1. Types of Models Marti Blad PhD PE
Handouts/supplemental stuff for workshop: Air pollution Dispersion Models; List of different models from epa’s scram site
Marti Blad PhD PE

2. EPA Definitions
Dispersion Models: Estimate pollutants at ground level receptors
Photochemical Models: Estimate regional air quality, predicts chemical reactions
Receptor Models: Estimate contribution of multiple sources to receptor location based on multiple measurements at receptor
Screening Models: applied 1st , determines if further modeling needed
Refined Models: req’d for SIP, NSR, and PSD
Regulatory requirement for permits

3. Models = Representations or pictures
Numerical algorithms
Sets of equations need inputs
Describe = quantify movement
Simplified representation of complex system
Box or Mass Balance
Used to study & understand the complex
Physical, chemical, and spatial, interactions

4. Types of Models Gaussian Plume Statistical & Stochastic Empirical
Analytical approximation of dispersion
more later
Statistical & Stochastic
Based on probability
Recall regression is linear model
Empirical
Based on experimental or field data
Actual numbers
Physical (scale models)
Flow visualization in wind tunnels, etc.
Understanding which model will help to understand its limitations. What is the model based on.

5. Recall bell shaped curve
Plume dispersion in lateral & horizontal planes characterized by a Gaussian distribution
Normal Distribution
Mu is median
Sigma is spread

6. Gaussian-Based Dispersion Models
Pollutant concentrations are calculated estimations at receptor
Uncertainty of input data values
Data quality, completeness
Steady state assumption
No change in source emissions over time
Screen3 will be end of the week

7. Gaussian Dispersion C(x,y,z) Downwind at (x,y,z) ? z ¤ Dh = plume rise
h = stack height Dh
H = effective stack
height
H = h + Dh H h x
C(x,y,z) Downwind at (x,y,z) ? y

8. Air Pollution Dispersion (cont.)
This assumption allows us to calculate concentrations downwind of source using this equation
where
     c(x,y,z) = contaminant concentration at the specified coordinate [ML-3],       x = downwind distance [L],       y = crosswind distance [L],       z = vertical distance above ground [L],       Q = contaminant emission rate [MT-1],       sy = lateral dispersion coefficient function [L],       sz = vertical dispersion coefficient function [L],       u = wind velocity in downwind direction [L T-1],       H = effective stack height [L]. 

9. Gaussian model picture
Predicted concentration map

10. The Gaussian Plume Model
The shape of the curve = Bell shaped = Gaussian curve hence the model is called by that name.
Now, we come to why did I have to torture you with bell shaped curves and Z numbers, probability and predictions etc. Remember the computer will do calculations but think about how to mathematically describe the movement of pollutant molecules, dancing,

11. Ways to think about math
Gaussian = “normal” curve math
Recall previous distribution picture
Dispersion & diffusion dominates
Eulerian
Assumes uniform concentrations in box
Assumes rapid vertical and horizontal mixing
Plume in a grid
Predicts species concentrations
Multi day scenarios

12. Eulerian Air Quality Models
Grid type models
Model simulates the species concentrations in an array of fixed computational cells (Zion)
AKA Plume in Grid
Figure from

13. Box idea: 1-D and 2-D Models

14. Dimensional Concept Variable is Time: t
Variable is Time and height: t, y
Variable is Time, height and length distance:
t, x, y
t, x, y, z

15. 3-Dimensional Models
Depth of boxes discussed under meteorology

16. Other choice: Lagrangian
“Puffs” of pollutants
Trajectory models
Follow the particle
Puff W2 W1
S.S. Plume

17. Lagrangian Air Quality Models
From “INTERNATIONAL AIR QUALITY ADVISORY BOARD PRIORITIES REPORT, the HYSPLIT Model” (

18. Assumptions & limitations
Physical conditions: Topography
Locations: buildings, source, community, receptor
Appropriate for the averaging time period
Statistics & math
Meteorology
Stack or source emission data
Pollutant emission data
Plume rise, Stack or source specific data
Location of source and receptors

19. EPA MODELS—Screening

20. EPA MODELS—Regulatory
CALPUFF
AERMOD

21. EPA Models—Other