United Stales Environmental Sciences Research
Environmental Protection Laboratory
A9encV Research Triangle Park NC 27711
EPA-600/8-80-018
May 1980
Research and Development
User's Guide for
HI WAY —2
A Highway Air
Pollution Model
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the SPECIAL REPORTS series. This series is
reserved for reports which are intended to meet the technical information needs
of specifically targeted user groups. Reports in this series include Problem Orient-
ed Reports, Research Application Reports, and Executive Summary Documents.
Typical of these reports include state-of-the-art analyses, technology assess-
ments, reports on the results of major research and development efforts, design
manuals, and user manuals.
EPA REVIEW NOTICE
This report has been reviewed by the U.S. Environmental Protection Agency, and
approved for publication. Approval does not signify that the contents necessarily
reflect the views and policy of the Agency, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/8-80-018
May 1980
USER'S GUIDE FDR HTWAY-2
A HIGHWAY AIR POLLUTION MODEL
William B. Petersen
Meteorology and Assessment Division
Research Triangle Park, North Carolina 27711
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
-------
DISCLAIMER
This report has been reviewed by the Environmental Sciences Research
Laboratory, U. S. Environmental Protection Agency, and approved for publication,
Mention of trade names or commercial products does not constitute endorsement
or recommendation for use.
AUTHOR'S AFFILIATION
The author, William B. Petersen, is on assignment with the U.S. Environ-
mental Protection Agency from the National Oceanic and Atmospheric Administra-
tion, U.S. Department of Commerce.
ii
-------
PREFACE
HIWAY-2 is intended as an update to the HIWAY nodel. The changes
necessary to update HIWAY were significant enough that a simple change in the
version number seemed inappropriate. The name of the mdoel was changed to
HIWAY-2 (version 80080) to reflect the major changes in the model. These
changes also warranted a rewriting of the user's guide.
The User's Guide for HIWAY-2 was written so that the model can be easily
executed without a thorough understanding of the mathematical formulation.
Although the User's Guide is complete in itself, the user may wish to avoid
the step of going from the printed page to a computer source program by
obtaining the source code on magnetic tape from the National Technical
Information Service, Springfield, VA 22151.
While attempts are made to thoroughly check out computer programs with a
wide variety of data, errors occur occasionally. In case this model needs to
be corrected, revised, or updated revisions will be distributed in the same
manner as this report. If your copy was obtained by purchase or special order,
you may obtain revisions as they are issued by completing the mailing form on
the last page of this report.
Comments and suggestions regarding this document should be directed to
Chief, Environmental Operations Branch, Meteorology and Assessment Division,
Mail Drop 80, U.S. Environmental Protection Agency, Research Triangle Park,
North Carolina 27711.
iii
-------
ABSTRACT
A computer model, called HIWAY-2, is described that can be used to esti-
mate the concentrations of nonreactive pollutants from highway traffic. This
steady-state Gaussian model can be applied to determine air pollution concentra-
tions at receptor locations downwind of "at-grade" and "cut section" highways
located in relatively uncomplicated terrain. For an at-grade highway, each
lane of traffic is modeled as though it were a finite, uniformly emitting line
source of pollution. For the cut section, the top of the cut is considered an
area source. The area source is simulated by using ten line sources of equal
source strength. The total source strength equals the total emissions from the
lanes in the cut.
The air pollution concentration representative of hourly averaging times
at a downwind receptor location is found by a numerical integration along the
length of each lane and a summing of the contributions from each lane, with
the exception of receptors directly on the highway or within the cut, the
model is applicable for any wind direction, highway orientation, and receptor
location. The model was developed for situations where horizontal wind flow
occurs. The model cannot consider complex terrain or large obstructions to
the flow such as buildings or large trees.
iv
-------
CONTENTS
Preface iii
Abstract iv
Figures vi
Tables vi
Abbreviations and Symbols vii
Acknowledgments x
1. Introduction 1
2. Description of Model 3
At-grade highway 3
Calculations 5
Cut section 9
3. Dispersion and Dilution 11
Initial dispersion 14
Aerodynamic drag 1?
4. Computer Aspects and Input Data Preparation 18
Computer model 18
Card input sequence 19
Interactive operation 19
References 24
Appendices
A. Example problem 25
B. Source code listing for HIWAY-2 35
C. Suggestions for improvement
of the EPA-HIWAY model 59
-------
FIGURES
Number Page
1 Overhead view of the geometry of
at-grade highway 4
2 Line source and receptor relationships ..6
3 Method of simulating dispersion from a
cut section 10
4 Vertical dispersion parameter as a
function of downwind distance 12
5 General flow diagram for HIWAY-2 20
6 Input data deck for the batch node of
operation for H3mY-2 21
TABLES
Number Page
1 Values of a and b Used to Compute o_ 13
za
2 Values of c and d Used to Compute e 14
3 Input Data Cards 22
vi
-------
LIST OF ABBREVIATIONS AND SYMBOLS
ABBREVIATIONS
EPA — U.S. Environmental Protection Agency
g — grams
GM — General Motors
hr — hours
km — kilometers
LIE — Long Island Expressway
m — meters
mi — miles
PG — Pasquill-Gifford
sec — seconds
UNAMAP — User's Network for Applied Modeling of Air Pollution
veh — vehicles
SYMBOLS
A — end point of line source
B — end point of line source
c — term to determine e , dependent upon
stability, degrees
d — factor to determine e , dependent
upon stability, degrees
n — Number of reflections in Equation 9 ,dimensionless
vii
-------
a — factor to determine a (depends upon stability and
distance range)
b — exponent to determine a (depends upon
stability and distance range), dimensionless
C — constant related to vehicle traffic
D — line source length, m
EF — emission factor, g veh~^ mi""''
f — point source dispersion function, m~2
H — effective source height, m
L — mixing height, m
4 — distance from point A to point R,S, m
q — emission rate from line source, g m~^ sec~1
Uc — cross-road wind component, m sec~^
U — wind speed corrected for aerodynamic drag, m
R — east coordinate, m
R — east coordinate of point A, m
R — east coordinate of point B, m
O
R — east coordinate of receptor k, m
S — north coordinate, m
S — north coordinate of point A, m
viii
-------
S_ — north coordinate of point B, m
D
S, — north coordinate of receptor k, m
TV — traffic volume, veh hr~1
u — wind speed, m sec~^
x — downwind distance, meters or km
x — normalizing distance, km
y — crosswind distance, meters or km
z — receptor height above ground, m
0 — direction, relative to north, of line from point A
to point B, degrees
0 — wind direction, relative to north, degrees
e — half angle of horizontal plume spreading, degrees
a — standard deviation of the concentration
^ distribution in the crosswind direction, m
o — initial a , m
yo Y
a — standard deviation of the concentration
z distribution in the vertical direction, m
a — initial a » m
zo z
a — vertical dispersion due to the ambient
a environment, m
o — total vertical dispersion, m
^T
o — crosswind dispersion due to the ambient environment,m
^a
a — total crosswind dispersion, m
yT
x — concentration, g m~3
$ — wind angle relative to the roadway, degrees
IX
-------
ACKNOWLEDGMENTS
The User's Guide for HIWAY-2 contains verbatim much of the information
recorded in the User's Guide to HIWAY. The author wishes to acknowledge John
Zimmerman and Roger Thompson for their work. The author also expresses his
appreciation to Dr. S.T. Rao and his staff at the New York State Department
of Environmental Conservation without whose help this document could not have
been written. Sincere appreciation also goes to Bruce Turner and George Schewe
for their comments and review, and to Joan Emory for her assistance.
-------
SECTION 1
INTRODUCTION
The National Environmental Policy Act of 1969 requires any federally
funded highway construction project to be preceded by an impact statement
analyzing the effect of the proposed roadway on air quality. This report
describes a computer program, called HIWAY-2, that calculates air quality
levels of nonreactive pollutants produced by highway automotive traffic at
distances tens to hundreds of meters downwind of the highway in relatively
uncomplicated terrain. HIWAY-2 provides the air quality specialist with
a valuable tool for projecting the air quality impacts of future highway
construction.
In making estimates of pollution concentrations for an "at-grade"
highway, highway emissions are considered to be equivalent to a series of
finite line sources. Each lane of traffic is modeled as though it were a
straight, continuous, finite line source with a uniform emission rate. Air
pollution concentrations downwind from a line source are found by a numerical
integration along the line source of a simple Gaussian point-source plume.
Although most applications of this model will be for ground-level sources and
receptors, and for receptors close bo the source where mixing height will have
almost no effect, the more general case of nonzero source and receptor heights
and inclusion of the effects of mixing height can be considered by the model.
The HIWAY-2 model is similar to the line-source equations (5.19 and 5.20)
in the Workbook of Atmospheric Dispersion Estimates (Turner, 1970) but can
also consider finite line sources at any angle to the wind.
-------
An estimate may also be made of air pollution concentrations downwind of
a "cut section" for which the top of the cut section is considered to be
equivalent to an area source. This area source is simulated by using a
series of ten equal line sources such that the total source strength is equal
to the total pollution emissions of the highway.
No pollution emissions module is included in this version of the model.
A value of the line-source strength, qt(g m"1 sec"1), for each lane of
traffic must be obtained from a separate computation. Line-source strength is
generally a function of traffic rate, average vehicle speed, and traffic mix
(fraction of heavy-duty vehicles, fraction of late models with emission
control devices, etc.). Data input for the HIWAY-2 program can be accomplished
in two ways: {1) through batch mode, with data cards that follow the program
deck (see Section 4 for format), and (2) through continuous mode, i.e.,inter-
actively on a time-share computer terminal. The term interactive refers to
the information exchange between the user and the computer program in asking
and answering questions.
-------
SECTION 2
DESCRIPTION OF MDDEL
AT-GRADE HIGHWAY
A view of an idealized four-lane at-grade highway is shown in Figure 1.
Traffic emissions from each lane are simulated in the computer model by a
straight line source of finite length. As shown in Figure 1 for a four-lane
highway, the location of the highway is specified by the coordinates at the
centerline (from edge to edge) of the highway (points 1 and 2). The ordering
of the lanes is from left to right when one looks from point 1 to point 2.
One lane or any even number of lanes from 2 to 24 can be used in the model.
The width of the highway and its center strip must also be entered as
input data. With this information, the computer program HIWAY-2 will assign a
finite uniform line source to each lane of traffic. These line sources are
placed at the center of each traffic lane.
A uniform emission rate, q , must be specified for each line source.
This line-source emission rate can be found if the emission factor, EF (g
veh~1 mi"1), and the traffic volume, TV (veh nr~1), are known:
. , EF (g veh~1 mi'1) TV (veh hr~1)
q (g sec n m-') = —
* 1609.3(m mr1) 3600 (sechr'1)
= 1.726 x 10-7
-------
RECEPTOR
Figure 1 . Overhead view of the geometry of at-grade highway as seen by the
computer model. The endpoints of the highway are specified by the
by
-------
A value of the emission factor for vehicles can be obtained from the most
current issue of mobile source emission factors (EPA, 1974 and 1978).
CALCULATIONS
The calculation of concentration is made by a numerical integration of
the Gaussian plume point-source equation over a finite length. The coordinates
(meters) of the end points of a line source of length D (meters), representing
a single lane extending from point A to point B (see Figure 2), are %,SA
and %,Sg. The direction of the line source from A to B from the north is
6 (degrees). The coordinates, R,S, of any point along the line at an arbitrary
distance, t (meters), from point A are given by:
R = RA + «, sin B (Eq.2)
S = SA + i cos B (Eq.3)
Given a receptor at Rfc'sk' the downwind distance, x (meters), and the
crosswind distance, y (meters), of the receptor from the point R,S for any
wind direction, e (degrees), is given by:
x = (S - Sk) cos e + (R - Rk) sin e (Eq.4)
y = (S - Sk) sin 9 - (R - Rfc) cos e (Eq.5)
Since R and S are functions of a, x and y are also functions of t. The
concentration, x (gm~3), from the line source is then given by:
q, f
= — I
u J
(Eq>6)
-------
NORTH
t
tsj
n
3
(RA,SA)
WIND
RECEPTOR
(Rk,Sk)
EAST
-------
where: u = wind speed, m
D = line source length, m
f = point source dispersion function (Equations 7 to 9), m~2
For application of this model to a highway segment in relatively open terrain,
an estimate of the wind speed, u, at approximately 2 meters height above
ground is suitable.
For stable conditions, or if the mixing height is > 5000 meters:
exp
2 '
+ exp
' / \21
l(z+H\
_ _ i i
(Eq.7)
where: av = standard deviation of the concentration distribution in
the crosswind direction, m
az = standard deviation of the concentration distribution in
the vertical direction, m
z = receptor height above ground, m
H = effective source height, m
In unstable or neutral conditions, if oz is greater than 1.6 times the mixing
height, L (meters), the distribution below the mixing height is uniform with
height regardless of source or receptor height, provided both are less than
the mixing height:
ex? r T I ~
(Eq.8)
-------
In all other unstable or neutral conditions:
r =-
exp --
r i /z + H
-;
N =
-H - 2NL\2
*z/
1 /z + H - 2NL\ 2 1 Iz - H +
exp--{ - - + exp--(
2 \ *z / 2 \
-------
estimate. If convergence is not reached by the time the number of intervals
Q
reaches 1536 (which is 3*(2) ), the estimated integral value is saved. A
9
new sequence of estimations for intervals equal to 4, 8,..., 4*(2) is performed.
9
Any new integral estimate for interval values of 4, 8,..., 2048 which is 4*(2) )
having a relative error from the saved integral estimate less than 2 percent
signals convergence. If convergence is not obtained after 2048 intervals the
program writes out a message indicating the relative error, the receptor
number, and the line source number. The program then uses the minimum of the
saved integral estimate and the current integral estimate as the integrated
results.
The above evaluation of the integral is repeated for each lane of traffic;
the resulting concentrations are summed to represent the total concentration
from the highway segment.
CUT SECTION
Estimates of air pollution concentrations at locations downwind of a
depressed highway (cut section) can be determined by considering the top of
the cut section to be an area source of pollution (Figure 3). In the model,
this area source is approximated by using ten line sources located at the top
of the cut section. The total emission rate for the highway is first found by
adding together the emission rates for each individual lane of traffic. Then
this emission rate is distributed equally over each of the ten line sources
used to simulate the area source at the top of the cut section. The procedure
used to determine pollutant concentrations downwind of the cut section is then
entirely similar to the procedure used to determine the concentrations for an
at-grate highway. It should be emphasized that these estimates of air pollution
concentrations should be made for receptors downwind of the cut section and
not for locations inside the cut section itself.
-------
WIND
DIRECTION
VERTICAL CROSS SECTION OF
POLLUTION SPREADING FROM
TOP OF CUT SECTION.
••.;.;.;.
X . • •
:^
c
i
¥
"2
iiiiTn.i
;.
x : : : : :
'3
T
";;
'4
*r
-------
SECTION 3
DISPERSION AND DILUTION
Recent studies by Rao et al. (1979, 1980) and Eskridge et al. (1979a,
1979b) have demonstrated that the dispersion near the roadway is dominated by
the turbulence generated by the moving traffic and that the ambient atmospheric
stability plays little role in dispersing the pollutant in the immediate
vicinity of the roadway. The dispersion parameters, ov and oz, indicate
the amount the pollutant plume has spread (dispersed) after leaving its
source. The values for these parameters near the roadway are documented in
Appendix C.
The vertical dispersion parameter, az, was evaluated from the tracer
data collected during the General Motors (GM) Experiment (Cadle et al., 1976)
and the Long Island Expressway (LIE) Experiment (Rao et al., 1978). Figure 4
is a plot of az versus downwind distance. The (Pasquill-Gifford) (PG)
curves are shown for reference. The data indicate that the dispersion downwind
of the highway is typically between stability class A and C, even though the
GM data represented a large nuinber of cases when the atmosphere was stable.
Also, considerable scatter in oz was found between stability classes. For
these reasons, the dispersion in HIWAY-2, is specified to be a function of
only three stability regimes. For PG stability classes A, B, and C, the
unstable curve is used. For stability class D, the neutral curve is used; for
stabilities E and F, the stable curve is used. The user specifies the PG
stability class (A-F) and the appropriate dispersion curve is chosen by the
model.
11
-------
22.0
20.0-
Il8.0
^.16.0
o>
114.0
Il2.0
110.0
v
8- 8.0
"TJ
15 6.0
y
£ 4.0
2.0
GM site
0 Local Pasquill stability
» a
a it
• "
it
it
it
it
C
D
E
F
NY site
Stability B
60°
-------
The total vertical dispersion paramter, o^' is composed of dispersion
due to ambient turbulence plus the initial dispersion due to the turbulence
generated by the vehicles, similar to that suggested by Pasquill (1976).
oz is computed as:
2
is of the form a x b. Table 1 snows the values for a and b for the
three stability regimes used in the model.
TABLE 1. VALUES OF a AND b USED TO COMPUTE oz
Stability Regime
Unstable
Neutral
Stable
a
110.62
86.49
61.14
b
0.93198
0.92332
0.91465
The total horizontal dispersion parameter, ay , is given by:
a - (a 2 + a V/2
-------
The factor 465.1 is 1000 m km~1 divided by 2.15, the number of standard
deviation of a Gaussian distribution from the oenterline to the point where
the distribution falls to 10 percent of the centerline value. The angle
9p is given by:
ep = c - d 1n (X/XQ) (Bq.13)
where c and d (degrees) are functions of stability and the normalizing distance,
XQ, is 1 km. Values of the parameters c and d are given in Table 2.
TABLE 2. VALUES OF c AND d USED TO CALCULATE, 0p
Stability Regimes
Unstable
Neutral
Stable
c
18.333
14.333
12.5
d
1.8096
1 .7706
1.0857
Sufficiently far downwind the atmospheric dispersion process dominates
the dispersion of the plume from the roadway. At 300 meters downwind the
dispersion curves are merged into the PG curves; the unstable curve is
merged into the PG class B curve, the neutral curve into the PG class D curve,
and the stable curve into the PG class E curve. At 300 meters downwind the
dispersion due to the roadway is computed by subtracting the square of the
ambient dispersion parameter (a function of stability class) from the square
2 9
of the total dispersion, o^, or oy^ . The initial dispersion due to the
roadway at 300 meters is then used as the initial dispersion and the ambient
dispersion is determined from the PG curves. cz and oy for distances
beyond 300 meters are then computed in a similar manner to Equations 10 and 11.
14
-------
INITIAL DISPERSION
Turbulence of the air produced by the notion of automobiles results in a
rapid mixing of the pollutants near the highway. This mixing is modeled by
assuming that an initial spreading of the pollutant plume occurs over the
highway. Zimmerman and Thompson (1975), using a limited data base, suggested
that the initial vertical dispersion, o^, should be 1.5 meters. The
value of 1.5 meters is a conservative estimate of the vertical standard
deviation of the plume at the downwind edge of the at-grade highway and was
considered as a tentative value by the authors.
In order to improve the estimate of the initial vertical dispersion,
a2- was calculated as a function of wind speed from the &l data for the
nearest roadside receptor. (See Appendix C). The initial vertical dispersion
parameter is specified as:
CTZO = 3.57 - 0.53 Uc (Eq.14)
The cross-road wind component is indicated by Uc. HIWAY-2 is programmed
such that Uc is computed and used in Equation 14 to estimate a-. However,
the smallest allowable value of a^ is 1.5 meters.
For at-grade highways, the initial horizontal dispersion, oy , has
an arbitrary value of twice the initial vertical dispersion. When a~
equals 1.5 meters, ayQ equals 3.0 meters, the same as in HIW&Y. However,
as a^ increases due to small cross-road winds, oy increases proportion-
ally. The value given to ay has little effect on the computation of air
pollution concentration when the wind direction has a component perpendicular
to the highway. The use of an initial ay accounts for a reasonable amount
of cross-road spreading caused by vehicle-generated turbulence when the wind
direction is parallel or nearly parallel to the highway.
15
-------
\fery few measurements have been published on air quality downwind of a
cut section. Nevertheless, the available data indicate that the cut section
configuration tends to increase the dispersion of the air pollution orginating
from the cut section. This effect occurs particularly when wind speeds are
light, for then the release of heat from combustion, the long travel time of
the pollutant to the receptor, and mechanical turbulence produced by the
cut section highway aid the dispersion. Thus, for the cut section case,
based upon very limited data, the initial s's for wind speeds less than 1 m
sec~1 were set at 10 meters for Oy and 5 meters for oz. it was assumed
that for wind speeds greater than 3 m sec"^ the cut section did not enhance
the initial dispersion. Therefore, 3 meters for Oy and 1.5 meters for az
were used. For speeds between 1 and 3 m sec"1, the initial sigmas are
linearly interpolated. These initial °'s are assumed for each of the ten
lanes used to represent the cut. The initial values of cry and oz (meters)
are found front:
yo
= 3
for u > 3 m sec~l
(Eq.15)
*yo = 10-7
jr — K. 1'
*ZQ ~ & - J-.
for 1 < u<3 m sec~l
(Bq.16)
and
*yo = 10
'zo = 5
/ for u < 1 m sec"
(Bq.17)
16
-------
AERODYNAMIC DRAG
HIWAY-2 has incorporated in it an aerodynamic drag factor that accounts
for the initial dilution of the pollutant on the roadway, and allows the
model to make reasonable concentration estimates during low wind speed
conditions. Analysis of the GM data revealed that the aerodynamic drag
factor must be a function of the wind-road orientation angle, because the
amount of acceleration in the lower layers is most significant under parallel
wind-road orientation (Appendix C). Hence, an aerodynamic drag factor that
is a function of wind-road angle was developed and incorporated into the
HIWAY-2 model. The relation developed is:
Uc = C u0-164 cos2 (Eq. 18)
u is the ambient wind speed (m sec~1), 41 is the wind-road angle, and C
is a constant related to the traffic speed. It is observed that C equals
1.85 for moderate to high traffic speed conditions. However, for low traffic
speeds data are not available at this time to evaluate the value of C. This
relation takes full effect for parallel wind U = 0) situations but has no
effect for perpendicular wind cases. If the ambient wind speed is less than
the wind speed computed according to the above relation, then only the
corrected wind speed will be applied. If the ambient wind speed is greater
than the corrected wind speed, no changes to the wind speed are made, thus,
allowing correction for only low wind speed situations (ambient wind speeds
less than 2 m sec"1).
17
-------
SECTION 4
COMPUTER ASPECTS AND INPUT DATA PREPARATION
COMPUTER MODEL
The FORTRAN computer program consists of a main program and four sub-
routines. Figure 5 depicts the general flow of the model. The main program
handles input and sets up a separate line source for each lane of traffic.
Subroutine HWYJUJE does the integration and provides printouts of results.
This subroutine calls HWYRCX, which evaluates Equations 7, 8, or 9, or
simplifications of these equations if H or z is zero. Evaluation of ov and
oz are done by subroutines HWYSIG and DBTSIG, which are called from IWYRCX.
The program is capable of processing multiple hours of meteorology for
multiple sources, (see Appendix A).
An east-north coordinate system (see figure 2) is used in the computer
model. The width of the highway and of its center strip, the coordinates of
the centerline of the highway, and the coordinates of the receptor(s) are
input parameters. However, in Equations 4 and 5, x and y refer to a coordinate
system aligned along the wind vector (x, the downwind direction, and y, the
crosswind direction). That system is distinct from the coordinate system
used for locating sources and receptors in the model.
In the basic equations given earlier (Equations 2 to 5), units of the
coordinate system have been specified as meters for dimensional balance.
However, for practicality units of the computer coordinate system, are in
kilometers. The user may use any convenient highway map unit if he enters an
appropriate scaling factor to convert those units to kilometers. For example,
18
-------
if units of meters are desired for highway coordinates, the scale factor
should be entered as 0.001. This section contains a list of the input variables,
including a brief description of each of the units by which the input parameters
must be expressed. An example of input data, as well as the output of a run
made with the example input data, is given in Appendix A.
CARD INPUT SEQUENCE
The sequence of input data cards is shown in Figure 6. The format of data on
the input cards for the batch mode of operation is given in Table 3. The
coordinates of the roadway are in the center of the highway (from edge to
edge). The ordering of the lanes is from left to right when looking from
point 1 to point 2 (Figure 1).
INTERACTIVE OPERATION
Tne self-explanatory listing produced by the model on a remote computer
terminal is shown in Appendix A to illustrate the operation of the model in
an interactive mode. The computer communicates to the user in upper case
letters, while the user replies in lower case letters.
Operation of the model in an interactive mode is similar to batch mode
operation. However, in the interactive mode the data for each line source
must be entered seperately. Thus, the impact of multiple line sources on
air quality is more easily assessed using the batch mode. The model is
capable of assessing the air quality impact for multiple sets of meteor-
ological data in both the batch and interactive modes.
19
-------
r
READ
PROBLEM
TITLE
r.
READ
METEOROLOGY
r
READ
SCALE
FACTOR
^ READ
RECEPTOR
COORDINATES
r
READ
SOURCE
DATA
r
READ
EMISSIONS
DATA
n
READ
CONTROL FOR
CUT SECTION
YES
HOURLY
OUT PUT
FOR ALL
SOURCES
Figure 5. General flow diagram for HIWAY-2.
20
-------
CUT SECTION
CARD
(7)
EMISSIONS
CARD
(6)
5, 6, 7 CARD SEQUENCE
FOR EACH SOURCE. ANY
NUMBER OF SOURCES.
SOURCE
CARD
(5)
RECEPTOR
CARDS !
(4)
SCALE
FACTOR
CARD
(3)
METEOROLOGY
CARD
(2)
TITLE
CARD
(1)
Figure 6. Input data deck for the batch node of operation for HIWAY-2.
Card type nunbers are in parenthesis.
21
-------
3. INPUT DATA CARDS
Name
Card type 1
Head
Card type 2
THETA
u
HL
XKST
Card type 3
OS*
Card type 4
cards)
XXRRb
XXSR
Z
Card type 5
of cards
REP1C
SEP1
REP2
SEP2
H
WIDTH
CNTR
XML
Card type 6
cards )
QIS
Card type 7
Columns
(1 card)
1-80
(1 card)
1-10
11-20
21-30
31-40
(1 card)
1-10
(up to 50
1-10
11-20
21-30
(any number
1-0
11-20
21-30
31-40
41-50
51-60
61-70
71-80
(up to 3
1-80
( 1 card,
Format
20A4
F10.0
P10.0
F10.0
F10.0
F10.0
F10.0
F10.0
F10.0
F10.0
F10.0
F10.0
F10.0
F10.0
F10.0
F10.0
F10.0
F10.0
Form
AAAA
XXX.
XX. X
XXXX.
X.
X.
XXXX. XXX
XXXX. XXX
XX.
XXXX. XXX
XXXX. XXX
xxxx.xxx
XXXX. XXX
xx.x
XX.
XX.
X.
.xxxxxxx
Variable
Alphanumeric data
for heading
Wind direction
Wind speed
Height of mixing layer
Pasquill stability
class
Scale factor
East coordinate of
receptor
North coordinate
of receptor
Height (above ground)
of receptor
East coordinate,
point 1
North coordinate
point 1
East coordinate,
point 2
North coordinate,
point 2
Height of line source
Total width of highway
Width of center strip
Number of traffic lanes
Emission rate for
each lane
Units
-
Degrees
m sec"1
Meters
-
-
Map units
Map units
Meters
Map units
Map units
Map units
Map units
Meters
Meters
Meters
—
g secern"1
can be blank for
at grade)
CUT
WIDTC
1-10
11-20
F10.0
F10.0
X.
XX.
1, if cut; 0, if at
grade
Width of top of cut
section
-
Meters
scale factor converts map units to kilometers:
If map units in kilometers, scale factor = 1.0
If map units in meters, scale factor = 0.001
If map units in feet, scale factor = 0.000305
If map units in miles, scale factor = 1.61
lto begin again with another set of data, a value 9999. is punched for XXRR (card type 4)
following the last receptor card. A value of -9999. for XXRR will cause the program to
terminate after data set.
cAny number of sources can be input. Card types 5-7 must be used for each source.
IF REP1=9999. end of source data, card types 6 and 7 should not follow.
22
-------
A complete listing of the computer code is given in Appendix B with
verification of HIV^Y-2 demonstrated in Appendix C. The HIWAY-2 model has
been placed on the EPA User's Network for Applied Modeling of Air Pollution
(UNAMAP) computer system. For information on this system contact: Chief,
Data Management, Meteorology Laboratory, U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina 27711.
23
-------
REFERENCES
Cadel, S.H., D. P. Chock, J. M. Heuse and P. R. Monson 1976. Results of the
General Motors Sulfate Dispersion Experiments, General Motors Research
Publication, GMR-2107.
Eskridge, R.E., and J.C.R. Hunt 1979a. Highway Modeling: Part I -
Prediction of Velocity and Turbulence Fields in the Wake of Vehicles.
J. Appl. Meteor. 36: 387.
Eskridge, R.E., F.S. Binkowski, J.C.R. Hunt, T.L. Clark and K.E. Demerjian
1979b. Highway Modeling Part II - Advection and Diffusion of SF5 Tracer
Gas. J. Appl. Meteor. 36: 401.
Pasquill, F. 1976. Atmospheric Dispersion Parameters in Gaussian Plume Modeling
Part II. Possible Requirements for Changes in the Turner Workbook Values.
EPA-600/4-76-030br U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina. 43 pp.
Rao, S.T., M. Chen, M. Keenan, G. Sistla, R. Peddada, G. Wotzak, and N. Kolak
1978. Dispersion of Pollutants Near Highways - Experimental Design and
Data Acquisition Procedures. EPA-600/4-78-037, U.S. Environmental Protection
Agency, Reseach Triangle Park, North Carolina. 56 pp.
Rao, S.T., M. Kennan, F. Sistla, and P. Sampson 1979. Dispersion of Pollutants
Near Highways, Data Analysis and Model Evaluation. EPA-600/4-79-011, U.S.
Environmental Protection Agency, Research Triangle Park, North Carolina.
158 pp.
Rao, S.T. Final Report of Grant 806017-01. Atmospheric Turbulence and Pollutant
Dispersion Near Highways. 1980. (Unpublished)
U.S. Environmental Protection Agency 1974. Automobile Exhaust Emission Modal
Analysis Model. EPA-460/3-74-005, U.S. Environmental Protection Agency,
Ann Arbor, Michigan.
U.S. Environmental Protection Agency 1978. Mobile Source Emission Factors.
EPA-400/9-78-005, U.S. Environmental Protection Agency, Washington, D.C.
Turner, D.B. 1970 Workbook of Atmospheric Dispersion Estimates. AP-26,
U.S. Environmental Protection Agency, Research Triangle Park, North
Carolina. 84 pp.
Zimmerman, J.R., and R.S. Thompson 1975 User's Guide for HIWAY, A Highway Air
Pollution Model, EPA-650/4-74-008, U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 59 pp.
24
-------
APPENDIX A
EXAMPLE PROBLEM
INTRODUCTION
In order to clarify the procedure for using both the batch and interactive
(continuous) versions of the HIWAY-2 model, the following test problem is
solved using both versions.
Given: Length of highway - 5 km.
Orientation - east-west.
Number of lanes - four.
Road width (edge to edge) - 46 meters.
Median width - 30 meters.
Emission rate in each lane from south to north
0.0112, 0.0103, 0.0106, and 0.0156 g sec"1 m~1.
Wind direction - 42 degrees.
Wind speed - 3.7 m sec"^.
Stability class - 3.
Find: The expected concentration at receptors along a line
perpendicular to the center of the highway segment at
distances 1, 5, 10, 30, and 50 meters from the downwind
edge of the roadway (1) if the road is an at-grade section,
and (2) if the road is a cut section with the top of the cut
being 50 meters in width.
SOLUTION USING THE INTERACTIVE VERSION
Assuming that you have already logged on the computer, etc.,you are then
given the choice of receiving a description of the model. Following that,
enter the input parameters as the model calls for them. Most of them are
self-explanatory; however, a few comments are in order:
25
-------
1. When entering the mixing height, never use the value 0.
2. If you do not want the effect of a limit to vertical mixing in your
calculation, use a large enough mixing height so that there is no
chance of its influencing your results, such as 5000 meters.
3. The coordinates for the ends of the roadway segment are assumed to
be in the center of the road (from edge to edge).
4. The ordering of emission rates is for lanes in order from left to
right when looking from point 1 to point 2.
An example problem using the interactive version of HIWAY-2 is shown in
Table A-1. The results for the at-grade section are given following the
entry of receptor coordinates. For convenience, the center of the roadway
has been placed 0.023 km north of the origin in this example, so that the edge
of the road is on the axis and the y coordinate of the receptor is the
distance from the edge of the road. The roadway and receptors could have
been placed at any location.
The option to run the model for a new receptor location (DOC), change
the road type (TYPE), or to end the program (END) is given after the results.
In the second part of the problem, the road type (cut), the width (50
meters), and the location of the road (to again place the edge of the road at
a "y" coordinate value of zero) are changed. The results for the cut section
are shown following the entry of data. Note that the concent rations are in
micrograms per cubic meter (UGM/M**3). The part per million (PPM) column is
a conversion from micrograms per cubic meter for the pollutant carbon monoxide.
The part per million column would be incorrect for any other pollutant.
SOLUTION USING THE BATCH VERSION
The batch version requires at least seven input cards. Depending upon
the number of receptor points, the number of sources, and number of problems
26
-------
to be run, nore cards may be necessary. The format for each card is given in
Table 3. Table A-2 lists the input for the example problem; Table A-3 lists
the results. Note that for the cut section, the sixth and seventh fields
(oolunns 51 to 70) in card type 5 were left blank. Also, note that for card
type 4, XXPR can have a value of -9999. or 9999. A -9999. indicates the end
of the problem after card types 5-7 are read in. A value of 9999. indicates
a new problem follows after card types 5-7. Card types 5-7 are always read in
sequence. Any number of sources can be input using the 5,6,7 card sequence to
describe the source and emissions. If REP1 = 9999., then the last source
has been read in and card types 6,7 are not read. The program will print out
the concentration estimates from each line source and the total contribution
from all the line sources.
27
-------
TABLE A-1. EXAMPLE PROBLEM USING INTERACTIVE VERSION OF HIWAY-2.
DO YOU WANT A DESCRIPTION C! ';'," FPA "-U^iCf-il" !•
iteFOKE APPLYING IT?(YES 01; lit:)
>yes
l.TilE EPA "HIWAY-2" MODEL COMPILES INEhT POLLUTANT CONCENTRATIONS IN THE
VICINITY OF A ROADWAY ON A SHORT TERM BASIS (HOURLY /WERAGES) USING
THE GAUSSIAN PLUME FORMULATION. IF MORE THAN ONE ROADWAY IS PRESENT,
SUPERPOSITION APPLIES. THE MODEL CAN EE USED FOR AT GRADE AND CUT
SECTIONS FOR RECEPTOR DISTANCES OF TENS TO HUNDREDS OF METERS DOWNWIND
OF THE LINE SOURCE IN RELATIVELY UNCOMPLICATED TERRAIN.
2.THE COORDINATE SYSTEM IS ARRANGED SUCH THAT THE X-AXIS INCREASES FROM
WEST TO EAST WHILE THE Y-AXIS INCREASES FROM SOUTH TO NORTH.THE UNITS
RELATED TO HIGHWAY MEASUREMENTS ARE INDICATED BY A SCALE FACTOR OF
USER UNITS TO KILOMETERS. THE MOST FREQUENTLY USED FACTORS ARE:
UNITS SCALE FACTOR
KILOMETERS 1.0
METERS 0.001
FEET 0.000305
MILES 1.61
SCALE FACTOR UNITS APPLY EXCEPT WHEN OTHER UNITS ARE SPECIFICALLY
REQUESTED.
3.THE EMISSION DATA IS DEPENDENT ON VEHICLE SPEED,TYPES AND NUMBER OF
VEHICLES,AND EMISSION CONTROL DEVICES. EMISSIONS (GM/SEC*M) ARE
ENTERED IN ORDER FROM LEFT TO RIGHT WHEN LOOKING FROM ROAD END
PT 1 TO END PT 2.
4.ROAD COORDINATES ARE THE ENDPOINTS OF THE HIGHWAY CENTER LINE.
WIND DIRECTION IS DERIVED BY MEASURING CLOCKWISE(EAST) FROM
DUE NORTH.(E.G..WIND FROM NORTH IS 0 DEGREES,"EASTERLY WIND IS 90.)
5.THE PROGRAM CONTAINS THE OPTION TO EVALUATE ANY NUMBER OF
RECEPTOR LOCATIONS AND/OR TYPES OF ROADS.
6.YOU MUST SEPARATE MULTIPLE INPUTS WITH COMMAS.
7.FOR MOST APPLICATIONS,THE HEIGHTS OF THE RECEPTOR AND
SOURCES ARE ASSUMED TO 3E THE SAME.
DO YOU WANT A DESCRIPTION OF THIS VERSION OF "HIWAY-2"?(YES OR NO)
>no
ENTER PROBLEM TITLE OF 61 CHARACTERS OR LESS
> example of interactive version of HIWAY-2
ENTER SCALE FACTOR (KILOMETERS/USER UNIT).
>1.
ENTER LIKE(ROAD) ENDPOINTS.(ORDERED PAIRS:X1,Y1,X2,Y2)
28
-------
TABLE A- 1 . ( cont inued )
2. 5, .023, -2. 5, .023
ENTER EMISSION HEIGHT. (METERS)
>0.
ENTER WIND DIRECTION (DEC). NORTH IS ZERO.
ENTER WIND SPEED (METERS/SEC).
>3.7
ENTER MIXING HEIGHT (METERS).
>1000.
ENTER PASQUILL-TURNER STABILITY CLASS (1-6).
>3
ENTER THE NUMBER OF LANES.
>U
ENTER LINE SOURCE STRENGTH VECTOR. (A VALUE FOR EACH LANE)
>. 0112, .0103, .0106, .0156
IS THIS A CUT SECTION? (YES OR NO)
>no
ENTER HIGHWAY WIDTH (METERS).
ENTER WIDTH OF CENTER STRIP (METERS).
>30.
ENTER NUMBER OF RECEPTOR LOCATIONS DESIRED. (MAXIMUM OF 25)
>5
ENTER RECEPTOR COORDINATE SETS.(X&Y IN SCALE FACTOR UNITS ;Z IN METERS)
>0. ,-. 001 , 0., 0. ,-. 005, O.,0.,-. 010,0.,
>0., -. 030, 0.,0.,-. 050,0.
HIWAY-2 VERSION: 80080
example of interactive version of HIWAY-2
ENDPOINTS OF THE LINE SOURCE
2.500, .023 AND -2.500, .023
EMISSION HEIGHT IS .000 METERS
EMISSION RATE (GRAMS/SECOND*METER) OF H LANE(S)
.112-001 .103-001 .106-001 .156-001
WIDTH OF AT-GRADE HIGHWAY IS M6.000 METERS
WIDTH OF CENTER STRIP IS 30.000 METERS
WIND DIRECTION IS 42. DEGREES
WIND SPEED IS 3.7 METERS/SEC
STABILITY CLASS IS 3
HEIGHT OF LIMITING LID IS 1000.0 METERS
THE SCALE FACTOR IS 1.0000 KM/USER UNIT.
29
-------
TABLE A-1. (continued)
RECEPTOK LOCATION HEIGHT CONCENTRATION
X Y Z (M) UGM/M**3 PPM*
.0000 -.0010 .0000 3258. 2.834
.0000 -.0050 .0000 3137. 2.730
.0000 -.0100 .0000 2631. 2.292
.0000 -.0300 .0000 1546. 1.345
.0000 -.0500 .0000 1106. .962
PPM CONCENTRATIONS CORRECT FOR CARBON MONOXIDE ONLY.
YOU HAVE THE OPTION TO RUN THE MODEL FOR A NEW RECEPTOR LOCATION
(LOG),OR TO CHANGE THE ROADWAY TYPE,OR TO END THE PROGRAM.
ENTEH LOG, OR TYPE, OR END.
>type
ENTER LINE(ROAD) ENDPOINTS.(ORDERED PAIHS:X1,Y1,X2,Y2)
>2.5,.025,-2.5,.025
ENTER EMISSION HEIGHT. (METERS)
>0.
ENTER WIND DIRECTION (DEC). NORTH IS ZERO.
>42.
ENTER WIND SPEED (METERS/SEC).
>3-7
ENTER MIXING HEIGHT (METERS).
>1000.
ENTER PASQUILL-TURNER STABILITY CLASS (1-6).
>3
ENTER THE NUMBER OF LANES.
>4
ENTER LIUE SOURCE STRENGTH VECTOR.(A VALUE FOR EACH LANE)
>.0112,.0103,.0106,.0156
IS THIS A CUT SECTION? (YES OR NO)
>yes
ENTER WIDTH OF TOP OF CUT. (KETEHS)
>50.
ENTER NUMBER OF RECEPTOR LOCATIONS DESIRED.(MAXIMUM OF 25)
>5
ENTER RECEPTOR COORDINATE SETS.U&Y IN SCALE FACTOR UNITS;Z IN METERS)
>0.,-.001,O.,0.,-.005,0.,
>0.,-.010,0.,0.,-.030,0.,
>0.,-.050,0.
30
-------
TABLE A-1. (continued)
HIWAY-2 VERSION:
example of interactive version of HIWAY-2
ENDPOINTS OF THE LINE SOURCE
2.500, .025 AND -2.500, .025
OMISSION HEIGHT IS .000 DETERS
EMISSION RATE (GRAKS/SECOND*METER) OF 4 LANE(S)
.112-001 .103-001 .106-001 .156-001
WIDTH OF TOP OF CUT SECTION IS 50.000 METF.RS
WIND DIRECTION IS 42. DEGREES
WIND SPEED IS 3.7 METERS/SEC
STABILITY CLASS IS 3
HEIGHT OF LIMITING LID IS 1000.0 METERS
THE SCALE FACTOR IS 1.0000 KM/USER UNIT.
RECEPTOR LOCATION HEIGHT CONCENTRATION
X Y Z (M) UGM/M*»3 PPM*
.0000 -.0010 .0000 3023. 2.630
.0000 -.0050 .0000 2740. 2.384
.0000 -.0100 .0000 2343. 2.039
.0000 -.0300 .0000 1465. 1.274
.0000 -.0500 .0000 1076. .937
* PPM CONCENTRATIONS CORRECT FOR CARBON MONOXIDE ONLY.
ENTER LOG, OR TYPE, OR END.
>end
31
-------
12345678
12345678901234567890123456789012345678901234567890123456789012345678901234567890
CO
to
EXAMPLE OF BATCH VERSION OF HIWAY-2
42.
1.
0.
0.
0.
0.
0.
-9999.
2.5
.0112
0.
2.5
.0112
1
9999.
3.7
-.001
-.005
-.010
-.030
-.050
0.023
.0103
0.025
.0103
50.
1000.
0.
0.
0.
0.
0.
-2.5
.0106
-2.5
.0106
3.
0.023 0.0 46.0 30.0 4.
.0156
0.025 0.0 4.
.0156
12345678901234567890123456789012345678901234567890123456789012345678901234567890
12345678
TABLE A-2. CARD INPUT FOR EXAMPLE PROBLEM
-------
TABLE A-3. EXAMPLE PROBLEM USING BATCH VERSION OF HIWAY-2.
EXAhPLE OF hATCH VEHSiCf! OF I1IWAY-2
HIWAY-2 VERSION: 80080
WlilD DIRECTION IS 42. DEGREES
WIND SPEED IS 3.7 METERS/SEC
STABILITY CLASS IS 3
HEIGHT OF LIMITING LID IS 1000.0 METERS
THE SCALE OF' 7HF COORDINATE AXES IS 1.0000 KM/USER UNIT.
ENDPOINTS OF LIME SOURCE 1
2.500, .023 AND -2.500, .023
EMISSION HEIGHT IS .000 MFTEHS
EMISSION RATE (GKAMS/SECO!lD*METER) OF 4 LAHE(S)
.H2-001 .103-001 .106-001 .155-001
WIDTH OF AT-GRADE HIGHWAY IS 46.0 M
WIDTH OF CENTER STRIP IS 30.0 K
RECEPTOR LOCATION HEIGHT CONCENTRATION
X Y Z(M) UGM/HETER»»3 PPM *
.0000
.0000
.0000
.0000
.0000
-.0010
-.0050
-.0100
-.0300
-.0500
.0000
.0000
.0000
.0000
.0000
3258.
3137.
2634.
1546.
1106.
2.834
2.730
2.292
1.345
.962
* PPM CONCENTRATIONS CORRECT FOR CARBON MONOXIDE ONLY.
EMDPOINTS OF LINE SOURCE 2
2.500, .025 AMD -2.500, .025
EMISSION HEIGHT IS .000 METERS
EMISSION RATE (GRAV.S/SECOND»METER) OF 4 LAME(S)
.112-001 .103-001 .106-001 .156-001
wIDTH OF TOP OF CUT SECTION IS 50.000 M
RECEPTOR LOCATION HEIGHT CONCENTRATION
X Y Z(K) UGM/KETER*»3 PPM *
.0000 -.0010 .0000 J023. 2.630
.0000 -.0050 .0000 2740. 2.384
.0000 -.0100 .0000 2343. 2.039
.0000 -.0300 .0000 1465. 1.274
.0000 -.0500 .0000 1076. .937
* PPK CONCENTRATION'S CORRECT FOR CARBON MONOXIDE ONLY.
33
-------
TABLE A-3. (continued)
TOTAL CONCENTRATION FROM ALL 2 LIKE SOURCE(S)
RECEPTOR LOCATION HEIGHT CONCENTRATION
X Y Z(M) UGM/METEK**3 PPM *
.0000 -.0010 .0000 6281. 5.464
.0000 -.0050 .0000 587«. 5.114
.0000 -.0100 .0000 4977. 4.330
.0000 -.0300 .0000 3011. 2.619
.0000 -.0500 .0000 2183. 1.899
* PPM CONCENTRATIONS CORRECT FOR CARBON MONOXIDE ONLY.
34
-------
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
APPENDIX B
SOURCE CODE LISTING
C HI.VAY-2 - NEW VERSION - /ARCH 1930
C THIS PROGRAM CALCULATES THE CONCENTRATION FROM A LIKE SOURCE
COKKON /SOL/ 3LN(25) ,HLN(25) ,RAQ(25) ,SAQ(2r;) ,l'BQ(25)
, CLSST(50)
IVERS=80080
IRD=5
IWRI=6
FORM OF INPUT TO HIWAY (BATCH)
13HIOOOOO
BHIOOQ10
VARIABLE
NAME
COLUMNS
FORMAT
C
C
C
C
C
C
C
C
C CARD TYPE 1 (1 CARD) HEADER OR TITLE CARD
C HEAD 1-80 20A4 AAAA ALPHANUMERIC DATA FOR HEADING.
C
C CARD TYPE 2 (1 CARD) METEOROLOGICAL CARD
,o!3Q(25),CO:-K5BHI00020
BHIU0030
BHI00040
31:100050
BHIOOObO
BHI00070
RH100080
BHI00090
BHI00100
BHIOQ110
BHI00120
BHI00130
FORM
VARIABLE
BHI00150
BHI00160
BHIOQ170
UNITS BHI00180
BHI00190
BHI00200
BHI00210
BHI00220
BHI00230
WIND DIRECTION (DEGREES)BHI00240
WIND SPEED (METERS)HII00250
HEIGHT OF MIXIMG LAYER (METERS)BHI00260
PASQUILL STABILITY CLASS(DIMENSIONLESS)BHI00270
BHI00280
CARD TYPE 3 (1 CARD) SCALE FACTOR (MAP 'JMITS TIMES SCALE FACTOR = KM)BHI00290
GS 1-10 F10.0 X.XXXX SCALE FACTOR BHI00300
THETA 1-10 F10.0 XXX.
U 11-20 F10.0 XX.X
HL 21-30 F10.0 XXXX.
XKST 31-40 F10.0 X.
C CARD TYPE U (UP TO 50 CARDS) RECEPTOR CARDS
BHI0031Q
XXRR 1-10 F10.0 XXXX.XXX EAST COORD. OF RECEPTOR (MAP UNITS)BH100320
XXSR 11-20 F10.0 XXXX.XXX NORTH COORD. OF RECEPTOR (MAP UNITS)BHI00330
Z 21-30 F10.0 X.XX HEIGHT OF RECEPTOR (ABOVE GROUND) (METEBHI00340
BHI00350
XXRR ON LAST RECEPTOR CARD SHOULD HAVE A VALUE OF -9999.
OR 9999.. IF XXRR = -9999. END OF PROBLEM . IF XXRR = 9999.
BEGIN A MEW PROBLEM AFTER CARD TYPES 5-7 ARE READ IN.
CARD TYPE 5 d CARD) SOURCE CARD
REP1 1-10 F10.0 XXXX.XXX EAST COORD..POINT 1
SEP1 11-20 F10.0 XXXX.XXX NORTH COORD..POINT 1
REP2 21-30 F10.0 XXXX.XXX EAST COORD..POINT 2
SEP2 31-40 F10.0 XXXX.XXX NORTH COORD.,POINT 2
H 41-50 F10.0 XX.X HEIGHT OF LINE SOURCE
WIDTH 51-60 F10.0 XX. TOTAL WIDTH OF HIWAY
CMTR 61-70 F10.0 XX.
XML 71-80 F10.0 X.
BHI00360
BHI00370
BHI00380
BHI00390
BHI00400
(MAP UNITS)BHIOOM10
(MAP UtlITS)BHIOOU20
(MAP UNTTS)BHI00130
(MAP UNITS)BHIOOUMO
(METERS) 311100450
(METERS )RH 1.00460
WIDTH OF CENTER STRIP (MEDIAN) (METEFS)HIIOOU70
NUMBER OF TRAFFIC LANES (DIMENSIONLESS)BHIOCW80
BHIOM90
35
-------
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
10
c
c
c
c
20
c
c
c
c
30
50
CARD TYPES 5-7 ARE ALWAYS READ IN SEQUENCE. ANY NUMBER BHI00500
OF SOURCES CAM BE INPUT USING THE 5,6,7 CARD SEQUENCE BHI00510
TO DESCRIBE THE SOURCE AND EMISSIONS. IF REP1 = 9999. I3HI00520
LAST SOURCE. CARD TYPES 6,7 ARE HOT READ. BHI00530
BH100540
CARD TYPE 6 (UP TO 3 CARDS) EMISSIONS FOR EACH LAME. BHI00550
LANES ORDERED LEFT TO RIGHT WHEN LOOKING FROM POINT 1 TO POINT 2BHI00560
QLS 1-80 F10.0 .XXXXXXXXX EMISSION RATE FOR FACH LAME (G/SEC-KOBHI00570
BHI00580
CARD TYPE 7 (1 CARD) AT-GRADE OR CUT? (CAM BE BLANK FOR AT-GRADE) BHI00590
CUT 1-10 F10.0 X. 1, IF CUT; 0, IF AT-GRADE(DIMENSIONLESS)BHI00600
WIETC 11-20 F10.0 XX.
WIDTH AT TCP OF CUT SECTION
READ HEADER CARD
READ (IRD,230,END=400) HEAD
WRITE (IWRI,240) HEAD
WRITE (IWRI,250) IVERS
READ (IRD.290) THETA,U,HL,XKST
KST=XKST
THETA IS THE WIND DIRECTION IN DEGREES.
U IS THE WIND SPEED IN METERS PER SECOND.
KST IS THE STABILITY CLASS
HL IS THE HEIGHT OF THE LIMITING LID
WRITE (IWRI.260) THETA,U,KST,HL
READ (IRD,290) GS
GS IS THE MEASURE BETWEEN COORDINATES (KM).
WRITE (IWRI.270) GS
ICHK=1
Nsi
READ (IRD.290) XXRR(N),XXSR(N),Z(N)
IF (XXRR(N).GE.9998.) GO TO 50
IF XXRR = -9999. END OF PROGRAM.
IF XXRR = 9999. NEW PROGRAM FOLLOWS AFTER CARD TYPES 5-7
IF (XXRR(N).LE.-9998.) ICHK=2
IF (XXRR(N).LE.-9998.) GO TO 50
IF (N-52) 40,30,30
WRITE (IWFI,280)
GO TO 400
RR(N)=XXRR(N)*GS
SR(M)=XXSR(N)«GS
ZR(N)=Z(N)
N=N+1
GO TO 20
N=N-1
NLINE=0
DO 60 I=\50
coirr(i)=o.
(METERS)BH100610
BHI00620
BHI00630
BHI00640
BHI00650
nHI00660
BHI00670
BH100680
BHI00690
BH100700
BHI00710
BHI00720
BHI00730
BH 100740
BHI00750
BH100760
BHI00770
BHI00780
BHI00790
BHI00800
BHI00810
BHI00820
BHI00830
BHI00840
BHI00850
BHI00860
BHI00870
BHI00880
BHI00890
BHI00900
BHI00910
BHI00920
BHI00930
BH100940
BHI00950
BHI00960
BH100970
BHI00980
BHI00990
36
-------
60
70
C
C
C
C
C
C
C
C
C
C
C
C
C
C
80
90
100
110
CLSST(I)=0.
READ (IRD,290,Eh'Ds200) 3FP1 .SFP1 ,REP2,SFP2,il, WIDTH,CNTR, XNL
IF (REP1.CE.999E.) 00 TO 200
NLINE=tJLIf!E+i
REP1,SEP1 ARE THE COORDINATES OF AM END POINT OF THE LINE
SOURCE IH SOURCE COORDINATES.
REP2.SEP2 ARE TliF COORDINATES OF THE OTHEh END POINT OF THE
LINE SOURCt It." SOURCE COORDINATES.
K IS THE EFFECTIVE EMISSION HEIGHT OF THE SOURCE IM METERS.
WIDTH IS THE HIGHWAY WIDTH (M) FOR AT GRADE
CNTR IS THE WIDTH OF THE CENTER STRIP (M)
XNL IS THE HUMBER OF LANES FOh THE AT-GRADE HIGHWAY.
IF REP1 = 9999- LAST SOURCE. CARD TYPES 6,7 ARE NOT READ.
WRITE (IWRI.300) NLINE,REP1,SEP1,REP2,SEP2
NLsXML
WRITE (IWRI,310) H
WRITE UWRI.320) NL
READ (IRC,290) (QLSU),I=1 ,NL)
QLS IS THE LINE SOURCE STRENGTH (GRAMS/SECOND*METER)
WRITE (IWRI.330) (QLS(I),I=1,NL)
READ (IRD.290) CUT.WIDTC
CUT SECTION.
WIDTC IS THE WIDTH OF THE TOP OF THE CUT SECTION (M)
IF (CUT.LE.0.0001) GO TO 100
DQLS IS THE CUT SECTION SOURCE STRENGTH
DQLS=0.
DO 80 I=1,NL
DQLS=DQLS-nXS(I)
XMDLslO.
NL=XNDL
DQLS=DQLS/XNDL
WRITE (IWRIo^O) WIDTC
DO 90 1=1,NL
QLS(I)=DQLS
WIETHsWICTC
XNL=XNDL
CNTRsO.
GO TO 110
WRITE (IWRI.350) WIDTH,CNTR
CONTINUE
WRITE (IWRI.360)
REP12=REP1-REP2
SEP12=SEP1-5EP2
RNGL=ATAN2(REP12,SEP12)
RMGL=RMGL*57.2958
IF (RNGL.LT.O.) RNGL=360.-AnS(RHGL)
DHNG=ABS(THETA-RNGL)
IF (DRMG.GT.1oO.) DPN
BHIOtOOO
BHI01010
BH I 01 020
OHIO 10 30
BH I 01 040
DHI01050
OHIO 1060
BHI01070
B11I01080
BHI01Q90
BHI01100
BHI01110
BHI01120
BI1I01130
BHIOH40
BHI01150
BHI01160
BHI01170
BHIOH80
BHI01 190
BHI01200
BHI01210
BHI01220
BHI01230
BHI012MO
BH 10 1250
BHI01260
BHI01280
BHI01290
BHIC1300
BHI01310
BHI01 320
BHI01330
BHI01340
BHI01350
BHI01360
BH 10 1370
OHIO 1380
BHI01390
BHI01400
BHI01420
BHI01430
BHI01440
BI1I01450
BHI01460
BHI01470
B1!I01480
37
-------
IF (DRNG.GT.90.) Dh;JG=''30.-DHMG
ANGLE=0.0175*ADS(DK:!G)
CO=COS(ANGLE)
CS=CO*CO
FU=1.85*CS*(U**0.l6iO
R=FU/U
RFU=1.
IF (R.GT.1.) RFU=R
U=RFU*U
WSP=ABS(U *SIN(ANGLE))
RA=REP1*GS
RB=KEP2*GS
SA=SEP1*GS
SB=SEP2*GS
WL=(WIDTH-CNTR)/XNL
IF (CUT.GT.0.00001) GO TO 130
120 SYON=3.
SZON=1.5
GO TO 150
130 IF (U.GT.3.) GO TO 120
IF (U.LT.1.) GO TO 140
DUh=('J-1.)/2.
SYCN=10.-7.*DUK
SZON=5.-3.5*DUK
GO TO 150
1^40 SYON=10.
SZON=5.
150 CONTINUE
IF (NL.EQ.1) WL=0.
IF (NL.EQ.1) CNTR=0.
DELR=RB-RA
DELS=SB-SA
DIST=SORT(DELS*DELS+DELR *DELR)
NLIM=NL/2
ALIMsNLIM
DO 160 ID=1,NLIM
A=ID
DLs(-0.5)»CNTR+((-l)*ALIM-0.5+A)«VIL
DUMsDL«0.001/DIST
RAQ(ID)=RA+DELS«DUM
RBQ(ID)=RB+DELS»DUM
SAQ(ID)=SA-DELR»DUM
SBQ(ID)=S5-DELR*DUM
QLN(ID)=QLS(ID)
HLN(ID)sH
160 CONTINUE
NSsNLIK+1
AS=NS
DO 1?0 ID=NS,NL
A=ID
BH 10 1500
BHI01510
OHI01520
BHI01530
BH 10 1540
BHI01550
BHI01560
3HI01570
BHI01580
BHI01590
BHI01600
BHI01610
BHI01620
BHI01630
BHI01650
BHI01660
BHI01670
BHI01680
BHIQ1690
BHI01700
BMI01710
BHI01720
3I1I01730
BHI01740
BHI01750
BHI01760
BHI01770
BHI01780
BHI01790
BHI01800
BHI01810
BHI01820
BHI01830
3H 10 1840
BHI01850
BHI01860
BHI01870
BMI01880
BHI01890
BHI01900
BHI01910
BHI01920
BHI01930
BHI01940
BHI01950
Bi 1 10 1960
BHI01970
BH 10 1980
BHI01990
38
-------
DL=0.5*CNTR+(0.5+A-AS)«WL
DUM=DL*O.OOVDICT
RAQ(ID)=RA+DELS*DUK
RBO(ID)=RB+DELS*DUM
SAQ(ID)=SA-DELB*OUM
SBQ(ID)=SE-DELR*DUM
QLN(ID)=QL5(IP)
HLN(ID)=H
170 CONTINUE
K=NL
DO 180 IDUMsl.N
180 CON(IDUM)=0.
C K IS NUMBER OF LIME SOURCES
C N IS NUMBER OF RECEPTORS
CALL HWYLNE (K,M)
WRITE (IWRI.370)
DO 190 1=1, N
CONT(I)=CONT(I)4-CON(I)
190 CLSST(I)=CLSST(I)+CLSS(I)
GO TO 70
200 CONTINUE
IF (NLINE.EQ.1) GO TO 220
WRITE (IWRI.380) NLINE
WRITE (IWRI.360)
DO 210 1=1, N
WRITE (IWRI,390) XXRRtt),XXSR(I),ZR(I),CONT(I),CLSST(I)
CONTINUE
WRITE (IWRI.370)
GO TO (10,400), ICHK
210
220
C
230
2MO
250
260
290
300
310
320
330
340
350
350
PHI02000
BHI02010
RHI02020
3HI02030
BHI02040
BHI02050
B1II02060
BHI02070
BHI02080
BHI02090
EH102100
BHI02110
BHI02120
BHI02130
BHI02140
BHI02150
BHI02160
BHI02170
BHI02180
BHI02190
BHI02200
BHI02210
BHI02220
BHI02230
BHI02240
BHI02250
BHI02260
BHI02270
BHI02280
BHI02290
FORMAT (20AU) BHI02300
FORKAT COf,/,20A4,/) BHI02310
FORMAT CO HIWAY-2 VERSION:',16) BHI02320
FORMAT C WIND DIRECTION IS',F7.0,' DEGREES',/,' WIND SPEED IS1,FBHI02330
17.V KETEKS/SECV STABILITY CLASS IS',15,/,1 HEIGHT OF LIMITING BHI02340
2LID IS',F8.1,' METERS') BHI02350
270 FORMAT C THE SCALE OF THE COORDINATE AXES IS '.F10.V KM/USER UNBHI02360
UT.' //) BHI02370
280 FORKAT (IHO.'THE NUMBER OF RECEPTORS IS LIMITED TO 50. YOU HAVE ATBHI02380
1 TEMPTED TO READ THE 51ST. NO COMPUTATIONS HADE.
FORMAT (8F10.0)
3HI02390
3HI02UOO
FORMAT C ENDPOINTS OF LINE SOURCE' ,13,/,F9.3,' ,' ,F9.3,' AND1,F9.3BHI02410
-, ,',F9.3>
FORMAT C EMISSION HEIGHT IS',F6.3,' METERS')
FORMAT C EMISSION RATE (GRAMS/SECOKD*METER> OF1, IV LANE(S)')
FORMAT (6E12.3)
FORMAT (' WIDTH CF TOP OF CUT SECTION IS',F10.3,
FORMAT C WIDTH OF AT-GRADE HIGHWAY IS',F10.1/
INTER STRIP IS'.FIO.V M',/)
FORMAT (1HU,' HECEPTOh LOCATION HEIGHT
BHI02M20
BHI02M30
BJ1I02440
BHI02450
BHI02460
M',/,' WIDTH OF CEBHI02470
BHI02480
COKCEMTRATIOM',/,BHI02490
M')
39
-------
V XVOX.'Y Z(M) UUM/.'€TER**3 PPM*1) BHI02500
370 FORMAT OHO,'* PPM CONCLIITKATIONS CORRECT FOR CARBON MONOXIDE OULBHI0251Q
1Y.',/////) DHI02520
330 FORMAT (///,5X,'TOTAL CONCENTRATION FROM ALL1,13,' LINE SOURCE(S)'BHI02530
1,/> BHI02540
390 FORMAT (1H , 3(F10.4,2X) ,nO.O,FlO. 3) BHI02550
MOO STOP BHI02560
C BHI02570
END BHI02580
C HIWAY-2 - ME1// UNIVAC INTERACTIVE VERSION - MARCH
C THIS PROGRAM CALCULATES THE CONCENTRATION FROM A
C AT EACH OF A NUMBER OF RECEPTORS.
COMMON /SOL/ QLN(25),HLN(25),RAQ(25),SAQ(25),RBQ(25)
10),CLSS(50),NLIHE
COMMON /REC/ RR(5^) ,SR(5'<) ,ZR(51)
COMMON /WEA/ THETA,U,KST,HL
COMMON /PUT/ XXRR(51),XXSR(51),QLS(25),HEAD(20),IWRI
COMMON /WS/ WSP,SYON,SZON
DIMENSION Z(51), SVF(48), QQLS(25)
DATA YES /'YES V ,YFSL /0171145163040/
DATA RLOC /'LOG V ,RLOCL /01541571430MO/
DATA CHA /'TYPE'/ ,CHAL /0164171160145/
C INITIALIZATION
IVERS=80080
NLINE=1
ISPEC=0
IRD=5
IWRI=6
READ (IHD,210) DESC
* * ABOVE READ IS PROCESSOR DUMMY READ * * *
!•
-MODEL AND VERSION DESCRIPTIONS.
•
WRITE (IWRI.200)
READ (IRD.210) DESC
IF (DESC.EQ.YES.OR.DESC.EQ.YESL) CALL TEXT1
WRITE (IWRI.220)
READ (IRD.210) DESC
IF (DESC.EQ.YES.OR.DESC.EQ.YESL) CALL TEXT2
C *
C—
C™*""
C—
C—
C—
C
-INTERACTIVE INPUT.
WRITE (6,230)
READ (IRD.210) HEAD
1980 HIWOOOOO
LINE SOURCE HIW00100
HIW00200
,SBQ(25),CON(5HIW00300
HIWOMOO
HIW00500
HIW00600
HIW00700
HIW00800
HIW00900
HIW01000
HIW01100
HIWQ1200
HIW01300
HIW01UOO
HIW01500
HIW01600
HIW01700
HIW01800
HIW01900
HIW02000
HIW02-»00
HIW02200
HIW02300
HIW02400
HIW02500
HIW02600
HIW02700
HIW02800
HIW02900
HIW03000
HIW03100
HIW03200
HIW03300
WRITE (IWRI,2'»0)
HIW03500
40
-------
READ (IRD,«) OS
10 WRITE (IWRI.250)
READ (IRD,») REP1,SEP1,REP2,SEP2
WRITE (IWRI.260)
READ (IRD,*) II
WRITE (IWRI.270)
READ (IRD,*) THETA
WRITE (IWRI,2bO)
READ (IFD,«) U
WRITE (IWRI.290)
READ (IRD,*) HL
WRITE (IWRI.300)
READ (IRD,*) XKST
KST=XKST
WRITE (IWRI.310)
READ (IRD,*) HL
XNL=NL
WRITE (IWRI,360)
READ (IRD,*) (QLS(I),I=1,NL)
C STORE DUMMY NL, QLS FOR PRINTOUT.
NNL=NL
DO 50 1=1,NL
50 QQLS(I)=QLS(I)
WRITE (IWRI.370)
READ (IRD,210) CUT
IF (CUT.NE.YES.AND.CUT.NE.YESL) GO TO 60
C CUT SECTION.
C WIDTC IS THE WIDTH OF THE TOP OF THE CUT SECTION (M)
WRITE (IWRI.380)
READ (IRD,*) WIDTC
GO TO 70
60 WRITE (IWRI.390)
READ (IRD,*) WIDTH
WRITE (IWRI.IOO)
READ (IRD,*) CMTR
70 WRITE (IWRI.niO)
READ (IRD,*) NRCO
IF (NRCO.GT.25) NHCO=25
WRITE (IWRI.420)
READ (IRD,*) (XXRR(N),XXSR(N),Z(M),N=<',NRCO)
C—
C LIST PARAMETERS FOR THIS RUN.
C
WRITE (IWRI.430) IVERS
WRITE (6,440) HEAD
WRITE (IWRI,4rjO) REP\SEP1,REP2,SEP2
WRITE (IWRI.160) H
C NLL=NL, QQLS=QLS.
WRITE (IWRI.470) NHL
C QLS IS THE LINE SOURCE STRENGTH (GRAKS/SECOND»METER)
HIW03600
HIW03700
HIW03300
HIW03900
HIW04000
HIWOU200
HIWOU300
HIW04400
HIWOU500
HIW04600
HIW04700
HIWOU800
HIWOU900
HIW05000
HIW05100
HIW05200
HIW05300
HIW05MOO
HIW05500
HIW05600
HIW05700
HIW05800
HIW05900
HIW06000
HIW06100
HIW06200
HIW06300
HIW06400
HIW06500
HIW06600
HIW06700
HIW06800
HIW06900
HIW07000
HIW07100
HIW07200
HIW07300
HIW07400
HIW07500
HIW07600
HIW07700
HIW07800
HIW07900
Krrt'08000
HIW08100
HIW08200
HIW08300
H1W08MOO
HIW08500
41
-------
WRITE UWRI,4<30)
-------
RB=REP2*GS
SA=SEP1*GS
SBaSEP2"GS
WL=(WIDTH-CNTR)/XNL
IF (CllT.EQ.YES.OR.CUT.EQ.YESL) GO TO 130
120 SYGN=3.
SZON=1' .5
GO TO 150
130 IF (U.GT.3.) GO TO 120
IF (U.LT.1.) GO TO 1140
DUM=(U-1.)/2.
SYON=10.-7.*DUM
SZON=5.-3.5*DUM
GO TO 150
140 SYON=10.
SZON=5.
150 CONTINUE
IF (NL.EQ.1) WUU.
IF (NL.EQ.1) CNTR=0.
DELR=RB-RA
DELS=SB-SA
DIST=SQRT(DELS*DELS+DELR *DELR)
NLIK=NL/2
ALIMaNLIM
DO 160 ID=1,NLIM
AsID
DL=(-0.5)»CNTR+((-1)«ALIM-0.54A)»WL
DUM=DL*O.OOVDIST
RAQ(ID)=RA+DELS«DUM
RBQ(ID)=RB+DELS*DUH
SAO(ID)=SA-DELR*DUM
SBQ(ID)=SB-DELR*DUH
QLN(ID)aQLSdD)
HLN(ID)=H
160 CONTINUE
NS=NLIM+1
AS=NS
DO 170 ID=NS,NL
AsID
DL=0.5*CNTR+(0.5+A-AS)*WL
DUHsDL«0.001/DIST
RAQ(ID)sRA+DELS*DUM
RDQ(ID)=RB+DELS*DUM
SAQ(ID)aSA-DELR«DUM
SBQ(ID)sSB-DELR«DUK
QLN(ID)=QLS(ID)
HLN(ID)=!1
170 CONTINUE
DO 180 IDUM=1,NRCO
HIW1 3600
HIW1 3700
HIW1 3800
MIW13900
Hivmooo
HIW11100
HIW1U200
HIW14300
HIW1M400
HIW1H500
HIW14600
HIW1U700
HIW1M800
HIW15000
HIW15100
HIW15200
HIW15300
HIW15MOO
HIW 15500
HIW15600
HIW15700
HIW15800
HIW15900
HIW16000
HIW16100
HIW16200
HIW16300
HIW16100
HIW16500
HIW16600
HIW16700
HIW16800
HIW16900
HIW17000
HIW17100
HIW17200
HIW17300
HIW17400
HIW17500
HIW17600
HIW17700
HIW17800
H1W17900
HIW18000
HIW18100
HIW18200
HIW18300
H1W8UOO
HIW18500
43
-------
RR(IDUM)=XXRR(IDIW)«GS
SR(IHUM)=XX3R(IDUM) *GS
ZR(IDUM)=Z(IDUM)
180 CON(IDUM)=0.
K=NL
NsKRCO
C K IS NUMBER OF LINE SOURCES
C N IS NUMBER OF RECEPTORS
CALL HWYLNE (K.K)
WRITE (IWRI.540)
C—
C RERUN OPTIONS.
C
WRITE (IWRI.550)
IF (ISPEC.GT.l) GO TO 190
WRITE (IWRI.560)
WHITE (IVRI.570)
ISPEC=2
190 WRITE (IWRI,58C)
RLAD (IRD.210) SPEC
IF (SPEC.EQ.RLOC.OR.SPEC.EQ.3LCCL) GO TO 70
IF (SPEC.EQ.CHA.OR.SPEC.EQ.CHAL) GO TO 10
HIV13GOO
HIW13700
HIW18BOO
HIW18900
HIW19000
HIW191OU
HIrf19200
HIW19300
HIW19400
HIW19500
HIW19600
HIW19700
HIW19800
HIW19900
HIW20000
HIW20100
HIW20200
HIW20300
HIW20400
HIW20500
HIW20600
HIW20700
C HIW20800
200 FORMAT (/' DO YOU WANT A DESCRIPTION OF THE EPA "HIWAY-2" MODEL'/'HIW20900
1 BEFORE APPLYING IT?(YES OR NO)'/) HIW21QOO
210 FORMAT (20A4) HIW21100
220 FORMAT (/' DO YOU WANT A DESCRIPTION OF THIS VERSION OF "IUWAY-2"?HIW2^200
UYES OR NO)1/) HIW21300
ENTER PROBLEM TITLE OF 64 CHARACTERS OR LESS') HIW21400
ENTER SCALE FACTOR (KILOMETERS/USER UNIT).') HIW2^500
ENTER LINE(ROAD) ENDPOINTS. (ORDERED PAIRS: XI ,Y1 ,X2,Y2)' )HIW21600
ENTER EMISSION HEIGHT. (METERS)1) HIW21700
ENTER WIND DIRECTION (DEG). NORTH IS ZERO.') HIW21800
ENTER WIND SPEED (METERS/SEC).') HIW21900
ENTER MIXING HEIGHT (METERS).') HIW22000
ENTER PASQUILL-TURNER STABILITY CLASS (1-6).') HIW22100
ENTER THE NUMBER OF LANES.') IIIW22200
ENTER LINE SOURCE STRENGTH VECTOR.(A VALUE FOR EACH LANEHIW22300
HIW22400
IS THIS A CUT SECTION? (YES OR NO)') HIW22500
ENTER WIDTH OF TOP OF CUT.- (METERS)') HIW22600
ENTER HIGHWAY WIDTH (METERS).') HIW22700
ENTER WIDTH OF CENTER STRIP (METERS).') HIW22800
ENTER NUMBER OF RECEPTOR LOCATIONS DESIRED.(MAXIMUM OF 2HIW22900
HIW23000
ENTEH RECEPTOR COORDINATE SETS.(X4Y IN SCALE FACTOR UNITHIW231QO
rERS)') HIW23200
'///' IHWAY-2 VERSION:',16) HIW23300
>',20AM) HIW23400
ENDPOINTS OF THE LINE SOURCE'/' ' ,F9.3,' ,' ,F9. 3,' AriD',HIW23500
230
240
250
260
270
280
290
300
310
360
370
380
390
400
410
420
430
440
45)0
FORMAT C
FORMAT C
FORMAT ('
FORMAT ('
FORMAT C
FORMAT C
FORMAT C
FORMAT ('
FORMAT C
FORMAT ('
1)')
FORMAT C
FORMAT C
FORMAT ('
FORMAT C
FORMAT ('
15)')
FORMAT C
1S;Z IN ME
FORMAT (/
FORMAT ('
FORMAT ('
44
-------
1F9.3,1,',F9.3) HIW23600
460 FORMAT (' EMISSION HEIGHT IS',F8.3,' MFTFKS1) HIW23700
470 FORMAT (' EMISSION HATE (GRAMS/SECONWFTEH) OF1,14,' LAHE(S)') 11IW23800
480 FORMAT C&E^.B) HIW23900
490 FOhMAT (' WIDTH OF TOP OF CUT SECTION IS',F10.3,' METERS') HIV24000
500 FORMAT (' WIDTH OF AT-GHADE HIGHWAY IS',FIG.3,' METERS',/,' WIDTH HIV24100
10F CENTER STRIP IS',F10.3,' METERS') HIW24200
510 FORMAT (' WIND DIRECTION IS',F7.0,' DEGREES',/,' WIND SPEED IS'.FHIW24300
17.V METERS/SEC'/' STABILITY CLASS IS',I5,/,' HEIGHT OF LIMITING HIW24400
2LID IS',F8.1,' METERS') HIW24500
520 FORMAT (' THE SCALE FACTOK IS ',Fi0.4,' KX/USER UNIT.1///) HIW24600
530 FORMAT OHO,' RECEPTOR LOCATION HEIGHT CONCENTRATIHIW24700
10M1,/,1 XVOX,' Y Z (M) UGK/M**3 PPM*') HIW24800
540 FORMAT (/,' * PPM CONCENTRATIONS CORRECT FOR CARBON MOliOXIDE ONLY.HIW24900
1•) HIW25000
550 FORMAT (/////) HIW25100
560 FORMAT (' YOU HAVE THE OPTION TO RUM THE MODEL FOR A NEW RECEPTOR HIW25200
1LOCATION') HIW25300
570 FORMAT (' (LOG),OR TO CHANGE THE ROADWAY TYPE,OR TO END THE PROGRAHIW25400
1M.1) HIW25500
580 FORMAT (' ENTER LOG, OR TYPE, OR END.') HIW25600
STOP HIW25700
C HIW25800
END HIW25900
SUBROUTINE DBTSIG (X,XY,KST,SY,SZ) SIGOOOOO
DIMENSION XA(7), XB(2), XD(5), XE(8), XF(9), AA(8), BA(8), AB(3), SIG00010
1BB(3), AD(6), 3D(6), AE(9), BE(9), AF(10), BF(10) SIG00020
DATA XA /.5,.4,.3,.25,.2,.15,.V SIG00030
DATA XB /.4,.2/ SIG00040
DATA XD /30.,10.,3-,1.,.3/ SIG00050
DATA XE /40.,20.,10.,4.,2.,1.t.3,.V SIG00060
DATA XF /60.,30.,15.,7.,3.,2.,1.,.7,.2/ SIG00070
DATA AA /453.85,3^6.75,258.89,217.41,179.52,170.22,158.08,122.8/ SIG00080
DATA BA /2.1166,1.7283,1.409^,1.2644, M262,1.0932,1.0542, .9447/ SIG00090
DATA AB /109.30,98.483,90.673/ SIG00100
DATA BB /L0971,0.98332,0.93^9S/ SIGOOI'O
DATA AD /44.053,36.650,33.504,32.093,32.093,3^-459/ SIG00120
DATA BD /0.51179,0.56539,0.60486,0.64403,0.31Q66,0.86974/ SIG00130
DATA AE /47.618,35.420,26.970,24.703,22.534,21.628,21.628,23.331,2SIG00140
1^26/ SIG00150
DATA BE /O.29592,0.37615,0.46713,0.50527,0.57154,0.63077,0.7566C,OSIG00160
1.81956,0.63&6/ SIG00170
DATA AF /3^.219,27.074,22.651,17.836,16.167,14.323,13.953,13.953,1SIG00180
14.457,15.209/ SIG00190
DATA BF /0.21716,0.27436,0.32631,0.41507,0.46490,0.54503,0.63227,OSIG00200
45
-------
1.68465,0.73407,0.8155b/
GO TO 00,40,70,80,110,140), ;
-------
180 IF (SZ.GT.5000.) SZ=rjOOO.
190 SY=1000.«XY«SIN(TH)/(2.15*COS(TH))
RETURN
C
END
SIG00710
SIG00720
SIG00730
SIG00740
SIG00750
CC
CC
CC
SUBROUTINE HWYLNE (NQ,NR)
COMMON /SOL/ QLN(25),HLN(25),RAQ(25),SAQ(25),RBQ(25),SBQ(25)
10),CLSS(50),NLINE
COMMON /REC/ RR(51),SR(51),ZR(51)
COMMON /WEA/ THETA.U.KST.HL
COMMON /PUT/ XXRRC51),XXSR<51),QLS(25),HEAD(20) ,IWRI
DIMENSION XSTOD, YST(11)
DIMENSION T(10)
DATA KMAX /9/
MODIFIED OCT. 1979 TO ADD ROMBERG INTEGRATION ENHANCEMENTS.
X(R,S)=(R-RREC)»SINT+(3-SREC)«COST
C X IS UPWIND DISTANCE OF R,S FROM RREC,SREC
Y(R,S)=(S-SREC)«SINT-(R-RREC)»COST
C Y IS CROSSWIND DISTANCE OF R,S FROM RREC.SREC
TR=THETA/57.2958
SIKT=SIN(TR)
COST=COS(TR)
PIN=0.02
UZ=U
C CALCULATE CONCENTRATIONS FOR EACH RECEPTOR.
DO 840 NC=1,NR
RREC=RR(NC)
SREC=SR(NC)
Z=ZR(NC)
C SUM CONCENTRATIONS OVER EACH LANE.
DO 830 NS=1,NQ
R1=RAQ(NS)
S1=SAQ(NS)
R2=RBQ(MS)
S2=SBQ(NS)
QL=QLN(NS)
HsliLN(NS)
X1=X(R1,Si)
X2=X(R2,S2)
IF (X1) 10,30,30
10 IF (X2) 20,30,30
20 RC=0.
GO TO 830
LNEOOOOO
,CON(5LHE00010
LNE00020
LNE00030
LNEOOOHO
LNE00050
LNE00060
LNE00070
LNE00080
LNE00090
LNE00100
LNE00110
LNE00120
LNE00130
LNE001UO
LNE00150
LNE00160
LNE00170
LNE00180
LNE00190
LNE00200
LNE00210
LNE00220
LNE00230
LNE00240
LNE00250
LNE00260
LNE00270
LNE00280
LNE00290
LNE00300
LNE00310
LNE00320
LNE00330
LNE00310
LNE00350
LNE00360
LNE00370
LNE00380
LNE00390
47
-------
30 IF (X1-100.) 40,40,50
40 IF (X2-100.) 60,60,50
50 WHITE (IWRI,850)
GO TO 840
60 DELRsR2-Ri
D£LS=S2-Si
Y1=Y(R1,S1)
Y2=Y(R2,S2)
IF (Y1-Y2) 70,370,70
C IF Y1 = Y2, LINE SOURCE- IS PARALLEL TO UPWIND AZIMUTH F
70 IF (COST+0.0001) 140,30,8U
80 IF CCOST-0.0001) 90,90,140
90 IF (DELR+0.0001) 120,100,100
100 IF (DELR-0.0001) 110,110,120
110 SLOC=SREC
RLOC=R1
GO TO 260
120 SLP=DELS/DELR
IF (SLP) 130,370,130
130 SLOCrSREC
RLOC=(SLOC-S1)/SLP+R1
GO TO 260
140 IF (SINT+0.0001) 190,150,150
150 IF (SINT-0.0001) 160,160,190
160 IF (DELR+0.0001) ISO, 170, HO
170 IF (DELR-0.0001) 370,370,180
180 SLP=DELS/DELR
RLOCsRREC
SLOC=SLP*(RLCC-R1 )+S1
GO TO 260
190 IF (DELR+0.0001) 220,200,200
200 IF (DELR-0.0001) 210,210,220
210 RLOC=R1
SLOC=(RLOC-RREC)"COST/SINT+SREC
GO TO 260
220 IF (DELS+O.OOQ-!) 250,230,230
230 IF (DELS-0.0001) 240,250,250
240 SLOC=S1
RLOCs(SLOC-SREC)«SIHT/COST+HREC
GO TO 260
250 TATHsSIKT/COST
C TATH IS TANGENT (THETA)
SLP=DELS/DELR
C SLP IS SLOPE OF LINE SOURCE.
RLOCs(RREC/TATH+Sl-SLP»R1-SREC)/(1./TATIl-3LP)
SLOCs(RLOC-RREC)/TATH+SREC
C RLOC, SLOC IS LOCUS OF UPrflND VECTOR FROM RECEPTOR AMD
C EXTENSION OF LI'lE SOURCE.
260 XLOC sX(RLOC,SLOC)
IF (XLOC) 370,370,270
LKEOOUOO
LHE00410
LNE00420
LNE00430
LMF00440
LUE00450
LNEOOU60
LNE00470
LNE00480
HGM RECELNE00490
LNE00500
LNE00510
LNE00520
LNE00530
LNE00540
LNE00550
LME00560
LNE00570
LN'E00580
LNE00590
LNE00600
LNE00610
LNE00620
LNE00630
LNE00640
LNE00650
LNE00660
LME00670
LNE00680
LNE00690
LNE00700
LKE00710
LNE00720
LNE00730
LNE00740
LNE00750
LNE00760
LHE00770
LNE00780
LNE00790
LNE00800
LNE00810
LME00820
LNE00330
LNE00840
LNEOOB50
LINEAR LNF.00860
LNEOOS70
LNE00880
LNE00390
48
-------
C XLOC IS POSITIVE IF LOCUS IS UPWIND.
270 IF (S2-S1) 280,280,290
280 SMAX=S1
SMIN=S2
GO TO 300
290 SMAX=S2
SMIN=S1
300 IF (R2-R1) 310,310,320
310 RMAX=R1
RMINrR2
GO TO 330
320 RMAX=R2
RMIN=R1
C SEE IF UPWIND LOCUS IS ON LINE SOURCE.
330 IF (RLOC-RMIN) 370,340,340
340 IF (RMAX-RLOC) 370,350,350
350 IF (SLOC-SMIN) 370,360,360
360 IF (SMAX-SLOC) 370,3BO,380
370 INDIC=1
C INDIC =1 FOR NO LOCUS ON LINE SOURCE.
XA=X1
YA=Y1
XBsX2
YB=Y2
GO TO 390
380 INDIC=2
C INDIC =2 FOR LOCUS ON LINE SOURCE.
XA=X1
YA=Y1
XBsXLCC
YB=0.
390 DISX=XB-XA
DISY=YB-YA
DISIsSQRT(DISX»DISX+DISY«DISY)
C DISI IS LEHGTH(KH) OF LINE CONSIDERED.
IF (DISI) 410,400,4^0
400 CURR=0.
GO TO 770
410 DDI=DISI*1000./2C.
C ONE-HALF IS INCLUDED IN THE 20.
C DDI IS ONE-HALF TIMES 1/10 OF DISI (M).
DX=DISX/10.
DY=DISY/10.
PREV=0.
KNTRL=1
XIsXA
YI=YA
KNTsO
DO 530 1=1,11
C STORE EACH XI,YI.
LNE00900
UJE00910
LNE00920
LNE00930
LNE00940
LNE0095C
LNE00960
LME00970
LNE00980
LNE00990
LNE0100Q
LNE01010
LNE01030
LNE01040
LNE01050
LNE01060
LNE01070
LNEO-'OBO
LNE01Q90
LNE01 100
LNE01MO
LNE01120
LNE01130
LNE01140
LNE01 150
LNE01160
LNE01170
LNE01 180
LNE01190
LNE01200
LNE01210
LNE01220
LNE01230
LNE01240
LNE01250
LNE01260
LNF.01270
LNE01280
LNE01290
LNE01300
LNE01320
LNE01330
LNE013HO
Lt;E01360
LNE01 370
LKE01380
LNE01390
49
-------
XST(I)=XI
YST(I)=YI
IF UST(I)) 420,420,430
'420 RC=0.
GO TO 440
430 xz=xi
XY=XI
CALL HWYRCX (UZ,Z,H,HL,XZ,XY,YI,KST,AU,K,SY,SZ,RC)
440 GO TO (450,490), KHTRL
C IF RC IS ZERO, CONTINUE UHTIL RC IS POSITIVE.
450 IF (RC) 520,520,460
460 IF (1-1) 470,470,480
470 KNTRL=2
GO TO 510
C RESET POINT A TO LAST ONE PREVIOUS.
480 XA=XST(I-1)
YA=YST(I-1)
KNTRL=2
GO TO 510
490 IF (RC) 500,500,510
C RESET POINT B IF REACH ZERO CONCEKTHATION.
500 XB=XI
YB=YI
GO TO 540
510 KNT=KNT+1
520 XI=XI+DX
YI=YI+DY
530 COKTINUE
540 IF (KNT) 560,560,550
550 IF (KNT-6) 390,390,600
C IF GET TO 560, CONC. FROM THIS SEGMENT IS 0.
560 GO TO (570,580,590), INDIC
570 RC=0.
GO TO 830
580 FIRST=0.
GO TO 800
590 RC=FIRST
GO TO 820
600 CONTINUE
C DO A TRAPEZOIDAL INTEGRATION FROM A TO B III TEN STEPS.
C IT IS LIKELY THAT A OR B HAVE BEEN REDEFINED.
DISX=XB-XA
DISY=YB-YA
DISI=SQRT(DISX*DISX+DISY*DISY)
C DISI IS DISTANCE(KM) FROM A TO B
LNDEX=0
ILIM=3
610 CONTINUE
FILIM=FLOAT(ILIM)
FAC=lOOO./FILIh
LNE01400
LME0141Q
LNE01420
LNE01430
LNE01440
LNE0145Q
LNEC1460
LNE01470
LME01480
LNE01490
LNE01510
LNE01520
LNE01530
LNE01540
LNE01550
LNE01560
LNE01570
LNE01580
LNE01590
LNE01600
LNE0 16-10
LNE01620
LNE01630
LNE016UO
LNEO-I650
LNE01660
LNE01670
LNE01680
LNE01690
LNE01700
LNE01710
LNE01720
LNE01730
LNE01740
LNE01750
LNE01760
LNE01770
LNE01780
LNE01790
LNE01810
LNE01820
LUE01830
LME01840
LNE01850
LNE01860
LNE01870
LNE01830
LME01890
50
-------
DELD=DISI*FAC
C DELD IS VFILIM DISI IN KETERS.
DX=DISX/FILIM
DY=DISY/FILIM
SUMsO.
XDUM=XA
YDUM=YA
IF (XDUM.LE.O.) GO TO 620
XZ=XDUM
XY=XDUM
CALL HWYRCX (UZ ,Z,H,HL,XZ,XY,YDLT1,K5T?A?i,y:,SY,SZ,RC)
SUM=SUM+RC/2.
ILIM1=ILIM-1
620 DO 630 I=1,ILIM1
XDUM=XDUM+DX
YDUM=YDUM+DY
IF (XDUM.LE.O.) GO TO 630
XZ=XDUM
XYsXDUM
CALL HWYRCX (UZ.Z.H.HL.XZ.XY.YDUM.KST.AN.M.SY.SZ.RC)
SUM=SUM-»-RC
630 CONTINUE
XDUM=XDUM+DX
YDUM=YDUK+DY
IF (XDUM.LE.O.) GO TO 6*40
XZrXDUM
XY=XDUM
CALL HWYRCX (UZ,Z,H,HL,XZ,XY,YDUM,KST,AN,M,SY,SZ,RC)
SUM=SUM+RC/2.
C INTEGRATED VALUE IS CURR.
6UO CURR=SUM»DELD
C
T(1)=CURR
K=0
DO 650 KK=2,10
650 T(KK)=0.
C
C FIRST ESTIMATE COMPLETED HERE.
660 PREV=CURR
C EVALUATE FOR POINTS IN BETWEEN THOSE ALREADY EVALUATED.
DELD=DELD/2.
XDUM=XA-»-DX/2.
YDUy=YA+DY/2.
DO 680 Isl.ILIM
IF (XDUM.LE.O.) GO TO 670
XZsXDUM
XY=XDUM
CALL HWYRCX (UZ,Z,H,HL,XZ,XY,YDUM,KST,AN,MtSY,SZ,RC)
C NOTE ADD THESE TO RC'S FOUND ABOVE.
SUM=SUM+RC
LHE01900
LNE01910
LNE01920
LNE01930
LNE019MO
LNE01950
LNE01960
LNE01970
LNE01990
LKE02000
LNE02010
LNE02020
LNE02030
LNE02CWO
LNE02050
LNE02060
LNE02070
LNE02080
LNE02090
LNE02100
LNE021 10
LNE02120
LNE02130
LNE02.1MO
LNE02150
LNE02160
LNE02170
LNE02180
LNF.02190
LNE02200
LNE02210
LNE02220
LNE02230
LNE02240
LNE02250
LNE02260
LNE02270
LNE02280
LNE02290
LNE02300
LNE02310
LNE02320
LNE02330
LNE02340
LNE02350
LNE02360
LNE02370
LNE02380
LNE02390
51
-------
670 XDUM=XDUM+DX LNE02HOO
680 YDUM=YBUK.+DY L'!E02410
CUhR=SUM«DELD LNt:0242G
C SECOND ESTIMATE COMPLETED IOC. ALSO FOURTH, SIXTH, FTC. LNE02430
c LNE02WO
K=K+1 LNE02450
OLDi=T(i) LNEQ2460
T(1)=CUKR LNE02470
DE.NOM=4 L^fE02'480
DO 690 KK=V< LNE02¥jO
KKKsKK+1 LNE02500
OLD2=T(KKK) LNE025''0
T(KKK )=T (KK )+ (T (KK )-OLDl ) / (DEMO-;-'' } LNE02520
OLD1=OLD2 LNE02530
DENOM=DENOK*4 LNE025MO
690 CONTINUE LNE02550
CURF=T(KK) LNE02560
c LNE02570
IF (INDEX. EQ.O) TEST=AB3((CUHR-PRE:V)/CURR) LNE02580
IF ( I NDEX . Eg . 1 ) TEST=A DS ( ( CUR K-CUR OLD ) /CUROLD ) L.N'E02590
C IF WITHIN PIN OF LAST VALUE (PRFV) , CONSIDER THIS AS FINAL VALULNE02600
IF (TEST-PIN) 770,700,700 LNE02610
7CO ILIM=ILIK»2 LNE02620
IF (K.GE.KMAX) GO TO 750 LNE02630
PRFV=CURR LNE02640
C EVALUATE POINTS IN BETWEEN. LNE02650
DELD=DELD/2. LNE02660
DX=DX/2. LNE02670
DY=DY/2. LNE02680
XDUM=XA+DX/2. LNE02690
YDUMsYA+DY/2. LNE02700
DO 720 I=1,ILIM LNE02710
IF (XDUM.LE.O.) GO TO 7^0 LNE02720
XZ=XDUM LNE02730
XYdCDUM LNE02740
CALL H'rfYRCX (UZ ,Z,H,HL,XZ,XYfYDUM,KST, Af;,M,SY,SZ,RC) LNE02750
SUMrSUK+RC LNE02760
71 C XDUM=XDUM+DX LNE02770
720 YDUM=YD'JM+DY LNE02780
CURR=SUM»DELD LHE02790
C „ ,. , LNE02800
,«t\./^ LNE02810
OLD=T(i) LNt:02820
T(1)=CUKh LNE02830
LME028MO
OLD2=T(KKK) LNE02870
T(KKK)=T(KK)-t-(T(KK)-OLD1 )/(DENOM-l ) LKE02380
OLD1 =OLI)2 LK'E02890
52
-------
730
C
C
7UO
750
760
C
770
780
790
800
810
820
830
840
C
850
860
870
C
DENOM=DENOK»4
CONTINUE
CURR=T(KK)
IF (INDEX. EQ.O) TEST=ABS((CUkR-PREV)/CURR)
IF (INDEX. EQ.1) TEST=AnS((CURR-CUROLD)/CUROLD)
THIRD ESTIMATE COMPLETED HERE. ALSO FIFTH, SEVENTH, ETC
IF (TEST-PIN) 770,740,740
ILIM=ILIM*2
DX=DX/2.
DY=DY/2.
IF (K.GE.KMAX) GO TO 750
GO TO 660
IF (INDEX. EQ.1) GO TO 760
CUROLD=CURR
IHDEXsl
LNE02900
LNE02910
LKE02920
LNE02930
GO TO 610
WRITE (6,860) NS,NC,NLINE,TEST
AT 770 HAVE FINAL VALUE OF INTEGRATION IN CURR.
IF (INDEX. EQ.1) CURR=AMIN1(CURR,CUROLD)
GO TO (730,790,8^0), INDIC
RCsCURR
GO TO 820
FIRST=CURR
INDIC=3
XJUXLX
YA=0.
XB=X2
YB=Y2
GO TO 390
RCsFIRST-nCURR
CON(NC)=CON(NC)+RC*QL
CONTINUE
CON(NC)=1 .OE+6*CON(NC)
CLSS(MC)s0.00087»OON(NC)
WRITE (IWRI.870) XXRR(NC) ,XXSR(HC),ZR(NC) ,CON(NC) ,CLSS(NC)
CONTINUE
FORMAT (1 HO, 'RECEPTOR IS 100KM OR MORE FROM SOURCE1)
FORMAT (/' ***** THE INTEGRATED VALUE FOR LANE ',12,' RECEPTOR
1 ',I3,/,' FROM LINE ',I3,f HAS1,' A RELATIVE ERROR OF'
2**')
FORMAT OH ,3(F10.4,2X),F10.0,F10.3)
RETURN
END
LNE02950
LNE02960
LIIE02970
LNE02980
LNE02990
LNE03000
LNE03010
LNE03Q20
LNE03030
LNE03MO
LHE03050
LNE03060
LNE03070
LNE03080
LNE03090
LNE03100
LNE031 10
LNE03120
LNE03130
LNE031UO
LNE03150
LNE03160
LNE03170
LNE03180
LNE03190
LNE03200
LNE03210
LNE03220
LNE03230
LNE032UO
LNE03250
LNE03260
LNE03270
LNE03280
LNE03290
NO.LNE03300
*»«LNE03310
LNE03320
LNE03330
LNE033MO
LNE03350
LNE03360
53
-------
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
SUBROUTINE MWYRCX (U ,Z,H,HL,X,XY,Y,K3T,Ah, ;.,,SY ,3Z ,RC) RCXQOOOU
TilI3 IS THE 1979 VERSION OF IIWYRCX. RCXOOQiQ
L. B. TUhJJEFi, HF.SF.AkCH METEOROLOGIST* MODF.L DEVFLOPMEHT BRANCH,RCX00020
DIVISION OF METEOROLOGY, ENVIRONMENTAL PROTECTION AGENCY. RCX00030
ROOM 314B, MCHS BUILDING, RTF. PHONE (919) 549-8411 EXT 4564RCX00040
MAILII.'G ADDRESS- DM, EPA, RESEARCH TRIANGLE PARK, NC 27711 RCX00050
* ON ASSIGNMENT FRO/ NATIONAL OCEANIC AMD ATMOSPHERIC RCX00060
ADMINISTRATION, DEPARTMENT OF COMMCRCE. RCX00070
SUBROUTINE HWYRCX CALCULATES CliI/Q CONCENTRATION VALUES, UWYRCXRCXOC08C
CALLS UPON
THE INPUT
U
z
H
HLsL
X
XY
Y
KST
SUBROUTINE HWYSIG TO OBTAIN STANDARD DEVIATION'S.
RIABLES ARE
',,'IND SPEED (M/SEC)
RECEPTOR HEIGHT (M)
EFFECTIVE STACK HEIGHT (M)
HHIGHT OF LIMITING LID (M)
DISTANCE RECEPTOR IS DOWNWIND OF SOURCE (KM)
X+VIRTUAL DISTANCE USED FOR AREA 30URCF APPROX.
(KM)
IS
CROSSWIMD
FROM SOUHCE (KM)
DISTANCE RECEPTOR
STAHILITY CLASS
THE OUTPUT VARIABLES ARE
AN THE NUMBER OF TIMES THE SUMMATION TERM IS EVALUATED
AND ADDED IN.
RC RELATIVE CONCENTRATION (SEC/M**3)
THE FOLLOWING EQUATION IS SOLVED —
RC = (V(2*PI*U*SIGMA Y "SIGMA Z))* (EXP(-0.5*(Y/SIGMA Y)**2) )RCX002>40
(EXP(-0.5*((Z-ll)/SIGMA Z)**2) + EXP(-0.5*((Z+H)/3IGMA Z)**2)RCX00250
RCX00090
RCX00100
RCX00110
RCX00120
RCX00130
Rcxoomo
RCX00150
RCX00160
RCX00170
RCX001SO
RCX00190
RCX00200
RCX00210
RCX0022Q
RCX00230
C
c
c
c
c
c
c
c
c
c
10
20
30
40
C
C
50
C
DO
C
PLUS THE SUM OF THE FOLLOWING 4 TERMS K TIMES (N=1
TERM 1- EXP (-0.5* (U-H-2KL) /SIGMA Z)**2)
TERM 2- EXP(-0.5»((Z+H-2NL)/SIGMA Z)**2)
TERM 3- EXP(-0.5*((Z-H+2NL)/SIGMA Z)**2)
TERM U- EXP(-0.5*((Z+H+2ML)/SIGMA Z)**2)
THE ABOVE EQUATION IS SIMILAR TO EQUATION (5.8) P 36 IN
WORKBOOK OF ATMOSPHERIC DISPERSION ESTIMATES WITH THE
OF THE EXPONENTIAL INVOLVING Y.
IWRI IS CONTROL CODE FOR OUTPUT
IWRI=6
IF (KST.GE.5) GC TO 5>0
IF THE SOURCE IS ABOVE THE LID, SET RC = 0. , AND RETURN
IF (H-HL) 10,10,20
IF (Z-HL) 50,50,40
IF (Z-HL) 40,30,30
WRITE (IWRI, 460)
RC=0.
RETURN
IF X IS LESS THAN 1 NETth, SET RC=0. AND RETUKN. THIS
PROULEK3 OF INCORRECT VALUES KEAR TiiE SOURCE.
IF (X-0.001) 40,60,60
CALL HWYSIC TO OET-MN VALUES FOR SY AND 32
CALL IIWYSIG (X,XY,?;ST,SY,SZ)
SY = SIGliA Y, TIIF STANDARD DEVIATION OF CONCENTRATION
,K) — RCX00260
RCX00270
RCX00280
RCX00290
RCX00300
BCX00310
ADDITIONRCX00320
RCX00330
RCX00340
RCX00350
RCX00360
RCX00370
RCX00380
RCX00390
HCX00400
RCX00410
HCX00420
RCX00430
AVOIDS RCX00440
RCX00450
RCX00460
RCX00470
HCX00480
IN THE HCX00490
54
-------
c
c
c
70
c
80
90
100
C
c
110
120
130
C
150
160
170
C
180
C
C
C
190
C
200
C
210
Y-DIRECTICN (M)
SZ = SIGMA Z, THE STANDARD DEVIATION OF CONCENTRATION IK THE
Z-DIRECTIOK (X.)
C1=1.
IF (Y) 70,90,70
YD=1000.*Y
YD IS CROSSWIKD DISTANCE IN METERS.
DUM=YD/SY
TEMP=0.5*DUM*DUM
IF (TEMP-50.) 80,40,40
C1=EXP(TEMP)
IF (KST-4) 100,100,110
IF (HL-5000.) 190,110,110
IF STABLE CONDITION OR UNLIMITED MIXING HEIGHT,
USE EQUATION 3.2 IF Z = 0, OR EQ 3.1 FOR NON-ZERO Z.
C2=2.*SZ»SZ
IF (Z) 40,120,140
C3=H*H/C2
IF (C3-50.) 130,40,40
A2=1./EXP(C3>
WADE EQUATION 3.2.
RC=A2/(3.14159*U*SY*SZ*C1)
M=1
RETURN
A2=0.
A3=0.
CA=Z-ii
CB=Z+H
C3=CA*CA/C2
C4=CB*CD/C2
IF (C3-50.) 150,160,160
A2=1./EXP(C3)
IF (C4-50.) 170,180,180
A3=1./EXP(C4)
WADE EQUATION 3.1.
RC=(A2+A3)/(6.2831B*U*SY*SZ*C1)
Ms2
RETURN
IF SIGKA-Z IS GREATER THAN 1.6 TIKES THE MIXING HEIGHT
THE DISTRIBUTION BELOW THE MIXING HEIGHT IS UNIFORM WITH
HEIGHT REGARDLESS OF SOURCE HEIGHT.
IF (SZ/HL-1.6) 210,2^0,200
WADE EQUATION 3.5.
RC=1./(2.5066*U*SY*HL*C1)
RETURN
INITIAL VALUE OF
AN=0.
IF (Z) 40,370,220
AN SET = 0.
rkcxoo5oo
RCX00510
RCX00520
RCX00530
RCX00540
RCX00550
RCX00560
RCX00570
RCX00580
RCX00590
RCX00600
RCX00610
RCX00620
RCX00630
RCX006MO
RCX00650
RCX00660
Ra00670
RCX00680
RCX00690
RCX00700
RCX00710
RCX00720
RCX00730
RCX007MO
RCX00750
RCX00760
RCX00770
HCX00780
RCX00790
RCX00800
RCX00810
RCX00820
RCX00830
RCX008MO
RCX00850
RCX00860
RCX00870
RCX00880
RCX00390
RCX00900
RCX00910
RCX00920
RCX00930
RCX00940
RCX00950
RCX00960
RCX00970
RCX00980
sTATF-MEhTS 220 TO 360 CALCULATE RC, THE RELATIVE CONCENTRATION,RCX00990
55
-------
C
C
C
C
C
C
C
220
230
240
250
260
270
280
290
300
B10
320
330
340
350
360
USING THE EQUATION DISCUSSED ABOVE. SEVERAL INTERMEDIATE
VARIABLES ARE USED TO AVOID REPEATING CALCULATIONS.
CHECKS ARE MADE TO BE SURE THAT THE ARGUMENT OF THE
EXPONENTIAL FUNCTION IS IOER GREATER THAN 53 (OR LESS THAN
-50). IF 'AN' BECOMES GREATEF THAN 45, A LINE OF OUTPUT IS
PRINTED L'JFORMING OF THIS.
CALCULATE MULTIPLE EDDY REFLECTIONS FOR RECEPTOR HEIGHT Z.
A1=1./(6.2831i5*U*SY*SZ*C'<)
C2=2.*SZ*SZ
A2=0 .
A3=0.
CA=Z-H
CD=Z+H
C3=CA*CA/C2
C4=CB*CB/C2
IF (C3-'JO.) 230,240,240
A2=1./EXP(C3)
IF (C4-50.) 250,260,260
A3=1./EXP(C4)
SUM=0.
THL=2.*HL
AOsO.
A7=0 .
C5=AN«T(L
CC=CA-C5
CD=CB-C5
CE=CA-*-C'j
CFsCD+CS
CO=CC*CC/C2
C7=CD*CD/C2
C8=CL*CE/C2
C9=CF*CF/C2
IF (C6-5G.) 280,290,290
A't=i./EXP(C6)
IF (C7-50.) 300,3^,310
A5=1./EXP(C7)
IF (C3-5C.) 320,330,330
A6=1./EXP(CS)
IF (C9-50.) 3'lO,350,350
A7=1./F.XP(C9)
T=A4+Ab+A6+A7
SUHsSUM+T
IF (T-0.01) 360,270,270
RC=A1*(A2+A3+3UM)
1-1=5
RCX01000
RCX01010
RCX01Q2C
RCX01030
RCX01040
RCXO^O'^C
KCXO-'OfjO
HCX01070
RCX010GO
RCX01090
HCXCHQO
RCXO^MO
RCX01120
RCX01130
RCXOi 140
RCXO-M50
RCX01170
RCXOUSO
RCXOH9Q
RCXC1200
RCX01210
RCXOI 220
RCXQ1230
RCX0124C
RCX01250
RCXOI 260
RCXOI 270
RCX01280
RCX01290
RCX01300
RCX01310
RCXOI 330
RCX013UO
RCX01350
RCX01360
RCX0137C
RCXOI 380
RCXOI 390
RCX01400
RCX01M10
RCXOI 421)
RCXOI 4 3()
CALCULATE MULTIPLE EDDY REFLECTIO.'JS FOR GROUND LEVEL RECEPTOR
RCXOI450
ilCXO 1U60
RCX01470
KCX01430
HRCX01490
56
-------
370
380
390
400
420
430
440
450
C
460
A1=1./(6.28318*U*SY*SZ*C1)
A2=0.
C2=2.*SZ*SZ
C3=H*H/C2
IF (C3-50.) 360,390,390
A2=2./EXP(C3)
SUM=0.
THL=2.*HL
ANsAfJ+1 •
A4=0.
A6=0.
C5=AN*THL
CC=H-C5
CE=H+C5
C6=CC*CC/C2
C8=CE*CE/C2
IF (C6-50.) 410,420,420
A4=2./EXP(C6)
IF (C8-50.) 430,440,440
A6s2./EXP(CS)
T=A4+A6
SUM=SUM+T
IF (T-0.01) 450,400,400
RC=A1*(A2+SUM)
M=4
FORMAT (1 HO,'BOTH H AND 7. ARE ABOVE
IE COMPUTATION CAN NOT BE MADE.')
RETURN
END
RCX01500
PCXO-<520
RCX01530
RCX01540
RCX01550
RCX01570
RCX01590
RCX01600
RCX01610
RCX01620
RCX01630
RCX01640
RCX01650
RCX01660
RCX01670
RCX01680
Raoiego
RCX01700
RQ01710
RCX01720
RCX01730
RCX01750
THE MIXING HEIGHT SO A RELIABLRCX01760
RCX01770
RQ01780
RCX01790
RCX01800
C
C
C
C
SUBROUTINE HWYSIG (X,XY,KST,SIGY,SIGZ)
COMMON /WS/ WSP,SYON,SZON
DIMENSION 3PGZ(3), SPGY(3)
DATA SPGZ /30.144,12.093,8.69B/
DATA SPGY /52.203,22.612,16.395/
IF (X.EQ.C.) X=0.0001
XP=X
SIGZOsSZON
RATsSYON/SZOK
IF (WSP.LE.3.91) SIGZO=3.57-0.53*WSP
300 KETERS USE CUHVES
AS GIVEN IH THIS SUBROUTINE.
IF X.LE.
IF X.GT. 300 METERS THEN DETSIG IS CALLED TO COMPUTE
THE SIGMAS USING THE DISPERSION AT 300 METERS DUE TO
HSGOOOOO
HSG00010
HSG00020
HSG00030
HSG00040
HSG00050
HSG00060
HSG00070
HSG00080
HSG00090
HSG00100
HSC00110
HSGOO^O
HSGOO^O
57
-------
c
c
10
20
THE KQADWAY A3 Ti!L INITIAL DISPERSION.
IF (X.GT.G.3) X=0.3
IF (:
-------
APPENDIX C
SUGGESTIONS FOR IMPROVEMENT OF THE EPA-fllGHWAY MODEL
Appendix C is a copy of the article which discusses the recommended changes to
the HIWAY model that was received from New York State Department of Environ-
mental Conservation. HIWAY-2, presented in the User's Guide, is essentially
the same as HIWAY #4, presented in this appendix. The performance of HIWAY-2
with the GM and New York State's data sets is indicated by the performance of
HIWAY |4.
59
-------
Suggestions for Improvement of
The EPA-HIWAY Model
S. Trivlkrama Rao and Michael T. Keenan
New York State Department of Environmental Conservation
Previous studies have Indicated that the EPA-HIWAY model signifi-
cantly overestimates the pollutant concentrations for stable atmo-
spheric conditions, especially under parallel wind-road orientation
angles with low wind speed. This over estimation Is due to the fact that
the model's dispersion parameters do not properly account for the
traffic-Induced turbulence near roadways. In this paper, the Pas-
quIII-Glfford dispersion curves used by the model are modified based
on the recent studies that have quantified the nature of the traffic-
Induced turbulence and Hs Influence on the pollutant dispersion In the
near-field. The results show that the model performance Is signifi-
cantly Improved when these new dispersion curves In conjunction
wtth an aerodynamic drag factor, which In a rough way accounts for
the change In the mean wind field due to the moving vehicles, are
used In the HIWAY model.
Previous investigations by Chock,1 Noll, et a/.,2 Sistla, et al.,'A
and Rao, et a/.,4 indicated that the EPA-HIWAY model5
overestimates pollutant concentrations adjacent to the
highway. This overestimation is more significant under stable
atmospheric conditions and for parallel wind-road orientation
angles with low wind speeds. Petersen6 used the wind fluc-
tuation data in a modified version of the original HIWAY
model which specifies the dispersion parameters as a function
of wind fluctuation statistics and found that there was sig-
nificant improvement in the model performance over the
current version of the HIWAY model. This modified version
of the model requires the standard deviations of the horizontal
wind direction and elevation angles as a function of averaging
time and sampling duration as input to the modeL These so-
phisticated data are not generally available and the model
cannot be applied without having this information. The most
important finding of Petersen6 is that the ambient turbulence
mechanisms will be properly represented when on-site tur-
bulence data are used. Rao, et a/.7'8 and Eskridge, et a/.9-10
clearly demonstrated that the dispersion in the near-field is
completely dictated by the locally generated turbulence and
that the ambient atmospheric stability plays an insignificant
role in dispersing pollutants in the immediate vicinity of the
roadway.
This paper presents a new set of dispersion curves appli-
cable for pollutant dispersion estimation near roadways based
on the data collected in the General Motors (GM) Study11 and
in the New York (NY) State Study.12 Further, an empirical
aerodynamic drag factor is developed to handle pollutant
dispersion under low wind speed conditions when traffic-
generated effects dominate dispersion. When the original
Pasquill-Gifford (P-G) curves used in the HIWAY model are
replaced by these new dispersion curves and the aerodynamic
drag factor is included, the performance of the HIWAY model
is significantly improved. Several statistical tests are made
to compare the simulation capabilities of the original HIWAY
model and the modified HIWAY model. These results are
compared to the results of the GM model,13 since it was
identified in the previous investigations that the GM model
was in very good agreement with the observed data. However,
the GM model has a tendency to underestimate the pollutant
concentrations. Although the modified HIWAY model does
not have as good regression statistics as the GM model in some
cases, the modified HIWAY has a slight tendency to be con-
servative. This is desirable since HIWAY is used by the reg-
ulatory agencies for their decision-making purposes.
Data Bases
During October, 1975, the General Motors Corporation
conducted several tracer gas experiments over a simulated
roadway. Cadle, etal.11 discussed the details of the experiment
and the 'data set. The meteorological data consisted of three
components of wind velocity at 1 sec intervals from 20 ane-
mometers located on six towers and two stands adjacent to the
test track. Sulfur hexafluoride (SFe) was used as a tracer and
samples were collected over a period of 30 minutes at 20 lo-
cations. A total of 61 tracer runs were used in this study. Figure
1 shows the locations of various instruments. The other data
set used in this study was obtained by the New York State
Department of Environmental Conservation in a similar ex-
periment on the Long Island Expressway. A total of 23 tracer
experiments were included in the New York study and Figure
2 shows the location of SF6 samplers in this study. The details
of this experiment can be found in Rao, et at.12
Copyright 1980 Air Pollution Control Association
March 1980 Volume 30, No. 3
-------
105
9.5
© 0
O Gill UVW anemometer
o SF6 sampling points
Temperature sensors
Numbers in circles
are site numbers
45
35
1.5
0.5
0
Figure
Distance 1rom roadway (m)
1. Instrument layout in the General Motors dispersion experiment.
Modifications to the HIWAY Model
Basically, three parameters are necessary in estimating
dispersion from a ground level line source. These are the mean
wind speed, and the standard deviations of the plume spread
in the horizontal and vertical directions. The gaussian dis-
persion equation has a singularity at zero wind speed, and
hence the gaussian assumptions become invalid at very low
wind speeds. Hence, all gaussian models perform poorly when
the wind speeds are less than 1 m/sec.
Rao, etal.1 studied the effects of aerodynamic drag due to
moving vehicles on the wind profiles adjacent to the roadway,
and determined that there is a pronounced acceleration of
wind in the lowest 8 m, especially in the cases of wind direc-
tions nearly parallel to the roadway. Thus, even when the
ambient wind speeds are quite low, near the roadway itself
there is an acceleration of wind. As for the plume spread, the
HIWAY model employs Pasquill-Gifford (P-G) curves ex-
trapolated upwind to 1 m from the source. These curves were
originally developed for downwind distances beginning at 100
m from a point source and, hence, do not properly describe the
dispersion in the near-field. Rao, et al.8 indicated that the
wake effects due to the moving vehicles on the roadway are
superimposed upon the naturally occuring turbulence and
play a dominant role in dispersing the pollutants in the
near-field.
Hence, there are two possible ways to modify the line source
dispersion equation to account for the enhanced dispersion
due to the traffic flow, namely, (1) application of an aerody-
namic drag factor, and (2) application of a new set of disper-
sion curves that properly reflect the turbulence characteristics
adjacent to the roadways.
Wind Speed Correction Factor
As previously indicated, the assumptions in the gaussian
equation are violated under low wind speed conditions. Since
the concentration of the pollutants is inversely proportional
to the wind speed, the concentration approaches infinity
asymptotically as the wind speed approaches zero. This sit-
uation is obviously unrealistic. Carpenter and Clemena14 ar-
gued that the inverse linear relationship is valid only for rel-
atively high wind speeds, and hence, requires a correction
factor for low wind speeds. Based on observational data, they
suggested that the concentration is inversely proportional to
(U + 1.92e~"---'') where (7 is the wind speed. This relation
suggests that the concentration is inversely proportional to
1.92 at zero wind speed, and as the wind speed increases the
concentration becomes inversely proportional to U. Since the
original HIWAY model performed poorly for low wind speeds,
the wind speed correction formulation as suggested by Car-
pen, "r and Clemena was included in the HIWAY model.
Aerodynamic Drag Factor
Analysis of the GM data revealed that the aerodynamic drag
factor must be a function of the wind-road orientation angle.
This is because the amount of acceleration in the lower layers
is most significant under parallel wind-road orientation (see
Rao, et al.1). Hence, an aerodynamic drag factor which is a
function of wind-road angle is developed and is incorporated
into the HIWAY model. The relation developed is 17 =
AUf,0 164cos20 where U is the adjusted wind speed used in the
model, t/o is the ambient wind speed (m/sec), ft is the wind-
road angle, and A is a constant related to the traffic speed. It
is observed that A equals 1.85 for the traffic speed conditions
such as those simulated in the GM experiment. This relation
takes its full effect for parallel wind (0 = 0) situations and has
no effect for perpendicular wind cases. If the ambient wind
speed, t/o, is less than the wind speed, U, computed according
to the above relation, then only the adjusted wind speed, U,
is used in the model. If the ambient wind speed is greater than
the adjusted wind speed, no changes to the wind speed are
made (i.e. if Uo ^ U, then U = t/o). Thus, this allows correc-
tion for only low ambient wind speed situations (when am-
bient wind speeds are less than 2 m/sec).
Dispersion Parameters
The applicability of the dispersion values used in the
HIWAY model can be evaluated by comparing the calculated
a:, standard deviation of the plume spread in the vertical, to
those used in the HIWAY model. Since measurements of
concentrations at various receptors downwind of the source
and meteorological conditions during the experiment are
available, the vertical dispersion parameter can be computed
from the line source equation:15
C
v/2 Q
1 Z-'
exp|--^
(1)
V/TT u sin Bas
Where C is the observed concentration, u is the mean wind
speed, 6 is the angle between the wind direction and the ori-
entation of the road (60° < 6 < 90°), Z is the height of the
receptor, and az is the vertical diffusion parameter. At each
tower location, using the gaussian plume assumption that a,
is not a function of height and that the wind is uniform in the
layer, o2 can be calculated by transposing Eq. (1) into the
form
Z./ - Zr
' 2 In (C,/C2)
(2)
Where Z2 and Zi are the two levels at a given downwind dis-
tance at which C2 and Ci are the concentrations measured.
In Eq. (2), the concentrations at heights 0.5 m (Zt) and 3.5 m
(Z2) at the nearest roadside tower are used to compute a:.
Once this value is known, the variation of oz with downwind
distance (x) is calculated from the relation
C(atxi) a,-
C(atx2)
-------
O CO "Y" Climet wind speed & direction
T SF6 x Gill UVW anemometer
wv "T Q Dichotomous paniculate sampler
Not to scale
D(16m)
C(8m|
B(4m
A(2m)
-JLJ
-16m-
2 3 456 7 89 10
Figure 2. Specific locations of various instruments in the New York experiment.
13
typically associated with P-G dispersion classes B to C. These
results clearly indicate that the present dispersion values in
the HIWAY model are unrealistic and need to be modified.
The calculated az from Eq. (2) at the nearest roadside re-
ceptor is plotted against the cross-road wind speed in Figure
4, where actual local stability for each data point is also indi-
cated in the diagram. These results show that az must be a
function of not only atmospheric stability but also cross-road
wind speeds compared to high cross-road wind speed situa-
tions. Under low cross-road wind speeds, the plume takes a
GM site NY site A
22.0
20.0
-§18.0
*
16 0
i '
|14.0
§12.0
1 10-°
1 8.0
6
3 6-°
£ 4.0
2.0
•> Local Pasquill stability A .Stabi|jtvB ° /'
; "t B 60°,; and
-------
original HIWAY model. The new set of dispersion curves,
computed according to the above equations, is compared to
the P-G curves of the HIWAY model in Figure 5.
Results
The following versions of the HIWAY model are employed
to simulate the tracer release experiments conducted in the
GM and NY studies.
HIWAY # 1—original HIWAY model,
HIWAY # 2—HIWAY model with the wind speed correc-
tion factor suggested by Carpenter and
Clemana,14
HIWAY #3—HIWAY model with the modified set of
dispersion curves,
HIWAY #4—HIWAY model with the modified dispersion
curves and the aerodynamic drag factor.
The performances of each of the above versions of the HIWAY
model are compared to the GM model developed by Chock,l:i
and the numerical wake model developed by Eskridge, et a/.10
cording to the stability classes (see Table II), again HIWAY
#3 has a better slope than HIWAY # 1. The correlations of
HIWAY #4 show even further improvement over HIWAY
#3. HIWAY #4 shows a considerable improvement over
HIWAY # 1 for stable atmospheric conditions. When the data
are grouped according to wind speed class (see Table III),
HIWAY #4 provides a better simulation than the other
HIWAY versions for low wind speeds.
The aerodynamic drag factor is applied only if the adjusted
wind speed according to the relation 1.85 I/a0-164 cos2 0 is
greater than the ambient wind speed (t/o). It should be noted
that this correction is not applicable when t/o approaches zero.
No attempts were made to make the correction valid for (Jo
= 0 since the basic gaussian dispersion equation becomes in-
valid at this point. The cut-off wind speed is about 30 cm/sec.
The reason for this cut-off is that commonly used wind in-
struments themselves cannot measure wind speeds reliably
when the wind speeds are less than about 30 cm/sec.
When the data with wind speeds greater than 1 m/sec are
used, the simulation of HIWAY #3 is quite close to that of
E
51
of n
UJ
JJ 14
2 a
2
tr
UJ
>
UNSTABLE
10 20 SO 40 90 W 70 »0 »0 IOO
DOWNWIND DISTANCE (m)
y
NEUTRAL
If
»r "
K> 20 » 40 SO «0 70 «0 90 IOO
DOWNWIND DISTANCE (m)
STABLE
10 20 JO 40 90 «0 70 «O *0
DOWNWIND DISTANCE (m)
u.
I -
UNSTABLE
DOWNWIND DISTANCE (m)
_
I
NEUTRAL
of
LJ ,2
K 4
Z
O
STABLE
^•£&X*^--===1
t&L
*0 20 SO 40 50 *0 70 10 K> KX>
DOWNWIND DISTANCE (m)
20 » 4O SO «0 TO CO
DOWNWIND DISTANCE (m)
Figure 5. Comparison of new dispersion curves (shaded areas) to the P-G dispersion curves (solid lines) used in the original HIWAY model The dispersion parameters
(ar, af) are given as a function of downwind distance for each stability class. The upper bound of the new dispersion curves represents the values used for zero
cross-road wind speed situations, while the lower bound is for cross-road wind speeds greater than 3.91 m/sec.
The same meteorological conditions described by Rao, et a/.4
for the GM data set and Sistla, et a/.3 for the NY data set are
used in these model evaluations.
Regression Analysis
The GM data are segregated according to wind-road or-
ientation angle and the results of regression analysis are shown
in Table I. Although the correlation when all data combined
for HIWAY #3 is about the same as HIWAY # 1, the slope
is about three times that of HIWAY #l.This indicates that
the overpredictions of the original HIWAY are reduced
somewhat in the modified HIWAY model. The simulation of
HIWAY #4, which employs an aerodynamic drag factor as
a function of wind-road orientation angle, is much better than
the other HIWAY versions, and is quite comparable to the
simulation of the GM model. When the data are divided ac-
HIWAY #4. Although the correlation for HIWAY #1 is
comparable to that of the modified versions, the slope is still
less desirable. These results also indicate that inclusion of a
wind speed correction alone such as the one suggested by
Carpenter and Clemana14 (HIWAY #2) does not significantly
improve the model's ability to simulate the dispersion process.
The spatial variation of the regression statistics of all the
HIWAY versions and the GM model is shown in Table IV.
These results show that the predictions of HIWAY #4 are
comparable to the GM model lor receptors close to the road-
way. In general, all models appear to overpredict as the dis-
tance from the road increases. The numerical wake model
developed by Eskridge, et al.'" has an r- of 0.63, slope of 0.77,
intercept of 0.18, and standard error of estimate of 0.47 for a
sample size of 551 data points. The regression results for
HIWAY #4 are comparable (see Table I) to those of the wake
model indicating that the modified HIWAY model is as good
63
Journal of the Air Pollution Control Association
-------
Table F. Ensemble regression statistics for dispersion models. Included in the table are the
explained variance (r'2), slope (ft), intercept (a), standard error of estimate (.S'o/p) between observed
(dependent variable) and predicted (independent variable), sample size (N) for each data subset. It
is the ratio of mean observed to mean predicted concentrations for that data subset. Here the C.M
data are segregated according to wind-road orientation angle.
Data subset
Perpendicular
60" < 0 < 90°
Oblique
30° < 9 < 60°
Parallel
0° < 6 < 30°
All data
Combined
Statistical
parameter
r2
b
a
So/P
R
N
r2
b
a
So/P
R
N
r2
6
a
S0/p
R
N
r*
b
a
So/P
R
N
HIWAY
#1
0.65
0.41
0.37
0.39
0.75
173
0.37
0.22
0.65
0.60
0.61
128
0.24
0.08
0.80
0.67
0.40
293
0.26
0.11
0.75
0.64
0.50
594
HIWAY
#2
0.73
1.17
0.21
0.34
1.57
173
0.52
0.78
0.44
0.52
1.39
128
0.64
0.53
0.27
0.46
0.73
293
0.54
0.54
0.42
0.50
0.96
594
HIWAY
#*
0.81
0.80
0.12
0.29
0.94
173
0.51
0.49
0.41
0.52
0.82
128
0.28
0.23
0.69
0.65
0.72
293
0.35
0.31
0.58
0.59
0.79
594
HIWAY
#•»
0.81
().«()
0.12
0.29
0.94
173
0.61
O.fi6
0.27
0.47
0.89
128
0.71
0.91
-0.04
0.41
0.88
293
070
0.81
0.10
0.40
0.89
594
CM
0.92
0.99
0.12
0.19
1.15
173
0.77
0.97
0.25
0.36
1.15
128
0.83
0.94
0.03
0.32
1.00
293
0.83
0.94
0.12
0.30
1.06
594
as the numerical model in predicting pollutant concentrations
adjacent to this simple at-grade roadway configuration.
Cumulative Frequency Distributions
In order further to assess the model performance, various
frequency distribution plots are developed. These plots depict
the model performance in an overall statistical sense rather
than the usual one-to-one relationship afforded by the re-
gression statistics. Figure 6 shows the frequency distribution
of observed concentration and the concentration distributions
predicted by the HIWAY # 1, HIWAY #3, and HIWAY #4
models. It is evident from this plot that HIWAY #3 and
HIWAY #4 simulate the observed concentration distribu-
tion quite well compared to the original HIWAY model. The
original HIWAY model consistently overestimates the con-
centrations.
The cumulative frequency plots of (() — P)/O where 0 is
the observed and P is the predicted concentration give more
insight into the model's behavior. Such plots for HIWAY # l,
HIWAY #4, and for the GM model are shown in Figures 7 to
9. It is preferable that these plots have a gaussian shape with
Table II. Same as Table I except that the data are divided according to the stability class.
Data subset
Unstable
UtoC)
Neutral
(D)
Stable
(£ & f)
Downwind ground level
(0.5m) receptors only
Statistical
parameter
T-2
6
a
SO/P
R
N
r>
b
a
.S'o/p
ft
N
r'1
b
a
So/p
ft
N
r*
b
a
SH/P
ft
N
HIWAY
#1
0.61
0.54
0.22
0.37
0.76
276
0.69
0.45
0.18
0.39
0.56
100
0.23
0.08
0.93
0.75
0.37
218
0.21
0.09
1.00
0.78
0.47
260
HIWAY
#2
0.68
1.10
0.11
0.34
1.29
276
0.57
0.65
0.25
0.46
0.89
100
0.52
0.46
0.51
0.60
0.82
218
0.53
0.57
0.50
0.60
0.95
260
HIWAY
#3
0.60
0.72
0.15
0.38
0.89
276
0.75
0.90
0.03
0.35
0.93
100
0.29
0.22
0.80
0.72
0.68
218
0.34
0.33
0.78
0.71
0.88
260
HIWAY
#4
0.76
0.99
-0.03
0.29
0.95
276
0.78
1.02
-0.05
0.32
0.97
100
0.65
0.73
0.15
0.50
0.83
218
0.68
0.84
0.18
0.50
0.98
260
OM
0.91
1.18
0.04
0.18
1.24
276
0.80
1.01
0.05
0.31
1.06
101)
0.82
0.90
0.07
0.36
0.95
218
0.84
0.99
0.12
0.35
1.09
260
March 1980 Volume 30, No. 3
64
-------
Table III. Same as Table I except that the data are grouped according to the wind speed class.
Data subset
U < 1 m/sec
1 < U < 2.5 m/sec
U > 2.5 m/sec
All data with
U > 1 m/sec
Statistical
parameters
r*
6
a
So/P
ft
N
r*
b
a
S0/p
ft
N
r*
6
a
So/P
ft
N
r2
6
a
So/P
ft
N
HI WAY
#1
0.24
0.05
1.03
0.74
0.29
85
0.65
0.40
0.30
0.43
0.59
339
0.69
0.41
0.24
0.37
0.58
170
0.66
0.41
0.28
0.41
0.59
509
HIVVAY
#2
0.36
0.36
0.84
0.68
1.12
85
0.60
0.66
0.34
0.46
1.02
339
0.68
0.53
0.25
0.38
0.79
170
0.61
0.61
0.32
0.44
0.94
509
H1WAY
#3
0.34
0.16
0.85
0.69
0.49
85
0.71
0.84
0.01
0.39
0.85
339
0.84
1.18
-0.07
0.27
1.09
170
0.72
0.88
0.03
0.38
0.91
509
HIWAY
#4
0.67
0.62
0.22
0.48
0.75
85
0.74
0.93
-0.05
0.37
0.88
339
0.84
1.18
-0.07
0.27
1.09
170
0.75
0.95
-0.02
0.36
0.94
509
GM
0.82
0.78
0.26
0.36
0.98
85
0.87
1.10'
0.02
0.26
1.12
339
0.82
0.86
0.12
0.29
1.01
170
0.84
1.00
0.07
0.28
1.09
509
the peak at zero and rapid fall off on either side. The HIWAY
# I yields a distribution pattern (see Figure 7) that deviates
very significantly from the gaussian shape. HIWAY #4 gives
a rather smooth distribution (see Figures 8) and is quite
similar to distribution from the GM model (see Figure 9).
It is possible to derive information as to how the model
simulates the dispersion mechanism by comparing (0 — P)/0
cumulative plots of normalized concentrations. The normal-
ization is done by dividing each data point predicted by the
model by the corresponding maximum predicted for that run.
The observed data for that run are also normalized by the
100
90
80
70
60
N 50
40
30
20
10
-1.4-1.2-1.0-0.8-0.6-0.4-0.2 0.0 0.2 0.4 0.6 0.8 1.0
ALL DATA IN=5941
• OBSERVED
i HIWAY 1
9 HIWAY 3
O HiWAY 4
JO 30 40 50
CONCENTRATION INTERVAL Ippo)
Figure 6. Cumulative frequency distribution of observed
concentration is compared to the distribution provided by the
original and modified HIWAY models.
observed maximum for the run. In this way, any uncertainties
in the estimation of source strength and meteorological vari-
ables will be removed since direct comparisons of observed and
predicted concentration profiles can be made. The spatial
variation of concentration is related only to the spatial vari-
ation of dispersion parameters. Hence, comparison of nor-
malized concentration profiles will provide information on
how well a model handles the diffusion process.
Figure 7. Cumulative plot of (O — P\IO, where O is the observed and P is the
predicted concentration for the original HIWAY model. 25% of the data lie to
the left of -1. 4 (N = 594).
Comparing the cumulative plots of (On - Pn)/0n, where
the subscript n denotes normalized concentrations, for
HIWAY # 1 (see Figure 10) and HIWAY #4 (see Figure 11),
it is seen that the distribution for the former is not as smooth
as the latter. The distribution for HIWAY #4 is skewed to the
negative side indicating that the predicted concentration
profile is more than the observed profile thereby yielding
conservative estimates of pollutant levels. The distribution
for the GM model (see Figure 12) has a gaussian shape while
HIWAY #4 resembles more of a log-normal curve with neg-
ative skewness. It is of particular interest to note that the
100
90
80
70
60
N 50
40
30
20
10
_1 4 _1.2-1.0-0.8 -0.6-0.4-0.2 0.0 0.2 0.4 0.6 0.8
Figure 8. Same as Figure 7 except for the HIWAY #4 model 5 % of
the data lie to the left of -1.4 (N = 594).
65
Journal of the Air Pollution Control Association
-------
100
90
80
70
60
N 50
40
30
20
10
0
1
__rTTi
6=. rrrn 1 1 1
—
r~i
_
,-
-
— i
!__
rn-r-i r-
— 1 1 1 1 1 (=3
O-P
o
Figure 9. Same as Figure 7 except for the GM model 2% of the data lie to
the left of -1.4(N = 594).
distribution for HIWAY #4 reaches its peak about zero and
rapidly falls off to the right of the peak. This feature is quite
important for a regulatory model since environmental health
decisions will be made using the model predictions. The re-
gression statistics for normalized predictions and observations
after removing mutual values of unity are given in Table V.
These results indicate that when all the data are considered,
HIWAY #3 and HIWAY #4 simulate the dispersion process
better than HIWAY # 1 and are as good as the GM model.
The original HIWAY and the modified HIWAY version are
employed to simulate the experiments conducted in the New
York study. Table VI presents the regression statististics for
the HIWAY #1, HIWAY #4, and the GM models for the
New York data set. These results also show significant im-
provement of the HIWAY #4 model over the original
HIWAY model. The results of HIWAY #4 (using the same
aerodynamic drag factor developed with the GM data set) are
quite comparable to those of the GM model.
More quantitative information as to the actual percentage
of prediction to within a factor of 2 of the observed concen-
tration is derived from these diagrams, and the results are
summarized in Table VII. For the C1M data set, it is evident
from this table that the original HIWAY model predicts
concentrations 56% of the time to within a factor of 2 of the
observed, while the modified models (HIWAY #3 and
HIWAY #4) show a prediction to within a factor of 2 of the
observed in excess of 80% of the time. However, it should be
noted that the predictions are skewed toward overestimation.
Although the GM model has by far the best percentage in the
category of within a factor of 2, it is skewed toward underes-
timation.
For regulatory purposes, the ability of a model in providing
accurate estimates of pollutant levels in the upper 50th per-
centile of concentration is of greater importance than the
overall predictability in the entire range of concentrations.
To show how the modified model behaves in the upper half
N.
130
120-
110
100
90
80
70
60
50
40
30
20
10
0
-1.4-1.2-1.0-0.8-0.6-0.4-0.2 0.0 0.2 0.4 0.6 0.8 10
On
Figure 10. Cumulative plots of normalized concentrations of (On — Pn)/On
for the original HIWAY model. The observed and predicted concentration at
various locations for a given run are divided by their corresponding maximums
for that run. 5% of the data lie to the left of -1.4 (N - 594).
Table IV. Same as Table I except that the data are divided according to the distances of the
receptors from the roadway.
Data subset
Towers 2 & 4
(4 meters
from highway)
Tower 5
(15 meters
from highway)
Towers 1 & 6
(30 meters
from highway)
Tower 7
(50 meters
from highway)
Tower 8
(100 meters
from highway)
Statistical
parameter
r*
b
a
So/p
R
N
r2
h
a
•S'o/P
R
N
r*
h
a
•Si/P
R
N
r'i
b
a
•Vp
R
.V
r-'
h
a
•Si//'
R
,v
HIWAY
#1
0.27
0.10
1.11
0.88
0.52
182
0.55
0.30
0.50
0.36
0.60
137
0.17
0.05
0.65
0.32
0.41
183
0.73
0.31
0.22
0.15
0.61
46
0.48
0.21
0.28
0.14
0.60
46
HIWAY
#3
0.31
0.26
0.93
0.86
0.80
182
0.54
0.58
0.33
0.37
0.87
137
0.27
0.17
0.55
0.30
0.68
183
0.31
0.33
0.39
0.20
0.83
46
0.41
0.45
0.16
0.14
0.80
46
HIWAY
#4
0.75
0.90
0.09
0.52
0.96
182
0.63
0.72
0.21
0.33
0.91
137
0.43
0.41
0.36
0.27
0.80
183
0.40
0.42
0,33
0.19
0.87
46
0.50
0.55
0.13
0.13
0.83
46
GM
0.86
0.92
0.22
0.39
1.09
182
0.75
0.89
0.18
0.27
1.09
137
0.56
0.73
0.19
0.23
0.99
183
0.77
0.60
0.17
0.14
1.00
46
0.68
0.55
0.17
0.11
0.93
46
March 1980
Volume 30, No. 3
66
-------
Table V. Regression statistics between normalized observed and
normalized predicted for the original HIWAY, HIWAY #3 and
HIWAY #4, and the GM model for the CM data set. The
regression analyses here show how well the models predict the
concentration profile.
Data
subset
All data
Statistical
parameter
r*
b
a
Sa/p
N
HIWAY
#1
0.61
0.65
0.08
0.12
533
HIWAY
#3
0.85
0.76
-0.01
0.07
533
GM
0.79
0.85
0.03
0.08
533
of the cumulative frequency distribution, measured concen-
trations in the range of the 50th percentile through the 100th
percentile are separated out and the capability of the models
in predicting concentrations to within a factor of 2 is presented
in Table VIII. The modified HIWAY models show consider-
able improvement over the original HIWAY version. For the
GM data set, the GM model shows an exact prediction of 12%
and an overprediction of 19% while HIWAY #4 shows 14%
and 37% respectively. A similar trend can be seen when the
analysis is applied to the NY Data set.
Model Sen*Hlvity to Wind Direction
In order to see the sensitivity of the model predictions to
the wind-road orientation angles, contours of normalized
concentration Xp"/Q (where XP is the predicted concentration)
for the original HIWAY model and HIWAY #3 for stable
atmospheric conditions as a function of wind-road angle and
distance from the median are developed and shown in Figure
13. Also included in this diagram are the contours of observed
Xou/Q (where xo is the measured concentration) and those
predicted by the GM model for stable atmospheric conditions
for the purpose of comparison with the original and modified
HIWAY models. Only results for HIWAY #3 are included
since it has been shown (see Table V) that the simulation of
dispersion process of HIWAY #3 and #4 are similar, except
that HIWAY #4 has an aerodynamic drag factor to handle
special situations such as low wind speed conditions. These
plots for the model predictions are developed for unit wind
speed and unit source strength for a receptor height of 2 m.
Whereas the maximum measured value is of the order of 900
90
80
70
60
N 50
40
30
20
10
0
-1
-n-n-Tu
,—
-
"h-r-n r-
4-1.2-1.0-0.8-0.6-0.4-0.2 0.0 0.2 0.4 0.6 0.
On
Flgur* 11. Same as Figure 10 except for the HIWAY #4 model. 6%
of the data lie to the toft of -1.4 (N = 594).
Table VI.
models for the New York data set.
Data subset
Parallel
Oblique
Perpendicular
U < 2 m/sec
2 < U < 5 m/sec
U > 5 m/sec
Stability 2
Stability 3
Stability 4
Stability 6
All data
Model
GM
HIWAY #4
HIWAY #1
GM
HIWAY #4
HIWAY #1
GM
HIWAY #4
HIWAY #1
GM
HIWAY #4
HIWAY#1
GM
HIWAY #4
HIWAYfl
GM
HIWAY #4
HIWAY #1
GM
HIWAY #4
HIWAY #1
GM
HIWAY #4
HIWAYfl
GM
HIWAY #4
HIWAY #1
GM
HIWAY #4
HIWAY #1
GM
HIWAY #4
HIWAY # I
T*
0.92
0.81
0.72
0.73
0.74
0.57
0.87
0.86
0.67
0.89
0.91
0.73
0.80
0.70
0.61
0.88
0.86
0.85
0.78
0.70
0.55
0.93
0.89
0.85
0.78
0.79
0.63
0.95
0.97
0.66
0.86
0.81
0.66
6
0.93
0.66
0.39
0.97
0.81
0.56
0.83
0.63
0.39
0.84
0.63
0.33
0.88
0.69
0.52
1.06
1.05
0.79
0.88
0.63
0.34
0.96
0.80
0.59
0.87
0.75
0.48
0.77
0.59
0.29
0.90
0.68
0.42
a
0.33
0.71
1.16
0.28
0.24
0.54
0.62
0.77
1.36
0.66
0.68
1.45
0.58
0.68
0.98
0.02
-0.03
-0.02
0.84
1.08
1.84
0.14
0.21
0.34
0.30
0.25
0.51
0.61
-0.34
1.95
0.43
0.59
1.07
So/P
0.65
1.02
1.24
0.88
0.86
1.12
0.75
0.79
1.21
0.75
0.67
1.17
0.95
1.14
1.30
0.40
0.43
0.45
0.94
1.10
1.35
0.52
0.64
0.76
0.68
0.67
0.86
0.68
0.53
1.86
0.78
0.90
1.20
Mean Mean
N observed predicted
39
39
39
53
53
53
41
41
41
34
34
34
59
59
59
40
40
40
46
46
46
54
54
54
27
27
27
6
6
6
133
133
133
3.34
3.34
3.34
2.16
2.16
2.16
2.89
2.89
2.89
3.48
3.48
3.48
3.09
3.09
3.09
1.54
1.54
1.54
3.42
3.42
3.42
2.33
2.33
2.33
1.89
1.89
1.89
4.66
4.66
4.66
2.73
2.73
2.73
3.25
3.97
5.52
1.93
2.36
2.89
2.72
3.35
3.95
3.36
4.47
6.24
2.86
3.48
4.06
1.43
1.50
1.96
2.93
3.72
4.67
2.29
2.66
3.36
1.84
2.18
2.87
5.28
7.35
9.36
2.56
3.14
3.98
67
Journal of the Air Pollution Control Association
-------
150
140
130
120
110
100
90
80
70
60
50
40
30
20
10
-1.4-1.2-1.0-0.8-0.6-0.4-0.2 0.0 0.2 0.4 0.6 0.8 1.0
Figure 12. Same as Figure 10 except for the GM model. 2% of the data lie
to the left of -1.4 (N = 594).
3. When the P-G curves are replaced by these new dispersion
curves, it is seen that the model predicts concentrations (GM
data set) to within a factor of two of the observed 82% of the
time compared to 56% of the time for the original HIWAY
model.
4. Although the model predictions are significantly im-
proved, a slight tendency to overpredict concentrations still
exists which makes the model very useful to the regulatory
agencies in their decision-making process.
5. The modified HIWAY model provides a better simulation
of the physics of the near-roadway dispersion compared to the
original HIWAY model.
6. The fact that the modified model shows improvement in
prediction with two different data sets adds to the confidence
with which this model can be applied to major roadways.
7. With the aerodynamic drag factor, HIWAY can be applied
with greater confidence to handle dispersion even under low
wind speeds.
m ' adjacent to the roadway for near parallel wind conditions,
the original HIWAY model predicts a value of 4000 m~' for
F-stability, and 2500 m-1 for E-stability. The modified
HIWAY model predicts 1100 m"1 for stable situations. The
GM model, on the other hand, predicts a value of 600 m~' in
the immediate vicinity of the roadway for near parallel wind
situations. These diagrams clearly indicate that the predic-
tions of the modified HIWAY model are much better than the
original HIWAY model and are in good agreement with the
measured data.
Summary and Conclusions
1. The computed vertical dispersion parameters are found
to lie between A and C of the P-G stability categories.
2. Based on the information on the characteristics of the
local turbulence mechanisms, a new set of dispersion curves
applicable to roadway dispersion problems and an aerody-
namic drag factor to handle dispersion under low wind speeds
are developed.
Acknowledgements
The authors are grateful to John Wilson and Gopal Sistla
for their help in the data analysis. The authors wish to thank
William Petersen and Bruce Turner for their helpful com-
ments on the manuscript. Thanks are due to Robert Eskridge
for giving us the results of his wake model. Thanks are ex-
tended to Carol Clas and Gary Lanphear for drafting the di-
agrams, and Catherine Cassidy and Nancy Gardner for typing
the manuscript. This research is supported by the United
States Environmental Protection Agency under Grant No.
R-806017-01.
References
1. D. P. Chock, "General Motors sulfate dispersion experiment:
assessment of the EPA-HIWAY model,''JAPCA 27:39 (1977).
2. K. E. Noll, T. L. Miller and M. Claggett, "A comparison of three
highway line source models," Atmos. Environ. 12:1323 (1978).
3. G. Sistla, P. Samson, M. Keenan and S. T. Rao, "A study of pol-
lutant dispersion near highways," At mas. Environ. 13:669
(1979).
HIWAY
- STABLE
40 50 6O TO 80
DISTANCE FROM MEDIAN (m)
HI WAY **3 - STABLE
40 50 60 7O 80
DISTANCE FROM MEDIAN(m)
OBSERVED DATA - STABLE
,«" .0°
45 55 65 75 85
DISTANCE FROM MEDIAN (
GM - STABLE
f I
30
40 50 60 70 80
DISTANCE FROM MEDIANlm]
90 100
Figure 13. Variations of normalized concentration (x u/Q) with horizontal distance and wind-road angle (in degrees) for E-stability
as computed from the original HIWAY (HIWAY #1). and observed data; modified HIWAY (HIWAY #3), and the GM model, for stable
conditions.
March 1980 Volume 30, No. 3
68
-------
Table VII. Quantitative evaluation of the dispersion models tested. Here, the models' ability to
predict concentrations to within a factor of two of the observed are compared.
%of
prediction % of
within a overproduction
Sample
Model
HIWAY
#1
HIWAY
#2
HIWAY
#3
HIWAY
#4
GM
HIWAY #1
HIWAY #4
GM
Mode
All GM data
U > I m/sec
All GM data
U > 1 m/sec
All GM data
U > I m/sec
All GM data
U > 1 m/sec
All GM data
U > 1 m/sec
All NY data
All NY data
All NY data
size
594
509
594
509
594
509
594
509
594
509
133
133
133
factor
of 2
56
61
62
64
82
86
85
87
88
87
72
87
80
% of exact
prediction
5
6
6
6
10
11
11
11
10
10
7
12
14
(Obs < Pred
< 2 Obs)
37
39
30
31
41
43
44
43
27
22
42
48
30
%of
underprediction
( V2 Obs < Pred
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
T7WEPORTNO. 2.
EPA-600/8-80-018
4. TITLE AND SUBTITLE
USER'S GUIDE FOR HIWAY-2
A Highway Air Pollution Model
7. AUTHOR(S)
William B. Petersen
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Environmental Sciences Research Laboratory - RTP, NC
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
T2. SPONSORING AGENCY NAME AND ADDRESS
Environmental Sciences Research Laboratory - RTP, NC
Office of Research and Development
U.S. Environmental Protection Agency
Research Trianale Park. NC 27711
. RECIPIENT'S ACCESSION-NO.
. REPORT DATE
Mav 1980
. PERFORMING ORGANIZATION CODE
. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
C05A1A 10-0008 (FY-80)
11. CONTRACT/GRANT NO.
13. TYPE OF REPORT AND PERIOD COVERED
nhni|«;p 7/7ft - VRn
14. SPONSORING AGENCY CODE
EPA/600/09
15. SUPPLEMENTARY NOTES
16. ABSTRACT
A computer model, called HIWAY-2, that can be used for estimating the concentra-
tions of nonreactive pollutants from highway traffic is described. This steady-state
Saussian model can be applied to determine air pollution concentrations at receptor
locations downwind of "at-grade" and "cut section" highways located in relatively
uncomplicated terrain. For an at-grade highway, each lane of traffic is modeled
as though it were a finite, uniformly emitting line source of pollution. For the
cut section, the top of the cut is considered an area source. The area source is
simulted by using ten line sources of equal source strength. The total source
strength equals the total emissions from the lanes in the cut.
The air pollution concentration representative of hourly averaging times at a
downwind receptor location is found by a numerical integration along the length of
sach lane and a summing of the contributions from each lane. With the exception
Df receptors directly on the highway or within the cut, the model is applicable for
any wind direction, highway orientation, and receptor location. The model was de-
Vc \ oueu lur :> i iua u luio in WNIV-II nur i^unuai winu i i un ui,k,ui3. IMC muue i (.annul cuilbiaer
complex terrain or large obstructions to the flow such as buildings or large trees.
|7^ KEY WORDS AND DOCUMENT ANALYSIS
|a DESCRIPTORS
F~
*Air Pollution
*Atmospheric Models
Algorithms
*Dispersion
*Highways
Traffic
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
b.lDENTIFIERS/OPEN ENDED TERMS
19. SECURITY CLASS (This Report)
UNCLASSIFIED
20. SECURITY CLASS (This page)
.UNCLASSIFIED
. COSATl Field/Gioup
13B
14A
12A
21. NO. OF PAGES
80
22. PRICE
EPA Form 2220-1 (9-73)
-------
Date
Chief, Environmental Applications Branch
Meteorology and Assessment Division (MD-80)
U.S. Environmental Protection Agency
TKL PK, NC 27711
I would lite to receive future revisions to the User's Guide for HIWAY-2
Name
Organization^
Address
City State Z ip_
Hwne (Optional) ( )
------- |