United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
EPA-450/2-80-082
November 1980
Air
&EFA
Interim Guidance
for Visibility Monitoring
-------
EPA-450/2-80-082
Interim Guidance for
Visibility Monitoring
Environmental Monitoring Systems Laboratory
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air, Noise, and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
November 1980
-------
DISCLAIMER
This report has been reviewed by the Environmental Monitoring Systems
Laboratory-Las Vegas, U.S. Environmental Protection Agency, and approved for
publication. Mention of trade names or commercial products does not consti-
tute endorsement or recommendation for use.
-------
CONTENTS
List of Tables v
List of Figures vi
1. Introduction 1
2. General Discussion 3
2.1 Visibility Definition 3
2.2 Visibility Theory 4
2.3 Instrumentation 8
2.3.1 Optical measurements 11
2.3.1.1 Contrast measurements 11
2.3.1.2 Scattering measurements 13
2.3.1.3 Transmission measurements 17
2.3.2 Additional measurements 19
2.3.2.1 Particulate measurements 19
2.3.2.2 Meteorological measurements 22
3. Program Design Considerations 23
3.1 General 23
3.2 Monitoring Duration 24
\
3.3 Monitoring Instrumentation . . , 24
3.3.1 Contrast measurement 25
3.3.1.1 Instrumentation 25
3.3.1.2 Sensitivity 26
3.3.1.3 Site selection 26
3.3.1.4 Target selection 27
3.3.1.5 Frequency 31
-------
3.3.2 Scattering measurement 32
3.3.2.1 Instrumentation 32
3.3.2.2 Sensitivity 32
3.3.2.3 Site selection 34
3.3.2.4 Probe siting 34
3.3.2.5 Operations 34
3.3.2.6 Frequency . 37
3.3.3 Photographic measurements 37
3.3.3.1 Instrumentation 37
3.3.3.2 Site and target selection 37
3.3.3.3 Frequency 37
3.3.4 Particulate measurements 37
3.3.4.1 Instrumentation 37
3.3.4.2 Site selection 38
3.3.4.3 Probe siting 38
3.3.4.4 Frequency 38
3.3.4.5 Analysis 38
3.3.5 Meteorological measurements 40
4. Non-Routine Monitoring 41
5. Quality Assurance For Visibility Monitoring Data 42
5.1 Introduction 42
5.2 Internal Quality Assurance Procedures 42
5.3 External Quality Assurance Procedures 45
5.4 Techniques for Data Assessment 45
5.4.1 Telephotometer 46
5.4.2 Nephelometer 49
5.4.3 Camera 49
IV
-------
5.4.4 Particulate measurement 50
5.4.5 Meteorological measurement 51
6. Data Reporting, Management and Interpretation 52
7. References 60
-------
LIST OF TABLES
Number Page
1 Recommended Minimum Visibility Monitoring Program 25
2 Minimum Contents of Visibility Monitoring Program Description . 44
3 Data Assessment Techniques for Visibility Instruments 47
4 Visibility Related Variables 53
VI
-------
LIST OF FIGURES
Number Page
1 Elements of Visibility 9
2 Telephotometer Schematic 10
3 Integrating Nephelometer Schematic 14
4 Laser Transmissometer 18
5 Bimodal Mass Distribution of Atmospheric Particles 20
6 Telephotometer Target Distance As A Function of Average
Extinction Coefficient 29
7 Visual Range Error Resulting from Inherent Contrast
Assumptions 30
8 Uncertainty in Predicted Apparent Contrast Resulting from
Extinction Coefficient Error 33
9 Nephelometer Clean Air Reference System 35
10 Effect of Relative Humidity on Scattering Coefficient 36
11 Log Probability Cumulative Frequency Distribution for
Visual Range ^
12 Normal Probability Cumulative Frequency Distribution for
Apparent Vista Contrast 58
13 Normal Probability Cumulative Frequency Distribution for
Plume Contrast 59
vn
-------
1. INTRODUCTION
Section 169A of the Clean Air Act requires the Environmental Protection
Agency (EPA) to promulgate regulations to assure reasonable progress towards
the congressionally declared goal of "The prevention of any future, and the
remedying of any existing, impairment of visibility in mandatory Class I
Federal areas, which impairment results from man-made air pollution." Visi-
bility analysis is also required under EPA's prevention of significant
deterioration (PSD) regulations. EPA has proposed regulations which would
require certain states to develop and implement programs to address the con-
gressionally declared national goal (FRL 1487-2, Docket No. A-79-40). These
regulations require the consideration of visibility monitoring and data in
three aspects:
1. Identification of visibility impact from existing sources.
2. Visibility assessment for New Source Review.
3. Evaluation of long term strategy for making reasonable progress
toward achieving the national goal.
In July 1978, EPA sponsored a Visibility Monitoring Workshop to address
technical and programmatic visibility monitoring needs . This workshop was
attended by representatives of the EPA, .Departments of Interior and Agricul-
ture, Electric Power Research Institute, industry consulting firms, and other
scientific and government organizations. The recommendations of this workshop
served to provide a focus for EPA's visibility monitoring methods development
program. EPA recognizes that continuing research is required in support of
visibility monitoring both in the area of instrumentation and in the use and
-------
interpretation of data obtained. Although the Agency has not yet promulgated
reference methods, there is substantial information available regarding visi-
bility monitoring methods presently in use. This document is intended to sum-
marize that information in terms of interim monitoring recommendations.
The first section of the document covers the general concept of visibility
and principles of measurement. A more thorough discussion is found in "Pro-
2
tecting Visibility, An EPA Report to Congress" . The remaining sections dis-
cuss technical considerations involving the design of visibility monitoring
programs, selection of instrumentation, quality assurance and data processing.
A more detailed visibility monitoring procedures manual will be available
in the near future which will detail procedures on operation and maintenance,
data handling, calibration, and quality assurance for visibility monitoring.
Over a longer time frame, EPA intends to develop a standardized reference
method (or methods) for visibility monitoring.
-------
2. GENERAL DISCUSSION
2.1 Visibility Definition
Visibility can be broadly defined as the degree or extent to which something
is visible. The study of visibility and its relationship to meteorology and
atmospheric aerosol content is a complex and, in many cases, a semi-quantitative
science. Traditionally, visibility has been defined in terms of visual range;
the distance from an object that corresponds to a minimum or threshold contrast
between that object and some appropriate background. Threshold contrast refers
to the smallest difference between two stimuli that the human eye can distinguish.
The measurement of these quantities depends on the nature of the observer, his
or her physical health, and mental attitudes of attention or distraction such
as effects of boredom and fatigue.
Although visibility defined in terms of visual range is a reasonably pre-
cise definition, visibility is really more than being able to see a black tar-
get, or any target, at a distance for which the contrast is reduced to the
threshold value. Visibility also includes seeing vistas at shorter distances
and being able to appreciate the details of line, texture, color, and form.
Since at this time it is not reasonable or even possible to define visibility
in terms of one physical variable, the recommended alternative is to measure
a set of variables that effectively characterize visibility, that is charac-
terize the perception of such things as color, line, texture and form.
Characterization of visibility and its impairment involves the measurement
of variables that: 1) relate directly to what the eye-brain system sees, 2)
can be monitored directly, and 3) can be related to the atmospheric constituents
3
-------
controlling visibility. Apparent target contrast is a variable which meets
the first two criteria and serves as the fundamental measure of visibility
impairment. Color change also meets the first two criteria. However,
current knowledge does not allow for a definition of the best way to express
vista color and color change and thus they will not be addressed in this
document. The measurements of fine particulate concentration and scattering
coefficient go a long way towards achieving the second and third criteria
and must be considered as an integral part of any visibility monitoring
program. Visual range cannot be directly measured by. any instrument.
However, when site intercomparisons are required (such as for establishing
regional trends) it will be necessary to use visual range as a normalizing
parameter. Also, because of its historical popularity, it remains a useful
concept to indicate atmospheric "clarity" to the lay person.
A visibility monitoring program should utilize a combination of these
measures depending on the specific objectives of the program.
2.2 Visibility Theory
In order to deal quantitatively with visibility impairment, it is necessary
to define the physical basis for light absorption and scattering.
The importance of air quality impact on visibility, "the seeing" of distant
objects, is based on the ability of aerosol to scatter and absorb image forming
light as it passes through the atmosphere (Figure 1). The loss of image forming
light is proportional to b$, and bgbs, the atmospheric scattering and absorption
coefficient. The combined effects of scattering and absorption will be referred
to as extinction, and b w-m be used to represent the extinction coefficient.
-------
Radiance, N, is a measure of the amount of monochromatic radiant energy
present at some point in space. Thus tNr, the apparent target radiance incident
at an observation point located a distance r from some target, is a measure of
radiant energy reaching an observer who is viewing a target in some specific
direction. ^Nr is then the sum of the attenuated inherent radiance of the
target, .N , and radiant energy scattered by the intervening atmosphere. The
u o
radiant energy scattered by the intervening atmosphere is a result of air mole-
cules or aerosols scattering direct sun, diffuse, or ground reflected light into
the sight path. The volume scattering function determines how much of the
radiant energy incident on the sight path is scattered toward the eye. It is
a minimum for radiant energy incident perpendicular to the sight path and a
maximum for radiant energy incident on the sight path from in front of or behind
the observer (forward and back scattering). The relative amount of forward and
back scattered radiation necessarily depends on the relationship between the
wavelength of incident radiation and particle size.
Apparent target contrast, C , is defined as the difference between target
radiance, .N , and some background radiance, .N (when the background is the
U i D i
sky, . N becomes N ) divided by the background radiance.
c -
s"r
In a similar manner, inherent contrast, C , is defined to be the contrast
of a target viewed at a distance r = 0, against a background sky:
4.N - N
r _ to so o N
° ~ '
-------
Apparent and inherent contrast of a coherent plume (or horizontally con-
strained layer) of aerosol as seen against a sky or vista background can be
expressed in a similar way:
r,P
and
_
°»P bNo
where C and C are the inherent and apparent plume (layer) contrast while
o ,p r ,p
N and ,N are the inherent plume (layer) and background radiance values
respectively. N and .N are the respective radiance values of the plume
(layer) and background at a distance r. The background radiance, . N , may be
for either the sky or a selected vista. It should be noted that contrast is a
unitless parameter.
The ratio of the apparent to inherent contrast (C /C ) is contrast trans
mittance, a measure of the ability of an intervening atmosphere to transmit
contrast. The equation that describes the reduction of contrast over a path
of length r is given by:
sNo e"bextr
o ~~W~
s r
where t5"ext is the average extinction coefficient over the distance r. Change
in apparent vista contrast, AC , the physical parameter that relates directly
to human perception of visual air quality, is calculated by comparing apparent.
-------
target contrast, C , to the apparent contrast for the same target when the
atmosphere is free of any air pollution (Rayleigh atmosphere), C :
r ,ra.y
"V ' Cr -
In most cases C will have to be calculated using equation 5 with N / N = 1
1 ) i fljr S O S I
and assumming F . is equal to Rayleigh scattering at the altitude of the obser-
vation point.
The quantity sNQ/sNr is equal to 1 if the earth is assumed to be flat, the
atmospheric aerosol and gas concentrations are assumed to be evenly dispersed
both in the vertical and horizontal, and the observation angle is equal to
zero (horizontal sight path). With these assumptions, equation 5 can be trans-
formed to an equation for the extinction coefficient b .
In addition, if the above assumptions are met, visual range can be calculated
from the extinction coefficient by:
Vr = 3'912/bext
If it is further assumed that the absorption coefficient is zero (b . = 0)
then bgxt = b and equation 8 becomes:
= 3.912/bs, 9.)
-------
where b is the scattering coefficient. While radiance values, extinction
and scattering coefficients can be measured at many different wavelengths
(colors), it is usually desirable to make one measurement at 550 nm since
the human eye response curve peaks near this wavelength. More detailed
234567
discussion of visibility theory is available elsewhere ' ' ' ' ' .
2.3 Instrumentation
Visibility; the seeing of distant objects, depends on properties of the
object, its background, the quality of the air along the sight path, the
length and illumination of the path, and the observation angle with respect
to the horizontal.
As indicated previously there are three basic criteria to be used when
evaluating candidate parameters to characterize visibility impairment.
1. Measured parameters should relate directly to what an observer perceives,
2. These parameters should be directly measurable.
3. Measured parameters should relate to pollutants causing visibility
impairment.
Instrumental measures that relate to visibility are generally of two
types: instruments that measure the optical properties of the atmosphere
and those instruments that measure physical characteristics of atmospheric
constituents.
-------
2.3.1 Optical Measurements
Optical visibility-related instruments are divided into three major
classes: contrast, scattering, and transmission type measurement.
2.3.1.1 Contrast Measurements: Contrast type instruments measure the amount
of radiant energy reaching a detector from selected targets and their surround-
ing background. These instruments are called telephotometers and directly meas-
ure the apparent spectral radiance of the sky, target or a plume and thus allow
for a calculation of target or plume contrast and its change. The apparent
contrast of targets or plumes can be easily calculated from the measurements
using equation 1 or 3. Visual range can also be calculated after making a
series of assumptions about the inherent contrast of the target, uniformity of
the atmosphere along the sight path and angle of observation.
Telephotometers make measurements in a way that is very similar to observa-
tions made by the human eye. In Figure 1, the eye could be replaced by a tele-
photometer (shown in Figure 2). Properties of the target, air quality
(homogeneity and concentration of visibility reducing aerosols), distance to
the target, illumination of the sight path, humidity, and observation angle
all affect the measurement. A disadvantage of making such measurements with
telephotometers is that it is difficult to separate the different effects.
The separation of effects is important if the goal is to isolate the effect
of anthropogenic air pollutants on visibility.
Currently telephotometer instrumentation can be broken into three classes:
1) telephotometers which measure radiance at one point in the sky and one point
in the vista (two-point telephotometer). The resulting radiance values can
9
-------
Background
Sky
Target
Figure 1. Elements of visibility.
The observer is at a distance r from the target; the inherent back-
ground and target radiance are represented by ,N and N respectively
while .N and ,N are the apparent background and target radiances.
U I U I
Point (1) represents the reduction of sky and target radiance resulting
from absorption; point (2) shows the reduction in sky and target radi-
ance resulting from scattering; point (3) represents the increase in
target and sky radiance resulting from sunlight scattered into the
sight path while point (4) represents increase in target and sky radi-
ance due to scattering of sky light and ground reflected light.
10
-------
Telescope (set for infinity focus)
Objective
Lens
2"F/15
Focusing Eyepiece
Target/Sky Selector Knob
Eyepiece Flip Mirror Mountjng plate
Altitude Adjuster
Liquid
Crystal
Display
Electronics Module
Figure 2. Telephotometer Diagram.
This diagram shows the typical configuration of the electronic and
optical components for a telephotometer. This instrument measures
sky and target radiance at a number of different wavelengths.
11
-------
then be used to calculate the apparent target contrast between these two
points in the vista. 2) Scanning telephotometers which measure the radiance
of a number of points (40 or more) within a vertical scan of the vista. This
type of measurement will allow for a calculation of plume or layered haze con-
trast as well as the contrast reducing capability of the layered haze. Cal-
culation of vista apparent contrast using only radiance values measured from
two points in the vista might miss the visual effect of a layered haze, while
a measure of radiance values at each point within a vertical scan would not.
3) Cameras; telephotometers which use photographic film as the detector. Pho-
tographic documentation can be utilized to "reasonably attribute" visibility
impairment to existing stationary sources as required under "Phase I" of the
proposed visibility regulation. It also provides a means of documenting the
effect of light absorption by NCL on plume color (the brown cloud effect).
Photographic data serves as a valuable supplement to other measurement tech-
niques. For example, cloud cover, snow cover, and vertical stratification as
documented through photography can be used to interpret telephotometer con-
trast data.
Human eye observation is another important contrast type approach to meas-
uring visibility. Human eye observations are useful in establishing whether
visibility impairment is "reasonably attributed" to an existing stationary
source. Also, these observations presented as visual range form the largest
Q
source of historical visibility data . By knowing distances to spatially dis-
tributed vistas or targets, trained observers are able to approximate the
furthest distance they can see, that is, determine the visual range. However,
the natural targets available for viewing at different locations are not uni-
form in directions and distances from the observation sites, making the com-
parison of sites difficult. The availability of targets at long distances
12
-------
(MOO km) has usually set an arbitrary upper bound on the visual range
reported for each location.
In summary, it is recommended that, when doing routine visibility monitoring,
contrast measurements should be made with a two-point multi-wavelength telephotometer
that has a capability of measuring apparent target and sky radiance. Color photog-
raphy should also be employed to document vista appearance.
2.3.1.2 Scattering Measurements: The second major class of visibility-related
instruments measures the light scattered from a relatively small volume of
air in specified directions and at one point in space. Scattering instruments
measure a basic optical property of the air sample, the volume-scattering
function, which is independent of target properties, natural illumination of
the atmosphere, and distance between the observer and the target. Many scat-
tering instruments enclose the air sample, allowing continuous day and night
operation by eliminating any need to use natural illumination. Enclosed
instruments also allow control of ambient air conditions in order to study the
influence of relative humidity, for example. Some unenclosed scattering
instruments modulate (vary) the intensity of the light source in order to
allow operation during daylight hours.
Scattering instruments measure the light scattered at various angles
from the air sample. The choice of scattering angle allows a classification
of the instruments into integrating nephelometers, backscatter meters, forward
scatter meters and polar nephelometers.
Integrating nephelometers (schematically shown in Figure 3) measure the
13
-------
Clean Air
Purge Marrow Band
, . /Optical Filter
Tungsten Filament
Light Source
Aerosol
Outlet
Collimating Disks
Aerosol
Inlet
v
Clean Air
Purge
Figure 3. Integrating Nephelometer Diagram.
Schematic shows a typical configuration of the electronic and optical
components of an integrating nephelometer. Notice how the relative
location of the phototube and light source allow this instrument to
detect radiant energy that has been scattered in the forward and back
directions. As a consequence, this instrument measures the total
scattering coefficient.
14
-------
light scattered over an angular range covering nearly 0° to 180°, hence, these
instruments approximate the total scattering coefficient b . The air sample
is enclosed, allowing automated, continuous day and night operation. The
instrument is calibrated to read out the total scattering coefficient of the
air sample. This variable can be translated into the visual range of a
black target if the atmosphere is uniform over a distance as long as the visual
range and if the target is viewed horizontally. Otherwise instrumental output
interpreted in terms of visual range is not valid.
Backscatter meters, measure the amount of light scattered backwards from a
volume of air (scattering angle between 90° and 180°). The instruments that
use a laser for the light source are usually called lidars. Several other
backscatter instruments use incandescent or spark lamps. Lidars have been
most commonly used to measure the elevation of aerosol layers in the atmosphere
by probing from the ground or aircraft with an upward or downward directed beam.
Except for the problem of eye safety, lidar could be used in a horizontal ori-
entation to measure the distance of plumes and other inhomogeneities. No back-
scatter instrument is suitable for measuring scattering coefficients less than
the 0.03 km (130 km visual range) found in many Western Class I areas.
Forward scatter meters similarly limited to measuring scattering coef-
ficients greater than 0.2 km are much too insensitive for use in most Class I
areas. These instruments measure the amount of light scattered from a col-
limated beam in a forward direction (scattering angle between 0° and 90°).
They are most commonly used for measuring visibility at airports and along
stretches of highway where fog is a danger.
15
-------
The polar nephelometer is a scattering instrument that measures the light
scattered from a collimated source in any specified direction, that is, the
volume scattering function. The volume-scattering function is usually measured
at a number of scattering angles between 10° and 120°. The polar nephelometer
is a powerful research instrument but it is not yet appropriate for routine
monitoring use.
It is important to note that the transfer of light through the atmosphere
depends on two aspects of air quality: scattering and the absorption of light
by gas molecules and aerosol. The total attenuation (loss) of light being
transferred through the atmosphere is equal to the sum of scattering and absorp-
tion. Typically, since scattering dominates absorption, especially in clean air,
it is acceptable to neglect absorption. However, it is not acceptable for urban
air or for plumes of rurally located large point sources like coal-fired power
plants.
The final type of scattering instruments measure sky radiation. Some of
these instruments are used just like a telephotometer, measuring the amount of
light (the apparent spectral radiance) reaching the detector from a small por-
tion of the sky. The measurement of sky radiation is important to a rigorous
calculation of the visual range of mountains and other natural targets viewed
against a sky background. Other sky radiation instruments (pyranometers)
measure the total amount of light coming from the sun and sky (the downwelling
irradiance). The total sky radiation (sky irradiance) can be measured without
the sun irradiance if the instrument is used with an occulting disk to block
the direct solar radiation.
The integrating nephelometer is the only instrument recommended for
16
-------
routine monitoring of scattering coefficient.
2.3.1.3 Transmission Measurements: Transmission instruments, shown schemat-
ically in Figure 4, measure the amount of light transmitted from a specified
source to a receiver, allowing the direct calculation of the average attenuation
coefficient of the air along the instrument path. These instruments are inde-
pendent of the effects of target properties and the natural illumination of
the sight path.
One class of transmission instruments called transmissometers uses arti-
ficial light sources, including incandescent lamps, xenon spark lamps and
lasers. Transmissometers require the placement of either the receiver or
reflectors at one end of a baseline and the transmitter at the other end. This
fixed baseline does not allow the instrument to easily measure visibility-related
variables in different directions. The path for transmission instruments is
long compared to the small volume measured by scattering instruments and
short when compared to a typical 50 to 90 km path used by a telephotometer.
Laser transmissometers are faced with a critical sensitivity to atmospheric
turbulence, which is a problem when attempting to measure transmission through
the very clean air characteristic of Western Class I areas. Additionally, a
laser transmissometer is limited to one wavelength and, in the case of a He-Ne
laser (663 nm), to a wavelength that is unrepresentative of the peak sensitivity
of the human eye (550 nm).
Another class of transmission instruments called pyrheliometers measures
the apparent sun radiance. These measurements allow the calculation of the
optical thickness of the total atmosphere along the path between the observer
17
-------
oo
Light
Source
Detector
Scattered Light
Figure 4. Laser Transmissometer.
Diagram shows the operating principle of a laser transmissometer. As light passes through the
atmosphere, it is removed through scattering and absorption processes (extinction). By knowing
the amount of radiant energy at the transmitter (laser) and measuring how much of that energy
is left after passing a certain distance through the atmosphere, extinction is measured.
-------
V
and the sun. Corrections for Rayleigh scattering and ozone absorption allow
a calculation of the optical thickness due to just aerosol scattering and
absorption.
Transmission type measurements are not recommended for routine visibility
monitoring.
2.3.2 Additional Measurements
2.3.2.1 Particulate Measurement: There are several instrumental methods that
measure the size distribution, mass concentration, or number concentration of
the aerosol that usually dominates the scattering and absorption of light in
air. Given the aerosol size distribution, the Mie theory of aerosol scattering
can be used to calculate the scattering coefficient and thus allow an approxi-
mate determination of contrast reduction, visual range, and color change.
More importantly, aerosol characterization (composition and morphology) may allow
the identification of aerosol sources so that the relative contribution of these
sources to visibility degradation can be evaluated and in turn the visibility
degradation resulting from anthropogenic activities can be determined.
Generally speaking, the mass distribution of atmospheric aerosol is bi-
modal in nature (see Figure 5). The coarse mode is defined as containing par-
ticles greater in size than about 2 m while smaller particles are said to be
in the fine mode. For the most part, the coarse particles are mechanically
produced and the fine particles result from condensation and coagulation proc-
esses. Traditionally particle sampling has been performed with filters col-
lecting particles over the the entire range from about 100 to 0.1 micrometers
in diameter. Ideally for visibility monitoring, particles should be size-
19
-------
180
Scale Change
i 1 1
Run Concentration Percent
Run 14
Run 4-14 Average
0.03
100 300
Particle Diameter - urn
Figure 5. An Example of a Bimodal Mass Distribution of Atmospheric Particles.
Bimodal mass distributions measured with a set of special impactors
and a cascade impactor are shown in this plot of change in mass con-
centration for a certain change in diameter versus particle diameter.
Run 14 contains many more coarse particles than the average because
of construction activity upwind. Note the negligible effect of this
increased concentration of coarse particles on the fine particle
mode
20
-------
segregated into many samples (e.g., ten size ranges less than 20 ytn) since
their efficiency for light scattering is size dependent. However, since air-
borne particles between 1 and 0.1 micrometers in diameter dominate other par-
ticles in their efficiency to scatter light, it is possible to sample in
fewer size ranges. As a minimum, particles should be collected in two size
ranges corresponding to the two ambient aerosol modes. The size ranges selec-
ted by EPA for the Inhalable Particluate Monitoring Network (15 to 2.5 ytn and
less than 2.5 urn) can be used for this purpose .
There are a number of size-segregating particle samplers available.
These include physical and virtual impactors, cyclones, and series filtration
samplers. In selecting a sampler several things must be considered. The
sample configuration and collection medium or substrate must be compatible
with the anticipated analysis. Sample duration and flow rate plus estimated
concentration within each size range can be used to estimate collected sample
mass. Collected sample mass must also be compatible with the analysis tech-
niques. Most manufacturers of sampling equipment include with their specifi-
cations a list of compatible analysis techniques.
There are potential problems associated with various types of size segre-
gating samplers which can result in erroneous capture of large particles on
smaller particle stages. The use of some types of sample substrates can pro-
mote the conversion of gaseous pollutants to particles on the substrate.
Detailed discussion of these problems and methods to avoid or minimize them
are beyond the scope of this document. A review of literatue pertaining to
particle sampling and analysis methodology is a prudent step to take before
selection of a sampling technique.
21
-------
2.3.2.2 Meteorological Measurements: Meteorological information is essential
for the thorough interpretation/evaluation of the physio-optical and supple-
mentary particulate data and for use with tools like models to ascertain the
effectiveness of proposed control strategies. For instance, information on the
speed and direction of the wind can help in the identification of sources of
pollutants contributing to visibility degradation. Also, relative humidity,
obtainable from measurements of a temperature and dewpoint temperature, can
assist in the understanding of local aerosol growth.
Ideally, detailed spatial and temporal information on the dynamic struc-
ture (e.g., wind speed and direction and mixing layer thickness) and attendant
turbulent properties of the atmospheric boundary layer should be collected.
For some monitoring sites, data collected at nearby locations by governmental
agencies such as the National Weather Service or by private agencies may be
sufficient. For other sites, especially in remote regions in complex terrain
or near large bodies of water, such data may not exist or may not be represent-
ative. The necessary meteorological monitoring thus should be determined on a
case-by-case basis and include consideration of existing data collection.
Guidance for meteorological measurements with regard to monitoring is
found in "Ambient Monitoring Guidelines for Prevention of Significant Deterior-
11 i ?
ation" and in "Guidance for NAQTS: Review of Meteorological Data Sources"
while that with regard to modeling is found in "Guideline on Air Quality
Models"13.
22
-------
3. PROGRAM DESIGN CONSIDERATIONS
3.1 General
Three broad objectives of routine visibility monitoring are considered
in this guideline. One is to determine the impact of existing emission sources
on visibility in mandatory Class I areas. Another is to document and evaluate
progress towards achieving the national goal in preventing and remedying visi-
bility impairment. Monitoring to achieve these two objectives is the respon-
sibility of the State and the appropriate Federal Land Manager (FLM). An
existing emitting facility may also monitor to document its impact on a
Class I area. The third objective is to develop a visibility data base as
might be required as part of new source review procedures for PSD permit
approval. This monitoring is conducted by the Prevention of Significant
Deterioration (PSD) applicant, at the discretion of the State or permit grant-
ing authority. Although these objectives are different, the visibility moni-
toring program developed to meet them must address two needs in each case.
1. Document visual air quality.
2. Ascertain a cause-effect relationship for the observed impair-
ment.
As a result, several different measurements must be made for a given
vista. Documenting visibility impairment requires contrast measurements using
a telephotometer, with supporting photographs. Supplemental measurements
23
-------
include fine-particulate collection and analysis, measurement of scattering
coefficient by an integrating nephelometer, and measurement of meteorological
parameters. These measurements are important for interpreting visibility data
and establishing the cause of visibility impairment.
3.2 Monitoring Duration
Regulatory language suggests that visibility assessment for new or exist-
ing major point sources be done on a seasonal basis. Seasons are defined as:
Winter, December to February
Spring, March to May
Summer, June to August
Fal1, September to November
It is recommended that a minimum of one full year of monitoring be conduc-
ted for visibility analysis of major point sources. Because of climatological
variability, it is recommended that five years of data be obtained to document
ambient conditions and to perform trend analysis. Selected monitoring stations
will be needed to monitor over a longer term (10 to 15 years) in order to
document progress towards achieving the national goal.
3.3 Monitoring Instrumentation
The minimum recommended sampling configuration for routine visibility
monitoring is shown in Table 1. Detailed guidance concerning instrumentation,
siting, measurement frequency, and operation is presented below.
24
-------
TABLE 1. RECOMMENDED MINIMUM VISIBILITY MONITORING PROGRAM
Instrument Parameter Frequency
Electro-optical Measurements
Manual or Continuous multi-wave- Target and sky radiance Manual:
length telephotometer 3 measurements/day
Continuous:
daylight hourly
averages
Camera (color photography) Vista appearance 3 photographs/day
Integrating nenhelometpr Scattering coefficient Continuous (hourly
average)
Supplemental measurements
Particulate monitor Mass concentration of 2 samples/week
particulates, elemental
constituents, in 2 size
ranges
Meteorological sensors See Section 2.3.2.2
3.3.1 Contrast Measurement
Target contrast provides a basic measure of visibility impairment. It
is directly measurable and relates directly to what people perceive in terms
14
of visual air quality .
3.3.1.1 Instrumentation: The multi-wavelength telephotometer is the fundamental
instrument for contrast measurement. The telephotometer should have the capa-
bility of recording sky and target spectral apparent radiance at wavelengths
of 400 nm, 450 nm, 550 nm, and 630 nm. The 550 nm wavelength corresponds to the
peak response of the human eye and is used as a measure of contrast and to cal-
culate visual range. The other wavelengths will be needed to evaluate vista
color. The full width at half maximum (FWHM) of each "color band" should not
25
-------
exceed 30 nm. The sensitivity of the instrument should be such that it is
2
capable of measuring a radiance of 2 ywatts/cm , steradian, nanometer. Depend-
ing on operating requirements, either a manual or continuous telephotometer
may be used. If a coherent plume can typically be seen, or if levels of
layered haze are observable, a scanning telephotometer should be considered.
A two-point telephotometer might look over or under the layered haze or plume.
3.3.1.2 Sensitivity: On-going studies are establishing the functional rela-
tionship between physical variables (apparent target contrast) and human per-
ception. There is strong evidence that the functional relationship between
perceived visual air quality and contrast is linear; furthermore, that an
observer can perceive differences between vista apparent contrast as small as
14 15
0.04 ' . Ideally then, an instrument should have a sensitivity such that
its output is capable of predicting vista apparent contrast changes of approxi-
mately 10 percent of the smallest perceptible value or a contrast change of
0.004. For most commercially available telephotometers the actual uncertainty
in measured contrast is less than 0.01 , which is satisfactory for routine
monitoring.
3.3.1.3 Site Selection: A number of criteria must be considered when selec-
ting a site for contrast monitoring.
a) Appropriate vistas: The visibility regulation requires the identification
by Federal Land Managers of vistas to be afforded visibility protection. Such
vistas become priority candidates for visibility monitoring sites.
b) Logistics: In the design, deployment, and operation of a monitori
ng
26
-------
station, the accessibility of the site and the availability of power are prime
considerations. Security and aesthetic impact are other practicalities which
must be considered.
c) Local Impact: Telephotometer measurements tend to be quite independent
of local sources if the vista of concern extends well beyond the immediate
vicinity of those sources. However, local sources should be avoided.
d) Target Diversity: If possible, it is recommended that multiple targets
be monitored from a single site. A minimum of three targets is recommended.
3.3.1.4 Target Selection: The four primary considerations for target selec-
tion are:
a. the distance from the observation point
b. size of target
c. the inherent color of the target
d. the observation angle.
a) Target Distance: The optimum target distance to minimize measurement error
and maximize sensitivity is directly dependent on the visual air quality.
The optimum distance between observer and vista should be between 10 percent
and 75 percent of the average visual range; the ideal distance being 25 percent
of the average visual range. Figure 6 graphically displays acceptable observer
target distances as a function of the average visual range and extinction
coefficient.
27
-------
100
0.0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Existing Average Extinction Coefficient (km-1)
0.09 0.1
400 200 10090 80 70 60 50
Existing Average Visual Range (km)
40
Figure 6.
Telephotometer Target Distance as a Function of Average Extinction
Coefficient.
This graph shows the acceptable telephotometer observer-target
distance as a function of the average existing visual range or
extinction coefficient. The solid center line is the ideal dis-
tance while the shaded area represents acceptable telephotometer
target distances.
28
-------
\
For example, for a condition typical of a Western Class I area, one can
assume an average visual range of 130 km (extinction coefficient = 0.03 km ).
For this average condition (from Figure 6), targets between 13 and 98 km would
be acceptable, with 33 km being optimum. In an Eastern situation with an average
visual range of 40 km (extinction coefficient = 0.1 km ), the optimum target
distance would be 10 km.
b) Target Color: The ideal (black) target normally does not occur in the real
world. However, targets of uniform dark color (conifer tree-covered mountain
or hill) are often available. Target color and inherent contrast must be cor-
rected for especially if the data are to be used for calculating visual range .
Figure 7 shows the effect that an erroneous assumption for inherent con-
trast, C will have on calculated visual range or extinction coefficient as a
function of the ratio of target distance (r) to visual range (V ). This figure
also points out the extreme importance of choosing a tarret that is greater
than 10 percent of the average visual range.
For example, tree covered mountains have an inherent contrast, C , at
550 nm, of -0.87 in shade and -0.72 in sunlight. Assuming this type of target
to be black (C = -1.0) would result in an error in C of 15 percent and 39
percent respectively. Fo" a short target-observer distance (r/V < 0.2) the
resulting error in calculated visual range or extinction coefficient is greater
than 25 percent. This error drops to less than 10 percent for r/V > 0.6. Thus
when visual range is calculated, it is important to know the appropriate C values;
targets cannot be assumed to be black. The development of C values specific
for individual targets under varying light conditions is an area of active
29
-------
0)
D)
C
CO
QC
1
50-i
C <
® t
'o °
£ 2
0) O
O r-
O x
O 0>
+= ^
o <
X
UJ
I.
o
>_
^
UJ
40-
30
x
0)
20-
10
00
0.0 0.2 0.4 0.6
r/Vr
0.8
1.0
Figure 7. Visual Range Error Resulting from Inherent Contrast Assumptions.
This graph shows the error in extinction or visual range as a func-
tion of the target distance divided by existing visual range (r/Vp)
for different inherent contrast errors. It should be evident that
it is important to keep the target distance greater than 10 percent
of the average visual range. Typically for tree-covered targets under
even illumination, the error in inherent contrast is less than a
few percent.
30
-------
research. The previously stated values of inherent contrast for tree-covered
mountains, -0.87 for shade and -0.72 in direct sun, have been shown to be
18
reasonable estimates .
c) Target Observation Angle: Target observation angle becomes important if
telephotometer data are to be used to derive visual range . This derivation
assumes that the sky radiance at the target and observation point are equal
and the average attenuation coefficient between the observer and target is the
same as that between the observer and a distance equal to the visual range.
Because scattering coefficient generally decreases with elevation, viewing an
object at some observation angle greater than 0° induces an error in the cal-
culation of visual range. Although it is recognized that all targets will not
be able to be viewed horizontally, it is recommended that viewing angles be
kept less than 3° from the horizontal plane.
3.3.1.5 Measurement Frequency: When siting permits, continuous instrumenta-
tion should be used. Telephotometer measurements of contrast and spectral
apparent target radiance should be made at the four different wavelengths con-
tinuously during daylight hours. Hourly averages should be computed. When
using a manual tel ephotometer, measurements of contrast and color should be
made at least three times a day: 9:00 a.m., 12:00 noon, 3:00 p.m. local time.
However, more measurement are desirable. At the time the measurements are
made, a color photograph should be taken of each vista that is being used as
a telephotometer target. Additionally, a record of time of day the measure-
ment was made, cloud cover and target condition (snow cover, sun illumination,
etc.) must be kept. Measurements should be made daily, irrespective of
weather conditions, as long as the targets are visible.
31
-------
3.3.2 Scattering Measurements
Scattering coefficient provides a measure that is directly related to the
primary cause (particulate scattering) of visibility impairment.
3.3.2.1 Instrumentation: The fundamental instrument for scattering measurement
is the integrating nephelometer. The integrating nephelometer should be a
single wavelength instrument with peak response at 550 nm, with a FWHM of 100 nm.
3.3.2.2 Sensitivity: As presented in section 3.3.1.2, the instrument should
be sufficiently sensitive to predict a contrast change of 0.004. Figure 8 shows
the uncertainty in C as a function of scattering coefficient error (Ab/b) or
visual range error (AV /V ) for various atmospheric fine particulate concentra-
tions (FPC) and for a black target (C = -1) located 50 km from an observation
point. The horizontal dashed line represents the acceptable error in the meas-
ured scattering coefficient to yield the sought after uncertainty of apparent
contrast of 0.004 or less. Currently available integrating nephelometers claim
an overall accuracy of +10 percent of reading, over an operating range 0 to
0.25 km" . For the case shown in Figure 8, this would not be within the
required accuracy to predict an apparent vista contrast change as small as 0.004
o
in a pristine atmosphere (FPC < 8 yg/m ). An error in the measured scattering
coefficient of 10 percent would correspond to an uncertainty of 0.03 in predicted
target contrast under Rayleigh scattering conditions. The same error corresponds
to a contrast uncertainty of only 0.005 under conditions equivalent to a FPC of
3 19
16 pg/m . It is important to recognize these limitations when interpreting
nephelometer data. An additional problem with currently available nephelom-
eters is zero drift. Corrective action for zero drift is discussed in section
3.3.2.4.
-------
0.08
0.0
0.05 0.1 0.15 0.2 0.25 0.3 0.35
Error in Scattering Coefficient or Visual Range
(Ab/b or
Figure 8. Uncertainty in Predicted Apparent Contrast Resulting from Extinction
Coefficient ^rror.
Plot of uncertainty in predicted apparent contrast resulting from
error in measured scattering coefficient for various atmospheric
fine particlate concentrations. The horizontal dashed line
corresponds to the acceptable error in scattering coefficient that
would yield a contrast change of 0.004 or less.
33
-------
3.3.2.3 Site Selection: Site selection for the nephelometer is critical.
The nephelometer, being a point measurement, will be insensitive to horizontal
pollutant gradients or inhomogeneities. If the monitoring objective is to
document regional haze, the site may be collocated with the telephotometer.
If, however, the objective is to measure a distinct or channeled plume then
a site or sites must be selected to be representative of plume impact. For
Western Class I areas this is a difficult task. The nephelometer is also very
sensitive to local sources, i.e., automobiles, space heaters, generators,
etc.19. Consequently, it must be located in an area free from local traffic
or population impact.
3.3.2.4 Probe Siting: Probe siting criteria should follow that specified for
particulate non-criteria pollutants under "Ambient Monitoring Guidelines for
Prevention of Significant Deterioration (PSD) .
3.3.2.5 Operations: When operating a nephelometer in an area of high relative
humidity (greater than 60 percent), care must be taken not to modify the aero-
20
sol as it passes through the inlet sampling tube (see Figure 9) . Specifi-
cally, the optional heater/dryer should not be used. Experience has shown
that the sensitivity of the instrument can be increased by : 1) controlling
the temperature at which the nephelometer operates and 2) monitoring the zero
point drift. Temperature can be maintained by housing the nephelometer in an
insulated, temperature controlled shelter. Zero drift can be monitored by
routinely recording the nephelometer output when its sample chamber has been
purged with clean air. The clean air reference system (Fiqure 10) consists of
a blower (brushless motor) with preceding and following glass fiber filters.
A clock timer interrupts the nephelometer sampling pump and engages the ex-
ternal blower, which is connected to the sample inlet through a tee fixture .
34
-------
M
CO
CO
II
00
3-
2-
1-
50%
Relative Humidity
100%
Figure 9. Effect of RtNative Humidity on Scattering Coefficient.
This figure shows the response of scattering coefficient to
relative humidity obtained for a number of locations in the
West and Midwest. The cross hatched area includes the entire
range of observations^.
35
-------
IT)
Clean Air Inlet
(PVC Pipe)
Sample Inlet
(Metal Pipe)
Figure 10. Nephelometer Clean Air Reference System.
The clean air reference system consists of a timer, relay, two glass
fiber filters, and a brushless blower. The blower must create a
sufficient air flow to purge the integrating nephelometer with
clean air despite the fact that the nephelometer is still open to
its own sample inlet.
36
-------
3.3.2.6 Frequency: The nephelometer is a continuous monitoring instrument.
Data should be reduced to hourly average values.
3.3.3 Photographic Measurements:
Photography provides a means for documenting overall vista appearance,
and the effect of NOg on plume color.
3.3.3.1 Instrumentation: The camera system should be equipped with a 135 mm
lens with a UV filter. Kodachrome 25 film is appropriate for use. The camera
must have an automatic exposure feature and be operated in the automatic mode.
3.3.3.2 Site and Target: The site arid target selection criteria are the same
as for the telephotometer.
3.3.3.3 Frequency: In conjunction with continuous telephotometers, a photograph
should be obtained of each vista a minimum of three times per day (morning, noon,
and afternoon). At manual telephotometer sites, photographs should be obtained
concurrent with contrast measurements.
3.3.4 Particulate Measurement
The value of collecting particulate samples in conjunction with visibility
monitoring comes from the development of statistical relationships between
particle characteristics and visibility.
3.3.4.1 Instrumentation: As a minimum, particles should be collected in two
size ranges corresponding to the two ambient aerosol modes (Figure 5). The
37
-------
Q
size ranges selected by EPA for the Inhalable Participate Monitoring Network
(15 to 2.5 pm and less than 2.5 ym) can be used for this purpose. There are a
number of size-segregating particle samplers available. These include physical
and virtual impactors, cyclones, and series filtration samplers. In selecting
a sampler, a number of things must be considered: 1) sample configuration and
collection medium or substrate must be compatible with the anticipated analysis
2) collected sample mass must also be compatible with the analysis techniques
of choice. Most manufacturers of sampling equipment include with their specifi-
cations a list of compatible analysis techniques.
3.3.4.2 Site Selection: The same criteria should be used for particulate
sampling site selection as for the nephelometer site.
3.3.4.3 Probe Siting: Probe siting criteria should follow that specified for
particulate non-criteria pollutants under "Ambient Monitoring Guidelines for
Prevention of Significant Deterioration (PSD) .
3.3.4.4 Frequency: Ideally sampling would be performed continuously for the
entire period of visibility data gathering, with sample durations of twenty-
four hours or less. The low concentrations in many visibility protected areas
coupled with the modest flow characteristics of most size-segregating samplers
and the sensitivities of many analytical techniques may dictate sample periods
of two or three days. Beyond a three-day sample duration many interesting
fine features of the aerosol temporal distribution are lost. Sampling should
at least be conducted twice weekly.
3.3.4.5 Analysis: The mass concentration of aerosols in various size ranges
38
-------
can be used to estimate visibility levels, however, this use is not suggested
since it is not nearly as reliable as direct optical measurements. More
important is the use of particulate data to establish statistical relationships
between particle characteristics and visibility. Often the composition and
morphology of particles can be used to identify their sources. In other words,
particle characterization data is used to form a bridge between ambient visi-
bility and pollution emission sources.
The characterization of the particle samples is the key to relating visi-
bility to sources. The more completely the samples are characterized, the
better the chance of defining the relationship. There are a variety of ana-
lytical techniques suitable for aerosol samples. X-ray fluorescence, neutron
activation and atomic absorption techniques are available for elemental analysis.
There are many chemical and instrumental techniques to analyze for individual
chemical compounds and ions. X-ray diffraction can be used to identify the
crystalline nature of samples. Optical and electron microscopy plus individual
particle elemental analysis by electron and ion microprobe can be powerful
tools for source identification. Because of its relatively low cost, high
information yield and non-destructive nature, a multi-element X-ray fluores-
cence type of analysis is recommended. When applied to all samples collected,
it provides a large data S3t for statistical and other interpretive schemes.
It also can be employed as a screening tool to identify samples for subsequent,
more complete, analysis. When the influence of a specific source (or sources)
is suspected the analysis scheme should be tailored to characteristics of the
particles emitted.
39
-------
3.3.5 Meteorological Measurement
Guidance for meteorological measurements is found in "Ambient Monitoring
Guidelines for Prevention of Significant Deterioration (PSD) .
40
-------
4. NON-ROUTINE MONITORING
Phase I of the proposed visibility regulation deals specifically with
the requirement to control visibility impairment that can be traced to a single
major source or small group of sources. It specifies that simple monitoring
techniques, such as visual observation or other methods, should be utilized to
"reasonably attribute" visibility impairment to a source. Several monitoring
methods are available to accomplish this. The objectives of such monitoring
would be to 1) demonstrate visibility impairment and, 2) establish the source
of the impairment.
The most straightforward method is human observation, where a plume and its
source are visible from the protected area. A minimum documentation for each
such sighting would include: time and date, observation location, direction and
estimated distance to the plume, and a description of plume apparent size, shape
and color. Color photography is another method to document impairment as reason-
ably attributable to a source. The time, date, observation location, and
sighting direction would have to accompany each observation.
If visual impairment of a protected vista is evident but the source of
the impact is not, other means to demonstrate that a specific source is reason-
ably attributable can be employed. These would include; 1) concurrent visual
observations made from other view points, 2) the use of aircraft observations
or aerial photography, or, 3) the application of airborne instrumental tech-
niques such as lidar, correlation spectrometers, or in situ sensors to establish
plume source and continuity. These data should be used to augment conventional,
ground based visibility monitoring data.
41
-------
5. QUALITY ASSURANCE FOR VISIBILITY MONITORING DATA
5.1 Introduction
The objective of the quality assurance program discussed here is to con-
trol and assess characteristics of the visibility data collected. Unfortunately,
many assessment techniques, such as would be used to determine precision and
accuracy, are still in the research or testing stage of development. Thus,
many of the procedures described here are control oriented and documentary in
nature. As such they represent the minimum quality assurance program for an
organization operating a visibility monitoring station or network to produce
acceptable monitoring data. (In this discussion, "organization" is defined
as a source owner/operator, a government agency, or a contractor who operates
a routine visibility monitoring network.) If an organization has or wants to
develop a quality assurance program more extensive than the one described here,
EPA encourages them to do so.
A more comprehensive discussion on quality assurance is available in
99
Volume I of the EPA Quality Assurance Handbook . Several aspects of quality
assurance unique to visibility monitoring, such as the criteria for choosing
monitoring locations and targets and specifications of instruments, have
already been discussed in this guideline. Additional aspects of both internal
and external quality assurance programs will be discussed in this section.
5.2 Internal Quality Assurance Procedures
Internal quality assurance is the responsibility of the organization per-
42
-------
forming the monitoring. It consists basically of using standardized and docu-
mented procedures when carrying out essential activities such as:
Installation and testing of equipment
Calibrations
Zero/span checks (as appropriate)
Control checks (as appropriate)
Corrective actions resulting from out-of-control conditions (as
appropriate)
Preventive and remedial maintenance
Recording and validating data
Documentation of quality control information.
Each visibility monitoring project should develop a project-specific operation
manual which details procedures for each of the activities listed above. This
manual shou.ld be based on manufacturer's instrument manuals and the detailed
information on span checks, calibrations, etc., discussed in Section 5.4.
Similarly, each project should also have a monitoring project description.
In the case of PSD monitoring, this description will be suitable for review by
the permit granting authority and the FLM prior to collection of data. In
other types of monitoring, it will serve as a summary of project information
provided by the monitorug organization when its data are released to other
users. The minimum contents of this description, listed in Table 2, generally
follow those required for a PSD ambient air monitoring plan, but with one key
difference. Target specifications, in addition to monitoring site specifications,
are necessary. Project review and data use will be facilitated by the informa-
tion provided by the monitoring description.
43
-------
TABLE 2. MINIMUM CONTENTS OF VISIBILITY MONITORING PROGRAM DESCRIPTION
I. NETWORK DESCRIPTION
Topographical description
Land-use description
Climatological description
Wind roses (if available)
Topographical map of area showing proposed and existing sources and
environs as appropriate with monitoring sites and targets marked
Rationale for visibility monitoring equipment locations
II. MONITORING SITE DESCRIPTION (Sites with fine particulate monitor, or
nephelometer, or meteorological instruments)
Universal Transverse Mercator (UTM) coordinates and elevation above sea
level
Height of probe(s) above ground
Distances from obstructions
Distances from pollutant sources such as generators or roadways
Photographs of each site, of each cardinal direction looking out from
probe, of the nephelometer intake, and of the temperature-humidity
sensor
III. OBSERVATION POINT AND TARGET DESCRIPTION
Universal Transverse Mercator (UTM) coordinates and elevation above sea
level for each observation point and target
Photographs of each vista, with telephotometer targets identified
Description of target surface features such as vegetation, rock color, etc.
Elevation angle from horizontal, azimuth angle from true north, and dis-
tance from observation point to target
IV. INSTRUMENT DESCRIPTION
Manufacturer make and model number, principle of operation
Age of instrument
Description of calibration system
V. SAMPLING PROGRAM DESCRIPTION
Time periods for which measurements will be made
Discussion of the use of existing data or model results in lieu of and/or
in addition to collecting ambient data
VI. QUALITY ASSURANCE PROGRAM
Internal quality control procedures
Calibration frequency; precision and accuracy calculation procedures
External audit program
44
-------
5.3 External Quality Assurance Procedures
External quality assurance for a project generally consists of two parts:
a systems audit for the total project, and performance audits for the individual
instruments.
The system audit includes a variety of activities designed primarily to
identify solutions to problems inherent in the entire data gathering/reporting
system. The key activity associated with the systems audit is a critical review
of the operational procedures employed. This review should be performed by
a person who is intimately familiar with proper visibility monitoring procedures
yet is not associated with the design or operation of the project in question.
Performance of this review requires monitoring site and laboratory evaluations,
interviews with the station operators, laboratory technicians, and managerial
personnel, and an inspection of log books, the procedures manual, and other
pertinent documentation. Another activity associated with the systems audit
should be the submission of test data into the data handling system to insure
that proper procedures are followed throughout the entire data system. The
systems audit should be performed on an annual basis, or at the beginning and
end of the project. The systems audit report and the organization's response
to it should be provided to the permitting authority along with the PSD moni-
toring data, or if the rr.. nitoring is for other purposes, the report and
response should be made available to data users.
Instrument performance audits should be conducted for telephotometer,
nephelometer, and particulate measurements. The specific auditing procedures
are described in section 5.4. Performance audits should be conducted annually,
or at the beginning and end of the project, whichever is more frequent.
45
-------
Results of the performance audits should accompany the corresponding data
report.
5.4 Data Assessment Techniques
Each visibility monitoring instrument requires specific data quality
control and evaluation procedures. Table 3 lists the recommended assessment
techniques, their frequency and the information that is obtained. The eval-
uation techniques listed in the table for external audits and internal procedures
are the same in some cases. However, external audits are to be conducted
annually, by someone not involved in routine operations, .using standards different
from those normally used in the project. Internal procedures are to be conducted
by regular project personnel. External audits should also be performed at the
beginning and end of a project. It is recognized that these procedures are not
as comprehensive as the precision and accuracy determinations required by
40CFR58 and the PSD regulations and guidelines for ambient air monitoring.
However, these interim guidelines may be revised as instrument evaluations
continue.
5.4.1 Telephotometer
Both external and internal methods to assess telephotometer data are
listed in Table 3. The external telephotometer instrument audit involves
side-by-side field performance comparisons between the routine monitoring
instruments and a calibrated instrument of the same type operated by the
person performing the audit. Target and sky radiance values and contrast
values for all targets routinely observed should be obtained with both the
field and audit instruments. The person operating the auditing instrument
46
-------
TABLE 3. DATA ASSESSMENT TECHNIQUES FOR VISIBILITY MONITORING INSTRUMENTS
Instrument
Method
Frequency
Result
Telephotometer
Nephelometer
Side by side, 1n-the-f1e1d com-
parison of field telephotometer
with auditor's telephotometer
External audit
Internal procedure
Camera
Fine Partlculate
Sampler'and Analy-
ses
Annual1y
Every 6 months
External audit with NBS traceable Annually
standard light source
Internal field test and calibra-
tion with NBS traceable light
source
Internal procedure: check and
align telephotometer optics
check physical Integrity of filters
External. 1 point audit with
with Freon-12" or Freon-22",
according to Instrument manu-
facturer's specification
Internal procedure: sample
clean, filtered air to determine
zero drift
Internal field span check and
calibration with Freon-12" or
Freon-22", according to instru-
ment manufacturer's specifica-
tions
Photograph color chart and gray
scale under standard lighting
conditions. Check for color uni-
formity for each roll of film.
Every 6 months
Every 6 months
Annually
Percent differences of radiance and
contrast values at 4 wavelengths
Accuracy as a percent difference
of radiance values at 4 wavelengths
Precision as a percent difference
of radiance values at 4 wavelengths
Properly aligned Instrument,
filter integrity verified
Accuracy as a percent difference
Every six hours or less Correct data appropriately
Every 2 weeks
At beginning of each
roll of film
External audit with NBS-traceable Three weights annually
weights
Meteorological
Sensors
External audit of flowrate with
flowrate audit device
Internal/external evaluation:
participate In interlaboratory
comparisons of analytical results
and audits when available
Internal procedure: construct a
control chart ,jr Initial flow-
rates fpr sampler; establish
appropriate control limits
Internal procedure: construct
a control chart for measurements
of a known weight
Calibrate sensors (same as PSD
requirements)
Annually
Annually
Use flowrate value for
each sample collected
(minimum of 2 per week)
Weigh sample after each
set of 20 filters or
before and after weigh-
ing session, whichever
Is more
Precision as a percent difference
between known and measured scat-
tering coefficient
Qualitative comparison
Weighing accuracy as a percent
difference
Accuracy as a percent difference
in flowrates
Report summarizing differences
between laboratories
Invalidate data outside control
limits
Invalidate data outside control
limits; reweigh filters
Every 6 months or less New calibration factors to use
as needed 1f appropriate
47
-------
should not be the station operator. For each filter, results of these com-
parisons are summarized as percent differences. Percent difference, d^, is
defined as
d. = Yf1e1Yd " Xaud1t x 100 9.)
1 Aaudit
where Yf. ld and Xaudit represent target contrast, and target and sky radiance
values measured for each filter by the field and auditing instruments respec-
tively. In addition, as part of the audit, a portable, standard National
Bureau of Standards (NBS) traceable light source should be used to check the
calibration of the telephotometer for each filter. The resulting accuracy data
should be expressed as percent differences between calibrator radiance and tele-
photometer-measured radiance values.
Two of the three internal procedures for the telephotometer are similar
to those for external audits. In the first procedure, the field telephotometer
is calibrated at one point for each filter using an NBS traceable standard lamp;
in the second, the field telephotometer is compared to another calibrated tele-
photometer as described previously. When the instrument is calibrated, the
response to the standard lamp should be recorded for each filter before any
instrument adjustments are made. Instrument precision for each wavelength can
then be computed using equation 9 in the same manner as for accuracy. The
results of this comparison are presented as percent differences. In the third
procedure, the internal alignment of the telephotometer should be checked and
adjusted every 6 months according to the manufacturer's instructions. At the
same time, the filters should be examined for physical integrity. A record of
these calibrations, instrument checks, and alignments should be maintained for
48
-------
each telephotometer. These three procedures should be conducted twice a year.
5.4.2 Nephelometer
Three data assessment procedures for the nephelometer are mentioned in
Table 3. The first procedure, the external nephelometer instrument audit,
involves introducing a test gas into the instrument to evaluate the response.
Freon-12™ or Freon-22", as specified by the instrument's manufacturer, should
be used. The audit should be conducted by someone other than the station
operator. A percent difference is calculated between the known scattering
coefficient of the standard gas and the instrument-measured scattering coef-
ficient using equation 9. The second, an internal assessment procedure, con-
sists of sampling filtered air a minimum of every 6 hours so corrections for
zero drift can be made. The last is a combination one point field span check
similar to the external audit above, and a calibration according to the
instrument manufacturer's specification. This is to be conducted once every
2 weeks. The span check values should be determined before the calibration
adjustments are made, and a percent difference should be calculated using
equation 9. This percent difference should be reported with the quarterly
data.
5.4.3 Camera
One data assessment procedure for the camera is given in Table 3. It
consists of photographing a standard color chart and gray scale under controlled
lighting conditions at the beginning of each roll of film. A controlled light-
ing condition is achieved by photographing the color chart and gray scales
49
-------
indoors, using an electronic flash located approximately a meter to the side
O O
of the camera to minimize glare . It is important to fill as much of the field
of view as possible with the gray and color scales so densitometer readings of
the slides may be taken at a later time (if desired). At present the tech-
nique involves only a visual review of the processed photographs of color
charts. A second photograph should be taken to document the film roll
number, site name, date, and time.
5.4.4 Particulate Measurement
Five techniques to assess different aspects of particulate measurement
are described in Table 3. The external particulate sampling audits involve
flowrate audits in the field and weighing audits for the laboratory. A "blind"
flow audit device is provided to the station operator who tests the sampler
and returns a sampler flowrate value and the associated flow audit device
value. The flowrate accuracy results should be presented as percent differences
between measured and known flowrates. In the weighing audit a set of three
"blind" masses is provided to the laboratory where each mass is weighed. The
result, the weighing accuracy, should be presented as percent differences
between known and measured weight values.
Several internal procedures are recommended for particulate sampling. To
document the status of the pump in the sampler, a control chart should be con-
structed using initial flowrate for each sample. Details on the construction
of control charts and the setting of control limits are discussed in Appendix H
of Reference 22. Data collected with flowrates outside the control limits should
be invalidated. To document the status of the balance used to weigh the filters,
a control chart based on measurements of a known mass should be used. The test
50
-------
mass should be weighed after each set of 20 filters is weighed, or before and
after a weighing session, whichever is greater. Filter weights determined when
the test mass weight is outside the control limits should be invalidated, and
the filters reweighed.
Finally, to check the analytical results, the laboratory should partici-
pate in interlaboratory comparisons and external audits where available for the
parameters of interest. Results of these comparisons should be summarized
and reported annually.
5.4.5 Meteorological Measurement
Procedures for assessment of meteorological measurements are described
in the PSD Ambient Monitoring Guidelines . In general, the guidelines call
for sensor calibrations every 6 months or more often as needed.
51
-------
6. DATA REPORTING MANAGEMENT AND INTERPRETATION
As discussed previously, nephelometer data (scattering coefficient) and
particulate data are most useful in ascertaining the cause of visibility
impairment while telephotometer measurements (contrast) are most useful in
quantifying the effect of visibility impairment. Visual range is a useful
parameter for interrelating scattering and contrast and to indicate atmospheric
clarity to the lay person.
Table 4 outlines, for either a telephotometer or nephelometer, the rela-
tionships necessary to calculate apparent target contrast, contrast change,
visual range and scattering/extinction coefficient. Only radiance measured at
550 nm should be used in these calculations. Data gathered at other wavelengths
will be used to assess vista color change when EPA has determined the most
appropriate formulae.
Terms in Table 4 which weren't explicitly defined previously are:
b = aerosol scattering coefficient
s ,a
bext a = extinction coefficient (less Rayleigh)
bray,h = Raylei9h scattering coefficient at altitude h of the
?4
observation point
0.01 km" is defined as the standard Rayleigh atmosphere
Calculations of visual range are all standardized to the standard Rayleigh
atmosphere.
52
-------
TABLE 4. VISIBILITY RELATED VARIABLES
Measures
Apparent
target
contrast
Contrast
change
Visual
Range
Extinction
coefficient
(less Rayleigh)
Aerosol
scattering
coefficient
Tel ephotometer (two point)
Apparent Sky Radiance ( N )
Apparent target Radiance (.N )
tNr " sNr
r - — Mn^
L ... (LV)
r sNr
C = C - C „., where
' y (12)
r.ray ~ o e ray'
v 3.9 (14)
V U/r)ln(C0/Cr)-bray>h + 0.01
, (16)
b = - In (C /C ) -b .
ext.a r v o' r' ray,h
Not applicable
Nephel ometer
Scattering coefficient (bf )
s,a
Cr = CQ e"(bray,h + bs,a) (11)
C=Cr-Cr,ray **ere (13)
r.ray o e ray>
„ _ 3.9 (15)
V bs>a + o.oi
Not applicable
b (measured) (17)
5,0.
-------
An example calculation using these equations follows:
Assume measurements are made at 550 nm and that:
Site/target characteristics
target distance r = 50 km
b . = 0.008 km
ray.h
CQ = -0.87 (shaded forest)
Telephotometer measurements yield
.N = 7 units of radiance
t r
N = 10 units of radiance
Nephelometer measurement yields
br a = 0.011 km"1
s,a
Then for a telephotometer:
Apparent target contrast
Cr = ~10~ = "°*3
Contrast change
C = _o.87e"(0-008km~1)(50km) = -0.58
r ,ray
AC = -0.3 - (-0.58) = 0.28 (Seven times perceptible limit)
Visual Range
v = 3^9
r [(1/50 km) In (-0.87/-0.3)] - 0.008 km"1 + 0.01 km"1
= 167 km
-------
Extinction Coefficient less Rayleigh
bext a = C(1/5° km) ln (-°-87/-°-3)3 - °-008 km"1 = 0.013 km"1
While for a nephelometer:
Apparent Target Contrast
C = (-0.87) e ~(°'008 km"1 + 0.011 km"1) 50 km = _Q .
Contrast Change
C = -0.87 e-
r,ray
AC = -0.34 - (-0.58) = 0.24 (six times perceptible limit)
Visual Range
Vr = (3.9)/(0.011 km"1 + 0.01 km"1) = 186 km
Aerosol Scattering Coefficient
be a = 0.011 km"1
s,a
Similar calculations can be carried out using data derived from a scanning
telephotometer; however, in addition, the contrast of a plume or layered haze as
seen against the sky or vista background can be calculated.
When extinction coefficient or apparent contrast are measured at a number
of locations and are to be intercompared or intracompared, readings should be
converted to visual range using the appropriate equations from Table 4. In
doing this, however, it is important to recognize the significance of back-
ground radiance to the interpretation of apparent target contrast in terms of
visual range. A target with a bright cumulus cloud behind it will indicate
55
-------
a higher contrast (larger visual range) than a clear sky situation. Therefore,
such comparisons should be considered only for those scenes which are cloud
free or have a cloud cover which results in uniform illumination (such as high
cirrus clouds or uniform overcast).
The proposed visibility regulation suggests that the visibility analysis
dealing with existing or new sources be done on a seasonal basis. Seasonal
statistical analysis of visual range, apparent target contrast change, and
layered haze contrast should include maximum, minimum, mean, geometric mean,
and standard deviation values, and a cumulative frequency distribution. Visual
range data should be plotted as a log normal probability plot while apparent
contrast should be presented on a normal probability plot.
Examples of hypothetical cumulative frequency plots (see Figures 11, 12,
and 13) for visual range, apparent target contrast change, and plume contrast
are included for reference. The visual range and contrast plots were
developed from telephotometer measurements of sky and target radiance for a
target 103 km distant from the observation point.
There will be days when certain telephotometer targets are not visible
and visual range and change in contrast cannot be calculated. However, if
target distance is, for example, 60 km and the target is not visible, the
visual range is evidently less than 60 km and the contrast change has reached
a maximum. This observation then becomes a data point that can be inte-
grated into the cumulative frequency distribution. Consequently, when a
target has a high occurrence of not being visible, the approximate geometric
or arithmetic mean derived from the probability plots is more meaningful than
the mean calculated analytically.
56
-------
600
500
400
300-
200-
Number of Standard Deviations From Median
21012
(0
DC
To
100-
80
60
50
40
30-
20-
10
0.1 1 5 102030 50 70809095
Cumulative Frequency (%)
99 99.9
Figure 11. Log Probability Cumulative Frequency Distribution for Visual Range.
This figure shows the percent occurrence of visual ranges equal to
or less than the specified value. Visual ranges equal to or less
than 200 km occur 50 percent of the time.
57
-------
o
3
•o
0)
M
2
§
O
0.50
0.45
0.40
0.35
0.30
0.20
0.15
0.10
0.05
Number of Standard Deviations From Median
321012
0.1
5 10 2030 50 70809095
Cumulative Frequency (%)
99 99.9
Figure 12. Normal Probability Cumulative Frequency Distribution for Apparent
Vista Contrast.
This figure shows the percent occurrence of vista contrast equal to
or less than the specified value. A vista contrast of equal to or
less than 0.20 occurs 50 percent of the time. A vista contrast
change of 0.04 may be perceptible.
58
-------
CO
2
o
O
3
E
0.40-
0.35-
0.30-
0.25
0.20
0.15-
0.10-
0.05
0.00
Number of Standard Deviations From Median
3210123
0.1
5 102030 50 70809095
Cumulative Frequency (%)
99 99.9
Figure 13. Normal Probability Cumulative Frequency Distribution for Plume
Contrast.
This figure shows the percent occurrence of plume contrast equal
to or less than a specified value. A plume contrast equal to or
less than 0.10 occurs 50 percent of the time. A plume contrast
as low as 0.02 may be perceptible.
59
-------
7. REFERENCES
1. Malm, W. C. "Summary of Visibility Monitoring Workshop." July 12-13,
1978. Available from author at EMSL-LV, U.S. Environmental
Protection Agency, Las Vegas, Nevada 89114.
2. Protecting Visibility, An EPA Report to Congress. EPA-450/5-79-008.
U.S. Environmental Protection Agency. 1979.
3. Middleton, W. E. K. Vision through the Atmosphere. Univ. of Toronto
Press. Toronto, Ontario, Canada. 1952.
4. Malm, W. C. "Considerations in the Measurement of Visibility." J. Air
Poll. Cont. Assoc., 29, 10, 1042-1052. 1979.
5. Malm, W. C. and Eric G. Walther. "A Review of Instruments Measuring
Visibility-Related Variables." EPA-600/480-016. Environmental
Protection Agency, Las Vegas, Nevada. 1980.
6. Duntley, S. Q., A. R. Boileau, and R. W. Preisendorfer. "Image Transmission
by the Troposphere I." J. Optical Soc. Amer. 47(6):499. 1957.
7. Tombach, I. "Intercomparison of Visibility Measurement Methods,"
In: Proceedings of Conference, View on Visibility—Regulatory and
Scientific, Air Pollution Control Association, 1979. pp. 197-221.
60
-------
*. THJonls, J. and D. Shapland. "Existing Visibility Levels 1n the U.S."
Prepared by Technology Service Corporation for U.S. Environmental ,»
Protection Agency, Grant No. 802815, Research Triangle Park, NC.
1979.
9. Rhodes, C. E. "Protocol for Establishment of a Nation Wide Inhalable
Particulate Network." May 15, 1979. Available from author at
EMSL-RTP, U.S. Environmental Protection Agency. Research Triangle
Park, NC 27711.
10. Whitby, K. T. "Modeling of Atmospheric Aerosol Particle Size Distributions."
A Progress Report on EPA Research Grant No. R800971, Sampling and
Analysis of Atmospheric Aerosols. Particle Technology Laboratory
Report No. 253. Environmental Division, Mechanical Engineering
Department, University of Minnesota, 1975.
11. "Ambient Monitoring Guidelines for Prevention of Significant Deteriora-
tion." EPA-450/2-78-019 (OAQPS No. 1.2-096) U.S. Environmental
Protection Agency, Research Triangle Park, NC. 1978.
12. Cole, H. S. "Guidance for NAQTS: Review of Meteorological Data Sources.
U.S. Environmental Protection Agency (OAQPS), Research Triangle
Park, N.C. December 1977.
13. "Guideline on Air Quality Models." U.S. Environmental Protection Agency
(OAQPS), Research Triangle Park, N.C. October 1980.
61
-------
14. Malm, W. C., K. K. Leiker, and J. V. Molenar. "Human Perception of Visual
Air Quality." In: Proceedings of Conference, View on Visibility--
Regulatory and Scientific, Air Pollution Control Association, 1979.
pp 36-69.
15. Malm, W. C., K. K. Kelle.y, J. V. Molenar, and T. Daniels. "Human
Perception of Visual Air Quality (Regional Haze)." In: Proceedings
of -Grand Canyon Research Symposium on Visibility, Nov. 1980
(Available from EPA-EMSL, Las Vegas, NV).
16. Bergstrum, R. W., K. E. O'Dell, and W. C. Malm. "Calibration and Error
Analysis of Multiwavelength Telephotometers." In: Proceedings of
Conference, View on Visibility—Regulatory and Scientific, APCA,
1979, pp. 165-180.
17. Robert, E. M. "Photometric Measurements of Visibility Using Non-black
Objects." In: Proceedings of Conference, View on Visibility--
Regulatory and Scientific, APCA, 1979, pp. 148-164.
18. Malm, W. C., E. Walther, K. O'dell, and M. Kleine. "Visibility in the
Southwest." Paper presented at 2nd Conference on Scientific Research
in the National Parks, San Francisco, CA. Nov. 26-30, 1979.
19. Malm, W. C., S. Archer, and M. Pitchford. "Comparison of Electro-optical
Measurements Made by Various Visibility Monitoring Instruments," In:
Proceedings of Conference, View on Visibility—Regulatory and
Scientific, Air Pollution Control Association, 1979. pp 222-242.
62
-------
20. Covert, D. S. "A Study of the Relationship of Chemical Composition and
Humidity to Light Scattering by Aerosols." Ph.D. dissertation,
University of Washington, Seattle, WA. 1974.
21. Archer, S. F. "Visibility Investigative Experiment in the West."
Prepared by Northrop Services Inc. for U.S. Environmental Protection
Agency, Contract No. 68-03-2591. Las Vegas, NV 89114. pp 20-22.
22. Quality Assurance Handbook for Air Pollution Measurement Systems; Volume I
—Principles. EPA-600/9-76-005. U.S. Environmental Protection Agency,
Research Triangle Park, NC, 1976.
23. Macbeth Color Checker Color Rendition Chart (Available through photographic
supply stores).
24. Elterman, L. 1968: UV, Visible, and IR Attenuation for Altitudes to
50 Kilometers. Rep. AFCRL-68-0153, Air Force Cambridge Research
Laboratories, Bedford, Mass, 1-49. [NTIS No. AD 671933]
62
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO. ' 2
EPA 450/2-80-082
4. TITLE AND SUBTITLE
Interim Guidance for Visibility Monitoring
7. AUTHOR(S)
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Environmental Monitoring Systems Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Las Vegas, Nevada
12. SPONSORING AGENCY NAME AND ADDRESS
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, N.C. 27711
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
November 1980
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
I
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This report is designed to summarize the substantial information available regarding
visibility monitoring methods presently in use. It does not specify a reference
method, but recommends measures for interim visibility monitoring.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS fa. IDENTIFIER
Aerosols Class I
Monitoring Monitori
Nitrogen oxides
Sulfur oxides ;
Visibility
.
S/OPEM ENDED TERMS C. COSATI Field/Group
Areas
ng
g O' •Tri.p-JfiC N '?" ATEME'-i"
Release to public
,19 SE- UhiTY CLASS !i'.iis'Re"por: 1 '-21. NO. O!= P~AGES
1 __Unclaj>slfJ_ed i 63
_
2. PRICE
..' ^>oN .- c B:-OLC~ f.
------- |