Spatial Microsimulation for City Modelling, Social Forecasting and Urban Policy Analysis. Presentation
Post on 19-Nov-2014
1230 Views
Preview:
DESCRIPTION
Transcript
School of GeographyFACULTY OF ENVIRONMENT
Spatial Microsimulation for City Modelling, Social Forecasting and
Urban Policy Analysis
Mark Birkin 6649386
Example: Urban Simulation
MoSeS Project
• Can we project the population of a city forwards in time over a 25 year period?
• technically & intellectually demanding
• policy relevant
• housing, transport, health care, education, …
• Three components• Population reconstruction
• Dynamic simulation
• Activity and behaviour modelling
Health and social care...
2001
20312016
2006
Health and Social Care…
2001Co-dependency
2031
LLTI20312001
Health and Social Care…2001
Ethnicity
2031
MultipleDeprivation
20312001
Moses Dynamic Model
Ageing/mortality
Fertility
Inmigration
Emigration
Household formation
Marital status
Health status
Local migration
Transition rates for fertility, mortality and migration are spatially disaggregated
E.g. fertility: rates by age, marital status and locationEvent is simulated as a Monte Carlo process
Example: married woman, aged 28, living in AireboroughProbability of maternity is 0.127Pull a probability from a distribution of random numbers; if <= 0.127 then the event occursAll events in discrete intervals of one year
MoSeS Data Sources
Census Small Area Statistics
Household and Individual SARS
ONS Vital Statistics
Special Migration Statistics
International Passenger Statistics
BHPS
Health Survey for England
National Travel Survey
General Household Survey
Hospital Episode Statistics
EASEL Housing Needs Study
Google Maps
Moses Dynamic Model
Simulation Year 1
Gender FAge 45
Location AMarital status W
Household status LHealth G
Gender FAge 18
Location AMarital status S
Household status LHealth G
Moses Dynamic Model
Simulation Year 1 2
Gender F FAge 45 46
Location A AMarital status W W
Household status L SHealth G G
Gender F FAge 18 19
Location A BMarital status S S
Household status L SHealth G G
Moses Dynamic Model
Simulation Year 1 2 3
Gender F F FAge 45 46 47
Location A A AMarital status W W W
Household status L S SHealth G G G
Gender F F FAge 18 19 20
Location A B BMarital status S S S
Household status L S SHealth G G G
Moses Dynamic Model
Simulation Year 1 2 3 4 5
Gender F F F F FAge 45 46 47 48 49
Location A A A A AMarital status W W W W W
Household status L S S S SHealth G G G G G
Gender F F F F FAge 18 19 20 21 22
Location A B B B CMarital status S S S S C
Household status L S S S CHealth G G G G G
Gender MAge 24
Location CMarital status C
Household status CHealth G
Moses Dynamic Model
Simulation Year 1 2 3 4 5 6 7
Gender F F F F F F FAge 45 46 47 48 49 50 51
Location A A A A A A AMarital status W W W W W W W
Household status L S S S S S SHealth G G G G G G G
Gender F F F F F F FAge 18 19 20 21 22 23 24
Location A B B B C C CMarital status S S S S C C C
Household status L S S S C F FHealth G G G G G G G
Gender M M MAge 24 25 26
Location C C CMarital status C C C
Household status C C FHealth G G G
Gender MAge 0
Location CMarital status S
Household status FHealth G
Moses Dynamic Model
Simulation Year 1 2 3 4 5 6 7 8 9 10
Gender F F F F F F F F F FAge 45 46 47 48 49 50 51 52 53 54
Location A A A A A A A A A AMarital status W W W W W W W W W W
Household status L S S S S S S S S SHealth G G G G G G G G G G
Gender F F F F F F F F F FAge 18 19 20 21 22 23 24 25 26 27
Location A B B B C C C C C CMarital status S S S S C C C C M M
Household status L S S S C F F F F FHealth G G G G G G G G G G
Gender M M M M M MAge 24 25 26 27 28 29
Location C C C C C CMarital status C C C C M M
Household status C C F F F FHealth G G G G G G
Gender M M M MAge 0 1 2 3
Location C C C CMarital status S S S S
Household status F F F FHealth G G G G
Gender FAge 0
Location CMarital status S
Household status FHealth G
Moses Dynamic Model
Simulation Year 1 2 3 4 5 6 7 8 9 10 11 12 13
Gender F F F F F F F F F F F F FAge 45 46 47 48 49 50 51 52 53 54 55 56 57
Location A A A A A A A A A A A A AMarital status W W W W W W W W W W W W W
Household status L S S S S S S S S S S S SHealth G G G G G G G G G G G G G
Gender F F F F F F F F F F F F FAge 18 19 20 21 22 23 24 25 26 27 28 29 30
Location A B B B C C C C C C C C DMarital status S S S S C C C C M M M M M
Household status L S S S C F F F F F F F FHealth G G G G G G G G G G G G G
Gender M M M M M M M M MAge 24 25 26 27 28 29 30 31 32
Location C C C C C C C C DMarital status C C C C M M M M M
Household status C C F F F F F F FHealth G G G G G G G G G
Gender M M M M M M MAge 0 1 2 3 4 5 6
Location C C C C C C DMarital status S S S S S S S
Household status F F F F F F FHealth G G G G G G G
Gender F F F FAge 0 1 2 3
Location C C C DMarital status S S S S
Household status F F F FHealth G G G G
Moses Dynamic Model
Simulation Year 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Gender F F F F F F F F F F F F F F F FAge 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Location A A A A A A A A A A A A A A A AMarital status W W W W W W W W W W W W W W W W
Household status L S S S S S S S S S S S S S S SHealth G G G G G G G G G G G G G G G G
Gender F F F F F F F F F F F F F F F FAge 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
Location A B B B C C C C C C C C D D D DMarital status S S S S C C C C M M M M M M M M
Household status L S S S C F F F F F F F F F F FHealth G G G G G G G G G G G G G G G G
Gender M M M M M M M M M M M MAge 24 25 26 27 28 29 30 31 32 33 34 35
Location C C C C C C C C D D D DMarital status C C C C M M M M M M M M
Household status C C F F F F F F F F F FHealth G G G G G G G G G G G P
Gender M M M M M M M M M MAge 0 1 2 3 4 5 6 7 8 9
Location C C C C C C D D D DMarital status S S S S S S S S S S
Household status F F F F F F F F F FHealth G G G G G G G G G G
Gender F F F F F F FAge 0 1 2 3 4 5 6
Location C C C D D D DMarital status S S S S S S S
Household status F F F F F F FHealth G G G G G G G
MoSeS Dynamic Model
Simulation Year 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
Gender F F F F F F F F F F F F F F F F F F F F F F F F F F F F FAge 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73
Location A A A A A A A A A A A A A A A A A A A A A A A A A A A A AMarital status W W W W W W W W W W W W W W W W W W W W W W W W W W W W W
Household status L S S S S S S S S S S S S S S S S S S S S S S S S S S S SHealth G G G G G G G G G G G G G G G G G G G G G G G G G G G G G
Gender F F F F F F F F F F F F F F F F F F F F F F F F F F F F FAge 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
Location A B B B C C C C C C C C D D D D D D D D D D D D D D D D DMarital status S S S S C C C C M M M M M M M M M M M M M M M M M M M M M
Household status L S S S C F F F F F F F F F F F F F F F F F F F F F F F FHealth G G G G G G G G G G G G G G G G G G G G G G G G G G G G G
Gender M M M M M M M M M M M M M M M M M M M M M M M M MAge 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
Location C C C C C C C C D D D D D D D D D D D D D D D D DMarital status C C C C M M M M M M M M M M M M M M M M M M M M M
Household status C C F F F F F F F F F F F F F F F F F F F F F F FHealth G G G G G G G G G G G P P M M G G G G G G G G G G
Gender M M M M M M M M M M M M M M M M M M M M M M MAge 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
Location C C C C C C D D D D D D D D D D D D D E E E EMarital status S S S S S S S S S S S S S S S S S S S S S S S
Household status F F F F F F F F F F F F F F F F F F F F S S SHealth G G G G G G G G G G G G G G G G G G G G G G G
Gender F F F F F F F F F F F F F F F F F F F FAge 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
Location C C C D D D D D D D D D D D D D D D D DMarital status S S S S S S S S S S S S S S S S S S S S
Household status F F F F F F F F F F F F F F F F F F F FHealth G G G G G G G G G G G G G G G G G G G G
0
200,000
400,000
600,000
800,000
1,000,000
2001 2005 2010 2015 2020 2025 2031
Pop
ulat
ion (
pers
on)
0
10
20
30
40
50
60
Ave
rage
spe
ed (
km/h
)
Population Average speed
Population and average speed changes in Leeds from 2001 to 2031
Transport…
2001 2031
2015
* Traffic Intensity=Traffic load/Road capacity
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Traffic Intensity *
Transport…
Scenario-based forecasting
Public Policy
Source: MAPS2030
Simulation of Epidemics
Ferguson et al, Nature, 2006
The El Farol Bar Problem
Everyone wants to go the bar
- unless it’s too crowded!
Must relax neoclassical economic assumptions (homogeneity of preferences, simultaneous decision-making)
Individual actors/ agent-based decision-making
- generic template for real markets
heterogeneous
out of equilibrium
(Arthur, 1994)
NeISS Architecture
NeISS Portal
NeISS Portal
Census Data Population
ReconstructionModel
Crowdsourced data
PersistentRepository
Policy
InfrastructureData
SpatialData
SurveyData
DynamicModel
Visualisation
Social Simulation
Models
Data FusionTool
Research Training Public
Planners
MapTube
NGSSyntheticData
Key IndicatorsReports
Provenance
What ifassumptions
EnhancedData
SyntheticData
SyntheticData
Synthetic Data Sets Provenance
Synthetic Data Sets Provenance
Data Issues and Questions
• Complexity
• Visualisation
• Integration
• Proliferation
• Generation
Complexity of data
Complexity, scale and volume of data inputs
Legend
Silverburn
Observed / Predicted0.00 - 0.50
0.51 - 1.00
1.01 - 1.50
1.51 - 2.00
2.01 - 3.40
µ
0 30 6015
Kilometers
Centre 1961 1971a 1971b Capture
Liverpool 54 77 69 10.4%Manchester 51 70 63 10.0%Haydock 0 0 47Warrington 6 11 7 36.4%Wigan 7 11 6 45.5%
Data integration
Modelling and simulation as data integration
• “Data diarrhoea, information constipation”• → data compression
• → missing data
Legend
Silverburn
Observed / Predicted
0.00 - 0.50
0.51 - 1.00
1.01 - 1.50
1.51 - 2.00
2.01 - 3.40
µ
0 30 6015
Kilometers
Proliferation of data domains
• “customer science”• public/ private/ commercial
• Crowd-sourced data
Data Generation
Example 1. (Silverburn)
• 400 post sectors
• 100 destinations
• 6 ages
• 4 ethnic groups
• 4 social/ income groups
• 2 car ownership• 516 inputs; 8 million model
flows (sparse matrix!)
Example 2. (MoSeS)
• 25 years of simulation
• 60 million individuals
• 200? characteristics
• 20? scenarios
Example 3. (Epstein, 2009)
• 8 billion agents!
• Dynamic resolution at 10 minute intervals?!!
Example 1. (Silverburn)
• 400 post sectors
• 100 destinations
• 6 ages
• 4 ethnic groups
• 4 social/ income groups
• 2 car ownership• 516 inputs; 8 million model
flows (sparse matrix!)
Example 2. (MoSeS)
• 25 years of simulation
• 60 million individuals
• 200? characteristics
• 20? scenarios
Example 3. (Epstein, 2009)
• 8 billion agents!
• Dynamic resolution at 10 minute intervals?!!
Conclusion
Social simulation involves quite a lot of data intensive research!!
Note that quite a lot of social scientists have so far failed to appreciate this important fact!!!
top related