MARCH 1985
SAI AIRSHED MODEL OPERATIONS MANUALS
Volume II -- Systems Manual
ATMOSPHERIC SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
-------
SAI AIRSHED MODEL OPERATIONS MANUALS
Volume II -- Systems Manual
by
J. Ames, S.R. Hayes, T.C. Myers, and D.C. Whitney
Systems Applications, Incorporated
101 Lucas Valley Road
San Rafael, California 94903
Contract No. 68-02-2429
Project Officers
Kenneth L. Demerjian
Kenneth L. Schere
Meteorology and Assessment Division
Atmospheric Sciences Research Laboratory
Research Triangle Park, North Carolina 27711
ATMOSPHERIC SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
-------
NOTICE
The information in this document has been funded
wholly or in part by the United States Environmental
Protection Agency under Contract No. 68-02-2429 to
Systems Applications, Inc. It has been subject to the
Agency's peer and administrative review, and it has
been approved for publication as an EPA document.
Mention of trade names or commercial products does
not constitute endorsement or recommendation for use.
-------
ABSTRACT
This report describes the Systems Applications, Inc. (SAI) Airshed Model
System from a programmer's point of view. Included are discussions of all
subroutines and how they fit together, run-time core allocation techniques,
internal methods of segment handling using secondary storage, and detailed
definitions of the structure of each file in the system. This manual also
includes suggestions for implementing the programs on different computers.
The SAI Airshed Model System consists of 17 programs that communicate
with each other through 17 data files; both the programs and files are des-
cribed in detail in this manual. Chapter II presents an overview of the
system, briefly describing the programs used to perform each of the five sys-
tem functions. Chapter III contains guidelines for implementing the programs,
and Chapters IV through VIII contain detailed descriptions of the programs by
function, including hierarchical subroutine tables and short module descrip-
tions. Chapter IX contains detailed file structure definitions for each file
in the system.
The User's Manual, the companion to this report, contains a general view
of the SAI Airshed Model and detailed operating instructions for the user.
Included are discussions of all the files needed to run the model, the data
preparation programs that produce each file, all the input formats and data
preparation methods, samples of input and output, and information flow diagrams
that illustrate the job stream control.
This report was submitted in fulfillment of Contract No. 68-02-2429 by
Systems Applications, Inc. under the sponsorship of the U.S. Environmental
Protection Agency. This report covers the period 23 August 1976 to 15 Septem-
ber 1979 and work was completed as of 15 September 1979.
-------
CONTENTS
ABSTRACT iii
LIST OF FIGURES vii
LIST OF TABLES ix
LIST OF EXHIBITS xi
I INTRODUCTION 1
II OVERVIEW 2
A. M.E.B. File Preparation 2
B. M.E.B. File Segmentation 5
C. Control Data Preparation 5
D. Airshed Simulation 6
E. Display and Analysis 6
III IMPLEMENTING THE SAI AIRSHED MODEL SYSTEM 7
A. Coding Conventions 7
B. Implementation Strategy 8
IV M.E.B. File Preparation Programs TO
A. Common Program Features 10
B. Utility Subroutines 12
C. Individual Program Structures T5
D. Implementing the Programs gl
V M.E.B. FILE SEGMENTATION PROGRAM 65
VI CONTROL DATA PREPARATION PROGRAMS 80
VII AIRSHED SIMULATION PROGRAM 81
A. Core Allocation 81
B. Temporary Secondary Storage Files 92
C. Boundary Handling 95
-------
D. Computing Average Concentrations 103
E. Flow of Control 104
F. COMMON Blocks 117
G. Overlays 121
H. Implementing the Program 125
I. Locations of Specific Algorithms 12g
VIII THE DISPLAY PROGRAM 132
A. Overall Logic Flow 132
B. Input/Output 132
C. Subroutine Definition and Hierarchy 124
D. Implementation Considerations 137
IX FILE STRUCTURES '• • 138
A. Input and Output File Structure Definitions 130
B. Internal File Structure Definitions 233
-------
FIGURES
Number
II-l Information Flow in the SAI Airshed Model System 3
IV-1 Typical Flow of Control for File Preparation
Programs 13
IV-2 Flow of Control for AIRQUALITY File Preparation
Program 25
IV-3 Flow of Control for BOUNDARY File Preparation
Program 28
IV-4 Flow of Control for DIFFBREAK File Preparation
Program 31
IV-5 Flow of Control for EMISSIONS File Preparation
Program 34
IV-6 Flow of Control for METSCALARS File Preparation
Program 37
IV-7 Flow of Control for PTSOURCE File Preparation
Program 40
IV-8 Flow of Control for REGIONTOP File Preparation
Program 43
IV-9 Flow of Control for ROADWAY File Preparation
Program 45
IV-10 Flow of Control for TEMPERATUR File Preparation
Program 49
IV-11 Flow of Control for TERRAIN File Preparation
Program 52
IV-12 Flow of Control for TOPCONC File Preparation
Program 55
IV-13 Flow of Control for WIND File Preparation Program 58
IV-14 Overlay Structure for Data Preparation Programs 62
-------
Number Page
V-l Flow of Control for File Segmentation 55
V-2 Flow of Control--AIRQUALITY File Segmentation Subroutine- • • 68
V-3 Flow of Control—BOUNDARY File Segmentation Subroutine • • • 59
V-4 Flow of Control—DIFFBREAK File Segmentation Subroutine • • • 70
V-5 Flow of Control—EMISSIONS File Segmentation Subroutine . . . 7-]
V-6 Flow of Control —PTSOURCE File Segmentation Subroutine • • • 72
V-7 Flow of Control—ROADWAY File Segmentation Subroutine • • • • 73
V-8 Flow of Control—REGIONTOP File Segmentation Subroutine ... 74
V-9 Flow of Control—TEMPERATUR File Segmentation Subroutine . . 75
V-10 Flow of Control—TOPCONC File Segmentation Subroutine . . . 75
V-ll Flow of Control—TERRAIN File Segmentation Subroutine .... 77
V-12 Flow of Control—WIND File Segmentation Subroutine 73
VII-1 Pointers to Variably Dimensioned Arrays g3
VII-2 The Four Tables Assigned to the Vector of
Available Space 25
VII-3 Segment Table Structure • 27
VII-4 Local Table Structure 29
VII-5 Boundary Table Structure 91
VI1-6 Segment Scratch Table Structure 93
VII-7 Core Allocation and File Storage 97
VII-8 Example of a Segmented Region 92
VII-9 Hierarchy of Top-Level Subroutines 105
VII-10 Airshed Simulation Program Overlay Structure 122
VIII-1 Schematic of the Display Program Structure 133
VIII-2 Subroutine Hierarchy Diagram 135
IX-1 Files in the SAI Airshed Model System 139
vm
-------
TABLES
Number page
II-l SAI Airshed Model Program 4
IV-1 UTILITY Modules Grouped by Usage Category 15
IV-2 Subroutines in UTILITY 18
IV-3 Subroutines in FILUTIL 23
IV-4 Routines Spec'ific to AIRQUALITY File Preparation 27
IV-5 Routines Specific to BOUNDARY File Preparation. ...'... 30
IV-6 Routines Specific to DIFFBREAK File Preparation 33
IV-7 Routines Specific to EMISSIONS File Preparation 35
IV-8 Routines Specific to METSCALARS File Preparation 39
IV-9 Routines Specific to PTSOURCE File Preparation 42
IV-10 Routines Specific to REGIONTOP File Preparation 45
IV-11 Routines Specific to ROADWAY File Preparation 43
IV-12 Routines Specific to TEMPERATUR File Preparation 51
IV-13 Routines Specific to TERRAIN File Preparation 54
IV-14 Routines Specific to TOPCONC File Preparation 57
IV-15 Routines Specific to WIND File Preparation 60
V-l File Segmentation Routines 79
VII-1 Local Variable Overlays 90
VI1-2 Arrays in the Segment Scratch Table 94
VII-3 Boundary Definition Indices 101
-------
Number
VII-4 Subroutine Hierarchy in the Airshed Simulation
Program 108
VII-5 Descriptions of Airshed Simulation Program
Subroutines 112
VII-6 COMMON Block/Subroutine Cross-Reference List 119
VI1-7 LBL Library Functions 126
VII-8 File Unit Assignments -]30
VII-9 Location of Specific Numerical Functions ]31
VIII-1 Subroutine Descriptions 135
IX-1 SAI Airshed Model System Files and Their Descriptions .... 140
IX-2 AIRQUALITYTile Stucture 145
IX-3 AVERAGE File Structure 151
IX-4 BOUNDARY File Structure 157
IX-5 CHEMPARAM File Structure 163
IX-6 DIFFBREAK File Structure 169
IX-7 EMISSIONS File Structure 174
IX-8 INSTANT File Structure 18Q
IX-9 METSCALARS File Structure 185
IX-10 PTSOURCE File Structure 191
IX-11 REGIONTOP File Structure 197
IX-12 ROADWAY File Structure • • • 202
IX-13 SIMCONTROL File Structure 207
IX-14 TEMPERATUR File Structure 211
IX-15 TERRAIN File Structure 216
IX-16 TOPCONC File Structure 221
IX-17 VARIATION File Structure 226
IX-18 WIND File Structure 231
-------
EXHIBITS
Number page
VII-1 Code Fragments Implementing the Core Allocation Scheme . ... 84
-------
I INTRODUCTION
This report describes the Systems Applications, Inc. (SAI) Airshed
Model System from a programmer's point of view. Included are discussions
of all subroutines and how they fit together, run-time core allocation
techniques, internal methods of segment handling using secondary storage,
and detailed definitions of the structure of each file in the system.
This manual also includes suggestions for implementing the programs on
different computers.
The SAI Airshed Model System consists of 17 programs that communi-
cate with each other through 17 data files; both the programs and files
are described in detail in this manual. Chapter II presents an over-
view of the system, briefly describing the programs used to perform
each of the five system functions. Chapter III contains guidelines for
implementing the programs, and Chapters IV through VIII contain detailed
descriptions of the programs by function, including hierarchical sub-
routine tables and short module descriptions. Chapter IX contains detailed
file structure definitions for each file in the system.
The User's Manual, the companion to this report, contains a general
view of the SAI Airshed Model and detailed operating instructions for
the user. Included are discussions of all the files needed to run the
model, the data preparation programs that produce each file, all the
input formats and data preparation methods, samples of input and output,
and information flow diagrams that illustrate the job stream control.
-------
II OVERVIEW
Figure II-l shows the flow of information through the SAI Airshed
Model System. Each rectangular box in the diagram represents one of the
five major functions of the system:
> Meteorology, emissions, and initial and boundary conditions
(M.E.B.) file preparation.
> M.E.B. file segmentation.
> Control data preparation.
> Airshed simulation.
> Display and analysis.
These five functions were described from the user's point of view in
the User's Manual. This report describes the characteristics of the
computer programs that perform each function.
There are 17 programs in the SAI Airshed Model System. Table II-l
lists the programs, grouped by function. The remainder of this chapter
presents general descriptions of the programs that constitute each of
the five functions and discusses additional programs. For a particular
application, additional programs can be developed to convert raw data
into the input format required by the data preparation programs, or to
create a data file directly, bypassing the existing data preparation
program.
A. M.E.B. FILE PREPARATION
The 12 programs that prepare the 12 meteorology, emissions, and
initial and boundary conditions (M.E.B.) files each perform the follow-
ing functions:
-------
Meteorological,
Emissions, and
Initial and
Boundary
Conditions Data
IV-1-13)
Control
Data
(Fig-. IV-D
(Fig. VII-9)
Display
Controls
f Predicted / L other |
Concentrations Concentration
V V VI Oata '
Display and
Analysis
(Fig. VIII-1)
Haps
Contours
Stations
Differences
Figure II-l. Information Flow in the SAI Airshed Model System
-------
Table II-l. SAI Airshed Model Programs
Function
M.E.B. File Preparation
Program
Purpose
M.E.B. File Segmentation
Control Data Preparation
Airshed Simulation
Display and Analysis
AIRQUL
BNDARY
DFSNBK
EMISSN
METSCL
PTSRCE
REGNTP
ROADWY
TEMPTR
TERAIN
TPCONC
WINDSET
Creates
Creates
Creates
Creates
Creates
Creates
Creates
Creates
Creates
Creates
Creates
Creates
SEGDEF
CPREP
SPREP
AIRSIM
DISPLAY
the AIRQUALITY file
the BOUNDARY file
the DIFFBREAK file
the EMISSIONS file
the METSCALARS file
the PTSOURCE file
the REGIONTOP file
the ROADWAY file
the TEMPERATUR file
the TERRAIN file
the TOPCONC file
the WIND file
Segments any spatially varying M.E.B.
file
Creates the CHEMPARAM file
Creates the SIMCONTROL file
Performs the airshed simulation and
writes the INSTANT, AVERAGE, and
VARIATIONS files
Generates concentration maps and
vertical profiles
-------
> Read and verify input card packets that control the data pre-
paration algorithms to be used, and supply observed data points.
> Allocate space for each necessary array in a "scratch vector."
> Assign "core extension" files, if necessary, for temporary
storage of large arrays.
> Perform required computations to generate gridded data.
> Write output data files.
The programs were designed together as a unified package with standard-
ized input, processing, and output modules. Thus, they all have the
same basic structure and share many general purpose utility and file
handling subroutines. These shared subroutines are kept in two separate
libraries where they need be compiled and stored only once, but can be
picked up by the loader (collector, linkage editor) when each program is
prepared for execution by the system.
B. M.E.B. FILE SEGMENTATION
M.E.B. files containing data that vary spatially may need to be
segmented before they are passed to the Airshed Simulation Program.
The file segmentation program--SE6DEF--can be used to segment all the
spatially varying M.E.B. files. This program reads control cards that
define the segment layout and constructs a segment grid, verifying the
proper overlapping of segment boundaries. It then reads the file to be
segmented and uses the segment grid to direct output of data arrays by
segment to a new file. The file segmentation program uses a scratch
vector for allocation of arrays and, when necessary, core extension
files for large arrays. It also uses many of the same utility subrou-
tines used by the M.E.B. file preparation programs.
C. CONTROL DATA PREPARATION
Two programs, CPREP and SPREP, prepare the two control data files,
CHEMPARAM and SIMCONTROL, respectively. They are short, simple pro-
grams, without subroutines, that read cards, perform simple validation
tests, and write files in a straightforward manner.
-------
D. AIRSHED SIMULATION
The Airshed Simulation Program, AIRSIM, reads the files created
in the first three steps. It performs the simulation and writes its
results, as instantaneous and time-averaged concentrations over the
entire grid, to the simulation results files. It also produces a trace
of its execution steps. This program contains the following features:
> Run-time allocation of arrays of variable-lengths within a
vector of available space.
> Overlay structure to maximize the length of that vector.
> Use of temporary (scratch) files for intermediate storage of
data arrays by segment.
E. DISPLAY AND ANALYSIS
The display and analysis function includes all programs that pro-
cess the results of the airshed simulation. Currently, there is one
such program—DISPLAY. It reads the simulation results files and a
control deck, including station locations; it produces concentration
maps at any selected level as well as vertical concentration profiles
at the stations.
-------
Ill IMPLEMENTING THE SAI AIRSHED MODEL SYSTEM
Since the SAI Airshed Model System contains so many separate pro-
grams and data files, its initial implementation on a different com-
puter may seem a formidable task. To simplify the conversion ta,sk,
this chapter describes some of the coding conventions used and pre-
sents an eight-step approach to the implementation of the system as a
whole.
A. CODING CONVENTIONS
All SAI Airshed Model programs are written in FORTRAN IV and were
originally implemented on the Lawrence Berkeley Laboratory (LBL) CDC 7600
computer system. The programs were subsequently modified to run on the
Environmental Protection Agency's UNIVAC 1100 computer. This report
discusses the version of the codes implemented on the EPA's UNIVAC computer.
The following are some features of the code for this system:
> There are no multiple entry points or returns.
> No arrays are greater than three-dimensional.
> There are no complex or double-precision variables.
> There is- no rereading of input data cards.
> All alphanumeric variables are represented by vectors
containing one character per word, left justified.
> Output format statements use Hollerith counts rather
than special delimiters for alphanumeric text fields.
> Routines accessing secondary storage files or large
core memory are isolated modules that can be easily
changed.
> All main programs begin with comments naming all files
that the program uses and, if necessary, providing
equivalences between local file names and FORTRAN unit
numbers.
7
-------
B. IMPLEMENTATION STRATEGY
This section briefly sketches the tasks necessary to make the
SAI Airshed Model System operational on a different computer.
1. First, determine the locations of the programs and test
decks. Read in each source file (program or data) and
store it by name.
2. Locate @ ELT cards. These cards precede each predefined
main program, COMMON block, and subroutine. Change or
remove these cards as necessary, and attempt to compile
each program module.*
3. Locate and change all language incompatibilities. Change
large core and secondary storage access routines as
required. Special implementation considerations for each
program are described as necessary in Chapters IV through
VIII.
4. Load (or link/edit or collect) each program, using sub-
routine libraries and overlay structures where necessary.
Adjust the scratch vector in each program to make opti-
mal use of available space.
5. Run each data preparation program with the test data
decks provided. Refer to the job control diagrams in
Chapters VII and IX of the User's Manual. Save the files
produced.
6. Determine whether the segment size defined in the test
data deck will fit in the Airshed Simulation Program
*Some compilers using the FORTRAN-77 standard will cause fatal errors to
occur at program execution time if real arrays are passed through integer
arrays from a main program to its subroutines. Because the Airshed Model
uses this technique, the variable-type checking should be turned off
during the compile step (usually by invoking the FORTRAN-66 compiler option)
8
-------
(see Section VIII.A of the User's Manual). If not, make
a new segment definition deck. Segment all spatially
varying data files and save the segmented versions.
7. Run the Airshed Simulation Program, saving the results.
8. Run the display program, using the sample control deck
and the simulation results files.
Chapter V of the User's Manual contains a worksheet that can be
used to keep track of the locations and special characteristics of each
program. It also contains a summary of file characteristics that can
be used to aid in the preparation of job control specifications.
-------
IV M.E.B, FILE PREPARATION PROGRAMS '
All of the 12 programs used for preparing the M.E.B. files have
the same basic structure and use many routines in common. This chapter first
describes the features common to all the programs, including general purpose
routines, scratch storage, core extension, and user subroutines. Next, a list
of the utility modules is presented, followed by a detailed discussion of the
structure of the programs with charts and tables showing the utility and
special purpose routines called by each one. Finally, some of the consider-
ations involved in implementing the programs are discussed, including source
deck structure, scratch vector length, overlaying, I/O error handling, and
core extension.
A. COMMON PROGRAM FEATURES
The M.E.B. file preparation programs were designed and written as a
unified system of programs that incorporate standardized techniques of data
transfer, naming conventions, subroutine interfaces, and program structure in
general. The following features are common to all the M.E.B. file preparation
programs.
1. General Purpose Routines
There are two libraries of utility modules that can be used by more than
one program. The first library--UTILITY—contains routines that read the
standard data packets, perform the method and vertical method calculations,
convert input units to internal units, allocate and initialize scratch storage,
perform data transfers involving core extension, if any, and control reads to
other data files. Each of the routines in the second library—FILUTIL--
performs either a single read or write to one of the 12 data files.
Included are reads and writes of header records, other time-invariant
records, time interval records, and other time-varying records.
10
-------
2. Scratch Allocation
To provide maximum flexibility for defining region size and item counts,
array sizes have not been preassigned in the FORTRAN programs. Instead, all
available storage is contained in a single "scratch" vector. Once the system
parameters are read from the CONTROL and REGION packets, the scratch storage
vector is partitioned by a subroutine—SCRPRT-- that constructs a pointer to
successively higher locations in the scratch vector for each array required.
The value of each pointer is calculated on the basis of the size of the preced-
ing array, with the final pointer indicating the beginning of "available scratch"
storage, which can be used by any module that needs more space for its calcula-
tions.
When the processing routine is called, the array variables are passed as
arguments by specifying their locations in the scratch vector. In the proces-
sing routine itself, the actual array names appear in the argument list, and
the appropriate set of variable dimensions is assigned. For a more detailed
description of run-time scratch storage allocation, see Section VILA.
3. Core Extension
Some of the programs require three-or four-dimensional arrays. In these
programs, a calculation is made to determine whether the scratch vector is
large enough to contain all of the arrays. If it is not, a special "core
extension" routine—MATSET--is called that creates a random access file for
storing the entire array in two-dimensional slices and allocates space in
scratch storage for a single two-dimensional slice. (Note: This feature
has been disabled for use on the UNIVAC; it is a CDC computer feature.)
Access to such arrays is provided through modules that transfer informa-
tion between the two-dimensional slice and the entire array, whether that
array is in core or in a special file. See Section D of this chapter for
suggestions on implementing this feature on different computers.
4. Typical Program Structure
All of the programs have the same basic structure. The main routine
reads the CONTROL and REGION data packets, reads the header records from any
11
-------
input files, and writes the header records to the output file. The main
program also partitions scratch storage for use by the rest of the program,
sets up core extension if necessary, and finally calls the processing rou-
tine. The processing routine reads the remainder of the user's time in-
variant and time-varying data packets, performs all calculations on the in-
put data, and writes the remainder of the output file. It uses a variety
of standard utility modules and, for most programs, a few special-purpose
modules. Figures IV-l{a) and (b) illustrate the general flow of control
for all of the M.E.B. file preparation programs. These figures are used
again in Section C of this chapter to show the specific subroutines used
by each program.
5. User Subroutines
Each of the data processing programs except that for METSCALARS has the
capability of calling one user-written subroutine for computations at ground
level and another for vertical computations, if any. The interfaces for these
routines are included in the code, and dummy user routines have been supplied.
There is no restriction on the functions of the user routines, and all vari-
ables from the processing routine are included in each subroutine call.
Access to such routines is accomplished by the specification of USER in the
METHOD packet or VERTUSER in the VERTICAL METHOD packet.
B. UTILITY SUBROUTINES
The two libraries of utility subroutines—UTILITY AND FILUTIL—are des-
cribed in this section.
1. UTILITY
The UTILITY library subroutines can be grouped into the following cate-
gories of uses:
> Text manipulation
> Input file read
> Scratch storage manipulation
> Core extension
> Packet read
12
-------
Read CONTROL
and REGION
packets
Read header
records from
input files,
if any
Write output
file header
records
Partition
scratch
storage
Set up core
extension,
if necessary
Call processing
routine to do
everything else
(a) Main Program
Figure IV-1.
Typical Flow of Control for
File Preparation Programs
13
-------
Read time-invariant
packets, if any
Write time-invariant
records, if any
Read time interval
and time-varying
input files,
if any
Read time-varying
packets until
ENDTIME
Perform
method
processing
Perform vertical
method processing,
if any
Write output
file and print
(b) Processing Routine
Figure IV-1 (Concluded)
14
-------
> Unit conversion
> Method computation
> Vertical method computation
> Grid printing.
Table IV-1 lists the names of the subroutines in each group. Table IV-2
lists alphabetically and describes all subroutines in UTILITY and shows
the other subroutines that each calls.
2. FILUTIL
The subroutines in the FILUTIL library perform reads or writes directly
to the data files. All subroutine names are six characters long, beginning
with "FL" and ending with "RD" for read or "WR" for write. The middle two
letters indicate the type of record being written. The subroutines in FILUTIL
can be called by main programs, processing routines, and subroutines in UTILITY.
Table IV-3 is an alphabetical list of all subroutines in FILUTIL with a short
description of each one.
C. INDIVIDUAL PROGRAM STRUCTURES
This section contains the following information for each program:
> A structure chart showing all subroutines directly called by the
main program and its processing routine (Figures IV-2 through IV-13).
> A table listing all routines not contained in either of the
libraries (Tables IV-4 through IV-15).
Each structure chart follows the same general pattern of processing steps
illustrated and described in Section A of this chapter. Next to each box
(indicating a processing step) in the diagram is a list of subroutines called
to perform that function. Other subroutines called by these subroutines are
not shown in the figure, but are listed in the corresponding subroutine table
or in the UTILITY subroutine table (IV-2). If any processing step is not
performed by a particular program, the word "none" appears next to the
corresponding box. The structure charts and tables appear in alphabetical
order according to the names of the files they create.
15
-------
Table IV-1. UTILITY Modules Grouped by Usage Category
Usage Category
Text manipulation
Input file read
Scratch storage manipulation
Core extension
Packet read
Module
COMPAR
TXTPAK
DFBKRD
MTSCRD
RGTPRD
TMPSRD
TPCNRD
WINDRD
SCRPRT
SCRSET
SINVAL
COREXT
MATGET
MATLOD
MATPUT
MATSET
MATSTR
CNSTRD
CNTLRD
GDVLRD
MTHDRD
REGNRD
SBRGRD
STATRD
STVLRD
TIMERD
UNITRD
VPRFRD
16
-------
Table IV-1 (Concluded)
Usage Category Module
Unit conversion UNTCNV
Method computation EWINTR
NSINTR
POISSP
POISVR
SPLITC
STASGN
Vertical method computation ABSPRF
ABPRFR
CELLHT
EWINTR
HTZCLL
NSINTR
RELPRF
RLPRFR
Grid printing GRIDPR
17
-------
Table IV-2. Subroutines in UTILITY
Name
ABSPRF
ABPRFR
Description
CELLHT
CNSTRD
Absolute profile method—calculates the value at level
aloft from user profile
Absolute profile ratio method—calculates the
value at level aloft from user profile, relative
to top and bottom values
Calculates the height of the center of a vertical cell
above ground level
•
Reads the CONSTANTS packet, performs unit conversions
if necessary, and stores values in the output matrix
CNTLRD Reads the CONTROL packet
COMPAR Compares two strings character by character for
equality
COREXT Sets up core extension for matrix if it will not fit
in core
•
DFBKRD Reads the time-variant Diffusion Break Record from
DIFFBREAK
FILVER Reads and verifies file header records
Subroutines
Called
CELLHT
CELLHT
TXTPAK
COMPAR
UNTCNV
MATGET
MATPUT
TXTPAK
COMPAR
TIMERD
SCRSET
MATSET
MATPUT
COMPAR
FLDBRD
FLTMRD
COMPAR
FLHDRD
FLRGRD
FLSGRD
FLSPRD
18
-------
Table IV-2 (Continued)
Name
EWINTR
GDVLRD
Description
Subroutines
Called
GRIDPR
HTZCLL
MATGET
MATLOD
MATPUT
MATSET
MATSTR
MTHDRD
MTSCRD
NSINTR
Interpolates a value on an east-west line from values
in adjacent subregions
Reads the GRID VALUES packet
Prints a grid of output values in accordance with a
selectable format
Calculates the vertical cell number for a given
height above ground level
Gets a matrix from scratch vector or from core
extension
Recovers a record from random access mass storage
Puts a matrix to scratch vector or to core extension
Sets up a random access mass storage file
Stores a record on random access mass storage
Reads the METHOD packet
Reads time-variant records on METSCALARS
Interpolates a value on a north-south line from
values in adjacent subregions
TXTPAK
COMPAR
UNTCNV
MATGET
MATPUT
MATLOD
MATSTR
TXTPAK
COMPAR
UNTCNV
COMPAR
FLMSRD
FLTMRD
19
-------
Table IV-2 (Continued)
Name
POISSP
Description
Calculates full grid distribution of station values
(concentrations) in one subregion by means of the
POISSON method for each variable
Subroutines
Called
SINVAL
COMPAR
POISVR Calculates full grid distribution of station values
(meterological and other variables) in one subregion
by means of the POISSON method for each variable
REGNRD Reads the REGION packet
RELPRF Calculates the value at level aloft from user profile,
adjusted for variation of diffusion break
RFTPRD Reads time-variant records on REGIONTOP
RLPRFR Calculates the value at level aloft from user profile,
relative to top and bottom values, adjusted for vari-
ation of diffusion break
SBRGRD Reads the SUBREGION packet
SCRPRT Allocates scratch vector to the arrays required
SCRSET Allocates space to a single array and initializes
space
SINVAL Places a single value into every location of a vector
SPLITC Calculates the value of a variable as a function of
the values of a set of other variables
SINVAL
TXTPAK
CELLHT
COMPAR
FLRTRD
FLTMRD
CELLHT
TXTPAK
COMPAR
SINVAL
GRIDPR
SCRSET
SINVAL
COMPAR
MATGET
MATPUT
20
-------
Table IV-2 (Continued)
Name
STASGN
STATRD
Description
Subroutines
Called
Calculates station-based interpolation values for a
single grid square
Reads the STATIONS packet
STVLRD- Reads the STATION VALUES packet
TIMERD Reads a time interval card (beginning and ending dates
and times) .
TMPSRD Reads time-varying records on TEMPERATUR
TPCNRD Reads time-varying records on TOPCONC
TXTPAK Left-justifies text string, removing blanks, with
blank fill at right
UNITRD Reads the UNITS packet
UNTCNV Uses unit conversion tables to convert from input
units to internal units
VPRFRD Reads the VERTICAL PROFILES packet
TXTPAK
COMPAR
UNTCNV
TXTPAK
COMPAR
UNTCNV
COMPAR
FLTPRD
FLTMRD
COMPAR
FLTCRD
FLTMRD
MATPUT
TXTPAK
COMPAR
TXTPAK
COMPAR
UNTCNV
21
-------
Table IV-2 (Concluded)
Subroutines
Name Description Called
WINDRD Reads time-varying records on WIND COMPAR
FLWDRD
FLTMRD
WNDSPD Calculates maximum absolute wind speeds in x- and MATGET
y-directions and average wind speeds at each boundary
22
-------
Table IV-3. Subroutines in FILUTIL
Name Description
FLAQRD Reads Air Quality Records for a segment
FLAQWR Writes Air Quality Records for a segment
FLBIRD Reads Boundary Definition Record for one edge
FLBIWR Writes Boundary Definition Record for one edge
FLBVRD Reads Boundary Concentration Record for one edge
FLBVWR Writes Boundary Concentration Record for one edge
FLDBRD Reads Diffusion Break Record for a segment
FLDBWR Writes Diffusion Break Record for a segment
FLEMRD Reads Emissions Records for a segment
FLEMWR Writes Emissions Records for a segment
FLHDRD Reads File Description Header Record
FLHDWR Writes File Description Header Record
FLMSRD Reads Meteorological Scalars Record
FLMSWR Writes Meteorological Scalars Record
FLPIRD Reads time-invariant Counter Record and Point
Source Definition Record
FLPIWR Writes time-invariant Counter Record and Point
Source Definition Record
FLPVRD Reads time-varying Counter Record, Point Source
Definition Record, and Point Source Emissions
Record
FLPVWR Writes time-varying Counter Record, Point Source
Definition Record, and Point Source Emissions
Record
23
-------
Table IV-3 (Concluded)
Name Description
FLRGRD Reads the Region Definition Header Record
FLRGWR Writes the Region Definition Header Record
FLRTRD Reads the Region Top Record
FLRTWR Writes the Region Top Record
FLRWRD Reads Roadway Emissions Records
FLRWWR Writes Roadway Emissions Records
FLSGRD Reads the Segment Definition Header Record
FLSGWR Writes the Segment Definition Header Record
FLSPRD Reads the Species Definition Header Record
FLSPWR Writes the Species Definition Header Record
FLTCRD Reads Top Concentration Records
FLTCWR Writes Top Concentration Records
FLTMRD Reads Time Interval Record
FLTMWR Writes Time Interval Record
FLTPRD Reads Temperature Record
FLTPWR Writes Temperature Record
FLTRRD Reads Terrain Records
FLTRWR Writes Terrain Records
FLWDRD Reads time-varying Wind Records
FLWDWR Writes time-varying Wind Records
24
-------
AIRQUL
(Startj
Read CONTROL
and REGION
packets
Read header
records from
input files,
if any
Write output
file header
records
Partition
scratch
storage
Set up core
extension,
if necessary
CNTLRD
REGNRD
FILVER
FLHDWR
FLRGWR
FLSGWR
FLSPWR
SCRPRT
COREXT
Call processing
routine to do
everything else
'AIRPRC
(a) Main Program
Figure IV-2. Flow of Control for AIRQUALITY
File Preparation Program
25
-------
AIRPBC
Read time-invariant
packets, if any
Write time-invariant
records, if any
Read time interval
and time-varying
input files,
if any ,
Read time-varying
packets until
ENDTIME
Perform
method
processing
Perform vertical
method processing,
if any
Write output
file and print
TXTPAK
COMPAR
UNITRD
STATRD
none
TIMERD
DFBKRD
RGTPRD
TPCNRD
TXTPAK
COMPAR
SBRGRD
MTHDRD
MATGET
STASGN
POISSP
SPLITC
MATGET
ABSPRF
RELPRF
ABPRFR
PLPRFR
FLTMWR
MATGET
FLAQWR
GRIDPR
(b) Processing Routine
Figure IV-2 (Concluded)
CNSTRD
GDVLRD
STVLRD
VPRFRD
AIRUSR
EWINTR
NSINTR
MATPUT
AIRVUS
EWINTR
NSINTR
MATPUT
26
-------
Table IV-4. Routines Specific to AIRQUALITY File Preparation
Name Description Routines Called
AIRPRC Processing routine See Figure IV-2(b)
AIRQUL Main program See Figure IV-2(a)
AIRUSR USER method dummy subroutine
AIRVUS VERTUSER vertical method dummy subroutine
27
-------
BNDARY
Read CONTROL
and REGION
packets
Read header
records from
input files,
if any
Write output
file header
records
Partition
scratch
storage
Set up core
extension,
if necessary
CNTLRD
REGNRD
FILVER
FLHDWR
FLHGWR
FLSGWR
FLSPWR
SCRPRT
COREXT
Call processing
routine to do
everything else
BNDPRC
(a) Main Program
Figure IV-3.
Flow of Control for BOUNDARY
File Preparation Program
28
-------
BNDPRC
Read time-invariant
packets, if any
Write time-invariant
records, if any
Read time interval
and time-varying
input files,
if any
Read time-varying
packets until
ENDTIME
Perform
method
processing
Perform vertical
method processing,
if any
Write output
file and print
TXTPAK
COMPAR
UNITRD
BLINRD
FLBIWR
TIMERD
DFBKRD
RGTPRD
TPCNRD
TXTPAK
COMPAR
MTHDRD
MATGET
SPLTBC
BNDUSR
ABSPRF
RELPRF
ABPRFR
FLTMWR
FLBVWR
GRIDPR
(b) Processing Routine
Figure IV-3 (Concluded)
BLVLRD
VPRFRD
RLPRFR
BNDVUS
29
-------
Table IV-5. Routines Specific to BOUNDARY File Preparation
Name
BLINRD
Description
BLVLRD
BNDARY
BNDPRC
BNDUSR
BNDVUS
SPLTBC
Reads BOUNDARIES packet and sets boundary
cells
Reads BOUNDARY READINGS or CONSTANTS
packet
Main program
Processing routine
USER method dummy subroutine
VERTUSER vertical method dummy subroutine
Performs split/combination method on
boundary concentrations
Routines Called
COMPAR
TXTPAK
UNTCVN
SINVAL
6RIDPR
TXTPAK
COMPAR
UNTCNV
See Figure IV-3(a)
See Figure IV-3(b)
30
-------
DFSNBK
Figure IV-4.
Read CONTROL
and REGION
packets
Read header
records from
input files,
if any
Write output
file header
records
Partition
scratch
storage
Set up core
extension,
if necessary
CNTLRD
REGNRD
none
FLHDWR
FLRGWR
FLSGWR
SCRPRT
COREXT
Call processing
routine to do
everything else
DFBPRC
&
(a) Main Program
Flow of Control for DIFFBREAK
File Preparation Program
31
-------
DFBPRC
Read time-invariant
packets, if any
Write time-invariant
records, if any
Read time interval
and time-varying
input files,
if any
Read time-varying
packets until
ENDTIME
Perform
method
processing
Perform vertical
method processing,
if any
Write output
file and print
TXTPAK
COMPAR
UNITRD
STATRD
none
TIMERD
TXTRAK
COMPAR
SBRGRD
MTHDRD
MATGET
STASGN
POISVR
none
FLTMWR
MATGET
FLDBWR
GRIDPR
(b) Processing Routine
Figure IV-4 (Concluded)
CNSTRD
GDVLRO
STVLRD
DFBUSR
EVJINTR
NSINTR
MATPUT
32
-------
Table VI-6. Routines Specific to DIFFBREAK File Preparation
Name Description Routines Called
DFBPRC Processing routine See Figure IV-4(b)
DFBUSR USER method dummy subroutine
DFSNBK Main program See Figure IV-4(a)
33
-------
EMSPRC
Read CONTROL
and REGION
packets
Read header
records from
input files,
if any
Write output
file header
records
Partition
scratch
storage
Set up core
extension,
if necessary
Call processing
routine to do
everything else
CNTLRD
REGNRD
none at present
FLHDWR
FLRGWR
FLSGWR
FLSPWR
SCRPRT
COREXT
EMSPRC
(a) Main Program
Figure IV-5. Flow of Control for EMISSIONS
File Preparation Program
34
-------
EMSPRC
Read time-invariant
packets, if any
Write time-invariant
records, if any
Read time interval
and time-varying
input files,
if any
Read time-varying
packets until
ENDTIME
Perform
method
processing
Perform vertical
method processing,
if any
Write output
file and print
TXTPAK
COMPAR
UNITRD
none
TIMERD
TXTPAK
COMPAR
SBRGRD
MTHDRD
MATGET
SPLITC
EMSUSR
MATPUT
none
FITMWR
MATGET
FLEMWR
GRIDPR
CNSTRD
GDVLRD
(b) Processing Routine
Figure IV-5 (Concluded)
35
-------
Table IV-7. Routines Specific to EMISSIONS File Preparation
Name Description Routines Called
EMISSN Main program See Figure IV-5(a)
EMSPRC Processing routine See Figure IV-5(b)
EMSUSR USER method dummy subroutine
36
-------
METSCL
Read CONTROL
and REGION
packets
CNTLRD
RE6NRD
Read header
records from
input files,
if any
none
Write output
file header
records
FLHDWR
FLRGWR
Partition
scratch
storage
SCRPRT
Set up core
extension,
if necessary
none
Call processing
routine to do
everything else
MTSPRC
(a)
Figure IV-6.
Main Program
Flow of Control for METSCALARS
File Preparation Program
37
-------
MTSPRC
Read time-invariant
packets, if any
Write time-invariant
records, if any
Read time interval
and time-varying
input files,
if any
Read time-varying
packets until
ENDTIME
Perform
method
processing
Perform vertical
method processing,
if any
Write output
file and print
TXTPAK
COMPAR
UNITRD
none
TIMERD
TXTPAK
COMPAR
UNTCNV
none
none
FLTMWR
FLMSWR
(b) Processing Routine
Figure IV-6 (Concluded)
33
-------
Table IV-8. Routines Specific to METSCALARS File Preparation
Name Description Routines Called
METSCL Main program See Figure IV-6(a)
MTSPRC Processing routine See Figure IV-6(b)
39
-------
PTSRCE
Read CONTROL
and REGION
packets
Read header
records from
input files,
if any
Write output
file header
records
Partition
scratch
storage
Set up core
extension,
if necessary
CNTLRD
REGNRD
FILVER
FLHDWR
FLRGWR
FLSGWR
FLSPWR
SCRPRT
Call processing
routine to do
everything else
PTSPRC
(a)
Figure IV-7.
©
Main Program
Flow of Control for PTSOURCE
File Preparation Program
40
-------
PTSPRC
Read time-invariant
packets, if any
Write time-invariant
records, if any
Read time interval
and time-varying
input files,
if any
Read time-varying
packets until
ENDTIME
Perform
method
processing
Perform vertical
method processing,
if any
Write output
file and print
(Return)
(.b) Processing Routine
Figure IV-7 (Concluded)
TXTPAK
COMPAR
UNITRD
PTSRRD
FLPIWR
TIMERD
DFBKRD
RGTPRD
TXTPAK
COMPAR
MTHDRD
SPLTPS
HTZCLL
PLMRIS
FLTMWR
FLPVWR
SINVAL
GRIDPR
TMPSRD
MTSCRD
WINDRD
PSEMRD
PSFCRD
PTSUSR
PTSVUS
41
-------
Table VI-9. Routines Specific to PTSOURCE File Preparation
Name Description Routines Called
PLMRIS Calculates plume rise using Briggs
formulae
PSEMRD Reads EMISSIONS VALUES packet TXTPAK
COMPAR
UNTCNV
PSFCRD Reads EMISSIONS FACTORS packet TXTPAK
COMPAR
PTSPRC Processing routine See Figure IV-7(b)
PTSRCE Main program See Figure IV-7(a)
PTSRRD Reads POINT SOURCES packet TXTPAK
COMPAR
UNTCNV
PTSVUS VERTUSER vertical method dummy subroutine
SPLTPS Performs split/combination method on point COMPAR
source emissions
42
-------
Read CONTROL
and REGION
packets
Read header
records from
input files,
if any
Write output
file header
records
Partition
scratch
storage
Set up core
extension,
if necessary
Call processing
routine to do
everything else
CNTLRD
REGNRD
FILVER
FLHDWR
FLRGWR
FLSGWR
SCRPRT
COREXT
RTPPRC
Main Program
Figure IV-8.
Flow of Control for REGIONTOP File
Preparation Program
43
-------
RTPPRC
Read time-invariant
packets, if any
Write time-invariant
records, if any
Read time interval
and time-varying
input files,
if any
Read time-varying
packets until
ENDTIME
Perform
method
processing
Perform vertical
method processing,
if any
Write output
file and print
(b) Processing Routine
Figure IV-8 (Concluded)
TXTPAK
COMPAR
UNITRD
STATRD
none
TIMERD
DFBKRD
TXTPAK
COMPAR
SBRGRD
MTHDRD
MATGET
STASGN
POISVR
FXDHGT
SAMHGT
none
FLTMWR
MATGET
FLRTWR
GRIDPR
CNSTRD
GDVLRD
STVLRD
RTPUSR
EWINTR
NSINTR
MATPUT
44
-------
Table IV-10. Routines Specific to REGIONTOP File Preparation
Name Description Routines Called
FXDHGT Performs FIXDHEIGHT method
REGNTP Main program See Figure IV-8(a)
RTPPRC Processing routine See Figure IV-8(b)
RTPUSR USER method dummy subroutine
SAMHGT Performs SAMEHEIGHT method
45
-------
ROADWY
Read CONTROL
and REGION
packets
Read header
records from
input files,
if any
Write output
file header
records
Partition
scratch
storage
Set up core
extension,
if necessary
Call processing
routine to do
everything else
CNTLRD
REGNRD
none at present
FLHDWR
FLRGWR
FLSGWR
FLSPWR
SCRPRT
COREXT
RDWPRC
(End)
(a) Main Program
Figure IV-9.
Flow of Control for ROADWAY File
Preparation Program
-------
RDWPRC
\
Read time-invariant
packets, if any
TXTPAK
COMPAR
UNITRD
RDWYRD
Write time- in variant
records, if any
none
Read time interval
and time- varying
input files,
if any
Read time- vary ing
packets until
ENDTIME
,_
Perform
method
processing
Perform vertical
method processing,
if any
Write output
file and print
1
TIMERD
TXTPAK CNSTRD
COMPAR RDVLRD
SBRGRD STVLRD
MTHDRD VPRFRD
MATGET RDWUSR
RDASGN MATPUT
SPLITC
none
FLTMWR
MATGET
FLRDWR
GRIDPR
(b) Processing Routine
Figure IV-9 (Concluded)
47
-------
Table IV-11. Routines Specific to ROADWAY File Preparation
Name Description Routines Called
RDASGN Assigns roadway values (dummy)
RDVLRD Reads ROADWAY VALUES packet (dummy)
RDWPRC Processing routine See Figure IV-9(b)
RDWUSR USER method dummy subroutine
RDWYRD Reads ROADWAY DEFINITION packet (dummy)
ROADWY Main program See Figure IV-9(a)
48
-------
TMPRTR
Read CONTROL
and REGION
packets
CNTLRD
REGNRD
Read header
records from
input files,
if any
none
Write output
file header
records
FLHDWR
FLRGWR
FLSGWR
Partition
scratch
storage
SCRPRT
Set up core
extension,
if necessary
COREXT
Call processing
routine to do
everything else
TMPPRC
Figure IV-10.
(a) Main Program
Flow of Control for TEMPERATUR File
Preparation Program
49
-------
Read time-invariant
packets, if any
Write time-invariant
records, if any
Read time interval
and time-varying
input files,
if any
Read time-varying
packets until
ENDTIME
Perform
method
processing
Perform vertical
method processing,
if any
Write output
file and print
(b) Processing Routine
Figure IV-10 (Concluded)
TXTPAK
COMPAR
UNITRD
STATRD
none
TXTPAK
COMPAR
SBRGRD
MTHDRD
MATGET
STASGN
POISVR
none
FLTHWR
MATGET
FLIPWR
GRIDPR
CNSTRD
GDVLRD
STVLRD
TMPUSR
EWINTR
NSINTR
MATPUT
50
-------
Table IV-12. Routines Specific to TEMPERATUR File Preparation
Name Description Routines Called
TMPPRC Processing routine See Figure IV-lO(b)
TMPRTR Main program See Figure IV-lO(a)
TMPUSR USER method durnny subroutine
51
-------
Figure IV-11.
Read CONTROL
and REGION
packets
Read header
records from
input files,
if any
Write output
file header
records
Partition
scratch
storage
Set up core
extension,
if necessary
CNTLRD
REGNRD
none
FLHDWR
FLRGWR
FLSGWR
SCRPRT
COREXT
Call processing
routine to do
everything else
TRNPRC
(a) Main Program
Flow of Control For TERRAIN File
Preparation Program
52
-------
Read time-invariant
packets, if any
TXTPAK
COMPAR
IIWTTDn
STATRD
Write time-invariant
records, if any
Read time interval
and time- varying
input files,
if any
Read packets
until
ENDTERRAIN
Perform
method
processing
Perform vertical
method processing,
if any
Write output
file and print
(all records are
time-invariant)
none
TXTPAK CN
COMPAR GD
SBRGRD ST
MTHDRD
MATGET TR
STASGN EW
POISVR NS
MA
none
FLTMWR
MATGET
FLTRWR
GRIDPR
(b) Processing Routine
Figure IV-11 (Concluded)
53
-------
Table IV-13. Routines Specific to TERRAIN File Preparation
Name Description Routines Called
TERAIN Main program See Figure IV-lHa)
TRNPRC Processing routine See Figure IV-ll(b)
TRNUSR USER method dummy subroutine
54
-------
Figure IV-12.
Read CONTROL
and REGION
packets
Read header
records from
input files,
if any
Write output
file header
records
Partition
scratch
storage
Set up core
extension,
if necessary
Call processing
routine to do
everything else
CNTLRD
REGNRD
FILVER
FLHDWR
FLRGWR
FLSGWR
FLSPWR
SCRPRT
COREXT
TPCPRC
(a) Main Program
Flow of Control for TOPCONC File
Preparation Program
55
-------
TPCPRC
Read time-Invariant
packets, if any
Write time-invariant
records, if any
Read time interval
and time-varying
input files,
if any
Read time-varying
packets until
ENDTIME
Perform
method
processing
Perform vertical
method processing,
if any
Write output
file and print
(b) Processing Routine
Figure IV-12 (Concluded)
TXTPAK
COMPAR
UNITRD
STATRD
none
TIMERD
DFBKRD
RGTPRD
TXTPAK
COMPAR
SBRGRD
MTHDRD
MATGET
STASGN
POISSP
SPLITC
ABSTPC
RELTPC
none
FLTMWR
MATGET
FLTCWR
GRIDPR
CNSTRD
GDVLRD
STVLRD
VPRFRD
TPCUSR
EWINTR
NSINTR
MATPUT
56
-------
Table IV-14. Routines Specific to TOPCONC File Preparation
Name Description Routines Called
ABSTPC Performs ABSTOPCONC method
RELTPC Performs RELTOPCONC method
TPCONC Main program See Figure IV-12(a)
TPCPRC Processing routine See Figure IV-12(b)
TPCUSR USER method dummy subroutine
57
-------
WNDSET
Read CONTROL
and REGION
packets
CNTLRD
RE6NRD
Read header
records from
input files,
if any
FILVER
Write output
file header
records
Partition
scratch
storage
FLHDWR
FLRGWR
FLSGWR
SCRPRT
Set up core
extension,
if necessary
COREXT
Call processing
routine to do
everything else
WNDPRC
Ca) Main Program
Figure IV-13.
Flow of Control for WIND File
Preparation Program
58
-------
WNDPRC
Read time-invariant
packets, if any
Write time-invariant
records, if any
Read time interval
and time-varying
input files,
if any
Read time-varying
packets until
ENDTIME
Perform
method
processing
Perform vertical
method processing,
if any
Write output
file and print
(b) Processing Routine
Figure IV-13 (Concluded)
TXTPAK
COMPAR
UNITRD
STATRD
none
TIMERD
DFBKRD
RGTPRD
TMPSRD
TXTPAK
COMPAR
SBRGRD
MTHDRD
MATGET
STASGN
POISVR
SCRSET
DIVERG
MATGET
ABSPRF
RELPRF
FLTMWR
WNDSPD
MATGET
FLWDWR
GRIDPR
CNSTRD
GDVLRD
STVLRD
VPRFRD
WNDUSR
EWINTR
NSINTR
MATPUT
WNDVUS
EWINTR
NSINTR
MATPUT
59
-------
Table IV-15. Routines Specific to WIND File Preparation
Name Description Routines Called
DIVERG Performs DIVERGENCE method CELLHT
POISVR
MATPUT
MATGET
WNDPRC Processing routine See Figure IV-13(b)
WNDSET Main program See Figure IV-13(a)
WNDUSR USER method dummy subroutine
WNDVUS VERTUSER vertical method dummy subroutine
60
-------
D. IMPLEMENTING THE PROGRAMS
1. Source Deck Structure
•
The source version of each program and each of the libraries consists of
a number of "decks", each of which is a main program or subroutine and is
preceded by an @ELT, I card.
2. Scratch Vector Length
Each program contains a scratch vector JSCRTC; its length is fixed in a
DIMENSION statement in the main program and is also assigned to the variable
NSCRTC in a DATA statement. For efficient execution of the program, the
scratch vector should be as long as possible. The length currently compiled
into each program is a function of the EPA UNIVAC computer system. After
the program has been recompiled on a new computer and loaded along with
the library subroutines it requires, the load map should be examined to
determine how much space is actually available for the scratch vector.
The DIMENSION and DATA statements should then be changed and the main
program recompiled.
3. Overlaying
Currently, no overlay structure is built into the data preparation pro-
grams. Since there is no provision for COMMON, and all information is passed
through argument lists, it is impossible to construct independent overlays as
required by the LBL system (see Section VII-G). On the UNIVAC system, however,
an overlay consists merely of subroutines called from a higher level overlay
segment; with this less stringent requirement overlay schemes become possible.
The simplest overlay scheme is illustrated in Figure IV-14(a). In this
figure, the main segment contains the main program; the first overlay segment
contains routines called by the main program before the processing routine, and
the second overlay segment contains the processing routine itself.
61
-------
MAIN
CNTLRD
REGNRD
FILVER
FLHDWR
FLRGWR
FLSGWR
FLSPWR
SCRPRT
COREXT
xxxPRC
(processing routine)
all subroutines it calls
(a) For Simple Overlaying
MAIN
CNTLRD
xxxPRC (processing routine)
UNITRD
time-invar read
time-invar write
TIMERD
other time-varying
processing
(b) For Further Overlaying
Figure IV-14. Overlay Structure for Data
Preparation Programs
62
-------
A greater savings in core space can be accomplished by further subdividing
the second overlay segment (processing routine) as shown in Figure IV-14(b).
Here, the processing of time-invariant data is in one lower-level overlay seg-
ment and the processing of time-varying data is in another.
In all cases, routines used by more than one parallel overlay segment
should be assigned to higher-level segments. For example, TXTPAK and COMPAR,
used by all packet-reading subroutines, should be placed in the main overlay
segment.
4. I/O Error Handling
The subroutines in the library FILUTIL contain provisions for setting
flags and retaining control if errors are detectable. These routines can
be modified to include end-of-file and error checks if these are available.
5. Core Extension
The most machine-dependent feature of the data preparation programs is
core extension, which at present involves the execution-time assignment of
random-access files. This section discusses each of the core extension routines
in some detail, and suggests alternatives for other implementations.
a. COREXT
COREXT controls the assignment of three-and four-dimensional arrays.
It assigns two two-dimensional arrays for working storage and then checks
as to whether the entire array will fit in the remainder of the available
scratch vector space. If the array will fit, it is assigned to that space.
If it will not fit, COREXT allocates space for an index table and calls MATSET
to set up the random-access file. It then calls MATPUT to initialize the data
in the file and returns. COREXT itself contains no machine-dependent features
other than the passing of buffers and an index table through arguments to
lower level subroutines. (Note: MATSET has been bypassed for UNIVAC
implementation. It may be useful for installation of the code on a CDC
machine.)
63
-------
b. MATSET
MATSET creates a FET (File Environment Table), causes large core buffers
to be allocated for a new random-access file, and initializes an index table.
Use of MATSET to perform these functions avoids the need to allocate files
and buffers if they are not going to be used.
C. MATPUT AND MATGET
These routines transfer two-dimensional slices of arrays between the
three- or four-dimensional arrays and the two-dimensional working storage.
If the large array is in the scratch vector, a simple transfer of values is
performed. If the large array is on a random-access file, MATPUT and MATGET
call MATSTR or MATLOD, respectively, to write or read the data on the file.
There is no machine-dependent code in MATPUT or MATGET.
d. MATSTR and MATLOD
These routines are called to store data to or to load data from a random-
access file. Each is passed the appropriate index table entry and the length
of the record, and each calls a system routine that performs the store or
load. (These are dummy routines on the UNIVAC version of the code.)
Implementing the core extension feature on another machine necessitates
the establishment of a medium for storing the oversized arrays and the means
for accessing them. However, if a large amount of core storage is available,
it may not be necessary to implement this feature at all. Furthermore, if
extended core memory is available, it, rather than a secondary storage file,
can be used to reduce I/O accesses.
64
-------
V M.E.B, FILE SEGMENTATION PROGRAM
One program—SEGDEF—Is used to segment the 11 spatially varying M.E.B.
data files. It is designed to segment any number of files in a single run,
but because of certain computer requirements for small core file buffers, it
has been necessary to restrict program usage to one file at a time.
The SEGDEF program contains many of the features described in Chapter IV:
it uses routines from the same two libraries, UTILITY and FILUTIL; it uses
the same scratch vector allocation scheme for variable data storage; it uses
core extension if necessary for storage of three- and four-dimensional arrays.
This section describes the unique features of the file segmentation
program. For information on utility routines, scratch allocation, core
extension, and implementation suggestions, see the appropriate sections of
Chapter IV.
The flow of control in the program is shown in Figures V-l(a) and V-l(b).
The main program reads the CONTROL and REGION packets, allocates scratch
storage for the regional definition parameters, and then calls the processing
routine, SEGPRC, to complete the file segmentation. SEGPRC reads the SEGMENT
DEFINITION packet, assigns a segment number to each grid cell, and constructs
the internal segment definition parameters — adding one or two boundary cells
as necessary on each edge. SEGPRC then reads the SEGMENT FILE packet and,
for each file, calls COREXT to set up the array storage required and calls
a file-specific segmentation subroutine. Figures V-2 through V-12 show the
flow of control in each subroutine. Table V-l lists all nonutility routines
in the file segmentation program.
65
-------
SEGDEF
Read CONTROL
and REGION
packets
CNTLRD
REGNRD
Allocate scratch
storage for
region parameters
SCRSET
Call processing
routine to do
everything else
SEGPRC
(a) Main Program
Figure V-l. Flow of Control for File Segmentation
66
-------
SEGPRC
Read SEGMENTDEF
packet—construct
and print segment
definition grid
TXTPAK
COMPAR
GRIDPR
Read file
name card from
SEGMENT FILE
TXTPAK
COMPAR
Read and write
file header
records
HDRCPY
Set up array
storage and
call file segmenter
nteturnj
COREXT
AIRSEG
BNDSEG
DFBSEG
EMSSEG
PTSSEG
RDWSEG
RTPSEG
TMPSEG
TPCSEG
TRNSEG
WNDSEG
(b) Segmentation Processing Routine
Figure V-l (Concluded)
67
-------
AIRSEG
Read time
invariant records
if any
none
Segment and
write time
invariant records
none
Read time
varying records
Segment and
write time
varying records
FLTMRD
FLAQRD
MATPUT
FLTMWR
MATGET
SEGCPY
FLAQWR
Return]
Figure V-2. Flow of Control--AIRQUALITY File
Segmentation Subroutine
68
-------
BNDSEG
Read time
invariant records
if any
FLBIRD
Segment and
write time
invariant records
FLBIWR
Read time
varying records
Segment and
write time
varying records
FLTMRD
FLBVRD
MATPUT
FLTMWR
MATGET
SEGCPY
FLBVWR
Figure V-3. Flow of Control—BOUNDARY File
Segmentation Subroutine
69
-------
DFBSEG
Read time
invariant records
if any
none
Segment and
write time
invariant records
none
Read time
varying records
FLTMRD
FLDBRD
Segment and
write time
varying records
FLTMWR
SEGCPY
FLDBWR
(Return]
Figure V-4. Flow of Control--DIFFBREAK File
Segmentation Subroutine
70
-------
EMSSEG
Read time
invariant records
if any
none
Segment and
write time
invariant records
none
Read time
varying records
Segment and
write time
varying records
FLTMRD
FLEMRD
MATPUT
RTMWR
MATGET
SEGCPY
FLEMWR
Figure V-5. Flow of Control—EMISSIONS File
Segmentation Subroutine
71
-------
PTSSEG
Read time
invariant records
if any
SCRSET (allocates space
for point source parameters )
Segment and
write time
invariant records
Read time
varying records
FLTMRD
Segment and
write time
varying records
FLTMWR
Figure V-6. Flow of Control--PTSOURCE File
Segmentation Subroutine
72
-------
ROWSEG
Read time
invariant records
if any
none
Segment and
write time
invariant records
none
Read time
varying records
Segment and
write time
varying records
FLTMRD
FLRWRD
MATPUT
FLTMWR
MATGET
SEGCPY
FLRWWR
Figure V-7. Flow of Control—ROADWAY File Segmentation
Subroutine
73
-------
RTPSEG
Read time
invariant records
if any
Segment and
write time
invariant records
Read time
varying records
Segment and
write time
varying records
(Return)
none
none
FLTMRD
FLRTRD
FLTMWR
SEGCPY
FLRTWR
Figure V-8. Flow of Control—REGIONTOP File
Segmentation Routine
74
-------
TMPSEG
Read time
invariant records
if any
none
Segment and
write time
invariant records
none
Read time
varying records
FLTMRD
FLTPRD
Segment and
write time
varying records
FLTMWR
SEGCPY
FLTPWR
Figure V-9. Flow of Control--TEMPERATUR File
Segmentation Subroutine
75
-------
TPCSE6
Read time
invariant records
if any
none
Segment and
write time
invariant records
none
Read time
varying records
Segment and
write time
varying records
FLTMRD
FLTCRD
MATPUT
FLTMWR
MATGET
SEGCPY
FLTCWR
(Return]
Figure V-10.
Flow of Control--TOPCONC File
Segmentation Subroutine
76
-------
TRNSEG
Read time
invariant
if any
; records
FLTRRD
MATPUT
Segment and
write time
invariant records
MATGET
SEGCPY
FLTRWR
Read time
varying records
none
Segment and
write time
varying records
none
(Return)
FIGURE V-ll.
Flow of Control—TERRAIN File
Segmentation Subroutine
77
-------
WNDSEG
Read time
invariant records
if any
none
Segment and
write time
invariant records
none
Read time
varying records
Segment and
write time
varying records
FLTMRD
FLWDRD
MATPUT
FLTMWR
MATGET
SEGCPY
MATPUT
WNDSPD
FLWDWR
Figure V-12.
Flow of Control—WIND File
Segmentation Subroutine .
78
-------
Table V-1. File Segmentation Routines
Name
AIRSEG
BNDSEG
DFBSEG
EMSSEG
HDRCPY
Description
PTSSEG
RDWSEG
RTPSEG
SEGCPY
SEGDEF
SEGPRC
TMPSEG
TPCSEG
TRNSEG
WNDSEG
Segments AIRQUALITY file
Segments BOUNDARY file
Segments DIFFBREAK file
Segments EMISSIONS file
Reads and checks headers from unsegmented
file and writes to segmented file
Segments PTSOURCE file
Segments ROADWAY file
Segments REGIONTOP file
Copies one segment of data from full array
Main program for file segmentation
Processing routine for file segmentation
Segments TEMPERATUR file
Segments TOPCONC file
Segments TERRAIN file
Segments WIND file
Routines Called
See Figure V-2
See Figure V-3
See Figure V-4
See Figure V-5
FLHDRD
COMPAR
FLRGRD
FLSGRD
SCRSET
FLSPRD
FLHDWR
FLRGWR
FLSGWR
FLSPWR
See Figure V-6
See Figure V-7
See Figure V-8
See Figure V-1(a)
See Figure V-l(b)
See Figure V-9
See Figure V-10
See Figure V-ll
See Figure V-12
79
-------
VI CONTROL DATA PREPARATION PROGRAMS
Two control data preparation programs—CPREP and SPREP—create the
control files CHEMPARAM and SIMCONTROL, respectively. Both programs are
quite simple and, therefore, should present no problems in implementation
on any machine. The flow of control is as follows:
> Each card is read and its values are printed.
> Packet header and terminator fields are tested to ensure the
presence of all cards.
> Numeric fields are validated where possible.
> After all cards have been read and validated, the output
control file is written and its contents are printed.
80
-------
VII AIRSHED SIMULATION PROGRAM
The Airshed Simulation Program is the most complex program in the
SAI Airshed Model System. It reads 10-14 input data files, predicts
concentrations forward in time, and writes up to three output data
files as well as a printed trace of its own progress. It has been
designed to handle regions of arbitrary sizes through the method of
segmentation. By means of this technique, all information for a given
segment is resident in core at once, and the numerical integration
proceeds for the segment for one time slice as if it were an independent
region.
The following sections of this chapter discuss the techniques of
core allocation, the use of secondary storage, and the maintenance of
boundary conditions, all of which enable the processing of multiple
segments of variable lengths. Next, the method used to compute con-
centration averages is briefly discussed. Then, a step-by-step
description of the flow of control in the program is presented with
references to hierarchical diagrams. Finally, some of the considera-
tions in implementing the program are discussed, including FORTRAN
anomalies, file assignments, overlays, and the locations of the
specific numerical algorithms.
A. CORE ALLOCATION
As discussed in Chapter VIII of the User's Manual, segments (i.e.,
rectangular subregions) of the region to be modeled are selected such
that all the data arrays for one segment will fit in core at once. The
maximum dimensions for a segment in the horizontal plane depend on the
number of vertical levels and chemical species, the simulation options
used, and the amount of core space available. The amount of space avail-
able is fixed for an implementation on a given computer, but the other
81
-------
parameters determining segment size vary from application to application.
Since the model will run most efficiently if the segments are as large
as possible, the program is designed to make optimal usage of core
space. To avoid the need for recompiling the program for each appli-
cation (a task that would require detailed knowledge of the program and
the special considerations for each case), a run-time core allocation
technique is employed that allows the user to set the dimensions of a
segment to obtain maximum core usage.
The essence of a run-time core allocation scheme lies in the estab-
lishment of pointers to the beginning of each variably dimensioned array
in a "vector of available space" (VOAS). This vector is assigned when
the program is implemented on a particular computer; it is given a
dimension as large as possible within the core requirements of the
computer system. The pointer to an array is the index of the first word
of the array within the VOAS. The array can be passed to a subroutine
by providing in the CALL statement the address of the beginning of the
array. Figure VII-1 is a simple illustration of the technique, which is
discussed below.
Three arrays—ART, AR2, and AR3—are allocated to the vector of
available space called AV. ART is dimensioned (NOX.NOY), AR2 is
dimensioned (NOX.NOY.NOZ), and AR3 is dimensioned (NOZ.NOX), where NOX,
NOY, and NOZ are read in at run time. The pointers JAR1, JAR2, and
JAR3 are the indices of the beginning of AR1, AR2, and AR3, respectively,
within AV.
Exhibit VII-1 shows code fragments implementing this scheme.
JNEXT is a pointer to the next available word. It begins at 1. JAR!,
the pointer to ART, is first set to JNEXT, i.e., to the beginning of AV.
Then, the number of words required by AR1 is calculated and JNEXT is set
to point to the next available word. JAR2 is set to JNEXT. The next
available word after AR2 is calculated and AR3 starts at that point.
The subroutine SUB! uses the arrays ART, AR2, and AR3 with their
dimensions passed as arguments. In the call to SUB1, the beginning of
each array is passed by naming the properly indexed word in AV.
82
-------
NOX
NOY
NOZ
JAR!
JAR2
JARS
ART ... AV(JARl)
AR2 ... AV(JAR2)
AR3 ... AV(JAR3)
K NOX*NOY
NOX*NOY*NOZ
NOZ*NOX
Figure VII-1. Pointers to Variably Dimensioned Arrays
83
-------
DIMENSION AV( 1*900)
C
C *** READ VALUES FOR DIMENSIONS HOX, HOY, HOZ
C
READ (...) HOX, HOY, HOZ
C
C *** SET POINTERS TO AR1, AR2, AND AR3
C
JNEXT = 1
JAR1 = JNEXT
JNEXT = JNEXT -t- NOX*NOY
C
JAR2 = JNEXT
JNEXT = JNEXT + NOX*NOY*NOZ
C
JARS = JNEXT
JNEXT = JNEXT + NOX*NOZ
C
C *** CALL STJB1, PASSING DIMENSIONS AND
C FIRST WORD OF EACH ARRAY
C
CALL SUB1 (NOX, HOY, ROZ, AVtJARl), AV(JAR2), AV(JAH3)>
END
SUBROUTINE SUB1 (NOX, HOY, ROZ, AR1, AR2, AR3)
DI HENS I ON ARK HOX, HOY) , AR2( HOX. HOY, HOZ), AR3 (HOX, HOZ)
C
DO 200 J-l.NOY
DO 100 1=1,NOX
ARKI.J) = ...
100 CONTINUE
~00 CONTINUE
RETURN
END
Exhibit VII-1. Code Fragments Implementing the
Core Allocation Scheme
84
-------
The Airshed Simulation Program makes extensive use of array pointers;
in some cases, there are pointers to other pointers. Each pointer to an
array is named by the array name prefixed by a 0, as in the previous
example.
The program contains four tables of variable lengths that are
assigned to AV:
> Segment
> Local
> Boundary
> Segment Scratch.
Figure VI1-2 illustrates the assignment of the four tables in AV. Each
table, which contains a number of arrays of variable lengths, is described
in more detail below.
1. Segment Table
The Segment Table contains 65 data items for each segment, including
the origin and dimensions of the segment, scalars and control values for
the segment, and pointers to the segment's data arrays in the Boundary and
Segment Scratch Tables. Since the Segment Table is of variable length
depending on the number of segments in the run, there is a table of
pointers to the beginning of each type of item in the Segment Table.
These pointers are in the COMMON block /SEGPTR/; they are set in sub-
routine STABAL. During the course of a segment loop, it is desirable
to have the values for the 65 data items for a given segment easily
accessible without extensive indexing. Therefore, another COMMON block
/SEGTAB/ contains the 65 values from the Segment Table for a given seg-
ment. These values are picked out of the Segment Table by the sub-
routine SEGSET. Figure VII-3 illustrates the relationship between
/SEGPTR/, the Segment Table, and /SEGTAB/.
85
-------
AV
AV(JLOC)
AV(JBDRY)
AV(JSCR)
Segment Table
Local Table
Boundary Table
Segment Scratch Table
Figure VII-2. The Four Tables Assigned to the Vector of
Available Space (AV)
86
-------
/SEGPTR/
Jiteml
Jitem2
Segment Table
/SEGTAB/
Jitem65
item 1
for each-
segment
item 2
for each'
segment
item 65 /
for each''
segment
item 1
item 2
item 65
Figure VI1-3. Segment Table Structure
87
-------
2. Local Table
The Local Table contains vectors that are used by the integration
routines. With the exception of the six STEP1 and STEP2 vectors, which
depend on the length of the longest x- or y-dimension of any segment,
all vectors in this table depend on the number of vertical cells and
the number of species, dimensions that do not change from segment to
segment. The pointers to the vectors in the Local Table are in COMMON
block /LCOPTR/; their values are set in subroutine LOCAL. Some over-
laying of local vectors is accomplished in LOCAL by resetting JNEXT
for groups of variables. Figure VII-4 illustrates the relationship
between /LOCPTR/ and the Local Table, and shows how vector overlays
are done. Table VII-1 names the vectors in the Local Table and the
subroutines that use them, and shows how the vectors are assigned.
3. Boundary Table
The Boundary Table contains the boundary cell definitions and the
boundary concentrations in the interface cells for a given segment.
In multisegment runs, the Boundary Table for all segments is saved on a
secondary storage file. The table comprises eight variable length
arrays. Pointers to these arrays are found in the Segment Table; they
are set in subroutine BTABAL. The first four arrays define the location
of the boundary cells on the west, east, south, and north edges,
respectively. The remaining four contain the vertically varying con-
centrations for each species at each interface cell on the four edges.
If any boundary edge is completely external—i.e., there are no other
segments adjacent to it—the boundary concentration array for that
edge will have zero length. Figure VI1-5 shows the Boundary Table
structure; see Section C of this chapter for a description of how the
Boundary Table is used.
4. Segment Scratch Table
The Segment Scratch Table contains all the other spatially varying
data arrays for a given segment; it is saved on a secondary file in all
-------
/LOCPTR/
global
vectors
Local Table
steps 1
and 2
vectors
steps 3
and 4
vectors
Figure VII-4. Local Table Structure
89
-------
Table VII-1. Local Variable Overlays
STEP 1 and STEP 2
(General Vectors)
STEP 3
WINDZ
STEP 3
SIMDIG
STEP 4
TH(NOZ,3)
HT(NOZ,3)
VD(NOSPEC)
SRC(NOSPEC)
R(NORS,NORS)
HSl(MAXDIM)
HCSl(MAXDIM)
WSl(MAXDIM)
CSl(MAXDIM)
ASl(MAXDIM)
DELSl(MAXDIM)
AER(NOZ)
FACT(NOZPl)
CON(NOSPEC)
CONT(NOSPEC.NOZ)
XKZ(NOZPl)
FLUX(.NOSPEC)
WZ(NOZPl)
THEDG(NOZ,4)
WAV(4,NOZ)
DHDT(NOZ)
JDUM(NOZ)
A(NOZ)
B(NOZ)
C(NOZ)
Y(NOZ)
X(NOZ)
CTEST(NOZPl) DC(NORS)
Z(NOZ)
AOMEGA(NOZ)
HOLD(NOSPEC)
RATE(NORS)
YGF(NORS)
GA(NORS.NORS)
GF(NORS)
90
-------
/SEGTAB/
Boundary
Table
INDW Boundary indices on West edge
INDE Boundary indices East edge
INDS Boundary indices South edge
INDN Boundary indices North edge
BCW Boundary concentrations, West
BCE Boundary concentrations, East
BCS Boundary concentrations, South
BCN Boundary concentrations, North
Figure VII-5. Boundary Table Structure
91
-------
airshed simulation runs, even those for only one segment, because portions
of the table are overlaid by other arrays when concentration averages are
calculated. (See Section H.4 of this chapter for details.)
The number and lengths of the arrays in the Segment Scratch Table
depend on the simulation options used and the dimensions of each segment.
The pointers to the arrays in this table are found in the Segment Table;
they are set in subroutine SCRAL. Figure VI1-6 shows the structure of
the Segment Scratch Table and Table VII-2 lists the arrays it contains.
B. TEMPORARY SECONDARY STORAGE FILES
The Segment Table and the Local Table, which are of variable length
and thus not allocated until run time, remain in core throughout a simula-
tion run. The Boundary Table and the Scratch Table, however, contain
information for only one segment at a time and may differ in length from
segment to segment if the segment dimensions differ. To handle this
situation, the contents of each table are saved for all segments on
temporary secondary storage files. Since the tables may be of different
lengths for different segments, and because the length of even the longest
record is not determined until run time, the DEFINE FILE feature could not
be used for random access of the tables.
Instead, a procedure was devised for using a pair of sequential
files to hold all data for each table. One of the pair is designated as
a "write" file and the other as a "read" file. During the course of a
loop over segments, data for a segment are read from the read file at
the beginning of the loop and revised data for the segment are written
to the write file at the end of the loop. When the loop has been
terminated, the read file, with old data, has been read completely and
the write file has been written with all new data. At that point, both
files are rewound and their designations are switched, i.e., the write
file with the new data becomes the read file for the next pass, and the
read file with obsolete data becomes the write file for the next pass.
92
-------
/SEGTAB/
Scratch Table
JCTOP
JCUM
JDBD
JDBE
JCONC
CTOP (and CUM)
DBD
DBE
CONC
Figure VII-6. Segment Scratch Table Structure
93
-------
Table VII-2. Arrays in the Segment Scratch Table
Name
CTOP
DBD
DBE
DB1
DB2
TOPD
TOPE
TOPI
TOP2
WX
WY
* TEMP
* ROUGH
* VEG
QT
* QR
* CARM
* UPS
* KPTS
* QPTS
CONG
* CVAR
CUM
* CUMV
Dimensions
(sp,x,y)
(x.y)
(x,y)
(x,y)
(x.y)
(x.y)
(x,y)
(x.y)
(x,y)
(x.y.z)
(x.y.z)
(x.y)
(x.y)
(x,y)
(sp.x.y)
(sp.x.y)
(x.y)
(pts)
(pts)
(sp.pts)
(x,y,z,sp)
(sp,x,y)
(x.y.z)
(x,y)
Meaning
Concentrations at the top of the region
Diffusion break slope during time interval
Diffusion break at end of interval
Diffusion break at beginning of time step
Diffusion break at end of time step
Top of the region slope during time interval
Top of the region at end of interval
Top of the region at beginning of time step
Top of the region at end of time step
u-component of wind
v-component of wind
Ground-level temperature
Surface roughness
Vegetation factor
Total ground-level emissions
Emissions from roadways
Roadway microscale parameter
Point source location index
Vertical cell to receive point source emissions
Point source emissions
Pollutant concentrations
Subgrid-scale concentration variations
Temporary instantaneous concentrations
Temporary instantaneous concentration
variations
Optional arrays.
94
-------
There are four such pairs of files in the Airshed Simulation Program:
> BON1 and BON2 contain all the data from the Boundary Table for
each segment. If there is only one segment, these files need
not be used, because the Boundary Table will not be overlaid
by anything else.
> CON1 and CON2 contain the four-dimensional array of concentration
predictions and the three-dimensional array of concentration
variations, if it exists.
> CUM! and CUM2 contain the four-dimensional array of cumulative
concentrations and the three-dimensional array of cumulative
concentration variations, if it exists. These arrays and those
in CON! and CON2 share space in small core and thus must always
be saved on secondary storage files. If there is only one
segment, however, only one file is necessary'for each pair of
arrays.
> SEG1 and SEG2 contain the data in the Segment Scratch Table
appearing before the above mentioned concentration arrays. In
other words, they contain all the rest of the spatially varying
data for each segment.
The Segment Table contains, for each segment, two entries to control
the usage of each of the four pairs of files:
> The number of words in the record for the segment (i.e., the
length of the table to be saved).
> The location of the record on the file.
For the general case of all sequential files, the second entry is not used.
However, it may be possible to use an access method other than sequential
for a particular host computer or application. In that case, the loca-
tion of the record might be a record number or the relative location of
the first word. The file control entries in the Segment Table are set
in subroutine SCRAL.
95
-------
One pair of subroutines is used for accessing each of the four pairs
of files within a segment loop: a GET routine to obtain data from a "read"
file, and a PUT routine to put data to a "write" file. The subroutine
names (not surprisingly) are:
> PUTBON > GETBON
> PUTCON > GETCON
> PUTCUM > GETCUM
> PUTSEG > GETSEG.
The segment number, the read or write file unit number, the two control
entries mentioned above, and the core location to which or from which
data are to be transferred are all passed to each subroutine. The sub-
routines can easily be changed to reflect different access techniques
without affecting the rest of the program. The unit numbers for the four
pairs of files are set in the BLOCK DATA subprogram. These numbers can
also be easily changed, if necessary.
Figure VII-7 is a summary of the core allocation and file storage
procedures we have discussed so far. /SEGPTR/ contains pointers to the
Segment Table and /LOCPTR/ contains pointers to the Local Table. These
two tables always remain in core. Pointers to arrays in the Boundary
Table are found in the Segment Table, and the Boundary Table can be
stored on secondary storage files BON1 and BON2 (if there is more than
one segment). The pointers to arrays in the Segment Scratch Table are
also found in the Segment Table. The Segment Scratch Table, not including
the concentration array, is saved on secondary storage files SEG1 and SEG2;
the instantaneous concentrations and variations on CON! and CON2; and the
cumulative concentrations and variations on CUM1 and CUM2.
C. BOUNDARY HANDLING
Figure VI1-8 shows a 7 X 12 cell region divided into two 7 X 8 cell
segments to illustrate the following discussion of segment interface
handling.
96
-------
/SEGPTR/
Segment
Table
Local
Table
Boundary
Table
Segment
Scratch
Table
CONC
GETBON
PUTBON
GETSEG
PUTSEG
GETCQN
PUTCON
6ETCUM
PUTCUM
J BON1 J
) BON2J
SEG1
) SEG2 )
CON1
J CON2 J
) CUM1 ]
) CUM2 J
Figure VII-7. Core Allocation and File Storage
97
-------
Segment 1
12
n
10
9
8
7
6
5
4
3
2
1
1 234567
8
7
6
5
4
3
2
1
1234567
8--
7 +
6
5
4
3
2
1
Segment 2
1234567
Figure VII-8. Example of a Segmented Region
98
-------
First, it is necessary to define the boundary indices. The rectangular
horizontal plane for each segment has four edges: West, East, South and
North. Each row of the region has a West and an East boundary and each
column has a South and a North boundary. The boundary indices are
defined as the indices of the first and last cells within each row or
column to be simulated, not including the extra cells at each edge. Thus,
for each row of both segments in Figure VII-8, the West index is 2 and
the East index is 6. In the North-South directions, however, the boundary
indices in the two segments differ because of the two-row overlap at the
segment interface. In Segment 1, the South index is 3 and the North index
is 7; in Segment 2, the South and North indices are 2 and 6, respectively.
These boundary indices are used to control the numerical integration for
each segment as if it were an independent region.
To answer the question of where the concentrations in the outer cells
come from, it is necessary to define external and internal boundaries:
> External boundary cells in a segment are those that corres-
pond to the edges of the unsegmented region.
> Internal boundary cells in a segment are those that corres-
pond to simulated cells in another segment.
Concentrations in the external boundary cells are read from the BOUNDARY
file, and steady-state calculations may be performed on them at the
beginning of each time slice. Concentrations in the internal boundary
cells must be filled in from the concentration arrays for other segments.
Four numbers are read from the BOUNDARY file for each cell on each
edge for each segment:
> Boundary index (index of the first or last cell to be simulated
in the row or column). A boundary index of zero indicates that
the row or column is an external edge.
99
-------
> Adjacent segment number. If the adjacent segment number is zero,
the last simulated cell is at an external boundary; if the number
is not zero, the last simulated cell is at an internal boundary
and the concentration values in the interface cells must come
from simulated cells in the adjacent segment.
> I and J indices, within the adjacent segment, of the cell
corresponding to the nearer interface cell at the edge.
Table VII-3 shows the complete set of boundary definition indices that
would be read from the BOUNDARY file for the sample region. After they
are read, the I and J indices for each cell are combined into one number
by the following formula:
INDEX = (J - 1)*NOX + JI
where NOX = number of cells in the x-direction in the adjacent segment.
This combined index is the location of word (I,J) in an array dimensioned
(NOX.NOY).
The Boundary Table contains for one segment the four sets of boundary
definition indices (INDW, INDE, INDS, INDN) and four sets of boundary con-
centrations in the interface cells, if any (BCW, BCE, BCS, BCN). During
the subroutine SETUP, which is called at the beginning of each time slice,
the concentrations at the external boundaries in the CONC array are up-
dated from the BOUNDARY file if necessary. For internal boundaries,
concentrations in the Boundary Table for interface cells are filled in
from the appropriate adjacent segment cells by the subroutines SEGINT and
BCFILL. During the subroutine ADVANC, which controls the integration
steps, the interface cells along the internal edges in CONC are filled by
the concentrations from the corresponding BC arrays (subroutine CFILL)
and steady-state computations are performed in external boundary cells if
required (subroutine SSBC).
100
-------
Table VII-3. Boundary Definition Indices
Segment 1
Segment 2
OS
,
o
CT
LU
West
1
2
3
4
5
6
7
8
East
1
2
3
4
5
6
7
8
X
•o
^>
to
T3
3
O
CO
0
2
2
2
2
2
2
2
0
6
6
6
6
6
6
6
c
b
4-> O>
•r- 0)
3: co
, ~ *. **,
I
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
^•B^.
J
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
X
O)
•o
c
h— t
i.
03
•o
3
o
CO
2
2
2
2
2
2
2
0
6
6
6
6
6
6
6
0
c
0>
CT
(U
co
c
01 L.
U 0)
ID -Q
T> E
-O 3
**• Z.
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
c.
Ol
u
re
•o
<£ ts>
c c
•t- O>
r* c:
^^B C;
-(-> 0)
•i— QJ
S CO
/ • • ^^
I
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
•III" V
J
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
101
-------
Table VII-3 (Concluded)
Segment 1
^
Q
o:
.^
o
(U
•o
LLJ
South
1
2
3
4
5
6
7
North
1
2
3
4
5
6
7
X
OJ
•o
c
*-4
1.
(/)
_xs__^^
•^ ^i^^^^^
J
0
6
6
6
6
6
0
0
0
0
0
0
0
0
Segment 2
O)
-o
c.
>*H
>^
^
TO
-a
c
o
CO
0
2
2
2
2
2
0
0
6
6
6
6
6
0
4-J
E
g
o-
.«£ Z •
0
0
0
0
0
0
0
0
1
1
1
1
1
0
.^Mm^^
^^^^^^^^>
i
0
0
0
0
0
0
0
0
2
3
4
5
6
0
E
OJ
(J
(O
•o
«C to
+->
E E
•i- O)
-E E
4J Cn
•^ O)
3 00
^^S^^M
J
0
0
0
0
0
0
0
0
3
3
3
3
3
0
102
-------
D. COMPUTING AVERAGE CONCENTRATIONS
Predicted pollutant concentrations are averaged over a given time
interval that is input by the user by means of the SIMCONTROL file.
To calculate the time averages, the cumulative concentration array must
be created; it contains the time-weighted sum of the concentrations
calculated during each time step from the beginning of the averaging
interval.
The Interim Instantaneous Concentrations (INTERINST) file keeps a
complete history of instantaneous concentrations calculated within each
time slice. See Chapter IX on File Structures for a more detailed
description of this file. For a given time slice, each segment is inte-
grated in turn by one or more time steps of equal size. The instantaneous
concentrations array for the segment is written to the INTERINST file at
the beginning of the time slice and at the end of each time step in the
slice. Thus, for n time steps there will be n + 1 sets of concentration
records for the segment. When the time slice has been completed for all
segments (subroutine ADVANC), the INTERINST file is rewound.
In the accumulation phase, controlled by subroutine ACCUM, the follow-
ing procedures are completed for each segment:
> The current cumulative concentration arrays are brought in
from CUMR (the secondary storage file to be read). These
arrays overlay the instantaneous concentration arrays in the
Scratch Table.
> Instantaneous concentrations are read in from INTERINST for
each time step, weighted by the time step size (as described
below) and added into the cumulative concentrations.
> The new cumulative concentrations are written to CUMW (the
secondary storage file to write).
103
-------
The concentrations at the beginning and end of the time slice are multiplied
by the time step used for the slice. If there was more than one step in a
slice, there will be one or more sets of concentration records between the
first and the last records; these are each weighted by the value of two times
the time step.
At the end of the averaging period (in subroutine FINISH), the cum-
ulative concentration arrays are again read in and the time-weiohted values
are divided by (2 x total elapsed averaging interval). The resulting
average concentrations are then written to the AVERAGE file, and the
cumulative concentration arrays are zeroed for the next averaging period.
If concentration variations are being calculated, the array containing
the variations follows the concentration array in the Segment Scratch Table.
To calculate the average concentration variation output to the VARIATION file,
the same procedure is performed. The instantaneous variations are also writ-
ten to the INTERINST file, and the cumulative variations are maintained in
the same way.
E. FLOW OF CONTROL
This section describes the operation of the Airshed Simulation Program
and includes figures and tables that supplement the text to aid in the compre-
hension of the program as a whole. Figures VII-9(a) through VII-9(c) are
top-down structure diagrams that illustrate the hierarchy of the highest level
subroutines. Table VII-4 is a hierarchical list of subroutine names, indicating
the complete calling structure of the program. Table VII-5 lists the sub-
routines alphabetically and includes a one-line description of the function
of each.
1. AAASIM
The main program AAASIM controls the calls to the two major controlling
routines [Figure VII-9(a)] . The first, AAINIT, performs all one-time
initialization functions and the second, AAMRCH, performs the simulation,
104
-------
1.
AAINIT
Initialize
0.
AAASIM
Main Program
2.
AAMRCH
March through
time
(a) AAASIM
Figure VI1-9. Hierarchy of Top-Level Subroutines
105
-------
IXI
1 j
i § !
l T; i
1 -o 1
1 ±! 1
0
_J
fO ID
OJ
S- E
O •!-
O
«t
LU
o:
evi
(A OJ
•o 'i (o
fO OJ i-
O) -C IB
ce o a.
T3
O)
3
O
O
en
i
ro E c
O) -r- O
EC O
106
-------
Cv
Qi 3 t/>
•M Q. CO
.,_ 4J r—
S_ 3 v-
304-
51 !
sg -& i
S| = |
«cl i~ 1
1 •*= 1
1 0 0) |
1 rtJ £ 1
u
o
"^
co
CM
01
3
3
U
U
U
c
o
'£
CO
C
CU
u
c
o
u
en
O •!-
4- O)
01 J-
E 0)
3 >
(A
Q
•a:
CM
e
CM
1/1
C
O
•9—
*J
u c
C O)
(O U
> c
•o o
f u
Ol
_E
4-^
c
• •> in
«3 C
•4-> 4J O
«J C t-
•o a; •*->
t- to
2 J- >> S-
O) 3 S- -U
C U 3 C
0) ro O O
a; co ja o
107
-------
Table VII-4. Subroutine Hierarchy in the
Airshed Simulation Program
0. AAASIM
1. AAINIT
1.1 SCREAD
CLFILE
RETURNS
1.2 CHREAD
SPECID
COEFID
CLFILE
1.3 ALLOC
STABAL
BTABAL
BDREAD
PUTBON
LOCAL
SCRAL
1.4 TZERO
HDRCK
SPETAB
FILSKP
MSREAD
MSREAD
SE6SET
RESDB
RESTMP
RESEM
RESPTS
RESTOP
RESRWY
RESTER
RESCT
RESWND
PUTSEG
PUTCUM
108
-------
Table VI1-4 (Continued)
CLFILE
HDRRIT
GETSEG
GETBON
ARTIME
RESAQ
RESBDY
SSIC
PHOTOL
VERTCL
PHFACT
TDRATE
STEADY
CHEM
CONPRT
PUTCON
SECOND
2.0 AAMRCH
2.1 SETUP
FILCK
FILCKE
SEGSET
GETBON
GETCON
RESBDY
PUTCON
MSREAD
GETSEG
RESDB
RESTMP
RESEM
RESPTS
RESTOP
RESRWY
RESCT
109
-------
Table VI1-4 (Continued)
RESWND
PUTSEG
GETCON
SEGINT
GETBON
BCFILL
PUTBON
SLICE
2.2 ADVANC
SEGSET
GETCON
GETSEG
GETBON
CFILL
SSBC
PHOTOL
VERTCL
PHFACT
TDRATE
STEADY
CHEM
WINTER
ARTIME
STEP1
VERTCL
SHASTA
STEP2
VERTCL
SHASTA
STEP34
PHOTOL
VERTCL
DIFCOF
SNKSRC
110
-------
Table VI1-4 (Concluded)
MICRO
WINDZ
VERTCL
PHFACT
TDRATE
STEADY
CHEM
STEP3
SIMDIG
STEP4
CHEM
CHEM2
DEBUG
CRASH
PUTCON
PUTSEG
2.3 ACCUM
SEGSET
GETCUM
ADDEM
PUTCUM
2.4 FINISH
SECOND
DISPLA
WARN
SEGSET
GETCUM
PUTCUM
GETCON
111
-------
Table VII-5.
Subroutine
Descriptions of Airshed
Simulation Program Subroutines
Description of Function
AAASIM Main program—controls calls to two major phases
AABDAT BLOCK DATA (not executable)—sets values for labeled
COMMON variables
AAINIT Controls calls in initialization phase
AAMRCH Controls calls in march through time
ACCUM Controls accumulation of concentration sum at the end of
each time slice
ADDEM For a segment, adds concentrations at each time step into
cumulative concentrations
ADVANC Controls advance of concentrations one time slice
ALLOC Controls allocation of all variable core space
ARTIME Calculates values in an array (DB or TOP) for a given time;
stores previous values
BCFILL Stores CONC in BC for a given segment and edge
BDREAD Reads boundary definition records for a segment
BTABAL Allocates Boundary Table
CFILL Stores BC in CONC at all interfaces for a segment
CHEM Computes steady-state values for radicals, reaction rates,
and partial derivatives for each species
CHEM2 Computes reaction rates and partial derivatives for night
chemistry
CHREAD Reads CHEMPARAM file
CLFILE Closes a file
COEFID Establishes coefficient values for chemistry
CONPRT Prints concentrations for a segment
CRASH Prints all relevant information at the time of a crash
DEBUG Controls call to crash print
DIFCOF Calculates vertical diffusivity
FILCK Checks current time against file time interval; if at the
end of the interval, reads next interval and sets flag;
adjusts time slice so that it won't cross time interval
112
-------
Table VII-5 (Continued)
Subroutine Description of Function
FILCKE Like FILCK, but tests current time against beginning of
the file time interval
FILSKP Reads time-varying data until first good time interval
FINISH Updates time and writes files at end of a time slice
GETBON Gets Boundary Table for a segment from secondary storage
file
GETCON Gets concentrations for a segment from secondary storage
file
GETCUM Gets cumulative concentrations for a segment-from secondary
storage file
GETSEG Gets scratch arrays for a segment from secondary storage
file
HDRCK Reads and checks header records
HDRRIT Writes header records to INSTANT and AVERAGE
LOCAL Allocates space for local variables
MICRO Performs surface layer microscale computations
MSREAD Reads data record from METSCALARS file
PHFACT Calculates vertically varying photolysis factors
PHOTOL Multiplies photolysis rate constants by factors
PUTBON Puts Boundary Table for a segment to secondary storage
file
PUTCON Puts concentrations for a segment to secondary storage
file
PUTCUM Puts cumulative concentrations for a segment to secondary
storage file
PUTSEG Puts scratch arrays for a segment to secondary storage
file
RESAQ Reads data for one segment from AIRQUALITY file
RESBDY Reads data for one segment from BOUNDARY file
RESCT Reads data for one segment from TOPCONC file
113
-------
Table VII-5 (Concluded)
Subroutine Description of Function
RESDB Reads data for one segment from DIFFBREAK file
RESEM Reads data for one segment from EMISSIONS file
RESPTS Reads data for one segment from PTSOURCE file
RESRWY Reads data for one segment from ROADWAY file
RESTER Reads data for one segment from TERRAIN file
RESTMP Reads data for one segment from TEMPERATUR file
RESTOP Reads data for one segment from REGIONTOP file
RESWND Reads data for one segment from WIND file
SCRAL Allocates scratch storage space
SCREAD Reads SIMCONTROL file
SEGINT Updates segment interfaces
SEGSET Sets segment table values for a given segment
SETUP Sets up data for a new time slice
SHASTA Performs SHASTA integration
SIMDIG Solves tridiagonal system of linear equations
SLICE Calculates time-slice size
SNKSRC Calculates deposition velocity and reemission
SPECID Sets species indices for chemistry
SPETAB Sets species index table for a file
SSBC Computes steady-state concentrations at external boundaries
SSIC Computes steady-state values for initial concentrations
STABAL Allocates segment table space
STEADY Calculates steady-state values for NO, N02, and Og
STEP1 Controls integration in x-direction
STEP2 Controls integration in y-direction
STEPS Performs advection integration in z-direction
STEP4 Performs chemistry integration for a single cell
STEP34 Controls vertical and chemistry integration
TDRATE Calculates temperature-dependent rate constants
TZERO Sets up all data for time = tQ (beginning of integration)
VERTCL Calculates vertical distribution of grid cells
WINDZ Calculates vertical winds
WINTER Writes INTERINST file
114
-------
marching through time by time slices. Each routine (both are described below)
calls four major subroutines. The subroutines used in AAINIT that are not
also used in AAMRCH can be overlaid when the initialization phase is completed
(see Section 6 of this chapter for more details on overlays).
2. AAINIT
AAINIT calls four subroutines: SCREAD, CHREAD, ALLOC, AND TZERO
[see Figure VII-9(b)].
SCREAD reads the SIMCO'lTROL file, prints its values, and closes it.
CHREAD reads the CHEMPARAM file, prints its values and closes it.
Subroutines SPECID and COEFID are called, if required, to verify that
the species and coefficient names, respectively, correspond to the names
required by the built-in chemical mechanism. CHREAD also sets internal
flags controlling optional program functions depending on the properties
of the species and reactions and on the simulation options specified on
SIMCONTROL.
ALLOC reads the BOUNDARY file for the overall dimensions of the region
and the number and size of its segments. It then calls four subroutines
to set up all core allocation pointers:
> STABAL sets up the pointers to the Segment Table in /SEGPTR/.
> LOCAL sets up the pointers to the Local Table in /LOCPTR/.
> BTABAL sets up the pointers to the Boundary Table in the Segment
Table.
> SCRAL sets up the pointers to the Scratch Table in the Segment
Table. SCRAL reads the PTSOURCE file to determine the maximum
number of point sources in each segment.
TZERO reads the remainder of the M.E.B. data files and establishes all
array values for the beginning of the simulation. For each file, it calls
HDRCK to read and check the header records, SPETAB to establish species
index tables based on species names (which allows species to appear in a
115
-------
different order on each file), and FILSKP to skip records, if necessary, up
to the first time interval applicable to the run. TZERO then reads all M.E.B.
data for each segment (the subroutines beginning with RES ) and zeros
arrays where appropriate. It calls SSIC, if required, to perform steady-
state calculations on the initial concentrations. It writes all data onto the
appropriate secondary storage files, calls HDRRIT to write headers to the
output files, and calls CONPRT, if required, to print the initial concentra-
tions.
3. AAMRCH
AAMRCH controls the march through time. It calls SETUP, ADVANC, ACCUM,
and FINISH for each time slice [see Figure VII-9(c}].
SETUP checks all file time intervals (FILCK and FILCKE) and reads new
data, if necessary. It also saves all boundary interface concentrations
(SEGINT) and establishes the size of the time slice that will be used (SLICE).
ADVANC calls the subroutines that perform the numerical integration.
These are listed in Section I of this chapter. ADVANCE processes each
segment completely for one time slice. For each segment, it first fills in
the concentrations at the segment interfaces (CFILL), performs steady-state
calculations, if necessary, on the external boundary concentrations (SSBC),
and writes the entire concentration array to the INTERINST file (WINTER).
Then, for each time step, it calculates the diffusion break and top of the
region (ARTIME), performs the horizontal transport steps (STEP! and STEP2),
performs the vertical transport and chemistry steps (STER34), and writes the
concentrations at the end of the time step to INTERINST (WINTER).
ACCUM adds the instantaneous concentrations calculated for the time
slice into the cumulative concentration arrays. For each segment, it calls
ADDEM, which reads INTERINST for each time step and does the actual time-
weighting and adding. (Details were previously given in Section D of this
chapter.)
116
-------
FINISH is called at the end of each time slice. It calculates and prints
the elapsed CPU time for the slice and then checks to see whether an instantaneous
or average output time has been reached. If so, concentrations are read from
the appropriate secondary storage file and written to the appropriate output
file. For averages, the time-weighting is divided out of the cumulative con-
centrations. FINISH also calls CONPRT if the concentrations are to be printed.
If the simulation end time has been reached or if it has been detected that
computer time is running out, FINISH will write end-of-file on the output files
and terminate the run.
F. COMMON BLOCKS
Eleven labeled COMMON blocks are used to share data among the many Airshed
Simulation Program subroutines; a short description of each one is given in
this section. Table VII-6 is a cross-reference list indicating the appearance
of each of the following COMMON blocks in each subroutine.
> /AVAIL/--contains the vector of available space, its length, and
pointers to the beginning of its tables.
> /CHDATA/—contains species indices, names, and coefficients required
by the Carbon-Bond Mechanism.
> /CHPARM/-- contains chemistry parameters input by means of the
CHEMPARAM file.
> /CNTROL/—contains many control parameters of global interest:
- Dimensions and region descriptors.
- Logical control flags.
- Tine controls.
- Integration controls and calculated scalar values.
- Trace and internal file unit numbers.
- Trace options and current location indicators.
- Descriptive information.
> /FILCON/~contains all file control variables.
> /LCM/--contains the large core memory vector used to store boundary
tables for all segments (LBL only).
117
-------
> /LOCPTR/--contains pointers to vectors in the Local Table.
> /MSCAL/—contains meteorological scalars read from METSCALARS or
calculated in MSREAD.
> /SEGPTR/--contains pointers to the Segment Table entries.
> /SEGTAB/--contains the values of the Segment Table entries for a
particular segment.
> /VPARAM/--contains the vertical parameters used by subroutine
VERTCL.
118
-------
Table VII-6. COMMON Block/Subroutine Cross-Reference List
COUCH Blocfc
Subroutine AVAIL CHOATA CHPARH CNTROL FILCON LCH LOCPTR KSCAL SEGPTR SEGTAB YPARAM
AAASIH XXXXXXXXXXX
AAflOAT X X X X XXX
MINIT X
AAKRCH X
ACClft X X X
ADOEH
AOVANC XX X X X X
ALLOC XXX X X
ARTIME
term
BOREAO
BTABAL
CF1LL
CHEH X X
CHEW X X
CHREAO X X X X X
CLFILE
COEFIO X
CONPOT XX X
CRASH XX XX
DEBUG X X X
DIFCOF X XX
FRCK
FILCK£
FILSKP X
FINISH XXX X
GETBCN X
GETCON
crrcun
GETSEG
HORCK X X
HORRIT XXX
LOCAL X
MICRO X X X
KSREAO X X
PHFACT
PHOTOL X
PUTBON . X
PUTCON
PUTOJM
WTSEG
RESAQ X X
KESBOY X
RESCT X X
KES08 X
KSEM X
RESPTS X X X
KSRMY X
HESTER
KSTHP
119
-------
Table VII-6 (Concluded)
COHMON Block
Subroutine AVAIL CHOATA CHPARM CMTROL FILCOM LCH LOCPTR HSCAL SEGPTR SE6TA8 YPARAH
•ESTOP X
KSUND X X X
SCRAL X
SCREAD X X
SEGINT X X
SEGSET x X X
SETUP XX XXX
SHASTA
SIMDIG
SLICE X
SMKSRC X X
SPECID X
SPETAB
SSBC XXX X
SSIC XXX X
STABAl X
STEADY XX X
STEP1 XX X X
STEP2 XX X X
STEP3 x X X
STEP4 XX X
STEP34 XXX XX X
TDRATE X X
TZERO XXX X X X X
VERTCl X
WINOZ X X
WINTER
120
-------
G. OVERLAYS
Core space requirements can be reduced by overlaying program or data
storage areas. Both of these techniques are used in the Airshed Simula-
tion Program and are described in the following sections.
1. Program Overlays
A number of subroutines are used only in the initialization phase of the
Airshed Simulation Program and are not required again. These subroutines may
be overlaid by subroutines used only in the "march through time" loop.
Figure VII-10(a) is a list of subroutines in the simulation program
arranged by overlay possibility. At the top is the main program, that calls
each major function. Next are the subroutines that are used both in the
initialization and simulation phases. These routines should not be over-
laid, or they must appear in both overlays. Finally, in two columns, are
the initialization and simulation routines that are completely independent
of each other and that may be overlaid. None of the lower-level subroutines
are named here; see Table VII-4 for the complete list of subroutines that
belong in each overlay. This type of program overlay should produce little
additional system overhead, since the second program overlay is brought in
only once, at the end of the initialization phase, and no further program
shuffling is necessary.
If necessary, the "march through time" overlay could be further divided,
with the ADVANC portion in one segment and the SETUP, ACCUM, and FINISH
routines in another, as shown in Figure VII-10(b). This would necessitate
two overlay swaps in each time slice.
Another overlay technique, which could be used with either of the two des-
cribed above, consists of separating the subroutines DEBUG and CRASH from
the other routines called by ADVANC. [See Figures VII-lO(c) and VII-10(d).]
These routines are called only when negative concentrations are predicted or
when the chemistry calculations do not converge. In normal situations, these
121
-------
AAMAIN
All Common Blocks
BLOCK DATA
SEGSET
GETxxx
PUTxxx
RESxxx
MSREAD
ART I ME
VERTCL
PHOTOL
PHFACT
TDRATE
STEADY
CHEM
CONPRT
AAINIT
SCREAD
CHREAD
ALLOC
TZERO
AAMRCH
SETUP -
ADVANC
ACCUM
FINISH
(a) Simple Overlay
AAMAIN
COMMON Blocks
BLOCK DATA
Common Subroutines
AAINIT
AAMRCH
SETUP
ACCUM
FINISH
ADVANC
(b) Overlay Scheme with Two Swaps in Each Time Slice
Figure VII-10. Airshed Simulation Program Overlay Structure
122
-------
AAMAIN
COMMON Blocks
BLOCK DATA
Common Subroutines
AAINIT
AAMRCH
SETUP
ACCUM
FINISH
ADVANC
other
subs.
DEBUG
CRASH
(c) A Possible Overlay Scheme with No Swaps
AAMAIN
COMMON Blocks
BLOCK DATA
Common Subroutines
AAINIT
•
•
AAMRCH
•
SETUP
ACCUM
FINISH
ADVANC
other
subs.
DEBUG
CRASH
(d) Another Possible Overlay Scheme with Two Swaps
Figure VII-10 (Concluded)
123
-------
routines would never be called and, therefore, no additional overlay swapping
would be required.
Some computer systems have more stringent requirements than others for
•
overlays. On the LBL system, for example, each overlay must be a free-standing
main program, called explicitly by the higher level overlays, sharing only
COMMON blocks and subroutines within overlays above it, and lacking the ability
to receive parameters passed as arguments. On the UNIVAC system, however,
overlay segments contain only subroutines, and these will be swapped in as nec-
essary when the subroutines are called. The Airshed Simulation Program has
been designed so that the simplest overlay [described in Figure VII-10(a)]
is possible on the LBL system. On other systems, to accommodate the largest
possible vector of available space, it might be necessary to design overlay
structures that are considerably more complex.
2. Data Overlays
The overlaying of data arrays is already handled within the program in
the following three ways:
> All data arrays for a single segment are in core at once, but they
share core space with corresponding arrays for other segments.
> Arrays containing cumulative concentrations (used to compute time-
averages) share space with the instantaneous concentration arrays.
Interim instantaneous concentrations read from INTERINST are stored
•in space ordinarily used by other data arrays.
> The local vectors allocated in subroutine LOCAL are implicitly over-
laid by the manner in which their pointers are set. See Section A.2
of this chapter for a further explanation of this procedure.
124
-------
H. IMPLEMENTING THE PROGRAM
1. Source Deck Structure
The source version of the Airshed Simulation Program consists of
labelled COMMON block definitions as well as the main program and sub-
routines.
For implementation on the UNIVAC, the main program and subroutines
are preceded by @ELT,I cards. The elements formed from labelled COMMON
block definitions should be processed by @PDP to become PROCs. The
insertion of predefined COMMON blocks within subprogram decks is
accomplished with INCLUDE statements.
2. LBL Library Functions
There are five LBL functions used in the program (Table VI1-7 shows where
they appear):
> The EOF function is used to test for end-of-file on file reads. It
can be replaced by END=nnn parameters on READ statements.
> The subroutine SECOND returns the elapsed CPU time, in seconds,
from the beginning of the job. If there is no comparable function
on the host computer, this function can be replaced by calls to
other clocking routines or can be removed.
> The subroutine DISPLA prints a message on the CDC Dayfile, a job
control stream printout. This function can be replaced by a com-
parable routine or can be eliminated.
> The function WARN returns a nonzero value when the computing time
limit is approaching. This function, too, can be replaced by a
comparable routine or can be eliminated.
125
-------
Table VII-7. LBL Library Functions
Function Subroutines That Call the Functions
EOF ADDEM, FILCK, FILCKE, FILSKP, MSREAD, RESxxx
SECOND TZERO, FINISH
WARN FINISH
DISPLA FINISH
RETURNS CLFILE
126
-------
> The subroutine RETURNS returns to the system an input file no
longer needed. On the UNIVAC, this function can be replaced by
a call to CLOSE (see Fortran V Library PRM, pp. 3 to 8).
Note that these functions have been bypassed for implementation
on the EPA's UNIVAC computer. They may be employed for CDC computer
operation of the Airshed Model.
3. Changing File Assignments
Table VII-8 shows the current assignment of unit numbers to the Airshed
Simulation Program's data files. These numbers are assigned in the BLOCK
DATA subprogram. If they conflict with standard unit assignments for other
purposes, they can be changed simply by changing the BLOCK DATA subprogram.
4. Overlaying
Section G.I of this chapter describes a number of overlay structures
that can be used. It is suggested that the program first be compiled and
loaded without overlays to determine the total core requirements of the
program and data. Then, the effects of the various program overlay structures
on core storage availability can be evaluated.
5. Vector of Available Space
The vector of available space is currently in Blank COMMON in order to
reduce the core requirements for loading on the LBL system. If necessary, it
can easily be changed by changing the COMMON declaration in the labeled COMMON
AVAIL. The dimension of the vector of available space (AV) is currently
26,000 words. After the program has been compiled and loaded, the load map
should be examined to determine how much space is actually available for
variable length arrays. The dimension of AV can then be changed and the three
modules that use it, AAASIM, AAINIT, and AAMRCH, can be recompiled.
6. Use of Large Core Memory
On the LBL system, large core memory rather than a pair of secondary
storage files is used to store the Boundary Tables for each segment. Boundary
127
-------
Table storage can be changed to secondary storage file accesses by using the
following steps:
> Remove the COMMON/LCM/ block from AAASIM, GETBON, and PUTBON.
> Change GETBON and PUTBON to read and write files, as GETCON and
PUTCON do.
> Restore statements that REWIND NUBONR and NUBONW, which appear after
segment loops in TZERO, SETUP, SEGINT, and ADVANC, and have been
disabled by C in Column 1.
On the UNIVAC, it may be desirable to use extended memory to store the
Boundary Table. In that case, the following steps should be taken:
> Calculate the desired length of the table storage vector by the
formula:
L =
+ yseg) + z*sp*(no. interface cells^l
> Edit the COMMON/LCM/ block to change the length of the table
storage vector and to remove the LEVEL statement.
> Change @MAP directives so that the COMMON/LCM/ block is loaded
last.
7. Flexibility Versus Efficiency
The Airshed Simulation Program has been designed to maximize flex-
ibility in choosing regional dimensions while optimizing the usage of
available core storage. This capability has been accomplished by assign-
ing all spatially varying arrays at run time, when spatial dimensions are
specified. This feature provides that the user can specify the largest
128
-------
region possible given the selection of program options, species and vertical
levels, and that the program need not be recompiled for each new use. This
provision of flexibility is not, however, without some cost in efficiency.
Whenever variably dimensioned arrays are passed as subroutine arguments
rather than being declared with fixed dimensions within COMMON blocks,
extra instructions are required for each reference. These instructions add
to program size and to operating time. It may be desirable, at some time,
to tailor the program for a specific application, where the program options,
numbers of species and vertical levels, and maximum segment size are all well
known. Then, to achieve maximum efficiency in program operation, the arrays
now assigned to available space could be fixed in dimensions, put in COMMON
blocks, and removed from argument lists. This task is not conceptually dif-
ficult but, practically, it could be quite time consuming, since the use of
variably dimensioned arrays in arguments is widespread within the program.
The increased efficiency that would result from these changes is not known.
Therefore, it is not recommended unless the number and length of computer
runs is great enough to justify the changes.
I. Locations of Specific Algorithms
In addition to all the.data handling processes that have occupied the
bulk of descriptive material in this chapter, the primary function of the .
Airshed Simulation Program is to perform the integration of the governing
equations described in Chapter II of the User's Manual. The routines that
perform the numerical integration are called by subroutine ADVANC; routines
that calculate steady-state values of initial and boundary conditions are
called from subroutine SSIC (called by TZERO) and SSBC (called by ADVANC).
Table VII-9 is a list of numerical functions used within the program and the
subroutines in which they are performed. See also the hierarchical list
of subroutines (Table VII-4) for indications of how the modules fit together.
129
-------
Table VII-8. File Unit Assignments
File
AIRQUALITY
BOUNDARY
DIFFBREAK
EMISSIONS
METSCALARS
PTSOURCE
REGIONTOP
ROADWAY
TEMPERATUR
TERRAIN
TOPCONC
WIND
CHEMPARAM
SIMCONTROL
EXECUTION TRACE
AVERAGE
INSTANT
VARIATION
INTERINST
BON1
BON2
CON1
CON 2
CUM1
CUM2
SEG1
SEG2
Number Unit
3
11
12
13
7
14
15
16
17
4
18
19
2
1
6
21
22
23
30
37
38
33
34
35
36
31
32
130
-------
Table VII-9. Location of Specific Numerical Functions
Numerical Function
Vertical distribution of grid cells
Horizontal advection and diffusion
Emissions unit conversion and scaling
Surface deposition
Surface-layer microscale
Vertical diffusivity
Vertical winds
Vertical advection integration
Photolysis rates
Temperature-dependent rates
Steady-state calculations
Chemical mechanism
Night chemistry
Chemistry integration
Subroutine
VERTCL
STEP1, STEP2, SHASTA
MSREAD, RESEM, RESPTS
SNKSRC
MICRO
DIFCOF
WINDZ
STEP3, SIMDIG
PHFACT, PHOTOL
TDRATE
STEADY
CHEM
CHEM2
STEP4
131
-------
VIII THE DISPLAY PROGRAM
Output from the Airshed Simulation Program is written in binary form
onto two library storage files that contain instantaneous and time-averaged
concentration predictions throughout the day in each grid cell for each
modeled chemical species. The display program, in turn, processes either simu-
lation output file; it can be run either back to back with the simulation
program or separately later. In the following discussion, important aspects
of display program organization and structure are considered: overall logic
flow, input/output, subroutine definition and hierarchy, and implementation
considerations.
A. OVERALL LOGIC FLOW
The executive control logic of the display program performs its
tasks in the following order. Card input data are processed; input files are
attached and positioned at the proper simulation hour; the full concentration
field for the user-specified vertical level is reconstructed from the
segmented simulation output file; and concentration maps as well as station
predictions are printed for all desired species and simulation hours. A
schematic of this logic flow is presented in Figure VIII-1.
As noted in that figure, the comparative output/analysis portion
of the display program is included only in skeletal form in the present
version. Although the executive control currently can interface an
observation file that is formatted in the same way as the simulation
output file, no analysis is performed. Extension of the display program
to do so, however, would be straightforward.
B. INPUT/OUTPUT
Input to the display program is of two types: cards and a binary
data file. The card input provides print information, desired output options,
132
-------
3
LU
X
o
LU
o
o
o
O
LU
O
O
I— O«C
LU =2
_J V> Q
O. I— -<
O O 00
I/O _J LU
i—i a. a:
GO
c;
•—i I— LU I—
t^ O > Z
>- i—i c: LU
_i a LU o
< LU (S) "Z.
z: o: CD o
< c. o c_>
LU
tj
O
LU
_j O on
a. i— i—
o z: o
oo c _j
•- <-> a.
o
a)
en
o
a.
c
QJ
U
OJ
a>
s_
3
O
3
CO
E
(O
s-
cn
o
Q.
(O
"a.
o
O>
-C
-t->
ii-
O
U
re
a>
jz
L>
CO
OJ
s_
3
CT
LU
O
o
«
133
-------
units conversion for printing, and the names and location of stations and
landmarks. Details about input parameters and their card formats can be found
in Chapter XI of the User's Manual. The data files required by the program
include: an airshed simulation output file, which contains segmented con-
centration predictions from the simulation program [these may be either time-
averaged (AVERAGE) or instantaneous (INSTANT) values]; a diffusion break
file (DIFFBREAK) which details the behavior of the diffusion break profile
during the modeling period; a region top file (REGIONTOP), which specifies
the spatial and temporal variation of the top of the modeling region during
the day; and a boundary file (BOUNDARY), which contains information necessary
to reconstruct from segmentation specifications the overall characteristics
of the modeling region.
Output from the display program is designed to assume two forms:
prediction and comparison. The first of these includes output that is based
solely on concentration values from one file. Such output may consist of
concentration maps and station prediction vertical profiles. Output of the
second form, though not available in the current program version, would be
based on direct comparison of two concentration files. The capability to
produce output of this latter type would require only a straightforward
extension of the program, since the file handling logic is already included
in the executive control.
C. SUBROUTINE DEFINITION AND HIERARCHY
Executive control as well as various subordinate functions (input,
file handling, and output control) are handled in the main routine (DISPLY)
of the display program. The subroutines used by DISPLY to produce the final
program output are listed in Table VIII-1.
As an illustration of subroutine precedence (which routines are called
by another routine), Figure VIII-2 presents a subroutine hierarchy diagram.
As each subroutine is called by another, its name is shown beneath and
indented from the calling routine.
134
-------
Routine
CALNDR
DISPLY
FMT4A1
MAPPER
PRINT]
PRMAP
XT I ME
Table VIII-1. Subroutine Descriptions
Description
Adds one to calendar date
Main program; handles executive control, input
Converts integer to alpha 4A1 for printing
Cleans up line for printing of concentration map
Produces ground-level concentration maps and vertical
concentration profiles above each station
Prints concentration map
Performs arithmetic operations on times in minutes
and 2400-hour time
135
-------
DISPLY
EOF*
INTf
AMIN11"
MINO1"
FLOATf
ABSf
PRINT1
IFIX"1"
FMT4A1
PRMAP
INTf
MAPPER
FMT4A1
* CDC-specific function.
t Inline function.
Figure VII1-2. Subroutine Hierarchy Diagram
136
-------
D. IMPLEMENTATION CONSIDERATIONS
The source deck form of this program is like that of the Airshed
Simulation Program; that is, it consists of labeled COMMON blocks, a
main program, and subprograms as described in Section VII.H. Unlike the
Airshed Simulation Program, the dimensions of its arrays are fixed and
are passed to the subroutines in COMMON blocks. Maximum dimensions can
be changed if necessary to fit the available core space.
The CDC function, EOF, can be replaced by the appropriate "END ="
clause in the READ statement. The subroutine FMT4A1 can be replaced by an
ENCODE statement if desired.
137
-------
IX FILE STRUCTURES
The 26 files in the SAI Airshed Model are divided into three categories
on the basis of their relationship to the Airshed Simulation Program: input
files, output files, and internal (scratch) files. The files input to the
Airshed Simulation Program are created by external data preparation programs
as described in Chapter IV. The output files are generated by the Airshed
Simulation Program for the storage of data by segment. The usage of these
files is discussed in Chapter VII. Figure IX-1 shows the position of each
file in the logical structure of the system. This figure is discussed in more
detail in Chapter III of the User's Manual; Table IX-1 lists, in alphabetical
order, first the input and output files and then the internal files. (The files
will also be discussed in this order in the following sections). The table
contains the ten-character code name for each file, its longer descriptive name,
and the dimensions over which file entries vary. Dimensions are not indicated
for the internal files, which are basically core images for each segment.
A. INPUT AND OUTPUT FILE STRUCTURE DEFINITIONS
This section contains, in alphabetical order, the formal descriptions of
the input and output files for the Airshed Simulation Program. Each descrip-
tion includes a table (IX-2 through IX-18) illustrating the structure and
listing the contents of each file. Also, the descriptions discuss
separately the three types of data that may appear: standard header records,
time-invariant data, and time-varying data. Furthermore, the contents of each
record are completely defined within the text by means of word definition
lists. Each data item is described by word number (order), item type (I =
integer, R = real, L = logical, A = alphanumeric), and by means of a definition
that includes units and dimensionality, if any.
138
-------
/• 1
I Simulation I
I Controls I
I , 1
i ,
SPREP I
r
T'
J_
SINCONTIDLI
I DIFF1REAK (
( RESIOKTOP ( ( CN1SSIOHS ( ( TCRMIN (
U1NO
TEHPERATUR I
ICTSCALARS
ROADWAY
AIRQUALITYI
k \
UUMOARY
TDPOMC
•ONI
CON1
cum
K
K
K
KM?
0*2
SEC1
IHST
I
/ I ( 1 ( I I «—«-
I IMSTANT I I AVERME I I VM1ATION I I Trace
! J
Figure IX-1. Files in the SAI Airshed Model System
139
-------
Code Name
Table IX-1. SAI Airshed Model System Files
and Their Dimensions
(a) Input and Output Files
Descriptive Name
Dimensions*
AIRQUALITY
AVERAGE
BOUNDARY
CHEMPARAM
DIFFBREAK
EMISSIONS
INSTANT
METSCALARS
PTSOURCE
REGIONTOP
ROADWAY
SIMCONTROL
TEMPERATUR
TERRAIN
TOPCONC
VARIATION
WIND
Air Quality
Average Concentrations
Boundary Definition and Concentrations
Chemistry Parameters
Diffusion Break
Ground-Level Emissions
Instantaneous Concentrations
Meteorological Scalars
Point Source Emissions
Top of the Region
Roadway Microscale Parameters
Simulation Controls
Temperature
Terrain
Concentrations at Top of Region
Subgrid-Scale Concentration Variation
Wind
X
X
(x
X
X
X
(x
X
X
X
X
X
X
X
y z s
y z s
y) z s
s
y
y s
y z s
y z) s
y
y s
y
y
y s
y s
y z
t
t
t
t
t
t
t
t
t
t
t
t
t
t
x = east-west axis of grid.
y = north-south axis of grid.
z = vertical axis of grid.
s = species.
t = time.
() = data that vary in space but are not present over the entire grid.
140
-------
Table IX-1 (Concluded)
(b) Internal Files
Code Name Descriptive Name
BON1 Boundary Table
BON2 Boundary Table
CON! Concentrations
CON2 Concentrations
CUM! Cumulative Concentrations
CUM2 Cumulative Concentrations
INTERINST Interim Instantaneous Concentrations
SEG1 Scratch Table
SEG2 Scratch Table
141
-------
1. Air Quality
The Air Quality (AIRQlJALITY) File contains observed concentration
values as a function of time for each species over the region (x,y,z).
a. Header Records
The AIRQUALITY File begins with four standard -header records.
1) The File Description Header Record contains 76 words:
1-10 A File name = 'AIRQUALITY1, ten characters,
one character per word
11-70 A File identifier, 60 characters, one char-
acter per word
71 I Number of segments
For unsegmented files, this number = 1.
72 I Number of chemical species
73 I Beginning date of the file (Julian)
74 R Beginning time of the file (hours)
75 I Ending date of the file (Julian)
76 R Ending time of the file (hours)
Words 73-76 describe the total time span contained on the
file.
2) The Region Description Header Record contains 15 words. The
first three words define the reference origin:
1 R x-coordinate (UTM units)
2 R y-coordinate (UTM units)
3 I UTM zone
The next two words define the location of the modeling region
with respect to the reference origin:
142
-------
4 R x-location (meters)
5 R y-location (meters)
The next two words define the size of each grid cell in
the x- and y-directions, respectively:
6 R Grid cell size —x-direction (meters)
7 R Grid cell size —y-direction (meters)
The next three words define the size of the modeling
region in grid cells:
8 I Number of grid cells—x-direction
9 I Number of grid cells—y-direction
10 I Number of grid cells—z-direction
The last five words describe the vertical distribution
of grid cells:
11 I Number of cells between surface layer
and diffusion break
12 I Number of cells between diffusion break
and top of region
13 R Height of surface layer (meters)
14 R Minimum height of cells between surface
layer and diffusion break (meters)
15 R Minimum height of cells between diffusion
break and top of region (meters)
3) The Segment Description Header Record contains one group of
four words for each segment; the number of segments appears
in the File Description Header Record:
1 I x-location of segment origin with respect
to origin of modeling region (grid units)
143
-------
2 I y-location of segment origin with respect
to origin of modeling region (grid units)
3 I Number of grid cells in segment—x-direction
4 I Number of grid cells in segment— y-direction
4) The Species Description Header Record contains ten words for
each species; the number of species is defined in the File
Description Header Record:
1-10 A Species name, ten characters, one char-
acter per word
b. Time-Invariant Data
The AIRQUALITY File contains no time-invariant data.
c. Time-Varying Data
The AIRQUALITY File contains one set of the following records for
each time interval.
1) The Time Interval Record contains four words:
1 I Beginning date (Julian)
2 R Beginning time (hours)
3 I Ending date (Julian)
4 R Ending time (hours)
2) For each segment of the region the AIRQUALITY File contains a
set of Concentration Records, grouped by species, and ordered
within each species by vertical level. The first eleven words
of the record identify segment and species:
1 I Segment number
2-11 A Species name, ten characters, one char-
acter per word
144
-------
The next group of words is the concentration array itself:
R Concentrations (ppm, or yg/m for AEROSOLS)
for each cell in one vertical level, varying
by x-, then y-directions.
145
-------
Table IX-2. AIRQUALITY File Structure
a. Header Records
1) File Description Header Record
File name
File Identifier
Number of segments
Number of species
Time span
2) Region Description Header Record
Reference origin
Grid origin
Grid cell size
Region size
Vertical distribution
3) Segment Description Header Record
For each segment:
Segment origin
Segment size
4) Species Description Header Record
For each species:
Species name
b. Time-Invariant Data—none
c. Time-Varying Data—one set per time interval
1) Time Interval Record
Beginning date and time
Ending date and time
2) For each segment:
For each species:
For each level:
Concentration Record
Segment Number
Species name
Concentrations (x,y)
146
-------
2. Average Concentrations
The Average Concentrations (AVERAGE) File contains time-averaged
concentrations for each species over the region.
a. Header Records
The AVERAGE File begins with four standard header records.
1) The File Description Header Record contains 76 words:
1-10 A File name = 'AVERAGE ', ten characters, one
character per word.
11-70 A File identifier, 60 characters, one char-
acter per word
71 I Number of segments
For unsegmented files, this number = 1.
72 I Number of chemical species
73 I Beginning date of the file (Julian)
74 R Beginning time of the file (hours)
75 I Ending date of the file (Julian)
76 R Ending time of the file (hours)
Words 73-76 describe the total time span contained on the file.
For the AVERAGE File, the ending date and time may not correspond
to the last date and time on the file.
2) The Region Description Header Record contains 15 words.
The first three words define the reference origin:
1 R x-coordinate (UTM units)
2 R y-coordinate (UTM units)
3 I UTM zone.
The next two words define the location of the modeling region
with respect to the reference origin:
147
-------
4 R x-location (meters)
5 R y-location (meters)
The next two words define the size of each grid cell in the
x- and y-directions, respectively:
6 R Grid cell size--x-direction (meters)
7 R Grid cell size--y-direction (meters)
The next three words define the size of the modeling region
in grid cells:
8 I Number of grid cells—x-direction
9 I Number of grid cells—y-direction
10 I Number of grid cells--z-direction
The last five words describe the vertical distribution of
grid cells:
11 I Number of cells between surface layer and
diffusion break
12 I Number of cells between diffusion break and
top of region
13 R Height of surface layer (meters)
14 R Minimum height of cells between surface layer
and diffusion break (meters)
15 R Minimum height of cells between diffusion break
and top of region (meters)
3) The Segment Description Header Record contains one group of
four words for each segment; the number of segments appears
in the File Description Header Record:
1 I x-location of segment origin with respect to
origin of modeling region (grid units)
148
-------
2 I y-location of segment origin with respect to
origin of modeling region (grid units)
3 I Number of grid cells in segment—x-direction
4 I Number of grid cells in segment—y-direction
4) The Species Description Header Record contains ten words for
each species; the number of species is defined in the File
Description Header Record:
1-10 A Species name, ten characters, one character
per word
b. Time Invariant Data
The AVERAGE File contains no time-invariant data.
c. Time-Varying Data
The AVERAGE File contains one set of the following records for each
time interval.
1) The Time Interval Record contains four words:
1 I Beginning date (Julian)
2 R Beginning time (hours)
3 I Ending date (Julian)
4 R Ending time (.hours)
2) For each segment of the region the AVERAGE File contains a set of
Average Concentration Records, grouped by species, and ordered within
each species by vertical level. The first 11 words of the record
identify segment and species:
1 I Segment number
2-11 A Species name, ten characters, one character
per word
149
-------
The next group of words is the concentration array itself:
12+ R Concentrations (ppm, or yg/m3 for AEROSOLS)
averaged over the time interval for each cell
in one vertical level, varying by x-, then
y-directions.
150
-------
Table IX-3. AVERAGE File Structure
a. Header Records
1) File Description Header Record
File name
File Identifier
Number of segments
Number of species
Time span
2) Region Description Header Record
Reference origin
Grid origin
Grid cell size
Region size
Vertical distribution
3) Segment Description Header Record
For each segment:
Segment origin
Segment size
4) Species Description Header Record
For each species:
Species name
b. Time-Invariant Data—none
c. Time-Varying Data-one set per time interval
1) Time Interval Record
Beginning date and time
Ending date and time
2) For each segment:
For each species:
For each level:
Average Concentration Record
-Segment number
Species name
Concentrations (x,y)
151
-------
3. Boundary Definition and Concentrations
The Boundary Definition and Concentrations (BOUNDARY) File contains the
definition of external boundaries and segment interfaces and the concent*-a-
tions for each species at each external boundary cell over the region (x,y,z)
a. Header Records
The BOUNDARY File begins with four standard header records.
1) The File Description Header Record contains 76 words.
1-10 A File name * 'BOUNDARY ', ten characters,
one character per word
11-70 A File identifier, 60 characters, one char-
acter per word
71 I Number of segments
For unsegmented files, this number = 1.
72 I Number of chemical species
73 I Beginning date of the file (Julian)
74 R Beginning time of the file (hours)
75 I Ending date of the file (Julian)
76 R Ending time of the file (hours)
Words 73-76 describe the total time span contained on the
file.
2) The Region Description Header Record contains 15 words. The
first three words define the reference origin:
1 R x-coordinate (UTM units)
2 R y-coordinate (UTM units)
3 I UTM zone
The next two words define the location of the modeling region
with respect to the reference origin:
152
-------
4 R x-location (meters)
5 R y-location (meters)
The next two words define the size of each grid cell In
the x- and y-directions, respectively:
6 R Grid cell size--x-direction (meters)
7 R Grid cell size—y-direction (meters)
The next three words define the size of the modeling
region in grid cells:
8 I Number of grid cells—x-direction
9 I Number of grid cells—y-direction
10 I Number of grid cells—z-direction
The last five words describe the vertical distribution
of grid cells:
11 I Number of cells between surface layer and
and diffusion break
12 I Number of cells between diffusion break
and top of region
13 R Height of surface layer (meters)
14 R Minimum height of cells between surface
layer and diffusion break (meters)
15 R Minimum height of cells between diffusion
break and top of region (meters)
3) The Segment Description Header Record contains one group of
four words for each segment; the number of segments appear
in the File Description Header Record:
1 I x-location of segment origin with respect to
origin of modeling region (grid units)
153
-------
2 I y-location of segment origin with respect
to origin of modeling region (grid units)
3 I Number of grid cells in segment--x-direction
4 I Number of grid cells in segment--y-direction
4) The Species Description Header Record contains ten words for
each species; the number of species is defined in the File
Description Header Record:
1-10 A Species name, ten characters, one character
per word.
b. Time-Invariant Data
The BOUNDARY File contains the definition of the boundaries
of the region, both the external boundaries defined in the data
preparation and the internal segment interfaces created when the
region is segmented.
For each segment there are four Boundary Definition Records,
one record for each edge. The edges are defined as follows:
1 Left West lower limit for each row
2 Right East upper limit for each row
3 Bottom South lower limit for each column
4 Top North upper limit for each column
Each Boundary Definition Record defines the location of the boundary
cells at an edge of a segment. The first three words identify
the edge and its dimensions;
1 I Segment number
2 I Edge number
3 I Number of cells on edge
(i.e., number of rows or columns)
154
-------
Then, for each grid index (row or column) along the edge, there is a
group of four words that define the boundary location:
4 I Index, within the segment, of the cell at the
edge of the region modeled (i.e., the first or
last cell simulated within the row or column).
If this number = 0, this row or column is to
be omitted from the simulation and the next
three numbers are ignored.
5 I Segment number in which adjacent cell is located.
If this number = 0, the boundary is an external
one, and the next two numbers are ignored.
6 I x-index of adjacent cell within segment defined in
Word 5.
7 I y-index of adjacent cell within segment defined in
Word 5.
c. Time-Varying Data
The BOUNDARY File contains one set of the following records for each
time interval.
1) The Time Interval Record contains four words:
1 I Beginning date (Julian)
2 R Beginning time (hours)
3 I Ending date (Julian)
4 R Ending Time (hours)
2) For each segment of the region, there is a set of Boundary
Concentration Records, grouped by species, with four records
(one for each edge) for each species. The first twelve words
of the record identify segment, species, and edge:
155
-------
1 I Segment number.
2-11 A Species name, ten characters, one character
per word.
12 I Edge number
The next group of words is the boundary concentration array on the
vertical plane along the edge:
13+ R Boundary concentrations (ppm, or vg/m3 for AEROSOLS)
at each vertical level for each cell along the edge.
For rows or columns that are not to be simulated and
for edges that represent internal segment boundaries,
these numbers must be present, but will be ignored.
156
-------
Table IX-4. BOUNDARY File Structure
a. Header Records
1) File Description Header Record
File name
File identifier
Number of segments
Number of species
Time span
2) Region Description Header Record
Reference origin
Grid origin
Grid cell size
Region size
Vertical distribution
3) Segment Description Header Record
For each segment:
Segment origin
Segment size
4) Species Description Header Record
For each species:
Species name
b. Time-Invariant Data
For each segment:
For each edge:
Boundary Definition Record
Segment number
Edge number
Number of cells on edoe
For each cell:
Index of boundary
Location of adjacent cell
157
-------
Table IX-4 (Concluded)
c. Time-Varying Data—one set per time interval
1) Time Interval Record
Beginning date and time
Ending date and time
2) For each segment:
For each species:
For each edge:
Boundary Concentration Record
Segment number
Species name
Edge number
Boundary concentrations (z.cell)
158
-------
4. Chemistry Parameters
The Chemistry Parameters (CHEMPARAM) File contains the definition of
species characteristics, reaction properties, and stoichiometric coef-
ficients.
a- Header Records
The CHEMPARAM File begins with two standard header records.
1) The File Description Header Record contains 76 words:
1-10 A File name = 'CHEMPARAM ', ten characters, one
character per word.
11-70 A File identifier, 60 characters, one character
per word.
71 I Number of segments
For the CHEMPARAM File, which does not vary
spatially, this number = 0 and the Segment
Definition Header Record is omitted.
72 I Number of chemical species
73 I Beginning date of the file (Julian)
74 R Beginning time of the file (hours)
75 I Ending date of the file (Julian)
76 R Ending time of the file (hours)
Words 73-76 describe the total time span contained on the file.
For the CHEMPARAM File, which does not vary in time, these
numbers are ignored.
2) The Region Description Header Record is omitted for the CHEMPARAM
File, which is independent of modeling region.
159
-------
3) The Segment Description Header Record is omitted for the
CHEMPARAM File, which does not vary spatially.
4) The Species Description Header Record contains ten words for
each species; the number of species is defined in the File
Description Header Record:
1-10 A Species name, ten characters, one character
per word.
b. Time-Invariant Data
The CHEMPARAM File contains two to four time-invariant data records.
1) The Chemistry Definition Record contains 62 words:
1-60 A Mechanism identifier, 60 characters, one
character per word
61 I Number of reactions
62 I Number of stoichiometric coefficients
2) The Species Parameters Record contains one group of 18 words
for each species; the number of species is defined in the
File Definition Header Record:
1-10 A Species name, ten characters, one character per word
11 L Reactive/Unreactive Flag
.TRUE. * Reactive
.FALSE. = Unreactive
12 L Steady-state for initial concentrations
.TRUE. » Compute steady-state values r
.FALSE. = Use input values
160
-------
13 L Steady-state for boundary concentrations
.TRUE. = Compute steady-state values
.FALSE. = Use input values
14 R Resistance to surface sinks (hr/m)
15 R Lower bound for steady-state calculations (ppm)
16 R Upper bound for steady-state calculations (ppm)
17 R Lower bound for numerical integration (ppm, or
ug/m3 for AEROSOLS)
18 R Upper bound for numerical integration (ppm, or
wg/m3 for AEROSOLS)
3) The Reaction Parameters Record contains one group of five
words for each reaction; the number of reactions is specified
in the Chemistry Definition Record (if the number of reactions = 0,
this record is omitted):
1 R Rate constant
2 L Photolysis flag
.TRUE. = Photolysis reaction
.FALSE. = Not a photolysis reaction
3 L Temperature-dependence flag
.TRUE. = Temperature-dependent reaction
.FALSE. = Not a temperature-dependent reaction
If this is FALSE the next two words will
be ignored
4 R Activation energy (deg K)
5 R Reference temperature for activation energy (deg K)
4) The Stoichiometric Coefficients Record contains one group
of eleven words for each Stoichiometric coefficient; the
number of these coefficients is specified in the Chemistry
Definition Record (if the number of coefficients = 0, this
record is omitted):
161
-------
1-10 A Coefficient name, ten characters, one character
per word
11 R Coefficient value
c. Time-Varying Data
The CHEMPARAM File contains no time-varying data.
162
-------
Table IX-5. CHEMPARAM File Structure
a. Header Records
1) File Description Header Record
File name
File identifier
Number of segments
Number of species
Time span
2) Region Descriotion Header Record—omitted
3) Segment Description Header Record—omitted
4) Species Description Header Record
For each species:
Species name
b. Time-Invariant Data
1) Chemistry Definition Record
Mechanism identifier
Number of reactions
Number of stoichiometric coefficients
2) Species Parameters Record
For each species:
Name
Reactive flag
Steady-state flags
Steady-state bounds
Numerical integration bounds
Resistance to surface sinks
163
-------
Table IX-5 (Concluded)
3) Reaction Parameters Record
For each reaction:
Rate constant
Photolysis flag
Temperature-dependent flags
Activation energy
Reference temperature
4) Stoichiometric Coefficients Record
For each coefficient:
Name
Value
c. Time-Varying Data—none
164
-------
5. Diffusion Break
The Diffusion Break (DIFFBREAK) File contains the height of the
change of diffusion characteristics over the region (x,y).
a. Header Records
The DIFFBREAK File begins with three standard header records.
1) The File Description Header Record contains 76 words:
1-10 A File name = 'DIFFBREAK ', ten characters, one
character per word.
11-70 A File identifier, 60 characters, one character
per word.
71 I Number of segments
For unsegmented files, this number =1.
72 I Number of chemical species
For the DIFFBREAK File, which does not vary with
species, this number = 0 and the Species Definition
Header Record is omitted.
73 I Beginning date of the file (Julian)
74 R Beginning time of the file (hours)
75 I Ending date of the file (Julian)
76 R Ending time of the file (hours)
Words 73-76 describe the total time span contained on the file.
2) The Region Description Header Record contains 15 words.
The first three words define the reference origin:
165
-------
T R x-coordinate (UTM units)
2 R y-coordinate (UTM units)
3 I UTM zone
The next two words define the location of the modeling region
with respect to the reference origin:
4 R x-location (meters)
5 R y-location (meters)
The next two words define the size of each grid cell in the
x- and y-directions, respectively:
6 R Grid cell size—x-direction (meters)
7 R Grid cell size—y-direction (meters)
The next three words define the size of the modeling region in
grid cells:
8 I Number of grid cells—x-direction
9 I Number of grid cells—y-direction
10 I Number of grid cells—z-direction
The last five words describe the vertical distribution of
grid cells. For the DIFFBREAK File, which does not vary
vertically, these words are ignored:
1l I Number of cells between surface layer and diffusion
break
12 I Number of cells between diffusion break and top of
region
13 R Height of surface layer (meters)
166
-------
74 R Minimum height of cells between surface layer and
diffusion break (meters)
15 R Minimum height of cells between diffusion break
and top of region (meters)
3) The Segment Description Header Record contains one group of
four words for each segment; the number of segments appears in
the File Description Header Record:
1 I x-location of segment origin with respect to
origin of modeling region (grid units)
2 I y-location of segment origin with respect to
origin of modeling region (grid units)
3 I Number of grid cells in segment—x-direction
4 I Number of grid cells in segment—y-direction
4) The Species Description Header Record is omitted for the
DIFFBREAK File, which does not vary with species.
b. Time-Invariant Data
The DIFFBREAK File contains no time-invariant data.
c. Time-Varying Data
The DIFFBREAK File contains one set of the following records for
each time interval.
1) The Time Interval Record contains four words:
1 I Beginning date (Julian)
2 R Beginning time (hours)
167
-------
3 I Ending date (Julian)
4 R Ending time (hours)
2) For each segment of the region there is one Diffusion
Break Record. The first eleven words of the record
identify segment and variable:
1 I Segment
2-11 A Variable name = 'DIFFBREAK '
ten characters, one character per word
The next group of words is the diffusion break array itself:
12+ R Diffusion break (meters) at the beginning of
the time interval at each cell in the row for
each row in the segment
168
-------
Table IX-6. DIFFBREAK File Structure
a. Header Records
1) File Description Header Record
File name
File identifier
Number of segments
Number of species
Time span
2) Region Description Header Record
Reference origin
Grid origin
Grid cell size
Region size
Vertical distribution
3) Segment Description Header Record
For each segment:
Segment origin
Segment size
4) Species Description Header Record—omitted
b. Time-Invariant Data—none
c. Time-Varying Data—one set per time interval
1) Time Interval Record
Beginning date and time
Ending date and time
2) For each segment:
Diffusion Break Record
Segment number
Variable name
Diffusion break (x,y)
169
-------
6. Ground-Level Emissions
The Ground-Level Emissions (EMISSIONS) File contains emissions
values for each species over the region (x, y).
a. Header Records
The EMISSIONS File begins with four standard header records.
1) The File Description Header Record contains 76 words:
1-10 A File name = 'EMISSIONS ', ten characters, one
character per word.
11-70 A File identifier, 60 characters, one character
per word.
71 I Number of segments
For unsegmented files, this number = 1.
72 I Number of chemical species
73 I Beginning date of the file (Julian)
74 R Beginning time of the file (hours)
75 I Ending date of the file (Julian)
76 R Ending time of the file (hours)
Words 73-76 describe the total time span contained on the file.
2) The Region Description Header Record contains 15 words.
The first three words define the reference origin:
1 R x-coordinate (UTM units)
2 R y-coordinate (UTM units)
3 I UTM zone
The next two words define the location of the modeling region
with respect to the reference origin:
178
-------
4 R x-location (meters)
5 R y-location (meters)
The next two words define the size of each grid cell in the
x- and y-directions, respectively:
6 R Grid cell size~x-direct ion (meters)
7 R Grid cell size--y-direct!on (meters)
The next three words define the size of the modeling region in
grid cells:
8 I Number of grid cells—x-direction
9 I Number of grid cells—y-direction
10 I Number of grid cells—z-direction
The last five words describe the vertical distribution of
grid cells. For the EMISSIONS File, which does not vary
vertically, these words are ignored:
11 I Number of cells between surface layer and
diffusion break
12 I Number of cells between diffusion break and top of
region
13 R Height of surface layer (meters)
14 R Minimum height of cells between surface layer and
diffusion break (meters)
15 R Minimum height of cells between diffusion break
and top of region (meters)
3) The Segment Description Header Record contains one group of
four words for each segment; the number of segments appears in
the File Description Header Record:
171
-------
1 I x-location of segment origin with respect to
origin of modeling region (grid units)
2 I y-location of segment origin with respect to
origin of modeling region (grid units)
3 I Number of grid cells in segment—x-direction
4 I Number of grid cells in segment—y-direction
4) The Species Description Header Record contains ten words for
each species; the number of species is defined in the File
Description Header Record:
1-10 A Species name, ten characters, one character
per word.
b. Time-Invar! ant Data
The EMISSIONS File contains no time-invariant data.
c. Time-Varying Data
The EMISSIONS File contains one set of the following records for
each time interval.
1) The Time Interval Record contains four words:
1 I Beginning date (Julian)
2 R Beginning time (hours)
3 I Ending date (Julian)
4 R Ending time (hours)
2) For each segment of the region there is a set of Emissions
Records, one record for each species. The first eleven words
of the record identify segment and species:
172
-------
1 I Segment number
2-11 A Species name, ten characters, one character per word
The next group of words is the emissions array itself:
12+ R Ground-level emissions (gram-moles/hour, or
grams/hour for AEROSOLS)
173
-------
Table IX-7. EMISSIONS File Structure
a. Header Records
1) File Description Header Record
File name
File identifier
Number of segments
Number of species
Time span
2) Region Description Header Record
Reference origin
Grid origin
Grid cell size
Region size
Vertical distribution
3) Segment Description Header Record
For each segment:
Segment origin
Segment size
4) Species Description Header Record
For each species:
Species name
b. Time-Invariant Data—none
c. Time-Varying Data—one set per time interval
1) Time Interval Record
Beginning date and time
Ending date and time
174
-------
Table IX-7 (Concluded)
2) For each segment:
For each species:
Emissions Record
Segment number
Species name
Ground-level emissions (x,y)
175
-------
7. Instantaneous Concentrations
The Instantaneous Concentrations (INSTANT) File contains calculated con-
centration values for each species over the region (x,y,z).
a- Header Records
The INSTANT File begins with four standard header records.
1) The File Description Header Record contains 76 words:
1-10 A File name = 'AIRQUALITY1; ten characters, one
character per word
11-70 A File identifier, 60 characters, one char-
acter per word
71 I Number of segments
For unsegmented files, this number = 1
72 I Number of chemical species
73 I Beginning date of the file (Julian)
74 R Beginning time of the file (hours)
75 I Ending date of the file (Julian)
76 R Ending time of the file (hours)
Words 73-76 describe the total time span contained on the file. For
the INSTANT File, the ending date and time may not correspond to the
last date and time on the file.
2) The Region Description Header Record contains 15 words.
The first three words define the reference origin:
1 R x-coordinate (UTM units)
2 R y-coordinate (UTM units)
3 I UTM zone
The next two words define the location of the modeling region with
respect to the reference origin:
176
-------
4 R x-location (meters)
5 R y-location (meters)
The next two words define the size of each grid cell in the
x- and y-directions, respectively:
6 R Grid cell size~x-direction (meters)
7 R Grid cell size—y-direction (meters)
The next three words define the size of the modeling region
in grid cells:
8 I Number of grid cells—x-direction
9 I Number of grid cells—y-direction
10 I Number of grid cells—z-direction
The last five words describe the vertical distribution of
grid cells:
11 I Number of cells between surface layer and
diffusion break
12 I Number of cells between diffusion break and
top of region
13 R Height of surface layer (meters)
14 R Minimum height of cells between surface layer
and diffusion break (meters)
15 R Minimum height of cells between diffusion
break and top of region (meters)
3) The Segment Description Header Record contains one group of
four words for each segment; the number of segments appears
in the File Description Header Record:
1 I x-location of segment origin with respect to
origin of modeling region (grid units)
2 I y-location of segment origin with respect to
origin of modeling region (grid units)
177
-------
3 I Number of grid cells in segment—x-direction
4 I Number of grid cells in segment--y-direction
4) The Species Description Header Record contains ten words for
each species; the number of species is defined in the File
Description Header Record:
1-10 A Species name, ten characters, one character
per word
b. Time-Invariant Data
•
The INSTANT File contains no time-invariant data.
c. Time-Varying Data
The INSTANT File contains one set of the following records for each
time interval.
1) The Time Interval Record contains four words:
1 I Beginning date (Julian)
2 R Beginning time (hours)
3 I Ending date (Julian)
4 R Ending time (hours)
2) For each segment of the region there is a set of Instantane-
ous Concentration Records, grouped by species, and ordered within
each species by vertical level. The first eleven words of the
record identify segment and species:
1 I Segment number
2-11 A Species name, ten characters, one character
per word
178
-------
The next group of words is the concentration array itself:
12+ R Concentrations (ppm, or pg/m3 for AEROSOLS) at
the beginning of the time interval for each
cell in one vertical level, varying by x-, then
y-directions.
179
-------
Table IX-8. INSTANT File Structure
a. Header Records
1) File Description Header Record
File name
File identifier
Number of segments
Number of species
Time span
2) Region Description Header Record
Reference origin
Grid origin
Grid cell size
Region size
Vertical distribution
3) Segment Description Header Record
For each segment:
Segment origin
Segment size
4) Species Description Header Record
For each species:
Species name
b. Time-Invariant Data—none
c. Time-Varying Data—one set per time interval
1) Time Interval Record
Beginning date and time
Ending date and time
2) For each segment:
For each species:
For each vertical level:
Instantaneous Concentration Record
Segment number
Species name
Concentrations (x,y)
180
-------
8. Meteorological Scalars
The Meteorological Scalars (METSCALARS) File contains values for
meteorological parameters that are assumed to be constant over the
entire region.
a. Header Records
The METSCALARS File begins with two standard header records.
1) The File Description Header Record contains 76 words:
1-10 A File name = 'METSCALARS1, ten characters, one
character per word.
11-70 A File identifier, 60 characters, one character
per word.
71 I Number of segments
For the METSCALARS File, which does not vary
spatially, this number = 0 and the Segment Definition
Header Record is omitted.
72 I Number of chemical species
For the METSCALARS File, which does not vary with
species, this number = 0 and the Species Definition
Header Record is omitted.
73 I Beginning date of the file (Julian)
74 R Beginning time of the file (hours)
75 I Ending date of the file (Julian)
76 R Ending time of the file (hours)
Words 73-76 describe the total time span contained on the file.
181
-------
2) The Region Description Header Record contains 15 words.
The first three words define the reference origin:
1 R x-coordinate (UTM units)
2 R y-coordinate (UTM units)
3 I UTM zone
The next two words define the location of the modeling region
with respect to the reference origin:
4 R x-location (meters)
5 R y-location (meters)
The next two words define the size of each grid cell in the
x- and y-directions, respectively:
6 R Grid cell size~x-direction (meters)
7 R Grid cell size—y-direct ion (meters)
The next three words define the size of the modeling region in
grid cells:
8 I Number of grid cells—x-direction
9 I Number of grid cells—y-direction
10 I Number of grid cells—z-direction
The last five words describe the Vertical distribution of
grid cells. For the METSCALARS File, which does not vary
vertically, these words are ignored:
11 I Number of cells between surface layer and
diffusion break
12 I Number of cells between diffusion break and
top of region
182
-------
13 R Height of surface layer (meters)
14 R Minimum height of cells between surface layer
and diffusion break (meters)
15 R Minimum height of cells between diffusion break
and top of region (meters)
3) The Segment Description Header Record is omitted for
the METSCALARS File, which does not vary spatially.
4) The Species Description Header Record is omitted for the
METSCALARS File, which does not vary with species.
b. Time-In van'ant Data
The METSCALARS File contains no time-invariant data.
c. Time-Varying Data
The METSCALARS File contains two records for each time interval.
1) The Time Interval Record contains four words:
1 I Beginning date (Julian)
2 R Beginning time (hours)
3 I Ending date (Julian)
4 R Ending time (hours)
2) The Meteorological Scalars Record contains 66 words, or six
name-number pairs:
1-10 A Name = 'TGRADBELOW
11 R Temperature gradient (deg K/meter) below the
diffusion break
183
-------
12-21 A Name « 'TGRADABOVE', ten characters, one
character per word
22 R Temperature gradient (deg K/meter) above
the diffusion break
23-32 A Name * 'EXPCLASS ', ten characters, one
character per word
33 R Exposure Class
34-43 A Name = 'RADFACTOR ', ten characters, one
character per word
44 R Radiation Factor at end of time interval
45-54 A Name = 'CONCWATER ', ten characters, one
character per word
55 R Concentration of H20 (ppm)
56-65 A Name = 'ATMOSPRESS', ten characters, one
character per word
66 R Atmospheric pressure (atm.)
184
-------
Table IX-9. METSCALARS File Structure
•
a. Header Records
1) File Description Header Record
File name
File identifier
Number of segments
Number of species
Time span
2) Region Description Header Record
Reference origin
Grid origin
Grid cell size
Region size
Vertical distribution
3) Segment Description Header Record—omitted
4) Species Description Header Record—omi tted
b. Time-In variant Data—none
c. Time-Varying Data—one set per time interval
1) Time Interval Record
Beginning date and time
Ending date and time
2) Meteorological Scalars Record
For each scalar:
Name
Value
185
-------
9. Paint Source Emissions
The Point Source Emissions (PTSOURCE) File contains the location of
each point source and the emissions for each species.
a. Header Records
The PTSOURCE File begins with four standard header records.
1) The File Description Header Record contains 76 words:
1-10 A File name - 'PTSOURCE ', ten characters, one
character per word.
11-70 A File identifier, 60 characters, one character
per word.
71 I Number of segments
For unsegmented files, this number = 1.
72 I Number of chemical species
73 I Beginning date of the file (Julian)
74 R Beginning time of the file (hours)
75 I Ending date of the file (Oulian)
76 R Ending time of the file (hours)
Words 73-76 describe the total time span contained on the file,
2) The Region Description Header Record contains 15 words.
The first three words define the reference origin:
1 R x-coordinate (UTM units)
2 R y-coordinate (UTM units)
3 I UTM zone
186
-------
The next two words define the location of the modeling region
with respect to the reference origin:
4 R x-location (meters)
5 R y-location (meters)
The next two words define the size of each grid cell in the
x- and y-directions, respectively:
6 R Grid cell size—x-direction (meters)
7 R Grid cell size—y-direction (meters)
The next three words define the size of the modeling region in
grid cells:
8 I Number of grid cells—x-direction
9 I Number of grid cells—y-direction
10 I Number of grid cells—z-direction
The last five words describe the vertical distribution of
grid cells:
11 I Number of cells between surface layer and
diffusion break
12 I Number of cells between diffusion break and top
of region
13 R Height of surface layer (meters)
14 R Minimum height of cells between surface layer
and diffusion break (meters)
15 R Minimum height of cells between diffusion break
and top of region (meters)
3) The Segment Description Header Record contains one group of
four words for each segment; the number of segments appears
in the File Description Header Record:
187
-------
1 I x-location of segment origin with respect to
origin of modeling region (grid units)
2 I y-location of segment origin with respect to
origin of modeling region (grid units)
3 I Number of grid cells in segment—x-direction
4 I Number of grid cells in segment—y-direction
4) The Species Description Header Record contains ten words for
each species; the number of species is defined in the File
Description Header Record:
1-10 A Species name, ten characters, one character
per word.
b. Time-Invariant Data
The PTSOURCE File contains the location and other fixed properties
of each point source. For each segment there are two records.
1) The Counter Record contains two words:
1 I Segment number
2 I Number of point sources in segment
2) The Point Source Definition Record contains the following
group of six words for each point source in the segment
If there are no point sources in the segment this record
does not appear;
1 R x-coordinate of point source with respect to
reference origin (meters)
2 R y-coordinate of point source with respect to
reference origin (meters)
3 I x-index of grid cell within the segment
188
-------
4 I y-index of grid cell within the segment
5 R Stack height (meters)
6 R Stack diameter (meters)
c. Time-Varying Data
The PTSOURCE File contains one set of the following records for each
time interval.
1) The Time Interval Record contains four words:
1 I Beginning date (Julian)
2 R Beginning time (hours)
3 I Ending date (Julian)
4 R Ending time (hours)
2) For each segment there is a Counter Record, a Point Source Location
Record, and one Point Source Emissions Record for each species. The
Counter Record contains two words:
1 I Segment number
2 I Number of point sources in segment for this time
interval. If the number of point sources defined
in the corresponding time-invariant Counter Record
is greater than zero, this number must also be greater
than zero.
The Point Source Location Record contains the following group of
five words for each point source in the segment. If there are no
point sources in the segment, this record does not appear:
1 I x-index within segment of cell to receive emissions
2 I y-index within segment of cell to receive emissions
3 I z-index of cell to receive emissions
4 R Flow rate (m3/hr)
5 R Effective plume height (meters)
189
-------
The Point Source Emissions Record contains the following group
of words:
1 I Segment number
2-11 A Species name, ten characters, one character
per word
12+ R Emissions (gm-moles/hr, or gm/hr for AEROSOLS)
from each point source
190
-------
Table IX-10. PTSOURCE File Structure
a. Header Records
1) File Description Header Record
File name
File identifier
Number of segments
Number of species
Time span
2) Region Description Header Record
Reference origin
Grid origin
Grid cell size
Region size
Vertical distribution
3) Segment Description Header Record
For each segment:
Segment origin
Segment size
4) Species Description Header Record
For each species:
Species name
b. Time-Invariant Data
1) For each segment:
Counter Record
Segment number
Number of point sources
191
-------
Table IX-10 (Concluded)
Point Source Definition Record
For each point source:
Location in region
Stack properties
c. Time-Varying Data—one set per time interval
1) Time Interval Record
Beginning date and time
Ending date and time
2) For each segment:
Counter Record
Segment number
Number of point sources
Point Source Location Record
For each point source:
Indices of cell to receive emissions
Flow rate
Effective plume height
For each species:
Point Source Emissions Record
Segment number
Species name
For each point source:
Emissions
192
-------
10. Top of Region
The Top of Region (REGIONTOP) File contains the height of the top
of the region (x,y).
a. Header Records
The REGIONTOP File begins with three standard header records.
1) The File Description Header Record contains 76 words:
1-10 A File name = 'REGIONTOP ', ten characters, one
character per word.
11-70 A File identifier, 60 characters, one character
per word.
71 I Number of segments
For unsegmented files, this number = 1.
72 I Number of chemical species
For the REGIONTOP File, which does not vary with
species, this number = 0 and the Species Definition
Header Record is omitted.
73 I Beginning date of the file (Julian)
74 R Beginning time of the file (hours)
75 I Ending date of the file (Julian)
76 R Ending time of the file (hours)
Words 73-76 describe the total time span contained on the file.
2) The Region Description Header Record contains 15 words.
The first three words define the reference origin:
1 R x-coordinate (UTM units)
2 R y-coordinate (UTM units)
3 I UTM zone
193
-------
The next two words define the location of the modeling region
with respect to the reference origin:
4 R x-location (meters)
5 R y-location (meters)
The next two words define the size of each grid cell in the
x- and y-directions, respectively:
6 R Grid cell size—x-direction (meters)
7 R Grid cell size—y-direction (meters)
The next three words define the size of the modeling region in
grid cells:
8 I Number of grid cells—x-direction
9 I Number of grid cells—y-direction
10 I Number of grid cells—z-direction
The last five words describe the vertical distribution of
grid cells. For the REGIONTOP File, which does not vary
vertically, these words are ignored:
11 I Number of cells between surface layer and
diffusion break
12 I Number of cells between diffusion break and
top of region
13 R Height of surface layer (meters)
14 R Minimum height of cells between surface layer
and diffusion break (meters)
15 R Minimum height of cells between diffusion break
and top of region (meters)
194
-------
3) The Segment Description Header Record contains one group of
four words for each segment; the number of segments appears
in the File Description Header Record:
1 I x-location of segment origin with respect to
origin of modeling region (grid units)
2 I y-location of segment origin with respect to
origin of modeling region (grid units)
3 I Number of grid cells in segment—x-direction
•
4 I Number of grid cells in segment—y-direction
4) The Species Description Header Record is omitted for the
REGIONTOP File, which does not vary with species.
b. Time-Invariant Data
The REGIONTOP File contains no time-invariant data.
c. Time-Varying Data
The REGIONTOP File contains one set of the following records for
each time interval.
1) The Time Interval Record contains four words:
1 I Beginning date (Julian)
2 R Beginning time (hours)
3 I Ending date (Julian)
4 R Ending time (hours)
2) For each segment of the region there is one Top of Region Record.
The first eleven words of the record identify segment and variable:
195
-------
T I Segment number
2-11 A Variable name = 'REGIONTOP ', ten characters,
one character per word.
The next group of words is the top of region array itself:
12+ R Top of region (meters) at the beginning of the
time interval at each cell in the row for each
row in the segment
196
-------
Table IX-11. REGIONTOP File Structure
a. Header Records
1) File Description Header Record
File name
File identifier
Number of segments
Number of species
Time span
2) Region Description Header Record
Reference origin
Grid origin
Grid cell size
Region size
Vertical distribution
3) Segment Description Header Record
For each segment:
Segment origin
Segment size
4) Species Description Header Record—omi tted
b. Time-Invariant Data—none
c. Time-Varying Data—one set per time interval
1) Time Interval Record
Beginning date and time
Ending date and time
2) For each segment:
Top of Region Record
Segment number
Variable name
Top of region (x, y)
197
-------
11. Roadway Microscale Parameters
The Roadway Microscale Parameters CROADWAY) File contains values
for parameters used in the surface layer microscale computations, varying
over the region (x,y).
a. Header Records
The ROADWAY File begins with three standard header records.
1) The File Description Header Records contains 76 words:
1-10 A File name » 'ROADWAY ', ten characters,
one character per word
11-70 A File identifier, 60 characters, one char-
acter per word
71 I Number of segments
For unsegmented files, this number = 1.
72 I Number of chemical species
73 I Beginning date of the file (Julian)
74 R Beginning time of the file (hours)
75 I Ending date of the file (Julian)
76 R Ending time of the file (hours)
Words 73-76 describe the total time span contained on the
file.
2) The Region Description Header Record contains 15 words. The
first three words define the reference origin:
1 R x-coordinate (UTM units)
2 R y-coordinate (UTM units)
3 I UTM zone
The next two words define the location of the modeling region
with respect to the reference origin.
198
-------
4 R x-location (meters)
5 R y-location (meters)
The next two words define the size of each grid cell in
the x- and y-directions, respectively:
6 R Grid cell size—x-direction (meters)
7 R Grid cell size—y-direction (meters)
The next three words define the size of the modeling
region in grid cells:
8 I Number of grid cells—x-direction
9 I Number of grid cells—y-direction
10 I Number of grid cells—z-direction
The last five words describe the vertical distribution
of grid cells. For the ROADWAY File, which does not vary
vertically, these words are ignored:
11 I Number of cells between surface layer
and diffusion break
12 I Number of cells between diffusion break
and top of region
13 R Height of surface layer (meters)
14 R Minimum height of cells between surface
layer and diffusion break (meters)
15 R Minimum height of cells between diffusion
break and top of region (meters)
3) The Segment Description Header Record contains one group of
four words for each segment; the number of segments appears
in the File Description Header Record:
1 I x-location of-segment origin with respect
to origin of modeling region (grid units)
199
-------
2 I y-location of segment origin with respect
to origin of modeling region (grid units)
3 I Number of grid cells in segment—x-direction
4 I Number of grid cells in segment--y-direction
b. Time-Invariant Data
The ROADWAY File contains no time-invariant data.
c. Time-Varying Data
The ROADWAY File contains one set of the 'following records for
each time interval.
1) The Time Interval Record contains four words:
1 I Beginning date (Julian)
2 R Beginning time (hours)
3 I Ending date (Julian)
4 R Ending time (hours)
2) For each segment of the region there is one Roadway Micro-
scale Parameter (CARM) Record and a set of Roadway Emissions
Records, one record for each species:
In the CARM Record the first 11 words identify the segment
•
number and variable name:
1 I Segment number
2-11 A Variable name = 'CARM ', ten characters,
one character per word
The next group of words is the CARM array itself:
R CARM over the segment
200
-------
In each Roadway Emissions Record the first 11 words identify
the segment number and species name:
1 I Segment number
2-11 A Species name, ten characters, one character
per word
The next group of words is the roadway emissions array itself:
12+ R Roadway emissions (gram-moles/hour, or grams/hour
for AEROSOLS) over the segment
201
-------
Table IX-12. ROADWAY File Structure
a. Header Records
1) File Description Header Record
File name
File identifier
Number of segments
Number of species
Time span
2) Region Description Header Record
Reference origin
Grid origin
Grid cell size
Region size
Vertical distribution
3) Segment Description Header Record
For each segment:
Segment origin
Segment size
4) Species Description Header Record
For each species:
Species name
b. Time-Invariant Data—none
c. Time-Varying Data—one set per time interval
1) Time Interval Record
Beginning date and time
Ending date and time
2) For each segment:
CARM Record
Segment number
Variable name
CARM (x,y)
-------
Table IX-12 (Concluded)
For each species:
Roadway Emissions Record
Segment number
Species name
Roadway emissions (x,y)
203
-------
12. Simulation Controls
The Simulation Controls (SIMCONTROL) File contains calculation
options, integration parameters, output controls, and other para-
meters required by the simulation program.
a. Header Records
The SIMCONTROL File begins with one standard header record.
1) The File Description Header Record contains 76 words:
1-10 A File name = 'SIMCONTROL1, ten characters,
one character per word
11-70 A File identifier, 60 characters, one char-
acter per word
71 I Number of segments
For the SIMCONTROL File, which does not vary
spatially, this number = 0 and the Segment
Definition Header Record is omitted.
72 I Number of chemical species
For the SIMCONTROL File, which does not vary
with species, this number = 0 and the Species
Definition Header Record is omitted.
73 I Beginning date of the file (Julian)
74 R Beginning time of the file (hours)
75 I Ending date of the file (Julian)
76 R Ending time of the file (hours)
Words 73-76 describe the total time span contained on the file.
For the SIMCONTROL File, which does not vary in time, these
numbers are ignored.
2) The Region Description Header Record is omitted for the SIMCONTROL
File, which is independent of region.
204
-------
3) The Segment Description Header Record is omitted for the
SIMCONTROL File, which does not vary spatially.
4) The Species Description Header Record is omitted for the
SIMCONTROL File, which does not vary with species.
b. Time-Invariant Data
The SIMCONTROL File contains one Simulation Controls Record that in-
cludes all simulation control variables.
The Simulation Controls Record contains 87 words:
1-60 A Run identifier—sixty characters, one character
per word
61 I Beginning date of the simulation (Julian)
62 R Beginning time of the simulation (hours)
63 I Ending date of the simulation (Julian)
64 R Ending time of the simulation (hours)
65 L Restart flag
66 L Surface sink flag
67 L Point source flag
68 L ROADWAY file flag
69 L TEMPERATUR file flag
70 L TERRAIN file flag
71 L Concentration variation flag
72 R Default surface roughness (meters)
73 R Default vegetation factor (fraction alfalfa)
74 R Maximum time slice size (hours)
75 I Maximum number of steps in a slice
76 R Minimum step size (hours)
77 I Maximum number of iterations in chemistry
78 R Relative error tolerance for convergence in chemistry
79 R Darkness criterion (same units as RADFACTOR on the
METSCALARS file)
205
-------
80 R Time interval for instantaneous concentrations output
81 R Time interval for averaging concentrations
82 I File history print flag
83 I Core allocation print flag
84 I Number of levels for instantaneous concentration print
85 I Number of levels for average concentration print
86 I Concentration variation print flag
87 I Print option not implemented
c. Time-Varying Data
The SIMCONTROL File contains no time-varying information.
206
-------
Table IX-13. SIMCONTROL File Structure
a. Header Records
1) File Description Header Record
File name
File identifier
Number of segments
Number of species
Time span
2) Region Description Header Record—omitted
3) Segment Description Header Record—omitted
4) Species Description Header Record—omitted
b. Time-Invariant Data
1) Simulation Control Record
Run identifier
Time span of simulation
Option flags
Default terrain values
Integration controls
Output intervals
Print options
c. Time-Varying Data—none
207
-------
13. Temperature
The Temperature (TEMPERATUR) File contains ground-level tempera-
tures over the region (x,y).
a. Header Records
The TEMPERATUR File begins with three standard header records.
1) The File Description Header Record contains 76 words:
1-10 A File name = 'TEMPERATUR', ten characters,
one character per word
11-70 A File identifier, 60 characters, one char-
acter per word
71 I Number of segments
For unsegmented files, this number = 1.
72 I Number of chemical species
For the TEMPERATUR File, which does not vary
with species, this number = 0 and the Species
Definition Header Record is omitted.
73 I Beginning date of the file (Julian)
74 R Beginning time of the file (hours)
75 I Ending date of the file (Julian)
76 R Ending time of the file (hours)
Words 73-76 describe the total time span contained on the
file.
2) The Region Description Header Record contains 15 words.
The first three words define the reference origin:
1 R x-coordinate (UTM units)
2 R y-coordinate (UTM units)
3 I UTM zone
The next two words define the location of the modeling region
with respect to the reference origin:
208
-------
4 R x-location (meters)
5 R y-location (meters)
The next two words define the size of each grid cell in the
x- and y-directions, respectively:
6 R Grid cell size—x-direction (meters)
7 R Grid cell size—y-direction (meters)
The next three words define the size of the modeling region
in grid cells:
8 I Number of grid cells—x-direction
9 I Number of grid cells—y-direction
10 I Number of grid cells—z-direction
The last five words describe the vertical distribution of
grid cells. For the TEMPERATUR File, which does not vary
vertically, these words are ignored:
11 I Number of cells between surface layer and
diffusion break
12 I Number of cells between diffusion break and
top of region
13 R Height of surface layer (meters)
14 R Minimum height of cells between surface layer
and diffusion break (meters)
15 R Minimum height of cells between diffusion break
and top of region (meters)
3) The Segment Description Header Record contains one group of
four words for each segment; the number of segments appears
in the File Description Header Record:
1 I x-location of segment origin with respect to
oriqin of modeling region (grid units)
209
-------
2 I y-location of segment origin with respect to
origin of modeling region (grid units)
3 I Number of grid cells in segment—x-directi on
4 I Number of grid cells in segment—y-directi on
4) The Species Description Header Record is omitted for the
TEMPERATUR File, which does not vary with species.
b. Time-Invariant Data
The TEMPERATUR File contains no time-invariant data.
c. Time-Varying Data
The TEMPERATUR File contains one set of the following records for
each time interval.
1) The Time Interval Record contains four words:
1 I Beginning date (Julian)
2 R Beginning time (hours)
3 I Ending date (Julian)
4 R Ending time (hours)
2) For each segment of the region the TEMPERATUR File contains
one Temperature Record. The first eleven words of the record
identify segment and variable:
1 I Segment number
2-11 A Variable name « 'TEMPERATUR1, ten characters,
one character per word.
The next group of words is the temperature array itself:
12+ R Temperature (deg K) at ground level at each
cell in the row for each row in the segment
210
-------
Table IX-14. TEMPERATUR File Structure
a. Header Records
1) File Description Header Record
File name
File identifier
Number of segments
Number of species
Time span
2) Region Description Header Record
Reference origin
Grid origin
Grid cell size
Region size
Vertical distribution
3) Segment Description Header Record
For each segment:
Segment origin
Segment size
4) Species Description Header Record—omitted
b. Time Invariant Data—none
c. Time Varying Data—one set per time interval
1) Time Interval Record
Beginning date and time
Ending date and time
2) For each segment:
Temperature Record
Segment number
Variable name
Temperature (x,y)
211
-------
14. Terrain
The Terrain (TERRAIN) File contains surface roughness values and
vegetation factors over the region (x,y).
a. Header Records
The TERRAIN File begins with three standard header records.
1) The File Description Header Record contains 76 words:
1-10 A File name * 'TERRAIN ', ten characters, one
character per word
11-70 A File identifier, 60 characters, one char-
acter per word
71 I Number of segments
For unsegmented files, this number = 1.
72 I Number of chemical species
For the TERRAIN File, which does not vary with
species, this number = 0 and the Species
Definition Header Record is omitted.
73 I Beginning date of the file (Julian)
74 R Beginning time of the file (hours)
75 I Ending date of the file (Julian)
76 R Ending time of the file (hours)
Words 73-76 describe the total time span contained on the
file. For the TERRAIN File, which does not vary in time,
these numbers are ignored.
2) The Region Description Header Record contains 15 words.
The first three words define the reference origin:
1 R x-coordinate (UTM units)
2 R y-coordinate (UTM units)
3 I UTM zone
212
-------
The next two words define the location of the modeling region
with respect to the reference origin:
4 R x-location (meters)
5 R y-location (meters)
The next two words define the size of each grid cell in the
x- and y-directions, respectively:
6 R Grid cell size--x-direction (meters)
7 R Grid cell size--y-direction (meters)
The next three words define the size of the modeling region
in grid cells:
8 I Number of grid cells—x-direction
9 I Number of grid cells—y-direction
10 I Number of grid cells--z-direction
The last five words describe the vertical distribution of
grid cells. For the TERRAIN File, which does not vary
vertically, these words are ignored:
11 I Number of cells between surface layer and
diffusion break
12 I Number of cells between diffusion break and
top of region
13 R Height of surface layer (meters)
14 R Minimum height of cells between surface
layer and diffusion break (meters)
15 R Minimum height of cells between diffusion
break and top of region (meters)
3) The Segment Description Header Record contains one group
of four words for each segment; the number of segments appears
in the File Description Header Record:
213
-------
1 I x-location of segment origin with respect to
origin of modeling region (grid units)
2 I y-location of segment origin with respect to
origin of modeling region (grid units)
3 I Number of grid cells in segment— x-direction
4 I Number of grid cells in segment--y-direction
4) The Species Description Header Record is omitted for the
TERRAIN File, which does not vary with species.
b. Time-Invariant Data
•
For each segment of the region the TERRAIN File contains two
records .
1) The Surface Roughness Record begins with eleven words that
identify segment and variable:
1 I Segment number
2-11 A Variable name - 'ROUGHNESS ', ten characters,
one character per word
The next group of words is the surface roughness array
itself:
R Surface roughness (meters) over the segment
2) The Vegetation Factor Record begins with eleven words that
identify segment and variable:
1 I Segment number
2-11 A Variable name = ' VEGFACTOR ', ten characters,
one character per word
The next group of words is the vegetation factor array
itself:
214
-------
12+ R Vegetation factors over the segment (fraction of
alfalfa absorption)
c. Time-Varying Data
The TERRAIN File contains no time-varying data.
215
-------
Table IX-15. TERRAIN File Structure
a. Header Records
1) File Description Header Record
File name
File identifier
Number of segments
Number of species
Time span
2) Region Description Header Record
Reference origin
Grid origin
Grid cell size
Region size
Vertical distribution
3) Segment Description Header Record
For each segment:
Segment origin
Segment size
4) Species Description Header Record—omitted
b. Time-Invariant Data
1) For each segment:
Surface Roughness Record
Segment number
Variable name
Surface roughness (x,y)
Vegetation Factor Record
Segment number
Variable name
Vegetation factor (x,y)
c. Time-Varying Data—none
216
-------
15. Concentrations at Top of Region
. The Concentrations at Top of Region (TOPCONC) File contains the concen-
tration values for each species at the top boundary of the region (x,y).
a. Header Records
The TOPCONC File begins, with four standard header records.
1) The File Description Header Record contains 76 words •'
1-10 A File name = 'TOPCONC ', ten characters,
one character per word.
11-70 A File identifier, 60 characters, one char-
acter per word
71 I Number of segments
For unsegmented files, this number = 1.
72 I Number of chemical species
73 I Beginning date of the file (Julian)
74 R Beginning time of the file (hours)
75 I Ending date of the file (Julian)
76 R Ending time of the file (hours)
Words 73-76 describe the total time span contained on the
file.
2) The Region Description Header Record contains 15 words.
The first three words define the reference origin:
1 R x-coordinate (UTM units)
2 R y-coordinate (UTM units)
3 I UTM zone
The next two words define the location of the modeling region
with respect to the reference origin:
217
-------
4 R x-location (meters)
5 R y-location (meters)
The next two words define the size of each grid cell in
the x- and y-directions, respectively:
6 R Grid cell size—x-direction (meters)
7 R Grid cell size—y-direction (meters)
The next three words define the size of the modeling
region in grid cells:
8 I Number of grid cells—x-direction
9 I Number of grid cells—y-direction
10 I Number of grid cells—z-direction
The last five words describe the vertical distribution
of grid cells. For the TOPCONC File, which does not vary
vertically, these words are ignored:
11 I Number of cells between surface layer
and diffusion break
12 I Number of cells between diffusion break
and top of region
13 R Height of surface layer (meters)
14 R Minimum height of cells between surface
layer and diffusion break (meters)
15 R Minimum height of cells between diffusion
break and top of region (meters)
3) The Segment Description Header Record contains one group of
four words for each segment; the number of segments appears
in the File Description Header Record:
1 I x-location of segment origin with respect
to origin of modeling region (grid units)
218
-------
2 I y-location of segment origin with respect
to origin of modeling region (grid units)
3 I Number of grid cells in segment--x-direction
4 I Number of grid cells in segment—y-direction
4) The Species Description Header Record contains ten words for
each species; the number of species is defined in the File
Description Header Record:
1-10 A Species name, ten characters, one char-
acter per word.
b. Time-Invariant Data
The TOPCONC File contains no time-invariant data.
c. Time-Varying Data
The TOPCONC File contains one set of the following records for
each time interval.
1) The Time Interval Record contains four words:
1 I Beginning date (Julian)
2 R Beginning time (hours)
3 I Ending date (Julian)
4 R Ending time (hours)
2) For each segment of the region there is a set of Top Concen-
tration Records, one record for each species. The first
eleven words of the record identify segment and species:
1 I Segment number
2-11 A Species name, ten characters, one char-
acter per word
219
-------
The next group of words is the concentration array itself:
R Concentration at top of region (ppm, or yg/m3
for AEROSOLS) at each cell in the row for each
row in the segment
220
-------
TABLE IX-16. TOPCONC File Structure
a. Header Records
1) File Description Header Record
File name
File identifier
Number of segments
Number of species
Time span
2) Region Description Header Record
Reference origin
Grid origin
Grid cell size
Region size
Vertical distribution
3) Segment Description Header Record
For each segment:
Segment origin
Segment size
4) Species Description Header Record
For each species:
Species name
b. Time-Invariant Data—none
c. Time-Varying Data—one set per time interval
1) Time Interval Record
Beginning date and time
Ending date and time
2) For each segment:
For each species:
Top Concentration Record
Segment number
Species name
Concentration at top (x,y)
221
-------
16. Concentration Variations
The Concentration Variations (VARIATION) File contains time-averaged
subgrid-scale concentration variations in the surface layer for each species,
a. Header Records
The VARIATION File begins with four standard header records.
1) The File Description Header Record contains 76 words:
1-10 A File name = 'VARIATION ', ten characters,
one character per word
11-70 A File identifier, 60 characters, one character
per word
71 I Number of segments
For unsegmented files, this number = 1
72 I Number of chemical species
73 I Beginning date of the file. (Julian)
74 R Beginning time of the file (hours)
75 I Ending date of the file (Julian)
76 R Ending time of the file (hours)
Words 73-76 describe the total time span contained on the file.
2) The Region Description Header Record contains 15 words. The
first three words define the reference origin:
1 R x-coordinate (UTM units)
2 R y-coordinate (UTM units)
3 I UTM zone
The next two words define the Icoation of the modeling region
with respect to the reference origin:
222
-------
4 R x-location (meters)
5 R y-location (meters)
The next two words define the size of each grid cell in
the x- and y-directions, respectively:
6 R Grid cell size--x-direction (meters)
7 R Grid cell size—y-direction (meters)
The next three words define the size of the modeling
region in grid cells:
8 I Number of grid cells—x-direction
9 I Number of grid cells—y-direction
10 I Number of grid cells—z-direction
The last five words describe the vertical distribution
of grid cells.
11 I Number of cells between surface layer
and diffusion break
12 I Number of cells between diffusion break
and top of region
13 R Height of surface layer (meters)
14 R Minimum height of cells between surface
layer and diffusion break (meters)
15 R Minimum height of cells between diffusion
break and top of region (meters)
3) The Segment Description Header Record contains one group of
four words for each segment; the number of segments appears
in the File Description Header Record:
1 I x-location of segment origin with respect
to origin of modeling region (grid units)
223
-------
2 I y-location of segment origin with respect
to origin of modeling region (grid units)
3 I Number of grid cells in segment--x-direction
4 I Number of grid cells in segment—y-direction
4) The Species Description Header Record contains ten words for
each species; the number of species is defined in the File
Description Header Record:
1-10 A Species name, ten characters, one char-
acter per word.
b. Time-Invariant Data
The VARIATION File contains no time-invariant data.
c. Time-Varying Data
The VARIATION File contains one set of the following records for
each time interval.
1) The Time Interval Record contains four words:
1 I Beginning date (Julian)
2 R Beginning time (hours)
3 I Ending date (Julian)
4 R Ending time (hours)
2) For each segment of the region there is a set of Concen-
tration Variation Records, one record for each species.
The first eleven words of the record identify segment and
species:
1 I Segment number
2-11 A Species name, ten characters, one char-
acter per word
224
-------
The next group of words is the concentration variation array
itself:
R Subgrid-scale concentration variation (ppm, or
pg/m3 for AEROSOLS), averaged over the time
interval, for each cell in the surface layer.
225
-------
Table IX-17. VARIATION File Structure
a. Header Records
1) File Description Header Record
File name
File identifier
Number of segments
Number cf species
Time span
2) Region Description Header Record
Reference origin
Grid origin
Grid cell size
Region size
Vertical distribution
3) Segment Description Header Record
For each segment:
Segment origin
Segment size
4) Species Description Header Record
For each species:
Species name
b. Time-Invariant Data—none
c. Time-Varying Data—one set per time interval
1) Time Interval Record
Beginning date and time
Ending date and time
2) For each segment:
For each species:
Concentration Variation Record
Segment number
Species name
Concentration variation (x,y)
226
-------
77. Wind
The Wind (WIND) File contains the x- and y-components of the
wind vector at each cell in the region (x,y,z).
a. Header Records
The WIND File begins with three standard header records.
1) The File Description Header Record contains 76 words:
•
1-10 A File name = 'WIND ', ten characters,
one character per word
11-70 A File identifier, 60 characters, one char-
acter per word.
71 I Number of segments
For unsegmented files, this number = 1.
72 I Number of chemical species
For the WIND File, which does not vary with
species, this number * 0 and the Species
Definition Header Record is omitted.
73 I Beginning date of the file (Julian)
74 R Beginning time of the file (hours)
75 I Ending date of the file (Julian)
76 R Ending time of the file (hours)
Words 73-76 describe the total time span contained on the
file.
2) The Region Description Header Record contains 15 words. The
first three words define the reference origin:
1 R x-coordinate (UTM units)
2 R y-coordinate (UTM units)
3 I UTM zone
227
-------
The next two words define the location of the modeling region
with respect to the reference origin:
4 R x-location (meters)
5 R y-location (meters)
The next two words define the size of each grid cell in
the x- and y-directions, respectively:
6 R Grid cell size—x-direction (meters)
. 7 R Grid cell size—y-direction (meters)
The next three words define the size of the modeling
region in grid cells:
8 I Number of grid cells—x-direction
9 I Number of grid cells—y-direction
10 I Number of grid cells—z-direction
The last five words describe the vertical distribution
of grid cells:
11 I Number of cells between surface layer
and diffusion break
12 I Number of cells between diffusion break
and top of region
13 R Height of surface layer (meters)
14 R Minimum height of cells between surface
layer and diffusion break (meters)
15 R Minimum height of cells between diffusion
break and top of region (meters)
3) The Segment Description Header Record contains one group of
four words for each segment; the number of segments appears
in the File Description Header Record:
228
-------
1 I x-location of segment origin with respect
to origin of modeling region (grid units)
2 I y-location of segment origin with respect
to origin of modeling region (grid units)
3 I Number of grid cells in segment—x-direction
4 I Number of grid cells in segment—y-direction
4) The Species Description Header Record is omitted for the WIND
File, which does not vary with species.
b. Time-Invariant Data
The WIND File contains no time-invariant data.
c. Time-Varying Data
The WIND File contains one set of the following records for each
time interval.
1) The Time Interval Record contains four words:
1 I Beginning date (Julian)
2 R Beginning time (hours)
3 I Ending date (Julian)
4 R Ending time (hours)
2) For each segment of the region, there is a set of records:
one for scalar values of interest, and, for each vertical
level, two records containing x- and y-components of wind,
respectively.
The Wind Scalars Record contains eight words:
1 I Segment number
2 R Reference height of wind measuring stations (meters)
3 R Maximum absolute value of wind speed in
x-direction (meters/hour)
229
-------
4 R Maximum absolute value of wind speed
1n y-direction (meters/hour)
5 R Average wind speed In x-direction
(meters/hour) at west boundary
6 R Average wind speed In x-direction
(meters/hour) at east boundary
7 R Average wind speed in y-direction
(meters/hour) at south boundary
8 R Average wind speed in y-direction
(meters/hour) at north boundary
In the x-Wind Record the first eleven words identify the
segment number and variable name:
1 I Segment number
2-11 A Variable name = 'WINDX ', ten characters,
one character per word
The next group of words is the X-wind array itself:
12+ R Wind in x-direction (meters/hour) over the
segment for one vertical level
In the Y-wind Record the first eleven words identify the
segment number and variable name:
1 I Segment number
2-11 A Variable name = 'WINDY ', ten characters,
one character per word
The next group of words is the Y-wind array itself:
12+ R Wind in y-direction (meters/hour) over the
segment for one vertical level
230
-------
Table IX-18. WIND File Structure
a. Header Records
1) File Description Header Record
File name
File identifier
Number of segments
Number of species
Time span
2) . Region Description Header Record
Reference origin
Grid origin
Grid cell size
Region size
Vertical distribution
3) Segment Description Header Record
For each segment:
Segment origin
Segment size
4) Species Description Header Record—omitted
b. Time-Invariant Data—none
c. Time-Varying Data—one set per time interval
1) Time Interval Record
Beginning date and time
Ending date and time
2) For each segment:
231
-------
Table IX-18 (Concluded)
Wind Scalars Record
Segment number
Maximum absolute wind speeds
Average wind speeds at each boundary
For each level:
X-wind Record
Segment number
Variable name
x-wind (x,y) .
Y-wind Record
Segment number
Variable name
y-wind (x,y)
232
-------
B. INTERNAL FILE STRUCTURE DEFINITIONS
The internal files are used to store data by segment; they are not
intended to be used outside the Airshed Simulation Program. With the
exception of the INTERINST file, the internal files appear in identically
formatted pairs. During the sequential processing of segments, one of each
pair will be an input file containing current values and the other will be
an output file containing new values. At the end of the segment loop, both
files will be rewound and their functions switched, so that the new values
become input for the next processing loop. A more detailed description of
the usage of these files is given in Chapter VII.
1. Boundary Table Files
The internal Boundary Table files (BON! and BON2) are binary files
containing one record for each segment. Each record is simply a core image
of the Boundary Table for a particular segment. (See Section VII.A.3 for
the structure of this table.)
The number of words in the record for a given segment is calculated
in the following way:
8(nx + ny) + nz ng n..
where n = number of cells in x-direction
^
n = number of cells in y-direction
n = number of vertical levels
n = number of chemical species
n. ' number of interface cells in the segment, i.e., two times
the length of all nonexternal edges.
2. Instantaneous Concentration Storage Files
The internal Instantaneous Concentration Storage files (CON! and CON2)
are binary files containing one record for each segment. Each record is simply
233
-------
a core image of the concentration and variation arrays in the Segment Scratch
Table for a particular segment (see Section VII.A.4 for the structure of this
table).
The number of words in the record for a given segment is calculated
in the following way:
nx r\y ns (nz + v),
where nx = number of cells in x-direction,
n = number of cells in y-direction,
r\z = number of vertical levels,
ng = number of species,
!1, if concentration variability is calculated;
0, otherwise.
3. Cumulative Concentration Storage Files
The internal Cumulative Concentration Storage files (CON! and CON2) are
binary files containing one record for each segment. Each record is simply
a core image of the cumulative concentration and variation arrays in the Seg-
ment Scratch Table for a particular segment (see Section VII.A.4 for the struc-
ture of this table).
The number of words in the record for a given segment is calculated in
the following way:
nx ny ns (nz + v),
where nx = number of cells in x-direction,
ny = number of cells in y-direction,
nz = number of vertical levels,
ns = number of species,
if concentration variability is calculated;
otherwise.
I1'
(0.
234
-------
4. Interim Instantaneous Concentrations File
The Interim Instantaneous Concentrations file (INTERINST) is a binary
file containing, for each segment, a history of instantaneous concentrations
calculated during a single time slice. At the end of the time slice, the
file is read, the concentrations are weighted and added into the cumulative
concentration arrays, and the file is rewound for the start of the next time
slice. A more detailed description of the usage of this file is given in
Section VII.D. The INTERINST file contains the following concentration his-
tory for each segment: There is one set of records written at the beginning
of the time slice, followed by one set of records written at the end of each
time step. Each set of records consists of one Time Record, followed by a
Concentration Record and a Variation Record (if required) for each species.
a. The Time Record contains three words:
1 I Segment number
2 I Date (Julian)
3 R Time (hours)
b. The Concentration Record contains the concentration array itself:
1+ R Concentrations (ppm, or yg/m3 for AEROSOLS) at the given
time for each cell in the segment, varying by x-, then y-,
then z-directions.
c. The Variation Record will follow the Concentration Record for each
species if concentration variation is calculated. It contains the
concentration variation array itself:
1+ R Concentration variation (ppm, or yg/m3 for AEROSOLS)
at the given time for each cell in the segment, varying
by x-, then y-directions.
235
-------
5. Segment Scratch Storage Files
The internal Segment Scratch Storage files (SEG1 and SEG2) are binary
files containing one record for each segment. Each record is simply a core
image of the data arrays (excluding concentrations) in the Segment Scratch
Table (see Section VII.A.4 for the structure of this table).
The number of words in the record for a given segment is calculated
as follows:
nx n [8 + 2terr + temp + road + 2nz + ng(2 + road)] + n (2 + n )
where n = number of cells in x-direction,
^
n = number of cells in y-direction,
n = number of vertical levels,
n = number of chemical species,
if TERRAIN file input;
terr
•£:
otherwise,
1, if TEMPERATUR file input;
0, otherwise,
\
(1, if ROADWAY file input;
road =(0, otherwise,
n = number of point sources.
236
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
, REPORT NO.
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
SAI AIRSHED MODEL OPERATIONS MANUALS
Volume II -- Systems Manual
5. REPORT DATE
6. PERFORMING ORGANIZATION CODE
. AUTHOR(S)
J. Ames, S.R. Hayes, T.C. Myers, D.D. Whitney
8. PERFORMING ORGANIZATION REPORT NO.
EM78-79R2
I. PERFORMING ORGANIZATION NAME ANO ADDRESS
Systems Applications, Inc.
101 Lucas Valley Road
San Rafael, CA 94903
10. PROGRAM ELEMENT NO.
CDWA1A/01-5006 (FY-85)
11. CONTRACT/GRANT NO.
Contract No. 68-02-2429
12. SPONSORING AGENCY NAME ANO ADDRESS
Atmospheric Sciences Research Laboratory -- RTP, NC
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
(8/76 - 9/79)
14. SPONSORING AGENCY CODE
EPA/600/09
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This report describes the Systems Applications, Inc. (SAI) Airshed Model
System from a programmer's point of view. Included are discussions of all sub-
routines and how they fit together, run-time core allocation techniques, internal
methods of segment handling using secondary storage, and detailed definitions of
the structure of each file in the system. This manual also includes suggestions
for implementing the programs on different computers.
The SAI Airshed Model System consists of 17 programs that communicate with
each other through 17 data files; both the programs and files are described in
detail in this manual. Chapter II presents an overview of the system, briefly
describing the programs used to perform each of the five system functions.
Chapter III contains guidelines for implementing the programs, and Chapters IV
through VIII contain detailed descriptions of the programs by function, including
hierarchical subroutine tables and short module descriptions. Chapter IX contains
detailed file structure definitions for each file in the system.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport/
UNCLASSIFIED
21. NO. OF PAGES
20. SECURITY CLASS (Tin's page I
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (R«v. 4-77) PREVIOUS EDITION is OBSOLETE
------- |