Final Report
REGIONAL AIR POLLUTION STUDY:
A PROSPECTUS
Part III — Research Facility
Prepared for:
THE ENVIRONMENTAL PROTECTION AGENCY
NATIONAL ENVIRONMENTAL RESEARCH CENTER
RESEARCH TRIANGLE PARK, NORTH CAROLINA
CONTRACT 68-02-0207
SRI Project 1365
STANFORD RESEARCH INSTITUTE
Menlo Park, California 94025 • U.S.A.
-------
Final Report
January 1972
REGIONAL AIR POLLUTION STUDY:
A PROSPECTUS
Part III - Research Facility
Prepared for:
THE ENVI RONMENTAL PROTECTION AGENCY
NATIONAL ENVIRONMENTAL RESEARCH CENTER
RESEARCH TRIANGLE PARK. NORTH CAROLINA
CONTRACT 68-02-0207
SR I Project 1365
R. T. H. COLLIS. Director
Atmospheric Sciences Laboratory (Project Director)
DON R. SCHEUCH. Vice President,
Office of Research Operations
-------
FOREWORD
This Prospectus was prepared by Stanford Research Institute for the
Environmental Protection Agency under Contract No. 68-02-0207. While
this Prospectus has been reviewed by the Environmental Protection Agency
and approved for publication, approval does not signify that the contents
necessarily reflect the views and policies of the Environmental Protec-
tion Agency, nor is it intended to describe the Agency's program.
The complete Prospectus for the Regional Air Pollution Study is pre-
sented in four parts.
Part I
Part II
Part III
Part IV
Summary
Research Plan
Research Facility
Management Plan
A table of contents for all parts is provided in each of the four
parts to facilitate the use of the Prospectus.
iii
-------
ACKNOWLEDGMENT
This Prospectus was prepared at the Institute by a project team
representing the full range of disciplines necessary for the comprehensive
analysis of problems of air pollution. Research team members were drawn
from four of the eight Institute Research Divisions, including the fol-
lowing:
Electronic and Radio Sciences
Physical Sciences
Information Science and Engineering
Engineering Systems
Because of the interdisciplinary nature of the effort, the contributions
and research findings of many team members are distributed throughout this
Prospectus rather than concentrated in one or more specific chapters. Ac-
cordingly, contributions are acknowledged below by general areas associ-
ated with the study of air pollution problems.
This Prospectus was prepared under the supervision of R.T.H. Collis,
Project Director. The Project Leader was Elmer Robinson (now of Washing-
ton State University) until 15 January, when Richard B. Bothun, who had
been Deputy Project Leader, succeeded him.
The main contributions were as follows:
.
Elmer Robinson--Project leadership and the formulation of the
Research Plan
.
Richard B. Bothun--Project leadership and administrative man-
agement and the formulation of the Management Plan.
Technically, the principal contributions were:
.
Richard B. Bothun--Management, scheduling, costing, planning
.
Leonard A. Cavanagh--Air quality instruments, atmospheric
chemistry
v
-------
.
Ronald T. H. Collis--Meteorology, remote sensing, research
planning
.
Walter F. Dabberdt--Transport and diffusion modeling, mete-
orology, instrumentation
.
Paul A. Davis--Solar radiation, tracer studies
.
Roy M. Endlich--Meteorological models, satellite systems
.
James L. Mackin--Helicopter and aircraft systems
.
Elmer Robinson--Meteorology, instrumentation, atmospheric
chemistry, research planning
.
Sylvin Rubin--Data processing systems
.
Konrad T. Semrau--Source inventory and emissions
.
Elmer B. Shapiro--Communication systems
.
James H. Smith--Atmospheric chemical transformation processes
.
Eldon J. Wiegman--Synoptic climatology
Valuable contributions were made in the latter stages of the project
by Dr. W. A. Perkins and Mr. J. S. Sandberg, consultants.
The Institute wishes to express its appreciation for the assistance
and provision of information by many staff members of the Environmental
Protection Agency, especially Charles R. Hosler, Contracting Officer's
Technical Representative; Dr. Warren B. Johnson, Jr., Chief, Model Devel-
opment Branch; Robert A. McCormick, Director, Division of Meteorology;
and Dr. A. P. Altshuller, Director, Division of Chemistry and Physics.
Additionally. the constructive criticism and comment provided by
members of the Meteorology Advisory Committee of the Environmental Pro-
tection Agency during the preparation of the Prospectus were of signifi-
cant value, and our indebtedness is hereby acknowledged.
vi
-------
CONTENTS
PART I - SUMMARY
FOREWO RD . . .
. . . . . . . . . . . . . . . . . . . . . . . . . .
ACKNOWLEDGMENT.
III
. . . .
. . . . . . . . . . . . . . . . . . . . .
I
THE BASIC PREMISE
. . . . . . . .
. . . . . . . . . . . . .
II
SCIENTIFIC AIR QUALITY MANAGEMENT
. . . . .
. . . . .
The Basic Tool--The Mathematical Model. . . . . . . . . .
The Processes To Be Modeled. . . . . . . . . .
Accuracy. . . . . . . . . . . . . . . . . . . . . . . . .
Current Limitations of Modeling. . . . . . . . . . .
Steps To Improve Models . . . . . . . . . . . . . . .
The Regional Scale. . . . . . . . . . . . . . . . .
THE REGIONAL AIR POLLUTION STUDY (RAPS)
. . . . . . . . . .
Concept. . . . . . . . . . . . . . . . . . . . . . .
Pur po se . . . . . . . . . . . .
Organization. . . . . . . . . . . . . . . . . . . . . . .
Objectives. . . . . . . . . . . . . . .
. . . . .
IV
SITE SELECTION
. . . . . . . . . . . . . . . . . . . . . .
V
THE RESEARCH PLAN.
. . . . . . . . .
. . . . .
Introduction . . . . . . . . . . .
Model Evaluation and Verification Program.
. . . . . .
. . . . .
Meteorological Factors. . . . . . . . . . . . . . . . .
Pollutant Source Estimates. . . . . . . . . . . . . . .
Air Quality Measurements . . . . . . . . . . . . .
Atmospheric, Chemical, and Biological Processes. . . . . .
Human Social and Economic Factors . . . . . . . . . .
RAPS Technology Transfer
. . . .
. . . .
. . . .
. . . . .
vii
iii
v
1
3
3
3
4
4
5
7
9
9
10
10
14
17
19
19
21
24
25
25
26
28
28
-------
CONTENTS
v
Continued
Schedules and Task Specifications for the Research Plan. .
Introduction. . . . . . . . . . . . . . . . . . . . . .
100 Model Verification . . . . . . . . . . . . . .
101 Boundary Layer Meteorology Program. . . . . . . . .
102 Emission Inventory. . . . . . . . . . . . . . . . .
103 Air Quality Measurements. . . . . . . . . . . . . .
104 Model Calculation and Verification . . . . . .
200 Atmospheric, Chemical, and Biological Processes
201 Gaseous Chemical Processes. . . . . . . . . . . . .
202 Atmospheric Aerosol Processes. . . . . . . . . . . .
203 Other Pollutant Related Atmospheric Processes.
204 Atmospheric Scavenging by Precipitation. . . .
205 Air Pollutant Scavenging by the Biosphere. . .
206 Atmospheric Processes. . . . . . . . . . . . .
300 Human, Social, and Economic Factors. . . . . . . . .
301 Human and Social Factors. . . . . . . . . . . . . . .
302 Economic Factors. . . . . . . . . . . . . . . . . .
400 Transfer of RAPS Technology for Control Agency
Applications and the Formulation of Control Strategies
401 Source Inventory Procedures. . . . . . . . . .
402 Atmospheric Monitoring . . . . . . . . .
403 Data Handling. . . . . . . . . . . . . .
404 Modeling Technology. . . . . . . . . . . . . . . . .
405 Other Significant Factors in Control Strategy
Formulation. . . . . . . . . . . . . . . . . . . . . . .
VI
THE FACILITY
. . . .
. . . " . .
. . . . . .
. . . .
Rationale. . . . . . . . . . . . . . . . . . . . . . . . .
Basic Operations. . . . . . . . . . . . . . . . .
Basis for Monitoring Network. . . . . . . . . . . . . .
The St. Louis Regional Monitoring Network. . . . .
viii
30
30
45
46
46
47
47
48
49
49
51
51
52
53
54
54
55
55
57
57
58
59
60
63
63
63
63
66
-------
VII
VIII
CONTENTS
MANAGEMENT AND SCHEDULING.
. . . .
. . . .
. . . . .
I nt roduct ion. . . . . . . . . . . . . . . . . . . .
Facility Activation Schedule. . . . . . . . . . . .
Permanent Management and Staffing. . . . . . . . . .
COST SUMMARY
. . . . .
. . . .
. . . . . . . . .
Permanent Facilities and Staff. . . . . . . . . . . . . .
Helicopter and Mixing Layer Observational Program. .
Research Plan. . . . . . . . . . . . . . . . . . . .
Personnel. . . . . . . . . . . . . . . . . .
Instrumentation and Equipment. . . . . . . . . . .
Operations. . . . . . . . . . . . . . . . .
Total Cost of Research Plan. . . . . . . . .
Total Costs of RAPS. . . . . . . . . . . . . .
ix
71
71
71
73
77
77
80
80
81
83
85
86
86
-------
CONTENTS
PART II - RESEARCH PLAN
FOREWORD. . .
. . . . . . . . . . . . . . . 8 . . . . . . . . .
ACKNOWLEDGMENT. . . . . .
III
. . . . . . .
. . . . . . . . . . . .
I
INTRODUCTION TO THE RESEARCH PLAN
. . . . . . . .
. . . .
II
RESEARCH PLAN--OVERVIEW OF AIR POLLUTION MODELING
. . . .
Introduction. . . . . . . . . . . . . . . . .
Model Evaluation and Verification Program
. . . . .
. . . . .
RESEARCH PLAN--METEOROLOGICAL PROCESSES
Introduction. . . . . . . . . . . . . . . . . . .
Atmospheric Dispersion Models. . . . . . . . . . .
Gaussian Formulae. . . . . . . . . . . . . . . . . . .
Gradient Transfer Theory. . . . . . . . . . . . . . .
Other Model s . . . . . . . . . . . . . . . . . . . . .
Numerical Weather Prediction Models. . . . . . . . . . .
Model Sensitivity to Meteorological Variables. . . . . .
Experimental Meteorology Program. . . . . . . . . . . .
General. . . . . . . . . . . . . . . . . . . . . . . .
Tracer Studies of Transport and Diffusion . . . .
Urban and Rural Radiation Budget Studies. . . . . . .
The Role of Remote Probing T~chniques . . . . . .
Upper Air Sampling Program. . . . . . . . . . . . . .
Plume Dispersion Studies. . . . . . . . . . . . . . .
Studies of Spatial Variabilities. . . . . . . . . . .
References. . . . . . . . . . . . . . . . . . . . . . .
IV
RESEARCH PLAN--ATMOSPHERIC CHEMISTRY AND TRANSPORTATION
PROCESSES. . . . . . . . . . . . . . . . . . . . .
Introduction
. . . . . . . . .
. . . . .
. . . . . . . .
x
iii
v
1-1
II-l
-1
-3
I II-l
-1
-4
-4
-11
-24
-27
-30
-46
-46
-48
-62
-67
-74
-79
-81
-84
IV-l
-1
-------
CONTENTS
IV
Continued
The S02 Cycle. . . . . . . . . . . . . . . . . .
Emission Sources of Sulfur Compounds. . . . . . . . .
Chemical Reactions of Importance to the SO -Cycle Model
2
Removal Mechanisms for Sulfur Compounds. . . . .
The Research Program. . . . . . . . . . . . . . . . .
The Photochemical Cycle--Hydrocarbons:Nitrogen Oxides:
Oxidant. . . . . . . . . . . . . . . . . . . . . . . . .
Sources of Nitrogen Oxides. . . . . . . .
Reactions of Nitrogen Oxides in the Absence of
Hydrocarbons. . . . . . . . . . . . . . .
Hydrocarbon Sources and Removal Processes. . . . .
The Hydrocarbon-Nitrogen Oxides Reactions.
The Research Program. . . . . . . . . . . . . .
The Particulate Cycle. . . . . . . . . . . . . . . . .
Background Haze. . . . . . . . . . . . . . . . . . . .
Natural Background Aerosols . . . . . . . . . . .
Particulate Emissions Inventory. . . . . . . . . . . .
The Chemistry of Particulate Formation in the
Atmo sphere. . . . . . . . . . . . . . . . . . . . . .
The Research Program. . . . . . . . . . . . . . . . .
Development of Continuous Rainfall pH Measurement and
Sequential Precipitation Collection. . . . .
Carbon Monoxide Cycle. . . . . . . . . . . . . . . . . .
Source of CO . . . . . . . . . . . . . . .
Important Chemical Reactions of CO . . . . . . . . . .
CO Sinks. . . . . . . . . .
The CO Research Program. . . . . . . . . . .
The Chemical Research Program Schedule. . . . . . . . .
Aerosol Research Program. . . . . . . . . . . . . .
The S02 Flux Measurement. . . . . . . .
The HC:NO Research Program. . . . . . . . . . . . . .
x
The CO Research Program. . . . . . . . . . . . . . . .
References. . . . . . . . . . . . . . . . . . . . . .
xi
IV-2
-3
-5
-8
-9
-12
-14
-15
-17
-18
-20
-24
-25
-27
-28
-30
-32
-34
-35
-37
-39
-40
-41
-42
-44
-46
-47
-51
-53
-------
CONTENTS
v
RESEARCH PLAN--EMISSION ESTIMATES. . . . . . .
. . . . .
Requirements of the Emission Inventory System. . .
Classification of Emission Sources. . . . . . . . . . .
Source Processes . . . . . . . . . . . . . . . . .
Source Uni t s . . . . . . . . . . . . . . . .
Stationary Sources. . . . . . . . . . . . . . . . . .
Mobile Sources. . . . . . . . . . . . . . . . . . . .
Po 11 u t ant s . . . . . . . . . . . . . . . . . . . .
Factors Affecting Emission Levels. . . . . . . . . . . .
Stationary Sources. . . . . . . . . . . . . . .
Mobile Sources. . . . . . . . . . . . . . . . .
Inventory Procedures and Accuracy. . . . . . . . .
Accuracy of Estimates . . . . . . . . . . . . . .
Procedures. . . . . . . . . . . . . . . . . . .
Emission Model. . . . . . . . . . . . . . . . . . . . .
Control Strategy Studies . . . .
Inventory Schedule. . . . . . . . . .
VI
RESEARCH PLAN--ECONOMIC AND SOCIAL IMPACT STUDIES.
Introduction
. . . D .
. . . . .
. . . .
. . . . .
Human and Social Factors. . . . . . . .
Economic Fac tors. . . . . . . . . . . .
. . . . .
. . . .
. . . .
VII
RESEARCH PLAN--TECHNOLOGY TRANSFER
. . . . . .
. . . . .
Introduction . . . . .
Technology Transfer Program. . .
. . .. .
. . . .
. . . " . .
VIII
OTHER AGENCY RESEARCH PROGRAMS
. . . . .
. . . . .
METROMEX . . . . . . .
NCAR Fate of Pollutants
NOAA's MESOMEX . . . .
. . . . 0 .
. . . .
. . . .
Study (FAPS)
. . . . . . .
. . . . . .
. . . . . . . .
APPENDIX
SCHEDULES AND TASK SPECIFICATIONS FOR THE RESEARCH
PLAN . . . . . . . . . . . . . . . . . . . . . . . .
xii
V-l
-1
-2
-5
-5
-5
-7
-7
-8
-8
-10
-11
-11
-13
-17
-18
-18
VI-l
-1
-2
-5
VII-l
-1
-1
VIII-l
-2
-3
-4
A-l
-------
CONTENTS
PART III - RESEARCH FACILITY
FOREWORD. . .
. . . . . . . . . . . . . . . . . . . . . . . . .
ACKNOWLEDGMENT. . . .
. . . . . . . . . .
. . . . .
. . . . . .
IX
INTRODUCTION TO FACILITY DESIGN AND OPERATIONS
. . . . .
X
ST. LOUIS SITE SELECTION
. . . . .
. . . . .
. . . . . .
Summary. . . . . . . .
Site Selection Criteria.
. . . . . .
. . . . .
. . . .
. . . .
. . . . .
Method of Analysis . . . . . . . . . .
National Summary Analysis. . . . . . .
Meteorology. . . . . . . . . . . . . . . . . . .
Fossil Fuels. . . . . . . . . . . . . . . . . . . . .
Regional Isolation. . . . . . . . . . . . . . . . . .
Res ul t s . . . . . . . . . . . . . . . . . . . . . . . .
General Analysis of Standard Metropolitan Statistical
Areas. . . . . . . . . . . . . . . . . . . . . . . . . .
Pollutants
. . . . .
. . . . .
. . . . . . . . .
Manufacturing . . . . . . . . . . . . . . . . . .
Geographical Separation. . . . . . . . . . . . . . . .
Sunshine . . . . . . . . . . . . . . . . .
Space Heating . . . . . . . . . . . .
Resul ts . . . . . . . . . . . . . . . . . . . . .
Agricul ture . . . . . . . . . . . . . . . . . . .
XI
NEAR-SURFACE ATMOSPHERIC RESEARCH FACILITY
. . . . . . .
General Considerations. . . . . . . . . . . . . . . . .
Horizontal Extent of the Network. . . .
Network Orientation. . . . . .
Characteristics of Stations. . . . . . . . . . . . . . .
Meteorological Instrumentation. . . . . . . . . .
. . . . .
. 0 . . . . .
xiii
iii
v
IX-l
X-l
-1
-4
-7
-8
-8
-9
-11
-12
-12
-15
-17
-21
-22
-22
-23
-24
XI-l
-1
-1
-14
-23
-27
-------
CONTENTS
XII
AIR QUALITY S~iPLING
. . . . .
. . . .
. . . . . .
Introduct ion . . . . . . . . . . . . . . . . . . . .
Permanent Network Monitoring Stations for Pollutants
Semipermanent Monitoring Station. . . . . . . . .
Pollutants To Be Monitored for Establishment of Air
Qual it Y . . . . . . . . . . . . . . . .
Carbon Monoxide. . . . . . . . . . . . . . . . .
Methane, Nonmethane Hydrocarbons, and Total Hydro-
carbons. . . . . . . . . . . . . . . . . . . . . . . .
Nitrogen Oxides. . . . . . . . . . . . . . . . . . .
Sulfur Oxides and Hydrogen Sulfide and Total Sulfur. .
Ozone. . . . . . . . . . . . . . . . . . . . . . . . .
Suspended Particulate Material. . . . . . . . . . . .
Other Pollutants of Interest Not Measured by Network
Pollutant Monitoring Techniques. . . . . . .
Carbon Monoxide, Methane, and Total Hydrocarbon.
Nitric Oxide, Total Nitrogen Oxides. . . .
Sulfur Dioxide, Hydrogen Sulfide, Total Sulfur. . . .
Ozone. . . . . . . . . . . . . . . . . . . . . .
Suspended Particulate Material. . . . . . . . . . . .
Summary of Pollutant Instruments
Instrument Calibration. . . . . . . . . .
Local Calibration. . . . . . . . . . . . . . . . . . .
Calibration Vans for Primary Calibration . . . .
Role of Research Programs. . . . . . . . . . . . . . . .
XIII
DATA ACQUISITION AND HANDLING.
. . D .
. . . . .
. . . .
Introduction. . . . . . . . . . . . . . . . . . .
Policies and Principles. . . . . . . . . . . . . . .
System Overview. . . . . . . . . . . . . . .
Instruments. . . . . . . . . . . . . . . . . . . . . . .
Data Acquisition Equipment at Stations . . . .
Data Formats. . . . . . . . . . . . . . . . . . .
The Central Data Collection Facility. . .. ....
Overview. . . . . . . . . . . .
Functional Tasks
Equipment Complement. . . . . . . . . . . . . . . . .
. . . . . .
. . . . . . . .
. . . .
xiv
XII-1
-1
-1
-4
-7
-8
-8
-8
-8
-9
-9
-9
-10
-12
-13
-15
-17
-18
-20
-21
-21
-23
-23
XIII-l
-1
-1
-2
-6
-8
-13
-15
-15
-16
-18
-------
XIII
XIV
CONTENTS
Continued
..
The Central Data Collection Facility (cont.)
Disk File Organization. . . . . . . . . . . . . . .
Manual Data Entry. . . . . . . . . . . . . . . . . .
Reliability Considerations and Maintenance. . . . .
Communications for the Data Collection Network. . . .
The Telephone Companies. . . . . . .
Facil it ies
Cost s . . .
Alternative Communication Approaches
. . . . . . . .
. . . .
. . . . . .
. . . .
. . . .
. . . . . . . .
. . . .
. . . . .
MIXING LAYER OBSERVATION PROGRAM
. . . . .
. . . . .
Introduct ion. . . . . . . . . . . . . . . . . . . . .
General Mixing Layer Observation Methods. . . . . . .
Primary Aircraft Support for the Regional Study. . . .
Helicopter Support Function. . . . . . . . . . .
Schedule of Operations. . . . . . . . . . . . . . .
Cost s . . . . . . . . . . . . . . . . . .
Helicopter Instrumentation Package . . . . . .
Balloon Tracking System. . . . . . . . . . . . .
System Design Concept. . . . . . . . . . . . .
Activation Schedule. . . . . . . . . . . .
Cost s . . . . . . . . . . . . . . . . . . . . .
Special Aircraft-Based Meteorological Observations
General Concepts. . . . . . . . . . . . . . . . . .
Instrumentation Considerations
Special Aircraft-Based Air Quality Observations.
Western Environmental Research Laboratory (WERL)
Selection of Parameters. . . . . . . .
Selection of Surveillance Techniques
. . . .
. . . . . . .
XV
GENERAL CLIMATOLOGY OF ST. LOUIS
. . . .
. . . .
Introduction. . . . . . . . . . . .
Principal Seasonal Meteorological Characteristics
Cold Season (Mid-October to Mid-April) . . . .
Warm Season (Mid-April to Mid-November) . . . . . .
xv
XIII-19
-20
-20
-21
-21
-22
-23
-24
XIV-l
-1
-1
-3
-4
-4
-9
-9
-14
-14
-17
-17
-18
-18
-19
-24
-24
-26
-26
XV-l
-1
-3
-3
-8
-------
XVI
CONTENTS
XV
Continued
Meteorological Parameters Relating to Pollution. . . . .
Stagnating Anticyclones. . . . . . . . . . . . .
Wind. . . . . . . . . . . . . . . . . . .
LAND AND BUILDING REQUIREMENTS
. . . . . .
. . . .
Central Facility. . . . . . . . . . . . . . . . .
Selection Criteria. . . . . . . . . . . . . . .
Interior Space Requirements. . . . . . . . . . .
Outdoor Facilities . . . . . . . . .
Instrument Station Sites . . . . . . . . . .
Selection Criteria. . . . . . . . . . . . . . . . . .
Implementation. . . . . . . . . . .
xvi
XV-9
-9
-13
XVI-l
-1
-1
-4
-6
-7
-7
-9
-------
CONTENTS
PART IV - MANAGEMENT PLAN
FOREWORD. . .
. . . . . . . . . . . . . . . . . . . . . . . . .
ACKNOWLEDGMENT. . . . .
XVII
XVIII
. . . . . . . " . . . . . . .
. . . . .
INTRODUCTION TO THE REGIONAL STUDY SCHEDULING,
STAFFING, AND COST . . . . . . . . . . . . . . .
Introduction. . . . . . . . . . . . . . . . . . . . .
The St. Louis Facility . . . . . . . . . .
Facility Activation Schedule. . . . . . . . . .
Permanent Management and Staffing. . . . . . . .
Cost s . . . . . . . . . . . . . . . . . . . . . . . . .
Permanent Facility and Staff. . . . . . . . .
Helicopter and Mixing Layer Observational Program. . .
Research Plan. . . . . . . . . . . . . . . . . . . .
Total Costs. . . . . . . . . . . . . . . . . . . . .
IMPLEMENTATION SCHEDULE OF THE ST. LOUIS FACILITY
Introduction. . . . . . . . . . . . . . . . . . . . .
Prototype Instrument Station and Central Facility. . .
Schedule Network and Estimated Activity Durations. .
Activity Descriptions. . . . . . . . . . . . . . . .
Network Critical Path. . . . . . . . . . . . . . . .
Activation Schedule of Class A and Class B Stations. .
Uni t Schedules. . . . . . . . . . . . . . . . . . .
Sequential Station Activation Schedule and Mainte-
nance Requirements. . . . . . . . . . . . . .
Full Facility Implementation. . . . . . . . . . . . .
General Scheduling Conditions. . . . . .
Class A and B Station Activation with Prior Proto-
type Station Acceptance. . . . . . . . . . . .
Class A and B Station Activation Without Prior
Prototype Station Acceptance. . . . . . . . . . . .
Class C Station Activation Schedule. . . . . . . . .
Aggregate Facility Activation Schedule. . . . . . .
xvii
iii
v
XVII-l
-1
-2
-5
-7
-9
-9
-12
-12
-14
XVIII-l
-1
-3
-3
-17
-27
-33
-33
-36
-44
-44
-53
-55
-59
-63
-------
XIX
xx
XXI
CONTENTS
PERMANENT MANAGEMENT AND STAFFING
. . . . .
. . . . . .
Introduction. . . . . . . . . . . . . . .
Regional Study Management. . . . . . . . .
Research Triangle Park Staff. . . . . . . . . .
St. Louis EPA Operating Staff with EPA Operation. . .
St. Louis EPA Staff with Prime Contractor Operation. .
St. Louis Facility Implementation . . . .
Staff Scheduling . . . .
ST. LOUIS FACILITY INITIAL COSTS AND ANNUAL OPERATING
COSTS. . . . . . . . . . . . . . . .
Introduct ion . . . . . . . . . . . . . . . . . .
Ini t ial Costs of the St. Louis Facili ty . . . . . . . .
Air Quality and Meteorological Instruments. . . . .
Instrument Station Preparation, Facilities, and
Appurtenances. . . . . . . . . . . . . . . . .
Ditigal Data Terminal and Communication Equipment. .
Central Facility and Equipment. . . . . . . . . . .
Vehicular Support Facilities . . . . . . . . .
Total Initial Costs. . . . . . . . . . . . . .
Annual Operating Costs. . . . . . . . . . . . . . . .
Personnel. . . . . . . . . . . . . . . . I . . . . .
Instrument Replacement and Spare Parts . . . .
Telephone Communication System . . . . .
Motor Vehicles. . . . . . . . . . . . .
Building and Land Rental . . . .
Miscellaneous. . . . . . . . . . . . . . . . . . . .
Total Estimated Annual Operating Costs . . . .
RESEARCH PLAN COSTS
. . . . .
. . . .
Introduction. . . . . . . . . . . . . . . . . .
Personnel. . . . . . . . . . . . . . . . . . . .
Requirements. . . . . . . . . . . . . . . . . . . .
Cost s . . . . . . . . . . . . . . . . . . . . .
Instrumentation and Equipment. . . . . . . . . . . . .
Operations. . . . . . . . . . . . . . . . . . .
Total Cost. . . . . . . . . . . . . . . . . . .
xviii
XIX-l
-1
-1
-6
-13
-25
-26
-32
XX-l
-1
-1
-3
-5
-6
-9
-11
-13
-15
-17
-17
-19
-20
-21
-22
-23
XXI-l
-1
-2
-2
-5
-6
-10
-11
-------
X-I
XI-l
-10
-11
ILLUSTRATIONS
PART III - RESEARCH FACILITY
General Pattern of Usage of Fossil Fuels
. . . .
Major Point Sources. . . . . .
. . . . . .
-2
Particulate Emission Density by Study Area Zone,
Winter Average. . . . . . . . . . . . . . . . .
-3
Sulfur Oxides Emission Density by Study Area Zone,
Winter Average. . . . . . . . . . . . . . . . .
-4
Normalized Values of Surface Concentration Computed
with a Gaussian Dispersion Model (Uniform Area Source,
Slightly Unstable Atmosphere) . . . . . . . . . . . . .
-5
Normalized Values of Surface Concentration Computed
with a Gaussian Dispersion Model (Gaussian Area Source,
Slightly Unstable Atmosphere) . . . . . . . . . . . . .
-6
Normalized Values of Surface Concentration Computed
with a Gaussian Dispersion Model (Uniform Area Source,
Slightly Stable Atmosphere) . . . . . . . . . . . . . .
-7
Normalized Values at Surface Concentration
with a Gaussian Dispersion Model (Gaussian
Slightly Stable Atmosphere) . . . . . . .
. . . .
Computed
Area Source,
-8
Surface Wind Roses for Lambert Field, St. Louis
1951-1960 . . . . . . . . . . . . . . . .
. . . .
-9
Combined Wind Roses: January, April, July, October,
Lambert Field, St. Louis, Surface Observations,
1951-1960 . . . . . . . . . . . . . . . . . . . . . . .
Wind Rose--l,lOO Ft., Lambert Field, St. Louis,
Daytime Observations, 1951-1959 . . . . . . . .
. . . .
Wind Rose--l,lOO Ft., Lambert Field, St. Louis,
Nighttime Observations, 1951-1959 . . . . . . .
. . . .
xix
X-IO
XI-4
-5
-6
-7
-8
-9
-10
-15
-16
-17
-18
-------
XI-12
-13
-14
-15
-16
XIII-l
XIV-l
XV-l
ILLUSTRATIONS
Ventilation Wind Rose St. Louis EMSU Data,
May 1969 - April 1971 . .
" . . . . .
Ventilation Wind Rose, St. Louis EMSU Data, Morning
Soundings, May 1969 - April 1971 . . . . . . .
Ventilation Wind Rose, St. Louis EMSU Data, Midday
Soundings, May 1969 - April 1971 .. . . . . . . . . .
Proposed Distribution of Class A Stations. . .
. . . .
Proposed Distribution of Class B Stations, Showing
Also the Location of Class A Stations. . . . . . . . .
Instrument Station Data and Control System
. . . . . .
-2
Central Facility Data Processing System
. . . . .
-3
Logic Diagram of Instrument Identifier
. . . . .
External Features of the Jet Ranger Model 206B
-2
Interior Dimensions of the Jet Ranger Model 206B
-3
Aerial Surveillance Systems Analysis Concept
. . . .
Typical Paths of Air Masses
. . . . " "
" " . " .
-2
Monthly Distribution of Prevailing Flow Patterns,
Central U.S. ......... . . . . . . .
-3
Generalized Tracks of Highs and Lows
" " . . .
" " . "
-4
Geographic Distribution of a Number of Low-Level Jet
Occurrences for a Total Period of Two Years, According
to Evaluations by Bonner (1965) of United States
Weather Bureau Synoptic Pilot-Balloon Network Data
xx
XI-20
-21
-22
-25
-26
XIII-3
-5
-10
XIV-5
-6
-27
XV-2
-5
-6
-12
-------
IX-l
X-I
XI-l
XIV-l
TABLES
PART III - RESEARCH FACILITY
Classification of the Regional Study In~trument
Stations. . . . . . . . . . . . . . . . . . . . . . .
Selected Characteristics of Candidate SMSAs . . .
-2
Range of Pollutant Levels of the Candidate SMSAs
-3
Candidate SMSAs Ranked by Pollutant Levels
-4
Ranking of Candidate SMSAs by Value Added by Manufac-
ture for Selected Three-Digit SIC Groups. . . . . . .
-5
Separation of Central Cities of Candidate SMSAs
. . . .
-6
Agricultural Land Use in Candidate SMSA Counties
Variation of the Concentration Ratio and the Range of
the Maximum Surface Concentration for a Continuous
Point Source
. . . .
. . . . . . . .
. . . . .
-2
Major Cities and Towns Within 100 km of Downtown
St. Lou is. . . . . . . . . . . . . . . . . . . . . . .
-3
Major Area Cities and Towns Beyond 100 km of Downtown
St. Louis. . . . . . . . . . . . . . . . . . .
Weight, Performance, and Power Ratings for the Jet
Ranger Model 206B . . . . . . . . . . . . . . . . . . .
-2
Sample Weight Calculation for Jet Ranger Model 206B . .
-3
Direct and Fixed Costs for Helicopter Support Function
-4
Helicopter Support Function Total Cost of Operation. .
-5
Vector Error of Wind (m/sec) at 5 km as a Function of
Tracking System and Ratio (Q) of Mean Wind to Ascent
Rate . . . . . . . . . . . . . . . .
-6
Estimated Initial Costs of the METRAC Six-Balloon
System. . . . . . . . . . . . . . . . . .
xxi
IX-3
X-13
-16
-16
-18
-21
-25
XI-2
-12
-13
XIV-7
-8
-10
-12
-16
-18
-------
XV-l
XVI-l
TABLES
Selected Climatological Characteristics of St. Louis
-2
Frequency Occurrence of Total Hours of Inversions. . .
Interior Space Requirements for the Central Facility
xxii
XV-4
-10
XVI-5
-------
Chapter IX
INTRODUCTION TO FACILITY DESIGN AND OPERATIONS
Part III of this Prospectus includes the conceptual definition and
detailed discussion of the permanent instrument, data-handling, and
processing facility planned for the Regional Study. The facility, by
its design and mode of operation, is intended to provide the basic sup-
port to the research efforts defined within the Research Plan presented
in Part II of this Prospectus. The facility capability, however, is
expected to be augmented for several of the research experiments requir-
ing unique or specialized instruments and other equipment. Such addi-
tional required instruments and equipment or personnel are considered as
an integral part of the experimental effort rather than of the St. Louis
facility itself. The facility as defined in Part III was used to pre-
pare the estimates covering schedules, staffing, and costs presented in
Part IV of this Prospectus.
Early in the course of preparing this Prospectus, consideration was
given to the possible locations in the nation that would be suitable as
the site for the Regional Study. Four alternative sites were recommended,
and the St. Louis urbanized area and rural environs was selected as the
most appropriate by the EPA.
The St. Louis facility is conceived in this Prospectus as consisting
of a system of air quality and meteorological instrument stations estab-
lished within an area roughly enclosed by a circle of IOO-km radius with
the St. Louis arch as its center. A central support facility is also
planned, which includes data-handling and processing equipment, office
and laboratory space, and repair and maintenance shops. Most instrument
stations are expected to be linked to the central facility by telephone
circuits to permit automated remote data recording at the central facility,
The St. Louis facility is planned as the basic instrument system
of the Regional Study. It could be operated in a fully continuous or
part-time mode to develop a comprehensive data base of air quality and
meteorological conditions. Detailed statistical and other analyses
could then be performed as required by the various elements of the
Research Plan. During the various field studies and data-acquisition
efforts covered in the Research Plan, the facility would be operated to
IX-l
-------
support these efforts most effectively. Support could include equipping
the instrument stations with additional instruments, locating transport-
able stations as specified by the research group, preparation and opera-
tion of specialized data processing programs, instrument and experimental
technician support, and other activities.
Six types or classes of instrument stations are included in the
facility. These range from the permanently installed Class Al stations
with 30-meter instrument towers equipped with a full complement of air
quality and meteorological instruments to the trailer-mounted Class C2
stations having no air quality instruments and a single meteorological
instrument. The principal characteristics of the stations are summarized
in Table IX-I.
Stations of Classes AI' A2, and Bl are visualized as permanently
sited throughout the Regional Study, although this is certainly not a
fixed requirement. These stations are considered the basic units for the
long-term observational program. The Class B2 stations have the same
instrument complement as the Class Bl stations, but they are transportable
units housed in trailers. The Class Cl stations are denoted as trans-
portable units since a trailer is used for instrument installation. The
fact that the station is equipped with a 30-meter tower, however, suggests
less frequent movement than the other transportable stations.
The Class C2 stations are a hybrid unit. As part of the central
facility they include the trailer unit, tower, and digital data terminal
equipment. They will be used by the various groups carrying out field
experiments and data-gathering efforts associated with the various
research efforts presented in Part II of this Prospectus. Any additional
instrumentation required at a Class C2 station in support of these field
efforts would be considered a part of the particular research effort
rather than of the St. Louis facility. Accordingly, the instrumentation
of the Class C2 stations would be expected to vary widely over the course
of the Regional Study. This same concept would also apply to instruments
added to other classes of stations set up in the support of field
activities.
Data acquisition and handling in the St. Louis facility is expected
to be automated to the greatest possible extent. Instrument observations
at all but the Class CI and C2 stations are planned to be transmitted by
telephone circuits to the central facility for automatic computer-
controlled recording. The Class Cl and C2 stations are currently planned
to have local data-recording facilities, but further analysis might
indeed indicate these too could efficiently utilize a remote reporting
capability.
IX-2
-------
Table IX-l
CLASSIFICATION OF THE REGIONAL STUDY
INSTRUMENT STATIONS
Class of Station
A
1
A
2
B
1
B2
C
1
C
2
Number of Instruments
Air quality instruments
Nephelometer
1
1 1 1
1 1 1
1
1 1 1
1 1 1
1 1 1
1
2 2 2
Carbon monoxide - methane - hydrocarbon
1
Hydrogen sulfide-sulfur dioxide
1
Total sulfur
1
Ozone
Nitrous oxide
oxides of nitrogen
Carbon monoxide
(NDIR)
1
Hi-vol sampler
2
Meteorological instruments
Temperature
1 1 1
3 3 1 1 3 1
1 1
1 1
1 1
1 1
1 1
1 1
Station Characteristics
Wind direction and speed
Pyranometer
Pressure transducer
Mercury barometer
Net radiometer
Dew point hygrometer
Rain - snow gauge
Tower height
3D-meter
x
x
lO-meter
x
x
x
Data recording
Remote
x
x
x
'(
Local
x
x
Mobili ty
Total quantities
x x x
x x x
9 8 24 8 4 24
IX-3
Fixed
Transportable
-------
The digital data terminal equipment of the remotely reporting in-
strument stations interrogates each air quality and meteorological instru-
ment at a predetermined frequency, converts the analog instrument output
to its digital equivalent, and stores the digital data in a relatively
small-capacity magnetic core memory. Upon command from the central
facility, expected to occur at approximately 15-minute intervals for
each station, the stored data are automatically transmitted to the
central facility. Calibration curves of all instruments are stored in
the data-processing system at the central facility, so that immediate
conversion is made from the digital data format to engineering units for
archiving.
Each air quality and meteorological instrument would be equipped
with a solid state nonerasable memory unit to serve as an identifier of
each instrument. A unique serial number, or equivalent, of each instru-
ment would be coded into the identifier with the identifier then mounted
on the instrument. At each interrogation of each instrument, the identi-
fier would respond immediately before or after the instrument reading
was acquired, so that the instrument reading and its identification would
always be together. This procedure should result in an absolute minimum
of data ambiguity and erroneous interpretation and is judged to be far
superior to customary procedures employing instrument log books and other
manual methods.
The principal data-handling and processing function at the central
facility would include the recording and archiving of all instrument
station data and other field and experiment information. The archival
tapes would be forwarded to the Research Triangle Park or other EPA
installations, as appropriate, or to contractors for use in their analyses
on a particular research project. Minimal detailed data analysis is
expected to be carried out at the St. Louis facility, and the electronic
data-processing equipment is sized accordingly. Selected research ex-
periments may require limited data processing during their execution,
and the St. Louis facility should have the required capability. But,
large scale data processing covering data acquired over a period of, say,
the spring and summer seasons would be expected to be undertaken on the
larger computer systems existing at the Research Triangle Park and
elsewhere.
The instrument stations comprising the St. Louis facility are
planned to bp. located within a circle of approximately IOO-km radius
centered generally on the St. Louis arch. Eight Class A stations are
symmetrically located around the IOO-km circle, while an additional eight
are symmetrically deployed on a 40-km square with the arch as its center.
The final Class A station is planned at the arch itself. Depending Upon
IX-4
-------
actual conditions at the arch site, advantage might well be taken of
nearby taller television or other towers.
The Class BI stations are planned for installation on a uniform
square grid about the arch with station spacings of approximately 12 km.
The remaining stations are considered to be transportable and would
be deployed as required to support a given field data-acquisition pro-
gram. This is particularly true for the Class C2 stations. The stations
other than the Class C2 would generally be expected to provide the over-
all or ambient observations of air quality, meteorological, and other
parameters of interest on an area-wide basis. Observations in detail
within a specific smaller area would be carried out by deployment of the
Class C2 stations within the area of interest.
The precise station locations will, of course, depend upon the
availability of suitable sites, so that the pattern presented here must
be regarded as tentative. Instrument station sites must be selected with
respect to several important factors, including freedom from unique or
overriding micrometeorological effects, general absence of nearby sig-
nificant pollutant sources, convenient access to electric power and com-
munication utility services, and free access at all times. Only a
detailed field survey will reveal sites possessing these and other
necessary characteristics.
The St. Louis ground-based instrument system is expected to be aug-
mented by an aerial observational capability. Aerial measurements are
expected to be made on both a routine basis and in support of special
research experiments using aircraft and the proposed METRAC system. The
existing EPA aerial observation capability has been examined along with
alternative methods. Significant advancement in the state of the art of
aerial measurement is expected during the course of the Regional Study,
so that planning of the aerial support program should be flexible.
IX-5
-------
Chapter X
ST. LOUIS SITE SELECTION
Summary
The preparation of the Regional Study Prospectus depended in part
upon the actual area of the nation in which the Regional Study would be
carried out. Accordingly, an early task entailed the development of site
selection criteria and the evaluation of the larger urbanized areas of
the nation with respect to their conformance to the criteria. Because of
the ready availability of statistical data, the Standard Metropolitan
Statistical Areas (SMSAs) were used as the basis for analysis. The eval-
uation resulted in the nomination of four candidate sites. These were:
Birmingham, Alabama
Cincinnati, Ohio
Pittsburgh, Pennsylvania
St. Louis, Missouri, including
the Illinois portion.
These candidate sites and the data supporting their nomination were sub-
mitted to the EPA for consideration. St. Louis was selected as the Re-
gional Study site by the Assistant Administrator for Research and Moni-
toring. The rating of the four sites for each criterion is provided in
the following tabulation.
Criterion Birmingham Cincinnati Pittsburgh St. Louis
Surrounding area Fair Poor Good Good
Heterogeneous emissions Fair Fair Fair Good
Area size Good Good Good Good
Control program Poor Good Good Good
Information Poor Good Fair Good
Climate Good Fair Fair Good
Birmingham compares reasonably well with most of the more important
criteria but in general it is lowest overall. The bulk of the surround-
ing area is essentially rural, although not heavily agricultural, and
the nearest candidate SMSA is approximately 140 miles away. Interarea
X-I
-------
pollution would tend to be small. The area has a fair degree of hetero-
geneous emissions but is strongly dominated by the steel industry. The
area has no petroleum refineries and only a small representation of the
chemical industry, chiefly agricultural fertilizers. No power plants are
within the urban area. Space heating is largely by natural gas, but this
tends to be true for all areas. Observed atmospheric concentrations of
suspended particles are quite high, but S02 concentrations are low. The
Birmingham area is moderate in size, so that the instrumentation and other
facilities necessary for the Regional Study ought not be excessive. The
pollution control program in the Birmingham area is so poorly defined as
to make the area attractive perhaps for the Regional Study. This advan-
tage would derive from the chance to follow the changes in the area as
control programs were initiated and sources were brought into compliance.
However, only minimal historical information is available and the Regional
Study would be faced with the task of developing a complete set of re-
quired information. Climatically, Birmingham would be a good site. The
area is sufficiently northerly to have the continental-type climate rep-
resentative of a large section of the nation but, at the same time, its
southerly location would permit virtually year-round experimental efforts.
The surrounding area is forested hills with not much area used for farm-
ing. On a scale of 1 to 3 for "poor" through "good," Birmingham scales
12 out of a possible 18.
Cincinnati is rated as poor with regard to the surrounding area.
The SMSA is contiguous with two other SMSAs and the central city is sepa-
rated from Dayton--a candidate region--by about 40 miles. Rural fringes
may be smaller than those in other areas and interarea pollution is a
greater possibility. The heterogeneity of emissions is fair. The area
has some ferrous and nonferrous metal industry and small levels of the
primary chemical and petroleum refining in.dustries. The Cincinnati area
size is rated as good but the influence of Dayton and other SMSAs in the
area must be considered. The pollution control program in the area has
been in existence for some time and has shown considerable effectiveness,
so that the area is ranked as good in this regard. The available histori-
cal information is appreciable; the programs conducted at the Taft Center
and in Cincinnati when the federal air pollution program was centered
there would be especially valuable in providing important historical in-
formation. Finally, the climate in the area is representative of the
continental type and should permit experimental efforts throughout most
of the year. Overall, Cincinnati scales 14 on this final list.
Pittsburgh appears overall to be a generally suitable area. Even
though the SMSA is contiguous with two others, they are relatively small
and should not cause significant interarea pollution. The nearest candi-
date SMSA--Youngstown-Warren--lies over 100 miles to the northwest, pro-
viding a reasonable degree of isolation. Appropriate rural areas appear
X-2
-------
to be in relative abundance. The area has a fair heterogeneity of emis-
sion sources, although dominated by the steel industry. The area does
have some industrial activity in chemicals, petroleum refining, and non-
ferrous metals. The size of the Pittsburgh area would not appear to
strain the resources expected to be available in the Regional Study. Pol-
lution control programs have been in existence for an appreciable period
and have had marked effect. However, the current pollution levels are
high enough that the continuing abatement and control program should pro-
vide important data for the Regional Study. Historical data are regarded
as fair; studies of various types having been conducted since 1960 usually
on a county or city basis within the SMSA. Climatically, Pittsburgh is
rated as fair. Winters in the area are sometimes severe and may restrict
year-round experimental activities. More importantly, however, Pittsburgh
lies in a distinct river valley which tends to make it less than ideally
representative of many areas and perhaps not immediately applicable for
many generalized air quality models. On an overall basis, Pittsburgh
earns a scale value of 15.
St. Louis is rated as good with respect to all points. The nearest
candidate SMSA is 250 miles distant, giving St. Louis the highest measure
of isolation of the four considered here. The rural fringe should be ex-
tensive. The heterogeneity of emissions is appreciable, probably greater
than that of any other candidate area. The area has large facilities for
steel production, oil refining, and primary chemical production; signifi-
cant facilities are also found for nonferrous metal production and agri-
cultural fertilizers. The area size is considered as good but likely
represents the upper limit. The air pollution control program of St.
Louis is one of the oldest in the nation, and it has been relatively ef-
fective. Observations of the effectiveness of abatement and control mea-
sures during the course of the Regional Study may be therefore more dif-
ficult to obtain than in other candidate areas. However, since the St.
Louis area was relatively highly polluted at the initiation of the program
and still seems to exceed air quality criteria, effectiveness measures
still might be readily developed. The available historical information
for St. Louis is probably more plentiful than that for any other candidate
area. Representative of these data is the large multiagency eight-volume
study of the St. Louis area published in 1967 by the Public Health Service
under the title "Interstate Air Pollution Study." Reports of this type,
plus the 20-year air pollution control program, provide impressive data.
Additionally, at least three studies of the St. Louis area are currently
under way that will add to the available information. Finally, the St.
Louis area climate tends to be representative of many candidate areas
and also permits virtually year-round experimental activity. The scale
value for St. Louis is 17.
X-3
-------
Site Selection Criteria
The extent to which the objectives and goals of the Regional Study
can be achieved may indeed depend in an important way upon the region
selected for the experimental effort. Accordingly, the objectives and
goals of the Regional Study should stand as first-order constraints on
the development of selection criteria for the region followed by
second-order constraints resulting from numerous factors such as required
logistic support, institutional arrangements, and organizational proce-
dures. In brief, the general objectives of the Regional Study as provided
in the contract covering the preparation of this Prospectus are as follows:
.
Develop and apply techniques to permit
of the distribution of both stable and
over distances of 150 kilometers.
meaningful predictions
reactive pollutants
.
Improve the understanding of
with S02' NOx' hydrocarbons,
dehydes.
atmospheric reactions associated
ozone, organic nitrates, and al-
.
Define an optimum network to
and meteorological phenomena
control region.
measure and monitor air quality
within a typical air quality
.
Prepare methods to obtain and update emission inventories
within a region and develop such an inventory for the se-
lected region.
The principal site selection criteria for the Regional Study include
the following:
.
Climate--The site should possess a climate representative of
a large proportion of the major urban regions in the nation.
It should permit experimental work throughout the greatest
possible portion of the year.
.
Heterogeneous Emissions--The site should have a broad spectrum
of emissions as determined by fuels consumed, industrial pro-
duction processes, and the current pollution mix. An impor-
tant specific factor is the extent of coal usage because of
its relationships to both sulfur oxide and particulate emis-
sions.
.
Surrounding Area--This criterion includes measures of the iso-
lation of the site from other major sources of pollutants and
X-4
-------
the presence of a rural fringe. Sites in close proximity to
large bodies of water should be avoided because of their cli-
matic influences and their constraint on the location of air
quality and meteorological instruments.
.
Area Size--The criterion of area size provides a general
sure of the expected scope and magnitude of the Regional
for each candidate site.
mea-
Study
.
Pollution Control Program--The various candidate areas have
considerable variation in their existing control programs.
It appears desirable for the Regional Study to be carried
out at a site where the control program is generally well de-
veloped but can be significantly improved. Such a site would
provide a background of data and general experience that could
be used to establish the Regional Study. In addition, an ex-
isting program could provide initial contacts with industrial
sources for the gathering of source inventory data.
.
Historical Information--Historical data, including meteorolog-
ical and pollution, applicable to a site also vary widely. In
general, a site tends to be more attractive as the quantity of
historical data increases. Care must be exercised, of course,
to distinguish between quantity and quality.
The selected region should represent to the greatest possible degree
all other regions in the nation. On the basis of climatology, atmospheric
chemistry and physics, and emission sources, a region must be selected
that combines the most important characteristics of all separate regions.
At the same time the region should be comparatively "simple" with respect
to these characteristics. That is, since the state of the art in pollu-
tion modeling and prediction has many limitations, some of which should
be eliminated by the Regional Study, the accepted scientific approach is
to proceed from the simple to the complex. Accordingly, the region ought
not encompass unusual and difficult to treat topographical and relief fea-
tures, which may markedly affect and interact with the synoptic and mezo-
scale meteorological factors or for any reason may cause the geographical
pattern of location of the emission sources to be unusually configured.
The selection of a simple area, moreover, may tend to permit the use of
a relatively small number of meteorological and air quality monitoring
instrument stations.
The Regional Study interaction with the local pollution control pro-
gram will be most beneficial. Clearly, however, the region must have all
pollutants in its atmosphere which are intended to be examined in the
course of the study. Moreover, the quantities of each pollutant should
X-5
-------
be present in substantial amounts to minimize the need for highly sensi-
tive instruments and elaborate facilities.
The criterion that the region should have a reasonably
set of areas to permit accurate measurements, inventorying,
activities associated with each of the study goals leads to
minimum requirements:
well-defined
and related
the following
.
One or more heavily industrialized sections containing a broad
spectrum of manufacturing and processing facilities producing
the required range of pollutants
.
A well-defined pattern of vehicular traffic flows, including
a distinct central business district
.
A generous rural fringe or annulus containing agricultural
crops and other vegetation, which may be affected by the
various pollutants. The rural fringe should be extensive
enough to permit pollutant monitoring at distances up to
100 kilometers from the principal pollutant sources.
These requirements on land usage within the region indicate that the
region should be relatively isolated from other principal pollution sources
Moreover, they tend to limit the size of the region. While difficult to
quantify at the present time, the diffusion models themselves and their
required computer capacities impose certain ceilings on the number of in-
put elements and the atmospheric volumes which can be treated at anyone
time. Additionally. design and funding requirements for meteorological
and air quality instrumentation must be recognized and can easily impose
limitations upon the size of the region. By the same token, the time and
effort needed to develop a comprehensive emission inventory of a region
will surely act as at least a secondary restraint on region selection.
In the event of modest rather than substantial separation between
the region and a second major emission complex, the need may arise for
selected instrumentation to monitor pollution levels in the region caused
by emission sources elsewhere. The character and extent of such facili-
ties are uncertain, but their design might well require procedures similar
to that for the regional system itself. However, selection of the region
should be made to minimize and hopefully eliminate the need for these fa-
cilities.
X-6
-------
Method of Analysis
The examination of all potentially suitable regions leading toward
selection was carried out in phases. First, the nation as a whole was
summarily viewed on a gross climatological, coal use, and regional sepa-
ration basis. This procedure permitted the elimination of both east and
west coastal regions from further consideration. Second, within the re-
maining area, all SMSAs with populations in excess of 400,000 for a total
of 33 were analyzed to determine the following characteristics:
.
Geographical separation from other candidate SMSAs and all
other SMSAs
.
Population concentrations and area size
.
Existing pollution levels of representative types.
The results of this analysis lead to the elimination of 20 SMSAs.
The remaining 13 were examined with respect to the manufacturing industry
existing in the SMSA to provide measures of the possible variation in the
types of pollutants. Industrial activity was also characterized by its
value added by manufacture to gain a qualitative measure of possible emis-
sion intensities. This review permitted the elimination of selected ad-
ditional SMSAs because of a relative lack of industrial concentration.
The final group was treated on a more subjective basis with respect
to the following criteria:
.
Area size
.
Climatological conditions affecting or prohibiting experimental
operations
.
Geographical and political conditions affecting instrument lo-
cations.
Four SMSAs remained at the conclusion of this final screening process.
Although the existence of a pronounced rural fringe was virtually insured
by the criteria of SMSA separation, the final four candidate SMSAs were
examined in more detail with regard to agricultural activity. Even though
agricultural activity was not examined within all counties expected to be
included within the Regional Study, the criterion of separation combined
with high agricultural activities within the SMSA itself provides a con-
fident expectation of a highly suitable rural fringe in the Regional
Study area.
X-7
-------
National Summary Analysis
Meteorology
The site of the Regional Study should be selected within an area
where the climate is generally representative of the greatest possible
portion of the nation. Translation of research findings and application
to other urban areas should be greatly facilitated by such a location.
The nation's climate can be generally classified into four types.
The first is termed the Continental. In general, the western limits of
the Continental climate are the Cascade-Sierra Nevada-Coast Range system,
while the eastern limit tends to follow the eastern slopes of the Appa-
lachian Chain. The southern limit is somewhat less distinct but roughly
parallels the Gulf Coast at distances 200-250 miles inland. Although
significant differences exist in virtually all climatological factors
from north to south and east to west within this area, there is neverthe-
less an overall sense of commonality within the area and sharp differences
with the remainder of the nation. The climatic homogeneity increases
markedly within the area east of the Rockies.
The second type of climate includes the Pacific Coast areas west of
the Cascade-Sierra Nevada-Coast Range system. The Pacific Coast climatic
region has several distinct components ranging from the California or Med-
iterranean type to the Marine of the Oregon and Washington Coasts. Never-
theless, all are influenced by the Pacific Ocean and have mild winters
and generally dry summers, and much less distinct seasonal variations are
observed than in the Continental areas.
The third type includes the Atlantic seaboard north of approximately
the 40th parallel. The general climate of the area is in some sense a
derivative of the Continental type due to the prevailing westerly winds.
However, the Atlantic Ocean constrains the seasonal variations so that
both winters and summers tend to be more moderate than those of the Con-
tinental type.
The fourth climatic type, sometimes termed the Humid Subtropical,
exists along the south Atlantic Coast and extends westward along the Gulf
Coast to southeastern Texas. It is generally characterized by very hot
humid summers and normally mild winters but with occasional very cold
periods. Rainfall occurs throughout the year. Seasonal changes are
far less pronounced than in the Continental climate. Storm disturbances
tend to be more frequent and severe than in other climatic areas of the
nation.
X-8
-------
Since the Continental climate indeed embraces the largest portion of
the nation, this area would appear most appropriate as the Regional Study
site. Care should be taken, however, in selecting the site, so that the
greatest possible period of the year can be used for the conduct of re-
search experiments. This provision tends to exclude the northern portions
of the area in which lengthy winter periods could hamper or indeed prevent
field work for periods of as much as four to five months.
Fossil Fuels
The types and quantities of fossil fuels used within the region will
influence the nature of the pollutants, the content of the emission
inventory, and the procedures developed for control. Accordingly, it is
eminently desirable that the region include emission sources using all
types of fossil fuels in significant quantities.
Representative fuel usage for stationary sources can be typified by
public utility practices. Fossil fuels used in electric power generation
consist almost exclusively of coal, fuel oil, and natural gas. Very small
quantities of diesel oil, gasoline, and other fuels are used for some
standby and peaking generating units. Nationally, coal accounts for ap-
proximately 60% of power generation, while fuel oil and natural gas stand
at 10 and 30%, respectively. Long term trends show natural gas claiming
an increasing share at the expense of coal, with fuel oil remaining rela-
tively constant.
The examination of the geographical pattern of fossil fuel usage
throughout the nation reveals significant differences among states and
regions. The variations stem primarily from the differing costs of fuels
from state to state but also results to some extent from pollution con-
trol programs which favor natural gas over coal. The general pattern of
fossil fuel usage is shown in Figure X-I. The New England states princi-
pally use fuel oil, since it is readily available at relatively low cost
by ocean tanker and barge transport. Most generating plants are near
tidewater. At the other extreme, the Pacific Northwest states use vir-
tually no fossil fuel because of the predominance of hydroelectric facil-
ities.
The remainder of
and natural gas-using
South Central states,
and California. The
nantly coal users.
the nation can be approximately divided into coal-
regions. The natural gas region includes the West
the Gulf Coast areas, portions of the Southwest,
remaining areas and states, therefore, are predomi-
X-9
-------
~
I
t-'
o
NONE
COAL
'- -----",
GAS
/'
SA-1365-6
FIGURE X-l
GENERAL PATTERN OF USAGE OF FOSSIL FUELS
-------
The use of fuels by manufacturing industries tends to follow the
same general pattern shown by utilities but with the coal-natural gas
interface displaced northward. On the other hand, essentially no coal
is used south of the interface. Likewise, commercial and residential
space heating by gas is very common throughout the nation with the ex-
ception of New England. In fact, over 60% of the space heating in resi-
dential structures is accomplished by natural gas.
Regional Isolation
In considering the criteria of regional isolation, a summary map
analysis should be sufficient for the purpose of focusing attention on
the most suitable area of the nation for which a more detailed analysis
will be appropriate. For purposes here as well as more detailed subse-
quent analysis, the use of the SMSA appears most appropriate as the basic
unit of measure. However; since the definition of an SMSA is based upon
population and county boundaries, some distortion in gaging isolation of
candidate regions may arise. For example, the Las Vegas SMSA is adjacent
to the San Bernardino-Riverside-Ontario SMSA. However; the population
and economic activities of both are concentrated within a very small area
within each SMSA and are separated by over 200 miles. Thus, in fact, a
high degree of isolation may actually exist in adjacent areas.
With this limitation in view, the geographical pattern of SMSAs
nevertheless reveals an exceptionally high concentration in two areas of
the nation. The first includes the well-known Washington, D.C.-Boston
corridor in which the SMSAs are virtually continuous and have heavy con-
centrations of industry, electric power central stations, transportation
facilities, and other emission sources throughout its length. Moreover,
the residential population is sufficiently dense and widespread to pre-
vent most of the area from having a suitably expansive rural fringe nec-
essary for the study. These conditions present extreme problems in de-
fining a region in terms of both a consolidated set of pollution sources
which can be considered as a unit and a pollution receptor area which is
clearly coupled with the sources. Accordingly; unless other compelling
reasons arise, the selected region should best not be within the
Washington-Boston corridor.
The second area showing somewhat the same characteristics is the
industrial crescent around the south and east of Lake Erie from Cleveland
through Toledo to north of Detroit. Although the pollution sources
through the area tend to be more closely grouped or form less of a con-
tinuous band than the Washington-Boston corridor, the possibility of
their interaction in receptor areas nevertheless is likely to occur.
X-ll
-------
However, rather than totally discard candidate regions in this area at
this point, they will be retained for more detailed analysis~
Results
The combination of noncoal usage and the lack of a representative
climate provides the basis to eliminate the Pacific Coast area from fur-
ther consideration. Lack of coal use west of the Rocky Mountains to the
Pacific Coast area discourages further consideration of this area as well.
(Of course, few, if any, suitable sites exist in this latter area.) The
southwest and south central regions and the southmost fringe of the south-
ern states also can be ruled from further consideration by the combination
of somewhat atypical climatic conditions and the almost complete exclusion
of coal. Finally, the Atlantic seaboard east of the Appalachian Mountains
is unsuitable due to the general complexity of the area, intense indus-
trial and urban development, and the lack of a clearcut representative
climate and well-defined rural fringe.
General Analysis of Standard Metropolitan Statistical Areas
The foregoing process of selection permits the narrowing of focus to
candidate sites situated east of the Rocky Mountains, west of the Appa-
lachian divide, and north of approximately the 32nd parallel, except for
those included in the northward excursion of the natural gas-coal inter-
face. A total of 33 SMSAs are situated within the area with populations
greater than 400,000. SMSAs smaller than 400,000 are excluded from fur-
ther consideration because they tend to lack an adequate industrial mix,
transport, and other features necessary for the Regional Study. These
SMSAs plus an additional four in the noncoal region are shown in Table X-I.
The latter four are intended to show certain variations in characteristics
from the prime candidate site.
Table X-I shows for each candidate SMSA a number of descriptors for
the following characteristics:
Levels of Representative Pollutants--The Regional Study is expected
to focus on a large number of pollutants. Accordingly, the region se-
lected for study must have them in sufficient concentrations to permit
thorough analysis. Comparatively high levels of pollutants are desirable
to reduce the need for highly sensitive instruments and associated physi-
cal and chemical analysis. Moreover, it is likely that all candidate re-
gions will be engaged in active control and abatement programs during the
X-12
-------
Table X-I
SELECTED CHARACTERISTICS OF CANDIDATE SMSAs
Oxides of Oxides of Nearest Candidate
Hydrocarbons Nitrogen Sulfur Annual Mean Nearest SMSA SMSA
T/mi 2/day T /mi 2/d ay T/mi 2/day Su spended Central Central Possible
Central Central Central Particles City City Sunshine Degree
~ SMSA ~ SMSA ~ SMSA Boundary Separation Boundary Separation P95 P50 Percent Days
- -
1. Akron 0.482 0.075 0.511 0.088 2.199 0.343 134 0 20 0 30 20 5 0.8. 6,345
2. Albany-Schenectady-Troy 0.642 0.034 1. 204 0.063 4.618 0.243 *
86 10 30 50 120 10 9 54 6,825
3. Atlanta 0.505 0.072 0.379 0.054 1.115 0.159 97 40 80 90 140 19 7 61 2,961
4. Birmingham 0.382 0.054 0.652 0.091 1.500 0.211 142 0 40 90 140 28 6 58 2,779
5. Buffalo 0.706 0.071 0.751 0.075 2.366 0.238 139 0 70 0 70 22 6 53 7,213
6. Chicago 0.636 0.164 0.705 0.182 3.100 0.800 124 0 0 0 0 36 10 58 6,113
7. Cincinnati 0.432 0.049 0.677 0.076 3.719 0.419 154 0 10 0 40 24 58 4,806
8. Cleveland 0.322 0.124 0.346 0.134 1. 068 0.413 116 0 10 0 30 27 52 6,432
9. Columbus 0.062 0.060 0.392 0.038 0.680 0.066 124 10 40 10 70 24 5 57 5.444
:><: 10. Dallas 0.210 0.037 0.170 0.030 0.013 0.002 101 0 30 0 30 35 7- 66 2,363
I
I-' 11. Dayton 0.631 0.046 0.666 0.049 2.776 0.203 143 0 30 0 40 22 5 58 5,606
t.o)
12. Denver 0.214 0.010 0.450 0.020 1.613 0.073 126 30 60 300 350 24 6 70 5,982
13. Detroit 0.590 0.221 0.610 0.228 2.145 0.804 161 0 30 0 50 26 10 54 6,275
14. Flint 0.641 0.037 0.416 0.024 1.0915 0.063 n.a. 0 50 0 50 21 5 n.a. 7,007
15. Fort Worth 0.253 0.043 0.210 0.035 0.010 0.002 133 0 30 0 30 20 5 n.3. 2,405
16. Gary-Hammond-East Chicago n.8. 0.072 0.3. 0.150 0.3. 0.616 164 0 0 0 0 17 7 58 6,113
17. Grand Rapid s 0.577 0.037 0.487 0.031 1.732 0.111 o.a. 20 50 70 100 26 5 48 6,894
18. Houston 0.409 0.028 0.556 0.039 0.019 0.001 102 . 0 40 100 200 42 7 66 1,278
19. Indianapolis 0.783 0.043 0.674 0.037 3.072 0.168 154 0 30 30 100 27 6 60 5,699
*
20. Kansas City 0.447 0.046 0.362 0.037 0.411 0.042 103 0 50 150 260 22 7 64 4,868
21. Louisville 0.614 0.093 0.970 0.145 4.950 0.741 148 40 70 40 100 13 5 58 4,718
22. Memphis 0.440 0.051 0.444 0.051 0.868 0.099 115 80 140 140 210 17 6 65 3,116
23. Mi lwaukee 0.328 0.125 0.349 0.133 1.890 0.721 150 0 20 20 100 25 6 56 7,635
*
24. Minneapolis-St. Paul 0.255 0.080 0.222 0.069 0.538 0.168 94 70 140 230 300 19 58 8,394
-------
Table X-I (Concluded)
Oxides of Ox~des of Nearest Candidate
Hydrocarbons Nitrogen Su lfur Annual Mean Nearest SMSA S'.ISA
T/mi2/day T/mi2/day T/mi 2 /day Suspended Central Central Posslble
Central Central CBntral Part i cles City City Sunshine Degree
~ SMSA ~ SMSA ~ SMSA Bound ary Separation Boundary Separation P95 P50 Percent Days
- -
25 Nashvllle 0.443 0.035 0.556 0.044 3.162 0.250 115 60 120 110 150 24 5 58 3,578
26. Oklahoma C>ty 107 30 100 30 100 21 68 3,844
27. Pittsburgh 0.425 0.073 0.886 0.153 3.948 0.680 151 0 30 30 100 30 9 55 5,987
28. Rochester 0.582 0.028 0.496 0.024 2.420 0.118 106 0 70 0 70 37 4 55 6,748
29. St. Louis 0.717 0.084 0.843 0.066 3.722 0.292 170 40 90 150 250 29 8 58 4,809
30. San Antonio n.3. n.a. n.a. o.a. n.a. n.a. 73 10 80 100 200 16 62 1,549
31. Syracuse n.a. n.a. n.a. O.a. n.a. o.a. 161 0 40 10 80 32 5 51 6,756
32. Toledo O.a. n.3. n.a. 11.3. n.a. n.a. 92 0 60 0 60 25 5 59 6,494
*
33. Youngstown-Warren 0.484 0.050 0.705 0.074 2.142 0.224 138 0 50 0 50 20 6 o.a. 6,212
~
I
1-"
~
/
0.3. = not available.
*
Average of stations in area.
-------
course of the Regional Study, so
high levels will perhaps provide
fectiveness of the programs.
that selection of a region with initially
indices for use in the study of the ef-
Arithmetic mean annual measures of three important pollutants are
shown in Table X-I and include emission levels in tons per square miles
of hydrocarbons, oxides of nitrogen, and oxides of sulfur for both SMSA
and the central city. Additionally, the annual mean levels of suspended
particles in micrograms per cubic meter are shown for the central cities.
Degree of Isolation--The region should be somewhat isolated from
other comparable areas to provide for a suitable rural fringe, to permit
a well-defined emission inventory, and otherwise to minimize interaction
of the study region with outside pollution sources. Three measures are
provided. The first is the distance from the candidate SMSA boundary to
the nearest SMSA and the corresponding distance between the principal
cities of each SMSA. The second is the distance from the candidate SMSA
boundary to the nearest other candidate SMSA and the corresponding dis-
tance between the principal cities of the two SMSAs. The last includes
the radius of a circle enclosing 50 and 95% of the population of the SMSA
which provides a general measure of concentration and the urban-rural mix.
Possible Sunshine--Since some of the pollutants to be considered in-
clude those resulting from photochemical reactions, a measure of sunshine
in the region is of interest. The percentage of actual sunshine to the
maximum possible is shown.
Degree Days--The mean annual degree days with a base of 65°F is the
mean of the annual sums of the differences between 65°F and the daily
mean below 65°F. This measure provides an indication of space heating
requirements and the possible corresponding pollutant levels from fuel
usage.
Pollutants
The ranges of pollutant emissions and suspended particles within the
regions show wide variations as summarized in Table X-2.
In general, areas having high emission levels for one pollutant show
correspondingly high levels for most others, although significant varia-
tions from this pattern are noted. To ensure adequate pollution levels
for the Regional Study, it is preferable to select a region with moderate
X-15
-------
Table X-2
RANGE OF POLLUTANT LEVELS OF THE CANDIDATE SMSAs
Pollutant
Central City SMSA
Maximum Minimum Maximum Minimum
0.783 0.062 0.221 0.010
1.204 0.176 0.228 0.020
4.950 0.010 0.804 0.002
170 73 n.a. n.a.
Hydrocarbons, tons/mi2/day
Oxides of nitrogen, tons/mi2/day
Oxides of sulfur, tons/mi2/day
Suspended particulates, ~g/m3
to severe levels of all pollutants. Accordingly, Table
ranking of the top 15 SMSAs and central cities for each
ies are denoted by their number as shown in Table X-I.)
X-3 provides a
po 11 u tan t . (C it.
Table X-3
CANDIDATE SMSAs RANKED BY POLLUTANT LEVELS
Oxides of Oxides of
Hydrocarbons Nitrogen Sulfur Particulates
Central Central Central Central
Rank City SMSA City SMSA City SMSA City
1 17 13 2 13 19 13 26
2 26 6 19 6 2 6 15
3 5 21 24 24 24 19 28
4 2 8 26 15 26 21 13
5 31 19 5 19 7 24 17
6 6 26 6 8 33 15 7
7 11 22 7 21 6 7 24
8 19 1 17 4 17 8 21
9 13 24 11 1 11 1 9
10 25 3 4 7 15 26 11
11 3 15 13 5 5 33 4
12 30 5 16 22 1 2 5
13 1 9 33 26 13 5 30
14 18 4 1 2 21 4 1
15 20 20 25 3 12 11 14
X-16
-------
As can be noted from Table X-l, pollutant data were not readily
available for seven candidate areas. However, within reasonable limits
the areas without such data can be taken as being equivalent to other
areas as shown below.
Area
Equivalent Area
Oklahoma City
Tulsa
San Antonio
Dallas
Salt Lake City
Denver
Syracuse
Rochester
Toledo
Cleveland
These top 15 areas can be narrowed for further consideration on the
basis that the SMSAs and central cities should each rank in the top 15
for at least all but one pollutant or show the highest level for at least
one pollutant. A total of 13 SMSAs meet this criterion as follows:
Akron
Albany-Schenectady-Troy
Birmingham
Buffalo
Chicago
Cincinnati
Dayton
Detroit
Gary-Hammond-East
Indianapolis
Louisville
Pittsburgh
St. Louis
Chicago
Manufacturing
The pollutant levels within the candidate SMSAs are generally related
to the sheer size of the population, the number and use patterns of motor
vehicles, and similar factors. Additionally, emission levels are poten-
tially related to the size and type of manufacturing industry located in
the SMSA. Table X-4 shows the type and magnitude of 15 different manu-
facturing industries in the 13 SMSAs having high emission levels. Manu-
facturing activity is classified by the Standard Industrial Classification
(SIC) three-digit groups for each SMSA. The table indicates the SMSA rank
X-17
-------
Table X-4
RANKING OF CANDIDATE SMSAs BY VALuE ADDED BY MANUFACTURE FOR SELECTED TIfREE-DIGIT SIC GROUPS
SMSA 262 265 281 282 287 291 301 321 324 327 331 ]32 333 334 336
-
1. Akron * /0,07 41/0,49 68/0,40 /0,01 1/31. 68 75/0.25 * /0,02 33/0,71 108/0,12
2. Albany-Schenectady-Troy 21/1. 41 41/0.50 76/0.21 33/1. 22 * /0.07 ll/1. 46 91/0.22 47/0,27 5/3.34 /0,02
3. BirminJ;;ham 95/0,18 86/0,17 13/1. 75 102/0.05 3/2 ,36 27/0.64 3/7.35 1l/2.39 13/1,33 86/0,17
4. Buffalo 44/0.79 15/1.16 5/3.80 31/1. 26 3/3 ,41 26/0.77 19/1. 20 21/1.01 19/0,77 7/3.71 10/2,47 29/0 . 88 43/0,43
5. Chicago 132/0.10 2/8.70 13/2.16 17/1.85 12/1. 80 4/8 ,20 * /0.01 19/1,19 3/2,71 1/14.76 1/6.84 41/0,53 1/16.37 1/11.27
6. C~nclnnati 73/0.37 11 /1. 33 34/0.59 52/0.56 31/0,78 60/0,20 35/0,54 42/0.33 37/0,62 14/1.56 18/1. 06 25/0,!!0
7. Dayton 74/0.37 52/0.36 * /0.04 127/0.12 16/1.29 14 /1. 38 34/0.55 . /0.03 15/1,64 27/0.78
8. Detroit 177/0. II 18/1. 09 19/1.15 65/0.42 71/0.26 1l/1. 96 6/4.13 4/7.46 6/2. 15 5/1. 69 4/5.81 9/2.89 75/0.05 8/2.16 5/3.75
9. Gary-Hammond-East Chicago t
10. Indianapolis 28/0.67 11!!/0. II 42/0.54 64/0.15 21/1. 17 36/0.50 * /0.02 25/0.90 49/0,34 82/0.20
11. Louisville 25/0.75 29/0.63 6/2.86 46/0.46 65/0,15 17/1. 37 57/0,35 39/0.60 64/0.13 * /0.01
X 12. Pittsburgh 139/0.05 14/1. 21 31/0.61 37/0. 9!! 61/0.15 * /0.01 2/9.29 22/1,01 17 /0 . 80 2/13.7 3/3.50 29/0.95 23/0.93 10/1.82
I
...... 13. St. Louis 7/2.34 7/3.61 * /0.05 9/2 . 12 9/2.66 35/0.34 9/5.55 7/1. 87 15/0.92 13/1. 58 8/2.!!9 13/1. 73 6/2.93 8/2 . 41
00
* Unavailable but lower ranking than all other SMSAs.
t Incorporated with Chicago SMSA.
Source: Location of Critical Industries, Stapford Research Institute, October 1969.
-------
in production for each group and its percentage of the national value
added by manufacture for each SMSA. The SIC three-digit groups are:
262
265
281
282
287
291
301
321
324
327
331
332
333
334
336
Paper mills, except building paper mills
Paperboard containers and boxes
Industrial inorganic and organic chemicals
Plastics materials and synthetic resins, synthetic,
rubber, synthetic and other man-made fibers, except
glass
Agricultural chemicals
Petroleum refining
Tires and inner tubes
Flat glass
Cement, hydraulic
Concrete, gypsum, and plaster products
Blast furnaces, steel works, and rolling and finishing
mills
Iron and steel foundries
Primary smelting and refining of nonferrous metals
Secondary smelting and refining of nonferrous metals
Nonferrous foundries.
Production facilities for the three-digit manufacturing groups are
found in most of the candidate SMSAs. Flat glass, SIC 321, however, is
limited to three SMSAs and primary smelting and refining of nonferrous
metals, SIC 333, is limited to five. Only the Detroit SMSA has manufac-
turing activity in all 14 three-digit groups.
Blast furnaces, steel works, and rolling and finishing mills, SIC 331,
exist in all but the Louisville area. The ranges of value added by manu-
facture lie between Chicago at 14.76% to Akron and Indianapolis at 0.02%
each. The development of a strict definition of the ideal mix of indus-
try in SMSA does not appear possible or necessary. It is nevertheless
desirable that the area have the widest possible mix with perhaps the in-
dustries somewhat in balance. Such a condition would permit a more com-
prehensive emission inventory analysis and ensure the largest possible
X-19
-------
number of pollutant types. In this regard, the St. Louis SMSA tends to
show perhaps the most satisfactory industrial mix without undue concen-
tration of a single industry. The area generally ranks in the range of
sixth to twelfth in the nation for most industrial groups represented.
On the other extreme is the Akron SMSA with very low levels of produc-
tion in all three-digit groups except for the overwhelming first-place
position or 31.68% of production for SIC 301, tires and inner tubes.
Broadly, on the basis of current manufacturing activities, the fol-
lowing SMSAs appear most suitable.
Birmingham--Although dominated by SIC 331, blast furnaces, steel
works, and rolling and finishing works, and related industrial groups
and lacking significant SIC 291, petroleum refineries, the SMSA shows
a reasonably comprehensive industrial mix.
Buffalo--The SMSA is not significantly dominated by any particular
industrial group and is one of the few candidate areas with any signifi-
cant capacity in SIC 262, paper mills, which is, of course primarily in
northern non-SMSA counties of the nation.
Chicago--The area presents a very attractive industrial mix with
almost all groups represented to a major degree but perhaps to an exces-
sive degree for the Regional Study resources.
Cincinnati--The area presents a relative balance among industrial
group sizes but with the general scope of operations somewhat limited,
except for appreciable production in SIC 333, nonferrous metals, not
shared by most candidate regions.
Detroit--As in the case of Chicago, the Detroit SMSA
advantageous set of industrial groups without significant
any of the principal polluting industries.
includes an
dominance by
Pittsburgh--As in the case of Birmingham, the industrial facilities
are dominated by SIC 331 and SIC 332. However; sufficient capacities
are noted in most other groups to maintain some balance.
X-20
-------
St. Louis--The SMSA tends to have the best combination of industry
of all candidate areas with no single three-digit group in dominance.
The area has significant petroleum refining, SIC 291; hydraulic cement,
SIC 324, industry exceeded only by Chicago and Birmingham; and has appre-
ciable capacity in both ferrous and nonferrous metals. The chemical group
is also present to a major extent.
Geographical Separation
The extent of candidate SMSA separation as provided in Table X-I is
summarized in Table X-5 for distances between central cities.
Table X-5
SEPARATION OF CENTRAL CITIES OF CANDIDATE SMSAs
(Miles)
500-200 200-100 100-50 50-0
Denver Albany-Schenectady- Buffalo Akron
Houston Troy Columbus Chicago
Kansas City Atlanta Detroit Cincinnati
Memphis Birmingham Flint Cleveland
Minneapolis-St. Paul Grand Rapids Rochester Dallas
Oklahoma City Indianapolis Syracuse Dayton
St. Louis Louisville Toledo Fort Worth
San Antonio Milwaukee Youngstown- Gary-Hammond-
Nashville Warren East Chicago
Pittsburgh
The SMSAs in the 500-200 mile group generally tend to be the most
appropriate for the Regional Study. Interarea pollution should be minimal.
and the larger distances suggest the existence of suitably extensive rural
areas. Similar but less pronounced characteristics likely can be attributed
X-21
-------
to the 200-100 mile group. The SMSAs with central city spacings less
than 100 miles generally appear as undesirable. Interarea pollution ef-
fects may be likely which might force the Regional Study to consider the
two adjoining SMSAs perhaps in complete detail. This requirement could
introduce needless complexity in the Regional Study and severely tax the
available resources of the Study.
The relationship of the radius of circles containing 95% and 50% of
the SMSA population provides a rough normalized measure of the proximity
of the truly rural fringe of the SMSA to the desired centralized emission
sources, or the extent of urban sprawl and the resulting diversity of
emission sources. This relationship might be taken as a simple ratio of
the two radii. As a general rule, the less the ratio the greater the de-
gree of concentration to be expected. Rounded to the nearest whole num-
ber, the most favorable SMSA is the Albany-Schenectady-Troy SMSA with a
ratio of unity. This case, however, is a statistical anomaly because the
SMSA consists of three central cities rather than one. Most of the re-
maining ratios range from three to five with Syracuse at six and Rochester
at an extreme of nine. On this general basis of comparison, therefore,
most candidate regions tend to exhibit the same characteristics.
Sunshine
The percentage of possible sunshine as shown in Table X-I ranges
from a low of 48 in Grand Rapids to a high of 70 in Denver, but the bulk
of the cities are in the high 50% to low 60% range. Since the extent and
severity of photochemical reactions and resultant pollutant levels depend
upon sunshine magnitudes, site selection should favor those areas with
high percentates.
Space Heating
Lastly, the magnitude of space heating as related to the measure of
degree days tends to have a wide range of values. The degree days vary
from a maximum of 7007 in Flint, Michigan to a low of 1278 in Houston
with a somewhat linear rate of increase from low to high. In general,
this factor does not appear of primary importance in site selection but
could become of significance in the selection between two otherwise vir-
tually identical sites.
X-22
-------
Results
Candidate SMSAs having features unsuitable for use as the site of
the Regional Study are as follows.
Akron--The surrounding area tends to be congested, especially since
the Cleveland SMSA is contiguous and its central city 30 miles north of
Arkon. Rural fringe areas may be scarce. The prevailing winds tend to
be southwesterly toward Cleveland and Lake Erie. Emission sources are
overwhelmingly dominated by the rubber industry.
Albany-Schenectady-Troy--The area is attractive because of its iso-
lation--120 miles to the nearest candidate SMSA--and its relatively high
pollution levels. However, it lacks a geographically concentrated set
of emission sources in that it has three central cities. Study in the
area could be correspondingly and needlessly complex.
Buffalo--The prevailing winds tend to be southwesterly, thereby
carrying the pollutants over Lake Ontario and largely eliminating sig-
nificant rural fringe analysis. Winter climatic conditions in Buffalo
are severe and could eliminate significant experimental efforts for about
40% of the year.
Chicago--The Chicago area is of considerable interest but must be
eliminated for two principal reasons. First, its sheer size is a major
disadvantage. Since Chicago is contiguous with the Gary-Hammond-East
Chicago SMSA and almost indistinguishable from it, the Regional Study
almost certainly would have to treat the entire complex. Such an area
would probably strain the resources available for the Regional Study.
Second and even more important, the prevailing winds in the area are
south to southwesterly during the greater part of the year; this moves
pollutants over Lake Michigan, largely eliminating the possibility of
the rural fringe analysis.
Dayton--The Dayton SMSA lies in close proximity to other SMSAs and
especially the significantly larger Cincinnati SMSA. The industrial mix
in the Dayton SMSA is poor as it lacks a wide variety of representative
installations.
X-23
-------
Detroit--Detroit suffers from its sheer size and its nearness to
other large and somewhat heavily industrialized SMSAs in both Michigan
and Ohio. Interarea pollution would be likely and rural fringe areas
may be limited. Additionally, the prevailing winds are to the east over
Canada. Canadian involvement, while perhaps possible, would introduce
an unnecessary complication in the Regional Study.
Indianapolis--The area has a limited industrial inventory.
industry is represented to only a slight degree.
Heavy
Louisville--As in the case of Indianapolis, the Louisville area lacks
the spectrum of emission categories displayed by other candidate areas.
New York City--East coast areas, such as New York City, were elimi-
nated because of a lack of individual industry and the predominant flow
of pollutants over the ocean.
Los Angeles--California and Arizona locations were eliminated be-
cause of a general lack of coal combustion and sulfur emissions along
with the dominance of automobile emissions over other sources.
Agriculture
The Regional Study should, if at all possible, be carried forth in
a region having significant crop lands in the rural fringe. The extent
to which a candidate area is isolated from other areas is suggestive of
the presence of crop lands. Obviously, however, the nonexistence of
SMSAs or similar land uses does not immediately imply usage to be allo-
cated to agriculture. The terrain, soil conditions, climate, and a host
of other factors can be such that the land is utilized for a wide range
of other purposes. To portray the urban-rural relationship in terms of
crop land usage, the chief candidate SMSAs are described in Table X-6
by county with respect to the total area used as farms and the percent
of the total county land area in farms.
Counties of the Birmingham SMSA show a relatively small area devoted
to agriculture although much of the area is somewhat rural in character.
Much of the area consists of wooded rolling hills not particularly suit-
able for intensive agricultural activity, and thereby causes the area to
be somewhat deficient for study of the interaction between urban pollu-
tion sources and rural crop lands.
X-24
-------
Table X-6
AGRICULTURAL LAND USE IN CANDIDATE SMSA COUNTIES
SMSA
Birmingham
Cincinnati
Pittsburgh
St. Louis
County and State
Jefferson, Ala.*
Shelby; Ala.
Tuscaloosa, Ala.
Walker; Ala.
Clermont, Ohio
Hamilton, Ohio*
Warren, Ohio
Boone, Kentucky
Campbell, Kentucky
Kenton, Kentucky
Dearborn, Indiana
*
Allegheny, Pa.
Beaver, Pa.
Washington, Pa.
Westmoreland, Pa.
Franklin, Mo.
Jefferson, Mo.
St. Charles, Mo.
St. Louis, Mo.
St. Louis City; Mo.
Madi son, Ill.
St. Claire, Ill.
*
Denotes central city county.
Included in St. Louis County.
t
X-25
Total
Acreage
(thousands)
84
122
235
114
164
53
187
114
53
61
143
55
80
328
239
388
166
255
83
t
346
312
Percent of
County Area
11.9%
23.8
27.3
22.0
55.9
20.1
71. 4
71. 6
55.8
57.7
72.9
11.8
28.4
59.8
36.4
65.0
38.8
71.0
23.2
t
74.1
72.8
-------
The Cincinnati SMSA consisting of seven counties shows comparatively
intensive agricultural activity with six of the counties having crop acre-
age in excess of 50% and three in excess of 7070. Principal crops include
corn for grain and for silage, and moderate plantings of clover, alfalfa,
and soybeans. Some fruits, nuts, and vegetables are also grown. Thus,
the Cincinnati SMSA appears to be quite suitable for the Regional Study
for purposes of the analysis of urban pollutants on rural agriculture.
Pittsburgh, as in the case of Birmingham, presents a comparative
lack of crop land with only one county of the SMSA supporting crop pro-
duction in excess of 50% of the area. Several factors appear to contrib-
ute to the lack of agricultural activity, including the topography of the
area which is characterized by flat-topped hills and steep-sided-stream
valleys and the significant areas that have been given over to strip mines.
Consequently; the Pittsburgh area does not appear altogether advantageous
for the study of pollution-crop interactions.
The St. Louis SMSA, consisting of six counties plus St. Louis City,
has four counties with land areas of 65% or more given to crop production.
The principal crops include corn for grain and for silage, soy beans, and
alfalfa. Additionally. considerable wheat harvest is included in the
field crops. Since wheat is one of the principal food crops of the nation,
its existence in the St. Louis area is regarded as particularly notable.
A modest harvest of fruits, nuts, and berries originates in the SMSA. In
light of these factors, the St. Louis SMSA appears to offer definite ad-
vantages over most areas in the analysis of the pollution impact on crops.
X-26
-------
Chapter XI
NEAR-SURFACE ATMOSPHERIC RESEARCH FACILITY
General Considerations
Data gathering on both a routine and a special basis will be a
fundamental activity carried out within the program of the Regional Air
Pollution Study. Detailed information will be derived on both the
meteorological and air quality conditions of the atmosphere throughout
the region. Facilities will be provided so that these input data can
be obtained both at ground level and at least to heights comparable with
the surface mixing layer.
This discussion will consider the types of observations to be
obtained for the Regional Study program. The design of the stationary
regional air quality-meteorological monitoring network must consider
three major factors: (1) the horizontal extent of the network, (2) sta-
tion density and location, and (3) station characteristics.
Horizontal Extent of the Network
The horizontal extent of the monitoring network should be dictated
by the physics of atmospheric transport and diffusion, and transformation
processes in conjunction with the nature and location of emission sources.
Ideally, the network should encompass the major center (or centers) of
pollutant emissions such that the level of pollutant concentrations at
the outer edge of the network is a very small fraction of the maximum
concentrations that occur in the immediate vicinity of the emission
sources.
Table XI-l quantifies these considerations by presenting the ratio
of theoretical (Gaussian model) values of predicted surface concentra-
tions of an inert pollutant at a given range to the maximum ground level
concentration from a continuous point source at each of three heights
under the six Pasquill stability categories. Wherever possible, the
distance is given where the ratio is reduced to 0.01; however, when this
criterion cannot be satisfied, the ratio is given at the 100-km range.
Table XI-l also lists the distance from the source to the point at the
XI-l
-------
Table XI-l
VARIATION OF THE CONCENTRATION RATIO AND THE RANGE OF THE MAXIMUM SURFACE CONCENTRATION
FOR A CONTINUOUS POINT SOURCE (FROM DATA PRESENTED IN SLADE, 1968)
><
t-I
I
tV
Source Height
10 m 30 m 100 m
Concn Max Concn Concn Max Concn Concn Max Concn
Ratio Range Range Ratio Range Range Ratio Range Range
Stabili ty ( percent) * (km) (km) ( percent) (km) (km) (percent) (km) (km)
S
A - extremely unstable < 1 1 < 0.1 1 1 0.2 1 2 0.4
B - moderately unstable 1 1 < 0.1 1 2 0.2 1 5 0.7
C - slightl y unstable 1 2 < 0.1 1 6 0.3 1 28 1.1
D - neutral 1 4 0.2 1 20 0.6 1.8 100 2.7
E - slightly stable 1 9 0.3 1 43 0.9 8 100 4.5
F - moderately stable 1 21 0.5 1.7 100 1.9 36 100 16
*
The ratio of the surface concentration at a given range to the maximum surface concentration; the ratio is given at the
range where it first equals 1% or at 100 km.
-------
ground where the concentration reaches a maximum. For all cases except
two, ground level concentrations at 100 km are reduced to at least 1 per-
cent of the maximum surface value. The two exceptions occur for the tall
stack case (lOO-meter height) under slightly and moderately stable atmo-
spheric conditions. Under these conditions, the concentration ratios are
8 and 36 percent, respectively.
With these guidelines in hand, it seems reasonable, as a first-order
approximation, to consider a nominal radius for the basic air monitoring
network (Class A stations) of the order of 100 km from the centroid of
the major emission sources. Figures XI-I, 2, and 3 are taken from the
1968 Report for Consultation on the Metropolitan St. Louis Interstate
Air Quality Control Region (Missouri-Illinois). These figures give:
(1) the location of major point sources of any single pollutant for the
general region within a 100-km radius of downtown St. Louis, (2) the
particulate emission density for the winter season, and (3) the sulfur
oxide emissions for winter. Examination of these figures will show that,
although the emission sources for the various pollutants are not neces-
sarily coaligned, they are fairly well concentrated within a 35-km radius
of downtown St. Louis at the Gateway Arch. Therefore, the minimum down-
wind distance between emission sources and the outer boundary of a 100 km-
radius circular grid is seen to be on the order of 65 km. When reference
is again made to Table XI-I, it will be seen that the large majority of
stability and source-height cases still satisfy the I-percent criterion
at a range of 65 km. In fact, there are only two additional cases which
do not; one case is again for the very tall stack situation (neutral
stability) and the other case is for moderate stability with a moderate
stack height. It therefore appears that the 100-km radius network is a
reasonable, first-order grid size.
To obtain some idea of the concentration distribution downwind of a
circular area source (35-km radius), some computations were made using a
simple Gaussian dispersion model. The emissions region was represented
by a series of approximately 500 point sources at a uniform height of 10 m.
The distribution of the source strength about the region was represented
in two ways: (1) all sources were assumed to be of similar strength
(Q = 1), and (2) a circular Gaussian distribution of source strength
was projected, where Q at the center was equal to unity and the one sigma
value occurred at 17.5 km. Computations were then made for both source
configurations under two classes of atmospheric stability--slightly stable
and slightly unstable. The results are depicted in Figures XI-4 through
XI-7. In all cases, the sharp decline in concentration immediately down-
wind of the source region is evident. Further downwind, computed concen-
trations with a slightly unstable atmosphere decrease almost logarithmic-
ally, while under slightly stable conditions the decrease is very gradual.
XI-3
-------
.
r--~-
I
~
I
,
I
i
( -----r--
',---f GREENE i i ----:
'\ 2 I . .f-.J-i MACOUPIN i MONTGOMERY ...-.1
. -----1 ~ (f V'-- i i !
. . «. JERSEY I , I
LINCOLN () \. _L'----~'--1
rL, -_.r~')-"~ ~ BONO i
~ L_-f '-, ~f"'-2-" MADISON I . 100 km
r I ST. CHARLES )' \ L J...
WARREN 1- 35 ----...,
. ';' . j ,
."\ ! (. ST. LOUIS L~SJ;' ---- - CLINTON!
-''-J .-./' lorT)' I I
.../'.. I
kr' -(-'0 i' ST. CLAIR I ,....---..........--.r-'I
" IN \ ~ I
. ''-, i WASHINGTON I
/ JEFFERSON" -~L I I
.1 \ MONROE r--- -~-r------
-r--- ". ". j , L
I \" .<.A.<: RANDOLPH ! ~
WASHINGTON i '----- ./ '''. J--- --~
I <. STE GENEVIEVE >./"'., ' I
t_-...., L' I, /' - 'I
' ro- '---. I FRANCOIS -" -
L- ~ ..,/' / PERRY"",- I
~----- ~ \ I
----'--._, ',-----
-,I
/.,./'...' \
(' ,
,
\
\
l-
PIKE
FRANKLIN
SCALE ~ km
~
N
I
EMISSION RATE ~ TONS/DAY
I~"""I I I. I I
o 10 20 30 40 50
. 10 100
.0> 100
of any single
pollutant
SA-1365-33
FIGURE XI-1
MAJOR POINT SOURCES
XI-4
-------
I""""" I I I I
o 10 20 30 40 50
/'....
./. \, r-----r---
( '\----1 GREENE I' I',
\ ----:
\ '\ z t' - --r1 MACOUPIN i MONTGOMERY -.I
l- j------l ~ r V-__r I' I' .
. ' «. JERSEY I
! LINCOLN (0 \. _L_____~---(
rLL_-f--.rk.,'-../, ~ BONO i
--l ' .
. - WARREN i ST. CHARLES ,/ L___...1._.....,
l-.-..I '" i (-' --.-J i
'--' . ST. LOUIS I CLINTON '
i ~t./-/ , i r'-' I
I'~. iJ-~''--- --,
FRANKLIN! "35 r WASHINGTON I
/ ' '\.-"\ I I
. JEFFERSON l MONROE L I
r' ---r-----\ . ;-- --L-y------- ,
~ i '\ ~.---.J . i PERRY ~
! . ~ ---.< -,- RANDOLPH i )
I CRAWFORO I WASHINGTON 1-./ "'-. f-- '--....0,
ii' <, STE GENEVIEVE .>/",y........' 100 i
i.--- L' I '\, / JACKSON I'
, . 'FRANCOIS". -
L..J- ---- ". / PERRY "'-. I
~-----~ \ I
L-.---....-...-- .~._._-
~ PARTICULATE DENSITY - (::~aY)/Mi2
J D <0.01
N ~ 0.01 .099
I =:',
SCALE - km
SA-1365-34
FIGURE XI-2
PARTICULATE EMISSION DENSITY BY STUDY AREA ZONE, WINTER AVERAGE
XI-5
-------
/'~\
(. .
.
\
\
l-
,,'----~ GREENE i i
'\ z i' .-l-i MACOUPIN i
-----...( ~ (.\,._.J, .
. ~. JERSEY I !
LINCOLN I U , .1-.-.-.
r-LL.-{._.I'~',->
r ~ " .
II WARREN. ~
i '", !
i '-'''-J
,
PIKE
~._---,
,
MONTGOMERY -.I
.
FRANKLIN
r
,
BOND I
L___~--,
, .
I
I
,r-.----r-'i
I
CLINTON
1"""'''1 I I I I
o 10 20 30 40 50
35
" WASHINGTON
-~ '
r-' ,r-_L_--L_,-__- -~
~ \ ~.,,,-.J i, ~
! n --'< ,. RANDOLPH , )f
I '1-- /. ~
CRAWFORD WASHINGTON 'J;-" ._-~
i 'I <, STE GENEVIEVE .>./"')<,,' 100 i
t - -....., L' '\ / JAC KSON 'I
'L. 1"'" .----' FRANCOIS ',. -,
~ '\,.,/" / PERRY '-. ,
~--~"-- L-._-_.--._-~~.____I
SOX DENSITY ~ (Ton/DayI/M?
D < 0.01
NI ~ 0.01 - 0.1
~ 0.1 1
.>1
SCALE ~ km
SA-1365-35
FIGURE XI-3
SULFUR OXIDES EMISSION DENSITY BY STUDY AREA ZONE, WINTER AVERAGE
XI-6
-------
RELATIVE SURFACE CONCENTRATIONS (Xou/Q)
UNIFORM AREA SOURCE AT 10m HEIGHT
SLIGHTLY UNSTABLE ATMOSPHERE
---------------------
---------
--
--
-...
.05 ...
-...
......
.....
......
....
......
....
....
...
...
..
...
...
...
...
...
...
...
,
...
...
...
:><
H
I
....:J
o
35
FIGURE XI-4
0.2
\
0.1
NORMALIZED VALUES OF SURFACE CONCENTRATION
COMPUTED WITH A GAUSSIAN DISPERSION MODEL
160 km
SA-1365-36
-------
RELATIVE SURFACE CONCENTRATIONS (Xou/Q)
GAUSSIAN AREA SOURCE AT 10 m HEIGHT
SLIGHTLY UNSTABLE ATMOSPHERE
-~:-------------------------------
---.025 "---
-- -- - ------...
--
......
"''''''''
,
...
...
...
,
,
,
,
,
,
,
\
\
\
>::
H
I
00
o
35
FIGURE XI-5
0.1
.05
NORMALIZED VALUES OF SURFACE CONCENTRATION
COMPUTED WITH A GAUSSIAN DISPERSION MODEL
160km
SA -1 365-37
"
-------
:x:
H
I
CD
o
35
FIGURE XI-6
3.0
\
0.5
1.0
2.0
RELATIVE SURFACE CONCENTRATIONS (Xou/Q)
UNIFORM AREA SOURCE AT 10 m HEIGHT
SLIGHTLY STABLE ATMOSPHERE
160 km
SA-1365-38
NORMALIZED VALUES OF SURFACE CONCENTRATION
COMPUTED WITH A GAUSSIAN DISPERSION MODEL
-------
X
H
I
.....
o
o
..... 1.0
" ......
" ..
" ~ .. .. ... 1 .5
~~ ~ 2.0 """
"3\ \ '\
35
RELATIVE SURFACE CONCENTRATIONS r'ou/Q)
GAUSSIAN AREA SOURCE AT 10m HEIGHT
SLIGHTLY STABLE ATMOSPHERE
0.1
0.2
160 km
SA-1365-39
FIGURE XI-7
NORMALIZED VALUES AT SURFACE CONCENTRATION
COMPUTED WITH A GAUSSIAN DISPERSION MODEL
-------
This simple illustration points out the complexities introduced by atmo-
spheric variations in locating fixed monitoring stations. A reasonable
compromise might be the selection of the slightly stable atmospheric
condition for establishing station spacing criteria due to the prime
interest in periods of poor air quality. Of course, the concept of a
concentrated circular area source is highly idealized and additional
practical considerations must also be evaluated.
Although a 100-km range defines the nominally satisfactory horizontal
scale (radius) for the surface network under most cases, it may be
desirable to extend the range for the study of transport and diffusion
from very tall source heights under stable conditions. The maximum
extent of the surface network will, of course, be dictated by practical
considerations rather than theoretical. The terrain in the St. Louis
area is relatively homogeneous for hundreds of kilometers. The major
restriction in extending the network is the location of cities and towns
with their own emission sources. It will be impossible to track St.
Louis pollutants with near-surface measurements in the vicinity of these
sources, and the airborne tracking of elevated pollutants will be made
more difficult.
Table XI-2 presents a survey of the population and locations of the
major cities and towns within a 100-km radius of the Gateway Arch. As
a working hypothesis, it has been assumed that locations of emissions
sources and population centers coincide. With the exception of four
relatively small towns, all population centers are within 50 km of the
Arch. It is particularly noteworthy that the area beyond 35 to 50 km
is devoid of significant, secondary emission cores. Table XI-3 lists
the principal cities and towns between 100 and approximately 200 km of
the Arch; also given are the range, bearing, and population. It will be
noted that for the important S-SE windflow cases (with their relatively
low ventilation factors and subsequent high pollution potential), Hannibal,
Quincy, and Jacksonville lie directly downwind and between 16 and 67 km
of a 100-km grid. Under NW flow, the industrial area of Cape Girardeau,
Carbondale, Murphysboro, Marion, Herrin, and West Frankfort is 13 to 54
km from the grid perimeter. And for SW flow, Springfield is within 29 km
and Decator 60 km.
In summary, we may conclude that the major emissions sources are
concentrated within a 35-km radius of the Arch and that a monitoring net-
work (Class A stations) of 100 km-radius from the Arch is satisfactory
under most conditions. However, for stable atmospheric conditions and
large source heights, attention should be directed toward possible exten-
sion of the network for, at least, those wind directions corresponding to
low ventilation factors (stable conditions).
XI-II
-------
Table XI-2
MAJOR CITIES AND TOWNS WITHIN 100 km OF DOWNTOWN
ST. LOUIS; POPULATION GIVEN IN THOUSANDS
Illinois
Alton
Bell evi 11 e
Cahokia
Centreville
* Centralia
* Chester
46
40
19
14
14
4
Missouri
Berkeley
Clayton
Ferguson
Flori ssant
Jennings
Kirkwood
Maplewood
5
16
22
38
15
29
13
*
Collinsville
East St. Louis
Edwardsville
Granite City
* Litchfield
* Vandalia
Overland
Pine Lawn
Richmond Heights
St. Charles
St. Louis
Wellston
Located within 50 to 100 km of downtown St. Louis;
others are within 50 km.
XI-12
15
82
12
42
7
5
11
6
15
21
750
9
-------
Table XI-3
MAJOR AREA CITIES AND TOWNS BEYOND 100 km OF OOWNTOWN ST. LOUIS
Population Bearing* Range
City, State (thousands) (ReI) (km)
Benton, Ill. 7
Cape Girardeau, Mo. 25 SSE 54
Carbondale, Ill. 19 \
SE x S 13
Murphysboro, Ill. 9
tColumbia, Mo. 37 WNW 77
Decatur, Ill. 85 NE 60
tFulton, Mo. 11
Hannibal, Mo. 20 NNW 48
Jacksonville, Ill. 22 N 16
tJefferson City, Mo. 28 W x S 64
Marion, Ill. 13 I
Herrin, Ill. 9 SE 29
W. Frankfort, Ill. 9
tMexico, Mo. 13
Mt. Vernon, Ill. 16
Quincy. Ill. 45 NW 67
Salem, Ill. 6
Springfield, Ill. 87 NE 29
*
Bearings are given for major, nearby cities relative to downtown
St. Louis; range is that beyond 100 km; population is given in
thousand s.
Cities outside of proposed network orientation.
t
XI-13
-------
Network Orientation
It will be advantageous to consider possible deviations from a
uniform circular network design and, also, the limitations which may be
imposed on potential modifications. The first item to consider is whether
the grid should encompass a circular region or be skewed or elongated in
some fashion--perhaps according to the nature of the prevailing winds.
Figure XI-8 is a series of four monthly surface wind roses (for all
hours of the day) recorded at Lambert Airport in St. Louis for the years
1951 through 1960. It will be seen that the distribution is not sym-
metrical but shows a preference for southerly and northwesterly winds.
The distribution is sufficiently skewed that a modified grid design
based on climatological data may more adequately monitor dispersion
processes in the St. Louis area.
To depict in greater detail the annual distribution of surface wind
direction, Figure XI-8 (a-d) is combined in Figure XI-9. The prevailing
bidirectional nature of the flow stands out clearly with 37 percent of
all winds from 180 ~ 560 and 24 percent from 315 ~ 340. However, in
biasing the location of monitoring stations according to wind direction,
an effective transport wind representative of conditions throughout the
entire mixing layer must be considered; the surface wind may not be a
good indicator--especially under stable conditions (of prime interest to
the study) when significant veering occurs near the surface.
Figures XI-IO and -11 illustrate the frequency of occurrence of
winds at 350 m at Lambert Field for four months (January, April, July,
and October) over the period 1951-59 during daytime and nighttime periods,
respectively. The prevailing winds at this height are more regularly
concentrated in a sector from about 1800 west through 3150. A more
meaningful approximation of the effective transport wind direction can
be obtained by considering the direction at some height above the surface
in order to describe realistically the manner in which a diffusing
"cloud"--ranging in height up to perhaps 1500 m--is transported, as well
as to minimize possible directional anomalies in the microscale (surface)
observations.
Meteorological data of the type required for a more objective
assessment of potential air pollution conditions are now being obtained
in St. Louis at the Environmental Meteorological Sounding Unit (EMSU)
of the National Weather Service. The EMSU station collects low-level,
slow-ascent radiosonde data during the early morning and midday periods
at the Gateway Arch. The objective of the program is to provide high
resolution data in the lower atmosphere (surface to approximately 3000 m)
XI-14
-------
10
10
Ie) JULY
~
N
I
(d) OCTOBER
SPEED CLASSES
CA~ I-
':::/0-3 4-78-1213-181~ 19
mph
CA~O 2 4 6 8 10
J OCCURRENCE, %
FIGURE XI-8
SA-1365-40
SURFACE WIND ROSES FOR LAMBERT FIELD, ST. LOUIS 1951-1960
XI-15
-------
~
N
I
SA-1365-41
FIGURE XI-9
COMBINED WIND ROSES: JANUARY, APRIL, JULY, OCTOBER, LAMBERT
FIELD, ST. LOUIS, SURFACE OBSERVATIONS, 1951-1960
XI-16
-------
~
N
I
9%
SA-1365-42
FIGURE XI-10
WIND ROSE - 1,100 FT, LAMBERT FIELD, ST. LOUIS, DAYTIME
OBSERVATIONS, 1951-1959
XI-17
-------
~
N
I
6
9%
SA-1365-43
FIGURE XI-11
WIND ROSE - 1,100 FT, LAMBERT FIELD, ST. LOUIS, NIGHTTIME
OBSERVATIONS, 1951-1959
XI-IS
-------
for the evalua~ion of the ventilation or dilution capability of the
boundar~ layer. Unfortunately, the tracking is done by single theodolite
and the accuracy of the results is sometimes questionable, as discussed
in the section (Chapter III) on upper air sounding systems. (In fact,
it is this limitation that has prompted the recommendation of a system
such as METRAC to supplement as well as replace the EMSU program.) In
spite of this problem, the data are more useful than any routinely avail-
able. The St. Louis EMSU has been in operation less than three years--
too short a period for strict climatological averages, yet long enough to
provide valuable insight into the problems of assessment of the trans-
port wind and urban mixing depth.
Mr. Donald Wuerch, who is in charge of the St. Louis operation, has
kindly provided data for a two-year period on the distribution of venti-
lation as a function of the transport wind direction and time of day.
The transport wind is defined as the mean vector wind over the depth of
the mixing layer, and the ventilation (or dilution) factor is the product
of the transport wind speed and the mixing layer depth. These data are
presented in Figure XI-12 as a ventilation wind rose for both observa-
tional periods and are stratified (in azimuth) by the transport wind
direction (100 increments). The length of the radial spurs indicates
the frequency of occurrence (as a percentage); the value at the tip of
each spur is the ventilation factor (sq m/sec) for each wind direction
interval. Figures XI-13 and -14 show the same information for the morning
and noon periods, respectively, but over 300 intervals. The data indi-
cate that for the morning period, 37 percent of the winds are from the
quadrant centered on 210° and the associated ventilation factors are
extremely small (on the order of 2.1 x 103 m2/sec). During the mid-
day period, low ventilation factors (~.5 x 103 m2/sec) are associated
with winds from 120°, but these occur relatively infrequently (14 percent)
On the other hand, the 210° quadrant shows moderately high midday venti-
lation (9.2 x 103 m2/sec) occurring quite often (36 percent).
The orientation of stations in the air monitoring network should,
in part, reflect the asymmetry in the mean annual distribution of wind
direction in the St. Louis area. This arrangement will restrict the area
in which the stations will be deployed, thereby actively increasing the
station density. The network may be broken into three primary segments:
(1)
A central core where most of the emissions sources and popula-
tion centers are located. Station spacing will be relatively
dense and--in the absence of a systematic pattern of source
locations--uniform. Preliminary results indicate this score
will be contained with a 35-km radius of the Arch.
XI-19
-------
~
N
I
10,200
VENTILATION (m2/s) TRANSPORT
WIND SPEED x MIXING DEPTH -4,680
SA-1365-44
FIGURE XI-12
VENTILATION WIND ROSE ST. LOUIS EMSU DATA,
MAY 1969 APRIL 1971
XT-20
-------
~
N
I
----
NO. OF OCCURRENCES*
- VENTILATION FACTOR (103 m2;s)t
SA-2465-45
FIGURE XI-13
VENTILATION WIND ROSE, ST. LOUIS EMSU DATA, MORNING SOUNDINGS,
MAY, 1969-APRIL, 1971
XI-21
-------
I
N
I
I
I
I
I
I
I
8 -----+--40
I
10----1-50
I
12~60
I
NUMBER OF OCCURRENCES*
VENTILATION FACTOR (103 m2/s)t
SA-1365-46
FIGURE XI-14
VENTILATION WIND ROSE, ST. LOUIS EMSU DATA, MIDDAY SOUNDINGS,
MAY 1969-APRIL 1971
XI-22
-------
(2)
An outer perimeter encircling the inner core and extending
to a radius of about 100 km. Azimuthal station spacing will
be uniform while radial spacing should reflect the expected
concentration distribution. Station density will be less than
that of the inner core.
(3 )
An augmented sector within the outer perimeter. Based on the
need for more detailed measurements under poor ventilation
conditions, the northeast quadrant should be supplemented with
additional stations. Although it would be desirable to expand
the range of measurements in this direction, the locations of
the cities of Springfield, Jacksonville, Hannibal, and Decatur
limit the maximum useful range to about 125 km.
Characteristics of Stations
It will be noted from the discussions on station design that three
distinct classes of station are recommended: Class A--permanent station
for the routine acquisition of meteorological and air quality profile
data; Class B--semipermanent station for air quality and meteorological
data at a single level; Class C--transportable, trailer station to be
used for special studies of meteorology and/or air quality. This section
provides a reasonable outline of the number of stations that can be
anticipated as well as their deployment. Specific details of recommended
meteorological instrumentation will follow in the next section.
In locating a fixed-point (or profile) monitoring station, it is
hoped that the conditions measured at that location will be representative
of the average conditions over a larger area. Upon acceptance of this
premise, inverse logic dictates that the station density must be greatest
in regions where air quality and/or meteorological conditions are expected
to vary most rapidly in space, and station spacing may safely be increased
in "flat regions." Figures XI-4 through -7 graphically illustrate how,
in principle, these concepts may be applied to the St. Louis region.
Furthermore, the limitations imposed by emissions cores surrounding the
St. Louis metropolitan area restrict the extent of the study region to
something on the order of a 100-km radius from downtown St. Louis. In
the specification of the total number of stations that will be required
to monitor adequately atmospheric conditions over the region, cost will
be a' constraining factor. However, the network should comprise the
minimum number of stations required to monitor atmospheric conditions
on a scale compatible with the evaluation of the more sophisticated air
pollution models. From a practical standpoint, the choice cannot be made
XI-23
-------
with the fixed stations alone as the number required would be astronom-
ical (~3 x 104). A reasonable compromise is the judicious placement of
a moderate number of fixed stations to monitor the larger-scale features
routinely, while utilizing the transportable station network to collect
data on the fine scale as it may be required. Many of these requirements
have been summarized in Chapter III.
The spacing that is proposed between the Class A stations is on the
order of 40 km or an order of magnitude less than the synoptic scale
network over the contiguous United States. Therefore, a series of nine
of these stations is recommended for deployment in a box pattern centered
at the Arch, while an additional eight may be uniformly distributed along
the 100-km radius. The mean characteristic spacing among the outer sta-
tions is increased to about 60 km, but the increase is justified in view
of the homogeneous topographic and emissions fields in this outer
perimeter. The proposed A-station distribution is depicted in Figure
XI-15. It will be noted that two types of Class A stations are shown:
types Al and A2. Specific differences are given in the following sec-
tion; the basic or essential difference is that the A2 type collects
information on the structure of the surface layer (i.e., absolute values
and gradients of items such as wind, temperature, humidity, and pollu-
tants) while the Al type additionally obtains data on the forcing param-
eters; namely, the surface energy budget and the mesoscale horizontal
pressure field. In view of the relatively small additional cost of the
Al station, it might be desirable to outfit all A-stations in this manner;
what has been proposed, however, is an adequate minimum.
While the Class A stations are designed to provide information on
lower atmospheric dynamics, thermodynamics, and pollutant fluxes, the
Class B stations will provide higher spatial resolution meteorological
and air quality data for kinematic analyses. The objective is the study
of the response of the atmosphere (in the form of, say, isotach, isotherm,
and pollutant isopleth distributions) to the atmospheric forces and
emissions sources. The Class B stations may be considered to be fixed,
although they have been designed to permit relocation on an infrequent
basis. When reference is again made to Figures XI-4 to -7; it follows
that a significant number of these stations should occupy the region of
maximum emissions. A suggested grid of B-type stations is depicted in
Figure XI-16. Twenty-four stations are located in a uniform, square
array within 35 km of the Arch. Initial tests using the transportable
Class C stations should probably be made first in order to solve first-
order siting problems. An addi tional number of B-type stations (per-
haps eight) should be located in the downwind "poor ventilation" sector--
the NE quadrant--discussed earlier in this chapter for the routine,
XI-24
-------
"Arch" J
40 km
100 km
o - CLASS A2
. CLASS A1
SA-1365-48
FIGURE XI-15
PROPOSED DISTRIBUTION OF CLASS A STATIONS
XI-25
-------
100 km
FIGURE XI-16
~
125 km
~
.
~
.
~
~
}I
"Arch"
o - CLASS A
. - CLASS B
'SA-1366-47
PROPOSED DISTRIBUTION OF CLASS B STATIONS,
SHOWING ALSO THE LOCATION Of CLASS A STATIONS
XI-26
-------
detailed measurement of air quality at medium-to-long ranges under atmo-
spheric conditions that retard dilution. These are also pictured in
Figure XI-16.
The Class C stations provide a wealth of flexibility to both the
meteorological and air quality programs. They will be used in a large
number of specific and special research tasks. They also give the basic
(fixed) system flexibility when they are used to supplement routine
observational network. Such flexibility may be required by unique atmo-
spheric conditions or as a result of gaps in the fixed network. The
potential of these stations lies in their mobility, adaptability (almost
any type or number of observations can be made), and compatibility (the
data format is identical to that of the fixed network). The absolute
number of C stations that will be required will most likely evolve as
the program progresses--an initial estimate would be 16 and a likely
projection 25.
In summary, the following number of stations is recommended:
Class AI' 9; Class A2' 8; Class B, 32; and Class C, 16 to 25.
Meteorological Instrumentation
The meteorological data that are to be obtained under the RAPS pro-
gram must serve to further the description and understanding of atmos-
spheric processes, and in so doing they will specifically be utilized for
input to the various air pollution models as well as for the thorough
evaluation of the meteorological aspects of these models. As in the
evaluation of station number, a reasonable compromise must be found
amongst the requirements of instrumentation, field effort, data proces-
sing, and analysis.
In evaluating the representativeness of measurements made at a fixed
point or points on a vertical tower, it should be kept in mind that each
element of surface structure produces an individual disturbance to the
air flow; however, it is the statistical effect of all such disturbances
that is important for the study of representative profile structure in
micrometeorology. The specific location of a meteorological tower and
the height to which it extends must be governed by this principle. As
a rule of thumb, the lowest level at which the wind should be measured
should be above the tops of the characteristic roughness elements of the
area, while the maximum level will be described by the thickness of the
internal boundary layer that develops upon a change in characteristic
surface roughness. As a rule of thumb, the maximum height of measurement
should be less than 2% of the wind fetch that is characteristic of the
XI-27
-------
area while the minimum height should be five times greater than the
characteristic height of roughness elements. In the latter case, when
reference is made to a forested region, the factor of five will apply to
variations in the canopy height and the measurement height itself will
be defined above the mean maximum height of the canopy.
In consideration of the number of levels required to be instrumented,
the computation of a single measure of mean profile curvature requires
data from three levels. Uncertainties in the representativeness or in
the determination of the parameters at the various heights will increase
the minimum number of levels to four. When vertical gradients and curva-
ture characteristics are to be studied as functions of height, the number
of instrument levels should be increased to five or more. When these
items are considered with reference to the objectives of RAPS, it will be
seen that a minimum number of levels for measurement at each tower will
be three, although four would be preferable. It is imperative that both
the mean and fluctuations of the u and v components be measured, as well
as the fluctuating vertical wind component and possibly also its mean
value. Fifteen-minute averages of the mean and variance of the horizontal
and vertical wind speed and direction (vector average) should be resolved.
Results of micrometeorological studies (such as the 1967 Cooperative Field
Experiment at Davis, California) have indicated the occurrence of mean
vertical motions persisting over periods of several hours near the sur-
face boundary. The effect on the transport and diffusion processes may
therefore be significant and should be examined systematically throughout
the study region. With regard to the measurement of the vertical wind,
it is perhaps unnecessary that the measurements be made at all three
heights--rather a single measurement at the uppermost level may be suf-
ficient. In view of this, estimates for the cost of such instrumentation
will be given in the form of two alternate schemes using instrumentation
manufactured by the R. M. Young Company as typical price standards.
The wind instrumentation will thus provide data on the mean and
variance of the three wind components and also aid in the evaluation of
a near-surface value of the eddy diffusivity; these parameters are of
direct interest to meteorological dispersion models. The aerodynamic
roughness length can also be computed and used to evaluate empirical
algorithms that estimate this parameter. Of course, the wind profile
is also necessary for the computation of the gradient Richardson number.
Additionally, the wind sensors will provide data compatible with commonly
available surface wind observations and will serve to evaluate the
representativeness of single observations (e.g., airport winds) for use
as input in diffusion models (as is the current fashion). Results will
very likely illustrate the extreme variability of measurements at the
lower heights over the region due to the strong influence of surface
XI-28
-------
discontinuities and the surface aerodynamic roughness in the immediate
vicinity of the sensor. Similarly, the upper observations should better
represent the features of the flow on a scale compatible with the hori-
zontal distance between stations.
An effect of urban structures will be to decrease the horizontal
length scale over which a wind measurement may be considered represen-
tative. Similarly, there may also be a need to increase the height to
which measurements are made in these built-up areas. Such need may arise
if the aerodynamic roughness and zero plane displacement at the station
are of the same order of magnitude as the height of the observation.
This condition most likely exists in the urban core, and the basic tower
network may need to be supplemented by one or several tall towers, in
addition to the balloon observations.
Measurements of air temperature should be made at least at the same
three heights as the wind measurements and preferably at four levels,
with the addition of a near-surface temperature measurement. The tem-
perature sensors must be housed in a temperature radiation shield to pre-
vent measurement errors resulting from the absorption of short-wave ra-
diation with subsequent element heating. The sampling period should be
the same as that for the winds (15 minutes). Estimates of variance do
not seem essential and the sampling frequency may be reduced. The tem-
perature data are to be used for the computation of the gradient Richard-
son number (stability index) and the atmospheric sensible heat flux, as
well as in the mapping of horizontal temperature gradients for the study
of such phenomena as the urban heat island and the thermal wind.
Humidity measurements should be made at the same levels as tempera-
ture and at a similar frequency. Moisture data are to be used in the
computation of the flux of latent heat, in the study of urban effects on
the moisture budget, and for the study of aerosol formation.
Other measurements that need to be made for the study of the surface
energy budget as well as other mesoscale meteorological features include:
(1) solar radiation, (2) terrestrial radiation, (3) pressure, and (4)
precipitation. It may also be desirable to make subsurface temperature
profile measurements for the computation of the subsurface sensible heat
flux. However, in view of lateral surface and subsurface variations,
representative sites may not be readily obtained. These measurements may
therefore be made on a special-task basis; routine estimates of the soil
heat flux can alternatively be made from an energy budget technique by
analyzing the surface temperature response to the solar forcing function.
The types of instruments that are recommended and the class of sta-
tion where they will be deployed are summarized in the following discussion,
XI-29
-------
Trade names are indicated only for costing purposes and no specific
recommendations are intended; prices are approximate values.
Wind System - Option A
Gill UVW anemometer (R. M. Young,* Model 27003) . .
Indicator-translator (Model 27302)
. . . . . . . .
Spare propellers (Model 27108)
. . . . .
. . . . .
Total (Class A--3 levels)
::::
$2880
Total (Class B or C--l level) :::: $960
Wind System - Option B
Gill propeller-vane (R. M. Young, Model 35002)
Spare propeller (Model 21180) . .
. . . . .
. .. . .
. . . . .
Spare tail (Model 35050)
Translator (Model 35402)
Indicator (Model 35602)
. . . . .
. . . .
. . . .
. . . . .
. . . . .
. . . . . .
. . . .
. . . . . . . . .
Gill propeller anemometer (Model 27101)
Spare propeller (Model 27108) . . . .
. . . . . . . . .
. . . .
. . . . . .
Indicator-translator (Model 27170)
. . . .
. . . . . . .
Total (Class A--3 UV, 1 W systems) :::: $2460
Total (Class B or C--l UV system)
::::
740
Wind System - Additional Equipment
Calibration unit (R. M. Young, Model 27230)
Mounting booms (estimated)
. . . . . . . . . .
. . . . .
*
$700.90 each
200.00 each
20.00 each
360.00 each
18.00 each
12.00 each
260.00 each
90.00 each
150.00 each
20.00 each
11 O. 00 each
$100.00 each
50.00 each
R. M. Young Company, 2801 Aero-Park Drive, Traverse City, Michigan 49684
XI-30
-------
Temperature System
Per Level
Gill aspirated radiation shield (R. M. Young)
Rosemount* temperature sensor (Model 104)
$250.00
65.00
Rosemount differential temp bridge (Model 400D)
Rosemount absolute temp bridge (Model 400A)
65.00
. . . . .
90.00
Per System
Rosemount bridge chassis (1 absolute, 3 diff, No. 421-BX)
$360.00
Class A Station Temperature System Costs
Option A
(1) 4 levels:
Shields . . .
Sensors . . . .
Bridges . . .
Chassis . .
Total
Option B
$1000
260
285
360
(2) 3 levels
. . . $750
. . . . . . 195
. . . . . . . .. 220
. . . . . . 360
-
$1905
Total
$1525
Additional equipment:
Rosemount (2501-4-AIO-SC64) digital
temperature indicator. . . .
. . . .
. . . . .
$ 735
Rosemount calibration baths.
. . . .
. . . .
. . . . .
1370-4675
Class B or C Station Temperature Costs
Shield.
Sensor. .
. . . .
. . . . . . . . $250
. . . . . . . .. 65
90
. . . . . . . 360
. . . . . .
Bridge
Chassis
. . . .
Total
$765
*
Rosemount Engineering Company, P.O. Box 35129, Minneapolis, Minnesota
55435.
XI-31
-------
Class A(l) Supplemental Meteorological Instrumentation
1.
Eppley precision spectral pyranometer (2)
0.285 to 2.800 ~
5 mV/ly/min
Impedance 300 [2
Temp compensation ~l%
Time constant 1 see
-20/ -t40oC
2. Pyranometer filters (optional)
GG 14 > 0.500 ~
OG 1 > 0.530 ~
RG 2 > 0.630 ~
RG 8 > 0.700 ~
3.
Barometric pressure transducer
NiSpan C diaphragm
Output ~ 1000 [2 (3 k 0 optional)
Range (1) 27.-31.5 in. Hg
(2) 28.-32 in. hg
Linearity 1: 0.3%
Sensitivity 0.2% (0.12% optional) [0.27-0.16 mb]
Hysteresis ~ 0.2%
Repeatability ~ 0.2%
Temperature operating range - 30/l80oF
10-point calibration
4.
Pressure transducer converter
Optional for item 3
0-10 mV output (0-100 mY, 0-10 V optional)
5.
Mercurial barometer (reference)
Accuracy 0.1 mm Hg (~.14 mb)
6.
Thermal radiometer (2)
Temperature compensation ~l%,
Sensitivity 2.5 mV/ly/min
Response time (95%) 12 sec
-20/+l60oF
*
$990 each (SA)*
$150 each (SA)
$385
( SA)
$160
( SA)
$190
145
(SA)
(WM)t
$850 each (SA)
Science Associates, Inc., 230 Nassau Street, Princeton, New Jersey 08540
WeatherMeasure Corporation, P. O. Box 41257, Sacramento, California 95841
t
XI-32
-------
7.
Dew p6int hygrometer, Cambridge Systems* Model 992 A (Std)
-40 to +800F dew point range
1200 dew point depression
Accuracy ~loF (nominal)
Platinum element accuracy! 0.50F
Maximum response 40F/sec
Flow rate 0.5 to 5 CFH
Pressure range 0 to 20 psia
Instrument operating range 40 to 1200F
Sensor temperature range -60 to 2000F
Mounting 19 -in. rack
Cabinet heater (to 1200F)
Optional:
Platinum resistance thermometer and output:
(a) 100 (2
(b) 0 to 50 mV dc
(c) 0 to 5 V dc
Pressure gage--optional
Fil ters
Switching
Tubing
8.
Tipping bucket rain-snow gage
Sensitivity 0.01 in.
Accuracy 0.5% at 0.5 in./hr
Heat control: thermostat 0 to 350C
Power: 500 watts
Output: mercury switch; discrete event/O.Ol in. precip.
Alternates:
MRI gage (105 watt heater)t
MRI gage (propane heater)
Windshield for precipitation gage, options:
*
Cambridge Systems, EG&G, Incorporated, 151
Massachusetts 02154
Meteorology Research, Incorporated, 464 W.
California 91001
$2995
200
295
490
1090
uncertain
$345
$370
$525
$ 89 (SA)
75 (WM)
85 (WM)
Bear Hill Road, Waltham,
t
Woodbury Road, Altadena,
XI-33
-------
Chapter XII
AIR QUALITY SAMPLING
Introduction
Data gathering on both a routine and a special basis will be a
fundamental activity carried out in the program of the RAPS. The data
will include detailed information on both the meteorological and air
quality conditions of the atmosphere throughout the region. Facilities
will be provided so that these input data can be obtained both at ground
level and at least to heights comparable with the surface mixing layer.
This discussion will consider the types of observations to be ob-
tained for the RAPS. Suggestions will also be made covering network
design in terms of station input capability and station locations. This
discussion will be organized on the basis of three types of surface
stations having generally different purposes within the network and one
upper air measurement facility.
Permanent Network Monitoring Stations for Pollutants
Three types of monitoring stations will be used to measure pollutant
concentrations and meteorological parameters during the course of the
RAPS. The stations, designated Class A, Class B, and Class C, will ful-
fill somewhat different functions in the monitoring network. The Class A
ground level monitoring stations will form the basic network for the
observation of both meteorological and air quality data throughout the
RAPS area. These station sites will be occupied for the duration of the
program and the observed data will be continually transmitted to a central
control facility for monitoring and subsequent processing. For the pur-
pose of initial estimates, a total of 17 locations is suggested for these
Class A stations. The Class A stations associated with a 30-meter
meteorological tower will obtain measurements for both meteorological
and air quality information at three levels above the ground.
Routine air quality sampling at heights of 3, 10, and 30 meters
is suggested. The three-meter height is low enough so that breathing
zone concentrations and other essentially ground level measurements can
be observed, while at the same time, this height is high enough so that
XII-l
-------
it can be protected from local extraneous activities. The 10-meter
sampling height probably represents most nearly a standard sampling
height since, except in locations having multistory buildings or tall
trees, a 10-meter sampling height will be well exposed to the low level
atmosphere. The 10-meter sampling height also conforms to the generally
recommended sampling height for ground level meteorological wind measure-
ments and thus should provide a useful standard height for comparing RAPS
meteorological measurements with standard synoptic weather observations.
From a practical standpoint, a 10-meter sampling height can be obtained
for both meteorological and air quality sensors through the use of rather
simple structures or portable towers and thus this height could be readily
adapted to less permanent stations. These Class A stations should also
provide measurements from at least one more upper level and here a height
of 30 meters is suggested. This 30-meter height is high enough so that,
except in downtown areas, it would be taller than the surface roughness
element and would thus provide measurements indicative of conditions in
the general transport wind stream. The 30-meter height would also provide
an indication of the nature of any vertical stratification, especially
that associated with nocturnal inversion conditions. A height of 30
meters can be obtained with commercially available towers that can be
maintained without the employment of special riggers or steeplejack per-
sonnel. If taller towers can be obtained for use at any of the stations
without inordinate costs to the program in terms of either facility or
maintenance, they should also be considered and measurements should be
extended to the height made available by such special facilities. These
special facilities could be one or more of the commercial radio or tele-
vision broadcasting towers presently situated in the St. Louis area, or
they could be special tower installations constructed and loaned to the
program by a military or other government agency.
The Class A stations are subdivided into Class Al and Class A2.
Nine stations, designated Class AI' will be equipped with the full
complement of meteorological instruments. The remaining eight will be
equipped with a limited complement of meteorological instrumentation.
The meteorological instrument complements of both Class Al and Class A2
stations are described in detail in Chapter X. Both Class Al and A2
stations will operate with the full complement of pollutant monitors
described in detail later in this chapter.
Class A stations will provide the skeleton of the monitoring network.
The 17 Class A stations will be located in a geometrical pattern encom-
passing the St. Louis city center, with the outermost stations located
on a 100-km radius from the arch. The locations of the Class A stations
and the rationale for their distribution through the area of interest
are described in detail in Chapter XI, Meteorological Network and Station
XII-2
-------
Location. Each Class A station site will require a land area of appro-
ximately 1/4 acre to accommodate tower installation, the instrument
shelter, compressed gas cylinder storage area, and access for the
calibration van. Class A stations located in the city center will require
self-supporting towers or access to existing structures of a suitable
height for vertical profile measurements. Class A station sites should
include perimeter fences of chain link or wire mesh to reduce equipment
replacement costs, downtime due to vandalism, and insurance liability.
The instrument shelters at Class A stations must be large enough to
house the full complement of pollutant monitors, the data acquisition
system, and the local secondary calibration sources with their associated
controls. In addition, bench space must be available for installation
and operation of additional monitoring instruments required by research
programs during the five-year RAPS. The recommended instrument shelter
facility will consist of a 15-ft by 24-ft prefabricated insulated build-
ing on a concrete slab foundation of 20 ft by 24 ft. The 8-ft by 24-ft
apron of the slab floor adjacent to the instrument shelter will be covered
to provide shade for compressed gas cylinder storage. Active compressed
gas cylinders for calibration, zero gas, and reagents will be plumbed
through the wall of the shelter. Three sample inlet lines of TFE Teflon
of approximately one in. I.D. enclosed in a protective armor will origi-
nate at the 30-meter, 10-meter, and three-meter levels of the meteor-
ological tower and terminate in a manifold within the instrument shelter.
A low pressure blower will exhaust the air from the manifolds and keep
a constant flow of air through the sample lines. The constant air flow
through the sample lines permits equilibration of the atmospheric con-
stituents with sample line surfaces and will reduce the apparent dead
volume of the lines. The pollutant monitors have access to sample air
from all three tower levels and access to calibration gases. Solenoid-
actuated valves would permit programmed access to the sample lines or
continuous sampling from a single level sample line. The sample line
manifolds will have a common vent to the exhaust blower. Access to
calibration gases would be initiated by the calibration controller on a
programmed schedule.
At each Class A station a Master Station Controller will control the
overall operation of the station. The Master Station Controller will
initiate calibration via a calibration controller on a programmed schedule,
and will sequentially actuate a solenoid valved system to complete a two-
point calibration with zero and span gas. The Master Station Controller
will switch pollutant monitors to sample sequentially from all three
sample levels. In addition, it will monitor parameters that are critical
to operation of the station, such as sample flow rates (where appropriate),
gaseous reagent flow rates, temperature of permeation tube baths, and
XII-3
-------
others. A warning system would be employed to initiate a flag to the
central computer facility when these parameters did not lie within preset
limits.
Data acquisition equipment required for temporary storage of data
and read-out of accumulated data upon interrogation by the central com-
puter facility would complete the instrumentation complement. At least
one telephone line at each station will be required for data transmission
to the central computer facility. The data acquisition equipment is
described in detail in the appropriate section of this report.
The instrument shelters must be provided with electrical heating of
about 40,000 Btu, and refrigeration cooling of 35,000 Btu capacity. A
standard 100-amp residence-type inlet electrical terminal will provide
adequate utility power for operation of the Class A shelter. No connec-
tion for water or sewer will be required.
Semipermanent Monitoring Station
It is recommended that 32 Class B stations be deployed in the monitor-
ing network. The Class B stations are usually similar in design and func-
tion to the Class A stations except that there will be only one sampling
inlet at a height of 10 meters, and wind direction and wind speed will
be recorded by conventional anemometers instead of the three directional
u, v, w techniques. Since sample access from one inlet height is proposed
for Class B stations, the instrumentation inlet manifold will be simpler
than that of the Class A stations with no switching required between
sampling inlets.
A simple 10-meter TV antenna tower could be substituted for the com-
plex and expensive sampling tower used for most of the Class A stations.
The instrument complement for the Class B stations would be similar to that
of Class A stations, but the rapid response total sulfur monitor and non-
dispersive infrared CO monitor would be eliminated. The rapid response
analyzers are recommended for the Class A stations to provide monitoring
of CO and total sulfur concentrations at all three levels within a five-
minute interval. The Class B stations are subdivided into Class BI and
Class B2' The Class BI stations are designated as fixed stations and use
the same instrument building and foundation as that recommended for
Class A stations. It is recommended that a total of 24 Class BI stations
could be used in the monitoring network. The Class B2 stations will
utilize an 8-ft X 20-ft shell house trailer to house the instruments.
The Class B2 trailer shelters would not require the slab foundation. A
total of eight Class B2 stations is recommended. Elimination of the
XII-4
-------
meteorological tower would reduce the land area requirements of the
Class B site to a plot of 50 ft by 50 ft. Meteorological instrumentation
employed at the Class B stations would measure wind speed and direction
rather than the wind vector component sensors of the Class A stations.
The data-acquisition system for the class B stations would be similar in
all respects to the Class A stations without meteorological towers.
Again, a function controller would be required to calibrate air quality
instruments at selected intervals and to monitor critical instrument
parameters and calibration temperatures. The Class B stations should be
perimeter fenced with agricultural mesh fencing where feasible. The semi-
permanent Class B stations would be considered transportable rather than
portable.
The Class C, or portable stations, would be housed in a trailer or
van and would be intended for measurements at a discrete site for six
months or less. Two types of Class C stations are recommended for the
RAPS monitoring network. Class Cl stations would consist of a small
trailer or van with a full meteorological complement and data-acquisition
system including a 30-meter tower. The Class Cl station, though not
portable, will be located and operated at more than one site during the
RAPS program. No air quality instruments would be used in the Class Cl
stations.
Class C2 stations are designed to be easily adapted to the specific
needs of a wide variety of research programs. The permanently installed
equipment in these stations would be limited to the simpler meteorological
instruments, a data-acquisition and local recording system, and a sample
inlet system to monitor at a 10-meter height. A standard sampling mani-
fold will be provided that can be used by a number of air quality sampling
instruments. The bench and/or rack space available for installation of
air quality sampling instruments should accommodate four to five instru-
ments. The instrument mounting technique should be compatible, wherever
possible, with the concept of monitor installation and check-out at the
central facility with the trailer and instrument complement then trans-
ported to the selected site.
Although most sites selected for Class C2 stations will have utility
power available, portable power generating systems should be available
as an optional power source. The trailers designed to use either a
portable power source or utility power should also be equipped with a
well-vented liquid propane gas heating system to reduce the power require-
ments of the portable generator. The power requirements of such a Class
C station, including instruments and air conditioning, could probably be
supplied by a generator of 5- to 8-kW capacity. Prime movers using liquid
propane gas are recommended with full storage capacity to operate the
station continuously for one week.
XII-5
-------
The meteorological sensors installed in each Class C2 station will
include wind direction and wind speed using conventional anemometers and
wind vanes, and possibly atmospheric temperature and relative humidity
sensors.
The data-acquisition system permanently installed in a Class C2
station should include all the basic equipment common to both on-board
recording and direct data linkage to the central control network. Package
assemblies then would complete the conversion to either on-board or remote
recording. Anyon-board recording should, of course, be compatible with
the system used for the rest of the network. Data from on-board recorded
tapes must be readily integrated into the network data records.
Class C2 stations will be located in areas defined by the specific
requirements of the experimental programs in which they will be used.
In some cases, the needs of several research programs can be realized
with one Class C2 station at a single location. Conversely, any program
having more extensive shelter space requirements on a temporary basis
could combine two Class C stations at a single location and use the data-
2
acquisition system of one station for data storage or linkage to the
central control facility.
Class C2 stations should also have perimeter agriculture fencing
enclosing the site where feasible to avoid unauthorized access.
To provide spatial resolution of the meteorological data acquired
from the permanent stations, the 24 Class C2 stations would be deployed
within an area where data augmentation is required. More detailed, spa-
tial resolution of the meteorological data is necessary, for example, for
development and validation of atmospheric models describing areas where
elements of surface roughness distort or modify circulation patterns.
The C2 mobile laboratories are also designed to accommodate a limited
complement of air quality instruments. The deployment of an array of C2
trailers containing one or more air quality monitors would provide valu-
able data to the research programs. These would provide within specific
areas finer resolution of pollutant concentrations with respect to dis-
tance and surface roughness. For example, the reactive chemistry program
could require improved spatial resolution in the studies of the transfor-
mation of sulfur dioxide downwind of urban emission sources. In this
instance, the number of mobile stations that would be outfitted with
sulfur dioxide monitors and the deployment of the stations would be based
largely on the incongruities in the data acquired by the fixed monitoring
network and on special program needs. Collection of pollutants other than
802' such as NO , CO, and possibly ozone and particulate material would
x
XII-6
-------
require deployment of C2 stations within a specific area of interest to
fill in the gaps observed in network concentration rate.
The C2 trailers may also be outfitted and deployed for specific
experiments. Tracer release studies using sulfur hexafluoride would
require outfitting with electron capture gas chromatographs. If dye or
phosphor tracers are released, then filter collection or other particulate
collection schemes would be employed at the individual sites.
It is not possible at this time to predict all pollutants or trace
compounds for which the C2 trailers may be outfitted, except in a general
way. The most likely pollutant candidates would be SO , NO , and CO.
2 x
The same measurement techniques recommended for the network stations would
be recommended for the mobile laboratories. The cost of pollutant moni-
tors for S02, NOx' CO, would be $4000, $5000, and $3000, respectively
per trailer. No changes in the data acquisition system would be required.
The number of trailers required must be determined on the basis of program
needs. Extensive spatial resolution studies could require a minimum of
16 outfitted trailers to a maximum of 24, located at specific sites for
one to two months. Less extensive studies may only require 5 to 10 out-
fitted trailers relocated on a weekly or biweekly basis.
Pollutants To Be Monitored for Establishment of Air Quality
The air quality monitoring equipment that forms the instrument com-
plement of the Class A and Class B stations will measure the following
pollutant concentrations in the atmosphere: carbon monoxide, methane,
nonmethane hydrocarbons, total hydrocarbons, nitric oxide, total nitrogen
oxides, sulfur dioxide, hydrogen sulfide, total sulfur, ozone, and sus-
pended particulate material. National ambient air quality criteria have
been issued for all of these pollutants either as a class of compounds,
such as hydrocarbons, or as a specific entity which must be distinguish-
able or definable from other class members commonly found in aerometric
surveys. The importance of these pollutants and their relationship to
overall air quality has been extensively documented in the technical
literature as well as in the formal Air Quality Criteria issued by the
Environmental Protection Agency and the U.S. Department of Health, Educa-
tion and Welfare. A very brief description of these pollutants, their
origin, and their significance follows.
XII-7
-------
Carbon Monoxide
Total emissions of 00 exceed those of all other pollutants combined.
Incomplete combustion of carbonaceous fuels represents the only signifi-
cant source of CO on a global basis. The major combustion source is the
gasoline engine. A number of geophysical and biological sources are
known; however, their contribution to urban atmospheric concentration
is not considered to be significant. CO can be used as a tracer pollu-
tant for automotive emissions and will provide data on the dilution of
pollutants as a function of source distance.
Methane, Nonmethane Hydrocarbons, and Total Hydrocarbons
Methane occurs as a natural constituent of the unpolluted atmosphere
at an approximate concentration of 1.5 ppm. Although CH4 is a major con-
stituent of automobile exhaust and other combustion gases, its nonreactive
role in atmospheric chemistry makes measurement of CH4 concentration in
the atmosphere of secondary concern. The measurement of THC, particularly
with the CH4 component subtracted, is of primary interest in air quality.
The measurement of primary emissions of nonmethane HC as well as the
secondary HC products formed as the result of chemical reactions in the
atmosphere is of extreme interest to the RAPS.
Nitrogen Oxides
Nitric oxide (NO) and nitrogen dioxide (N02) are the only significant
pollutants of the nitrogen compounds. Their importance as pollutants
arises mainly from their participation in photochemical reactions involv-
ing reactive organics and sulfur dioxide. NO and significantly less N02
are produced as a result of high temperature combustion. Although sig-
nificant quantities of NOx are produced from biological sources, most of
the NOx produced from anthropogenic sources originates as NO and is
oxidized in the atmosphere to N02. In air quality measurements it is
important to differentiate the NO component from total NO to characterize
x
the photochemical reactions occurring in the atmosphere.
Sulfur Oxides and Hydrogen Sulfide and Total Sulfur
The oxides of sulfur are common atmospheric pollutants that have
become one of the major pollution problems of the later 1960s. The oxides
of sulfur originate primarily from combustion processes. Solid and liquid
fossil fuels contain appreciable amounts of sulfur in the form of organic
XII-8
-------
sUlfur-containing compounds and inorganic sulfides. The oxides of sulfur
in combustion products are predictable from the sulfur content of the
fuel. Insignificant quantities of S02 are released to the global atmo-
sphere from natural sources. Conversely, the worldwide distribution
of H2S is primarily the result of natural processes, such as decay of
organic matter under anerobic conditions.
Most aerometric surveys have considered that a total sulfur measure-
ment is primarily a measure of S02. However, recent data have confirmed
that a significant portion of the total atmospheric sulfur may be present
as H2S in areas where coal combustion is minimal. It is recommended that
the air quality monitoring instruments of the RAPS differentiate the con-
tribution of H2S and S02 to the total sulfur burden.
Ozone
Photochemical reactions between NO and HC produce secondary prod-
x
ucts classified as oxidants. A major component of these oxidant products
is ozone. Other components present as oxidants are nitrogen dioxide and
peroxyacyl nitrates. The ozone concentration of the atmosphere is an
important indicator of the extent of photochemical processes occurring
in the atmosphere.
Suspended Particulate Material
Natural sources of atmospheric aerosols include dust, sea salt par-
ticles, and smoke from combustion and volcanoes. In addition, naturally
emitted gaseous materials can form aerosols through atmospheric reaction
processes. Sulfate aerosols are formed by sulfur compound reactions,
nitrate aerosols by nitrogen oxide reactions, ammonium aerosols by am-
monia reaction, and photochemical aerosols by atmospheric reaction of
natural emanations of terpenes and terpene-type hydrocarbons. Haze,
as a result of particulate in the atmosphere, is often the first notice-
able phenomenon associated with the development of air pollution in an
area. Thus, measurement of particulate and atmospheric turbidity is an
important parameter of air quality.
Other Pollutants of Interest Not Measured by Network
Measurements of atmospheric concentration of the pollutants for which
Air Quality Criteria have been issued will provide a general index as to
air quality within the instrumented region. However, measurement of
XII-9
-------
specific pollutants other than those scheduled for network monitoring
can be integrated into the Class A and Class B stations during the program.
These stations are designed to expand the instrument complement with only
a modest expenditure in expense and effort. The financial responsibility
for specific pollutant monitors other than the standard complement and
their installation costs should fall within the role of research programs.
As an example, if a specific research program requires atmospheric moni-
toring of beryllium, then clearly the cost of acquiring and analyzing
samples should be borne by the research program. Many other research
programs may require support from the permanent RAPS staff without the
associated costs being borne entirely by the research program. Expanding
the capability of network monitors to measure ammonia concentrations can
be achieved by modification of the operating procedures of existing net-
work monitors. This type of input data expansion is less clearly defin-
able as being within the role of research programs. The other end of
the spectrum is illustrated by the specific requirements for additional
pollutant concentration measurements to define the atmospheric chemistry
of the region. To define the reactive chemical processes within the area
the suspended particulate material should be characterized as to sulfate,
nitrate, and chloride content. In addition, the benzene or hexane soluble
fraction of the collected particulate must be determined. Atmospheric
concentration of peroxyacetyl nitrate, other organic nitrates, oxygenated
organics, total aldehydes, and formaldehyde should be defined to unify and
improve atmospheric models. Sample collection and specific analysis of
these atmospheric constituents do not lend themselves to automated proce-
dures and were not included in the network instrument complement. The
atmospheric chemistry measurement requirements are not considered to be
within the role of a research program but rather as a separate function
within the framework of the RAPS. The details of the sampling procedures
and analysis requirements for the atmospheric chemistry phase are described
elsewhere in this report.
Pollutant Monitoring Techniques
The types of air quality sampling instruments incorporated into the
station instrument complement for a large network must be evaluated based
on the specific requirements of the RAPS. The following list of param-
eters must be considered in the evaluation of commercially available instru-
ments to measure atmospheric pollutants:
Reliability--a five-year lifetime with minimum downtime
.
Suitability to unattended operation
XII-IO
-------
Accuracy and precision--within state-of-the-art
.
Sensitivity--must measure pollutants in rural areas
Specificity--must be specific or reasonably so
.
Stability--minimal zero drift and span drift
Ease of calibration--frequency and time required to calibrate
.
Sample cycle time
.
Reponse lag time
.
Operating temperature extremes
.
Operating humidity range
.
Linearity of response
.
Instrument noise
.
Supplementary reagents or gases required
.
Utility services other than power required--water, sewer
.
Physical size
.
Ease of maintenance--type of construction
.
Initial cost of instrument.
These factors and others must be weighed in selecting the most
suitable instrumentation for pollution monitoring. Selection or recom-
mendation of specific instruments by vendor is not within the scope of
this report; however, recommendations will be made as to the pollutant
measurement technique or techniques that are suitable for inclusion in
the network instrument complement. A committee within EPA should be
formed to select the individual pollutant monitors, by vendor, to com-
prise the basic instrument complement of the network stations.
XII-II
-------
Carbon Monoxide, Methane, and Total Hydrocarbon
A single automated gas chromatograph using hydrogen flame detection
and an appropriate inlet valving and column system can measure the atmo-
spheric concentration of CO, CH4' and THC. Nonmethane HC concentrations
can be determined by the THC and CH4 difference. Total HC are commonly
measured by direct sample gas insertion into the hydrogen flame detector.
The separation of atmospheric constituents by gas chromatography prior
to detection in this multipurpose instrument permits specific measurements
of CO and CH4 as well. The measurements of CO, CH4' and THC are standard
with this type of gas chromatographic instrument, but additional HC
(ethylene and acetylene) can also be measured if need warrants the addi-
tional expense. The measurement sensitivity (1 ppm full scale for CO,
CH4' and THC) is suitable for ambient air monitoring in rural areas.
One disadvantage of gas chromatographic measurement techniques is that
the measurements are cyclical, with five minutes required for analysis
of all three components. The use of the monitor with its five-minute
sample cycle at all three levels of the Class A stations would permit
only one concentration measurement per level for each IS-minute data
interrogation by the central data acquisition system. It is recommended
that a supplementary monitor for CO with rapid response be included in
the basic instrument complement of each Class A station associated with
a meteorological tower. A monitor for atmospheric CO based on the non-
dispersive infrared (NDIR) technique is recommended to supplement the
gas chromatographic CO measurements. The NDIR CO monitor would be
sequentially valved to measure all three vertical profile levels in a
single five-minute interval. The inherent poorer sensitivity of the
NDIR CO instruments (generally 50 ppm full scale) makes this approach
attractive only for those towers located in urban or semi rural areas
where CO concentrations are of such levels that meaningful data can be
obtained. The gas chromatographic CO would then be limited to measure-
ment at one level only, that is, 10 meters. The Class A stations without
towers and Class B stations would not require the supplementary NDIR CO
measurements.
The only other technique for the measurement of CO in the ambient
atmosphere is based on mercury replacement with detection of mercury vapor
by ultraviolet absorption. Commercial instrumentation employing this
technique is not considered as reliable or as specific as either the gas
chromatograph or NDIR. Therefore, atmospheric monitors using the mercury
replacement technique are not recommended for the basic instrument com-
plement of the network stations.
XII-12
-------
Recommended Monitors
CO, CH4' THC:
Gas chromatograph, hydrogen flame detection
CO:
Nondispersive infrared detection.
Nitric Oxide, Total Nitrogen Oxides
The measurement of NO and total NO probably presents the most dif-
x
ficult problem in evaluation of suitable instrumentation. The state-of-
the-art of NO measurement has not progressed as rapidly as the monitor-
x
ing equipment for other pollutants. Although very promising techniques
for the specific measurement of ambient concentrations of NO have been
developed recently, information concerning their reliability and stabil-
ity in unattended operation is scanty. The chemiluminescent technique
for NO is based on the gas phase reaction of NO and internally generated
ozone with detection of the emitted light by a photomultiplier. The
intensity of the chemiluminescence is related directly to concentration
of NO in the sample. The chemical conversion of NOx to NO permits the
determination of NOx by difference. The NOx conversion technique is
commercially available, but has barely graduated from the laboratory
development. Little information is available to substantiate conversion
parameters, such as catalyst lifetime and efficiency. A recent develop-
ment in the chemiluminescent NO monitor permits operation of the reaction
chamber at atmospheric pressure. Early models of the chemiluminescent
NO monitor used reaction chamber pressures of one to five torr achieved
by a mechanical vacuum pump. Measurement at atmospheric pressure simp-
lifies the operation of the instrument and reduces the space and power
requi rement s.
The chemiluminescent NO monitor is a second-generation development
of NOx instrumentation and is currently in a state of flux with improve-
ments and/or changes appearing frequently in vendor claims. It is recom-
mended that the potential applicability of this measurement technique
be examined critically by EPA at the time of vendor selection. If at
that time the chemiluminescent NO technique has realized its potential,
then this would be the recommended technique for inclusion in the basic
instrument complement of the network stations.
The coulometric technique for analysis of NO and N02 is recommended
as a second choice if the chemiluminescent technique does not appear
feasible. In this technique for N02' a series of scrubbers remove inter-
fering chemicals prior to sample oxidation of a potassium iodide solution
XII -13
-------
producing free iodine. The iodine is reduced back to iodide coulometric-
ally. with the required current directly related to pollutant concentra-
tion. The technique for NO is similar, except that N02 is removed by a
scrubber system and the NO oxidized to N02 by internally generated ozone.
The resultant N02 is then measured coulometrically as described. The
coulometric analyzer for NO and N02 is not as specific as the chemi-
luminescent technique and may require somewhat more routine maintenance.
The response time of the coulometric NO and N02 is much slower (10 min-
utes) than that of the chemiluminescent technique (2 seconds). Although
the instrumentation for coulometric NO and N02 is nearly identical, the
response time prohibits alternating NO and N02 measurements on a single
instrument. Therefore, two coulometric analyzers would be required in
each station instrument complement. The full scale sensitivity of the
coulometric analyzer is an order of magnitude lower than that of the
chemiluminescent (0.2 ppm coulometric versus 0.02 ppm chemiluminescent).
The membrane-solid electrolyte electrochemical NOx monitor could
also be considered a second-generation technique. This technique was
initially developed for stack sampling rather than ambient air monitoring.
Early instruments of this type lacked sensitivity to measure the low
concentrations of NO and NOx present in ambient air. Although consider-
able improvements in this electrochemical scheme have been made, frequent
replacement of the sensor module and poor specificity make this unsuit-
able for inclusion in the instrument complement of the network station.
Another technique for measurement of NOx is that of ultraviolet
absorption using a second-derivative measurement scheme. This third-
generation instrument is quite new and as yet unproven in its capability.
The measurement technique is very versatile and can be used to monitor
sulfur dioxide and ozone, as well as NO and N02' The full scale sensi-
tivity of the second-derivative (d2) spectrometer is 0.12 ppm for NO and
0.8 ppm for N02' These specifications indicate that at the current state-
of-the-art, the technique is feasible for NO but of marginal sensitivity
for ambient monitoring of N02' This d2 spectrometer shows definite
promise but, without additional information as to its performance both
in the laboratory and in the field, its applicability to the RAPS cannot
be assessed. The multicomponent measurement capability of the d2 spec-
trometer is of considerable interest and therefore it is recommended
that this technique be field evaluated, possibly within a RAPS research
program.
The last technique to be discussed for monitoring of NO in the
x
atmosphere is that of colorimetric measurement, the basis for the first-
generation atmospheric monitors. This scheme has been in use for many
years and its inherent problems and limitations are well known. Although
XII-14
-------
the NO-NOx colorimetric techniques have been continuously improved and
modifieQ to meet monitoring requirements dictated by more stringent air
quality regulations, the limitations of colorimetric techniques stimu-
lated the development of new second- and third-generation analysis
schemes. The problems associated with colorimetric analysis of NOx are
often those of a physical or mechanical nature, failure of pump tubing,
unreliable operation after cleaning, and formation of mold. Nonlinear
response to different concentrations of N02 is also often encountered.
Therefore, based on the long history of usage of the colorimetric tech-
nique, this monitoring technique for NO cannot be recommended.
x
Recommended Monitors for NO and NOx
Chemiluminescent Method:
of RAPS
subject to evaluation at initiation
Coulometric:
negative.
second choice if chemiluminescent evaluation is
Sulfur Dioxide, Hydrogen Sulfide, Total Sulfur
The first-generation and most commonly used pollutant monitors for
sulfur compounds measure only the S02 concentrations. The use of
instrumentation to monitor only S02 in the atmosphere is based on the
premise that the emission of sulfur compounds other than S02 represents
only a small fraction of the total sulfur burden. However, recent sur-
veys have shown that in areas of low coal usage significant concentra-
tions of hydrogen sulfide (H2S) may be present in the atmosphere. There-
fore, it is recommended that second-generation monitors employing tech-
niques that can distinguish the individual contribution of S02 and H2S
to the total sulfur burden be employed in the RAPS. The monitoring tech-
nique for S02 and H2S is similar in many respects to the technique used
for CO, CH4' and THC. A single automated gas chromatograph using flame
photometric detection and an appropriate inlet valving and column system
can specifically measure the atmospheric concentrations of S02 and H2S.
The automatic gas chromatographic measurement technique has the disad-
vantage that a sample cycle must be used rather than continuous measure-
ments. A sample cycle of three-minutes' duration is required for meas-
urements of S02 and H2S. The use of this monitor with its three-minute
sample cycle for all three levels of Class A stations would permit only
one measurement of each level for each five-minute interrogation by the
central data-acquisition system. Therefore, it is recommended that a
supplementary monitor for total sulfur with rapid response be included
XII-IS
-------
in the basic instrument complement of each Class A station associated
with a tower for vertical profile measurements. A monitor for atmospheric
total sulfur based on a flame photometric measurement technique without
separation by chromatography is recommended to supplement the gas
chromatographic 802 and H28 measurements. The flame photometric total
sulfur instrument would be valved to measure all three vertical profile
levels in a single five-minute interval. The gas chromatographic 802-H28
measurement would be limited to only one level--that at 10 meters. The
total sulfur monitor using flame photometric detection without separation
gives a high degree of specificity for measurement of all sulfur-
containing compounds. These techniques are reliable and require little
maintenance in unattended operation. The gas chromatograph-flame photo-
metric 802 and H28 technique and the flame photometric total sulfur have
been evaluated in both laboratory and field conditions. They are cur-
rently recommended as equivalent to the reference method published in
the Federal Register.
Another technique for the measurement of 802' discussed previously
in the section on nitrogen oxides, is that of ultraviolet absorption
using a second-derivative measurement. This technique is a third-
generation technique that has not been in use sufficiently long to eval-
uate. The sensitivity to measurement of 802 (0.06 ppm full scale) is
competitive with that of the flame photometric method. This instrument
should be evaluated further, possibly within the role of a research pro-
gram of the RAP8.
Another second-generation technique for the measurement of 802 is
that of membrane-solid electrolyte electrochemical measurement. Poor
specificity and frequent sensor module replacement make this technique
unsuitable for unattended monitoring of ambient concentrations of 802.
The last techniques for measurement of 802 to be discussed are those
of the first-generation 802 monitors. These techniques include colori-
metric, coulometric, and conductimetric means of sensing the concentra-
tion of 802 in the atmosphere.
The instruments employing colorimetric analysis, particularly the
conventional West-Gaeke scheme, can provide acceptable analyses for 802.
The pararosaniline colorimetric method has been collaboratively tested
and is recognized as valid. The instruments using colorimetric analysis,
however, are subject to considerable downtime due to mechanical problems
and lengthy calibration procedures.
Coulometric methods of sensing atmospheric 802 concentrations are
probably the best of these first-generation schemes. The correlation
XII-16
-------
of the observed concentrations of 80 by coulometry with respect to the
2
West-Gaeke method is fairly good, and fair reliability under field con-
ditions can be obtained with this instrumentation. However, neither
the reliability nor the accuracy of the coulometric 802 measurement is
equivalent to the gas chromatograph-flame photometric or the flame photo-
metric technique without prior gas chromatographic separation.
Concentration measurements of 802 by conductometric schemes exhibit
the poorest correlation with respect to the West-Gaeke method of all the
first-generation techniques. Although the conductometric instrumentation
can be reliable in field service, the inaccuracy of the measurement of
802 concentrations makes this technique unsuitable.
Recommended Monitor for 802 and H28:
Gas chromatograph-flame photometric detection
Recommended Monitor for Total 8ulfur Concentrations
Flame photometric detection without prior gas chromatographic
separation.
Ozone
The specific measurement of ozone concentrations can be accomplished
through a second-generation technique--measurement of the chemilumines-
cence resulting from the gas phase reaction of ozone and ethylene. This
measurement can be made with the reaction chamber at atmospheric pressure
and relatively simple instrumentation. In the presence of a known con-
centration of ethylene supplied by a compressed gas cylinder, the inten-
sity of the chemiluminescence is a linear function of the ozone concentra-
tion in the sample gas. The instrumentation is reliable, accurate, and
suitable for unattended operation in field installations. Chemilumines-
cent ozone measurement is recommended as equivalent to the reference
method published in the Federal Register.
A third-generation scheme for the measurement of ozone concentra-
tions in the atmosphere is based on ultraviolet absorption. Instruments
based on ultraviolet absorption using a second-derivative detection tech-
nique and conventional Beer's Law absorption are available commercially.
The ultraviolet absorption techniques give a specific measurement of
ozone concentration and do not require a supplementary reagent, as does
XII-17
-------
the chemiluminescent method. The ozone monitors based on ultraviolet
absorption have not been on the market a sufficient length of time to
evaluate their performance under field conditions. However, it is recom-
mended that this technique be evaluated under field conditions within a
research program of the RAPS.
The first-generation techniques for the measurement of total oxidant
are based on colorimetric and coulometric measurements. Both the colori-
metric and the coulometric measurements utilize the oxidation of buffered
potassium iodide solution and a measurement of the free iodine released.
These techniques do not give a specific measure of ozone but respond in
a variable way to the concentration of N02 in the atmosphere. The
presence of S02 also interferes with the colorimetric or coulometric
total oxidant measurement and must be removed by a filter or scrubbers
prior to analysis.
The direct measure of ozone concentration as with chemiluminescence
or ultraviolet absorption is a significant improvement over the methods
based on total oxidant using an applied correction factor for N02
concentrations.
Recommended Monitor for Ozone
Chemiluminescent ozone-ethylene reaction.
Suspended Particulate Material
The usual method of measuring the concentration of atmospheric
particulate material is by high volume sample collection of aerosol on
filters. The concentration of the aerosol is determined from the volume
of air passed through the filter and the weight of collected particulate.
This technique does not lend itself to unattended operation. Many manual
steps are required--weigh filter before particulate collection at a
known humidity, package filter for transport to collection site, place
filter in high volume sampler, sample for known time interval at a known
air flow rate, remove filter and package for transport to laboratory for
weighing, weigh filter and collected particulate at a known humidity,
calculate particulate concentration from sample flow rate and weight of
collected particulate. The manpower requirements to obtain data on par-
ticulate concentrations with high volume samplers within a time frame
similar to that of other pollutants is not feasible.
XII-IS
-------
An alternative technique for particulate measurements is through
the uSe of a particle mass monitor. Here an electrostatic precipitator
deposits suspended particulate material on a quartz crystal oscillating
at its resonant frequency. The resonant frequency decreases linearly
with the total mass of the particulate deposition. The frequency of
the circuit containing the sample crystal is compared to the frequency
of a reference oscillator and the difference frequency is directly
relatable to the mass of the particulate deposited on the sample quart
crystal. If the sampler is cycled to sample for ten minutes every four
hours, the sample crystal needs to be cleaned on a once-a-week basis.
A respirable range attachment can be cycled such that the alternate
readings will be total particulate and respirable particulate. However,
though the principle of this automated technique makes particle mass
monitors attractive for inclusion in the monitoring network, the com-
mercially available mass monitors have proven to be unsatisfactory in
service. The design and engineering of this type of instrumentation
must be considerably improved before mass monitors can be recommended
for inclusion in the instrument complement of the RAPS monitoring network.
Specifically, fluorescent tracer studies indicate that particulate depo-
sition other than on the sample crystal is experienced. Inefficient
collection with respect to high volume samplers corroborates the find-
ings of tracer studies. Measurements of negative masses or a weight loss
of collected particulate during operation of the particle mass monitor
not explainable by volatile aerosol or changes in humidity tends to
indicate loss of particulate after deposition can occur. This technique
should be reevaluated during the RAPS program to determine if adequate
instrumentation can be or has been developed to fill the need of auto-
mated particulate measurements.
Another approach to particle mass measurement is through the attenu-
ation of beta particles. A radioactive source emitting low-energy elec-
trons in conjunction with a scintillator/phototube detector can determine
masses of collected particulate as a function of the electron density
of collected material. This technique, though not as sensitive as the
oscillating crystal technique, does show promise for automation of par-
ticulate measurements. Commercially available instrumentation based on
the beta-particle attenuation is or will be commercially available soon.
The beta gage should be evaluated as to its potential application to
the RAPS.
Since no automated procedure seems suitable for inclusion in the
pollutant monitoring network, a compromise manual system of particulate
collection and measurement is recommended for the RAPS. It is recommended
that two conventional high volume samplers be employed at each station
to collect aerosol on a sequential or concurrent basis. These samplers
XII-19
-------
would be serviced by filter replacement during the normal twice-weekly
station service call. Thus two high volume samples would be obtained
per week per station. It is recommended that initially each sampler
would operate for 24 hours to obtain an integrated sample on each of two
selected days. Alternatively, the samplers might be run concurrently
with one sampler operating during an interval of four hours during the
day to characterize particulate concentrations as a function of time of
day.
The second particulate measurement at each monitoring site will be
that of atmospheric turbidity using an integrating nephelometer. This
instrument will give a response related to the degradation of visibility.
The integrating nephelometer through the use of a pulsed flash lamp
detects the total scattered light from all scattering angles. The output
of the nephelometer, the scattering coefficient, can be related to visi-
bility in units of miles. The integrating nephelometer is a simple,
reliable instrument, well suited to unattended operation in the monitor-
ing station network.
Vertical profile particulate concentrations are not within the realm
of practicality due to the losses of particulate while traversing long
sample lines. Therefore, the particulate measurement with the particle
mass monitor and the integrating nephelometer will be made at only one
level--either ground or three meters above ground.
In addition, the results of the high volume
taken by the Atmospheric Chemistry program would
data collection system.
sampler measurements
be integrated into the
Recommended Monitors for Suspended Particulate Material
High volume sample collection on filters (2 required at each
monitoring station)
Integrating nephelometer integrating-measurement of atmospheric
turbidity
Summary of Pollutant Instruments
The following instruments are recommended for inclusion in the basic
pollutant complement located in each of the network monitoring stations:
XII-20
-------
Carbon Monoxide, Methane, and Total Hydrocarbons
Gas chromatograph, hydrogen flame detection
Carbon Monoxide
Nondispersive infrared detection
Nitric Oxide-Total Nitrogen Oxides
Chemiluminescent method, subject to evaluation at initiation
of Regional Study
Coulometric, second choice if chemiluminescent evaluation
negative
Sulfur Dioxide-Hydrogen Sulfide
Gas chromatograph-flame photometric detection
Total Sulfur Compounds
Flame photometric detection
Ozone
Chemiluminescent method, ozone ethylene reaction
Suspended Particulate Material
High volume sampler collection on filters
Integrating nephelometer, measure of atmospheric turbidity.
Instrument Calibration
Two separate systems of calibration are recommended for each network
station to verify the accuracy of the pollutant monitors. The first con-
sists of a local calibration capability permanently installed at each
station. The second provides for mobile calibration vans to visit each
station to more accurately calibrate the instruments and to maintain
instrument standardization.
Local Calibration
The local capability consists of a two-point system for calibration
of the pollutant monitors to be initiated upon an interrogation command
XII-21
-------
or on a cyclical basis. This two-point calibration will establish the
operating zero and the span of the pollutant monitor. A cylinder of high
purity nitrogen simulating an air sample will serve as a zero gas to
establish the true operating zero of the monitors. Compressed gas cylin-
ders containing a known concentration of individual pollutants will be
maintained at the station site to calibrate the span of the monitors. The
concentration of the calibration gas must be chosen such that nearly full
scale response can be obtained on one of the less sensitive ranges. The
use of compressed gas cylinders containing very low concentrations of
pollutants is to be avoided. The stability of ppm-pollutant concentra-
tions during long-term cylinder storage is questionable with the uncer-
tainty increasing at lower concentrations. The actual pollutant concen-
tration appropriate for the calibration cylinders must be based on the
stability of the pollutant. Compressed gas cylinders with known concen-
trations of the pollutant of interest will serve as span gas for CO, CH4,
NO, and H28. Of these pollutants, H28 would be the least stable under
long-term storage conditions.
The total HC monitor would be span calibrated based on its response
to CH4'
The 802 monitor would be calibrated using a permeation tube system.
A permeation tube containing 802 and maintained at a known temperature
would lose a known and predictable weight of 802 as a function of time
through the walls of the permeation tube. A constant, low diluent flow
over the permeation tube with subsequent dilution to an appropriate con-
centration will serve as a span gas for the 802 monitor. A commercially
available automatic permeation tube calibration system would be used at
each network station for 802 calibration. The same system could be used
for N02 calibrations if required. The total sulfur monitor would be span
calibrated based on its response to 802.
An ozone generator would also be used at each network station to
span calibrate the ozone monitor. These generators are commercially
available and are well suited to automatic calibrations.
The integrating nephelometer contains a built-in calibration system
which would be used in this application.
It is recommended that all monitors be calibrated by a zero-span,
two-point calibration at least once a week. It is also recommended that
all instruments of the station complement be calibrated simultaneously.
The calibration manifolds and/or plumbing must be fabricated of
material inert to the specific pollutant. The calibration schedule of
XII-22
-------
of the network stations should be established such that only one network
station will be in a calibration sequence while others are collecting
data. The time required for calibration is estimated to be 1/2 to 1
hour. A digital flag will indicate on the data tape at the central
facility when the calibration sequence is in progress at any network
station.
Calibration Vans for Primary Calibration
It is recommended that three mobile vans be equipped with complete
calibration facilities to perform four-point calibrations of all monitors
located at network stations on a regular schedule. The calibration vans
will intercalibrate monitors at all network stations. This procedure
will ensure a continuity when standard calibration cylinders are replaced
and will verify the accuracy of automatic calibration procedures.
The calibration equipment of the van will be similar to that of the
integral station system, but will be operated manually by experienced
technicians. This dual calibration procedure should permit recognition
and correction of minor instrument malfunctions without significant data
loss. With appropriate scheduling of the mobile vans, each network sta-
tion would be calibrated once each month. It is estimated that one day
or less would be required for a primary calibration of each station.
Role of Research Programs
Although specific research programs or areas of research interest
are discussed elsewhere in this report, the relationship between research
programs and the network air quality measurements should be clarified.
Although the Class A stations will provide vertical concentration profiles
of pollutants within the surface layer, measurements of both meteorolog-
ical and air quality conditions above the surface layer will be needed
by the on-going monitoring programs and a significant number of the
research programs. These data needs can be provided by the use of
balloon-borne instrumentation but more likely by special instrumented
aircraft, especially helicopters.
It is unlikely that significant air quality measurements can be made
by balloon-borne sampling techniques although equipment is available for
ozone and total particulate measurements from balloons. In a congested
urban area such as St. Louis, it will be necessary to rely upon high
performance helicopter operations to provide the detailed vertical sam-
pling for air quality data that will be necessary in the research program.
XII-23
-------
The helicopter to be used in this program will have to be of at least
medium size so that the carrying capacity can include both a technical
observer and several hundred pounds of instrumentation. With such a
load, the performance of the helicopter should be sufficient to permit
it to climb to an altitude of 4000 or 5000 ft in a time period of 15 or
20 minutes. There is no standard instrument package available for air-
borne sampling and thus an initial research project associated with the
RAPS should be the design and development of a research instrumentation
package to be used aboard an available helicopter. This research project
would obviously require specification of the type of helicopter to be made
available to the program.
Another research area of interest to the monitoring network is the
establishment of measurement correlations between the first-generation
pollutant monitors with the specific detection of the second-generation
monitors used in the network stations. A wealth of data has been
accumulated from a variety of aerometric studies by pollutant monitors
employing first-generation measurement techniques. The RAPS provides a
unique opportunity to correlate and intercompare measurements obtained
by the older equipment with modern monitors under real conditions.
Precise establishment of correlations could give new meaning to old data
and provide updated information about pollutant concentrations without
the expense of new aerometric surveys. These correlations could be of
direct help to control districts using first-generation monitors and
could identify monitors and equipment that need to be upgraded.
A similar need exists for evaluation of new instruments and tech-
niques. The measurement correlation program established for new instru-
ments would be primarily a field evaluation and not a feasibility study
for new concepts. Establishing the feasibility of instrument concepts
in the development of new concepts is an ongoing program in the EPA.
A case in point is the Fourier Transform Infrared Spectroscopic
system. This measurement system uses a Block interferometer in conjunc-
tion with a dedicated computer to analyze trace gases in the atmosphere.
This measurement technique, a development of EPA, has undergone limited
field testing to assess its potential as an air quality monitoring in-
strument. The Fourier Transform Spectrometer appears to be a very prom-
ising measurement technique. The sites and back-up support provided by
RAPS could provide the vehicle in terms of field experience to complete
this and other new concepts in instrumentation.
XII-24
-------
Chapter XIII
DATA ACQUISITION AND HANDLING
Introduction
The Regional Study will require data from an extensive network of
air measurement stations distributed over a large urban and rural area.
Extensive measurements of meteorological parameters and of a number of
specific and general air contaminants will be made at these stations at
intervals of less than 1 minute to several hours. A rough estimate of
the total number of measurements per day is about 1-1/2 million.
The Regional Study is intended to provide a continuous, organized
data base for a wide variety of research programs. The data collected
must therefore meet the usual standards and criteria for a good set of
experimental data. They must be reliably recorded with sufficient reso-
lution; every item of data must be unambiguously identified, and all
measuring devices must be periodically calibrated at intervals based on
operating experience. Because of the large number of instruments and
the geographical spread of the stations, all of these activities should
be performed automatically.
The goal will be to provide sufficient data identification and docu.
mentation on the primary data recording tapes so that they can be de-
ciphered and used without any supplementary documentation.
Policies and Principles
The design of the recording system and the data formats will be
guided by these principles:
All data must be individually
Each instrument should have a
be recorded together with the
and automatically identified.
unique identifier code which will
value of the measurement.
XIII-I
-------
Information should be recorded in "free field"* formats. This
will permit unrestricted extensibility of all available levels
of information on the tapes without obsoleting the formats of
any previously recorded data.
Tape is cheap; it should be used freely for descriptive comments
about any special conditions. In particular, there should be a
fully descriptive header on each tape, in English, describing
any nonroutine research instrumentation that may be recorded on
that tape.
All codes should be translated into understandable quantities as
early as possible. The data transmitted from the stations will
be binary-coded representations of output voltages and instrument
identifier codes. These should all be immediately recorded on a
primary archival tape. The central computer should prepare sec-
ondary archival tapes in which all the measurements are in en-
gineering units, corrected for the instrument calibrations, and
the instrument identifiers are translated into readable descriptors
All manually collected data should be recorded on tape in the
same formats as the automated data. The standby computer and a
teletype can be used to transcribe such data via a computer pro-
gram that will write it on the tape in the proper formats for
future analysis.
System Overview
The system will comprise a large number of discrete stations dis-
tributed over the instrumented area. It is presently estimated that
there will be at least 49 stations (see Figure XIII-1) linked to a cen-
tral data acquisition system, of which 17 or more will be provided with
30-meter tall towers for vertical profile sampling, and the others will
be semimobile or transportable units that may be relocated occasionally
as the program develops. There will also be a group of self-contained
instrument stations, recording data locally on tape, and not connected
to the central computer.
The stations will contain varying complements of meteorological and
air sampling instruments. The tower stations will especially concentrate
*
See data formats.
XII 1-2
-------
:><1
H
H
H
I
w
MEASURING -L.
INSTRUMENT IDENTIFIER SERIAL TO
D - - ... PARALLEL .
.. INTERROGATOR
CONNECTION 5 MINUTE CONVERTER READ AND
PANEL - SCANNER ' WRITE
Range Control MEMORY
I ANALOG TO I-
-
- DIGITAL
~ CONVERTER ~
I
I CLOCK MEMORY
CONTROL CONTROLLER
I 1
INSTRUMENT MASTE R
- CALIBRATION
.... STATION
CONNECTION I 30 SECOND CONTROLLER CONTROLLER
I -
PANEL D SCANNER
~~
I '
I ' , BIDIRECTION
PARALLEL-5ERIAL
I CODE
,I. CONVERTER
---
I = Identifier Channel
D= Data Channel Calibration Equipment Telephone
COMMUNICATION System
MODEM
SA-1365-2
FIGURE XIII-1
INSTRUMENT STATION DATA AND CONTROL SYSTEM
-------
on gathering meteorological data, and the semipermanent
focused on air pollution measurements, but all stations
least a basic set of both kinds of data.
stations will be
will record at
The wind components, air temperature, and humidity will be recorded
at 30-second intervals and at 3 different elevations at each of the 20
tower stations. Most of the pollution measurements will be taken at
5-minute (or longer) intervals. The meteorological data will be the
largest set of data to be transmitted.
The stations will be connected to the regional central computer by
telephone. The lines may be leased or dial-up, depending on cost factors
not yet known. The computer will query all the stations at 15-minute
intervals to gather the data. Because the instruments will be sampled
locally at more frequent intervals, each station must have a temporary
data storage device to accumulate data for 15 minutes. This arrangement
will simplify and speed up the data transmission to the central computer,
because it will not have to stay connected to a station long enough for
an electromechanical scanner to scan all the instrument channels, but
only long enough to read out the memory device at the station. This
can be read at data rates limited only by the telephone line capacity,
which should be at least 2400 bits per second.
The tower stations are estimated to make about 500 measurements in
15 minutes. Each value will require an 8-bit transmission code. Iden-
tification of the type of instrument, the instrument number, and the
position, should require 2 or 3 more 8-bit codes per value. The entire
15-minute data block can be transmitted in 5-10 seconds. The data from
the smaller stations will require about 2 seconds for transmission.
The computer at the central facility (shown in Figure XIII-2) will
convert the raw primary data to secondary data in engineering units. It
can do this during the slack time in each query cycle. The data from
each station will be written as it is received on a primary archival
magnetic tape and also temporarily on a fast access disk memory. The
disk memory will be large enough to hold in a more permanent file area
all the current calibration data for all instruments, and also a the-
saurus table of readable identifiers corresponding to the transmitted
instrument identification codes. At the end of each station transmis-
sion, the computer will access the data block on the disk for that sta-
tion, compute the calibration corrections, convert values to engineering
units and identifier codes to descriptors, and write the secondary record
tape while doing this.
XIII-4
-------
:><
I-t
I-t
I-t
I
tJ1
Telephone
Circuits
15K 15K
DISK A DISK B TAPE TAPE TAPE TAPE
DRIVE A DRIVE B DRIVE C DRIVE D
I I
CONTROLLER
I I
10K 10K
COMPUTER A COMPUTER B
MULTIPLEXER ..
I I TELETYPE
COMMUNICATION COMMUNICATION WITH LINE CRT
MODEM --- KEYBOARD PRINTER TERMINAL
MODEM
AND PAPER (LOW SPEED) ?
t TAPE
-
4
..
SA-1365-1
FIGURE XIII-2
CENTRAL FACILITY DATA PROCESSING SYSTEM
-------
The most recent portion of the secondary record will also be re-
corded on another portion of the disk for system monitoring purposes.
This will be accessible by the stand-by computer as requested by staff
personnel and can be printed or displayed on a CRT output terminal.
The main reason for having a stand-by computer is to take over the data
acquisition task whenever the first computer needs maintenance. It can
be used for data monitoring, entry of manual data to tape, and other
computational tasks when both computers are operational.
The disk memory at the central site will require sufficient capacity
for at least two full sets of calibration data, at 2 values per installed
instrument, so that each new calibration, as it is received, can be com-
pared with the previous one to detect any anomalous behavior of either
the instrument or the calibration system. It must also contain a table
of conversion factors to engineering units, the library of the identifier
codes, and the English language descriptors for about 2000 separate ac-
tive instruments, plus room for about one hour's worth of monitoring
data, and a "scratch" area for writing the input data for each station.
Instruments
Most of the meteorological and air sampling instruments will be
devices that measure continuously and generate a DC output voltage pro-
portional to the quantity measured. Instruments of this kind are easily
recorded by connecting their outputs to a multiinput scanner device which
periodically connects an analog-to-digital converter to each input ter-
minal.
A few instruments, such as automatic gas chromatographic analyzers,
operate intermittently. They may require a fixed quantity of air to be
processed through an analytical c.ycle. or may have some other internal
cycling operation, so that the measurement appears at the output ter-
minals at periodic intervals. Data recording from instruments of this
kind can be done with either a synchronous control system, in which the
instrument is initiated by an €xternal timer, or by an asynchronous poll-
ing system which either keeps polling the instrument to find out whether
it has a new datum, or is signaled by a code initiated at the instrument
to record the data. It appears that the best choice for data acquisition
control in these stations will be the use of a single-cycle controller
for the entire station. This method will not require any modification
of most of the instruments, but will require a capability in the inter-
mittent instruments of being reset and started on a new cycle by an ex-
ternal signal.
XIII-6
-------
A few instruments can measure more than one quantity during a single
instrument cycle. One gas chromatographic analyzer will measure CH4' CO,
and total hydrocarbons as it goes through its analysis of an air sample.
The most straightforward way to fit such an instrument into the scanning
pattern is to require that the instrument be equipped with separate out-
put circuits for each component that it measures. Since the timing or
recognition of the separate peaks will be a function of internal instru-
ment parameters, such as flow rates, it would be very difficult to syn-
chronize an external scanner to the characteristics of each instrument.
A better procedure is to connect an instrument that measures three dif-
ferent components to three different scanner input terminals, with three
different identifiers. This application may require that the instrument
be extensively modified. It will require internal capability for iden-
tifying each component peak, and converting the peak height (or the in-
tegral of the peak, depending on the nature of the device) to a voltage
that can be routed to the appropriate output terminal and held constant
there until the reset signal starts the instrument on the next measure-
ment cycle.
Some instruments provide output voltages in the 0-100 millivolt
range, and others produce 0-5 or 0-10 volts. An analog-to-digital con-
verter with automatic range switching can be used in the system. This
will be much less expensive than equipping each of the low output in-
struments with a stable DC amplifier. The range setting of the converter
can be provided as part of its digital output code. It may increase the
number of bits associated with the measurement by 2 or 3, depending on
the number of range settings needed. Alternatively, the instrument iden-
tification code can also specify the converter range setting, and avoid
the necessity of providing it as data.
One of the most serious problems that must be faced in any system
for handling very large amounts of data, whether automated or not, is
the identification of each datum. For this reason, every instrument
should be self-identifying to the system, so that the researcher who is
using the data much later does not have to find log books and laboriously
correlate the notes made by instrument installers with the recorded data
sets. In addition, human errors in recording channel numbers, or acci-
dentally interchanging a pair of signal leads in a large bundle out of
sight of the actual instruments can ruin a large amount of data. The
problem essentially is that there are many information transfers in the
identification process, and everyone must be correct.
We are proposing a method for instrument identification that will
be automated in complete parallelism to the data itself. The data from
a given instrument come out of a cable into a particular input channel
XIII-7
-------
connector on a data scanner. The instrument identifier device should be
physically mounted on the instrument and connected by the same cable car-
rying the measurement signals to a multicontact scanner connector. The
identifier unit might consist of a 16-bit shift register and a 16-bit
read-only memory. These are inexpensive solid state devices. The read-
only memory should contain the instrument type code and instrument num-
ber, which would be written into it when it is installed in the instru-
ment, and also painted or stamped on the instrument panel where it can
be easily seen.
Information, such as sampling height or other static data, may be
implied by or included in the instrument identifier code. If a more
dynamic indication is needed, as, for example, a flow rate indication
for a specific instrument, that can be regarded as a separate measure-
ment and assigned to a "pseudoinstrument channel."
The master controller is a device that interacts directly with the
central computer via the telephone system. It has the following func-
tions:
Establish communication with the central computer by responding
to a signal from the telephone data set with the station iden-
tification code.
Recognize the control commands sent by the central computer after
the computer recognizes the station.
Send control signals to the local necessary unit,
the calibration controller, and any other station
instruments specified by the computer.
the scanners,
systems or
Data Acquisition Equipment at Stations
The measurement and identification data from every instrument must
be periodically collected, digitized, and forwarded to the central com-
puter or stored locally. This equipment is described below primarily
in terms of the various functional units that are required. Most of
these units are available as commercial items, with the exception of
the instrument-mounted identifier and the interrogator, for which a
specific detailed design is described. However, it may be advantageous
to have the data acquisition systems custom-manufactured, because of the
number of them that will be required to equip all the stations.
XII I-8
-------
It is assumed that all instruments will fall into one of two classes.
In one class, the output of each instrument will be sampled once each
30 seconds. These instruments will be sampled in sequence and one com-
plete cycle will take 30 seconds. Instruments in the second class will
be interrogated once each 5 minutes and, consequently, a complete sampling
cycle will take 5 minutes. The fast and slow instruments would be con-
nected to separate scanners that provide paths from each instrument to
the common equipment that generates interrogation commands and stores
the resulting measurement and identification code data in a memory unit.
It is assumed that the scanner provides three paths between each instru-
ment and the common control circuits. One path is provided for the
analog voltage output of the instrument, one for command information
from the interrogator to the identifier, and the third for output of
the identifier that is sent back to the common control.
The identifier unit would generate a 16-bit serial identification
code. In one simple implementation of the identifier circuit, an inte-
grated circuit shift register and a few gating circuits would be required.
The read-only memory that contains the instrument identification could
be realized in a set of metallic conductors on the printed circuit board
that mounts the shift register and logic ICs. The conducting paths would
be broken (by physical removal of the metal) according to the desired
identification code. DC power for the identifiers would be provided
from the centralized control circuits on a continuous basis (there is
no need to go through the scanner paths). An initial pulse to load the
shift register with the ID code and subsequent clock pulses to shift the
code out of the register would be provided from the centralized control
circuits through the scanner paths. It is reasonable to expect that an
identifier unit might occupy a volume of 4 or 5 cubic inches and be made
for less than $50 in quantities of several thousand units.
The identifier function could be realized by a simple circuit like
that shown in Figure XIII-3. Its input is a series of 17 pulses, the
first being an initiate pulse and the next 16 being clock pulses for the
shift register. A control flip-flop, whose function is to direct the
incoming pulses either to the set parallel input or to the shift terminal
of the shift register, is normally held at the output of a one-shot cir-
cuit in the condition for sending the incoming pulse to the set input.
Consequently, the first of a series of pulses sets the identification
code into the shift register. The trailing edge of the first pulse
triggers the one-shot circuit and removes the holding signal on the
flip-flop. The trailing edge of this pulse, stretched out by a series
capacitor, also causes the flip-flop to change state so that the follow-
ing input pulses are sent to the shift terminal of the shift register.
The serial input of the shift register (connected at the most significant
XIII-9
-------
INPUT FOR PULSES
FROM INTERROGATOR
FF
:x:
1-4
1-4
1-4
I
J-'
o
SHIFT
SET
SERIAL
INPUT
FIGURE XIII-3
16 STAGE SHIFT REGISTER
MOST SIGNIFICANT END
LOGIC DIAGRAM OF INSTRUMENT IDENTIFIER
/
PARALLEL INPUTS (1m
4
SERIAL OUTPUT
(I.D. Code)
SA-1365-16
-------
end) is grounded and, as the ID code is shifted out, zeros are shifted
in. At the end of 16 cycles, the register has been cleared and is ready
for the next interrogation. The one-shot circuit returns to its normal
state shortly after the end of the 16th pulse and the control flip-flop
is again held in its quiescent state. Off-the-shelf ICs needed to build
the circuit cost about $20 in quantities of 100, at the present time.
One set of centralized control and storage circuits could be used
to service all instruments. It appears that solid-state random access
memories are becoming cheap enough to provide an attractive method of
storing measurement and identification data during the 15-minute interval
in which data are accumulated (at present prices, storage for 4,000 8-bit
words would cost about $1500). Since the input to such a memory is in
parallel form, the serial identification code that comes from the iden-
tifier would have to be converted to parallel form. This could easily
be done in an 8-bit shift register. After the first eight bits of the
ID code have been shifted from the identifier to the register in the
centralized circuit, the parallel output of this register could be gated
to the memory. When the next eight bits have arrived in the shift reg-
ister, they can be gated to the memory. Then the output of the AID con-
verter, which is normally in parallel form, could be gated to the memory.
Thus, for each instrument interrogation, three 8-bit words would be put
into the solid-state memory.
In addition to generating control pulses for the instrument identi-
fiers and transferring data to memory, the central control circuits could
control the connection of the high and low speed scanners to the common
equipment and provide stepping pulses for the scanners. It appears that
the required complexity of control may be substantially less than the
capability provided by commercially available scanners. There appears
to be a high probability that a system including scanners, AID converters,
control circuits, and memory that is specifically designed to meet the
requirements of the EPA regional operation might be substantially less
expensive than a system designed around commercially available equipment,
in particular commercially available scanners. This approach would ap-
pear especially attractive if no exotic requirements were added to the
basic requirements of sequentially interrogating all instruments in the
station. It would not be difficult, for example, to provide the capa-
bility of stopping the interrogation sequence on any selected instrument
and displaying the measured output visually. Further, a relatively sim-
ple AID converter could be used with instruments having a wide range of
output voltages (100 mV to 10 V, for example) by using a small part of
the ID code to set automatically the input range of the AID converter.
Compatibility with a serial communication modem could easily be achieved
by transferring the contents of the memory in parallel form to an 11-stage
XlII-ll
-------
shift register in which the conventional start and stop bits of the 11-
bit ASCII code are automatically attached to the eight bits of data.
Shifting out of this register would give the ASCII characters in the
serial form that is required by communication modems.
The central control
be a part of the station
of all the operations at
circuits mentioned above may be considered to
master controller which provides for control
each station.
The master controller will contain the read-only memory device that
holds the station identification code, so that it can verify to the com-
puter that it has called an instrument station and can proceed to collect
data. About 4-6 different commands may be used to control the station,
so the controller can be a hard-wired logic device, with fixed output
signal wiring.
The time at which measurements are taken need not be transmitted
in the data stream, because the scanners can be reset to the beginning
of a cycle when the computer calls for data, or by computer command even
if the data transmission interval is chosen to be different from the
data recording interval. The accuracy of the clock controlling the
scanners will be adequate for determining time relative to the time
value recorded on the tape by the computer each time it resets a sta-
tion scanning system.
Each station must also have a calibration controller- This equip-
ment will operate all the relays and valves required to perform the
physical calibrations of the instruments. Each instrument that can be
calibrated should be read for zero and full scale. The logical procedure
is to calibrate all the instruments in the station at the same time, be-
cause only one command is needed, and because data analysis that uses
several items of data from a station will be least interfered with if
all measurements are available or all missing. The calibration procedure
for a typical station may take two 15-minute measurement cycles, because
many of the chemical analysis instruments require several minutes to
equilibrate after a large change in the sample gas.
The on-line stations also require
into the telephone network. These are
vices and are readily available.
a modem and a data set to couple
standard data communication de-
The self-contained stations will require similar data acquisition
equipment, except that the station controller will have its own clock,
rather than be controlled by signals from a computer. The controller
will otherwise perform the same functions as it does in the on-line
XIII-12
-------
stations. The local recorder will also be
capacity to accumulate several days' data,
is manually replaced. It will not require
different. It must have
if necessary. before the
any read-out capability.
tape
tape
Data Formats
The standardization of data formats is always a troublesome problem.
For efficient computer operation, the data formats should be rigorously
defined, including the exact number of bits in every value, and the number
of values in every record. Identifier codes and labels are often packed
into the smallest possible number of bits, and the sequence of numbers
is standardized so that identification is implied by the position of a
value in a record. In spite of the advantages of minimizing storage
space and computer running time, there are very serious drawbacks to
this approach. Once the format is established, there is no room for
expansion without making a new format. All measurements are limited to
the predetermined resolution, and there is usually no way to deal with
special research requirements that were not initially planned for.
A much better approach for handling data in a research environment
is the concept of "free-field" recording. Free-field information record-
ing is usually character-oriented, rather than bit- or word-oriented.
A few special characters are chosen to have special meanings. These are
not part of the data, but serve as identifiers to the computer to recog-
nize elements in the data stream, or changes in the data type. Gener-
ally, a computer does not read free-field data as quickly as it reads
formatted data, because it must examine every character to test whether
it is one of the special characters. The process can be made somewhat
more efficient at little added cost in storage space by choosing only
one such "escape" or signal character. Its significance to the computer
is that the next following single character is a data code descriptor.
This speeds up reading because the computer makes only one test on each
character to determine whether or not it is the escape character. It
makes more detailed tests only on the character following the escape
character, and otherwise reads the stream of characters into appropriate
formats determined by the current data group that it is set for. The
escape character also is automatically the terminator of any data group.
This free-field coding will, for example, accept a sequence of 8-bit
identifiers and values, then perhaps a 16-bit value, then maybe a comment
in English describing some special experiment recorded in that portion
of the tape, back to some more data, and so on.
XIII-13
-------
The standard ASCII (American Standard Code for Information Inter-
change) code is a defined set of 128 characters, each using a different
7-bit code. The character set includes all the upper and lower-case
letters, the 10 digits, many punctuation, arithmetic, and logical char-
acters (such as !, ", $, @, +, and #), and a group of 32 special communi-
cations control characters. An eighth bit is normally transmitted as
part of each character code block for use in parity testing for trans-
mission errors.
An 8-bit character can be used to represent integer values from
-127 to +128, thus providing at least 1% resolution of either positive
or negative values, depending on where the effective value of zero is
defined to be. One of these characters will be taken as the escape
character and another as null (not zero), reducing the number set to
254 characters. There are then also 254 possible different data code
identifiers, even if this function is restricted to only one character.
This is probably many more than will be needed, but the reserve will be
helpful whenever new code categories need to be introduced. The most
important code categories are:
16-Bit identifier (two 8-bit characters)
8-Bit numeric data
16-Bit identifier + 8-bit numeric (three 8-bit characters)
16-Bit identifier + 16-bit numeric (four 8-bit characters)
16-Bit identifier + "no-of-characters in following numeric"
+ numeric
Zero-point calibration state
Full-scale calibration state
Measurement state
Comment
16-Bit unsigned integer
32-Bit signed floating point number.
These category codes are essentially announcements to the computer
that is reading the tape to tell it how to interpret the following record
XI II -14
-------
up to the next occurrence of the escape character- For practical reasons
there needs to also be a null character, corresponding to no information
or blank space. A convenient representation of null is 00000000, and an
obvious representation for an escape character then is 11111111.
With this free-field system, an unbounded number of instruments,
resolution levels, types of data, and lengths of data records can be
managed. We strongly recommend that any standardization of data formats
be established at this information processing level.
The Central Data Collection Facility
Overview
It is envisioned that there will exist a single data collection
computer facility to serve the entire data collection network. Data
from all the measuring sites would be transmitted via telephone communi-
cation links to the facility. The computers would be in continuous op-
eration on a 24-hour basis, every day of the year. During normal working
hours, it is expected that the facility would be attended, and also rou-
tinely used by some research personnel. Outside of these hours, the fa-
cility might be unattended or supported by a single person; this should
not preclude the use of the facility by research personnel during the
out-of-normal hours.
Two major outputs from the facility would be (1) "clean" data tapes
for use by research personnel, and (2) "raw" and clean data tapes,for
archival purposes.
Reliability of the data collection process would be promoted by
proper configuration of the equipment and the limited use of redundant
equipment. The unavailability of any critical element (including the
source of electrical power) whether due to a failure or routine mainte-
nance, should not interrupt the data collection process. For example,
it is likely that a single minicomputer will be able to perform all
critical data collection tasks, but will lack the capacity to process
the data to produce clean tapes. This latter activity, plus others,
might be handled by another such computer. If these two computers are
chosen to be identical, that computer engaged in the nonessential activi-
ties can be switched to the data collection task, if the other computer
fails while collecting data.
XII 1-15
-------
Functional Tasks
Thus far, five major functional tasks have been identified for the
data collection facility, (1) control of data collection, (2) maintenance
of data quality, (3) production of clean output, (4) archiving of data,
and (5) real-time observation of data. These tasks are grossly described
below. At this time, it is premature to discuss either the detailed data
processing activities, or the details of the facilities hardware.
Control of Data Collection
It is necessary that the data collected at each site be col-
lected periodically, in an orderly and timely manner. Present thinking
about the computer facility provides for a high degree of sharing of the
computer's time, storage, and programs in the data collection activity.
This form of sharing is common practice in large, medium, and small sys-
tems as a means of achieving substantial economies with little or no
sacrifice in performance.
To effect this sharing, the computer must control the order
in which the instrumentation sites are allowed to talk to the computer.
When the computer determines it is prepared to accept more data, it
selects the site to be so permitted and then sends a message to that
site informing it to proceed. After the site has completed transmission,
the computer can repeat the cycle with another site. If transmission
difficulties are encountered, the computer may request retransmission
of the data; in the case of a bad dial-up circuit it may redial the site
or, if a persistent problem is encountered, the computer can immediately
produce a trouble report.
Maintenance of Data Quality
The computer can check the values of each received datum for
reasonableness. Such a check could compare the datum or its rate of
change with an upper or lower bound that is associated with that meas-
urement. A datum can be compared with other data values, for instance,
the difference between two independently measured temperatures can be
compared with the directly measured difference.
Where a check yields an out-of-bound result, the computer can
immediately produce a trouble report so that an investigation, and pos-
sibly corrective action, can be undertaken by the proper persons.
XIII-16
-------
Production of "Clean" Output
It is anticipated that the data desired by the research per-
sonnel should be easy to work with. The data as received from the site
are expected to be "raw" in that they will be (1) in unusual units of
measure, (2) without calibration correction, (3) confusingly related
through codes to the specific instrument and quantity measured, and
(4) interspersed with items of "control" information. The computer
facility should remove these undesirable features and produce clean
data: actually the researcher could and, in some instances, might un-
dertake this task himself. It is expected, though, that the complexity
and burden of the effort will make this an undesirable task for him.
The clean output will have the data in such a form that (1)
each datum is in the units commonly used for the specific application,
(2) each datum will have been corrected on the basis of the latest cali-
bration data, (3) each datum will be clearly identified as to the nature
of the quantity measured and the source instrument, and (4) no extraneous
information will be present.
Archiving of Data
A data-archiving process will be used. As a minimum, the raw
data obtained from the sites, including calibration information, will be
archived. These data will likely not have been processed significantly
by the computer as it records them on magnetic tape after receipt from
the sites. The clean data can also be placed in the archive, although
it will be possible to derive the clean data from the raw data in the
archi ve .
Real-Time Observation of Data
As an aid to the conduct and design of experiments and the
maintenance of the instrumentation network, it should be possible for
research and operational personnel to monitor the data being collected
by the central facility. The monitoring should be under the control of
the observer(s). Thus, the observer should be able to specify which
specific measurements and specific sites he desires to observe. He
should be able to examine the data just gathered at the facility, or
the last value received, or the values over a short previous period
(e.g., during the preceding hour). Some simple cathode ray tube dis-
plays may greatly aid in the presentation of data plots of data simul-
taneously gathered from several sources.
XII 1-17
-------
Equipment Complement
We can identify the major components of the computer facility.
Communication I/O Multiplexors
These items serve to couple the many communication lines from
the instrument sites to the computer system. Under computer control,
the device can effect the simultaneous exchange of information with one
or more lines.
CPUs and Main Memories
The CPUs (central processing units) and their memories do all
the information processing tasks, including the control of information
flow to display terminals and disk storage units. Two CPUs are envi-
sioned, one concerned with data collection activities, and the other
with the production of clean data tapes and the displays. Different
programs serve to differentiate these two identical CPUs.
Disk storage Units and Tape Units
The disk storage units provide storage of the most recently
collected measurement information, instrument calibration data, and data
validation limits. The tape units provide the long-term storage of all
the collected data. When one reel of tape is filled, it is placed in
the archive or delivered to a researcher, and a new reel is mounted on
the tape unit.
Displays and Other Output Units
Cathode-ray tube display units are for use by researchers and
the operational staff, to observe the current and recently captured data.
These data are kept on the disk units. Via the keyboard of the display
unit, a person can send messages to the CPU, directing the CPU with re-
spect to the particular data items to be shown, the formats, and so on.
XIII-18
-------
Disk File Organization
A iarge amount of data must be kept on the disk, at several differ-
ent levels of permanence, and for different functions, as described pre-
viously. Although the exact details of the data organization will depend
on specific implementation factors, such as the hardware and operating
system and programming language capabilities of the computer that will
be chosen, a few general statements and suggestions about data organiza-
tion can be made.
The reason for using a disk memory is to provide random access to
data. However, that capability must be used with the total system op-
eration in mind, because it requires about 16 milliseconds per access
on a typical head-per-track disk memory, and a longer time with a moving
head disk system. A very large number of disk access requests for proc-
essing one item of data can use more time than is allowable for this
system. Therefore, it is necessary to organize the data on the disk so
that the number of accesses used per operation is not excessive.
The data stream, as it is received from each station, can be written
sequentially in a reserved "scratch" area of the disk, on top of the data
from the previous station transmission. This requires only one position-
ing operation for writing the data and also for reading it, because it
can be processed in the same order it was received.
The processing of the data will require access to the calibration
and the descriptor for each instrument. These can be located together
in a single contiguous table entry whose "disk address" is computable
from the instrument identifier code. Thus, as the received record for
each instrument is read by the computer, it can compute the location of
all the information needed to calculate the corrected reading in suitable
units, and also the readable description of the instrument, for recording
on the secondary archival tape and on the monitor portion of the disk.
To facilitate the use of free formatting of measurements and de-
scriptors, and also to conserve disk space, it is probably better not
to preassign file space to specific instruments, or to predetermine the
length of field needed for each instrument, but rather to write compactly
on the disk, and maintain an integer-coded "disk address directory" to
the locations of all information relating to specific instruments, both
in the reference files and in the monitoring data. This indirect address-
ing can be organized hierarchically by stations and by instrument types,
and can be utilized by the monitoring display program, as well as by the
on-line data processing.
XI II-19
-------
Manual Data Entry
Because the Regional Facility will be used to support a wide variety
of short-term research programs in addition to its routine data collec-
tion operation, there will be many occasions on which experimenters will
make measurements without using the automatic data collection system.
In order for such data to be available for later computer analysis in
combination with the basic meteorological and air quality data, they
must be entered on data tapes in appropriate form. This data entry op-
eration can be done by using the standby computer, the terminal, and the
tape equipment. There are several approaches that can be followed to
assure that all data are entered in a readable form.
One way is to depend on a trained data clerk, who would know all
the details of the various data formats, and be skilled at copying data.
This procedure probably would require, however, that each experimenter
first transcribe all his data from the laboratory notebook form to some
standardized tabular form. This may be necessary if there is a large
amount of such manual data entry, but if there is only sporadic activity
of this type, it may be more practical for the experimenters to learn
to enter the data themselves.
Two kinds of support will be needed to assist visiting experimenters
to correctly record their data. First. a well-written manual, containing
a variety of examples of data sets from various kinds of experiments,
should be provided for study and reference. Second, the data should be
entered through a computer program that is designed to assist the user
by requesting data in the correct sequence, doing whatever verification
can be done easily, and generating the correct formats and format codes
for writing on the tape. This program should be written in a high-level
language so that it can quickly be modified as needed for special cases.
The staff programmer should be available to advise or instruct those
who need to use the manual data entry procedure, and to progressively im-
prove the data entry program to provide most of the assistance.
Other output units include line printers and console typewriters.
Via these devices, data and messages can be printed.
Reliability Considerations and Maintenance
We assume that it is intended for the data collection to proceed on
a continuous basis. This being the case, provisions must be made to ac-
commodate the incidence of the more frequent types of malfunctions and
XIII-20
-------
the needs for routine maintenance. Only two of the five functional tasks
are critical--control of data collection, and maintenance of data quality.
When the other functions are sacrificed, then the CPU, storage, and other
system resources thus freed are available for use by the critical tasks.
Consider the simple case where it is necessary to remove from service
the CPU controlling the communication lines. By loading a new program
into the other CPU (which had been producing the clean data tapes and
running the CRTs) the CPU can assume the communication control task.
With these constraints, we see the need for dual communication I/O
multiplexors, dual CPUs and memories, and backup disk units and tape units.
Two independent sources of commercial a.c. power should be provided.
A separate, independent standby power source seems unwarranted for, if
a serious power failure does occur, it is likely that many instrument
sites could be affected too.
If air conditioning is needed, consideration should be given to the
use of a backup unit.
Communications for the Data Collection Network
The Telephone Companies
A set of about 12 telephone companies have been identified as serv-
ing the area around st. Louis that is encompassed by the data collection
network. Some are very small, such as the Orchard Farm Telephone Co. of
Orchard Farm, Mo., which serves a population of 706. The Staunton Tele-
phone Company of Staunton, Illinois does not even have a listed telephone
number. Actually, we need only be concerned with the three major com-
panies--Illinois Bell Telephone Co., Southwestern Bell Telephone Co.,
and Continental Telephone Corporation.
In practice, it is likely that those responsible for the operation
of the network will need to deal with only Southwestern Bell, especially
if all the communication lines terminate in st. Louis. Where the facili-
ties do not reach to st. Louis--consider an instrumentation hut in Or-
chard Farm, connected only to a local dial-up line--then these many com-
panies may need to be individually dealt with.
XIII-21
-------
Facilities
Only a limited number of communication facilities have been con-
sidered at this time. These are: (1) full duplex leased circuits,
(2) circuit treatment, (3) modems, and (4) data couplers. Not consid-
ered were (1) dial-up service, (2) automatic dialing units, and (3) auto-
matic answering units. A very brief functional description of each of
these items is listed below.
Full Duplex Leased Circuit--This is a voice-bandwidth transmission
facility that interconnects two specific points; this circuit is avail-
able 24 hours per day every day of the month for the sole use of the
customer. Being full duplex, it is possible to transmit simultaneously
a different signal in each direction. Depending upon the terminal equip-
ment, either voice or data signals can be sent.
Circuit Treatment--This is a special adjustment made to the circuit
to improve its frequency response. This type of adjustment permits data
to be transmitted at rates in excess of 2000 bits per second, though less
than about 10,000. Special telephone company equipment is attached to
each end of the circuit to provide the treatment.
Modems--These are items of equipment that couple data equipment to
the communication circuit. They transform the digital signals received
from the data equipment into a signal of a form compatible with the com-
munication circuit; and vice versa. Typically, a single modem will pro-
vide both functions simultaneously. One modem is required at each end
of a circuit. The rate at which data are transmitted is controlled by
the modem; our interest is in data transmission rates in the range of
1200 to 2400 bits per second, though rates very much slower and very
much faster are also in common use.
Data Couplers--If the equipment connected to the circuit (the modems)
is not telephone company property, then this device, supplied by that
company, is required by law. Such a condition arises when nontelephone
company modems are used.
ular
must
Dial-Up Service--This is
telephone service. With
"dial" the number of the
the type of service provided users of reg-
this service, the originator of the call
desired party. A circuit between them is
XI II-22
-------
set up, if possible, and maintained for the duration of the call. If
the called party is busy or does not answer, the circuit cannot be com-
pleted. There is also a finite probability, the order of 0.05, some-
times greater, that a shortage of circuits will also prevent the call's
completion. The dialing time, and the switching time at the central of-
fice, can delay the set-up of the circuit by as much as 15 seconds. The
quali ty of the dialed circuit may be poor--typically only data rates of
less than 2000 bits per second can be used.
Automatic Dialing Units--This is a device that is required to enable
an item of automatic equipment, such as a computer, to "dial" the call.
With such a unit, automatic placement and termination of calls is possible.
Automatic Answering Units--This is a device that is required
enable an item of automatic equipment, such as an unattended data
lection system, to answer a call.
to
col-
Costs
Detailed cost analyses are not possible at this moment, primarily
due to the complexity of the rate structures. Factors in this complexity
are (1) whether or not the service crosses a state line; (2) if entirely
within a state, the particular state of concern, i.e., Missouri or Illinois;
(3) for a given state, whether the service crosses telephone company boun-
daries; and (4) for a metropolitan area, whether the service is entirely
within that area or not.
For the moment, we can ignore
ing costs. These are conservative
be lower.
these many factors and use the follow-
because the actual costs will probably
Full Duplex Leased Circuits--$4 per
wholly within the st. Louis metropolitan
for all other circuits, and $30 one-time
mile per month for circuits
area, $3 per mile per month
installation charge per circuit.
Circuit Treatment--$28 per month per circuit, and $20 one-time in-
stallation charge per circuit.
XIII-23
-------
Modems--$146 per month per pair of modems, and $200 one-time instal-
lation charges per pair.
Data Couplers--$16 per month per pair, and $20 one-time installation
charge per pair.
Alternative Communication Approaches
We first observe that the average data rates from the instrumenta-
tion sites are expected to be quite low, the order of 10 bits per second.
This can be contrasted with the 110 bits per second for 100-word per
minute teletypewriter service, or the 1000 bits per second easily ob-
tainable over leased or dial-up circuits.
Conceptually and technically, the simplest communication arrange-
ment would provide a leased circuit from each data collection site to
the central computer site. The poor utilization of the line, 1%, indi-
cates that this may not be the most economical arrangement, however.
Experience has shown that one can effect a trade-off between communica-
tion volume (and speed) and equipment complexity; pending an examination
of that trade-off, we cannot dismiss this one circuit per site approach.
A second approach would tie several sites. perhaps 10 or more, to
one single leased circuit that threads its way across the area (several
such circuits might be needed). By suitable additional equipment at
each site and at the computer site, the several sites could effect a
sharing of the one circuit. With this technique, the utilization of
the circuit could be substantially increased, and the number of circuit
miles substantially reduced. Because of the additional costs of the ad-
ditional equipment. such an approach is most beneficial when applied to
sites distant from the computer site. An operational concern with this
approach is reliability, in that a circuit malfunction, or a severe site
malfunction, could disable all data-gathering activities for all stations
on the affected circuit. Where repair times, primarily of the circuit,
may be long (days to tens of hours may not be uncommon), this could be
serious.
A third approach would have the computer periodically dialing each
site. This is the most complex technique thus far considered, and the
least examined. Cost economy is potentially possible here but, until
the actual values are ascertained, we can only talk in generalities.
In the Bay Area, a dialed call may cost on the order of $172.80 per mile
per month; this is equivalent to $0.04 per call, plus $0.004 per mile
XIII-24
-------
per minute. Strong nonlinearities tend to reduce the cost per mile as
the distance increases. If the circuit is fully utilized when in use,
then it need be occupied on 0.01 month per calendar month, at best, thus
the equivalent cost can be considered to be $1.73 per mile per calendar
month, which is competitive with $3 to $4 per mile per month for a lease.
Additional costs to be considered, however, are those of the automatic
dialing and calling units, plus other auxiliary equipment. Dialed con-
nections, we should also note, should prove to be more reliable than the
leased, for with dialed connections a multiplicity of circuits is usually
available to set up a call.
XIII-25
-------
Chapter XIV
MIXING LAYER OBSERVATION PROGRAM
Introduction
Air quality and meteorological measurements made at the ground-based
instrument stations provide atmospheric vertical profile data within the
surface layer. However, both meteorological and air quality data will
also be required above the surface extending through the mixing layer
to a level on the order of one kilometer. These measurements, as well
as specially designed procedures to support specific research experi-
ments, will be required on a periodic and repeating basis similar to the
ground-based operational scheme.
This chapter provides a discussion of the mixing layer observation
program and presents several sets of concepts and recommendations for
its development. The first covers instrumentation platforms suggested
to function as the primary observational system and includes a helicopter
for air quality measurements and the balloon-tracking system known as
METRAC for meteorological measurements. The second set discusses special
aircraft-based meteorological observations and their required instrumen-
tation. The third set of concepts and recommendations covers the devel-
opment of aircraft instrument packages with emphasis on air quality
instruments. The last set reviews the current and anticipated future
capability of the Western Environmental Research Laboratory (WERL) to
conduct airborne air quality and meteorological observations and their
potential involvement in the Regional Study.
General Mixing Layer Observation Methods
Two general methods can be employed for the acquisition of mixing
layer measurements. The first includes the use of manned fixed-wing
aircraft or helicopters, and the second covers balloon-borne systems.
The characteristics of these two types of measurement systems are, of
course, markedly different and each system tends to offer selected
advantages and disadvantages, depending upon the nature of the measure-
ment to be made. In the case of meteorological observations, for example,
the advantages and drawbacks of aircraft measurements are fairly well
known. To sum a few of the more important considerations, it should be
XIV-l
-------
recognized that control of the measurement platform is a distinct advan-
tage. Measurements can be made at the desired location and can be
repeated in short order. On the other hand, the accuracy of the wind
measurements in particular is open to some question because of aerodynamic
effects, and, equally important, the operating costs of the system are
quite high. As such, it seems that use of aircraft would be highly
desirable for special, short-duration meteorological research studies,
but perhaps not for daily operation.
The collection of meteorological data by balloon techniques, on the
other hand, is restricted to the more-or-Iess uncontrollable trajectory
of the balloon. This can be an advantage when studying Lagrangian or air
parcel processes, but is a disadvantage when objective cross sections of
Eulerian conditions over the region are desired. Some flexibility is
introduced in the use of both rising and constant level balloons, espe-
cially if it is possible to track simultaneously several balloons of one
or both types. In this manner, one could observe the vertical structure
of the atmosphere at several locations while also observing the nature
of wind trajectories over the region at various altitudes and/or geo-
graphical locales. Operating costs can differ widely among various
balloon systems and may this be an advantage for one and a drawback for
another. These considerations have led to the recommendation that a
newly developed balloon-borne instrument system known as METRAC be
seriously considered for the Regional Study to provide routine, upper-air
meteorological data. Aircraft or helicopters would be utilized primarily
for air quality measurements alone on a routine basis, although the use
of either for special meteorological observations is encouraged.
The selection between a fixed-wing aircraft and a helicopter for air
quality measurements depends upon the number of factors with the choice
significantly affected by the instrumentation package to be carried.
Fixed-wing aircraft offer several advantages over helicopters as a means
of making airborne measurements. The typical vibrations encountered
aboard a helicopter are of low frequency and large amplitude, making
isolation of vibration-sensitive instruments very difficult. Shock-
mounting techniques are most effective against the higher frequency,
small amplitude vibrations encountered in fixed-wing aircraft. In addi-
tion, considerable forward speed must be maintained in a helicopter to
obtain air quality samples uncontaminated by the helicopter exhaust prod-
ucts in the downwash. The distinctive, annoying sound produced by heli-
copters in low level flight may result in frequent complaints when over-
flying populated areas. Conversely, fixed-wing aircraft are subject to
more constraints when overflying urban areas than are heli.copters.
XIV-2
-------
The use of a helicopter as both an airborne laboratory and logistical
support for emergency servicing of remote monitoring stations makes the
rotary-wing aircraft an attractive choice. In addition, vertical profile,
air quality measurements are more easily achieved over a fixed point
with a helicopter than with conventional fixed-wing aircraft. On these
bases it is recommended that a helicopter serve as the primary platform
for an airborne laboratory. Support of routine helicopter operations
could quite likely be provided by WERL during the execution of large field
experimental programs. The expected increase in the capability of WERL
over time, as discussed later, should prove to be a significant resource
of the EPA, and its use in the Regional Study for the support of special
programs should be of great value.
Primary Aircraft Support for the Regional Study
At least three important tasks can be defined for the use of
craft in support of the ground-based activities that comprise the
work to be performed during the Regional Study. These are
air-
major
(1)
Logistic support during all phases of the research program
(2)
Design, acquisition, testing, and operation of appropriate
instrument packages in support of selected research experiments.
(3)
Conduct of selected air quality and possibly meteorological
research studies that require relatively sophisticated instru-
mentation and techniques.
The first two of these tasks should be initiated during the first
six months of the program. It is recommended that a helicopter be ac-
quired as a fixed cost to the research facility, and that operational
costs be included during the life of the Regional Study. It is suggested
that the third task be undertaken in collaboration with WERL, Environ-
mental Surveillance Division. Mechanisms for allocation of aircraft,
equipment, and personnel from WERL to the Regional Study should be
specified by memorandum of understanding issued by the appropriate EPA
office and acknowledged by WERL.
To accomplish the above and related tasks, two major systems need
to be defined--a medium-sized helicopter with appropriate performance
characteristics, and a fixed-wing aircraft to support planned upper air
quality and meteorological studies. The helicopter is specified in
terms of commercially available equipment that can accommodate a medium-
sized instrumentation package subsystem. The fixed-wing aircraft and
XIV-3
-------
operating personnel can be provided by WERL. Integration of the WERL
aircraft with the air quality and meteorological subsystems defined by
the Research Plans can proceed concurrently with implementation of the
ground station network.
Helicopter Support Function
The Regional Study provides a unique opportunity to critically
examine airborne monitoring concepts and systems under controlled con-
ditions. Moreover, the large scope of the study in terms of ground sta-
tions, personnel, equipment, and geographical coverage provides a reason-
able justification, if not a logistic requirement, for aerial support.
The versatility of a helicopter makes it ideally suited to provide
emergency support in consideration of the number of ground stations that
will require maintenance, calibration, and other functions. To provide
sufficient capability to transport personnel, equipment, and an airborne
instrument package, a medium-sized helicopter is recommended. The Bell
Jet Ranger (Model 206B), or its equivalent, is considered to meet the
requirements of the Regional Study. In comparison with smaller craft,
the Jet Ranger offers an increased work area, it is less restrictive as
to weight and balance or total load weight, and it provides more power
for faster surveys, yet has the landing advantages of the small models.
Additional factors include a reasonable fuel capacity and range, and
availability of 110 to 115-volt outlets without the necessity for con-
version from a 12- or 24-volt system. General features of the Jet
Ranger are indicated by Figures XIV-l and XIV-2 and by Table XIV-I. A
sample weight calculation is shown by Table XIV-2.
Schedule of Operations
The scheduling of helicopter operations will vary to some extent
depending upon the schedule of research experiments, weather conditions,
and helicopter maintenance requirements. In general, more frequent and
longer operational periods would be expected during active experimental
programs in the spring-summer season than in the winter months. Major
maintenance could be scheduled during the less active winter season.
For planning purposes here, the operational schedule will be assumed to
include 18 hours per week during the period March through October and
12 h9urs per week for the balance of the year. Thus a total of appro-
ximately 830 hours of operation could be expected.
XIV-4
-------
r
DIM D
I
DIM C
~
H
<:
I
CJI
11.7
DIM D
1- DIM C----1
1
'==
FIGURE XIV-1
31.2
R;
Pa-1
D~~=r~
I
33.3
k~ =st. -
3.6 .1 DIM B 1
21.4
i 39.1
I.
8.8
,I,
I
I
12.6
19.5
-6.0
DIMENSIONS A B C D
STANDARD SKID 1.0 8.3 6.4 9.5
HIGH SKID 1.9 8.3 6.8 10.4
STANDARD FLOATS 2.2 13.4 11.5 11.0
EMERGENCY FLOATS 2.1 10.9 8.8 10.5
10.5
SA-1365-30
EXTERNAL FEATURES OF THE JET RANGER MODEL 2068
-------
:><
H
<:
I
(j)
MAIN CABIN CARGO SPACE
APPROXIMATELY 40 CUBIC FEET
A. FORWARD CABIN WIDTH
B. FORWARD SEAT-WIDTH (each)
C. - AND LEG-ROOM
D. AFT SEAT- AND LEG-ROOM
E. -WIDTH
-TO FLOOR
-TO ROOF OF CABIN
F. BAGGAGE SHELF-WIDTH
G. -DEPTH
H. BAGGAGE COMPARTMENT-FWD WIDTH
I. -AFT WIDTH
J. DOOR OPENING HEIGHT
K. CEI LI NG TO FLOOR
L. BAGGAGE SHELF HEIGHT
M. BAGGAGE COMPARTMENT-HEIGHT
N. -LENGTH
O. AFT DOOR OPENING-WIDTH
[J}
;;.----
-N-
BAGGAGE COMPARTMENT
16 CUBIC FEET
FEET
4.0
1.3
3.6
3.3
3.9
1.0
3.3
3.4
0.8
3.4
1.6
3.8
4.3
1.3
1.8
3.0
2.9
FIGURE XIV-2
INTERIOR DIMENSIONS OF THE JET RANGER MODEL 2068
SA-1365-31
-------
Table XIV-l
WEIGHT, PERFORMANCE, AND POWER RATINGS FOR THE
JET RANGER MODEL 206B
WEIGHTS:
LB
*F.A.A. empty wt
Basic configuration wt (F.A.A. empty wt + usable oil)
tStandard configuration wt
F.A.A. normal gross wt
F.A.A. external load gross wt
Max useful load (gross wt F.A.A. empty wt)
Conf useful load (gross wt std conf wt)
Max external load
1455
1466
1515
3200
3350
1745
1685
1200
*
t
The 206B empty weight includes 70 pounds for 5-place upholstered interior
with cushioned seats, soundproofing, rug, and seat belts.
The 206B standard configuration weight includes 29 pounds of additional
standard accessories not included in the empty weight plus 11 pounds of
usable oil and 20 pounds of ballast (allowance for avionics).
PERFORMANCE: (Standard Day)
Take-off, gross wt 2200 3000 3200
I.G.E. hovering ceiling (2 ft skid ht) ft 20,000 13,000 11 , 300
O.G.E. hovering ceiling ft 17,800 8500 5800
Certificated altitude ft 20,000 20,000 13,500
Service ceiling (TOP) ft 20,000 20,000 13,500
Service ceiling (MCP) ft 20,000 19,000 13,500
Sea level max rate of climb (TOP) ft/mi n 2730 1540 1260
Sea level vert rate of climb (TOP) ft/mi n 2160 690 280
Max allowable A.S. (at 5000 ft) mph 153 153 134
Max allowable A.S. (at sea level) mph 150 150 140
Max cont cruise A. S. (sea level) mph 142 136 133
tRange cruise A.S. (sea 1 evel) miles 365 350 345
Max cont cruise A.S. (5000 ft) mph 145 138 138
tRange cruise A.S. (5000 ft) miles 417 393 383
t
7 Ibs fuel ~llowed for warm-up and take-off but no reserve.
POWER RATINGS
Engine (Allison 250-C20)
Take-off horsepower, SHP
Max continuous, SHP
Airframe ratings
Take-off horsepower, SHP
Max continuous, SHP
400
346
317
270
FUEL
Type
Capacity
Aviation turbine
76 gallons
XIV-7
-------
Table XIV-2
SAMPLE WEIGHT CALCULATION FOR JET RANGER
MODEL 206B
Configuration Weight
Weight
in Ib
Weight empty
* Engine oil (usable)
1455
11
* Flap restraint
3
* Night lighting
10
* Door locks
1
* Hour meter and clock
2
* First aid kit
1
* Fire extinguisher (cabin)
8
* Fuel filter kit
4
* Ballast or equipment allowance
20
Total configuration weight
1515
Mission Weight
Standard configuration weight 1515
Pil 0 t 170
Full fuel 494
Passengers (4) 680
Baggage 250
Additional payload or equi pment 91
Total (gross weight) 3200
*
Accessories, kits, included in standard configuration.
XIV-8
-------
Costs
A general breakdown of estimates for direct and fixed costs is
provided by Table XIV-3. Certain additional costs will need to be eval-
uated in terms of the actual conditions found to exist in the St. Louis
area; these include rental costs for ground space at an airport and
provisions for helistop areas near some ground stations. Helistop pads
will require essentially the same road access and local approval as the
ground station. A 100-foot diameter area is suitable, assuming the
absence of tall buildings or other obstructions in the vicinity of the
pad. Both the direct and fixed costs are developed on the basis of an
hourly cost of operation. The principal components of the costs of
direct operation include consumables: fuel and lubrication, maintenance,
labor; and parts. Direct costs are estimated at $45.12 per hour of
operation.
Estimates of hourly operating costs derived from fixed costs require
assumptions concerning depreciation, insurance, pilot salary, and annual
flight hours. Values for depreciation and insurance are based on an
assumed helicopter configuration with a list price of $150,000. A
depreciation schedule is assumed such that the helicopter will have a
30% residual value at the end of five years. The total hourly fixed
cost is estimated at $56.62. The total annual estimated costs are shown
in Table XIV-4 under the assumption of 830 hours of operation per year.
Annual costs average $84,400, with five-year costs at $422,000.
Helicopter Instrumentation Package
The airborne instrumentation complement should include both air
quality and meteorological instruments. The basic instrument complement
would be installed as a package in a helicopter for measurements to be
made on a routine or a scheduled basis. The recommended meteorological
parameters to be measured on a routine basis would be dew point, temper-
ature, and ground surface temperature. Moisture measurement techniques
are probably less satisfactory and less well advanced in the state of
the art than most other meteorological instrumentation techniques.
Several types of moisture analyzers are commercially available based on
different moisture measuring techniques. The most common techniques
are based on (1) an electrical resistance change as a desiccant material
absorbs moisture, (2) condensation of moisture on a mirror at a known
temperature, and (3) absorption of moisture by a desiccant deposited on
an oscillating quartz crystal. Of these techniques, the one best suited
to airborne operation would be a dew point hygrometer based on change in
reflectivity as moisture condenses on a mirror. This technique has been
used successfully in airborne operation by NCAR.
XIV-9
-------
Table XIV-3
DIRECT AND FIXED COSTS FOR HELICOPTER SUPPORT FUNCTION
Direct Cost of Operation
Consumables
Fuel - 25 gal/hr at 30~ gal
Lubrication - grease and oil/hr
Maintenance labor
Airframe maintenance man-hours
Periodic Required per Approx
Insp 1200 hr M/H ea
100 hr 11 8.0
1200 hr 1 128.0
Unscheduled maintenance estimate
Engine maintenance man-hours
Scheduled and unscheduled maintenance
Total maintenance man-hours
Labor cost =
453 M/H X $6.00 hr
1200 hr
Allowance for airframe spare parts
Replacement
Replacement
Unscheduled
Replacement
at scheduled
at overhauls
replacements
of retirement
inspections
parts
XIV-IO
7.50
0.18
Total
M/H
-
88.0
128.0
87.0
150.0
453.0
=
$0.37
3.29
0.98
6.03
-
Cost/hr
$ 7.68
2.27
10.67
-------
Table XIV-3 (Concluded)
Cost/hr
Allowance for engine spare parts
Replacement at scheduled inspections
and unscheduled replacements
$ 1.50
Allowance for engine overhaul
Based on 1000 hr TBO (price estimated at $13,000)
13.00
Allowance for contingencies
Primarily experimental functional modifications
10.00
Total direct cost per hour
45.12
Fixed Cost of Operation
Fixed cost on annual basis
Depreciation
An equal apportionment over 5 years is assumed
with a 30% residual
$21,000
Insurance
Direct government operation would imply self-
insurance; for illustration, however, 8% of the
list price is assumed based on private operator
experience
12,000
Pi 1 ot
Average salary experience
14,000
Total
47,000
Total fixed cost per hour
56.62
XIV-11
-------
Table XIV-4
HELICOPTER SUPPORT FUNCTION
TOTAL COST OF OPERATION
(830 Hours per Year)
Hourly direct costs $ '45.12
Hourly fixed costs 56.62
Total hourly costs 101.74
Total annual costs 84,400.00
Total five-year cost 422,000.00
The measurement of temperature does not present any particular prob-
lem for airborne operation. Therefore, either the commercial Rosemount
probe or the technique used by NCAR for airborne operation would be
recommended.
Ground surface temperature is measured through the use
red radiation measurement. A simple narrow-field bolometer
mapping of the surface radiative temperatures.
of an infra-
would permit
The space and weight restrictions of the airborne laboratory platform
limit the number of pollutants that can be monitored on a routine basis.
The pollutants that are recommended for routine monitoring would be sulfur
dioxide, carbon monoxide, ozone, and atmospheric turbidity. Sulfur
dioxide monitoring is recommended for several reasons. First, it is a
pollutant of considerable interest. As it will be studied extensively
within the reactive chemistry program, the additional data concerning
its vertical distribution and transport will be of great value in the
development of atmospheric models. The concentration of sulfur dioxide
within the ambient rural atmosphere will be significantly lower than the
concentration anticipated within the plume of St. Louis, but, even at
100 kilometers downwind, the boundaries of the plume can be defined
through airborne measurements. The airborne measurements of carbon
monoxide will be used to indicate dilution processes within the plume
of St. Louis. The ozone measurement and atmospheric turbidity will be
used as an indicator of photochemical activity and general air quality
within the flight patterns.
XIV-12
-------
The research project to design an instrument package for use aboard
a helicopter must evaluate the available pollutant monitors to determine
those measurement techniques that are most suitable for this application.
The monitoring techniques should be evaluated on the basis of speed of
response, size and weight, power requirements, type of reagent required
(gases, liquid), sensitivity to vibration, and safety in airborne opera-
tion. The constraints of airborne operation are such that techniques
not recommended for the ground-based monitoring network may prove to be
the most suitable for airborne measurements.
A preliminary recommendation of air quality monitors, based on the
constraints of airborne service, and subject to a more detailed examina-
tion in the evaluation research project would be as follows:
Total sulfur gases - flame photometric detection
Carbon monoxide - mercury displacement
Ozone - ultraviolet absorption.
The sulfur dioxide monitor based on flame photometric detection may
not be feasible due to safety considerations of an open flame during
flight. However, if the flame is extinguished during take-off and land-
ings, this may be an acceptable procedure. The use of an on-board hydro-
gen generator with a limited production capacity could be substituted
for a compressed hydrogen cylinder. The stability of the flame during
airborne operation should be evaluated to determine if adequate detec-
tion sensitivity is maintained. A second choice for sulfur gases would
be a microcoulometric detection technique, which uses liquid reagents
instead of flame photometric detection.
The mercury displacement technique is recommended for the measure-
ment of carbon monoxide for several reasons. Through modification of
available instrumentation, a monitor of adequate sensitivity for rural
concentration of carbon monoxide can be obtained. In addition, size and
weight are suitable for airborne operation.
The ultraviolet absorption technique recommended for ozone is
primarily based on the elimination of the reagent ethylene as is required
by chemiluminescent ozone monitors. A second choice would be the electro-
mechanical technique for the measurement of ozones.
The airborne data acquisition system should record the measurement
by an on-board digital recorder in a format compatible with the central
data archives of the ground-based monitoring network. Calibration of
the equipment can be made on the ground prior to and after surveillance
flights.
XIV-13
-------
Balloon Tracking System
A balloon system designed primarily for meteorological observations
that appears to have especially high potential is known as METRAC and
previously known as DlLOSONDE. The METRAC system is currently a propri-
etary system under final development by the Control Data Corporation.
System Design Concept
METRAC employs Doppler tracking techniques to determine a highly
accurate three-dimensional position of an inexpensive radio transmitter.
This light weight transmitter, carried either on a rising or horizontally
floating small balloon, permits fine-scale measurements of the wind, and
as many as four measurements from additional sensors on the balloon.
Radio signals are received by a minimum of four portable ground stations,
and the information is relayed by radio or landline communication to a
central data terminal. The data are then recorded for future use or
processed immediately by a small computer.
Primary METRAC advantages are low cost and high accuracy. Flight
packages are expendable, as in radiosondes, and the cost is comparable
or less. Ground station designs are much less complex than the usual
instrumented towers, rawinsonde sounding system, or radar-positioning
system. The master station is fully automated and the only operator
requirement is starting the receivers before launch. As data generated
by the master station are formatted for direct computer processing, no
intermediate processing preparation is required. A small computer can
process an entire profile in less than five minutes. The high accuracy
in three-dimensional positioning permits detailed calculation of resultant
winds. This high resolution of data in time and space thus enables the
determination of small-scale features of vertical or horizontal profiles.
Several other advantages are basic to the system. By using a network of
these systems, the inherent accuracy and range is limited only by the
boundary of the network. Thus tracking over hundreds of kilometers is
possible as long as at least four receivers remain in radio line-of-sight.
Observations can be taken day or night in all weather conditions. Simul-
taneous tracking of multiple balloons is possible with special modifica-
tion of the system. Finally, the flexibility provided by the multi sensor
capability should not be underestimated. Simultaneous point measurements
(gradient values if more than one balloon is tracked) of wind, tempera-
ture, humidity, pressure, and net radiation can be obtained at precisely
determined locations.
XIV-14
-------
The determination of space coordinates at any instant of time is
based upon the system's ability to monitor differential Doppler shifts
between the received frequencies at the various ground-based receivers.
At each ground station, a receiver translates both the NBFM telemetry
information and its Doppler frequency shift from a radio frequency car-
rier to an aUdio-frequency carrier signal, which comes unmodulated from
the central data terminal. This data-carrying signal from each receiver
is then transmitted to the central data terminal via a voice channel
communications link.
While a minimum of four ground stations is required, it is recom-
mended that seven stations, arranged in a hexagonal pattern with one in
the center, be used in order to provide some redundancy and thereby
reduce the vulnerability of the system to equipment failures or multi-
path and other propagation effects. Use of seven stations allows the
calculation of 35 coordinate positions for the mobile transmitter at any
given instant. Should propagation problems affect the signal at one
ground station, a total of 15 coordinate positions is available; whereas
if the signals at two of the ground stations are affected, a total of
five coordinate positions is still available.
At the data terminal, the signal from each ground station is sampled
a minimum of 50 times per second. This high rate is necessary to recover
the telemetry data even though the modulation deviation is low. Each
balloon-borne sensor is sampled once each second; however, 50 positions
per second may be computed if this fine a time scale is desired. Ordi-
narily, and especially in on-line computing, it is not desirable to
obtain position information at this rate, and digital filtering of the
data may be advantageous.
Indications are that the use of a six-balloon METRAC system could
provide six l5-second averages of vector winds every 15 seconds with a
resolution of better than 0.1 m/sec. Winds for a one-second average
(single balloon) can be resolved to 0.3 m/sec. Therefore, the system
is capable of resolving the vector velocity of the balloon to a degree
equal to or better than the representativeness of the balloon in describ-
ing the ambient wind fields. As a comparison of alternate optical and
radio tracking techniques, the following table from the HMSO Handbook
of Meteorological Instruments (1961) is presented below. (Table XIV-5)
The ground-based receivers are housed in light-weight weatherproof
packages and can be mounted on a 10-foot pole that is used to support
the receiver antenna system. The pole itself is sectioned for easy
portability and is supported by a guy system that also functions as a
reflecting screen for the antenna. The antennas are simple and stationary
since the use of directional arrays is not required.
XIV-15
-------
Table XIV-5
VECTOR ERROR OF WIND (m/sec) AT 5 km AS A FUNCTION OF
TRACKING SYSTEM AND RATIO (Q) OF MEAN WIND TO ASCENT RATE
Q A B C D
1 0.5 0.5 0.5 1.0
2 0.5 1.0 1.0 2.0
3 0.5 2.0 1.5 3.6
4 1.0 3.6 2.5 5.1
5 1.0 5.6 3.6 7.1
A - Radar
B - Optical or radio-theodolite and radiosonde height
C - Double theodolite or direction finder
D - Optical or radio-theodolite and assumed ascent rate.
Simultaneous tracking of more than one balloon requires equipment to
be added at both the receiver and data terminal sites in addition to the
use of separate transmitting frequencies for the mobile transmitters.
Each receiver will require an additional IF unit for each channel used
as well as a multiplex unit for the data link. A multiplex adapter will
also be needed at the data terminal unit in addition to either a high
speed incremental tape unit or a high speed data transmission link to a
computer.
Apart from the high-resolution features of METRAC, the flexibility
of the system is particularly attractive. For example, it would be
extremely valuable to have the capability to collect simultaneously five
or six vertical soundings throughout the study region at, say, a one-hour
periodicity during intensive model test periods or, say. four-to-six per
day (perhaps at fewer locations) on a routine basis. The coincidental
measurement of wind trajectories over the region would also provide neces-
sary meteorological data. With a system such as METRAC, both programs
could be undertaken simultaneously or perhaps the vertical soundings could
be interspersed in the horizontal tracking program.
XIV-16
-------
Activation Schedule
The basic METRAC system has been tested on a limited basis but the
system developers have estimated that an additional six-month preliminary
design and test period is required for implementation as an operational
system. This program would evaluate and resolve design criteria for
particular space and time scales (resolution), available radio frequen-
cies, and sensor requirements.
If the METRAC proves successful after the six-month test period, it
is estimated that one year will be required to finalize software and sta-
tion design, fabricate the system, and complete installation and accep-
tance tests. Thus, a total of 18 to 24 months is estimated subsequent
to authorization of the Regional Study to bring the METRAC system to full
operational status.
Costs
The costs of the METRAC system will vary to a limited extent with
the number of ground stations and the number of balloons which the system
can simultaneously track. For example, a single balloon tracking capa-
bility having six remote stations and one central station is estimated at
$395,000, whereas a six-balloon tracking capability with 18 remote and
one central station is estimated at $476,000. Both estimated costs
include final development costs of $100,000. Table XIV-6 provides these
estimated costs in greater detail. For the St. Louis region approximately
18 slave stations would be required to adequately track up to six balloon
instrument packages simultaneously within a 350-km radius. Remote sta-
tion costs for a single-balloon system would be approximately one-half
the six-balloon system.
The METRAC system operating costs have been estimated at approximately
$8000 per balloon release location per month for the weighted manpower
costs of an intensive round-the-clock operation with all data processing
completed in real time. Thi s cost compares favorably with the NOAA
radar-tracking tetroon system, for example, which is estimated to incur
costs up to $100,000, for a one-month intensive period of operations and
subsequent data processing.
XIV-17
-------
Table XIV-6
ESTIMATED INITIAL COSTS OF THE METRAC SIX-BALLOON SYSTEM
(Thousands of Dollars)
Preliminary design and test
$100.00
Final design, fabrication, installation
250.00
Central station
50.00
Remote stations at $2
36.00
Computer
40.00
Total
$476.00
Special Aircraft-Based Meteorological Observations
General Concepts
While the proposed balloon-tracking system will provide upper-air
meteorological observations on a routine basis as well as in conjunction
with some special programs, an aircraft-based meteorological observing
system is also recommended for some of the special problems. In partic-
ular, the use of aircraft is desirable and almost necessary for studies
of urban-rural radiation differences and basic studies of large-scale
atmospheric turbulence structure in relation to regional, upper-level
diffusion processes.
The intermittent sequencing of these special programs might require
the use of several different aircraft. Existing operational systems
maintained by other agencies, institutions, or firms might be brought in
as necessary on either a cooperative or contractual basis. As an example,
the flight facility at NCAR might be engaged through a cooperative arange-
ment with their FAPS program (see Chapter VIII); or the aerial capabilities
of NOAA or various university and private institutions could be solicited
on a contractual basis. Alternately, WERL may recognize the task under
their charter of responsibility and seek to expand their capabilities to
undertake these tasks. Above all, the choice should be based on the
ability of the party to provide the necessary meteorological data with a
satisfactory resolution and processing time, as well as their ability to
provide aircraft of sufficient size to accommodate air quality instrumen-
tation.
XIV-18
-------
Vertical and horizontal profiles of the mean wind are essential.
Horizontal profiles should be obtainable on a distance scale on the order
of five kilometers; vertical profiles to 1500 meters above ground level
are anticipated. As part of the tasks of the detailed evaluation of
gradient transfer diffusion models, turbulence measurements may also be
required to compute eddy diffusi vi ty values. If this is a major require-
ment, an inertial guidance system will be necessary rather than the more
conventional Doppler system. Ambient air temperatures will be necessary
for several programs; horizontal mapping of mean temperatures at several
altitudes will be necessary for heat island and radiation studies, while
high-frequency measurements will provide input to sensible heat flux com-
putations for heat budget and gradient transfer studies. Dew point
measurements are also advantageous for the aforementioned studies as well
as investigations of urban precipitation modification and tall-stack plume
rise phenomena. Radiation measurements that are required have been out-
lined in the earlier discussions of both meteorological and photochemical
research tasks. Measurements should be made of the outgoing longwave
radiation in the atmospheric "window" to permit mapping of the surface,
radiative temperature. Space-averaged values from a narrow-field bolom-
eter should be used to obtain the desired spatial resolution. Precision
observations of the broad spectrum incoming direct and diffuse and the
outgoing total (diffuse) shortwave radiation should also be made. Addi-
tionally, the capability for making spectral radiation measurements and
collecting particle samples is also desirable.
In summary, there are three capabilities which a fixed-wing airborne
system has over either rotary aircraft or balloon-tracking systems. First,
airplanes provide a platform for making difficult turbulence and radia-
tion measurements. Secondly, balloons may be capable of providing some
of these data, but not with the positional flexibility of the airplane,
and helicopters may have greater positional flexibility, but they cannot
be effectively utilized for precision radiation measurements and can
provide virtually no wind data. .Thus, airplanes have a common "radiation
advantage" and some "wind advantages" over the two alternatives, while
providing positional flexibility only in comparison with balloons. These
advantages are sufficient to warrant limited use of a fixed-wing system,
but probably not on an extended basis.
Instrumentation Considerations
There is a long history and extensive literature concerning the use
of aircraft as meteorological platforms. No attempt is made here to sum-
marize this field. We note, however, that continuing aircraft programs
XIV-I9
-------
exist at NCAR, at the research flight facility of NOAA, and elsewhere.
Improved instruments are being developed by these groups.
The main problem in measuring meteorological variables from a moving
aircraft is that conditions at sensor locations close to the aircraft are
generally disturbed and are different from ambient conditions. Thus cor-
rections must be made, or nose booms may be used as sensor platforms.
For the Buffalo aircraft used at NCAR, the nose boom is five meters in
length with a modified conical shape. It consists of an inner stainless
steel circular boom or probe that is directly aligned with the inertial
system, and an outer fiberglass shell which provides the airfoil shape
as well as the housing for some of the other sensors on the craft.
Temperature can be measured with an absolute accuracy of approxi-
mately O.50F and a relative accuracy of O.loF. A well-known instrument
is the commercial Rosemount probe. It is adequate for the low-frequency
temperature measurements that need to be made for mesoscale analysis.
Two other thermometers have recently been developed at NCAR. They are
a reverse flow thermometer and a tungsten fine wire resistance thermom-
eter. The latter gives very high resolution data for turbulence studies.
Infrared thermometers are also being built. The Rosemount instrument is
probably adequate for the st. Louis study.
For measuring atmospheric water vapor, aircraft dew point hygrometers
are available from EG&G, Waltham, Massachusetts, that are accurate to 1°F,
and sensitive to O.loF. This type of instrument should be quite suitable
for the present use. For faster response data, a microwave refractometer
is used at NCAR.
Pressure can be measured adequately using the aircraft altimeter or
a special transducer. To measure horizontal pressure gradients, however,
simultaneous pressure and radar altitudes must be recorded, and the height
of the underlying terrain must also be known. For the relatively flat
terrain of the st. Louis area, it should be possible to measure horizontal
pressure gradients (geostrophic winds) with useful accuracy. It appears
too that a measurement of the vertical pressure gradient can be obtained
by comparing the radar altimeter readings with those of the barometric
altimeter along a vertical axis.
The measurement of wind from an aircraft is difficult because the
wind is deduced as a relatively small vector difference between two
measured vectors: the ground speed vector (along the true course) and
the true airspeed vector (along the true heading). Both Doppler and
inertial navigation systems measure the ground speed vector and the drift
angle. The main uncertainty in using either system for measuring winds
is in the length of the true airspeed vector, which is difficult to
XIV-20
-------
measure with the same high precision achieved in measuring ground speed.
The error in true airspeed causes a small uncertainty in the measured
wind vector. For the NCAR Buffalo aircraft, a special true airspeed
system has been installed to maximize accuracy. Such a system may add
a precision of 1 to 2 m sec-l to the true airspeed measurement. The
NCAR system uses a set of vanes for the alignment of the pitot pressure
sensor with the relative wind. This serves to decrease the error in
pitot pressure introduced by side-slip of the aircraft. True airspeed
is computed from pitot and static pressure, and temperature.
The Doppler and inertial systems have some significant differences
which should be noted. The inertial system, built around accelerometers
along three axes, is capable of very high response that makes it espe-
cially appropriate for use in measuring turbulence. The Doppler system
is of much slower response and gives integrated values of ground speed
over approximately 10 to 20 seconds. In contrast to the inertial navi-
gator, it does not give a measurement of speed along a vertical axis.
According to Dr. Donald Lenchow of NCAR, the inertial system (with the
special true airspeed measurement) gives winds accurate to 1 m sec-l,
while a Doppler system has a comparable accuracy of 2 m sec-l. Both of
these values are better than the accuracy of standard rawinsonde measure-
ments of winds. On the other hand, some experienced personnel at SRI
do not agree that inertial systems are inherently more accurate than
Doppler systems.
In further consideration of the two systems, the following opera-
tional aspects should be considered. The inertial system requires a
15-minute alignment time after takeoff, thereby increasing the total
length of time of a projected flight. However, for operational considera-
tions this l5-minute delay would probably not mean very much. The in-
ertial equipment is more difficult to maintain than the Doppler; there-
fore two (or three) independent inertial units are normally combined in
a single system for use on commercial aircraft. The inertial system
cannot be turned on and off during flight unless the aircraft returns
to a fixed position and initiates a second alignment procedure. The
inertial system, however, is capable of operation at any height while
the Doppler system will not operate at extremely low altitudes. This
lower limit is on the order of 200 feet.
It appears that this potential height limitation in the Doppler
system will not be of any serious consequence for its consideration in
the Regional Study Program. The output from both the inertial and Doppler
systems is of the following: position, ground speed, drift angle, true
heading, and track.
XIV~21
-------
In the inertial system these items are routinely determined once
per second although higher resolution is possible. The inertial system
comes with its own computer and display consoles and, with the measure-
ment of true air speed, will also provide an output of the two horizontal
components of the wind. Comparable components are available for Doppler
equipment. The inertial system also provides one second output of verti-
cal acceleration and, upon integration with respect to time, will provide
the vertical component of the wind. In practice the vertical wind com-
ponent is determined with the inertial system for frequencies less than
100 seconds and for the longer term mean wind (that is frequencies less
than 1 every 100 seconds) a pressure altimeter measurement is used. To
display the wind components an auxiliary computer is necessary. The
computer must accept readings of the pitot, static pressures, and tem-
perature to compute true air speed, as mentioned previously. This re-
quirement for a computer is in reality of no great significance as it
will always be highly desirable to have an onboard computer to preprocess
the meteorological and air quality data being collected on the aircraft.
The accuracy of the wind measurement with an inertial system is improved
by use of frequent fixes on ground navigation points because of the so-
called Schuler period of 84 minutes over which the inertial system has
a sinusoidal drift. This drift, if uncorrected, will not permit the
accuracy levels mentioned previously to be met.
In considering the relative merits of
tion equipment for the Regional Study, the
should be borne in mind:
Doppler and inertial naviga-
following considerations
(1)
The accuracy of the systems is approximately equivalent,
but some authorities favor one system over the other.
(2)
The inertial is a rapid response system and is much more
suitable for turbulence measurements, if these are needed.
(3)
Both systems are small and compact.
(4)
The costs of an inertial system are approximately double those
of Doppler equipment.
(5)
The inertial system requires alignment at the beginning of a
flight.
If it is decided that the aircraft should be capable of providing
both average and turbulent wind conditions along the flight path, then
selection of inertial equipment is favored. However, if high frequency
data are not needed routinely, the decision should not be made without a
careful engineering evaluation of the factors mentioned above.
XIV-22
-------
In meteorological flights made with the NCAR Buffalo aircraft, alti-
tudes as low as 500 feet have been followed at speeds between 140 and
160 knots. This range provides a safe flying speed as well as maximum
resolution in the wind and temperature measurements. That is, from a
safety standpoint, it is desirable to fly above a minimum speed, while
for the measurement of high frequency meteorological parameters and in
consideration of the sources for measurement errors it is desirable to
fly at the lowest possible speed. The indicated speed range represents
a compromise on the two considerations.
In using aircraft to measure vertical profiles of meteorological
variables, they can be flown in spirals with a 3600 turn in two or three
minutes, and a rate of climb (or descent) of approximately 1000 feet per
minute. The diameter of the spiral would be approximately three to five
miles. The Doppler system can accommodate this type of maneuver; and the
inertial system will function in even tighter turns.
Projected long-range flights by the aircraft for monitoring around
the Regional Study area will probably have to encompass a circular grid
having a diameter of approximately 120 kilometers. Therefore, if one
defines a relatively square pattern to be flown about the area, each leg
of the flight would probably have a length on the order of 80 nautical
miles. It will be desirable to make routine flights at a minimum of two
different altitudes within as short a time period as possible. It will
also be necessary to make vertical profile measurements with the aircraft
at different times and at different locations in the grid.
The data, in summary, that would result from such a flight program
would include spiral measurements (vertical profiles) at upwind and down-
wind locations on the grid as well as the measurement of the vertical and
horizontal wind structure around the perimeter of the grid at one or more
heights. The program will also provide some input on temporal variations
in the wind flow field. Such variations will further be indicated by
frequent (perhaps three-hourly) low level balloon measurements within the
grid.
Various flight plans will be desirable for special programs. One
purpose might be to sample along the general flow direction at several
altitudes, with the path extending both upwind and downwind from pollu-
tion sources. A flight which starts at a given altitude on one side of
the city could be made to the opposite point where the altitude is
changed, and the route is traversed in the opposite direction. This
would provide data for cross-sectional analysis; for example, along the
axis of a plume of effluent. Other flight programs will be dictated by
the requirements of the special research studies.
XIV-23
-------
Special Aircraft-Based Air Quality Observations
The fixed-wing airborne laboratory recommended for special meteor-
ological observations can serve as a vehicle for nonroutine air quality
measurement of nitrogen oxides, aerosol particles size distribution, and
total particulates. The recommended monitors for the nonroutine air
quality measurements should be quite suitable for airborne operation and
are as follows:
Nitrogen oxides - chemiluminescence
Particle size distribution - Royco-type particle counter
Total particulate - filter collection.
If space limitations aboard the fixed-wing aircraft are not severe,
the entire air quality instrument package used in the helicopter could
be installed on a temporary basis in fixed-wing aircraft in addition to
the nonroutine air quality instrumentation. The data acquisition system
used aboard the NCAR Buffalo Aircraft, for example, may not be compatible
with the RAPS system. However, the infrequency of flights would make
the manual transfer or transcribing of data to the appropriate format
a feasible procedure.
Western Environmental Research Laboratory (WERL)
The current and planned airborne monitoring capability of the WERL
represents an important potential resource for the Regional Study. WERL
currently operates several aircraft instrumented for detecting and
tracking airborne and ground-deposited radioactive particles. The air-
craft are also fitted with a variety of sampling systems which have
application to a broad range of radioactive and nonradioactive air pol-
lutants. A staff with up to ten years' experience in aerial surveillance
provides personnel with the ability to perform diverse functions in this
field. Current objectives include (1) acquisition of additional equip-
ment and personnel to provide aerial surveillance services, (2) provision
of aerial surveillance of air, water, and terrestrial pollution and pol-
lution effects to requesting offices in EPA, including collection,
interpretation, and reporting of data, and (3) maintenance of equipment
and personnel capability to respond to environmental crises which might
occur any place in the nation.
WERL currently plans to make available to the Regional Study, as
required, one fixed-wing aircraft, one helicopter, and appropriate
XIV-24
-------
operational and maintenance staff. WERL aircraft wil~ eventually include
the following:
Two each Mohawk aircraft (A)-OVIB and (B)-OVIC*
One T-34
.
One C-45
One C-123
.
Two each Huey Model P, II-place helicopters (200-minute flight
time with 15-minute reserve time with speeds up to 140 knots).
The C-123 will be able to transport an experimental team and mobile
laboratory to distant sites, if required. ESD's analytical and sensing
hardware is at present limited to particulate sampling, radiological
monitoring, and aerial photographic equipment.
The ESD staff foresees a need to be prepared for a full range of
environmental problems and to offer support to other EPA divisions in
pollution-control monitoring, enforcement, and research, and is beginning
to assemble the remote-sensing, sampling, analytical, and information-
processing equipment to perform this broad range of functions. The
Regional Study will provide an opportunity to assemble an air quality
surveillance system that is based on demonstrated needs as defined by
the upper air meteorological and air quality research plans. The system
will be composed of subsystems that have specific functions and require-
ments. The surveillance system functions will include the following:
(1)
Selection of environmental parameters
(2)
Selection of surveillance techniques
(3)
Data collection
(4)
Data recording, processing, and analysis
(5)
Dissemination to project personnel and other users.
*
Two-place Army observation
payload, top speed 300 mph
external tanks-1400 miles.
plane. Two turboprop engines, 2500-pound
at 4000-foot altitude. Maximum range with
XIV-25
-------
Effective execution of each of these functions will depend on a number
of elements, including skilled personnel, appropriate technology, and
efficient organization. This set of interrelated functions must be car-
ried out in a unified, balanced manner if the system is to be productive
in terms of project objectives.
A generalized and simplified picture of systems interrelationships
in the aerial surveillance program is presented by Figure XIV-3, in which
the approach to air quality surveillance is outlined.
Selection of Parameters
Selection of the type of data to be collected is critical to any
program that aims to improve project operations in a significant way. It
is important to choose parameters that are directly related to the goals
to be attained. It is also necessary to consider surveillance techniques
in selecting parameters. Deciding on where and when to sample or measure
and what quantity of data is desirable are critical questions that are
inextricably tied to the design of the data-collection subsystems. The
desired parameters with respect to meteorology and air quality are dis-
cussed in Chapters XI and XII, as well as throughout Part II, and provide
the key factors for initiation of systems integration.
Selection of Surveillance Techniques
This is the subtask that encompasses the hardware
study. The approach will be to consider the following
hardware evaluation and selection:
aspects of the
three aspects of
(1 )
Identify the surveillance functions that can be fulfilled
with ESD's present stock of instruments
(2)
Review the candidate surveillance instrumentation, with a
view to acquiring the specific hardware suitable for ESD's
functions in the Regional Study
(3)
Identify the more important surveillance applications.
Project personnel will probably
field work or outside contacts other
Las Vegas and other laboratories.
perform these activities without
than liaison with EPA personnel in
XIV-26
-------
r---------
I
I
I
I
I
I
I
I
I
RECOMMEND
REVISION OF
CRITERIA AND
PROCEDURES
EVALUATE
EFFECTIVENESS
OPERATION OF
SURVEI LLANCE
SYSTEM
PARAMETERS
NUMBER OF
SAMPLING POINTS
DEFINING DATA CHOOSE DATA
NEEDS FOR
REFINING QUANTITY ACQUISITION
PROCEDURES OF DATA SYSTEMS
::><
H
<:
I IDENTI FY
l:\:I EVALUATE FREQUENCY DESIGN DATA
-..J SURVEILLANCE
CAPABILITIES OF SAMPLING TRANSMISSION
CAPABI LlTI ES AND PROCESSING
SYSTEMS DESI GN
INTEGRATED
NUMBER SURVEI LLANCE
IDENTIFY OF USERS SYSTEM
PROJECT USERS FORMULATE
DATA ANALYSIS
PROCEDURES
DEFINE DATA OUTPUT FORMAT
NEEDS AND USES
DESIGN DATA
IDENTIFY DISSEMINATION
OTHER USERS ALLOWABLE AND FORMAT
TRANSMISSION
DELAY
SA-1365-32
FIGURE XIV-3
AERIAL SURVEILLANCE SYSTEMS ANALYSIS CONCEPT
-------
(1 )
Upper air meteorology
(2)
Relation of individual pollutant sources to substandard air
quality conditions in regions characterized by multiple
sources or diffuse emission
(3 )
Surveillance of geographical areas difficult to reach by sur-
face monitors
( 4)
Surveillance of mobile sources
(5)
Impact of pollutants on agricultural lands, forests, and grass
lands.
The emphasis is on measurements in which rapid response, mobility,
and/or a synoptic view are key elements. These requirements also must
be respected in designing subsystems for data recording and information
processing and analysis, as discussed in subsequent sections.
The overall capability of the particular subsystem design for the
aerial surveillance function will depend on the sensor/aircraft inter-
action as well as the sensor's characteristics per see The subsystem
capabilities may be divided into the following general categories:
(1)
Capabilities defined by the particular sensor
(2)
Capabilities defined by the sensor platform only (sensitivity
of altitude and altitude control)
(3 )
Capabilities defined by both the sensor and platform (the
ground resolution as determined by the angular resolution of
the sensor and the altitude of the aircraft)
(4 )
Suitability for both day and night operation under a variety
of meteorological conditions.
Preliminary discussions with ESD personnel have indicated the
availability, now or in the near future, of the following types of hard-
ware:
.
Photographic equipment--70-millimeter
cameras, four Hasselbad multispectral
gyrostabilized camera
forward and 5-inch vertical
cameras, and a 9-inch
Side-looking radar
XIV-28
-------
.
Infrared scanning system
Accelerometer
Nephelometer
Vortex thermometer
Black-body references for infrared system
PRT-5 Barnes infrared thermometer (narrow view)
Turbulence instrumentation
Air quality instrumentation--ozone, sulfur dioxide, total hydro-
carbon, and condensation nuclei instruments.
One of the turbojet aircraft is equipped with a radio altimeter and
a Doppler guidance system. ESD is now in the process of incorporating
the Doppler guidance system into an airborne wind measuring system. The
group has also looked at the feasibility of acquiring inertial guidance
systems to replace the Doppler systems, and they may proceed along these
lines if sufficient needs develop. The current Doppler system, which is
being implemented or incorporated into a wind measuring capability, has
the ability to resolve aircraft drift angle to within 0.1 degree and wind
speed to within x 2 or 3 knots.
Based on work at the NCAR, it appears that the inertial system
provides greater accuracy. The accuracy of the inertial system is limited
6nly by the resolution with which the air speed of the aircraft can be
measured; furthermore, this will be a function of the type of aircraft
being used. The inertial system is capable of resolving the mean wind-
speed components (u, v, w) to approximately one meter per second. Greater
accuracy can be obtained but the one meter per second value is a con-
servative estimate. Wind-speed fluctuations can be obtained with a
frequency of 100 per second and can be resolved to 10 cm per second.
comparison, the Doppler type system indicates that the mean winds can
be resolved only to within a factor of two meters per second. If
this is the case, it appears that the inertial system may be given a
higher priority in spite of the larger initial cost.
In
It is estimated that an inertial guidance system alone (at commercial
prices) will cost in the neighborhood of $100,000 to $150,000; the pur-
chase and fabrication of the true air speed measuring equipment will cost
approximately $25,000, while the recording system and miscellaneous
XIV-29
-------
sensors should cost in the neighborhood of an additional $10,000. The
overall cost of the system, therefore, will be between $150,000 and
$200,000. The Doppler system, on the other hand, would cost approximately
$75,000 to $125,000 less than the inertial system.
Assuming the continued use of the acquired instrumentation for
WERL's broader mission for EPA, these costs should be only partially
allocated to the Regional Study. The details for funding WERL's activi-
ties within the Study will be negotiated by interagency agreement under
the overall memorandum of understanding.
With respect to service facilities for the fixed-wing and rotary
aircraft in the St. Louis area, WERL has an arrangement with the Air
Force to utilize their facilities throughout the country. Therefore,
they expect to be able to base their aircraft at Scott Air Force Base,
Belleville, Illinois. WERL also expects to have available the C123 air-
craft to be used both for transport and as an instrumentation platform,
and the C-45 aircraft with a capability of carrying eight persons or a
l500-pound payload. Final selection of aircraft and instrumentation
should be accomplished by the end of the first year of the Study.
XIV-30
-------
Chapte:", XV
GENERAL CLIMATOLOGY OF ST. LOUIS
Introduction
Seasonal climate, and even the daily weather, of an area essentially
is conditioned by the character of the prevailing air mass or the inter-
action of air masses over that region. An air mass can be defined as a
large body of air (approximately 1000 miles or greater in horizontal ex-
tent) which takes on the particular temperature and humidity properties
of the land or sea area (source region) over which it resides--the cold-
ness of polar regions, the heat of the tropics, the moisture of oceans,
and the dryness of continents. Because of the uniform distribution of
temperature and humidity properties, the weather over an area associated
with a given air mass tends to be quite similar, with local variations
due to geographical features. The four basic types of air mass can be
characterized as "cold moist," "cold dry," "warm moist," or "warm dry"
depending on the geography of the source region (polar or tropical) and
the underlying surface, maritime (water) or continental (land). Air
masses over a given source region vary in their properties of tempera-
ture and moisture from season to season, as the properties of the source
region undergo seasonal change.
These air masses are exported from their source region by the action
of large pressure systems--cyclonic storms and anticyclones, and it is the
interaction of these air masses that produces weather--fair or foul. The
midlatitude of the United States is a favored locale for these interactions.
The area of eastern Missouri and southern Illinois, by virtue of its
location, is subject to a variety of weather, as well as rapid changes in
weather resulting from the intrusions and interplay of cold, rather dry
air moving south from Canada and the Great Plains, as well as warm moist
air moving northward from the Gulf of Mexico (see Figure XV-I).
St. Louis and environs, which is approximately 400-800 ft above MSL,
is characterized by flat to gently rolling terrain and is free of large
bodies of water; therefore, there are no major physiographic influences
on these air masses. Solar change during the year is pronounced and does
contribute to the seasonal climate through the amount of radiation received,
XV-l
-------
:;>'
o
~~(
'I'\~({(\
FIGURE XV-l
...~
. ..~
COOl ... ~J,1... .
. . ~I)(y 'QI6ry..
~(J St. .~\9:
""kt. Louis ...>('.
SA-1365-20
TYPICAL PATHS OF AIR MASSES
XV-2
-------
e.g., the height of the noonday sun changes about 47° between solstices;
the length of the shortest day (Dec. 22) is 9 hr 26 min, and the longest
day (June 21) is 14 hr 53 min. Selected climatological characteristics
are summarized by month in Table XV-I.
Principal Seasonal Meteorological Characteristics
Cold Season (Mid-October to Mid-April)
Between mid-October and mid-April, the area is subject to frequent
changes in the prevailing air masses (Figure XV-2). Intrusions of cold
air from the north (which may drop the temperature some 20° in a 24-hr
period) are not uncommon but such cold spells are not long-lasting--
generally much less than a week. By and large, winter is characterized
by mild temperatures (seasonal average maximum temperature) with occa-
sional short spells of extreme cold. The rapidity of the weather changes
is due to the rapid movement of the cyclones. The warm inflow from the
south in front of a depression is of almost as pronounced a character as
the cold wave which suddenly takes its place when the trough of the de-
pression has passed. Cold waves are coldest in the north of the United
States. In the mid-course of the Mississippi the changes are especially
noteworthy since the temperature often drops suddenly from well above,
to well below, freezing-point, and such a change entails the maximum of
inconvenience and discomfort. St. Louis has recorded 74°F and -22°F in
January, that is to say, 42° above, and 54° below, freezing-point.
Some indication of the severity of the winter is given by the fact
°
that about 70% of the January nights have a minimum of 32 F or lower and
about 25% of the January days are continuously below freezing; somewhat
smaller percentages prevail in December and February. Rapid warming
. °
takes place by March. The average date of the flrst 32 F temperature
. °
observed is about mid-October, the last 32 F temperature is observed
about the beginning of April. Between October and April periods of warm
tempera~ures extending into the high 80s are observed but the incidence
is small.
Precipitation can be expected during all months in the period, largely
in association with the passage of cyclonic storms or fronts. Generally,
such storms move in a southwest to northeast path over the area and pri-
marily during the winter and spring months. Such passages will bring
precipitation in the form of rain, although periods of snow, sleet, or
freezing rain are not uncommon. Post-storm conditions typically show
fair skies with moderate to brisk surface winds. Typical tracks of low
and high pressure areas are shown in Figure XV-3.
X~3
-------
Table XV-l
SELECTED CL1MATOLOGICAL CHARACTERISTICS OF ST. LOUIS
STA NO. 72434 (IN AREA NUMBER 13) LATITUDE 3844N LONGITUDE 09021W ELEVATION(FT) 00571
POR NO.
PARAMETER DESCRIPTION JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC ANN (YRS) OBS
ABS MAX TMP (F) 77 78 86 91 98 105 115 104 104 94 86 76 115 23 -113
MEAN MAX TMP (F) 40 44 53 66 75 85 89 87 81 70 54 43 66 30 -116
MEAN MIN TMP (F) 24 25 32 44 53 63 67 66 58 47 35 27 45 30 -116
ABS MIN TMP (F) -14 -7 -5 24 33 47 52 50 32 22 3 -4 -14 23 -113
MEAN NO DYS TMP = OR GTR 90 (F) 0.0 0.0 0.0 0.3 2.0 11.0 16.0 16.0 7.0 0.3 0.0 0.0 52.6 10 -113
MEAN NO DYS TMP = OR LES 32 (F) 26.0 19.0 17.0 2.0 0.0 0.0 0.0 0.0 0.0 1.0 13.0 22.0 100.0 10 -113
MEAN NO DYS TMP = OR LES 0 (F) 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 23 -29
MEAN DEW PT TMP (F) 21 23 30 39 49 60 65 64 58 44 34 26 43 22 -29
MEAN REL HUM (PCT) 68 66 64 60 63 66 68 68 71 63 69 72 67 6 -116
:><: MEAN PRESS ALT (FT) 386 392 470 501 529 537 513 508 459 431 421 392 462 0 -50
<: MEAN PRECIP (IN) 1. 98 2.04 3.08 3.71 3.73 4.29 3.30 3.02 2.76 2.86 2.57 1. 97 35.3 30 -116
J. MEAN SNOW FALL (IN) 4.4 4.1 4.8 0.1 0.0 0.0 0.0 0.0 0.0 0.0 1.2 2.9 17.5 30 -116
~lliAN NO DYS PRCP = OR GTR 0.1 IN 4.6 4.7 6.2 6.6 6.7 7.1 6.0 5.6 4.7 4.8 4.4 4.6 66.0 30 -29
MEAN NO DYS SNFL = OR GTR 1.5 IN 1.0 0.9 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.3 0.7 3.9 30 -29
MEAN NO DYS W/OCUR VSBY LES 1/2 MI
MEAN NO DYS TSTMS 1.0 1.0 3.0 5.0 7.0 9.0 7.0 7.0 5.0 3.0 1.0 0.0 49.0 59 -24
Source:
U.S. Naval Weather Service World-Wide Airfield Summaries, Volume VIII, Part 5, United States of America (Mississippi Valley Area),
December 1969.
-------
SA-1365-17
FIGURE XV-2
MONTHLY DISTRIBUTION OF PREVAILING FLOW PATTERNS
CENTRAL U.S.
XV-5
-------
GENERALIZED TRACKS
TAKEN BY MANY
HIGHS OR ANTICYCLONES
TRACKS TAKEN BY
MANY LOWS; WIDTH OF
LINE SUGGESTS
RELATIVE ABUNDANCE
FIGURE XV-3
SA-1365-18
GENERALIZED TRACKS OF HIGHS AND LOWS
XV-6
-------
The prepassage of the lows is often accompanied by extensive cloudi-
ness so that the weather experienced at St. Louis will be common to its
environs as well. Rainfall amounts vary from a little over 2 inches in
December to slightly over 3 inches in March. However, in any given year
the precipitation may vary as much as 1.2 to 1.5 times the seasonal long-
term average. Snowfall amounts range from 3 inches in December to a maxi-
mum of 4.5 inches in February. Continuous precipitation episodes rarely
last longer than 1 day. Copious precipitation does occur on occasion, for
example, precipitation ~0.01 inch occurs on an average of 8-11 days a month
in this season; precipitation ~0.50 inch on 1 to 2 days per month. The
normal annual daily rainfall probability is about 25%. Thunderstorms are
recorded on an average of 1 day per month December through February, 3
days during March and about 8 days in April, concomitant with the onset
of spring.
The mean sky cover in this season remains virtually the same in all
months (approximately 6/10 coverage), the average percentage of possible
sunshine varies from 44% in December to 59% in April. The average rela-
tive humidity during this season varies from 70-75%; however, the average
relative humidity between 1 AM and 7 AM is 77%; between 1 PM and 7 PM it
is about 66%. Fog is observed in all the cold months with maximum occur-
rence December through February.
The prevailing (most often observed) surface winds are from a north-
westerly direction in winter and early spring. The typical storm tracks
(Figure XV-3) are such that a bi-model frequency of directions (from the
south and from the northwest) appears in the wind roses from October
through April. Wind speeds are generally less than 10 mph on the aver-
age, with most of the wind speed ranging from 4-18 mph. Strong winds
(in excess of 30 mph) occur infrequently (less than 15% of the observa-
tions), but speeds reaching nearly 60 mph have been recorded in the
months December through February and a peak value of 82 mph in April.
All the high speeds were observed for winds from a southwesterly direc-
tion.
Figure XV-3 indicates that the central United States is a preferred
region for anticyclonic tracks. This, in turn, generated a rather high
incidence of stable thermal conditions over this area--primarily dur~ng
the night and early morning hours, concomitant with clear skies and light
winds. Inversions are reported on about 35% of the total reported obser-
vation hours in winter and 28% of the total hours in early spring.
Except for short periods, the weather in this season does not re-
strict airport activities. Airport observations indicate a ceiling above
1000 ft on about 85% of the observations; visibility is above one mile
X~7
-------
about 85% of the time. Taken in combination, a ceiling of less than
300 ft and/or a visibility of less than one mile occurs less than 6%
of the time.
Warm Season (Mid-April to Mid-November)
Beginning in late April and extending through about mid-October,
the area is dominated by warm and moisture-laden air from the Gulf of
Mexico (see Figure XV-2). Summers remain consistently warm (average
daytime temperatures in the 80s) with short spells of hot weather; hav-
ing temperatures over 100°F. The cold waves of winter have a summer
parallel in the hot waves of the south and east of the United States,
which are spells of hot weather with very moist air, brought by the
south and south-east winds which blow when an anticyclone is situated
off the east coast, and a low-pressure system lies over the Mississippi
valley. The moist heat is very enervating though the thermometer may
°
not rise above 100. It even causes many cases of heat stroke and
prostration. Considerable cooling takes place at night; however, mini-
°
mum temperatures of less than 32 F are not observed from May to Septem-
ber. Maximum temperatures in the high 90s begin to occur in May and
readings exceeding 1000F are observed in all summer months and into
September. Marked cooling takes place through October. The average
monthly temperatures between June and September fall within the range
° . . °
70-80 F, the average d1urnal range 1S about 22 .
The onset of spring brings with it spells of thunderstorms and humid
conditions. In fact, during late spring, summer, and early fall, most
of the precipitation is the result of thunderstorm activity. Such rain-
fall is usually of short duration, but may be of heavy intensity. Monthly
precipitation amounts average between 3 and 4 inches. An occurrence of
5 thunderstorms in April, increasing to 10 days in June, and receding
to 5 days in September has been recorded in the long-time averages. Put
another way, the normal number of thunderstorms is 17 in spring, 27 in
summer; and 9 in fall. In summer about 40% of the thunderstorms occur
between noon and 6 PM; about 24% occur between 6 PM and midnight. As
in winter; the range of precipitation for a given year can vary as much
as 1.5 times normal precipitation value. Hail is common in April and
May, averaging nearly 1 occurrence per month. Tornadoes are not uncom-
mon and can be expected during the months from May through September,
in particular. About 7 are reported annually and May normally has the
greatest frequency. The mean relative humidity in the period April to
October varies between 65 and 72%; the diurnal range is marked, averag-
ing 74% between 1 AM and 7 AM and decreasing to 50-58% in afternoon hours
1 PM to 7 PM.
X~8
-------
Summer is characterized by rather fair skies. The average sky cover
varies from an average of 6/10 in April to a minimum of slightly over
4/10 from July through October. Sunshine is recorded for 60% (April) to
72% (July) of the available sunshine hours. Light to moderate fog is
reported rather infrequently in summer and dense fog hardly at all.
Surface winds during the period May through October are primarily
from the south, at a mean speed of 7 to 8 mph. Occasional strong winds
occur, however, usually from southwesterly directions, and have been re-
ported as strong as 60 mph in April through August, with a maximum of
73 mph being recorded in September. Speeds generally reach a maximum
during the afternoon.
In summer there is a definite tendency for nocturnal stability and
daytime instability in the lower levels; inversions are reported on
28-35% of the total hours during spring and summer and increasing to
43% in the fall months. Of these occurrences, some 60-80% are reported
at 9 PM and 7-15% at 9 AM, illustrating the effect of nighttime cooling
and generally fair skies. Flying weather during this period is good.
Ceilings greater than 1000 ft and/or visibility greater than 1 mile are
reported in excess of 97% of the observations. Short periods of reduced
ceilings and visibilities will occur with the presence of thunderstorm
acti vi ty.
Meteorological Parameters Relating to Pollution
The meteorological parameters that, perhaps, contribute most to in-
cidents of air pollution are stability and wind speed. Therefore, the
climatology which delineated the simultaneous occurrence of these epi-
sodes--and other parameters in which these factors are involved--is of
special interest.
Stagnating Anticyclones
Over the Great Plains anticyclones generally are characterized by
strong stability (inversions) in the lower levels, clear skies (which,
through radiation loss, contribute to stability) and sluggish surface
circulation, particularly if the anticyclones remain over an area for
an appreciable amount of time. A study by Korshovevl places this period
1.
Julius Korshovev, "Climatology of Stagnating
Rocky Mountains, 1936-1965," National Center
trol, Cincinnati, Ohio (1967).
Anticyclones East of the
for Air Pollution Con-
XV-9
-------
at 4 days or more. His data indicate that the St. Louis area experienced
nine cases of stagnation lasting four days or more (for a total of 42
stagnation days) in a 30-year period; there were two cases where the
stagnation lasted seven days or more. Seasonally, the frequency shows
a minimum in winter and a maximum from April to June, followed by a sharp
decrease in July and a continuous increase to an absolute maximum in the
month of October. Hosler2 has also pointed out that, while "precise fre-
quency values of low-level inversions may not be obtained for a specific
location from wind speed and cloud cover data alone, isopleths of percent
frequencies of cloud cover and wind speed indicate that the general areas
having a higher frequency of inversions are also characterized by a higher
frequency of relatively clear nights with light winds and vice-versa."
His data show that the percent frequency of nighttime cloud cover of 3/10
or less is about 36% in winter, 44% in spring, 55% in summer, and 60% in
the fall. Likewise the percent frequency of nighttime wind speed 7 mph
or less is at a minimum (20%) in winter; 40% in spring; 65% in summer
and fall. Taken together, the occurrences of nighttime cloud cover
~3/l0 and wind speed ~7 mph follow the same trend; a minimum of 31% in
winter, 42% in spring, 62% in summer, and 60% in autumn. Specific data
on inversion frequency at St. Louis are given in Table XV-2. Hosler
Table XV-2
FREQUENCY OCCURRENCE OF TOTAL HOURS OF INVERSIONS
Winter Spring Summer Fall
Percent of Total Hours 35 28 35 43
Percent frequency (maximum
observed) 61 60 81 83
Percent cloud cover ~3/lO,
percent wind speed ~7 mph 30 40 58 60
Percent frequency at GMT
03(9PM LST) 63 60 82 81
l5(9AM LST) 24 11 7 13
2.
Charles Hosler, "Low Level Inversion in the Contiguous United States,"
Monthly Weather Review, 319-339, September 1961.
XV-10
-------
also remarks that
" .
the higher frequency during fall and winter probably is a
reflection of minimum storminess in fall and maximum length
of a stable nocturnal period in Ninter. The opposite is true
for the spring and summer months. During the colder months,
particularly if snow cover exists, warm southerly advection
over an existing cold surface may enhance low-level stability
in these areas; however; such advection regimes are usually
the incipient stages of warm fronts and cyclones with their
associated precipitation and storminess, which ultimately
produce less stable lapse rates below 500-foot elevations
above ground."
A parameter that expresses the potential for vertical mixing of the
air is the "mean maximum mixing depths." It is apparent from Hosler's
data that at night the mean extent of vertical mixing over the Great Plains
generally is small (less than 1000 ft). Holzworth3 indicates that where
these depths are shallow (less than 3000 ft and preferably less than
1000 ft) the likelihood of extended periods of limited vertical mixing
is large; where they are deep (greater than 3000 ft) such likelihood is
small. The mean maximum mixing depths over the Great Plains vary greatly
by season (see Figure XV-4). The computed values for the St. Louis area
average about 1200-1500 ft for winter; 3000 ft in spring; 3000-4000 ft
in summer and early fall followed by a marked reduction to 2000-2500 ft
in October and November.
In another study Holzworth4 presents the mean annual morning and
afternoon mixing heights. He defines the "mixing layer" as "the surface-
bounded layer of the atmosphere through which relatively vigorous verti-
cal mixing occurs." Solar heating results in mixing layers that are com-
mon over mid-latitude regions during the day. Such mixing layers are
also present over cities at night and Holzworth suggests that "it is not
uncommon for mixing heights over urban areas to vary diurnally by more
than two orders of magnitude. Over rural areas the nighttime loss of
height results in inversions in which vertical mixing is minimal."
3.
George C. Holzworth, "Estimates of
Contiguous United States," Monthly
242, May 1964.
George C. Holzworth, "Meteorological Potential for Urban Air Pollu-
tion In the Contiguous United States," paper No. ME-20C presented at
the Second International Clean Air Congress, Wash. D.C., Dec. 6-11,
1970.
Mean Maximum Mixing Depths in the
Weather Review, 92, No.5, 235-
4.
XV-ll
-------
'0°
IICIII
SA-1365-19
FIGURE XV-4
GEOGRAPHIC DISTRIBUTION OF A NUMBER OF LOW-LEVEL JET
OCCURRENCES FOR A TOTAL PERIOD OF TWO YEARS, ACCORDING
TO EVALUATIONS BY BONNER (1965) OF UNITED STATES WEATHER
BUREAU SYNOPTIC PILOT-BALLOON NETII\ORK DATA
In the St. Louis area, the mean annual morning mixing height is
about 1200 ft; the mean annual afternoon mixing height is about 4500 ft.
Holzworth states that the morning mixing heights, for the most part, do
not vary greatly on a seasonal basis; they are roughly 350-700 ft higher
than the annual mean in spring and about 350 ft lower than the annual
mean in summer and autumn. Afternoon mixing heights display considera-
ble differences in their seasonal upper limits, ranging from a value of
5000 ft in winter to values above 10,000 ft in summer.
XV-12
-------
Wind
Dispersion of pollutants, of course, is also dependent on wind speed.
The mean annual wind speed averaged through the morning mixing layer for
the St. Louis area is about 5 m/sec-l and about 6-7 m/sec-l averaged
through the afternoon mixing layer. The predominant direction for light
winds is from the southeast. In general, the faster speeds occur in
spring. Maximum seasonal afternoon speeds (averaged through the layer)
is 10 m/sec-l and minimum seasonal morning speeds are less than 2 m/sec -1
Holzworth also presents data on episodes of limited dispersion* which
indicate that stated criteria in eastern Missouri and western Illinois oc-
curred a total of approximately 30 episode-days in a five-year period;
the autumn season contained most of the total occurrences.
A parameter of importance to pollution considerations is the so-
called "low-level jet stream." It is generally a summertime phenomenon,
occurring during the nighttime hours. Characteristically, the wind field
in the lower 1500 ft above the surface begins to increase in speed from
a southerly quarter, and to decrease in speed above the 3500 ft level.
The speed increases to a maximum of about 45 knots at about 1500 ft
shortly after midnight after which unstable lapse-rates prevail. Data
presented by Bonner (see Figure XV-4) show about 125 observations (12
and 6) of the low-level jet in the St. Louis area during a two-year pe-
riod. Also, it has been shown by Bonner that the nocturnal thunderstorms,
not associated with frontal activity, are significantly correlated (in
time and space) with the presence of a low-level jet stream.
*
Defined as follows: (1) all mixing heights 5000 ft or less, (2) all
mixing layer average wind speeds 4 m/sec-l or less, (3) no significant
occurrence of precipitation during l2-hr periods covering each mixing
height calculation, and (4) above conditions satisfied continuously
for at least 2 days.
XV-13
-------
Chapter XVI
LAND AND BUILDING REQUIREMENTS
The Regional Study requirements for land and building space have two
principal components. The first covers the central facility itself, and
the second includes land for the instrument stations. The general char-
acteristics and selection criteria for building space and instrument sta-
tion sites are provided in the following discussion.
Central Facility
Selection Criteria
The building space requirements for the St. Louis central facility
consist of two types. The first covers administrative and other office
space necessary for the bulk of the permanent staff, temporary research
personnel, and the data-processing equipment. The second includes space
necessary for instrument repair and outfitting of the transportable in-
strument stations, the chemical laboratory, and related functions. Ide-
ally, these two functions would be located within the same structure but
a modest separation ought not hamper efficient operations. Since the
metropolitan St. Louis area has many types of structures potentially
available, satisfactory space in existing structures should be readily
acquired. The precise location will, of course, depend upon availability,
rent or lease terms, and similar local factors. Several basic technical
factors should be considered in location selection.
First, since marked utilization of aircraft is anticipated in many
of the research experiments as well as for routine data gathering, serious
consideration should be given to locating the central facility near an
airport. The airport need be of only modest size but should have mainte-
nance and fueling facilities and should be in continuous 24 hour operation.
Facility location near an airport would clearly facilitate the transfer of
airborne-acquired data to the computer facility for logging. Additionally,
air sample containers could be quickly transported to the laboratory facil-
ities for expedient analysis. Finally; in the event that helicopters are
included in the system, personnel and equipment may well be transported to
the more remote rural sites far more efficiently and quickly than by sur-
face transportation.
XVI-l
-------
A total of 16 commercial airports are situated within about 30 miles
of the central St. Louis area. Twelve of these, however, are judged to
be inappropriately located for purposes of the Regional Study because
they are generally located south and west of the central area. Lambert
Field is included in the twelve, and it could be eliminated because of
expected operational constraints, and one--Lakeside Airport east of Gran-
ite City, Illinois--tends to be too small and lacks facilities. Since
the bulk of the airborne and surface monitoring will likely be carried
out in the northerly and easterly directions from the central area, the
airport location should lie in the same general area. The remaining three
possible sites include: the Bi-States Park Airport at Centreville, Illi-
nois; Civic Memorial Airport at Bethalto, Illinois; and Collinsville Air-
port at Collinsville, Illinois. In addition to these three commercial
airports, Scott Air Force Base should be included as a significant can-
didate. The most centrally located airport is perhaps the Civic Memorial
Airport near Bethalto.
Second, the facility should be situated roughly at the center of the
instrumented area. This location may tend to minimize data transmission
costs and reduce the difficulties of moving personnel and equipment be-
tween the central facility and the instrument sites. Since the instru-
mented area may be somewhat biased to the north and east of St. Louis,
the central facility should be similarly displaced.
Third, the location should be sufficiently near the St. Louis central
area to permit expedient negotiations with local equipment vendors and
suppliers, especially during the conduct of a research experiment when un-
foreseen requirements may arise for additional equipment, laboratory chem-
icals, or similar reasons.
Fourth, the central facility is intended to serve as the site for a
prototype instrument station and as a test bed for subsequently developed
meteorological and air quality instruments prior to their incorporation
in the instrument system. Consequently, the site should incorporate an
open unobstructed area of at least one-fourth an acre. Moreover, because
of the possible interference of the facility building with wind flow con-
ditions, the instrument station site should be slightly removed from the
facility building site itself. Thus, housing of the facility in a multi-
floored structure should be avoided if at all possible. Any appreciable
separation of the instrument site from the laboratory and shop facilities
will obviously create difficulties in day-to-day operation. Zoning regu-
lations at the site naturally must be such that a 30-meter tower can be
erected. This factor may control the proximity of the facility and tower
site to the airport.
XVI-2
-------
Finally, the facility should be located near one or more of the
cipal freeway routes in the St. Louis area to reduce travel times to
instrument sites.
prin-
the
On the basis of these criteria, the most appropriate general location
in regard to commercial airports appears to be in the Alton-East Alton
vicinity near the Civic Memorial Airport at Bethalto. The location tends
to be to the north and near the center of the instrumented area and is well
served by the freeway system. The Civic Memorial Airport is perhaps the
most fully developed airport in the area, short of Lambert Field, so that
all appropriate services should be available. The Alton-East Alton area
is sufficiently close to St. Louis itself to facilitate procurement of
equipment, supplies, and spare parts without undue delay and difficulty.
Scott Air Force Base stands as a strong alternative to the commercial
airport. Appropriate facilities should exist for aircraft maintenance,
fueling, and other tasks. The base likely will have existing buildings
which can be utilized for office and data-processing use as well as suit-
able shop areas. The base tends to be slightly isolated in terms of the
freeway network and is also somewhat further than East Alton from the
comparatively dense instrument system in the urbanized area of St. Louis.
Some difficulty is likely to be encountered with the day-to-day prob-
lems of housing, transport, and the like, of the temporarily located re-
search experimental groups. Entry and exit of such groups is likely to
be at Lambert Field in St. Louis, about 35-40 surface miles from Scott
Air Force Base. Public accommodations may on occasion be difficult to
find with the nearest large community being Belleville, Illinois, about
seven miles to the west. Perhaps accommodations could be located on the
base itself, but this may involve difficult administrative procedures,
especially if contractor personnel are involved in the research experi-
ments. All these drawbacks notwithstanding, Scott Air Force Base deserves
serious consideration, especially since EPA is currently authorized to use
the base for aircraft operations.
An additional set of factors should also be considered for the loca-
tion of the central facility. First, undoubtedly requirements will de-
velop for coordination and liaison with local, state, and federal agencies
in the St. Louis area, who have responsibilities and interests in air qual-
ity and meteorology. Most of these groups are likely to be situated in the
urbanized St. Louis area at some distance from the airport sites, so that
efficient coordination might be difficult.
Second, public support of the Regional Study would clearly be of con-
siderable value. The development of the emission inventory, for example,
XVI-3
-------
could likely be facilitated by such support, and, perhaps, the difficul-
ties of acquiring the instrument station sites could be markedly lessened.
The location of the central facility at a site near or even within the ur-
banized St. Louis area could well enhance the general visibility of the
Regional Study and aid in the development and maintenance of public sup-
port. Such benefits, of course, should also flow to other air quality and
meteorological studies in the St. Louis area. The extent to which the cen-
tral facility might feature special exhibits and displays to develop aware-
ness and understanding of the Regional Study is difficult to assess at this
time but might well be considered in subsequent planning.
Thus, final selection of the central facility site should merge the
strictly technical factors favoring an airport location with those suggest-
ing a more urbanized location. If, on balance, the technical considera-
tions predominate, then either the Alton-East Alton or Scott Air Force Base
areas should be given the highest priority in site selection. On the other
hand, if the need for coordination, high visibility, and related factors
outrank technical considerations, then sites in central and northern St.
Louis should be favored for selection.
Interior Space Requirements
The estimated requirements for space at the central facility are pre-
sented in Table XVI-I. The general office and data-processing areas have
the normal requirements for heating, lighting, acoustics, and related fac-
tors. Additionally, the area must be suited to necessary modification to
accommodate the special requirements of the data-processing equipment,
including any cooling equipment and ducting, cabling, and the like. Space
allocation is primarily based upon the number of staff members estimated
for the facility, as discussed in Chapter XIX with an allocation approxi-
mating 150 square feet per office. An additional 400 square feet has been
established for use by the temporary experimental research groups when
working in the facility. A total of 500 square feet has been allocated
for the data-processing center which should be sufficient for all equip-
ment as well as storage space for tapes, cards, and other materials. This
area should also be adequate to provide for maintenance and work space
around the equipment. The total office/computer space is estimated at ap-
proximately 3,900 square feet. This includes a contingency allowance of
775 square feet for hallways, waiting rooms, and the like.
The requirements for shop and laboratory space total 1,800 square
feet. The experimental research program requires a well equipped chemi-
cal laboratory. Appropriately, it should be a "wet" laboratory with all
basic laboratory utilities--municipal and deionized water, electric power,
gas, air, and the like. Additionally, it would be provided with standard
XVI-4
-------
Table XVI-l
INTERIOR SPACE REQUIREMENTS FOR THE CENTRAL FACILITY
Office/computer space
System chief
Research program coordinator and staff
Meteorologist
Instrument system engineers
Data processing
Chemical engineer and field staff
Clerical, files, office supplies
Conference room
Library, engineering files
Effects research
Experimental research group
Subtotal
Hallways, waiting room, and miscel-
laneous @ 25%
Total
Shop/laboratory space
GC mass spectrometer laboratory
Chemistry laboratory
Humidity controlled storage
Instrument installation
Electronic service
Other service
Parts storage
Total
Grand total
XVI-5
Square
Feet
200
300
150
350
500
250
400
200
200
150
400
3100
775
3875
400
400
100
500
100
100
200
1800
5675
<"-'5700)
-------
laboratory reagents, glassware, and laboratory bench facilities
ficient space for specialized equipment provided on a temporary
the experimental research groups.
with suf-
basis for
Bench space will be required for instrument calibration, maintenance
and modification, and similar tasks. These tasks can be divided into
electronic and mechanical activities, and they should be provided with
separate work areas. The former would be fitted with the necessary diag-
nostic electronic instrumentation and related equipment. Mechanical ac-
tivities would require facilities to accommodate a variety of tasks, in-
cluding glass blowing for special instrument repair and fabrication, metal
cutting and forming, and similar functions which will clearly arise during
the course of the Regional Study.
An appreciable area will also be required for servicing and modifying
the transportable instrument stations of the facility. The space should
be sufficiently large to house at least one of the mobile calibration ve-
hicles, so that access from the outside can be gained by means of large
garage-type doors; interior ceiling heights should be in the range of
10-12 feet. The servicing of the transportable instrument stations will
consist primarily of the installation and change of air quality instru-
ments as dictated by the experimental research program. Minor modifica-
tions of the instruments likely can be carried out in the field as part
of the routine maintenance program. However, at the conclusion of major
research experiments when significant equipment changes would be in order
and probably most appropriately carried out during the winter months of
minimal field activity, the service space will be in continuous use. The
space requirements, accordingly, are based upon housing two transportable
instrument stations at one time for simultaneous servicing. These sta-
tions are estimated to be approximately 90 to 120 square feet in area.
The simultaneous housing of two transportable stations and one mobile
calibration vehicle is estimated to require 500 square feet including
service and working space about each unit.
Outdoor Facilities
The outdoor facilities are expected to consist mainly of parking and
storage space. Space, of course, will be required for the prototype in-
strument station and tower. A guyed 30-meter tower, for example, requires
approximately one-fourth acre and should be near electric power and tele-
phone facilities. All prototype station test activities likely can be
carried out within this area with no problem.
Adequate parking facilities will be required for the permanent staff,
experimental research personnel, vendor representatives, and others.
XVI-6
-------
Approximately 40 spaces should be provided for an aggregate area, includ-
ing movement space, of approximately one-half acre or 20,000 square feet.
A securely fenced outdoor equipment storage area is required for the
calibration trailers, Class C instrument stations, and miscellaneous gen-
eral equipment. Storage space of the order of 10,000 square feet should
be adequate. However, on occasion additional temporary space may be re-
quired.
Instrument Station Sites
Selection Criteria
The instrument sites associated with the Regional Study should
generally be selected with respect to at least eight criteria; these are:
.
Freedom from unique or overriding micrometeorological effects
caused by structures, vegetation, and topography
.
Minimal deviation from analytically defined location
.
Absence of nearby significant pollutant sources
.
Convenient availability of electric power and communication
utility services
.
Area stability with respect to land usage, emission sources,
and the like
.
Free access at all times
.
Minimal zoning restrictions and institutional constraints
.
Suitable site size.
The requirement for absence of unique micrometeorological effects is per-
haps one of the most important. Since the measurements taken at each site
will be used as the basis to define the meteorological and pollutant con-
ditions or fields throughout the entire region, the measurements should
represent insofar as possible the free field conditions in the area of
each station. Small-scale anomalous conditions can indeed seriously im-
pair the value of the data acquired. Difficult compromises in the urban-
ized area may be required, because available sites may be limited.
XVI-7
-------
The need to achieve the least deviation from the analytically defined
station location clearly is of importance. Minor adjustments to conform
with other criteria should be permissible but large displacements could
necessitate review and adjustment of numerous other stations as well.
The criterion of absence of nearby significant pollutant sources will
vary with the type of instrument station. The stations intended to moni-
tor the ambient air quality must generally meet this criterion. On the
other hand, the more specialized Class C transportable stations sited for
certain unique research experiments may indeed appropriately ignore this
criterion.
The nearby availability of electric power and communication utility
services will be required at the Class A and B stations, and at least
electric power would be desirable at the Class C stations. At a cost,
both services can be made available at any point in the Regional Study
area. Clearly, however, in the absence of compelling reasons, especially
lengthy power and communication drops should be avoided in favor of ad-
justment of location. Lengthy drops increase both the costs and expected
outage times and should be avoided. The Class C stations are designed to
be equipped with engine-generator sets in the event that electric power
is not available. Although these are suitable power sources, their use
tends to increase servicing and maintenance times and their exhaust and
noise levels may introduce site location problems. On the other hand,
Class C stations for certain research experiments may be emplaced for
periods as short as two to four weeks, in which case the installation of
commercial power drops may not be justified because of their cost. As
noted in the discussion of digital data links, a number of the telephone
utilities in the Regional Study area have plant facilities not altogether
suitable for data transmission and are not firmly interconnected with the
larger utilities in the area. Accordingly, these factors should be
balanced--in terms of locating the site within the larger integrated tele-
phone utility service area with perhaps lengthy drop distances--against
the special engineering design and modification required of the smaller
utility if it is to achieve the proper transmission and switching charac-
teristics.
The criterion of area stability covers the general pattern of land
usage in the vicinity of the instrument site. The essential aim is to
select instrument sites in neighborhoods which appear to offer the small-
est possible change in land usage over the course of the Regional Study.
For example, a well-developed residential area with moderately new struc-
tures or city parks would tend to remain much the same over a five-year
period except, perhaps, for the growth of vegetation. Areas to be par-
ticularly avoided would be rapidly developing suburban residential and
XVI-8
-------
commercial areas and areas of the central city planned for large urban
renewal projects. The purpose of this criterion is simply to ensure that
changes in the instrument station measurements over time be caused by
meteorological variations and emission levels rather than local changes
in land usage causing micrometeorological anomalies or the establishment
of a new pollutant source near the instrument station.
The need for free access to the site at all times is, of course, of
extreme importance. With proper arrangements, this criterion should pre-
sent little if any difficulty but nevertheless should not be overlooked.
The rural stations should be selected so that winter weather conditions
do not hamper or prohibit access.
The criterion pertaining to zoning restrictions and institutional
constraints will likely vary somewhat directly with the degree of urban-
ization. The constraint of particular importance is any limitation on
structure height and the manner in which instrument towers are inter-
preted within such limitations. Other constraints of importance include
exterior structural design and its possible impact on instrument shelter
design, site security fencing, materials storage including chemicals and
high pressure cylinders, and related factors.
Site size will vary with the class of station and the tower selected
for use. Sites having 30-meter guyed towers will require a square one-
fourth acre plot in order to enclose the guy anchors. Sites of this size
may be difficult to acquire in the urbanized area, in which case an un-
guyed self-supporting tower will be necessary. In this event the site
may be reduced to approximately 50 X 50 feet. The Class B instrument
stations without towers also require a square 2500 square-foot site.
The Class C stations with the smaller instrument trailers require
approximately 200 square feet (10 X 20 feet). Formal negotiation and
acquisition of these sties is not anticipated. Since the Class C sta-
tions will be emplaced for relatively brief periods, somewhat less for-
mal authorization procedures for site usage with property owners should
be possible.
Implementation
The process of actual site selection and acquisition will consist
of six principal tasks. First, analytically derived pattern of station
location, as provided in this Prospectus, should be comprehensively re-
viewed and reassessed in the light of the actual allocation of funds for
the Regional Study.
XVI-9
-------
Second, a survey party composed of the Research Program Coordinator
and assistants, would carry out a detailed field inspection of each
potential site and the immediate surrounding area. The survey team
would compile detailed information with respect to the existence of
sites meeting the selection criteria and would note engineering design
features, such as terrain slope, drainage, and general soil conditions.
Where evident by field inspection, site ownership should be noted. This
would likely be possible for public parks, schools, federal agencies, and
the like. Ownership of other sites likely will not be obvious by inspec-
tion. A photographic record would be desirable. In the heavily urbanized
area where appropriate sites may be limited, a systematic street-by-street
search may be necessary up to one mile from the analytically defined loca-
tion. Increased visibility in the rural areas probably will ease the sur-
vey procedure appreciably for those areas.
Insofar as possible, on-the-spot assessment of each site should in-
clude consideration of the future trends in land usage in the vicinity.
Such assessment should focus mainly on the possibility of new construc-
tion and related activities in the area which may tend to affect air qual-
ity or meteorological measurements of the nearby instruments. Such assess-
ments should be subsequently augmented by consultation with city and county
planning commissions, highway departments, and perhaps realty boards.
Third, the potential sites within each area would be ranked with re-
spect to their conformance with the physical selection criteria. Ranking
would also acknowledge ownership, where known. Ranking by ownership is
of value from the viewpoint of subsequent acquisition tasks and possibly
facility operation throughout the course of the Regional Study. That is,
it appears most appropriate to limit insofar as possible the number of
organizations and individuals with which negotiations are conducted con-
cerning the instrument sites. If, for example, all St. Louis City sites
could be selected within city parks, negotiation would be essentially
limited to a single group rather than up to perhaps 12 groups in a mul-
tiple site ownership site pattern.
Fourth, the ownership, where unknown from the field survey, of the
most desirable sites would be determined. Review of land ownership rec-
ords at the appropriate county offices will reveal this information.
Fifth, zoning regulations and institutional constraints affecting
the use of the potential sites should be reviewed. Probably the most
important constraint will be associated with the erection of the 30-
meter tower within urban areas. Where zoning regulations prohibit such
structures, a variance must be sought. Such procedures can be time-
consuming and therefore should be initiated at a very early point in
XVI-IO
-------
the implementation schedule. In certain difficult situations, the use of
the more sightly unguyed tower may more easily win a zoning variance.
Of course, if these efforts fail, alternative sites must be sought even
though they may be less desirable for other reasons.
Sixth, acquisition negotiations with the site owners would be ini-
tiated. The nature of the negotiations and the method of acquisition
likely will depend largely on the character of the owner. In some cases,
site acquisition may be in the form of a no-cost outright loan. In others
however, rental or lease agreements likely will be required. From a point
of view of cost, the former method is more desirable but may involve a
risk factor of use revocation by the site owner during the Regional Study.
Long-term lease avoids this risk. In the event the instrument site is a
part of a larger parcel under one ownership, plans for use of the parcel
should be reviewed to ensure that the planned use or area stability will
not affect the instrument installation.
It should be noted that site selection for the Class C stations will
usually be on an ad hoc or perhaps an opportunistic basis determined by
the specific research needs fulfilled by the station at the time. Thus,
many of the siting criteria would not apply to the Class C stations.
XVI-ll
------- |