United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park. NC 27711
EPA-454/B-94-020
(Revises EPA-600/3-88-043)
October 1994
Air
ERA METEOROLOGICAL PROCESSOR
FOR REGULATORY MODELS
(MPRM) USER'S GUIDE
(REVISED)
-------
EPA-454/B-94-020
(Revises EPA-600/3-88-043)
METEOROLOGICAL PROCESSOR FOR
REGULATORY MODELS
(MPRM)
USER'S GUIDE
(REVISED)
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Technical Support Division
Research Triangle Park, NC 27711
October 1994
-------
EPA-454/B-94-020
(Revises EPA-600/3-88-043)
DISCLAIMER
The information in this guide has been reviewed by the
U.S. Environmental Protection Agency (EPA), and approved for
publication as an EPA document. Mention of trade names or
commercial products does not constitute endorsement or
recommendation for use.
The following trademarks appear in this guide:
IBM is a registered trademark of International Business
Machines Corp.
Microsoft is a registered trademark of Microsoft Corp.
UNIVAC is a registered trademark of Unisys Corp.
VAX/VMS is a registered trademark of Digital Equipment Corp.
11
-------
EPA-454/B-94-020
(Revises EPA-600/3-88-043)
PREFACE
This document revises and replaces the previous MPRM
User's Guide (MPRM-1.2) dated January 1990. The revisions
reflect necessary changes associated with the May 20, 1993
update of MPRM (version dated 93140) on the EPA Technology
Transfer Network (TTN), Support Center for Regulatory Air
Models (SCRAM) electronic bulletin board. These include the
addition of trace capability for missing data, direct
processing of SCRAM formatted surface meteorological data, and
output of ASCII meteorological files for use in ISCST2. Most
of the remaining changes to the document, for example its
conversion to WordPerfect 5.1, are cosmetic.
The modular design of MPRM facilitates the ready
adaptation of the processor to changes in technology-
Consequently, it is anticipated that MPRM will be updated as
necessary to accommodate new input/output formats for
meteorological data, new dispersion models, and new processing
techniques.
This guide was prepared using WordPerfect 5.1 word
processing software and, as such, is available in both hard and
soft (computer compatible) copy. Hard copies of the user's
guide are available from the National Technical Information
Services (NTIS), Springfield, VA 22161; phone (703) 487-4650.
Soft copies of the user's guide in WordPerfect 5.1 format may
be obtained from the TTN SCRAM bulletin board; phone (919) 541-
5742.
111 October 1994
-------
Although all efforts are made to ensure an error-free
code, errors inevitably do occur. If you find errors in the
code or documentation, please report them to:
Desmond T. Bailey (MD-14)
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Phone: (919) 541-5248
Fax: (919) 541^2357
Internet: TTZ@EPAVAX.RTPNC.EPA.GOV
iv October 1994
-------
ABSTRACT
The Meteorological Processor for Regulatory Models (MPRM)
is a general purpose program used to process meteorological
data for use in EPA recommended air quality dispersion models.
MPRM includes new capabilities not available in its
predecessor, the RAMMET meteorological processor. These new
capabilities include quality assurance procedures, detailed
report generation, and the ability to process both on-site
(user collected) and off-site [National Weather Service (NWS)]
meteorological data.
The MPRM processor consists of three stages. Stage 1
(extraction and quality assurance) retrieves meteorological
data provided by the user on magnetic tape or disk and conducts
the quality assurance of these data. The stage 1 report files
provide listings of missing, suspect, and invalid data. These
reports provide necessary information allowing users to correct
problem data prior to its use in modeling. Stage 2 merges the
quality assured and corrected meteorological data obtained from
the various MPRM pathways, upper air, on-site, and surface
(NWS). The third and final stage performs the necessary
processing to create a meteorological data file for use in a
dispersion model selected by the user.
MPRM supports the following dispersion models listed in
the Guideline on Air Quality Models Revised (EPA, 1986): 1)
Those requiring unformatted data: BLP, RAM, ISCST2, MPTER,
CRSTER, and COMPLEX1. 2) Those requiring ASCII (formatted)
data: ISCST2. 3) Those requiring STAR formatted data: CDM
(with either 16 or 36 wind direction sectors), ISCLT2, and
VALLEY (long term). 4) Those requiring special formats:
CALINE-3 and RTDM (default).
v October 1994
-------
CONTENTS
Disclaimer
Preface
Abstract v
Figures x
Tables xiv
Revision History xv
Acknowledgments xvi
1. Introduction 1-1
Why is this processor needed? 1-1
What does MPRM do? 11
Stage 1 (Extraction and Quality Assurance) . . . 1-2
Stage 2 (Merging the Data) 1-3
Stage 3 (Creating a Model Input File) 1-4
Relationship of MPRM to EPA Air Pollution Modeling
Guidance 1-5
Missing Data 1-5
Document overview 1-7
2. Getting Started on a Personal Computer 2-1
Installation of MPRM 2-2
3. Tutorial 3-1
Introduction 3-1
MPRM input syntax 3-3
Processing on a PC 3-5
Overview 3-6
Stage 1 (Extract and Quality Assurance) 3-6
Description of example input for Stage 1 .... 3-6
Summary of Stage 1 output 3-11
Stage 2 (Merging the Data) 3-12
Process data for a RAMMET type dispersion model
Stage 3 3-14
V1 October 1994
-------
CONTENTS (Continued)
Stage 3 output options 3-18
A closer look at the general report file content 3-19
4. Processing On-site Data and Other Special Options . . 4-1
Introduction 4-1
Stage 1 processing 4-1
Extract UA data (Step 1) 4-2
Quality assess UA data (Step 2) 4-6
Extract SF data (Step 3) 4-13
Quality assess SF data (Step 4) 4-15
Extract and quality assess OS data (Step 5) . . . 4-19
Stage 2 processing 4-33
Stage 3 processing 4-36
Selecting dispersion models 4-38
Selecting processing methodologies 4-38
Selecting reporting procedures 4-39
5. Scientific Notes 5-1
Introduction 5-1
Stage 1 5-1
Averaging subhourly values 5-1
Quality assessment 5-2
Upper air data modification 5-5
Stage 3 5-8
Wind 5-8
Temperature 5-10
Stability category 5-11
Mixing height 5-14
Roughness length 5-14
Output formats 5-15
6. Program Notes 6-1
Introduction 6-1
FORTRAN compatibility and processor design .... 6-1
System specific source code 6-2
Creating executable code from the source code ... 6-5
vii October 1994
-------
CONTENTS (Continued)
On a personal computer 6-5
On a VAX 6-7
On an IBM 6-8
Checklist 6-9
Processor notes 6-10
Standard input and output devices 6-10
End-of-file conditions 6-10
File header records 6-10
Upper air sounding format 6-11
Mounting magnetic tapes on the VAX 6-11
Running MPRM on an IBM mainframe 6-12
Scan reports 6-17
MPRM run time and file statistics 6-17
7. Interpreting Processing Errors 7-1
Appendices
A. Summary of input keywords A-l
Introduction A-l
How to use these tables A-2
B. Summary of input syntax B-l
Introduction B-l
How to use these tables B-l
C. Variable names and default range checks C-l
D. Examples of reports D-l
E. Summary of error and warning messages E-l
Input image processing, 00-19 E-3
File header and library processing, 20 - 29 .... E-5
Upper air processing, 30-39 E-5
Surface observations processing, 40-49 E-7
On-site observations processing, 50-59 E-8
Merge processing, 60-69 E-9
Stage 3 processing, 70-79 E-10
October 1994
-------
CONTENTS (Continued)
F. Format of data files F-l
Upper air F-l
Surface observations F-3
RAMMET meteorological data F-7
ISCST2 ASCII meteorological data F-9
CALINE-3 meteorological data F-9
RTDM meteorological data F-10
16-sector Joint Frequency Function F-ll
36-sector Joint Frequency Function F-12
Glossary GLOSSARY-1
References REFERENCE-1
Index INDEX-1
ix October 1994
-------
FIGURES
Number Page
3-1 Overview of processing stages within MPRM 3-2
3-2 Run stream for extraction and quality assessment
of mixing height data for a 6-day period and
surface observations for a 4-day period 3-7
3-3 Overview of MPRM output 3-11
3-4 Run stream for combining (merging) data files of
mixing height data for a 6-day period and surface
observations for a 4-day period 3-14
3-5 Run stream for developing a meteorological data
file in RAMMET format from a merged data file of
mixing height data for a 6-day period and surface
observations for a 4-day period 3-15
3-6 Excerpts from the first page of general report,
for example where a RAMMET type output file is
generated 3-20
3-7 Third page of general report, for example where a
RAMMET type output file is generated 3-21
3-8 Fourth page of general report, for example where a
RAMMET type output file is generated 3-22
3-9 Fifth page of general report, for example where a
RAMMET type output file is generated 3-23
3-10 Sixth page of general report, for example where a
RAMMET type output file is generated 3-24
3-11 Seventh page of general report, for example where
a RAMMET type output file is generated 3-26
4-1 Second page of general report, for example where
upper air soundings and mixing height data are to
be extracted for a 6-day period 4-4
4-2 Error/message file generated during a 6-day
extraction period for upper air soundings and
*
mixing height data 4-5
x October 1994
-------
FIGURES (Continued)
Number Page
4-3 First page of general report, for example where
UA-pathway data are submitted for quality
assessment 4-9
4-4 Second page of general report, for example where
UA-pathway data are submitted for quality
assessment 4-10
4-5 Last page of general report, for example where
UA-pathway data are submitted for quality
assessment 4-11
4-6 Partial listing of error/message file generated
during a 6-day quality assessment period for
upper air soundings and mixing height data .... 4-12
4-7 Error/message file, for example where data are
extracted for use on the SF-pathway 4-14
4-8 Second page of general report, for example where
data are extracted for use on the SF-pathway ... 4-15
4-9 Last page of general report, for example where SF
data are checked for integrity 4-18
4-10 Example of on-site data 4-23
4-11 First page of general report, for example where
on-site data are extracted and submitted for
quality assessment 4-29
4-12 Second page of general report, for example where
on-site data are extracted and submitted for
quality assessment 4-30
4-13 Final page of general report, for example where
on-site data are extracted and submitted for
quality assessment 4-31
4-14 Partial listing of error/message file generated
during the processing of the on-site data 4-32
xi October 1994
-------
FIGURES (Continued)
Number Page
4-15 Second page of general report, for example where
data from all three data pathways are combined
(merged) 4-35
4-16 Partial listing of error/message file, for example
with trace option enabled 4-40
6-1 Flow diagram of MPRM Stage 1 and Stage 2 processing 6-2
6-2 Overview of processing stages within MPRM 6-7
7-1 First page of general report 7-2
7-2 Error/message file 7-3
7-3 Second page of general report 7-5
D-la First page of general report for 4-day example of
Stage 1 processing used in Tutorial D-2
D-lb Second page of general report for 4-day example of
Stage 1 processing used in Tutorial D-3
D-lc Third page of general report for 4-day example of
Stage 1 processing used in Tutorial D-4
D-2a First page of general report for 4-day example of
Stage 2 processing used in Tutorial D-5
D-2b Second page of general report for 4-day example of
Stage 2 processing used in Tutorial D-6
D-2c Third page of general report for 4-day example of
Stage 2 processing used in Tutorial D-7
D-3a Portion of listing of meteorological data
generated for the CDM16 dispersion model from
Stage 3 processing (4-day example for Tutorial) . . D-8
D-3b Portion of listing of meteorological data generated
for the CDM36 dispersion model from Stage 3
processing (4-day example for Tutorial) D-9
D-3c Portion of listing of meteorological data generated
for the RTDM dispersion model from Stage 3
processing (4-day example for Tutorial) D-10
October 1994
-------
FIGURES (Continued)
Number Page
D-3d Portion of listing of meteorological data generated
for the CALINE-3 dispersion model from Stage 3
processing (4-day example for Tutorial) D-ll
D-4 Listing of an error/message file for Stage 1
processing with a report of tape contents for NWS
upper air data D-12
XI11 October 1994
-------
TABLES
Number Page
4-1 Summary of MP VBL input keywords ITEM and ACTION . 4-39
4-2 Subroutines employed for various processing
options 4-41
6-1 Data set attributes on the IBM for the four day
test case 6-14
6-2a VAX run time and file statistics 6-18
6-2b Personal computer run time and file statistics . . 6-19
C-l Variable names, units, and quality assessment
default settings for the UA-pathway C-3
C-2 Variable names, units, and quality assessment
default settings for the SF-pathway C-4
C-3 Variable names, units, and quality assessment
default settings for the OS-pathway C-5
X1V October 1994
-------
REVISION HISTORY
Version 1.1 of MPRM (dated 89142) was released and made
available on the SCRAM Bulletin Board in May 1989.
Version 1.2 (dated 90045) was released in February 1990-
This release corrects a problem in the STAGES processing of
calm conditions. Details are provided in MPRM Model Change
Bulletin #2 (MCB#2) on the SCRAM Bulletin Board.
The current version (dated 93140) was released in May
1993. This release corrected 'bugs' in all three stages of
MPRM processing and incorporated revisions to enhance user-
friendliness. Details are provided in MPRM MCB#3 on the SCRAM
Bulletin Board.
October 1994
-------
ACKNOWLEDGMENTS
John S. Irwin and James 0. Paumier coauthored the first
release of the user's guide in July 1988. Other contributors
have included: Roger Erode, who was coauthor of version 1.1;
Carol Brown, who expended a large amount of effort in editing
and assembling the document; Dr. P.M. Barlow, who aided in the
planning of the document; Phil Boone, who helped in creating
code for MPRM; and Russell F. Lee, who aided in the overall
design of MPRM. Desmond T. Bailey served as editor for the
latest update to the guide and as project officer on the
conversion of the document to WordPerfect.
XV1 October 1994
-------
SECTION 1
INTRODUCTION
WHY IS THIS PROCESSOR NEEDED?
EPA approved software for processing of user collected on-
site meteorological data did not exist prior to the release of
MPRM. The only approved software available at that time,
RAMMET (Turner and Bender, 1986), was designed for processing
of National Weather Service (NWS) meteorological data only.
MRPM was created as a general purpose meteorological
processor with capability for processing on-site meteorological
data as well as NWS surface and upper air data. Output files
provided by MRPM accommodate dispersion models recommended for
use in the Guideline on Air Quality Models (Revised) (EPA,
1986).
WHAT DOES MPRM DO?
MPRM can be envisioned as a three stage processing system.
In Stage 1, the processor extracts surface and upper air data
and processed mixing height data (obtained from NWS data bases)
and/or surface data (obtained from on-site meteorological
monitoring data bases). The extracted data are then processed
through a series of quality assurance (QA) checks and reports
of missing and suspect values are generated. In Stage 2, the
processor combines the data from each of the three pathways:
the upper air (UA) pathway, the NWS surface (SF) pathway, and
the on-site (OS) pathway. These data are merged and written to
an unformatted (binary) file for subsequent use by Stage 3. In
Stage 3, the processor reads the merged data and develops a
11 October 1994
-------
meteorological data file for the dispersion model selected by
the user.
Stage 1 [Extraction and Quality Assurance (QA) 1
The goal of Stage 1 processing is to:
o Extract NWS and on-site meteorological data as specified
by the user
o Quality assure these data, report missing data, and flag
suspect data
o Output the data in a format suitable for editing with a
standard text editor
NWS meteorological data are available from the National
Climatic Data Center (NCDC) on a variety of storage media
(magnetic tape, floppy diskette, and CD ROM) and in two codes:
American Standard Code for Information Exchange (ASCII) or
Extended Binary Coded Decimal Interchange Code (EBCDIC). MPRM
currently can process surface data provided by the NCDC in the
older CD-144 format (see Appendix F); the newer NCDC format
for surface data (the TD-3280 format) is not currently
implemented in MPRM. NWS surface data that have been archived
on the EPA SCRAM Bulletin Board can also be processed by MPRM.
These data are identical to the NCDC CD-144 data, but are
stored in a compressed format. MPRM can also process upper air
data from NCDC in the TD-5600 format and processed mixing
height estimates from NCDC in the TD-9689 format. The syntax
for specifying the above types of data for use in MPRM is
described in Appendix B.
MPRM, since it was designed to process a 'generic1 on-site
meteorological data file, places few restrictions on the format
1-2 October 1994
-------
of the on-site data file. The generic capability is provided
by specifying the format of the on-site data file in the input
runstream. The only restrictions to consider in processing on-
site data are 1) the observations (records) must be ordered
sequentially with no missing records, 2) each record must
contain data for the same variables in the same order, and 3)
the data must be in a form such that it can be read using a
FORTRAN Format Statement. This capability is discussed in more
detail in Section 4.
The most important function of Stage 1 processing is the
QA function. QA is accomplished by comparing the data values
for each variable to user specified upper and lower bounds;
data values that fail this range check are flagged as suspect.
Missing data, identified by a user specified code, are also
flagged during the Stage 1 processing. All missing and suspect
data are identified in an error report file which is generated
by the Stage 1 processor. These data should be filled or
otherwise corrected before proceeding to Stage 2 processing. A
discussion of the QA function is presented in Section 5.
The output files from Stage 1 can be edited using standard
text editors routinely available on computer systems.
Stage 2 (Merging the Data)
The goal of Stage 2 processing is to merge the data from
the three pathways (UA, SF, and OS) and to output the data for
subsequent use in Stage 3.
For the specification of the dispersive state of the
atmosphere, it is most convenient to consider the physics of
the atmosphere on a daily basis, i.e., a 24-hour period.
Estimation of the depth of convective mixing is in reality the
summation of effects starting with the heating of the surface
1-3 October 1994
-------
shortly after sunrise. Thus, the merging of the available data
for each 24-hour period is the next logical step in the
processing before developing the characterization of the input
meteorological data files for the dispersion models.
The merged data are stored in unformatted data files
because this provides a more efficient use of storage than the
formatted ASCII data files used during the first stage of
processing. The ASCII files are convenient for text editors
but are no longer needed once the quality assessment and
editing have been completed. The merged data can be used to
develop data sets for a variety of dispersion models and should
be retained for possible future use.
Stage 3 (Creating a Model Input File)
The goal of Stage 3 processing is to create a
meteorological data file for use with a regulatory dispersion
model chosen by the user.
MPRM can generate any one of several output formats to
meet the input requirements of the regulatory dispersion model
chosen by the user. The RAMMET format has been selected as the
default output. In addition, RAMMET procedures have been
selected as the default methods for processing wind,
temperature, stability category and mixing heights. A detailed
discussion of Stage 3 processing is presented in Section 5.
Additional output options are discussed in Sections 3 and 4.
It is expected that future advances in dispersion modeling
will have the most impact on Stage 3 processing. For example,
acceptance of new algorithms for mixing height estimation, or
new methods for characterizing the variation of wind speed and
wind direction with height, will probably require modifications
1-4 October 1994
-------
and/or additions to the Stage 3 software. Additionally,
acceptance of a new dispersion model will probably require
changes to the output subroutine within MPRM in order to
accomodate the data format required by the new model. MPRM,
because of its modularized design, is especially suited for
such updates.
RELATIONSHIP OF MPRM TO EPA AIR POLLUTION MODELING GUIDANCE
The data processing methods incorporated in MPRM implement
the recommendations of the "On-Site Meteorological Program
Guidance for Regulatory Modeling Applications" (EPA, 1987),
hereafter, referred to as the "On-Site Guidance". These
recommendations include, methods for estimating Pasquill-
Gifford (P-G) stability categories from on-site measurements,
methods for processing wind data to obtain the sealer averaged
wind direction, and methods for estimating surface roughness.
As data processing recommendations are modified, MPRM will be
revised to reflect the latest guidance. In cases where
discrepancies exist between MPRM and current guidance, the
guidance should be regarded as the authority.
The user's guide is not intended to provide comprehensive
guidance on all aspects of regulatory dispersion modeling.
Users should refer to the appropriate model user's guide for
specific guidance on dispersion modeling, to the On-Site
Guidance document for guidance on meteorological monitoring,
and to the "Quality Assurance Handbook for Air Pollution
Measurement Systems, Volume IV Meteorological Measurements" for
guidance on quality assurance of meteorological measurements.
Missing Data
Short-term dispersion models require hourly meteorological
data and in most cases will not accept missing data; i.e.,
1~5 October 1994
-------
there must be a valid record for every hour in the analysis
period. Before proceeding with stage 3 processing, users
should check to see that the meteorological file is complete
(does not contain missing data) and if not, that the model they
intend to use allows missing data. If the file is not
complete, and the model does not allow missing data then
procedures to complete the data base will need to be
implemented. Such procedures are often case specific and may
require prior approval by the permit granting authority. Users
are referred to the section on "Completeness Requirements" in
On-Site Guidance Document for more detailed guidance on
handling missing data.
DOCUMENT OVERVIEW
Section 1
Explains why this meteorological processor was
designed and summarizes its overall structure.
Section 2
Describes how to install the processor on an IBM
compatible PC.
Section 3
Section 4
Presents a tutorial of the processor.
Describes the use of several optional processing
commands and the commands for processing on-site
data.
Section 5
Summarizes the methods used for the quality
assessment checks and the processing of
meteorological data for developing estimates of
mixing height and stability category.
Section 6
Summarizes the program structure and
installation requirements for use of the
1-6
October 1994
-------
Section 7
Appendix A
Appendix B
Appendix C
Appendix D
Appendix E -
Appendix F -
Glossary
References
Index
processor on IBM and VAX mainframe computers and
personal computers.
Describes how to interpret the error messages.
Presents a summary of the keywords used in
defining the input to the processor and denotes
those that are mandatory and those that are
optional.
Describes usage, limitations, and syntax of each
keyword.
For each meteorological variable, lists naming
conventions used, and default upper and lower
range check bounds used in the quality
assessment checks.
Provides examples of various reports that can be
generated by the processor.
Describes various types of messages that are
generated by the processor.
Describes the computer file formats used for the
storage of the extracted data.
Lists and defines unfamiliar technical
terminology used within the guide.
Lists material referenced in the guide.
Lists pertinent statements within the guide.
1-7
October 1994
-------
SECTION 2
GETTING STARTED ON A PERSONAL COMPUTER
This section describes what you need to do in order to run
MPRM on an IBM compatible PC. The discussion assumes that you
have obtained the necessary files from the SCRAM Bulletin Board
and have installed them on your hard drive. The following MPRM
files are available on SCRAM:
SCRAM BBS Contents
FILE
MPRM1.ZIP STAGE1N2.EXE, STAGE3.EXE, IRND.DAT, README.TXT
MPRM2.ZIP Source code
MPRM3.ZIP Tutorial test cases see sections 3 and 4 of
user's guide
MPRM4.ZIP SCRAM BBS data - test cases
MPRM5.ZIP Runstream generator
These files are in a compressed 'ZIP1 format. The
software to expand (unzip) these files 'PKUNZIP.EXE' is
available from the Systems Utilities area of the OAQPS TTN BBS
To operate the processor you will need to download
MPRM1.ZIP, MPRM3.ZIP, and MPRM4.ZIP- MPRM1.ZIP contains the
stage 1 and 2 executibles (STAGE1N2.EXE), the stage 3
executable (STAGE3.EXE), and a file of random numbers
(IRND.DAT). MPRM3.ZIP contains the tutorial test cases
described in sections 3 and 4 of the user's guide. Use these
files to become familiar with the several processing steps in
MPRM, e.g., extraction, quality assurance, merge, and creation
of meteorological files for use in regulatory dispersion
2-1
October 1994
-------
modeling. MPRM4.ZIP contains additional test case files
associated with processing of SCRAM BBS data.
First-time users should also download the runstream
generator (MPRM5.ZIP) which introduces new users to the
development of 'runstream1 control data. Use the runstream
generator initially, to become familiar with what is needed to
construct proper input runstreams. Once you are familiar with
what is needed, you can then begin creating your own runstreams
using a text editor.
DISK2.ZIP contains the FORTRAN source code for MPRM (dated
93140). This file is not needed unless you plan to install
MPRM on a main-frame computer. See section 6 of the user's
guide for a discussion of the installation procedures for VAX
and IBM mainframes.
INSTALLATION OF MPRM
The following steps are used to install the MPRM files on
a hard disk drive designated as drive C:
1. If you have not done so already, create subdirectories
UTILITY and MPRM on your hard drive. If you already have
a subdirectory for general purpose utilities, use that
directory in place of UTILITY.
2. Copy the utility 'PKUNZIP.EXE1 to the utility
subdirectory- Copy necessary MPRM 'ZIP1 files to the MPRM
subdirectory.
3. As necessary, revise the default DOS path to include
C:\utility- Change to subdirectory MPRM and enter the
command: PKUNZIP . This will provide the syntax
for the PKUNZIP utility. For example to expand (unzip)
2-2 October 1994
-------
the MPRM1.ZIP file use the following: PKUNZIP -X MPRM1
. Expand other files as necessary.
October 1994
-------
SECTION 3
TUTORIAL
INTRODUCTION
This section provides a demonstration of the three
processing stages outlined in Section 1. Examples to
illustrate the input and the output associated with each of the
three stages are presented.
For discussion purposes, the activities of extracting data
and processing these data through the QA function, are labeled
as Stage 1. Data extraction is a one-time activity, only
repeated when additional data are required; whereas QA may be
performed several times on the same file. The QA process flags
suspect data which must then be reviewed and corrected as
necessary; a text editor is required for this activity.
Corrections to suspect data should be entered in the extracted
data file only. The original data files should never be
altered, as they represent the original archive as delivered.
Whenever corrections are made to the extracted data file,
the data should be reprocessed through the Stage 1 QA. This
provides a final listing of data, and may call attention to new
problems that need to be fixed. When as many of the suspect
and missing values as possible have been eliminated, Stage 2
processing (merging) can begin.
At the completion of Stage 2 processing, the data have
been combined (merged) into one data file. Stage 3 processing
can then be used to construct the data files needed to
accommodate dispersion modeling.
3-1
October 1994
-------
IMPORTANT NOTE FOR PROCESSING DATA
FOR USE WITH SHORT-TERM DISPERSION MODELS
MPRM can only flag missing variables in a record -it does
not incorporate provisions for filling the missing data.
Short-term dispersion models, which require hourly
meteorological data, generally will not accept missing data.
As such, users should resolve all missing data items before
proceeding to Stage 2 processing.
The three stages of processing provided for within MPRM
are summarized in Figure 3-1. As shown in the figure, there
are two executables providing access to the three stages of
processing. The first executable, STAGE1N2.EXE, supports Stage
1 and Stage 2 processing. As discussed in Section 1, Stages 1
and 2 processing are concerned with the extraction of the data
from the raw data files, the quality assessment checks of these
data, and the combination of these data into a single data file
for subsequent processing. The second executable, STAGE3.EXE,
supports Stage 3 processing, which processes the available
meteorological data for use by a particular dispersion model.
METEOROLOGICAL PROCESSOR
FOR
REGULATORY MODELS
STAGE1N2.EXE
STAGE3.EXE
STAGE 1
PROCESSING
Extract and Quality
Assess Data
STAGE 2
PROCESSING
Combine (Merge)
Data Files
STAGE 3
PROCESSING
Create Data
File for Modeling
Figure 3-1. Overview of processing stages within MPRM.
3-2
October 1994
-------
MPRM Input Syntax
We refer to the collection of all input necessary to run
the MPRM processor as the input run stream or simply run
stream. The run stream consists of several 80-character
images, each of which begins with a 2-character group, called a
pathway identification, followed by a 3-character group, called
a keyword. The keywords define the general process to be
defined, while the pathway identifications define the overall
processing goals.
Each input run stream can be thought of as a sequence of 80-
character images. Each image consists of two or more
fields. The fields presented within the 80-character image
are considered FREE format; that is, proper interpretation
of a field is not dependent on column position. However,
the fields must be separated by commas or spaces.
The logic within the processor, and hence the input to the
processor, can be functionally divided into six major areas.
These areas are called pathways and derive their names
(pathwayID) from the functions performed. Each pathway has one
of the following two-letter acronyms:
o JB - processes that affect or pertain to the entire job
o UA processing NWS upper air data and NCDC mixing height
data
o SF - processing NWS hourly surface weather observations
o OS - processing user supplied on-site meteorological data
3-3
October 1994
-------
o MR - process of combining (merging) all available
meteorological data
o MP - processing for a particular dispersion model.
Input images in the run stream are always directed to one of
the pathways listed.
There is a certain structure to the input. Once familiar
with the "language," the input images will almost read in plain
language. A summary of the input possibilities is presented in
two appendices. Appendix A lists and defines the mandatory and
optional input for each process and Appendix B discusses in
more detail how to define the fields associated with each input
image. Together these two appendices enable the user to
construct customized run streams.
One of the first and obvious questions that arises is,
what information is needed and how is this information conveyed
to the MPRM processor? To answer this question, one first must
define what processing is to be accomplished. Basically, we
have three processing options to choose from, corresponding to
the three stages of processing:
1. To extract and/or quality assess data.
2. To combine (merge) data files that have been quality
assessed.
3. To create an input meteorological data file for a
dispersion model.
For any run stream to be complete, we must at the very
least define an error/message file, which is done using a JB
ERR input. Hence, every run stream has input directed towards
the JB pathway. Whether we have input directed to one or more
of the other pathways depends on the data to be processed and
3-4 October 1994
-------
the processing to be accomplished. Appendix A summarizes what
input is mandatory and what input is optional for each pathway
as a function of the processing to be accomplished.
Processing on a PC
On a PC, the default input is the keyboard (UNIT 5) and
the default output is the screen (UNIT 6). For MPRM, the
default input is used for reading the input run stream, and the
default output is used for writing the general report
information summarizing the results of the processing.
On a PC, the output directed to UNIT 6 can be redirected
to a file using a DOS command. Suppose the run stream for
Stage 1 processing is stored in a file entitled "STAGE1.INP",
and suppose that we wish the general report information to be
written to a file entitled "REPORT.LIS".
Shown below is an example of how one might initiate Stage
1 processing on a PC
STAGE1N2 < STAGE1.INP > REPORT.LIS
This statement directs the run stream to be read from the file
STAGE1.INP and the general report output to be written to the
file REPORT.LIS. The syntax for file names used in the above
example and in all subsequent examples follows the PC naming
conventions. In addition, it is assumed that all the files,
including the executables, reside in the current directory.
The file names will need to be changed if processing is done on
a mainframe computer. For example, for VAX processing, square
brackets are used to enclose the directory information that is
necessary for complete specification of the filename.
5 October 1994
-------
Overview
The tutorial discussion begins with a Stage 1 example
involving the extraction and processing of 4 days (96 hours) of
NWS hourly surface observations and 6 days of twice-daily
mixing heights. The data are then combined (Stage 2) into one
data file for final processing. Lastly a Stage 3 example
generates an output file identical to the RAMMET processor.
RAMMET output can be used with dispersion models BLP, RAM,
ISCST2 (short term), MPTER, CRSTER, and COMPLEX1.
If you have not already done so, MPRM must first be
installed on your system; the procedures outlined in Section 2
should be followed. The examples can then be executed
following the discussion in the tutorial.
It should be noted that the input run streams are
sufficiently long and complex to preclude direct input from the
keyboard. Instead, a text editor should be used to create the
necessary input files. These files should contain standard
ASCII characters only. Note the some text editors
differentiate between document files, which contain special
non-ASCII control characters, and non-document files, which are
composed of ASCII characters only. For such editors, be sure
to select the non-document mode of editing.
EXTRACT AND QUALITY ASSESS DATA (STAGE 1)
Description of example input for Stage 1
As only several days of data are involved and as the
purpose of this tutorial is to provide an overview of the
operation of the processor, the extraction and quality
3-6 October 1994
-------
assessment of both the surface and mixing height data will be
performed in one step. Presented in Figure 3-2 is an annotated
1
JB STA
JB ERR
JB FIN
UA STA
UA IN2
UA LOG
UA EXT
UA IQA
UA OQA
UA FIN
SF STA
SF IN2
SF LOG
SF EXT
SF IQA
SF OQA
SF FIN
Starts general definitions
(JB-pathway)
error/message file
mixing height data (UA-
pathway)
1 KFP PAM7 T HAT f A1* "^T? 1Y T ^ 1^>V 1*5^ Q^ftl1! . Hnf i nn^ mi vi no
height input data
93815 84 68U 39 07N +0 - .... Defines location of mixing
height data
63 12 31 64 01 05 - Defines time period to
extract mixing height data
DT ^If TQA7 TRRFRAT Rp-fi no? nut nut 1 nrat i nn nf
extracted mixing height
data
HT ^If no A 7 TRPFDAT . - He-f i n»Q niitnut 1 nrnt i nn n-f
mixing height data
following quality
assessment checks
(SF-pathway)
DISK RAMSFC DAT CD144FB 93S1ii Dofino'- t-urfnrp inout ei-it-\
9381/1 8/1 QOU 3° nnN +fl nn-finnr 1 ni itinn nf cnr-f-irn
data
extract surface data
D ISK IOASFBRF DAT - fin-f i nno ni itm it 1 ni it ion n-f
extracted surface data
surface data following
quality assessment checks
Figure 3-2.
Run stream for extraction and quality assessment
of mixing height data for a 6-day period and
surface observations for a 4-day period.
3-7
October 1994
-------
sample run stream for performing these activities. This run
stream is file EXAMPLE1.INP on DISKS. Note, in a real
application employing larger data files, it might be more
practical and manageable to extract and quality assess the data
for each pathway separately. Example run streams for such
processing are presented in Section 4.
The input images to the processor have a structure, which
when understood, facilitates construction of customized run
streams. The three-character keywords are, for the most part,
abbreviations of the actions that this input should invoke.
Consider the first two lines of input shown in Figure 3-2.
JB STA
2-character pathway identifications (PathwaylD)
3-character action identifications (Keyword)
JB ERR DISK ERRBRF.LIS
The STA is derived from "start". The JB STA input defines the
beginning (or start) of those inputs that have general meaning
to the processing job to be defined. The ERR input is derived
from "error". The JB ERR input defines the file name where the
error, warning, informational, and quality assessment messages
produced by the processor are written.
The third line (JB FIN), terminates the general job
information. The fourth line (UA STA), starts defining the
parameters and actions to be performed on the upper air and
mixing height data. This is referred to as the UA-pathway.
The fifth line, starting with UA IN2, defines the file storage
medium, file name, data format and station identification.
3-8 October 1994
-------
The fields following the keywords can vary somewhat
depending on the pathway. For instance, compare the usage of
the keyword IN2, as shown in the example in Figure 3-2, on the
two pathways UA and SF.
2-character PathwavID: UA = NWS upper air/mixing height
SF = NWS hourly surface
i 3-character Keyword: IN2 = Original (ran) input data
UA IN2 USER RAMZI.DAT (A5.3I2.1X,I5,13X,15) 93815
SF IN2 DISK RAMSFC.DAT CDU4FB 93814
I I
Fields following keyword
In this example, the mixing height data, stored in a file
named RAMZI.DAT, is not in the traditional format (TD-9689) as
delivered from NCDC. To alert the processor that the format is
being provided, the data field just after the keyword IN2 was
set to USER. To inform the processor of the format, the
FORTRAN FORMAT statement (A5,312,IX,15,13X,15) has been
provided. The last item in the UA IN2 input, 93815, is the
station number for the NWS observation site at the Dayton, Ohio
airport.
The sixth line, starting with UA LOG, provides the station
identification number, latitude and longitude (in decimal
degrees), and the number of hours to be subtracted to convert
the time given on the pathway to Local Standard Time (LST).
The latter adjustment depends on the location and the type of
data being extracted. For example, unprocessed upper air data
(TD-5600 data) are reported in Greenwich Mean Time (GMT);
thus, for an east coast location, the conversion to LST would
require an adjustment of 5 hours (i.e., EST = GMT - 5).
Processed mixing height data, on-the other-hand, are normally
reported in local time (actually only two times are reported,
morning and afternoon). In this case, an adjustment of +0
hours is indicated.
October 1994
-------
The seventh line beginning with, UA EXT, defines the
starting and ending times for the extraction. The sequence
year, month, and day must be observed in specifying extraction
dates. The year may be defined by all four digits or by the
last two digits only.
The eight and ninth records beginning with, UA IQA and UA
OQA, respectively, define the input and output file names for
the QA processing. The format for the OQA file is identical to
that for the IQA file. It should be noted, although currently
the IQA and OQA files are identical, that in the future this
may not be the case. The additional OQA file has been included
in MPRM to facilitate the possible incorporation of automatic
replacement procedures for missing data in the future. The two
files, IQA and OQA, provide a logical design for QA of the
data, reporting suspect or missing values, and storing the
revised data. The latter can then be examined for consistency
and reasonableness.
It is not the intent of this tutorial to fully explain
each of the items shown in Figure 3-2. Hopefully, with actual
experience and with the descriptions provided here and within
the appendices, the input images will begin to make sense. The
options available in the input are summarized in Appendix A,
where the keywords available to each pathway are defined along
with whether the input is mandatory or optional. Appendix B
provides detailed descriptions of the data fields following the
keywords and can be used to decipher those shown in Figure 3-2.
3-10 October 1994
-------
Summary of Stage 1 Output
The MPRM produces three types of output files as outlined
in Figure 3 - 3 .
Meteorological Processor
for
Regulatory Modeling
J.
Output
Data
File(s)
General
Report
File
Error/
Message
File
Figure 3 - 3. Overview of MPRM output.
Four output data files are produced, in the example shown
in Figure 3-2. Two are from processing the mixing height data,
the raw extracted data file, defined by the UA IQA input, and
the quality assessed data file, defined by the UA OQA input.
Two similar output data files are produced associated with the
processing of the hourly surface weather observations.
The general output report summarizes the processing
results and is typically only several pages long.
Within the error/message file reside all the messages
generated during processing. Appendix E summarizes the various
messages that can be generated by the processor. There are
nearly 100 different messages possible. Understandably, the
error/message file may be rather long. It is not advisable to
print this file, unless it is necessary. The user should view
the contents using a general purpose text editor. A discussion
of the structure and content of the error messages written to
the error/message file is provided in Section 7. Also within
Section 7 is provided a discussion of how the general report is
altered when errors are encountered.
3-11
October 1994
-------
output files and reports will not be discussed in
detail within this tutorial. The general report generated by
this Stage 1 example is the first example report shown in
Appendix D. Basically, the report lists the input and output
files, a table summarizing the kinds and frequency of
occurrences of the messages generated, and a table (or tables,
if more than one data pathway is processed) summarizing the
results of the quality assessment of the data. A more complete
discussion of the content of the general report appears in
Section 4.
COMBINE (MERGE) DATA FILES (STAGE 2)
At the completion of Stage 1 processing, there exist from
one to three data files. The three possible files are:
o NWS upper air observations combined with NCDC twice-daily
mixing height values (UA-pathway data)
o NWS hourly surface weather observations (SF-pathway data)
o User supplied on-site meteorological data (OS-pathway
data).
The data files resulting from Stage 1 processing are
standard ASCII data. This means that routine text editors can
be used to view the data, or to fill in values for missing time
periods of data. This is all quite convenient but it is also
quite wasteful of computer storage. A more compact and
efficient storage format is unformatted. FORTRAN programs can
process unformatted data faster than they can ASCII data. The
two goals of Stage 2 processing are to combine and organize the
data (so the user knows what data are available each day) and
to convert to unformatted storage.
3-12 October 1994
-------
Presented in Figure 3-4 is an annotated run stream
performing these activities. Comparison of the two run streams
shown in Figures 3-2 and 3-4 reveals that the only new items in
Figure 3-4 are those specific to the OS-pathway and the
MR-pathway. Basically, this run stream identifies the data
files to be merged (OQA input images), the filename for the
merged data file (MR OUT image), the location of interest for
dispersion modeling (OS LOG image), and the time period to be
merged (MR EXT image).
The two data files OQAZIBRF.DAT and OQASFBRF.DAT will be
combined (merged) into one data file named MERGEBRF.DAT for the
period from January 1, 1964 through and including January 4,
1964.
When creating a merged data set, it is required that the
OS LOC input be provided. This longitude and latitude will be
stored in the header of the MR OUT data file. It is unlikely
that the sites where the UA, SF, and OS data were collected are
collocated. It may be that the location of interest for the
dispersion modeling is not located at any one of the locations
where the UA, SF, and OS data were collected. The longitude
and latitude given in the OS LOC input, when the data are
merged, is interpreted as the location of interest in the
dispersion modeling.
The general report created by the example shown in Figure
3-4 is the second example report shown in Appendix D. This
report is quite similar to that generated by the example shown
in Figure 3-2.
3-13 October 1994
-------
2-character pathway identifications (PathwaylD)
3-character action identifications (Keyword)
JB STA
JB ERR DISK MRBRFERR.LIS
JB FIN
UA STA
UA OQA DISK OQAZIBRF.DAT
UA FIN
SF STA
SF OQA DISK OQASFBRF.DAT
SF FIN
OS STA
OS LOG 9381464 84.68U 39.07N 0
OS FIN
MR STA
MR OUT DISK MERGEBRF.DAT
MR EXT 64 01 01 64 01 04
MR FIN
Starts general definitions
(JB-pathway)
Defines file name for
error/message file
Terminates JB-pathway data
Starts upper air and
mixing height data (UA-
pathway)
Defines output location of
mixing height data
following quality
assessment checks
Terminates UA-pathway data
Starts hourly surface data
(SF-pathway)
Defines output location of
surface data following
quality assessment checks
Terminates SF-pathway data
Starts hourly on-site data
(OS-pathway)
Defines location associated
with on-site data and/or
dispersion modeling
Terminates OS-pathway data
Starts merged data (MR-
pathway)
Defines output location for
merged data
Defines time period for
combined (merged) data
Terminates MR-pathway
Figure 3-4.
Run stream for combining (merging) data files of
mixing height data for a 6-day period and hourly
surface observations for a 4-day period.
PROCESS DATA FOR A RAMMET TYPE DISPERSION MODEL (STAGE 3)
The goal of Stage 3 processing is to process the merged
data and to produce an output file for use by a particular
dispersion model. In default mode, Stage 3 processing produces
3-14
October 1994
-------
an output file in the RAMMET format. This format accommodates
several of the hourly models currently approved for routine use
in regulatory decision making. Presented in Figure 3-5 is an
annotated run stream for performing the Stage 3 processing.
2-character pathway identifications (PathwaylD)
3-character action identifications (Keyword)
JB STA
JB ERR DISK ST3AERR.LIS
JB FIN
MP STA
MP MET
MP MMP
MP EXT
MP 1ST
MP TRA
MP FIN
DISK MERGEBRF.DAT 5
DISK STAGE3A.DAT CRSTER
64 1 1 64 1 4
Starts general definitions
(JB-pathway)
file name for
error/message file
Terminates JB-pathway data
Starts MP-pathway data
Defines merged input
meteorological data file
output file for
use by dispersion model
Defines period of data to
be processed
(Optional) Turns on listing
of dispersion model
meteorological data to
general report file
(Optional) Turns on
detailed listing of
messages to error/message
file
Terminates MP-pathway data
Figure 3-5.
Run stream for developing a meteorological data
file in RAMMET format from a merged data file of
mixing height data for a 6-day period and
surface observations for a 4-day period.
As a minimum, Stage 3 processing requires the file name of
the input (merged) data file and the output file name for use
by the dispersion model. There are various methodologies for
generating the output meteorological data file. The RAMMET
format has been selected as the default. This is an
unformatted file containing a header record and one record for
each 24-hour period. The default selections for processing
wind, tempe'rature, stability category, and mixing height employ
3-15
October 1994
-------
the NWS.hourly surface weather observations and the NCDC
twice-daily mixing height values. A full discussion of the
processing assumptions is given in Section 5. In essence,
processing with the default selections duplicates that
performed by the RAMMET meteorological processor.
Two of the new input records shown in Figure 3-5 are MP
MET and MP MMP- They are used to identify the file name of
combined (merged) input meteorological data (MP MET) and the
file name for storage of the output meteorological data (MP
MMP) for use by a particular dispersion model. The number 5,
given after the file name in the MP MET input, is the number of
hours to be added to local standard time (LST) to convert LST
to Greenwich Mean Time (GMT).
Included as an option within the MP MMP input is the
ability to define the dispersion model anticipated to be
accessing the output meteorological data file. Appendix F
provides specific details regarding the various output formats.
If no dispersion model name is included in the MP MMP input
image, the default selection is CRSTER. The available choices
are:
BLP, RAM, ISCST2, Selection of these models results
MPTER, CRSTER, the output file having a RAMMET
COMPLEXl storage format. This format has
24 hours of data in each record.
CALINE-3 This selection results in an
output file format unique to
CALINE-3. This format has 1 hour
of data in each record.
ISCST2A Results in a formatted file for
use with ISCST2.
3-16 October 1994
-------
RTDM
This selection results in an
output file format unique to
RTDM. This format has 1 hour of
data in each record.
VALLEY, ISCLT, CDM16
This selection results in an
output file containing the joint
frequency function of the
meteorological conditions.
Sixteen wind sectors are used in
developing the frequency
function, with the first 22.5°
sector centered on winds from the
North.
CDM36
This selection results in an
output file containing the joint
frequency function of the
meteorological conditions.
Thirty-six wind sectors are used
in developing the frequency
function, with the first 10°
sector centered on winds from the
North.
The CDM-2.0 dispersion model can process meteorological
frequency function data (often referred to as STabiltiy ARray
data, STAR), constructed using either 16 or 36 wind direction
sectors. To provide for this flexibility, we have used CDM16
and CDM36 to identify whether 16 or 36 wind direction sectors
are desired in constructing the STAR data. STAR output for
ISCLT and VALLEY models use the six stability categories: A, B,
C, D, E, F. STAR output for CDM use the six stability
categories A, B, C, D-day, D-night, E-F.
3-17
October 1994
-------
Stage 3 Output Options
In the example shown in Figure 3-5, there are two optional
input items; 'MP LSI" and 'MP TRA'. Keyword 'LST1 activates
the listing of the generated meteorological daza to the general
report file and keyword 'TRA1 enhances the detail provided in
the Stage 3 processing messages. The extra output produced by
these two options are of little concern in this example, since
there are only 4 days of output. If a longer period of record
were involved, say a 1-year period, one might question the
wisdom of employing these options. As will be seen in the
general report for this example, the listing of the RAMMET type
meteorological data requires one page of output for every 3
days listed. This would result in 122 pages of output in
listing 1 year's output of RAMMET data. If output for RTDM or
CALINE-3 was being generated, the list option would require one
page for each 2 days listed (or 183 pages for 1 year).
The MP TRA input image currently has meaning only with the
determination of the Pasquill stability category. As is
described in Section 5, there are five different methodologies
available for specifying the stability category: one is
equivalent to the RAMMET processor and employs only NWS data,
the other four employ on-site data. Since all of the on-site
data methods are acceptable, MPRM searches for an alternate
on-site method if the primary method is incapable of working
for a given hour due to lack of data. The TRA option reports
the results of such searches whenever they occur. If these
searches were to occur frequently, the error/message file might
well become quite large. Indiscriminate use of the TRA option
is not recommended. A better use of the TRA option would be to
process short periods of record that are known to contain
frequent periods of missing data. The MP EXT input can be used
to control the dates processed. In this manner, the search
results can be reviewed in a manageable fashion. Obviously, if
3-18 October 1994
-------
an alternate method for specifying the stability category is
employed too often, it may invalidate the use of the data for
certain regulatory actions. (Refer to the discussion on
missing data in Section 1 for further details.)
A Closer Look at the General Report Content
The Stage 3 general report is structured somewhat
differently than that generated during Stages 1 and 2. With
the MP LST input, the first page of the general report (shown
in Figure 3-6) provides the header information written to the
beginning of the output file (defined within the MMP input).
Then follows the listing of the generated meteorological data
values. This listing continues on as many succeeding pages as
necessary to complete the listing (Figure 3-6). For the
example shown in Figure 3-5, the listing is completed in two
pages.
Following the listing of the generated values, the general
report lists the files accessed, and the processing selections
as determined from the run stream data (Figures 3-7 and 3-8).
It is beyond the scope of this tutorial to discuss the various
options available for specifying the wind speed, wind
direction, stability category.- and mixing height. This is
discussed in Section 4. (Table 4-1 provides a quick summary of
the processing options available.) Within item 6 of the
general report are listed the file names of the output files.
For the example shown, the generated meteorological data file,
in RAMMET format, was stored in STAGE3A.DAT.
The next two pages of the general report, shown in Figures
3-9 and 3-10, summarize the job termination status, the status
of files referenced, the dispersion model selected (CRSTER is
default), and the processing selections for generating the
output meteorological file. In item 3 of the report, NWSWXX
3"19 October 1994
-------
indicates that the meteorological variable (process) will be
determined using NCDC mixing heights and NWS hourly weather
observations. This is exactly what the RAMMET processor does.
The processor assigns a default value of 10 m as the
measurement height of wind and turbulence data (item 4 in
Figure 3-8) . Within the processor this is referred to as the
anemometer height, ANEHGT. When processing in default mode,
there are no modeling decisions that employ ANEHGT. Hence, the
METEOROLOGICAL PROCESSOR FOR REGULATORY MODELS (MPRM), VERSION 1.2
TODAY'S DATE AND TIME: 18-APR-88 AT 07:42:56
PROCESSING OF MERGED METEOROLOGICAL DATA
HEADER ON OUTPUT MP-DATA FILE:
93814
64
93815
64
YEAR= 64 MONTH= 1 DAY= 1 JULIAN DAY= 1 SUNRISE= 8.051 SUNSET= 17.353
PGSTAB= 444444444444
SPEED= 4.1 6.7 6.7 5.7 5.7 6.2 9.3 7.7 6.2 4.6 5.7 6.2
TEMP= 268. 268. 268. 268. 269. 269. 270. 270. 269. 269. 270. 270.
FLWVEC= 210. 230. 220. 220. 210. 220. 230. 230. 190. 150. 160. 180.
RANFLW= 211. 228. 224. 223. 213. 222. 235. 233. 187. 151. 164. 176.
RURAL= 511. 483. 456. 429. 401. 374. 347. 319. 292. 264. 237. 210.
URBAN= 511. 483. 456. 429. 401. 374. 347. 319. 292. 264. 237. 210.
PGSTAB= 444445544444
SPEED= 4.6 3.6 5.7 6.2 4.6 3.6 4.1 5.7 8.2 7.2 6.2 7.7
YEAR= 64 MONTH= 1 DAY= 3 JULIAN DAY= 3 SUNRISE= 8.054 SUNSET= 17.379
PGSTAB= 444444444444
SPEED= 8.8 8.8 8.8 7.7 6.7 6.2 6.2 6.2 7.2 7.7 7.2 7.7
TEMP= 278. 279. 279. 279. 279. 279. 279. 279. 279. 279. 280. 281.
FLWVEC= 40. 40. 40. 40. 30. 30. 30. 30. 40. 40. 40. 40.
RANFLW= 40. 36. 44. 40. 27. 29. 26. 29. 38. 44. 42. 38.
RURAL= 507. 497. 487. 477. 467. 457. 447. 437. 426. 416. 406. 396.
URBAN= 507. 497. 487. 477. 467. 457. 447. 437. 426. 416. 406. 396.
PGSTAB= 444444544445
SPEED= 6.7 7.2 7.2 6.7 4.6 6.2 5.1 5.7 4.1 8.8 7.2 5.1
TEMP= 281. 282. 283. 283. 282. 281. 279. 279. 278. 278. 278. 276.
FLUVEC= 50. 60. 50. 60. 60. 90. 100. 110. 90. 110. 110. 100.
RANFLW= 51. 62. 47. 60. 59. 91. 97. 108. 87. 109. 111. 97.
RURAL= 386. 376. 376. 376. 376. 398. 434. 469. 505. 540. 576. 611.
URBAN= 386. 376. 376. 376. 376. 398. 555. 469. 505. 540. 576. 1108.
Figure 3-6. Excerpts from the first page of general report,
for example where a RAMMET type output file is
generated.
3-20
October 1994
-------
value assigned to ANEHGT is of little significance. The value
assigned to ANEHGT becomes of significance for those processing
options employing use of on-site data, which will be discussed
in Section 4. Item 5 of the report summarizes the location
information determined from the header records within the
merged file. Currently the longitude and latitude values given
for the on-site data are used in processing. The only
processing decisions that require longitude and latitude are in
the determination of the position of the sun. The elevation
angle of the sun with respect to the horizon is used to
METEOROLOGICAL PROCESSOR FOR REGULATORY MODELS (MPRM), VERSION 1.2
TODAY'S DATE AND TIME: 18-APR-88 AT 07:42:58
PROCESSING OF MERGED METEOROLOGICAL DATA
ATTEMPT PROCESSING OF METEOROLOGICAL
DATA FOR USE BY A DISPERSION MODEL.
********************************
*** JOB TERMINATED NORMALLY ***
********************************
1. INPUT/OUTPUT FILENAMES AS DETERMINED
DURING SETUP PROCESSING.
GENERAL REPORT FILE
ST3AERR.LIS
MERGEBRF.DAT
STAGE3A.DAT
2. DISPERSION MODEL DEFINED DURING SETUP
*****************************************
* CRSTER *
*****************************************
3. PROCESSING DEFINITIONS AS DETERMINED
DURING SETUP PROCESSING.
STANDARD OUTPUT UNIT
OPENED SUCCESSFULLY
OPENED SUCCESSFULLY
OPENED SUCCESSFULLY
PROCESS
SCHEME
WIND NWSWXX
TEMPERATURE NUSWXX
MIXING HEIGHTS NWSWXX
STABILITY NWSWXX
Figure 3-7.
Third page of general report, for example where
a RAMMET type output file is generated.
3-21
October 1994
-------
determine sunrise, sunset, and insolation class, all of which
are significant when processing stability category using the
default scheme (NWSWXX). The sixth and final item is a listing
of the file names used for the output files containing the
error messages and the generated meteorological data.
Since missing values are an anathema to most of the
dispersion models routinely employed in regulatory analyses, a
table is presented indicating the number (if any) of hours
specified with missing values, see Figure 3-9. For the 96
hours of data processed in this example, there were no missing
METEOROLOGICAL PROCESSOR FOR REGULATORY MODELS (HPRM), VERSION 1.2
TODAY'S DATE AND TIME: 18-APR-88 AT 07:42:58
PROCESSING OF MERGED METEOROLOGICAL DATA
PROCESSING ASSUMPTIONS AS DETERMINED
DURING SETUP PROCESSING.
WIND SPEED/TURB. MEASUREMENT HEIGHT (M): 10.00
STACK HEIGHT (M) 10.00
TEMPERATURE HEIGHT (M) 2.00
5. LOCATIONS OF METEOROLOGICAL DATA DETERMINED
DURING SETUP PROCESSING.
DATA
PATHWAY
UA
SF
OS
SITE
ID
93815
93814
9381464
LONGITUDE
(DEGREES)
84.68W
84.00W
84.68W
LATITUDE
(DEGREES)
39.07N
39.00N
39.07N
*****************************************
* LONGITUDE AND LATITUDE FOR PROCESSING *
* 84.68 39.07 *
*****************************************
6. FILENAMES OF OUTPUT DISK FILES.
THE LISTING OF WARNINGS AND ERROR MESSAGES
GENERATED BY THIS RUN IS STORED IN FILE:
ST3AERR.LIS
THE METEOROLOGY GENERATED FOR THE DISPERSION
MODEL IS STORED IN FILE:
STAGE3A.DAT
Figure 3 - 8.
Fourth page of general report, for example where
a RAMMET type output file is generated.
3-22
October 1994
-------
values. Next is presented a table summarizing the wind speed
values. Presented in the table are the frequency of
occurrences and the harmonic average value (m/s) for six wind
speed ranges. The ranges are 0-3, 4-6, 7-10, 11-16, 17-21, and
>2l kts. In the example presented, there were nine cases with
wind speeds greater than 16 and less than (or equal to) 21 kts,
and the harmonic average wind speed for the nine cases was 9.07
m/s. Also presented is a table summarizing the methods used in
specifying the stability category.
METEOROLOGICAL PROCESSOR FOR REGULATORY MODELS (MPRM), VERSION 1.2
TODAY'S DATE AND TIME: 18-APR-88 AT 07:42:58
PROCESSING OF MERGED METEOROLOGICAL DATA
SUMMARY OF DATA PROCESSING RESULTS
NUMBER
VARIABLE PRESENT
STABILITY
CATEGORY 96
WIND
SPEED 96
WIND
DIREC. 96
RURAL
MIX HGT. 96
URBAN
MIX HGT. 96
AMBIENT
TEMPER. 96
SUMMARY OF
WIND
1 2
NUMBER
PRESENT 0.00 9.00
AVERAGE
SPEED 0.00 3.04
SUMMARY OF
NWSWXX ONSITE
96 0
NUMBER
MISSING
0
0 NUMBER CALMS:
0
0
0
0 AVERAGE (C) 0
WIND SPEED RESULTS
SPEED CLASS
345
21.00 57.00 9.00
4.30 6.67 9.07
STABILITY COMPUTATIONS
SESITE SASITE
0 0
0
.93
6
0.00
0.00
WNDUXX
0
Figure 3-9. Fifth page of general report, for example where
a RAMMET type output file is generated.
3-23
October 1994
-------
The final tables (Figure 3-10) summarize the mixing height
values present within the output meteorological data file. If
both rural and urban mixing heights can be employed with the
requested dispersion model, tables are provided for each.
Presented within the tables is the number of cases within
METEOROLOGICAL PROCESSOR FOR REGULATORY MODELS (MPRM), VERSION 1.2
TODAY'S DATE AND TIME: 18-APR-88 AT 07:42:58
PROCESSING OF MERGED METEOROLOGICAL DATA
RURAL MIXING HEIGHT RESULTS
HEIGHT STABILITY CATEGORY
RANGE
0-850
500
1000
1500
2000
>2000
PRESENT
MISSING
AVERAGE
A
0.00
0.00
0.00
0.00
0.00
0.00
0
0
0.00
B
0.00
0.00
0.00
0.00
0.00
0.00
0
0
0.00
c
0.00
0.00
0.00
1.00
0.00
0.00
1
0
1108.00
DD
8.00
13.00
11.00
3.00
0.00
0.00
35
0
467.24
DN
2.
27.
12.
0.
0.
0.
41
00
00
00
00
00
00
0
433.
24
EF
2.00
1.00
14.00
2.00
0.00
0.00
19
0
750.67
TOTAL
12
41
37
6
0
0
.00
.00
.00
.00
.00
.00
URBAN MIXING HEIGHT RESULTS
HEIGHT
RANGE
0-250
500
1000
1500
2000
>2000
PRESENT
MISSING
AVERAGE
STABILITY CATEGORY
A
0.00
0.00
0.00
0.00
0.00
0.00
0
0
0.00
B
0.00
0.00
0.00
0.00
0.00
0.00
0
0
0.00
c
0.00
0.00
0.00
1.00
0.00
0.00
1
0
1108.00
DD
7.00
12.00
8.00
8.00
0.00
0.00
35
0
547.11
DN
2.
27.
12.
0.
0.
0.
41
00
00
00
00
00
00
0
433.
24
EF
3.00
2.00
4.00
10.00
0.00
0.00
19
0
783.43
TOTAL
12
41
24
19
0
0
.00
.00
.00
.00
.00
.00
Figure 3-10
Sixth page of general report, for example where
a RAMMET type output file is generated.
height ranges (in meters) as a function of stability category.
The DD and DN indicate daytime-neutral and nighttime-neutral.
All moderate to stable cases are summarized together under the
heading EF. The table provides a means for assessing the
average mixing height as a function of stability, as well as
3-24
October 1994
-------
providing a means for assessing the variation in mixing
heights.
The summary presented on the fifth and sixth pages of the
general report is especially useful in developing the input for
dispersion models such as CDM and ISCLT. This summary can also
be used in developing a sense for the variation in dispersive
conditions within the generated output meteorological data
file.
The final page of the general report (Figure 3-11)
summarizes the messages written to the error/message file
(defined by the JB ERR input image). Warning and error
messages are listed beneath the message summary table. Various
checks are made in an effort to provide useful warning and
error messages to enable detection and correction of input run
stream errors. Section 7 provides a full discussion on how to
identify and understand the error messages.
October 1994
-------
METEOROLOGICAL PROCESSOR FOR REGULATORY MODELS (MPRM), VERSION 1
TODAY'S DATE AND TIME: 18-APR-88 AT 07:42:58
PROCESSING OF MERGED METEOROLOGICAL DATA
********************************************************
*** JOB TERMINATED NORMALLY ***
********************************************************
.2
**** MPRM MESSAGE SUMMARY TABLE ****
0- 9
MP
E 0
W 0
I 1
T 0
1
****
4 MP
****
10-19
0
0
1
0
1
20-29
0
0
0
0
0
30-39
0
0
0
0
0
WARNING MESSAGES ****
W70 FETCH: SWAPPED HR 23
ERROR
MESSAGES
****
40-49
0
0
0
0
0
INTO HR
50-59
0
0
0
0
0
24 FOR
60-69 70-79
0 0
0 1
0 1
0 0
0 2
SF-DATA
TOTAL
0
1
3
0
4
--- NONE ---
Figure 3-11.
Seventh page of general report, for example
where a RAMMET type output file is generated.
3-26
October 1994
-------
SECTION 4
PROCESSING ON-SITE DATA AND OTHER SPECIAL OPTIONS
INTRODUCTION
This section presents examples illustrating the processing
possibilities available through the use of the optional input
images and the interpretation of the reports generated by MPRM.
In particular,
o Example run streams are presented illustrating how to
extract and quality assess the data associated with each
pathway separately. Such processing is especially useful
when processing large data files.
o Alternate processing choices available through the
optional input images are illustrated.
o The contents of the general report and error/message file
are discussed.
o The on-site data pathway is fully discussed.
STAGE 1 PROCESSING
o Extraction of upper air soundings, mixing height data,
surface observations, and on-site data from the raw data
files
o Processing these data through a series of quality
assessment checks.
October 1994
-------
Suppose data are available for a 4-day period. What might the
sequence of steps be in processing these data through data
extraction and quality assessment?
Extract UA Data (Step 1)
We extract the upper air soundings and mixing height data
from their respective raw data files (tape for upper air
soundings, tape or disk for mixing height data). This step
creates one data file containing both upper air and mixing
height values. Also discussed are the optional input images:
JB RUN, JB END, UA INI, UA TOP and UA OFF.
JB STA
JB ERR DISK ERROR.LIS
JB RUN
JB FIN
UA STA
UA INI TAPE TAPE10 56OOVB ASCII 13840
UA IN2 USER RAMZI.DAT (A5,312,IX,15,13X,15) 93815
UA IQA DISK IQAUA002.DAT
UA TOP 7000
UA LOG 93815 39.83N 84.05W +5
UA EXT 63 12 31 64 1 5
UA FIN
JB END
The JB RUN is useful for checking the syntax of the run
stream input. When this input is present, the MPRM processor
reads the entire run stream for syntax and then stop
processing. This is a quick way to check the run stream for
syntax errors. To actually process data remove this input
image from the run stream.
The UA INI input is used for input of upper air
observations. In the example shown above, the INI input opens
access to a computer magnetic tape whose logical tape name is
TAPE10. The logical tape name is assigned by the instructions
4-2 October 1994
-------
given for mounting the tape. These instructions differ
depending on the computer. For a VAX they would likely be
DIGITAL Command Language (DCL) instructions, for a UNIVAC they
would be Executive Control Language (ECL) instructions, and for
an IBM they would be Job Control Language (JCL) instructions.
The procedures for mounting tapes differ from one computer site
to the next, but the basic functions remain constant. An
example of the DCL that performs this task is shown in Section
6. The data on TAPE10 are in the NWS TD-5600 format with a
variable blocked structure, which is indicated by the VB
following the 5600. (Note, variable blocking pertains to the
NCDC format and has nothing to do with IBM variable blocking
for data storage.) The information is stored on the tape in
ASCII characters (as opposed to the IBM Extended Binary Coded
Decimal Interchange Code (EBCDIC) characters). The 13840 is
the station identification number.
The UA TOP input instructs the processor to store all
upper air data at or below 7000 meters. This is thought of as
a clipping height. All data above 7000 meters are "clipped,"
i.e., disregarded. The default value for the clipping height
is 5000 meters.
The JB END input is an alternate means for identifying the
end of the run stream input. The processor assumes that the end
of the input data occurs whenever it encounters an end-of-file
(EOF) or the JB END input image.
As discussed in Appendix A the general report file is
optional, while the error/message report file is mandatory.
What happens if no general report file is given? The printed
output from MPRM, by default, is routed to logical UNIT 6, as
defined in FORTRAN-77. On a mainframe computer, this is the
default printer device for batch processing; on a PC, this is
the screen for interactive processing. Each computer system
4-3
October 1994
-------
has various means for controlling the output directed towards
logical UNIT 6. In Section 3, we discussed how to direct UNIT
6 output on a PC.
The portion of the general report summarizing the messages
within the error/message file is presented in Figure 4-1. In
this example, there are 19 informational messages shown for
UA-pathway data processing.
[[[
METEOROLOGICAL PROCESSOR FOR REGULATORY MODELS (MPRM), VERSION 1.2
TODAY'S DATE AND TIME: 14-APR-88 AT 08:59:48
INITIAL PROCESSING ON RAW METEOROLOGICAL DATA
********************************************************
*** JOB TERMINATED NORMALLY ***
********************************************************
**** MPRM MESSAGE SUMMARY TABLE ****
0- 9 10-19 20-29 30-39 40-49 50-59 60-69 70-79 TOTAL .
JB
E
, U
I
UA
E
, W
, I
0
0
0
0
0
0
0
1
7
0
0
0
0
0
0
0
0
0
0
0
0
0
0
19
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
7
0
0
19
080 19 0000 27
**** WARNING MESSAGES ****
0 JB W12 UAQAST: SUMMARY: MISSING/ERRORS IN UA-OQA CARD
**** ERROR MESSAGES
--- NONE ---
Figure 4-1. Second page of general report, for example where
upper air soundings and mixing height data are
extracted for a 6-day period.
The bulk of these informational messages are the result of
automatic upper air data modifications made to the soundings
4-4
-------
during the extraction process. The automatic modifications are
described in more detail in Section 5. In brief, modifications
are made for unreasonable temperature changes in the vertical,
resulting from an incorrect sign being assigned to a
temperature value. Modifications are also made to remove
unnecessary levels of data, resulting from inclusion of the
extra standard (mandatory) reporting levels of data (such as
the 900, 850, 500 mb levels of data). These mandatory levels
are determined by interpolation from the original data. As
«
indicated in the error/message file (Figure 4-2), 13 mandatory
levels were deleted and the signs changed for two temperature
values.
12
0
0
0
0
0
0
0
0
0
0
1
1
2
1
2
3
2
1
1
1
2
3
1
1
6
0
JB
JB
JB
JB
JB
JB
JB
JB
UA
UA
UA
UA
UA
UA
UA
UA
UA
UA
UA
UA
UA
UA
UA
UA
UA
UA
UA
119
U12
110
111
112
no
111
112
130
136
137
137
137
137
137
137
137
137
137
137
137
137
137
137
137
139
130
SETUP:
UAQAST:
TEST:
TEST:
TEST:
TEST:
TEST:
TEST:
UAEXT:
UAEXT:
UAAUTO:
UAAUTO:
UAAUTO:
UAAUTO:
UAAUTO:
UAAUTO:
UAAUTO:
UAAUTO:
UAAUTO:
UAAUTO:
UAAUTO:
UAAUTO:
UAAUTO:
UAAUTO:
UAAUTO:
GETMIX:
UAEXT:
ENCOUNTERED END OF "JOB/RUN CARD"
SUMMARY: MISSING/ERRORS IN UA-OQA CARD
SUMMARY: NO SF-EXT CARD, NULL EXTRACT
SUMMARY: NO SF-IQA CARD, NULL QA
SUMMARY: NO SF-OQA CARD, NULL MERGE
SUMMARY: NO OS-EXT CARD, NULL EXTRACT
SUMMARY: NO OS-EXT CARD, NULL QA
SUMMARY: NO OS-EXT CARD, NULL MERGE
**** UPPER AIR EXTRACTION ****
* *** AUTOMATIC SDG. CHECKS ARE ON
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
1
1
1
1
1
1
1
1
1/7;
1/19;
2/7;
2/7;
2/19;
2/19;
2/19;
2/19;
3/7;
3/19;
4/7;
4/ 7;
4/ 7;
5/ 7;
5/19;
END-OF-FILE
11
LVL 6 -TEMP. SIGN
450.MB
900. MB
850.
950.
900.
600.
MB
MB
MB
MB
-MAND.
-MAND.
-MAND.
-MAND.
-MAND.
-MAND.
LVL
LVL
LVL
LVL
LVL
LVL
LVL 8 -TEMP. SIGN
700.
900.
850.
500.
450.
950.
900.
MB
MB
MB
MB
MB
MB
MB
-MAND.
-MAND.
-MAND.
-MAND.
-MAND.
-MAND.
-MAND.
LVL
LVL
LVL
LVL
LVL
LVL
LVL
CHANGE
DELETED
DELETED
DELETED
DELETED
DELETED
DELETED
CHANGE
DELETED
DELETED
DELETED
DELETED
DELETED
DELETED
DELETED
, END-OF-DATA
SDGS AND 6 MIXING HTS EXTRACTED
Figure 4-2.
Error/message file generated during a 6-day
extraction period for upper air soundings and
mixing height data.
Experience has shown that the modifications outlined are
frequently useful, and for this reason, MPRM includes them as
automatic procedures. However, these modifications can be
4-5
October 1994
-------
deactivated through the use of the optional input image UA OFF.
Quality Assess UA Data (Step 21
Having extracted the UA data, we now process these data
through Stage 1 quality assessment. With the information from
the reports generated, one can use a text editor to alter
suspect values. This process is continued until the UA data
are ready for Stage 2 processing. Also discussed are the
optional input images: JB OUT and UA CHK.
JB STA
JB OUT DISK REPORT.LIS
JB ERR DISK ERROR.LIS
JB FIN
UA STA
UA IQA DISK IQAUA002.DAT
UA OQA DISK OQAUA002.DAT
UA CHK UAPR 1 -9999 4000 10999
UA LOG 93815 75.97W 33.37N +5
UA FIN
JB END
There are not many differences between the above input and
that given for step 1. The JB OUT image allows redirection of
UNIT 6 output to a file. In this case, the file is REPORT.LIS.
The most important differences are that the INI and IN2 input
are not present. These are only needed for initial data
extraction from the raw data files. For processing the UA data
through quality assessment, the input data file is defined by
the Input-to-Quality-Assessment (IQA) data. The file defined
by the Output-from-Quality-Assessment (OQA) input is used to
store the data read from the IQA file as it is processed
through the quality assessment checks. As discussed in Section
3, the format of the data written to the OQA file is identical
to that for the IQA file.
4-6 October 1994
-------
Inspection of the example given above reveals another
input that is unique to the assessment phase of Stage 1
processing, the CHK input, which is interpreted as:
UA CHK UAPR 1 -9999 4000 10999
Upper bound of range
Lower bound of range
Missing value indicator
Range check switch
Name of variable
The CHK input provides a means for redefining (overriding) the
default settings used in during quality assessment. Each
variable processed through the range checks is given a unique
four-character name. A list of the names for the UA-pathway is
given in Table C-l in Appendix C. The units of values entered
for the upper and lower bounds must be the same as those
employed within the processor and defined in Appendix C.
In the example given, UAPR refers to the upper air
atmospheric pressure. These are stored as tenths of millibars
within the processor. Hence, the lower bound of 4000 is
equivalent to 400 mb. The upper bound given is equivalent to
1099.9 mb.
The range check switch tells the processor how to perform
the range check. A switch value of 1 excludes as valid values
the upper and lower bounds. Hence, a value equal to the lower
bound would be considered a violation, and a warning message
would be listed to the error/message file. A switch value of 2
includes as valid values the upper and lower bounds. As an
example of where the switch eases the specification of the
range check bounds, consider the difference between mixing
heights and snow amounts. A value for snow amount of 0 is
logical, but a mixing height value of 0 is illogical.
4-7
October 1994
-------
The CHK input has a very special quality: the MPRM
processor will 'remember' this input.
The processor incorporates the CHK input into the header
record that is written to the OQA file. If these data are
reprocessed, and the OQA file (output file from a previous
quality assessment run) is the file defined as the IQA file
(input file for a quality assessment run), the MPRM processor
will pick up the previously defined CHK input and redefine the
range check information accordingly, while reading the header
record, thus 'remembering' the input. Therefore, the user does
not have to continually repeat the CHK input. However, the CHK
input can be given for the same variable as many times as
desired. The processor will use the latest definition in
performing the range checks, allowing the range checks to be
altered as needed.
The first page of the general report (Figure 4-3) is a
status report which basically presents a view of the processing
environment within MPRM just before any attempts are made to
read or process data from any of the data files. The first
item on the status report is the MPRM heading, which includes
the version number and the date the processor was run. This
heading is standard and appears at the top of all output pages.
Following the heading, within the body of the report, are the
names of files to be used and the status of the pathway data
files (OPEN or NOT OPEN). Also reported are other pertinent
information such as the extraction dates, the site identifier,
and the latitude and longitude. In the example presented,
since a file was designated for the general report, the report
was written to the file named REPORT.LIS. Within item 2 (UPPER
AIR DATA) is given the station identification and the latitude
and longitude for the UA-pathway data. Below this is a message
4-8 October 1994
-------
summarizing the processes that MPRM is anticipated to perform
on the UA-pathway data.
METEOROLOGICAL PROCESSOR FOR REGULATORY MODELS (MPRM), VERSION 1.2
TODAY'S DATE AND TIME: U-APR-88 AT 09:28:20
INITIAL PROCESSING ON RAW METEOROLOGICAL DATA
STATUS REPORT PRIOR TO BEGINNING PROCESSOR RUN
1. REPORT FILE NAMES
ERROR MESSAGES: ERROR.LIS
SUMMARY OF RUN: REPORT.LIS
2. UPPER AIR DATA
SITE ID LATITUDE(DEC.) LONGITUDE(DEG.)
93815 33.37N 75.97H
THE PROCESS(ES) MPRM ANTICIPATES TO PERFORM ARE
QUALITY ASSESSMENT ONLY
Figure 4-3.
EXTRACT OUTPUT-
QA OUTPUT
OPEN: IQAUA002.DAT
OPEN: OQAUA002.DAT
3. NWS SURFACE DATA
THE PROCESS(ES) MPRM ANTICIPATES TO PERFORM ARE
NONE, NO DATA TO BE PROCESSED ON THIS PATH
4. ON-SITE DATA
THE PROCESS(ES) MPRM ANTICIPATES TO PERFORM ARE
NONE, NO DATA TO BE PROCESSED ON THIS PATH
First page of general report, for example where
UA-pathway data are submitted for quality
assessment.
The second page (Figure 4-4) summarizes the messages
listed to the error/message file. In this case, 135 quality
assessment messages were generated.
4-9
October 1994
-------
METEOROLOGICAL PROCESSOR FOR REGULATORY MODELS (MPRH), VERSION
TODAY'S DATE AND TIME: 14-APR-88 AT 09:28:24
INITIAL PROCESSING ON RAW METEOROLOGICAL DATA
********************************************************
*** JOB TERMINATED NORMALLY ***
********************************************************
JB
E
W
I
UA
E
U
I
Q
0- 9
0
0
0
0
0
0
0
0
****
****
10-19
0
0
8
0
0
0
0
8
**** MPRH
1.2
MESSAGE SUMMARY TABLE ****
20-29 30-39
0
0
0
0
0
0
0
0
WARNING MESSAGES
...
NONE ---
ERROR MESSAGES
...
NONE
0
0
0
0
0
5
135
HO
****
****
40-49
0
0
0
0
0
0
0
0
50-59
0
0
0
0
0
0
0
0
60-69
0
0
0
0
0
0
0
0
70-79
0
0
0
0
0
0
0
0
TOTAL
0
0
8
0
0
5
135
148
Figure 4-4.
Second page of general report, for example where
UA-pathway data are submitted for quality
assessment.
The last page of the general report (Figure 4-5)
summarizes the quality assessment checks, or audit summary. In
this case, only the twice-daily mixing height values were
summarized. These are the only variables audited by default on
the UA-pathway. One of the six morning urban mixing height
values, UAMl, was found to be greater than the range check
upper bound (500 m) and three of the afternoon mixing height
values, UAM2, were found to be lower than the range check lower
bound (500 m). Whether any of these mixing height values are
unreasonable can only be determined through inspection of the
values coupled with expert judgment.
4-10
October 1994
-------
METEOROLOGICAL PROCESSOR FOR REGULATORY MODELS (MPRM), VERSION 1.2
TODAY'S DATE AND TIME: 14-APR-88 AT 09:28:26
********************************************************
*** JOB TERMINATED NORMALLY ***
********************************************************
**** SUMMARY OF THE QA AUDIT ****
MIXING HTS | VIOLATION SUMMARY |
TOTAL # LOWER UPPER X
# DBS MISSING BOUND BOUND ACCEPTED
UAM1 6 001 83.33
UAM2 6 030 50.00
THERE IS NO AUDIT TRAIL FOR SOUNDINGS
THIS CONCLUDES THE AUDIT TRAIL
| TEST VALUES |
MISSING LOWER UPPER
FLAG BOUND BOUND
-9999.0. 50.0, 500.0
-9999.0. 500.0, 3500.0
Figure 4-5.
Last page of general report, for example where
UA-pathway data are submitted for quality
assessment.
The range check messages (refer to Figure 4-6) provide
sufficient information to identify each occurrence of a range
check violation - the bound violated, LB for lower bound and UB
for upper bound, the value of the bound, and the data value
causing the message. The second line of the message provides
the year, month, day, and hour.
4-11
October 1994
-------
Figure 4-6.
11 JB 119 SETUP: ENCOUNTERED END OF "JOB/RUN CARD"
0 JB 110 TEST: SUMMARY: NO UA-EXT CARD, NULL EXTRACT
0 JB 110 TEST: SUMMARY: NO SF-EXT CARD, NULL EXTRACT
0 JB 111 TEST: SUMMARY: NO SF-IQA CARD, NULL QA
0 JB 112 TEST: SUMMARY: NO SF-OQA CARD, NULL MERGE
0 JB 110 TEST: SUMMARY: NO OS-EXT CARD, NULL EXTRACT
0 JB 111 TEST: SUMMARY: NO OS-IQA CARD, NULL QA
0 JB 112 TEST: SUMMARY: NO OS-OQA CARD, NULL MERC5E
107 UA Q37 INTECK: LB: 500 UAM2: 155
107 UA : ON 64/01/01/07
119 UA 137 INTECK: LB: 500 UAM2: 155
119 UA : ON 64/01/01/19
407 UA 038 INTECK: UB: 500 UAM1: 1108
407 UA : ON 64/01/04/07
12 UA 139 UAQASM: EOF AFTER UA REPORT # 11 (NORMAL)
Partial listing of error/message file generated
during a 6-day quality assessment period for
upper air soundings and mixing height data.
The message code in the second range check message starts
with an I rather than a Q. When the daily sounding and mixing
height data were extracted, the mixing heights were repeated
with each sounding during the day; therefore, any mixing height
violations after the first one for the day were labeled as
informational messages. This provides an accurate count of the
number of acceptable mixing heights in the audit summary and,
at the same time, reminds the user that the mixing heights
appear in the input file to quality assessment in multiple
places.
As shown and discussed above, part of the information
given within the general report is a summary of the occurrences
of range check violations for several of the meteorological
variables. Through the use of AUD input, other variables on
the UA-pathway can be added to the list of variables summarized
within the audit report. On the UA-pathway, only the
4-12
October 1994
-------
twice-daily mixing height values (NCDC TD-9689 data) are
audited by default. The syntax of the AUD input is given in
Appendix B. An example will be presented later within the
discussion of the example run stream for quality assessing the
NWS hourly surface weather observations (SF-pathway).
Extract SF Data (Step 3)
In this step we extract the hourly surface data from their
raw data file (disk). This creates a new data file containing
the extracted NWS hourly weather observations. Each hour's
observation requires two lines in the output file.
In comparing the run stream given below with that given
for step 1, there are many similarities. The reuse of the
general report and error files reduces the number of files
created. Furthermore, these files are of little use once the
user has successfully completed each processing step.
JB STA
JB ERR DISK ERROR.LIS
JB OUT DISK REPORT.LIS
JB RUN
JB FIN
SF STA
SF IN2 DISK RAMSFC.DAT CD144FB 93814
SF IQA DISK IQASF002.DAT
SF LOG 93814 84.05W 39.83N 0
SF EXT 64 01 01 64 01 04
SF FIN
As no new input images are illustrated in this run stream,
portions of the reports are illustrated for comparison with the
examples discussed in previous examples. As shown in Figure
4~7 / there were several informational messages and one warning
message. This error/message file is similar in content and
October 1994
-------
interpretation to the error/message file generated for the
second upper air example (refer to Figure 4-6).
11
0
0
0
0
0
0
0
0
96
96
JB
JB
JB
JB
JB
JB
JB
JB
SF
SF
SF
119
110
111
112
U12
110
111
112
140
149
149
SETUP:
TEST:
TEST:
TEST:
SFQAST:
TEST:
TEST:
TEST:
SFEXT:
GETSFC:
SFEXT:
FOUND "END OF FILE" ON DEVICE DEVIN 5
SUMMARY: NO UA-EXT CARD, NULL EXTRACT
SUMMARY: NO UA-IQA CARD. NULL CA
SUMMARY: NO UA-OQA CARD, NULL MERGE
SUMMARY: MISSING/ERRORS IN SF-OQA CARD
SUMMARY: NO OS-EXT CARD, NULL EXTRACT
SUMMARY: NO OS-IQA CARD, NULL QA
SUMMARY: NO OS-OQA CARD. NULL MERGE
*** SURFACE OBSERVATION EXTRACTION ***
END-OF-FILE, END-OF-DATA
96 SURFACE OBSERVATIONS EXTRACTED
Figure 4-7.
Error/message file, for example where data are
extracted for use on the SF-pathway.
The final summary report (Figure 4-8) is quite similar to
that for the second upper air example (refer to Figure 4-4).
The three messages specifically issued for the SF-pathway were
only informational. They were concerned with accessing the
data files for the surface observations and reporting how many
observations were processed.
4-14
October 1994
-------
METEOROLOGICAL PROCESSOR FOR REGULATORY MODELS (MPRM), VERSION
TODAY'S DATE AND TIME: U-APR-88 AT 09:25:07
INITIAL PROCESSING ON RAW METEOROLOGICAL DATA
********************************************************
*** JOB TERMINATED NORMALLY ***
A*****************************************************'**
0- 9
JB
E 0
W 0
I 0
SF
E 0
U 0
I 0
0
****
0 JB
****
10-19
0
1
7
0
0
0
8
**** MPRM
20-29
0
0
0
0
0
0
0
1.2
MESSAGE SUMMARY TABLE ****
30-39
0
0
0
0
0
0
0
40-49 50-59
0 0
0 0
0 0
0 0
0 0
3 0
3 0
WARNING MESSAGES ****
W12 SFQAST: SUMMARY: MISSING/ERRORS IN
ERROR MESSAGES
...
NONE ---
****
60-69
0
0
0
0
0
0
0
SF-OQA
70-79
0
0
0
0
0
0
0
CARD
TOTAL
0
1
7
0
0
3
11
Figure 4-8.
Second page of general report, for example where
data are extracted for use on the SF-pathway.
Looking at the run stream SF EXT input, we see that the
intention was to extract 4 days of data (January 1 - January 4,
inclusive). With 24 hours in a day that translates to 96
hours. In Figure 4-7, the message that 96 observations were
processed is an indication that the data requested were indeed
found and extracted.
Quality Assess SF Data fStep 41
The SF-pathway data are now processed through the quality
assessment portion of Stage 1 processing. As with the UA data,
we review the reports generated, and using a text editor, alter
suspect values. This process is continued until the SF-pathway
data are ready for Stage 2 processing.
4-15
October 1994
-------
JB STA
JB OUT DISK REPORT.LIS
JB ERR DISK ERROR.LIS
JB FIN
SF STA
SF IQA DISK IQASFOO2.DAT
SF AUD RHUM PWTH
SF LOG 93814 84.05W 39.83N +O
SF FIN
JB END
The new optional input illustrated in this example is SF
AUD. Comparison of the above run stream with that given in
step 2 for the UA-pathway quality assessment reveals that the
two run streams are almost identical. This is no surprise,
especially since the logic associated with quality assessment
is identical regardless of the data involved. The data given
as input (IQA input) is to be checked and rewritten to the
output file (OQA input). Warnings and messages are written to
the error report file (JB ERR input) and summarized in the
general report file (JB OUT input).
In the upper air example, the UA CHK input was discussed.
This input can also be used when processing SF and OS data as
needed. Similarly, the AUD keyword to be discussed now can be
used in processing UA and OS data.
SF AUD RHUM PWTH
' Name of variable
Name of variable
The AUD input can be repeated as often as needed in order
to list all the variables to be added to the audit report.
Several variables can appear on one input image, with the
restriction that no field extends beyond column 80. The
default list of audit variables on the SF-pathway are:
sea-level pressure, station pressure, ceiling height, sky cover
4-16 October 1994
-------
(total and opaque), horizontal visibility, dry bulb
temperature, wind direction, and wind speed. The above AUD
input adds RHUM and PWTH to this default list of audit
variables. RHUM is the variable name for relative humidity and
PWTH is the variable name for present weather. Present weather
information might be useful in a dispersion analysis if one
anticipated that a pollutant might be "washed out" of the
atmosphere during a precipitation event, like a rain shower,
and therefore it would no longer be available for transport
further downwind. Such information would be relevant in
attempting to compare values of gaseous concentrations observed
versus concentration values estimated by a dispersion model
which makes no provision for pollutant washout.
The AUD input differs from the CHK input in that the AUD
input is not 'remembered' by the MPRM processor. Hence, any
additional variables that are to be added to the audit report
must be repeated with the AUD input each time quality
assessment is performed.
The audit summary (Figure 4-9) is similar in structure to
the one discussed for the UA data (refer to Figure 4-5) but
different in content. The two primary differences between the
upper air audit and the surface audit are that there are no
multilevel SF data and that there are SF variables containing
two pieces of information. These 'double duty' variables are
two integer variables combined to form a single integer
variable. Table C-2 shows which SF variables are composed of
two variables. These combined variables are separated before
use in quality assessment checks or use in development of
meteorological data files.
October 1994
-------
METEC
SURFACE DA
NOTE:
SLVP
PRES
CLHT
TS
KC
PW
TH
KZVS
TMPD
RHUH
WD16
WIND
TEST
(SEE
WOLOGICAL PROCESSOR FOR
TODAY'S DATE AND T
************************
*** JOB TERM
************************
**** SUMMARY
FA 1 i/fm ATI
TOTAL # LOWER
# DBS MISSING BOUND
96 00
96 00
96 00
96 00
96 00
96 00
96 00
96 00
96 00
96 00
96 00
96 00
IEGULATI
IME: 14
*******
INATED
*******
3F THE
ON SUMH
UPPER
BOUND
0
0
0
0
0
0
0
0
0
0
0
0
DRY MODELS (MPRM),
-APR-88 AT 09:48:2
******************
NORMALLY
******************
3A AUDIT ****
any 1 1
X
\ I
MISSI
VERSION 1.
7
*******
***
*******
-TEST VALUE
NG LOWER
ACCEPTED FLAG
100
100
100
100
100
100
100
100
100
100
100
100
VALUES MATCH INTERNAL SCALING APPLIED
APPENDIX C OF THE USER'S
THE FOLLOWING CHECKS WERE ALSO
OF
96 REPORTS, THERE WERE
0 CALM WIND CONDITIONS
GUIDE)
PERFORMED FOR
(WS=0,
WD=0)
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
TO
-9999
-9999
-9999
99
99
99
99
-9999
-9999
-9999
-9999
-9999
o.
o,
0.
0,
.0.
o.
o,
0,
0,
.0.
o.
o.
BOUND
9000
9000
0
0
0
0
0.
-o.
o,
0,
o.
0,
0.0.
0
-300
0
0
0
.0.
o,
o.
o,
.0,
2
i
I
UPPER
BOUND
10999
10999
300
10
10
92
92
1640
350
100
36
500
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
VARIABLES
THE SURFACE
QA
0 ZERO WIND SPEEDS WITH NONZERO WIND DIRECTIONS
0 DEW-POINT GREATER THAN DRY
THE TIMES OF THESE OCCURRENCES
WITH
QUALIFIERS CLM, WDS, TDT
CAN BE
(RESP.)
BULB TEMPERATURES
FOUND
IN
THE MESSAGE
FILE
THIS CONCLUDES THE AUDIT TRAIL
Figure 4-9.
Last page of general report, for example where
SF data are checked for integrity.
In Figure 4-9, this latter difference can be seen near
the middle of the report where the four-character variable TSKC
has been split into two variables, TS and KC. The variable
TSKC contains information regarding both total sky cover and
opaque sky cover. In the audit report, the line called TS
reports the audit results for total sky cover. The line called
KC reports the audit results for opaque cloud cover. The
variable PWTH is another example where two fields of
information are stored together. The PWTH variable contains
information regarding both the prevailing and secondary weather
occurring during the observation. In the audit report, the
4-18
October 1994
-------
line called PW reports the audit results for the prevailing
weather data, and the line called TH reports the audit results
for the secondary weather data.
Because there were no quality assessment violations and no
missing data, all data were accepted. While somewhat
uninteresting, this is what is hoped for when reasonable upper
and lower bounds are specified. The upper and lower bounds and
the missing value indicator employed in the quality assessment
checks are shown for each variable on the right side of the
table. The default values and multipliers used for each
variable are shown in Appendix C.
Besides checking individual variables against upper and
lower bounds, several other checks are made involving
comparison of two variables within the same observation. The
results of these checks are presented below the audit summary
table. The additional comparison checks are:
o Dew-point temperature exceeding dry bulb temperature
o Defined wind direction with zero wind speed
o Occurrences of calm wind conditions.
Extract and Quality Assess OS Data (Step 5)
In this step, we extract and quality assess the on-site
data from their raw data files. All data extracted will
automatically be processed through the quality assessment
checks. Notice that no provisions have been made for
processing these data from a computer tape.
Unique to the above examples for processing the NWS
meteorological data, the format of the original data files was
known, as NCDC supplies these data in only several well-defined
formats. This assumption is not valid in the context of user
4~"19 October 1994
-------
collected on-site meteorological data. There is no standard
for storage of these data, and furthermore, there cannot be.
The variety of measurements possible preclude development of
such a standard.
Therefore, the MPRM must be "told" the storage format. To
reduce the pain to the user, MPRM "remembers" the format. This
occurs when MPRM first reads the on-site data. The first time
the data are read, the data format is defined using special
input records (MAP and FMT) . MPRM stores, as part of the
header records to these data, the MAP and FMT input images and
uses these header records in all successive reads.
Below is an example of the run stream for the initial
extraction and processing of on-site data. Comparison of this
run stream with that given in steps 1 and 3 reveals that the
run streams are very similar, down to the point where the MAP
and FMT input are given. This is no surprise, since the logic
associated with initially reading a data file is identical
regardless of the data involved.
The one difference is that for the UA and SF data, INI and
IN2 input were used to define the original source files of
data, whereas for the OS data, the IQA input is used to define
the source file of OS data.
The data given as input (IQA input) are checked, hourly
averages are computed if needed, and the hourly averages are
written to the output file (OQA input) . Warnings and messages
are written to the error/message file (JB ERR input) and
summarized in the general report file (JB OUT input) .
October 1994
-------
JB STA
JB OUT DISK REPORT.LIS
JB ERR DISK ERROR.LIS
JB FIN
OS STA
OS IQA DISK IQAOS001.DAT
OS OQA DISK OQAOS001.DAT
OS EXT 64 1 1 64 1 4
OS LOG 9381464 39.07N 84.68W 0
OS AVG 4
OS DTI 2 10
OS MAP DAT01 OSYR OSMO OSDY OSHR OSMN DT01
OS MAP .DAT02 HT01 WS01
OS MAP DAT03 HT02 TT02 WS02 WD02 SA02
OS FMT DAT01 (5X,3(12.2),5X,212,10X,F10.4)
OS FMT DAT02 (1X,2F10.4)
OS FMT DAT03 (1X,5F10.4)
OS SFC SETUP ANNUAL 2
OS SFC VALUES 1 1 0.6 0.5 0.3
OS SFC VALUES 1 2 0.8 0.6 0.1
OS SFC SECTORS 1 30 150
OS SFC SECTORS 2 150 30
OS CHK DT01 1 -999 -2 10
OS CHK WS 1 -999 0 50
OS CHK WD 1 -999 0 360
OS CHK TT 1 -999 -30 35
OS CHK SA 1 -999 0 35
OS FIN
JB END
The OS AVG input image, unique to the OS-pathway, defines
the maximum number of data records expected for each hour,
OSAVG. The default value for OSAVG is 1. In the example
given, the AVG keyword instructs MPRM that up to four data
records are provided each hour (OSAVG is set to 4) . This would
be the case if the original source files of OS data contained
data for each 15-minute period. The logic within MPRM to
accomplish the averaging is quite simple. It reads the data
records until a new hour is encountered. The program compares
the number of valid data values encountered versus the minimum
number of values accepted for computation of an hourly average.
The minimum number is computed as OSAVG/2 (if OSAVG is an even
number) or as OSAVG/2 + 1 (if OSAVG is an odd number). The AVG
October 1994
-------
input image is only needed the first time the OS data are
processed. Thereafter, only hourly averages are available for
processing.
The DTI input signals that there are temperature
difference measurements available, Delta-T, and that the
measurements were made at 2 and 10 m above the ground. The
convention followed in MPRM is that the temperature difference
value represents the upper measurement height temperature minus
the lower measurement height temperature. There can be up to
three temperature differences (DTI, DT2, and DT3) given as
input. The measurement heights for each would be entered as
they were for DTI. Temperature differences are treated as
scalar input as they represent a single measurement (typically
a voltage potential or change in resistance) .
The next three input images employ the MAP keyword. They
describe to the processor the sequence and format (in the
FORTRAN sense) of meteorological variables available for input
for one time period. All the possible variables have been
given 4-character names which are listed in Table C-3 in
Appendix C. The MAP DATxx defines the sequence of variables
available for input to the processor. The xx indicates the
record number, as DAT01 is the first record of data. It is
required that each observation be labeled with a date and time,
and that this be given before the rest of the data values.
Below is an example of what the first six data records of
on-site data might look like (Figure 4-10) . As the data have a
specific format, scales are given at the top and bottom of the
table to ease interpretation.
4~22 October 1994
-------
1 2345678901 2345678901 234567890123456789012345678901 234567890
Record
number
1
2
3
4
5
6
640201
2.0
10.0
640201
2.0
10.0
Sample On-Site
1005
2.1
3.2
1010
2.3
3.7
(OS) Input Data
-0.01
2.0 181.2
-0.01
2.1 181.0
10.3
9.7
1 234567890 1 234567890 1 234567890 1 234567890 1 2345678901 234567890
Figure 4-10. Example of on-site data.
In the example, records 1 through 3 are for the first time
period and the next three records are for the next time period.
The first record for each time period begins with a date and
time group (OSYR, OSMO, OSDY, OSHR OSMN), followed by the
Delta-T value (DT01). Record 1 deciphered is February 2, 1964,
at 10:05 a.m. The temperature difference value is -0.01 °C
(given the DTI input, this measurement was taken between 10 and
2 m) .
The DAT02 input declares that the second read of input
data consists of a measurement height and wind speed. HTyy
indicates measurement height, where the yy indicates level
above ground. For instance, HT02 is above HT01. Likewise,
WSyy indicates wind speed, and the 01 means this value was
measured at the lowest measurement height, HT01. These is no
relationship assumed between the data record sequence number
xx, given in DATxx, and the measurement level yy, given in the
variable name. If all the wind speed measurements had been
reported on the second record, then the DAT02 input would have
looked like:
OS MAP DAT02 WS01 WS02
or
OS MAP DAT02 WS02 WS01
4-23
October 1994
-------
depending on the order the wind speed values are given within
the file.
Obviously, there are many variations possible. Another
example is discussed in Appendix B. The primary constraints to
remember are:
o Date and time (the order is not important) must be given
first for each on-site observation. The time must, at the
very least, define the hour to be associated with the
observation. All date and time data must be capable of
being read using INTEGER format.
o The other data are given next, following the date and
time. There can be up to 20 records (DATxx records) of
input data. No more than 40 values can be given in any
one record.
The FMT input images define the FORTRAN FORMAT to be
employed for reading each input record. A FMT DAT01 image is
used to define the format for the first scalar record.
The FMT DAT01 record shown in the example demonstrates how
the date and time, made up of five fields of information -
year, month, day, hour, and minutes - may appear in the input
data set as two integer values. The year, month, and day are
given as 640201 and the hour and minutes are given as 1005. By
appropriate specification of the FORTRAN FORMAT, the MPRM can
handle various data file structures. For instance, if the time
were stored as 10:05, a format with 12.2,IX,12.2 rather than
212.2 would be used.
The formats given in the FMT input are used by the
processor in listing the data values to the file specified in
the OQA input. By using 312.2 rather than 312 in our format
&.?&.
* October 1994
-------
specification, the date group, 640201, would be written to the
OQA file as 640201. Using 312 would result in the date group,
640201, becoming 64 2 1 in the OQA file. The leading zeros are
lost with the 312 format. It is easier to read with the
leading zeros; therefore, use of 12.2 style formats is
recommended in specifying the date and time FMT input.
Not all variables appearing in the input data file have to
be mapped. Only those variables the user believes will be
useful in the final application are required, whether it be
dispersion modeling or other analyses. With proper usage of
the X or T format specifier, variables can be skipped; with
careful usage of the / format specifier, entire records can be
skipped. The format statements defined by the FMT input images
are used in all processing involving reading or writing of the
on-site data from data files. Hence, if X, T or / format
specifiers have been employed, the information skipped in the
original raw data file will appear as blanks in subsequent
files.
The format statements defined by the FMT input images
are used in all processing involving reading or
writing of the on-site data from data files. If in
the course of quality assessment it is found necessary
to edit the on-site data file, be sure not to alter
the data format, or read errors will occur in
subsequent attempts to process the data.
The SFC input is optional. However, it is recommended
that the SFC input be given whenever on-site data are present.
The SFC input provides the processor with a description of the
characteristics of the measurement site. These characteristics
are described in terms of the surface albedo, the Bowen ratio,
and the surface roughness length. The albedo is the fraction
of radiation reflected at the surface (mostly incoming solar
4-25 October 1994
-------
radiation). Typical values range from 0.1 for thick deciduous
forests to 0.65 for fresh snow. The Bowen ratio is H/He, where
H is the upward sensible heat flux and H6 is the heat flux used
in evaporation. The Bowen ratio varies from 0.1 over water to
10.0 in desert. In principle, the surface roughness length is
the height at which the horizontal wind speed is typically
zero. Roughness length values range from less than 1 cm over a
calm water surface to 1 m or more over a forest. The default
values for albedo, Bowen ratio, and roughness length are 0.25,
0.75, and 0.15 m, respectively. These are typical of
cultivated land with average moisture content.
Through the use of the SFC input, the specification of
these site characteristics can be varied according to the wind
direction and the month of the year. The SFC SETUP input
specifies whether the new values apply annually, seasonally, or
monthly and how many wind direction sectors will be employed in
specifying the values. In the example, ANNUAL values are given
for two wind sectors. The two SFC VALUES input images define
the values for the albedo, the Bowen ratio, and the roughness
length (in that order). The first SFC VALUES input is
deciphered as:
OS SFC VALUES 1 1 0.6 0.5 0.3
Roughness length
Bowen ratio
Albedo
Wind sector number
Time period key
The time period key is defined with respect to how frequently
the values are varied throughout the year. For annual
definitions, the time period key is always 1. For cases where
the values are to be varied either seasonally or monthly, refer
to the discussion and example given in Appendix B for the SFC
October 1994
-------
keyword. The wind sector number defines which wind sector, as
defined on the SFC SETUP input image, the site characteristics
given are applicable.
The last two SFC SECTOR input images define the beginning
and ending azimuths of the wind direction sectors. The sectors
are defined in terms of wind direction (in the meteorological
sense, i.e., the direction from which the wind is blowing with
North associated with the angle 360°). The beginning and
ending directions are defined in a clockwise sense. Hence,
for the example run stream, the first sector is for wind
directions from the east to southeast, between 30 and 150
degrees. For hours having winds from 30 to 150 degrees, the
albedo, the Bowen ratio, and the roughness length are to be
0.6, 0.5 and 0.3 m, respectively.
For the example given, the measurement heights for the
multilevel variables were defined within the data set, as part
of the data record. This is not always the case. In fact, it
is likely the exception rather than the rule. Since
measurement heights are fixed, they need not be repeated within
each data record. For data sets where the measurement heights
for the multilevel variables are not given within the data
records, use the OS HGT input. For the above example, the
heights could have been defined as:
OS HGT 2 2.0 10
This indicates that there are two measurement heights, the
first (and lowest) at 2 m above ground and the second at 10 m
above ground.
The general report resulting from the example discussed is
shown in Figures 4-11, 4-12 and 4-13. As can be seen from
inspection of Figure 4-12, there were 35 range violation
4-27 October 1994
-------
messages generated and 141 informational messages. Of the 141
informational messages, 140 were due to range violations during
quality assessment. If the input data are more frequent than
one observation per hour, then all intrahour range (i.e.,
quarter-hour intervals in the example) violations are reported
as informational messages (I). If the hourly average violates
the prescribed range, the violation is reported as a quality
assessment message (Q). Of the default audit variables for the
on-site data, two were not found in the data map: mixing
height, and standard deviation of the vertical wind direction
fluctuations. These were automatically removed from the audit
list, as indicated by the warning messages listed beneath the
message summary table (Figure 4-12). The audit summary (Figure
4-13) reports that there are several values missing within the
hourly data. For those data present within the audit report,
all but two passed the range checks for each variable.
October 1994
-------
METEOROLOGICAL PROCESSOR FOR REGULATORY MODELS (MPRM), VERSION 1.2
TODAY'S DATE AND TIME: 14-APR-88 AT 09:55:48
INITIAL PROCESSING ON RAW METEOROLOGICAL DATA
STATUS REPORT PRIOR TO BEGINNING PROCESSOR RUN
1. REPORT FILE NAMES
ERROR MESSAGES: ERROR.LIS
SUMMARY OF RUN: REPORT.LIS
2. UPPER AIR DATA
THE PROCESS(ES) MPRM ANTICIPATES TO PERFORM ARE
NONE, NO DATA TO BE PROCESSED ON THIS PATH
3. NWS SURFACE DATA
THE PROCESS(ES) MPRM ANTICIPATES TO PERFORM ARE
NONE, NO DATA TO BE PROCESSED ON THIS PATH
4. ON-SITE DATA
SITE ID
KINCAID
LATITUDE(DEC.) LONGITUDE(DEC.)
39.07N 84.68U
THE PROCESS(ES) MPRM ANTICIPATES TO PERFORM ARE
EXTRACT AND QUALITY ASSESSMENT
EXTRACT OUTPUT-
QA OUTPUT
OPEN: IQAOS001.DAT
OPEN: OQAOS001.DAT
THE EXTRACT DATES ARE:
STARTING: 1-JAN-64
ENDING: 4-JAN-64
Figure 4-11. First page of general report, for example where
on-site data are extracted and submitted for
quality assessment.
4-29
October 1994
-------
METEOROLOGICAL PROCESSOR FOR REGULATORY MODELS (MPRM), VERSION
TODAY'S DATE AND TIME: H-APR-88 AT 09:56:04
INITIAL PROCESSING ON RAW METEOROLOGICAL DATA
********************************************************
*** JOB TERMINATED NORMALLY ***
********************************************************
**** MPRM MESSAGE SUMMARY TABLE ****
0- 9 10-19 20-29 30-39 40-49 50-59 60-69 70-79
JB
E 0
U 0
I 0
OS
. E 0
U 0
I 0
Q 0
0
0
0
7
0
2
0
0
9
0
0
0
0
0
0
0
0
**** WARNING MESSAGES
o os
0 OS
****
W15 AUTCHK:
W15 AUTCHK:
MHGT
SE
ERROR MESSAGES
NONE
...
0
0
0
0
0
0
0
0
****
NOT IN
NOT IN
****
0
0
0
0
0
0
0
0
INPUT
INPUT
0
0
0
0
0
141
35
176
LIST: AUDIT
LIST: AUDIT
0
0
0
0
0
0
0
0
DISABLED
DISABLED
0
0
0
0
0
0
0
0
1.2
TOTAL
0
0
7
0
3
141
35
185
Figure 4-12.
Second page of general report, for example where
on-site data are extracted and submitted for
quality assessment.
4-30
October 1994
-------
METEOROLOGICAL PROCESSOR FOR REGULATORY MODELS (MPRM), VERSION 1.2
TODAY'S DATE AND TIME: H-APR-88 AT 09:56:06
********************************************************
***
JOB TERMINATED NORMALLY ***
********************************************************
THERE
IS NO AUDIT
SITE VECTORS |
TOTAL
**** SUMMARY OF THE QA AUDIT ****
TRAIL FOR SITE SCALARS
......wini AT inu ci UIUADY. __-__) I _ _ _ . .TCCT WAI I icc
# LOWER UPPER X MISSING LOWER
# DBS MISSING BOUND BOUND ACCEPTED FLAG BOUND
10.00
US
30.00
SA
TT
WD
US
M
95
M
95
95
95
95
THIS CONCLUDES THE
200 97.89 -999.0, 0.0,
200 97.89 -999.0, 0.0,
200 97.89 -999.0, -30.0,
200 97.89 -999.0, 0.0,
200 97.89 -999.0, 0.0,
AUDIT TRAIL
1
1
UPPER
BOUND
50.0
35.0
35.0
360.0
50.0
Figure 4-13.
Final page of general report, for example where
on-site data are extracted and submitted for
quality assessment.
The message summary table (Figure 4-12) states that there
are 176 messages generated concerning the on-site variables
with error codes in the range of 50 to 59. From an inspection
of the error message file, we quickly see that most of these
messages resulted from range check violations involving the
temperature differences. Figure 4-14 presents a partial
listing of these messages.
4-31
October 1994
-------
29 JB 119 SETUP: ENCOUNTERED END OF "JOB/RUN CARD"
0 JB 110 TEST: SUMMARY: NO UA-EXT CARD, NULL EXTRACT
0 JB 111 TEST: SUMMARY: NO UA-IQA CARD, NULL QA
0 JB 112 TEST: SUMMARY: NO UA-OQA CARD, NULL MERGE
0 JB 110 TEST: SUMMARY: NO SF-EXT CARD, NULL EXTRACT
0 JB 111 TEST: SUMMARY: NO SF-IQA CARD, NULL QA
0 JB 112 TEST: SUMMARY: NO SF-OQA CARD, NULL MERGE
0 OS U15 AUTCHK: MHGT NOT IN INPUT LIST: AUDIT DISABLED
0 OS U15 AUTCHK: NRAD NOT IN INPUT LIST: AUDIT DISABLED
0 OS U15 AUTCHK: SE NOT IN INPUT LIST:
100 OS 158 REALCK: UB: 10.00 DT01:
100 OS 158 REALCK: UB: 10.00 DT01:
100 OS 158 REALCK: UB: 10.00 DT01:
101 OS 158 REALCK: UB: 10.00 DT01:
100 OS 058 REALCK: UB: 10.00 DT01:
100 OS : ON 64/01/01/00
101 OS Q58 REALCK: UB: 10.00 DT01:
101 OS : ON 64/01/01/01
106 OS 157 REALCK: LB: -2.00 DT01:
107 OS 157 REALCK: LB: -2.00 DT01:
107 OS 157 REALCK: LB: -2.00 DT01:
107 OS 157 REALCK: LB: -2.00 DT01 :
107 OS 157 REALCK: LB: -2.00 DT01:
108 OS 157 REALCK: LB: -2.00 DT01:
107 OS 057 REALCK: LB: -2.00 DT01:
414 OS : ON 64/01/04/14
415 OS 157 REALCK: LB: -2.00 DT01:
416 OS 157 REALCK: LB: -2.00 DT01:
417 OS 157 REALCK: LB: -2.00 DT01 :
1128 OS 159 OSFILL: FOUND EOF FILE
AUDIT DISABLED
13.96
15.40
12.90
13.57
14.09
10.05
-2.96
-3.55
-3.77
-3.58
-4.09
-4.40
-3.75
-4.41
-2.60
-2.16
Figure 4-14.
Partial listing of error/message file generated
during the processing of on-site data.
Inspection of these messages suggest that the reported
temperature differences sometimes exceed the range check
bounds. These messages are best understood in light of how the
on-site data are extracted and processed. As the data are read
in, checks are made to determine if the data are within the
dates specified for extraction. Once it is determined that the
data are acceptable for inclusion within the extracted data
set, each variable is passed through the range checks. In the
example, this would occur for each of the 15-minute periods.
Any range check messages generated during this phase of the
processing are preceded with a number which is a concatenation
of the Julian day and hour. For instance, the 100 at the
beginning of the range check message for a Delta-T value of
13.96 (refer to Figure 4-14) stands for Julian day 1 hour 00.
The message is not followed with a continuation line supplying
4-32
October 1994
-------
a complete date. This, in combination with the I on the
message code, is an indication that the message is for one of
the 15-minute observations within hour 00. When it is detected
that an observation has been encountered that belongs to a
succeeding hour, the processor computes the hourly averages for
the current hour. The processor then passes each hourly
average through the range checks. In summary, the range check
messages can result from either the within hour observations or
the hourly averages, or both. Having messages from both
provides a rapid means for finding suspect values within an
hour, which may in turn result in suspect hourly values.
STAGE 2 PROCESSING
o Combine into one file the available on-site and NWS
meteorological data files created during Stage 1
processing
o Store the data in a more compact format.
Having retrieved (extracted) the data from the original
storage files and having processed these data through the
quality assessment checks, the next major activity is to
combine the data into one file in a compact data storage
format.
4~33 October 1994
-------
The run stream input for Stage 2 processing is
JB STA
JB OUT DISK REPORT.LIS
JB ERR DISK ERROR.LIS
JB FIN
UA STA
UA OQA DISK OQAUA002.DAT
UA FIN
SF STA
SF OQA DISK OQASF002.DAT
SF FIN
OS STA
OS OQA DISK OQAOS001.DAT
OS LOG 9381464 84.68W 39.07N 0
OS FIN
MR STA
MR OUT DISK MERGE1.DAT
MR EXT 63 12 31 64 01 04
MR FIN
No new input images are encountered in this example,
however, it does illustrate for the first time the merging of
data from all three pathways. The JB-pathway input should be
familiar by now. It defines the general files to be used for
listing the error messages and summarizing the processing
results. In this example, there are data available for all
three pathways, UA, SF, and OS. These data files were created
during steps 2, 4, and 5 in the Stage 1 examples discussed
previously.
The three data files OQAUA002.DAT, OQASF002.DAT, and
OQASOS001.DAT will be combined (merged) into one data file
named MERGE1.DAT. The MR OUT input image is used to provide
the name of this file.
The MR EXT input is optional. Specifying start and stop
dates to be associated with Stage 2 processing, this input
provides a means for selecting the data to be included in the
4-34
October 1994
-------
merged data file. This might be useful if one wanted to create
a merged data file for a specific few days or months. If the
MR EXT input is omitted, MPRM searches the OQA files on each
pathway for the earliest date. This date becomes the first day
of merged data. The last day of merged data acceptable is then
defined as the first day plus 367. In other words, omission of
the MR EXT input allows up to 368 days of data to be merged,
starting from the earliest date encountered on all the input
files.
The second page of the general report is shown in Figure
4-15. The time period for which merged data is created and the
site latitude and longitude to be used by the regulatory
dispersion models follow the standard MPRM general report
header.
METEOROLOGICAL PROCESSOR FOR REGULATORY MODELS (MPRM), VERSION 1.2
TODAY'S DATE AND TIME: 14-APR-88 AT 09:59:35
INITIAL PROCESSING ON RAW METEOROLOGICAL DATA
***** USER INPUT PARAMETERS FOR MERGE *****
MERGED DATA BEGIN (YR/MO/DA) 63/12/31
AND END 64/1/4
THE ON-SITE LATITUDE AND LONGITUDE ARE:
LATITUDE = 39.07N
LONGITUDE = 84.68W
***** DAILY OUTPUT STATISTICS *****
MO/DA 12/31 1/1 1/2 1/3 1/4
NWS UPPER AIR SDGS 24444
NCDC MIXING HEIGHTS 46666
NUS SFC OBSERVATIONS 1 24 24 24 23
ON-SITE OBSERVATIONS 1 24 24 24 22
UPPER AIR OBS. READ: 10
SFC. OBS. READ: 96
ON-SITE OBS. READ: 95
***** MERGE PROCESS COMPLETED *****
Figure 4-15. Second page of general report, for example where
data from all three data pathways are combined
(merged).
4-35
October 1994
-------
The daily output statistics shown in Figure 4-15 are
important. Provided in this summary are the number of NWS
upper air soundings, NCDC mixing heights, NWS hourly surface
observations and on-site observations that were combined for
each day. The are a maximum of 6 NCDC mixing heights available
for each day, as Stage 2 processing combines the mixing heights
for the current day with previous and next day's values. This
allows development and use of interpolation routines for
developing hourly estimates of mixing height. We have provided
for a maximum of 24 soundings per day. As with the treatment
of mixing heights, the last sounding from the previous day and
the first sounding from the next day, if available, are
combined with the current day's soundings, unless there are 24
values present for the current day. This allows development
and use of interpolation routines for developing hourly
estimates of upper air soundings. We have provided for a
maximum of 24 observations of NWS surface observations and on-
site observations. If an entire year's data were merged, the
daily statistics summary reguires five pages of printed output.
The last item of the summary is the number of observations read
from each input (OQA) file.
STAGE 3 PROCESSING
o Create a meteorological data file for use with a
regulatory dispersion model chosen by the user.
At this point, summaries have been constructed reporting
suspect and missing data values (Stage 1) and these
meteorological data have been stored in a combined format
(Stage 2). The goal of Stage 3 processing is to construct a
meteorological data file for use with a particular dispersion
model given the merged meteorological data file resulting from
Stage 2 processing.
4~"36 October 1994
-------
As a minimum, Stage 3 processing requires the file name of
the input (merged) data file and the output file name for use
by the dispersion model. There are various methodologies for
generating the output meteorological data file. The RAMMET
format has been selected as the default output. This is an
unformatted file containing a header record and one record for
each 24-hour period. The default selections for processing
wind, temperature, stability category, and mixing height employ
the NWS hourly surface weather observations and the NCDC twice-
daily mixing height values. A full discussion of the
processing assumptions is given in Section 5. In essence,
processing with the default selections duplicates that
performed by the RAMMET meteorological processor.
The run stream input for Stage 3 processing, with the
default dispersion model, is:
JB STA
JB ERR DISK ERROR.LIS
JB OUT DISK REPORT.LIS
JB FIN
MP STA
MP MET DISK MERGE1.DAT
MP MMP DISK MPOUT.DAT
MP FIN
The optional controls provided for Stage 3 processing fall
within one of the following three groups:
o Options for selecting the dispersion model
o Options for altering the methods employed in generating
the output meteorological file
o Options for altering the reporting procedures.
October 1994
-------
Selecting Dispersion Models
Included as an option within the MP MHP input is the
ability to define the dispersion model. This was discussed in
Section 3. Appendix F provides specific details regarding the
various output formats. If no dispersion model name is
included in the MP MMP input image, the default selection is
CRSTER, which produces RAMMET output.
Selecting Processing Methodologies
The MP VBL (optional) input image is provided for
controlling the methodologies employed in generating the output
meteorological data. The syntax is outlined as:
HP VBL ITEM ACTION XXXX X
L
Input needed only with
Item "STAB" and Action
"USERIN"
Additional input,
as required
6-character keyword
instructing processor
on how ITEN is to be
accomplished
4-character keyword
defining process
(Methodology)
Table 4-1 summarizes the keywords available for ITEM and
ACTION, as well as the additional input that is sometimes
required.
4-38
October 1994
-------
TABLE 4-1. SUMMARY OF MP VBL INPUT KEYWORDS ITEM AND ACTION
Item
Action
Description
WIND
NUSWXX
ONSITE
(Default). The wind direction and speed are determined
using the wind direction and speed given in the hourly NUS
weather observation. As the observations are reported to
the nearest 10°, a standard set of random numbers is
used, as in RAHMET, for randomizing the wind directions.
Additional input is a value for STKHGT. Wind direction
and speed are determined from on-site observations. Values
selected are from on-site tower level nearest to value
given for STKHGT.
TEMP
NUSUXX
ONSITE
(Default). The temperature is determined using the values
given in the hourly NUS weather observation.
Additional input is a value for TMPHGT. Temperature is
determined from on-site observations. Values selected are
from on-site tower level nearest to value given for TMPHGT.
MHGT
NUSUXX
ONSITE
(Default). The mixing height is determined using the NCDC
twice-daily mixing height values and the stability category
for the hour. The procedure is that employed in RAMMET.
The mixing height given in the hourly on-site observation
is used.
STAB
NUSUXX
ONSITE
SESITE
SASITE
USER IN
UNDUXX
TTDIFF
(Default). The stability category is determined using the
cloud cover, ceiling height, and wind speed (from NUS
weather observations) coupled with sun's position. The
procedure is that employed in RAMMET.
Additional input is a value for ANEHGT. The stability
category is determined using the same procedure as NUSUXX,
with all data taken from the on-site observation. Wind
speed values are from on-site tower level nearest to value
given for ANEHGT.
Additional input is a value for ANEHGT. The stability
category is determined using standard deviation of vertical
wind direction fluctuations at tower level nearest to value
given for ANEHGT.
Additional input is a value for ANEHGT. The stability
category is determined using standard deviation of
horizontal wind direction fluctuations at tower level
nearest to value given for ANEHGT.
Additional input is a value for ANEHGT and [ISCVAR =1,2,
or 31. This action indicates that the user is providing
the stability directly in the variable US01 (when ISCVAR
= 1).
Additional input is a value given for ANEHGT. The
stability category is determined using on-site wind speed
from tower level nearest to ANEHGT, and NUS observations
of cloud amount and ceiling height.
Additional input is a value for ANEHGT. The stability
category is determined using the solar radiation
temperature difference method.
Selecting Reporting Procedures
The default reports have been described in the examples
given above. The user has two means to amplify the information
reported. First, the meteorological data stored in the output
file for the dispersion model can be listed to the general
report file. This is accomplished by use of the JB LST input.
4-39
October 1994
-------
Examples of such listings are provided within Appendix D.
Second, more detailed messages can be obtained within the
error/message file. This is accomplished by use the UA TRA, SF
TRA, OS TRA, or MP TRA inputs.
Figure 4-16 presents a partial listing of an error/message
file with the MP TRA input. For Julian day 365 hour 23, it
shows there was insufficient data to compute a stability
category. The reason why each scheme was unable to compute a
stability category is listed. In order to understand the
messages, the user must know which subroutines are used to
compute stability category. Table 4-1 presented the keywords
used to select the various processing options; Table 4-2
presents a list of the subroutines that are employed as a
function of the processing options selected.
30
8
0
0
0
365
365
365
365
365
365
365
365
365
365
365
365
365
365
JB
JB
OS
OS
OS
JB
MP
MP
MP
MP
MP
MP
MP
MP
MP
MP
MP
MP
MP
103
119
U15
W15
U15
U06
T75
T76
T75
T75
T76
T75
T75
T75
T75
T75
T75
140
W75
MPPROC:
MPSTUP:
AUTCHK:
AUTCHK:
AUTCHK:
HTKEY :
OS1PGT:
ROUGH:
OSSEPG:
OSSEPG:
ROUGH:
OSSAPG:
OSSAPG:
OS2NUS:
SFSTAB:
OS1PGT:
OSSEPG:
SFSTAB:
SFSTAB:
IOSTAT= 3047 REACHED END OF HEADERS
FOUND "END OF FILE" ON DEVICE DEVIN 5
MHGT NOT IN INPUT LIST: AUDIT DISABLED
NRAD NOT IN INPUT LIST: AUDIT DISABLED
SE NOT IN INPUT LIST: AUDIT DISABLED
NO TT AT LVL01. TMPHGT IS: 2.0
ISPD MISSING, HR 23 PGSTAB .EQ. 0
NO UD FOR ZO, HOUR 23
ZO MISSING FOR HOUR: 23
ISPD MISSING, HR 23 PGSTAB .EQ. 0
NO UD FOR ZO, HOUR 23
ZO MISSING FOR HOUR: 23
ISPD MISSING, HR 23 PGSTAB .EQ. 0
ISPD MISSING, HR 23 PGSTAB .EQ. 0
HOUR 23 NO PG CATEGORY POSSIBLE
OSTSKY MISSING, HR 24 PGSTAB .EQ.
SE & SW MISSING, HR 24 PGSTAB .EQ
HOUR 24 USED STAB. SCHEME SASITE
23 HOURS HAVE 0 FOR PG CATEGORY
0
. 0
Figure 4-16. Partial listing of the error/message file, for
example with trace option enabled.
4-40
October 1994
-------
TABLE 4-2.
SUBROUTINES EMPLOYED FOR
VARIOUS PROCESSING OPTIONS
ITEM
WIND
TEMP
MHGT
STAB
ACTION
NWSWXX
ONSITE
NWSWXX
ONSITE
NWSWXX
ONSITE
NWSWXX
ONSITE
SESITE
SASITE
WNDWXX
TTDIFF
USERIN
Subroutine
WS1NWS
WS1OS
TT1NWS
TT1OS
ZI1NWS
ZI1OS
PGTNWS
OS1PGT
OSSEPG
OSSAPG
OS2PGT
OSSRPG
OSINPG
Shown in Figure 4-16 is a portion of an error/message file
developed from a test case example having on-site data, in
which a
MP VBL STAB ONSITE 10
image was used to define the methodology for computation of the
stability categories. As explained in Section 5, an automatic
search is performed in computing stability category. Hence,
when subroutine OS1PGT (refer to Table 4-2) detected a missing
on-site wind speed at the designated ANEHGT, subroutine OSSEPG
was called to attempt computing the stability category using
the SESITE option. In this manner, each of the methods
employing on-site data were attempted for hour 23 to compute
the stability category. For hour 24, the search ended in
subroutine OSSAPG, where it was determined that sufficient data
were present to employ the SASITE option for computing
stability category.
4-41
October 1994
-------
SECTION 5
SCIENTIFIC NOTES
INTRODUCTION
This section provides a brief technical description of the
methods employed by MPRM during processing. All of the methods
are documented in the indicated references to which the reader
is referred for details (EPA, 1986; EPA, 1987). The methods
employed during Stage 1 (Extraction and Quality Assurance) are
described first, this is followed by descriptions of the
methods employed in Stage 3 processing.
STAGE 1
Averaging Subhourly Values
The initial data file of on-site data may consist of
subhourly values. The number of observations per hour is
specified with the optional run stream input OS AVG, which can
be used to specify the number of on-site observations expected
for an hour. The default value is one observation per hour.
During the extraction processing of the on-site data, the
subhourly values are averaged to produce hourly values. For
most variables the hourly value is computed as the arithmetic
mean. The wind speed and the wind direction are treated in a
special manner, in order to properly differentiate between
5-1 Octch >: 1994
-------
cases when values are missing and cases when values are present
but suspect since they are below instrument threshold.
The threshold wind speed by default is 1.0 m/s, but can be
redefined (but only to a lower value) through the use of the OS
CLM run stream input image. Wind speeds less than threshold
are given a value of one-half the threshold wind speed and the
wind direction is treated as missing. The hourly wind speed is
computed as an arithmetic mean. The hourly wind direction is
computed according to the method given in Section 6.1 of the
On-Site Guidance Document (EPA, 1987) to properly account for
the 0-360 degree crossover.
For all standard deviations of wind fluctuations (both
speed and direction), hourly values are computed as the
root-mean-square of the subhourly values, in accordinance with
the recomendations of the On-Site Guidance Document (EPA,
1987).
Quality Assurance
In Stage 1 processing, QA is performed on each pathway by
comparing data values to the upper and lower bounds defined for
each variable. Default QA bounds are defined in Appendix C,
but these values can be overridden by the user with the CHK
input image. The endpoints of the interval, i.e., the boundary
values, are either included as acceptable data or excluded as
questionable data according to the Range Check Switch field.
This parameter can also be changed with the CHK input image.
The default value of the Range Check Switch for each variable
is also defined in Appendix C.
Prior to performing the quality assessment on upper air
soundings, the processor recomputes the heights reported in the
5-2 October 1994
-------
soundings using the hypsometric formula. If the surface height
is missing, the heights are not recomputed.
Because the upper air soundings contain multiple levels of
data, vertical gradients of several variables can also be
checked. This poses a guestion of how to report the audit
results as there are a variable number of levels in a sounding
and the heights of the levels differ from sounding to sounding.
Our solution was to define ten height categories into which we
accumulate the results. The categories are defined as the
surface (the first level alone), eight height intervals, and
all heights above the uppermost interval. The interval
thickness is defined to be 500 meters. The eight height
intervals are, therefore, 0 - 500, 500 - 1000, ..., 3500 - 4000
meters. The lower bound is excluded and the upper bound is
included in each interval. All heights above 4000 meters are
placed in the last category. The interval thickness is
controlled by the variable UAINC, which is specified in a DATA
statement and cannot be changed without changing the value in
the DATA statement, recompiling and relinking the processor.
The gradient of temperature, or lapse rate, between two
levels in the upper air data is checked against an upper and
lower bound. The default maximum value is 5 °C/100 meters and
the default minimum is -2 °C/100 meters. Note that these
values are expressed as integer values per 100 meters and any
changes to the default values, via UA CHK input image, must
also be expressed in degrees C/100 meters. The variable name
for use with CHK to alter the range check parameters for the
temperature lapse rate is UALR.
Wind velocity has two components: speed and direction.
The vertical gradient of the wind, or shear, can be expressed
either as a vector shear in which both speed and direction are
combined to yield one shear value, or speed and direction
5~3 October 1994
-------
separately. The wind speed and direction shear are expressed
separately in MPRM. The default maximum wind speed shear is 5
(m/s)/100 meters. Because the absolute difference of the speed
is considered, i.e. the computation is independent of which
level has the higher wind speed, the default minimum speed
shear is 0 (m/s)/100 meters. The variable name for use with
CHK to alter the range check parameters for the wind speed
shear is UASS. The default maximum wind direction shear is 90
degrees/100 meters and the minimum is 0 degrees/100 meters.
The directional shear is independent of which way the wind
changes with height (i.e., clockwise or counterclockwise). The
variable name for use with CHK to alter the range check
parameters for the wind direction shear is UADS.
The vertical gradient of dew-point temperature is treated
a little differently than the lapse rate and wind shear. Three
consecutive values are required for this evaluation. An
estimate from the line drawn between the upper and lower points
is made at the height of the middle point. The absolute
difference between the estimate and the actual dew-point is
divided by the height difference between the upper and lower
points. This value is compared to the upper and lower bounds
defined by UADD. The default upper and lower bounds are 2
°C/100 meters and 0 °C/100 meters, respectively-
The violations for the lapse rate and shear are tallied
into the height category containing the upper height, whereas
the dew-point violations are tallied into the height category
containing the middle point. Therefore, if none of the data
required to perform the gradient calculations are missing, then
there are (N-l) checks on lapse rate and shear and (N-2) checks
on dew-point, where N represents the number of levels in a
sounding.
5-4
October 1994
-------
Upper Air Data Modification
During the extraction of upper air soundings in Stage 1
processing, the MPRM processor can check the data for possible
errors in each sounding and perform certain automatic data
modifications. These actions are an option controlled by the
user with the UA OFF input image. If this image is omitted,
then the soundings are checked for automatic substitution. If
this image is present, then the soundings are NOT checked for
modification. The syntax of this input image is simply UA OFF.
No other fields are required.
Data modifications can occur for:
o The sign of temperature above 1000 meters
o The lapse rate between two levels
o A mandatory level within 1 percent of a significant level
o A zero wind speed with a corresponding non-zero wind
direction
o Missing values of dry bulb temperature and dew-point
temperature
o Height above mean sea level.
There is no way to turn on and off individual actions. Either
all the actions are performed or none of them are performed.
The warning messages that are written if the data are
modified include the date and time. The format is YYMMDD/HH
where YY = 2-digit year, MM = month (1-12), DD = day (1-31) and
HH = hour (1-24).
Temperature above 1000 meters
For heights above 1000 meters above ground level (AGL),
temperatures greater than 10 °C are checked to insure they have
5-5
October 1994
-------
the correct sign. The sign of the temperature immediately
below the level in question is checked. If the sign is
negative, then the sign of the temperature at the level in
question is changed to negative; if the sign is positive or if
the temperature is missing, then no action is taken. This
action was introduced to reduce the number of sign errors that
occur in the TD-5600 format data. The primary emphasis is on
levels away from the surface where it is obvious that the signs
were in error. No attempt is made here to correct the data
near the surface. Checks of temperatures fluctuating around 0
°C are also avoided, where the temperature can switch signs
from one level to the next, by only considering temperatures
greater than 10 °C.
Lapse rate
If the temperature lapse rate between two levels is
superadiabatic, i.e., less than -0.0098 °C/meter, and the lower
level temperature is greater than 0 °C, the sign of the lower
level temperature is changed. This sign change is not
performed if it would create a temperature inversion greater
than the maximum defined with the keyword UALR, i.e., the lapse
rate upper bound for quality assessment (see Appendix C for
default value).
Mandatory levels
If a mandatory sounding level is within one percent of a
significant level (with respect to pressure) then the mandatory
level is deleted. Because the mandatory levels were originally
computed from the data at the significant levels, there is no
loss of information in the sounding.
Deleting mandatory levels occurs after the data are
retrieved from tape, which results in reducing the number of
5~6 October 1994
-------
levels in a sounding. If the maximum number of levels is
retrieved, then this process will produce a sounding with fewer
than the maximum number of levels, i.e., the processor does not
return to the tape to retrieve additional levels. At present
the maximum number of levels that can be retrieved is 20.
Calm wind conditions
The wind speed and direction at each level are checked to
insure that there are no levels with a zero wind speed and a
non-zero wind direction. If one is found, the wind direction
is set to zero to represent calm conditions.
Missing dry bulb and dew-point temperatures
If temperature or dew-point at a level is represented by a
missing value indicator, then an estimate for the missing
observation is made by linearly interpolating to the level in
question. The data from the level immediately below and the
nearest valid data from above the level in question are used.
If data that are required for the interpolation are also
missing, then no interpolation is performed.
Sounding heights
The sounding heights on magnetic tape are stored as meters
above mean sea level. With the sounding modification actions
enabled, the heights are converted to meters above ground
level. The first level in a sounding is for the surface. The
height at this level is subtracted from all levels including
the surface, so that the heights start at 0 meters. If the
height is missing at the surface, then a value of zero is
assumed in performing the subtractions.
October 1994
-------
STAGE 3
Wind
The default method for handling wind data by MPRM is to
use NWS surface observations. The data are hourly and result
from observations made at a single level. No input is required
for implementation. If the data are missing, then a missing
value indicator is written to the output file, and a message is
inserted in the error/message file to warn the user that this
has occurred.
Calm winds are handled somewhat differently for dispersion
models requiring hourly specification of the meteorological
conditions versus those requiring STAR output. When processing
NWS surface observations for hourly output, if the wind speed
is less than 1 m/s, then the speed is set to 1 m/s and the wind
direction is set equal to the last valid wind direction value.
When processing on-site observations for hourly output, if the
wind speed is less than threshold (OSCALM) , then the speed is
set to 1 m/s and the wind direction is set equal to the last
valid wind direction value. If the on-site speed is less than
1 m/s but greater than or equal to threshold, the speed is set
to 1 m/s. For those models requiring STAR output, the
occurrences of calm winds are distributed within the lowest
wind speed classes in accordance with the directional
distribution of non-calm observations within these classes.
For on-site data during Stage 1 processing, the user
specified the data available and the format of the data
(through the OS MAP input), which may have included several
levels of wind data. The current list of dispersion models
accommodated by MPRM process wind at only one level for use
primarily in estimates of plume rise and transport. The
current guidance recommends use of wind data at the height of
5~8 October 1994
-------
the stack top. Hence, to implement use of on-site data in
defining the the wind, the user must identify the measurement
level to be used in defining the wind. The value given on the
MP VBL input is stored within MPRM in a variable called STKHGT.
MPRM determines the measurement level closest to the value of
STKHGT. This measurement level is then used in all
determinations of wind. In defining the measurement level,
MPRM interrogates the OS MAP input images to insure that wind
data are available at the tower level nearest to the value
given for STKHGT. Processing is stopped if the data map of
on-site variables indicates that no wind data are available on
the tower level nearest to the value given for STKHGT. Since
any value greater than zero for STKHGT is allowed, the user can
specify the level that is best for the given dispersion
analysis. This may or may not be the level nearest to the
actual height of the stack being modeled, as the data may not
be representative for one reason or another, or more likely,
the data may not be sufficiently complete at all levels to
satisfy the guidance recommendations of a 90 percent data
capture rate.
Consider the case of on-site wind data being available at
10, 60 and 100 m on a meteorological mast, and the stack height
to be modeled is 300 m. The user could include as input during
Stage 3 processing:
MP VBL WIND ONSITE 300
MPRM would then process the 100 m wind data, the level closest
to 300 m, for generating the modeling output file. What would
happen if the 100 m data were only available during a small
portion of the period to be analyzed? Perhaps, the instruments
at 100 m were damaged and replacement was delayed. Then the
input could be modified to force MPRM to select the wind data
5~9 October 1994
-------
at a lower level where the observation record is more complete
as:
MP VBL WIND ONSITE 60.
Temperature
The default method for processing temperature data by MPRM
is to use NWS surface observations. The data are hourly and
result from observations made at a single level. No input is
required for implementation. If the data are missing, then a
missing value indicator is written to the output file, and a
message is inserted in the error/message file to warn the user
that this has occurred.
For on-site data during Stage 1 processing, the user has
specified the data available and the format of the data
(through the OS MAP input), which may have included several
levels of temperature. The current list of dispersion models
accommodated by MPRM only process temperatures at one level.
To implement use of on-site data in defining the temperature,
the user must identify the measurement level to be used within
the on-site observations having temperature data. The value
given on the MP VBL input is stored within MPRM in a variable
called TMPHGT. MPRM determines the measurement level closest
to the value of TMPHGT. This measurement level is then used in
all determinations of temperature. In defining the measurement
level, MPRM interrogates the OS MAP input to insure that
temperature data are available at the tower level nearest to
the value given for TMPHGT. Processing is stopped if the data
map of on-site variables indicates that no data are available
on the tower level nearest to the value given for TMPHGT.
No explicit height is recommended within the guidance on
processing on-site meteorological observations for use in
5-10 October 1994
-------
regulatory dispersion models. Given the usage made of
temperatures within the dispersion models currently accepted
for regulatory use, we believe the temperature should be
representative of what is typically reported within the NWS
hourly observations, which is a near-surface value.
As with defining the value of STKHGT for use in specifying
the wind, any value greater than zero for TMPHGT is allowed.
Thus, the user can specify the level that is best for the given
dispersion analysis.
Consider the case of on-site temperature data being
available at 10 and 100 m on a meteorological mast. The user
could include as input during Stage 3 processing:
MP VBL TEMP ONSITE 2
MPRM would then process the 10 m temperature data, the level
closest to 2 m, for generating the modeling output file.
Stability Category
There are five options in MPRM for processing data to
obtain Pasquill stability categories, a user defined option and
an option which is not yet accepted methodology for regulatory
applications. The default option is to use the method of
Turner (1964) as implemented within RAMMET. This method
employs wind speed, cloud cover and ceiling height data from a
nearby NWS surface station. Opaque sky cover is preferred, but
total sky cover may also be employed. This method is described
in the guidance document on processing on-site meteorological
observations for use in regulatory applications (EPA, 1987).
The next four methods require at least some on-site data.
The methods, in order of preference, are:
5~H October 1994
-------
1. Turner's (1964) method using on-site data, which include
sky cover, ceiling height and near-surface ("10 m) wind
speed data
2. a. and wind speed (u) from on-site data (cr. may be
determined either from direct observation of elevation
angle fluctuations, or from the transformation o. « ff,/u)
3. a, and wind speed from on-site data
4. Turner's (1964) method using on-site wind speed data
coupled with sky cover and ceiling height observations
from a nearby NWS station.
These methods are described more fully in the guidance
document on processing on-site meteorological observations
(EPA, 1987). The wind speed and turbulence data used in
stability category determinations should be taken from a
near-surface level, nominally 10 meters.
The user identifies the measurement level to be used in
the on-site observations within the MP VBL input. The value
given on the MP VBL input is stored within MPRM in a variable
called ANEHGT. MPRM determines the measurement level closest
td the value of ANEHGT. This measurement level is then used in
all determinations of stability categories for definition of
wind speed and turbulence values.
All four of the on-site methods are acceptable, according
to the modeling guidance. If the data necessary for the method
chosen by the user are missing, MPRM has been designed to cycle
through the methods, in the order of preference. For example,
the user may specify the o, method (method 3) as the primary
method for stability category determination. If o, data are
missing, MPRM will first attempt to determine stability from
5-12 October 1994
-------
method (1) above, and if the requisite data are missing for
method (1) , then MPRM will attempt to employ method (2) , and if
need be method (4). If the stability category can not be
determined by any of the four methods, then it is encoded as
missing for that hour. MPRM maintains a count of the number of
hours of data processed for each of the methods and includes
this information in the general report file. Specific
instances of stability substitutions can be traced through the
error/message file, by including the MP TRA input image during
Stage 3 processing. Note, since all four on-site methods
require near-surface wind speed data, if the wind speed is
missing, then the stability category will be considered
missing. Furthermore, whenever the a. method is attempted, the
transformation using o. is automatically invoked if MPRM
detects that a. is missing for the hour.
The default method for stability category determination,
which relies solely on NWS hourly observations, is excluded
from use if the primary method of choice is one of the methods
that employs on-site data. Since the default method and
method (4) differ only in location of wind speed data, use of
the default methodology, when on-site data methods are the
primary methods of choice, is equivalent to data substitution
of off-site wind speed data for missing on-site wind speed
data. As mentioned earlier in the first section, we have
avoided automatic data substitutions until there is a clear
consensus as to how best to implement such procedures.
5-13 October 1994
-------
Mixing Height
The default method for processing mixing height is to use
the interpolation scheme employed in the RAMMET meteorological
processor, which uses the twice-daily mixing heights from the
nearest NWS upper air observation site, coupled with the
stability category determined for the hour. This method is
described in more detail in the RAM model user's guide
(Catalano et al., 1987).
The user may also designate the on-site mixing height to
be employed. MPRM simply uses the value for the hour given in
the on-site observation.
Roughness Length
MPRM allows the user to specify .a direction dependent
surface roughness length for up to twelve wind direction
sectors. As discussed earlier, the roughness length values can
be varied by month of year as well as wind direction sector.
The roughness length values are incorporated into the header
information of the OQA-file created while processing the
on-site data. These header records are in turn incorporated
into the header information of the combined (merged) file
created during Stage 2 processing.
The roughness length is used in Stage 3 processing to
adjust the
-------
1987) for recommendations on estimating site-specific roughness
lengths.
Output Formats
The output format of the meteorological file generated
during Stage 3 processing can be characterized into two
classes, short term (hourly) and long term (seasonal or
annual). The hourly formats available are the RAMMET format,
used by several models, the CALINE-3 format, and the RTDM
format. The long term formats are joint frequency functions
also know known as STability ARrays (STAR) . The STAR formats
for the ISCLT and VALLEY models employ six stability categories
(A, B, C, D, E, F). The STAR formats for the CDM-2.0 model
employ six stability categories (A, B, C, D-day, D-night, E-F) .
Appendix F describes in more detail the various output formats.
5-15 October 1994
-------
SECTION 6
PROGRAM NOTES
INTRODUCTION
The MPRM processor is designed and coded to allow it to
run on most operating systems. The processor is coded entirely
in FORTRAN. Although this increases the amount of time for I/O
operations (for example, reading data from magnetic tape is
greatly increased), the user does not have to make major
modifications to the code to create an executable processor.
This section describes briefly the processor design, system
dependent source code, creating executable code from the source
code and some miscellaneous, but important, processor notes.
FORTRAN COMPATIBILITY AND PROCESSOR DESIGN
To retain portability, the MPRM processor was coded using
American National Standards Institute (ANSI) FORTRAN X3.9-1978
or, as it is more commonly known, FORTRAN-77. Every attempt
was made to use only the standard FORTRAN available without
additional computer specific features (extensions). However,
there are a few subroutines where these extensions became
necessary and they are described below.
The MPRM processor was designed to be used on the most
limiting machine, in terms of memory, the personal computer.
To this end, the processor is composed of units that are easily
overlayed when linking the processor with the FORTRAN languages
available for personal computer systems. The system was
compiled and tested on a PC (compatible with the IBM PC-XT)
that has 640 Kb of main memory using Ryan-McFarland
°~^ October 1994
-------
Corporation's RM/FORTRAN, version 2.10. The program processing
for Stages 1 and 2 is shown in Figure 6-1. All operations
involving the setup and writing summaries are separate from
pathway processing, which are in themselves separate from each
other.
,,,,,,,,,,,,,,,,,,, J.,
STAGES
1 & 2 .
. PROCESSING .
T,,, ,,7, ,, , , .6
,,,,,,,&,,,,,,,,
,,7,,,,,,,,,,,,,,,,,,,,
DOT
Ktliftf°»I t t tI l
Setup
* 111111 * 11111 .**
DAT
Kiiiiit°iiiiii
. Process all .
.input images .
and
. establish .
. functions .
. to perform .
F.,,,......,.,G
pftTpflTDAT
Write . . Process . . Write final .
status . . pathways . . general
of setup . . report
F.............6 F,f,fff7,.,.f,G Ff...,,.,..,,.G
*^ii,,11111,iii,,'iiiiiii^iitiiii'iiiiiiiiiitiiii*
Rft TD ft TD ft TD ft T
fi11Il°,i11ii' Kf11I Il°lI 11ii' RffffIl°iIiiii' KiiiiIl°ltI 11 I '
. Upper Air . . Surface . . On-site . . Merge
fiinii7,,,,,,G F,,,,,,7,,,(>>G f,,,,,,7,,,,,,G F,,,,, ,7,,,,, ,G
Rft TD ft TD R TD ft T
.I I IiI°.iiii' Kiiiiii°iiiiIt l KlI I I Ii°lI Ii11' *i*f*f«u».I I I I l
.NWS upper air. . NWS . .User supplied. . Combine data.
.soundings and. . surface . . on-site . . in an unfor-.
.mixing height. . observations. . observations. . matted file .
F.,.,,,,,.,,,,G F.............G F.............G F..,..,.,,,,,.6
Figure 6-1.
Flow diagram of MPRM Stage 1 and Stage 2
processing.
SYSTEM SPECIFIC SOURCE CODE
Nearly all the subroutines in the MPRM processor contain
statements to include named COMMON blocks. The named COMMON
blocks were divided into 12 separate files. The method for
incorporating these statements into the source code varies
according to the compiler in use, but all use some form of the
nonstandard INCLUDE statement. These statements appear as
extensions to VAX-11 FORTRAN (version 3.0), IBM VS FORTRAN
(version 2),RM/FORTRAN (version 2.10) and Microsoft FORTRAN
(version 4.0). The syntax for each of these is shown below.
The punctuation is required where shown except for the left and
6-2
October 1994
-------
right bracket, which indicates an optional parameter. In the
accompanying examples, the assumption is made that the INCLUDE
files are in the same subdirectory or partitioned data set as
the main programs and subroutines that use them.
VAX-11 FORTRAN
INCLUDE 'filename[/LIST]', where filename is any valid VAX
file specification and /LIST, which is optional, indicates
that the statements are to be listed in the compilation
source listing. If /LIST is omitted then no listing of
the included files appears in the compilation listing.
Example: INCLUDE 'MAIN1.INC/LIST'
IBM VS FORTRAN
INCLUDE (name) [n], where name is the member name in the
partitioned data set and n is a value used to decide
whether or not to include the file during compilation.
This parameter can be omitted, in which case the file is
included, or can take on a value from 1 to 255. If the
value of n appears in the CI compile option, then the file
is included, otherwise it is omitted. This is not the
simple list or no list option, rather it includes or
excludes the named member in the compilation.
Example: INCLUDE (MAIN1)
RM/FORTRAN, version 2.10
INCLUDE 'filename', where filename is any valid DOS file
specification.
Example: INCLUDE 'MAIN1.INC'
6~3 October 1994
-------
Microsoft FORTRAN, version 4.0
$INCLUDE:'filename' where filename is a valid DOS file
specification.
Example: $INCLUDE:'MAIN1.INC'
There are several other system dependent statements that
have been placed in separate subroutines to expedite the
transfer of source code between systems. The first group of
subroutines are concerned with the OPEN statements for disk
files and magnetic tapes. The second group contains a
subroutine for character conversion and a subroutine to return
the system date and time.
The first group of subroutines is retained in a file named
SETUPVX.FOR for VAX applications, SETUPPC.FOR for PC
applications and SETUPIBM.FOR for IBM mainframe applications.
These control the opening of disk files (subroutine FLOPEN) and
magnetic tapes (subroutine TPOPEN) and perform operations on
ASCII and EBCDIC positional codes.
The second group is retained in LIBVX.FOR (for VAX),
LIBPC.FOR (for PC) and LIBIBM.FOR (for IBM). The character
conversion subroutine (CHCONV) is used in converting EBCDIC
characters on magnetic tape to ASCII characters. It uses the
VAX octal representation of characters. This conversion is
only required on the VAX. Because magnetic tape drives are not
normally available on personal computers and because the IBM
operates with EBCDIC, a dummy subroutine has been provided for
these systems.
Also included in the second group is a subroutine (DATER)
to return the operating system date and time. The system
6-4 October 1994
-------
routines used to perform this function vary, so the correct
calls to the system routines are provided.
Within the second group is a subroutine for use on IBM
systems. The routine accesses an extended error handling
routine. This routine allows uninterrupted processing of data
files typically available from NCDC. This is discussed in more
detail under "Processor Notes". Dummy subroutines are provided
for VAX and PC applications.
CREATING EXECUTABLE CODE FROM THE SOURCE CODE
On a Personal Computer
MPRM requires nearly 540 Kb of RAM for Stage 1 and 2
processing and 500 Kb to for Stage 3 processing. For testing,
the source code was compiled using the /N option available in
RM/FORTRAN. This option uses the math co-processor (8087 or
80287) if present, but will run on a system that does not
contain the co-processor.
For Stage 1 and 2 processing, the following files are
needed:
Main program and subroutines:
STAGE1N2.FOR OSFILE.FOR
SETUPPC.FOR UAFILE.FOR
COMPLETE.FOR HEADER.FOR
SETUP.FOR
OSSETUP.FOR
LIBFILE.FOR
SFFILE.FOR
MERGE.FOR
LIBPC.FOR
Files for INCLUDE statements:
MAIN1.INC SF1.INC
OS1.INC UA1.INC
MASTER.INC WORK1.INC
MAIN2.INC
OS2.INC
SF2.INC
UA2.INC
6-5
October 1994
-------
Only the main program and subroutines have to be compiled.
The source code in the INCLUDE files is automatically included
in the compilation. An object file library should be created
using OSSETUP.OBJ, COMPLETE.OBJ, HEADER.OBJ, LIBFILE.OBJ and
LIBPC.OBJ. This library is also used in creating the Stage 3
executable code. It is recommended that an overlay structure
be used. An overlay is a subprogram that resides on disk until
needed. When an overlay is loaded into memory, it replaces any
previous overlay. The Stage 1 and 2 executable (STAGE1N2.EXE)
should be linked with the following overlay structure:
Main program and subroutines:
STAGE1N2.0BJ and object file library
SETUP.OBJ, SETUPPC.OBJ
OSFILE.OBJ
SFFILE.OBJ
UAFILE.OBJ
MERGE.OBJ
always memory resident
overlay 1
overlay 2
overlay 3
overlay 4
overlay 5
The five overlays are independent of each other and can be
linked in any order. Consult your FORTRAN manual for
additional information on overlaying files.
The Stage 3 requires the following files:
Main program and subroutines:
STAGE3.FOR MP2 XFOR.FOR
SETUPPC.FOR MP4XFOR.FOR
HEADER.FOR
SETUP.FOR
OSSETUP.FOR
MP3XFOR.FOR
COMPLETE.FOR
Files for INCLUDE statements:
MAIN1.INC SF1.INC
SF2.INC OS1.INC
UA2.INC MASTER.INC
MP1.INC
UA1.INC
WORK1.INC
MAIN2.INC
OS2.INC
6-6
October 1994
-------
The Stage 3 executable (STAGE3.EXE) should be linked with the
following overlay structure:
Main program and subroutines:
STAGES.OBJ and object file library
SETUP.OBJ, SETUPPC.OBJ, MP2XFOR.OBJ
MP3XFOR.OBJ, MP4XFOR.OBJ
always memory resident
overlay 1
overlay 2
Figure 6-2 summarizes the structure of the processor after
the executable code has been created for the three stages of
processing.
METEOROLOGICAL PROCESSOR
FOR
REGULATORY MODELS
STAGE1N2.EXE
STAGE3.EXE
STAGE 1
PROCESSING
Extract and Quality
Assess Data
STAGE 2
PROCESSING
Combine (Merge)
Data Files
STAGE 3
PROCESSING
Create Formatted Data
File for Modeling
Figure 6-2. Overview of processing stages within MPRM.
On a VAX
The following instructions assume that each file is
separate and all files reside in the same directory.
For Stage 1 and 2 processing, the following source code
files are used:
6-7
October 1994
-------
STAGE1N2.FOR SETUP.FOR SETUPVX.FOR OSSETUP.FOR
COMPLETE.FOR HEADER.FOR OSFILE.FOR SFFILE.FOR
UAFILE.FOR MERGE.FOR LIBFILE.FOR LIBVX.FOR
The INCLUDE files are the same as for the PC. Because the VAX
is a virtual memory machine, the memory restrictions imposed by
a personal computer are not applicable on the VAX. Therefore,
each file can be compiled separately (we suggest using the
/OPTIMIZE and /CHECKS=ALL options) and an executable code
created with a simple LINK statement, e.g.,
LINK STAGE1N2, SETUP, ..., LIBFILE, LIBVX.
Note that, unlike the PC, no object file library is created
with this method.
In creating the executable code for Stage 3, the same
source code used are:
STAGE3.FOR SETUP.FOR SETUPVX.FOR OSSETUP.FOR
COMPLETE.FOR HEADER.FOR MP2XFOR.FOR MP3XFOR.FOR
MP4XFOR.FOR LIBFILE.FOR LIBVX.FOR
The INCLUDE files used for the PC are also used for the VAX.
Again each file is compiled separately and the executable code
is created with a simple LINK statement.
On an IBM
The following instructions assume that each file resides
as a member in one partitioned data set. A partitioned data
set allocated with a fixed length (80 bytes), blocked record
format and a block size of 3120 bytes required 51 tracks on the
IBM 3090 at the National Computer Center (NCC) in Research
Triangle Park, NC.
6-8
October 1994
-------
The same source code files used for the VAX are used on
the IBM except that SETUPIBM and LIBIBM replace SETUPVX.FOR and
LIBVX.FOR, respectively (note that SETUPIBM and LIBIBM are
referenced as members of a partitioned data set). The INCLUDE
files are the same as for the PC and VAX. A load module is
created in one step (the code compiled and link edited) by
concatenating the data set members in one DD statement and
placed in a partitioned data set separate from the source code.
This procedure is applicable to both STAGE1N2 and STAGES.
Additional considerations for running MPRM on the IBM can be
found later in this section.
Checklist
The source code, as delivered, is ready for use on a PC.
Several changes to the source code may be required prior to
running MPRM on another system. The following is a list of
items to assist the user in preparing MPRM for another computer
system.
o Check the INCLUDE syntax (discussed earlier in this
section)
o Check how end-of-file conditions are handled and set the
logical variable BACK40 accordingly (discussed later in
this section or see comments in MASTER.FOR)
o Verify that the standard default input and output logical
device numbers are 5 and 6, respectively (discussed later
in this section)
o If upper air soundings are to be extracted from the
TD-5600 variable block format, verify that the variable
controlling the number of bytes to skip at the beginning
6~9 October 1994
-------
of each block (UABSIZ) is set correctly (see discussion in
this section or see comments in MASTER.FOR)
PROCESSOR NOTES
Standard Input and Output Devices
The MPRM processor uses logical unit 5 as the standard
input (console or keyboard) and unit 6 as the standard output
(screen) devices. These are defined in DATA statements as
variables DEVIN and DEVIO, respectively, in the master common
blocks in files MASTER.FOR. These unit numbers can easily be
changed to accommodate other operating system's default values.
End-of-file Conditions
Because end-of-files are handled differently on the PC
than on the VAX, a logical variable was defined to either
backspace or not backspace when the end-of-file was
encountered. This logical variable, BACK40, is defined in
MASTER.FOR at the end of the file. For the VAX, BACK40 should
be .FALSE, and for personal computer and IBM applications,
BACK40 should be .TRUE..
File Header Records
The MPRM processor writes pathway information as a series
of records at the top of output files used during data
extraction, data quality assessment and merging. These
records, referred to as header records, are a maximum of 83
characters in length. The first 3 characters have special
meaning to MPRM. If the additional informational header
records are added to the top of these files, the first
character should be an asterisk, *, followed by 2 blank
characters. The 2 characters following the asterisk are
6-10 October 1994
-------
reserved for use exclusively by MPRM and should not be altered,
except as altered by MPRM during processing. The only
restriction currently imposed on the last 80 characters of the
header records is that characters should be limited to
parenthesis, plus and minus sign, comma, period, numbers and
letters (upper or lower case).
Upper Air Sounding Format
Another variable whose value may require changing has to
do with extracting upper air soundings from magnetic tape. The
variable, UABSIZ, is either a 0 as currently set or 4. This
variable controls the number of bytes to skip at the beginning
of each physical block of data on the tape and pertains only to
the TD-5600 variable blocked format available from NCDC. If
the tape pre-dates the introduction of the newer TD-6200 series
formats, then UABSIZ must be set to 4. If the tape was ordered
after the introduction of the TD-6200 series format, then
UABSIZ must be set to 0.
Mounting Magnetic Tapes on the VAX
On mainframe systems such as the VAX, tape drives are
usually available. National Weather Service upper air
soundings and surface data can be retrieved from magnetic tape
using the MPRM processor. Prior to running the processor,
however, the tape must be mounted on a tape drive. One
possible sequence of VAX command (DCL) statements to do this
for the upper air soundings is shown below. Variations on this
DCL may be available or necessary.
$ ! ALLOCATE THE TAPE DRIVE
$ i
$ ALLOCATE $TAPE1
$ !
$' ! SEND A MESSAGE TO THE OPERATOR WHO MOUNTS
^ October 1994
-------
$ ! THE TAPE (QUOTES REQUIRED)
$ 1
$ REQUEST/TO=TAPES 'your message'
$ !
$ MOUNT/FOREIGN/NOWRITE/DENSITY=1600/BLOCKSIZE=8000 -
/NOLABEL $TAPE1 B20101 TAPE10
The ALLOCATE command allocates a tape drive with the
device name $TAPE1. The REQUEST command, which is optional,
sends a message to the tape operator informing him a request
for a tape to be mounted is forthcoming.
The MOUNT statement sets several parameters, but the most
important piece of information in this command is the last
parameter, TAPE10. This field is the logical name associated
with the tape and must match the name given in the run stream
on the INI image. Likewise, when NWS surface observations or
mixing height estimates are to be extracted from tape, the
logical name of the tape and the tape name parameter on the IN2
image must match (see the table in Appendix B on Defining Tape
File Names, Keywords INI and IN2, parameter Tapename).
The remaining parameters in the MOUNT statement indicate
that the tape is not a VAX system tape (FOREIGN and NOLABEL)
and not to write to the tape (NOWRITE) , that the data are
stored at 1600 bpi (DENSITY) and the maximum block size is 8000
bytes (BLOCKSIZE).
Running MPRM on an IBM mainframe
MPRM was tested on an IBM 3090-200 running under Release
3.8 of OS/VS2 MVS. A discussion on creating the executable
code on the IBM appears earlier in this section. Some of the
information appearing here may be too restrictive or
inefficient, but accurately describes how MPRM was tested on
the IBM. Suggestions on improving the use of the IBM should be
sent to the authors.
6-12 October 1994
-------
If MPRM requires a magnetic tape as input (e.g., NWS
surface weather observations), a scanning utility should be
used to determine the correct attributes of the tape, such as
the maximum size of the blocks, if a label exists, fixed or
variable length records. This insures the correct job control
language (JCL) when running MPRM.
For MPRM to work properly on the IBM, the following
guidelines should be observed.
1) When specifying disk file names in a run stream, the run
stream file name must match the 1-7 character DD statement
name in the JCL. This statement defines the data set name
where the input/output data are read/written.
2) When allocating space for a new data set, whether through
the JCL or TSO or ISPF, allocate sufficient space to store
the entire data set. Data set attributes for several data
types are shown below in Table 6-1. Other attributes can
be inferred from Table 6-2b.
October 1994
-------
TABLE 6-1. DATA SET ATTRIBUTES ON THE IBM FOR THE FOUR DAY TEST CASE.
Data Organ i -
zation
1 year NWS hourly PS
surface data,
CD- 144 format
1 year NCDC mixing PS
heights, 44 bytes/
record
1 year merged mixing PS
heights and surface
observations
1 year RAMMET format PS
4 days on-site, PS
4 observations/hour,
3 records/observation
Error message file* PS
General report file* PS
Record
Format
FB
FB
VBS
FBS
FB
FB
FB
Block Size
(bytes)
3120
3120
6604
6000
85
8000
7920
Record Length Storage
(bytes) Required
80
44
2200
1000
85
80
132
18 tracks
1 track
5 tracks
1 track
15 tracks
3 tracks
3 tracks
PS = physical sequential
FB = fixed length, blocked
VBS = variable length, blocked, spanned
FBS = fixed length, blocked, spanned
* storage requirements depend on options used.
3) MPRM makes use of a temporary file attached to logical
unit (device) 70. The DD statement name TEMP70 must be
used. Because the data set is small and temporary, a
suggested DD statement and subparameters are
//TEMP7 0 DD DSN=prefix.data_set_name
// DISP=(NEW,DELETE),
// DCB=(DSORG=PS,RECFM=FB,LRECL=100,
BLKSIZE=100),
// SPACE=(TRK,(1,1))
The data set name (DSN) is completely arbitrary and can be
any name the user chooses. For the IBM 3090 at NCC, the
prefix is the concatenation of the user ID and account ID.
The important name is the DD statement name - it must be
TEMP70.
6-14
October 1994
-------
4) The default DD statement name FT05F001 connects the input
images in the run stream to the load module and FT06F001
connects the output to the printer. If the default input
and output logical unit numbers (5 and 6, respectively)
are changed in MPRM, then the DD statement names must also
change.
5) To extract data from magnetic tape, the JCL must reflect
the 'unknown' nature of the data. Because the same
subroutine, and, therefore, FORTRAN READ statement, is
used to read all magnetic tape formats, the record format
should be specified as U (for undefined length records)
and the block size (BLKSIZE) should be specified as the
largest block, in bytes, in the file on tape. This is
where the tape scanning utility is most helpful. An
example DD statement for extracting upper air soundings in
an EBCDIC TD-5600VB format (note: do not confuse the VB
here with IBM's variable length, blocked record format -
the two are different) is
//TAPE10 DD DSN=prefix.data_set_name
// DISP=(OLD),
// DCB=(RECFM=U,BLKSIZE=6000),
// UNIT=(TAPE,,DEFER)
// LABEL=(1,NL,EXPDT=98000)
The DD statement name, TAPE10 in this example, must match
the tape file name in the run stream defined on the UA INI
input image (see the table in Appendix B on Defining Tape
File Names).
6) ASCII tapes require additional information to be added to
the DD statement. The subparameter OPTCD=Q should be
added to the DCB parameter. This informs the computer to
expect ASCII characters. NOTE: Some systems may restrict
6~15 October 1994
-------
the maximum number of bytes per block (the BLKSIZE).
Consult your IBM manuals or user support services before
proceeding with an ASCII tape.
7) The IBM extended error-handling subroutine ERRSET is
employed to provide uninterupted processing of magnetic
tapes and the merged data file.
Because the length of a block of data on tape is not
constant from one format to another (e.g., TD-5600 and
CD-144) and possibly within in a format (e.g., TD-5600VB),
each read of the tape causes a non-fatal error condition
(number 212) on the IBM. The default number of such
errors is 10 and there are normally several hundred blocks
of data on one tape. The subroutine ERRSET is called to
increment by one the number of allowable errors prior to
each time the tape is read. This allows the tape to be
read but MPRM will stop if the same error condition occurs
elsewhere in the processing.
When processing the unformatted merged data, MPRM
distinguishes between the file header records and the data
by the asterisk (*) in column 1 of the headers. The
headers are read as character strings with a check that an
asterisk is found in column 1. When the first record of
data are encountered, an error condition occurs. MPRM
responds to this by branching to the section of code that
backspaces on the merged data file and begins reading the
integer and real data. The IBM issues a non-fatal error
message (for error 213) and increments its internal error
counter. The default number of 213 errors is 10. Actions
identical to those taken for 212 errors are employed.
However, in any one run of Stage 3, the error for this
particular reason can occur only twice. MPRM will stop
for any other reason if error 213 is encountered.
6~16 October 1994
-------
8) The DD statement name for the data set of random numbers
used during Stage 3 processing for RAMMET processing must
be IRND.
Scan Reports
If data to be extracted from magnetic tape are not found
on the tape, then a short report of the tape contents is
written to the end of the error/message file. An example of
the report is shown in Figure D-4 in Appendix D. A separate
line is written whenever a station identification number
changes or a year changes. To speed the search for data, only
the first upper air sounding, mixing height or surface
observation is read from a data block on tape (there can be
several observations per block). The dates in the report,
therefore, do not necessarily correspond to the beginning and
end of the year.
For data on disk, no report is written but a warning
message is written to the error/message file, which is repeated
in the general report file, indicating that no data were
retrieved.
MPRM RUN TIME AND FILE STATISTICS
The following tables present the amount of time it took to
run various scenarios and the resulting file sizes. The VAX
CPU time is the time that appears in a VAX log-file and the
file sizes are expressed in blocks, where one block equals 512
bytes. Several factors affect the size of these files. For
example, in extracting upper air soundings from magnetic tape,
if the automatic data modification option is turned on and
there are several modifications, then the error/message file
may become quite large. This is evident in case A in Table
6-2a. Also, the more heights there are in the soundings, the
6"~17 October 1994
-------
larger the output file, as seen in cases A and B. The number
of quality assessment violations affects the size of the
error/message file. In case E there were a significant number
of violations reported, whereas in case F, the surface
observation quality
TABLE 6-2a. VAX RUN TIME AND FILE STATISTICS
A.
B.
C.
D.
E.
F.
G.
H.
I.
J.
K.
L.
Case
Extract Upper Air
Soundings (1st year
of 7 years on tape)
Extract Upper Air
Soundings (7th year
of 7 years on tape)
Extract Mixing Heights
(1 year, on disk)
Extract Upper Air
Soundings (as in case A)
and Mixing Heights
(as in case C) (1 year)
Quality Assess data
from case D
Extract NUS Surface
Obs. (1 year, on disk)
Quality Assess data
from case f
Extract/Quality
Assess On- site data
(4 days, on disk)
Merge data from
cases C and G (1 year)
Merge data from
cases E, G, and H (4 days)
RAMMET format from
Stage 3 from case I
RAMMET format from
Stage 3 from case J
CPU Time
(mm:ss)
4:19
14:13
0:10
4:15
1:38
3:51
4:45
0:24
4:28
0:19
1:35
0:15
| Report Files I
Error/Message General
(blocks) (blocks)
115
131
2
115
346
2
286
30
2
2
1
27
6
6
6
7
8
6
11
9
39
9
16
3
Output
File
(blocks)
798
849
22
799
800
2644
2644
27
1966
59
492
6
6-18
October 1994
-------
TABLE 6-2b. PERSONAL COMPUTER RUN TIME AND FILE STATISTICS
A.
B.
C.
D.
E.
f.
G.
H.
I.
J.
K.
L.
Case
Extract Upper Air
Soundings (1st year
of 7 years on tape)
Extract Upper Air
Soundings (7th year
of 7 years on tape)
Extract Mixing Heights
(1 year, on disk)
Upper Air Soundings
and Mixing Heights
(1 year, file down-
loaded from VAX)
Quality Assess data
from case D
Extract NWS Surface
Obs. (1 year, on disk)
Quality Assess data
from case f
Extract/Quality
Assess On-site data
(4 days, on disk)
Merge data from
cases C and G (1 year)
Merge data from
cases E, G, and H (4 days)
RAMMET format from
Stage 3 from case I
RAMMET format from
Stage 3 from case J
Real
Time
(rnnrss)
.
.
0:44
.
15:07
38:47
56:22
2:20
34:48
2:06
9:36
1:23
J Report Files -4
Error/Message General
(Kb) (Kb)
.
.
0.7
.
174.5
0.8
144.0
14.8
0.7
0.8
0.3
0.6
.
.
2.9
-
3.8
3.0
5.4
4.4
19.4
4.5
7.8
8.4
Output
File
(Kb)
-
-
10.8
408.7
409.3
1344.5
1344.9
13.3
1075.5
29.8
252.6
2.8
6-19
October 1994
-------
assessment, there were very few violations. Both data sets are
for one year of data.
When running the MPRM processor on a personal computer,
the user should try to have sufficient disk space to store
these files before starting a run. The times in Table 6-2b
represent the difference in time between calls to the system
clock when the processor is started and upon completion of a
test case. The system used for these tests was an
XT-compatible running at 8.0 MHz with an 8087 math co-processor
present.
Note that the general report file is nearly constant for
Stage 1 and 2 processing. The size of this file for Stage 3
processing can be affected by the options chosen. In these
examples processing was requested for RAMMET output with no
listing of the generated meteorological values. The Stage 3
general report for case K, where one year of data were
processed, was 7.8 Kb. If the output meteorology are listed,
using the JB LST option, the general report becomes 489.7 Kb,
which translates to about 125 printed pages.
The upper air soundings were extracted from a magnetic
tape that contained seven years (1964 - 1970) from Dayton, Ohio
in the TD-5600 format. One year of mixing heights, also for
Dayton, Ohio, resided on disk and were extracted on both the
VAX and a personal computer. To compare run-time statistics
between the VAX and personal computer, the extracted soundings
and mixing heights file generated by test case D was downloaded
from the VAX. One year of NWS surface observations also
resided on disk and were downloaded from the VAX to the
personal computer. The data were in an unblocked CD-144
format, i.e., there were 80 characters/record with each record
containing one hourly observation. See Sections 1 and 3 for
additional discussions of the data formats. The on-site data
6-20 October 1994
-------
used here consisted of four days of data with four observations
per hour. The size of the input file for test case H was 89
blocks on the VAX and 44.4 Kb on the personal computer. The
output file consisted of one observation per hour. The data
consisted of a date and time group and a single temperature
difference in the first record followed by two records of
vector data. The first vector record consisted of a height and
a wind speed, and the second record consisted of a height, wind
speed, wind direction, temperature and a..
In test case I, the mixing height file was not quality
assessed because these runs were to accumulate time and file
statistics. For regulatory use, quality assessment of the
mixing heights would be required prior to merging the data.
For Stage 3 processing, cases K and L, output was in
RAMMET format and the processor was run in default mode, i.e.,
the LST and TRA options were not activated, to produce the
smallest general report file possible.
October 1994
-------
SECTION 7
INTERPRETING PROCESSING ERRORS
Suppose, as happens on occasion, an improper input run
stream is fashioned. To investigate such an event, we will
purposely leave out a necessary input image from the first run
stream example shown in Section 4. Recall, that in this
example the goal was to extract mixing height data and upper
air observations and store these data in a file for subsequent
quality assessment. The file for storage of the extracted data
is defined in the UA IQA image. How will MPRM react to
omission of the UA IQA image? The MPRM processor checks the
input data for completeness before any data processing is
attempted. It will detect that the UA IQA data input image is
missing and will stop processing after all the input images
have been processed. A message will be placed in the
error/message file; this message is repeated within the general
report file (beneath the message summary table) indicating that
the output file for data extracted was not specified.
Note that MPRM will read and attempt to decipher all the
input images. When an end-of-file (or JB END statement) is
encountered on the input file, the processor reviews the images
processed to see if errors were detected. If any errors were
detected, further processing is inhibited except to generate a
general report. If the JB RUN image is included in the run
stream, MPRM will only read the run stream for syntax errors or
omissions, but will stop before actual data processing occurs.
The JB RUN image can be very useful in debugging run streams.
October 1994
-------
If the UA IQA input was omitted, the status report on
first page of the general report would appear as shown in
Figure 7-1. Since an error was detected in the UA-pathway run
stream data, the message reports that no UA data will be
processed. This message was generated because the IQA image,
which defines the output file for storage of the extracted
data, was omitted from the run stream for this example.
METEOROLOGICAL PROCESSOR FOR REGULATORY MODELS (MPRM), VERSION 1.2
TODAY'S DATE AND TIME: 14-APR-88 AT 08:12:06
INITIAL PROCESSING ON RAW METEOROLOGICAL DATA
STATUS REPORT PRIOR TO BEGINNING PROCESSOR RUN
1. REPORT FILE NAMES
ERROR MESSAGES: [OPJ.MPRH.DOC]ERROR.LIS
SUMMARY OF RUN: STANDARD OUTPUT DEVICE, UNIT 6
2. UPPER AIR DATA
SITE ID
93815
LATITUDE(DEG.)
39.83N
LONGITUDE(DEG.)
84.05U
THE PROCESS(ES) MPRM ANTICIPATES TO PERFORM ARE
NONE, ERROR(S) ON INPUT IMAGES FOR THIS PATH
EXTRACT INPUT -
EXTRACT INPUT -
OPEN: TAPE10
OPEN: [OPJ.MPRM.I01RAMZI.DAT
THE EXTRACT DATES ARE: STARTING: 31-DEC-63
ENDING: 5-JAN-64
ALL UPPER AIR DATA ABOVE 7000 METERS ARE CLIPPED, (DISREGARDED)
UPPER AIR AUTOMATIC DATA CHECKS ARE: ON (DEFAULT: ON)
3. NUS SURFACE DATA
THE PROCESS(ES) MPRM ANTICIPATES TO PERFORM ARE
NONE, NO DATA TO BE PROCESSED ON THIS PATH
4. ON-SITE DATA
THE PROCESS(ES) MPRM ANTICIPATES TO PERFORM ARE
NONE, NO DATA TO BE PROCESSED ON THIS PATH
Figure 7-1. First page of general report.
Even though the processing of UA data was inhibited, the file
names, the status of each file name, and the extraction dates
are still reported for verification by the user. Because there
7-2
October 1994
-------
were no surface observations or on-site data to process, a
message to that effect is written for their respective
pathways.
All messages generated during processing can be found in
the error/message file, defined in the JB ERR input image.
This file can be very long if there are a substantial number of
messages; therefore, the file's contents should be reviewed
before the file is printed. The messages generated by this
example are shown in Figure 7-2.
11
0
0
0
0
0
0
0
JB
JB
JB
JB
JB
JB
JB
JB
119
SETUP:
U12 UAEXST:
110
111
112
no
111
112
TEST:
TEST:
TEST:
TEST:
TEST:
TEST:
ENCOUNTERED
SUMMARY:
SUMMARY:
SUMMARY:
SUMMARY:
SUMMARY:
SUMMARY:
SUMMARY:
END OF
"JOB/RUN CARD"
MISSING/ERRORS
NO
NO
NO
NO
NO
NO
SF-EXT
SF-IQA
SF-OQA
OS -EXT
OS-IQA
OS-OQA
CARD
CARD
CARD
CARD
CARD
CARD
IN UA-IQA CARD
t
i
t
t
t
t
NULL
NULL
NULL
NULL
NULL
NULL
EXTRACT
QA
MERGE
EXTRACT
QA
MERGE
Figure 7-2. Error/message file.
The second message in the file (generated by the omission
of the UA IQA image) stopped data processing. This message is
deciphered below.
0 JB U12 UAEXST: SUMMARY: MISSING/ERRORS IN UA-IQA CARD
L- Message (up to 40 characters)
Subroutine where message was generated
Message code (Appendix E)
Pathway identification
Counter
The structure of all messages is identical. In the first
field of the message is a counter, which in this particular
case has no significance. The second field indicates that the
message was generated at the JB level of logic within the
processor. The conditions for the generation of messages are:
7-3
October 1994
-------
o JB messages - most often when problems are encountered in
deciphering the input run stream or when the processor
detects incomplete run stream information.
o UA, SF, OS, and MR messages - when problems are
encountered in deciphering a input run stream for the
respective pathway, when data cannot be properly read, or
when suspect data are encountered during a quality
assessment check.
The third field is a three-character message code. The first
character indicates the type of message: informational,
warning, error, quality assessment, trace. In the example, the
W indicates that this message is a warning. Warnings do not
necessarily prohibit data processing; error messages do. The
two numbers following the first character are used to provide
additional information concerning this message. This
additional information is provided in Appendix E. In general,
messages concerning input run stream have message numbers from
0 to 20. Message numbers greater than 20 are generated while
reading data files or during quality assessment checks.
The second page of the general report summarizes the
processor activities (Figure 7-3). Beneath the heading is a
message indicating the status when processing was completed.
For this example, the message reports an ABNORMAL JOB
TERMINATION. Following this message is the tabular summary of
the messages codes generated during processing. The report
concludes with a list of the warning and error messages, sorted
by pathway ID. These messages are identical to the ones in the
error file. Informational messages (codes that begin with I)
and quality assessment violations (codes that begin with Q) are
not listed within the summary report.
7-4 October 1994
-------
METEOROLOGICAL PROCESSOR FOR REGULATORY MODELS (MPRM), VERSION
TODAY'S DATE AND TIME: U- APR -88 AT 08:12:07
INITIAL PROCESSING ON RAW METEOROLOGICAL DATA
********************************************************
*** ABNORMAL JOB TERMINATION ***
********************************************************
1.2
**** MPRM MESSAGE SUMMARY TABLE ****
JB
E
U
I
0- 9
0
0
0
0
****
0 JB
****
10-19
0
1
7
8
20-29
0
0
0
0
30-39 40-49 50-59 60-69 70-79
00000
00000
00000
00000
TOTAL
0
1
7
8
WARNING MESSAGES ****
U12 UAEXST: SUMMARY: MISSING/ERRORS IN UA-IQA CARD
ERROR MESSAGES
****
--- NONE ---
Figure 7-3. Second page of general report.
In this example with the missing UA IQA image, information
on a single error can be traced through both the error/message
file and the general report. The first sign of the error was
the generation of the message indicating that no UA data would
be processed because errors were detected in the input run
stream. The next indication of the error was the display of
the "ABNORMAL JOB TERMINATION" message just above the table
summarizing the types of messages generated. One warning
message was listed below the summary table, which reported that
the IQA input for the UA-pathway either was missing or had been
entered in error. This warning, which is also listed within
the error/message file, prompts a check of the IQA input for
the UA-pathway- Once some familiarity is gained with the
processor and with developing the input data, the user should
begin to see the relationships between the abnormal job
terminations and the messages generated by the processor.
7-5
October 1994
-------
APPENDIX A
SUMMARY OF INPUT KEYWORDS
INTRODUCTION
The following tables define the keywords used in the input
to the MPRM processor. The logic within the processor, and
hence the input to the processor, can be functionally divided
into one of five major areas. These areas are called pathways
and derive their names from the functions performed. Each
pathway has one of the following two-letter acronym
designations:
o JB - processing that pertains to the entire processing job
o UA - processing of NWS upper air data and NCDC mixing
height data
o SF - processing of NWS surface weather observations
o OS - processing of user supplied (on-site) meteorological
data
o MR - processing that combines all available meteorological
data for each day
o MP - processing that pertains to dispersion model
meteorology.
Separate tables are presented for each pathway. Input data to
the processor is always directed to one of these five logical
pathways. There are basically four processing tasks that might
A-l
October 1994
-------
be involved. Each of these tasks has been assigned one of the
following two-letter acronyms:
o EX - input data that pertains to reading raw data files
and storing data in files for later processing checks of
data
o QA - input data that pertains to performing quality
assessment
o MR - input data that pertains to combining (merging) data
from various sources into one data file
o MP - input that pertains to developing a meteorological
data file for use with a specific dispersion model.
In the tables, the keywords used for each pathway are
listed along with a brief description of the input associated
with each keyword. Beneath the columns, labeled by task, is
shown whether the input defined by a particular keyword is
mandatory (M) or optional (O) . If neither an M or O is shown,
that keyword is not relevant or used for that task.
The syntax of the input data is:
AA BBB C-
Dependent on pathway and keyword
3-character Keyword
1 2-character Pathway ID (i.e., JB, UA, SF, OS, or MR)
HOW TO USE THESE TABLES
Appendices A and B are designed to be used together.
Given that the user comes to this appendix with a specific task
A~"2 October 1994
-------
in mind, the user can determine which keywords will be needed
to develop the input for each pathway.
The user can consult Appendix B to see the syntax
associated with each keyword, and any variations particular to
a given pathway.
JB keywords
FIN
ERR
STA
OUT
RUN
END
EX
H
H
0
0
0
0
QA
N
H
0
0
0
0
MR
H
H
0
0
0
0
HP
N
M
0
0
0
0
Description and usage
Signals the completion of input data
for this pathway.
Defines disk file name for processor
generated warning and error messages.
Defines the beginning of input data
for pathway.
Defines disk file name for the general
report.
If present, processing STOPS following
completion of processing input images.
Useful for checking input images.
Alternate method for signaling the end
of the input data. Default is
encountering an End-Of-File (EOF) in
reading the input data.
A-3
October 1994
-------
UA keywords
EX
QA
MR
MP
Description and usage
FIN
STA
LOG
IQA
OQA
EXT
IN1
IN2
TOP
OFF
CHK
AUD
TRA
M
(M)
(M)
Signals the completion of input data
for this pathway.
Defines the beginning of input data
for pathway.
Defines station ID, station longitude
and latitude, and number of hours to
convert times given in data to 1ST.
Defines disk file name for storage of
extracted data and input data for QA
processing.
Defines disk file name for storage of
QA processed data and input data for
MERGE.
Defines EXTRACT start and stop dates.
Defines tape name, format, and
characteristics for upper air data.
Defines tape name and characteristics,
or disk file name and format for
morning and afternoon mixing height
data.
Upper air data for altitudes above
ground greater than UATOP are ignored.
Default value of UATOP is 5000 m.
Turns off automatic data modifications
during EXTRACT of upper air soundings.
By default adjustments are made to
correct suspect temperatures, redefine
directions associated with calm winds,
delete (when possible) data given for
mandatory pressure levels, fill in
missing dew-point temperatures, and
adjust heights to AGL.
Redefines QA range checks and missing
value flag for a variable.
Adds variable to general AUDIT report.
Default variables are UAM1 and UAM2,
morning and evening mixing heights.
Turns on trace for missing data during
QA processing.
(M) One or both of IH1 and IN2 must be present, if EXT is present.
A-4
October 1994
-------
SF keywords
EX
QA
HR
HP
Description and usage
FIN
STA
LOG
IQA
OQA
EXT
IN2
CHK
AUD
TRA
M
M
0
0
Signals the completion of input data
for this pathway.
Defines the beginning of input data
for pathway.
Defines station ID, station longitude
and latitude, and number of hours to
convert times given in data to 1ST.
Defines disk file name for storage of
extracted data and input data for QA
processing.
Defines disk file name for storage of
QA processed data and input data for
MERGE.
Defines EXTRACT start and stop dates.
Defines tape name and characteristics,
or disk file name and format of hourly
weather observation data.
Redefines QA range checks and missing
value flag for a variable.
Adds variable to general AUDIT report.
Default variables are:
SLVP Sea-IeveI station pressure
PRES Uncorrected station pressure
CLHT Ceiling height
TSKC Total and Opaque sky cover
HZVS Horizontal visibility
TMPD Dry-bulb temperature
WD16 Wind direction
WIND Wind speed.
Turns on trace for missing data during
QA processing.
A-5
October 1994
-------
OS keywords
FIN
STA
LOC
IQA
OQA
EXT
NAP
FHT
AVG
DT1
DT2
DT3
HGT
CLH
CHK
AUD
TRA
SFC
EX
M
0
M
M
0
M
M
0
0
0
0
0
0
QA
M
0
H
M
M
M
H
0
0
0
0
0
0
0
0
0
0
MR
N
0
N
H
M
M
MP
Description and usage
Signals the completion of input data
for this pathway.
Defines the beginning of input data
for pathway.
Defines site ID, site longitude
and latitude, and number of hours to
convert times given in data to 1ST.
Defines disk file name for storage of
input data for QA processing. May
have more than one observation
per hour.
Defines disk file name for storage of
QA processed data and input data for
MERGE. Always hourly averages.
Defines EXTRACT start and stop dates.
Defines order of OS input variables as
they appear within IQA file.
Defines FORTRAN format statements for
reading IQA file.
Defines maximum number of observations
to be expected per hour for input data
provided within IQA file.
Defines lower and upper measurement
heights associated with first
temperature difference.
Defines lower and upper measurement
heights associated with second
temperature difference.
Defines lower and upper measurement
heights associated with third
temperature difference.
Defines meteorological mast
configuration, number of levels, and
height associated with each.
Defines valid minimum wind speed for
use in definition of calm wind.
Redefines QA range checks and missing
value flag for a variable.
Adds variable to general AUDIT report.
Default variables are:
MHGT Mixing height
SA SD of horizontal wind direction
SE SD of vertical wind direction
TT Dry-bulb temperature
WD Horizontal mean wind direction
US Horizontal mean wind speed.
Turns on trace for missing data during
QA processing.
Defines surface characteristics.
Default values are albedo, 0.25; Bowen
ratio, 0.75; and surface roughness
length, 0.15 m.
A-6
October 1994
-------
MR keywords
OUT
FIN
EXT
STA
EX
QA
MR
M
M
0
0
HP
Description and usage
Defines disk file name for storage
of combined (merged) data.
Signals the completion of input data
for this pathway.
Defines start and stop dates for
MERGED data.
Defines the beginning of input data
for pathway.
A-7
October 1994
-------
HP keywords
FIN
STA
MET
HMP
EXT
VBL
TRA
LSI
EX
QA
MR
HP
M
0
M
H
0
0
0
0
Description and usage
Signals the completion of input data
for this pathway.
Defines the beginning of input data
for pathway.
Defines disk file name associated with
combined (merged) meteorological data
file.
Defines disk file name associated with
output meteorological file created by
this run. Included as an option is
the ability to define the dispersion
model that will be accessing this
file.
Defines start and stop dates for
meteorological data file to be
created by this run.
Redefines (override default)
processing methodology to be employed
in generating output meteorological
data file. Currently there are
selection options available for
processing wind, temperature.
stability category, and mixing
height. Choice is largely whether to
use NWS or on-site meteorological
data.
Turns on more detailed trace of errors
encountered during processing.
Default is to provide daily summaries.
Turns on listing of generated
meteorology to general report file.
A-8
October 1994
-------
APPENDIX B
SUMMARY OF INPUT SYNTAX
INTRODUCTION
In the following tables, the keywords used in the input to
the MPRM processor are presented in more detail. The first two
tables explain the keywords used in defining the input files
and tapes and the output files. These are presented first, as
they are likely to be consulted most often.
Presented next are tables for the rest of the keywords,
arranged in alphabetical order by keyword.
The specific information provided for each keyword is the
pathway(s) employing the keyword and the syntax of the input
image employing the keyword. The description for each keyword
concludes with one or more examples to illustrate typical
usage.
HOW TO USE THESE TABLES
Appendices A and B are designed to be used together.
Since the user comes to this appendix with a specific task in
mind, the user can determine what information is defined by
each keyword (Appendix A) and the syntax of the input image
employing the keyword (Appendix B).
The user can consult Appendix A to see which keywords are
mandatory and which are optional, dependent on tasks to be
performed and pathway involved.
B-l
October 1994
-------
Defining Diskfile Names
Keywords: ERR, IQA, OUT. OQA on PathwaylDs JB, UA, SF and OS
Purpose: Define diskfile name.
PathwaylD Keyword Keywordl Filename
Keywordl always equals DISK.
IBM PC example: JB OUT DISK REPORT.LIS
VAX example: JB OUT DISK [JSI.MPRM.FORT]REPORT.LIS
Keyword: IN2 on PathwaylD UA
Purpose: Define diskfile name containing mixing height data.
PathwaylD Keyword Keywordl Filename Form-or-Format SitelD
Station identification
Keywordl is either
DISK or USER.
If DISK, then Form-or-Format must be 9689FB.
If USER, then Form-or-Format must be a valid FORTRAN format. The
input list to be read is of the form: AAAA,YEAR,MONTH,DAY,UAM1.UAM2.
Where AAAA is station ID (note even though this is typically a
5-digit number, it is input as a CHARACTER variable).
YEAR is last 2-digits of year.
MONTH is month (1 = January, 2 = February, etc).
DAY is day of month.
UAM1 is the morning urban mixing height to nearest meter.
UAH2 is the afternoon mixing height to nearest meter.
Note, YEAR, MONTH, DAY, UAM1 and UAM2 are input as INTEGER variables.
Examples: UA IN2 DISK CSCZIDAT.DAT 9689FB 13840
UA IN2 USER CSCZIDAT.DAT (A5.3I2.2X,I4,14X,14) 13840
Keyword: IN2 on PathwaylD SF
Purpose: Define diskfile name containing hourly surface data.
PathwaylD Keyword Keywordl Filename Form SitelD
Station identification
Form equals CD144FB for the
standard format or equals
SCRAMFB for the compressed
'SCRAM' format.
Keywordl always equals DISK.
Example: SF IN2 DISK CSCSFDAT.DAT CD144FB 13840
B-2
October 1994
-------
Defining Tapefile Names
Keyword: IN1 on PathwaylD UA
Purpose: Define tapefile containing upper air data.
PathwaylD Keyword Keywordl Tapename Form Format SitelD
Station
identification
Either ASCII or EBCDIC,
(always ASCII on IBM)
Form is 5600FB or 5600VB.
FB stands for Fixed-block and
VB stands for Variable-block
formatting on tape.
Keywordl always equals TAPE
Example: UA IN1 TAPE TAPE1 5600VB EBCDIC 13840
Keyword: IN2 on PathwaylD UA
Purpose: Define Tapefile containing mixing height data
PathwaylD Keyword Keywordl Tapename Form Format SitelD
Station
identification
Either ASCII or EBCDIC,
(always ASCII on IBM).
Form is always 9689FB.
FB stands for Fixed-block
formatting on tape.
Keywordl always equals TAPE
Example: UA IN2 TAPE TAPE2 9689FB ASCII 13840
Keyword: IN2 PathwaylD SF
Purpose: Define tapefile containing hourly surface data.
PathwaylD Keyword Keywordl Tapename Form Format SitelD
L
Station
identification
Either ASCII or EBCDIC,
(always ASCII on IBM).
Form is always CD144FB.
FB stands for Fixed-block
formatting on tape.
Keywordl always equals TAPE
Example: SF IN2 TAPE TAPE3 CD144FB EBCDIC 13840
B-3
October 1994
-------
Keyword: AUD used on PathwaylDs UA, SF, and OS
Purpose: Add variables to audit summary report.
PathwaylD Keyword PRM1, PRH2, ...
I 4-character variable name
1 4-character variable name
PRM1, PRM2 - the available 4-character names of the variables
are provided in Appendix C.
The AUD input can be repeated as often as needed in order to list
all the variables to be added to the audit summary.
The default list of audit variables on the UA-pathway are the
twice-daily mixing height values.
The default list of audit variables on the SF-pathway are: sea-
level pressure, station pressure, ceiling height, total sky cover,
horizontal visibility, dry-bulb temperature, wind direction and
wind speed.
The default list of audit variables on the OS-pathway are: mixing
height, wind speed, wind direction, temperature, aa, and ?e.
Example: OS AUD DT01.DT02
OS AUD PRES.TSKC
Keyword: AVG used only on PathwaylD OS
Purpose: Define maximum number of observations per hour.
PathwaylD Keyword VALUE
VALUE is the maximum number of on-site observations expected
during an hour. The default value is 1. The maximum value
allowed is 12.
Example: OS AVG 12
B-4
October 1994
-------
Keyword: CHK used on PathwaylDs UA, SF, and OS
Purpose: Alter quality assessment range check parameters.
PathwaylD Keyword PRM1 PRH2 PRH3 PRM4 PRH5
I Upper bound of
range
Lower bound of
range
Missing value
indicator
Range check switch
Name of variable
PRM1 - 4-character name for sealers variables and 2-character
names for multi-level on-site variables
(names listed in Appendix C)
PRM2 - Integer input, either 1 or 2
PRM3 - Integer input, be sure missing value indicator is not also
an acceptable value for the variable
PRH4 - Integer input, be realistic to insure useful quality assessment
checks
PRM5 - Integer input, be realistic to insure useful quality assessment
checks
Examples: OS CHK DT01 1 -999 -2 10
OS CHK US 1-99 0 50
Keyword: CLM used only on PathwaylD OS
Purpose: Define wind threshold speed.
PathwaylD Keyword VALUE
VALUE is the threshold wind speed for valid wind measurements,
in meters per seconds, used in defining calms.
Example: OS CLM 0.45
B-5
October 1994
-------
Keywords: DT1, DT2 and DT3 used only on PathwaylD OS
Purpose: Define heights associated with Delt-T measurements.
PathwaylD Keyword LHEIGHT UHEIGHT
LHEIGHT Lower height of temperature difference measurement in meters.
UHEIGHT Upper height of temperature difference measurement in meters.
(Values may be entered as real or integer as shown in example below.)
Example: OS DT2 2.0 10
Keyword: END used only PathwaylD JB
Purpose: Signals end of input run stream for entire job.
PathwaylD Keyword
This keyword has no parameters.
Example: JB END
Keyword: EXT used on PathwaylDs HP, UA, SF, OS and MR
Purpose: Define start and stop dates for processing.
PathwaylD Keyword YY1 MM1 DD1 YY2 MM2 DD2
YY1, MM1 and DD1 define the starting year, month and day of the data
to be retrieved (merged on PathwaylD MR). YY2, MM2 and DD2 define the
ending year, month and day of the data to be retrieved (merged on
PathwaylD MR). Note, these dates are inclusive, so data for day DD2
will be included. All values are input as INTEGERS. YY1 and YY2 can
be expressed fully or by last two digits, for example 1987 or 87.
Example: UA EXT 84 1 1 84 12 31
Keyword: FIN used on PathwaylDs JB, MP, UA, SF, OS AND MR
Purpose: Signals end of input run stream for pathway.
PathwaylD Keyword
This keyword has no parameters.
Example: JB FIN
B-6
October 1994
-------
Keyword: FMT used only on PathwaylD OS
Purpose: Define FORTRAN formats for reading input data.
PathwayID Keyword Keywordl PRH
DATxx
PRH
Example:
A valid FORTRAN format
Keywordl equals DATxx , where xx
are INTEGERS such as 01, 02, 03 etc.
- the xx refers to sequence number. For instance, if
there are three READs (in the FORTRAN sense) needed
for reading the data, then there would be DAT01,
DAT02 and DAT03 definitions within the OS NAP input
and there would likewise be DAT01, DAT02 and DAT03
definitions within the OS MT input.
- this includes the right and left parentheses.
OS MAP DAT02 INSO TSKC CLHT
OS FMT DAT02 (5X,F4.2,2X,2F5.2)
I- Note the specification of the
number of decimal places. This
is not necessary for proper input
of the value, but since we use
this format in writing to the
OQA-file, we NEED the decimal
specification.
Keyword: HGT used only on PathwaylD OS
Purpose: Define heights associated with multilevel input data.
PathwaylD Keyword NN HGT1 HGT2
NN
HGTNN
Number of heights to be read in as input.
(Maximum: 10)
HGT1 ... HGTNN N values of heights, in meters.
(You can not repeat this card, so all values must be listed.)
Example: OS HGT 5 2.0 4.0 10 30.0 50
B-7
October 1994
-------
Keyword: LOG used on PathwaylDs UA, SF and OS
Purpose: Define location parameters for site.
PathwaylD Keyword SitelO AAAAAA BBBBBB GMTLST
SitelD Site identification number. Typically, this is a 5-digit
number. To avoid conflicts with current dispersion models
expecting RAMMET type meteorological input, we suggest the
OS pathway SitelD contain only numbers, no letters. Note,
SitelD MUST agree with that given in the INPUT data.
AAAAAA These two data entries are the longitude and latitude in
and decimal degrees. Longitude and latitude can appear in
BBBBBB either position AAAAAA or BBBBBB. The important point is to
define both. North and south latitude are entered as 30.DON
and 30.00S. Likewise, the longitude is entered as 170.13E.*
GMTLST The number of hours to be subtracted to convert times given on
this pathway to Local Standard Time (LST). If times are given
as Greenwich Mean Time (likely on UA pathway), for the
east coast of the United States, GMTLST would be 5. Note,
if times are given in LST, a zero must still be given as
input for GMTLST.
Example: UA LOG 03820 34.37N 81.97E 5
Keyword: LST used only on PathwaylD JB
Purpose: Turn on printed listing of generated meteorological data.
PathwaylD Keyword
This Keyword has no parameters.
Example: MP LST
*NOTE: Entering a latitude north of the Arctic Circle (63.5
degrees N) where the sun does not rise during all or
part of the winter will cause a run-time error in
STAGES processing.
B-8
October 1994
-------
Keyword: MAP used only on PathwaylD OS
Purpose: Define sequence of variables for data records.
PathwaylD Keyword Keywordl
PRM1, PRM2, PRN3,
L PRM1, PRM2, etc. are the 4-cheracter
variable names for the OS variables to
be processed.
Keywordl equals DATxx, where xx
are INTEGERS such as 01, 02, 03 etc.
The MAP data describe the order and structure of the OS data to be
processed. There are several assumptions made, but given these rules
are followed, a variety of data sets can be processed with little
to no restructuring.
Rule 1. Each OS observation is completely labeled with a date and
time, which will at least define the year, month, day and hour of the
OS observation.
Rule 2. The date and time data are given as INTEGERS and are the first
data presented in the OS observation. Note, the order is unimportant,
presenting year, month and day is just as acceptable as presenting day,
month and year.
Rule 3. The rest of the OS data, such as mixing heights, insolation,
wind speed etc.), are presented FOLLOWING the DATE and TIME data.
Limits: There can be up to 20 records of input in one observation.
Each record is limited to have no more than 40 variables. The
multi-level variables are limited to no more than 10 levels.
B-9
October 1994
-------
Keyword: NAP used only on PathwayID OS
(continuation of discussion):
PathwaylD Keyword Keywordl PRM1, PRH2, PRM3,
L PRM1, PRH2, etc. are the 4-character
variable names for the OS variables to
be processed.
Keywordl equals DATxx, where xx
are INTEGERS such as 01, 02, 03 etc.
Appendix C lists the 4-character names to be used in defining
PRH1, PRM2. ... etc.
DATxx defines the sequence for the input. The xx indicates
the record number, as DAT01 is the first record of data and
DAT02 is the second record of data.
Example:
OS MAP DAT01 OSYR OSMO OSDY OSHR OSMN
OS MAP DAT02 TT01 US01
OS MAP DAT03 TT02 US02 UD02 SA02
The above three input images employ the MAP keyword. They
describe to the processor the input data records for one time
period.
B-10
October 1994
-------
Keyword: MAP used only on PathwayID OS
(continuation of discussion):
PathwaylD Keyword Keywordl PRH1, PRH2, PRIG,
L PRM1, PRH2, etc. are the 4-character
variable names for the OS variables to
be processed.
Keywordl equals DATxx, where xx
are INTEGERS such as 01, 02, 03 etc.
DATxx rules:
DAT01 must begin with a date and time definition. The date must
provide year, month and day. The time must provide hour of day,
minutes is optional. The order of the date and time variables is
not important.
PRH rules:
The scalar variable names are: ALTP, SLVP, PRES, CLHT, TSKC, HFLX,
USTR, HHGT, ZOHT, SAHT, PAMT, INSO, NRAD, DT01, DT02, DT03, US01,
US02, US03, OSDY, OSNO, OSYR, OSHR, OSMN
The multi-level variable names are: HT, SA, SE. SV, SU, SU, TT,
WD, US, W, DP, RH, V1, V2, V3. These are the FIRST two letters of
the 4-character names. The LAST 2 characters are the level on the
tower to be associated with the variable (maximum: 10).
B-ll
October 1994
-------
Keyword: MAP used only on PathuaylD OS
(continuation of discussion):
Pathway ID Keyword Keywordl PRH1. PRH2, PRH3,
PRM1, PRM2, etc. are the 4-character
variable names for the OS variables to
be processed.
Keywordl equals DATxx, where xx
are INTEGERS such as 01, 02, 03 etc.
Below is an example of what the first data record of on-site
data might look like (see example page B-11). As the data
have a specific format, scales are given at the top and bottom
to ease interpretation.
In the example, records 1 through 3 are for one time period
The first record for each time period begins with a date and
time group. For instance, deciphering record 1 we have,
January 1, 1984 at 11:05 AM. Deciphering record 2 we have
temperature 2° C and wind speed of 2.1 m/s. The third record
is, temperature 1.99° C, wind speed 3.2 m/s, wind direction
112° and OB 32 degrees.
1 2345678901 2345678901 2345678901 2345678901 234567890
Record
number
1
2
3
84 01 01
2.0
1.99
Sample On- Site
1105
2.1
3.2 112.
(OS) Input Data
32.
12345678901234567890123456789012345678901234567890
There is no relationship assumed between the sequence
number,xx, given in DATxx with the measurement level, yy,
given in the multi-level variable name, as HT02.
B-12
October 1994
-------
Keyword: MET used only on PathwaylD MP
Purpose: Define diskfite of merged data.
PathwaylD Keyword Keywordl Filename PRM
Keywordl always equal to DISK.
Filename be sure to fully specify the filename.
PRM - Integer number of hours needed to be subtracted
from Greenwich Mean Time to convert to Local
Standard Time.
Examples: MP MET DISK [JSI.MPRM.DAT1RAMSTL.DAT 5
MP MET DISK RAMSTL.DAT 5
Keyword: MMP used only on PathwaylD MP
Purpose: Define dispersion model and diskfile name for output.
PathwaylD Keyword Keywordl Filename PRM
Keywordl always equal to DISK.
Filename - be sure to fully specify the filename.
PRM - (optional) Name of dispersion model. Current list
of dispersion model names are BLP, RAM, I SCSI,
ISCST2A, MPTER, CRSTER, COMPLEX1, CALINE-3, RTDM,
VALLEY, ISCLT, CDM16, CDM36. The default
dispersion model is CRSTER. This output
is equivalent to that produced by RAMMET.
Examples: MP MMP DISK [JSI.MPRM.DATJRAMMET.DAT
MP MMP DISK RAMMET.DAT CDM36
Keyword: OFF used only on PathwaylD UA
Purpose: Turn off automatic modification checks to upper air data.
PathwaylD Keyword
This keyword has no parameters.
Example: UA OFF
B-13
October 1994
-------
Keyword: RUN used only on PathwayID JB
Purpose: Inhibit data processing; check run stream snytax and stop.
PathwayID Keyword
This keyword has no parameters.
Example: JB RUN
B-14
October 1994
-------
Keyword: SFC used only on PathwaylD OS
Purpose: Define site characteristics.
PathwaylD Keyword Keywordl PRM1... PRH5
L
Meaning and number of parameters is
dependent of definition of Keywordl.
- Keywordl is either SETUP, VALUES or SECTOR.
The first time Keyword SFC data are given,
the value of Keywordl MUST be SETUP.
When Keywordl is SETUP:
PRM1 is either ANNUAL, SEASONAL or MONTHLY. This defines the
frequency (number of times) throughout the year that the
Albedo, Bowen Ratio and Surface Roughness length will be
defined.
PRM2 is the number of wind direction sectors to be used in defining
values for the Albedo, Bowen Ratio and Roughness length.
PRM2 can be any value 1 through 12.
When Keywordl is VALUES:
PRM1 is the time period being defined.
If PRM1 is ANNUAL on the SETUP image, then PRM1 is always 1
on the VALUES image.
If PRM1 is SEASONAL on the SETUP image, then PRM1 is
1 through 4 on the VALUES image (1 = winter months: 12, 1
and 2; 2 = spring months: 3, 4 and 5 etc.).
If PRM1 is MONTHLY on the SETUP image, then PRM1 is 1 through
12 on the VALUES image (1 = January, 2 = February etc.).
PRM2 is the wind direction sector being defined.
PRM3 is the Albedo for this time (PRM1) and wind
sector (PRM2).
PRM4 is the Bowen ratio for this time (PRM1) and
wind sector (PRM2).
PRM5 is the Roughness length in meters for this time (PRM1) and
wind sector (PRM2).
(continued)
B-15
October 1994
-------
Keyword: SFC used only on PathwaylD OS
(continuation of discussion):
PathwayID Keyword Keyword1
PRH1... PRM5
I Meaning and number of parameters is
dependent of definition of Keywordl.
- Keywordl is either SETUP, VALUES or SECTOR.
The first time keyword SFC data are given,
the value of Keywordl MUST be SETUP.
When Keywordl is SECTOR:
PRM1 is the wind sector number, it must be less than
or equal to the value given for PRM2 on the SETUP image.
PRM2 is the beginning azimuth in degrees for this wind sector,
reading in a clockwise sense. PRM2 must be a value between
0 through 360.
PRH3 is the ending azimuth in degrees for this wind sector,
reading in a clockwise sense. PRH2 must be a value between
0 through 360.
Note, presently no checks are made to insure that the wind sectors are
defined with no overlapping boundaries, and that a sufficient number
of sectors are defined to cover all possible directions.
No checks are made to insure that the Albedo, Bowen ratio and
Roughness length have been defined for all months and all wind sectors.
Special cases can arise when such is not necessary. To define values
for all months and sectors requires: (frequency)x(number of sectors)
VALUES data cards, where frequency is 1, 4 or 12 for ANNUAL, SEASONAL
and MONTHLY, respectively.
Example: OS SFC SETUP ANNUAL 2
OS SFC VALUES 1 1 0.6 0.5 0.3
OS SFC VALUES 1 2 0.8 0.6 0.1
OS SFC SECTORS 1 30 150
OS SFC SECTORS 2 150 30
B-16
October 1994
-------
Keyword: STA used PathwaylDs JB, HP, UA, SF, OS AND MR
Purpose: Signals beginning of input run stream data for pathway.
PathwayID Keyword
This keyword has no parameters.
Example: JB STA
Keyword: TOP used only on PathwaylD UA
Purpose: Define 'clipping height' for upper air data.
PathwaylD Keyword UATOP
UATOP is height in meters. Upper air data given for heights above
ground greater than UATOP are ignored. Note, value must be entered
as an INTEGER.
Example: UA TOP 7500
Keyword: TRA as used on PathwaylD MP
Purpose: Turn on trace notes to provide details of processing.
PathwaylD Keyword
This Keyword has no parameters.
Example: MP TRA
Keyword: TRA as used on PathwaylDs UA, SF and OS
Purpose: Turn on trace for missing data during QA processing.
PathwaylD Keyword PRM1, PRM2, ...
PRM1, PRM2 - the available 4-character names of the variables are
provided in Appendix C.
Example: OS TRA OT01 OT02
Keyword: VBL used only on PathwaylD MP
Purpose: Define methodology for generating output variable.
PathwaylD Keyword ITEM ACTION XXXX X
' for use only
with additional
input 'USERIN'
additional input,
as required.
6-character keyword
instructing processor
how 'ITEM' is to be
accomplished.
4-character keyword
defining process
(methodology).
See Table 4-1 in Section 4 for input values acceptable for
ITEM, Action and XXXX.
Examples: MP VBL WIND ONSITE 35.0
MP VBL STAB SASITE 10.0
B-17
October 1994
-------
APPENDIX C
VARIABLE NAMES AND DEFAULT RANGE CHECKS
MPRM gives the user the ability to override the internal
definitions for upper and lower bounds, missing value
indicator, and treatment of endpoints during quality assessment
checks during Stage 1 processing. This appendix presents the
following:
o Variable names used to override the parameters
o A description of each variable and the units used
o Quality assessment default parameters for each variable
o Variables automatically audited during quality assessment.
There are seven fields in each table: variable name,
description, units, range check switch, missing flag, lower
bound, and upper bound. Each of these fields is described
below.
Variable name
This is the four-character variable name used in the input
images for redefining quality assessment parameters (the
CHK image on each pathway), activating auditing of
variables not automatically audited (the AUD image on each
pathway) and defining the on-site data map (the DAT and
LVL images only on the OS-pathway).
If an asterisk (*) appears before the variable name, then
the variable is automatically audited during quality
assessment. These variables are always audited on the
upper air and surface pathways. However, for the on-site
pathway if the variable is not in the data map, then the
C-l
October 1994
-------
variable is omitted from the audit. If a person wants to
audit additional variables on any pathway, the AUD input
image is used.
Description and units
A brief description and the units of each variable follows
the name. For several variables, a multiplier also
appears in the units field. This can be identified by the
*10 or *100 (the only two multipliers used) that appears
after the units. Because the upper air and surface
observations are treated as integers within MPRM,
multipliers are used to retain significant digits prior to
rounding the value to the nearest integer.
Range Check Switch
The Range Check Switch field indicates whether to exclude
the lower and upper bound ( = 1) or include the bounds ( =
2) in determining if the variable violates the prescribed
limits. This value can be changed by using the CHK input
image.
Missing Value Indicator
The missing value indicator is the value used in the
processor to represent missing data for the variable.
This value can be changed by the user on the CHK input
image. This option is particularly useful if data are
already extracted and a different missing value flag was
used.
Bounds
The last two fields, the Lower and Upper bounds, are the
limits against which the value of a variable is checked.
If the value lies outside this interval, the endpoints
either included or excluded according to the Range Check
Switch, then a quality assessment violation is recorded.
C-2
October 1994
-------
As in the Range Check Switch and Missing Flag, these
values can be modified using the CHK card.
TABLE C-1. VARIABLE NAMES, UNITS, AND QUALITY ASSESSMENT DEFAULT SETTINGS FOR THE UA-PATHUAY
Variable
name Description
UAPR
UAHT
UATT
UATD
UAWD
UAWS
UASS
UADS
UALR
UADD
UAM1*
UAM2*
Atmospheric pressure
Height above ground level
Dry bulb temperature
Dew-point temperature
Wind direction
Wind speed
Wind speed shear
Wind direction shear
Temperature lapse rate
Dew point deviation
A.M. Mixing height
P.M. Mixing height
Units
millibars *10
meters
°C *10
°C *10
degrees from north
meters/second *10
(m/s)/(100 meters)
degrees/dOO meters)
°C/(100 meters)
°C/(100 meters)
meters
meters
Range
check
switch
1
2
1
1
2
1
2
2
2
2
2
2
Missing
value
indicator
-9999
-9999
-9999
-9999
-9999
-9999
-9999
-9999
-9999
-9999
-9999
-9999
Bounds
Lower
5000
0
-350
-350
0
0
0
0
-2
0
50
500
Upper
10999
5000
+350
+350
360
500
5
90
5
2
500
3500
^Automatically included in audit report.
C-3
October 1994
-------
TABLE C-2. VARIABLE NAMES, UNITS, AND QUALITY ASSESSMENT DEFAULT SETTINGS FOR THE SF-PATHUAY
Variable
name
ALTP
SLVP*
PRES*
CLHT*
TSKC*
C2C3
CLC1
CLC2
CLC3
CLC4
CLT1
CLT2
CLT3
CLT4
PUTH
HZVS*
TMPD*
TMPU
DPTP
RHUM
WD16*
WIND*
Description
Altimeter pressure
Sea level pressure
Station pressure
Ceiling height
Tot a I //opaque sky cover
2 level//3 level cloud cover
Sky condition//cover, level 1
Sky condition//cover, level 2
Sky condition//cover, level 3
Sky condition//cover, level 4
Cloud type//height, level 1
Cloud type//height, level 2
Cloud type//height, level 3
Cloud type//height, level 4
Present weather (2 types)
Horizontal visibility
Dry bulb temperature
Wet bulb temperature
Dew-point temperature
Relative humidity
Wind direction
Wind speed
Units
inches of mercury
millibars *10
millibars *10
kilometers *10
tenths//tenths
tenths//tenths
// + An+ K c
______ ///Urn *in\
kilometers *10
"C *10
°C *10
"C *10
whole percent
tens of degrees
meters/second *10
Range
check
switch
2
1
1
2
2
2
2
2
2
2
2
2
2
2
2
2
1
1
1
2
2
1
Missing
value
indicator
-9999
-9999
-9999
-9999
9999
9999
9999
9999
9999
9999
99999
99999
99999
99999
9999
-9999
-9999
-9999
-9999
-9999
-9999
-9999
Bounds
Lower
2700
9000
9000
0
0
0
0
0
0
0
0
0
0
0
0
0
-300
-650
-650
0
0
0
Upper
3200
10999
10999
300
1010
1010
910
910
910
910
98300
98300
98300
98300
9292
1640
350
350
350
100
36
500
NOTE: The three pressure variables (ALTP, SLVP, and PRES), the ceiling height (CLHT) and the
sky cover (TSKC) are also available on the OS-pathway.
Automatically included in audit report.
//The two variables described have been combined to form one variable.
C-4
October 1994
-------
TABLE C-3. VARIABLE NAMES, UNITS, AND QUALITY ASSESSMENT DEFAULT SETTINGS FOR THE OS-PATHWAY
Variable
name Description
HFLX
USTR
MHGT*
ZOHT
SAMT
PAMT
INSO
NRAD
DT01
DT02
DT03
US01
US02
US03
HTnn
SAnn*
SEnn*
SVnn
SUnn
SUnn
TTnn*
WDnn*
WSrm*
Wnn
DPnn
RHnn
V1nn
V2nn
V3nn
ALTP
SLVP*
PRES*
CLHT*
TSKC*
OSDY
OSMO
OSYR
OSHR
OSMN
Surface heat flux
Surface friction velocity
Mixing height
Surface roughness length
Snow amount
Precipitation amount
Insolation
Net radiation
Temperature diff.(U - L)'
Temperature diff.(U - L)'
Temperature diff.(U - L)>
User's scalar #1
User's scalar #2
User's scalar #3
Height
Std. dev. horizontal wind
Std. dev. vertical wind
Std. dev. v-comp. of wind
Std. dev. w-comp. of wind
Std. dev. u-comp. of wind
Temperature
Wind direction
Wind speed
Vertical component of wind
Dew-point temperature
Relative humidity
User's vector #1
User's vector #2
User's vector #3
Altimeter pressure
Sea level pressure
Station pressure
Ceiling height
Sky cover (total or opaque)
Day
Month
Year
Hour
Minute
Units
watts/square meter
meters/second
meters
meters
centimeters
centimeters
watts/square meter
watts/square meter
"C
°C
°C
user's units
user's units
user's units
meters
degrees
degrees
meters/second
meters/second
meters/second
CC
degrees from north
meters/second
meters/second
°C
whole percent
user's units
user's units
user's units
inches of mercury *
millibars *10
millibars *10
kilometers *10
tenths
Range
check
switch
1
1
1
1
2
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
1
1
1
2
1
1
1
100 2
1
1
2
2
2
2
2
2
2
Missing
value
indicator
999
999
9999
999
999
999
9999
999
9
9
9
999
999
999
9999
99
99
99
99
99
99
999
999
999
99
999
999
999
999
-9999
-9999
-9999
-9999
99
-9
-9
-9
-9
-9
Bounds
Lower
-50
0
0
0
0
0
0
-50
-2
-2
-2
0
0
0
0
0
0
0
0
0
-30
0
0
0
-65
0
0
0
0
2700
9000
9000
0
0
1
1
0
0
0
Upper
800
2
4000
2
250
100
1250
800
5
5
5
100
100
100
4000
35
25
3
3
3
35
360
50
5
35
100
100
100
100
3200
10999
10999
300
10
31
12
99
24
60
'(U - L) indicates (upper level) - (lower level).
'Automatically included in audit report.
Note that the units for temperature difference are °C. However, the range check for
temperature differences is based on the temperature gradient in °C/(100 meters).
Notes: The nn in variables HT to V3 refers to the level at which the observation was
taken; e.g., TT01 is temperature at the first level, WS02 is wind speed at the
second level.
The three pressure variables (ALTP, SLVP, and PRES), the ceiling height (CLHT),
and the sky cover (TSKC) are also available on the surface pathway.
C-5
October 1994
-------
If upper air soundings are to be extracted, an upper
height limit is used above which no data are extracted. The
default height limit is 5000 meters. The value of the height
limit is stored in a variable named UATOP- The user can
override this value by specifying a new height on the UA TOP
input image for the upper air pathway. A description of this
image and its syntax can be found in Appendix B.
NOTE: The maxinun number of levels that can be extracted is
set in the processor to 20. If 20 levels of data are
extracted and UATOP has not been reached, no additional
data are extracted.
On the OS-pathway there are several default variables
having default values that can be redefined by the user through
the run stream input data. The number of observations per hour
is assumed to be 1. This can be changed via the OS AVG input
image. The number can range from 1 to 12. Albedo, Bowen
ratio, and surface roughness also have default settings. The
default albedo is 0.25 for all months of the year; the default
Bowen ratio is 0.75; and the surface roughness length defaults
to 0.15 meters for all wind directions and months. All of
these can be expanded or modified by using the OS SFC input
image. The tutorial in Section 4 and Appendix B have more
details on the syntax of these images. The default threshold
wind speed for use in definition of calm winds is 1.0
meters/second. This value can be changed with the OS CLM input
image.
C~6 October 1994
-------
APPENDIX D
EXAMPLES OF REPORTS
October 1994
-------
METEOROLOGICAL PROCESSOR FOR REGULATORY MODELS (MPRM), VERSION 1.2
TODAY'S DATE AND TIME: 18-APR-88 AT 14:03:29
INITIAL PROCESSING ON RAW METEOROLOGICAL DATA
STATUS REPORT PRIOR TO BEGINNING PROCESSOR RUN
1. REPORT FILE NAMES
ERROR MESSAGES: ERROR.LIS
SUMMARY OF RUN: STANDARD OUTPUT DEVICE, UNIT 6
2. UPPER AIR DATA
SITE ID LATITUDE(DEG.) LONGITUDE(DEC.)
93815 39.07N 84.68U
THE PROCESS(ES) MPRM ANTICIPATES TO PERFORM ARE
EXTRACT AND QUALITY ASSESSMENT
EXTRACT INPUT -
EXTRACT OUTPUT-
QA OUTPUT
OPEN: RAMZI.DAT
OPEN: IQAZIBRF.DAT
OPEN: OQAZIBRF.DAT
STARTING: 31-DEC-63
ENDING: 5-JAN-64
THE EXTRACT DATES ARE:
3. NUS SURFACE DATA
SITE ID LATITUDE(DEC.) LONGITUDE(DEC.)
93814 39.DON 84.00U
THE PROCESS(ES) MPRM ANTICIPATES TO PERFORM ARE
EXTRACT AND QUALITY ASSESSMENT
EXTRACT INPUT -
EXTRACT OUTPUT-
QA OUTPUT
OPEN: RAMSFC.DAT
OPEN: IQASFBRF.DAT
OPEN: OQASFBRF.DAT
THE EXTRACT DATES ARE:
ON-SITE DATA
STARTING: 1-JAN-64
ENDING: 4-JAN-64
THE PROCESS(ES) MPRM ANTICIPATES TO PERFORM ARE
NONE, NO DATA TO BE PROCESSED ON THIS PATH
Figure D-la. First page of general report for 4-day example
of stage 1 processing used in Tutorial.
D-2
October 1994
-------
METEOROLOGICAL PROCESSOR FOR REGULATORY MODELS (MPRM), VERSION 1.2
TODAY'S DATE AND TIME: 18-APR-88 AT 14:03:38
INITIAL PROCESSING ON RAW METEOROLOGICAL DATA
********************************************************
*** JOB TERMINATED NORMALLY ***
********************************************************
JB
E
U
I
UA
E
W
I
Q
SF
E
U
I
Q
0- 9
0
0
0
0
0
0
0
0
0
0
0
0
****
10-19
0
0
4
0
0
. 0
0
0
0
0
0
4
**** MPRM
20-29
0
0
0
0
0
0
0
0
0
0
0
0
WARNING MESSAGES
MESSAGE SUMMARY TABLE ****
30-39
0
0
0
0
0
4
4
0
0
0
0
8
****
40-49
0
0
0
0
0
0
0
0
0
4
0
4
50-59
0
0
0
0
0
0
0
0
0
0
0
0
60-69
0
0
0
0
0
0
0
0
0
0
0
0
70-79
0
0
0
0
0
0
0
0
0
0
0
0
TOTAL
0
0
4
0
0
4
4
0
0
4
0
16
--- NONE ---
****
ERROR
MESSAGES
****
--- NONE ---
Figure D-lb. Second page of general report for 4-day example
of Stage 1 processing used in Tutorial.
D-3
October 1994
-------
METEOROLOGICAL PROCESSOR FOR REGULATORY MODELS (MPRH), VERSION 1.2
TODAY'S DATE AND TIME: 18-APR-88 AT 14:03:40
********************************************************
*** JOB TERMINATED NORMALLY ***
******************************************************<*
**** SUMMARY OF THE QA AUDIT ****
MIXING NTS | VIOLATION SUMMARY |
TOTAL # LOWER UPPER X
# OBS MISSING BOUND BOUND ACCEPTED
UAM1 6 001 83.33
UAM2 6 030 50.00
THERE IS NO AUDIT TRAIL FOR
SOUNDINGS
SURFACE DATA
SLVP
PRES
CLHT
TS
KC
HZVS
TMPD
WD16
WIND
TOTAL
# OBS
96
96
96
96
96
96
96
96
96
I-
-VIOLATION SUMMARY-
LOWER UPPER
MISSING
0
0
0
0
0
0
0
0
0
BOUND
0
0
0
0
0
0
0
0
0
BOUND
0
0
0
0
0
0
0
0
0
ACCEPTED
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
j TEST VALUES |
MISSING LOWER UPPER
FLAG BOUND BOUND
-9999.0, 50.0, 500.0
-9999.0, 500.0, 3500.0
| TEST VALUES |
MISSING LOWER UPPER
FLAG BOUND BOUND
-9999.0, 9000.0,10999.0
-9999.0, 9000.0,10999.0
-9999.0,
99.0,
99.0,
-9999.0,
-9999.0,
-9999.0,
-9999.0,
0.0,
0.0,
0.0.
0.0,
-300.0,
o.o,
0.0,
300.0
10.0
10.0
1640.0
350.0
36.0
500.0
NOTE: TEST VALUES MATCH INTERNAL SCALING APPLIED TO VARIABLES
(SEE APPENDIX C OF THE USER'S GUIDE)
THE FOLLOWING CHECKS WERE ALSO PERFORMED FOR THE SURFACE QA
OF 96 REPORTS, THERE WERE
0 CALM WIND CONDITIONS (WS=0, WD=0)
0 ZERO WIND SPEEDS WITH NONZERO WIND DIRECTIONS
0 DEW-POINT GREATER THAN DRY BULB TEMPERATURES
THE TIMES OF THESE OCCURRENCES CAN BE FOUND IN THE MESSAGE FILE
WITH QUALIFIERS CLM, WDS, TDT (RESP.)
THIS CONCLUDES THE AUDIT TRAIL
Figure D-lc. Third page of general report for 4-day example
of Stage 1 processing used in Tutorial.
D-4
October 1994
-------
METEOROLOGICAL PROCESSOR FOR REGULATORY MODELS (MPRM), VERSION 1.2
TODAY'S DATE AND TIME: 22-APR-88 AT 10:09:03
INITIAL PROCESSING ON RAW METEOROLOGICAL DATA
STATUS REPORT PRIOR TO BEGINNING PROCESSOR RUN
1. REPORT FILE NAMES
ERROR MESSAGES: ERROR.LIS
SUMMARY OF RUN: STANDARD OUTPUT DEVICE, UNIT 6
2. UPPER AIR DATA
THE PROCESS(ES) MPRM ANTICIPATES TO PERFORM ARE
MERGE ONLY
QA OUTPUT
OPEN: OQAZIBRF.DAT
3. NWS SURFACE DATA
THE PROCESS(ES) MPRM ANTICIPATES TO PERFORM ARE
MERGE ONLY
QA OUTPUT
4. ON-SITE DATA
SITE ID
9381464
OPEN: OQASFBRF.DAT
LATITUDE(DEC.)
39.07N
LONGITUDE(DEC.)
84.68U
THE PROCESS(ES) MPRM ANTICIPATES TO PERFORM ARE
NONE, NO DATA TO BE PROCESSED ON THIS PATH
5.
MERGED DATA
MERGE OUTPUT
OPEN: MERGEBRF.DAT
Figure D-2a.
First page of general report for 4-day example
of stage 2 processing used in Tutorial.
D-5
October 1994
-------
METEOROLOGICAL PROCESSOR FOR REGULATORY MODELS (MPRH), VERSION 1.2
TODAY'S DATE AND TIME: 22-APR-88 AT 10:09:03
***** USER INPUT PARAMETERS FOR MERGE *****
MERGED DATA BEGIN (YR/MO/DA) 64/ M 1
AND END 64/ 1/ 4
THE ON-SITE LATITUDE AND LONGITUDE ARE:
LATITUDE 39.07N
LONGITUDE 84.68U
DAILY OUTPUT STATISTICS *****
1/1 1/2 1/3
MO/DA
NUS UPPER AIR SDGS 0
NCDC MIXING HEIGHTS 6
NWS SFC OBSERVATIONS 24
ON-SITE OBSERVATIONS 0
UPPER AIR OBS. READ:
SURFACE OBS. READ:
ON-SITE OBS. READ:
0
6
24
0
6
96
0
0
6
24
0
0
6
23
0
Figure D-2b.
Second page of general report for 4-day example
of Stage 2 processing used in Tutorial.
D-6
October 1994
-------
METEOROLOGICAL PROCESSOR FOR REGULATORY MODELS (MPRM), VERSION
1.2
TODAY'S DATE AND TIME: 22-APR-88 AT 10:09:09
INITIAL PROCESSING ON RAW METEOROLOGICAL DATA
********************************************************
*** JOB TERMINATED NORMALLY ***
********************************************************
**** MPRM
0- 9 10-19 20-29
JB
E 0 0 0
U 0 0 0
10 8 0
080
**** WARNING MESSAGES
--- NONE ---
**** ERROR MESSAGES
--- NONE ---
MESSAGE SUMMARY TABLE ****
30-39 40-49 50-59 60-69 70-79
00000
00000
00000
00000
****
****
TOTAL
0
0
8
8
Figure D-2c. Third page of general report for 4-day example
of Stage 2 processing used in Tutorial.
D-7
October 1994
-------
METEOROLOGICAL PROCESSOR FOR REGULATORY MODELS (MPRM), VERSION 1.2
TODAY'S DATE AND TIME: 28-APR-88 AT 15:23:37
PROCESSING OF MERGED METEOROLOGICAL DATA
METEOROLOGICAL JOINT FREQUENCY
STABILITY CATEGORY
WIND DIRECTION
N
NNE
NE
ENE
E
ESE
SE
SSE
S
ssw
su
usu
u
UNU
NU
NNU
STABILITY CATEGORY
WIND DIRECTION
N
NNE
NE
ENE
E
ESE
SE
SSE
S
SSW
su
wsu
w
UNU
NU
NNU
DN
SECTOR
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
EF
SECTOR
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
FUNCTION
WIND SPEED CLASS
1
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
2
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
3
0.000000
0.010526
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.010526
0.000000
0.000000
0.000000
4
0.000000
0.010526
0.052632
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.115789
0.105263
0.000000
0.010526
0.021053
0.000000
0.000000
5
0.000000
0.000000
0.010526
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.073684
0.000000
0.000000
0.010526
0.000000
0.000000
6
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
UIND SPEED CLASS
1
0.000000
o.oooood
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
2
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.052632
0.021053
0.000000
0.000000
0.000000
0.000000
0.000000
3
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.010526
0.010526
0.063158
0.031579
0.000000
0.000000
0.000000
4
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
5
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
6
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
Figure D-3a. Portion of listing of meteorological data
generated for the CDM16 dispersion model from
Stage 3 processing (4-day example for Tutorial)
D-8
October 1994
-------
METEOROLOGICAL PROCESSOR FOR REGULATORY MODELS (MPRM), VERSION 1.2
TODAY'S DATE AMD TIME: 28-APR-88 AT 15:23:37
PROCESSING OF MERGED METEOROLOGICAL DATA
METEOROLOGICAL JOINT FREQUENCY
STABILITY CATEGORY DN
UIND DIRECTION
355-005
005-015
015-025
025-035
035-045
045-055
055-065
065-075
075-085
085-095
095-105
105-115
115-125
125-135
135-145
145-155
155-165
165-175
175-185
185-195
195-205
205-215
215-225
225-235
235-245
245-255
255-265
265-275
275-285
285-295
295-305
305-315
315-325
325-335
335-345
345-355
SECTOR
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
1
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
2
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
FUNCTION
WIND SPEED CLASS
3
0.000000
0.000000
0.000000
0.010526
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.010526
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
4
0.000000
0.000000
0.000000
0.010526
0.031579
0.021053
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.031579
0.073684
0.073684
0.031579
0.000000
0.000000
0.000000
0.010526
0.000000
0.021053
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
5
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.010526
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.010526
0.063158
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.010526
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
6
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0/000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
Figure D-3b.
Portion of listing of meteorological data
generated for the COM36 dispersion model from
Stage 3 processing (4-day example for Tutorial)
D-9
October 1994
-------
METEOROLOGICAL PROCESSOR FOR REGULATORY MODELS (MPRM), VERSION 1.2
TODAY'S DATE AND TIME: 18-APR-88 AT 07:56:39
PROCESSING OF MERGED METEOROLOGICAL DATA
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
1 1
1 2
1 3
1 4
1 5
1 6
1 7
1 8
1 9
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
2 1
2 2
2 3
2 4
2 5
2 6
2 7
2 8
2 9
210
211
212
213
214
215
31.
48.
44.
43.
33.
42.
55.
53.
7.
331.
344.
356.
343.
299.
292.
274.
261.
237.
224.
227.
210.
232.
210.
200.
206.
222.
222.
230.
216.
217.
219.
216.
220.
221.
235.
211.
209.
207.
200.
9.2
15.0
15.0
12.7
12.7
13.9
20.8
17.2
13.9
10.3
12.7
13.9
10.3
8.1
12.7
13.9
10.3
8.1
9.2
12.7
18.3
16.1
13.9
17.2
16.1
13.9
16.1
18.3
15.0
16.1
15.0
12.7
15.0
15.0
15.0
15.0
16.1
16.1
15.0
511.
483.
456.
429.
401.
374.
347.
319.
292.
264.
237.
210.
182.
155.
155.
155.
155.
168.
189.
210.
231.
252.
272.
293.
314.
335.
355.
376.
397.
418.
438.
459.
480.
501.
522.
542.
563.
584.
584.
4.
4.
4.
4.
4.
4.
4.
4.
4.
4.
4.
4.
4.
4.
4.
4.
4.
5.
5.
4.
4.
4.
4.
4.
4.
4.
4.
4.
4.
4.
4.
4.
4.
4.
4.
4.
4.
4.
4.
23.0
21.9
23.0
23.0
24.1
24.1
26.1
26.1
25.0
24.1
26.1
27.0
27.0
28.9
28.9
28.9
28.0
21.9
21.9
21.9
24.1
25.0
23.0
24.1
24.1
23.0
23.0
23.0
23.0
23.0
23.0
21.9
23.0
26.1
28.9
33.1
35.1
37.0
39.9
Figure D-3c. Portion of listing of meteorological data
generated for the RTDM dispersion model from
Stage 3 processing (4-day example for Tutorial)
D-10
October 1994
-------
METEOROLOGICAL PROCESSOR FOR REGULATORY MODELS (MPRM), VERSION 1.2
TODAY'S DATE AND TIME: 18-APR-88 AT 07:58:02
PROCESSING OF MERGED METEOROLOGICAL DATA
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
1 1
1 2
1 3
1 4
1 5
1 6
1 7
1 8
1 9
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
2 1
2 2
2 3
2 4
2 5
2 6
2 7
2 8
2 9
210
211
212
213
214
215
216
217
218
219
220
221
222
223
4.1
6.7
6.7
5.7
5.7
6.2
9.3
7.7
6.2
4.6
5.7
6.2
4.6
3.6
5.7
6.2
4.6
3.6
4.1
5.7
8.2
7.2
6.2
7.7
7.2
6.2
7.2
8.2
6.7
7.2
6.7
5.7
6.7
6.7
6.7
6.7
7.2
7.2
6.7
7.2
8.2
6.2
6.7
6.2
8.8
9.3
9.8
31.0
48.0
44.0
43.0
33.0
42.0
55.0
53.0
7.0
331.0
344.0
356.0
343.0
299.0
292.0
274.0
261.0
237.0
224.0
227.0
210.0
232.0
210.0
200.0
206.0
222.0
222.0
230.0
216.0
217.0
219.0
216.0
220.0
221.0
235.0
211.0
209.0
207.0
200.0
204.0
198.0
201.0
202.0
205.0
214.0
215.0
217.0
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
5
5
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
1000.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Figure D-3d. Portion of listing of meteorological data
generated for the CALINE-3 dispersion model from
Stage 3 processing (4-day example for Tutorial).
D-ll
October 1994
-------
13 JB 119 SETUP: ENCOUNTERED END OF "JOB/RUN CARD"
0 JB U12 UAQAST: SUMMARY: MISSING/ERRORS IN UA-OQA CARD
0 JB 110 TEST: SUMMARY: NO SF-EXT CARD, NULL EXTRACT
0 JB 111 TEST: SUMMARY: NO SF-IQA CARD, NULL QA
0 JB 112 TEST: SUMMARY: NO SF-OQA CARD, NULL MERGE
0 JB 110 TEST: SUMMARY: NO OS-EXT CARD, NULL EXTRACT
0 JB 111 TEST: SUMMARY: NO OS-IQA CARD, NULL QA
0 JB 112 TEST: SUMMARY: NO OS-OQA CARD, NULL MERGE
0 UA 130 UAEXT: **** UPPER AIR EXTRACTION ****
0 UA 136 UAEXT:* *** AUTOMATIC SDG. CHECKS ARE ON
0 UA 139 GETSDG: END-OF-FILE, END-OF-DATA
0 UA W38 GETSDG: NO OBS. RETRIEVED; SEE UNIT 60 FOR SCAN
0 UA 139 GETMIX: END-OF-FILE, END-OF-DATA
0 UA W38 GETMIX: NO OBS RETRIEVED-CHECK INPUT STN/DATES
0 UA 130 UAEXT: 0 SDGS AND 0 MIXING HTS EXTRACTED
TAPE CONTENTS FOR UNIT 10, FILE:TAPE10
FROM
STATION YR MO DA HR JDAY
13840
13840
13840
13840
13840
13840
13840
64 1 1 0
65 1 3 0
66 1 2 0
67 1 3 12
68 1 1 12
69 1 2
70 1 3
1
3
2
3
1
2
3
TO
MO DA HR
12 31 12
12 30
12 31
12 30
12 30 12
12 31 12
12 30 12
12
12
0
JDAY
366
364
365
364
365
365
364
Figure D-4.
Listing of an error/message file for Stage 1
processing with a report of tape contents for
NWS upper air data. Tape contents are reported
only if the specified station and dates are not
located.
D-12
October 1994
-------
APPENDIX E
SUMMARY OF ERROR AND WARNING MESSAGES
During processing of the input images and the data, the
processor writes messages to the error/message file defined in
the JB ERR input image (see Appendix B for a description of
this image). Five types of messages can be generated:
o An error that stops the processor from completing the
original request for data processing; considered a fatal
error
o An error that may not stop processing
o Status of the processing
o Quality assessment violation
o A computation could not be performed during Stage 3
processing AND the trace option is on.
October 1994
-------
A message from the processor has the form:
N PP 8,8283 SSSSSS: message
* I
code (up to 40 characters)
Subroutine from which the message is generated
More detailed description of the message
Alphanumeric message code
Pathway ID (JB, UA, SF, OS, MR, MP)
Counter
The message code is composed of two parts - a leading
alphabetic character and a trailing 2-digit code. The
alphabetic character (aj above) can be:
E fatal error; if the error occurs during processing of
of input images, the remainder of the images are
processed to locate other possible problems with the
images; if the error occurs during processing of data
on a pathway, processing ceases on that pathway and
the next step defined by the input images begins
W warning; further data processing may or may not be
prohibited depending on the processing requested
through the input images;
I information on the status of the processing - these
messages monitor the progress of a processor run;
Q quality assessment violation - a value for a variable
was outside the interval defined by the upper and
lower bounds;
T an hourly computation could not be performed during
Stage 3 processing AND the hourly trace key was
turned on using the input image MP TRA.
E-2
October 1994
-------
The 2-digit codes (a2a3 in the message code) are grouped into
general categories corresponding to the processing in Stages 1,
2, and 3. These categories are
00-19 Input image processing
20-29 File header and library processing
30-39 Upper air soundings and mixing heights processing
40-49 Surface observations processing
50 - 59 On-site observations processing
60 - 69 Merge processing
70-79 Stage 3 processing
** program trap - there is an error in program
logic; the processor stops immediately
Within the general categories are codes pertaining to
processing in MPRM. These codes are more specific than the
general categories but do not completely specify the reason for
the message. That is left to the 40-character message. The
codes are summarized by pathway and severity and category at
the end of each processor run in a table that is written to the
file defined by the JB OUT input image or, if this image is
omitted, the default output device. The MPRM processor uses
device number (logical unit) 6 for this purpose.
INPUT IMAGE PROCESSING, 00 - 19
If an error is detected on an input image, a message is
written to the message file and any attempt to process data is
prohibited. If a warning occurs, data may or may not be
processed, depending on the processing requirements specified
within the run stream input data. For example, a warning with
a message code of W12 is written if there is no IQA image on a
pathway. If it is detected that the user wants to extract
E-3 October 1994
-------
data, processing will be halted as there is not output file
(defined by use of the IQA input) for the extracted data.
EDO A keyword field is blank
E01 Repeated keyword/Improper keyword
E02 Error reading an input image
E03 Error decoding a field on an input image
E04 Incomplete or superfluous information on an input
image
EOS Error in a field - a keyword could not be determined
E06 A value or character is not within bounds or is
unreasonable; improper information specified in a
field; no match with allowable names (see also code
W06 below)
E07 Error opening a tape or file; file not open; the file
name was specified for more than one device (logical
unit) number
E10 Fatal write error to the temporary file
Ell Pathway status (status = -1) does not allow further
processing
E12 Unable to proceed; previous errors on a pathway card
(from subroutine COMPLETE)
E13 The on-site data map and formats were not specified
E14 An attempt was made to change a previously
established value for Stage 3 processing
WOO A blank image was encountered in the input images
WO6 Value on an input card image may not be reasonable
(see also code E06 above)
WlO Non-fatal write error writing to the temporary file
W12 Missing/errors on an input image - may or may not be
fatal depending on the processing requested
W15 Audit disabled for the on-site variable specified
110 No extraction on the pathway specified
E~4 October 1994
-------
Ill No quality assessment on the pathway specified
112 No file on the pathway specified to merge
119 End-of-file or JB END card encountered
FILE HEADER AND LIBRARY PROCESSING, 20-29
Any messages that pertain to writing file headers during
the initial processing or messages issued by the library
routines (those routines accessed by more than one subroutine)
are in this category.
E20 Header read error
E21 Header write error
E22 From subroutine FLHEAD: there were errors on the
input file, so there are no data to process
E23 From subroutine FLHEAD: error reading the headers
placed in the output file by input image processing
E24 From subroutines CHROND or ICHRND: there were
problems in computing the chronological day from
year, month and day, or computing the year, month and
day from the chronological day; the subroutine that
generates the message is identified
W22 From subroutine FLHEAD: An end-of-file was
encountered reading the headers on an input file,
there is no processing of data from this pathway
123 From subroutine FLHEAD: an end-of-file reading the
headers from the output file - a correct condition
UPPER AIR PROCESSING, 30-39
Any messages that pertain to the upper air pathway, and
issued after the input images are processed, are in this
category.
E-5
October 1994
-------
E32 There is an error reading or decoding the data and
the count now exceeds the maximum number of errors
allowed
E35 Data blocking type specified incorrectly - allowable
specifications are VB (variable block) and FB (fixed
block)
E36 Error reading the file headers during quality
assessment - no data processed
W32 There is an error reading or decoding the data and
the count is less than the maximum number allowed -
processing continues
W33 Sounding surface height is less than zero; the height
is set to zero
W38 No soundings or mixing heights were retrieved because
there was no match found in the data with the station
ID specified in the run stream input.
130 A message indicating that the point has been reached
where processing of the UA-data can begin
131 Automatic data modification for upper air soundings
is enabled (see section 6 for a discussion of these
modifications)
132 No data were extracted; a report of the tape/file
contents is written to the error/message file
137 A mixing height quality assessment lower bound
violation; this message code is written for the
second and subsequent encounters of the mixing
heights within a day
138 A mixing height quality assessment upper bound
violation; this message code is written for the
second and subsequent encounters of the mixing
heights within a day
139 An end-of-file was encountered on the input file in
the -expected position in the file
E-6
October 1994
-------
Q34 The vertical gradients cannot be computed at one or
more heights because one or more heights are missing
Q35 The sounding height is not in any of the height
intervals defined for the upper air audit; the
processor STOPS immediately and the last message in
the message file will identify the location of the
problem; neither a summary table nor an audit table
is generated
Q36 From subroutine HTCALC: the heights have not been
recomputed due to missing data
Q37 A lower bound quality assessment violation
Q38 An upper bound quality assessment violation
SURFACE OBSERVATIONS PROCESSING, 40-49
Any messages that pertain to the surface pathway, and
issued after the input images are processed, are in this
category-
E42 There is an error reading or decoding the data and
the count now exceeds the maximum number of errors
allowed
E46 Error reading the file headers during quality
assessment - no data processed
W42 There is an error reading or decoding the data and
the count is less than the maximum number allowed -
processing continues
W43 There is an error decoding an over-punch character;
the missing value indicator for that variable will be
inserted into the output file
W48 No surface observations were retrieved because there
was no match found in the data with the station ID
specified in the run stream input.
E-7
October 1994
-------
140 A message indicating that the point has been reached
that processing of the SF-data can begin
148 No data were extracted; a report of the tape/file
contents is written to the error/message file
149 An end-of-file was encountered on the input file in
the expected position in the file
Q47 A lower bound quality assessment violation
Q48 An upper bound quality assessment violation
ON-SITE OBSERVATIONS PROCESSING, 50 - 59
Any messages that pertain to the on-site pathway, and
issued after the input images are processed, are in this
category
E50 There is an error reading an input file header
E51 There is an error writing an input file header to the
output file
E52 There is an error reading or decoding the data and
the count now exceeds the maximum number of errors
allowed
E53 There is an error writing data to the output file
E54 The observations are not sequential in time
E55 The number of observations exceeds the number
expected for the hour, defined as 1 (default) or on
the input image OS AVG (note maximum of 12 allowed)
E56 An end-of-file on the input data was encountered
before one complete observation was read
W52 There is an error reading the data and the count is
less than the maximum number allowed - processing
continues
E-8
October 1994
-------
157 An intra-hour observation violated a quality
assessment lower bound
158 An intra-hour observation violated a quality
assessment upper bound
159 An end-of-file was encountered on the input file in
the expected position in the file
Q57 Lower bound quality assessment violation
Q58 Upper bound quality assessment violation
MERGE PROCESSING, 60 - 69
Any messages pertaining to combining the three data types
(merge), and issued after the input images are processed, are
in this category.
E60 There is an error computing the chronological day
from Julian day and year
E61 There is an error computing the Julian day and year
from the chronological day
E62 There is an error reading the upper air data
E63 There is an error reading the surface data
E64 There is an error reading the on-site data
E65 There is an error writing the on-site data to the
output file
E66 There is an error processing an input file's header
cards
E67 No chronological days for merging were computed
167 The beginning chronological day is computed from the
earliest available date on the three pathways; the
ending chronological day = beginning day + 367. This
message is generated if no MR EXT input is defined.
E-9 October 1994
-------
STAGE 3 PROCESSING, 70-79
Any messages that pertain to Stage 3 processing, and
issued after the input images are processed, are in this
category.
E70 The preliminary processing (such as the latitude and
longitude) has produced an error; the input file has
no data; the 40-character message further identifies
the source of the error
E71 There is an error reading the input data; the data
are not a 1 - 24 hour clock; the hour for on-site
data is represented by the missing value indicator
W70 During preliminary processing, a value does not
appear correct (e.g., the GMT to LST conversion
factor); hour 23 data have been swapped in for hour
24 data for the surface pathway
W72 The mixing heights were not computed for the
specified number of hours
W73 The temperatures cannot be determined for the
specified number of hours
W74 The winds cannot be determined for the specified
number of hours
W75 The stability categories cannot be determined for the
specified number of hours
W76 The surface roughness length cannot be determined for
the number of hours specified
170 End of header group reached on input file
171 Station location for pathway changed from characters
to 0 (zero) to conform to RAMMET output
specifications
October 1994
-------
175 Missing data has resulted in use of an alternate
estimation scheme for stability, the alternate scheme
is reported in message
179 The end of the processing window, defined by the JB
EXT image, was encountered or, if no window was
specified, the end-of-file was encountered
The following are written only if the JB TRA image is
present for Stage 3 processing
T72 The mixing height cannot be computed for the
specified hour
T73 The temperature cannot be determined for the
specified hour
T74 The winds cannot be determined for the specified hour
T75 The stability category cannot be computed for the
specified hour and methodology
T76 The surface roughness length cannot be determined for
the specified hour
October 1994
-------
APPENDIX F
FORMAT OF DATA FILES
The upper air and surface observations are written in a
specific format after the data are extracted. These formats
are retained for the remainder of the processing until the
merging of the data (Stage 2 processing). Because the user
specifies the data map (format) of the on-site data, this
discussion does not apply to the on-site data.
The format of meteorological data files generated by Stage
3 processing are specific to the dispersion models. Details
regarding the format should be obtained from the respective
user's guides. A brief summary of the format is provided for
informational purposes.
UPPER AIR
An extracted upper air data report is composed of two
parts:
o A header record consisting of a year, month, day, hour
group, the number of sounding levels, and the morning and
evening mixing heights
o Sounding data, if soundings were extracted, consisting of
a pressure, height above ground level, temperature,
dew-point temperature, wind speed, and wind direction.
In the READ statements that accompany the FORMAT
statements below, u refers to the device (logical unit) number
and f refers to the format statement number. The variable
F-l
October 1994
-------
names are also shown for those variables that appear in
Appendix C. The names given for the date/time group and the
number of levels in an upper air sounding are generic names and
do not correspond to any of the names in Appendix C.
The format of the header record is:
READ(u,
FORMAT
f)
yr, mo, da, hr, lev, UAH1, UAH2
<1X. 12, 1
year '
(2-digit)
month
da\
2.
t
2.
2.
5, 1X,
* 01
5, 1X.I5)
I p.m. mixing
(meters)
a.m. mixing
(meters)
: sound ina Leve
height
Is in tt
height
lis PCD
(this number is 0 if no soundings were
extracted or there are no levels of data)
hour of the observation in Local Standard
Time (1ST) (the hour is set to 12 if only
mixing heights are extracted)
The format of the levels of data, if they exist, is;
READ (u.f) UAPR, UAHT, UATT, UATD, UAWD. UAUS
FORMAT (1X.I5. 1X.I5, 1X.I5, 1X.I5. 1X.I5, 1X.I5)
atmospheric
pressure
(millibars)*
J
height above
ground level
(AGL) (meters)
L
wind speed
(meters/second)*
wind direction
(tens of degrees from north)
dew-point temperature
(degrees Celsius)*
dry bulb temperature
(degrees Celsius)*
All values on the upper air pathway are written as
integers. The asterisk (*) following four of the descriptions
above indicate that the values were multiplied by 10 to retain
one significant digit after the decimal point prior to rounding
the result to the nearest whole number. The READ uses
4-character names to identify the variables. Tables of these
names with the default parameters are in Appendix C.
F-2
October 1994
-------
SURFACE OBSERVATIONS
Each hourly surface observation is written as two records.
As with the upper air data, all values are reported as integers
with several variables being multiplied by 10 or 100 to retain
significant digits. Several of the variables are two variables
combined to form one integer value. These are recognized by
the // in the variable name and units below. The single
asterisk next to the units indicates that the value written to
the output file is multiplied by 10; the double asterisk next
to the altimeter pressure indicates that the value is
multiplied by 100.
The format of the first record of the surface observations
is:
READ(u,f) yr,mo,da,hr,ALTP,SLVP,PRES,CLHT,TSKC.C2C3,CLCl,CLC2,CLC3,CLC4
FORMAT(1X,I2,I2,I2,I2,1X,I5,1X,I5,1X.I5,1X,I5,1X,I5.5,1X.I5.5,4(1X,I5.5))
year, month,
day, hour (LST)
altimeter pressure
(in. Hg)**
sea level pressure
(millibars)*
station pressure
(millibars)*
cloud ceiling height '
(kilometers)*
sky cond.//
coverage, 4
layers
(--//tenths)
sky cover,
2//3 layers
(tenths//tenths)
sky cover,
total//opaque
(tenths//tenths)
F-3
October 1994
-------
The format of the second record of the surface
observations is:
READ (u.f) CLT1,CLT2,CLT3,CLT4,PWTHfHZVS,TMPD,THPU,DPTP,RHUM,UDl6,WIND
FORMAT(8X,4(1X,I5.5),1X,I5.5f1X.I5f1XfI5flXfI5f1X,I5.1XfI5,1Xf!5,1X,l5)
cloud type//height,.
4 layers
(---//kilometers)*
present weather,
2 types - (no units
horizontal visibility
(kilometers)*
dry bulb temperature
(degrees Celsius)*
lire
*
iroerati
jre
wind speed
(meters/
second)*
wind direction
(tens of degrees
from north)
- relative
humidity
(percent)
- dew-point
temperature
(degrees Celsius)*
All reports of sky conditions, cloud types and present
weather are converted to the TD-3280 numeric codes. These
conversions are performed automatically as a part of the
extraction process on the SF-pathway.
The following codes are used to report the sky conditions.
The corresponding code used in the CD-144 format is shown in
parentheses.
00 = clear or less than 0.1 coverage (0)
01 = thin scattered 0.1 to 0.5 coverage (1)
02 - scattered 0.1 to 0.5 coverage (2)
03 = thin broken 0.6 to 0.9 coverage (4)
04 = broken 0.6 to 0.9 coverage (5)
05 = thin overcast 1.0 coverage (7)
06 = overcast 1.0 coverage (8)
07 = obscuration 1.0 coverage (X or -)
08 = partial obscuration <1.0 coverage (blank)
09 = unknown
The following codes are used to report the cloud types for
all levels and obscuring phenomena. The corresponding code
used in the CD-144 format is shown in parentheses. If no code
follows the description, then there is no corresponding code in
the CD-144 format. The overpunch characters in the CD-144
F-4
October 1994
-------
format are represented by X/n where n is an integer. An
overpunch character as it appears in an ASCII file is also
shown.
Cloud types
00 = none (0)
11 = cumulus (4)
12 = towering cumulus
13 = stratus fractus (X/2 or K>
14 = stratus cumulus lenticular
15 = stratus cumulus (3)
16 - stratus (2)
17 = cumulus fractus (X/4 or N)
18 = cumulonimbus (5)
19 = cumulonimbus mammatus (X/5 or N)
21 = altostratus (6)
22 = nimbostratus (X/6 or 0)
23 = altocumulus (7)
24 = altocumulus lenticular
28 = altocumulus caste11anus (X/7 or P)
29 = altocumulus mamma tus
32 = cirrus (8)
35 = cirrocumulus lenticular
37 = cirrostratus (9)
39 = cirrocumulus (X/9 or R)
Obscuring phenomena
01 = blowing spray
03 = smoke and haze
04 = smoke
05 = haze
06 = dust
07 = blowing dust
30 = blowing sand
36 = blowing snow
44 = ground fog
45 = fog
48 = ice fog
50 = drizzle
60 = rain
70 = snow
76 = ice crystals
98 = obscuring phenomena other than fog, prior to 1984 (X or -)
The code definitions for present weather conditions are
presented below. They are divided into general categories and
each category is further divided for specific weather
conditions. Dashes in a field indicate that there is no
definition for that code. The 8-digit CD-144 format weather
conditions are converted to the 2-digit TD-3280 category.
F~5 October 1994
-------
1X - Thunderstorm, Tornado,
Squall
2X - Rain, Rain Shower,
Freezing Rain
= 2
= 3
= 4
= 5
= 6
= 7
= 8
= 9
thunderstorm - lightning
and thunder
severe thunderstorm -
frequent intense lightning
and thunder
report of tornado or
water spout
light squall
moderate squall
heavy squall
water spout
funnel cloud
tornado
unknown
light rain
moderate rain
heavy rain
light rain showers
moderate rain showers
heavy rain showers
light freezing rain
moderate freezing rain
heavy freezing rain
unknown
3X - Rain Squalls, Drizzle,
Freezing Drizzle
4X - Snow, Snow Pellets,
Ice crystals
X =
light rain squalls
moderate rain squalls
heavy rain squalls
light drizzle
moderate drizzle
heavy drizzle
light freezing drizzle
moderate freezing drizzle
heavy freezing drizzle
unknown
light snow
moderate snow
heavy snow
light snow pellets
moderate snow pellets
heavy snow pellets
light snow crystals
moderate snow crystals
heavy snow crystals
unknown
5X - Snow Shower, Snow
Squalls, Snow Grains
6X - Sleet, Sleet Shower, Hail
X =
light snow showers
moderate snow showers
heavy snow showers
light snow squalls
moderate snow squalls
heavy snow squalls
light snow grains
moderate snow grains
heavy snow grains
unknown
light ice pellet showers
moderate ice pellet showers
heavy ice pellet showers
light hail
moderate hail
heavy hail
light small hail
moderate small hail
heavy small hail
unknown
7X - Fog, Blowing Dust,
Blowing Sand
8X - Smoke, Haze, Blowing Snow,
Blowing Spray, Dust
X = 0
= 1
= 2
= 3
= 4
= 5
= 6
= 7
= 8
= 9
fog
ice fog
ground fog
blowing dust
blowing sand
heavy fog
glaze
heavy ice fog
heavy ground fog
unknown
smoke
haze
smoke and haze
dust
blowing snow
blowing spray
dust storm
--
--
unknown
F-6
October 1994
-------
9X - Ice Pellets
X = 0
= 1
= 2
= 3
= 4
= 5
= 6
= 7
= 8
= 9
light ice pellets
moderate ice pellets
heavy ice pellets
--
--
--
--
--
--
unknown
RAMMET METEOROLOGICAL DATA
This format accommodates several dispersion models: BLP,
RAM, ISCST, MPTER, CRSTER, and COMPLEX1. The file contains two
types of records, the first is a header record and the second
is the meteorological data. The second contains the data for
one 24-hour period (midnight to midnight) and is repeated until
all data are listed. The data are written unformatted to the
file.
The format of the header record is:
READ(u) ID1.IYEAR1.ID2.IYEAR2
Last 2 digits of beginning year of mixing
height data.
5-digit station identification of mixing
height data.
Last 2 digits of beginning year of hourly
surface data.
- 5-digit station identification of hourly
surface data.
F-7
October 1994
-------
The format of the meteorological records are:
REAO(u) IYEAR,MONTH,IDAY,PGSTAB,SPEED,TEMP,FLUVEC.RANFLU,MIXHGT
*
f ,
w
I
Array of mixing
heights (m)
| Array of randomized
flow vectors (to
nearest degree)
- Array of flow vectors (to
nearest 10 degrees)
- Array of temperatures (degrees
Kelvin)
- Array of wind speeds (m/s)
- Array of Pasquill stability categories
- Day of month (1-31)
- Month of year (1-12)
- Last 2 digits of year
The DIMENSION statements used to define the arrays are:
DIMENSION PGSTAB(24),SPEED(24).TEMP(24).FLUVEC(24),RANFLU(24).MIXHGT(2f24)
The first index in the MIXHGT array controls which of the two
mixing height values is referenced. MIXHGT(l,i) refers to the
rural mixing height values, where i equals from 1 to 24 and
refers to hour of day in local standard time. MIXHGT(2,i)
refers to the urban mixing height values.
Preset values are used to indicate missing data; they are;
PGSTAB
SPEED
TEMP
FLWVEC
RANFLW
MIXHGT
0
-9
-99
-99
-99
-999
F-8
October 1994
-------
ISCST2 ASCII METEOROLOGICAL DATA
This format is an option available with ISCST2. The
format of the record is:
Variable
Year (2 digits)
Month
Day
Hour
Flow Vector (deg.)
Wind Speed (m/s)
Ambient Temperature
Stability Class
Rural Mixing Height
Urban Mixing Height
Format
(K)
(m)
(m)
12
12
12
12
F9.4
F9.4
F6.1
12
F7.1
F7.1
Columns
1-2
3-4
5-6
7-8
9-17
18-26
27-32
33-34
35-41
42-48
CALINE-3 METEOROLOGICAL DATA
This format is specific to the CALINE-3 dispersion model,
The file contains only one type of formatted data record, one
for each hour. The format of the record is:
READ(u.f) WNDSPDfUNDDIR,PGSTAB,NIXHGT,BCKGRD
I Background concentration (ppm)
Nixing height (m)
L Pasquill Stability category
Wind direction (to nearest degree)
i- Wind speed (m/s)
FORMATC F3.0,F4.0,I1.F6.0,F4.0 )
The preset values to indicate missing date are:
WNDSPD
WNDDIR
PGSTAB
MIXHGT
BCKGRD
-9
-99
0
1000
0
F-9
October 1994
-------
RTDM METEOROLOGICAL DATA
This format is specific to the RTDM dispersion model
(default). The file contains only one type of formatted data
record, one for each hour. The format of the record is:
READ(u,f) IYEAR,IJDAY,IHOUR,UNDOIR,UNDSPD,MIXHGT,PGSTAB,TEMP
I Temperature (F)
! Pasquill category
L Mixing height
L- Wind speed (miles/hr)
L- Wind direction (degrees)
L- Hour of day (1ST)
! Julian day of year
Last 2 digits of year
FORMATC 12,13,12.1X,F6.0,F6.1,2F6.0.F6.1 )
The input variables listed above are the only ones allowed
by current regulatory guidance. The RTDM dispersion model
provides for specification of other meteorological variables
but these require quite special meteorological observations, or
at the very least an intimate knowledge of the meteorological
conditions appropriate to the dispersion problem to be modeled.
The preset values to indicate missing data are:
WNDDIR
WNDSPD
MIXHGT
PGSTAB
TEMP
-999
-999
-999
-999
-999
F-10
October 1994
-------
16-SECTOR JOINT FREQUENCY FUNCTION
For the dispersion models VALLEY, ISCLT and CDM16, the
input file describing the meteorological conditions is a joint
frequency function. The frequency function is constructed
using 16 wind direction sectors, with the first 22.5° sector
centered on winds from the North (increasing clockwise), six
wind speed classes and six stability classes. The wind speed
classes are 0-3, 3-6, 6-10, 10-16, 16-21 and >21 kts. The
Pasquill stability categories for the CDM16 dispersion model
are grouped into classes as:
Pasqui11
Class category Remarks
1 A Very unstable conditions
2 B Moderately unstable conditions
3 C Slightly unstable conditions
4 D (daytime) Neutral conditions (sunrise to sunset)
5 D (nighttime) Neutral conditions (sunset to sunrise)
6 E-F All stable conditions
The Pasquill stability categories for the ISCLT and VALLEY
dispersion models are grouped into classes as:
Pasqui11
Class category Remarks
1 A Very unstable conditions
2 B Moderately unstable conditions
3 C Slightly unstable conditions
4 D Neutral conditions
5 E Stable conditions
6 F Very stable conditions
The format of the meteorological file is:
LOOP ON 1=1,6
LOOP ON K=l,16
October 1994
-------
READ(u.f) FREQC (I,J,K),J=1,6 )
I Index associated with wind speed class
Index associated with wind direction sector
>- Index associated with stability class
- Frequency of occurrence (decimal), of stability class I, with
wind speed class J, for wind from wind sector K
FORMAT( 9X.6F9.6 )
Hence the meteorological files consist of 96 records, the first
16 are for stability class 1, the next 16 are for stability
class 2, and so forth.
36-SECTOR JOINT FREQUENCY FUNCTION
For the dispersion model COM36, the input file describing
the meteorological conditions is a joint frequency function.
The frequency function is constructed using 36 wind direction
sectors, with the first 10° sector centered on winds from the
North (increasing clockwise), six wind speed classes and six
stability classes. The wind speed classes and stability
classes are the same as described above for the construction of
the 16-sector joint frequency function.
The format of the meteorological file is:
LOOP ON 1=1,6
LOOP ON K=l,36
REAO(u,f) FREQ( (I,J.K),J=1,6 )
I
Index associated with wind speed class
Index associated with wind direction sector
Index associated with stability class
Frequency of occurrence (decimal), of stability class I, with
wind speed class J, for wind from wind sector K
FORMAT( 9X.6F9.6 )
F-12
October 1994
-------
Hence the meteorological files consist of 216 records, the
first 36 are for stability class 1, the next 36 are for
stability class 2, and so forth.
F-13
October 1994
-------
GLOSSARY
ABNORMAL JOB TERMINATION this statement, if found in the
general report file, indicates that an error condition was
detected and further processing has been inhibited.
ASCII American Standard Code for Information Interchange.
AUD Input an input image used to add variables to the
default list of variables being tracked on the UA, SF or OS
pathway during quality assessment.
Audit Summary a written summary of the results for the
variables tracked (audited) during quality assessment.
Audit Variables variables that are tracked during quality
assessment.
BSS Bulletin Board System
BLP Buoyant Line and Point source dispersion model, from
Appendix A of the modeling guideline (EPA, 1986).
Bowen Ratio ratio of the upward flux of sensible heat to the
energy flux used in evaporation; a measure of the relative
evaporative power of the atmosphere.
CD-144 Format Card Deck-144 data format available from NCDC
for National Weather Service surface observations commonly used
for dispersion models. Each record represents an 80-column
"card image"-
GLOSSARY-1 October 1994
-------
COM Climatelogical Dispersion Model, from Appendix A of the
modeling guideline (EPA, 1986).
COMPLEXl A multiple source complex terrain dispersion model,
used as a second level screening model in regulatory modeling,
(EPA, 1986).
Convective Mixing mixing of atmospheric properties as a
result of surface heating.
CRSTER A single source dispersion model, from Appendix A of
the modeling guideline (EPA, 1986).
Dispersion Model A group of related mathematical algorithms
used to estimate (model) the dispersion of pollutants in the
atmosphere due to transport by the mean (average) wind and
small scale turbulence.
DOS Disk Operating System.
EBCDIC Extended Binary Coded Decimal Interchange Code.
EOF End-of-File.
EPA U. S. Environmental Protection Agency.
Error message a message written by the processor to the
error/message file whenever an error is encountered that will
inhibit data processing.
Error/Message File a file used in all stages of processing
for storage of messages written by the processor.
GLOSSARY-2 October 1994
-------
Extracted Data File the file resulting from Stage 1
processing for storage of data retrieved from a magnetic medium
(disk or tape).
Extraction Process the process of retrieving data from a
magnetic medium.
Fatal Error any error which inhibits further data processing
on a pathway or stops the MPRM processor.
File Headers records written by the processor at the top of
files during Stage 1 and Stage 2 processing. These records
contain the input images from individual pathways in addition
to supplementary records tracking the history of the data set.
Flow Vector The direction towards which the wind is blowing.
General Report File a file written either to the default
output device or a disk file summarizing the processor results.
GMT Greenwich Mean Time, the time at the 0° meridian.
Harmonic Average Wind Speed [S(l/Uj) /N]'1, where N is the
total number of observations and Uj is the i* wind speed
observation. The harmonic average wind speed is used by the
COM dispersion model in computing the effects of dilution.
Height Intervals heights used for reporting the results of
quality assessment of upper air data (see also Interval
Thickness).
Hypsometric Formula a determination of the height difference
between any two pressure levels based on hydrostatic balance,
which requires the mean virtual temperature of the layer.
GLOSSARY-3 October 1994
-------
Information(al) message any message written to the
error/message file that reports the status of the processing
and further data processing is not affected.
Initial Status Report the first page of the general report
for Stage 1 and Stage 2 processing.
Input Command Structure the syntax and sequence of the input
images.
Input Image user supplied input, read through the default
input device, controlling MPRM data processing.
Interval Thickness the height difference used in summarizing
the quality assessment results of upper air data (see also
Height Interval).
IQA Input to Quality Assessment, an input image that defines
the output file to receive extracted data. This file also
serves as the input file for the first pass through quality
assessment of the data.
ISCST Industrial Source Complex - Short Term dispersion
model, from Appendix A of the modeling guideline (EPA, 1986) .
JB JoB, the 2-character code indicating that all fields on
the input image pertain to the overall operation of the
processor.
JB Data collective term for all input images that begin with
the 2-character code JB.
JB Pathway collective term for logic associated with
deciphering input images beginning with the JB character code.
GLOSSARY-4 October 1994
-------
JCL Job Control Language, an IBM mainframe's operating
system control language for batch jobs.
Joint Frequency Function the joint frequency of wind
direction sector, wind speed class and stability category (see
also STAR).
Kb kilobyte, 1000 bytes
Keyword the 3-character codes that follow immediately after
the pathway ID in the input run stream data.
Library Routines a collection of subroutines that are called
by two or more subroutines and/or main program.
LST Local Standard Time.
Math Co-processor a computer chip used to speed up floating
point arithmetic in a personal computer.
Mb megabyte, one million bytes.
Merged Data File the file produced by Stage 2 processing
consisting of available upper air and mixing height data,
surface observations and on-site data for a specified period of
time.
Merge Processing the process by which data from the 3
pathways (UA, SF, OS) are combined to produce a merged data
file.
Meteorological Data File any file containing meteorological
data, whether it be upper air soundings, mixing heights,
surface observations or on-site data, or any combination of
these.
GLOSSARY-5 October 1994
-------
Missing Value alphanumeric character(s) that represent
breaks in the temporal or spatial record of an atmospheric
variable.
Mixing Height the depth through which atmospheric pollutants
are typically mixed by dispersive processes.
Monthly Mean Value a one-month arithmetic average of a
meteorological variable.
MPDA Meteorological Processor for Diffusion Analysis, the
predecessor to MPRM, available in UNAMAP version 6.
MPRM Meteorological Processor for Regulatory Models, the
software described in this document.
MPTER Multiple Point source dispersion model with TERrain
adjustment, from Appendix A of the modeling guideline (EPA,
1986).
MR MeRge, the 2-character code indicating that all fields on
the input image pertain to combining of the data from the three
pathways into a single unformatted file.
MR Data collective term for all input images that begin with
the 2-character code MR.
MR Pathway collective term for logic associated with
deciphering input images beginning with the MR character code.
NCDC National Climatic Data Center, the federal agency
responsible for distribution of the National Weather Service
upper air, mixing height and surface observation data.
GLOSSARY-6 October 1994
-------
NTIS National Technical Information Services, the agency
responsible for distribution of technical information,
including UNAMAP products.
NWS National Weather Service.
OAQPS Office of Air Quality Planning and Standards
OQA Output from Quality Assessment, an input image that
defines the output file to receive data that have gone through
quality assessment. This file is also used as the input file
for Stage 2 processing.
On-site Data data collected from a meteorological
measurement program operated in the vicinity of the site to be
modeled in the dispersion analysis.
Opaque Sky Cover the amount of sky cover, expressed in
tenths, that completely obscures all that might be above it.
OS On-Site, the 2-character code indicating that all fields
on the input image pertain to the processing of on-site data
OS Data collective term for all input images that begin with
the 2-character code OS, also used to collectively refer to
on-site data processed.
OS Pathway collective term for logic associated with
deciphering input images beginning with the OS character code.
Overlay one or more subprograms that reside on disk and are
loaded into memory only when needed.
Pasquill Stability Categories a classification of the
dispersive capacity of the atmosphere, originally defined using
GLOSSARY-7 October 1994
-------
surface wind speed, solar insolation (daytime) and cloudiness
(nighttime). They have since been reinterpreted using various
other meteorological variables.
Pathway one of the five major processing areas in MPRM.
These are JB, OS, SF, UA, and MR (see these entries in this
section for a description).
PC Personal Computer.
Processing Methodologies options controlling Stage 3
processing.
Quality Assessment judgment of the quality of the data.
Quality Assessment Check determining if the reported value
of a variable is reasonable (see also Range Check).
Quality Assessment Message message written to the
error/message file when a data value is determined to be
suspect.
Quality Assessment Violation occurrences when data values
are determined to be suspect (see also Range Check Violation).
RAM (1) Random Access Memory on a personal computer.
(2) a multiple source dispersion model from Appendix A
of the modeling guidance, (EPA, 1986).
RAMMET Meteorological processor program used for regulatory
applications capable of processing twice-daily mixing heights
(TD-9689 format) and hourly surface weather observations
(CD-144 format) for use in dispersion models such as CRSTER,
MPTER and RAM, (EPA, 1986).
GLOSSARY-8 October 1994
-------
Range Check determining if an observation of a variable
falls within predefined upper and lower bounds.
Range Check Switch parameter whose value indicates whether
to include or exclude the upper and lower bounds during range
checks.
Range Check Violation determination that the value of a
variable is outside range defined by upper and lower bound
values (see also Quality Assessment Violation).
Raw Data File any file which has not been processed by MPRM
Regulatory Applications dispersion modeling involving
regulatory decision-making as described in the Guideline on Air
Quality Models (Revised), (EPA, 1986).
Regulatory Model a dispersion model that has been approved
for use by the regulatory offices of the EPA (EPA, 1986).
Reporting Procedures options available in Stage 3
processing for reporting the availability of meteorological
data for the selected dispersion model.
Roughness Length see Surface Roughness Length
RTDM Rough Terrain Dispersion Model, complex terrain
dispersion model used as a third level screening model used in
regulatory dispersion modeling (EPA, 1986).
Run Stream collectively, all input images required to
process data in MPRM.
GLOSSARY-9 October 1994
-------
aa - - standard deviation of the horizontal wind direction
fluctuations.
-------
assessment of the data, and all reports and files generated
during this process.
Stage 2 Processing the process of combining or merging the
three types of meteorological data into an unformatted file,
and all reports and files generated during this process.
Stage 3 Processing the process of preparing meteorological
data, processed in Stage 2, for use by a dispersion model, and
all reports and files generated during this process.
Standard Reporting Levels mandatory pressure levels (and
corresponding measured atmospheric quantities) in a NWS upper
air sounding.
STAR (STability ARray) stability and wind rose summary, a
joint frequency distribution summary of stability category,
wind speed and wind direction. The STAR data are used as input
for several long-term dispersion models such as COM and ISCLT.
Station Identification an integer or character string used
to uniquely identify a station or site as provided in the upper
air (TD-5600 and TD-6201), mixing height (TD-9689), and surface
weather (CD-144 and TD-3280) data formats available from NCDC.
There are no standard station numbers for on-site data and the
user may include any integer or character string up to eight
digits or characters.
Storage Formats For magnetic tapes, the formats available
from NCDC for storing upper air and surface observations. See
TD-1440, TD-3280, TD-5600, TD-6201, TD-9689 and CD-144. For
on-site data 'storage formats' refers to the format of the data
on the input file.
GLOSSARY-11 October 1994
-------
Subdirectory a directory below the root, or highest level,
directory or another subdirectory, used for organization of
files on a storage medium such as a PC hard disk.
Surface Albedo fractional amount of radiation incident on a
surface that is reflected away from the surface.
Surface Weather Observations a collection of atmospheric
data on the state of the atmosphere as observed from the
earth's surface. In the U.S. the National Weather Service
collect these data on a regular basis at selected locations.
Surface Roughness Length height at which the wind speed
extrapolated from a near-surface wind speed profile becomes
zero.
TD-1440 Format a format available from NCDC for summarizing
NWS surface observations in an 80-column format; the CD-144
format is a subset of this format. This format has been
superseded by the TD-3280 format.
TD-3280 Format the current format available from NCDC for
summarizing NWS surface weather observations in an elemental
structure, i.e., observations of a single atmospheric variable
are grouped together for a designated period of time.
TD-5600 Format a format available from NCDC for reporting
NWS upper air sounding data. This format has been superseded
by the TD-6201 format.
TD-6201 Format the current format available from NCDC for
reporting NWS upper air data. The file structure is
essentially the same as the TD-5600 format except that there is
more quality assurance information.
GLOSSARY-12 October 1994
-------
TD-9689 Format the format available from NCDC for mixing
heights estimated from morning upper air temperature and
pressure data and hourly surface observations of temperature.
Temperature lapse rate the fall of temperature per unit
height, and is taken as positive when temperature decreases
with height.
TTN Technology Transfer Network
Total Sky Cover the amount of sky, expressed in tenths,
covered by a combination of transparent and opaque clouds or
obscuring phenomena.
Turbulence The irregular "eddy" motions in fluids, whether
liquid or gaseous, which cause an irreversible mixing of fluid
properties between neighboring parcels.
UA Upper-Air, the 2-character code indicating that all
fields on the input image pertain to the processing of the
twice-daily mixing height data and the upper air data.
UA Data collective term for all input images that begin with
the 2-character code UA, also used to collectively refer to
mixing height and upper air data processed.
UA Pathway collective term for logic associated with
deciphering input images beginning with the UA character code.
UNAMAP User's Network for Applied Modeling of Air Pollution,
a collection of dispersion models and closely related support
utilities available through NTIS.
Unformatted File a file written without the use of a FORTRAN
FORMAT statement.
GLOSSARY-13 October 1994
-------
Upper Air Data (or soundings) meteorological data obtained
from balloon-borne instrumentation that provides information on
pressure, temperature, humidity, and wind away from the surface
of the earth.
VALLEY a complex terrain dispersion model used as a first
level screening model in regulatory dispersion modeling, (EPA,
1986).
Warning Message a message written by the processor to the
error/message file whenever a problem arises that may inhibit
further data processing.
Wind Shear the change in wind velocity with height.
GLOSSARY-14 October 1994
-------
REFERENCES
Catalan©, J.A., D.B. Turner, and J.H. Novak. User's Guide for
RAM--Second Edition. EPA-600/8-87-046, U.S. Environmental
Protection Agency, Research Triangle Park, NC, 1987. 200 pp.
Heim, R., Jr., 1988. Personal Computers, Weather Observations,
and the National Climatic Data Center. Bull. Amer. Meteor.
Soc., 69:490-495.
Paumier, J., D. Stinson, T. Kelly, C. Ballinger, and J.S.
Irwin. MPDA-1: A Meteorological Processor for Diffusion
Analysis--User's Guide. EPA-600/8-86-011, U.S. Environmental
Protection Agency, Research Triangle Park, NC, 1986. 197 pp.
Pierce, T.E. and D.B. Turner. Developments in National Weather
Service Meteorological Data Collection Programs as Related to
EPA Air Pollution Models. EPA-600/3-87-048, U.S. Environmental
Protection Agency, Research Triangle Park, NC, 1987. 120 pp.
Turner, D.B. A Diffusion Model for an Urban Area. J. Appl.
Meteor., 3:83-91, 1964.
Turner, D.B. and L.W. Bender. Environmental Research
Brief--Description of UNAMAP (Version 6) (Revised). EPA-
600/M-86-027, U.S. Environmental Protection Agency, Research
Triangle Park, NC, 1988. 13 pp.
U.S. Environmental Protection Agency. Guideline on Air Quality
Models (Revised). EPA-450/2-78-027R, U.S. Environmental
Protection Agency, Research Triangle Park, NC, 1986. 283 pp.
REFERENCE-1 October 1994
-------
U.S. Environmental Protection Agency. On-Site Meteorological
Program Guidance for Regulatory Modeling Applications. EPA-
450/4-87-013, U.S. Environmental Protection Agency, Research
Triangle Park, NC, 1987. 187 pp.
REFERENCE-2 October 1994
-------
INDEX
A
abnormal job termination 7-5
additional quality assessment checks . 4-19
albedo 4-25, A-6, B-15, C-6
ANSI FORTRAN 6-1
ASCII data file 1-2, 1-4
audit summary (report) 4-10, 4-12, 4-16 to
4-18, 4-28, B-4
B
BLP 3-9
backspace 6-10
bounds, upper and lower for quality . . 1-8, 4-7, 4-19
assessment 5-2, 5-4, C-l
default 5-2, C-3 to C-5
Bowen ratio 4-25 to 4-27, A-6,
B-16, C-6
C
CALINE-3 3-21
calm wind conditions 5-7, 5-8
CD-144 1-2, 3-10, 6-14, 6-20,
F-4 to F-5
CDM16 3-22, F-ll
CDM36 3-22, F-12
ceiling height 5-11
character conversion 6-4
clipping height 4-3
combining data 1-3, 3-16, 4-33
COMMON blocks .' \ . . 6-2
INDEX-1 October 1994
-------
COMPLEXl 3-21
CRSTER 3-21, 4-38
D
data extraction 1-1, 3-1,4-1
data substitution 1-5 to 1-6
default audit variables 4-16 to 4-17, A-4,
A-5, A-6
default dispersion model 3-21, 4-38
default printer device 3-5
dew-point
estimates of 5-7
gradient . . . . , 5-4
DISK (keyword) 3-8 to 3-9. B-2
disk file name A-2 to A-9, B-2
on the IBM 6-13
dispersion modeling location 3-14
dispersion models supported 3-16 to 3-19
E
end of file conditions 6-10
error message 7-lto7-2,E-l
error/message file 3-13, 7-3, E-l
extracting data see data extraction
F
FB (fixed block) tape structure .... B-3
file header records 6-10
FORTRAN-77 4-3, 6-1
extensions 6-2
frequency of occurrence, wind speed . . 3-27
G
general report file 3-13
INDEX-2 October 1994
-------
Stage 1 processing 4-8 to 4-10, D-2 to
D-4
Stage 2 processing 4-35 to 4-36, D-5 to
D-7
Stage 3 processing 3-21 to 3-29
Greenwich Mean Time (GMT) 3-18
H
harmonic average value 3-27
height
above ground level 5-7
above mean sea level 5-7
height categories, upper air quality
assessment 5-2, 5-3
height intervals (UAINC) 5-3
hypsometric formula 5-2
I
informational message 7-4, E-l
input syntax 3-3 to 3-5, B-l
INCLUDE, FORTRAN extension 6-2 to 6-5
ISCLT 3-19
ISCST 3-17
J
JB (keywords) A-3
END 4-2, A-3, B-6
ERR 3-5, 3-8, A-3, B-2
FIN A-3, B-7
OUT 4-6, A-3, B-2
RUN 4-2, A-3, B-14
STA 3-8, A-3, B-17
JB-pathway 3-4, A-1
joint frequency function 5-15
* "*
INDEX-3 October 1994
-------
K
keyword 3-3, 3-8, A-1
L
lapse rate 5-3 to 5-4, 5-6
superadiabatic 5-6
Local Standard Time (LST) 3-11
M
MP (keywords) A-8
EXT 3-11, 3-20, A-8, B-6
FIN A-8, B-6
LST - 3-17, A-8, B-9
MET 3-17, A-8, B-13
MMP 3-17, A-8, B-13
STA A-8, B-17
TRA 3-17, 4-40, 5-13, A-8,
B-17
VBL 4-39, 5-9, 5-11, 5-12,
A-8, B-17
MP-pathway 3-4, A-1, A-2
MR (keywords) A-8
EXT 3-14, 3-15, 4-34, A-8,
B-7
FIN 4-34, A-8, B-6
OUT 3-14, 3-15, 4-34, A-8,
B-2
STA 4-34, A-8, B-17
MR-pathway 3-4, A-l, A-2
magnetic tapes
mounting on a VAX 6-11
on the IBM 6-13
mandatory keyword A-2
mandatory level, upper air sounding . . 4-5, 5-5 to 5-6
math co-processor 6-5
INDEX-4 October 1994
-------
merging data 1-5, 3-13,4-33 to
4-36
message code 7-3, E-2
message structure 7-3,E-2
message types 7-4,E-l
missing values 1-6 to 1-7, 3-1 to
3-2
mixing heights 1-2, 3-6, 4-36
output from Stage 3 3-27 to 3-28
modeling guidance 1-5
MPTER 3-18
N
NCDC 1-1, 1-2, 1-4
0
OS (keywords) A-6 to A-7
ADD A-6, B-4
AVG 4-21, 5-1, A-6, B-4,
C-7
CHK A-6, B-5
CLM A-6, B-5, C-7
DTI, DT2, DT3 4-21, A-6, B-6
ERR B-2
EXT A-6, B-6
FIN 4-34, A-6, B-6
FMT 4-20, 4-21, 4-24, A-6,
B-7
HGT 4-27, A-6, B-7
IQA A-6, B-2
LOG 3-14 to 3-15, 4-34,
A-6, B-8
MAP 4-20 to 4-23, A-6,
B-7, B-9 to B-12
OQA ....'/ \ . . 4-34, A-6, B-2
INDEX-5 October 1994
-------
OUT B-2
SFC 4-21, 4-25 to 4-26,
A-6, B-15 to B-16
SECTOR 4-21, B-16
SETUP 4-21, 4-25, B-15
VALUES 4-21, B-15
STA 4-34, A-6, B-18
TRA A-6
OS-pathway 3-4, A-1
on-site data 1-1, 4-20
on-site data extraction and quality . . 4-19 to 4-33
assessment
on-site data format 1-4, 4-19,
see also OS MAP
on-site data hourly averages 4-20, 5-1, B-4, C-6
on-site data map 4-20 to 4-25, B-9 to
B-12
on-site guidance document 5-1
on-site measurement heights 4-27
anemometer height (ANEHGT) 4-39, 5-9
stack height (STKHGT) 4-39, 5-12
temperature height (TMPHGT) 4-39, 5-10
on-site surface characteristics
(albedo, surface roughness length,
Bowen ratio) 4-26, A-6, C-6
on-site temperature difference . . . . 4-22, A-6, B-6
output files 3-11 to 3-12, 5-15
opaque sky cover 5-11
OPEN, FORTRAN 6-4
operating system date and time . . . . 6-4 to 6-5
optional keyword A-2
overlay 6-1,6-6 to 6-7
P
partitioned data set name 6-3, 6-8
INDEX-6 October 1994
-------
pathway 3-3, A-1
personal computer
installation on 2-1 to 2-3
processing data on 1-4, 3-3 to 3-4, 6-5
plume rise estimates 5-8
Q
quality assessment 1-3, 1-4, 3-9, 3-11,
5-2 to 5-4
on-site data 4-27 to 4-33
surface data 4-15 to 4-19
upper air data 4-5 to 4-13
quality assessment message 4-8 to 4-9, E-l
R
RAM (dispersion model) 3-17
RAMMET meteorological processor . . . . 3-17, F-7
range checks 4-7,5-2
range check messages 4-8
range check switch 4-7, B-5
redefining quality assessment
parameters 4-5, 5-2
redirecting I/O on a personal
computer 3-5
roughness length see surface
roughness length
RTDM 3-18
run stream 3-3 to 3-4
run-time statistics 6-17 to 6-21
S
cra 5-12, 5-13
ae 5-12, 5-13
ffw 5-12, 5-13
»
INDEX-7 October 1994
-------
SF (keywords) A-5
AUD 4-16, A-5, B-4
CHK A-5, B-5
ERR B-2
EXT A-5, 3-6
FIN 4-34, A-5, B-6
IN2 3-8, A-5, B-2, B-4
IQA A-5, B-2
LOG A-5, B-8
OQA 4-34, A-5, B-2
OUT B-2
STA 4-34, A-5, B-18
SF-pathway 3-4, 4-13 to 4-19, A-1
scan report 6-17, D-12
significant level, upper air sounding . 5-5, 5-6
STability ARray (STAR) 5-15
stability category 4-39 to 4-41
stack top 5-9
Stage . . l-l
Stage 1 processing 11, 1-2 to 1-3, 3-7,
4-1
Stage 2 processing 11, 13, 3-13, 4-33
Stage 3 processing 1-2, 1-4 to 1-5, 3-16,
4-36
cycling through stability methods . . 4-41, 5-11 to 5-13
excluding the default method .... 5-13
default processing 3-16
processing options 4-39
standard input and output devices . . . 3-5, 6-10
station identification number 3-10
station location 3-10 to 3-11, B-8
status report 4-8
superadiabatic lapse rate . 5-6
surface data formats 11, 3-10, F-3 to F-7
surface data extraction 4-13 to 4-15
INDEX-8
October 1994
-------
surface roughness length 4-26, 5-14 to 5-15
surface weather observations l-l/ 3-4, 3-10
system dependent statements 6-2 to 6-5
T
TD-3280 F-4, F-5
TD-5600 1-2, 4-3, 5-6, 6-9,
6-11, 6-20, F-4, F-5
TD-9689 1-2, 3-9, 4-13, B-2 to
B-3
tape file name 6-12, B-3
on the IBM 6-15
TAPE keyword B-3
temperature
difference measurement, on-site . . . 4-22, B-6
estimates of, in an upper air
sounding 5-6
on-site measurement level 5-10 to 5-11
sign of, in upper air data 5-5 to 5-6
terminating an input run stream .... 4-3
trace message A-8, E-l
U
UA (keywords) A-4
AUD 4-12, 4-13, 4-16 to
4-17, A-4, B-4
CHK 4-6 to 4-8, 5-2 to
5-3, A-4, B-5
ERR B-2
EXT A-4, B-6
FIN 4-34, A-4, B-6
INI A-4, B-3
IN2 . . . 3-8 to 3-10, A-4, B-2,
B-3
INDEX-9 October 1994
-------
IQA 3-8, 3-11, 3-13, A-4,
B-2
LOG A-4, B-8
OFF 4-5, A-4, B-13
OQA 3-8, 3-11, 3-13, 4-6,
4-34, A-4, B-2
OUT B-2
STA 4-34, A-4, B-17
TOP A-4, B-17
UA-pathway 3-4, A-1
unformatted data 1-4
unit 6 3-5, 4-3 to 4-4, 6-10
upper air automatic data modification . 4-4 to 4-6, 5-5 to 5-7
upper air data extraction 4-2 to 4-6, 6-11
upper air data format 1-4, F-ltoF-3
upper air quality assessment 4-6 to 4-13
upper air sounding 11, 1-3, 3-11, 4-1,
4-2, 4-36
USER (keyword) 3-10, B-2
V
VB (variable blocked) tape structure . 4-3, B-3
VALLEY 3-19
variable names used in MPRM C-3toC-5
vertical gradients, upper air quality . 5-3 to 5-4
assessment
W
warning message 7-4 to 7-5, A-3, E-l
wind, on-site measurement level .... 5-8
wind direction sectors 4-27 to 4-28, 5-13
wind direction shear 5-3 to 5-4
wind speed, minimum 5-1, 5-8
wind speed shear 5-3 to 5-4
INDEX-10
October 19.94
-------
TECHNICAL REPORT DATA
(Please read Instructions on reverse before completing)
1 REPORT NO.
EPA-454/B-94-020
J RECIPIENT'S ACCESSION NO
4 TITLE AND SUBTITLE
Meteorological Processor for Regulatory Models (MPRM)
User's Guide (Revised)
S REPORT DATE
October 1994
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Desmond T. Bailey
8. PERFORMING ORGANIZATION REPORT NO
9. PERFORMING ORGANIZATION NAME AND ADDRESS
10. PROGRAM ELEMENT NO
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Research Triangle Park, NC 27711
II. CONTRACT'ORANT NO
12. SPONSORING AGENCY NAME AND ADDRESS
13. TYPE.OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
13. SUPPLEMENTARY NOTES
16. ABSTRACT
The Meteorological Processor for Regulatory Models (MPRM) is a general purpose program used
to process meteorological data for use in EPA recommended air quality dispersion models. MPRM
provides capabilities not available in the RAMMET meteorological processor. These capabilities include
quality assurance procedures, detailed report generation, and the ability to process both on-site (user
collected) and off-site [National Weather Service (NWS)] meteorological data.
The MPRM processor consists of three stages. Stage 1 (extraction and quality assurance)
retrieves meteorological data provided by the user on magnetic tape or disk and conducts the quality
assurance of these data. The stage 1 report files provide listings of missing, suspect, and invalid data.
These reports provide necessary information allowing users to correct problem data prior to its use in
modeling. Stage 2 merges the quality assured and corrected meteorological data obtained from the
various MPRM pathways, upper air, on-site, and surface (NWS). The third and final stage performs the
necessary processing to create a meteorological data file for use in a dispersion model selected by the
user.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Fidl/Ontip
Air Pollution
Atmospheric Dispersion Modeling
Meteorological Processors
Meteorological Monitoring
Meteorology
18. DISTRIBUTION STATEMENT
Release Unlimited
19. SECURITY CLASS (Report)
Unclassified
21 NO OF PAGES
236
20. SECURITY CLASS tPaff)
Unclassified
------- |