john bambrick advisor: frank hardisty
DESCRIPTION
Capstone Proposal: Development of a Crowd Flow Model Coupled with Concentration Fields for Consequence Assessments. John Bambrick Advisor: Frank Hardisty. Outline. What is Consequence Assessments Models for Coupling Proposed Model Development Case Study Journal Article. - PowerPoint PPT PresentationTRANSCRIPT
![Page 1: John Bambrick Advisor: Frank Hardisty](https://reader036.vdocuments.site/reader036/viewer/2022081502/568164d3550346895dd70605/html5/thumbnails/1.jpg)
1
Capstone Proposal:Development of a Crowd Flow
Model Coupled with Concentration Fields for
Consequence AssessmentsJohn BambrickAdvisor: Frank Hardisty
![Page 2: John Bambrick Advisor: Frank Hardisty](https://reader036.vdocuments.site/reader036/viewer/2022081502/568164d3550346895dd70605/html5/thumbnails/2.jpg)
2
Outline
• What is Consequence Assessments• Models for Coupling • Proposed Model Development• Case Study• Journal Article
![Page 3: John Bambrick Advisor: Frank Hardisty](https://reader036.vdocuments.site/reader036/viewer/2022081502/568164d3550346895dd70605/html5/thumbnails/3.jpg)
3
WHAT IS CONSEQUENCE ASSESSMENTS (CA)?
![Page 4: John Bambrick Advisor: Frank Hardisty](https://reader036.vdocuments.site/reader036/viewer/2022081502/568164d3550346895dd70605/html5/thumbnails/4.jpg)
4
Analysis of Risk/Threat from Hazardous Atmospheric Releases
2002 Accidental Release of Chlorine Gas from a Railcar in Festus, MO. http://www.aristatek.com/Newsletter/OCT07/OCT07ts.aspx
![Page 5: John Bambrick Advisor: Frank Hardisty](https://reader036.vdocuments.site/reader036/viewer/2022081502/568164d3550346895dd70605/html5/thumbnails/5.jpg)
5
MODELS FOR COUPLING
![Page 6: John Bambrick Advisor: Frank Hardisty](https://reader036.vdocuments.site/reader036/viewer/2022081502/568164d3550346895dd70605/html5/thumbnails/6.jpg)
6
CA Models
HPAC • Is commonly used by civil
government and military• Couples hazard plume with
static population dataset for casualty estimate
• Has potential for over prediction of casualties
HPACCATS VLSTRACK
ALOHAQUIC
EpiSimS
![Page 7: John Bambrick Advisor: Frank Hardisty](https://reader036.vdocuments.site/reader036/viewer/2022081502/568164d3550346895dd70605/html5/thumbnails/7.jpg)
7
Crowd Flow Models (2 Categories)
Microscopic• Simulates individuals• Discrete position• Computationally “heavy” • Applied within EpiSimS
Macroscopic • Models density flow• 1D and 2D cellular• Computationally “light” • Applied within ???
Simulated Individuals Moving Around an Obstaclehttp://www.cs.virginia.edu/~gfx/pubs/group_behaviors_2/jar_herd.pdf
Cellar Flow of Densityhttp://scholar.lib.vt.edu/theses/available/etd-12152006-143454/unrestricted/NewCM.pdf
![Page 8: John Bambrick Advisor: Frank Hardisty](https://reader036.vdocuments.site/reader036/viewer/2022081502/568164d3550346895dd70605/html5/thumbnails/8.jpg)
8
PROPOSED MODEL DEVELOPMENT
![Page 9: John Bambrick Advisor: Frank Hardisty](https://reader036.vdocuments.site/reader036/viewer/2022081502/568164d3550346895dd70605/html5/thumbnails/9.jpg)
9
Proposal
• Develop a prototype macroscopic crowd flow model coupled with CA outputs to predict casualty estimates
• Develop a 2D pedestrian flow model• Develop for specific scenario• Build in ArcGIS 9.3 with VBA• Couple HPAC outputs with 2D flow model• Document in journal submission
![Page 10: John Bambrick Advisor: Frank Hardisty](https://reader036.vdocuments.site/reader036/viewer/2022081502/568164d3550346895dd70605/html5/thumbnails/10.jpg)
10
Concepts Behind Model
http://www.timtim.com/drawing/view/drawing_id/398
http://www.theipinionsjournal.com/index.php/2006/07/its-official-second-hand-smoke-kills/
Running with the Bulls http://www.flickr.com/photos/mcmahon/4788415049/
![Page 11: John Bambrick Advisor: Frank Hardisty](https://reader036.vdocuments.site/reader036/viewer/2022081502/568164d3550346895dd70605/html5/thumbnails/11.jpg)
11
Key Components
• Concentration Fields – From HPAC
• Environment– Crowd– Obstacles– Terrain– Egresses
• Crowd Flow Model
![Page 12: John Bambrick Advisor: Frank Hardisty](https://reader036.vdocuments.site/reader036/viewer/2022081502/568164d3550346895dd70605/html5/thumbnails/12.jpg)
12
Key Functions• Dosage calculation– Dosage = ∑ Concentration × Time Step
• Crowd flow model– Primary flow driver
• Recognize hazard• Determine directional flow based on hazard
– Unique for differing hazard types
– Secondary flow drivers• Model crowd density• Find path of least resistance• Find egress for escape from hazardous environment
– Selection between primary and secondary flow factors
![Page 13: John Bambrick Advisor: Frank Hardisty](https://reader036.vdocuments.site/reader036/viewer/2022081502/568164d3550346895dd70605/html5/thumbnails/13.jpg)
13
Moved out Moved in Stayed in
Quasi-Macroscopic
True Macroscopic
• Model density flow• Know density but not who went
where• No way to know who got what
exposure
Quasi-Macroscopic
• Model “individual” in flow• Track who went where• Track exposure of each
“individual”
ρ=7 ρ=4ρ=2
ρ=9 ρ=5ρ=5
ρ=3 ρ=2 ρ=3
ρ=3 ρ=5 ρ=6
ρ=4 ρ=6 ρ=4
ρ=3 ρ=2 ρ=2
Concentrations
T1 T2 T1 T2
![Page 14: John Bambrick Advisor: Frank Hardisty](https://reader036.vdocuments.site/reader036/viewer/2022081502/568164d3550346895dd70605/html5/thumbnails/14.jpg)
14
Primary vs. Secondary Flow Direction
Flow
Rat
e
Fast
Slow
Prim
ary
Prim
ary
Secondary
Prim
ary
Secondary
Hazard
Primary Only Secondary OnlySplit Primary and
Secondary
T1
T2
![Page 15: John Bambrick Advisor: Frank Hardisty](https://reader036.vdocuments.site/reader036/viewer/2022081502/568164d3550346895dd70605/html5/thumbnails/15.jpg)
15
Data• Population shapefile
– Point geometry– Attributes
• Characteristics• Status• Variables
• Concentration fields & aspect– Source: HPAC– Point geometry interpolated to raster
• Environmental rasters– Density & aspect, slope– Distance from obstacles & aspect of slope– Terrain
![Page 16: John Bambrick Advisor: Frank Hardisty](https://reader036.vdocuments.site/reader036/viewer/2022081502/568164d3550346895dd70605/html5/thumbnails/16.jpg)
Crowd Flow Model
Flow Direction
Population Data
Population Density
Primary Flow:Hazard Induced Flow Direction
Secondary Flow:Path of Least
Resistance
Concentration Field / Flow Driver
Integrated Exposure
People Rate
Urgency: Hazard Induced
HPAC
User: Study Area Definitions
EnvironmentLocation,
tn-1, tn
Input
Output
Data Flow
16
![Page 17: John Bambrick Advisor: Frank Hardisty](https://reader036.vdocuments.site/reader036/viewer/2022081502/568164d3550346895dd70605/html5/thumbnails/17.jpg)
17
Variable Sources for Primary and Secondary Flow (Conceptual)
• Primary direction– Aspect (downward direction) of concentration
• Secondary direction– Aspect of density (weighted by slope)– Perpendicular to aspect of the slope of distance to the obstacle
(weighted by inverse of distance)• Rate
– Urgency• Function of concentration above detectable levels but limited to individual’s max
speed– Resistance
• Function of density• Friction factor classified by terrain
![Page 18: John Bambrick Advisor: Frank Hardisty](https://reader036.vdocuments.site/reader036/viewer/2022081502/568164d3550346895dd70605/html5/thumbnails/18.jpg)
Constants
Time Step Sequence
Pop X,Y tn-1
Population Density
AspectConc.Field
HPAC
AspectIDW
Kernel Density
SlopeSlope
AspectAspect
Pop X,Ytn
Conc.FieldIDW
Obstacles
Terrain
Output tn-1
Output tn
ReclassifyFrictionFactor
Distance Distance
Flow Direction Calculation
Rate Calculation
AspectAspect
TimeStep
Intake
Extract from Raster at Point
Internal Calculation
Geoprocess
Point Data
Raster
FlowVector
Tn-1 Tn
<<dependant on>>
18
![Page 19: John Bambrick Advisor: Frank Hardisty](https://reader036.vdocuments.site/reader036/viewer/2022081502/568164d3550346895dd70605/html5/thumbnails/19.jpg)
19
Development Priorities• Calculation of individual’s intakes from concentration fields at Tn
location (Required)• Create individual’s primary flow vector at Tn and new location
(Required)• Create secondary flow direction based on density and selection
algorithm (Required)• Create secondary flow direction based on obstacle avoidance and
selection algorithm (Optional - probable)• Apply terrain based friction factors to flow rate (Optional – possible)• Create secondary flow direction toward egresses and selection
algorithm (Optional - doubtful)
![Page 20: John Bambrick Advisor: Frank Hardisty](https://reader036.vdocuments.site/reader036/viewer/2022081502/568164d3550346895dd70605/html5/thumbnails/20.jpg)
20
CASE STUDY
![Page 21: John Bambrick Advisor: Frank Hardisty](https://reader036.vdocuments.site/reader036/viewer/2022081502/568164d3550346895dd70605/html5/thumbnails/21.jpg)
21
Case Study• Accident results in Cl2 released near pedestrian mall• Cl2
– Visible and odorous– Irritation and harmful effects
• Cl2 acts as primary flow driver (crowd responds)– Flow from high to low concentrations– Urgency determined by concentration level
![Page 22: John Bambrick Advisor: Frank Hardisty](https://reader036.vdocuments.site/reader036/viewer/2022081502/568164d3550346895dd70605/html5/thumbnails/22.jpg)
22
Expected Results
• Dosage developing over time• Casualty estimates provided• Results comparatively different from a static
population• Each run of model unique• Exploration of scale adaptability
![Page 23: John Bambrick Advisor: Frank Hardisty](https://reader036.vdocuments.site/reader036/viewer/2022081502/568164d3550346895dd70605/html5/thumbnails/23.jpg)
23
JOURNAL ARTICLE
![Page 24: John Bambrick Advisor: Frank Hardisty](https://reader036.vdocuments.site/reader036/viewer/2022081502/568164d3550346895dd70605/html5/thumbnails/24.jpg)
24
Article Submission• Purpose: To propose and promote the use of dynamic
population flow models in consequence assessments• Contents
– Present concepts of model– Demonstrate the working conceptual model– Suggest uses– Identify areas for future development
• Candidate journals– Risk Management– Journal of Homeland Security– Journal of Homeland Security and Emergency Management
![Page 25: John Bambrick Advisor: Frank Hardisty](https://reader036.vdocuments.site/reader036/viewer/2022081502/568164d3550346895dd70605/html5/thumbnails/25.jpg)
25
QUESTIONS?