Technical Design Proposal
Clean Air
Status and Trends Network
(CASTNET)
By The Environmental Protection Agency
In Cooperation With
The National Oceanic And Atmospheric Administration
And Other Federal Agencies
EPA
450/
1992.1
February 1992
-------
Technical Design Proposal
Clean Air Status and Trends Network
(CASTNET)
4
February 1992
Contributors Are Listed In Appendix A
.
!' :§ '')
Atmospheric Research and Exposure Assessment Laboratory
Office of Research and Development
. U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
JEEA Headquarters library
-------
-------
Notice
The use of trade names or commercial products in this document does not constitute endorsement
or recommendation for use.
Acknowledgements
This document was prepared jointly by the U.S. Environmental Protection Agency; other federal
agencies; state and local agencies; universities; and the Canadian environmental community. Specific
coordination was provided by the National Oceanic and Atmospheric Administration. A complete listing
of contributors in the planning and review of this document is presented in Appendix A.
-------
February 1992
Contents
Section Page
List of Tables vii
List of Figures viii
List of Acronyms and Chemical Symbols ix
Executive Summary xiii
Introduction 1
CASTNET Organization 1
Total Deposition 3
Aquatic and Terrestrial Effects 3
Visibility/Acid Aerosols 4
Air Toxics 4
Statistical Network Design for Status and Trends .. .. 5
Data Management 5
Instrumentation/Methods 5
Contract Acquisition ; 5
Drawing on Existing Resources 6
The "80% Network" 8
Reporting Results 8
Quality Assurance 9
This Report 9
Total Deposition Monitoring 11
Monitoring Objectives 11
Technical Approach , 12
Description of MeasurementsWet Deposition 17
Description of MeasurementsDry Deposition . 18
Treatment of Confidence and Uncertainty 21
Sites and Implementation 24
Relationship with Existing Networks ... 30
Report Formats 32
HI
-------
February 1992
Future Research Needs 33
Sample-Averaging Times 33
Wet Deposition 34
Dry Deposition 35
Regional Deposition Model Evaluation 36
Aquatic and Terrestrial Effects Monitoring 39
Monitoring Objectives 39
Three Types of Monitoring Required 40
Technical ApproachAquatics Effects and Stressor Monitoring 42
Description of Measurements 51
Additional Aquatics Research and Monitoring Programs 52
The Adirondacks Aquatics Effects Program 52
The EPA EMAPWetlands Program 53
Technical ApproachTerrestrial Effects and Stressor Monitoring 53
Description of Measurements 60
Deposition Sites and Implementation 61
Relationship with Existing Networks 61
Report Formats . 62
Future Research Needs 63
Visibility/Add Aerosols Monitoring 65
Monitoring Objectives 65
Technical Approach 66
Description of Measurements 69
Treatment of Confidence and Uncertainty 72
Sites and Implementation 72
Relationship with Existing Networks 76
Report Formats , 77
Future Research Needs . 77
Air Toxics Monitoring 79
Monitoring Objectives 79
iv
-------
February 1992
Great Waters Program 79
Area Source Program 80
Technical Approach 82
Great Waters Program 82
Area Source Program 84
Identification of Affected Areas ; : ... 86
Description of Measurements 87
Treatment of Confidence and Uncertainty 90
Sites and Implementation . 91
Relationship with Existing Networks 92
Future Research Needs 93
Statistical Network Design for Status and Trends 97
Air and Deposition Variables 97
Integration of Existing Monitoring Network Data 98
Frequency of Field Sampling 98
Media Representation and Area! Coverage 99
Design of Optimal Network .; 99
Preliminary Findings 101
Wet Deposition 101
Dry Deposition ..'; 104
Ozone 104
Data Management 105
Recommendation '. 105
Data Collection and Storage 106
Initial System Concept 107
Quality Assurance/Quality Control 107
AIRS Selection Rationale 110
Advantages 110
Disadvantages .. . . 110
Implementation Schedule HI
Instrumentation/Methods . 113
Current Projects 113
Developing Passive Sampling Devices (PSDs) 113
Evaluating the Versatile Air Pollution Sampler (VAPS) 114
-------
February 1992
Real-Time Acid Aerosol Instrument Development 115
Initiating Experiments with Filterpacks US
Monitoring of Hazardous VOCs with Automated Gas Chromatographs (autoGCs) 116
Sorbent-Based Sampling and Analysis of Hazardous Polar VOCs 116
Projects for Future Work 117
EstablishING Comparability of SO2 PSD Results with Filterpack Results at Four
Additional Geographically Diverse Locations 117
Deploying a Better Sample Collection Device 117
Deploying a Screening Device for Acid Sulfate 117
Testing PSDs for Other Trace Gases 118
Developing a VOC Sampling System 118
General Recommendations 118
Specific Recommendations 119
Total Deposition 119
Aquatic and Terrestrial Effects 119
Visibility/Acid Aerosols . 119
Air Toxics 119
Appendix A - CASTNET Work Group Members and Workshop Participants A-l
Appendix B - CASTNET CAAA Mandate Summary and Reporting Requirements B-l
Appendix C - Selected References C-l
i
Appendix D - Sample Instrumentation/Methods "White" Paper D-l
VI
-------
February 1992
Table
1
2
3
4
5
List of Tables
Method and Spatial Interpolation Uncertainties 22
Precision, Accuracy, and Site-Specific Uncertainty Estimates for Major
Species Measured 23
Characteristics of Some Nonurban Monitoring Networks in North America 31
Summary of Visibility Equipment Types and Analyses 71
System Inputs, Processes, and Outputs 108
vii
-------
February 1992
List of Figures
Figure Page
1 Acid Deposition and Ozone Regions Based on Topographical, Chemical
Concentration, and Climatic Characteristics 16
2 Currently Operating Wet Deposition Monitors as of March 1991 25
3 Currently Operating Dry Deposition Monitors as of March 1991 26
4 Possible CASTNET Deposition Monitoring Sites 28
5 Possible Site Types at Proposed CASTNET Monitoring Locations 29
6 Map of High-Interest Areas 46
7 Sensitive Terrestrial Ecosystems With Defined or Likely Effects From Air
Pollution Stress 57
8 Trends in U.S. Haze Based on Extinction Coefficients From Airport
Observations 68
9 Priority Areas for Visibility Monitoring 73
10 Combined Existing and Proposed Visibility Monitoring 74
11 Wet Sulfate Regional Trend Assessment 103
12 CASTNET Data Flow 109
via
-------
February 1992
List of Acronyms and Chemical Symbols
Acronyms
ADS Acid Deposition System
AIRMoN Atmospheric Integrated Research Monitoring Network
AIRS Aerometric Information and Retrieval System
ANC Acid-neutralizing capacity
ANL Argonne National Laboratory
AQS AIRS Air Quality Subsystem
AREAL Atmospheric Research and Exposure Assessment Laboratory
CAAA Clean Air Act Amendments of 1990
CASTNET Clean Air Status and Trends Network
DIC Dissolved inorganic carbon
DOC Dissolved organic carbon
DOE Department of Energy
DQOs Data quality objectives
EAM Engineering Aerosol Model
EC Canada: Environment Canada
EMAP Environmental Monitoring and Assessment Program
EPA Environmental Protection Agency
FHMP Forest Health Monitoring Program
FTIR Fourier transform infrared
GCRP Global Change Research Program
CIS Geographic Information System
GLWQA Great Lakes Water Quality Agreement
HAPs Hazardous air pollutants
IAA Infrared aerosol analyzer
IADN Integrated Atmospheric Deposition Network
IMPROVE Interagency Monitoring of Protected Visual Environments
LAI Leaf area index
LaMP Lake-wide Management Plan
LJDAR Light Detection and Ranging
LOD Level of detection
LOQ Level of quantitation
NAAQS National Ambient Air Quality Standards
NADP/NTN National Atmospheric Deposition Program/National Trends Network
NAPAP National Acid Precipitation Assessment Program
NDDN National Dry Deposition Network
NEP National Estuary Program
NERR National Estuarine Research Reserves
NESCAUM Northeast States for Coordinated Air Use Management
IX
-------
February 1992
NOAA National Oceanic and Atmospheric Administration
NFS National Park Service
NSWS National Surface Water Survey
NWS National Weather Service
OAQPS Office of Air Quality Planning and Standards
OME Canada: Ontario Ministry of the Environment
PAHs Polycyclic aromatic hydrocarbons
PCBs Polychlorinated biphenyls
PICs Products of incomplete combustion
PSDs Passive sampling devices
QA Quality assurance
QC Quality control
RADM Regional Acid Deposition Model
RFP Request for proposals
RLTM Regionalized Long-Term Monitoring
SAMWG Standing Air Monitoring Work Group
SGCP Southern Global Change Program
SIP State Implementation Plan
SOPs Standard operating procedures
SVOCs Semi-volatile organic compounds
TAMS Toxics Air Monitoring System
TIME Temporally Integrated Monitoring of Ecosystems
TVA Tennessee Valley Authority :
USFS U.S. Forest Service
USGS U.S. Geological Survey
VAPS Versatile air pollution sampler
VOCs Volatile organic compounds
XAD/PUF Sorbent/polyurethane foam
Chemical Symbols
Aluminum ion
Calcium ion
Chlorine ion
Hydrogen ion
- Hydrogen peroxide
Nitric acid
Potassium ion
Magnesium ion
- Sodium ion
-Ammonia
Ammonium ion
Nitrogen oxide
Oxides of nitrogen
NO2 Nitrogen dioxide
ci--
H*-
HA-
HNO3
Na*-
NH,-
NH/
NO-
-------
February 1992
NO3' Paniculate nitrate
Oj Ozone
Pb-Lead
PO4*" Phosphate ion
SO2 Sulfur dioxide
S0«* Paniculate sulfate
XI
-------
February 1992
xii
-------
February 1992
Executive Summary
The Clean Air Act Amendments (CAAA) of 1990 will have profound and far-reaching
effects on air pollution and public health
during the next decade and beyond. The
CAAA incorporate significant new
regulatory requirements designed to reduce
risks to public health and welfare as well
as broad new research provisions that can
substantially improve our scientific -
understanding of the causes and effects of
damage caused by air pollution.
Among the most significant of the
new research provisions are the
requirements for enhanced environmental
measurement and monitoring contained in
Titles I, m, IV, vm, and K of the
CAAA. These provisions require
measurement and monitoring of acid
deposition, ozone, acid aerosols, and
CAAA MONITORING AND RESEARCH
PROVISIONS ADDRESSED BY
CASTNET
Total Deposition
- National network and reports every five yean
- NAPAP (EPA and othera) monitoring and two year
report*
- Western high elevation monitoring and annual
report!
- National network and report* every five yean
- Ecomtem research on came*, effects, and trends
(forest* , crop*, toils; surface water, groundwater,
estuaries, wetlands)
- NAPAP (EPA) report every two yean on effects, and
every four yean on preventive deposition level*
. Annual report oa Western high elevation effect*
- EPA report on SO, (or NO. trades; and feasibility of
preventive standard in three yean
- National Acid Late* Regirtrv
- Adirondack* watershed research
Vribifiry/Acid Aerasoh
- National network, and report* every five yean
- NAPAP (EPA) reports every two yean
~ da** I ana monitoniig ud research
Air Toxics
- EPA/NOAA assessment, and reports every two yean
--Rve station* in Great Lakes, one per Great Lake
- Regulations in five yean for air deposition to Great
I,aktSi Chesapeake Bay, Lake Champlain, and coastal
w»t*r* a* deemed neeeatary by the Report due to
Congrea* in three yean
Urban toitics research and monitoring, with report in
three yean and rational control strategy in five yean
hazardous air pollutants.
Examined collectively, these requirements impose the need for a substantial expansion of the
existing national program for urban air measurements and monitoring; this current program has
xiu
-------
February 1992
provided the historical scientific basis for the establishment and subsequent revision of die National
Ambient Air Quality Standards (NAAQS). For the first tune, the CAAA also mandate a broad
measurement and monitoring program beyond our populated urban areas that will advance our
understanding of die effects of air pollutants in rural areas and in aquatic and terrestrial ecosystems.
The CAAA provisions also impose a
requirement to enhance and accelerate
environmental research and determination of
status and trends in crucial areas. In general,
the provisions also present two significant
scientific challenges. The initial challenge is
to develop 3 research and monitoring program
that integrates as many measurements as
possible into a single monitoring program that *
in turn integrates network design, trans-
jurisdictional operation, and data analysis.
The second challenge is to establish an adequate number of sites in appropriate locations to ensure
mat the resulting measurements can be interpreted with confidence and accuracy.
Neither of these scientific issues has been addressed explicitly and scientifically as our
national air pollutant monitoring networks have evolved over the past twenty years. Resolutions must
be generated, however, if we are to advance our research on the effects of air pollution and the
impacts of the CAAA of 1990 in reducing this pollution.
This document presents the preliminary design for a comprehensive research monitoring and
measurement network that meets the objectives established in the CAAA. An important element of
REPRESENTATIVE LIST
OF WORK GROUP PARTICIPANTS
Environmental Protection Agency (EPA)
National Oceanic and Atmospheric
Administration (NOAA)
U.S. Forest Service (USFS)
* National Park Service (NPS)
U.S. Geological Survey (USGS)
Canada: Ontario Ministry of the
Environment (OME)
Canada: Environment Canada (EC)
Colorado State University
Rensselaer Polytechnic Institute
Standing Air Monitoring Work Groin)
(SAMWG)
Northeast States for Coordinated Air Use
Management (NESCAUM)
XIV
-------
February 1992
this design is a rigorous analysis of alternative approaches that could be taken, with emphasis on
deciding the appropriate number and distribution of sites required to provide interpretable
measurements mat are both scientifically rigorous and suitable for policy and regulatory analysis. The
value of the air pollution related status and trends data of the kind to be collected and coordinated
through CASTNET is expected to extend beyond state and federal regulators to resource managers,
policy makers, and others.
Because of the underlying continental scale and scientific complexity of the scientific issues
that must be addressed in mis design, mis document necessarily is the product of extensive
collaboration within the United States and Canadian community of environmental scientists and
engineers. The workshops on which this document is based were attended by representatives of all
major North American organizations currently involved in environmental monitoring, including the
Environmental Protection Agency (EPA), the National Oceanic and Atmospheric Administration
(NOAA), the U.S. Forest Service (USFS), the National Park Service (NFS), the U.S. Geological
Survey (USGS), the Ontario Ministry of the Environment (OME), Environment Canada (EC),
Colorado State University, Rensselaer Polytechnic Institute, the Standing Air Monitoring Work Group
(SAMWG), and the Northeast States for Coordinated Air Use Management (NESCAUM). Appendix
A identifies the individual scientists who have contributed in the preparation and review of this
document.
The design overview on this page is a brief synopsis of the design of the environmental
research network that would respond to the CAAA. It is essential mat mis overview be interpreted in
the context of the detailed information contained in each of the subsequent chapters of this document.
Briefly, the network design incorporates the following monitoring sites that are essential to achieving
the Clean Air Act's research goals during the next ten to fifteen years: 51 additional site locations to
xv
-------
February 1992
monitor wet deposition, dry deposition,
ozone, or a combination thereof; 20
aerosol-only and 12 "full" visibility
monitoring sites; one air toxics site per
Great Lake with piloting and eventual
expansion to Great Waters and appropriate
urban areas; and integration of several
existing monitoring networks data. While
much of this network is regional or
national in scope, it also includes a focus
on specific regions and sensitive
ecosystems which require more intensive
research to elucidate the processes, rates,
and effects that are active there. The
network also includes a focus on urban
areas which represent elements of the
expanded urban air monitoring and .
research program mandated by the Amendments.
A small network of more intensive (daily sampling) measurement sites operated by NOAA is
important to the development of CASTNET. It is designed to test and evaluate new instruments and
monitoring techniques. The shorter sampling period is also valuable in developing a better
understanding of atmospheric phenomena associated with deposition processes and model validation,
and may allow the rapid detection of emission reductions from specific source areas.
DESIGN OVERVIEW
Total Deposition
. 19 additional sites added to existing
NADP/NTN
- 31 additional sites added to existing
NDDN
- 45 ozone sites added
Aquatic and Terrestrial Effects
- An application of EMAP's TIME and
RLTM programs in the NE lakes and Mid-
Appalachian streams
- An application of the USFS FHMP in the
NE, SE/S, and W
- 10 to 15 Intensive Monitoring Sites
Visibility/Acid Aerosols
- 12 "foil* and 20 aerosol-only sites
complementing existing IMPROVE
network
Air Toxics
1 station per Great Lake by 12/31/91
- Integrated pilot study on Lake Michigan
- Preliminary monitoring in diverse locations
- Expansion of piloted method to other Great
Waters
- Monitoring and source characterization in a
. "representative* number of cities
XVI
-------
February 1992
All of these networks are designed to be extensions of current activities, necessitated to
answer the questions posed in the Clean Air Act Amendments in the tune that is available.
xvn
-------
February 1992
XVIll
-------
February 1992
Introduction
On November 15, 1990, the President signed the Clean Air Act Amendments (CAAA) which
were designed to address the problems of acid rain, ozone nonattainment, and air toxics. The CAAA
included a call for an unprecedented, extensive national monitoring program to assess improvements
in the nation's air quality and overall environment. The challenge was considerable, not only because
of the sheer size of the effort but also because of the logistical, technical, and scientific requirements
involved. In addition to air quality monitoring, the CAAA address a wide range of interrelated
monitoring, assessment, and research efforts ranging from atmospheric pollutant monitoring, effects
of air pollution on aquatic and terrestrial ecosystems, and visibility and air toxics monitoring.
To meet part of the challenge, an integrated program has been developed to reflect the
interagency, intergovernmental, and international requirements of large-scale monitoring and
assessment. The key component of this multi-agency activity is the Clean Air Status and Trends
Network (CASTNET) developed by the Environmental Protection Agency (EPA) in specific
coordination with the National Oceanic and Atmospheric Administration (NOAA).
CASTNET's goal is to establish an effective monitoring and assessment network to:
* Determine the status and trends of air pollutant levels and their environmental effects.
Develop a scientific database to better understand causality for policy considerations.
These two goals are integral parts of the ultimate question that must be answered:
How well are the Clean Ah* Act goals being met?
CASTNET Organization
CASTNET has two separate and distinct functions: that of creator of the proposed network
design, capable of fulfilling the CAAA mandates, and that of the entity that will
-------
CASTNET WORK GROUPS
Program Work Groups:
Total Deposition
Aquatic and Terrestrial Effects
Visibility/Acid Aerosols
Air Toxics
Support Work Groups:
Statistical Network Design for
Status and Trends
Data Management
Instrumentation/Methods
Contract Acquisition
- February 1992
implement the proposed CASTNET monitoring
sites and perform the integrated assessments for
the reports to Congress. This document is a
product of the first function. Its
recommendations, however, are linked to the
second function, which is responsible for the
action items described.
Eight work groups were created for
CASTNET. Four program areas were defined to
design monitoring networks in their respective
areas: Total Deposition, Aquatic and Terrestrial Effects, Visibility/Acid Aerosols, and Air Toxics;
At the first of two all-hands workshops held in November 1990 (the second being hi February 1991),
these program groups were charged with designing an optimum network with alternatives for their
program areas, with the following specifics to be identified:
1. The Clean Air Act monitoring objectives being served.
2. An approach that identifies affected areas.
3. A description of measurements, to be made by:
a. location
b. type
c. frequency
d. instrumentation
4- Proposed relationship with existing networks: state, local, and Canadian networks, and
networks run by other federal agencies.
-------
February 1992
S. Sites to satisfy the CAAA requirements and implementation schedule.
6. Cost implications.
7. Treatment of uncertainties and confidence limits associated with data.
8. Formats for reports generated by the network.
9. Future research needs.
The other four work groups were created to provide the program groups the support
necessary for their network design: Statistical Network Design for Status and Trends, Data
Management, Instrumentation/Methods, and Contract Acquisition. Each of the eight work groups is
identified below, along with a summary of salient considerations affecting network design. The
specific mandates for each program area will be described in the appropriate section. See Appendix
A for a list of work group and workshop participants.
Total Deposition
The work group was to design a monitoring network to determine how wet and dry
deposition levels for acid deposition and rural ozone concentrations have changed on a regional,
national, and critical ecosystem scale. Spatial and temporal coverage were required to be adequate
for annual and seasonal resolution both for direct evaluation of deposition changes and for relating
ecosystem changes to this deposition. The dry deposition component was presumed to involve
monitoring concentrations in air rather than deposition levels, while techniques for this are being
developed. Monitoring for terrestrial effects needed to include ozone.
Aquatic and Terrestrial Effects
The work group was charged with designing an aquatics and terrestrial effects monitoring
program and an atmospheric monitoring network to allow for an assessment of effects of atmospheric
constituents on surface water quality and forest condition. The monitoring of effects of air pollution
-------
February 1992
on surface water, (including wetlands and estuaries) and groundwater, as well as on forests, crops, and
soils was to be provided for. The proposed network also was to include determination of the
occurrence and effects of acid deposition on Western and high-elevation ecosystems, including forests
and surface waters.
Visibility/Acid Aerosols
The work group's initial task was to design a network to determine how visibility levels
effected by acid aerosols have changed in those areas most affected and how they have changed over
large regions. Because visibility can vary greatly within the season, appropriate resolution on a
temporal scale was necessary. The network also had to supply information regarding the causes of
visibility problems and improvements. Ultimately, the group will consider the design of the network
to monitor changes in atmospheric acid aerosol related both to environmental and human health
effects.
Air Toxics
There are two separate mandates for this work group: the "Great Waters" program and the
Area Source Program. The "Great Waters" program (which includes the Great Lakes, Chesapeake
Bay, Lake Cbamplain, and approximately 40 to 50 estuarine systems) is concerned with monitoring
hazardous air pollutants (HAPs) deposition to these waters and with the hazards to water quality (and
indirectly to people) and the environment (aquatic life and terrestrial wildlife). The CAAA mandate
one monitoring facility per Great Lake by December 31, 1991. The Area Source Program focuses on
ambient air monitoring of HAPs from urban area sources and the resultant hazards to human health.
The list of potential HAPs that must be monitored is long, and the task is complicated by the lack of
measurement technology for many of the pollutants.
-------
February 1992
Statistical Network Design for Status and Trends
This work group will play an integral part in selecting areas for potential future monitoring
sites and in evaluating the statistical contributions of existing monitoring sites. Its statistical
evaluations of network design will help ensure that the generated data will allow adequate estimations
of ecosystem exposure as well as adequate detection and quantification of trends to a desired level of
precision within a specified period.
Data Management
In large integration efforts, and CASTNET is no exception, the importance of data
management cannot be overstated. Hits work group was charged with developing a comprehensive
standardization program to ensure mat all CASTNET data, which will be provided by a myriad of
entities, are of comparable quality and are available in the format required by the various user groups.
Instrumentation/Methods
This work group surveyed current instrumentation and methodologies and recommended
specific methods to choose from for each of the monitoring networks being designed.
Recommendations were arrived at by analyses encompassing reliability, accuracy and precision, and
cost/benefits for each option. Research projects considered most crucial to CASTNET needs were
pursued.
Contract Acquisition
Creating the statement of work for the CASTNET contract was this work group's
responsibility. The contract will be the instrument for conducting monitoring, not provided by other
agencies, at sites that CASTNET establishes. Considerable work was required to make the contract
flexible enough to encompass the .variety and scope of monitoring that may be needed. Contract
proposals are currently being evaluated, with final selection of a contractor pending.
-------
February 1992
REPRESENTATIVE LIST
OF WORK GROUP PARTICIPANTS
Environmental Protection Agency (EPA)
National Oceanic and Atmospheric
Administration (NOAA)
U.S. Forest Service (USFS)
National Park Service (NFS)
U.S. Geological Survey (USGS)
Canada: Ontario Ministry of the
Environment (OME)
Canada: Environment Canada (EC)
Colorado State University
Rensselaer Polytechnic Institute
Standing Air Monitoring Work Group
(SAMWG)
Northeast States for Coordinated Air Use
Management (NESCAUM)
Drawing on Existing Resources
To meet the CAAA objectives in the
most efficient, cost-effective manner,
CASTNET will use the resources of other
programs, agencies, and networks at the
federal, state, local, and international levels.
The importance of these existing programs and
their integration with the CASTNET effort
cannot be overstated. This document, for ease
of presentation, is structured to present each
work group's recommendations as a separate
entity.
The details of implementing each proposed network design make it clear, however, that
CASTNET itself will be involved only in air and deposition monitoring. In most cases, it will merely
supplement and/or upgrade an existing monitoring site, while relying on other programs for the
biological and chemical effects monitoring required. Some of the organizations included in the
CASTNET design, as well as their expected contributions, are described on the next page. It should
be noted mat many of the people involved in CASTNET also are involved in one or more of the
listed organizations, which should facilitate coordination in these partnerships.
State networks are not discussed extensively in this document, because of the number of such
networks and die variations among the different state programs. These networks will be incorporated
into the CASTNET program as indicated by considerations of network siting and installation.
Preference will be given to existing sites and networks where the methods and data generated are
-------
February 1992
CASTNET Design Participant
research
NADP/NTN
NDDN
NAPAP
EMAP/TIME
EMAP-Surface Waters
EMAP-Forests
FHMP
GCRP
NFS
USFS
IMPROVE/NFS
IADN
CASTNET Supporting Networks
NOAA-CORE
NOAA-AIRMoN
State networks
Contribution
Wet deposition data
Dry deposition (concentration) data
Data relating to acid deposition
Aquatic chemical data
Aquatic biological data
Terrestrial biological data
Terrestrial chemical and biological data
Terrestrial cause and effects research
Terrestrial cause and effects research
and some
national air monitoring data
Aquatic and terrestrial cause and effects
and some national air monitoring data
Visibility data
Air toxics monitoring at Great Lakes
Contribution
Dry deposition reference data
Short-term data for atmospheric effects
Air quality and deposition data
compatible with CASTNET requirements. ....'.
Identifying site locations during all phases of siting will involve making.a preliminary
assessment based on a particular need (that is, to improve spatial or temporal resolution, to reduce
interpolation errors, or to provide ecosystem specific and critical ecoregion monitoring). After a
potential site location has been identified, maps of existing sites (wet, dry, ozone, visibility, air
toxics) will be overlaid to discover whether mere are existing sites that could be used with minimal
modification or whether any could be augmented or upgraded to fulfill the monitoring need in that
-------
February 1992
area. The appropriate site-operating authorities will be contacted and included in the site location
process from the outset. In the interest of creating the most efficient and cost-effective national
monitoring network possible, sites will be collocated at every available opportunity. This strategy is
the keystone of CASTNET.
The "80% Network"
CASTNET participants are charged with establishing the most comprehensive national air
pollution effects environmental monitoring program to date. They generally agree that the current
planned levels of resources and the level of scientific and engineering knowledge will allow about 80
percent of the network to be established, operated, and maintained. Future resource and research
needs, challenges in statistical design, data gathering needs, and needed instrument and protocol
development in some areas preclude the ability to complete a 100 percent network in the short term.
CASTNET will be able to complete its mission of fully determining all related CAAA environmental
status and trends when the remaining 20 percent of the network is in place. This objective depends
on resource availability and will require improvements in methods, siting, statistics, assessment
protocols, and quality assurance. Best estimates are that the advances necessary to implement the full
network will be made over the next three to five years.
Reporting Results
Although the more pressing aim of CASTNET is to add monitoring sites across the country
and to increase the assessment of environmental data, the results of these activities will only become
useful when they are reported. Consequently, the CAAA require each program area to provide
regular reports of its findings. The frequency and description of these reports are provided in
Appendix B.
-------
February 1992
Quality Assurance
EPA has stringent quality assurance (QA) requirements for all of its field monitoring
programs that cover all aspects from sampling to analysis and reporting. QA/SOPs (Standard
Operating Procedures) exist for most measurements. CASTNET will develop appropriate QA/SOPs
for those measurements that currently lack them, as well as for any new procedure that is adopted.
Most, if not all, of the large networks CASTNET will be incorporating operate under an approved
plan (see Appendix C for selected QA, implementation, and field operation references).
Before any monitoring data is incorporated into CASTNET, there will have to be an approved
QA plan as well as a review of existing data and their associated QA procedures. Comparability of
data relating to the chemical composition of the air and precipitation will be established in part by
periodic analytical testing with known samples. Comparability of data relating to the deposition of
chemicals would involve additional tests of the flux-determining techniques. This is done with
intercomparisons and by deployment of research-grade methodologies at a subset of locations, and
would include a periodic rotation among CASTNET sites. Such intercomparison programs are the
raison d'etre of the existing research networks operated by NOAA, with which the CASTNET
program will be tightly linked. By maintaining a well quality-assured federal network, establishment
of data comparability between states will be facilitated.
This Report
This is a combined report of the CASTNET work groups, and consists of a series of
recommendations. It is not meant as a scientific review or summary on network design and
operations. Familiarity of the reader with modern monitoring methods is assumed.
-------
February 1992
THIS PAGE INTENTIONALLY
LEFT BLANK
10
-------
February 1992
Total Deposition Monitoring
Monitoring Objectives
In the present context, the topic of "total deposition" is considered to include acid deposition,
deposition of related air chemicals, and information on their concentrations in both air and
precipitation. The Clean Air Act Amendments (CAAA) require monitoring of acid deposition, which
occurs when sulfur dioxide and nitrogen oxide emissions are transformed in the atmosphere and
return to earth, and ozone, which occurs as a result of nitrogen oxide reactions with volatile organic
compounds (VOCs) in the atmosphere. Both of these components are suspected health hazards and
are implicated in forest damage. In addition, acid rain damages lakes and buildings and indirectly
contributes to reduced visibility:
Monitoring objectives addressed in Title IV (Section 404) of the CAAA include identification
of sensitive and critically sensitive aquatic and terrestrial resources, and a report on the feasibility and
effectiveness of an acid deposition standard or standards to protect them.
Title DC of die CAAA [Sections 103(c), (e), Q)» and 901 (g)] requires the establishment of a
national network to assume eight responsibilities associated with total deposition:
1. Monitor, collect, and compile data, quantifying certainty in the status and trends of air
emissions, deposition, and air quality.
2. Ensure comparable air quality data from different states and nations.
3. Determine trends by region and effects on water quality, forests and crops, sensitive
ecosystems, and materials.
4. Provide data for model maintenance and application and for estimating transboundary
impacts.
11
-------
February 1992
5. Develop improved atmospheric dispersion models.
6. Develop monitoring systems and networks for evaluating and quantifying exposure to and
effects of multiple environmental stresses associated with air pollution.
7. Conduct research on the occurrence and effects of acid deposition on surface waters in the
United States west of the Mississippi River and on high-elevation ecosystems; conduct
research on the occurrence and effects of episodic acidification, particularly for high-
elevation watersheds.
8. Continue the National Acid Precipitation Assessment Program (NAPAP) and report
information about acid deposition hi a format appropriate for policy makers and the
public.
The Total Deposition work group reviewed existing wet and dry deposition networks and
generally found mem inadequate to meet the objectives of the Clean Air Act Amendments.
particularly in regard to dry deposition. Although NAPAP ultimately received a mandate to continue
under the CAAA, several of the major networks that were designed and funded during the period of
the NAPAP program were reduced in size or terminated in 1990 (such as the MAP3S and OEN
networks), resulting in a reduced field of monitoring sites from which CASTNET could draw. The
group proposed die following approach to establishing wet and dry networks to allow the CAAA
objectives to be met hi the most efficient and cost-effective manner.
Technical Approach
The CAAA requires a total deposition monitoring network to determine spatial and temporal
trends of acid deposition and rural ozone and to address the effects of deposition and
ozone on aquatic and terrestrial ecosystems. Data on the concentration and deposition of acidic
materials and ozone will be collected to:
12
-------
February 1992
* Support water-quality
determinations.
Perform aquatic and
terrestrial effects
research.
Determine regions of
the country at risk.
* Assess materials
damage.
Perform model
maintenance and
application.
Assess the effectiveness
of emissions reductions on air chemistry and deposition.
* Determine transboundary impacts.
Pollutants are deposited to the earth's surface through wet processes (precipitation), and dry
processes associated with the atmosphere's interaction with the surface. Characterizing wet deposition
patterns is much simpler than characterizing dry deposition patterns because of differences in available
measurement systems. Precipitation processes mat remove pollutants from the atmosphere are
comparatively easily studied by analyzing precipitation samples.
Such samples have been analyzed at many sites in North America to characterize the spatial
patterns and temporal trends for wet deposition. Dry deposition patterns have not been similarly
characterized because cost-effective direct measurement systems are not available. Dry deposition is
WET DEPOSITION MONITORING
Network Participation Proposed:
NADP/NTN
States
fannHign networks
Others
Proposed Number of Wet Deposition Sites:
19 additional wet deposition sites
Proposed Wet Deposition Site Locations:
Eastern high elevations (Appalachians)
Western high elevations (Rockies and Cascades)
East and West coasts
Underrepresented low elevation areas in Northeast
Proposed Wet Deposition Variables:
In rain or snow - SO4*. NO/, H+, NH/, O% K+,
Na+, Mg2*, Ca3+, PO4*. pH, and conductance
In clouds or fog - SO4*, NO/, H+, NH«\ CT,
Mg*% and Ca1*
13
-------
February 1992
estimated using an inferential
model approach tor the 50-station
National Dry Deposition Network
(NDDN) and for the nine-station
National Oceanic and Atmospheric
Administration (NOAA) CORE
research network. Fifteen of the
NDDN sites are collocated with
NADP/NTN wet monitoring sites,
and five of the NOAA CORE sites
are collocated with NDDN sites.
The inferential model
approach determines deposition
flux as the product of a measured
concentration and a modeled
deposition velocity based on
measurements of meteorological
variables and physical/biological
DRY DEPOSITION AND OZONE
MONITORING
Network Participation Proposed:
NDDN
NFS
NOAA
States
Canadian networks
Others
Proposed Number of Dry Deposition and Ozone Sites:
An additional 31 dry deposition and 45 ozone
monitoring sites
Proposed Dry Deposition and Ozone Site Locations:
Eastern high elevations (Appalachians)
Western high elevations (Rockies, Cascades, Sierra
Nevadas, and San Bernardinos)
Intennountain West
Great Lakes
East and West coasts
Underrepresented low elevation areas in Northeast,
South, Mississippi Delta, and Plains
Proposed Dry Deposition Variables:
» Air concentrations - SO4*, NO,*, HNOj, O,, and
SO,
Meteorological parameters - wind velocity,
temperature, temperature lapse rate (delta T),
relative humidity, solar radiation, precipitation,
and surface wetness
Vegetation parameters - vegetation status and leaf
area index
surface conditions. The Total
Deposition work group recommends this approach, which can provide relevant information to the
Clean Air Status and Trends Network (CASTNET) at a reasonable cost. Other approaches such as
eddy .correlation, throughfall, and gradient processes, are in an exploratory phase of development.
14
-------
February 1992
Although these "direct" flux measurements have not been used in routine monitoring programs, they
should nevertheless be used at selected sites for comparison to inferential measurements.
For wet deposition monitoring, the principal goal of CASTNET relates to the need to assess
deposition trends and inputs for effects studies. For this purpose, monthly or seasonal averages are
adequate. The wet deposition monitoring community concerned with ecological-effects and trends in
the United States has settled on a weekly sampling protocol. CASTNET will continue this practice,
relying on the existing National Atmospheric Deposition Program/National Trends Network
(NADP/NTN) led by the USGS.
For studies of more localized atmospheric consequences and consequences more closely tied
to individual source regions, shorter-period samples are beneficial. The atmospheric deposition
research community has settled on a daily sampling protocol for mis purpose. CASTNET will rely
on the NOAA AIRMoN (Atmospheric Integrated Research Monitoring Network) program for input of
this kind.
In discussions of the weekly-sampling network, the work group divided the 48 contiguous
states into separate regions based on topographical, chemical concentration, and climatic
characteristics for both acid deposition and ozone for the purpose of establishing reporting regions for
statistical changes. These regions are shown in Figure 1. The monitoring requirements for bom the
wet deposition and dry deposition networks are addressed separately. CASTNET will provide
estimates of total nitrogen and sulfur deposition and of ozone and sulfur dioxide concentrations.
It should be noted here that the approach adopted is most readily applicable to the wet
deposition component. Dry deposition varies from site to site within any given region, and so
regionalization as described above is more imprecise. While this problem is widely appreciated, the
scientific capability to fully address it and to take the corresponding complexity into proper account is
15
-------
-------
February 1992
still under development. It is, in part, for this reason that the integrated network that is anticipated
will need different arrays of stations to address questions of wet and dry deposition.
Description of MeasurementsWet Deposition
Wet depositioD data are needed; (a) to contribute to the assessment of total deposition, as
affects the terrestrial and aquatic environment, and (b) to indicate by simple measurement practice
changes in air quality as may be related to changes hi emissions.
The chemical species or variables proposed for monitoring at the wet deposition sites are
SO,2", NOj', H+, NH4+, C1-, K+, Na+, Mg2*, Ca2+, PO4*. pH^ and conductance. These are the
species and variables currently being monitored by NADP/NTN.
A wet/dry collector manufactured by Aerochem Metrics has been used exclusively hi
NADP/NTN since the late 1970s to monitor wet deposition. Precipitation is measured with a Belfort
recording rain gauge with event marker; pH and conductivity are measured with standard instruments.
The major cations and anions are measured at a central analytical laboratory using proven ion
chromatographic and atomic absorption techniques.
For monitoring cloud/fog deposition (S04*. NO,% H4, NH/, Cl', Mga% and Ca*+) which is
limited to areas above cloud base and along some coastal areas, an automated cloud collection system
developed by the Environmental Protection Agency (EPA), Tennessee Valley Authority (TVA),
ManTech, and North Carolina State University is recommended. This system, chosen after careful
revjew of several designs, provides a level of accuracy and precision comparable to manual methods,
but at lower cost. Manual methods entail collecting droplets on a series of fine strings and analyzing
the droplets for cations and anions.
Field instrumentation and analytical methods are generally well established and have proven
satisfactory for meeting the majority of wet deposition monitoring objectives given a stringent quality
17
-------
February 1992
assurance (QA) program. Some important technical issues, however, heed to be addressed that relate
to data quality and the ability to obtain complete spatial coverage. For example, rain and snow
samples are more easily collected at low elevations than high elevations, particularly at sites receiving
significant snow and cloud/fog water deposition and wind. High elevations often restrict site access,
and severe whiter conditions frequently cause equipment to operate poorly, resulting hi inadequate
sampling of rainfall, snowfall, and cloud/fog deposition. To improve sample collection at high-
elevation and remote sites, establishing alternate sampling protocols may be required. Snow core
sampling may provide the total deposition to an area over a specified time period, but would not be
adequate should separate wet and dry determinations be required. If snow core sampling results
prove inadequate, then improved equipment and sampling methods will be required.
Description of MeasurementsDry Deposition
The chemical species and meteorological and vegetation parameters proposed to be
measured/monitored at the dry deposition sites are:
Chemical species: SO**, NO,% HNQ,, O,, and SO*
Meteorological parameters: Wind speed, wind direction, standard deviation of wind
direction, temperature, temperature lapse rate (delta T), relative humidity, solar radiation,
precipitation, and surface wetness.
Vegetation parameters: Vegetation status and leaf area index (LAI).
Currently, the NDDN and the National Park Service (NFS) monitor these species and variables, with
the exception of vegetation status and leaf area index at all NFS sites, and HNQ, at some of the NFS
sites. The NOAA dry deposition CORE program makes all of these measurements, but also focusses
on improving the methodologies for inferring dry deposition rates from such information and on
testing the quality of the answers by comparison against research grade techniques.
18
-------
February 1992
The NDDN measures concentrations of SO,*, N0j% HN03, and SOa using a filterpack.
Ozone is measured on a continuous basis using a UV photometric analyzer. Because ozone is one of
the atmospheric constituents for which a National Ambient Air Quality Standard (NAAQS) has been
set, the detection limits, measurement range, precision, and accuracy of ozone measurement methods
are well defined. Meteorological measurements use systems that meet established stringent
requirements. Filterpack techniques based on initial developments at Brookhaven National Laboratory,
and like the methods popularized in Canadian networks, have been used in several networks to
measure sulfur dioxide and nitric acid and is the technique currently used by the NDDN. Several
variations on this general design have been developed, with slightly differing performance
characteristics that are still being explored. The critical uncertainty relates to the precision of
measurement of HNO3 and SO2, bom of which are often present in sufficiently low concentrations mat
simple filterpack methods do not always work well. The accuracy of filterpack techniques has been
estimated to be of the order of 25 percent for HNO,, and somewhat better for S02 - currently
adequate for estimating deposition rates, but not sufficiently precise for applications such as the
testing of regional models..
A real-time UV fluorescence technique for continuous measurement of low-level sulfur
dioxide is commercially available. Although this is the ideal measurement method for sulfur dioxide,
it is expensive. Various types of filterpack or annular denuder integrated collection systems have
been used in some networks. On a 24-hour collected sample, the minimum detection limit, precision,
and accuracy of the integrated techniques compare favorably to real-time monitors. Samples collected
for periods of a week are subject to problems inherent in the system (such as compound interactions
and gas loss) and therefore tend to yield greater uncertainties. Considering all of these factors, the
filterpack technique remains the recommended method for measuring SO2 at this time, as the
19
-------
February 1992
projected uncertainty is within required limits. The work group also recommends placing low-level
continuous S03 monitors at several sites with varying concentrations for comparison purposes.
Although many techniques are available to measure ambient nitric acid, a major end product
of nitrogen oxide emissions, few of those techniques are applicable to network use. Concentrations of
nitric acid are low in rural areas, but dry deposition of nitric acid can add to the total nitrogen
deposition to forests. For. nitric acid vapor, annual denuder systems (integrated 24-hour) provide
improved speciation, however, use in networks has been limited because the applicability to weekly
sampling is not well defined and because additional evaluation is needed. While the filterpack
technique does not provide an unambiguous measurement of nitric acid, it does provide a
measurement of total nitrate. Thus, the filterpack technique is also recommended for measuring nitric
acid in the network at this tune.
Adding to total deposition are coarse and fine particles containing sulfate, calcium, nitrates,
ammonium, aluminum, and other metals. The particles usually are collected on a substrate or filter, a
filter being the case in the filterpack technique, These samplers may or may not have devices to
separate particles into size ranges before analysis. Chemical Con chrpmatographic) or physical (X-ray
fluorescence) techniques will be used for analysis.
In general, recent developments in measurement methodology indicate a shift toward multi-
constituent samplers (such as annular denuders), especially for dry deposition. Multi-constituent
samplers, as well as integrated sampling methods (passive samplers) for ozone, could have a
significant effect on network design because of the cost savings potential. Research continues on such
measurement technology. As the methods become more suitable for large- scale implementation, they
will be considered for use in the network.
20
-------
February 1992
Observations of vegetation status and leaf area index are made periodically at all dry
deposition sites so that dry deposition loadings to vegetation can be calculated using the inferential
approach.
Weekly sampling is proposed for both of the wet and dry deposition routine networks, with
the exception of ozone to be measured continuously and cloud/fog deposition to be measured on an
episodic basis. More frequent sampling, such as daily/event sampling, may be needed for special
studies and is the focus of the NOAA/AIRMoN program. Diurnal sampling (daytime vs. nighttime)
integrated over a week may give better estimates of dry deposition than a single weekly sample.
Instrument and sampling protocol development will be an integral part of CASTNET. New
instruments such as denuders and new protocols such as day/night sampling will be evaluated with
regard to precision and accuracy, cost, comparability, and network suitability. Changes in network
protocol will be made only after thorough evaluation and consultation with the work groups and
concerned parties, including the states and other federal agencies.
Treatment of Confidence and Uncertainty
Table 1 shows method and spatial interpolation uncertainty considerations for wet and dry
deposition; the numbers for dry deposition represent weekly integrated values for inferential as well
as measurement (eddy correlation) approaches. Table 2 shows the precision,
accuracy, and site-specific uncertainty estimates for the major species measured. The precision and
accuracy estimates for. site-specific measurements are acceptable in regional stressor monitoring for all
constituents measured by the NADP/NTN and the NDDN except HNQ, in dry measurements and H*
in wet measurements, which are marginal. The variability of sulfate wet deposition trend predictions
has been reduced by adjusting the data for the effects of precipitation (as measured by the
NADP/NTN). Future research will incorporate National Weather Service (NWS) precipitation data
21
-------
February 1992
TABLE 1. METHOD AND SPATIAL INTERPOLATION UNCERTAINTIES
Species
Wet
Dry
Gases
Aerosols
Deposition
(Inferential)
Method
±5-15%
±5-20%
±5%
±30-50%
}
Deposition
(Eddy Correlation/
Gradient Approach) ± 15-30%
Spatial Interpolation
(current vs. goal)
±20-200% vs. ±25-40%
±25-200% vs. ±25-40%
±50-200% vs. ±25-40%
Meteorology
±10-50% N/A
(the NWS has a denser network of rain gauges than the NADP/NTN) to improve the accuracy and
precision of spatial interpolation estimates. ...
Statistical analyses conducted to date on the spatial interpolation of air concentration and wet
deposition data indicate that uncertainties in concentration estimates are lower than those of
deposition. It is difficult to determine empirically the accuracy of measurement and inferential
approaches to monitoring dry deposition because accuracy varies according to the species,
vegetation, terrain, and the nature of the major sources affecting the site. By all methods, the
uncertainty generally should decrease as integration time increases. In other words, monthly,
seasonal, and annual values should yield decreased uncertainty compared to weekly values.
22
-------
Q
CM
CO
a g
'-S3 '5
fll
CO
I
-------
February 1992
The gradient approach for HN03 is considered promising because of the large gradient often
encountered. Uncertainties in the eddy correlation approach are estimated to be 15 percent for O, and
=20 percent for S02. Uncertainties in the inferential approach are greater in all cases: approximately
50 percent or greater for HNO,, NO,', and SO,2", and 30 percent for SO2 and O,,
Hie numbers generally reflect the level of experience and accuracy in measuring
concentrations. For example, direct eddy correlation measurements of Q, flux have been made for at
least 10 years, resulting in a refined technique, decreased uncertainty, and a good database for
parameterizing the inferential method.
Although eddy correlation flux measurements for S02 are becoming more common, direct
measurements for SO/ and HNO, are not, which is reflected in the high uncertainty of the inferential
approach. The statistics for dry deposition uncertainty in Table 1 are for relatively uniform terrain of
low and uniform vegetation cover. Uncertainty is greater for tree-covered, complicated terrain
because few direct measurements have been made in such environments and because some of the
processes are poorly understood. Extending the inferential approach to estimate deposition on a large
. . . «
area adds additional uncertainty because of the difficulty in characterizing the vegetation cover over a
large area and because the meteorology and ambient concentrations are more likely to vary
significantly. Uncertainty estimates for warm-region cloud deposition are based on the Modified
w'
Cloud Deposition Model. Estimates are for a site with a uniform forest canopy, with discrete cloud
water collection periods in which the physical input parameters for the model are well characterized.
Sites and Implementation
The Total Deposition work group's initial recommendation is that 51 additional site locations
(measuring wet deposition, dry deposition* ozone, or a combination thereof) will be required to fulfill
the mandates of the CAAA. Figures 2 and 3 show maps of existing wet deposition sites
24
-------
r*
I
-------
-------
February 1992
(NADP/NTN and some Canadian sites) and dry deposition sites (NDDN, NOAA, Wisconsin, Florida,
and some Canadian sites).
Figure 4 reflects potential locations of these 51 additional sites and is intended merely as a
starting point for further analyses and deliberations on specific site selections. Figure 5 shows the
site types proposed based on preliminary analyses. The initial recommendation is that of the 51
proposed sites, 6 would monitor dry deposition only; 14 would monitor dry deposition and ozone; 11
would monitor dry deposition, wet deposition, and ozone; 8 would monitor wet deposition and ozone;
and 12 would monitor ozone only. This results in 31 dry deposition monitoring sites, 19 wet
deposition monitoring sites, and 45 ozone monitoring sites. To create the most efficient and cost-
effective monitoring network possible, it is assumed that wet, dry, and ozone sites will be collocated
whenever possible. Although this discussion is focusing on acid deposition and ozone monitoring
sites, it is not meant to preclude the possibility of collocation with either visibility or air toxics
monitoring sites. Further, during actual siting the proximity and availability of existing comparable
sites will influence final site selection.
Figure 4 was created from the integration of the monitoring needs for the Total Deposition,
Aquatic and Terrestrial Effects, and Statistical Network Design work groups. Implementation of the
51 proposed monitoring sites is planned in two stages reflecting the likelihood of a multi-year build-
up to fit within available resources.
Stage I is composed of 13 sites whose immediate implementation is deemed most crucial to
CASTNET. Stage n is composed of the remaining 38 key sites. The goal for the Statistical Network
Design group, was to systematically fill gaps in under-represented areas, whether for sensitive
ecosystems such as the Appalachians or for providing spatial coverage adequate to reduce
interpolation errors to acceptable levels. This latter goal requires placing some sites within 150 km of
27
-------
a
00
O
fe
60
S
V)
jfi
£
-------
29
-------
February 1992
Total implementation of the needed sites will take from one to three years, depending on the
number of sites in existing networks mat can be upgraded and incorporated into the NADP/NTN or
NDDN.
STAGE I
3 EutemcoiM
S Northent AppthcbiaD Mwntriiv
4 Mid-Southern ApptUduan Mountain*
1 Uader-fcpneoud ireu in Nonhead
(MB
2 and in South (NC and G A)
TOTAL 13
Number of She*
3
2
2
1
4
3
4
4
1
1
2
2
3
C
TOTAL 31
STAGED
Location
Eotem (2) cod Weftrm (1) coach
Cascade Mountain
Siena Nevada Mountain*
g f-t |fL _!>« |m
Rocky Mcmduai
Gre«tLike»
Untar-KpmHntod mat oi-
Noi1neut/Nof«teutemApi«iMhBBi
Mounlu»(VT. NH. CT, NY)
Mid-Api»bdkaa MMBHUDI (PA)
Midwen(OH)
Phum(MN«ndAR)
Mbnntnu Delta (LA. TX. MS)
South (1 tnek m SC ml PU 2 «ch it
ALadGA)
Relationship with Existing Networks
In addition to the NADP/NTN wet deposition network and the NDDN dry deposition
network, other networks in the United States will be considered for integration into the CASTNET
program. Some of these networks are listed in Table 3, "Characteristics of Nonurban Monitoring
i
Networks in Norm America." Another potential rich source of sites not listed in Table 3 are those
belonging to state-operated networks.
As noted in the previous section, specific locations of new sites cannot be identified precisely
at this time, nor can sites from other networks be identified that would be appropriate to include in
30
-------
TABLE 3. CHARACTERISTICS OF SOME NONURBAN MONITORING NETWORKS IN
NORTH AMERICA
NETWORK
NADP/NTN
MAP3S/PCN
UAPSP(OEN)
EPA-SON
TVA
CARB(CA)
NEW YORK
PADMN (PA)
CLAD (Gnat
Lakes)
NDDN
OEN(UAPSP)
NPS(GASES)
NFS/IMPROVE
NFS
FADMP(FL)
CAPMoN
APIOS-C
APIOS-D
QUEBEC
SITES
200
9
25
27
32
13
7
7
8
55
25
42
16
17
7
4
24
11
38
i
17
4
51
TYPE
Wet only
Wet only
Wet only
Wet only
Wet only
Wet only
We«/Diy
Wet
Wet
WeJ/Diy
Wet/Dry
Diy
Dry/Vis
Dry/Vis
Wet
Dry
Wet
Dry
Wet only
Wet
Dry
Wet only
VARIABLES
(1)
(1)
(3)
(2)
(2)
(1)
(")
(1)
(12)
(1).(5)
W)
O
(8)
(8) .
(«)
(5)
(2)
(10)
<<)
(2)
(5)
0)
FREQUENCY
Weekly
Dairy
Daily
Weekly
14 Day
Weekly
Weekly
Weekly
Weekly
Weekly(cont.Oj)
Daily(cont.NO2)
continuous
continuous
continuous
4-Daily
- 3-Dairy
2-3 Day
2-Daity
Daily
Daily
28 day
Dairy
Dairy
' Weekly
INITIATED
1978
1976
1981
1982
1978
1985
1987
1983
1981-82
1987
1988
1988
1980
1980
1982
1981
1983
1980
1980
1981
STOPPED
Continuing
1991
1988
Continuing
1987
1993
Continuing
Continuing
Continuing
Continuing
1990
Continuing
Continuing
Continuing
Continuing
Continuing
Continuing
Continuing
Continuing
Continuing
AGENCY
NADP/USGS
DOE/NOAA
EPR1
EPA
TVA
CA/ARB
NY/DEC
PA/DER
EPA
EPA
EPRI
NPS
NPS/EPA
NPS
FCG
(Florida Power
Companies)
AES (Canada)
OME (Ontario,
Canada)
Quebec.
Canada
(1) SO^ NOj, Q, PO4> H, NH* Ca, Mg, Na. K, spedfic conductance and precipitation depth.
(2) Same as (1) including SO3.
(3) Same as (1) except conductance not measured. '
(4) Same as (2) including strong acid, total acid, and total organic carton.
(5) Same as (1) except total P and N and including Zn. Fe, Ni, Cu, Pb, Cd, Mn, and V.
(6) 03 (continuous); Our pack SO2 and HNOj vapor, paniculate SO4 and NOj; met data includes windspeed, wind direction,
. temperature, relative humidity, solar radiation, precipitation, and delta temperature.
(7) Same as (1) including At (total), organic carbon, acidity.
(8) O3 (41 sites continuous); SOj (14 sites continuous. 17 sites 2-24 hr. samples/week); met data at 36 sites (see (6)).
(9) Fine Particles and Visibility - 33 sites (16-NPS/IMPROVE, 17 NFS) measuring mass, SO* NO,, H, C, N, O, Na through Pb,
total organic and elemental carbon, all with cameras and IS sites with transmissometers.
(10) Same as (6) plus Q, NH* Na, K. but no met data.
(11) Same as (1) plus NO* Br, F, HPO4 ions for wet samples.
(12) Same as (1) except total P and including Cd, Hg, Pb, Ni. Cr. As, Cu, PC, Al, B, Be, Ba, Co, U, Mn, Sr, Ti, V, Zn.
1A5/92
31
-------
February 1992
the future monitoring program. Because of differences in siting and sampling criteria, many of the
sites in existing networks are not suitable for inclusion in CASTNET. Identifying the exact location
of new sites will require an analysis of the needs in under-represented areas and an analysis of current
data to determine whether present site density is adequate to meet future data-quality objectives. It is
hoped that many new site needs will be met by matching CASTNET requirements with operating
characteristics of some existing sites under one or more of die referenced networks. Subsequent to
design, a matching process will be conducted with existing networks to discuss and arrange sites on a
case-by-case basis.
The Statistical Network Design work group will be involved in determining the areas and
numbers of new sites needed for wet and dry deposition. Assessing data comparability of existing
networks is essential before historical data sets can be merged. Concerns focus on different levels of
uncertainty within and among networks which could confound attempts to determine spatial patterns
and temporal trends. , .' '
. ',../* -
Report Formats _ ...
Existing protocols for data reporting will be followed. All concentration and meteorological
data will be provided in electronic format for incorporation into the Aerometric Information and
Retrieval System (AIRS). The AIRS system is administered by the EPA's Office of Air Quality
Planning and Standards (OAQPS) and was put in production in July 1987 as the database management
system for the national database for ambient air quality, emissions, and compliance data. The wet
deposition program will be an enhancement of the existing NADP/NTN program. NADP/NTN will
continue producing an annual report that lists weekly precipitation chemistry data and precipitation
amounts for each site. NADP/NTN also produces annual and seasonal statistical summaries for each
32
-------
February 1992
site. EPA will, combine these annual summaries with dry deposition data and summaries to produce
national concentration and deposition isopleth maps.
The inferential deposition values for dry deposition will be calculated for weekly periods and
integrated to seasonal and yearly values for inclusion in the AIRS database. When combined with wet
deposition data, the air concentration and dry deposition data will be used to produce annual and
seasonal summaries, spatial maps (when appropriate), and trend statistics for the CASTNET program.
Future Research Needs
There are several research needs for total deposition monitoring. Investigations should focus
on:
Evaluating the benefit/cost for shorter sample-averaging times,
Evaluating inferential deposition models at sites with different topographical and ecological
structures. > . .
Improving wet deposition sampling methods at high-elevation sites.
Incorporating hydrogen and ammonium ion stabilizing techniques into wet deposition
samples. : ,: ,
Determining the applicability of integrated techniques for measuring ozone.
Sample-Averaging Times
An evaluation of sample-averaging times is recommended to determine whether weekly
sampling is adequate. Research has indicated mat biases of up to 20 percent in the calculated dry
deposition may result when a weekly (168-hour) integrated concentration is used in the calculation
rather than two separate, 84-hour, day-and-night integrated periods. Methods have been developed to'
adjust for such effects, but as yet the overall accuracy of the inferential method does not warrant such
fine tuning. If the need for day/night (or some similarly more detailed) dry deposition sampling is
33
-------
February 1992
indicated, collocation with weekly samplers at a limited number of sites should be a first step. Hie
NOAA dry deposition CORE network was designed with the intent to provide side-by-side
comparisons of routine weekly sampling protocols with weekly integrals of dry deposition
determinations based on hourly data. These stations provide an ongoing research basis for tests of
different sampling protocols.
For wet deposition sampling, weekly sampling is the U.S. continental norm, but other
networks in the United States and especially in Canada use daily sampling. In practice, the two major
goals for wet deposition sampling impose different sampling criteria: for annual and seasonal trends
detection and for ecological effects studies, weekly samples are adequate, but for relating deposition
to specific atmospheric events and for detecting short-term trends from monitoring data, daily samples
are preferable. The integrated national program of which CASTNET is the keystone must utilize both
approaches, since a number of different goals must be addressed. If the need for day/night dry
deposition sampling or event/daily wet deposition monitoring is indicated, collocation with weekly
samplers at a limited number of sites will be a first step in determining the effect of sample-averaging
times. Further research underway now will help determine the optimum mix of weekly and daily
samples in the network. ......
Wet Deposit/on
Improved sampling methods for high elevations are urgently needed; snow and cloud water
might be me most significant routes for chemical deposition at high elevations. Existing equipment
makes collection and measurement of these samples difficult, especially in remote areas. In areas
where sampling on a weekly basis is impossible due to severe weather conditions and access
problems, development of alternative sampling methodologies may be the only viable option. Current
networks, with the exception of six NADP/NTN sites in Colorado and Wyoming, do not measure wet
34
-------
February 1992
deposition in high-elevation areas of either the eastern or western United States, even though these
areas represent some of the nation's most sensitive ecosystems and have the greatest amounts of
precipitation.
Currently, wet samples can undergo significant hydrogen and ammonium ton loss over very
short time periods as a result of chemical reactions whose mechanisms are not fully understood.
Most of the change in collected samples occurs in the first day. The Global Precipitation Chemistry
Program of NOAA has established a method, by the addition of a biocide, for preserving nitrogen and
hydrogen ions in precipitation. This methodology has been widely tested, and is available for
exploitation if further analysis reveals that the chemical change of deposited materials during storage,
prior to chemical analysis, is an important factor.
Dry Deposition .
The inferential approach currently used to assess dry deposition is a modeling technique
through which surface deposition is calculated from ambient concentrations. Additional research is
needed to unprove the model's ability to make accurate estimates. Such research should focus on
developing unproved empirical and theoretical characterizations of the dry deposition processes and
on refining appropriate measurements of the governing variables. Past cooperative efforts hi mis area
by EPA, NOAA, and the Department of Energy (DOE) should be built upon.
Concerns of die current inferential model are:
Biological interaction of pollutants with vegetation species. .
* Impact of water stress on plants.
Parameterization of surface turbulence processes.
Observations and impact of surface wetness.
Deposition to snow cover.
35
-------
February 1992
A viable research program to address these concerns would involve further analyses of existing
databases of collocated inferential and eddy correlation measurements. Additional measurements of
dry deposition are therefore recommended by direct techniques, such as eddy correlation, and indirect
methods such as throughfall, and gradient/Bowen approach.
Research should also determine the most efficient and cost-effective method of collecting the
concentration measurements on which the inferential approach is based. Currently, the filterpack
method is used to determine pollutant concentrations, which are multiplied by the modeled deposition
velocity based on measurements of physical and biological variables. Improving the accuracy of
pollutant concentration measurements would increase (he confidence in the inferential approach.
Measurements of nitric acid vapor appear especially vulnerable.
Because of the potential acute effects of ozone on biological systems, past development of
measurement methodology has focused on measurement in real tune. Recent developmental efforts,
on the other hand, focus on daily or weekly integrated techniques for measuring ozone. If integrated
techniques are determined to be applicable to effects monitoring, spatial coverages could be attained at
reduced cost. - "
Regional Deposition Modal Evaluation .
Mathematical models requiring a variety of physical and chemical measurements are used to
provide spatial estimates of dry deposition and cloud deposition. Uncertainty estimates can be
assigned to model outputs only for specific sites at which chemical and physical parameters are very
well defined, or at which independent verification has been achieved. Information is available on
model development, sensitivity analysis of input variables, and evaluation of models at sites meeting
certain criteria. Such sites generally have uncomplicated terrain and low and uniform vegetation.
Additional research is needed to evaluate the models at sites with different topographical and
36
-------
February 1992
ecological structures, such as forests in complex terrain for the dry deposition model and non-
homogenous forests for the cloud deposition model.
37
-------
February 1992
THIS PAGE INTENTIONALLY
LEFT BLANK
38
-------
February 1992
Aquatic and Terrestrial Effects Monitoring
Monitoring Objectives
The Clean Air Act Amendments (CAAA) define two categories of aquatic and terrestrial
effects monitoring: effects on water quality, and biological effects on aquatic or terrestrial
ecosystems. Further, Title IV, Section 404 mandates identifying sensitive and critically sensitive
aquatic and terrestrial resources and evaluating the feasibility/effectiveness of an acid deposition
standard or standards to protect them.
Title DC, Sections 103(c), (e), (j), and 901 (g), of the CAAA requires aquatic and terrestrial
effects monitoring and research and specifically mandates the following:
1. Establishing a national network to monitor status and trends of surface water quality and
forest conditions and to develop monitoring to characterize regional ozone trends.
2. Evaluating risks to ecosystems exposed to air pollutants, including characterizing the
causes and effects of chronic and episodic exposures and determining the reversibility of
those effects.
3. Evaluating air pollution effects on water quality, including the short- and long-term
ecological effects of acid deposition and other atmospherically derived pollutants on
surface water (including wetlands and estuaries) and groundwater; and evaluating air
pollution effects on forests, crops, and soil.
4. Continuing the National Acid Precipitation Assessment Program (NAPAP) and providing
reports on the status of ecosystems (including forests and surface waters) affected by acid
deposition, as well as reports on the causes and effects of such deposition, including
changes hi surface water quality, and forest and soil conditions.
39
-------
February 1992
5. Researching the occurrence and effects of acid deposition on high-elevation ecosystems,
including forests and surface waters.
The work group has identified sensitive and critically sensitive aquatic and terrestrial
ecosystems and also has outlined the appropriate biological and stressor measurements to relate to
effects.
Three Types of Monitoring Required
NAPAP addressed the biological effects of acid deposition on lakes and streams and the
biological effects of acid deposition and ozone on forests. The results of the aquatics biological
effects research produced more definitive answers than the forest research, allowing inferences of fish
health when changes in water chemistry are known. Considerable research is continuing on the
effects of ozone and acid deposition on forested ecosystems; this research will be supplemented
directly through CASTNET. The monitoring of effects refers in this context to indirect chemical
measures, such as acid neutralizing capacity (ANC) and pH, and direct biological measures, such as
fish species, for aquatics; and indirect chemical measures, such as soil chemistry, and direct measures
such as visual damage, growth, and nutrient uptake for forests.
The Aquatic and Terrestrial Effects work group determined the need for three types of
monitoring. Two types are effects monitoring: Chemical indicators monitored will be used to
measure the acidity of surface waters and the chemical balance of forest soils, and biological
indicators monitored will be used to assess ecological health. In addition, atmospheric monitoring
will be necessary to support effects research in sensitive ecosystems as well as to establish a sufficient
database to allow for the interpretive and integrative status and trends assessments required.
Regional-scale monitoring will be needed to address die many different ecosystem types and different
levels of atmospheric inputs.
40
-------
February 1992
AQUATIC EFFECTS
Water chemical measurements EMAP/TIME
Water biological measurements EMAP-Surface Waters
Atmospheric chemistry CASTNET (with others)
monitoring
TERRESTRIAL EFFECTS
FHMP (wi«fa USFS)
Biological and chemical
measurements
Cause and effects research
Cause and effects research
and some national
atmospheric monitoring
Atmospheric chemistry
monitoring
GCRP (through USFS)
NFS
CASTNET (with others)
For aquatic effects
monitoring, the Environmental
Monitoring and Assessment
Program (EMAPyTemporally
Integrated Monitoring of
Ecosystems (TIME) will conduct
water chemical measurements,
EMAP-Surface Waters will
conduct water biological
measurements, and CASTNET
(with others as described
previously, including the
NADP/NTN and NDDN) will
conduct atmospheric chemistry and deposition monitoring. In addition, data on water chemical and
water biological measurements from federal Class I areas will be available from the USFS's Air
Resource Management Monitoring program. For terrestrial effects monitoring, the joint USFS-EPA-
EMAP-NPS Forest Health Monitoring Program (FHMP) will conduct biological and chemical
measurements; the USFS Global Change Research Program (GCRP) will conduct cause and effects
research; the National Park Service (NPS) will conduct cause and effects research and some national
atmospheric monitoring; and CASTNET (and others) will conduct atmospheric chemistry and
deposition monitoring.
Because the issues addressed by the Effects work group are complex and interrelated, some
aspects of aquatic and terrestrial.effects will be discussed separately.
41
-------
February 1992
Technical ApproachAquatics Effects and Stressor Monitoring
To meet the needs of Title DC for
aquatics effects assessment, largely created
for acid deposition control, the monitoring
design under EPA's Environmental
Monitoring and Assessment Program
(EMAP) was enhanced in the form of the
Temporally Integrated Monitoring of
Ecosystems (TIME) project. TIME'S goal
is to estimate, through long-term
monitoring, the changes and trends in
chemical conditions in acid-sensitive
surface waters of die United States that
result from changes in acidic deposition.
The TIME project will achieve mis goal
through a combination of annual,
probability-based surveys of lakes and streams in selected regions and intensive monitoring of a few
regionally representative lakes and streams. Although TIME itself is a chemical monitoring project, it
has been designed in parallel with the surface waters component of EMAP, which has a broader
ecological focus. The goals of EMAP-Surtace Waters are to:
Estimate the current extent (location, number, surface area or length) of the nation's lakes
and streams on regional and national scales, with known confidence.
ENVIRONMENTAL MONITORING
AND ASSESSMENT PROGRAM (EMAP)
EMAP (the Environmental Monitoring and
Assessment Program) was established by EPA as a
program aimed at monitoring for results; that is,
confirming that the nation's environmental
protection efforts are truly maintaining or
improving environmental quality. Planning is
being conducted in cooperation with other agencies
and organizations that share responsibilities for
renewable natural resources or environmental
quality. Over the next five years, integrated
monitoring networks will be designed and
implemented with the following objectives:
1. Estimate current status, extent, changes,
and trends in indicators of the condition of
the nation's ecological resources on a
regional basis with known confidence.
2. Monitor indicators of pollutant exposure
and habitat condition and seek associations
between human-induced stresses and
ecological condition.
3. Provide periodic statistical summaries and
interpretative reports on status and trends
to the EPA Administrator and the public.
42
-------
February 1992
Estimate the current status, changes, and trends in indicators of the ecological condition of
the nation's lakes and streams, on regional and national scales, with known confidence.
Monitor indicators of pollutant exposure and habitat condition within lakes and streams
and seek associations between human-induced stressors and ecological condition that
identify possible causes of adverse effects.
Publish annual statistical summaries on the extent and the status of indicators of ecological
condition of lakes and streams, and publish periodic interpretative reports on the status
and trends of indicators of ecological condition of lakes and streams to die EPA
Administrator and the public.
EMAP draws on the success of the National Surface Water Survey (NSWS) in applying a
probability design to regional surveys. With this design, sites (for example, lakes and streams) are
randomly chosen from a target population (for example, all lakes greater than one hectare in size and
greater than one meter in depth) within a given region. This subsample of sites is then sampled, and
results are extrapolated to the target population that the subsample represents. As an example, the
NSWS utilized a probability design to make estimates (with known confidence) of the number of
acidic lakes (acidic lakes are defined as those with acid neutralizing capacity, or ANC, <_ 0 jteqL'1)
in seven major regions of the United States. EMAP will extend this approach to monitor long-term
indicators of ecological health for all major resources (surface waters, forests, wetlands, etc.) of the
United States, without focussing on any specific environmental stress (such as acid deposition or
habitat modification). Indicators implemented as part of EMAP-Surface Waters will be primarily
biological (species composition of fish, diatoms, zooplankton, etc.), in keeping, with the EMAP goal
of monitoring ecological condition. Probability sites hi EMAP will be sampled on a 4-year rotation,
with one quarter of the sites being sampled each year during a summer index period (July-August),
43
-------
February 1992
and returning to the same set of sites in the
fifth year as in the first year.
The objectives of the TIME project
are a specific subset of the objectives of
EMAP. The first objective is that TIME
must be able to make regional estimates of
changes (or trends) in acid/base status.
Secondly, the project must be able to
detect regional trends in acid-sensitive
systems in a policy-relevant time span
(< 10 years). Thirdly, TIME must be able
to relate changes in acidic deposition (of
nitrogen and sulfur) to regional changes in
the acid/base status of surface waters.
And lastly, the network should be national
in scale, but regional in implementation,
allowing phased implementation in line
with funding availability.
In order to make regional estimates
of trends in the acid/base status of surface
AQUATIC EFFECTS MONITORING
Network Participation Proposed:
EMAP-TTMEandRLTM
EMAP-Surface Waters
CASTNET (with others)
Proposed Number of Deposition and Ozone Sites:
A* described in the Total Deposition section, 51
ute location* monitoring wet depoiiiion, dry deposition,
ozone, or a combination thereof
Proposed Effects Monitoring Locations!
TIME- Northeast lake* in 1991. Mid-Atlantic Highland
ttieamt in 1992 and Upper Midwest streams in 1993.
Episodic monitoring (RLTM sites) in the Sierra Nevada
Mountain* and Colorado Rockiea. Northeastern
dreams, Southern Blue Ridge stream*, Mid-Atlantic
Coastal Plain streams, Florida streams, sad Florida and
Northern Cascade lakes later
Surface Waters - Northeast lakes in 1992 and Northeast
streams in 1993
USFS - Federal Class I areas
CASTNET (MA others) - Highest-priority areas are
identical to thaw of TIME. Proposed site locations
are detailed in the Total Deposition section
Proposed Effects Variables:
TIME - Dissolved oxygen, temperature, site depth, and
secchi depth in the field. Air-equilibrated pH, closed
htadspace pH, ANC, specific conductance, SO,*'.
NO,'. 0-, Ca*.*.- Nt*. Mg»*; K*. NH,*, silica, total .
mononieric aluminum, organic monomeric ^Iviriinyni,
total dissolved aluminum, dissolved organic carbon
. - (DOC), dissolved inorganic carbon (DIG), color,
turbidity, total nitrogen, and total phosphorus in die
laboratory
KLTM - All major ions, pH, ANC, conductance,
ahimiiMlm Mid DOC
Surface Water* Chlorophyll; fish, macroinvertebrate,
ijrf zooplankton SMTIM MJ fltvinrfflnff^i sediment
diatoms; physical habitat; and possible sediment
toxiciry, ^*K tiffut c.flntnp*"t«tifin- itwf shoreline bird
species and abundance
CASTNET fivlth other*) - Proposed wet and dry
deposition variables to be monitored are detailed in
the Total Deposition section
waters, TIME will use the same
probability design as EMAP uses. This probability design, like that of the NSWS, allows results to
be expressed as estimates of population characteristics (such as the number or proportion of acidic
44
-------
February 1992
lakes in a region, or the change in mean ANC of lakes within a region over time). This regional
focus is important, as effects of emissions reductions will be felt on a regional scale. Regions for
TIME will be identified on the basis of two important criteria: (1) the selected areas should exhibit a
uniform response (change in acidic deposition) as a result of a change in emissions reductions (that is,
the regions should be airsheds); and (2) the region should contain surface waters with similar gross
characteristics that will make them sensitive (or insensitive) to acidic deposition. By the first
criterion, for example, the Northeast is considered a region, but the Adirondack Mountains are not;
the response to emissions reductions would be expected to be uniform across the entire northeastern
United States because prevailing wind patterns expose the region to emissions from the same source
areas. By the second criterion, the Northeast would be separated from, for example, the Southern
Blue Ridge Province, even if the two areas were expected to undergo the same changes in deposition,
because the two regions have dissimilar surface water characteristics mat can be expected to make
mem react very differently to changes in acidic deposition. High-interest areas will be those where a
large proportion of surface waters are known to be susceptible to acidic deposition and where
acid/base conditions are expected to change. This change in acid/base status might be a response to:
(1) a future decrease in deposition (for example, as a result of the CAAA emissions reductions); (2)
high levels of deposition in die past (for example, in regions with delayed responses); or (3) future
increases in deposition (for example, in areas like the mountainous West, unaffected by the CAAA).
High-interest regions for TIME are shown in Figure 6.
A key factor in the design of TIME is the ability to detect trends in a policy-relevant time
span. The design of TIME, particularly the selection of a sample size (the number of sites monitored
in a single region over a 4-year sampling rotation), is geared toward detecting trends in regional mean
values of acid/base indicators (for example, ANC, pH, SO**, NO3~) in 10 years. The strategy for
45
-------
o c
I
-------
February 1992
trend detection in TIME requires that all sites in each region be assigned to a unique subpopulation or
class. The easiest way to conceptualize this process is to think in terms of a geographical
classification. According to this scenario, all sites in one contiguous geographic area (for example,
the Adirondack Mountains or the Maine coastal plain) would be grouped together in a single class.
Trends would then be identified separately for each class or subpopulation of sites. .There are two
objectives in the classification exercise: (1) to lower the site-to-site variability in each class, thereby
improving the ability to detect significant trends; and (2) to group together sites that are likely to
show the same response to a change in deposition. Monitoring for trends in subpopulations has the
added advantage that TIME can determine which groups of sites (for example, the most sensitive) are
exhibiting improvements and which sites are not. The actual method used to identify subpopulations
for TIME will not be simply geographical, as in the example given here, but will include more
explicit geochemical and biogeochemical information, and will be designed to make more refined
distinctions between classes of sites man can be done simply with geographical location. The sum
total of trend test results for each of the subpopulations may then give a more precise picture of
improvement, or lack of improvement, for the entire region than if trends are examined only for the
entire regional population. Initial results for the Northeast suggest that TIME will be able to detect
(with 90 percent certainty and approximately 90 percent power) regional trends of ±5 /teqL"1 in
t
ANC, SO/, and NO,' in 10 years (for example, trends of ±0.5 peqL*1).
TIME must also relate observed changes in regional acid/base status to regional changes in
deposition. This suggests that TIME and CASTNET will need to be able to detect trends in at least
two key acidifying elements, sulfur and nitrogen, and in two separate media, deposition and surface
waters. For most of the country, we can confidently expect to be able to measure changes in SO**
concentrations in surface waters by sampling regional probability sites on an annual basis. This is
47
-------
February 1992
because the relationships between sulfur emissions, sulfur deposition, and surface water SO*
concentrations are fairly straightforward and direct, and because S04* concentrations are well-
characterized by sampling in a single index period (that is, SO*2" concentrations exhibit little seasonal
variability). In some regions, however, ANC and NO3* undergo pronounced short-term changes
during snow melt or seasons of heavy rainfall.
The incidence and severity of short-term, or episodic, acidification of lakes and streams is one
of the largest unanswered questions in the area of aquatic effects of acidic deposition, and one that is
specifically addressed in the CAAA. Biological damage may result in systems that are not chronically
acidified, but that do experience episodic acidification, and so the assessment of episodic acidification
has been made a priority for TIME. Long-term changes hi ions mat show high seasonal variability
will not be adequately assessed by sampling probability sites during a summer index period, or during
any other single index period. In order to monitor changes in the occurrence of episodic events,
TIME will utilize a second tier of sampling sites, where monitoring will be conducted on a monthly
(during most of the year) to weekly (during snow melt) basis. These sites (known as Regionalized
Long-Term Monitoring, or RLTM, sites), will be hand-chosen to be representative of each of the
subpopulations of sites identified for regional trend testing (described above). The TIME design will
therefore consist of a number of large regions, each of which is broken up into a number of coherent
subpopulations. Within each subpopulation, a large number of probability sites will be monitored
annually during a summer index period, and a small number of RLTM sites will be monitored more
frequently. Results from the probability sites will be used to estimate regional trends, while results
from RLTM sites will be used to estimate the occurrence of short-term events. By explicitly .
associating RLTM sites with classes or subpopulations of sites hi the probability sample, TIME will
also be able to make strong inferences about the behavior of the probability sites (and therefore the
48
-------
February 1992
target population) outside of the sampled index period. Examples of questions that could be answered
with this design are: (1) What proportion of non-acidic lakes or streams in each subpopulation
undergo episodic acidification during some seasons? (2) How much do spring NO,* concentrations
differ from summer NO}* concentrations in each subpopulation? and (3) How much do organic acids
contribute to acidification at different seasons in each subpopulation?
The temporal changes mat TIME will detect in surface waters must also be related to regional
changes in sulfur and nitrogen in deposition. This can best be accomplished by testing whether trends
in the two media (air and surface waters) are of the same direction and magnitude. Work with the
TIME design is based on trends hi surface water condition expected to result from a 30-40 percent
decline in emissions. The success of the TIME project depends on a similar ability to detect regional
trends in deposition. Experience with current methods of deposition monitoring suggests mat changes
on the order of ±25 percent over a 10-year period are detectable in SO4* concentrations in deposition
at single sites. This level of precision should be adequate if the emissions reductions mandated by the
CAAA (30-40 percent) are as expected, but will need to be expanded to the detection of regional
trends, rather man site-specific trends, and will need to include both sulfur and nitrogen species. If
CASTNET can detect regional changes in deposition of ±25 percent, then TIME should have the
capability to make strong inferential statements about changes in surface water condition attributable
to changes hi deposition.
The final key objective of TIME is to design a network mat is national in scale, but that is
regional in implementation. The results of the NSWS put TIME in a good position to identify those
regions of the country where concerns for the effects of acidic deposition are warranted. The ideal
monitoring network would include all of these regions of potential concern. The financial resources
needed to support a national network would be substantial, however, and reason dictates that TIME
49
-------
February 1992
will be implemented in the highest-priority regions. Additional regions may be added as funding
permits. High-priority regions include those where a large proportion of chronically acidic lakes and
streams currently exist, and where improvements in condition can be expected to result from
emissions reductions (for example, the Northeast and Mid-Atlantic Highland regions). Chronic
acidification is rare or absent in some regions, but surface waters have a high probability of
experiencing episodic acidification; these regions are also of high priority. Other regions may not
exhibit chronic acidification at current levels of deposition, but might be expected to acidify in the
future despite emissions reductions. Hie setting of priorities for TIME has included all of these
criteria, and does not focus only on those regions with high concentrations of chronically acidic
surface waters.
The TIME project was implemented in the highest-priority region (Northeast lakes) in 1991.
TIME is scheduled to be extended to streams in the Mid-Atlantic Highlands in 1992, and to the Upper
Midwest in 1993. In addition, some regions will be monitored using a modified TIME design, which
includes only RLTM sites. Regions included in this latter category are the Sierra Nevada Mountains
in California and the Colorado Rockies. Regions mat are not currently funded, but would rank as
high priorities, include Northeastern streams, Southern Blue Ridge streams, Mid-Atlantic Coastal
Plain streams, Florida streams, and lakes in Florida and the Northern Cascades. EMAP-Surface
Waters is scheduled to be implemented in the Northeast lakes in 1992 and in the Northeast streams in
1993.
Although TIME has been designed as a chemical monitoring project, its integration into
EMAP allows for substantial inference of biological effects. TIME and EMAP use the same method
for selecting probability sites, with the result that many sites are common to both monitoring projects.
TIME does, however, have a higher density of sites than EMAP (where required to achieve sufficient
50
-------
February 1992
sample sizes for trend detection), so that not all TIME sites within each region are also EMAP sites.
In the Northeast, for example, TIME will sample approximately 350 sites over each 4-year cycle, of
which 250 will also be EMAP sites. At each EMAP site, there will be data available for fish,
invertebrate, and algal communities, collected during the same summer index period as TIME data.
The existence of biological data for a high proportion of TIME sites (for approximately 70 percent of
the sites in this example) will create important opportunities to interpret TIME results in terms of both
chemical and biological responses.
The Ah- Resource Management Monitoring program is another aquatic effects monitoring
effort operated by the USFS Regions primarily for the purpose of evaluating ah* pollution effects on
resources in Class I areas. This effort is similar to those conducted by TIME and EMAP-Surface
Waters, incorporating determinations of lake and stream chemistry with macroinvertebrate and flora
and fauna surveys. Their monitoring efforts will provide another potential source for aquatic effects
data. ...
Description of Measurements
At each TIME probability site, a single epilimnetic (m the case of lakes) or grab (hi the case
of streams) sample is collected during the summer index period (July-August) by EMAP crews. Field
measurements of dissolved oxygen, temperature, site depth, and Secchi depth are made at the same
tune as samples are collected. Water samples are shipped on ice to a central laboratory, where
analyses are begun within 48 hours of sample collection. Laboratory analyses include air-equilibrated
pH, closed headspace pH, ANC, specific conductance, SO,*, N03% Cl', Ca2*, Na*, Mg2+, K+,
NH4+, silica, total monomeric aluminum, organic monomeric aluminum, total dissolved aluminum,
dissolved organic carbon (DOC), dissolved inorganic carbon (DIG), color, turbidity, total nitrogen,
51
-------
February 1992
and total phosphorus. Quality assurance guidelines have been developed to ensure that measurement
precision is adequate for TIME'S trend detection goals.
A modified list of variables is monitored at each RLTM site by individual cooperators in each
region. Approximately 15 tunes per year, samples for all major ions, pH, ANC, conductance,
aluminum, and DOC are collected at each RLTM site. In addition, continuous measurements of
stream discharge are made at stream sites.
The EMAP-Surface Water sites will have additional information on chlorophyll, fish species
and abundance, macroinvertebrate species and abundance, sediment diatoms, physical habitat,
zooplankton species and abundance, and possible sediment toxicity, fish tissue contamination, and
shoreline bird species and abundance.
The Aquatic and Terrestrial Effects work group designed air chemistry monitoring
specifications, based on the need to tie effects to causes, and passed on its requirements to the Total
Deposition work group. (Refer to Total Deposition sections on Description of Measurements and
Treatment of Confidence and Uncertainty for details.)
Additional Aquatics Research and Monitoring Programs
The Adlrondacks Aquatics Effects Program
The Clean Air Act Amendments require establishing a program to conduct research on acid
deposition effects on aquatic ecosystems in New York's Adirondack Mountains. The Adirondack
Aquatics Effects Program, a multi-year effort, will concentrate on the condition of biota in the lakes
and streams of the Adirondack and on determining how changes in deposition levels impact these
waters.
52
-------
February 1992
The EPA EMAP-Wetlands Program
Wetlands is one of seven broad ecological resource categories EMAP has identified for
monitoring to determine the current status and long-term trends of conditions. The objectives of the
EMAP-Wetlands Program are to:
Quantify the regional status of wetlands by measuring indicators of wetlands ecological
condition and stressors on wetlands systems.
Monitor changes of these indicators and stressors.
Assess the effectiveness of mitigation of stressors on wetlands condition.
* Identify the causes of change in wetlands condition.
Provide annual summaries and interpretive assessments of wetlands status and trends.
These objectives are a combination of long-term, intermediate, and short-term goals. Over
the short term, EMAP-Wetlands will provide standardized protocols for measuring wetlands
condition, as well as estimates of wetlands condition in several regions. Trends detection is an
intermediate goal. Over the long-term, EMAP-Wetlands will implement programs to identify or
eliminate causes of change in wetlands condition. CASTNET has deferred the wetlands issue to
EMAP as a future research objective and is relying on EMAP for planning wetlands monitoring to
address this area. . .
Technical ApproachTerrestrial Effects and Stressor Monitoring
The seven main programs being relied upon for needed terrestrial effects monitoring and
research are the Forest Health Monitoring Program (FHMP), the EMAP-Forest program, CASTNET
Regional Stressor Monitoring, the Global Change Research Program (GCRP), the NFS monitoring
and effects research programs, the USFS Air Resource Management Monitoring program, and
CASTNET Intensive Stressor Monitoring. These programs conduct activities for biological,
53
-------
February 1992
chemical, stressor monitoring, and cause and effects research. Each of these programs will be
discussed separately below.
The U.S. Forest Service (USFS), EPA, EMAP-Forests, and the National Park Service (NFS)
are combining efforts to develop and implement the FHMP to measure the long-term status and trends
of the condition (health) of the nation's forests. Objectives of this program are to:
Estimate the current status and determine the trends of forest health using a series of
biological and physical/chemical indicators as a measure of condition.
Monitor pollutant exposure to determine associations between stresses and forest
condition.
Provide statistical summaries of status and trends and interpretive reports assessing forest
health.
Hie FHMP relies on the permanent national sampling framework designed by EMAP,
supplemented with a survey approach used by the USFS. The program will provide regional
estimates of forest condition by sampling all of EMAP's forest plots every year over the next four
years, and 25 percent every year thereafter. Currently used indicators of forest health are crown
assessments, visual symptoms, growth, and soil characterization. Indicators under evaluation for
possible implementation are soil chemistry, foliar nutrients, lichen biomonitoring, and vertical
vegetation structure. CASTNET's deposition monitoring (which incorporates the EMAP-Air and
Deposition program) will provide the pollutant air concentration/deposition information required by
the FHMP.
The FHMP, like TIME for surface waters, is a national program being implemented on a
regional basis. Data on crown assessments, visual symptoms, growth, and soil characterization will
become available for the Northeast in 1991, the Southeast/South in 1992, and the West in 1993. The
54
-------
February 1992
Midwest and Intennountain regions
will be incorporated into this program
in following years.
EMAP-Forests, one of the
major players in the FHMP, is the
second program being relied on for
terrestrial effects monitoring. The
overall goal of EMAP-Forests is to
develop and implement a program to
monitor, evaluate, and report on the
long-term status and trends of the
nation's forest ecological resources as
these resources relate to changes in
and among natural phenomena,
resource management practices, and
pollutants across the landscape. . .
EMAP-Forests will assess the health
TERRESTRIAL EFFECTS MONITORING
Network Participation Proposed:
USFS, EPA, EMAP-Forests, and NFS - FHMP
USFS (with EPA) - GCRP
NFS
CASTNET (with other.)
Proposed Number of Deposition and Ozone Sites:
As described in the Total Deposition lection, 51
site location* monitoring wet deposition, dry depoiition,
ozone, or a combination thereof; and 10-15 Intensive
Monitoring Site*
Proposed Effects Monitoring Locations:
FHMP - Northeast in 1991, Southeast/South in 1992, and Wect
in 1993. Mid-West and Intennounttin region* later
GCRP- South in 1991, North in 1992, and Interior West and
Pacific Coast in 1993
NFS - National parks and wilderness areas
CASTNET (wUh others) - Priority region* are the Southeast,
Northeast, Southern California. Pacific Cascades, and the
Rockies. Sensitive ecosystems of priority for Intensive
Monitoring Site* an the Appalachian Mountains; die San
Bernardino, San Gabriel, and Santa Rosa Mountains; and the
Siem Nevada Mountains
out - Crown assessments.
Proposed Effects Variables:
Biological and chemical meat
visual symptoms, growth, and aoil characterization. Possibly
soil chemistry,, foliar nutrients, lichen biomonhoring, and
vertical vegetation structure
Came and ejffeeo research - Secondary streason being evaluated
include fire occurrence sad severity, insects and disease, and
climate cbanfie ~
CASTNET
-------
February 1992
forested ecosystems. Ozone is an important regional pollutant stressor; it is ubiquitous in the eastern
United States and in parts of the West. Acid deposition is the predominate regional stressor in the
Northeast, Upper Midwest, and Mid-Atlantic regions. Figure 7 shows regions of the United States
that are sensitive terrestrial ecosystems with defined or likely effects from air pollution stress. The
work group has compiled a list of prioritized regions for pollution monitoring to be conducted by
CASTNET and other networks based on their likelihood to experience effects from acid deposition
and ozone on their terrestrial ecosystems. These regions also represent areas in which gaps exist in
the spatial coverage required to meet the CAAA requirements for regional deposition monitoring.
These regions, in order of priority, are:
1. Northeast
2. Southeast
3. Southern California
4. Pacific Cascades
5. Rocky Mountains .. .
The fourth program, initiated by the USFS, is the GCRP. The GCRP is a nationally
coordinated research effort which will be managed on a regional basis, the four regions being the
North, South, Interior West, and Pacific Coast. In the southern United States, the GCRP conducts
research and monitoring to determine the interactive relationship among forest ecosystems,
atmospheric pollution, and climate change, and uses mis information to manage and protect forest
environments and their associated resources. The four geographic programs have the following
common objectives: :
Perform integrated ecosystem-level research. ;
Study the biotic and abiotic interactions that cause changes in natural ecosystems.
56
-------
§
*s
I
5
I
I
I
8
I
57
-------
February 1992
» Understand how climate can enhance/mitigate air pollution impacts and bow climate and
air pollution can combine to cause multiple stresses on vegetation.
Understand how global change might influence the availability of water.
Each regional program will be tailored to the specific resources at risk. Research topics for
each region are discussed in the Global Change Research Program Plan published by the USFS in
May 1990. The first region, to be implemented in 1991, is the South.
Through CASTNET, EPA and the USFS Southern Global Change Program (SGCP) are
cooperating in research and in monitoring effects of potential changes in environmental conditions on
forest ecosystems. The SGCP emphasizes research on the interaction of stressors and their effect on
forests. Highest priority is given to ozone, carbon dioxide, temperature, and moisture, whereas
secondary stressors being evaluated include fire occurrence and severity and insects and diseases.
Cooperative efforts between CASTNET and the SGCP include understanding the role of ozone as a
stressor; determining/evaluating the link of acid deposition, soils, and forest health in the Southern
Appalachian Mountains; supplying regional climate analyses and climate change scenarios; and
identifying locations for new monitoring sites for the NDDN and rural ozone monitors. EPA
(through CASTNET) will provide regional monitoring and site-specific atmospheric data for cause
and effect research. EPA (through CASTNET) will provide similar support to the Northeast GCRP
region in 1992 and to the other GCRP regions starting in 1993.
The fifth program comprises NPS efforts. The NFS conducts some national atmospheric
monitoring as well as cause and effects research. CASTNET will incorporate these data into its
program and may supplement the NPS network by monitoring at their research sites. The NPS is
currently expanding their research efforts on air pollution effects on national parks and wilderness
areas on an individual park basis. For example, the NPS is implementing a Great Smokies watershed
58
-------
February 1992
study, In which NFS will incorporate atmospheric monitoring being conducted by EPA (through
CASTNET and others) to determine the total deposition on a throughput basis.
The sixth program is the Air Resource Management Monitoring program mentioned in the
Aquatic Effects section which is operated by the USFS Regions primarily for the purpose of
evaluating air pollution effects on resources in Class I areas. The USFS conducts both wet and dry
deposition monitoring, but of specific interest to CASTNET, conducts research on the relationship
and long-term effects of ambient ozone levels on vegetation.
The final major monitoring effort recommended by CASTNET, Intensive Stressor
Monitoring, will respond to specific concerns cited in the amended Clean Air Act regarding sensitive
ecosystems, such as high elevation forests. In the NAPAP State-of-Science 16 dealing with changes
in forest health and productivity, concern was expressed about the contribution of cloud/fog
deposition in addition to ozone and possibly acid deposition to the decline of forested ecosystems.
Monitoring at these Intensive Sites will support research on the contribution of total (cloud,
precipitation, and dry) acid deposition and/or ozone to effects observed in these sensitive ecosystems.
The aim is to address such issues as the long-term effects of acid deposition on forest element cycles
and tree nutrition, including: the potential extent, magnitude, and time scale of accelerated
acidification and nitrogen saturation; and the interaction of these atmospheric stressors with other
biotic and abiotic stressors such as pests and climate.
Monitoring at these Intensive Sites for ozone, wet deposition, dry deposition, and cloud/fog
deposition will follow the same requirements as those planned for the regional stressor monitoring
sites for constituents measured, sample frequency, measurement methodology, and precision and
accuracy. Determining the effects of acid deposition on sensitive ecosystems such as forests requires
characterizing the deposition profile through the forest canopy and through the tree root-soil structure.
59
-------
February 1992
Consequently, in addition to the measurements to be conducted at Regional Stressor Monitoring Sites,
stemflow, throughfall, and soil chemistry/nutrient cycling measurements also may be conducted at the
Intensive Monitoring Sites. Actual CASTNET measurements and siting will depend on the efforts of
others at any particular site to take advantage of USFS and NFS efforts wherever possible and to
avoid duplication of efforts. Sensitive ecosystems under stress because of acid deposition and/or
ozone and considered priorities for Intensive Monitoring Sites are:
Appalachian Mountains
San Bernardino, San Gabriel, and Santa Rosa Mountains
* Sierra Nevada Mountains
* Cascade Mountains
Rocky Mountains
All five of these regions are mountainous areas, where deposition is intensified, or near large urban
centers, where ozone settles on the mountain slopes. Site locations for these areas will be coordinated
with federal and state agencies conducting research. Most, if not all, of these sites will be part of the
CASTNET Regional Stressor Monitoring network and will serve multiple purposes.
Description of Measurements
The Aquatic and Terrestrial Effects work group designed air chemistry monitoring
specifications, based on the need to tie effects to causes, and passed on its requirements to the Total
Deposition work group. (Refer to Total Deposition sections on Description of Measurements and
Treatment of Confidence and Uncertainty .for details.)
60
-------
February 1992
Deposition Sites and implementation
To provide the coverage needed for a regional ecosystem assessment of conditions and causes,
SI additional deposition monitoring site locations (monitoring dry deposition, wet deposition, ozone,
or a combination thereof) are recommended, as detailed in the Total Deposition section.
The Effects work group also recommends an additional 10 to 15 Intensive Monitoring Sites,
to be collocated (if possible) with sites from among the proposed 51 regional deposition monitoring
sites, to complete the requirements mandated by the CAAA. For Stage I in 1992, the following sites
are recommended:
1-3 sites in the Appalachian Mountains (priority ecosystem number 1).
As funding allows, the proposed 10 to 15 Intensive Monitoring Sites will be distributed as follows:
Number of Sites Location
5-6
1-2
1-2
1-2
2-3
Appalachian Mountains
San Bernardino, San Gabriel, and Santa Rosa Mountains
Sierra Nevada Mountains
Cascade Mountains
Rocky Mountains
Relationship with Existing Networks
A thorough review of current monitoring in rural areas has been completed. Other agencies
and states are being contacted to determine whether mutual monitoring is possible at some of their
existing sites. Gaps exist, so additional air and deposition sites will be needed to provide spatial
coverage to meet regional stressor monitoring and intensive monitoring requirements. The Statistical
61
-------
February 1992
Network Design work group will provide consultation on the number and location of additional sites.
Report Formats
Two levels of reports will be generated by CASTNET; annual statistical summaries and
biennial interpretative assessments. Annual summaries of air concentration/deposition data from each
site on a seasonal and yearly basis, summaries of ecological data, and interpolated data will be used to
provide regional estimates. Every two years, an assessment linking stressor and ecological effects
will be produced.
For reports on aquatic effects, TIME will provide statistical summaries of chemical data to
CASTNET, and EMAP-Surface Waters will provide statistical summaries of biological data.
Regional forest health monitoring data and reports will be provided under the joint USFS-EPA
FHMP. The FHMF annual statistical summaries will provide descriptions of regional status and
trends of forest indicators. Biennial interpretive reports will provide analysis and assessment of forest
health. As part of these interpretive reports, various stresses on forested ecosystems will be assessed.
Stresses to be included are natural, climatic, pests, and anthropogenic stresses such as acid deposition
and ozone exposure. CASTNET deposition monitoring will monitor deposition in high-priority
regions and will provide regional data on the status and trends of acid deposition. CASTNET also
will be responsible for interpretative and integrated assessment reports tying together the effects
and atmospheric stressor data produced biennially and used as the basis for required reports to
Congress.
For reports on sensitive ecosystems, detailed analyses of atmospheric constituent exposure and
deposition will be provided annually for each site. The analysis will include time series plots,
monthly summaries, growing season summaries, and seasonal and yearly trends. The organizations
62
-------
February 1992
jointly conducting the effects research and CASTNET will cooperate in producing reports and articles
relating atmospheric exposure and deposition to observed effects.
Future Research Needs
A series of watershed studies are needed. These are high-priority projects for both aquatic
(TIME) and terrestrial effects monitoring (CASTNET); they will provide information on the
relationship of surface water, forest, and soils relating to nitrogen saturation, causes of episodic
acidification, and responses of forests to acid deposition. The studies will include verifying and
improving existing models, establishing critical loads, and regionalizing site-specific information.
Integrated surface water/forest/soils monitoring sites will provide maximum information at greatest
cost savings.
Tracking deposition, nutrient flux, soil condition, and surface water chemistry will provide
the first quantifiable indication of ecosystem change from a change in acid deposition. Priorities for
regional implementation of these studies are the Northeast, Mid-Appalachian Highlands, and the
Southern Blue Ridge/Great Smoky Mountains.
63
-------
February 1992
THIS PAGE INTENTIONALLY
LEFT BLANK
64
-------
February 1992
Visibility/Acid Aerosols Monitoring
Monitoring Objectives
The legislative mandate for visibility protection began with the Clean Air Act Amendments of
1977, which called for the prevention of any future and the remedying of any existing impairment
from man-made air pollution of visibility in Class I federal areas. In Title IX, Section 103(c) of the
1990 Amendments, the establishment of a national network to monitor (beyond Class I areas), collect,
and compile data on visibility impairment and air quality is required. The network will characterize
trends in visibility and acid aerosol as indicators of changing air quality and will provide explanations
of why these changes occur or, in some cases where expected, why they do not occur. Planned
aerosol measurements at mis time are only for surrogate (optical) visibility measurements. Research
on methods and siting is needed before an acid aerosol network related to health effects can be
designed. This research is ongoing with results expected in time for CASTNET 1993 network
enhancement. ~
In Title Vm, Section 169B was added to specifically address visibility research issues for
Class I areas, giving EPA the lead role. Sources of visibility impairment as well as regions with
predominantly clean air were to be studied, including expanding current visibility-related monitoring
in largely western Class I areas; assessing current sources of visibility-impairing pollution and clean
air corridors; adapting regional air quality models for visibility assessment; and studying the
atmospheric chemistry and physics of visibility.
The Visibility work group explored issues concerning visibility and acid aerosol, including
their significance, measurement methods, design considerations, regulatory requirements, and model
evaluations. The work group recommends a network design that will use existing air quality
65
-------
February 1992
monitoring programs to meet CASTNET objectives in the most comprehensive, efficient, and cost-
effective manner, building largely upon the NFS and IMPROVE (Interagency Monitoring of Protected
Visual Environments) programs developed for federal Class I areas and national parks (described
below).
Technical Approach
Hie national monitoring network is required to track the trends of pollution resulting from
emission reductions. Two key elements expected from emissions reductions are improved visibility, a
welfare effect, and acid aerosol reduction, which has health effects. Both health and welfare effects
research issues are addressed in the Title DC mandate.
Visibility responds readily to changes in air quality. In virtually every attempt to quantify
economic effects or benefits of air pollution, visibility has been highly valued by the public and is
perceived as an index of health risk (visibility is closely related to acid aerosols and organics). Acid
aerosol is an emerging health effect issue, and although it is currently under consideration for
designation as a criteria pollutant, virtually no data exist for status and trends analysis of acid aerosol.
Visibility impairment in the United States, other than mat caused by precipitation events and
fog, is primarily attributable to anthropogenic fine particles and gases, such as sulfur oxides, nitrogen
oxides, and hydrocarbons, emitted from urban and industrial sources. Upon emission, many of the
gases convert to aerosols. Haze is a mixture of gases and secondary aerosols which form within 100
km of a source such as an urban center, power plant, or other industrial facility. After pollutants are
transported hundreds of kilometers, the haze which is formed is composed mainly of fine primary and
secondary aerosols such as sulfate and organic aerosols. Forms of natural aerosols that also affect
visibility are condensed water vapor (water droplets), wind-blown dust, and their primary and
secondary aerosols (including natural organics).
66
-------
February 1992
The Visibility work group
reviewed past trends in baze using
airport data (Figure 8) and discussed
model output scenarios from the
Regional Acid Deposition Model
(RADM) as well as other related
sources of information. EPA
regional offices polled their states
for visibility or acid aerosol
monitoring efforts and for
recommended areas suitable for
visibility monitoring sites. These
polls revealed that only a few states
have acid aerosol or visibility
monitoring programs.
After this review, the
Visibility work group established a design for visibility and acid aerosol monitoring intended to detect
a significant change within the seasonal or annual averages of visibility and its indicators. The design
goal is to provide an estimate of selected visibility indicators for any region of the country with a
relative error of ±50 percent, with a 95-percent confidence level.
VISIBILITY/ACID AEROSOLS MONITORING
Network Participation Proposed:
NFS
IMPROVE
NESCAUM
Other States
Others
Proposed Number of Visibility Sites:
, 12 "full" and 20 aerosol-only sites
Proposed Visibility Monitoring Locations:
Full sites will be located in: (1) Areas where significant
change is expected (in sulfate concentrations in New
York and the Ohio River Valley, and in mobile
source concentrations in California); (2) Areas east
of the 105th meridian (Rockies) and west of the
Mississippi River where there is a dearth of data; and
(3) The central states located along the east-west
transition. corridor
Aerosol-only sites will provide more uniform spatial
coverage and refinement in areas with complex
pollution scenarios at substantial cost savings
Proposed Visibility Variables:
Atmospheric optical properties (total light extinction),
visual scene (photographs), and fine particle
composition (to include elemental, organic, and total
carbon; acidity; mass; sulfate; nitrate; and the trace
dements sodium through lead)
67
-------
-------
February 1992
Description of Measurements
There are three types of visibility measurements:
Optical optical properties of the atmosphere (total light extinction) are monitored to
provide a scene-independent measure of air quality.
Scene characteristics of .a scene, viewed at a distance, are monitored to document the
scene-specific visibility.
Aerosol the aerosol characteristics are determined so that atmospheric optical properties
can be associated with the responsible pollutants.
Acid aerosols can be measured by adding denuders to the aerosol instrumentation used for
visibility. Supplemental data such as meteorology and continuous gaseous concentration data are
important for correct interpretation.
Optical Monitoring: The best choice of optical equipment depends on the characteristics of
each site. Although several methods are used for monitoring optical properties of the atmosphere, the
two methods most commonly employed are integrating nephelometry and transmissometry, each of
which has advantages and disadvantages. The atmospheric extinction coefficient is essential to
understanding how atmospheric aerosols affect the visibility of targets. The nephelometer measures
the scattering coefficient of particles, which is the major portion of the total extinction coefficient.
The nephelometer, however, may modify the particles by heating them, causing evaporation of
associated moisture and resulting in an over-estimation of die visual range. Nephelometers also suffer
from the fact that they yield only point measurements. Calculations of atmospheric extinction using
the transmissometer are sensitive to both scattering and absorption, and they measure effects of
different-size particles in the ambient aerosols. Methods do not yet exist to routinely account for light
scattered into the sight path of the transmissometer, which hi remote locations without continuous
69
-------
February 1992
visual observations, make it difficult to remove the effects of cloud and precipitation from the data
base. The major drawback to the transmissometer is that it has difficult siting requirements in that a
long visual path is usually required. The ultimate choice between the two measurement options is
therefore left to be decided by the characteristics of the site involved.
Scene Monitoring: Photos taken with an automated camera are recommended for scene or
view monitoring. This method allows for perception studies, trend evaluations, and sensitivity
assessments of a view to changes in air quality. Additional information yielded from this method
includes frequency and intensity of plumes and elevated haze levels associated with local sources.
This approach, however, requires accurate recording of brightness, color, and spatial detail of the
scene. Without these data, it is difficult to make meaningful comparisons among different scene-
specific conditions. The work group recommends that photos be taken three times a day, every day
at 9:00 a.m., 12:00 p.m., and 3:00 p.m. .
A erosoi Monitoring: The work group also recommends that CASTNET measure fine
particle composition with the Modular Aerosol Monitoring Sampler currently used by the IMPROVE
Program to determine which atmospheric constituents are responsible for changes in visibility
impairment. The aerosols associated with visibility reduction are mainly fine particles, with light
scattering attributed primarily to particles of less man 2.5 /tm in diameter. The aerosol measurements
required are elemental, organic, and total carbon; acidity; mass; sulfate; nitrate; and the trace
elements sodium through lead. Of the four filter modules in me IMPROVE sampler, three collect
fine particle samples (2.5 fan), whereas the fourth samples respirable particles (< 10 /tin). A
summary of the equipment and analyses for the optical, scene, and aerosol measurements is presented
in Table 4. Keep in mind that aerosol monitoring is being planned as a visibility surrogate only, and
that an aerosol health network is not being proposed at this time.
70
-------
February 1992
TABLE 4. SUMMARY OF VISIBILITY EQUIPMENT TYPES AND ANALYSES
Device
Nephelometer
Transmissometer
35-mm camera
Fine particle
sampler:
Module A: fine
Module B:fiae
Module C:fine
Module D.-PMIO
Measurement
>
Fine particle
scattering
Atmospheric
extinction
Visual scene
25-mm teflon
47-mm nylasorb
25-mm quartz
25-mm teflon
Frequency
Hourly
Hourly
3 times daily
24-hour, every
third day
24-hour, every
third day
24-hour, every
mud day
24-hour, every
third day
Analyses
Mass, absorption
elements (H, Na-Pb)
Nitrate
Organic and elemental
carbon
Mass
Research continues to enhance the IMPROVE aerosol sampler recommended by die work
group. It should be noted that some researchers have experienced problems in the field with critical
orifices clogging on some samplers. As well as continuing to study and document the current
IMPROVE system, the CASTNET Instrumentation/Methods work group recommends using a mass
flow controller-based flow system at a few sites, which will provide good
control and give a continuous flow output readable by a data logger. After adequate data have been
collected for comparison of the two configurations, a benefit/cost analysis can be performed.
Additional modification to die IMPROVE sampler protocol, such as use of a larger-diameter filter,
will reduce potential filter clogging problems at more polluted sites. Using the IMPROVE sampler
71
-------
February 1992
will preclude die need for parallel sampling at a variety of sites to ensure data comparability with the
National Park Service (NFS) sites, which, as discussed below, will be an integral part of this
monitoring effort. National Weather Service (NWS) meteorological data (for example, relative
humidity and temperature) will be used to help validate local instrumentation data that may be used in
modeling efforts.
Treatment of Confidence and Uncertainty
The inherent uncertainty in optical measurements falls within 10 to 15 percent. Uncertainty
of aerosol data varies depending on the element. Uncertainty cannot be quantified for camera data.
In developing trends in extinction owing to different species, uncertainty levels are expected to be 10
to 15 percent for sulfur, 20 to 30 percent for organics, and 20 to 30 percent for water/particle
interactions. No estimates exist for the uncertainty in regional acid aerosol trends, although the level
of uncertainty is expected to emerge from the Summer 1991 Acid Aerosol Study.
Sites and Implementation
The Visibility work group designated high-priority areas for visibility monitoring (Figure 9)
after reviewing siting of existing visibility networks, such as IMPROVE, reviewing model outputs
from RADM and the Engineering Aerosol Model (EAM), exploring the consequences of expected
shifts in urban sources and the possible trading of emissions, and defining the east-west transition area
and other areas that are simply lacking available visibility data.
The work group determined that augmenting the fully equipped IMPROVE and NFS networks
("full" sites) would yield the best and most cost-effective national coverage. Collocating sites with
existing air quality monitoring sites, such as those for wet (NADP/NTN) and dry (NDDN)
deposition, would result hi substantial cost savings. IMPROVE and NFS sites are located in national
parks and in federal Class I areas and can provide only limited coverage between the 95th and 105th
72
-------
a
i
t
Ch
73
-------
f
£
I
1
I
I
74
-------
February 1992
meridians, but they leave few inadequately covered areas in the West. The work group recommends
that 12 "full" sites and 20 aerosol-only sites (see Figure 10) be added to provide complete national
coverage in sufficient detail to conclusively delineate the relationship between visibility changes and
air quality.
An individual site that does not represent the surrounding area can bias spatial pattern and
trend estimates, leading to inappropriate inferences. Sites in remote regions exert a disproportionate
influence on spatial isopleth patterns. Because many areas in the country do not have enough
quantifiable visibility monitoring data for a statistical analysis of network design, the work group used
expert opinion in designating areas for recommended site locations. As the Statistical Network
Design work group determines sulfate distributions, this information will be used to locate new sites
and evaluate existing ones.
Sites were chosen to complement, not duplicate, data collected by the IMPROVE and NFS
networks. Preference was given to nonurban sites; areas with little other monitoring coverage; areas
with high current emissions levels or those predicted to have the greatest change in emissions (by
RADM); and areas in which spatial and statistical patterns are most complex. The 12 "full" sites
were located in: , -
1. Areas in which a significant change is expected: in sulfate concentrations in New York
and the Ohio River Valley; and in mobile source concentrations in California.
2. Areas east of the 105th meridian (Rockies) and west of the Mississippi River where there
is a dearth of data.
3. The central states located along the east-west transition corridor.
The 20 aerosol.sites will provide more uniform spatial coverage and refinement in areas with complex
pollution scenarios at a substantial cost savings.
75
-------
February 1992
An alternate design for siting is to implement sufficient coverage in some areas predicted to
change most (by RADM, EAM, etc.), so that trends can be detected on a smaller regional scale, with
large gaps between regions. This alternative design is not a viable option, however, because it
provides insufficient spatial coverages to achieve the national goals of the CAAA mandate.
The goal is to implement the visibility network during a two-year period, half being
implemented each year. If funding is not available for the complete network, the work group has
assigned priorities to site locations so that expected changes can be tracked for specific regions. To
determine trends on a national scale, the full CASTNET network is required, as is continuation of the
other networks on which CASTNET depends. An oversight committee will be formed to coordinate
visibility monitoring efforts, such as IMPROVE and NFS hi federal Class I areas, to monitor other
networks' program viability and direction.
Relationship with Existing Networks
In 1980, EPA issued regulations requiring visibility monitoring, new source review, and long-
term strategies for protecting visibility in Class I areas. EPA has implemented these monitoring
requirements through a cooperative effort with Federal Land Managers and certain state
organizations, called the Interagency Monitoring of Protected Visual Environments (IMPROVE),
which is funded by EPA State Grants and the NFS.
In addition to IMPROVE, other significant air quality monitoring efforts include the NFS,
USFS, NDDN, NADP/NTN, and NESCAUM, all of which vary substantially in monitoring,
analysis, and regional coverage. The USFS sites are camera-only, and the NESCAUM sites are
aerosol-only. The IMPROVE and NPS networks meet CASTNET recommendations because they
have complete visibility measurement equipment, including transmissometer or nephelometer, aerosol
sampler, and 35-mm camera.
76
-------
February 1992
CASTNET's success depends on continued operation of current IMPROVE, NFS, and
proposed Eastern IMPROVE sites for the next 10 years. CASTNET coordination with IMPROVE,
NFS, and NESCAUM sites will enhance each of these programs. For example, establishing a
CASTNET site in parallel with California's visibility network will result in operational savings and
comparability with other California sites. It is anticipated that other state programs will contribute
significantly. NESCAUM will continue to operate its aerosol-only sites, for which CASTNET will
conduct the filter analyses. Moreover, EPA Health Studies and various Visibility Transport
Commissions (as authorized hi the CAAA) will benefit from sulfate/nitrate and acid aerosol data
collected at CASTNET sites.
Report Formats
CASTNET will follow existing protocols for data reporting. Reports will be issued annually,
with quarterly breakouts, following the current IMPROVE schedule. All aerosol and optical data
will be deposited in AIRS. For special research needs, (invalidated optical data will be available the
next day, camera data will be available quarterly, and aerosol data will be available six months after
the close of a quarter. The work group recommends annual reports to document and summarize
available data for the science community. Biennial interpretive reports, using graphics such as those
in Figure 8, are also recommended to exhibit national trends.
The network data will be used to indicate die average seasonal visibility for die nation, with a
resolution of about one state area. Data from the network will be used with other CASTNET data to
study the impact of specific source groups on air quality and visibility.
Future Research Needs
Because the CASTNET visibility network design depends on the continuation of other
networks, a contingency plan is needed to prepare for possible failure of one or more of these
77
-------
February 1992
networks. The aforementioned oversight committee will coordinate with other visibility efforts to
monitor other networks' program viability and direction.
To ensure data quality and comparability, the work group recommends developing an external
quality assurance program that works with all networks. A centralized database that includes
mechanisms for data distribution will need to be established between the different networks and the
scientific community.
Acid aerosol methods and procedures research and site characterization studies must be
initiated to provide complete pollutant characterization and size distribution profiles that will help
determine the sources and sulfate species responsible for visibility impairment. For example, it is
known that the particles most efficient at scattering visible light (light scattering is usually going to
dominate over light absorption in remote areas) are those with diameters between 0.1 and 0.9 /on.
But additional instrumentation, with an on-site operator, is required to classify these fine particles,
which makes routine network deployment prohibitively expensive at this time. Special studies to
classify the distribution of particle sizes could provide important information on the causes of
visibility impairment at a greatly reduced cost. Other studies to support needed research are the
EPA's Acid Aerosol Study at Union Town and the NFS Acid Aerosol Summer Study at Shenandoah
National Park. It is hoped that mis research will lead to an acid aerosol-health component within
CASTNET.
78
-------
February 1992
Air Toxics Monitoring
Monitoring Objectives
Hie Clean Air Act Amendments (CAAA) specify two monitoring requirements that are of
concern to the Air Toxics work group: the "Great Waters" program, which monitors deposition of
hazardous air pollutants (HAPs) to specified water bodies for assessing the resultant environmental
and human health effects, and the Area Source Program, which monitors ambient concentrations,
sources, transformation processes, and public health risks from HAPs.
Great Waters Program
Title m, Section 112(m), of the CAAA requires that EPA, in cooperation with the National
Oceanic and Atmospheric Administration (NOAA), conduct a program to identify and assess the
extent of atmospheric deposition of hazardous air pollutants, and other pollutants at the
Administrator's discretion, to the Great Lakes, Chesapeake Bay, Lake Charaplain, and coastal waters.
Because the study covers many significant national public trust waters, and to distinguish it from other
ongoing, related efforts, the study is referred to as "the Great Waters Program."
The Great Waters Program is required to include:
Atmospheric deposition monitoring.
Investigations of sources and deposition rates of air pollutants as well as their
transformation precursors/products.
Monitoring methodology research.
Determination of relative contribution of ah* deposition to total loading in the waters.
Evaluations of adverse effects from deposition, including indirect effects to health and to
the environment.
79
-------
Febniaiy 1992
Assessments of contribution of such deposition to violations of water quality standards.
» Biological sampling to identify the presence of HAPs mat deposit from the air.
Some water body-specific requirements also are listed, the most important being the
requirement for one hazardous air pollutant sampling station per Great Lake by 12/31/91. The
requirements of Section 112(m) either directly dictate the needs of the Great Waters Program or make
them inherently necessary for a mandated report to Congress. A report is due to Congress, in
cooperation with NOAA, in 1993 and biennially thereafter and is required to include, at minimum:
The relative contribution of HAPs to total loading of the listed waters.
HAP-caused environmental and human health effects to listed waters.
Sources of HAPs.
Determination of whether HAP loadings cause or contribute to water quality violations or '
violation of the Great Lakes Water Quality Agreement (GLWQA).
Description of regulatory revisions (to the Clean Air Act and other applicable federal
.... . laws) necessary to ensure protection of the waters. -
Based on mis report and within five years of enactment, the Administrator of EPA is required to
promulgate any further regulations deemed necessary and appropriate.
Area Source Program
Title HI, Section 112(k), of the CAAA requires the EPA to conduct a research program, the
Area Source Program (also known as the Urban Area Source Program), to support a national strategy
to reduce risks from hazardous air pollutants from area sources in urban areas. The research program
must include:
80
-------
February 1992
* Ambient monitoring in a representative number of urban locations for a broad range of
HAPs, including, but not limited to, volatile organic compounds (VOCs), metals,
pesticides, and products of incomplete combustion (PICs).
Characterizing the sources of the HAPs by focusing on area sources and their contribution
to public health risks. ,. .
Considering transformations and other factors that may elevate the risks to public health.
The legislation requires a preliminary research report to Congress in November 1993 and a
comprehensive national strategy by 1995.
The national strategy must identify:
* At least 30 pollutants that pose the greatest threat to public health in the largest number of
urban areas.
* Source categories that individually or together present a threat of adverse effects to human
health or the environment.
Source categories that together represent 90 percent of the area source emissions of the 30
. aforementioned HAPs. .
Schedule of actions to reduce urban cancer risk by at least 75 percent by controlling HAPs
emissions from all stationary sources. '
Research needs in sampling and analysis methodology, modeling, or pollution control.
Recommended regulatory changes to further the CAAA goals.
Provision for ambient monitoring and emissions modeling to demonstrate progress toward
strategy goals.
81
-------
February 1992
As a part of the strategy, ambient monitoring and emissions modeling also are required, as
appropriate, to demonstrate that the goals and objectives of the strategy are met. Future CASTNET
monitoring and network refinements are the intended means of meeting these latter needs.
The Air Toxics work group's goal is to construct a program to meet the mandated CAAA
deadlines on schedule and to provide the essential information for sound regulatory decisions. To
design an efficient, cost-effective program, the work group recommends supplementing existing or
planned stations and expanding as possible to meet legislative deadlines.
Technical Approach
To design a program for air toxics monitoring, the work group was required to address both
die Great Waters and Urban Area Source perspectives. The following are specific objectives and
recommended approaches of each of these program areas:
Great Waters Program
A national monitoring program should be implemented consisting of the following steps On
order of priority): .
1. Establish the initial network of comprehensive stations (Integrated Atmospheric Deposition
Network (IADN) master sites), one per lake (two in Canada and three in the United
States), which will measure nutrients, precipitation, pH, conductivity, particles, VOCs,
semi-volatile organic compounds (SVOCs: pesticides, polychlorinated biphenyls (PCBs),
etc., both in precipitation and in air), polycyclic aromatic hydrocarbons (PAHs),
meteorology (such as temperature, wind speed, and relative humidity), and elemental
compounds (in both precipitation and in air).
2. A Lake Michigan integrated, pilot, mass balance study should be established to answer
questions on siting, network design, and basic deposition processes.
82
-------
INTEGRATED ATMOSPHERIC
DEPOSITION NETWORK
The Integrated Atmospheric Deposition Network
(IADN) is a joint effort between the U.S. and
Canada and was created to fulfill Annex IS of the
Great Lakes Water Quality Agreement (GLWQA).
The objective is to acquire sufficient, quality-
assured data to estimate with a specified degree of
confidence the loading to the Great Lakes Basin of
selected toxic substances. The network will consist
of several master (research grade) stations
augmented by a number of satellite (routine)
stations.
February 1992
3. Existing deposition monitoring
sites on Lake Champlain and
Chesapeake Bay should be
expanded to measure particle
mass, the ratio of coarse/fine
particles, elemental
composition, and other
compounds of interest, using a
dichotomous (or similar)
sampler.
4. New preliminary sites should be added for coastal estuaries measuring particle mass, the
ratio of coarse/fine particles, and elemental composition.
5. The program should be expanded to add more sites, or to add additional important
pollutants, or altered to correct problems.
The use of a dichotomous sampler for preliminary monitoring of Lake Champlain,
Chesapeake Bay, and the coastal estuaries allows for a relatively low-cost, reliable means of attaining
a broad geographic representation of elemental deposition to the Great Waters.
The contribution of atmospheric pollutants relative to total loading can be determined by
calculating atmospheric loading through deposition monitoring (with adjustments for meteorological
variability), verified by periodic intensive studies, and by incorporating the atmospheric contribution
into mass balance modeling. The adverse effects of HAPs loading on human health and the
environment can be determined in three steps. The first is to determine human exposure to HAPs
through the water, the second and third steps are to determine the atmospheric contribution to
83
-------
February 1992
exposure of aquatic life to HAPs, then
determine bioaccumulation and subsequent
health and environmental effects.
Modeling and other techniques, with
monitoring results for trace chemicals, can
be used to identify sources of HAPs. The
approach recommended to determine
departures from water quality criteria
(established by the Federal Water Pollution
Control Act, the Safe Drinking Water Act,
or GLWQA) is to determine the
atmospheric contribution to water column
concentrations, using atmospheric loading
information. Hie final objective of the
Great Waters air toxics monitoring is to supply the information needed so mat EPA can recommend
regulatory revisions to prevent adverse effects. This objective will be met by determining the
predicted improvement attributable to existing regulations and by identifying additional load
reductions needed.
Area Source Program
A national monitoring network will be developed by installing monitoring sites in populous
urban areas, as feasible. Recommended steps for establishing the monitoring network, in order of
priority, are:
AIR TOXICS MONITORING
GREAT WATERS PROGRAM
Network Participation Proposed:
IADN
Proposed Number of Toxics Sites:
Five IADN Master Stations, one site per
Great Lake
Proposed Toxics Site Locations:
One site in Canada on Lake Huron (Burnt
Island), one on Lake Ontario (Pt. Petre),
and, in the U.S., one each on Lake
Michigan (Sleeping Bear Dunes), Lake
Erie (Sturgeon Point) and Lake Superior
(Keweenaw Peninsula). Chesapeake Bay,
Lake Champlain, and coastal estuaries later
Proposed Toxics Variables:
Rain (pH, conductivity, nutrients); particle
mass and size; trace elements (in particles);
pesticides, PCBs, and PAHs (in air and
precipitation)
84
-------
. . February 1992
1. Identify and characterize the scope of existing urban toxics monitoring stations and
programs.
2. Establish an Urban Area Source monitoring network by augmenting selected enhanced
ozone monitoring sites with sampling for hazardous VOCs, metals, pesticides, and PICs.
Consideration will be given to sites that can also serve as part of the Great Waters
network. .
3. Enhance measurements as needed to identify sources and human health risks, or to
demonstrate progress toward goals of the national strategy.
4. Expand the network as feasible.
An EPA multi-office work group will assess existing urban monitoring data to identify the
broad range of HAPs present in urban areas, and then to identify the 30 pollutants that pose the
greatest threat to public health in the
largest number of urban areas.. They will
also identify the source categories that
represent 90 percent of me aggregate
emissions of those pollutants. Exposure .
and risk will be estimated from available
data on emissions, concentrations,
populations, and the pollutants' adverse .
effects on human health. Sources of the
pollutants will be identified both by
emission inventories and by apportionment modeling. As research provides more information, these
lists will be refined.
AIR TOXICS MONITORING
URBAN AREA SOURCE PROGRAM
Network Participation Proposed:
. r Enhanced Ozone Monitoring Network .
[ff> . . _, ,
Proposed Number of Toxics Sites:
Supplement selected "**&*****& Ozone sites
for air toxics measurements
-~-* . ,* . ' *
Proposed Toxics Site Locations:
" Selected serious, severe, and extreme ozone
latiainmtvit cities
Proposed Toxics Variables:
VOCs, metals, pesticides, and PICs
85
-------
February 1992
The ultimate objective of the Area Source program and its ambient air toxics monitoring
network is to supply the information needed to develop regulatory revisions to reduce adverse effects.
Monitoring and modeling studies are also required to assess progress as the national strategy is
implemented.
Identification of Affected Areas
The Great Waters program will address the significant public trust waters listed in the Act,
including the Great Lakes, Chesapeake Bay, Lake Champlain, and coastal waters. Coastal waters are
defined as those areas designated as either EPA National Estuary Program (NEP) estuaries or as
NOAA National Estuarine Research Reserves (NERR). These coastal waters comprise 19 NEP
estuaries and 27 NERR estuaries, geographically ranging from Maine to Washington, touching nearly
every coastal state on the continent, and including sites in Hawaii and Puerto Rico.
The limitless nature of the CAAA mandates and budget realities concerned preclude the initial
study of all the designated waters and will force the work group to consider the minimum needs of
the toxics programs. To meet the purposes of the CAAA in identifying problems and assessing
regulatory needs, initial monitoring will emphasize methodology assessment and priority issues.
Consequently, in defining a national atmospheric deposition network, monitoring will be focused on
the better-studied waters and on existing (such as: the three New York and three Massachusetts
Institute of Technology sites, one each on Lake Erie, Lake Ontario, and Lake Champlain; the two
Vermont sites on Lake Champlain; and the five sites on the Chesapeake Bay operated by the
University of Maryland (two), Virginia Institute of Marine Science (one), and the University of
Delaware (two)) or planned stations and will be expanded as budget allows to the other water bodies,
specifically the less-studied Great Lakes, Lake Champlain, Chesapeake Bay, and coastal waters.
Coastal stations not on Chesapeake Bay will be selected to represent the continental U.S. coastal
86
-------
February 1992
regions: North Atlantic, Mid-Atlantic, South Atlantic, East Gulf of Mexico, Central Gulf, West Gulf,
South Pacific, Mid-Pacific, and North Pacific.
The Area Source program is concerned with human exposure. As a result, the areas of
interest are urban locations with large populations. The first step of the Area Source Program is to
identify the broad range of HAPs (from which will be determined 30 or more) mat present the
greatest human health threat. Urban areas to be studied will overlap those areas that are affected by
the enhanced ozone monitoring regulations since ozone exceedences also tend to occur in the nation's
most populous areas. Under the CAAA, VOC monitoring will be required for ozone nonattainment
areas classified as serious, severe, or extreme. These 25 areas will present a ready opportunity for
collocation of sites for toxics monitoring.
Description of Measurements
A program to measure toxic air pollutants must address their various forms, including volatile
organic compounds, semi-volatile organic compounds, metals, and particle-bound species. Critical to
the calculation of deposition estimates are the size distribution of particles, as well as the phase
distribution, or the proportion of the toxic pollutant that is in the gas vs. particle phase.
Great Waters Program: To support modeling calculations, several "non-target" tracer
compounds must be measured. For example, to separate the contribution of residential wood
combustion from mobile source emissions, "non-target" tracers like potassium or isooctane may be
measured. . . .
Great Waters preliminary stations should provide information on rain (pH, conductivity,
nutrients), particle mass, trace materials (in particles), and particle sizing. The standard, low-volume
dichotomous sampler gives mass, size distribution (coarse from 2.5 pm to 10 /un, and fines below 2.5
jim), and trace elements from analysis of 37-mm teflon filters. Measuring trace elements determines
87
-------
February 1992
the concentration of some toxic compounds (for example, lead, cadmium, arsenic, and so forth) and
provides tracers for source identification. A high-volume sampler also gives the total paniculate
mass, a sample for determining total organic carbon, and elemental and volatilizable carbon
measurements that are useful for source apportionment.
Semi-volatile organic carbon compounds, which exist in gas and particle phase at ambient
conditions, are collected using an XAD/PUF (sorbent/polyurethane foam) cartridge in line with a
filter so that both gases and particles are available for later analysis. Hie number and size of samples
required will depend directly on the number of pollutants of interest and their expected
concentrations.
Gases or VOCs are collected in 6-liter, summa-polished, stainless-steel canisters and analyzed
at a central laboratory. Although mere are a number of otter equally acceptable methods for
collecting VOC samples, this method is recommended for its reasonable cost and ease of use in
remote sampling locations. Hie fewest number of different analytical laboratories should be used to
assure quality assurance measures are met.
Hie work group has established protocols for sampling. Integrated samples should be
collected over 24 hours. When some special study samples are being collected, two 12-hour samples
are preferred over one 24-hour sample to allow determinations of significant differences in daytime
and nighttime conditions, especially with respect to particle concentration and size distribution.
Weekly samples are generally taken at random, one 12- or 24-hour sample every 6 days, to allow for
random sampling of different days of the week.
Area Source Program: The sampling schedule for areas affected by the enhanced ozone
monitoring regulations (The Enhanced Ozone Monitoring Network) will be eight 3-hour samples for
VOCs and four 6-hour samples for aldehydes, either every day or every third day, depending on
-------
February 1992
population. The Air Toxics work group recommends supplementing this VOC monitoring with toxics
sampling at several sites, with preference given to those sites located on or near water bodies of
interest. The VOC sampling schedule may necessitate a different approach to air toxics monitoring
than in the Great Waters program. The VOC monitoring would be expanded to include more toxic
VOCs. Monitoring for pesticides, toxic metals, and PICs would be identical to methods under the
Great Waters program.
Meteorological Factors: Integrated samples during periods of intensive field research
should be collected daily for 12-hour periods (or less) for the reasons discussed above. Because
sampling across multiple meteorological conditions is undesirable, allowances should be made for
changes in synoptic conditions. Despite the difficulty hi coordinating all field programs to make use
of meteorological data, sample protocols should be adjusted when it becomes obvious through the use
of simple isobaric trajectory forecast models, National Weather Service products, or other field-tested
techniques that sampling is likely to occur across multiple air masses in a single sampling period. If
adjustments are not possible, such periods of record should be flagged in the final data sets using post
facto meteorological information so mat data users are aware of potential limitations.
Station S'rtina: Because the initial five stations in the Great Waters program also will be
IADN sites, station siting will follow existing IADN protocols, which specify nonurban settings with
no significant sources of particles, SVOCs, or VOCs near enough to the monitoring stations to have
any measurable impact. IADN also specifies location requirements including acceptable height above
ground, distance from trees and other obstructions, and location of roadways or sources of traffic.
Future Great Waters site locations should include urban and nonurban settings.
89
-------
February 1992
Urban-influenced sampler siting protocols can be studied in the 1991 Lake Michigan Air
Toxics Study and the Lake Michigan integrated pilot study. This information will be integrated into
future site selection procedures.
Treatment of Confidence and Uncertainty
A number of factors complicate establishing precision and accuracy levels for toxic air
pollutants:
Data for these compounds have either been collected relatively recently, are not fully
evaluated, or are non-existent.
Analytical costs of collocated sampling have prevented establishing a sufficient database.
Reference samples for blind laboratory analysis do not exist for many of the toxic
compounds.
Accuracy determinations are more difficult for complex mixtures of related compounds,
such as PCBs and dioxins.
Although precision and accuracy levels have not been determined for many of the toxic
compounds of interest, they can be estimated to establish Data Quality Objectives (DQOs) once
sufficient information on variability is established. The level of detection (LOD) or level of
quanthation (LOQ) of the analysis procedure is a reasonable starting point, particularly in the absence
of any other information on method variability. Precision and accuracy levels will be established
when quality assurance blind reference standards are available, when laboratory analysts gain
experience with the toxic compounds, and when sufficient information on method variability is
obtained. The data collected under the Area Source Program is to be used in conjunction with
modeling to evaluate progress toward reducing the cancer risk from area sources by 75 percent.
90
-------
February 1992
DQOs must be established to permit this determination in the presence of significant meteorological
variability.
Sites and Implementation
To meet the requirement of one site per Great Lake for measuring HAPs deposition by the
end of 1991, IADN placed one site in Canada on Lake Huron (Burnt Island), one on Lake Ontario
(Ft. Petre), and, in the United States, one each on Lake Michigan (Sleeping Bear Dunes), Lake Erie
(Sturgeon Point), and Lake Superior (Keweenaw Peninsula). Additional sites are recommended in the
Chesapeake Bay, Lake Champlain, and the coastal estuaries. .It is anticipated that the existing state
and university operated sites previously described will be excellent candidates for monitoring of both
the Chesapeake Bay and Lake Champlain. The CAAA do not specify the number of sites or
deadlines for these non-Great Lake monitors.
The Air Toxics work group recommends that the initial Great Waters sites adopt a minimum-
sampling approach as is in place for the IADN master stations, sufficient to meet the CAAA-required
deadlines. The work group determined that implementing a large network covering all potential toxic
pollutants would not be cost-effective, considering the complex nature of the research issues. In
addition to meeting the deadlines mandated by the CAAA, the work group recommends a minimum
sampling approach in other sites for geographic representation coupled with a research program in the
form of a comprehensive mass-balance pilot study on Lake Michigan. Lake Michigan was chosen for
the pilot study to support development of the Lake-wide Management Plan (LaMP), because a
significant preliminary urban air toxics study was conducted for 1991, and because additional
sampling is being planned for the first of the lake-wide mass balances. (Like IADN, LaMP was
spawned by the GLWQA. The mass balance of each Great Lake is an important part of these
management plans.)
91
-------
February 1992
The Lake Michigan pilot study can provide critical information on siting, network design, and
representativeness (urban vs. rural, land vs. water). Intensive, daily sampling will be conducted
during a three- to four-month period at multiple sites in and on the lake. It would be highly desirable
to monitor during different seasons of the year to clarify the pronounced seasonal concentration
patterns of pollutants such as PCBs.
Urban area source measurements of toxic air pollutants can overlap with VOC measurements
to be made by the enhanced ozone monitoring network for State Implementation Plan (SIP)
development as required by CAAA. Sites are to be located within the urban area of each of the 25
serious, severe, and extreme ozone nonattainment areas. To monitor urban area source pollution,
CASTNET will use at least one enhanced ozone monitoring site in each urban area design. The
measurements at the enhanced ozone site will be supplemented by other types of samplers (for PICs,
metals, and pesticides) and by increased analysis of VOCs to include toxic compounds such as carbon
tetrachloride and vinyl chloride as well as the VOCs needed for ozone modeling. Sites located near
Great Waters areas would be given high priority, because they would serve dual purposes.
In addition to technical considerations, the Air Toxics work group must coordinate both
monitoring and research resources and activities. For example, the Urban Area Source Program will
require coordination with the Enhanced Ozone Network. Such a well-defined, designed, and
integrated approach will allow a systematic examination of the complex research issues as well as an
efficient utilization of resources.
Relationship with Existing Networks
The proposed CASTNET network will share resources to the maximum extent possible with
existing and planned networks of all federal, state, or local agencies. Samplers will be added as
necessary in each of the networks for a common sampling array at all locations, but this overlap will
92
-------
February 1992
Network Minimum Number of Sites
Great Lakes 15
Pilot Mass-Balance on Lake Michigan 8-10
Lake Champlain
Chesapeake Bay
Coastal Estuaries
Urban Air Toxics
TOTAL
2
3
8-10
8-10
44-50
be allowed only when important common needs and quality concerns can be met.
Plans call for using the data from the EPA Toxics Air Monitoring System (TAMS) sites in
Boston and Houston, which were recently terminated (8/91), as the base of the Urban Area Source
program. These site locations also are being considered for inclusion in the enhanced ozone
monitoring network. The Air Toxics work group will incorporate planned enhanced ozone
monitoring sites into their efforts, if at all possible.
The monitoring requirement for five sites in the Great Lakes by December 31, 1991, has been
met by accelerating the schedule for the U.S.-Canadian-funded IADN. Existing state-operated sites
can be used as initial sites in the Chesapeake Bay and Lake Champlain areas if arrangements suitable
to all parties can be agreed upon.
Future Research Needs
There are a number of early research needs associated with Great Waters monitoring to
ensure mat monitors are appropriately sited and models are based on realistic processes.
Investigations should focus on but are not limited to:
93
-------
February 1992
Determining the representativeness of data from land-based sites used to characterize
deposition over water.
Determining the representativeness of data from nonurban locales used to characterize the
deposition to the entire lake surface.
Collecting sufficient data to assess parameters and define physical and chemical processes,
such as dry deposition rates (velocities), transmedia flux, particle sizing (including
pollutant concentration by particle size), and phase distribution.
Determining atmospheric reactivity and transformation of toxics.
Determining transboundary and long-range transport.
Evaluating natural source impacts, such as erosion.
Assessing environmental and human health effects.
Urban Area Source Program research needs which are to be addressed in part or in whole
tinder CASTNET include, but are not limited to:
Determining initial broad range of HAPs to be monitored.
Developing meteorological methodologies for determining mixing heights.
Developing data analysis procedures for source strength attribution.
Establishing human health risk assessment methods for the HAPs to be monitored.
Compiling health assessment data for urban pollutants.
Determining whether one sampling site per city is representative of the urban toxics
mixture.
* Characterizing the effect of mixtures in urban exposure scenarios for non-cancer
endpoints.
94
-------
February 1992
Key unresolved issues for network design are the total number of sites needed in the final
networks, the frequency of sampling necessary, and site locations. Data management issues include
formatting and archiving data, and establishing procedures for obtaining a good emissions inventory.
Consideration of these data management issues is necessary for modeling activities, for evaluating the
effectiveness of implementation measures taken for control of sources, and for predicting the impact
of future control measures.
The aforementioned research issues are unresolved because of a dearth of information on the
spatial and temporal patterns of these pollutants and on the extent to which vegetation and bodies of
water remove or introduce pollutants from or into the atmosphere. In addition, technical issues need
to be addressed regarding sampling/analysis limitations, sampling media, procedures/protocols,
sampling time, sample handling, shipping, analysis, and data reporting. Particular focus should be
directed at developing methods for sampling and analysis of toxic compounds of interest which cannot
currently be measured, and at standardizing and improving presently available methodologies.
The work group recommends a mmmqim monitoring approach until these research issues are
resolved, enabling the work group to complete network design criteria. At mat time, the network
should be expanded in order of the priority steps previously described.
95
-------
THIS PAGE INTENTIONALLY
LEFT BLANK
February 1992
96
-------
February 1992
Statistical Network Design for Status and Trends
The process of air and deposition network siting and data collection will be the critical factor
in obtaining valid statistical information on deposition status and trends. Network design is difficult
for air and deposition variables that vary in both time and space because these variations are usually
neither random nor well behaved. Examining the following aspects of network design will ensure the
ultimate provision of data that allow estimates of ecosystem exposure and adequate assessment of
trends to a desired level of precision:
Air and deposition variables of interest.
Integration of existing network data.
Frequency of field sampling.
Media representation and area! coverage.
Criteria in each of these areas form the basis for an optimal CASTNET network design. As
there is a threshold level of sites required to perform any sort of meaningful analysis, this
methodology has been applied only to wet deposition, dry deposition, and ozone. There are simply
too few toxics sites to allow for reliable statistical assessment of network design at mis time. The
statistical design of the visibility network is underway with assessment of candidate variables and a
review of existing networks for possible inclusion in the design.
Air and Deposition Variables
Core variables represent those species that are considered to affect adversely or beneficially
the health of ecosystems or to represent criteria air pollutants whose levels should respond to
emissions reductions. The Statistical Network Design work group has been advised by the Total
Deposition and Effects work groups that the core wet deposition variables are SO/, NO,', H+, and
97
-------
February 1992
rural O3 for ecosystem exposure and temporal trends analysis. Proposed core air concentration
variables are SO4*, Q,, HNQj, S02, and NO,'. As tbe CASTNET program objectives evolve,
variables may be added following a thorough evaluation of available monitoring and analytical
methods for each new variable.
Integration of Existing Monitoring Network Data
Several factors justify focusing on information from existing networks for which established
quality assurance and data comparability exist. First, such collaboration significantly reduces the
effort required to implement an integrated monitoring network, and it reflects CASTNETs role to
supplement and integrate, not replace, these networks. Second, the data can provide current status
and trend information for most of the atmospheric variables. All data will be evaluated for potential
contribution to the final, long-term monitoring network.
Determining which network sites to include is part of network design and requires considering
the objectives and spatial coverage of each network, quality assurance/quality control (QA/QQ
associated with the data, and compatibility of network monitoring and data validation protocols. Site
selection for spatial pattern and trend analyses may greatly affect the statistical results. Expert
judgment will be used hi conjunction with quantitative assessment in designing the network.
Frequency of Field Sampling
To establish a valid monitoring network, the effect of averaging times requires careful study.
Monitoring must be conducted on time and distance scales of die same magnitude as the duration of
changes in variable deposition/concentration levels. If possible, monitoring on a more refined time
scale is preferable to that intrinsic to the trend or spatial pattern under study. Even large-scale
changes can be misinterpreted when observations are averaged over long periods.
98
-------
February 1992
Successive weekly sampling periods should provide the level of accuracy and precision
required to satisfy the estimation goals for seasonal and annual trends for most of the core wet and
dry variables. Rural ozone measurements will be computed according to the definition of the SUM06
variable the variable frequently used to characterize rural ozone, which is the sum of successive
hourly ozone measurements above 0.06 ppm from April through October.
For purposes of detecting changes in air quality and in deposition that can be directly related
to changes occurring hi emissions from specific regions on a smaller scale, daily sampling appears
optimal. The NOAA monitoring programs are constructed around this premise, and will be designed
accordingly, in conjunction with the EPA CASTNET design.
Media Representation and Areal Coverage
Although atmospheric wet deposition variables should be monitored over the continental
United States, budget limitations may initially restrict complete implementation to high-priority
regions or sensitive ecosystems. The primary factors to be considered in network design are the
spatial distribution of sensitive ecosystems and sources of pollutant emissions. Because of scarcity of
siting or sampling equipment limitations, current deposition networks have under-represented areas,
such as high-elevation lakes and streams in the West and Eastern coastal areas. Correcting these
monitoring problems is the first priority for the Statistical Design Network work group.
Desion ot Optimal Network
Because the environment can never be sampled without error, determining the degree of
sampling uncertainty mat can be tolerated is an immediate concern. The network design must satisfy
design goals and meet the estimation requirements of several user groups. CASTNET serves two
main user groups within the scientific community: monitoring users who are interested primarily in
status and trends information; and ecosystem groups that require spatial estimates of
99
-------
February 1992
concentration/deposition for relating pollutant exposure to ecological health. CASTNET has three
broad monitoring goals hi meeting the needs of these user groups, and they are to:
1. Characterize in nonurban areas the status and trends of near-surface air concentrations and
deposition for regional/ecosystem areas.
2. Confirm that predicted results of control programs are achieved for chemical pollutant air
concentration and deposition.
3. Establish specialized concentration/deposition monitoring at ecosystem process study
research sites.
The ah- and deposition network should allow the detection of significant, long-term (seasonal
and annual) temporal trends and spatial patterns for the core monitoring variables. Determining the
spatial distribution of the core variables will require interpolation of data among monitoring sites.
Uncertainty for these estimates will depend on the natural variability, measurement precision, and
network designthe latter depending primarily on site density and location. If the uncertainties of
interpolated values provided by existing networks are too great, site density may need to be increased.
In addition, CASTNET should determine the response of the core variables to changing emissions
levels, a process mat requires assessing the sensitivity of existing networks hi detecting changes over
time. Short-term trends caused by meteorological variations must be distinguished from long-term
trends caused by changes in emissions. Long-term trends caused by changes hi climate will be
difficult to separate from anthropogenic trends, but need to be studied.
For the core monitoring variables, the specific estimation and network design objectives are
to:
1. Estimate seasonal and annual spatial patterns over nonurban areas within a relative
interpolation error of less than 50 percent at systematic grid locations.
100
-------
February 1992
2. Estimate average seasonal and annual deposition for die specified deposition areas with
average relative errors of less than 40 percent.
3. Estimate total change in deposition from data observed to date at existing monitoring sites
and for average deposition within the regional areas.
4. Assess the capabilities of deposition and air quality monitoring networks to detect
hypothesized future reduction scenarios for the core variables. This analysis will
determine the scope of monitoring needed to detect a trend of particular type and will
provide guidance on the potential need for additional monitoring to reduce time to
detection. .
The final network design must reflect consideration of data collection economics and intended
uses of the data by deposition trends and ecosystem groups. The current location of sites may not
prove sensitive to trends and may yield unacceptably large uncertainty levels associated with point or
ecosystem/regional estimates. Either deficiency may dictate adding or relocating monitoring sites to
improve CASTNETs ecosystem estimation capabilities. .
Preliminary Findings
Wet Deposition
Soatial Interpolation: The work group recommends locating additional sites in under-
represented high elevation and coastal ecosystem areas. The analysis presented below necessarily
excludes these areas, and the interpolation results are valid for areas representative of the
NADP/NTN siting criteria. An analysis of NADP/NTN annual 1989 wet sulfate, nitrate, and pH
data indicated that the relative errors of interpolation across most of the eastern United States were
less than SO percent and that the regional errors were below 40 percent, except in the West. Kriging
predictions are based on the correlation in measurements of neighboring sites (currently, mere are
101
-------
February 1992
very few sites within 150 km of one another, particularly in the West). Consequently, it is
anticipated that the error of interpolation could be reduced in regions characterized by high relative
errors by locating sites within 150 km of existing ones. To provide critical information on non-
monitored sensitive ecosystems, the work group recommends locating new sites at high-elevation
remote areas and coastal areas. Future research will incorporate NWS precipitation data to improve
the precision of spatial interpolation on annual and seasonal time scales.
Trend Detection: One of CASTNET's primary concerns is whether future deposition
monitoring will be capable of detecting changes in annual sulfate deposition in an accurate, timely
fashion. Familiarity with existing networks' detection and quantification capabilities is required for
evaluating current networks and for designing new or extended networks. An analysis of existing wet
deposition NADP/NTN sites in the eastern United States was conducted to determine whether this
network could detect and quantify the total RADM-projected sulfate deposition reduction (reduction of
10 million tons with no trading) in seven steps at 1994, 1995, 1996, 2000, 2001, 2002, and 2003 (see
Figure 11). The probability of correctly detecting a decreasing change in deposition is at least 80
percent by the end of the CAAA Phase I reductions in 1996 for the eastern half of me country.
These results do not apply to the Mississippi delta, where the probability is low («40 percent by
1996) due to the inadequate number of sites in this region and the Appalachians where little
monitoring information exists at high elevation areas. The probability of estimating the change in
sulfate (quantifying the trend) to within 20 percent of the projected change ranges from 40 to 65
percent by 1999. Increasing this probability to 80 percent would require monitoring to 2001 for die
Midwest, South, Mid-Atlantic, and Northeast regions. The variability of annual sulfate wet
deposition has been reduced by adjusting the data for the effects of precipitation (as measured by the
NADP/NTN). The work group recommends an additional 19 wet deposition monitoring sites in high
102
-------
Reliable
Detection
y
93 94 95 96 97 98 99
Reliable
Quantification
, T
00 01 02 03
5-
10
Percent
Reduction
15
20
25
30
i t
Figure 11. Wet sulfate regional trend assessment
103
-------
February 1992
elevation and coastal areas to provide the critical information required on these non-monitored
sensitive ecosystems for adequate spatial resolution.
Dry Deposition
An extremely limited pool of data exists for dry deposition analyses. Analyses similar to
those for wet deposition are being attempted using air concentration variables (SO3, SO^, and HNO3)
as surrogates for dry deposition. In the case of dry deposition, spatial fields are not as easily defined
as for wet deposition, because dry deposition varies greatly from place to place even when air
concentrations are the same. With so few existing sites, there is very poor characterization of the
spatial fields, resulting in large gaps. Because of the spatial irregularity of dry deposition (much like
the temporal irregularity of wet deposition), interpolation among existing data is particularly
vulnerable to error. Detection of trends in dry deposition can easiest be considered for each specific
site at which appropriate air concentration measurements are made. For trends over large areas,
detection relies on a modeled translation introducing additional error. This is complicated by the lack
of a historical data set. The work group recommends 31 additional dry deposition monitoring sites to
enhance trend quantification. At the same time, efforts will be ongoing to develop methods for
constructing improved area! averages (through improved inferential modeling) on the basis of these
data, on annual and seasonal scales.
Ozone
To increase the spatial resolution of ozone in sensitive ecosystems of interest and hi large
forested areas (terrestrial ecosystems) in the Southeast, the work group recommends an additional 45
ambient rural ozone monitoring sites. The data from these sites will also provide the concurrent
ozone and deposition data needed to differentiate if potentially adverse effects are due primarily to
ozone or to dry deposition.
104
-------
February 1992
Data Management
Recommendation
The Data Management group's most important role is to develop a comprehensive program of
standardization to ensure mat all CASTNET data, which will be provided by myriad entities, are of
comparable quality. Comparable-quality data are relatively easy to achieve for NAAQS variables for
which standard monitoring guidelines exist. For the other variables that will be monitored, collection
and analytical methods are not standardized. Although similar hi concept, the methodologies diverge
in important ways in bom the collection and analytical phases. A comprehensive data standardization
program is the first step in building a uniform database.
The Data Management group recommends EPA's Aerometric Information Retrieval System
(AIRS) as the eventual archival and reporting system for atmospheric measurements. AIRS is
currently limited to the NAAQS, meteorological, some toxic, and visibility measurements, but it can
be expanded and modified to manage CASTNET data.
Meanwhile, the existing NDDN and Acid Deposition System (ADS) database systems will be
merged into a new system that will serve as die archival and retrieval system for acid deposition and
acid aerosol measurements. The Atmospheric Research and Exposure Assessment Laboratory
(AREAL) will oversee receipt and validation of contractor-supplied data during the AIRS transition.
Additional databases will be managed separately, such as dry deposition calculations and interpolated
(kriged) area deposition grids. AIRS will be the source database for all acidic deposition calculations.
Databases external to AIRS will use common designs, architecture, and methodologies to the extent
possible to permit close integration with AIRS and other distributed systems, such as die
Environmental Monitoring and Assessment Program (EMAP).
105
-------
February 1992
Initial users of CASTNET data will be EPA regional offices, program offices, and scientists
interested in the extent of current deposition in particular regions. Congress has called for periodic
updates on the status, trends, and potential impact on ecosystems of acid deposition, ozone, and air
toxics through a series of mandated reports. Future regulatory actions may be based on the
broadening database, necessitating provable data quality. Generally, it is the intent of CASTNET
managers to prepare biennial assessment reports based on this data, and release the data itself within
one year of its being quality controlled and entered into the AIRS system.
Data Collection and Storage
Data are collected and stored for ongoing projects and several special-purpose studies.
Sampling intervals include hourly, weekly, bi-weekly, monthly, seasonal, yearly, and episodic, or by
event. Camera data are three 35-mm slides per day. Sample averaging times include daily, weekly,
monthly, quarterly, seasonal, and yearly. Both sampling interval and sample averaging times are
pollutant-specific and collection-method-specific.
Scientists retrieve information from a number of databases, including AIRS/Air Quality
Subsystem (AQS) and EMAP. Because certain information is available only in hard copy, the group
recommends*
Developing a database "clearinghouse" of acid rain reference material.
Developing an abstract bibliography of related materials with key word-search capability,
especially for acid deposition, air toxics, and visibility.
* Providing on-line laboratory and field standard operating procedure manuals.
Reporting criteria important to researchers include ad hoc, or customized, reporting format
capability; easy access; standardized reports for pollutants such as ozone and acid deposition,
106
-------
February 1992
including accepted statistical summaries for the data; and selection by monitoring network, site, and
special study name.
Frequency of existing database access varies from daily, weekly, quarterly, to yearly, and is
expected to increase with the CAAA implementation efforts underway.
The CASTNET data management system should include interactive and batch processing
capabilities; help screens; menu-driven software; PC front-end, report-generator, graphics/Geographic
Information System (GIS) capability; and a technical and user-support hotline. The work group
indicated that such a system would store related data in a single database, which would provide quick
access and better comparability of data as standards are defined.
Initial System Concent
The data management system will be a collection of databases, screening and acceptance
criteria, retrieval options, reports and associated analyses, and graphics capabilities (see Table 5).
The database consists of a catalog of metadata (descriptive and summary information) that facilitates
access to information stored in the main data repository.
Figure 12 shows the CASTNET information flow and basic system components required to
satisfy the needs of the CASTNET user community. Acquisition equipment for air pollutants is used
to obtain raw data, which will be converted to a format for input into the repository database. The
converted data will be stored in a temporary hold file for quality assurance/quality control (QA/QC)
processing.
Quality Assurance/Quality Control . ;
The two forms of QA/QC are data processing, which verifies the data integrity of the
converted data, and scientific, which verifies the quality of raw data by flagging the values and
ensuring proper Standard Operating Procedures (SOPs) are followed. Reports will show the status of
107
-------
TABLE 5. SYSTEM INPUTS, PROCESSES, AND OUTPUTS
Inputs
Processes
Outputs
Acid deposition data
APIOS-C
APIOS-D
APN
CAPMoN
NADP/NTN
MAP3S (historical)
TVA
UAPSP (historical)
GARB
FADS
GLAD
MGCP
NDDN
Aquatic and terrestrial effects
NADP/NTN
- NDDN
MAP3S (historical)
NPS
EMAP
NFHM
TIME
Visibility impairment . *
NPS
TVA
IMPROVE
Acid aerosol data
NPS
NDDN
IMPROVE
Air toxics
TAMS
Extended TAMS
IADN
QA/QC indicators
User interface
QA/QC
Data entry
Add
Modify
Delete
Database extraction
Algorithms
Data conversion
Input
Output
Query construction
Report generation
Raw data
Detailed reports
Summary reports
Error reports (QA)
Study abstracts
Study guidelines
Monitoring information
Methods
Stations
QA/QC standards
Graphical reports
108
-------
ill
LULL:
EEC
ES
CK--OH
II
*-
o
Q.
0)
DC
V)
o
+*
to
+5
CO
+"*
CO
CO
0
O CO
f*t
T3
O
e-
0)
4^
0)
LU
DC
O
CO
o
Q.
LU
CC
CO
CC
i«
£*<
^
- '
O
Q
O
CO
fe
^^
6
p
CO
08
t
INPUTS
Q.
O
O
"CD
CO
o
CO
CO
O
Aerosols
Met
CO
o
X
£
o
o -
>
I
CO
3
ri
109
-------
February 1992
these QA/QC checks. The raw data will then be input to the repository database along with die
QA/QC flags and will be available to scientists. CASTNET users will require raw data reports,
summary reports, and graphic output. Hie output function is the process by which data are extracted
and transferred to an external repository (for example, magnetic tape or disk file) in a specified
format.
AIRS Selection Rationale
The choice of AIRS to store measurement data was not preordained. The work group
considered various options before making its recommendation, ultimately developing a list of relative
advantages and disadvantages of using AIRS, which appears below:
Advantages
1. AIRS is accepted as the official national air quality database.
2. AIRS has an established infrastructure, staff, and budget.
3. User training is already established and easily accessible.
4. Some CASTNET users have already used AIRS to retrieve data.
5. Use of edit checking can improve overall data quality and reduce manual review of
ambient data.
6. AIRS access is available in EPA Regional Offices and Research Laboratories within EPA
and in a majority of state environmental offices. Extension to other agencies presents few
problems.
7. AIRS already contains some visibility data.
Disadvantages
1. Modification of AIRS is complicated by the need to avoid changing the original function
of the system.
110
-------
February 1992
2. The AIRS User's Guide initially will need to be used frequently. Planning is underway
for on-line help so the user can be less dependent on manuals.
3. Data submitted to AIRS that do not pass edit checks will be rejected and won't be loaded
into the AIRS files. The data manager will need to correct the data or override the AIRS
edit checks.
4. Graphic representation of data is not currently available but is being developed.
S. Cost of modifying large systems can be more than creating a new system. In the long
run, operating costs of a large system may be greater for a new system than for a
modified system.
Implementation Schedule
Assuming AIRS is used, the Data Management work group recommends the following
implementation schedule:
Phase I
1. Establish edit checks based on EPA-defined acceptance criteria.
2. Establish QA/QC for data of suspect quality.
3. Design separate CASTNET file (input routines and storage file).
4. Enhance output routines to provide CASTNET standard reports, summary reports, and
other EPA-desired outputs.
5. Modify AIRS Geo-Common to contain CASTNET-specific codes.
Phase n
1. Provide additional enhancements to AIRS/AQS as CASTNET evolves.
.. . 2. Provide additional standard reports. The ad hoc/custom report option may be included in
Phase I if it is a priority for CASTNET users.
Ill
-------
February 1992
3. Enhance QA/QC flagging capability.
Phase m
1. Design on-line help capability.
Phase IV
1. Provide additional graphic outputs hi the form of standard SAS/GRAPH reports and
customize graphic report capabilities.
2, Provide additional on-line help as user community expands.
112
-------
February 1992
Instrumentation/Methods
The Instrumentation/Methods work group is responsible for providing advice on
instrumentation and monitoring methods when questions arise in the other CASTNET committees.
Hie group has reviewed the status of instrumentation for the Total Deposition, Aquatic and Terrestrial
Effects, Visibility/Acid Aerosols, and Air Toxics program work groups. Written status reports have
been provided in these areas during the instrumentation selection process, and several projects critical
to CASTNET monitoring goals have been identified as methods development projects. Some of these
projects are being funded as part of other methods development programs, the results from which will
become part of the CASTNET information base for methods. Methods development activities for
monitoring of hydrocarbons and hazardous VOCs, as required by Titles I and m of the CAAA, are
examples (see Current Projects below). Other projects have been or will be funded as part of the
CASTNET program. Proposed future projects, general recommendations, and specific
recommendations by program work group are discussed in mis section. -
Current Projects
Developing Passive Sampling Devices (PSDs)
A set of passive sampling devices (PSDs) is needed for supplemental sampling sites around
the main CASTNET sites. These units require no electricity and may provide cost-effective
alternatives to fully configured monitoring sites. They appear especially well-suited to supplementing
spatial information on the ambient air concentrations of O,, NO,, and SO,. Field evaluation is
required to determine their ruggedness during week-long sampling intervals and to establish data
comparability with the real-time instruments that are the currently accepted monitoring standards for
network stations.
113
-------
February 1992
CASTNET has provided funding for continuing field evaluations of PSDs at the NDDN site
in Prince Edward, Virginia. These tests were begun in the fall of 1990 and included PSD
comparability tests for SO2, NO, NO2, and 0,. Maintaining this evaluation during the summer
months of 1991 was considered essential to monitor the effects of seasonal changes in atmospheric
chemistry on the integrity of samples. Atmospheric chemistry is also dependent on the part of the
country in which the monitoring is done, for example, paniculate nitrate is significantly higher at
west coast locations. For this reason, the NFS has arranged to deploy Q, PSDs at several of then-
sites across the country. Similar deployment of PSDs for SOj, NO, and NOj must be carried out in
the future.
As a continuation of the work on PSDs, a cooperative agreement with Harvard University has
been initiated to design a simplified network monitoring station to operate free of power utilities. He
design will be similar to the prototype station used at the NDDN site in Prince Edward, Virginia.
Evaluating the Versatile Air Pollution Sampler fVAPSJ
The Versatile Air Pollution Sampler (VAPS) is being evaluated as an alternative to the
IMPROVE sampler in use at network stations for characterizing visibility-related atmospheric
constituents. The VAPS is a new type of sampler that combines the features of a dichotomous
sampler and an annular denuder. The fine particles are split into two equal fractions and collected on
separate filterpacks, which allows different filter materials - such as teflon and quartz - to be used for
collecting fine particles. Annular denuder sections are used to remove individual gas species. An
additional feature of VAPS is mat die coarse fraction (containing coarse particles and a small fraction
of fine particles) can be used to determine the chemical properties of individual particles by scanning
electron microscopy. . .
114
-------
February 1992
The VAPS is being tested in the laboratory and in the field at a network monitoring station
maintained by the NFS in the Shenandoah Valley, Virginia. Comparisons between the VAPS and
IMPROVE sampler will be noted and may lead to simplification of sampler design. The ability of the
two samplers to provide acid aerosol concentration values is a major criterion hi their evaluation.
Real-Time Acid Aerosol Instrument Development
The infrared aerosol analyzer (IAA) has been developed by the Argonne National Laboratory
(ANL) to measure ammonium sulfate, ammonium bisulfate, and ammonium nitrate in near real tune.
The instrument provides the infrared absorption spectra of particulate matter collected by impaction
onto special substrates. The instrument is unique in the nondestructive measurement of the frequency,
temporal duration, and intensity of acid aerosol events. EPA requires this information to develop the
appropriate sampling strategy to relate health and ecological effects to acute and chronic exposure to
acid aerosols. . -
The IAA was used at a common monitoring site near State College, Pennsylvania, with other
instruments for one month in 1991. The study should yield a comparison of die IAA with 12- to 24-
hour tune-integrated sampling systems that require post-sampling analysis at a laboratory site. The
project also included the use of an FTIR (Fourier transform infrared) -based system for nondestructive
analysis of particles collected on in-line filters over 24 hours.
Initiating Experiments with Rherpacks
The filter pack used in NDDN differs from mat used in some other networks; the NOAA
filterpack, for example, exposes the incoming air to a small amount of heating, so as to eliminate
problems associated with liquid water forming on the particulate (teflon) and HNOj (nylon) filters.
The NDDN filterpack is similar to that used in Canada, which lacks heating of the inlet air (a nicety
that is likely unimportant at high latitudes but which could cause sampling differences in more humid
115
-------
February 1992
environments - e.g. the U.S. southern states). Hie NDDN sampler is known to read low for SO2
because SO2 is lost to the in-line nylon filter placed just before the carbonate-impregnated filter that
collects SO2. (This does not appear to be the case for the NOAA filterpack.) Even though the
NDDN nylon filter is extracted to retrieve captured S02, 10 to IS percent is not recoverable. To the
extent that the bias is constant, an average percent correction can be made. The bias is unknown,
however, unless it is measured each time a sample is taken. Special precautions must be taken to
detect any changes in the retentive properties of the nylon filters, such as those associated with
unannounced changes in die filter manufacturing process. Without these precautions, filter pack
analyses could indicate spurious trends.
To address this issue, arrangements have been made to check the bias periodically by loading
a second filter pack side-by-side, using a carbonate-impregnated filter in place of the nylon filter.
This procedure is currently in place at the five NDDN sites that operate duplicate filter packs.
Monitoring of Hazardous VOCs with Automated Gas Chromatographs (autoGCsJ
As mentioned in the Air Toxics section, automated gas chromatographs (autoOCs) will be
deployed at state-maintained monitoring stations to measure gaseous hydrocarbons (ozone precursors)
as required in Title I. At a subset of these stations, chromatograms from the autoGCs will be
examined for evidence of hazardous VOCs as an initial attempt at complying with Title m
requirements. Three methods development tasks are being carried out to strengthen autoGC design to
improve data capture, mlnipnigft use of consummables such as liquid cryogens, and to improve
chromatographic resolution for target VOCs of interest.
Sorbem-Based Sampling and Analysis of Hazardous Polar VOCs
Two methods development tasks are in progress to develop sampling and analysis methods for
polar VOCs using carbon-based solid sorbents for sample collection. This approach is receiving
116
-------
February 1992
renewed attention due to the availability of new sorbents and the acceptance of multi-sorbent tubes for
sample collection. In multi-sorbent tubes, two or three sorbents are packed in series to increase
retentive power. Heating the tube and backflushing with purge gas is one example of how adsorbed
compounds are retrieved. An alternative to thermal desorption of adsorbed compounds under
consideration is supercritical fluid extraction. The advantage to this latter method is that it would
avoid thermal decomposition of adsorbed compounds.
Projects for Future Work
Establishing Comparability of SO, PSD Resu/ts with Filterpack Results at Four Additional
Geographically Diverse Locations
Tests are needed to establish the comparability of data obtained by SO2 PSDs with data
obtained by filterpacks at different locations and during different seasons. Once such data are
available, a simpler monitoring scheme can be deployed. That is, a filterpack with only two filters
(for paniculate matter and HNO3) and two PSDs (for O3 and SO^ can obtain the same data currently
being collected at the NDDN stations, but without the question surrounding the SO2 values.
Deploying a Better,Sample Collection Device
A sampler design using a particle sizing inlet and annular denuder system sections could be
adapted to preserve and measure aerosol acidity and additional trace gases. The coarse particles are
collected on an inlet impactor. Annular denuder sections are added in front of the filterpack to collect
acid gases and ammonia. If fine particle acidity is required, then a minimum of a particle sizing inlet
and an ammonia denuder are needed. It may be necessary in some parts of the country to include in-
line filters (downstream of the fine particle filter) for HNO3 and NH3 to account for decomposition of
ammonium nitrate. This design option would be complemented by the use of PSDs for Q, and
possibly for NO and NO2.
117
-------
February 1992
Deploying a Screening Device for A cid Sulfate
Based on recent in-house experimental results at EPA, FTTR transmission spectroscopy can be
used as a screening device for acid sulfate collected on teflon filters. In-line filters, such as those
used in filterpacks, can be nondestructively analyzed for acid sulfate with a simple, compact,
commercially available FTIR system. This system needs to be field-demonstrated to gain the
credibility necessary for acceptance as a routine monitoring tool.
Testing PSDs for Other Trace Oases
The use of PSDs can be extended to other gases, such as NO, NOj, HNO3, and NH3. This
option depends on the perceived need by other CASTNET committees.
Developing a VOC Sampling System
Monitoring certain categories of VOCs may be desirable at CASTNET sites. Either a
sequential, battery-operated sampler (for multiple 24-hour samples) or an event sampler can be
considered. Designs which have incorporated specially prepared canisters for whole air collection or
the new generation carbon-based solid adsorbents are available commercially. Adequate field testing,
however, has not been completed and QA/QC procedures must be developed.
General Recommendations
The Instrumentation/Methods work group recommends several guidelines for improved
instrumentation:
Establish data comparability when changes in instrumentation are made so that information
on concentration trends is preserved.
Standardize instrumentation intended for gathering data of the same type in order to
minimize efforts needed to substantiate data comparability.
118
-------
February 1992
Design network instrumentation to facilitate internal quality assurance procedures and
thereby ensure high-quality data.
Compile and update the "white" papers developed by this work group for the program
work groups on the status of monitoring methods. Appendix D contains an example of a
"white" paper on methods for nitrogen containing compounds.
Specific Recommendations
Specific recommendations for each of the program work groups are as follows:
Total Deposition
Continue to establish data comparability between PSDs and samplers used for dry
deposition purposes.
Develop and test the throughfall method for dry deposition.
Develop/choose sampler for cloud impact sampling at high elevation sites. If acceptable,
deploy at high elevation sites hi the wet deposition network.
Aquatic and Terrestrial Effects
Continue interference testing of PSDs for sampling of NO2, SO2, and O3.
Consider the LEDAR (Light Detection and Ranging) fluorescence technique for airborne
measurement and survey of algae formation in estuaries.
* Add daily wet deposition sampling at selected sites to compare with longer-term sampling,
particularly for H+ and NH/.
VisibiKt y/Acid Aerosols
Finish characterization of the VAPS with respect to particle and acid gas transmission
characteristics of the cyclone and virtual impactor.
Run data comparability tests between the VAPS and IMPROVE samplers.
119
-------
February 1992
* Establish the frequency, duration, and intensity of acid aerosol events with the IAA at the
Shenandoah IMPROVE site.
Air Toxics
* Continue efforts to improve the autoGC design.
Continue efforts to improve sampling and analysis methods for polar VOCs using carbon-
based solid sorbents for sample collection.
Finish characterization of the UMASS sampler (the four channel, high volume sampler for
organics) and establish data comparability with PS-1 samplers.
Select and evaluate the most promising monitoring techniques for airborne mercury.
120
-------
Appendix A: CASTNET Work Group Members and Workshop Participants
lommittee Co-Chairs:
Dale A. Pahl
Atmospheric Research and Exposure Assessment Labontoiy
U.S. Environmental Protection Agency
MD-56
Research Triangle Park, NC 27711
Junes S. Vickecy
Atmospheric Research and Exposure Assessment Labontoiy
U.S. Environmental Protection Agency
MD-75
Research Triangle Park, NC 27711
1 Deposition:
Julian Chazin
Wisconsin Department of Natural Reaoutcet
Box 7921
Madison, WI 53707
loon F. Clarke
Atmospheric Research and Exposure Assessment Laboratory
U.S. Bnvironmenlal Protection Agency
MD-80
Research Triangle Park, NC 27711
Miguel Flores
National Park Service - Air
P.O. Box 25287
Denver, CO 80225
Jim Gibson
Colorado State University
Natural Resources Ecology Labontoiy
Fort Collins, CO 80525
William F. Hunt
OfBceof Air Quality Planning and Standards
U.S. Environmental Protection Agency
MD-14
Research Triangle Park, NC 27711
PaiUKapinos
U.S. Geological Survey
416 National Center
Reaton, VA 22092
Office of Atmospheric and Indoor Air Program*
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington, D.C. 20460
Barry E. Martin
Atmospheric Research and Exposure Assessment Laboratory
U.S. Environmental Protection Agency
MD-76
Research Triangle Park, NC 27711
Peter Summers
Atmospheric Environmental Service
4905 Dufferin Street
Downsview
Ontario, Canada M3H5T4
Aquatic and Terrestrial Effects:
* Ralph E. Battfngardner
Atmospheric Research and Exposure Assessment Laboratory
U.S. Environmental Protection Agency
MD-76
Research Triangle Park, NC 27711
Charles Boylea
RcnMciiicr Polytechnic
Dennis A. Leaf (ANR-445)
Clean Water Institute
Troy, NY 12181
Robert Hannah
Department of Environmental Quality
11720 Airline Highway
Baton Rouge, LA 70817
Dennis A. Leaf (ANR-445)
Office of Atmospheric and Indoor Air Programs
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington, D.C. 20460
A-l
-------
February 1992
PaulRingoId
Office of Environmental Processes and Effects Research
U.S. Environmental Protection Agency
RD-682
401M Street, S.W.
Washington, D.C. 20460
Rosaliaa Rodriguez
Office of Air Quality Planning and Standard*
U.S. Environmental Protection Agency
MD-I2
Research Triangle Park, NC 27711
John Stoddard
ManTech EnviroamenUl Technology, Inc.
200 SW 35th Street
Corvallit, OR 97333
KenSiolte
National Park Service - Air
P.O. Box 25287
Denver, CO 80225
Rick Slrassznan
Air Quality Division
Acid Deposition Division
Minneaota Pollution Control Agency
520 Lafayette Avenue
St. Paul, MN 55155
Alan Van Arsdale
NESCAUM
85 Merrimsc Street
Boston, MA 02114
Visibility/AcidAerosols;
VicfciAtwell
Office of Air Quality Planning and Standard*
U.S. Environmental Protection Agency
MD-11
Research Triangle Park, NC 27711
Neil J. Berg, Jr.
Office of Air Quality Planning and Standards
U.S. Eovironmental Protection Agency
MD-14
Reaeareh Triangle Park. NC 27711
Robert M. Burton
Atmospheric Reaeareh and Exposure Atariimmt Laboratory
U.S. Environmental Protection Agency
MD-56
Research Triangle Park, NC 27711
E. Gardner Evani
Atmospheric Research and Exposure Aaaeeement Laboratory
VS. Bnviromnental Protection Agency
MD-56
Reiearcn Triangle Park, NC 27711
Rich Fisher
Rocky Mountain Station
USDA Forest Service
240 West Prospect
Fort Coliint, CO 80526
Bill Malm
National Park Service - Air
CIRA - Foothills Campua
Colorado State University
Fort Collioi, CO 80523
Jan H. Moneysmitb
Environmental Services Division
U.S. Environmental Protection Agency
MD6E-SA
First Interstate Bank Tower at Fountain Place
1445 Ross Avenue 12th Floor, Suite 1200
Dallas, TX 75202
BillOshmd
California Air Resources Board
Monitoring Laboratory Division
P.O. Box 2815
Sacramento, CA 95812
MarkPitchford
Environmental Monitoring Systems Laboratory (EMSL)/ORD
U.S. Environmental Protection Agency
P.O. Box 15027
Las Vegas, NV 89114
Richard Potrot
Vermont Air Program
Building 3 South
109 S. Main Street
Weterbury.VT 05676
Bruce V. Polkowafcy
Office of Air Quafity Planning and Standards
U.S. Environmental Protection Agency
MD-12
Research Triangle Park, NC 27711
Larry J. Purdue
Atmospheric Research and Exposure Assessment Laboratory
U.S. Environmental Protection Agency
MD-77
Research Triangle Park. NC 27711
Mark Scruggs
National Park Service - Air
P.O. Box 25287
Denver, CO 80225
Riiuell W. Wiener
Atmospheric Research and Exposure Assessment Laboratory
A-2
-------
February 1992
U.S. Environmental Protection Agency
MD-77
Reaearch Triangle Park, NC 27711
RiekArtz
NOAA-Air Resources Laboratory (R/B/AR)
SSMCJl, Rm. 9358
1325 Eait Wen Highway
Silver Spring, MD 20910
Terry L. Clait
Atmospheric Research and Exposure Assessment Laboratory
U.S. Environmental Protection Agency
MD-80
Research Triangle Park, NC 27711
Larry T. Cupitt
Atmospheric Reaearcb and Exposure Assessment Laboratory
U.S. Environmenial Protection Agency
MD-75
Research Triangle Park, NC 27711
Geraldine J. Dorosz
Office of Air Quality FUnning and Standards
U.S. Environmental Protection Agency
MD-14
Rewarch Triangle Park, NC 27711
Robert B. Faoro
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency ...
MD-14
Reaearch Triangle Park, NC 27711
joanne Foy * >.
Michigan Department of Natural Resources.,. . . -
Air Quality Division , - -
Box 30028
Laming, MI 48909
Steve Hedtke . .
Environmental Reaearch Laboratory
U.S. Eavironmental Protection Agency '
6201 Congdon Boulevard
Duhrth, MN 55804
Dwight Hlustick
Office of Reaearch and Development
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington, D.C. 20460
AlaoJ. Hoffiaan
Atmoapheric Reaearcb and Expomre Aueaiment Laboratory
U.S. Environmental Protection Agency
MD-76
Research Triangle Park, NC 27711
Meluaa W. McCullough
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
MD-13
Reaearcb Triangle Park, NC 27711
Neville Reid
Air Resource* Branch, OME
880 Bay Street, 4th Floor
Toronto, Canada M5S 1Z8
Wayne Wilford
Great Lake* National Program Office
U.S. Environmental Protection Agency
230 South Dearborn Street
Chicago, IL 60604
Statistical Network Design for Status and Trends;
Thomas C. Cuma
Office of Air Quality Planning and Standardi
U.S. Environmental Protection Agency
MD-14
Research Triangle Park, NC 27711
Robin L. Dennia
Atmospheric Reaearch and Exposure Aaaeatroent Laboratory
U.S. Environmental Protection Agency
MD-80
Reaearch Triangle Park, NC 27711
Michele Ferland
Director, Atmospheric Networks
2360 Chemin Ste-Foy
Sto-Foy, Quebec
OIV4H2
ntalCo
Philip I. GaMn
New York State Department of Envir
30 Wold Road
Room 140
Albany, NY 12233
David M. Holland
Atmospheric Reaearch and Exposure Attewmeot Laboratory
U.S. Environmental Protection Agency
MD-56
Reaearch Triangle Park, NC 27711
Larry Lanen
California Air Reaoureea Board
Technical Support DivUion
P.O. Box 2815
1131 S. Street
Sacramento, CA 95812
Tony Olsen
Environmental R
aearch Laboratory /ORD
U.S. Environmental Protection Agency
A-3
-------
February 1992
200 SW 35ft Street
CorvaIIii,OR 97333
Bob Vet
Atmospheric Environmental Service*
4905 Duflerin Street
Downsview
Ontario, Canada M3H5T4
Data Management:
John C. Bosch
Office of Air Quality Panning and Standards
U.S. Environmental Protection Agency
MD-14
Research Triangle Park, NC 27711
Thomas C. Lawless
Aunoipheric Research and Exposure Assessment Laboratory
U.S. Environmental Protection Agency
MD-56
Research Triangle Park, NC 27711
lameaA. Reagan
Atmospheric Research and Exposure Aaaeanneat Laboratory
U.S. Enviroiiniental Protection Agency
MD-56
Research Triangle Park, NC 27711
IMC M. Sune
Atmospheric Research and Exposure Aitfiismntt Laboratory
U.S. Environmental Protection Agency
MD-76
Reaearch Triangle Park, NC 27711
Instrumentation/Methods:
^
William F. Barnard
Atmoapheric Research and Exposure Assessment Laboratory
U.S. Eovironinental Protection Agency
MD-77B
Reaearch Triangle Park, NC 27711
Harold M. Barnes
Atmospheric Rewarch and Exposure Assessment Laboratory
U.S. Environmental Protection Agency
MD-46
Reaearch Triangle Park, NC 27711
Thomas C. BUeatad
Atmospheric Rewarch and Expocure Assegsmeni Laboratory
U.S. Environmental Protection Agency
MD-S7
Research Triangle Park, NC 27711
Thomas A. Hanlage
Atmospheric Research and Exposure Assessment Laboratory
U.S. Environment^] Protection Agency
MD-76
Research Triangle Park, NC 27711
William F. Hunt
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
MD-14
Research Triangle Park, NC 27711
snt Laboratory
William A. McClenny
Atmospheric Reaearch aod Exposure Asses!
U.S. Environmental Protection Agency
MD-M
Research Triangle Park, NC 27711
StuMcNair
Atmospheric Enviroomental Services
4905 Dufferin Street
Downsview
Ontario, Canada M3H 5T4
James D. Mulik
Atmospheric Reaearch and Exposure Assessment Laboratory
U.S. Environmental Protection Agency
MD-44
Reaearch Triangle Park, NC 27711
AlanW.Oi
Environmental Services Division
U.S. Environmental Protection Agency
60 Wearview Street
Lexington, MA 02173
Joseph B. Sickles, U.
Atmospheric Research and Exposure Assessment Laboratory
U.S. Envtronmental Protection Agency
MD-7S
Reaearch Triangle Park, NC 27711
Robert K. Stevens
Atmoapheric Reaearch and Exposure Assessment Laboratory
U.S. EnvironmcDtal Protection Agency
MD-47
Reaearch Triangle Park, NC 27711
Robert M. Burton
Atmospheric Reaearch and Exposure
U.S. Environmental Protection Agency
MD-56
Reaearch Triangle Park. NC 27711
Contract Acquisition:
Assessment Laboratory
Rudolph P. Bokaleitner
Atmospheric Research and Exposure Assessment Laboratory
VJS. Environmental Protection Agency
A-4
-------
February 1992
MD-76
Research Triangle Park, NC 27711
Andrew E. Bond
Atmospheric Research and Exposure Assessment Laboratory
U.S. Environmental Protection Agency
MD-76
Research Triangle Park, NC 27711
Steven M. Bromberg
Atmospheric Research and Exposure Assessment Laboratory
U.S. Environmental Protection Agency
MD-75
Research Triangle Park, NC 27711
lohnB. Clioe
Office of Administration and Resources Management
U.S. Environmental Protection Agency
MD-33
Research Triangle Park, NC 27711
William P. Hunt
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
MD-M
Research Triangle Park, NC 27711
er Workshop Participants
John D* Bachmann
Office of Air Quality Planning and Standard!
U.S. Environmental Protection Agency
MD-11
Research Triangle Park, NC 27711
Joe Barnard
USDA Forest Service
Forestry Science Laboratory
Box 12254
Research Triangle Park, NC 27709
Jack A. Bowen
Atmospheric Research and Exposure Assessment Laboratory
U.S. Eoviromncntal Protection Agency
MD-77B
Research Triangle Park, NC 27711
Ronald L. Bndow
Atmospheric Research and Exposure Assessment Laboratory
U.S. Eavinuunental Protection Agency
MD-56
Research Triangle Park, NC 27711
Robert S. Chapman
Health Effects Research Laboratory
U.S. Environmental Protection Agency
MD-58
Research Triangle Park, NC 27711
Nancy Cobb
NOAA/OAR
1335 East-West Highway
Silver Spring, MD 20910
William M. Cox
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
MD-14
Research Triangle Park. NC 27711
Ed deSteiguer
U.S. Forest Service
1509 Vanity Drive
Raleigh, NC 27606
Ronald J. Drago
Atmospheric Research and Exposure Assessment Laboratory
U.S. Environmental Protection Agency
MD-76
Research Triangle Park, NC 27711
Gary F. Evans
Atmospheric Research and Exposure Assessment Laboratory
U.S. Environmental Protection Agency
MD-56
Research Triangle Park, NC 27711
Gary J. Foley
Atmospheric Research and Exposure Assessment Laboratory
VS. Environmental Protection Agency
MD-75
Research Triangle Park, NC 27711
William G.'
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
MD-14
Research Triangle Park, NC 27711
Robert G. Lewis
Atmospheric Research and Exposure Assessment Laboratory
U.S. Environmental Protection Agency
MEM4
Research Triangle Park, NC 27711
RickLii
Atmospheric Research and Exposure Assessment Laboratory
U.S. Environmental Protection Agency
MD-75
Research Triangle Park, NC 27711
Lester Machta
NOAA-Air Resources Laboratory (R/E/AR)
SSMCn. Rm. 9358
1325 East West Highway
Silver Spring, MD 20910
Frank McEiroy (RD-680)
A-5
-------
February 1992
Office of Reaearch and Development
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington, D.C. 20460
Jay I. Metier
Atmospheric Research and Exposure Assessment Laboratory
U.S. Environmental Protection Agency
MD-75
Research Triangle Pa*, NC 27711
William J. Mitchell
Atmospheric Research and Exposure Assessment Laboratory
U.S. Environmental Protection Agency
MD-75
Research Triangle Park, NC 27711
Peter K. Mueller
EPRI
P.O. Box 10412
Palo Alto, CA
Tom Pheiffer
Office of Program Management
VS. Environmental Protection Agency
OS-110
401 M Street, S.W.
Washington, D.C. 20460
Linda P. Porter
Atmospheric Research and Exposure Aneiiment Laboratory
VS. Environmental Protection Agency
MD-77B
Research Triangle Park, NC 27711
Allen Wiebe
Atmoapheric Environmental Service
4905 Dufferin Street
Downjview
Ontario, Canada M3H5T4
Nancy K. Wilson
Atmoapberie Research and Exposure Aaseaament Laboratory
U.S. Environmental Protection Agency
MD-44
Research Triangle Park, NC 27711
Lisa I. Smith
Atmospheric Research and Exposure
U.S. Environmental Protection Agency
MD-S6
Rewarch Triangle Park. NC 27711
L&bonttny
Jack C.Sugg*
Atmoapheric Research and Exposure Asaeaameot Laboratory
U^. Environmental Protection Agency
MD-77B
Reaeareb Triangle Park, NC 27711
Hemy C. Thomas, Jr.
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
MD-12
Reaearch Triangle Park, NC 27711
Dan Vallero
Atmoapheric Research and Exposure Assessment Laboratory
U.S. Environmental Protection Agency
MD-75
Research Triangle Park, NC 27711
A-6
-------
February 1992
* Denotes work group co-chain.
'Retired.
' On uignment to the Office of Air Quality Planning and Standard!.
A-7
-------
-------
February 1992
Appendix B: CASTNET CAAA Mandate Summary and Reporting Requirements
Total Deposition and Aquatic and Terrestrial Effects
Title IV:
Section 404
Report in 36 months on the identification of sensitive and critically sensitive aquatic and
terrestrial resources, and the feasibility and effectiveness of an acid deposition standard or
standards to protect them.
Title IX:
Section I03(c}
Establishment of a national network to monitor, collect, and compile data with quantification
of certainty in the status and trends of deposition, surface water quality, and forest condition.
Develop improved methods and technologies to increase understanding of the sources of
ozone precursors, and on ozone formation and transport. Reports to Congress every 5 years
which evaluate and assess the effectiveness of air pollution control regulations.
Section 203(e)
EPA, in cooperation with NOAA, shall conduct a research program on the short-term and
long-term causes, effects, and trends of ecosystems damage from air pollutants. The program
is to include: an evaluation of the risks to ecosystems, with causes and effects of chronic and
episodic exposures to air pollutants; an evaluation of air pollution effects on water quality,
including ecological effects of air pollution on surface water (including wetlands and estuaries)
and groundwater; and an evaluation of air pollution effects on forests, crops, biological
diversity, soils, and other aquatic and terrestrial systems.
Section 1030)
Continuation of the National Acid Precipitation Assessment Program (NAPAP). Reports to
Congress starting in 1992 and biennially thereafter on actual and projected emissions and acid
deposition trends; ambient concentrations; status of ecosystems (including forests and surface
waters), materials, and visibility; causes and effects of such deposition, including changes in
surface water quality and forest and soil conditions; and the occurrence and effects of episodic
acidification, particularly to high elevation watersheds. In 1996, and every 4 years thereafter,
the report shall include: deposition rates necessary to prevent adverse ecological effects; and
the costs and benefits of the acid deposition control program.
B-l
-------
February 1992
Section 901 (g)
Report to Congress annually, with periodic assessment reports, on the occurrence and effects
of: acid deposition on surface waters west of the Mississippi River; acid deposition on high
elevation ecosystems (including forests and surface waters); and on episodic acidification.
Visibility/Acid Aerosols
We VIII
Section 169B
EPA, in conjunction with the NFS, is to conduct research on sources and source regions of
both visibility impairment and regions that provide predominately clean air in Class I areas.
To include: expansion of current visibility monitoring in Class I areas; assessment of current
sources of visibility impairing pollution; adaptation of regional air quality models; and studies
of the atmospheric chemistry and physics of visibility. Report on interim findings in 3 years.
In 24 months, assess likely improvement in visibility in Class I areas due to CAAA. Every 5
years thereafter, assess the actual progress and improvement in visibility in Class I areas.
Tide IX
Section 103(c)
Establishment of a national network to monitor, collect, and compile data with quantification
of certainty in the status and trends of visibility impairment. Reports to Congress every 5
years which evaluate and assess the effectiveness of air pollution control regulations.
B-2
-------
February 1992
Air Toxics
We III
Section 112(k)
EPA to conduct research on the sources of HAPs in urban areas. To include: ambient
monitoring of VOCs, metals, pesticides, PICs, and other toxic pollutants; sources of
pollution, focusing on area sources and their contribution to public health risks; consideration
of atmospheric transformation and other factors which can elevate public health risks; and the
role of such pollutants as precursors of ozone or acid aerosol formation. Preliminary results
of such research is due in 3 years. A comprehensive strategy to control emissions of HAPs
from area sources in urban areas is due to Congress in 5 years.
Section 112(m)
EPA, in cooperation with NOAA, to identify and assess the extent of atmospheric deposition
of HAPs (and other pollutants) to the Great Lakes, Chesapeake Bay, Lake Champlain, and
coastal waters. To include: monitoring; sources and deposition rates of air pollutants and
precursors; relative contribution of HAPs to total pollution loadings; the contribution of mis
deposition to violations of water quality standards; and sampling for such pollutants hi biota,
fish, and wildlife, with characterization of the sources of such pollutants. One monitoring
site, for both wet and dry deposition, to be established in each of the Great Lakes by
12/31/91. Report in 3 years, then biennially, to include at a minimum: 1) relative pollutant
loading from the atmosphere compared to total loading; 2) adverse effects on human health
and the environment; 3) sources of those HAPs; 4) contribution to violations of water quality
standards or criteria; and 5) recommended regulatory revisions to prevent adverse effects. In
5 years, promulgate any further standards deemed necessary (per the report to Congress) to
prevent adverse effects, including effects due to bioaccumulation and indirect exposure
pathways.
B-3
-------
-------
February 1992
Appendix C: Selected References
Aubertin, G.M.; Bigelow, D.S.; Malo, B.A. (editors) 1990. Quality Assurance Plan: NADP/NTN
Deposition Monitoring. National Atmospheric Deposition Program. Natural Resources
Ecology Laboratory, Colorado State University; Fort Collins, CO.
Bigelow, D.S. 1984. Instruction Manual: NADP/NTN Site Selection and Installation. Natural
Resources Ecology Laboratory, Colorado State University; Fort Collins, CO.
Bigelow, D.S.; Dossett, S.R. 1988. NADP/NTN Instruction Manual: Site Operation. Natural
Resources Ecology Laboratory, Colorado State University; Fort Collins, CO.
Environmental Sciences and Engineering, Inc. 1990. National Dry Deposition Network. Project
Work Plan. Gainesville, FL.
Environmental Sciences and Engineering, Inc. 1990. National Dry Deposition Network. Field
Operations Manual. Gainesville, FL.
Environmental Sciences and Engineering, Inc. 1989. National Dry Deposition Network. Project
Quality Assurance Plan. GainesvUle, FL.
Morrison, M.; Newell, A.D.; Hjort, R. 1990. Data User's Guide to the United States
Environmental Protection Agency's Long-Term Monitoring Project: Quality Assurance Plan
and Data Dictionary. EPA 600/_/_. Office of Research and Development, Washington,
D.C.
Peck, D. 1991. EMAP Integrated Quality Assurance Project Plan for the Surface Waters
Resource Group - Fiscal Year 1991. EPA 600/_/_. Office of Research and Development,
Washington, D.C.
U.S. Environmental Protection Agency. 1990. IMPROVE Progress Report: Appendices A-H.
EPA 4SO/4-90-008a and 008b.
U.S. Environmental Protection Agency. 1987. Handbook of Methods for Acid Deposition
Studies, Laboratory Analyses for Surface Water Chemistry. EPA 600/4-87/026. Office of
Research and Development, Washington, D.C.
U.S. Environmental Protection Agency/Canada Ontario Coordinated Committee on Annex 15.
IADN Implementation Plan. 1990.
C-l
-------
-------
February 1992
Appendix D: Sample Instrumentation/Methods "White" Paper
Status of Measurement Methodology for Nitrogen Oxides
by J.E. Sickles, II
Oxides of nitrogen include many individual chemical species: NO. N02. HNO3, HN02,
NO3 radical, N2O5, PAN. other organic nitrates, and paniculate nitrates (NO3~). Table I
summarizes selected characteristics ol instruments and methods currently available for
sampling and analyzing many of these compounds. N2OS is not included because tt has not
been determined in the troposphere, tt has, however, been observed in the stratosphere
using spectroscopic methods.
Measuring the chemiluminescence (CLM) following the reaction of NO and O3 is the
method of choice (or NO. Commercially available instrumentation can measure NO
specifically with a nominal detection limit of 1 ppb. As noted in Table I, refinements of CLM
as well as other technologies exist for the determination of NO. These methods are generally
research or laboratory tools and as a result must undergo refinement prior to deployment in
routine monitoring networks. *'.-.--
Nitrogen dioxide may be measured as NO by CLM following catalytic or photolytic
conversion. Commercially available CLM instruments have detection Dmits in the range of 1
to 10 ppb. This type of device is not specific for N0? and other gaseous oxides of nitrogen
may. act as direct interferents. The Luminox NO2 Analyzer is sensitive but has a nonlinear
response below 10 ppb and is subject to interferences by PAN, HNOj, and O3.
Spectroscopic methods (e.g. TOLAS and DOAS) are specific and sensitive, but their cost is
a major limitation to widespread deployment Integratjve methods employing a chemical
reaction with NO2 are available and range from bubbler (EPA equivalent) methods to passive
sampling devices (PSD). These methods are subject to interferences from other nitrogen
oxides, but with development may prove useful.
D-l
-------
February 1992
It should be noted that in the case of integrative methods the sensitivity depends on the
sampled air volume and the sensitivity of the analytical finish. As an example, for a sampler
operating at 1 m3/h for 24-h, extracted in 10 ml, and analyzed by 1C with a sensitivity of
0.05 pg/ml, the method sensitivity is 0.02 pg/m3 (0.01 ppb NO2). This sensitivity would be
0.3 ppb for a 1-h sample at the same flow rate and 0.2 ppb for a 24-h sample at 1 ipm.
PAN and other organic nitrates are most commonly determined using integrative
sampling with analysis by gas chromatography with electron capture detection (GC-ECD).
Other methods are in early stages of development.
The simplest method for HN03 employs collection downstream of a paniculate filter on
a nylon or chemically impregnated filter followed by extraction and analysis for NO3- by ion
chromatography (1C). Although the method is appealing because of its simplicity, it is subject
to artifacts/interferences (e.g., volatilizing NH^NO-j). The denuder avoids these problems, but
Is more cumbersome and complex to use. Spectroscopic methods are available, but their
expense is a major limitatioa
The denuder approach (employing a NaCI-coaied denuder followed by two NdjCO^
coated denuders) can provide sensitive determinations of HNO^ but is limited by its
integrative approach. Other methods (i.e., differential optical absorption spectroscopy and
laser induced fluorescence) are research tools and may not be well suited for widespread
deployment.
The nitrate radical may be determined sensitively using remote sensing (le.. OOAS). it
should be noted that the sensitivity of methods that rely on fight absorption increases with the
path length employed. Path lengths for the various absorption methods are given in Table I.
Paniculate NO3- is easily collected on filters. Ambient aerosols including NH4NO3 have
significant vapor pressures at ambient temperature and pressure, and as a result may
volatilize during sampling. Ammonium nitrate may be in equilibrium with gaseous HN03 and
NHg in ambient air. The use of denuders upstream to remove HNO9 and NHg before filter(s)
to collect paniculate N03~ permits reconciliation of ambient gaseous and paniculate N03~
As an aside, pulsed fluorescence and flame photometric based instruments may be used
to monitor SOg. Commercially available instruments have 10-20 ppb detection limits, with
D-2
-------
February 1992
precision of 2-6 ppb, response times of 2-5 min. and minor interferences from aromatic
hydrocarbons. Newer, more sensitive models have shown a detection limit of 0.1 ppb.
precision of 0.03 ppb, response time of 4 min, and interferences of 2-3% for NO, 1% for
xylene, and 0.1% for H2S. Sulfur dioxide can also be collected using solutions of water,
glycerine, and alkaline salts (e.g., NagCOg) as coatings for denuders or filters. The
performance of these methods for 24-h samples shows detection limits of 8 ppt and precision
of Jess than 5%.
Paniculate SO4*~ is easily collected on filters. Suttates are not known to volatilize. As a
result denuders are not required to sample S042' unambiguously. However, Teflon or quartz
fitters are preferred since drawing S02-laden air through alkaline (i.e., glass fiber) filters can
cause the formation of artifact SO4Z~ on the filters. For 24-h samples, filtration shows
performance similar to that for the integrative methods for SO2 noted earlier: detection limit
of 0.02 jtg/m3 and precision of less than 5%.
units = ppb
Lab
Lab
Field
NO,
-S0.2
LMA-3 CSI-1600 TECO 43s
LoD (ppb)
LoQ (ppb)
0.01
0.03
0.3
0.1
0:3
0.03
Precision (ppb) 0.6
Interferences
PAN 65%
HN02 50%
HN03 0%
NO NA
HjS NA
Xylene NA
Response Time (min) 25 4
Reference WfigM. R.S., etal. (1989). Laboratory and Field Evaluations oJ ExtruensMve NO2 end SO, Analyze*.
Paper 89-65.1 AWMA 62nd Annual Meeting, Anaheim. CA.
NA
NA
100%
NA
NA
NA
NA
NA
NA
2-3%
0.1%
1%
D-3
-------
February 1992
TABLE I - Selected Instruments and Methods for Determining Ambient Oxides of Nitrogen
Species
NO
NOZ
PAN
Other Nitrates
HNO,
*
Method*
CLM(1)
CLM(1)
LIF
TOLAS
TTFMS
Passive
CLM(1)
CLM(1)
CLM(2)
Uf
TOLAS
TTFMS
OOAS
DIAL
Bubbler
TEA Fitter
Gualacol Denuder
DPACart
TEA PSD
GC-ECD
GC-CLM(f)
AlkaBne Hydrolysis
TTFMS
OC-ECD
FDtor
Denuder
TOLAS
TTFMS
Tree"
tAC
IAC
IAC
IAC
IAC
IP.I
IAC
IAC
IAC
IAC
IAC
IAC
RAC
RAC
IAI
IAI
UU
IAI
IP.I
IAI
IAt
IAI
IAC
IAI
IAI
IAI
IAC
IAC
Development Measure
Stage0 Time
C
R
R
R.C
L
C
C
R
C
R
ac
L
ac
H
RM
L
L
L
L
aRM
L
L
L
R
aRM
aRM
ac
L
Srnhi
2-606
30s
60S
60S
24 hd .
5 mh>
<100S
1008
2ffiin
60s
60S
12rr*i
.
24 h"
24 h-
1h«
9hd
24 h*
IS min*
*
-
60S
24 h<
24 h-
24 h"
Snrin
60s
*nf .
Precision
0.3 ppb
20ppl
-
.
.
-
10%
20ppt
0.6 ppb
*
20ppt
.
*
.
,
6 ppb
15%
4%
e%
30%
~
-
'
»
10%
8%
.
- .
Performance
Accuracy
^
ZOppt
16%
-
.
-
20%
30%
.
16%
15%
.
10%
20%
10%
.
.
ippb
30%
-
20%
*
20%
-
MDL
ippb
Sppt
IDppt
0.5 ppb
4ppt
70 ppb-h
9 ppb
10-25 ppt
10ppt
12ppl
100 ppt
0.3 ppt
4 ppb
10 ppb <
eppb
0.2 ppb
0.1 ppb
0.1 ppb
30 ppb-h
10ppl
10 ppb
-
2ppt
1ppt
Sppt
eppt
too ppt
0.3 ppJ
Comments
f
-
Two-photon
40-m path length
100-rn path length
Uses oxidizer plus TEA
EPA reference method; many interferences
Uses thermal or phototytic converters
Interferences: PAN. HNO,. Oa:
Non linear response below 3 ppb
Two-photon
ISO-m path length
.
800-m path length
6-km path tength
EPA equivalent methods'
Interferences: PAN and HNOZ*
Stability of extract uncertain*
DPA may volatilize;
Interferences: KNO2 and PAN
SimBar lo Palmes Tube; interferences as above*
Sensitivity can be enhanced by using
cryogenic sampling and capillary columns
*
e
100-m path length
Sample collected on charcoal
May be nylon or Nad-Impregnated titter;
Subject lo artifacts*
Not subject lo above artifacts*
150-m path length
100-m path length
D-4
-------
February 1992
ill I
!*
TO ^
§V
------- |