Challenges in Research and Paul Waddell

Professor and Chair

Practice in Land Use and Transport Department of City and Regional Planning
Aalto University, August 12, 2013 UC Berkeley
 ")
•   

   
– )/"-/
)
.
)
-).+*-/0*)
'))$)"
– )/"-/
)
.
)
-).+*-/0*)
*'$)"
• -)$(
.

.
/37

H
 )/.
*-
)/"-/
'))$)"
• <<
"$.'0*);
:
?HG:

• '$*-)$
"$.'0*);

IH:

IMK
–-"/
-.0
-30*)
$)
 
($..$*).:

$(+'$/.
3-)
.+-5'
–,3$-
3.
*
4)
/-4'
(*'.
)
')

3.
(*'.
• *'$7
A'"$.'0*)B
#
*
+-0<<<

I
#7
*))0)"
)
.
)
-).+*-/
$.
..)0'
A4)
$
$/
5-
)*/
'"''7
-,3$-B

Highway -$0*)'
-*%/
4'30*)
Project

Travel Speed

Air Quality
#7
*))0)"
)
.
)
-).+*-/
$.
..)0'
A4)
$
$/
5-
)*/
'"''7
-,3$-B

-$0*)'
-*%/
4'30*)
$.
*+0($.0''7
$.
*3/
+-*%/
Highway $(+/.
.$)
$/
6'3.
$)3

Project
()
/.
+

+ New Real Estate

Travel Speed
Development

+
-
+
Household
+
Traffic Location
? On New
+ Firm
Facility Location

Air Quality
 #'')".
/*
)/"-/
'))$)"
• *) $0)"
).0/30*).
• *) $0)"
'3.
• *) $0)"
+$./(*'*"$.
• *) $0)"
*'$$.
• *)/)0*3.
*'
*'$0.

L
#'')".
*
)/"-/
'))$)"

M
 )/.
*-
)/"-/
*'$)"
• 
-*3./
(/#**'*"7
*-
+*'$7
#*$.
– -"?.'
/-).+*-/
$)4./()/.
– *'
)
-"$*)'
')
3.
+*'$$.
• 
/*
*)3/
6?)/
4'30*)
–
)7
*($)0*).
*
+*'$$.
–
(+*..$'
/*
)'78
5$/#*3/
/#
$
*
(*'.
• 
/*

-.+*).$4
/*
+*'$0'
*)/6/
– *040)"
+*'$7
,3.0*).
–*'$0'
)
.'
*)./-$)/.
N
)/.
*-
)/"-/
*'$)"

O
#'')".
/*
)/"-/
*'$)"
Transparency

Behavioral Computational
Validity P f
Performance

Flexibility
Empirical APPROPRIATION
V ld
Validity

Data Availability
Ease of Use anddQQuality
l

Uncertainty

GF
 ")
• )/.
)
#'')".
– )/"-/
)
.
)
-).+*-/0*)
'))$)"
– )/"-/
)
.
)
-).+*-/0*)
*'$)"
• 



 
–
4-4$5
*
-)$(
–
')
7
-
++'$0*)
–I
$.3'$80*)
*-
*((3)$/7
)""()/

GG
UrbanSim Background

• UrbanSim is a model system to support land use, transportation and

environmental planning
• Licensed as an Open Source software system since 1998, and freely
downloadable from the project website (www.urbansim.org)
• Supported professionally by Synthicity (www.synthicity.com)
• Funded mainly by NSF, with additional grants from EPA, FHWA, state and local
governments, and the European Research Council
• Recent surveys show that UrbanSim has become the most widely used land use
model system by planning agencies in U.S.
• Synthicity created as a startup in 2012 to provide professional support for
UrbanSim, and to create UrbanCanvas, a cloud-based, interactive, user-friendly
3D GIS platform
 -)$(
.
.$")

/*
-..
/#.
#'')".

Transparency

Behavioral Computational
Validity P f
Performance

Flexibility
Empirical APPROPRIATION
V ld
Validity

Data Availability
Ease of Use anddQQuality
l

Uncertainty

GI
Selected UrbanSim Applications (Completed or In Progress)

United States: International:

• Albuquerque, NM • Accra
• Denver, CO • Amsterdam
• Detroit, MI • Beijing
• Durham, NC • Brussels
• Eugene-Springfield, OR • Durban
• Honolulu, HI • Paris
• Houston, TX • Rome
• Phoenix, AZ • Seoul
• Salt Lake City, UT • Taipei
• San Antonio, TX • Tel Aviv
• San Francisco, CA • Turin
• Seattle, WA • Zürich
• Tucson, AZ • Downloaded by users in over 70 countries
 Evolution of Land Use Model Frameworks
Alonso/Mills/Muth:
Lowry: Leontieff: Input- Orcutt:
1960 Urban Economic
Gravity Model Output Model Microsimulation
Bid-Rent Theory

Spatial Interaction McFadden:

Spatial Input-Output
1970 DRAM/EMPAL Discrete-Choice
MEPLAN; TRANUS
HLFM II+ Models

Aggregate Equilibrium Aggregate Dynamic

1980 Discrete Choice Discrete Choice
METROSIM; MUSSA HUDS

Geographic
Information Systems
1990

Spatially Detailed
Microsimulation Dynamic
Rule-based Planning
2000 PECAS Discrete Choice
Tools
UrbanSim
Index; Places; WhatIf?

2010
 Paul Waddell, 2013

Detail in 2 Dimensions
Geography

Fully Aggregate: Microsimulated Space:

Employment by Sector, Employment by Sector,
Households by Income, Households by Income,
Aggregated Zones Buildings, Parcels
Agents

Microsimulated Agents: Microsimulation in 2D:

Persons, Households Persons, Households
Jobs, Businesses, Jobs, Businesses, UrbanSim
Aggregated Zones Buildings, Parcels
 Paul Waddell, 2013

Detail in 3 Dimensions

Geography
Static Equilibrium:
Time Not Represented
Time

UrbanSim
Dynamic Microsimulation:
Time Explicitly Represented,
n ts
Path Dependent e
Ag
 Paul Waddell, 2013

Model System Based on Parcels and Buildings

 Paul Waddell, 2013

Theoretical Basis of UrbanSim

• Random Utility Theory (McFadden)
- Theoretical basis for discrete choice models (such as mode choice): agents choose
among alternatives based on the relative utility of the available options
- Multinomial and nested logit models, more general forms relaxing the IIA property
• Urban Economics/Bid Rent Theory (Alonso, Mills, Muth)
- Explains land use outcomes: density, rents, as outcome of bidding process based
on willingness to pay for locational amenities: trade-off travel time vs housing cost
• Hedonic Price Theory (Rosen)
- Observed prices of composite goods like housing can be ‘decomposed’ to the
implicit prices of their attributes; widely used to measure the market valuation of
locational amenities or disamenities
 Paul Waddell, 2013

Theoretical Basis of UrbanSim

• Dynamic Market Equilibrium, Price Adjustment, Disequilibrium
- Markets are rarely in equilibrium; especially real estate due to supply lags and high
transactions costs; information is imperfect; speculation can lead to bubbles
- Current housing market crisis is an excellent example of market disequilibrium
• Microsimulation (Orcutt)
- Initially a computational framework to explore individual level impacts of policies;
later found more computationally efficient than complex aggregate models
- Strong empirical support for microsimulation to avoid aggregation bias and
ecological fallacy (attributing to individuals behavior from aggregate patterns)
• Geographic Information Systems and Spatial Analysis (Tobler, Anselin)
- Radical advance in spatial analysis capabilities;
- Representation of walking scale accessibility
UrbanSim Models Agents Interacting in Urban Markets
 Paul Waddell, 2013

UrbanSim Models Choices of Agents in Urban System

Macroeconomic Model

Households Governments Businesses

Demographic Processes Land Use Regulation Economic Processes

Aging Economic structure

Land use plans
Household structure Output goods/services
Growth management
Migration Inter-regional trade

Long-term Choices Transportation Long-term Choices

Residential Mobility Infrastructure
Housing Choice Pricing Mobility
Labor Supply Location Choice
Workplace Choice Labor Demand
Vehicle Ownership
Developers

Short-term Choices Real Estate Processes Short-term Choices

Activity Generation Land development

Activity Location Housing development
Goods movement
Mode Choice Non-res development
Route Choice Redevelopment

Travel Model
 Paul Waddell, 2011

Software Architecture for UrbanSim:

Modular Choice Models

1 2 3 4

Implementation of a choice model involves selecting options for each step in

the model process, and setting its configuration. In most cases, this can be
done in the GUI, without the need to edit program code. New models can be
created from templates, specified, and estimated interactively.
UrbanSim: A System of Interacting Models
Land
Development Real Estate Price Model
Models Residential Development Project
Location Choice Model
Nonresidential Development Project
Household Location Choice Model Employment
Location Location
Building Construction Model Models
Models

Household Transition Model Employment Transition Model

Household Relocation Model Employment Relocation Model

Household Location Choice Model Employment Location Choice Model

Economic Transition Model

Workplace Home-based Job Choice Model

Location
Models Workplace Location Choice Model

Job Change Model

 




• Transportation
– -).$/
$)4./()/.
A$':
3.B
– *57
$)4./()/.
A:

:
:
$&:
./-$)B
– -$$)"
A*''.:
*)".0*)B

• Land Use Regulations

– $/7
*(+-#).$4
').
– -).$/
-$)/

4'*+()/:
-)
$''".

C
)/-.
– 3.$$.:
(+/
.
– -)
-*5/#
*3)-$.
– -*/0*)
*

)4$-*)()/''7?.).$04

-.




• Land and Development
– *3.$)"
3)$/.
7
/7+:
).$/7:
+-$

A*-$'$/7B
– *)?-.$)0'
3$'$)".
7
/7+:

).$/7:
+-$
– -"
$)
"-$3'/3-'
'):
*-./:

*+)
.+
• Demographics;
#*3.#*'.
7

$)*(:
.$8:
'$
7'
• Economics;
(+'*7()/
7

./*-
)
3$'$)"
/7+
• Transportation
– ..$$'$/7:
*
#-.:
#$'

$'.
-4':
*)".0*)
'7

• Environment
– -)#*3.
.
($..$*).
– *''30*)
Plan Bay Area UrbanSim Application
 -$/.
/*

)

*-
.4-'
*
/#.
.'$.

• $-./
-"$*)'
+')
/*
$)/"-/
/-).+*-/0*):
')

3.:
)
#*3.$)"
  
  
 
• )$0/
7
'$*-)$
)/
$''
IMK
 5
//
5.;
IH:
IMK
3$)"
-)#*3.

.
($..$*).
"$*)'
-*5/#
/-/"7;
*3.
-*5/#

*)?3-)$8
')
-)$8
')
.
■
..
/#)
KR
*
-"$*)=.
')
■
-'7
NFR
*
)5
#*(.
■
4-
LFR
*
)5
%*.
')
 /.
#$!$)"
-$"(;

*$'$/7
/*
..$$'$/7
   
  


  
  


  
   
  


  



 

  


 
 



 )- .
)*)<0/*
(*

 )- .
"-*..
- "$*)'

.#-
 
0

+ -
+$/
$
+-*0/
   $)/$) /#

$)/$)
/#
  
   
   /-).+*-//$*)
.4./ (

 0
+ -<+$/

"- )#*0.
".
($..$*).
 0
+- (/0-
 /#.
!-*(

 

!-*(
-.
)
'$"#/<0/4
3+*.0-
/*
+-/$0'/
($..$*).

 
/-0&.
   0
$)%0-$ .
)
!/'$/$ .

!-*(
*''$.$*).
  
$- /
''
)*)< )- .
1 -""
$'4
/$(
4
.+
+ )/

 


"-$0'/0-'
    2'&$)"
*-
$&$)"
'&$ $&$


 
 1 '*+( )/
2$/#$)
/#

  
0-)
!**/+-$)/
 

 - .
#*0.$)"
)

 

/-).+*-//$*)
*./.
.

*0.
''
*!
/#
- "$*)=.


.#-
*!
'*2<$)*(

+-*% / 
#*0.$)"
"-*2/#

#*0. #*'
0" /.


    

 
 


• /-/.
5$/#
+*'$$.
)
+-*%/.
.

$)+3/.
$)/*
/#
(*'.
• 6($).
+*/)0'
*).,3).

*
+*'$$.>+-*%/.
*)
*3/*(.

.3#
.
')
3.
+2-).:
')
3.

($6:
).$/7:
)
/-4'
+2-).
• ''*5.
$)/-04
/.0)"
*
#*5

$-)/
+*'$7
./-/"$.
-
$)

#$4$)"
)
*3/*(
• .....
"-*5/#
$)3$)"
)

3(3'04
$(+/.
@
/5*
&7

-.
*-

IMK

./-('$)$)"

IJ
  
)'4/$'
(* '

3+'$$/
')
0.
 /$' 

*!
- ' 1)/



)/ "-/  +*'$$ . *0/*( .
 #1$*-.



)'4/$'
(* '

3+'$$/
/-).+*-/
 /$' 

*!
- ' 1)/

+-*% /.
8
+*'$$ . *0/*( .
 #1$*-.

-" ./
.
.#*0'
>0$'
!*-('
($-* *)*($
')
0.
(* '.9
.
.**)
.
$.
+-/$'9
.*
/#/
/# 4
)

0. 

/*
)'45
)
1'0/
/#
6 /.
*!
"-*2/#
. )-$*.
*)
*)*($
2 '!-
@0/$'$/4A9
$)'0$)"
')
+-$ .9
#*(

6*-$'$/49
%*.<#*0.$)"
7/9
/#
*($) 
#*0.$)"</-).+*-//$*)
*./
0- )9
)
*)*($
 1 '*+( )/

@2" .9
%*.9
3+*-/.A;?
*0- :
  
 
  


   
 *)./-0/$)"
)
 
'/ -)/$1






 
      


 
 

 
     



    
  

 




   
 
  

 
 

  
 




  



 
 

 
 



  


  

•  ,0$- 
4

• ..0( .
/#
*)/$)0/$*)
*!
'*''4<*+/ 
" ) -'
+').
 
  

 <  
 

<   





  






   




 

E
    
E  
 

     

• ++-*1 
.
/#
- ! -- 
 )-$*
!*-
')
4
- 
4

8

$)

4
DBCD


 !


UrbanCanvas 3D Visualization
 
#)$,3.
*-
I
$.3'$80*)
3--)/
   

--$-.
/*
3.$)"
I
4$.3'$80*)
• *./'7
A0(
)
(*)7B
• )$)/
A/
)
./3+B
• ) 6$'
A./0
4$5.B
• **
.'*5
*-
'$4
$)/-0*)
3-

 UrbanCanvas
• Prototype Funded by Metropolitan Transportation Commission, NSF
- MTC Plan Bay Area, a Sustainable Communities Strategies Planning Process
- Involves visioning, modeling, and visualization for community engagement
• Commercial Version Being Used by Albuquerque and Denver, with more
soon after initial release in 2013
• Summary:
- Develop an extension to UrbanSim to visualize simulation results in 3D
- Adds flexible indicator and accessibility framework
- Adds intuitive scenario editing/creation interface in 3D environment
- Shared metropolitan data infrastructure for coordination of planning and evaluation
- Dramatically increases productivity of creating input data and using UrbanSim
Data Sources
• Digital terrain models
• Digital orthophotos
• Street Network
- openstreetmap.org
• Parcels
• Building attributes
• Google buildings where
available and of good qualityy
• Zoning
• General Plans
• Planned Developments
• Business establishments
• Synthesized population
• Planning boundaries
Connecting Urban Geometry, Computing Accessibilities
UrbanCanvas: Interactive 3D Visualization
Editing in UrbanCanvas
UrbanCanvas: Interactive Design
'
.//

Summary and Status

• UrbanSim
- Mature platform for urban simulation gaining wide adoption for planning use
- New opportunities:
• Modeling energy use
• Modeling impacts of climate change
• Adapting for developing world

• Developed an integrated UrbanCanvas platform for

- Visioning: engaging communities in designing their future
- Modeling: analyzing alternative land use and transportation policy/design scenarios
- Evaluating: Computing Return on Investment (ROI) of development projects
- Visualizing: 3D representation of alternative scenarios, with indicators
 Thank you!

Questions?

waddell@berkeley.edu