Article: Porta S, Latora V. (2008), The spatial analysis of urban systems: Multiple Centrality Assessment and the dynamics on street networks, in Hasic T (ed), ‘New Urbanism and beyond: the future of urban design’, Rizzoli International, New York, NY.

Disclaimer: This paper not necessarily reflects the final definitive publication: it might be a pre-copy-editing or a post-print author-produced .pdf or in any case a different version of that. Therefore the reader is advised to refer to the publishing house’s archive system for the original authenticated version of this paper.

CONTENTS

NEW URBANISM & BEYOND' 9 TIGRAN HAAS

PART I THEORIES OF URBAN FORM I 14

1.1 GENERATIVE CODES:Thc Path to BUilding Welcoming.

Beautiful. Sustainable Neighborhoods f 14

C H RISTOPHER ALEXANDER

1.2 THE NEW SCIENCE OF SPACE AND THE ART OF PLACE:

Toward a Space· led Paradigm for Researching and Designing the City I 30

BILL H ILLI ER

I.l THREE URBANI$MS: New. Everyday. and Post I 40

DOUGLAS KELBAUGH

I.~ URBAN RENAISSANCE. URBAN VILLAGES. SMART GROWT H:

Find the Dif(e,'ences I 48

PETE R H AI.L

1.5 SETILEMENTS OFTHE FUTURE I 52 LEON KR IIER

1.6 SOMETHING LIVED. SOMETHING DREAMED:

Principles and Poetics In Urban Design I 58

WI LL IAM Mc DONOUGH

PART 2 EXPLOitiNG NEW URBANISM I 64

2. 1 THE TRADITIONAL NEIGHBORHOOD

AND URB,l\N SPRAWL I 64

ANORtS CUANY AND ELI ZABETH PLATE R- ZYBE RK

2.2 THE URBAN N ETW ORK I 67

PETER CALT H O RP E

2.3 NEW URBANISM: A Forum. Not a Formula I 70

elLEN DUNHAM-JONES

H MAKING CLOTH FROM THREADS I 74

DANIEL SOLOMON

2.S THE UNBEARABLE LIGHTNESS OF NEW URBANISM I 77

EMILY TALEN

2..6 THE CHALLENGES OF ACHIEVING SOCIAL OBJECTIVES

THROUGH MIXED USE I 80

JILL GRANT

PART :) SUBURBIA , SPRAWL, AND URBAN DECLINE I 86

3. 1 THE SUBURBA N CITY I 86

DOLORES H AYDEN

3.2 SPEED, SIZE. AND THE DESTRUCTION OF CITIES I 89

DO M NOZZI

1.J RESTRAINING SPRAWL: A Common Interest to Enhance

the Quality of life for All I 9]

LOUISE NYSTR O M

3.4 THE EMERGENCE OF MIXED-USE TOWN CENTERS

IN THE UNITED STATES I 98

TO M MA RTINEAU

3.S URBAN DESIGN AND THE METROPOLIS I 103

ROBERT BEAUREGARD

PART 4 6.3 THE NEED FOR PATIEN T EQUITY

STREET·S , TRANSPORT, IN CREATIN G GREAT PLACES I 167

AND PlIBLIC REALM I 106 CHRI STO PHER B. LEINBERGER

i .1 lIVELY.ATIRACTIVE.AND SAFE CITIES- BUT HOWl I 106 6.<4 A CHANGED fOCUS ON RETAIL JAN GEHL AND ITS IMPLICATIONS ON PLANNING I 173

ANDERS ALMtR

4.1 GREAT STREETS AND CITY PLANNING I 109

ALLAN JACOBS 6.S THE ROLE OF CULTURE IN URBAN DEVElOPMENT I 176

GDRAN CARS

i .3 TRUE URBANISM AND THE EUROPEAN SQUARE: Ciltilly~t fOJ" Social Engagement and Oemouatlc Dialogue I 112 6.6 HERITAGE MANAGEMENT IN URBAN DEVELOPMENT

SUZANNE CROWHURST LENNARD PLANNING I 182 KRISTER OLSSON

i .i URBANISM ANDl HE ARTICULATION OFTHE BOUNDARY, 117

All MADANI POUR

i .S THE NEW URBANISM AND PUBLIC SPACE I 120

GEORGE I!AIRD

4.6 TRANSIT-O RIENTED DEVELOPMENT IN AM ERICA:

Strategies. Issues. Policy Directions f 12'1

ROBERT CERVERO

PART S. THE EL. EMENTS OF URBAN DESIGN I 1)0

5.1 RECOMBINANT URBANISM I 130

DAVID GItAH AME SHANE

S.l SPATIAL CAPITAL AND HOWTO MEASURE IT:

An Outline of an AnalytICal Theory of Urban Form ( 135

LARS MAR CUS

S.3 CENTRALITY AND CITIES: Multiple Centrality Assessment

as a Tool for Urban AnalySIS and Design' 1'10

SERGIO PORTA AND VITO LATORA

PART 7 SUSTAINABILITY. TECHNOLOGY, AND THE ENVIRONMENT 1186

7.1 DOES N EW U RBANISM REALLY OVERCOME

AUTO MOBILE DEPEN DENCE? , 186

PETER NEWMAN

7.1 GREEN URBANISM: A Manifesto for Re-Earth,ng Cttle~ I 189

TIMOTHY BEATlEY

7.3 ECO-TECH URBANISM:

Towards the Green and Smart CIty I J97

OUSHKO BO GUNOVICH

7.4 ON COMMUNITY (from Zoon Pohtikon to Multitude) I 201

LARS LERUP

7.S PETROCOLLAPSE AND THE LONG EMERGENCY I 20'1

JAMES HOWARD KUNSTlER

PART 8

5.4 INTEGRATED APPROACHES AND DYNAMIC PROCESSES 1 116 URBAN DIGITAL SPACES ARUN JAIN AND CYSER CITIES 1208

5.S PLANNING fOR WALKABLE STREETS I 15 3 S.! CONNECTIVITY AND URBAN SPACE I 208

RICK HALL W ILLI AM M!TCHELL

PART -6 REAL E STATE , CITY MARKETING , AND CULTURE I 158

6.1 URBAN RETAIL PLANNING PRINCIPLES

FOR TRADITIONAL NEIGHBORHOODS I 158

ROB ERT GI BBS

6.2 CONFRONTING THE QUESTION OF MARKET DEMAND

FOR URBAN RESIDENTIAL DEVELOPMENT I 163

LAURIE VOLK AND TODD ZIMMERMAN

6 Contents

S.l URBAN NETWORK ARCHITECTURES

AND THE STRUCTURING OF fUTURE CITIES I 212

STEPHEN GRAHAM

8.3 LOCATIVE MEOlA URBANISM I 218

MALCOLM McCULLOUGH

8.4 WHAT IS TH( INTERNET DOING TO COMMUNITY

ANDVISAVER$A I 221

BARRY WE LLMAN

8,S THE SANTA. FE-ING OF THE URBAN AND URBANE 1 225

JOEl GARREAU

PART 9 SOCIAL CAPITAL AND MUTUAL BENEFIT 12)2

9.1 CREATING COMMON SPACES: Urban Planning.

Local Media, and Technology 1 212

ROBERT PUTNAM AND LEWIS FELDSTEIN

9.1 THE THIRD PLACE: A Belated Concept 1 234

RAY OLDENBURG

9.3 DEMOCRACY AND "NEIGHBORLY COMMUNITIES":

Some Theoretical ConSiderations on

the BUIlt Environment 1 238

KEVIN LEYDEN AND PHILIP MICHELBACH

9.4 SPRAWLING CITIES 1 24'1

KARL OLOV ARNSTBERG

9.5 BEYOND THE NEIGHBORHOOD:

New Urbanism as CIVIC Renewal N9

DAVID BRAIN

9.6 URBAN DESIGN AND DEVELOPMENT

IN THE SWEDISH TRADiTION f 255

KNUT STR OMBERG

PART 10 COMPLI:: XITY SCIENCE AND NI:W URBAN FORMS f 2S8

10,1 HIERARCHY, SCALE, AND COMPLEXITY

IN URBAN DESIGN I 2S8

MICHAEL BATTY

10,2 GROWING SUSTAINABLE SUBURBS: An Incremental

Strategy for Reconstructing Modern Sprawl 1 262

LUCIEN STEIL. NIKOS A . SALINGAROS.

AND MICHAEL MEHAFFY

10.3 COMPLEX SYSTEMS THINKING AND NEW URBANISM I 275

T. IRENE SANDERS

10.4 NEW SCIENCE. NEW ARCHITEaURE .. NEW URBANISM! f 280

MICHAEL MEHAFFY

10.5 QUANTUJ"1 URBANISM:

Urban Design in the Post-Cartesian Paradigm I 288

AYSSAR ARIDA

PART II BEYOND URBANISM AND THE FUTURE OF CITIES 1292

11.1 ANOTHER NEW URBANISM I 292

EDWARD SOJA

11.2 NEW URBANISM IN THE AGE Of RE~URBANISM I 296

ROBERT FIS HMAN

I U THE NEW URBANISM IN THE TWENTY-FIRST CENTURY:

Progress or Problem? I 299

EDWARD ROBBIN S

11.4 MAKING PUBLIC INTERVENTIONS

IN TODAY'S MASSIVE CITIES I ]0]

SASKIA SASSEN

11.5 THE WORLD IS SPIKY: Globalization Has Changed

the Economic Playing Field, but Hasn't Leveled It I ]09

RICHARD FLORIDA

11.6 SPACE OF FLOWS, SPACE OF PLACES: Materials for

a Theory of Urbanism In the Information Age 1 ] I '\

MANUEL CASTE LLS

NOTES I 122

CONTRIBUTORS I 338

ACKNOWLEDGMENTS ]40

INDEX' ]41

CREDITS' 3'\8

Contems 7

5 . 3

CENTRALITY AND CITIES

MULTIPLE CENTRALITY ASSESSMENT AS ATOOL FOR URBAN ANALYSIS AND DESIGN

SERGIO PORTA AND VITO LATORA

Int rodu ction: Centrality and C ities

If YOll ,lsk :l grocer wherc he would loc:ltc his new shop YOll will probably see him mumbling a liule bit in search of something. He is exploring his Illcntalmap of rhe city

or the neighborhood. Then he will come om pointing his fin­ger on the Illap and saying: ""yes, [hac's the best location.'" Why that? "You know-he will point o ur- here's where all rhe people pass when going to work evcry morning, then here,;][ the corner between Massave and Pearl (I am pick­ing lip an example from Boston), you are quite close to eve rywhere in th e neighborhood , then Massave is re~t1 ly a spine that everyo:ne knows in the city, for it is straight down to the river, then to the South End and downtown as well." There is a lesson here anom how a real city works. It asserts the crucial importance of one bctor, among many others, what we ca ll "ce ntrality." Everyone knows that a p lace which is central has some speci:ll featu res to offer in man y ways to those who live or work in cities: it is more visible, morc accessible from the immediate surroundings as well ;15

fro m far ;1\vay, it is more pop ular in terms of peoplc wa lk­ing around a nd anracting potential customers, it has a greater probability of developi ng as an urban landmark and a social catalyst or offer first 1cvel functions like theaters or office headquarters plus a larger diversity of o pportunities and goods. Tha t's why central locations arc mo re expensive in terms of real {'statc va lues and tend to bc socia lly selec­tive : because such special features make them provide a reasonahle trade off to la rger in vestments than in less cen­tral spots of the same urban area.

If you look at where ;1 c ity center is located, you will mostly-if not always-find that it sprou ts from the inter­secti on of two main routes, where some special configura ­tion of the [err:lin or some particular shape of the river system or (he w:lterfrOIll makes that place compulsory to pass thro ugh. That's where cities begin. Then, departing from such centra l locations, they grow up in time adding buildings and activit ies here and there. firstly along the main routes, then filling the in-between areas, then adding

,<0

streets that provide cycl ic routes and poims of return, then, as the structu re becomes more complex, forming new cen­tral streets an d places and adding bu ildings ~round them agai n. It is an evolutiona ry process that has been driving the forma tion of o ur urban bodies, the heart of human c ivil iza­ti on through the ages, for most of the seven millenniums of rhe hisrory of cities until the da wn of modernity and the very beg inning of the industrial age. Thus, we understand that centr~ l ity is not just at work ~t the heart of contemporary urban I ife, linking spatial for ms and collective behaviors, bur it is rather, at the heart of the evo lutionary process that made our cities what they have always been , with a strong impact on how they still a re today.

At this point, the reader should be aware of the idea that urban planners and designers ha ve spent most 'of their time trying to understand and manage how centrality works in cities. Nothing could be less true. Despite the relevance of the issue, studies 011 centrality in cities have seldom been under­ta ken and never as a comprehensive approach to the subject. Geographers a nd transport planners have used centrality as a means to understand the location of uses and activities at the regional scale o r the level of convenience to reach a place from all other poin ts in an urban system I : in so doing, they have built their understanding on some hidden assumptions, particularly 011 the notion of centrality that, putting it shortl y, is limited to the fo llowing: the more a place is dose to all oth­ers, the more centr'll.

Since the early eighties space sYlitax, a methodology of spatia l ana lysis based on visibility and integra tion, has bcen open to urban designers and has offered them a whole range of opportunities to develop a deeper understanding o f some structu ra l properties of ci ty spaces, but such opportunities ha ve se ldom been underSTOod, often perceived as a quantita­tive threa t TO the creativity embedded in the art of city design . From a technical point of view, however, Sf){/ce SYlitax does not e.<;C:lpe SOTlle o f rhe shortcomings of regional analysis and tmllsport planning: it is nor a study of centra lity in its own sak e, it is mai nly based on the sa me concept of being central as being close 10 all o thers. It does not focus on the different

geograph ies that different w:ays of being central acru:a Jl y "shape" a ciry structure and how such different geographies correla te with the dYllamics th:at take place in a city (like commerci:'ll loc:lt ion, real estate values, o r information flows). However, recent developments from wit hi n the sl)(lce sy"tax community seem to have :acknmvledgcd centml ity and have opened to :alllore diversified concept of spa ti:a l config­uration in relation 10 approaches of soci:a l psychology and cognition. In any case, space s)"'lax aside, urban designers ha ve in general simply skipped the problem: they basic:llly deal with the aesthetics of rhe urba n landscape :'Ind the com­ponents of urb:m form rh ,lt embed regu larory issues li ke the height of buildings or the EA.R. (floor-area ratio). Mo re recently. urban designers have investigated a whole range of new issues rll:'lt :'Iddress the qUL'Stion of Ihe liveability of places under [he great umbrella of urban susrainahility: that's where this paper actually comes from.

In this paper we presel\( a discussion of cenrral ity in cities and a 1ll0(lel, named Mu ltiple Cenrra lity Assessment (l" ICA) , wh ich helps to m:l1lage centra lity fo r urban planning and design purposes. In so doing, we summarize a research 111M we have undertaken fo r the la st couple of years and is now being published in a journal of urban plann ing and physics.1

In add ition, we arc hereby offeri ng the first results of a new effort aimed at puning some light on how celltrality. in its various forms. correlates to sever~l l dYllamics of city life, research which is actua ll y slill undcr way over larger cases and datasets.

Representing the City

Gur reseMch on ccmral ity III cities is hascd on representing the stmctllre of tile city (which we idenrify as the system of ciry streets and their intcrsections ) by means of a mathema t­ical object ca lled a gra/)h. wh ich is a set of points, or nodes, linked by a set of lines, or edges. Graphs have been widely used for developing a deeper understanding o f how complex systems of all son s actually work. The property of a graph is, in fact. to focus on rhc re lationshi ps betwecn ind i\'idu~ll components mlher than the properties of such componentS themselves: in this sense, nodes might well represent compo­nents as different :as coamhors in scicnrific systcms (linked by coamhorsh ips), routers in internet systems (linked by tele­phone cables). airports in ai r-communication systems (linked by air TOmes), or individua ls in social systems (l inked by prey­predator, acq uai ntanceship, profess iona l or sexual relation­ships, contagious (Iisease transmissions) and many, Illany others. ln our case, we chose to represenr street inrerSl.'Ctions as nodes and strects as edges between nodes; one street as such is JUSt betwcen twO imerSl.'Ctions. 110 matter the nu mber of turns or curves it presents. Intuitive as it might appear, th is forma t is IH lt Ihe onl r possible one; space sylltax, for instance, is base<1 on an opposi te represema tion where !lodes are streets and intersections an' edgcs; morL·ovcr, in SIHICC S)'II­

tax one StrL'C t such as JUSt between twO turns. no maner the number of intersections it presents. following what geogra-

phers call a "seneralization process·' of geographic en tities. Howcver, the poi nt is th is: our represenration (figure I, image I). which we call ill/gelleralized ,)fimal, is the st:r.ndard format for all transport pla nners and gco-m:lpping profes­sionals on earth: that makes our MCA model cap:lbl!' of tak­ing adva nt:1ge of an immense amount of informatiOIl already marketed and constantly updat!'d worl dwide, i.e .• the road graphs of ,II I traffic departments of municipal or region:al :'Idmi nistration bodies.

Defi ning Centrality

Basicall y, Ihe professionals who have addressed a specific work on central ity itself arc rhe strw.::tural sociologists: they actuall y have been doi ng since the early fifties as a means to manage challenges coming from human. economic, or institutional organizations. Structur:11 sociologists ha ve exam in e(1 s llch problems in terms of how cach person (organ iza tion) is related to each other person (organization) in the reference system (the org:lllization or the syStem of organ izations), and they have done this in a qllalltitlltive way. They developed (hree main measurcs; on one hand, they counted how many edgcs each node has, and called (his degree celltrality, or CD; on the other hand. they counred how close each node is to all others and called th is closelless ccl/trality , or C C; fina ll y, they measured to what extent each node is traversed by the shortest paths that link each other co uple of nodes and ca lled this betweelllless eel/tral­ity, or CH. We have added (wo ot her indices: the first meas­ures how much each node is cri tical to the system as a who le, in terms of the drop of the system's efficiency that wou ld fo llow the removal of the node wi th all associatcd edges, and we have named this ill(omwtioll cCl/tmlity or 0; the second measures how much the real paths that connect each node ro a ll others diverge to the virtual straight paths, and we have na med this straightlless celltrality, or 0.

In order to deal wi th such indice...;; we shou ld agree on a definition of distance: not surprisingly, sociologists measu red the d istance between twO nodes (pcrsons or organiz:uions) in terms of Ihe number of other nodes rhat separate them; in fact, they stu dicd nOllspatial systems, i.e., systems where the position of nodes is not defined in space (spatial distance docs not make a difference). In ou r case, we measur('d the dista nce in terms of the number of meters that se parate two nodes (street intersections) along the connecting edge (street); i ll fac t, we have embedded ou r street gra phs in sp:1ce rooting the whole process in a Geographic Informa­tion System (G IS ) environ ment.

Centrality Over Planned and Self-organized Cities

The first result of our studies might not strike the :l\Id ience like an intcllectu :al bomb but it acwally emerges as the foun­dation of many interesting practica l uses especially in urban

11 1

planning and design: despi te the extreme fragmenrarion of rhe graph, ccnrraliry docs not SproUl up here and there in a sca ttered m:l nner, but r:l ther it forms legible rou[('s and :lreas consistentl y ordered ;n a hiemrchical spar;.11 distriburion (figure I, im:lgcs 3-6). Roures and areas of similar levels of centrality emerge over the complex urban system according to some imerna l rules thaI ,'aT)' from index ro index giving place to a multifaceted geogr.lphy of cities. Urban places tll:u arc centr:ll in terms of closeness may not be as central in terms of betweenlless (very offen rhey arc not ce ntral at a ll ). In shorr: there an" apparently many ways for a place to bc cemral in a ciry. Bur there is more: Tht: geography of cemra l­ity that emerges for a given index calcubrcd at the global sca le (i.e., relating; cach node with all others in the systcm) rypieally diverges a lot from thar calcu la tion, for the sa me index, at the local scales (i.e., relating each node to a subser of nodes loc:n ed withi n a certai n distance d from it (figure I, images 2, 3). Tha t mea ns th:u a place nlay nOt just be differ­ent according to different killds of ccmmliry, bu t also to dif­ferent s{Jatial scales of behaviors: if you are shoppi ng for your ordinary dai ly needs you usc a cerram city, if you need r:lre services (l ike arlending a university course or a dance pc:or­forma nee), you usc a differem city.

Not all cities' structures are complex in the same way and to the sallle extenr. Planned and self-organized citiL'S look ,'err different. the forme r ex hibit a geometric flavor for thei r streets which arc very often regu larly spaced and oriented so that we recognize shapes of Euclidean geometry like triangles, recta ngles, or pentagons, wh ile in the larrer, i.c., nwdieval urban fabrics grown gradual ly and "sponraneously" rhrough history without an y cemral control, we hardly recognize any­thing bur allless (fi!,;url' I, image 1). Hut is that rea ll Y:l mess, chaos, or is it jusr thar we arc not ca pable of seeing an order of a different ki nd ? Orher biologica l, social, technological, economic, or cultu ral self-organized systems-virtually a ll other systems that hav(.· passed through an evolutionary process and, as such, are nor the product of one Single .lgent­have been fou nd to ex hibit some astOnishingly similar prop­erties . The basics of suc h a shared property is a largely heterogeneous, -scale-free" distribution of cenrrality among the nodes. YOLL have very few nodes thar take high centrality and a lot of nodes that t:lkc quite a bit. We have shown that the sa me ru le actually applies to organic, historic:l1. sel f­organized urban patterns while it does not to planned ones. So, even if we do not sec it ar first glance, historical patterns do bave all order-though nor one o f a Euclidean kind: they

FI G. I One ~qua~ mile of cent",1 Ahmedabad (India). I . the primal graph: 2. local closeness (C::.",; d=400mt): 3. global closeness (C::ace). 4.1:10001 betweenness (C'ace). S. Slobal stnightness (G'ace); 6.global infof1Tl3tion (Caat).

1-42 SERG IO PORTA AND VI TO LATORA

BO L OGNA

Correlation between centrality and community commerce/service locations. From left to right: variables of Central ity CommercelService location dynamics;

I: extension or the bandwidth considered in KOE: 2: ratio or cells with both Central ity (C) and Dynamic (0)=0: 3: both C and DooO: 4: C,«) and 0 =0:

5: c=o and O,«): 6: Pearson index of linear correlation.

",,,,, CorrelJledV,mAb1e5

Central,ty """'" , c_ eo-m.s.r,

2 c_ e-

J C_ Comm+5efv

, C_ eo-m

5 ""- eo.,..,

, CQ Comm l 5e<v

7 C_ eo-m.s.r,

B "-0 Cornm ~S''IV

, c;. <-'SoN

" C_ ",,"",. SoN

" """ Co""""'Serv

" C_ ""'"" " <!- Comm

" C_ Corrm"'s..rv

" <!- """'" TA BLE I . Top 15 ~~rson correlations between kernel densities or street centr.llities Md commerdallservice ~ctivities in 8oIogna

actually exhibit the same order of most other organic systcms in nature, an order that ensures cxtraordinary performances to such syStems in terms of adaptabil ity to external threats, (:apability to modi fy and react, diffuse inform:uion effici ently, and strengthen local ties. Thc "small-world " propcrry, which emergcs in self-organized systems as the capaciry to achieve good performances both at the local and the global level, ;11so

"'" """-"" \..inear Com!Ia1K>l1

""= ~-

lOO om

lOO 0.701

200 0.673

200 'MJ

JOO 0.641

lOO 0.620

JOO OhlS

JOO D.«J8

200 ,saJ

'00 0567

300 0565

'00 om

200 o.5~ 7

200 0;<'

JOO om

is detectable a t the highesr grrtde in medieval urban patterns. Here, no il1rtHCr if Arabic or Europcrt~, rich or poor, hor or cold, Islamic or Ch ristian, rhe urban structure cnsures a cost compamblc to that of a tree-like Structure bur with rt ll effi­ciency comparable TO thar of;1 completed, interConnecTed ner­work. The S;1me "'small world" behavior, rhough ;11 a much lower grade, is present in gridiron rtnd mixed fabri cs; COIl-

S.3 Centrality ~nd Cities 1~3

f iG. 1 Bologna (northern Italy): Global Betweenness centrality on the primal gr.lph of the road system. one outcome of the MeA ~pplication ( I); Kernel Density Evaluation (b~ndwidth " 200 mt) of GI0b31 Between­ness (2); community commerce ~nd service Icutlons (3); Kernel Density Evaluation (bandwidth '" 200 mt) of commen:e ~nd service 'OClItions (2). The linear correlation between density v~l ues of panels 2 and 4 reaches the highest :sec.-e (Peuson"'O.67J) among those mentioned in the table.

versciy, modernist or suburban lollipop fabrics do not exh ibit such properties as they minimize the cost of the network generating on the other side, a fai lure in the dficiency of the system as a whole. \

Centrality and tke location of Commun ity Fleta ll and Services

Cities have grown in history following (to a cert:lin exten t) rules of centralities that Me facturs of efficiency in making complex urban systems competitive ill evolution. The idea is

144 SERGIO PO'RTA AND VITO LATORA

that we Gill actually sec some clues of this in how several dynamics of particular releva nce corrcl:uc on the ground to the spatial distribution of centrality. We have worked om a first test over thc city of Bologna, northcrn Italy. Here, we correlate the celltrality indices of closeness, betweenness, and straightness, computed at the global and local1cvc1s, to

(/YI/alllies inherenr to dail y urban life. i.e. , the location of community-level commcrcial activities and services.

Rather than comparing cClltr:l1ity alld cOlllmercclservice dynamics on a strcct-by-streer basis,4 w~ have explored in Hologna a different methodology based on the Kernel Den­sity Eva luation (KDE), a density funcrion based on Euclidean

distance widely used in gcogmphic information science and spa tial an ;llysi:i.sThe resu lting KDC methodology presems a twofold adva lHage: o n o ne hand , it ni cely ca ptures th e decreasi ng influence that centrality shows in the urban space as the distance increases from a given location {not only shops facing a Street take advantage of that street's central­ity, bm also thoSt' located nea rby on the parallel sn eets, though at a lower level}; on the o ther hand, it overcomes the problem of ach ievi ng fine cross-referenced dara needed for a streer-by-stree·t correlation analysis.

In this study, a fter a prelimi na ry MeA step, we -covered " the Bologna urban region with a grid of 2,771,956 square cells (edge: 10m ). We then calculated rhe Kernel Density of each cell of tht: study region in terms of all centra lity indices (de rived by M CA) and, on the other side, all commerce and service locations. Then we produced a table where, for each cell (reco rd ), Kernel density va lues of both kinds were asso­ciated. In rhe ta ble we report some structural data of cell val ­ues and the resu lting PCMson index of linear correlation for e<lch couple of considered vari<lbles. Such index expresses how much lWO quamities are co rrelated by gi ving a number between - '1 (most negative correl<ltion ) ;lI1d I (most positive); however, rhe v" lue of til{' Pearson correlation decreas.cs as the dataset size increases due to sta tistical fl uctuations.' We chose ro limit t he dataset to cells con taining at least one non­zero va lue. Re sulting Pearson va lues :Ire positive for all cou­ples of variabl es (sec Table I ) varyi ng from 0.35 1 to 0.727; the index reaches the highest scon.'--t'xtrcmely significam for a dataset of such magnitude-for global betweenness ccntral-

ity. Such find ings suggest that centraliry, espec ia lly global hetwecnness, arc driving forces as crucial factors in the evo­lution of city lifc as the location ofcommuniry rerail and serv­ices, a correlation that also emerges through the simple visual comparison o f spa ria) maps (figure 2).

Conclus ion: C entrality and City DeSign

How ca n we understand where to locate shops and retail services in a proposed development? Where can we place the main ha ll in a company headqua rter? What impact can we expect from the real izing a new bridge or road on the rest of a city system? How can we berrer integrate a neighborhood (l ik e a social housi ng estate) in its immediate and globa l sur­ro undings? These ;Ire a ll crucial qucstions in susmiTlable urban design processes and cemmliry, it rurns out, is a driv­ing force in all ofrhis. M eA, as a tool fo r man<lgingcentra l­ity in real spatial systems at .111 scales, can help decision makers dirccrly ,111d efficiently with scientific-based assistance in processin g architectural and urban design .? T he sa me approach is also being experimemed with in more specific areas of transpon planning and management: how ca n we improve the performance of a public-transport system by adding just a certain amount of new lines or srops ? Is a pro­posed transport-syStem extension plan actually going ro give the right answers to one community'S needs? The applicarion of M CA to "classic" transport planning issues is one of the main directions of development for Ollr current research.'

5.3 Cen tra lity and Cities J"5