Meteorology and air quality modelling

OBJECTIVES

1.    To understand the meteorological factors that influence atmospheric dispersion

2.    To understand key concepts like temperature inversions and atmospheric stability and how they influence dispersion

3     To develop an understanding of how meteorology is used in the modelling of point sources of air pollution

4.    To understand the types of output of air quality models

INTRODUCTION

It is not always possible or practical to monitor exposure to air pollutants.  Numerous models are available to determine concentrations of pollutants under varying circumstances, from simple  point sources (such as industrial stacks) and line sources (such as a line of traffic), to more complex modelling scenarios, such as airshed modelling where emissions from numerous sources are included and exposure over large areas estimated.  In order to understand how air pollution can be modelled, it is important to firstly have an understanding of the way in which the atmosphere behaves.

This workshop is structured into 4 parts:
Part 1 –  provides information on the types of factors that influence atmospheric dispersion.
Part 2 – specifically focuses on one of the atmospheric dispersion influences: atmospheric stability.
Part 3 – examines how ground level concentrations can change on the basis of changes in atmospheric stability using a simplified application of the Gaussian plume equation in an Excel spreadsheet.
Part 4 – uses a commercially available model AUSPLUME  to predict air pollution levels within a real context and plot ground level concentrations using example detailed meteorology data sets.

Part 1 – ATMOSPHERIC DISPERSION

Up to now we have considered only the types of air pollutants that can occur and how to monitor them.  Here we study the meteorological conditions which influence how air pollutants once emitted are dispersed.  In particular, we will isolate the meteorological factors which are important in transporting air pollutants. The factors affecting the microscale dispersion of air pollutants include diffusion, atmospheric stability (which is a combined effect of the turbulence in the horizontal and the vertical directions), plume rise, temperature inversions, topographical characteristics and other factors.

1.1 Diffusion
Everything has the natural tendency to diffuse from a high concentration region (e.g. close to the source of emission) to a lower concentration region (e.g. further away from the source). However, chemical diffusion is a relatively slow process and so is not the major mechanism for the dispersion of pollutants, except in indoor environment. In air pollution, ‘diffusion’ is usually used to refer to the dispersion of air pollutants by turbulence.

1.2 Horizontal turbulence
Wind blows mainly in the horizontal direction. Apart from carrying the air pollutants to the downwind direction, changes in wind direction also disperse the pollutants effectively in the horizontal direction.

1.3 Vertical turbulence
Solar radiation heats up the land surface and the air close to the ground and generates turbulence in the vertical direction (similar to the convection generated when boiling water in a kettle). The dispersion of pollutants in the vertical direction is the major mechanism of dispersion. The stronger the sunlight and the lighter the wind, the stronger the vertical turbulence and dispersion (why is turbulence stronger for lower wind speeds?). In urban areas, heat from fuel use will also enhance vertical turbulence. This phenomenon is called the heat island effect.

The combined effect of horizontal turbulence and vertical turbulence is expressed as atmospheric stability class. As shown in the wind tunnel experiment in Figure 1, stable, neutral and unstable atmospheric conditions may result in different shape of the pollutant plume, namely the fanning plume, coning plume and looping plume, respectively. (Why are the shapes of the plume called “fanning”, “coning” and “looping”?)

Figure 1 Atmospheric stability and dispersion in wind tunnel experiment. Top: ‘fanning’ plume under stable conditions. Middle: ‘coning’ plume under neutral conditions. Bottom: ‘looping’ plume under unstable conditions.
Source: Environmental Engineering Research Laboratories, New York University

In Brisbane for example, the more unstable conditions in summer enhance the dispersion of air pollutants in the vertical direction (Figure 2).

Figure 2- Seasonal Variations in Stabilities
Sourced from: A Study in Air Pollution from Brisbane Urban Traffic. Xu (1996)

1.4 Plume rise
The temperature and exit velocity of the flue gas may increase the buoyancy of the plume. Therefore the plume may rise to a height greater than the physical height of the stack before it is dispersed (Figure 4.21). (In general, is plume rise the higher the better or the lower the better?) Apart from flue gas temperature and velocity, plume rise can also be influenced by the stack diameter and the ambient temperature and stability. The total height of the stack and plume rise is called the effective stack height (Figure 3).

Figure 3 Plume rise

1.5 Temperature inversions
Normally temperature drops as altitude increases. However, under certain conditions, the temperature rises with altitude from a certain height above the ground. This results in a layer of warmer air sitting on top of a layer of cooler air. This phenomenon is called temperature inversion and the height at which the temperature profile reverses is called the inversion height (Figure 4). Temperature inversions will generate an extremely stable atmosphere within the inversion layer.

Figure 4 Inversion layer

If the pollutants are emitted within the inversion layer, they can only disperse very slightly in the vertical direction (refer to the previous discussions on atmospheric stability), resulting in a lofting plume. If the pollutants are emitted under the inversion layer, they can only disperse within the atmosphere under the inversion layer (therefore the region under the inversion layer is also called the mixing layer). In this case, the pollutants are ‘trapped’ under the inversion layer resulting in a fumigation plume (Figure 5). Also the lower the inversion height, the higher the ambient concentration of the pollutants.

Figure 4.5 Lofting plume and fumigation plume
Source: http://nptel.iitm.ac.in/courses/Webcourse-contents/IIT-Delhi/Environmental%20Air%20Pollution/air%20pollution%20(Civil)/Module-4/1.htm

The main types of temperature inversions include:
(1) Radiation inversions: occurs during cold, calm winter mornings after whole night of radiation of heat from the ground.
(2) Subsidence inversions: occurs as a result of a warmer layer of air subsiding on the layer of well-mixed air underneath under anticyclonic conditions.
(3) Frontal inversions: occurs when the warmer air mass from the coast meets with the cooler air mass from inland under cold front conditions.
(4) Sea breezes: occurs in coastal areas where the inflow of cooler sea breeze is hindered by high mountains close to the coastline. The pollution trapped underneath is often referred to as the “chimney effect”.
(5) Foggy valley: occurs in valley terrain in the morning when the layer of air within the valley is still cold and well mixed, but the layer of air above the valley is heated up by solar radiation.

In Brisbane for example, radiation inversions are most common from midnight to early morning hours in winter, leading to low altitude inversion layers and trapping air pollutants emitted from ground level sources (Figure 4). (What are the examples of emission sources on the ground, in particular in morning hours?)
1.6 Mechanical turbulence

Mechanical turbulence is related to factors such as surface roughness, aerodynamic downwash (Figure 6), obstacles (Figure 7) and street canyon effect (Figure 8). Rough surface tends to deter dispersion. Aerodynamic downwash typically increases the pollution under the buildings in the downwind direction due to emissions from stacks which are lower than 2.5 times the height of the highest building. When the plume is challenged by an obstacle such as a hill, the plume may rise above the obstacle or get around the obstacle, depending on the dimension of the obstacle. Street canyon effect tends to increase the pollution on the leeward side of the street, but decrease the pollution on the windward side of the street.

Figure 6 Aerodynamic downwash
Source: http://www.michiganair.com/newsletters/2005-3/section2.htm

Figure.7 Obstacle effect
Source: Huq et al. (2007) The shear layer above and in urban canopies. Journal of Applied Meteorology and Climatology, 46, 368-276.

Figure 8 Street canyon effect
1.7 Topographical characteristics
In areas such as the Californian Basin and the South East Queensland region that have high mountains further inland, cold inland air flows over the mountains then drains down resulting in a layer of cooler air above the coastal areas from midnight to early morning. This cold drainage winds phenomenon strengthens the trapping of pollution emitted from the ground in the early morning, before the atmosphere becomes well mixed in the day time.

1.8 Other factors
Examples of other factors which may also influence dispersion of air pollutants are:

•    Wind shearing: Wind speed and direction may vary with altitude
•    Tilting plume: Air pollutants such as sulphur dioxide and dust are heavier than air and therefore the plume will gradually tilt down to the ground
•    Deposition of air pollutants to the ground: When air pollutants such as acidic gases and particulate matter are transported to the ground, some of them may deposit and not bounce back into the air
•    Interaction between air pollutants: The interaction between air pollutants may increase or decrease the concentration of air pollutants in the atmosphere
•    Partial penetration of air pollutants into the inversion layer

Part 2:  ATMOSPHERIC STABILITY
2.1 Determination of atmospheric stability from temperature lapse rates
2.1.1 Adiabatic lapse rate (ALR) and environmental lapse rate (ELR)

Temperature in the atmosphere normally decreases with increase of height. The rate of temperature decrease with altitude, ?T/?Z, is called the lapse rate. If a parcel of air moves upward in the atmosphere without exchange of mass and energy with the surrounding air (adiabatic), the lapse rate is called the adiabatic lapse rate (ALR).  The theoretical adiabatic lapse rate for a parcel of dry air is -0.01 oC/m (or -1 oC/100m) and is called the dry adiabatic lapse rate (DALR) (Figure 9). For example, if an air parcel starts off at 30 oC on the ground and rises up adiabatically, its temperature will be 20 oC at 1000 m altitude, and 10 oC at 2000 m (Figure 9). The adiabatic lapse rate for a parcel of moist air is lower at -0.0035 oC/m.

Figure 9 Adiabatic lapse rates

The actual lapse rate of the surrounding air is called the environmental lapse rate (ELR) and can be greater, equal or less than the adiabatic lapse rate of the pollutant parcel. ELR also can vary with altitude. An inversion layer with very stable atmosphere is formed when the ELR is positive in value (i.e. temperature increases with height) (Figure 10). The importance of these two lapse rates in determining whether vertical motion in the atmosphere (and have vertical movement of pollutants) is enhanced or suppressed will now be examined.

Figure 10  Environmental lapse rates

2.1.2 Lapse rates and atmospheric stability
The vigorousness of dispersion of pollutants can be roughly estimated from the stability of the atmosphere.  Atmospheric stability is closely related to the relative magnitude of the adiabatic lapse rate (ALR; dry air ALR = -0.01 oC/m) and environmental lapse rate (ELR). In the following discussions we will consider the pollutants emitted from a stack as a huge number of tiny, insulated balloons (or parcels/puffs). Since the balloon loses very little heat to their surroundings, its upward or downward movement in the atmosphere can be assumed to be adiabatic.

Unstable conditions (ELR < 0 and |ELR| > |ALR|)
Consider the insulated balloons are dispersing at the effective stack height level in an atmosphere with environmental lapse rate steeper than the adiabatic lapse rate (i.e. temperature in the environment decreases faster than the temperature inside the balloon; ELR < 0 and |ELR| > |ALR|) (Figure 11 left). Since the balloon is stationary at the effective stack height (otherwise the balloon should rise further in the atmosphere due to buoyancy), the temperature inside the balloon should roughly equal to those in the atmosphere (Figure 11 middle).

Suppose some of the balloons are slightly higher in temperature than the atmosphere and start to rise. Once the balloons became higher in altitude the temperature difference between inside the balloons and the atmosphere is now even bigger, therefore the balloons rise further and further. On the other hand those balloons which are slightly lower in temperature than the atmosphere will sink further and further (Figure 11 middle). This results in vigorous vertical dispersion of the pollutants in the atmosphere and a looping plume shape (Figure 11 right). Unstable atmospheric conditions usually occur at midday in summer with strong sunlight.

Figure 11 Pollutants emitted in unstable conditions

Neutral conditions (ELR < 0 and |ELR| ~ |ALR|)
Now consider the insulated balloons are dispersing in an atmosphere with environmental lapse rate roughly the same as the adiabatic lapse rate (i.e. temperature in the environment decreases at roughly the same rate as the temperature inside the balloon; ELR < 0 and |ELR| ~ |ALR|) (Figure 12 left). Similar to the unstable conditions, under these conditions the balloons which are slightly higher in temperature than the atmosphere will rise, while those which are slightly lower in temperature will sink. But since the temperature difference between inside the balloons and the atmosphere remains small (Figure 12 middle), the vertical dispersion of the pollutants is moderate and the shape of the plume is coning shape (Figure 12 right). Neutral atmospheric conditions usually occur in clear afternoons.

Figure 12  Pollutants emitted in neutral conditions

Stable conditions (ELR < 0 and |ELR| < |ALR|, or ELR>0 (temperature inversion))
Now consider the insulated balloons are dispersing in an atmosphere with environmental lapse rate less steep than the adiabatic lapse rate or even positive in value (i.e. temperature in the environment decreases slower than the temperature inside the balloon; ELR < 0 and |ELR| < |ALR|, or ELR>0) (Figure 13left). Those balloons which are slightly higher in temperature than the atmosphere start to rise. But once the balloons became higher in altitude the temperature inside the balloons is now lower than the temperature in the atmosphere, therefore the balloons will drop back to the starting altitude. On the other hand those balloons which are slightly lower in temperature than the atmosphere and sink will rise back to the starting altitude (Figure 13 middle). This results in minimal vertical dispersion of the pollutants in the atmosphere and a fanning plume shape (Figure 13 right). Stable atmospheric conditions usually occur in night time and early mornings with calm wind conditions.

Figure 13  Pollutants emitted in stable conditions

In summary, the relative magnitude of the ELR comparing to that of the ALR (dry air ALR = -0.01 oC/m ) can be used to estimate atmospheric stability and the plume shape (Figure 14 and 15).

Dry air ALR = -0.01 oC/m, or ?ALR?= 0.01 oC/m
Very stable (F) when        ELR > 0 (temperature inversions)
Stable (E) when         ELR < 0 and ?ALR?> ?ELR?
Neutral (C,D) when         ELR < 0 and ?ALR?= ?ELR?
Unstable (A,B) when         ELR < 0 and ?ALR?< ?ELR?

Figure 14 Lapse rate and atmospheric stability

2.2 Determination of atmospheric stability – Pasquill-Gifford method

Environmental lapse rate (ELR) can be determined by measuring the temperature profile of the atmosphere using radiosondes. Adiabatic lapse rate (ALR) can be determined by measurements of the humidity conditions of the atmosphere, and as mentioned in above ALR in dry air = -0.01 oC/m. Then the atmospheric stability can be determined according to Figure 14. Atmospheric stability can also be determined by modelling based on synoptic wind data from satellite measurements.

If only ground level solar radiation, wind speed and cloud cover data, usually measured at 10 m above the ground, are available, the atmospheric stability can also be roughly estimated using the classification scheme proposed by Pasquill and Gifford (Table 1).

Table 1  Pasquill-Gifford classification of atmospheric stability

Surface wind speed at 10 m (m sec-1)    Solar radiation    Night time
Cloud cover fraction
Strong    Moderate    Slight    > 3/8    ? 3/8
< 2    A    A-B    B    F    F
2-3    A-B    B    C    E    F
3-5    B    B-C    C    D    E
5-6    C    C-D    D    D    D
> 6    C    D    D    D    D
Incoming Radiation    Solar radiation (W/m2)
Strong    I > 700
Moderate    350 ? I ? 700
Slight    I ? 350
Remark: A-B (unstable or convective); C-D (neutral); E-F (stable)

Note that atmospheric stability in the daytime decreases (more unstable) with solar radiation but increases (more stable) with wind speed (Why?). In the night time atmospheric stability decreases (more unstable) with cloud cover and wind speed (Why?). Strong solar radiation conditions usually occur on sunny midday in summer. Night time is defined as 1 hour before sunset to 1 hour after dawn and so varies from season to season. While the Pasquill-Gifford method is useful for rural conditions, there are also other schemes available to determine stability class in non-rural conditions.

Figure 15 summarises examples of the various smoke plumes that can occur for different environmental lapse rates. Note that emissions from different height could result in very different shapes of plume.

Figure 15  Examples of plumes under different atmospheric stability patterns
(—– ALR, ?? ELR)

Activity A.
A1.  Estimation of stability category from weather conditions:

The stability of the atmosphere can be estimated based on the weather conditions observed on the ground level using the Pasquill-Gifford scheme (Table 1).

e.g. For a mid-morning in January (summer in Brisbane) on a cloudless day with a wind speed of 2 to 3 m/s, the solar radiation is expected to be strong to moderate. Therefore according to Table 1, the stability class is A or B, with class B more likely.

Question 1:  By matching the following information on the time of year, time of day and wind speed with those in Table 1, estimate the stability category for each of the situations described below.

(a) Midday in December clear sky and a wind speed of 2 m/s………………………………………………..

(b) Midnight in June on a completely overcast night with a wind speed 5 to 6 m/s…………………………

(c) Midday in July with approximately 50% cloud and a wind speed of 3 to 5 m/s…………………………

Question 2: On the basis of the situations described below, calculate the average ELR (show your workings) and classify the degree of stability in the atmosphere for:

(a) Ground level temperature is 20 oC, temperature at 650 m is 25 oC

………………………………………………………………………………………………………………………….

………………………………………………………………………………………………………………………….

(b) Ground level temperature is 24 oC, temperature at 1400 m is 10 oC

………………………………………………………………………………………………………………………….

………………………………………………………………………………………………………………………….

A2. Determination of plume shape:

As discussed in the preliminary material and the lectures, the shape of the pollutant plume is typically looping, coning and fanning under the unstable (A,B), neutral (C,D) and stable (E,F) conditions, respectively. As discussed in above, the stability in the environment can be determined by comparing the ELR to the ALR (Figure 14).

Question 3: For each of the stability categories: A, D and F
(a)  draw the expected shape of plume emitted from a stack

(b) describe a set of typical weather conditions (time of the year, time of the day, wind conditions, etc) for each of these stability categories (hints: refer to Table , pg 10)

A:

D:

F:

A3. Wind sectors:

In dispersion models, wind direction is usually divided into 16 sectors. Only receptors located in the correct downwind sector are affected by the emissions. For example, if the wind is coming from NNE, only the receptors located in the SSW sector will be affected.

Question 4 Assuming a source in the middle of the above Figure:

4.1 If a receptor was located, south, south east of the pollution source what would be the most important prevailing wind direction?

4.2 where would the highest concentration be if the wind direction was SW?

PART 3:  MODELLING POINT SOURCES: a Gaussian method

A point source is a source of air pollution for which emissions come from one position in space.  A common example of a point source is a stack which can range from a chimney stack of a residential house, to the stack of a  power plant or a refinery.   Numerous theories exist as to how pollution from a point source is dispersed in the atmosphere.  The most widely accepted modelling theory is the Gaussian method (you may have heard of the Gaussian (or normal) distribution before in statistics).

Gaussian method
If  turbulence is homogeneous and stationary and only a point source is considered, then the pollutant concentration at a given x, y, z coordinate can be expressed is:

(3.1)

where:
Q is the source strength (mass emission rate),
u – the mean wind speed,
?y, ?z the standard deviations in concentration in the crosswind (y) and vertical (z) directions respectively, and H the effective height of emission.

The wind is assumed to be in the x direction.  The assumptions required to derive this equation are given in  Table 2.1.

Table 2.1.  Assumptions for derivation of Gaussian Plume Formulae
(a)    No initial concentration.

(b)    Steady-state emission from the point source, i.e. Q = constant.

(c)    No absorption or generation by the ground.

(d)    Constant wind in one direction, i.e.  = constant and so is its direction.

(e)    No inversion layer.

(f)    There is no downwind diffusion and the diffusivities in the crosswind and vertical directions vary only with downwind distance, x, and are constant in the diffusion domain.

In spite of the simplifying assumptions made in deriving the Gaussian plume formula and its weak verification by observational data, the formula given by equation above  is widely employed; indeed, it is the basis for the United States Environmental Protection Agency (USEPA) models recommended for use by air quality managers.  One undoubted reason for this is the simplicity of the formula; it requires little data to be collected for its use compared to other models which seek to improve on it.

Point Sources
To calculate pollutant concentrations using the Gaussian Plume formula, the source strength, Q, effective height of emission, H, mean wind speed, u, and dispersion parameters,? y and ?z are required.  The effective height H is given by

H = h + ?H

where h is the physical height of the source and ?H is the plume rise due to the buoyancy and momentum of the exit gas stream. If the plume is emitted free from these turbulent zones, a number of emission factors and meteorological factors influence the rise of the plume.  The emission factors are the velocity of the effluent at the top of the stack, and the temperature of the effluent at the top of stack. The meteorological factors affecting plume rise are the wind speed (u), temperature of the air, the shear of the wind with height, and atmospheric stability.  The plume rise can be calculated using Brigg’s plume rise formula which is described in Equation 1.

Equation  1  Briggs’ plume rise formula

Where u = wind speed (m/s-1),
h = physical height of stack (m),
H = plume rise (m), and

with     ?T =  Ts  – T, Ts = stack gas temperature (ºK),
T    =  ambient air temperature (ºK),
g    =  9.8  m/s-2
d  =  stack diameter (m)
Vs  =  gas exit velocity (m/s-1)

and p is given by:

Class    p (urban)    p(rural)
A
B
C
D
E
F    0.15
0.15
0.20
0.25
0.40
0.60    0.07
0.07
0.10
0.15
0.35
0.55

In the Gaussian Plume formula, the probability distribution of pollutant concentrations has been found to be approximately Gaussian with standard deviations in the crosswind (y) and vertical (z) of  ,? y and ?z respectively.  As a consequence, the shape of the predicted distribution of pollutant concentrations in the x,y,z directions has a conical form, with the central axis of the cone being in the x direction.  The dispersion parameters, ,? y and ?z have been found to depend on the turbulent state of the atmosphere and on the distance (in the direction of the prevailing wind) from the source.

The various turbulence states of the atmosphere are classified by six (Pasquill) categories, A-F, shown in Table 1 (pg 10).  The spread of the cone in the y and z directions depends on the displacement, x, downwind and on the stability category.  Once a stability class has been determined it is necessary to use Table 2.2 to determine the dispersion in the y and z directions that might be expected for each of these stability classes. Various forms of the Gaussian equation are shown in Box 1, to show how it can be used for different source scenarios.
Table 2.2  – Briggs’ formulae for ?y , ?z

OPEN COUNTRY CONDITIONS

Pasquill type    ?y (m)
(Recommended for 102 = x  =104)    ?z (m)
A
0.22x (1 + 0.0001x)-2    0.20x
B
0.16x (1 + 0.0001x)-2    0.12x
C
0.11x (1 + 0.0001x)-2    0.08x (1 + 0.0002x)-2
D
0.08x (1 + 0.0001x)-2    0.06x (1 + 0.0015x)-2
E
0.06x (1 + 0.0001x)-2    0.03x (1 + 0.0003x)-1
F
0.04x (1 + 0.0001x)-2    0.016x (1 + 0.0003x)-1

URBAN AREAS (102 = x  =104 m)
Pasquill type    ?y (m)    ?z (m)
A-B
0.32x (1 + 0.0004x)-?    0.24x (1 + 0.001x)+2
C
0.22x (1 + 0.0004x)-?    0.20x
D
0.16x (1 + 0.0004x)-?    0.14x (1 + 0.0003x)-2
E-F
0.11x (1 + 0.0004x)-?    0.08x (1 + 0.0015x)-2

Representative values of Zo and u* for natural surfaces
(Neutral stability, values of u* corresponding to  = 5 m s-1 at 2 m height)

Type of surface
Zo (cm)    u* cm s-1
Very smooth (mud flats, ice)
Lawn, grass up to 1 cm high
Downland, thin grass up to 10 cm high
Thick grass, up to 10 cm high
Thick grass, up to 30 cm high
Thick grass, up to 50 cm high    0.001
0.1
0.7
2.3
5
9    16
26
36
45
55
63

Box 1.  Gaussian plume formulae for different scenarios

For a source on the ground:

where    ? (x,y,z)    =    conc. at point (x,y,z) (g m-3)
Q    =    source strength (g sec-1)
?y , ?z    =    standard deviations in crosswind and vertical directions respectively (m)
u    =    mean wind speed (m sec-1) (normally 10 m above ground)

For an elevated source:

where    H = effective height of emission
u = wind speed at height H =
where u10 is wind speed at height 10 m above ground, and p is given in Table 2.2.

From an elevated source, the ground-level concentration directly downwind from the source,

2

For a ground-level source (H = 0),

Gaussian plume model assumptions and problems

•    Used to estimate atmospheric dispersion when mean wind speed and direction can be determined, but measurements of turbulence (e.g. standard deviation of wind direction fluctuations) are not available.

•    The approach is most applicable to ground level or low level releases (up to 20m), but is commonly used for higher elevations.

•    It is assumed that stability is the same throughout the diffusing layer.

•    Because mean values of wind speed and direction are required, neither the variation of wind speed nor the variation of wind direction with height in the mixing layer is taken into account.  This is not a problem in neutral or unstable (e.g. daytime) conditions, but can cause overestimations of downwind concentrations in stable conditions.

For the ?y and ?z parameter values given here, the sampling time is assumed to be 10mins, the height to be the lowest several hundred metres of the atmosphere, and the surface to be relatively open country.

The wind speed (u) is a mean through the vertical extent of the plume.  However, “surface” winds may be all that are available.

Uncertainties in calculating ?y are in general less than those of ?z.  In practice, ?z may be expected to be correct within a factor of 2 in the following circumstances:

(i)    all stabilities for distance of travel <200m;
(ii)    neutral to moderately stable conditions for distances < a few kilometres;
(iii)    unstable conditions in the lower 1000m of the atmosphere, with a marked inversion above for distances out to ?10km

Activity B: MGLC AND ATMOSPHERIC STABILITY; MIXING HEIGHT; A SIMPLE DISPERSION MODEL: GAUSSIAN PLUME EQUATION EXCEL SPREADSHEET

In the following activity you will use a simple Gaussian plume equation model to determine the impact of a point source on the hourly ground level concentration of SO2.

The Gaussian Plume Equation Excel Spreadsheet is a simple teaching program to calculate the downwind ground level concentration (GLC) of pollutants emitted from point sources. In this exercise you will use the spreadsheet to:

(1)    Investigate the influence of atmospheric conditions (unstable, neutral and stable conditions, respectively) on the downwind GLC of SO2 emitted from the power station.
(2)    Investigate the influence of mixing height on the downwind GLC.

Step 1. To run the Gaussian Plume Equation Spreadsheet
In the Air Pollution Week 8 assessment folder of the study schedule in the 3111ENV L@G page, click on the ‘Gaussian Plume Equation.xls’ file.

Step 2. To calculate the ground level concentration (GLC) at a certain downwind distance
Refer to figure above, enter the source, environmental and receptor parameters by clicking inside the parameter boxes.

What is the predicted effective stack height? (Unit also!)  _________________

Why is it much higher than the actual stack height you entered?____________________________

What is the predicted 1-hr average GLC at 3,000 m downwind from the source? (Unit also!) ________

Is the 1-hr guideline of 522 ?g/m3 for SO2 violated in this case? _____________

Step 3. To calculate the maximum ground level concentration (MGLC) under typical unstable conditions
Atmospheric stability classes A and B represent unstable conditions. Change the atmospheric stability class to B. The graph appears on the screen shows the 1-hr average ground level concentration against the downwind distance.

The predicted MGLC is (unit also!) _______________  and occurs at (unit also!) ____________

The predicted MGLC under this typical unstable condition (exceeds / does not exceed) the SO2 guideline.

Roughly sketch and label the distribution graph for this unstable condition in Figure 2 below.

Step 4. To calculate the maximum ground level concentration (MGLC) under typical neutral conditions
Stability classes C and D represent neutral conditions. Change the atmospheric stability class to D.

The predicted MGLC is (unit also!) _______________  and occurs at (unit also!) _____________

The predicted MGLC under this typical neutral condition (exceeds / does not exceed) the SO2 guideline.

Roughly sketch and label the distribution graph for this neutral condition in Figure 2.

Step 5. To calculate the maximum ground level concentration (MGLC) under typical stable conditions
Stability classes E and F represent stable conditions. Change the atmospheric stability class to E.

The predicted MGLC is (unit also!) _______________  and occurs at (unit also!) _____________

The predicted MGLC under this typical stable condition (exceed / does not exceed) the SO2 guideline.

Roughly sketch and label the distribution graph for this stable condition in Figure 2.

Figure 2 GLC against downwind distance under unstable, neutral and stable conditions

Based on your sketched graphs in Figure 2, circle the correct answers:

When atmospheric condition becomes more stable, MGLC (increases / decreases) and occurs (closer to / further away) downwind of the source.

Explain the effect of atmospheric stability on MGLC and the downwind distance when MGLC occurs. (Hints: Think about the shape of the pollutant plume under unstable, neutral and stable conditions and how this would influence the timing and concentration when the plume hits the ground.)

……………………………………………………………………………………………………………………….

………………………………………………………………………………………………………………………..

………………………………………………………………………………………………………

Step 6. Effect of inversion height (mixing height)
Change the inversion height to 50 m (other parameters unchanged).

What is the message shown on the screen?________________________________________________

What is the effective stack height as shown in the ‘effective height’ box? ____________

Is this mixing height above or below the effective stack height? __________

Compare the results with those obtained from Step 5 (in which the inversion height was 1,000 m). Comment on the effect of inversion height and the use of tall-stack strategy in the control of pollutants from point sources.

……………………………………………………………………………………………………….

……………………………………………………………………………………………………….

……………………………………………………………………………………………………….

Step 7. To leave the Gaussian Plume Equation Spreadsheet
Close the spreadsheet window.

PART 4  – USING A REGULATORY MODEL FOR POINT, AREA AND VOLUME SOURCES: AUSPLUME

The lectures and the first three components of this workshop have introduced the basic principles of dispersion of air pollutants. The concentration of a pollutant when it hits a receptor (usually located on the ground) is dependent on the emission parameters of the source, the dispersion parameters of the atmosphere and the position of the receptor.

Examples of regulatory dispersion models include the AERMOD model (US EPA) and the AUSPLUME model (Victorian EPA, Australia). These regulatory models allow the users to have more control on factors such as deposition of air pollutants onto the ground, penetration of pollutants through the inversion layer, terrain effect, building wake effect and algorithms to estimate horizontal and vertical variations.

Exercise
In the following experiment you will use AUSPLUME to determine the impact of a typical Australian power station on the hourly ground level concentration of SO2 within 10 km of the station in the Melbourne area in 1995.

NOTE: As a first time user, you may find AUSPLUME a bit complicated to run. So be patient and follow the following 21 steps closely (and efficiently) with reference to the attached figures!!

Consider a 1600 MW Australian power station which uses black coal with an average sulfur content of 1%.  At peak load, assume the power station operates at 38% efficiency.

Stack diameter = 5 m
Temperature of exit gas = 100 oC
Stack gas exit velocity = 5 m/s
Stack height = 200 m
Emission rate =2963m/s (this is calculated based on the details of the power station identified above)

(a) AUSPLUME dispersion model

Loading AUSPLUME and Surfer

1.    First you will need to install a program called Surfer which is for plotting contour maps using output from the AUSPLUME model. Find Surfer (go into search programs menu and type it in) and Double click ‘Surfer 8’ icon. Surfer will now be transferred onto your computer. After installation if the Surfer window pops up, close the Surfer window..

2.    Find AUSPLUME (go into search programs menu and type it in) and Double click the AUSPLUME icon to install it. The AUSPLUME logo will appear briefly, followed by the Main page.

Select meteorology data set

3.    In the Main page, click ‘Edit’ then ‘Sequential Inputs’. This will lead you into the ‘Simulation’ page. With reference to the following figure, enter the information, check the appropriate boxes and choose the Melbourne 1995 meteorological data set (by clicking on the ‘File’ button then choose ‘MELBOURN.MET’). Click ‘OK’ after entering the information.

What meteorology data do we need to run a regulatory dispersion model?
The meteorological data from midnight to 8 am on 1/1/1995 in the MELBOURN.MET file is shown in below to illustrate the typical hourly data we will need to run a regulatory dispersion model: temperature, wind speed, wind direction, stability class, mixing height, wind direction variation, frictional velocity, Moninobukov length, weather code and precipitation.

EPAV
95010101 17  3.2 202 E  629  15.   0.2710     74.50     4    0.0
95010102 17  2.9 197 E  564  16.   0.2710     74.50     4    0.0
95010103 17  2.5 201 E  485  14.   0.2710     74.50     4    0.0
95010104 17  2.0 190 E  406  22.   0.1800     74.50     4    0.0
95010105 17  1.7 199 F  406  26.   0.1670     50.00     4    0.0
95010106 17  2.5 230 E  465  14.   0.2710     74.50     4    0.0
95010107 17  2.9 239 C  583  12.   0.3250 100000.00     4    0.0
95010108 16  3.5 236 C  695  14.   0.4330 100000.00     4    0.0

Enter the emission source data

4.    With reference to the following figure, check the appropriate box in the ‘Emission Source’ page. Click ‘OK’.

5.    In the ‘Source Information (Stack Sources)’ page, enter the information for the power station with reference to the following figure. The x and y-coordinates (0,0) you entered will put the power station at the centre of the modelling area. Check the ‘Constant’ emission rate box and enter the emission rate you calculated in Part (a). Click the ‘Add’ button after you have entered the information to add this source to the calculation. (Although we have only one source in this exercise, AUSPLUME can handle up to 100 sources). Click ‘OK’ when finished.

6.    Skip the following ‘Area Sources’ and ‘Volume Sources’ pages by clicking ‘OK’ on these pages. (Apart from point sources, AUSPLUME can also model area sources and volume sources.)

7.    In the ‘Source Groups’ page, click the ‘Add’ button then click ‘OK’. (AUSPLUME allows you to group the sources into different groups to examine the impact of each group of sources)

Enter the receptor data

8.    In the ‘Gridded Receptors’ page, enter the following information. But do NOT click ‘Enter’ or ‘OK’ yet!!

9.    Then click on the two ‘Enter’ buttons. This creates a grid with 200 m size grid cells within 10 km (10000 m) of the power station. Later AUSPLUME will calculate the pollutant concentration in each grid cell. Click ‘OK’.

10. Skip the ‘Discrete Receptors Location’ page by clicking ‘OK’. (AUSPLUME allows you to look at the pollution levels at particular locations, such as a kindergarten or a hospital.)

Enter the output results requested

11.    In the Averaging Times page, uncheck the ‘less than 1-hr’ option and check the ‘1-hr’ averaging time, then click ‘OK’.

12.    With reference to the following 2 figures, in the ‘Output Options’ page, check the ‘Print 100 worst case table’ and ‘Plot concentration’ boxes, on the pop up page enter ‘example1.dat’ as the name for the concentration contour map file then click ‘Save’. You are now back in the ‘Output Options’ page, click ‘OK’.

Enter the land use and other advanced options

13.    In the subsequent ‘Dispersion Curves’, ‘Plume Rise’, ‘Wind Speed Categories’ and ‘Wind Profile Exponents’ pages, accept the default settings in AUSPLUME by clicking ‘OK’ on these pages. (These four pages are for more advanced users. For the purpose of this exercise we will just use the basic default values and settings.)

14.    In the ‘Land Use’ page, click on the ‘Rural flat’ button. Click ‘OK’.

15.    In the ‘Miscellaneous Parameters’ page, uncheck the ‘Including building wake effects’ option. Click ‘OK’. Click ‘OK’ for the ‘Building wake effect will be ignored’ and the ‘Sequential input is completed’ questions.

Save the entered data and run the model

16.    In the Main page, from ‘File’ and ‘Save Configuration’, save all the parameters you have entered in the configuration file as ‘example1.cfg’.

17.    Congratulations, you have finally completed the data input process and can now run the AUSPLUME model!! Click ‘Run’. Answer ‘Yes’ if some questions about overwriting of existing files come up. Click ‘OK’ for the ‘Calculation completed’ message.

18.    View the output in text format from ‘View’ and ‘View Text File Output’. This file lists out the 100 worst cases of SO2 concentration throughout 1995 and where did these cases occur.

NOW, based on the results of coordinates of the hot spots and date/time when the high pollution episodes occurred in the text output file, answer the following question:

From the text output file, what is the worst particle concentration and when and where does it occur?

………………………………………………………………………………………………………………..

………………………………………………………………………………………………………………..

………………………………………………………………………………………………………………..

19.    View the concentration contour map from ‘Plot’ and ‘Plot Run Results’. In the ‘Plot Run Results’ page check the ‘Concentration contour’ box as shown below.

20.    Click ‘OK’ if a question on ‘Undesirable contour smoothing’ comes up. After a few seconds a contour map of the maximum 1-hour average SO2 concentrations in the area around the power station will pop up.

Based on the results of approximate distance and direction of the hot spots from the power station in the concentration contour plot, answer the following question:
From the concentration contour plot, which parts of the surrounding area are the hot spots and how would you minimise the impact?
………………………………………………………………………………………………………………..

………………………………………………………………………………………………………………..

………………………………………………………………………………………………………………..

Also based on the observed location of the hot spots, comment on the prevailing wind directions in Melbourne.

………………………………………………………………………………………………………………..

………………………………………………………………………………………………………………..

21.    Close the ‘Surfer’ window. Answer ‘No’ to the ‘Save changes’ question. Click ‘Cancel’ to go back to the Main page. You may close the AUSPLUME window.