Polycentrism in the Spanish metropolitan system: An analysis for 7 metro areas Carlos, Marmolejo; Carlos, Aguirre; Jaume Masip; Eduardo, Chica & Claudia Pérez Centre for Land Policy and Valuations Barcelona s School of Architecture Polytechnic University of Catalonia 51 st European Congress of the Regional Science Association Barcelona, September 2011
The research question Polycentrism in the Spanish metropolitan system
Research question Whether or not the biggest metropolitan systems in Spain are advancing towards polycentrism? 3
Case study 7 Biggest metropolitan areas Number of municipalities Built up land LTL (working places) Population Density a b c.=(b+c)/a Madrid 183 860 2.446.400 5.542.843 9.291 Barcelona 184 745 1.903.867 4.530.164 8.636 Valencia 104 308 686.247 1.792.375 8.046 Sevilla 52 237 447.849 1.381.531 7.719 Bilbao 123 112 445.666 1.231.367 15.024 Zaragoza 88 127 301.860 724.335 8.066 Málaga 32 194 366.525 994.984 7.032 Source: Corine Land Cover & National Census 2001 (ICN, INE) Not considering the system formed by Oviedo-Gijón-Avilés found by Feria (2009) bigger than Málaga 4
Work in progress In this paper we present the first stage of the research consisting in the delimitation of metropolitan systems and characterization of their structure by means of the subcentre detection 5
Metropolis delimitation Polycentrism in the Spanish metropolitan system
El Delimitation ruido mensurado approaches in literature Nurec, 1994 (200m built up gap) Morphological Serra et al., 2002 (density 250 p/ sq km) Functional USA Census Bureau (since 1951) France (Julien, 2000); Italy (Martinotti, 1991); Canada (Murphy, 2003); UK (GEMACA; Hall & Pain; 2006); GEMACA Spain (Roca et al., 1997, 2005) Feria (2008, 2009) 1. Arbitrary thresholds 2. Periphery-centre problem
El Delimitation ruido mensurado approach followed Functional (Roca et al 2009; 2011) Vi ij = F 2 ij POR LTL i j + F POR 2 ji j LTL i Vi= Interaction value Fij= Residence (i) to work (j) commuting flows POR= employed population LTL= Jobs
El Delimitation ruido mensurado approach followed Functional (Roca et al 2009; 2011) x A x B D x x C Protosystem x + contiguity and 50% self containment = Subsystem
Integration of subsystems Subsystems can be integrated into metropolitan areas by means of the IV calculated among subsystems 1/1000 IV 10
Integration of subsystems 11
Methods of subcentre Identification Polycentrism in the Spanish metropolitan system
El Subcentre ruido mensurado identification Methods based on density analysis 1) Density peaks (using GIS analysis) McDonald (1997), MacDonald & McMillen (1990), Craig & Ng (2001). 2) Density cutoffs (density and mass) Giuliano & Small (1991), Song (1994), Cervero & Wu (1997), McMillen & McDonald (1997), Bogart & Ferry (1999), Anderson & Bogart (2001), Sheamur & Coffey (2002), Giuliano & Readfearn (2007), Garcia López (2007), Hall & Pain (2006). 3) Parametric models (density is a function of distance to CBD) McDonald & Prather (1994). LnD = k + B i D CBD i 4) Non-parametric models (locally or geographically weighted regressions) McMillen, (2001, 2002); Craig & Ng (2001), Readfearn (2006, 2007), Suarez & Delgado (2009).
El Subcentre ruido mensurado identification Methods based on functional analysis 1) Attraction ratio (multipurpose travelers attracted / employed people) Gordon, Richardson & Giuliano (1989); Gordon & Richardson (1996); Ruiz & Marmolejo (2008) 2) Positive residuals on (double and simple constrained) gravity models Emanuel & Dematteis (1990), Trullén & Boix (2000). 3) Interaction Value Roca, Marmolejo & Moix (2009); Roca, Arellano & Moix (2011)
El Centre ruido mensurado identification Hybrid method based on functional and morphological criteria In this paper the centre (Central Economical defined as: Continuous) has been 1) those municipalities inscribed in the functional subsystem of the central municipality 2) and at the same time having at least 700 employees per sq km (Gemaca and Polynet Criteria) 3) and being continuous in a 200 m gap framework (UN criteria)
Used data Polycentrism in the Spanish metropolitan system
Used data Metropolitan Area of Madrid 1. Travel to work commuting data (Census 2001) 2. Teleatlas road network (2000) 3. Corine Land Cover (2000)
Results of subcentre identification Polycentrism in the Spanish metropolitan system
El Result ruido mensurado of parametric models The parametric models have failed in explaining the employment density in half of the metropolitan areas Parametric methods CL R2 adj B dist CBD Madrid 0,365-0,041 Barcelona 0,298-0,034 Valencia 0,231-0,03 Sevilla 0,046-0,012 Bilbao -,009 0,002 LnD = k + B i D CBD i Zaragoza 0,149-0,015 Málaga 0,028-0,007 Dependent variable: Ln employment density Madrid Barcelona Valencia
El Result ruido mensurado of parametric models Madrid Barcelona Valencia
El Result ruido mensurado of cut-off and functional approaches Barcelona, Valencia and Bilbao are the metropolises with the biggest number of subcentres beyond the central economic continuous Cut-off Functional Metrópoli LTL LTL en CEC (%) Subcentros pot. fuera del CEC LTL en subcentros pot. fuera del CEC (%) Pob en subcentros pot. Fuera del CEC (%) Subcentros pot. fuera del CEC Madrid 2.446 72% 4 7% 8% 8,0 Barcelona 1.904 55% 7 15% 14% 23,0 Valencia 689 47% 5 10% 7% 17,0 Sevilla 448 63% 0% 0% 7,0 Bilbao 438 56% 3 7% 5% 14,0 Zaragoza 302 81% 1 3% 0% 7,0 Málaga 367 52% 1 6% 5% 3,0 LTl en miles de personas
Polycentrism (Share of subcentre employment) El Monocentrism ruido mensurado vs Polycentrism according functional approach Barcelona, Valencia, Bilbao and Málaga stand out as the most polycentric metropolises 90% 85% Perfectly monocentric system Monocentrism (Share of CEC employment) <<- Menos - Monocentrismo-Más -->> % LTL en el CEC Zaragoza 80% 75% Madrid 70% 65% Bilbao Barcelona Sevilla 60% 55% 50% 45% Distopic dispersed system Utopic Equipotential System Valencia Málaga 40% 0% 5% 10% 15% 20% 25% % LTL en los subcentros potenciales fuera del CEC <<- Menos - Policentrismo-Más -->> El tamaño de la esfera es significativo del número de subcentros potenciales Fuente: elaboración propia
El Polycentrism ruido mensurado according to Shannon s entropy index n H = 1* PLTL ln( PLTL ) m i i i Where centres (CEC and subcenters) are i And H is the complexity of employment distribution among centres in a given metropolitan system m 0,90 0,80 Entropía en la distribución de los LTL entre los núcleos 0,70 0,60 0,50 0,40 0,30 0,20 0,10 1. For a dispersed system is not possible to calculate H 2. In a perfect monocentric system H is 0 3. The bigger is H the more polycentric is the system - Barcelona Valencia Bilbao Málaga Sevilla Madrid Zaragoza Fuente: Elaboración propia
Results of Metropolitan complexity analysis Polycentrism in the Spanish metropolitan system
Complexity of metropolitan structuration Subsystem Metro area A CEC Subsystem Metro area B Subsystem Subsystem CEC Subsystem Subsystem
Complexity of metropolitan structuration
Complexity of metropolitan structuration Number of steeps that each municipality needs to intreract with CBD
Final remarks Polycentrism in the Spanish metropolitan system
Final remarks 1. In terms of methodology the approach devised by Roca et al (2005, 2009 & 2011) allows for simultaneously: 1) delimit the metropolitan subcentre, 2) identify subcentres; and 3) depict the topology of metropolitan systems 2. In Spain it seems to be two groups of metropolises with divergent structure patterns. Barcelona, Valencia and Bilbao stand out as the most policentrical 3. On the other hand Madrid, Sevilla and Zaragoza stand out as most monocentrical 4. Màlaga is a particular case because having few subcentres they do concentrate an important share of employment (Equipotential polycentrism?) 5. Polycentrical metropolises seems to have been produced both by the descentralization of employment and the integration of formerly independent urban systems 29
Thank you carlos.marmolejo@upc.edu Centre for Land Policy and Valuations Polytechnic University of Catalonia 30