Modeling urban expansion
VincentViguié*,BasilePfeiffer,QuentinLepetit
.
5.References4.Summary
2.Model(NEDUM-2D)
across apanelofdiversecities
3.Applications
Consequencesoftransport,landplanningandeconomicspoliciesonincomerepartitionandslumdevelopment.(workinprogress)
SimulationoftheshareofthehouseholdsearninglessthanthemedianincomeinCapeTown,SouthAfrica
SimulatedimpactofGrandParisexpressmetroline
constructiononrents
Consequencesofnewtransportinfrastructuresonurbandevelopment[9]
Comparisonofthespatialconfigurationsofcities,andofthelinkwithtransport/landplanningpolicies.(workinprogress)
Comparisonofthespatialvariationofrentsin7Chinesecities
[email protected] - CIRED(CNRS,EHESS,EcoledesPontsParisTech,AgroParisTech,CIRAD)Nogent-sur-Marne,France
Taxation,buildingconstraintsandlandplanningpoliciesconsequencesonrealestatepricesandurbandevelopment[2,4]
Scenariosoncitiesfuturespatialexpansion[6,8]
ExampleofascenarioforParismetropolitanareaexpansionbetween2010and2100[8]
• Economic-basedurbanexpansionmodelscaninformdecisionmaking,andderiveprospectivescenariosaboutcitiesfutureexpansion/structuremodification
• Suchmodelscanbecoupledwithenvironmentalmodules(flooding-pronezones,urbanmicro-climate,airpollutionemissionanddispersion…)
• Themodelwehavedeveloped,NEDUM2D,isabletodynamicallyassessvariationsinrealestatepricesassociatedwithpublicinvestmentsorchangesofurbanplanningregulations.
• Thismodelisrelativelyeasytocalibrate,andisbasedonrobustandverifiableassumptions :itallowstheusertoeasilyunderstandthemechanismsinvolvedandtounderstandclearlytheuncertaintyandthevalidityoftheresultsobtained.
Acknowledgements :Thesestudiesreceivedfinancialsupport fromtheAgence Nationale delaRercherchethrough theprojectsVITE(ANR-14-CE22-0013-03)andDRAGON(ANR-14-ORAR-0005).ThisworkwasalsosupportedbytheWorldBankandtheFrenchministryofenvironment.FormermembersoftheteamincludeStéphaneHallegatteandPaoloAvner.Muchofthisworkwasdonethankstothem.
MaincitiesstudiedusingNEDUM-2Dmodel
Thespatialstructureofcitiesplaysakeyroleontheirenergyconsumptionlevelsandontheirvulnerabilitiestoenvironmentalhazards.Thetypeofurbangrowththatcitieswillexperienceinthenextdecadeswillthereforehavemajorimplicationsforclimatechangemitigationandadaptation.
ObjectiveInourwork,westudyandmodelthemechanismsdrivingtheurbanexpansionofcities.Weanalyzeeconomic,environmentalandsocialconsequencesofpoliciesaimingatimpactingurbansprawl.
ApproachUrbanshapeistheresultof2forces:• Statedecisions:Land-useconstraints,zoning,urbanismpolicies...• Aggregationofmultipleindividualdecisionstakenbytheinhabitants,andoften
reflectedinalandmarket(thesedecisionscanbeinfluencedbypolicies,e.g.transportpolicies).
Thesecondforce(themarket)canbeanalyzedthrougheconomicmodels.Weusesuchamodeltosimulateprospectivescenariosofcitygrowthandtoassesstheconsequencesofvariouspolicies.
1.Contextandmainideas
Usingonlythemostfundamentaleconomicprinciplesfromurbaneconomicsliterature,NEDUM-2Dmodelenablestomodeltheseinteractionsandtobuildscenariosoncityconceivablefutureevolutions.Itusesasinputsscenariosonthecity’sfuturedemography,transportsystemandlanduseconstraints.Thismodelisbynatureanidealizationofreality,butimplementationsonseveralcitiesondifferentcontinentshaveshownthatitreproducesfaithfullymaincharacteristicsofinhabitantsresidentialchoices,buildingsconstructionandrealestatepricesacrossanurbanarea.
Transport
Land pricesLandplanning
Transport,landplanningpoliciesandrealestatepriceseachinteractwiteachother.Eachofthemimpactsresidentiallocationchoicesofcityinhabitants,whichthemselvesactonlandprices,andontransportdemand.
Implicationsofcitygrowthscenariosintermsofgreenhousegasesemissions,airpollutionandnaturalhazardsvulnerability[5,6,7]
Example:simulationofairtemperaturechangeincaseofheatwave,inasimulatedscenarioinwhichParisregionbecomeslesscompact[5]
Example:simulationofpotentialimpactsonpopulationdensityofanovelconstructiontaxinParisregion.
1.Viguié,V.&Hallegatte,S.Trade-offsandsynergiesinurbanclimatepolicies.Nat.Clim.Change2,334–337(2012).2.Avner,P.,Viguié,V.&Hallegatte,S.Modélisation del’effet d’une taxe sur laconstruction.Rev.OFCEN° 128,341–364(2013).3.Avner,P.,Rentschler,J.E.&Hallegatte,S.Carbonpriceefficiency:lock-inandpathdependenceinurbanformsandtransportinfrastructure.WorldBankPolicyRes.Work.Pap.(2014).4.Avner,P.,Mehndiratta,S.R.,Viguie,V.&Hallegatte,S.Buses,housesorcash ?socio-economic,spatialandenvironmentalconsequencesofreformingpublictransportsubsidiesinBuenosAires.1–54(TheWorldBank,2017).5.Lemonsu,A.,Viguié,V.,Daniel,M.&Masson,V.Vulnerabilitytoheatwaves:Impactofurbanexpansionscenariosonurbanheatisland andheatstressinParis(France).UrbanClim.14,586–605(2015).6.Houet,T.etal.Combiningnarrativesandmodelling approachestosimulatefinescaleandlong-termurbangrowthscenariosforclimateadaptation.Environ.Model.Softw.86,1–13(2016).7.Masson,V.etal.Adaptingcitiestoclimatechange:Asystemicmodellingapproach.UrbanClim.10,407–429(2014).8.Viguié,V.,Hallegatte,S.&Rozenberg,J.Downscalinglongtermsocio-economicscenariosatcityscale:AcasestudyonParis.Technol.Forecast.Soc.Change87,305–324(2014).9.Viguie,V.&Hallegatte,S.Urbaninfrastructureinvestmentandrent-capturepotentials.WorldBankPolicyRes.Work.Pap.(2014).
Analysisoftheconsequencesofdelaysintheimplementationofemissionreductionatcityscale[3]
Relativeimpactsofcarbontaxesoncommuting-
relatedemissionlevelsinParisregionin2020for
scenarioswithandwithoutpublictransport [3]
Analysisofthetrade-offsandsynergiesbetweenmitigationandadaptationpoliciesatcityscale[1]
ExampleofanalysisinParismetropolitanarea[1]
NATURE CLIMATE CHANGE DOI: 10.1038/NCLIMATE1434 LETTERS
Adaptation and natural risk reduction
Natural area and biodiversity protectionPolicy neutrality
Housing affordability
Do-nothing scenario
Greenbelt policy
Public-transport subsidy
Zoning policy to reduce the riskof flooding
Climate change mitigation
Figure 2 | Consequences of a greenbelt policy, a public-transport subsidyand a zoning policy to reduce the risk of flooding compared with thedo-nothing scenario. Axes orientation is such that directions towards theexterior of the radar plot represent positive outcomes. Preferred outcomesare a decrease in average distance travelled by car, in the population livingin flood-prone areas, in total urbanized area, in spatial Gini index, and anincrease in average dwelling size.
Figure 2 presents graphically the positive and negative impactsof the three policies. The impact of each policy on each indicatorhas been assigned a score and is located along one of the five axesof the figure. The �100% score is in the middle of the figure;the +100% score is at the extremity of each axis. All scores aremeasured relative to the do-nothing scenario, which is assigneda 0% score. The +100% score goes to the preferred outcomeamong all considered policies. Each policy is thus ranked bestwhen the corresponding coloured area is biggest. For instance,Fig. 2 shows that a public-transport subsidy improves the situationcompared with the do-nothing scenario for three policy goals(climate change mitigation, housing affordability, and adaptationand disaster risk reduction), and makes it worse with respectto two policy goals (natural area and biodiversity protectionand policy neutrality).
As Fig. 2 shows, each policy causes both positive and negativeoutcomeswith respect to different policy goals when comparedwiththe do-nothing scenario. Each policy thus seems to be undesirablebecause it has negative consequences with respect to at least onepolicy goal. This result can explain, for instance, why it is sodifficult to implement efficient flood-zoning or greenbelt policieson a local scale, even when it is required by national law14. Indeed,negative side effects on housing availability and on developmentopportunities understandably create political resistance.
However, as Fig. 3 shows, a mix that includes the three policiescan mitigate the adverse consequences of each individual policy.For instance, public-transport subsidies decrease the real-estatepressures caused by a greenbelt or a flood-zoning policy. Floodzoning also prevents the greenbelt from increasing the populationat risk of floods. When all three policies are applied together,the situation is improved as measured along all policy goalscompared with the do-nothing scenario. In particular, theseresults indicate that flood-zoning and greenbelt policies need tobe combined with transportation policies to gain real politicalmomentum and effectiveness.
Note that in a policy mix, the consequences of each policyare not simply additive. For instance, when all three policies areimplemented, the decrease in population in flood-prone areas
Adaptation and natural risk reduction
Public-transport subsidy, greenbelt policy and zoning to reducethe risk of flooding
Natural area and biodiversity protectionPolicy neutrality
Housing affordability
Do-nothing scenario
Climate change mitigation
Figure 3 | Consequences of a policy mix including all three policies. Axesorientation is such that directions towards the exterior of the radar plotrepresent positive outcomes.
is smaller than the sum of the variation caused by each policytaken separately. This nonlinearity and the complexity of policyinteractions explainwhy it is useful to analyse various urban policiestogether, in a consistent framework.
Our analysis shows that building win–win solutions by combin-ing policies is possible and leads to more efficient outcomes thana set of policies developed independently. Climate goals can thusbe reached more efficiently and with higher social acceptability,if they are implemented through taking into account existingstrategic urban planning, rather than by creating new independentclimate-specific plans. Such mainstreaming of climate objectiveswith other policy goals is found to help design better policies in ourmodel, confirming previous findings in other domains15–17.
Obviously, it does not mean that win–win strategies are alwaysavailable. In some cases, trade-offs will remain unavoidable andurban decision makers will need to make tough choices. Thiscan be done by associating different weights to our indicators—through a process of stakeholder engagement—and by maximizingthe resulting weighted sum of indicators. But our analysis showsthat a mainstreamed approach can allow for the design ofpolicies that are robust to differences in the weights of differentindicators, as they improve all indicators. Such policies areparticularly easy to implement, because they aremore likely to seemdesirable to all stakeholders, in spite of their different priorities,objectives and world views.
MethodsWe use the NEDUM-2D model to simulate the evolution of the Paris urban areabetween 2010 and 2030. This model is an extension of that described in previouspapers18,19, which is based on classical economic theory20–22. This theory explainsthe spatial distribution of land and real-estate values, dwelling sizes, populationdensity, and building height and density.
Our approach aims to bridge the gap between high-complexityland-use–transport interaction models23 and theoretical urban-economicsmodels. To do so, we propose a model that is fully based on microeconomicfoundations describing economic agent behaviours (like theoretical models), butthat can be calibrated on realistic transportation networks and include preciseland-use regulations and natural land characteristics (for example, rivers andother natural areas).
Two main mechanisms drive the model. First, households choose wherethey live and the size of their accommodation by assessing the trade-off betweenproximity to the city centre and housing costs. Living close to the city centrereduces transportation costs, but housing costs (per unit of area) are higher there.Theoretical extensions to account for decentralized production have been proposed,but are not included in this analysis24–26. Second, we assume that landownerscombine land with capital to produce housing: they choose to build more orless housing (that is, larger or smaller buildings) at a specific location dependingon local real-estate prices and construction costs. We assume that households
NATURE CLIMATE CHANGE | ADVANCE ONLINE PUBLICATION | www.nature.com/natureclimatechange 3
NATURE CLIMATE CHANGE DOI: 10.1038/NCLIMATE1434 LETTERS
Adaptation and natural risk reduction
Natural area and biodiversity protectionPolicy neutrality
Housing affordability
Do-nothing scenario
Greenbelt policy
Public-transport subsidy
Zoning policy to reduce the riskof flooding
Climate change mitigation
Figure 2 | Consequences of a greenbelt policy, a public-transport subsidyand a zoning policy to reduce the risk of flooding compared with thedo-nothing scenario. Axes orientation is such that directions towards theexterior of the radar plot represent positive outcomes. Preferred outcomesare a decrease in average distance travelled by car, in the population livingin flood-prone areas, in total urbanized area, in spatial Gini index, and anincrease in average dwelling size.
Figure 2 presents graphically the positive and negative impactsof the three policies. The impact of each policy on each indicatorhas been assigned a score and is located along one of the five axesof the figure. The �100% score is in the middle of the figure;the +100% score is at the extremity of each axis. All scores aremeasured relative to the do-nothing scenario, which is assigneda 0% score. The +100% score goes to the preferred outcomeamong all considered policies. Each policy is thus ranked bestwhen the corresponding coloured area is biggest. For instance,Fig. 2 shows that a public-transport subsidy improves the situationcompared with the do-nothing scenario for three policy goals(climate change mitigation, housing affordability, and adaptationand disaster risk reduction), and makes it worse with respectto two policy goals (natural area and biodiversity protectionand policy neutrality).
As Fig. 2 shows, each policy causes both positive and negativeoutcomeswith respect to different policy goals when comparedwiththe do-nothing scenario. Each policy thus seems to be undesirablebecause it has negative consequences with respect to at least onepolicy goal. This result can explain, for instance, why it is sodifficult to implement efficient flood-zoning or greenbelt policieson a local scale, even when it is required by national law14. Indeed,negative side effects on housing availability and on developmentopportunities understandably create political resistance.
However, as Fig. 3 shows, a mix that includes the three policiescan mitigate the adverse consequences of each individual policy.For instance, public-transport subsidies decrease the real-estatepressures caused by a greenbelt or a flood-zoning policy. Floodzoning also prevents the greenbelt from increasing the populationat risk of floods. When all three policies are applied together,the situation is improved as measured along all policy goalscompared with the do-nothing scenario. In particular, theseresults indicate that flood-zoning and greenbelt policies need tobe combined with transportation policies to gain real politicalmomentum and effectiveness.
Note that in a policy mix, the consequences of each policyare not simply additive. For instance, when all three policies areimplemented, the decrease in population in flood-prone areas
Adaptation and natural risk reduction
Public-transport subsidy, greenbelt policy and zoning to reducethe risk of flooding
Natural area and biodiversity protectionPolicy neutrality
Housing affordability
Do-nothing scenario
Climate change mitigation
Figure 3 | Consequences of a policy mix including all three policies. Axesorientation is such that directions towards the exterior of the radar plotrepresent positive outcomes.
is smaller than the sum of the variation caused by each policytaken separately. This nonlinearity and the complexity of policyinteractions explainwhy it is useful to analyse various urban policiestogether, in a consistent framework.
Our analysis shows that building win–win solutions by combin-ing policies is possible and leads to more efficient outcomes thana set of policies developed independently. Climate goals can thusbe reached more efficiently and with higher social acceptability,if they are implemented through taking into account existingstrategic urban planning, rather than by creating new independentclimate-specific plans. Such mainstreaming of climate objectiveswith other policy goals is found to help design better policies in ourmodel, confirming previous findings in other domains15–17.
Obviously, it does not mean that win–win strategies are alwaysavailable. In some cases, trade-offs will remain unavoidable andurban decision makers will need to make tough choices. Thiscan be done by associating different weights to our indicators—through a process of stakeholder engagement—and by maximizingthe resulting weighted sum of indicators. But our analysis showsthat a mainstreamed approach can allow for the design ofpolicies that are robust to differences in the weights of differentindicators, as they improve all indicators. Such policies areparticularly easy to implement, because they aremore likely to seemdesirable to all stakeholders, in spite of their different priorities,objectives and world views.
MethodsWe use the NEDUM-2D model to simulate the evolution of the Paris urban areabetween 2010 and 2030. This model is an extension of that described in previouspapers18,19, which is based on classical economic theory20–22. This theory explainsthe spatial distribution of land and real-estate values, dwelling sizes, populationdensity, and building height and density.
Our approach aims to bridge the gap between high-complexityland-use–transport interaction models23 and theoretical urban-economicsmodels. To do so, we propose a model that is fully based on microeconomicfoundations describing economic agent behaviours (like theoretical models), butthat can be calibrated on realistic transportation networks and include preciseland-use regulations and natural land characteristics (for example, rivers andother natural areas).
Two main mechanisms drive the model. First, households choose wherethey live and the size of their accommodation by assessing the trade-off betweenproximity to the city centre and housing costs. Living close to the city centrereduces transportation costs, but housing costs (per unit of area) are higher there.Theoretical extensions to account for decentralized production have been proposed,but are not included in this analysis24–26. Second, we assume that landownerscombine land with capital to produce housing: they choose to build more orless housing (that is, larger or smaller buildings) at a specific location dependingon local real-estate prices and construction costs. We assume that households
NATURE CLIMATE CHANGE | ADVANCE ONLINE PUBLICATION | www.nature.com/natureclimatechange 3
