, it is also possible to observe a
(a) (b)
Figure 17. Example of flood risk maps for vehicles for synthetic 10 year return period projected storms
Sustainability
related2020, 12,baseline
to (a) 5638 and (b) BAU scenarios. 16 of 25
In this case, the assessment has also been broken down into districts in order to observe the
In this case, the assessment has also been broken down into districts in order to observe the riskiest
riskiest districts in terms of vehicles’ stability. Moreover, in order to highlight the effect of climate
districts in terms of vehicles’ stability. Moreover, in order to highlight the effect of climate change in
change in terms of the increase of high-risk areas in Barcelona, we present the variation of high-flood
terms of the increase of high-risk areas in Barcelona, we present the variation of high-flood risk areas
risk areas for vehicles in all of the districts (Figure 18). In this case, it is also possible to observe a
for vehicles in all of the districts (Figure 18). In this case, it is also possible to observe a major increase
major increase of high flood risk areas (from 20% to 40% for the whole city area) with respect to the
of high flood risk areas (from 20% to 40% for the whole city area) with respect to the climate change
climate change coefficients (from 12% to 16%) for the same return periods.
coefficients (from 12% to 16%) for the same return periods.
Increase of high
100% 20%
Climate change
coefficient
risk area
80% 18%
60% 16%
40% 14%
20% 12%
0% 10%
T1 T10 T50 T100 T500
Floods related to return periods
Ciutat Vella Eixample
Sants-Montjuic Les Corts
Sarrià-St. Gervasi Gràcia
Horta-Guinardó Nou Barris
St. Andreu St. Martí
Barcelona Coef. Canvi climàtic (%)
Figure 18. Expected

Expected increase
increase of high-risk areas according to the future conditions.
3.2. Assessment of Economic Impacts Produced by Pluvial Floods

For the estimation of tangible direct damages caused by pluvial floods generated by urban floods,
both properties and vehicles were considered in the economic assessment. According to the claims data
provided by the Spanish re-insurance company (CSS), these two risk categories are the most significant.
According to the developed methodology to estimate property damage, flow depths on the streets
provided by the 1D/2D USM were properly reduced to achieve flood depths for properties using
specific sealing coefficients, which were collected for 14 land uses in Barcelona [35,36]. As a second
step, flood damages suffered by the properties were evaluated on the basis of tailored flood depth
damage curves for all the 14 land uses; a detailed flood damage model was developed and validated in
previous studies [28,30]. Models considered different typologies of properties: without basements,
with a basement and with up to two basements. On the other hand, configurations with or without
parking access were considered [28,30].
Regarding the evaluation of vehicle damage, a novel methodology—also based on the concept
of damage curves—was implemented. The methodology tried to reduce the uncertainty due to the
mobility of vehicles, proposing heterogeneous vehicular occupation for several areas of the city based
on the information provided by aerial photographs [28,37]. For this assessment, flood damage curves
developed by the Army Corps of the United States of America [44] for five types of vehicles were
adapted for the case study of Barcelona [28,37]. These curves were converted into a single damage
curve weighted according to the percentage of vehicle types in Barcelona, also taking into account
their depreciation according to statistical information concerning vehicle types and their age [28,30,37].
For both properties’ and vehicles’ flood damage assessment, damage maps were achieved for
the return periods T1, T10, T50, 100 and T500 and current (baseline) and future (BAU) scenarios and
aggregated for each district of the city (Figures 19 and 20). These figures show how future rainfall
conditions for a projected storm of 10 years significantly worsen the situation in several districts of the
city. Specifically, it can be observed that all the districts of the downtown would suffer high losses,
and the better situation of several districts upstream would be exacerbated due to climate change.
Moreover, for both scenarios, the expected annual damage (EAD) [29] for the whole city including
flood damages related to properties and vehicles [30] was calculated. The results indicate that, due to
climate change, the EAD would grow from € 39.8 M to € 54.7 M [28].
Finally, the methodology for the estimation of indirect damages produced by pluvial floods based
on an econometric method of input–output (IO) tables indicated a linear relationship between direct
and tangible
Sustainability losses.
2020, Specifically,
12, x FOR PEER REVIEWaccording to the obtained results, indirect tangible damages produced
18 of 26
by pluvial floods in Barcelona could represent around 29% of direct damages. This increase could be
taken into account in the previously reported EAD [28].
(a) (b)
Figure 19. Example of (a)economic flood damage maps for properties for synthetic (b) projected storms of
10 years19.
Figure
Figure related
19. to baseline
Example
Example of (a) and
ofeconomic
economic BAU
flood
flood scenarios
damage
damage (b) for
maps
maps indicating aggregated
forproperties
properties damages
forsynthetic
for synthetic for districts.
projected
projected stormsof
storms of
10 years
10 years related
related to
to baseline
baseline (a)
(a) and
and BAU
BAU scenarios
scenarios (b)
(b) indicating
indicating aggregated damages for districts.
(a) (b)
Figure 20.
Figure Exampleof(a)
20.Example ofeconomic
economicflood
flood damage
damage maps
maps forfor vehicles
vehicles forfor (b)projected
synthetic
synthetic projected storms
storms of
of 10
10 years related
years related
Figure to baseline
to baseline
20. Example (a) and
(a) and BAU
of economic BAU scenarios
floodscenarios (b) indicating
(b) indicating
damage maps aggregated
aggregated
for vehicles damages
damages
for synthetic for districts.
for districts.
projected storms of 10
years related to baseline (a) and BAU scenarios (b) indicating aggregated damages for districts.
3.3. Assessment of the Effects of Pluvial Floods on the Surface Traffic Service
3.3. Assessment of the Effects
The climate-related of Pluvial
resilience of aFloods on the Surface
city depends on its Traffic
capacityService
to maintain the correct functioning
of theThe
main urban services
climate-related duringof
resilience extreme weatheron
a city depends events such astopluvial
its capacity floods.
maintain The results
the correct of the
functioning
3.3. T500
and Assessment of the Effects
and current of Pluvial
(baseline) andFloods
futureon(BAU)
the Surface Traffic Service
scenarios. Examples of flood hazard maps are
shownThe in climate-related
Figure 21. Comparing
resiliencethe
of aresults for both
city depends onscenarios,
its capacityit to
can be observed
maintain that,functioning
the correct for the total
amount of 1492
of the main urbankm,services
the increase
duringofextreme
the roadweather
links that could
events suchbe as
affected
pluvialby speedThe
floods. reduction ranged
results of the
between 3% and 30% depending on the return period, while the increase in terms
impacts produced by this kind of floods on the surface traffic system were analyzed according to the of closed road links
could be around
methodology 20% for in
presented allSection
the considered return
2.7. In this case,periods (Figure
flood hazard was22).
assessed through flood hazard
maps elaborated on the basis of flood depths provided by the 1D/2D USM and the specific hazard
criteria
and T500previously
and current presented. Hazard
(baseline) and maps
futurewere
(BAU)elaborated for Examples
scenarios. the return periods
of floodT1, T10, T50,
hazard mapsT100are
and T500 and current (baseline) and future (BAU) scenarios. Examples of flood hazard maps are shown
shown in Figure 21. Comparing the results for both scenarios, it can be observed that, for the total
in Figure 21. Comparing the results for both scenarios, it can be observed that, for the total amount of
amount of 1492 km, the increase of the road links that could be affected by speed reduction ranged
1492 km, the increase of the road links that could be affected by speed reduction ranged between 3%
between 3% and 30% depending on the return period, while the increase in terms of closed road links
and 30% depending on the return period, while the increase in terms of closed road links could be
could be around 20% for all the considered return periods (Figure 22).
around 20% for all the considered return periods (Figure 22).
(a) (b)
Figure 21. Example of flood hazard maps for surface traffic for synthetic projected storms of 10 years
related to baseline (a) and BAU scenarios (b).
Finally, through the TransCAD mesoscalar traffic model, the increase in transit time for all the
synthetic storm events was assessed and monetized following the methodology proposed by the
Multi-Color Handbook (a) [45]. The monetization of the increase of traveling (b) time for the whole city
allowed the21.
Figure
Figure estimation
21.Example of
Exampleof a specific
offlood
flood hazard
hazard EAD
mapsfor
maps forbaseline
for surface (1.82for
surface traffic
traffic M€)
for and BAU
synthetic
synthetic (2.0 M€)
projected
projected [28,38].
storms
storms of
of10
10years
years
relatedto
related tobaseline
baseline (a)
(a) and
and BAU
BAU scenarios
scenarios (b).
(b).
Finally,
km through the
of roads TransCAD
with reducedmesoscalar
speed traffic model, the increase in transit
km of closed roadstime for all the
500 storm events was assessed and monetized following
synthetic 600 the methodology proposed by the
Multi-Color
400 Handbook [45]. The monetization of the increase of traveling time for the whole city
allowed
300the estimation of a specific EAD for baseline (1.82400
M€) and BAU (2.0 M€) [28,38].
km
km
200
200
100
km of roads with reduced speed km of closed roads
0
500 6000
400 Baseline BAU Baseline BAU
T10 T50 T100 T500 400 T10 T50 T100 T500
300
km
km
200
(a) 200 (b)
100
Figure22.
Figure Representation of
22.Representation of the
the effects
effects produced
produced by
by pluvial
pluvialflood
floodononthe
thesurface
surfacetransport
transportsystem inin
system
0
Barcelona for
for current 0
Barcelona current(baseline)
(baseline)andandfuture (BAU)
future scenarios
(BAU) in terms
scenarios of kmof
in terms ofkm
roads
of with
roadsreduced speed
with reduced
(a) and km Baseline
of closed roads (b). BAU Baseline BAU
speed (a) and km of closed roads (b)
T10 T50 T100 T500 T10 T50 T100 T500
3.4. Assessment of the Effects of Pluvial Floods on the Electric System

(a) (b)
Figure 22. Representation of the effects produced by pluvial flood on the surface transport system in
Barcelona for current (baseline) and future (BAU) scenarios in terms of km of roads with reduced
speed (a) and km of closed roads (b)
Finally, through the TransCAD mesoscalar traffic model, the increase in transit time for all the
synthetic storm events was assessed and monetized following the methodology proposed by the
Multi-Color Handbook [45]. The monetization of the increase of traveling time for the whole city
allowed the
Sustainability 2020,estimation of a REVIEW
12, x FOR PEER specific EAD for baseline (1.82 M€) and BAU (2.0 M€) [28,38]. 20 of 26
3.4. Assessment of the Effects of Pluvial Floods on the Electric System

Through the maximum flow depths provided by the 1D/2D USM and the geolocation of
Through
electrical the maximum
infrastructures, flow depths
an impact analysis provided by theout
was carried 1D/2D USM and
according geolocation of presented
to the methodology electrical
in infrastructures,
Section 2.6. an impact analysis was carried out according to the methodology presented in
Section
Figure 2.8.23 shows an example of the risk assessment carried out for all the electrical assets for
Figure 23
return periods ofshows
T10 and anT100.
example
Here,ofitthe risk assessment
is possible carried
to see the most out for all
affected the and
areas electrical assets
electrical for
assets,
return periods of T10 and T100. Here, it is possible to see the most affected areas and
which are classified within the categories specified in Table 3 and sized with respect to their failure electrical assets,
which
risk. It isare classified
possible to within
observethethat
categories
the area specified in Table
near the Besòs3 and sized with
riverside respect
is clearly theto most
their failure risk.
affected by
It is possible to observe that the area near the Besòs riverside is clearly the most
pluvial flooding, showing the densest cloud of affected locations. Additionally, the figure shows the affected by pluvial
flooding,
increase showing
of the failurethe densest cloud
probability fromofthe affected locations.
blue-colored Additionally,
baseline scenario the(BAS)
figuretoshows the increase
the BAU scenario,
of the in
colored failure
green.probability from the blue-colored baseline scenario (BAS) to the BAU scenario, colored
in green.
(a) (b)
Figure Figure
23. Example of risk of
23. Example maps
risk of all the
maps electrical
of all assetsassets
the electrical studied for T10
studied for (a)
T10and T 100
(a) and (b). (b).
T 100
Table4 4shows
Table showsthethenumber
numberof ofelectrical
electrical infrastructures
infrastructures potentially
potentially affected
affectedby bypluvial
pluvialflood
floodinin
Barcelonafor
Barcelona forbaseline
baselineand
and BAU
BAU scenarios
scenariosandandtheir
theirpotential
potentiallevel of impact.
level TheThe
of impact. tabletable
also also
shows the
shows
social impact that each type of flooding provokes in society by counting the number
the social impact that each type of flooding provokes in society by counting the number of people of people affected
in eachincase
affected each(reaching, in the worst
case (reaching, case,
in the 725,119
worst case,out 1,620,343
725,119 total people
out 1,620,343 in Barcelona)
total and the losses
people in Barcelona) and
provoked for each case, which in the worst scenario amounts to 771,129.01 €. It should be noted that
the losses provoked for each case, which in the worst scenario amounts to 771,129.01 €. It should be
the high (HV) and medium-voltage (MV) substations with a potential flood risk have been studied
noted that the high (HV) and medium-voltage (MV) substations with a potential flood risk have been
throughout the city, while only the distribution centers (DCs) in the vicinity of Besós and Llobregat
studied throughout the city, while only the distribution centers (DCs) in the vicinity of Besós and
rivers and coastal areas were considered [28,40].
Llobregat rivers and coastal areas were considered [28,40].
Table 4. Electrical infrastructure potentially affected by pluvial flood in Barcelona for baseline (BAS)
and BAU scenarios. DC: distribution center; HV: high-voltage substation; MV: medium-voltage
substation.
Return Scenar Type of Number of Locations Customers Costs

Period io Location Affected Affected Provoked
DC 165 14,984 90,403.68 €
Table 4. Electrical infrastructure potentially affected by pluvial flood in Barcelona for baseline (BAS) and
BAU scenarios. DC: distribution center; HV: high-voltage substation; MV: medium-voltage substation.
Type of Number of Customers Costs

Return Period Scenario
Location Locations Affected Affected Provoked
DC 165 14,984 90,403.68 €
BAS HV 6 116,872 2377.23 €
MV 11 94,231 5585.61 €
T10
DC 187 290,613 192,823.10 €
BAU HV 6 116,872 3709.35 €
MV 13 150,723 2231.57 €
DC 227 295,490 304,720.21 €
BAS HV 6 116,872 11,267.27 €
MV 13 372,311 6627.44 €
T50
DC 254 314,932 476,756.76 €
BAU HV 7 116,872 20,367.44 €
MV 15 372,311 18,549.21 €
DC 249 314,044 451,294.19 €
BAS HV 7 116,872 19,438.12 €
MV 13 372,311 12,771.98 €
T100
DC 272 315,991 556,183.29 €
BAU HV 8 116,872 28,873.49 €
MV 15 581,566 41,375.86 €
DC 296 318,232 633,795.69 €
BAS HV 9 215,368 56,870.91 €
MV 17 582,487 28,035.45 €
T500
DC 324 320,679 771,129.01 €
BAU HV 11 215,368 66,869.66 €
MV 18 725,119 53,948.15 €
3.5. Assessment of the Effects of Pluvial Floods on Waste Collection System

The integrated flood–waste collection model allowed the estimation of the potential number of
unstable containers and their location on specific hazard maps based on their typology and degree of
filling for the return periods T1, T10 and T50 [27,28,42]. This analysis showed that, for the most extreme
episode (T50), some districts of the city such as Ciutat Vella and l’Eixample could have between 20%
and 25% of their containers dragged due to the flow and that, in some cases, this amount could increase
to values above 30% for the BAU scenario. Figure 24 shows, as an example, the computed number of
containers which are potentially unstable for each district under the assumptions of current and future
rainfall conditions for a designed 10 year return period storm.
extreme episode (T50), some districts of the city such as Ciutat Vella and l’Eixample could have
between 20% and 25% of their containers dragged due to the flow and that, in some cases, this amount
could increase to values above 30% for the BAU scenario. Figure 24 shows, as an example, the
computed number of containers which are potentially unstable for each district under the
Sustainability
assumptions of12,current
2020, 5638 21 of 25
and future rainfall conditions for a designed 10 year return period storm.
St. Martí
St. Andreu
Nou Barris
Horta-Guinardó
Gràcia
Sarrià-St. Gervasi
Les Corts
Sants-Montjuic
Eixample
Ciutat Vella
0 100 200 300 400 500 600
T10 Business As Usual (BAU) T10 Baseline
Figure 24. Distribution of computed number of containers which are potentially unstable for each
Figure under
district 24. Distribution of computed
current (baseline) number
and future of rainfall
(BAU) containers which are
conditions duepotentially unstable
to a flooding for each
corresponding
district
to under10
a designed current (baseline) and future (BAU) rainfall conditions due to a flooding corresponding
year storm.
to a designed 10 year storm.
3.6. Assessment of Flood Resilience through a Holistic Approach
The holistic model was used to determine the recovery time of the city in the case of extreme
episodes of pluvial flooding produced by extreme rain events. The analysis of the holistic simulations
allowed the estimation of a recovery time of approximately 1.5 h (calculated as an average value for
all the events with return periods T1, T10, T50, 100 and T500), while for the BAU scenario, this value
increased up to 2 h.
4. Discussion
The potential increase of maximum rainfall intensities in Barcelona due to climate change could
produce a significant increase of tangible and intangibles impacts due to pluvial floods. This paper
aimed to perform a comprehensive multi-risk assessment using a detailed 1D/2D USM and several
loosely coupled models in order to estimate direct impacts not only due to the poor efficiency of the
drainage systems of the city but also due to several cascading effects on other critical urban services.
This kind of analysis represents a key tool for decision makers to achieve a reliable estimation of the cost
of not acting and to propose and justify correct adaptation measures which are able to reduce a large
set of tangible and intangible impacts. For the case of Barcelona, the development and calibration of a
1D/2D USM and its integration in several loosely coupled (or integrated) models allowed us to perform
a multi-risk analysis whose main important outputs are shown in Table 5. Moreover, the geographic
detailed analysis of the potential flood impacts could help in the prioritization of the implementation
of adaptation measures [46]. For example, the results provided by some impact models concerning
intangible (safety for pedestrians and vehicles, stability of containers) and tangible (economic losses for
properties and vehicles) damage indicate that the highest economic and social risks are concentrated in
the districts located in the downtown of the city (near the sea).
Table 5. Potential pluvial flood impacts due to climate change assessed by loosely coupled models.
Indicator (BAU
Model Type of Impact Values for T/EAD
vs. Baseline)
Pedestrians: +30 (T10), +34
Increase (%) of high
(T50), +32 (T100), +30 (T500)
1D/2D USM Intangible flood risk area for
Vehicles: +38 (T10), +42 (T50),
pedestrian and vehicles
+34 (T100), +25 (T500)
Increase (%) of EAD
1D/2D USM + (including properties,
Tangible +42%
Damage model vehicles and indirect
damages)
Increase (%) of km of
1D/2D USM + +31 (T10), +60 (T50), +66
Tangible and Intangible closed roads; EAD due to
Traffic model (T100), +116 (T500); +0.18 M€
travelling time rise
Increase (%) of the
1D/2D USM + number of flooded +13 (T10), +12 (T50), +11
Tangible and Intangible
Electric model electric infrastructures; (T100), +10 (T500); +0.12M€
related EAD
Increase (%) of the Empty: +27 (T10), +28 (T50)
1D/2D USM +
Intangible number of unstable 50% full: +28 (T10), +32 (T50)
Waste model
waste containers 100% full: +28 (T10), +36 (T50)
5. Conclusions
This paper demonstrates how the integration of a detailed and calibrated 1D/2D USM with other
models and tools which are able to describe the behavior of other urban services can be useful to
simulate the response of these services during pluvial floods produced by heavy storm events.
Furthermore, through the development of these loosely coupled models, socio-economic impacts
related to these events can be estimated and the cascading effects can be fully analyzed, as well as the
interrelationships between services and critical infrastructures.
In this study, the effects of floods in the potential context of climate change for the city of Barcelona
have been analyzed through a multi-risk approach, and the results of this assessment, in terms
of tangible and intangible impacts, have been presented for the whole city and with a geographic
discretization (i.e., in terms of city districts).
The results demonstrate that Barcelona could suffer a significant increase in these impacts due
to climate change if adaptation measures are not adopted. It was demonstrated that increments of
maximum rainfall intensity of 12–16% could cause increments of more than 25–30% in terms of social
impacts (e.g., intangible damages such as the increase of areas classified with high hazard conditions
in case of pluvial flood events) and of 42% of economic losses (including tangible direct and indirect
damages) expressed in monetary terms through the concept of EAD that has been calculated for each
analyzed urban district. Economic losses related to traffic disruption due to pluvial floods could
also increase by 9%, while for the electric system, the increase of economic damage could be 70%,
although the final EAD result was shown to be quite low.
Moreover, the average recovery time of the city (defined as the time in which urban services do
not recover their normal functioning) could increase from 1.5 to 2 h due to climate change effects.
Finally, the paper shows the geographical distribution of the socio-economic impacts.
This information could be very useful for the prioritization of implementation of adaptation measures.
Author Contributions: Conceptualization, B.R.; methodology, B.R., M.V., L.L., E.M.-G., R.M., B.E. and D.S. and
D.S.-M; validation, B.R. and M.V.; formal analysis, B.R. and E.M.-G.; investigation, B.R., M.V., L.L., R.M., D.S.-M.,
E.M.-G., and B.E.; resources, D.S., E.M.-G., B.E. and D.S.-M.; data curation, D.Y., D.S., E.M.-G., E.F.-O., B.E., D.S.-M.,
A.G.G., writing—original draft preparation, B.R.; writing—review and editing, B.R., M.V., L.L., D.S., E.M.-G., B.E.
and D.S.-M; visualization, E.M.-G.; supervision, M.V.; project administration, B.R.; funding acquisition, B.R. and
E.M.-G. All authors have read and agreed to the published version of the manuscript.
Funding: This research was funded by Horizon2020 Programme, Grant Agreement No. 700174.
Acknowledgments: This paper presents some of the results achieved in the framework of the RESCCUE
project (Resilience to Cope with Climate Change in Urban Areas—a multisectoral approach focusing on water)
(www.resccue.eu). RESCCUE is a research project funded by the European Commission under the H2020 program,
and its main goal is to provide methodologies and tools for the evaluation, planning and management of urban
resilience in the context of climate change.
Conflicts of Interest: The authors declare no conflict of interest.
References
1. Meerow, S.; Newell, J.P.; Stults, M. Defining urban resilience: A review. Landsc. Urban Plan. 2016, 147, 38–49.
[CrossRef]
2. Velasco, M.; Russo, B.; Martínez, M.; Malgrat, P.; Monjo, R.; Djordjevic, S.; Fontanals, I.; Vela, S.; Cardoso, M.A.;
Buskute, A. Resilience to cope with climate change in urban areas—A multisectorial approach focusing on
water—The RESCCUE project. Water 2018, 10, 1356. [CrossRef]
3. Walloth, C.; Gurr, J.M.; Schmidt, J.A. Understanding Complex Urban Systems: Multidisciplinary Approaches to
Modeling; Springer International Publishing: Cham, Switzerland, 2014; ISBN 978-3-319-02996-2.
4. Visvizi, A.; Lytras, M. Smart Cities: Issues and Challenges. Mapping Political, Social and Economic Risks and Threats,
1st ed.; Lytras, A., Visvizi, M., Eds.; Elsevier: Amsterdam, The Netherlands, 2019; ISBN 9780128166390.
5. Calzada, I.; Almirall, E. Data ecosystems for protecting European citizens’ digital rights. Transform. Gov.
People Process Policy 2020, 14, 133–147. [CrossRef]
6. Monjo, R.; Paradinas, C.; Gaitán, E.; Redolat, D.; Prado, C.; Pórtoles, J.; Torres, L.; Russo, B.; Velasco, M.;
Pouget, L.; et al. Report on Extreme Events Prediction. Deliverable 1.3, RESCCUE EU H2020 Project.
Available online: http://www.resccue.eu/ (accessed on 24 April 2020).
7. Hammond, M.J.; Chen, A.S.; Djordjević, S.; Butler, D.; Mark, O. Urban flood impact assessment:
A state-of-the-art review. Urban Water J. 2015, 12, 14–29. [CrossRef]
8. Barcelona City Council. Climate Plan 2018–2030; Urban Ecology: Barcelona, Spain, 2018.
9. Monjo, R.; Gaitán, E.; Pórtoles, J.; Ribalaygua, J.; Torres, T. Changes in extreme precipitation over Spain using
statistical downscaling of CMIP5 projections. Int. J. Climatol. 2016, 36, 757–769. [CrossRef]
10. Arnbjerg-Nielsen, K. Quantification of climate change effects on extreme precipitation used for high resolution
hydrologic design. Urban Water J. 2012, 9, 57–65. [CrossRef]
11. Russo, B. Multi-Hazards Assessment Related to Water Cycle Extreme Events for Future Scenarios-Business As Usual.
Deliverable 2.3; RESCCUE EU H2020 Project. Internal report.
12. Hénonin, J.; Russo, B.; Mark, O.; Goubersville, P. Real-time urban flood forecasting and modelling—A state
of the art. J. Hydroinform. 2013, 15, 717–736. [CrossRef]
13. Phillips, B.C.; Yu, S.; Thompson, G.R.; Silva, N. De 1D and 2D Modelling of Urban Drainage Systems
using XP-SWMM and TUFLOW. In Proceedings of the 10th International Conference on Urban Drainage,
Copenhagen, Denmark, 21–26 August 2005; pp. 21–26.
14. Lipeme Kouyi, G.; Fraisse, D.; Rivière, N.; Guinot, V.; Chocat, B. 1D modelling of the interactions between
heavy rainfall-runoff in urban area and flooding flows from sewer network and river. In Proceedings of the
11th International Conference on Urban Drainage, Edinburgh, UK, 31 August–5 September 2008.
15. Obermayer, A.; Guenthert, F.W.; Angermair, G.; Tandler, R.; Braunschmidt, S.; Milojevic, N. Different
approaches for modelling of sewer caused urban flooding. Water Sci. Technol. 2010, 62, 2175–2182. [CrossRef]
16. Leandro, J.; Chen, A.S.; Djordjević, S.; Savić, D.A. Comparison of 1D/1D and 1D/2D coupled (sewer/surface)
hydraulic models for urban flood simulation. J. Hydraul. Eng. 2009, 135, 495–504. [CrossRef]
17. Kandori, C.; Willems, P. Impact of the two-directional interaction of sewer and river systems on the flood
risk. In Proceedings of the 11th International Conference of Urban Drainage, Edinburgh, UK, 31 August–5
September 2008; pp. 1–10.
18. Mark, O.; Weesakul, S.; Apirumanekul, C.; Aroonnet, S.B.; Djordjevic, S. Potential and limitations of 1D
modelling of urban flooding. J. Hydrol. 2004, 299, 284–299. [CrossRef]
19. Pina, R.D.; Ochoa-Rodriguez, S.; Simões, N.E.; Mijic, A.; Marques, A.S.; Maksimović, Č. Semi- vs.
Fully-distributed urban stormwater models: Model set up and comparison with two real case studies. Water
2016, 8, 58. [CrossRef]
20. Russo, B.; Sunyer, D.; Velasco, M.; Djordjevic, S. Analysis of extreme flooding events through a calibrated
1D/2D coupled model: The case of Barcelona (Spain). J. Hydroinform. 2015, 17, 473–491. [CrossRef]
21. Gómez, M.; Russo, B. Methodology to estimate hydraulic efficiency of drain inlets. Proc. Inst. Civ. Eng.
Water Manag. 2011, 164, 81–90. [CrossRef]
22. Gómez, M.; Russo, B. Hydraulic efficiency of continuous transverse grates for paved areas. J. Irrig. Drain. Eng.
2009, 135. [CrossRef]
23. Gómez, M.; Parés, J.; Russo, B.; Martínez-Gomariz, E. Methodology to quantify clogging coefficients for
grated inlets. Application to SANT MARTI catchment (Barcelona). J. Flood Risk Manag. 2019, 12. [CrossRef]
24. Gómez, M.; Rabasseda, G.H.; Russo, B. Experimental campaign to determine grated inlet clogging factors in
an urban catchment of Barcelona. Urban Water J. 2013, 10. [CrossRef]
25. Innovyze. InfoWorks Integrated Catchment Modelling (ICM), version 10.0; Innovyze: Monrovia, CA, USA, 2020.
26. Russo, B.; Pouget, L.; Malgrat, P. Evaluacion del impacto del cambio climatico en un caso de estudio de
Barcelona a través de una modelización 2D-1D del drenaje dual. In Proceedings of the Jornadas de Ingeniería
del Agua (JIA), Barcelona, Spain, 5–6 October 2011.
27. Russo, B. Multi-Hazards Assessment Related to Water Cycle Extreme Events for Current Scenario. (Public Summary)
Deliverable 2.4. RESCCUE EU H2020 Project. Available online: http://www.resccue.eu/sites/default/files/d2.4.
_multi-hazards_assessment_related_to_water_cycle_extreme.pdf (accessed on 24 April 2020).
28. Evans, B. Impact Assessments of Multiple Hazards in Case Study Areas. Deliverable 3.4; RESCCUE EU H2020
Project. Internal report.
29. Velasco, M.; Russo, B.; Cabello, À.; Termes, M.; Sunyer, D.; Malgrat, P. Assessment of the effectiveness of
structural and nonstructural measures to cope with global change impacts in Barcelona. J. Flood Risk Manag.
2018, 11. [CrossRef]
30. Martínez-Gomariz, E.; Locatelli, L.; Guerrero, M.; Russo, B.; Martínez, M. Socio-Economic Potential Impacts
Due to Urban Pluvial Floods in Badalona (Spain) in a Context of Climate Change. Water 2019, 11, 2658.
[CrossRef]
31. Turner, B.L.; Kaspersonb, R.C.; Matsone, P.A.; McCarthy, J.; Corell, R.; Christensene, L.; Eckley, N.;
Kasperson, J.X.; Luers, A.; Martello, M.L.; et al. A framework for vulnerability analysis in sustainability
science. Proc. Natl. Acad. Sci. USA 2003, 100, 8074–8079. [CrossRef]
32. Russo, B.; Gómez, M.; Macchione, F. Pedestrian hazard criteria for flooded urban areas. Nat. Hazards 2013,
69, 251–265. [CrossRef]
33. Martínez-Gomariz, E.; Gómez, M.; Russo, B. Experimental study of the stability of pedestrians exposed to
urban pluvial flooding. Nat. Hazards 2016, 82, 1259–1278. [CrossRef]
34. Martínez-Gomariz, E.; Gómez, M.; Russo, B.; Djordjević, S. A new experiments-based methodology to define
the stability threshold for any vehicle exposed to flooding. Urban Water J. 2017, 14, 930–939. [CrossRef]
35. Martínez-Gomariz, E.; Guerrero-Hidalga, M.; Russo, B.; Yubero, D.; Gómez, M.; Castán, S. Desarrollo y
aplicación de curvas de daño y estanqueidad para la estimación del impacto económico de las inundaciones
en zonas urbanas españolas. Ingenieria del Agua 2019, 23, 229. [CrossRef]
36. Martínez-Gomariz, E.; Forero-Ortiz, E.; Guerrero-Hidalga, M.; Castán, S.; Gómez, M. Flood Depth—Damage
Curves for Spanish Urban Areas. Sustainability 2020, 12, 2666. [CrossRef]
37. Martínez-Gomariz, E.; Gómez, M.; Russo, B.; Sánchez, P.; Montes, J.A. Methodology for the damage
assessment of vehicles exposed to flooding in urban areas. J. Flood Risk Manag. 2019, 12, 1–15. [CrossRef]
38. Evans, B.; Chen, A.S.; Djordjevi, S.; Webber, J.; Gonzalez, A.; Stevens, J. Investigating the Effects of Pluvial
Flooding and Climate Change on Traffic Flows in Barcelona and Bristol. Sustainability 2020, 12, 2330.
[CrossRef]
39. Pyatkova, K.H. Flood Impacts on Road Transportation; University of Exeter: Exeter, UK, 2019.
40. Sánchez-Muñoz, D.; Domínguez-García, J.L.; Martínez-Gomariz, E.; Russo, B.; Stevens, J.; Pardo, M. Electrical
grid risk assessment against flooding in Barcelona and Bristol cities. Sustainability 2020, 12, 1527. [CrossRef]
41. FEMA. Multi-Hazard Loss Estimation Methodology, Flood Model: Hazus-MH MR4 Technical Manual; FEMA:
Washington, DC, USA, 2009.
42. Martínez-Gomariz, E.; Russo, B.; Gómez, M.; Plumed, A. An approach to the modelling of stability of waste
containers during urban flooding. J. Flood Risk Manag. 2019. [CrossRef]
43. Russo, B.; Velasco, M.; Monjo, R.; Martínez-Gomariz, E.; Sánchez, D.; Domínguez, J.L.; Gabàs, A.; Gonzalez, A.
Evaluación de la resiliencia de los servicios urbanos frente a episodios de inundación en Barcelona. El Proyecto
RESCCUE. Ingenieria del Agua 2020, 24, 101. [CrossRef]
44. U.S. Army Corps of Engineers (USACE). Economic Guidance Memorandum, 09-04, Generic Depth-Damage
Relationships for Vehicles; USACE: Washington, DC, USA, 2009.
45. Penning-Rowsell, E.; Viavattene, C.; Pardoe, J.; Chatterton, J.; Parker, D.; Morris, J. The Benefits of Flood and
Coastal Risk Management: A Handbook ofAssessment Techniques; Flood Hazard Research Centre, Middlesex
University: London, UK, 2010.
46. Guerrero-Hidalga, M.; Martínez-Gomariz, E.; Evans, B.; Webber, J.; Termes-Rifé, M.; Russo, B.; Locatelli, L.
Methodology to Prioritize Climate Adaptation Measures in Urban Areas. Barcelona and Bristol Case Studies.
Sustainability 2020, 12, 4807. [CrossRef]
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

Assessment of Urban Flood Resilience in Barcelona For Current and Future Scenarios. The RESCCUE Project

Uploaded by

Copyright:

Available Formats

Assessment of Urban Flood Resilience in Barcelona For Current and Future Scenarios. The RESCCUE Project

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Assessment of Urban Flood Resilience in Barcelona For Current and Future Scenarios. The RESCCUE Project

Uploaded by

Copyright:

Available Formats

sustainability

Sustainability 2020, 12, 5638; doi:10.3390/su12145638 www.mdpi.com/journal/sustainability

Sustainability 2020, 12, 5638 2 of 25

Strategic urban services

Hazard impacts &

Resilience & adaptation

Holistic resilience Initial holistic resilience

T1 T2 T5 T10 T25 T50 T100 T500

model electric system on electric system

(a) SD (b) FD (c) H

2.3. Barcelona Semi-Distributed 1D/2D USM

2.4. Modeling of the Effects of Pluvial Floods on Several Urban Services

Loosely Coupled Model Involved Sectors Main Purposes

2.5. Social Flood Impacts Model

2.6. Economic Flood

2.7. Integrated Flooding–Surface

Table 2. Flood hazard effects on traffic flow (from [38,39]).

Here, flood hazard analysis

2.8. Integrated Flooding–Electric

Energy Non Supplied

Bussines Cost (BC)

MAIN INPUTS HAZARD ASSESSMENT RISK ASSESSMENT COST ASSESSMENT

Table 3. Failure probabilities for electrical assets.

Probability Range Categorical Description

2.9. Integrated Flooding–Waste Collection System Model

3.10. Holistic Model of Urban Resilience

FigureFigure 12. Cascading

3.1. Assessment of Social Impacts Produced by Pluvial Floods

3.1.1. Flood Risk for Pedestrians

Sustainability 2020, 12, x FOR PEER REVIEW 16 of 26

Figure 18. Expected

3.2. Assessment of Economic Impacts Produced by Pluvial Floods

3.4. Assessment of the Effects of Pluvial Floods on the Electric System

3.4. Assessment of the Effects of Pluvial Floods on the Electric System

Return Scenar Type of Number of Locations Customers Costs

Type of Number of Customers Costs

3.5. Assessment of the Effects of Pluvial Floods on Waste Collection System

0 100 200 300 400 500 600

T10 Business As Usual (BAU) T10 Baseline

You might also like