Quality Assurance Handbook for Air
Pollution Measurement Systems
Volume IV: Meteorological Measurements
Version 1.0 (Draft)

-------
                                                      EPA-454/D-06-001
                                                           October 2006
Quality Assurance Handbook for Air Pollution Measurement Systems
          Volume IV: Meteorological Measurements (Draft)
               U.S. Environmental Protection Agency
             Office of Air Quality Planning and Standards
                   Air Quality Analysis Division
                   Measurement Technology and
                  Ambient Air Monitoring Groups
            Research Triangle Park, North Carolina, 27711

-------
                                                                                       Volume 4
                                                                                  Re vision No: 1.0
                                                                                    Date: 10/2006
                                                                                      Pageiiofxi
                                          Foreword
This document represents Volume 4 of a 5-volume quality assurance (QA) handbook series. This volume
is dedicated to meteorological measurement systems and their support equipment. Volume I provides
general QA guidance that is pertinent to the remaining volumes. Volume II is dedicated to the Ambient
Air Quality Surveillance Program and the data collection activities of that program.  Volume III pertains
to Source and Emission monitoring methods. Volume V pertains to ambient dry deposition.

This document, Volume IV, is designed to provide clear and concise information and guidance to the
State/Local/Tribal (SLT) air pollution control agencies that operate meteorological monitoring equipment
and systems.  Recently, the new monitoring rule was published, which establishes the requirements for
meteorological monitoring in support of National Core (NCore) network. The new monitoring rules
require that meteorological data be collected at all NCore stations, as stated in the Code of Federal
Regulations (CFR) Chapter 40 Section 58, Appendix D.S.b. Thus, there is a need for updated information
to guide the SLT agencies as they implement the NCore network.

Since the last Volume IV was written, there are been a number of breakthroughs in instrument
development and support equipment. The new "sonic" anemometer systems have been on the market for
several years and this document will provide guidance on how best to operate those systems. In addition,
there have been advancements in digital data acquisition where the signal from the sensor or the sensor's
translator box is a purely digital signal. Support equipment,  such as data acquisition systems (DAS) have
also changed in support to these new "state of the art" instruments that are now available.

As you read through this document, please be careful to note the references in the manual, as these may
have World Wide Web Internet links associated with them.  Where possible,  the authors placed Internet
links into the document so that if the reader wishes to get more in-depth or background material, then it is
available through the link.

Another addition to this manual that is not in the other QA Handbooks is the  links to audio/video files.
This document will have Internet links in the calibration sections  of 4.2 through 4.8 that will direct the
reader to audio/video files posted on the EPA's AMTIC website.  These audio/video files are short, but
in-depth movies on how to calibrate/audit common meteorological instruments that are in use today.
                               Disclaimer of Endorsement


Mention of or referral to commercial products or services, and/or links to non-EPA sites does not imply
official EPA endorsement of or responsibility for the opinions, ideas, data, or products presented at those
locations, or guarantee the validity of the information provided. Mention of commercial products/services
on non-EPA servers is provided soley as a pointer to information on topics related to environmental
protection that may be useful to EPA staff and the public.

-------
                                           Contents
                                                                                         Volume 4
                                                                                    Re vision No: 1.0
                                                                                      Date:  10/2006
                                                                                       Page iiiofxi
Section
Page
Revision
Date
Foreword	ii

Contents	iii

Acknowledgments	vii

Figures and Tables	ix

0.  Introduction
    0.1   Contents of the Handbook	1/15
    0.2   EPA Quality System	2/15

1.  Tower Guidance and Siting
    1.1   Type of Towers	1/17
    1.2   Installation and Setup	3/17
    1.3   Tower Wiring	12/17
    1/4   Tower Siting	14/17

2.  Wind Speed and Direction
    2.1   Introduction	1/27
    2.2   Types of Instruments and Specifications	1/27
    2.3   Acceptance Testing	4/27
    2.4   Installation, Instrument Exposure, and
         Wiring	5/27
    2.5   Calibration and Alignment	6/27
    2.6   Operation and Maintenance	15/27
    2.7   Auditing	17/27
    2.8   Scalar, Vector and Sigma Calculations	24/27
    2.9   Estimating Accuracy and Precision	26/27

3.  Temperature and Temperature Gradient
    3.1   Types of Instruments and Specifications	1/5
    3.2   Acceptance Testing	2/5
    3.3   Installation and Wiring	2/5
    3.4   Calibration	3/5
    3.5   Operation and Maintenance	5/5
    3.6   Auditing	5/5

4.  Rainfall and Precipitation
    4.1   Types of Instruments and Specifications	1/6
    4.2   Acceptance Testing	4/6
    4.3   Calibration	4/6
    4.4   Operation and Maintenance	5/6
                       1.0

                       1.0

                       1.0

                       1.0

                       1.0



                       1.0
                       1.0
                       1.0
                       1.0
                                               in

-------
Section
Page
Revision
     Volume 4
Re vision No: 1.0
  Date: 10/2006
   Pageivofxi

     Date
    4.5  Auditing	6/6

5.  Relative Humidity and Dew Point
    Determination
    5.1  Types of Instruments and Specifications	1/4
    5.2  Acceptance Testing	2/4
    5.3  Installation and Wiring	2/4
    5.4  Calibration	3/4
    5.5  Operation and Maintenance	4/4
    5.6  Auditing	4/4


6.  Quality Assurance of Solar Radiation
    Measurements
    6.1  Introduction	1/9
    6.2  Solar Radiation	1/9
    6.3  Types of Instruments	2/9
    6.4  Specifications	6/9
    6.5  Acceptance Testing	6/9
    6.6  Installation, Instrument Exposure, and
        Wiring	6/9
    6.7  Calibration	7/9
    6.8  Operations and Maintenance	8/9
    6.9  Auditing	9/9

7.  Quality Assurance for Atmospheric Pressure
    Measurements
    7.1  Units and Scales	1/5
    7.2  Types of Instrumentation	1/5
    7.3  Acceptance Testing	2/5
    7.4  Installation and Instrument Exposure	3/5
    7.5  Calibration	4/5
    7.6  Operation and Maintenance	4/5
    7.7  Auditing	5/5


8.  Quality Assurance for Ground-Based Remote
    Sensing Devices
    8.1  Types of Instruments and Specifications	2/26
    8.2  Acceptance Testing	10/26
    8.3  Installation and Siting	13/26
    8.4  Calibration	18/26
    8.5  Operation, Maintenance, and Quality
        Control	19/26
    8.6  Auditing	23/26
                       1.0
                       1.0
                       1.0
                       1.0
                                                IV

-------
Section                                        Page

9.  Data Acquisition Systems and Meteorology
    9.1  Introduction	1/7
    9.2  DAS Data Acquisition in Analog
        Layout- Signal Conditioning	1/7
    9.3  Instrument Connectivity	1/7
    9.4  Data Communication	4/7
    9.5  Sampling Rates	4/7
    9.6  Meteorological Data Generated by DAS	5/7

10. Meteorological Data Validation and Verification
    10.1 General Approach	1/13
    10.2 Data Verification Methods	3/13
    10.3 Manual Review Methods	7/13
    10.4 Data Validation Methods	7/13

11. References

Appendix A:  Meteorological Systems Audit
Evaluation Form	1/5

Appendix B:  Examples of Meteorological Sensor
Calibration Forms	1/6
Revision

    1.0
     Volume 4
Re vision No: 1.0
  Date: 10/2006
   Page vofxi

     Date
    1.0
    1.0
    1.0
    1.0

-------
                                                                                   Volume 4
                                                                              Re vision No: 1.0
                                                                                Date: 10/2006
                                                                                 Page viofxi
                                  Acknowledgments

Work on a document such as this requires the work and dedication of many people. This section will
acknowledge those that have provided their time and effort to create this document.

Team Lead: Dennis K. Mikel, EPA - OAQPS, AQAD
Team Co-lead:  Nealson Watkins, EPA - OAQPS, AQAD

Authors of the Chapters:

Chapter 0: Dennis Mikel, EPA - OAQPS, AQAD, Joey Landreneau, Sonoma Technologies Inc.
Chapter 1: Daniel Fields, Maricopa County Air Pollution District
Chapter 2: David Bush, T+B Systems, Inc., Paul Roberts, Sonoma Technologies Inc.
Chapter 3: Paul Fransiola, T+B Systems Inc.
Chapter 4: Tammy Eagan, Florida Department of Environmental Protection
Chapter 5: Paul Fransiola, T+B Systems Inc.
Chapter 6: Dennis Mikel, EPA - OAQPS, AQAD, Joey Landreneau, Sonoma Technologies Inc.
Chapter 7: Kent Field, Ventura County APCD, Joey Landreneau, Sonoma Technologies Inc.
Chapter 8: Bob Baxter, T+B Systems Inc., Tim Dye, Sonoma Technologies Inc.
Chapter 9: Dennis Mikel, EPA - OAQPS, AQAD
Chapter 10: Dennis Mikel, EPA - OAQPS, AQAD

Editorial review and formatting were provided by Sonoma Technologies Inc. under contract with the
EPA.

Reviewers:

Phil Lorang, EPA,-OAQPS-AQAD
Comments and questions can be directed to:

Dennis Mikel or,
Nealson Watkins
EPA-OAQPS-AQAD
109 Alexander Drive
Research Triangle Park, NC 27711
Email: mikel.dennisk@epa.gov
      Watkins.nealson@epa.gov
                                            VI

-------
Volume 4
Re vision No: 1.0
Date: 10/2006
Page viiofxi

Number
0.1
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
Figures
Title
The EPA Quality System
Telescopic Pole
Triangular Fixed
Triangular Adjustable
Pneumatic
Ground Installation Procedure
Roof Installation Procedure
(a) Fascia Board Support; (b) Fence Post Guy Wire Anchor
(a) Angle Iron Support — Expanded View;

Section Page
3/0
2/17
2/17
3/17
3/17
4/17
5/17
7/17

         (b) Angle Iron Support—Close-Up View
1.9      (a) Metal Wall Support—Telescopic Tower (Expanded View);
         (b) Metal Wall Support—Telescopic Tower (Close-Up View)
1.10     (a) Expanded View Guy Wire Harness—Telescopic Tower;
         (b) Close-Up View Guy Wire Harness—Telescopic Tower
1.11     Lightning Protection Installation
1.12     Signal Cable at Tower Base
1.13     Temperature Cable Installation and Expanded/Close-Up Views
1.14     Wiring on a Crank-Down Tower
1.15     Example of a Tower Location
1.16     Example of a Tower Attachment
2.1      Example of Cup and Vane System
2.2      Propeller Anemometer and Vane
2.3      2-D Sonic Anemometer
2.4      3-D Sonic Anemometer
2.5      Siting Wind Instruments
2.6      DC Motor Calibration
2.7      Torque Disk
2.8      Torque Disk in Testing Position
2.9      Magnetic Declination in North America
2.10     Magnetic Declinations for the United States
2.11     Side View of the Placement of a Compass or Transit for Measuring the
         Crossarm Direction
 8/17

 9/17

10/17
11/17
12/17
13/17
14/17
16/17
17/17
 2/27
 3/27
 3/27
 3/27
 6/27
 7/27
 7/27
 7/27
 9/27
10/27

11/27
                                       vn

-------
Volume 4
Re vision No: 1.0
Date: 10/2006
Page viiiofxi
Number
2.12
2.13
2.14
2.15
2.16
2.17
2.18
2.19
2.20
3.1
3.2
3.3
4.1
4.2
4.3
4.4
4.5
5.1
5.2
6.1
6.2
6.3
6.4
6.5
6.6
6.7
6.8
6.9
7.1
Title
Top View Illustrating the Measurement of the Relative Direction of the
Crossarm
Measurement of the Relative Direction of the Sun
Projection of Sun's Reflection Using a Brunton Pocket Transit
Linearity Test Fixture
Sensor Mounting for the Testing and Evaluation of the Audit Wind
Sensor Against a Sonic Anemometer
Wind Speed Plot Showing the Mechanical Sensor (AQ) vs. the
Sonic Anemometer for Wind Speeds Greater Than 1 ms"1
Wind Direction Plot Showing the Mechanical Sensor (AQ) vs. the Sonic
Anemometer for Wind Speed Greater Than 1 ms"1
Typical Mounting of the Audit Sensors on the Site Tower
Calculated Wind Direction Averages from a Simulated 3,600-Sample
Data Set
Example of a platinum wire thermistor
A motor aspirated shield
A naturally ventilated shield
A Typical Non-recording Rain Gauge
Automatic Wet/Dry Precipitation Collector
Universal-Weighing Gauge
Tipping Bucket without the Shield
Example of an Alter wind shield
A motor aspirated RH shield
A typical RH probe
Solar Irradiance Versus Wavelength of Light Emitted by the Sun
Eppley Pyranometer PSP
Illustration of an Epply Pyranometer
LiCor Pyranometer
NovaLynx Pyranometer
Epply Normal Incidence Pyrheliometer
Epply Solar Tracker
Net Radiometer
Absolute Cavity Pyrheliometer
Electronic Barometer
Section Page
12/27
12/27
13/27
20/27
22/27
22/27
22/27
23/27
25/27
1/5
3/5
3/5
1/6
1/6
2/6
2/6
3/6
1/4
1/4
2/9
3/9
3/9
5/9
5/9
5/9
5/9
7/9
8/9
2/5
Vlll

-------
Volume 4
Re vision No: 1.0
Date: 10/2006
Pageixofxi
7.2
7.3
7.4
8.1
8.2
Number
8.3
8.4
8.5
8.6
8.7
8.8
8.9
8.10
9.1
9.2
10.1
10.2
10.3
10.4
10.5
Portable Digital Barometer
NEMA 4X Enclosure
Tower-Mounted Barometer with Pressure Port
Schematic Showing the Differences Between In-Situ (Point) and Remote Senor
(Volume) Measurements
Schematic Showing the Transmitted and Received Signals From Sodars,
Radar Wind Profilers, and RASS
Title Section
Schematic Showing the Vertical and Oblique Beams
Schematic Showing a Beam Pattern for an Oblique Beam and its
Associated Side Lobes
Schematic Showing a Monostatic and Bistatic Sodar System
Pictures of Different Types of Sodars: (a) Mini-Sodar, (b) Multi-Axis Sodar,
(c) Phased-Array Sodar
Photographs of Several Types of Radar Wind Profilers:
(a) Phased-Array System and (b) Fixed-Axis Antenna System
Photographs of (a) a Radar Wind Profiler with a RASS and
(b) a Sodar with a RASS
Example Site Layout Diagram
Example Site Vista Diagram
DAS Rear Panel with 8-Channel Differential Analog Terminal Strip
DAS Rear Panel with RS-232 Signal Interface
Generalized Data Validation and Verification Process Flow
Example of Meteorological Sensor Visual Check List
Graphic Example of Temperature vs. Relative Humidity
Graphic Example of Ozone vs. Temperature
Example of a Meteorological Data Screening Checklist
2/5
3/5
4/5
2/26
3/26
Page
4/26
5/26
6/26
7/26
8/26
10/26
14/26
15/26
3/7
4/7
2/13
5/13
9/13
9/13
13/13

-------
Volume 4
Re vision No: 1.0
Date: 10/2006
Pagexofxi

Number
0-1
0-2
0-3
0-4
0-5
0-6
0-7
0-8
0-9
0-10
1-1
1-2
2-1
2-2
8-1
8-2
8-3
10-1
Tables
Title
PAMS Meteorological Measurement Quality Objectives
PAMS Calibration and Accuracy Criteria
NCore Meteorological Measurement Quality Objectives
NCore Calibration and Accuracy Criteria
PSD Measurement Quality Objectives
PSD Calibration and Accuracy Criteria
Modeling Application Measurement Quality Objectives
Modeling Application Calibration and Accuracy Criteria
National Weather Service Measurements Quality Objectives
Siting and Exposure for Meteorological Sensors
Description of Different Towers
Limits on Terrain and Obstacles Near Towers
Time Constant Effects
Proposed Audit Criteria for the Sonic Systems
Typical Specifications for Meteorological Remote Sensors
Characteristics of Radar Wind Profilers
Recommended Audit Criteria for Sodar, Radar Wind Profilers,
DAS Screening Techniques

Section Page
6/15
7/15
8/15
9/15
10/15
11/15
12/15
13/15
14/15
15/15
1/17
15/17
19/27
23/27
3/26
8/26
and RASS 23/26
6/13

-------
                                                                                Volume 4, Section 0
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
                                                                                      Page 1 of 15
                                   0.    Introduction
The purpose of the Quality Assurance Handbook for Air Pollution Measurement Systems, VolIV:
Meteorological Measurements (hereinafter called Handbook) is to provide information and guidance for
meteorologists (applied and research), State/Local/Tribal (SLT) air pollution agency staff who operate
meteorological equipment, SLT data reviewers who need validation guidance, and users of
meteorological data.  Methods that objectively define the quality of measurements needed for the
intended use of the data are described in this version of the Handbook.

This version of the Handbook follows two previous versions that, for their time, were groundbreaking
documents and paved the way for the air pollution monitoring community to begin and to continue
collecting valid meteorological data. The first version of this Handbook1 was published in 1983, and it
described the different meteorological systems that were available at that time.  An updated version2 was
published in 1990, with revisions and updates added in 1995. The second version discussed methods and
the equipment used within those methods in much greater detail as well as the calculation of vector and
sigma data in detail, information that was lacking in the first version. Later revisions included a section
on upper-air measurements and an appendix on Photochemical Assessment Monitoring Stations (PAMS)
meteorological guidance. However, newer technologies have been developed since the second edition of
the Handbook was published. This third version of the Handbook has a slightly different focus than the
second version. This document is intended to be more "user friendly"—it will have as much practical
information for those not trained to be meteorologists as those who are. This Handbook will discuss the
practical uses and operation of meteorological equipment and data.  Some of the very technical
information has been removed or clarified, and illustrations have been updated. Internet links and
references have been added throughout the document to allow the reader to research information quickly.
0.1 Contents of the Handbook

The first section of this Handbook describes the U.S. Environmental Protection Agency's (EPA) Quality
System (QS) and how it can be used to create a data collection system that gathers data of sufficient
quality for its intended use.  The tables of Measurement Quality Objectives (MQOs) in Section 0.2.2 will
be useful to organizations planning meteorological monitoring programs.  The tables will help users to
quickly review the requirements of a particular program (i.e., PAMS versus NCore). If an agency is
required to perform a particular type of monitoring, such as PAMS, Tables 0-1 and 0-9 clearly list all the
MQOs and calibration and accuracy criteria so that the agency can make the right choices when
purchasing or upgrading its equipment for a specific program.

   >   Section 1 focuses on meteorological towers, on which most equipment is mounted.  The different
       types of towers and their application, including installation, setup, wiring, and lightning
       protection, are discussed in  this section.
   >   Section 2 discusses wind speed and direction.
   >   Section 3 details temperature and temperature gradient.
   >   Section 4 deals with rainfall and precipitation.
   >   Section 5 illustrates relative humidity and dew point determinations.
   >   Section 6 discusses solar and total radiation.

   >   Section 7 discusses atmospheric pressure measurement.

-------
                                                                                Volume 4, Section 0
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
	Page 2 of 15
   >    Section 8 describes upper-air systems, which include radar wind profiler (RWP), Sodar (Sound
        Detection and Ranging) and RASS (Radio Acoustic Sounding Systems).

   >    Section 9 provides guidance on the advantages and disadvantages of analog and digital data
        acquisition.

   >    Section 10 provides a discussion of data validation and verification.

Sections 2 through 8 describe the types of instruments currently available; acceptance testing, installation
and wiring; calibration and alignment; operation and maintenance; and auditing. Sections 9 and 10 are
new to this version of the Handbook and describe meteorological data acquisition systems (DAS) and
meteorological data validation and verification.
0.2 EPA Quality System

The EPA document, "Guidance for the Data Quality Objective Process"3 states, "EPA Order 5360.1 A2
and the applicable Federal regulations establish mandatory QS that applies to all EPA organizations and
organizations funded by EPA." The guidance document describes the requirements, logic, and reasoning
for establishing a QS: "Organizations must ensure that data collected for the characterization of
environmental processes and conditions are of the appropriate type and quality for their intended use and
which environmental technologies are designed, constructed, and operated according to defined
expectations." Systematic planning is a key project-level component of the EPA QS. Components of the
QS are shown in Figure 0.1.

EPA policy is based on the national consensus standard, ANSI/ASQC E4-1994, Specifications and
Guidelines for Environmental Data Collection and Environmental Technology Programs, developed by
the American National Standards Institute and the American Society for Quality. This document
describes the necessary management and technical elements for developing and implementing a QS by
using a tiered approach.  The standard recommends documenting (1) each organization-wide QS in a
Quality Management Plan (QMP) or Quality Manual (to address requirements of Part A: Management
Systems of the standard) and (2) the applicability of the QS to technical activity-specific efforts in a
Quality Assurance Project Plan (QAPP) or similar document (to address the requirements  of Part B:
Collection and Evaluation of Environmental Data of the standard). EPA has adopted this  tiered approach
for its mandatory agency-wide QS. This document addresses Part B requirements of the standard for
systematic planning for environmental data operations.

In accordance with EPA Order 5360.1 A2, EPA requires that environmental programs performed for or
by the Agency be supported by data of the type and quality appropriate to their expected use.  EPA
defines environmental data as information collected directly from measurements, produced from models,
or compiled from other sources such as databases or literature.
0.2.1 Data Quality Objectives

As stated in Section 0.2, EPA Order 5360.1 A2 requires that all EPA organizations (and organizations
with extramural agreements with EPA) follow a systematic planning process to develop acceptance or
performance criteria for the collection, evaluation, or use of environmental data. A systematic planning
process is the first component in the planning phase of the project tier (see the bottom tier of Figure 0.1),
while the actual data collection activities take place in the implementation phase.

-------
                                                                           Volume 4, Section 0
                                                                              Revision No: 1.0
                                                                                Date: 10/2006
                                                                                 Page 3 of 15
                                          EPAQAPolicy
                                              a
                                          Prog-am Policy
                                 Program Tier
luppnrlng Ifikm Biminti
   f q., P DEinmiifc,
Can p i^ r Hi HIE E aofho E)
TnlnlngCnm murteitDn
  fg., Trail lig Pbi,
   CBItFEIEEt)
Monitoring Org.
Overall
Quality System
                                               Project Tier
                            Conduct Study/X,
                                     /
     PLANNING
                          IMPLEMEhTTATION
                            Reporting
                                                   ASSESSMENT
                                                                Monitoring Org.
                                                                Project Specific
                                                                Quality System
     Defensible Products  and Decisions
  Figure 0.1 The EPA Quality System

-------
                                                                                 Volume 4, Section 0
                                                                                    Revision No: 1.0
                                                                                      Date: 10/2006
                                                                                       Page 4 of 15

Systematic planning is a planning process based on the scientific method and includes concepts such as
objectivity of approach and acceptability of results.  Systematic planning is a common-sense, graded
approach to ensure that the level of detail in planning is commensurate with the importance and intended
use of the work and available resources.  This framework promotes communication among all
organizations and individuals involved in an environmental program.  Through a systematic planning
process, a team can develop acceptance or performance criteria for the quality of the data collected and
for the quality of the decision. When these data are being used in decision making by selecting between
two clear alternative conditions (e.g., compliance/non-compliance with a standard), the EPA's
recommended systematic planning tool is called the Data Quality Objective Process (DQO Process).

The DQO Process is a seven-step planning approach to develop sampling designs for data collection
activities that support decision making. This process uses systematic planning and statistical hypothesis
testing to differentiate between two or more clearly defined alternatives.
    Step 1. Define the problem
    Step 2. Identify the problem
    Step 3. Identify information needed for the decision

    Step 4. Define the boundaries of the study
    Step 5. Develop a decision rule
    Step 6. Specify limits on decision  errors
    Step 7. Optimize the design for obtaining data

The DQO Process is iterative and allows the planning team to incorporate new information and modify
outputs from previous steps as inputs for a subsequent step. Although the principles of systematic
planning and the DQO Process are applicable to all scientific studies, the DQO Process is particularly
designed to address problems that require making a decision between two clear alternatives.  The final
outcome of the DQO Process is a design for collecting data (e.g., the number of samples to collect, and
when, where, and how to collect samples).

The development and implementation of a quality system should be based on a "graded approach," that is,
the components and tools of a quality system (Figure 0.1) apply according to the scope and nature of an
organization, program, or project and the intended use of its products or services.  This approach
recognizes that a "one size fits all" approach to  quality management is not appropriate and that the quality
system of different organizations and programs  should (and will) vary according to the specific needs of
the organization.  For example, the quality expectations of a research program are different from those of
a regulatory compliance program because the intended use of the products differs.  The same applies to
meteorological data. Meteorological data can be used for a variety of reasons. The EPA has set forth a
number of regulatory programs to understand the effects that pollution has on the health of the population.
These programs include PAMS,  National Core (NCore), Prevention of Significant Deterioration (PSD),
and comparison of pollution data to the National Ambient Air Quality Standards (NAAQS).  Monitoring
agencies that use this guidance document are strongly encouraged to understand their data objectives,
perform the DQO Process, if needed, and use the Measurement Quality Objectives (MQOs) described in
Section 0.2.2 if they are applicable to an agency's program.

When an agency or entity is monitoring for non-regulatory purposes (e.g., background concentrations,
modeling applications, or exposure), these MQOs are recommended guidance. Meeting MQOs for non-
regulatory meteorological monitoring is strongly advised.

-------
                                                                               Volume 4, Section 0
                                                                                 Revision No: 1.0
                                                                                   Date: 10/2006
                                                                                    Page 5 of 15
0.2.2 Measurement Quality Objectives

Once DQOs are designated for a program or project, measurement indicators must be determined to
understand if the DQOs are being met. Most SLT agencies that collect data do so to support programs,
such as PAMS, NCore, or PSD, that are federally mandated or that need to meet federal requirements.
However, other non-regulatory applications exist, such as modeling applications, state implementation
plan development, and forecasting. These programs require different meteorological MQOs since the
application is different (i.e., different DQOs).  Many SLT agencies also utilize National Weather Service
(NWS) data available from nearby airports. The MQOs of the NWS data are included in Table 0-9.
Agencies that use NWS data can reference these MQOs.  The following prescribed objectives are listed in
the MQO tables:
   >   Measurement - Type of parameter needed to be collected
   >   Method - Recommended; however, other methods are available and newer technologies will be
       developed. For regulatory programs, any alternative methods that are employed must meet these
       minimum requirements or recommendations.
   >   Reporting Units - Mandatory for regulatory programs (PAMS, PSD, and NCore); for non-
       regulatory programs, they are recommended only.
   >   Recommended Operating Range(s) - Typically employed but are not required.
   >   Detection Limits - Required for regulatory programs (PAMS, PSD, and NCore); the detection
       limit levels are mandatory. For non-regulatory programs, they are recommended only.
   >   Minimum Sample Frequency - Recommended.
   >   Raw Data Collection Frequency - Recommended.
   >   Completeness - Required for regulatory programs (PAMS, PSD, and NCore); the levels of
       completeness are mandatory.  For non-regulatory programs, they are recommended only.
   >   Calibration - Recommended methods. Other methods may be employed; however, the levels of
       accuracy of other methods must meet the acceptance criteria.
   >   Accuracy - Required for regulatory programs (PAMS, PSD, and NCore); the levels of accuracy
       are mandatory.  For non-regulatory programs, they are recommended only.

   >   Representativeness - Table 0-10 summarizes the  recommended siting and exposure for most
       meteorological instruments.

-------
                                                                                                                          Volume 4, Section 0
                                                                                                                             Re vision No: 0
                                                                                                                              Date: 10/2006
                                                                                                                               Page 6 of 15
Table 0-1 PAMS Meteorological Measurement Quality Objectives
Measurement
Ambient Temperature
Relative Humidity
Barometric Pressure
Wind Speed
Wind Direction
Solar Radiation
UV Radiation
Precipitation
Upper-Air Meteorology
Temp
Relative Humidity
Direction
Wind speed
Method
Thermistor
Psychrometer/
Hygrometer
Aneroid
Barometer
Cup, Blade or
sonic anemometer
Vane or sonic
anemometer
Pyranometer
UV A and B
Radiometer
Tipping Bucket
RWP, Sodar,
RASS
Reporting
Units
°C
%
hPa
m/s
Degrees
Watts/m2
Watts/m2
mm/hr
°C
%
Degrees
m/s
Operating
Range
-20 to 40
0-100
800-1100
0.5-50.0
0-360
0 - 1200
0-12
0-30
0-35
0-100
0-360
0-45
Resolution
0.1
0.5% RH
0.1
0.1
1.0
1.0
0.01
0.25
0.2
5.0
10
0.5
Minimum
Sample
Frequency
Hourly
Hourly
Hourly
Hourly
Hourly
Hourly
Hourly
Hourly
4 profiles
per day
Raw Data
Collection
Frequency
1 minute
1 minute
1 minute
1 minute
1 minute
1 minute
1 minute
1 minute
Per sounding
Completeness
75%
75%
75%
75%
75%
75%
75%
75%
75%

-------
                                                                                                                             Volume 4, Section 0
                                                                                                                                 Re vision No: 0
                                                                                                                                  Date: 10/2006
                                                                                                                                   Page 7 of 15
Table 0-2 PAMS Calibration and Accuracy Criteria
Measurement
Ambient Temperature
Relative Humidity
Barometric Pressure
Wind Speed
Wind Direction
Solar Radiation
UV Radiation
Precipitation
Upper-Air Meteorology
Temp
Relative Humidity
Direction
Wind speed
Calibration
Type
3 pt. Water Bath with
NIST-traceable
thermistor or
thermometer
NIST-traceable
Psychrometer or
standard solutions
NIST-traceable
Aneroid Barometer
NIST-traceable
Synchronous Motor
Solar Noon, GPS or
Magnetic Compass
NIST-traceable
Pyranometer
NIST-traceable
Radiometer
Separatory runnel and
graduated cylinder
Tethered or Balloon
Sonde with NIST-
traceable Sensors
Acceptance
Criteria
±0.5 °C
±3% RH
±lhPa
±0.2 m/s +
5%
±5 degrees
±5%
±5%
±10% of
input
volume
0.2 °C
5%
10 degrees
1 m/s
Frequency
Semi-Annually
Semi-Annually
Semi-Annually
Semi-Annually
Semi-Annually
Semi-Annually
Semi-Annually
Semi-Annually
Semi-Annually
Accuracy
Type
3 pt. Water Bath With
NIST-traceable
thermistor or
Thermometer
NIST-traceable
Psychrometer or
standard solutions
NIST-traceable Aneroid
Barometer
NIST-traceable
Synchronous Motor
Solar Noon, GPS, or
Magnetic Compass
NIST-traceable
Pyranometer
NIST-traceable
Pyranometer
Separatory runnel and
graduated cylinder
Tethered or Balloon
Sonde with NIST-
traceable Sensors
Acceptance
Criteria
±0.5 °C
±3% RH
±lhPa
0.2 m/s +
5%
±5 degrees
±5%
±5%
±10% of
input
volume
0.2 °C
5%
10 degrees
1 m/s
Frequency
Annually
Annually
Annually
Annually
Annually
Annually
Annually
Annually
Annually

-------
                                                                                                                          Volume 4, Section 0
                                                                                                                              Re vision No: 0
                                                                                                                              Date: 10/2006
                                                                                                                                Page 8 of 15
Table 0-3 NCore Meteorological Measurement Quality Objectives
Measurement
(Required)
Ambient
Temperature
Relative Humidity
Wind Speed
Wind Direction
Vector Data
Wind Speed
Wind Direction
(Optional)
Solar Radiation
Precipitation
Barometric Pressure
Method

Thermistor
Psychrometer/
Hygrometer
Cup, prop or sonic
anemometer
Vane or sonic
anemometer
DAS Calculations

Pyranometer
Tipping Bucket
Aneroid Barometer
Reporting
Units

°C
%
m/s
Degrees
m/s
degrees

Watts/m2
mm/hr
mb
Operating
Range

-30-50
0- 100
0.5-50.0
0 - 360 (540)
0-50.0
0-360

0- 1100
0-25 mm/hr
600-1100
Resolution

0.1
0.5
0.1
1.0
0.1
1.0

10
0.2 mm
0.5
Minimum
Sample
Frequency

Hourly
Hourly
Hourly
Hourly
Hourly

Hourly
Hourly
Hourly
Raw Data
Collection
Frequency

1 minute
1 minute
1 minute
1 minute
1 minute
1 minute

1 minute
1 minute
1 minute
Completeness

75%
75%
75%
75%
75%

75%
75%
75%

-------
                                                                                                                              Volume 4, Section 0
                                                                                                                                  Re vision No: 0
                                                                                                                                  Date: 10/2006
                                                                                                                                    Page 9 of 15
Table 0-4 NCore Calibration and Accuracy Criteria
Measurement
Ambient
Temperature
Relative Humidity
Wind Speed
Wind Direction
Solar Radiation
Barometric
Pressure
Precipitation
Calibration
Type
3 pt. Water Bath
with NIST-
traceable
thermistor or
thermometer
NIST-traceable
Psychrometer or
standards solution
NIST-traceable
Synchronous
Motor
Solar Noon, GPS
Magnetic
Compass
NIST-traceable
Pyranometer
NIST-traceable
Aneroid
Barometer
Separatory runnel
and graduated
cylinder
Acceptance
Criteria
±0.5 °C
±7% RH
±0.25m/s <5m/s;
5%>2m/snotto
exceed 2.5m/s
±5 degrees;
includes
orientation error
±5% of mean
observed interval
±3mb
±10% of input
volume
Frequency
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Accuracy
Type
3 pt. Water Bath
With NIST-
traceable
thermistor or
Thermometer
NIST-traceable
Psychrometer or
standards
solution
NIST-traceable
Synchronous
Motor
Solar Noon,
GPS or
Magnetic
Compass
NIST-traceable
Pyranometer
NIST-traceable
Aneroid
Barometer
Separatory
funnel and
graduated
cylinder
Acceptance
Criteria
±0.5 °C
±7% RH
0.25m/s <5m/s;
5%>2m/snotto
exceed 2.5m/s
±5 degrees;
includes
orientation error
±5% of mean
observed interval
±3mb
±10% of input
volume
Frequency
Semi-
Annually
Semi-
Annually
Semi-
Annually
Semi-
Annually
Semi-
Annually
Semi-
Annually
Semi-
Annually

-------
                                                                                                                          Volume 4, Section 0
                                                                                                                              Re vision No: 0
                                                                                                                               Date: 10/2006
                                                                                                                               Page 10 of 15
Table 0-5 PSD Measurement Quality Objectives4
Measurement
Ambient Temperature
Vertical Temperature
Difference
Relative Humidity
Dew Point
Barometric Pressure
Wind Speed
Wind Direction
Solar Radiation
Vertical Wind Speed
Vector Data
Wind Speed
Wind Direction
Sigma Theta
Sigma W
Precipitation
Method
Thermistor
Thermistor
Psychrometer/
Hygrometer
Psychrometer/
Hygrometer
Aneroid Barometer
Cup or sonic
anemometer
Vane or sonic
anemometer
Pyranometer
Vane or sonic
anemometer
DAS Calculations
Tipping Bucket
Reporting
Units
°C
°C
%
°c
Mb
m/s
Degrees
Watts/m2
m/s
m/s
degrees
degrees
m/s
mm/hr
Operating
Range
-30-50
-3-7
0- 100
-30 - +30
600-1100
0.5-50.0
0 - 360 (540)
0- 1300
-25.0 to 25.0
0-50.0
0-360
0- 105
0-10
0-50 mm /hr
Resolution
0.1
0.1
0.5
0.1
0.5
0.25
1.0
10
0.1
0.1
1.0
1.0
0.1
0.2 mm/hr
Minimum
Sample
Frequency
Hourly
Hourly
Hourly
Hourly
Hourly
Hourly
Hourly
Hourly
Hourly
Hourly
Hourly
Raw Data
Collection
Frequency
1 minute
1 minute
1 minute
1 minute
1 minute
1 minute
1 minute
1 minute
1 minute
1 minute
1 minute
15 minute
1 minute
5 minute
Completeness
90%
90%
90%
90%
90%
90%
90%
90%
90%
90%
90%

-------
                                                                                                                             Volume 4, Section 0
                                                                                                                                 Re vision No: 0
                                                                                                                                 Date: 10/2006
                                                                                                                                  Page 11 of 15
Table 0-6 PSD Calibration and Accuracy Criteria
Measurement
Ambient
Temperature
Vertical Temperature
Difference
Relative Humidity
Dew Point
Barometric Pressure
Wind Speed
Wind Direction
Solar Radiation
Vertical Wind Speed
Vector Data
Wind Speed
Wind Direction
Sigma Theta
Sigma W
Precipitation
Calibration
Type
3 pt. Water Bath with NIST-
traceable thermistor or
thermometer
3 pt. Water Bath with NIST-
traceable thermistor or
thermometer
NIST-traceable
Psychrometer or standard
solutions
NIST-traceable
Psychrometer or standard
solutions
NIST-traceable Aneroid
Barometer
NIST-traceable
Synchronous Motor
Solar Noon, GPS Magnetic
Compass
NIST-traceable
Pyranometer
NIST-traceable
Synchronous Motor
Voltmeter and Voltage
Generator
Separatory funnel and
graduated cylinder
Acceptance
Criteria
±0.5 °C
±O.PC
±7%RH
±1.5°C
±3 mb
±0.2 m/s
±5 degrees
includes
orientation error
5% of mean
observed
interval
±0.2 m/s
±0.2 m/s
±5 degrees
±5 degrees
±0.2 m/s
±10% of input
volume
Frequency
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Accuracy
Type
3 pt. Water Bath With
NIST-traceable thermistor
or Thermometer
3 pt. Water Bath with NIST-
traceable thermistor or
thermometer
NIST-traceable
Psychrometer or standard
solutions
NIST-traceable
Psychrometer or standard
solutions
NIST-traceable Aneroid
Barometer
NIST-traceable
Synchronous Motor
Solar Noon, GPS, or
Magnetic Compass
NIST-traceable
Pyranometer
NIST-traceable
Synchronous Motor
Voltmeter and Voltage
Generator
Separatory funnel and
graduated cylinder
Acceptance
Criteria
±0.5 deg. C
±O.PC
±7% RH
±1.5°C
±3mb
±0.2 m/s
±5 degrees
includes
orientation
error
5% of mean
observed
interval
±0.2 m/s
±0.2 m/s
±5 degrees
±5 degrees
±0.2 m/s
±10% of
input volume
Frequency
Within 60 days of
startup and 6 month
intervals
Within 60 days of
startup and 6 month
intervals
Within 60 days of
startup and 6 month
intervals
Within 60 days of
startup and 6 month
intervals
Within 60 days of
startup and 6 month
intervals
Within 60 days of
startup and 6 month
intervals
Within 60 days of
startup and 6 months
thereafter
Within 60 days of
startup and 6 month
intervals
Within 60 days of
startup and 6 month
intervals
Within 60 days of
startup and 6 month
intervals
Within 60 days of
startup and 6 month
intervals

-------
                                                                                                                           Volume 4, Section 0
                                                                                                                               Re vision No: 0
                                                                                                                                Date: 10/2006
                                                                                                                                Page 12 of 15
Table 0-7 Modeling Application Measurement Quality Objectives5
Measurement
Ambient
Temperature
Dew Point
Temperature
Vertical
Temperature Diff.
Barometric Pressure
Wind Speed
Wind Direction
Solar Radiation
Upper-Air
Meteorology
Temp
Direction
Wind speed
Vector Data
Wind Speed
Wind Direction
Sigma Theta
Sigma W
Precipitation
Method
Thermistor
Psychrometer/
Hygrometer
Thermistor
Aneroid Barometer
Cup, blade, or sonic
anemometer
Vane or sonic
anemometer
Pyranometer
RWP, Sodar, RASS
DAS Calculations
Tipping Bucket
Reporting
Units
°C
°C
°c
mb
m/sec
Degrees
Watts/m2
°C
degrees
m/sec
m/s
degrees
degrees
m/s
mm/hr
Operating
Range
-40 - 60
-40 - 60
-5- 15
600-1100
0.5-50.0
0 - 360 (540)
0-1300
0-35
0-360
0-45
0-50.0
0-360
0-105
0-10
0-25 mm/hr
Resolution
0.1
0.1
0.02
0.5
0.1
0.5
10
1.0
10
1.0
0.1
1.0
1.0
0.1
0.25 mm
Minimum
Sample
Frequency
Hourly
Hourly
Hourly
Hourly
Hourly
Hourly
Hourly
15 minute/
Hourly
Soundings
Hourly
Hourly
Raw Data
Collection
Frequency
1 minute
1 minute
1 minute
1 minute
1 minute
1 minute
1 minute
15 minute
1 minute
1 minute
Completeness
90%
90%
90%
90%
90%
90%
90%
90%
90%
90%

-------
Table 0-8 Modeling Application Calibration and Accuracy Criteria
                                                                                                                             Volume 4, Section 0
                                                                                                                                 Re vision No: 0
                                                                                                                                  Date: 10/2006
                                                                                                                                  Page 13 of 15
Measurement
Ambient
Temperature
Dew Point
Temperature
Vertical Temp.
Diff.
Precipitation
Wind Speed
Wind Direction
Solar Radiation
Vertical Wind
Speed
Upper-Air
Meteorology
Temp
Direction
Wind speed
Vector Data
Wind Speed
Wind Direction
Sigma Theta
Sigma W
Calibration
Type
3 pt. Water Bath with NIST-
traceable thermistor or
thermometer
NIST-traceable Psychrometer
or standards solution
3pt. Water Bath with NIST-
traceable thermistor or
thermometer
Separatory runnel and
graduated cylinder
NIST-traceable Synchronous
Motor
Solar Noon, GPS, Magnetic
Compass
NIST traceable Pyranometer
NIST-traceable Synchronous
Motor
Tethered or Balloon Sonde
with NIST-traceable Sensors
Voltmeter and Voltage
Generator
Acceptance
Criteria
±0.5 °C
±0.5 °C
±0.1 °C
±10% of
input
volume
±0.2 m/s
±3-5
degrees
±5% of
mean
observed
interval
±0.2 m/s
±0.2 °C
±10 degrees
±lm/s
±0.2 m/s
±5 degrees
±5 degrees
±0.2 m/s
Frequency
Semi-
Annually
Semi-
Annually
Semi-
Annually
Semi-
Annually
Semi-
Annually
Semi
Annually
Semi-
Annually
Semi-
Annually
Semi-
Annually
Semi-
Annually
Accuracy
Type
3 pt. Water Bath With
NIST-traceable
thermistor or
Thermometer
NIST-traceable
Psychrometer or
standards solution
NIST-traceable
Psychrometer
Separatory funnel and
graduated cylinder
NIST-traceable
Synchronous Motor
Solar Noon, GPS or
Magnetic Compass
NIST-traceable
Pyranometer
NIST-traceable
Synchronous Motor
Tethered or Balloon
Sonde with NIST-
traceable Sensors
Voltmeter and Voltage
Generator
Acceptance
Criteria
±0.5 °C
±0.5 °C
±0.1 °C
±10% of
input volume
±0.2 m/s
±3-5
degrees
±5% of mean
observed
interval
±0.2 m/s
±0.2 °C
±10 degrees
±lm/s
±0.2 m/s
±5 degrees
±5 degrees
±0.2 m/s
Frequency
Annually
Annually
Annually
Annually
Annually
Annually
Annually
Annually
Annually
Annually

-------
                                                                                                                           Volume 4, Section 0
                                                                                                                               Re vision No: 0
                                                                                                                                Date: 10/2006
                                                                                                                                Page 14 of 15
Table 0-9 National Weather Service Measurements Quality Objectives6
Measurement
Ambient
temperature
Dew point
temperature
Barometric
pressure
Wind speed and
character
Wind direction
Sunshine sensor
Cloud height
Precipitation
(liquid)
Altimeter
Method
Liquid in glass
or electronic
Psychrometer
Barometer
Cup or sonic
anemometer
Vane or sonic
anemometer
Pyranometer
Ceilometer
Tipping bucket
Altimeter
Reporting
Units
°C
°C
inHg
knots
Degrees
Watts/m2
feet
in/hr
inHg
Operating Range
-62 to -50
-50 to +50
+50 to +54
-34 to -24
-24 to -01
-01 to +30
22-35
2 to 90
0-360
0 - 1200
0- 12,000
0- 10
22-35
Resolution
0.
0.
0.
0.
0.
0.
0.005
1.0
10
1 minute
Height of Cloud Base
0.01
0.01
Accuracy
±1.1
±0.6
±1.1
±2.2
±1.7
±1.1
±0.02
±lupto 10
±10% above 10
±5 when speed is > 5 knots
±10%
±3%
±0.02 or 4% of hourly amount
±0.02

-------
                                                                                                                          Volume 4, Section 0
                                                                                                                              Re vision No: 0
                                                                                                                               Date: 10/2006
                                                                                                                               Page 15 of 15
Table 0.10 Siting and Exposure for Meteorological Sensors8
Measurement
Wind
Speed/Direction
Temperature/Dew
Point
Vertical
Temperature
Difference
Solar Radiation
Barometric
pressure
Precipitation
Distance from
Obstruction
1 Ox the height
of the
obstruction
1.5x the tower
diameter
l.Sxthe tower
diameter
2 meters
1 meter
2x to 4x the
obstruction
height
Distance Above
Ground
10 meters
1.25 to 2 meters
2 meters and 10
meters
2 to 10 meters
1 to 10 meters
30 cm, minimum
Recommended
Group Cover
Grass or gravel
Non-irrigated or
un-watered short
grass, or natural
earth
Non-irrigated or
un-watered short
grass, or natural
earth
No requirements
No requirements
Natural
vegetation or
gravel
Comments
The standard exposure of the wind instruments over
level, open terrain is 10 meters above ground11
The surface should not be concrete or asphalt or oil-
soaked. Reflection from these surfaces may affect
the performance of the sensor.
The surface should not be concrete or asphalt or oil-
soaked. Reflection from these surfaces may affect
the performance of the sensor.
Sensor should be free from obstructions above the
plain of the sensor. It should be located so that
shadows will not cast on the device.
The location should have uniform, constant
temperature, shielded from the sun, away from drafts
or heaters
Asphalt or concrete should be avoided to avoid
splashing the gage. The gage should be high enough
to avoid it being covered by snow.
"Note:  More details on siting and exposure are available in the individual sections of this Handbook. Please see the installation section
 of each chapter for more information.

-------
                                                                               Volume 4, Section 1
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
	Page 1 of 18


                          1.    Tower Guidance and Siting

Meteorological towers house various types of meteorological equipment. There are many types of towers
and several ways to install them. Proper installation and siting of towers determine the effectiveness of a
system. Additionally, proper installation and siting ensure ease of maintenance and reliability. This
section focuses on stationary meteorological towers that are installed at permanent air monitoring sites.
The different towers and the methods used to install them are described in the following sections.
 1.1 Types of Towers

Numerous types of towers may be used to host a system. The type of tower that should be used will be
determined by the location, the type of support structures available, and the desired height of the tower.
In most cases, to accommodate wind speed/wind direction sensors, a meteorological tower must be able
to reach a height of 10 m. Therefore, if a tower will be mounted on an existing structure, the structure
should be measured first to determine what height the tower should be.  Once the tower height has been
determined, the tower type may be identified.  The most common types of towers are listed in Table 1-1.
Figures 1.1 through 1.4 illustrate tower examples.

Table 1-1 Description of Different Towers
Type of Tower
Telescopic pole
Triangular fixed
Triangular adjustable (crank-up
towers)
Pneumatic
Advantages
Inexpensive, good for
applications of 10 m or less
Simple to wire, easy to align,
mid-price range
Simple to wire, easy to align,
easy to raise and lower
Easy to raise and lower, good for
mobile applications
Disadvantages
Difficult to install, hard to align,
hard to raise and lower
Hard to raise and lower
Expensive, does not work well
with delta-t systems
Very expensive, not practical for
stationary monitoring sites

-------
                                                                        Volume 4, Section 1
                                                                           Revision No: 1.0
                                                                             Date: 10/2006
                                                                               Page 2 of 18
                    20.1 m when
                    fully extended
      2.4m
                      This tower
                      comes with
                      three sections
                      but is also
                      configurable to
                      use only one or
                      two sections
                                                                 10'
Figure 1.1 Telescopic Pole7
Figure 1.2 Triangular Fixed

-------
                                                                                Volume 4, Section 1
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
                                                                                      Page 3 of 18
                 (Top view)

                   (Hand crank)
                                                                       -Retracted
                                                                  Compressor used
                                                                  to extend mast
                                                                          i
                                                                    (Hand pumps may
                                                                    also be used)
                                                                     Coiled wire conduit
           Figure 1.3 Triangular Adjustable8       Figure 1.4 Pneumatic9
1.2  Installation and Setup

Installing and setting up a meteorological tower will vary greatly depending on the location of the tower.
Some locations allow a tower to be anchored into the ground, while others do not.  For example, if a
tower is needed on the roof of a building, anchoring the tower by digging a hole and pouring a concrete
base is not possible. Because there are so many installation scenarios, only the most common will be
discussed in this document. It is recommended that all installation scenarios be considered before
choosing a particular method. These scenarios are described in Sections 1.2.1 through 1.2.3.

-------
                                                                                 Volume 4, Section 1
                                                                                   Revision No: 1.0
                                                                                      Date: 10/2006
                                                                                       Page 4 of 18
In addition to location variations, climate plays a large role in the installation process.  For example, in a
location that experiences freezing rain, extra support devices may be used to ensure the tower does not
collapse due to the extra weight of the ice on the tower.
1.2.1 Ground Installations

Ground installation is intended for free-standing 10-m aluminum triangular towers that will be installed
on the earth's surface in an area that does not experience severe/extreme weather conditions.  Additional
guy wires (wires that extend from the upper portion of the tower to the ground to provide additional
support) and a larger concrete footing may be necessary depending on the circumstances of each
installation, especially in climates that experience permafrost conditions.

The procedure illustrated in Figure 1.510 explains, step by step, how to create a concrete footing, secure
the tower base to the footing, erect the tower, and secure the tower using guy wires and other methods.
For guidance on choosing a location and siting the tower, see Section 1.4.
Step 1. Dig a 3-ft x 3-ft
x4-ft deep hole.
Approximately 1.5 cubic
yards of concrete will
be needed to fill the
hole.  Rebarwill also be
needed for towers
greater than 10 m in
height.
Step 2. Place the tower's
base into the hole and
ensure that all three base
legs are lined up so that
the tower may be
lowered and raised
without obstruction. Pour
the concrete.
Step 3. (Perform this step while the concrete is
setting up).  It is very important to ensure that a
triangular fixed tower is
plumb. Use the bottom
section of the tower to
ensure proper positioning
of the base by vertically
placing a carpenter's
level on the  leg of the
tower.  Crank-up towers
will require at least three     /'~
people to perform this      /  .
step.                       ~~
Step 4. Allow sufficient time for the concrete to set
up.
Step 5. Ensure the tower is
completely assembled and attach
three guy wires to the guy wire
holes on  the tower using the
hardware included with the tower.
If hardware is not included, use an
anchor shackle fastener to secure
the guy wires to the tower.
Figure 1.5 Ground Installation Procedure (Steps 1-5)

-------
Step 6. Dig a 1 -ft x1 -ft x 1 -ft hole for each guy wire.
Ideally, holes should be approximately 25 ft from
the tower's base and should enter the earth at a
45° angle.
Step 7. Pour concrete into the hole and insert a
12-inch eyebolt.  Let the concrete set
                                                                                Volume 4, Section 1
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
                                                                                      Page 5 of 18
Step 8. Erect the tower.  Secure the tower to the
base. Attach the guy wires to the eyebolts and
make any necessary adjustments. It is
recommended that two people be present when
raising and lowering a tower.
Figure 1.5 Ground Installation Procedure (Steps 6-8)
1.2.2 Roof Installations

Roof installation is intended for triangular towers that will be installed on an existing structure in an area
that does not experience severe/extreme weather conditions. Roof installations differ from ground
installations because there is no earth available to dig a footing. This procedure assumes that penetrating
a roofs surface is not permitted. However, if penetration is permitted, fastening the tower directly to the
roof of the building will be much easier than creating a non-penetrating base. Additional guy wires and a
larger base may be necessary depending on the circumstances of each installation.

The procedure illustrated in Figure 1.610 explains, step by step, how to set up a non-penetrating base,
secure the tower to the base, erect the tower, and secure the tower using guy wires and other methods.
For guidance on choosing a location and siting the tower, see Section 1.4.
Step 1. Construct or purchase a 10-ft x10-ft support
base. The support base should be made of wood or
metal. The base should be constructed in a manner
that is similar to a freight pallet. However, empty
spaces should be left in the pallet to accommodate
for sandbag placement.  Also, the support base
needs to be constructed so that the tower's base
can be properly mounted on the top of the support
base.


                           Example support base
Figure 1.6 Roof Installation Procedure (Step 1)

-------
                                                                                 Volume 4, Section 1
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
                                                                                      Page 6 of 18
Step 2. Use bolts to secure the tower's base to
the support base. It is very important that the
tower is plumb. For
triangular fixed towers,
use the bottom section
of the tower to ensure
proper positioning of
the base by vertically
placing a carpenter's
level on the leg of the
tower.  Crank-up
towers will require at
least three people to
perform this step.	
Step 3. Ensure the tower is completely
assembled and attach three guy wires to the guy
wire holes on the tower using the hardware
included with the tower.  If hardware is not
included, use an anchor
shackle fastener to secure the
guy wires to the tower.  (Note:
fewer than three guy wires
may be used if the tower is
fastened  to a nearby structure
by other means).
Step 4. Locate nearby structures or retaining walls that are rigid and strong. These structures
may be used to host eyebolts for guy wire installation.
Figure 1.6 Roof Installation Procedure (Steps 2-4)

-------
                                                                                 Volume 4, Section 1
                                                                                    Revision No: 1.0
                                                                                      Date: 10/2006
                                                                                       Page 7 of 18
 Step 5. Install eyebolts into nearby
 structures. The eyebolts need to be arranged
 so that the guy wires on the tower form a
 strait line from the tower. Because the
 availability of nearby structures may be
 limited, this step may be problematic.
             (Top View of Tower)
                             'Direction of
                             tower tilt-down
      Guy anchor, eye-bolt. 3 places
Step 6. Erect the tower.  Secure the tower to the
base. Attach the guy wires to the eyebolts and
make any necessary adjustments.  It is
recommended that two people be present when
raising and lowering a tower.
Figure 1.6 Roof Installation Procedure (Steps 5-6)
1.2.3 Other Installations

Every site is unique and installation options will vary. Therefore, it is recommended that a site's layout
and available support structures be considered before commencing installation. Some unique installation
methods are illustrated in the Figures 1.7 through  1.10 below.
Figure 1.7 (a) Fascia Board Support; (b) Fence Post Guy Wire Anchor

-------
                                                              Volume 4, Section 1
                                                                 Revision No: 1.0
                                                                   Date: 10/2006
                                                                    Page 8 of 18
Figure 1.8 (a) Angle Iron Support—Expanded View; (b) Angle Iron
Support—Close-Up View

-------
                                                              Volume 4, Section 1
                                                                Revision No: 1.0
                                                                  Date: 10/2006
                                                                   Page 9 of 18
Figure 1.9  (a) Metal Wall Support—Telescopic Tower (Expanded View);
(b) Metal Wall Support—Telescopic Tower (Close-Up View)

-------
                                                                             Volume 4, Section 1
                                                                                Revision No: 1.0
                                                                                  Date: 10/2006
                                                                                  Page 10 of 18
                Figure 1.10  (a) Expanded View Guy Wire Harness—Telescopic Tower;
                (b) Close-Up View Guy Wire Harness—Telescopic Tower
1.2.4 Lightning Protection and Grounding

When tower setup is complete, ground rods and wires must be installed to protect the tower components
from damage caused by lightning.  The grounding mast mount should be installed on the tower mast
before the wind sensor crossarm. For ground installations, soil composition should be considered to

-------
                                                                                Volume 4, Section 1
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
	Page 11 of 18

determine the most beneficial grounding technique. Additional technical information about lightning
grounding specifications can be obtained from the Lightning Protection Institute (www.lightning.org).

The typical 10-m tower grounding kit consists of the following parts:
    >    Ground rod
    >    Ground rod clamp
    >    Point
    >    Railing point bracket
    >    Grounding cable—8# copper or 4# aluminum (purchased separately)
    >    Horizontal support
    >    Mast mount

Whether the tower is installed at ground level or on a roof, the grounding kit components should be
installed from the upper part of the tower to base level. The location of the point mast should be taken
into consideration so as to not interfere with proper installation and alignment of the wind sensor
crossarm. The grounding cable should be attached to the point base and fastened to the tower every 2 to
3 feet to one of the pivoting legs of the tower. Isolating the ground cable on one leg of the tower is
advised to reduce the possibility of damage to sensors and signal conditioning equipment in the event of a
strike. The grounding rod is driven into the ground at the base of the tower, leaving 3 inches of the rod
above ground to which to fasten the rod clamp and cable. For roof installations, the cable should run
down the side of the structure from the roof and into the earth's ground (Figure 1.11).
                                                                   /CLJtmTHHICS
                                                                    ^^^,~^^^
                                                         F&.ET3 U-5T OH
                                                                        tl&HTNING
    Figure 1.11 Lightning Protection Installation

-------
                                                                                Volume 4, Section 1
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
                                                                                    Page 12 of 18
1.3 Tower Wiring

The tower wiring scheme depends to a large extent on the type of tower and instrumentation.  Typically, a
meteorological sensor signal cable runs from the sensor to the base of the tower.  Outdoor-grade cable ties
work well to fasten the signal cables to the tower. Signal cables should be installed, keeping in mind the
potential disruption of signal cable placement during instrument audits and maintenance. If a tilt-down
tower is used, the signal cable should be secured to one of the pivoting legs of the tower with sufficient
slack at the tower base to avoid damaging the cable when the tower is raised or lowered.  Some examples
of signal cable installations provided in Figures 1.12 and 1.13 show signal cables secured at the base of a
tilt-down tower.
          Figure 1.12 Signal Cable at Tower Base

-------
                                                                                Volume 4, Section 1
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
                                                                                    Page 13 of 18
                Figure 1.13 Temperature Cable Installation and Expanded/Close-Up Views

Triangular adjustable (crank-down) towers require that signal cable(s) be fastened to the guide holes at
the end of each section of the tower to avoid damaging the cable when lowering or raising the tower.
Figure 1.14 shows the cable guide-hole for an adjustable tower.

-------
                                                                                  Volume 4, Section 1
                                                                                    Revision No: 1.0
                                                                                      Date: 10/2006
                                                                                      Page 14 of 18
          Figure 1.14 Wiring on a Crank-Down Tower
1.4 Tower Siting

This section provides guidance on the siting and exposure of meteorological towers and sensors for the in
situ measurement of primary meteorological variables.

As a general rule, meteorological sensors should be sited at a distance beyond the influence of
obstructions, such as buildings and trees; this distance depends on both the variable to be measured and
the type of obstruction. The other general rule is that the measurements should be representative of
meteorological conditions in the area of interest. Secondary considerations, such as accessibility and
security, must be taken into account, but should not compromise the quality of the data. In addition to
routine quality assurance  activities, annual site inspections should be made to verify the siting and
exposure of the sensors. Approval for a particular site selection should be obtained from the permit
granting agency prior to any site preparation activities or installation of any equipment.5
1.4.1 Representativeness

Site selection for tower placement should address the question, "Is the site (are the data) representative?"
Representativeness is defined as "the extent to which a set of measurements taken in a space-time domain
reflects the actual conditions in the same or different space-time domain taken on a scale appropriate for a
specific application". In general, the location of the tower should be representative of meteorological
conditions in the "area of interest".5

Proper site selection is critical to obtaining representative meteorological data. In order to minimize
absolute error, site selection is much more important than proper placement of individual pieces of air
monitoring equipment.  Poor placement can cause wind direction errors of up to  180° and can cause major
errors in any other meteorological variable, including wind speed, temperature, humidity, and solar
radiation.

-------
                                                                                 Volume 4, Section 1
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
	Page 15 of 18

Proper siting is part of a total quality assurance program. Ideal siting may not always be attainable; in
fact, in many urban areas where air quality studies are traditionally made, it will be impossible to find a
site that meets air quality and meteorological siting criteria. It is incumbent upon an agency gathering data
to carefully describe the meteorological siting deficiencies in a site and, if possible, quantify or at least
evaluate the probable consequences of the siting deficiencies on the data. Additional information about
siting of meteorological sensors in urban areas can be obtained from the WMO, Instruments and
Observing Methods, Report No. 81, "Initial Guidance to Obtain Representative Meteorological
Observations at Urban Sites"11.
 1.4.2 Meteorological Towers

 Meteorological variables are frequently measured at more than one height. Towers are the most
 advantageous platforms for continuous measurement. Height restrictions can be a factor with 60m towers
 in areas close to airports.

 Towers should be located in an open, level area (see Table 1-2) representative of the area. In terrain with
 significant topographic features, different levels of the tower may be influenced simultaneously by
 different meteorological regimes.  Such conditions should be documented well.

 Table 1-2 Limits on Terrain and Obstacles Near Towers
Distance from Tower (m)
0-15
15-30
30-100
100-300
Source: TVA 1977
Slope Between (%)
±2
±3
±7
±11

Maximum Obstruction or
Vegetation Height (m)
0.3
0.5 - 1.0 (most vegetation
<0.3)
3.0
10 x height (must be less than
distance to obstruction?

Tower construction should be open grid, similar to that of most television and radio broadcast towers.
Enclosed towers, stacks, water storage tanks, grain elevators, cooling towers, and similar structures
should not be used. Towers must be rugged enough to be climbed safely to install and service the
instruments. Folding or collapsible towers that enable servicing or calibration of instruments at ground
level are desirable, but they must be sufficiently rigid to hold the instruments in the proper orientation and
altitude during all seasonal weather conditions for that location.

Wind instruments should be mounted above the top of the tower or on booms projecting horizontally out
from the tower.  If a boom is used, it should support the sensor at a distance equal to twice the maximum
diameter (or diagonal) of the tower away from the nearest point on the tower. The boom should project
into the direction that provides the least distortion of the most important wind direction. For example, a
boom mounted to the east of a tower will provide the least distortion of northerly or southerly winds.
Two sets of instruments at each level may be appropriate, located on opposite sides of the tower. A
simple automatic switch can  facilitate the choice of which set of data to use. Orientation of the booms
should be included in the tower setup documentation.
Temperature sensors must be mounted on booms to hold them away from the tower, but a boom length
equal to the diameter of the tower is sufficient. Temperature sensors should have downward-facing

-------
                                                                                Volume 4, Section 1
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
                                                                                     Page 16 of 18
aspirated shields.  The booms must be strong enough so that they will not sway or vibrate excessively in
strong winds. The best vertical location of temperature sensors on a tower is a point with a minimum
number of diagonal cross-members and away from major horizontal cross members. Even with these
precautions, data obtained while the wind blows from the sector transected by the tower may not be free
from error.

Choosing a good location for tower placement is key to ensuring ease of maintenance.  The location
should permit unobstructed tower raising and lowering and particularly take into account how protruding
meteorological equipment such as a temperature probe may be affected. The location should also allow
placement of meteorological equipment at a proper distance from surrounding objects.  Triangular towers,
for example, may only be raised or lowered in one direction; and this limitation must be considered before
installation. Figure 1.15  shows that maintenance considerations were made when a tower was installed.
The tower is located so that it can be lowered between the gate opening of a chain-link fence. If the tower
had been installed in any other direction, the tower would have rested on the fence and a ladder would
have been required to gain access to the top of the tower.  Figure 1.16 illustrates the use of Superstrut®
metal framing attached to the shelter to support the tower. Using the shelter to support the tower at two
heights eliminates guy wires.
                           Figure 1.15 Example of a Tower Location

-------
                                                                  Volume 4, Section 1
                                                                    Revision No: 1.0
                                                                      Date: 10/2006
                                                                       Page 17 of 18
Figure 1.16 Example of a Tower Attachment

-------
                                                                                Volume 4, Section 2
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
                         	Page 1 of27
                          2.   Wind Speed and Direction
2.1  Introduction

All aspects of the task of monitoring wind at a particular site with an emphasis on quality assurance are
discussed in this section. This section includes background information describing the nature of wind and
the kinds of instruments commonly used to measure speed and direction. The important aspects of the
operation of conventional  anemometers and wind vanes are detailed. In addition, a discussion of the
emerging use of sonic anemometers is included. The background and detailed information provided are
necessary for two kinds of tasks. The first task is to collect valid data representative of the project
objectives. The second task is to audit or judge how well the first task was performed within the goals or
regulations requiring the measurements.

Section 2.2 on instruments and specifications stresses the understanding of how common sensors work.
That understanding is necessary before purchasing, installing and operating instruments. Specifications
set the performance standards for an instrument or system. This section provides wind instrument
definitions along with test methods that will enable the user to verify or judge the work of others who
verify conformance to specifications.

Once the specifications are clearly understood, the process of purchasing and acceptance testing can be
considered. Quality assurance is  a vital aspect of defining the systems or instruments to be purchased and
verifying their performance. Measurement quality objectives (MQOs) play the principal role in
determining what system should be purchased. Once the wind system is chosen based on specifications,
the installation can be planned and implemented.

Calibration is the foundation of data validity. This important function may be practiced in a number of
phases of the  monitoring program. Documentation of calibration findings and methodology is vital. The
use of inclusive calibration methods in field conditions is recommended. After a calibrated system has
been installed, the routine performance of operational checks, preventive and corrective maintenance, and
quality control operations  begin. The documentation of calibration results, procedures, and field
operations provides the framework to support data validity.

Performance audits document the accuracy and precision of a wind system and confirm that quality
control procedures have been properly implemented. Comprehensive performance audit methods should
be used to challenge the wind system. Recommended audit methods are described in detail in this section.

All the details or background information that might be needed or desired is not included in this section;
however, references are listed in Section 11.  If the reader needs additional information or is curious about
peripheral subjects, the references will provide the answers or a start in search of the answer.
2.2  Types of Instruments and Specifications

Manufacturer's specifications define performance criteria of wind instruments and systems needed to
meet project objectives. Procurement specifications ensure that instrumentation will meet data and
measurement quality objectives. Performance and procurement specifications provide the basis of
inspection and testing when instrumentation is received at a site. Wind speed and wind direction are
typically the most important parameters measured, and the specifications of the instrumentation used to
measure them are the most complicated.

-------
                                                                                 Volume 4, Section 2
                                                                                    Revision No: 1.0
                                                                                      Date: 10/2006
	Page 2 of27
Project and application requirements vary. To make this handbook as specific as possible, the discussion
will concentrate on requirements typically encountered for air quality monitoring, consistent with those
presented in the On-Site Meteorological Program Guidance for Regulatory Modeling Applications.5
These specifications in general meet the most demanding goals encountered in air monitoring.
2.2.1 Cup Anemometer and Vane Systems

The cup anemometer has cups that relate the rate of rotation and the wind speed. The cup anemometer's
dynamic performance characteristics (starting threshold and distance constant) are density-dependent, but
its transfer function (rate of rotation vs. wind speed) is independent of density. The cup is not very
efficient and creates turbulence as the air flows through and around it. The cup anemometer is omni-
directional in a horizontal flow situation but exhibits a complicated reaction to vertical flow situations.  It
                                              may indicate speed slightly greater than the total speed
                                              when the flow is not horizontal.12

                                              Figure 2.113 illustrates a classic wind cup anemometer
                                              and vane system. The wind vane is perhaps the
                                              simplest of instruments.  A fin is tied to a vertical shaft
                                              so that when force is applied to the area by the wind, it
                                              will turn the shaft, seeking a minimum force position.
                                              The relationship of the shape, size, and distance from
                                              the fin's axis of rotation to the bearing assembly and
                                              transducer torque recommendations determines the
                                              starting threshold. These attributes of the fin area,
                                              along with its counterweight, determine the dynamic
                                              performance characteristics of overshoot (damping
                                             ratio) and delay distance (distance constant) of the
                                             direction vane.
Figure 2.1 Example of Cup and Vane
System
                                             Vane design is of little importance if average wind
direction is the only recommendation. If turbulence parameters are of interest, the design of the vane
becomes important. The vane transducer is usually a potentiometer. It is common that the range of the
sensor is 350° rather than the physically true 360°. The reason is related to the problem of a continuous
range (a circle) with a discontinuous output (0 to n volts). Knowing how the transducer works is
important if the performance of the wind vane will be challenged for QA purposes.

A special direction vane—the bi-vane—has a vertical range of 45 to 60 degrees in addition to the full
azimuth circle. The additional range brings with it the need to neutralize gravity by having a perfectly
balanced vane assembly. A bi-vane can be conditionally out of balance, which can happen when dew
forms and then evaporates from the tail fins. The effect of this imbalance on threshold and performance is
complicated.  Bi-vanes are rarely used in air monitoring applications and will not be discussed further.

-------
                                                                               Volume 4, Section 2
                                                                                 Revision No: 1.0
                                                                                   Date: 10/2006
                                                                                    Page 3 of27
                                    2.2.2 Propeller Anemometer and Vane Systems

                                    The propeller anemometer, Figure 2.214, is a more efficient
                                    shape. The helicoid propeller is so efficient that its transfer
                                    function can be specified from theory.15 It creates little
                                    turbulence because the air flows mostly through it. The propeller
                                    measures wind speed when it is oriented into the wind by a vane.
                                    Its errors from imperfect alignment with some mean vector are
                                    small and are nearly proportional to the cosine of the angle of
                                    misalignment.
2.2.3 Sonic Anemometers
Sonic anemometer systems are based on the principle that wind changes the transit time of a sound pulse
across a fixed distance.  Sonic systems can be designed in two dimensions for horizontal wind speed and
direction as a replacement for the cup and vane or propeller units, or in three dimensions for both
horizontal and vertical wind measurements. For those applications where the contribution of small eddies
is important, sonic systems are an excellent choice. Sonic anemometers are being used in routine air
monitoring networks; however, because they are based on newer technology, sonic anemometer systems
are not as widespread as conventional systems. Because the measurements are based on a different
principle, sonic systems can produce different results, and more comparisons with conventional systems
are needed to understand these differences. As with any method in a routine air quality monitoring
network, it is critical to perform an audit of a sonic system.  Figures 2.3 and 2.414 show examples of sonic
anemometer systems.
      Figure 2.3 2-D Sonic Anemometer
                                       16
Figure 2.4  3-D Sonic Anemometer

-------
                                                                                 Volume 4, Section 2
                                                                                    Revision No: 1.0
                                                                                      Date: 10/2006
	Page 4 of27
2.3 Acceptance Testing

The instrumentation procurement document, purchase order, or contract should be specific in terms of
required performance specifications.  "Required" in this context may only mean that an instrument meets
the suggested or specified regulatory performance. Beyond the scope of this handbook is whether
"necessary" relates to the application  for which the data will be used.

There are two kinds of performance specifications: those that can be verified by simple inspection testing
and those that require unusual test equipment and experience. The former should be tested and the latter
certified by the manufacturer. The manufacturer should have either performed tests on one or more
samples of the model design or arranged for such tests to have been conducted by a calibration facility. In
either case, a test report should be available for any purpose that requires the documentation.

Acceptance testing procedures for checking sensor threshold, accuracy, distance constant, and overshoot
are detailed in the 1995 version of the quality assurance handbook (Volume IV, Section 4.2.32). Methods
for performing acceptance testing  are  in many ways similar or even identical to those for calibrating the
sensors. Thus, the following discussions on acceptance testing only summarize the methods.  The section
on calibration offers a more detailed discussion of the methods.
2.3.1 Wind Speed

2.3.1.1 Threshold

The key measurement for threshold is starting torque. Starting torque requires knowledge of the K value
(cup or propeller aerodynamic shape constant), which should be available from the manufacturer.

2.3.1.2 Accuracy

The acceptance test for accuracy is the conversion of rate of rotation to output in units of wind speed. The
transfer function, supplied by the manufacturer, should be in terms of rate of rotation (rps) versus wind
speed (m/s). The accuracy can be checked by turning the anemometer shaft at a few known rates of
rotation to see if the system output compared to the predicted output is within the tolerance specification.

2.3.1.3 Distance  Constant

The distance constant determination requires a special wind tunnel test and is beyond normal receiving
inspection capability.
2.3.2 Wind Direction

2.3.2.1 Threshold

The threshold acceptance test is a starting torque measurement. To relate the torque measured to wind
speed and offset angle, a K value is required, either from the manufacturer or from an independent test.
The torque measurement may be made with the vane assembly removed or with the vane assembly in
place.  If the latter is chosen, verticality is essential to negate any out-of-balance in the vane assembly
from biasing the test.  Also, there must be no air motion. Very small air motions will bias the test. A
smoke puff should be  used to ensure that the air is still and the inspector should refrain from breathing in
the direction of the vane surface.

-------
                                                                                 Volume 4, Section 2
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
	Page 5 of27
2.3.2.2 Accuracy

The acceptance inspection is the best time to establish true non-linearity.  The test using some circle-
dividing fixture capable of fine resolution of one degree should be utilized to provide a record that can be
referenced in future field spot-checks. Without such a test, it is difficult to prove wind direction sensor
accuracy within three degrees

The acceptance inspection cannot include orientation error with respect to the sensor siting to true north.
There may be orientation fixtures, however, that assume that an optical centerline is parallel to the line set
by an orientation pin. Field orientation may be based on the orientation of a crossarm with the
assumption that the output angle when the vane is parallel to the crossarm is known. This assumption can
be tested or the alignment fixture set in laboratory conditions to the desired output.

2.3.2.3 Delay Distance and Overshoot

These dynamic characteristics require a special wind tunnel test and their determination is beyond normal
acceptance inspection capability.
2.3.3 Measurement System

All the elements of a system of signal conditioners, recorders, and monitors will require checking for
correct function. The receiving inspection should include testing these various sub-systems.

After the calibration inputs have been adjusted and the "output" has shown the system to be in calibration,
a parallel analog recorder may show incorrect values. This event could be caused by an incorrect
adjustment in the interface that drives the analog recorders from the output.
2.4 Installation, Instrument Exposure, and Wiring

"The standard exposure of wind instruments over level, open terrain is 10 m above the ground"11,
however optimum measurement height may vary according to data needs. Open terrain is defined as an
area where the horizontal distance between the instrument and any obstruction is at least 10 times the
height of that obstruction. An obstruction may be man-made (e.g., a building) or natural (a tree)
(Figure 2.5). A wind instrument should be securely mounted on a mast that will not twist, rotate, or sway.
If a wind instrument must be mounted on the roof of a building, it should be mounted high enough to be
out of the wake of an obstruction. Roof mounting is not a good practice, however, and should only be
resorted to when absolutely necessary. Sensor height and its height above the obstructions, as well as the
character of nearby obstructions, should be documented.

-------
                                                                               Volume 4, Section 2
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
                                                                                     Page 6 of27
Figure 2.5 Siting Wind Instruments—a 10-m Tower Located at Least 10 Times
the Height of Obstructions Away From Those Obstructions (Not to Scale).17
                                                      4

2.5 Calibration and Alignment

Calibration qualifies as both a measurement and an adjustment, if necessary, of the performance of the
wind system and its components. Manufacturers usually include in their manuals the details of all
available calibration or adjustment points. The important consideration is how the system works as a
whole. Since only parts of a system are adjustable, the relationship of these adjustments to the whole
system must be known.  This section focuses on documenting calibrations and methods to verify system
response to subcomponent adjustments.
2.5.1 Wind Speed Calibration

Wind speed sensor bearings deteriorate, and that deterioration influences the performance of the sensors.
While bearing deterioration can typically be identified by "feel", true torque measurements for data
validity are nevertheless required.

Calibration that challenges the entire system, except for the coupling or sensor reaction to wind, compares
the rate of rotation of the anemometer shaft to output speed. The rate of rotation is generated by a
synchronous or a direct current (DC) motor. The average rate of rotation of the motor must be known to
convert the rotation rate to wind speed equivalency. The averaging period of the wind system must be
known to challenge the system with a known wind speed equivalency within the averaging time period.  It
is recommended that the calibration be performed with the sensor installed on the tower.

Accuracy determination depends on both the method used in the challenge and the accuracy of the
measurement of the input. Figure 2-6 illustrates one method of determining the  accuracy of the propeller
systems. The first step in calibrating a wind cup or propeller anemometer is to remove the vane or cup.
All vane or cup anemometers will have a starting threshold value, which is generally very low, usually
around 0.2 to 0.5 m/s.  When the vane or cup is removed, the value should be recorded from the DAS.

-------
                                                                                 Volume 4, Section 2
                                                                                    Revision No: 1.0
                                                                                      Date: 10/2006
                                             	Page 7 of27
                                              This recording should be checked against vendor
                                              specifications for starting threshold. A synchronous or
                                              DC battery-operated motor/controller should be
                                              attached to the propeller shaft. The motor in the
                                              illustration is attached to the propeller shaft by stiff
                                              tubing or a vendor-provided coupling. Once the motor
                                              is attached to the propeller shaft, the motor is turned on
                                              and allowed to spin the shaft.  Care should be taken at
                                              this step so that the propeller shaft turns at exactly the
                                              same speed as the motor. Any drag will produce
                                              erroneous readings. Once the motor and controller
                                              shows the shaft is spinning at the correct speed, the
                                              DAS should be read and compared against the vendor-
                                              specified rpm versus rotation table. Next, the speed of
                                              the motor needs to be adjusted. The sensor needs time
                                              to adjust to the next calibration point before recording
                                              this value. After the wind speed sensor has been
                                              challenged at several speeds, the response of the sensor
                                              should be compared to the vendor-specified results at
                                              that speed. The responses should be within the MQOs
                                              that are  detailed in Tables 0.1 through 0.9 in Section
Figure 2.6 DC Motor Calibration (source: http://www.youngusa.com)          ° 2  is manual

2.5.1.1 Wind Speed Sensor Threshold Testing

Sensor starting threshold is a shaft bearing efficiency measurement only. If the wind speed sensor
threshold value is above the vendor-recommended value, the instrument must be disassembled and the
shaft bearings replaced.  Starting thresholds are measured with  a torque watch or a torque disk, with a
range of 0.01 to 1.0 gram-centimeter (gm-cm). Figure 2.7 illustrates a torque disk for determining the
starting threshold.  A torque disk is a simple device that uses gravity to exert torque on the wind speed
sensor. The torque disk has a center hole with threaded holes that radiate out from the center.  The center
hole is used to attach the torque disk to the propeller or cup anemometer shaft. The holes that radiate out
from the center are utilized to attach screws of varying weights. First, the sensor is removed from the
tower and brought inside a shelter where there are no wind influences.  The sensor is placed on a level
surface. Figure 2.8 shows the torque disk attached to the propeller shaft.
Figure 2.7 Torque Disk
Figure 2.8 Torque Disk in Testing Position

-------
                                                                                 Volume 4, Section 2
                                                                                    Revision No: 1.0
                                                                                      Date: 10/2006
	Page 8 of27
The following steps constitute the procedure to test the starting threshold of a propeller-type wind speed
sensor:
    >    Place a 1-gm-cm screw in the first hole from the center on the torque disk.
    >    Orient the torque disk so that the screw is in a horizontal position.
    >    Release the torque disk and observe the response of the propeller shaft.
    >    If the screw moves from horizontal, the starting threshold is less than 1 gm-cm.
    >    If the screw does not move from horizontal, the starting threshold is greater than 1 gm-cm. The
        weight is either moved to a hole outward from the center to increase the torque or another weight
        is added.
    >    The procedure is repeated until the torque disk moves from horizontal determining the torque in
        gm-cm.

Starting threshold speed is calculated by using the following equation:

                                           T = Ku2                                          (2-1)

Where:   T =  torque (gm-cm)
         u2 =  the square of the wind speed
         K =  the aerodynamic constant supplied by the manufacturer documentation

The torque formula is converted to provide the starting threshold speed by the following relationship:

                                        u = (T/K)1/2                                        (2-2)
Where:   u =  starting threshold
         T =  torque (gm-cm)
         K =  aerodynamic constant

The wind speed starting threshold should be less than or equal to the manufacturer's specifications for
starting threshold.


2.5.2 Wind Direction

2.5.2.1 Orientation

The largest source of error in a wind direction measurement is the wind vane orientation to true north.
Orientation error is a fixed bias that can be removed from a data set. The method of wind vane
orientation must be capable of 1° accuracy with 2° as the upper limit of the error. Two steps are necessary
to achieve wind vane orientation:
    >    true north location must be determined accurate to <1 degree, and
    >    wind vane "reference position" must be fixed to true north accurate to <2 degrees.

2.5.2.2 Magnetic Declination Methods

The U.S. Geological Survey (USGS) indicates the deviation of magnetic north from true north on its
topographic maps. Because the earth's magnetic field is constantly changing, the magnetic deviation, or
declination, is periodically updated when maps are revised, and the year of the applicable  declination is

-------
                                                                                  Volume 4, Section 2
                                                                                    Revision No: 1.0
                                                                                      Date: 10/2006
                                                                                        Page 9 of27
 shown on each map. Several maps from the USGS site illustrate the magnetic declination lines throughout
 the United States and North America.

                                                                Figure 2.9 illustrates the magnetic
                                                                declination lines for North America
                                                                in 2000. Real-time magnetic
                                                                declination can be obtained using
                                                                the National Geomagnetism
                                                                Program from the USGS web site
                                                                (http://geomag.usgs.gov/models).
                                                                Figure 2.10, illustrates the magnetic
                                                                declination for the United States in
                                                                1995
                                                                (http://geology.isu.edu/geostac/Field
                                                                _Exercise/topomaps/mag_dec.htm).
                                                                The National Geographic Data
                                                                Center web site
                                                                (http://www.ngdc.noaa.gov/seg/geo
                                                                mag/jsp/struts/declZip) allows the
                                                                user to enter the longitude and
                                                                latitude of a location to determine
                                                                the actual declination.
Figure 2.9 Magnetic Declination in North America (source:
http://geomag.usgs.gov/charts)
 aberrations in the local magnetic field.  These aberrations could be due to soil types (high ferrous content)
 or ferrous metal debris buried underground.
The degree of magnetic declination
is used to correct a magnetic
compass reading to obtain true
north. The magnetic compass
reading is subject to errors due to

-------
                                                                                  Volume 4, Section 2
                                                                                    Revision No: 1.0
                                                                                      Date: 10/2006
                                                                                       Page 10 of 27
                         «0e
                       Shaded relief map copyright 1995 by Ray S
-------
                                                                                Volume 4, Section 2
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
	Page 11 of 27
time of day are known. For the 1995 revision to the EPA quality assurance guidance,2 Lockhart extracted
the subroutines from the Blackadar program that calculated the sun's azimuth, elevation, and solar noon
values based on the required inputs. The information from this revised program could then be compared
to the magnetic azimuth measurements of the sun to calculate the local magnetic deviation at the point of
measurement.  Presently, a number of comparable programs can be found on the Internet. The U.S. Navy
Observatory web site (http://aa.usno.navy.mil/data/docs/AltAz.html) offers a particularly easy program.
This web site requests the date, location, and state in order to calculate the solar noon angles in ten-minute
increments.

The calculation methodology requires the following equipment:
       Transit- or tripod-mounted compass, accurate to at least 1 degree. The 1-degree accuracy meets
gree accurc
        the EPA wind system alignment of 5 degrees
   >   Site location in latitude and longitude recorded from a hand-held GPS.  An accuracy of about one
        minute is sufficient as long as the readings are not taken at the time of solar noon with high sun
        elevation angles.
   >   Accurate time standard, correct to the nearest 5 seconds (the handheld GPS provides such a time
        standard). At the time of solar noon, with high solar angles, the sun's azimuth angle may change
        rapidly.  For example, at an 88-degree elevation angle, the azimuth angle will change 7 degrees
        per minute of time when low solar angles and times well away from solar noon, the azimuth
        angle change is slower therefore, time accuracy is not as critical.
   >   A program to calculate the sun's azimuth direction at the time of measurement.

The  following step-by-step procedure describes the measurement of the alignment of a sensor relative to
true  north.

STEP 1 - MEASURE THE RELA TIVE POINTING DIRECTION

Position a compass or transit to align the cross-hairs through the crossarm or alignment rod of the wind
direction sensor. Note the indicated direction of the transit when it is aimed through the crossarm.
Independently aligning the transit or compass to a known direction is not important because the crossarm
and  sun measurements are relative to each other.  The measurement is the pointing angle of the alignment
crossarm. This angle is called Apomtmg. Figure 2.11 illustrates a side view of the field setup. Figure 2.12
shows the measurement using a magnetic pocket transit that allows the body to be rotated while the
needle continuously points to magnetic north.
                 Figure 2.11 Side View of the Placement of a Compass or Transit for
                 Measuring the Crossarm Direction

-------
                                                                                 Volume 4, Section 2
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
                                                                                     Page 12 of 27
                                          Relative cross arm pointing direction Apointing
                                                (example shown is ~330°)
                 Figure 2.12 Top View Illustrating the Measurement of the Relative
                 Direction of the Crossarm
STEP 2 - MEASURE THE SUN'S RELATIVE AZIMUTH ANGLE

Without physically changing the ground location of the transit or compass, rotate the head to obtain a
direct measure of the sun's azimuth angle, as shown in Figure 2.13.  Do not look directly into the sun; eye
damage may result without suitable protection. If a "pocket transit" such as one made by Brunton is used,
the mirror can be set to project the sun and the sighting points and lines on a white piece of paper or other
flat object. Figure 2.14 shows the use of the Brunton transit to perform the projection.  When the solar
azimuth angle is identified, the exact time of the measurement is noted. This time  is used to calculate the
actual azimuth angle to the sun.  The measured angle is called Suncorrected.
                                          Relative sun azimuth angle Suncorrected
                                              (example shown is -295°)
                 Figure 2.13 Measurement of the Relative Direction of the Sun

-------
                                                                               Volume 4, Section 2
                                                                                 Revision No: 1.0
                                                                                   Date: 10/2006
                                                                                   Page 13 of 27
                Figure 2.14 Projection of Sun's Reflection Using a Brunton Pocket
                Transit
STEPS- CALCULATE THE SUN'S ACTUAL AZIMUTH ANGLE

Using the U.S. Navy Observatory web site19 calculator, calculate the true azimuth angle of the sun at the
exact date, time, and location that the reading of sun's azimuth angle was measured (Suncon-ected).  This
angle is called Strue.

STEP 4 - CALCULATE THE LOCAL DEVIATION

Calculate the local deviation by subtracting the uncorrected sun angle (Suncorrected) from the true sun angle
(Stnie).  This difference is the local deviation (Dlocal.) (Equation 2-3):
                                        D_ c    c
                                     local   ^true ~ ^uncorrected
STEP 5 - CALCULATE THE CROSSARM TRUE ALIGNMENT
(2-3)
Calculate the true pointing direction (Atme) of the sensor crossarm or alignment rod using the uncorrected
pointing angle (Apomtmg) and the calculated local deviation (Diocai) (Equation 2-4):

                                   Atrue ~ Apointing ' Dlocal                                  v^~v

The calculated pointing direction is now referenced to true north based on the known azimuth angle of the
sun.

STEP 6 - ALIGN THE WIND DIRECTION SENSOR

The wind direction vane is positioned parallel with the crossarm referenced to Atme.  The wind sensor
housing is adjusted until the DAS wind direction reading equals Atme. The wind sensor housing is locked
into position to maintain alignment to true north.

-------
                                                                                 Volume 4, Section 2
                                                                                    Revision No: 1.0
                                                                                      Date: 10/2006
	Page 14 of 27
2.5.2.4 Global Positioning System (GPS) Alignment Method

In May 2000, the Department of Defense removed the Selective Availability (SA) encoding on the GPS
satellite constellation which increased the position accuracy of 12-channel GPS within 5 m to 10 m. This
degree of accuracy of a GPS is capable of measuring the direction of travel, or bearing, over short
distances.

The following steps describe the verification procedure of wind sensor alignment using a GPS:
    >    Establish a reference point 20 m to 30 m from the crossarm in line with the direction of the
        crossarm.
    >    At the crossarm reference point, turn on the GPS and allow the GPS to obtain a reference
        waypoint.
    >    Walk from the reference point to the ground location directly under the crossarm.
    >    The GPS display provides a bearing or direction of travel directly related to the crossarm
        direction.

    >    The procedure is repeated several times in directions toward and away from the crossarm ground
        location to assure the GPS bearings are consistently 180 degrees from each other.

The accuracy of this method is about ±1 degree; therefore, the simple solar method could be replaced with
an even simpler "walking" method.

2.5.2.5 System Accuracy

The system calibration of a wind vane can be checked on the tower by aiming the vane to and from
known directions, such as a distant mountain peak or similar feature. If checks are made with respect to a
mounted component, such as a crossarm, the orientation of the crossarm also needs to be checked. A
single distant feature should be an orientation target that has a known bearing with respect to true north.
Other targets can be secondary checks that challenge both the orientation and the performance of the wind
direction transducer. For estimates using the 540 format, the targets should be reached after a clockwise
revolution and then again after a counter-clockwise revolution to challenge both parts of the transducer.

Before the wind direction transducer is removed from the tower, documentation of the as-found reading
with the vane aligned with the orientation target is essential. This single check provides the basis for data
validity for the period beginning with the previous calibration record and ending with the as-found
reading. Wind data validity is based on a data set bracketed between two valid calibration checks.  The
first calibration check is termed "as left" and the second calibration check is termed "as found". If the
sensor has not been removed from the tower or the orientation adjusted, the "as-left" and "as-found"
readings should be the same within ±2 degrees. If the accuracy of the two checks is not within 2 degrees,
data quality is suspect, and  further investigation is  required to determine the cause and timing of the
discrepancy. When the sensor is removed for calibration, replacement in a keyed fixture will cause the
"as-left" value to be the  same as the "as found". If there is no keyed fixture, the full orientation procedure
will be required to ensure the alignment of the crossarm and wind direction sensor.

A simple orientation procedure requires using a hose clamp to prevent the vane from turning.  Tape that
does not stretch is marginally useful. Stretchy tape like duct tape or electrician's tape will  only work on a
calm  day. The vane should  be set so that the output is the correct value for the orientation target.  If the
angle of the orientation target is coincident with an error relative to the average error of 0 degrees, the
output should reflect this error.  For example, if an error of 2 degrees is noted at the  orientation angle, the

-------
                                                                                  Volume 4, Section 2
                                                                                    Revision No: 1.0
                                                                                      Date: 10/2006
	Page 15 of 27
system reading should be 2 degrees higher than the bearing of the orientation target. Only in this way will
the relative error of the sensor be distributed equally for true directions. The clamp should be tightened so
the output is both correct and constant. The clamped sensor should be mounted on the tower and turned
until the vane points at the orientation target. The vane should be clamped in place. The output should be
verified that it is still correct before the vane clamp is  removed.

2.5.2.6 Component Accuracy

The same comments about calibration circuits, parallel recorders, and panel meters that apply to wind
direction also apply to wind speed, as mentioned in Section 2.5.1 above. With the sensor next to the
signal conditioner (attached with either the operating cable or a suitable substitute) and with a fixture that
holds known relative directions, the signal conditioner can be adjusted if required. The 540  offset voltage,
if one is used,  can be tested and adjusted. The output voltage versus position can be set. The open space in
the potentiometer, if one is used, can be measured and adjusted for the open spot in the potentiometer.

A single potentiometer has an electrical range of 355 degrees with a mechanical range of 360 degrees.
The transfer function of relative direction to voltage output is shown in Equation 2-5.
                                         e = 360xV                                       (2-5)
where:   e= angle (degrees)
         V = output voltage (0 - 1 V scale)
The maximum "full scale" voltage output, set by shorting the potentiometer wiper to the high side of the
potentiometer, is 1.000V (a small error will have been set into the system). The error will be +1.4 percent
of the voltage reading; therefore, the potentiometer electrical range of 355 degrees results in a 360 degree
DAS response. An electrical range of 180 degrees results in a 182.5 degree DAS response. The
adjustment error added to the potentiometer linearity error is not acceptable.  If the signal conditioner is
set to a voltage output of 0.986V when the vane is set to 355 degrees, the DAS response with be 355
degrees (360 x 0.986). At 180 degrees, the DAS response is 180 degrees (assuming no linearity error).
All the error between 355 degrees and 360 degrees is in the 5 degree sector.2

The starting threshold of the wind vane is important to record accurate directions  at low wind speed. The
design of the vane along with the offset angle (or error tolerance, typically 10 degrees) provides a K
value. The K value along with the starting torque of the vane assembly provides a threshold wind speed
using the following equation:

                                 Starting Threshold = (t / K) °5                                (2-6)

where the starting threshold is in m/s, and the torque (t) is gm-cm.


2.6 Operation and Maintenance


2.6.1 Operations

The important aspects of operations, from the standpoint of quality assurance, are planning and
documentation. The purpose of operations is  to acquire valid data. For wind measurements, acquisition of
valid data requires frequent (weekly, if possible) visual examination of the sensors. This examination is
not "hands-on"; it is simply a visual inspection of the active shapes, cups, propellers, and vanes to ensure

-------
                                                                                 Volume 4, Section 2
                                                                                    Revision No: 1.0
                                                                                      Date: 10/2006
	Page 16 of 27
no physical damage has occurred. Sensitive wind instruments can be damaged by hail and by birds. The
nature of an analog recording or plot of collected data, if used routinely, will tell how the sensor is
performing. Routine entries in the station log will provide the evidence of attention to support validity
claims.

Calibrations are a part of operations. A member of the operating organization needs to become the
"expert" on how the measurement system works and what it needs to continue "in control" performance.
Regularly scheduled calibrations build the expertise and the documentation showing measurement
accuracy. For a new installation, a calibration during the installation is necessary. A careful look at the
first week of operation will reveal early failures. If no problems surface, a full calibration at the end of
the first quarter is advisable. For some site  environments and some applications, quarterly calibrations are
recommended.  Semi-annual calibration is the minimum frequency.  If problems are found, they must be
documented and corrected as quickly as possible. The PSD requirement for data collection is 90 percent,
this does not permit much down time. The frequency of calibrations or calibration checks is ultimately
determined by the performance of the instrument system. If problems occur, more frequent calibrations
may be necessary.
2.6.2 Maintenance

2.6.2.1 Routine and Preventive Maintenance

The only routine maintenance required for a wind system should occur during routine calibrations.
Sensors exposed to the elements need the application of cleaning and protective lubricants to their
mounting hardware. When a sensor needs to be removed for close inspection or calibration and it cannot
easily be removed because set screws or nuts are locked to their threads by corrosion, a failure in routine
maintenance is probably the reason.

Preventive maintenance must at minimum follow the manufacturer's recommendations. Considerable
damage can result by ignoring this guidance. Sensitive wind sensors require  specific care if the threshold
is to be maintained.

2.6.2.2 Corrective Maintenance

When a part fails or wears out, a new part usually must come from the manufacturer. The replacement
may take a week or two depending on the part and the manufacturer. It is therefore prudent to have spare
parts available to cover predictable failures. A component plug-in philosophy is the quickest way to
correct failures.  If a bearing or a potentiometer fails in  a sensor, a new calibrated sensor is simply
plugged in while the failed one is repaired.

The next level of spare-part strategy is the sub-component level. Critical and difficult-to-purchase parts
should be stocked and used to repair sensors or circuit cards. Conventional sensor parts, typically bearings
and direction potentiometers, will always need repair at some point.

-------
                                                                               Volume 4, Section 2
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
	Page 17 of 27
2.7 Auditing
2.7.1 General Considerations

A performance audit is the determination of instrument system accuracy made with an independently
selected method and by a person who is independent of the operating organization. To make this
determination for wind measurements, knowledge of the input conditions imposed on the sensors is
required. With knowledge of these input conditions, the transfer functions, and the system's data handling
method, the output can be predicted. The difference between the predicted output and the system output is
the error of the system or its accuracy.

The methodology starts with ways of controlling and/or measuring input conditions. When controlled
inputs are used—as should always be the case for starting thresholds, anemometer rate of rotation versus
output, and relative vane position versus output—the accuracy of the output is easily determined. The
accuracy of the anemometer transfer function is not a part of this determination. When the input
conditions are not controlled, as with the collocated transfer standard (CTS) method, the accuracy
determination has a larger uncertainty. The CTS method, however, challenges the anemometer transfer
function. The best performance audit uses both methods when appropriate.
2.7.2 Wind Speed (Propeller or Cup Anemometer)

2.7.2.1 Sensor Control

Cup Anemometer

This method compares the transfer function used with the system to the system's output. The anemometer
shaft is turned at a known rate of rotation and the output is observed. The means of turning the shaft and
metering the rate of rotation are provided by the auditor and are completely independent of the operating
system. This method does not challenge the transfer function.

The torque measurement may be used as an indication of the bearing condition and, hence, the starting
threshold of the anemometer. The time constant is of use if turbulence is measured.

The following procedure is used.
    1.   Remove the cup assembly.
    2.   Mount a coupler to the anemometer shaft. A one-eighth inch shaft is required. If the anemometer
        does not use that size or it is not accessible, an interface fitting is required.
    3.   Clamp the drive motor to the support column of the shaft so that the coupler is engaged with the
        drive wheel. Determine if the cup assembly turns the shaft in a clockwise or counter-clockwise
        direction, when viewed from above.
    4.   Operate the drive motor from the transfer function at two speeds within the desired revolutions
        per second (rps). Use a time period synchronous with the system output. An average of one
        minute or longer is suggested.

This method requires that the system be operating with all cables in place. An rps of zero must be
measured (or observed) with the anemometer in place, the cup assembly removed, and the shaft taped to
ensure non-rotation.

-------
                                                                                Volume 4, Section 2
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
	Page 18 of 27

Fixed Axis Propeller

Similar to the cup anemometer method, the fixed axis propeller method compares the transfer function
used with the system to the system's output. A separate form is provided for the vertical component (W)
because a different transfer function is often used for this direction than is used for wind speed (U) and
output voltage (V). This method causes the propeller shaft to turn at a known rps while the output is
observed. The means of turning the shaft and measuring the rate of rotation are provided by the auditor
and are completely independent of the operating system. The method does not challenge the transfer
function. The designation of clockwise and counter-clockwise is determined by the system being
challenged. Differences are calculated by subtracting the audit challenge value from the system output.
Arithmetic convention is followed, even though the minus sign is used as an indicator of direction.

The torque measurement may be used as an indication of the bearing condition and, hence, the starting
threshold of the propeller. The time constant is of use if turbulence is measured.

The following procedure is used.
    1.  Remove the propeller.
    2.  Mount a coupler to the propeller shaft. A one-eighth inch shaft is required. If the propeller does
       not use that size or it is not accessible, an interface fitting is required.
    3.  Clamp the drive motor to the  support column of the shaft so that the coupler is engaged with the
       drive wheel.
    4.  Operate the drive motor in both a clockwise and counter-clockwise direction, when viewed from
       in front of the propeller.
    5.  Operate the drive motor at two speeds (find the desired rpm from the transfer function) that are
       important to the application of the wind speed data. Use a time period synchronous with the
       system output. An average of one minute or longer is required.

This method requires that the system be operating with all cables in place. An rps of zero must be
measured (or observed) with the propeller shaft in place, the propeller removed, and the shaft taped to
ensure non-rotation. A second observation may be either a motor-driven measured rate of rotation for the
operating period of the system or a normal non-zero operation to ensure that the signal reaches the signal
conditioners when the system is in the operating position.

2.7.2.2 Collocated Transfer System Method

The CTS method for comparing wind speed involves mounting a carefully calibrated anemometer in the
vicinity of the subject anemometer being audited. The  CTS should have NIST-traceable certificates. If the
ASTM method 10 for comparability is being used, the CTS needs to be within 10m of the subject
anemometer in the horizontal and the lesser of 1 m or H/10, where H is the height above the ground in
meters. It is important to site the CTS to be representative of the flow at the subject anemometer. Mutual
interference should be minimized through siting and through editing out data where the direction shows
the wind-passing through one to reach the  other. The accuracy potential of the CTS method is based on
data taken in  1982 at the Boulder Atmospheric Observatory (BAO) and published by Finkelstein et al.20
and Lockhart.21

The result of the CTS audit is the difference in speed calculated by subtracting the CTS speed from the
subject speed. This method requires a sufficient number of simultaneous and independent differences. A

-------
                                                                                 Volume 4, Section 2
                                                                                    Revision No: 1.0
                                                                                      Date: 10/2006
	Page 19 of 27
detailed discussion of the required sample number size is provided in ASTM.22 The optimum duration of
a CTS audit is 24 hours (one diurnal cycle).

The CTS method provides a measure of accuracy that can be related to wind tunnel tests. Some field audit
devices that claim this capability must be used with caution;23 however, the CTS method does not provide
a measure of starting threshold. It is possible to get threshold data from a CTS audit if the CTS has a low
threshold of 0.5 m/s, and if samples from the CTS sensors are found with periods in the 0.6 m/s to 1.6 m/s
range.
2.7.3 Wind Direction (Vane)

2.7.3.1 Sensor Control

The performance audit of a wind direction vane begins by determining the "as-found" orientation value.
    1.   Align the vane with the distant orientation target.
    2.   Use field glasses or a theodolite to confirm the alignment of the vane with the orientation target.
    3.   Hold or clamp the vane until a constant output exists for a few minutes.
    4.   Record this value.

A wind vane's controlled condition is its position relative to the sensor housing. Several ways exist to
impose a series of known relative positions on the vane-sensor combination, however, their accuracy
varies. It is critical to know the time constant of the direction circuit before starting the performance audit.
Set the vane to a known direction, simulate a wind from 90 degrees, and hold the vane until the 90-degree
(or voltage equivalent) output is steady. Then, move the vane quickly (< 1 s) to 270 degrees and measure
the time constant of the system. Assume that a time constant of 3 s is measured. Table 2-1 shows the
change  in output angle and voltage (assuming a 540-degree format and 5V output) as a function of time.
         Table 2-1 Time Constant Effects
Time
Constant
(no.)
0
0.2
1
2.3
3
4.6
6.9
Time
Angle
(sec.)
0
0.5
3
6.9
9
13.8
20.7
Vane
Angle
(deg.)
090
270
270
270
270
270
270
Output
Angle
(deg.)
090
106
204
252
261
268
270
Error
(deg.)
0
164
66
18
9
2
0
(540@5)
(volts)
0.833
0.981
1.889
2.333
2.417
2.483
2.498
Change
(%)
0.0
9.0
63.2
90.0
95.0
99.0
99.9
Notice that in this example, a 150-degree shift requires waiting 20 seconds for the reading to be
representative of the new position. If a 90-degree shift is used, then 14 seconds will provide an output
within 1 degree of the final value. If time constants are greater than the known constant of the direction
circuit, the operator should contact the manufacturer to discuss options to modify the circuit to a
measuring time suitable for 60 Hertz noise filtering.

The least accurate method for challenging the relative position accuracy of a wind vane is to point the
vane in various directions while it is still mounted on the tower. This can provide positions related to
external objects rather than constant angle changes.  It is estimated that the accuracy of this method is two

-------
                                                                                 Volume 4, Section 2
                                                                                    Revision No: 1.0
                                                                                      Date: 10/2006
	Page 20 of 27
to five degrees, with the exception of a parallel alignment. The tail vane can be located parallel to a
crossarm to within one degree and held parallel on a calm day.

Another recommended method is to have the operating sensor be placed in an environmentally controlled
room at the center of a template with radial lines every 60 degrees. With the sensor oriented to the
template, the vane is moved and clamped when it is parallel to the radial line. Care should be taken to
avoid parallax errors (i.e., non-parallel or non-perpendicular observations). The relative accuracy of this
method is on the order of 1 degree.

A third method for challenging the relative position accuracy of a wind vane replaces the vane with a
fixture capable of holding the shaft in position with respect to the sensor housing. Fixtures of this type can
provide repeatable position accuracy within 0.1 degree. A different application of this precise method
   ^^^^^^^^^^^_  uses a theodolite base as the mount for the sensor. With the vane or vane
                          restitute held in one position, the base is rotated in very accurate steps.
                          Theodolite worn gear assemblies divide a circle in whole degrees and have
                          vernier adjustments with 0.1 degree index marks far enough apart to allow
                          easy interpolation to zero degrees.

                          Figure 2.1513 shows a linearity  test fixture that accomplished the audit
                          procedures described above with a repeatable position accuracy of one
                          degree.

                          The audit report form2 should contain the transfer function used to convert
                          output voltage to azimuth degrees. This may include a 540-degree format
                          were azimuth values greater than 360 degrees are reduced by subtracting
                          360. The report form should also contain the challenge regression used by
                          the selected wind vane audit method.
                          The bearing to the orientation target should be independently challenged with
                          a method capable of better-than-compass accuracy. A theodolite is ideal for
                          finding the bearing to other distant objects.
                          The last activity of the sensor control audit is to repeat the orientation test
 Figure 2.15 Linearity
 Test Fixture

described above for the as-found value. The as-left value will represent any changes the operator may
have made and the new orientation if the sensor was not keyed for orientation.

2.7.3.2 CTS Method

There  is no technical need for a CTS audit of wind direction. No new information is added by
this method to that gained in the sensor control method.2
2.7.4 Sonic Anemometers

2.7.3.1 Sensor Control

Using a sensor control method to audit sonic anemometers in the field achieves little. To verify a response
of zero, the few control options available include placing a box over the anemometer or in some other
way keeping the sensors from observing any air movement.

-------
                                                                                 Volume 4, Section 2
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
	Page 21 of 27
2.7.3.2 CTS Method

For audits of sonic anemometer systems, some discussion is provided here, but field implementation of
these techniques is very limited. Additionally, there are standards for testing and evaluation of the
performance of sonic systems,24 but the methods described would not be practical for field
implementation. However, these methods have been adapted for field audits and implemented in a
routine air quality monitoring network in a cost-effective and timely manner.25 The focus of these
demonstration audits was on the variables of wind speed and wind direction from the sonic anemometer;
however, other variables (temperature, relative humidity, pressure, solar radiation, and UV radiation)
were handled as well.

For the CTS demonstration audit, the objective is to directly compare a collocated mechanical sensor
(such as a cup/vane or aerovane anemometer) with a sonic sensor system and evaluate the results against
standard criteria. The following section discusses field tests performed by Baxter, et. al.25

To allow rapid deployment and retrieval of the audit package in the field, the sensor wiring of a data
logger is converted so that both the from the screw type panel mount to a standard 25-pin connector used
for computers.  The required channels on the data logger were assigned specific pins in the connecting
cable and a seven-connector junction box was used as the main connector interface for multiple sensors.
All numbered pins in the junction box were wired in parallel, which allowed a sensor to be plugged into
any of the seven connectors and be operational. The assignments of power, ground, excitation, and signal
lines were performed in the wiring of the pins in the cable for the individual sensors. The length of the
main cable connection between the data logger and the interface was kept short to minimize electrical
noise and ground loop problems associated with the distances from  common ground connections.

Prior to the start of the demonstration audit program, the mechanical audit wind system was collocated
with an RM Young Model 81000 sonic anemometer. Data were collected over a 72-hr period and 5-min
horizontal scalar and vector averages of wind speed and wind direction were compared. The sensors were
approximately 4 meters above roof height and 1 meter apart, making the sampling height less than idea.
Figure 2.16  shows  the sensor mounting. Scatter plots for the scalar wind speed and unit vector wind
direction data sets for wind speeds greater than  1.0 ms"1, as measured on the mechanical sensor, are
shown in Figures 2.17 and 2.18, respectively. The wind speed plot  showed excellent agreement between
the sensors;  the sonic anemometer averaged  wind speeds 0.04 ms"1 higher than the mechanical sensor.
The  standard deviation of the differences was 0.07 ms"1.  Wind direction differences averaged 6 degrees
with a standard deviation of 7 degrees. These results were higher than what is recommended by
Lockhart;21 he indicates that the standard deviation of the differences for good agreement should be better
than 2 degrees. It is suspected that two factors caused this higher difference: the shorter time duration (5
minutes versus Lockhart's 20 minutes) and the less-than-ideal siting, which would induce more
turbulence over the rooftop.  It should also be noted that a regression of the measurement pairs for wind
direction was not done because there were no wind directions less than 135 degrees observed on the
mechanical sensor. Another comparison of mechanical and sonic sensor systems was reported by
Robertson and Katz26; Their results were similar to those given by Baxter et. al.: for  15-minute averages,
the wind speed and direction results did not quite meet Lockhart's21 criteria.

-------
                                                                               Volume 4, Section 2
                                                                                 Revision No: 1.0
                                                                                   Date: 10/2006
                                                                                   Page 22 of 27
                            Figure 2.16  Sensor Mounting for the Testing
                            and Evaluation of the Audit Wind Sensor
                            Against a Sonic Anemometer
                                                    360
               Wind Monitor AQ (ms  )
Figure 2.17 Wind Speed Plot Showing the
Mechanical Sensor (AQ) vs. the Sonic Anemometer
for Wind Speeds Greater Than 1 ms"1
Wind Speeds Gi
                                                    270 -
                                                    180 -
                                                  c  90 --
                                                            90       180       270
                                                              Wind Monitor AQ (°)
                                                                                          360
Figure 2.18 Wind Direction Plot Showing the
Mechanical Sensor (AQ) vs. the Sonic Anemometer
for Wind Speed Greater Than 1 ms"1
  Even with the observed differences above, the CTS method was to be used to audit multiple sonic
  systems in an air quality network. Based on the results of the available intercomparisons, proposed
  criteria for evaluation of the sonic sensors were developed; they are shown in Table 2-2.  These criteria
  should be modified as more audit and comparison data from systems installed in air quality networks are
  collected.

-------
                                                                                Volume 4, Section 2
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
                                                                                    Page 23 of 27
Table 2-2 Proposed Audit Criteria for the Sonic Systems
Wind Variable
Speed
Direction
Average Difference
±0.2 ms'1 + 5% of
observed
±5°
Standard Deviation of the
Differences
0.2ms'1
2°
Qualifications
Wind speeds greater than
1ms'1
Wind speeds greater than
1ms'1
Recommended procedures for the CTS audit of sonic anemometer systems are as follows.
        1.   The site sonic anemometer systems should not be removed from the mounting tower during
           the audit process and all checks should be conducted with the sensors in place.
        2.   Assuming the towers have movable carriages, then the entire crossarm and mounting
           assembly can be lowered from the measurement height to the surface. Before lowering the
           sensor, its orientation relative to true north should be measured using either a solar method or
           alignment walked effusing a hand-held GPS receiver. These methods are described in
           Sections 2.5.2.3 and 2.5.2.4. Figure 2.19 shows a typical mounting of the audit system on the
           carriage structure.
                                                                      ,.,
           The wind sensor should be placed on one end of the audit boom while the
           temperature/relative humidity sensor should be placed on the other end adjacent to the site
           sensors.
                          Figure 2.19 Typical Mounting of the Audit
                          Sensors on the Site Tower
         A zero point with no wind flow around the sensor can be established using a simple box lined
    with "egg-crate" type foam to absorb acoustic signals. This type of enclosure is a simple version of
    what is recommended in ASTM (2001).24 To seal the box, additional foam can be placed in the
    opening around the bottom and around the mounting mast.  The response of the sensor can then be
    observed over 5- to 10-minute periods and the wind speed and direction can be noted.

    5.    A collocated mechanical sensor can then be attached to the carriage on a separate cross-arm
    with the south-facing direction of the sensor aligned with the cross-arm. Once mounted and raised to
    the normal measurement height, the cross-arm direction can be measured and that direction used for
    the adjustment of the collected wind direction data to a true north alignment.

-------
                                                                                Volume 4, Section 2
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
	Page 24 of 27
These procedures would allow field audits to be conducted on multiple stations over a several-day period.
To aid in the efficiency of the audits, the audits should be performed for two days: this allows night-time
meteorological data to be collected, providing a larger comparison database before it is moved to the next
site.
2.8 Scalar, Vector and Sigma Calculations
2.8.1 Discussion of Calculation Methods

There are a number of commercially available data loggers that collect, process, and store wind speed and
wind direction information. Wind direction data have traditionally provided a challenge to those who
want to define an average of the circular function. The EPA has provided guidance for calculating wind
direction in regulatory driven and other monitoring programs.  These procedures have been incorporated
in one form or another into the available data loggers as standard calculation algorithms. Since the
release of the guidance in 1987, the scalar calculation has generally been accepted as a good estimate of
the average wind direction.  However, experiences gained in using data collection systems implementing
both the original EPA scalar average calculation and the unit vector method have raised questions about
the validity of the scalar method.

To illustrate the problem with the original algorithm, and to belter understand the behavior of it, a model
was developed to generate test wind data and perform the wind direction calculations. One-second values
comprising the test data were generated by specifying various characteristics about the desired data.
These characteristics included the starting direction, the maximum direction swing during the hour, and
the maximum rate of change from point to point. A random number generator was then used to create
values that fit within the criteria. After the data set was generated, a specified number of 360-degree
rotations were added to the set using rotation criteria such as direction and rate. Each of the generated
data sets was then saved to a file for analysis by the wind direction calculation algorithms.
Following the creation of the hourly wind data sets, each set was read by the model and average wind
directions were calculated. The simple average wind direction was calculated using
                                                                                           (2-7)
where Aarith is the simple arithmetic average, Ai is the azimuth angle of the wind vane for the i"1 sample,
and  N is the number of samples used (3,600).

The  scalar wind direction AS was calculated using the EPA algorithm
                                             1 ^
                                        ^^ZA                                       (2-8)

where:
          D,, = AI ;  for / = 1
          Dj = D{_^,+ 3; -+- 36O:   for o,- < - 1 SO and / > 1
          D, = D,_-! ; — 1
          D, = Dj_-\\ + 3, - 36O:   for J, > 180 and / :> 1
          3, =A, -D;_^.         for  /> 1

 ^ is the azimuth angle of the wind vane for the i"1 sample.

-------
D,  is undefined for Ł,=180 and />!.
The unit vector wind direction was calculated using
                                      A, ,,^90-tan-1  I
where:
                                       -   1 ^  . , .,
                                       v = — >  sin(A)
                                          N
                                             i=\
                                       ~   1 x^
                                       u=— Ycos(4)
                                                                                  Volume 4, Section 2
                                                                                    Revision No: 1.0
                                                                                      Date: 10/2006
                                                                                      Page 25 of 27
 (2-9)


(2-10)

(2-11)
Auv is the resultant unit vector direction angle in meteorological coordinates. Values less than zero are
corrected by adding 360 degrees, v is the average v component (north/south) of the unit vector wind and
u  is the average u component (east/west) of the unit vector wind.  At is the wind direction azimuth angle
in degrees for sample /'.  For each of the test data sets the number of samples (TV) was 3,600.

The simplest of the model test runs was to look at the calculations when no rotations were present.
Multiple runs were made varying the  allowance for rate of change, swings through north, and maximum
swing during the 3,600-sample period. In each case, both the scalar and unit vector algorithms produced
comparable results.

Then, one 360-degree rotation was injected in the middle of the profile. Figure 2.20  shows a generated
profile with the individual calculations from each of the wind direction algorithms. As shown in the
figure, the scalar calculation produced erroneous results for the data set while the  unit vector produced a
reasonable average. The general agreement with the simple average was due to the selected range of
values with no crossover in the north  direction.
             33C
                  S- m e artifice': z 4VS--B3S- —
        Figure 2.20 Calculated Wind Direction Averages from a Simulated 3,600-Sample Data Set

To evaluate why such significant differences occur between the scalar and unit vector methods, one needs
to recognize that the scalar calculation adjusts the wind direction values in a manner to account for the
360-degree circular function.  If the difference between successive values is greater than ±180 degrees,
then 360 degrees are added or subtracted to bring the result to the closest rotational direction. If a full
360-degree rotation is experienced with the rotation rate less than 180 degrees per increment, then values
outside ±360 degrees can exist in multiples of 360.  During this rotation adjustment process, it is possible
for equivalent directions with numerical values in multiples of 360 to be present in the same data set.
When the final data set is averaged, these multiples then produce erroneous results.  Furthermore,

-------
                                                                                 Volume 4, Section 2
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
	Page 26 of 27
depending on when the shifts occur during the averaging period, the resulting effect can show erroneous
results up to 180 degrees from a correctly interpreted average wind direction.

Based on the above discussion, a unit-vector algorithm should be used for calculating scalar wind speed
and wind direction.
2.8.2 Stability Classes

Standard deviation (i.e., sigma values, e or $) result from how the samples are combined to estimate the
statistical parameters. To audit these values, a determination of how the algorithm works and a method to
challenge that process with a known input is required. This is also a functional way to document the
impact of the signal conditioning time constant on the measurement of direction variability.

The challenge to the process should be realistic or at least within some realistic range. It must take into
consideration the wave shape of the variable direction imposed on the system when calculating the true
sigma value with which the output will be compared. The effective time constant of the direction system,
calculated from the delay distance of the sensor and nominal wind speed important to air pollution
applications, should define the maximum frequency used in the sigma challenge.

Wind Vane
The relative performance of the wind vane shaft position transducer is determined with a linearity test
fixture, part of which is mounted to the transducer body and part of which is mounted to the shaft in place
of the vane.

The following procedures are used.
    1.   Remove the wind vane assembly (vane, shaft, and counterweight). A one-eighth inch shaft is
        required to mount the linearity calibration disc to the sensor. If the sensor shaft is not one-eighth
        inches, an interface fitting is required.
    2.   Mount the linearity calibration disc on the vertical shaft of the senor.
    3.   Mount the clamp to the support column for the shaft so that the pin engages the disc and the disc
        is free to move when the pin is withdrawn. Figure 2.15 shows a sensor mounted on a linearity
        fixture.
    4.   Set the fixture parts with the pin in the 180-degree slot.
    5.   Rotate the clamp until the sensor output indicates 180, either by equivalent voltage or digital
        printout. This is a position measurement;, the challenge is constant and instantaneous values may
        be used However, the time constant needed for stable readings must be taken into account.
    6.   Rotate the disc to the  following positions taking  data at each degree marking: 120, 60, 360, 300,
        240, 180, 120, 180, 240, 300, 360, 60, 120, 180,  and 240. This rotates the sensor shaft 420
        degrees counter-clockwise and than 480  degrees clockwise to test '540' strategies for the angle
        discontinuity.
2.9 Estimating Accuracy and Precision

The 40 CFR 58, Appendix A.461 contains a equations and methods for estimating precision, and bias.
These calculations can be utilized to understand the precision and bias of wind speed and wind direction.
These equations can be applied to situation where one of the systems is considered the "primary" sampler,

-------
                                                                               Volume 4, Section 2
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
	Page 27 of 27
such as a cup/vane anemometer system.  This section of 40 CFR 58 also has equations for collocated
instruments where the operator does not have a "primary" sampling device.
2.9.1 Summarized Data

Appendix 1 contains a meteorological systems audit evaluation form to be used as a guideline to evaluate
the operation and exposure of meteorological sensors and overall condition of a monitoring site.

Summarization schemes are many and preclude a full discussion here. The auditor should define the
methods used and comment on the appropriateness of the methods in regards to the summarized data.
There may be concurrent summarizations such as a scalar wind speed, a vector wind speed, and
summarized wind direction. The accuracy of the data system should reflect estimated errors in case of an
inappropriate summarization program.

-------
                                                                                Volume 4, Section 3
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
                                                                                      Page 1 of 5

                 3.    Temperature and Temperature Gradient

Air temperature measurements are one of simplest meteorological measurements, but accurate and
representative measurements of air temperature require no less attention to quality than others.
Temperature gradient measurements between two levels above ground on a tower demand greater
accuracy and precision than measurements of temperature alone. Temperature gradient is commonly
known as delta-temperature and is abbreviated AT.

High-quality, economical sensors based on electrical resistance that changes with temperature are readily
available and adaptable to many data recording and display systems. The challenge is to place the sensors
in suitable locations and provide proper protection from moisture, wind, and radiation energy
interferences. ASTM International Standard Practice27 for atmospheric temperature measurements
describes the types of shields needed to properly protect electrical sensors. Naturally ventilated shields
can provide adequate protection for many simple air temperature measurements, but the additional
accuracy required for AT measurement necessitates using mechanically aspirated shields to provide
uniform airflow across the sensor in the shield.
3.1 Types of Instruments and Specifications

The most popular method of air temperature measurement is using devices whose resistance changes with
temperature—resistance temperature detectors (RTD). Thermistors and platinum wires are readily
included in resistance bridges that allow acceptably linear and accurate voltage measurement directly by
modern data acquisition system (DAS) or temperature-indicating instruments. Thermistors are often
incorporated into standard portable instruments that are used to check the operation of air temperature
measuring devices used in the field setting.
                                                 Thermistors are electronic semiconductors made
                                                 from certain metallic oxides, such as nickel,
                                                 manganese, iron, cobalt, copper, magnesium,
                                                 titanium and other metals. Individual thermistor
                                                 beads have non-linear properties relating
                                                 temperature and resistance, but suitable
                                                 combinations of beads can provide an adequately
                                                 linear response. Platinum wire properties relating
                                                 temperature and resistance are considerably more
                                                 linear than thermistors. Typical thermistors used
                                                 in meteorological measurements are in the range of
                                                 thousands of ohms range, compared to the
                                                 platinum wires being in hundreds of ohms. This
                                                 difference makes the electrical resistance in the
                                                 cable connections from the sensor to the
                                                 measurement point more sensitive for platinum
                                                 wires than for thermistors.
Figure 3.1 Example of a platinum wire thermistor
(source:  http://www.climatronics.com/)
In years before the routine use of RTD sensors, glass thermometers containing liquid were the standard
instrument for air temperature measurements.  Glass thermometers are still useful tools for testing RTD

-------
                                                                                Volume 4, Section 3
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
                                                                                       Page 2 of 5

sensors, though the need for at least partial immersion in the medium containing the test RTD device and
the fragility of the glass thermometers make them difficult to use in a field environment. More
information about glass thermometers is available from ASTM International.28

The easy availability of reasonably priced on-site digital signal processing makes separate analog signal
conditioning of temperature signals unnecessary. However, RTD sensors are components of bridge
circuits that include fixed resistors and an excitation voltage to produce a measurable voltage signal.
Equipment and system vendors are the best source of proper wiring and signal processing information.

The required system accuracy for air temperature measurements described in EPA monitoring guidance 4
is ±0.5 degrees C. Hence, the reasonable accuracy expected from the air temperature sensor is ±0.3
degrees C. The temperature range over which this accuracy applies depends on the location and purpose
of the measurements. Typical systems can meet this recommendation from -30 to +50 degrees C, which
is ample for many locations.

Sensors used for AT measurements should have at least the same absolute accuracy, with two temperature
sensors having a relative accuracy across the working range of the instruments of less than ±0.1 degrees
C. Suitable pairing of thermistors with the same non-linear characteristics can provide the more rigorous
relative accuracy needed for that measurement.

The exposed temperature sensors and related equipment must be  able to operate throughout the expected
range of temperatures encountered at a site, including likely extreme values.  In addition, ancillary
equipment and related signal processing or recording equipment must be capable of operating in the range
of temperatures corresponding to their location, such as in an instrument box or in a climate-controlled
structure.

When specifying cable length and the accessibility of the sensors, keep in mind that field QC procedures
require removal of the sensor from the shield and its placement in a temperature bath while still connected
in its normal configuration.
3.2 Acceptance Testing

New equipment or equipment that is returned from maintenance or calibration by vendors located
elsewhere should be checked for proper operation. Damage that could affect response can occur in
shipping and handling.  Simple checks for reasonable responses can be an adequate initial acceptance test;
precautions identified in the formal field QC procedures (Section 3.4) should be observed when testing.
A formal check is recommended during installation at the operating site.
3.3  Installation and Wiring

Temperature sensors should be mounted over a plot of open level ground at least 9 m in diameter. The
ground surface should be covered with non-irrigated or un-watered short grass or, in areas where grass
does not grow, natural earth. The surface must not be concrete or asphalt or oil-soaked.  The standard
height for climatological purposes is 1.25 m to 2 m, but different heights may frequently be required in air
quality studies. For general purposes, the primary temperature sensor is mounted 2 m above ground level,
with the inlet facing away, and at a distance of approximately 1.5 times the tower diameter, from the
tower.

-------
                                                                                 Volume 4, Section 3
                                                                                    Revision No: 1.0
                                                                                      Date: 10/2006
                                                                                        Page 3 of 5

The sensors should not be closer to obstructions, such as trees and buildings, than a distance equal to four
times their height. They should be at least 30m from large paved areas and not close to steep slopes,
ridges, or hollows. Areas of standing water should also be avoided.  Louvered instrument shelters should
be oriented so that the door opens toward true north in the northern hemisphere.  Motor-aspirated shields
should also be oriented with the sensors toward true north in the northern hemisphere.

Proper planning will assure that the mounting hardware, cables, power supply, and so forth are all
compatible and available. Planning also assures that an installation will proceed without difficulty.
Purchasing equipment through a vendor who  provides everything the system needs to readily attach
cables can save considerable technical time and effort of having to fabricate mounting hardware and
prepare cables. Additional lightning protection should be considered in areas where lightning occurs.
Figure 3.2 A motor aspirated shield
(source: http://www.metone.com/)
 Figure 3.3 A naturally ventilated shield
(source: http://www.voungusa.com/)
Most temperature sensors do not require additional power beyond the excitation voltages in the bridge
circuit, though motor-aspirated temperature shields require electrical power. Figures 3.2 and 3.3 are
examples of motor and naturally aspirated shields. Ambient temperature sensors must be shielded,
otherwise, solar radiation will cause errors in readings. Shields can be powered by direct current (DC)
fans operating on batteries recharged by solar panels or an alternating current (AC) trickle charging unit.
Operating the system on DC, even when AC power is available, can reduce missing data periods when
electrical power is unavailable.

A complete quality assurance plan will prescribe appropriate installation and testing procedures.  These
procedures are developed from manufacturer's specifications and guidance and instrument exposure
recommendations listed in monitoring guidance documents.

-------
                                                                                 Volume 4, Section 3
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
                                                                                       Page 4 of 5
3.4 Calibration
Calibrating a temperature measurement system consists of comparing the output of the device being
calibrated to a known value, and determining if the difference is within acceptable tolerance limits. Most
modern temperature measuring systems do not need adjusting to match known values if all components
are working properly, so acceptance is pass or fail. Improper signal cable connections or signal
processing instructions are more likely sources of problems producing unacceptable results than the
sensors themselves. Additional information about standardized testing for resistance temperature
measuring devices is available from ASTM International.29'30

EPA guidance4 specifies a tolerance limit of the difference between known and observed  values within
±0.5 degrees C. Calibrations for pairs of AT sensors are discussed separately in this section. Calibrations
should be made over as much of the measurement range encountered in operational service as possible.
Typical sensors are designed for a range from -30 degrees C to +50 degrees C; most stations experience a
lesser range.

Following manufacturer's instructions and using the on-site signal processing  method ensures that the
calibration is representative of the device's response. Note that some temperature probes may not be
submersible.  If this is the case, then follow the manufacturer's calibration procedures. Below is as  step by
step description of a water submersible temperature sensor.

    1.  The calibration test should be performed at three or more temperature levels spaced across the
       range of the sensor, such as 0 degrees C, 20 degrees C and 40  degrees C.

    2.  Prepare three test baths as stated in step 1.

    3.  Remove the sensor from the shield.

    4.  Place the probe and a NIST traceable temperature device in the water  bath with the lowest point,
       a thermal mass  in an ice slurry mixture.

    5.   To reduce physical stress on the sensors, the probes should be allowed to reach equilibrium with
       room conditions when they are removed from one thermal mass and before subjecting them to
       another thermal mass.

    6.  Record the temperature of the bath from the NIST traceable temperature device and the reading
       from the station temperature probe.

    7.  Remove the NIST traceable temperature device and station temperature probe and place them in
       the 20 degree C water bath. Repeat steps 6 and 7.

    8.  Remove the NIST traceable temperature device and station temperature probe and place them in
       the 40 degree C water bath. Repeat steps 6 and 7.

-------
                                                                                 Volume 4, Section 3
                                                                                    Revision No: 1.0
                                                                                      Date: 10/2006
                                                                                           : 5 of 5
3.4.1 Temperature Environment

Producing stable temperature environments for both the standard device and the sensor that is being
calibrated can be a challenge.  In order that both sensors be at the same temperature, both need to be
immersed adequately in the same medium at a stable temperature for enough time to reach thermal
equilibrium.  Most meteorological sensor response times are fast enough that thermal equilibrium
between the thermal environment itself, the standard, and sensor being calibrated can often be reached
quickly; waiting a few minutes for a stable output indication to persist is necessary to ensure that
equilibrium has been achieved.

Most thermistor beads are covered in a metal sheath for physical protection. If this cover is not
adequately  sealed, it will not be suitable for immersion in a liquid temperature bath. Assuming that the
sensor is not adequately sealed is the safer course. The recommended method for producing a stable
temperature environment is to place the temperature probes in a thermal mass, such as a cylindrical
aluminum block. A block about six inches long and four inches in diameter with holes drilled along the
axis of the cylinder is large  and heavy enough to accommodate typical temperature sensors. Minimizing
the difference between the size of the holes and the diameter of the probes reduces the air around the
probe; air at a temperature different from the thermal mass can affect the temperature of the sensor. The
holes should be spaced the same distance from the outside; the holes should be nearly the length of the
temperature probe to immerse the probe in the thermal mass. The mass can be drilled with various holes
to accommodate different sensors, particularly when a AT system is being calibrated.

The block is partially immersed in a water bath or ice slurry to reduce equilibration time and to minimize
horizontal temperature gradients in the cylinder. The block should be placed in an insulated container to
stabilize the temperature of the thermal mass and the temperature sensors. Placing the insulated container
on a magnetic stirring table  will reduce the time needed for the water bath to reach equilibrium.  The ice
slurry should be made from distilled water because the presence of foreign material can alter the freezing
point of water.

An alternative method is to  place the standard and sensor being calibrated in a protective waterproof
sheath.  The sheath should be made of a thin material to allow for a sensitive response of the sensor to the
surroundings, such as a water bath or ice slurry. This approach can reduce the response time, providing
the water bath is in an insulated container and is at a reasonably stable temperature.

If the metal sheath properly protects the thermistor to allow immersion in a liquid temperature bath, the
sensor and  standard device should be simultaneously placed in a water bath. Drill holes in the top of a
plastic insulated container to match the diameter of the respective probes.  Place the plastic insulated
container filled  with water near 20 degrees C on a stirring table, place the sensor and standard device in
the respective holes to allow simultaneous exposure to the water bath. Allow the water bath to reach
equilibrium and record the stable sensor response from the DAS for comparison to the stable standard
device response. Repeat this procedure using an ice bath near 0 degrees C and a warm water bath near
40 degrees  C.

For simple  comparison tests at one ambient temperature level, an adequate result can often be achieved by
placing  the sensor of an electronic thermometer that is similar to the sensor of the system being checked
inside the aspirated shield adjacent to the system sensor, taking care to minimize contact with the shield.

-------
                                                                                Volume 4, Section 3
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
                                                                                      Page 6 of 5
3.4.2 Delta-Temperature
Modern AT systems have eliminated the step of requiring a separate analog signal conditioning system to
produce a signal related to the temperature difference. A pair of simple temperature measurements made
at the two (or more) vertical levels can be digitally processed within the on-site DAS to produce a AT
result that is readily recorded with other measurements in an output data array.

The recommended calibration technique is to place both sensors being calibrated in the same thermal
environment, producing a known AT value of 0.0 degrees C. This test should be performed at the three
temperature levels used for the absolute calibration of the temperature sensors to assure that the sensors'
responses are within tolerance of each other across the working range of the system.

The challenge of AT calibration is to ensure that both temperature sensors are truly at the same
temperature because the acceptable tolerance limit for AT measurements  is  only ±0.1 degree C. The use
of the insulated container and magnetic stirring table will result in the water bath reaching equilibrium.
3.5  Operation and Maintenance

Modern temperature (and AT) measuring systems can be very reliable, but they still require occasional
physical inspection and data-checking to ensure that the sensors are accurately measuring air temperature.
The radiation shields can become clogged with dirt, vegetation, or small animals, so much so that airflow
or electrical connections can be adversely affected.  Changes to the airflow can be gradual, masking the
problem when typical software algorithms search for spurious results. The frequency of the checks
should be based on the environment of the system and operating experience.

Another source of potential measurement interference that can be minimized with routine maintenance is
the paint covering on the temperature shield. Most paint material degrades overtime, producing a dull
finish with reflective properties different from those of the fresh, shiny shield when it was new. Birds are
notorious for contributing to the changes on the surface of the shields; routinely cleaning the droppings
from the shield is advised.

Changes in the fan motors providing aspiration to the  shields can be monitored by measuring the current
flow to the motors. Airflow through the shield can be sensed by switches, but false problem signals can
be produced by unusual wind events that affect the airflow.
3.6 Auditing

Temperature and AT measurements should be included in routine performance and system audits at least
once every six months.  The performance audit should consist of challenging the temperature and AT
sensors to three test atmospheres at 0 degrees C, ambient temperature, and 40 degrees C using a water
bath compared to a NIST-traceable temperature standard. The temperature, AT, and NIST-traceable
temperature probes are simultaneously placed in a water bath until equilibrium is reached for each desired
temperature range.  The temperature sensor's response is compared to the audit transfer standard response
for each test atmosphere, and the AT response is recorded at the same time.

The recommended tolerance for an audit is the same as that for a calibration—±0.5 degrees C for
temperature and ±0.1 degrees C for AT.

-------
                                                                                 Volume 4, Section 4
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
                                                                                       Page 1 of 6
                           4.   Rainfall and Precipitation
Precipitation is defined as, "the total amount of precipitation which reaches the ground in a stated period
is expressed as the depth to which it would cover a horizontal projection of the earth's surface if there
were no loss by evaporation or run-off and if any part of the precipitation falling as snow or ice were
melted".11 In any method of precipitation measurement, the goal should be to obtain a sample that is
representative  of the fall in the area. At the outset, it should be recognized that the extrapolation of
precipitation amounts from a single location to represent an entire region is an assumption that is
statistically questionable. A network of stations with a density suitable to an investigation is preferable.
4.1  Types of Instruments and Specifications

There are two basic types of precipitation collectors: non-recording and recording.


4.1.1 Non-recording Rain Gauges

In its simplest form, a precipitation gauge consists of a
cylinder, such as a can with straight sides, closed at one end
and open at the other. The depth of the liquid in the can is
measured with a measuring stick calibrated in subdivisions of
centimeters or inches (Figure 4.1).
To obtain greater resolution, an NWS-specified standard 8-
inch gauge is constructed with a ratio of 10:1 between the area
of the outside collector cylinder and the inside measuring tube.
The funnel attached to the collector both directs the
precipitation into the tube and minimizes evaporation loss.
Amounts in excess of two inches of rainfall overflow into the
outer can, and all liquid and melted precipitation
measurements are made in the measuring tube with a
measuring stick.

Figure 4.2 Automatic Wet/Dry Precipitation
Collector
 The automatic wet/dry precipitation collector
 (Figure 4.2), available in several designs,
 represents a specialized non-recording
 instrument designed for programs involving
 the chemical and/or radioactive analysis of
 precipitation. The collector is built with a
 sensor that detects the onset and cessation of
 precipitation and automatically releases a lid to
 open and cover the collector.  In one design,
the lid can be made to remain open during
either wet or dry periods. Another model is

-------
                                                                                Volume 4, Section 4
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
	Page 2 of 6
fabricated with two collectors; the lid is made to cover one bucket during periods of rain and snow. In
equipment designed for precipitation chemistry, the volume of water, in proportion to the constituents
collected with the water, is important, so evaporation must be kept to a minimum.
4.1.2 Recording Rain Gauges

The two basic designs of recording gauges—the weighing-type gauge (Figure 4.3) and the tipping bucket-
type gauge (Figure 4.4)—are determined by their operating principles The former, when made to NWS
Specification No. 450.2201, is the Universal Gauge, indicating use for both liquid and frozen
precipitation.  Options for the remote transmission of signals from this type of gauge are available. The
standard NWS Tipping Bucket Rain Gauge is designed with a 12-inch collector funnel that directs
precipitation to a small outlet directly over two equal compartments, or buckets, that tilt in sequence with
each 0.01  inches of rainfall. The motion of the buckets causes a mercury switch closure.  Normally
operated on 6 VDC, the contact closure can be monitored on a visual counter and/or by one of several
recorders. The digital-type impulse can also be used with computer-compatible equipment.

Some new automatic gauges that measure precipitation without moving parts are  available. These gauges
use devices such as capacitance probes, pressure transducers, and optical or small radar devices to provide
electronic signal that is proportional to the precipitation equivalent.
Figure 4.3 Universal-Weighing Gauge
Figure 4.4 Tipping Bucket without the Shield
4.1.3 Instrument Characteristics

The recording-type gauge records rainfall begin and end times and measures the rate of fall. The
universal-weighing gauge incorporates a chart drum that is made to rotate either by an eight-day spring-
wound clock or a battery-powered clock. Recent developments include a unit with a quartz crystal
mechanism and gear shafts for a wide range of rotation periods from a half-day to one month.

The weighing gauge is sometimes identified by the name of its designer (Fergusson) and comes with one
of two recording mechanisms. In the single traverse unit, the pen moves from the base of the drum to the
top, typically a water equivalent of 6 inches. In a dual traverse unit, the pen moves up and then down for

-------
                                                                                Volume 4, Section 4
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
                                                                                       Page 3 of 6
a total of 12 inches of precipitation.  A variation of the weighing gauge, a "high capacity" design with
dual traverse, will collect as much as 760 mm or 30  inches.

To minimize the oscillations incurred by the influence of strong winds on the balance mechanisms,
weighing gauges are fitted with a damper immersed in silicone fluid.  By incorporating a potentiometer in
the mechanism, the gauge is  capable of providing a resistance or, with another refinement, a voltage
proportional to the amount of precipitation collected. Linearity of response is usually a factory
adjustment involving the use of calibrated weight to simulate precipitation amounts. Despite
manufacturer's specifications, it is doubtful that the  gauge can resolve 0.01 inches, especially when the
bucket is nearly empty.

In the tipping-bucket gauge, the balance of the buckets and the leveling of the bucket frame are critical.
Low voltage at the gauge is imperative for reasons of safety. Power is typically 6 VDC. The signal is
provided by a switch closure each time the bucket assembly tips (0.01 inches of rainfall per bucket). Rain
rates are calculated from an event recorder with pens energized sequentially to improve resolution. The
tipping bucket (a mechanical device) takes time to tilt from one position to the next. When the rate of fall
is high, spillage occurs and the unmeasured precipitation falls into the reservoir.  When greater accuracy
is needed, the collected water is measured manually and excess amounts are allocated proportionately in
the record. The accuracy of the gauge is 1 percent for rainfall rates of 1 in/hr or less; 4 percent for rates of
3 in/hr; and 6 percent for rates up to 6 in/hr.
4.1.4 Accessories -Windshields and Heaters
Measurement accuracy for all types of gauges is influenced by exposure more than by variation in design.
Windshields represent an essential accessory to improve the catch of precipitation, especially snow under
windy conditions.  The improved Alter design, made of 32 free-swinging but separated leaves supported
1/2 inch above the level of the gauge's collecting orifice, is an effective way to improve the catch. In a
comparison of shielded and unshielded 8-inch gauges at a wind speed of 5 mph, the efficiency of the
unshielded gauge decreases by 25 percent, and at 10 mph, the efficiency of the gauge decreases by
40 percent.31

                                                       In below-freezing conditions when the catch
                                                       in a gauge is snow or some other form of
                                                       solid precipitation, the collector/funnel of
                                                       non-recording gauges and the funnel in
                                                       recording gauges must be removed. Some
                                                       instruments are available with built-in heater
                                                       elements that are thermostatically controlled.
                                                       An effective heater for conditions that are not
                                                       too severe is an incandescent lamp installed
                                                       in the housing of the gauge. Caution should
                                                       be exercised because too much  heat will
                                                       result in evaporative loss.
Figure 4.5 Example of an Alter wind shield
(source: http://www.rap.ucar.edu)

-------
                                                                                Volume 4, Section 4
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
                                                                                      Page 4 of 6
4.1.5 Precipitation Data Recommendations

In research studies, especially those related to acid rain, the instrument used most frequently is the
automatic precipitation collector with one or two collecting buckets and a cover to prevent evaporation.
In operational activities, the choice is the weighing gauge or the tipping-bucket gauge.  For climatological
surveys, the choice might include both recording and non-recording type gauges.  The use of a windshield
is recommended to minimize errors that result from windy conditions if the application requires
maximum accuracy. The precipitation measurement made at air quality monitoring stations is frequently
used for descriptive purposes or for episodic analysis.  If effort is required to achieve accuracy levels that
are greater than the manufacturer's specifications for electrical recording gauges, then a 10 percent
tolerance limit may be adequate.
4.1.6 Procurement

Purchasing a suitable precipitation measuring system requires specifying the type of system that fits the
data application and the accuracy consistent with that application. A variety of gauges are available
commercially. In general, NWS-specified standards result in the fewest problems. For example,
numerous 8-inch gauges are available, but those conforming to NWS specifications are made only of
brass and copper, are more durable, and are reported to rupture less frequently under extended freezing
conditions than those made of galvanized steel.

The procurement of a weighing- type gauge should include a tripod mounting base as well as a set of
calibration weights. For locations that are not readily accessible or locations with heavy precipitation, the
bucket of the weighing gauge should have an overflow tube. If time resolution is not important, drum-
type, recording rain gauges can be obtained with monthly rather than weekly mechanisms. The tipping -
bucket gauge must be equipped with a heater for use when precipitation is frozen.
4.2 Acceptance Testing

Except for visual inspection, non-recording gauges do not require acceptance testing.  The weighing
gauges should be assembled and given a quick "bench-top" calibration check with standard weights or a
measured volume of water. In addition, the clock mechanism supplied with the gauge should be checked
for at least a couple of days, preferably a week. The tipping bucket gauge should also be bench-tested,
primarily to be certain that the bucket mechanism assembly is balanced and that the switch is operational.
4.3  Calibration

Bench calibrations should follow the manufacturer's recommendation.  The electrical output gauge or the
drum recording gauge measures weight, whether total weight in the case of the weighing gauge or
increments of weight in the case of the tipping-bucket gauge. Density of water is assumed so the weight
can be expressed in units of volume or depth assuming the area of the collector opening. Calibrations of
the measurement apparatus can be based on the introduction of known volumes of water. The area of the
collection surface must be known so that the volume collected can be expressed as a depth. For example,
an 8- inch collector may feed a tipping bucket which tips when 7.95 cc of water has arrived.  If this
volume of water represents 0.01 inches of rainfall, the effective collection area must be 48.51 square
inches, using the following calculations:

-------
                                                                                Volume 4, Section 4
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
                                                                                      Page 5 of 6
                           7.95 cc = 0.485 in.3 = 0.01 in. * 48.51 in.2                          (4-1)

If the area is a circle, the diameter should be 7.86 inches.

                                  (48.5l/7i)1/2 - 3.93 in. radius                                 (4-2)

For rate-sensitive systems such as the tipping bucket, the rate of simulated precipitation should be kept
less than 1 inch per hour. Calibrations require properly leveled weighing systems (gauges) whether on
the bench or in the field.


4.4  Operation and Maintenance


4.4.1 Installation

The support, or base, of any gauge must be firmly anchored, preferably on a level surface so that the sides
of the gauge are vertical and the collector is horizontal.  The collector can be checked with a carpenter's
level placed at two intersecting positions.  The level of the bucket assembly on the tipping bucket gauge is
also critical and should be checked along its length and width.

The gauge should be shielded from the wind but not placed in an area where there will be excessive
turbulence caused by the shield. For example, a good location is an opening in an orchard or grove of
trees where the  wind speed near the ground is reduced by the canopy effect. A location open but for a
few trees would be less desirable because of strong eddies that can be caused by the trees.  Obstructions to
the wind should not be closer than two to four times the obstruction height from the instrument. In open
areas, a wind shield  such as that specified by the NWS should be used.  The ground surface around the
rain gauge may be natural vegetation or gravel.  It should not be paved so as to avoid splashing the gauge.
The gauge should be mounted a minimum of 30 cm (approximately 1 foot) above the ground and should
be high enough that it  will not be covered by snow.

After the weighing gauge is installed, the silicone fluid should be poured into the damping cylinder as
required. The hygroscopic ink-filled pen of the  drum recording type is inked to less than capacity because
the ink expands with increasing humidity and can easily spill over the chart.  The final calibration check
with standard weights  or suitable substitute should be made.

To check the operation of the tipping bucket, a known quantity of water equivalent to 10 tips should be
placed in a separatory  funnel. The separatory funnel is adjusted to allow the water to flow into the tipping
bucket at a rate  of 1  tip every 15 seconds. It may be necessary to adjust the set screws, which act as limits
to the travel of the tilting buckets. Adjustment is required if there is a 10 percent or greater error or if
greater accuracy is needed.


4.4.2 Field Operation of a Precipitation Measurement System

Calibration checks for weighing and tipping bucket gauges using the techniques described above are
recommended at six-month intervals.  Non-recording gauges, whether used alone or in a network, should
be read daily at a standard time. Although the weighing gauge is used for liquid and frozen precipitation,
it requires special attention during winter operations.  The funnel must be removed when snow is
expected, and the bucket must be charged with an antifreeze—24 oz of ethylene glycol mixed with 8 oz

-------
                                                                                Volume 4, Section 4
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
                                                                                      Page 6 of 6
of oil.  The weight of this mixture represents the baseline for which precipitation amounts are to be noted.
The bucket should be emptied and recharged when necessary, at about 5 inches in the universal gauge.
Antifreeze mixture classified as hazardous should be disposed of properly. All operational activities
should be recorded in the station log.
4.4.3 Preventive Maintenance

Possible leaks in the measuring tube or the overflow container of the gauge can be easily checked.  The
receptacles, partially filled with water colored with red ink, can be placed over a piece of newspaper.
This procedure is especially applicable to clear plastic 4-inch gauges which are more easily damaged.
Repairs can be performed by soldering an 8-inch gauge and by applying a solvent to the plastic gauge.

A number of pens, some with greater capacity than others, can be used with the universal gauge. All
gauges require occasional cleaning by a good soaking and wiping in a mixture of water and detergent.

The chart drive is another source of problems; but they can sometimes be avoided by lubricating the clock
drive for the environmental conditions expected. Keeping spare clocks in stock is good practice.

Routine visual checks of the performance of weighing-type gauges should be made every time there is a
chart change. The time and date of change and site location should be documented. Routine maintenance
should include inking the pen and winding the clock. Battery-powered chart drives will require periodic
replacement of batteries based on the manufacturer's recommendations. All preventive maintenance
activities should  be noted in the log book.
4.5 Auditing

Audits of precipitation measuring systems every six months are adequate.  The irregular occurrence of
precipitation makes the use of a certified transfer standard impractical.  The performance audit should
depend on challenging the gauge with amounts of water known to an accuracy of at least 1 percent of the
total used. This method determines the measurement system accuracy but not the collection efficiency of
the gauge in natural precipitation.  For tipping bucket gauges, a rate of less than one inch per hour should
be used and an amount which will cause a minimum of 10 tips. For weighing gauges, using calibration
weights to challenge the weighing mechanism is more convenient rather than using of the quantities of
water necessary for full-scale testing.

-------
                                                                              Volume 4, Section 5
                                                                                Revision No: 1.0
                                                                                  Date: 10/2006
                                                                                    Page 1 of4
            5.    Relative Humidity and Dew Point Determination
Of the many atmospheric variables describing water vapor content in the atmosphere, relative humidity is
the most common for routine monitoring programs. Relative humidity is the ratio (percent) of actual
vapor pressure of moist air to the saturation vapor pressure at the same temperature. Dew-point
temperature (or dew point) is the temperature to which a moist air parcel must be cooled to achieve
saturation over water at constant pressure and water vapor content.  The corresponding temperature with
respect to ice is the frost point.

Dew point measurement was more reliable than relative humidity measurement before the invention of
modern hygrometers.  Dew point measurement equipment is now more expensive and often requires more
electric power and routine maintenance than is practical for remote  stations.
5.1 Types of Instruments and Specifications

The discussion of relative humidity instrumentation is limited to equipment most frequently used for
routine environmental monitoring and/or standards used to test monitoring equipment. As with
temperature measurement, relative humidity instruments measuring in outside air must be protected from
solar and terrestrial radiation, precipitation, and wind influences.  Hence, relative humidity  sensors,
similar to temperature sensors, are typically mounted in naturally or mechanically aspirated shields.
Examples of these shields can be seen in Figures 3.1 and 3.2. Another examples is shown in Figure 5.1
below.  An example of a combination relative humidity/temperature probe is illustrated in Figure 5.2.
Figure 5.1 A motor aspirated RH shield
(source: http://www.voungusa.com/)
Figure 5.2 A typical RH probe
(source: http://www.metone.com/)
5.1.1 Electrical hygrometer
Advances in electronic manufacturing have provided the meteorological community with alternatives to
chilled-mirrors, wet-bulb thermometers, and wire-wound salt-coated bobbins. The resistance and
capacitance of thin hygroscopic films on modern hygrometers are affected by the presence of moisture.
Measurement circuits provide the instrument output with scaled voltages and readouts of atmospheric
moisture content.  Corresponding temperature measurement is included within the instruments to
calculate the results expressed in variables other than actual moisture content.

-------
                                                                                Volume 4, Section 5
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
	Page 2 of4
Subtle differences in the type of sensor can create measurement advantages for certain moisture content
levels.32 Capacitive and resistive sensors respond better to relative humidity than to dew point.
Specifications for both types of sensors are similar.  Capacitive sensors are most linear at low humidity
levels and can tolerate condensation, although calibration shifts can occur. Resistive sensors are most
linear at high humidity levels and cannot tolerate condensation, although some have automatic protection
from saturation conditions. Dew point impedance sensors use a slightly different element; they measure
absolute rather than relative humidity.  The sensors are covered by membranes that are readily porous to
moisture, although the membranes thermally insulate the sensor, causing some lag time in measurement.
Electrical hygrometers are considerably less expensive than some other automated relative humidity
measurement methods and are readily adaptable to portable, hand-held units suitable for temporary
measurements and transfer standard use.
5.1.2 Chilled mirror

Dew point (or frost point) can be measured directly using thermoelectric cooling and precise temperature
measurement and control. A mirror surface is cooled until dew (or frost) forms on the surface. The
temperature of the mirror surface is measured, and that measurement is the dew (frost) point temperature.
The engineering aspects of airflow, temperature control, and optical identification have been refined in
modern equipment. Optical identification improvements have reduced the occurrences of mistaking
contamination on the mirror for condensed moisture, but mirror cleaning remains a necessary activity for
reliable dew point measurement.

An excellent reference for chilled-mirror measurements of dew point and frost point is the ASTM
International test method, D4230.29 This standard includes analytic expressions for saturation vapor
pressure as functions of temperature and relative humidity, which can be used to convert between these
variables.
5.1.3 Psychrometer

The psychrometer contains two identical thermometers—dry-bulb and wet-bulb. Dry-bulb temperature is
air temperature.  Wet-bulb temperature is the temperature of an air parcel if cooled adiabatically (no
external heat transfer) at constant pressure to saturation by evaporation of water into the parcel.  The wet-
bulb thermometer has a small cotton cover on the thermometer's bulb; the cover is soaked in distilled
water and spun around, or otherwise ventilated, until the reading stabilizes at the wet-bulb temperature.
ASTM International standard E33733 covers psychrometer measurements and the associated calculations
of other humidity variables.  Engineers still use wet-bulb temperature for heating, ventilating, and air
conditioning calculations.
5.2 Acceptance Testing

After equipment is newly installed or returned from maintenance or calibration by vendors located
elsewhere, it is prudent to check the equipment for proper operation. Damage can occur in shipping and
handling that can affect the response. Simple checks for reasonable responses can be adequate initial
acceptance testing; precautions identified in the formal field check should be observed during testing.
The formal check is recommended during installation at the operating site.

-------
                                                                                Volume 4, Section 5
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
	Page 3 of4
5.3 Installation and Wiring

Relative humidity and dew point sensors should be mounted over a plot of open level ground at least 9 m
in diameter. The ground surface should be covered with non-irrigated or unwatered short grass or, in
areas where grass does not grow, natural earth.  The surface must not be concrete or asphalt or oil-soaked.
The standard height for climatological purposes is 1.25 m to 2 m, but required heights may frequently be
different in air quality studies.

The sensors should not be  closer to obstructions, such as trees and/or buildings, than a distance equal to
four times their height. They should be at least 30m from large paved areas and not close to steep slopes,
ridges, or hollows. Areas of standing water should also be avoided. Louvered instrument shelters should
be oriented so that the door opens toward true north in the northern hemisphere. Motor-aspirated shields
should also be oriented with the sensors toward true north in the northern hemisphere.

Proper planning assures that the mounting hardware,  cables, power supply, and so forth are all compatible
and available and helps an installation proceed without difficulty.  Purchasing equipment through a
vendor who provides everything the system needs to  readily attach cables can save considerable technical
time and effort of having to fabricate mounting hardware and prepare cables. Additional lightning
protection should be considered in areas where  lightning occurs.

Most relative humidity sensors do not require additional power beyond the excitation voltages in the
bridge circuit, although motor-aspirated temperature  shields require electrical power. Shields can be
powered by direct current  (DC) fans operating on batteries being recharged by solar panels or an
alternating current (AC) trickle charging unit. Operating the system on DC, even when AC power is
available, can reduce the missing data periods when electrical power is unavailable.

A complete  quality assurance plan will prescribe appropriate installation and testing procedures.  These
procedures are developed from manufacturers' specifications and guidance and instrument exposure
recommendations listed in monitoring guidance documents. For general purposes, the relative humidity
sensor is mounted 2 m above ground level, with the inlet facing away, and at a distance of approximately
1.5 times the tower diameter, from the tower. Influences from nearby artificial or natural  moisture
sources can  adversely influence relative humidity measurement so that it will not be representative of the
surrounding area.
5.4 Calibration

Calibrating a relative humidity measurement system consists of comparing the output of the device being
calibrated to a known value and determining if the difference is within acceptable tolerance limits.
Modern relative humidity measurement systems may include software calibration adjustment capability.
The quality assurance plan for the monitoring program should offer guidance about when to make
adjustments and when to leave an instrument in the as-found mode. An essential factor in obtaining two
comparable relative humidity measurements is that the sensors be reasonably close to the same
temperature.  The basic sensor measurement is molecular water vapor. Data displays in relative humidity
and dew point involve algorithms that include temperature, so two sensors at significantly different
temperatures would provide different output values for the same moisture exposure.

EPA guidance5 specifies a tolerance limit of the difference between known and observed values within
±1.5 degrees  C in terms of dew-point temperature. For relative humidity values less than about
40 percent, the acceptable dew point difference translates to a relative humidity value smaller than most

-------
                                                                                Volume 4, Section 5
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
                                                                                      Page 4 of4
instruments can provide. Hence, a two-tier system of an acceptable relative humidity of ±7 percent when
less than 40 percent and using the recommended dew point difference above that level can provide
consistent criteria across the range of relative humidity levels.

Calibration tests performed in the field using the full system for relative humidity measurement and signal
processing used during routine operation reduce testing uncertainty, although the trade-off can be the
difficulty in providing a stable humidity environment at a field site. The calibration test should be
performed at three or more humidity levels spaced across the range of the sensor within the range of the
environment producing the stable atmosphere.  The typical calibration ranges are 35 percent, 50 percent,
75 percent, and 90 percent, respectively.

ASTM International standard  E10434 describes methods to produce stable humidity levels using aqueous
salt solutions.  These solutions are sensitive to temperature, so reliable tests in an exposed location can be
difficult. Small commercial chambers capable of maintaining preset humidity levels can provide the
stable environment needed for calibration checks across a range of conditions.  The need for temperature
stability often necessitates using these chambers in a reasonably well-controlled environment.

The high accuracy and quality of hand-held sensors provides a readily available resource for field tests of
relative humidity systems, providing the tests can be made in reasonably stable atmospheric conditions.
5.5  Operation and Maintenance

Modern relative humidity measurement systems can be very reliable systems, but they still require regular
physical inspection and data-checking to ensure that the sensors are accurately measuring the relative
humidity of the air.  The radiation shields can become clogged with dirt, vegetation, or small animals, so
much so that airflow or electrical connections can be adversely affected. Changes to the airflow can be
gradual, masking the problem when typical software algorithms search for spurious results.  The
frequency of the checks should be based on the environment of the system and operating experience.
Checks of the shields typically are the same as those for the corresponding temperature measurement.
5.6 Auditing

Relative humidity measurements should be included in routine performance and system audits at six-
month intervals.  The performance audit can consist of a simple one-point check against a hand-held
relative humidity measuring standard; however, a more complete challenge of the relative humidity
sensor using standard salt solutions or a portable humidity chamber is recommended. Using an electric
cooler can create a sufficiently stable environment in which to conduct a three-point relative humidity
audit using standard salt solutions referenced to a NIST-traceable transfer standard.  Another option for
conducting the in situ relative humidity audit is to use a battery-powered, portable humidity generator to
produce multiple humidity ranges.

-------
                                                                                Volume 4, Section 6
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
                                                                                       Page 1 of 9
                              6.    Quality Assurance of
                          Solar Radiation Measurements
6.1  Introduction
Solar energy is the driving force behind large-scale atmospheric motion. Many air pollution specialists
consider the measurement of solar radiation secondary to wind and temperature measurements; however,
solar radiation is directly related to atmospheric stability. It is measured as total incoming global
radiation, as outgoing reflected and terrestrial radiation, and as net total radiation.35

Solar and/or net radiation data are used (1) to determine atmospheric stability for calculating various
surface-layer parameters,  (2) in dispersion modeling for estimating convective (daytime) mixing heights,
and (3) for modeling photochemical reactions.5  Solar radiation refers to the electromagnetic energy in the
solar spectrum (0.10 to 4.0 urn wavelength).  The solar spectrum comprises ultraviolet light (0.10 to 0.40
urn), visible light (0.40 to 0.73 um), and near-infrared radiation (0.73 to 4.0 urn). Net radiation includes
both solar radiation (also referred to as short-wave radiation) and terrestrial or long-wave radiation. The
sign of the net radiation indicates the direction of the flux (a negative value indicates a net upward flux of
energy).
6.2  Solar Radiation

The sun generates about 3.9xl026 Watts of energy.  This energy is radiated into space uniformly.
Radiation decreases as the inverse square of the distance from the Sun.  The solar constant (S0) is the
average energy per unit area of solar radiation falling on the surface of a sphere of radius R around the
Sun (see Equation 6-1).

                                 S0 = E/(47i R2) = 1370 W/m2                                (6-1)

Where  R     =   the distance between the Earth and Sun,-150,000,000 km
        E      =   Total Solar Energy 3. 9x 1026W
        W/m2 =   Watts/meter2

The solar "constant" actually fluctuates and the energy the planet receives varies with the seasonal change
in the Earth/Sun distance. If one astronomical  unit (AU) is the average Earth/Sun distance, then the
amount of solar radiation reaching Earth varies according to Equation 6-2.
                         Smax = S0/(l - e)2  = S0/(1.017)2 = 1417 W/m2                        (6-2)
                         Smin = S0/(l + e)2  = S0/(0.983)2 = 1324 W/m2
Where e is the eccentricity (a measure of departure from a circle) of Earth's orbit around the Sun.

Earth's eccentricity varies. The current value is about 0.017. The maximum and minimum values vary
slightly more than 3 percent from the mean. Earth is closest to the Sun in early January and receives the
maximum amount of radiation during this time. The minimum amount of radiation is received about six
months later.

-------
                                                                                  Volume 4, Section 6
                                                                                    Revision No: 1.0
                                                                                      Date: 10/2006
	Page 2 of 9

Visible, infrared (IR), and ultraviolet light (UV) and heat are important constituents of solar radiation.
The Sun's energy is distributed over a broad range of the electromagnetic spectrum. It behaves
approximately like a "blackbody" radiating at a temperature of about 5,800 degrees Kelvin. Its  maximum
output is in the green-yellow part of the visible spectrum.  Figure 6.1 illustrates the energy versus
wavelength of light emitted by our Sun.
                2500
                                                   1.S         2
                                               wavelength (|jm)
                                                                         2.5
                          Spectrum AMQ (extraterrestrial)
Spectrum AM 1.5 (terrestrial)
            Figure 6.1 Solar Irradiance Versus Wavelength of Light Emitted by the Sun
Radiation is not emitted by the Sun in a uniform manner. Irregularities result from processes in the Sun's
interior and on its surface.

Quantitatively, solar radiation is described in units of energy flux, usually W/m2.  When measured in
specific, narrow wavelength bands, solar radiation may be used to evaluate such air pollution indicators as
turbulence and an indicator of photochemical processes. This section describes instruments that measure
broadband radiation and sunshine duration.  Specifications, acceptance testing, installation, calibration,
operations/maintenance, and auditing procedures are described for the different instrument types.
6.3 Types of Instruments

Instruments used to measure the transmission of sunlight through Earth's atmosphere fall into two
categories:  instruments that measure radiation from the entire sky (pyranometers) and instruments that
measure only direct solar radiation (pyrheliometers). For each instrument category there are two
measurement methods: thermal (i.e., thermopiles) and photovoltaic detectors.
6.3.1 Pyranometers

Pyranometers are instruments that measure solar radiation received from a hemispherical section of the
atmosphere.  A pyranometer measures solar radiation, including the total Sun and sky shortwave radiation
on a horizontal surface.  Pyranometers that measure net total radiation are termed net radiometers.  Most
pyranometers incorporate a thermopile as a sensor.  Some use a silicon photovoltaic cell as a sensor. The

-------
                                                                               Volume 4, Section 6
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
                                                                                      Page 3 of 9
precision spectral pyranometer (PSP) is made by Eppley Laboratories (see Figure 6.2) and has two
hemispherical domes designed to measure Sun and sky radiation on a horizontal surface in defined
wavelengths.
                 Figure 6.2 Eppley Pyranometer PSP

The Eppley Model PSP pyranometer is a widely used "first class" reference instrument as defined by the
World Meteorological Organization.  This instrument is about 15 cm in diameter. The sensor is under the
hemispherical glass dome. The glass is specially formulated to transmit solar radiation over a wide range
of wavelengths. Figure 6.3 is an illustration of the Epply pyranometer.
                                                1 60.O-
                               outcr
                               dom

                               ski
                     adjust-in
                     thread
                               vent  hoi
                                                                Kermopile
                                                         compensation
                                                         thermistor
                Figure 6.3 Illustration of an Epply Pyranometer

-------
                                                                                  Volume 4, Section 6
                                                                                    Revision No: 1.0
                                                                                      Date: 10/2006
	Page 4 of 9

Because pyranometers measure solar radiation from the sky, it is imperative that monitoring sites have a
360° view of the horizon, without significant obstacles.  Corrections can be made for some obstructions
but, the clearer the horizon is, the more accurate measurements will be.

A less obvious recommendation is that pyranometers have excellent "cosine response" to direct sunlight.
If sunlight has intensity I0 when the Sun is directly above a horizontal surface (zenith angle of 0°), then
the intensity Iz at some other zenith angle z is a function of the angle (Equation 6-3).
                                         Iz = I0cos(z)                                       (6-3)

If an ideal detector on a horizontal surface is illuminated by direct light, then its response should be
proportional to the cosine of the zenith angle of the light source.

Pyranometers usually do  not have perfect cosine response. Cosine response corrections can be
determined and applied for a direct light source, but this issue becomes much more complicated under
partly cloudy skies, when the radiation incident on a detector is an unknown combination of direct
sunlight and diffuse sky radiation, as is the case  for full-sky solar radiation.

High-quality reference pyranometers, such as the Eppley pyranometer shown in Figure 6.3, use
thermopiles, which are collections of thermocouples.  Thermocouples consist of dissimilar metals placed
together or joined. At the interface of the thermocouple are two dissimilar metals with different
electronic valence configurations. They produce a small current proportional to their temperature. When
thermopiles are appropriately arranged and coated with a dull black finish, they serve as nearly perfect
"black body" detectors that absorb energy across the range of the electromagnetic spectrum. Ideally, the
response of the thermopile sensor in the pyranometer is proportional to the cosine of the angle of the solar
beam and is constant at all azimuth angles. This characteristic is known as the Lambert Cosine Response,
an important characteristic of pyranometers.

Most net radiometers now available commercially are made with a small disc-shaped thermopile covered
by polyethylene hemispheres.  In most units the material used for shielding the element from the wind and
weather is very thin and is transparent to wavelengths of 0.3 to 60 (im.  Until recently, the internal
ventilation and positive pressure required to maintain the shape of the hemispheres of net radiometers was
considered critical; however, new designs have eliminated this problem.

The LiCor pyranometer (Figure 6.4) is a popular thermopile radiometer with a silicon photovoltaic
detector mounted in a fully cosine-corrected miniature head. The current output is directly proportional to
solar radiation.

The NovaLynx pyranometer (Figure  6.5) operates on the principle of temperature difference created by
light absorption of light material (white) and dark material (blank) when exposed to solar radiation.  The
temperature difference is proportional to the radiation intensity.

-------
                                                                                Volume 4, Section 6
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
                                                                                      Page 5 of 9
Figure 6.4 LiCor Pyranometer
Figure 6.5 NovaLynx Pyranometer
6.3.2 Pyrheliometers

A pyrheliometer is an instrument that measures the intensity of direct solar radiation at normal incidence.
In other words, it measures the direct radiation from the Sun, not total or incident solar radiation.
Pyrheliometers work on the same physical principles as pyranometers, i.e., using thermopiles to create a
current that can be measured by an electronic circuit.
A number of different vendors manufacture pyrheliometers.  Perhaps the most common pyrheliometer is
the Epply Normal Incidence Pyrheliometer (Figure 6.6).
A pyrheliometer is mounted in a solar tracker, or equatorial mount, automatically tracking the Sun as it
moves across the sky.  In contrast, a pyranometer is mounted facing the zenith (i.e., facing a point on the
celestial sphere directly above the observer). Figure 6.7 shows an example of a solar tracker.  The solar
tracker must be placed in a location that has line of sight of both horizons.

 Figure 6.6 Epply Normal Incidence
 Pyrheliometer
             Figure 6.7 Epply Solar Tracker

-------
                                                                                Volume 4, Section 6
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
	Page 6 of 9


6.4 Specifications

When purchasing a solar radiation measurement system, match the data recommendations to the
instrument selection. Refer to Tables 0-1 through 0-9 to match the sensor performance with the type of
sensor needed for your circumstance. The measurement quality objectives and calibration/auditing
recommendations are detailed in those tables. Specify the required performance on the purchase order.
Be sure to note which test method you will use to verify performance and test the instrument after receipt.

Class 2  sensors (as defined by the World Meteorological Organization) offer the advantage  of providing
data comparable to those collected at National Weather Service stations and at key Department of Energy
(DOE) locations. Specified sensors should be commercially available and meet the technical
recommendations established by the measurement quality objectives detailed in Tables 0-1 through 0-9.
American Society for Testing and Materials (ASTM) standards are available.22 When purchasing a
recorder or data acquisition system (DAS), your agency should match the calibration factor or sensitivity
of the sensor to the readout equipment. Note that the signals from pyrheliometers (in contrast to
pyranometers) require zero-offset capability to accommodate both negative and positive voltage outputs.
6.5 Acceptance Testing

Physical inspection of the relatively fragile pyranometers and pyrheliometers should be done immediately
after delivery of an instrument. Upon delivery, a pyranometer or pyrheliometer should be accompanied
by a calibration certificate that states that the instrument has been calibrated to a NIST-traceable
radiometer.  Be sure that the calibration data have been received and that these data correspond to the
serial number of the instrument. Storage of calibration information at the main office and in the field will
prove helpful when instrument calibration needs to be checked. A quick determination can be made
indoors as to whether the sensor is operating by exposing the sensor to the light of a tungsten lamp. It
may be necessary to place the instrument fairly close to the lamp.  Covering the sensor for several hours
will ensure that the system is not "dark counting". Zero response confirms that the sensor baseline
response is acceptable and the sensor can be used to collect data. If the sensor response is greater than
zero, the sensor should not be used to collect data and the manufacturer should be contacted.
6.6 Installation, Instrument Exposure, and Wiring

The site selected for an upward-looking pyranometer should be free from any obstruction above the plain
of the sensor and should allow easy access for cleaning and maintaining the instrument.  It should be
located so that shadows will not be cast on the device and away from light-colored walls or other objects
likely to reflect sunlight. A flat roof is usually a good choice; but if such a site is not possible, a rigid
stand with a horizontal surface some distance from buildings or other obstructions should be used. A site
survey of the angular elevation above the plane of the radiometer surface should be made through
360 degrees.

The same procedures and precautions should be followed for net radiometers that are both upward- and
downward-facing.  Figure 6.8 shows a net radiometer.  However, the net radiometer must be supported on
an arm extending from a vertical support about 1  m above the ground. Except for net radiometers with
heavy-duty domes, which are installed with a desiccant tube in series with the sensor chamber, most other
hemispherical net radiometers require the positive pressure of a gas, usually nitrogen, to both maintain the
shape of the polyethylene domes and purge the area surrounding the thermopile. Some of the more

-------
                                                                                 Volume 4, Section 6
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
                                                                                       Page 7 of 9
popular net radiometers incorporate internal purging with nitrogen and external ventilation with
compressed dry air through holes on the frame.  The compressed air supply minimizes fogging and
condensation.
        Figure 6.8. Net Radiometer

Precautions must be taken to avoid subjecting net radiometers to mechanical shock during installation.
They should be installed securely and leveled using the circular spirit level attached to the instrument.
Net radiometers are difficult to mount and to maintain free of vibration.  Pyranometers of the Moll-
Gorzynski design, manufactured by Kipp & Zonen (www.kippzonen.com), are oriented so that the
emerging leads face north.  This minimizes solar heat on the electrical connections of an instrument that is
not temperature compensated.  The thermopiles of these instruments should be oriented so that the long
side of the thermopile points east and west.36  The cable used to connect the pyranometer to the readout
device, recorder, or integrator should be between 16 and 20 gauge and made of shielded, waterproofed
2-conductor copper wire. The sensor, shield, and readout device should be connected to a common
ground. Potentiometric milivolt recorders are to be used with most high-impedance, low-signal
radiometers. Cable lengths of 300 m or more are practical.
6.7 Calibration
                                should
Pyranometers and net radiometers should be subjected to field calibration checks on two consecutive
cloudless days.  These checks involve a side-by-side comparison of the on-site reporting sensor to a
transfer standard sensor of similar design (it is recommended that it be the same make to eliminate any
bias). The transfer standard sensor must have a NIST-traceable calibration within a year of the date of the
calibration. If a side-by-side calibration is not possible, the device must be returned to the manufacturer
or to a laboratory that has facilities to check the calibration.  Pyranometers and net radiometers should be
calibrated once every six months. Any indication of discoloration or peeling of a blackened surface or of
scratches on the hemispheres of a pyranometer warrants recalibration and/or service.

Calibrating the recorder or integrator is an easy task.  The standard method involves the use of a precision
potentiometer to impress known voltages into the circuit.  The linearity of the readout instrument may be
checked by introducing a series of voltages covering the full scale, checking first up-scale and then down-
scale. Adjustments should be made as necessary. In the absence of a precision potentiometer, it may be
possible to introduce a calibrated millivolt source capable of checking the up-scale and down-scale
responses of the recorder. Integrators can be checked the  same way, except that the input value must also
be timed.

-------
                                                                                 Volume 4, Section 6
                                                                                   Revision No: 1.0
                                                                                      Date: 10/2006
	Page 8 of 9

The self-calibrating Absolute Cavity Pyrheliometer (Figure 6.9), Model HF, has been a reference  standard
level device for many years. The sensor consists of a balanced cavity receiver pair attached to a circular
wire-wound and plated thermopile. The blackened cavity receivers are fitted with heater windings which
allow for absolute operation using the electrical substitution method, which relates radiant power to
electrical power in international system (SI) units.  The forward cavity views the direct beam through a
precision aperture. The precision aperture area is nominally 50 mm5 and is measured for each unit.  The
rear receiver views an ambient temperature blackbody. The Model HF radiometer element with baffle
tube and blackbody fit into an outer tube which acts as the  enclosure of the instrument. The Model AHF
has an automatic shutter attached to the outer tube.
                           Figure 6.9 Absolute Cavity Pyrheliometer

The operation of the cavity radiometer and the measurement of the required parameters are performed
using an appropriate control box. The control functions include setting of the calibration heater power
level, activation of the calibration heater, selection of the signals to be measured, and control of the meter
measurement functions and ranges. The measured parameters include the thermopile signal, the heater
voltage, and the heater current which is measured as the voltage drops across a 10-Ohms precision
resistor.  The instrument temperature may also be measured using an internally mounted thermistor. The
meter resolution of 100 mV allows for a thermopile signal equivalent in radiation to approximately 0.1
W/m2.
6.8 Operations and Maintenance

As part of the quality assurance program, a field calibration check of the solar radiation sensor should be
performed at least once every six months according to the procedures outlined in Tables 0-1 through 0-9.
The data should be inspected for a reasonable diurnal pattern and the absence of dark counting. Where
strip chart or digital printers are used, daily time checks are desirable to confirm proper time sequence of
the chart or printer.  Frequency of data retrieval will depend upon program objectives; but even for
climatological programs, data should be collected monthly. All operational activities during a site visit
should be logged.

-------
                                                                                 Volume 4, Section 6
                                                                                    Revision No: 1.0
                                                                                      Date: 10/2006
                                                                                       Page 9 of 9
6.8.1 Preventive Maintenance
All types of radiometers require frequent cleaning to remove any material deposited on the surface that
may intercept the radiation. Ideally, this operation is daily.  The outer hemisphere should be wiped clean
and dry with a lint-free soft cloth and alcohol.  Any scratching of the surface will alter the transmission
properties of the glass, so cleaning must be done with care.  If frozen snow, glazed ice, hoarfrost, or rime
ice is present, an attempt should be made to remove the deposit carefully with warmed cloths.

Should the internal surface of a pyranometer's outer hemisphere become coated with moisture, it can be
cleaned by carefully removing the outer hemisphere on a dry day and allowing the air to evaporate the
moisture, then checking the desiccant.  If removal of a hemisphere exposes the thermopile element,
extreme care should be taken because it is fragile and easily damaged.  About once each month, the
desiccant installed in most pyranometers should be inspected. Whenever the silica gel drying agent is
pink or white instead of blue, it should be replaced or rejuvenated by drying it out on a pan in a
135 degrees C oven.  The level alignment should be checked after each servicing of the pyranometer, or at
least monthly.  Significant errors can result from misalignment.

Pyrheliometers require maintenance more frequently than pyranometers. It is necessary to replace the
polyethylene domes in pyrheliometers  as often as twice a year or more before the domes become
discolored, distorted, or cracked. More frequent replacement is necessary in polluted environments due to
accelerated degradation of plastic hemispheres when exposed to pollutants. A daily maintenance
schedule is essential to check on the proper flow of gas in instruments that are inflated and purged  with
nitrogen. All maintenance activities should be recorded.
6.9 Auditing

Installation of a certified transfer standard (CTS) is the only practical means of conducting a performance
audit on a solar radiation system. The CTS must have the spectral response and exposure equivalent to
the on-site sensor being audited. One diurnal cycle will establish an estimate of accuracy sufficient for
most air quality monitoring applications. If one diurnal cycle is not possible, the audit should be
conducted several hours prior to and after the peak solar radiation at the time of the audit. The CTS and
the on-site solar radiation sensor should be covered to determine the zero response of each instrument.  If
the meteorological site is equipped with a DAS, the CTS should be interfaced with a spare channel and
the DAS initialized to represent the accurate full scale and zero values of the CTS. Data from the CTS
and the on-site solar radiation sensor should be reported as daily integrated values, hourly integrated
values, and average intensity per hour to provide a meaningful comparison.  An audit frequency of at least
six months is recommended.

-------
                                                                                Volume 4, Section 7
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
                                                                                      Page 1 of 5
      7.    Quality Assurance for Atmospheric Pressure Measurements

The atmospheric pressure on a given surface is the force per unit area exerted by virtue of the weight of
the atmosphere above. The pressure is thus equal to the weight of a vertical column of air above a
horizontal projection of the surface, extending to the outer limit of the atmosphere.37
7.1 Units and Scales

The basic unit for atmospheric pressure measurements is the Pascal (Pa).  It is accepted practice to add the
prefix "hector" to this unit when reporting pressure, making the hectopascal (hPa), equal to 100 Pa, the
preferred terminology. This is largely because one hPa equals one millibar (mb), the formerly used unit.

The scales of all barometers used to measure atmospheric pressure should be graduated to hPa.  Some
barometers are graduated in millimeters (mm Hg) or inches of mercury under standard conditions (in.
Hg). Under these standard conditions (0 degrees C, 760 mm Hg), a column of mercury having a true
scale height of 760 mm Hg exerts a pressure of 1,013.250 hPa.

The following conversion factors apply:
        1 hPa = 0.750 mm Hg
        lin. Hg = 33.863hPa
        1 mm Hg = 0.039 in. Hg

Where 1 in. = 25.4 mm, the following conversion factors are obtained:
        1 hPa = 0.029 in. Hg
        lin. Hg = 33.863hPa
        1 mm Hg = 0.039 in. Hg
7.2 Types of Instrumentation

For air quality and meteorological purposes, atmospheric pressure is generally measured with mercury,
aneroid, or electronic barometers. Most, if not all of the atmospheric pressure sensors available provide
analog or serial output that is directly interfaced with a data acquisition system.

A mercury barometer measures the height of a column of mercury that is supported by the atmospheric
pressure. It is a standard instrument for many climatological observation stations, but it does not afford
automated data recording.

An aneroid barometer consists of two circular disks bounding an evacuated volume. As the pressure
changes, the disks flex, changing their relative spacing which is sensed by  a mechanical or electrical
element and transmitted to a transducer.

Most electronic barometers of recent design use transducers which transform the sensor response into a
pressure-related electrical quantity in the form of either analog or digital signals.  Current digital
barometer technology employs various levels of redundancy to achieve long-term stability and accuracy
of the measurements. One technique is to use three independently operating sensors under centralized
microprocessor control. Even higher stability and reliability can be  achieved by using three completely
independent barometers, incorporating three sets of pressure transducers and microprocessors. Each

-------
                                                                               Volume 4, Section 7
                                                                                  Revision No: 1.0
                                                                                   Date: 10/2006
                                                                                     Page 2 of 5
configuration has automatic temperature compensation from internal-mounted temperature sensors.
Triple redundancy ensures excellent long-term stability and measurement accuracy, even in the most
demanding applications.37 Figure 7.1 depicts an electronic barometer with three independent transducers.
                            Figure 7.1 Electronic Barometer (source:
      http://www.vaisala.com/businessareas/instruments/products/barometricpressure/ptb220)
7.3 Acceptance Testing

New barometers and barometers that have been sent to vendors for maintenance or calibration should be
checked for proper operation upon receipt. Damage can occur during shipping and handling that could
affect response. The barometric pressure reading from a new or repaired barometer should be compared
to the reading from a CTS barometer such as a portable electronic barometer. Figure 7.2 shows a portable
digital barometer.  To ensure proper operation of a station barometer, multiple pressure readings from the
station barometer and the CTS barometer should be compared over a period of several days. The readings
should be made with both barometers at the same height and in similar environmental conditions.  An
electronic barometer with a mean difference from the CTS that exceeds 0.25 hPa should be regarded as
unserviceable and returned to the calibration facility for recalibration.
                         Figure 7.2 Portable Digital Barometer (source:
    http://www.vaisala.com/businessareas/instruments/products/barometricpressure/ptb220ts)

-------
                                                                                 Volume 4, Section 7
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
                                                                                       Page 3 of 5

7.4 Installation and Instrument Exposure

The location of a barometer should be carefully considered in order for the equipment to accurately
measure atmospheric pressure. A barometer should be placed in a location
   >   That has uniform, constant temperature
   >   That has good general lighting but is shielded from direct sunshine
   >   That is away from drafts and heaters
   >   Where it will have a solid, vertical mounting
   >   Where it will be protected against rough handling

Wind can cause dynamic changes in air pressure, therefore causing barometric readings to be inaccurate.
Fluctuations from wind are superimposed on the static pressure and, with strong and gusty wind, may
amount to 2 or 3 hPa.  It is usually impractical to correct for such fluctuations because the "pumping"
effect on the mercury surface is dependent on both the direction and force of the wind, as well as on the
barometer's location. Thus the "mean value" will not represent the true static pressure. More information
on wind effects is found in Liu and Darkow.38

It is possible to overcome the effect of wind to a very large extent by inserting a static head between the
exterior atmosphere and the inlet port of the sensor. Details concerning the principles of operation of
static heads can be found in several publications.39'40 The cistern of a mercury barometer must be made
airtight except for a lead to a special head exposed to the atmosphere and designed to ensure that the
pressure inside is true static pressure.  Aneroid and electronic barometers usually have  simple connections
to allow for the use of a static  head which should be located in an open environment not affected by the
proximity of buildings.

Air conditioning may create a significant pressure differential between the inside and outside of a room.
Therefore, if a barometer is to be installed in an air conditioned room, it is advisable to use a static head
with the barometer that will couple the barometer to the air outside the building.

Figure 7.3 shows a small vented environmental enclosure (NEMA 4X) for applications where another
suitable shelter is not available. Figure 7.4  shows a tower-mounted barometer with a pressure port to
minimize dynamic pressure errors  caused by wind.

-------
                                                                               Volume 4, Section 7
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
                                                                                      Page 4 of 5
                            Figure 7.3 NEMA 4X Enclosure (source:
                 http://www.climatronics.com/pdf/products/sensors/102270.pdf)
   Figure 7.4 Tower-Mounted Barometer with Pressure Port (source: http://www.youngusa.com)
7.5 Calibration

Electronic barometers should be returned to a calibration facility annually for calibration. Upon receipt of
a barometer at a meteorological station a comparison test should be run.  Pressure readings from an
electronic barometer should be compared to pressure readings from a CTS over a period of several days.
The readings should be taken with both barometers at the same height, when the wind is less than 12ms-
1, and when the pressure is either steady or changing by less than 1 hPa.  An electronic barometer with a

-------
                                                                                Volume 4, Section 7
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
                                                                                          : 5 of 5
mean difference from the CTS that exceeds 0.5 hPa should be regarded as unserviceable and returned to
the calibration facility for recalibration.

Every six months, readings from an electronic barometer collected over several consecutive hours should
be compared to readings from a CTS under similar circumstances; a mean difference should be
established. If the mean difference is more than 3 mb, the station barometer should be returned to the
manufacturer for calibration.
7.6  Operation and Maintenance

A barometric sensor should meet the specifications listed in the MQO tables in Section 0. The minimum
reporting resolution should be 0.1 mb. The data should be at least hourly averaged referenced to local
standard time representing the actual hour the data were recorded. If the hourly average does not
represent the actual hour, then the data need to be flagged and noted so later comparisons will be accurate.

Routine maintenance procedures should include physical integrity checks of NEMA 4 enclosures to
ensure proper ventilation. Signal cables should be in good condition. Indoor sensors should be dusted to
prevent dust accumulation on the sensors.
7.7 Auditing

Performance audits should be conducted every six months.  A performance audit should entail a
comparison of atmospheric pressure sensor readings to a CTS. The elevation settings on the CTS and the
sensor should be equivalent to eliminate elevation bias.  Pressure readings should be compared once per
hour for the duration of time the auditor is on site and a mean difference should be calculated.  Audit
result acceptance criteria for pass-fail should be <3 mb.

-------
                                                                              Volume 4, Section 8
                                                                                 Revision No: 1.0
                                                                                   Date: 10/2006
	Page 1 of26
                    8.    Quality Assurance for Ground-Based
                              Remote Sensing Devices

Over the past few years, developments in remote sensing technology have made it possible to obtain
three-dimensional wind velocity (u, v, w) and virtual air temperature (Tv) profiles with the precision and
accuracy suitable for regulatory applications. Three types of commercially available remote sensors exist:
Sodar (Sound Detection And Ranging), which uses acoustic pulses to measure horizontal and vertical
wind profiles; radar (Radio Detection And Ranging), which uses electromagnetic (EM) pulses to measure
horizontal and vertical winds; and RASS (Radio Acoustic Sounding System), which uses both acoustic
and EM waves to measure Tv.  These remote sensors can also provide estimates of the height of the mixed
layer and elevated inversions by measuring the parameters listed here. Detailed descriptions of these
instruments are included in this section.

Wind. Upper-air wind speeds and wind directions are vector-averaged measurements. None of the
measurement systems described in the following sections provides a means to measure winds as scalar
quantities, as is done with cup and vane sensors mounted on an instrumented tower. The vertical beam of
the remote  sensor can measure vertical velocity. Upper-air wind data comprise either point measurements
(radiosondes) or volume averages (remote sensors). The altitude at which the winds are reported is
assumed to be the midpoint of the layer over which the winds are averaged. Averaging periods for upper-
air wind data also vary depending on the instrument system  used. The averaging interval for winds
measured by Sodars and radar profilers is typically 15-60 minutes.

Virtual Temperature. Upper-air temperature measurements are most commonly obtained using
National Weather Service (NWS) radiosonde sounding systems. Radiosonde temperature measurements
are point measurements.  RASS measures the Tv of the air rather than the dry-bulb temperature (T). The
Tv of an air parcel is the temperature that dry air would have if its pressure and  density were equal to
those of a parcel of moist air, and thus Tv is always higher than the dry-bulb temperature. Under hot and
humid conditions, the difference between Tvand T is  usually on the order of a few (2-3) degrees C; at low
humidity, differences between Tvand T are small. Given representative moisture and pressure profiles,
temperature can be estimated from the Tv measurements. RASS temperature measurements are volume
averages with a vertical resolution comparable to that of the wind measurements reported by the remote
sensing systems (i.e., 60-100 m).

Mixing Height.  For the purposes of this guidance, mixing height is defined as the height of the layer
adjacent to the ground over which an emitted or entrained inert non-buoyant tracer will be mixed (by
turbulence) within a time scale of about one hour or less.41  Mixing heights can be estimated using
reflectivity profiles from the radar wind profiler and Sodars. In addition, RASS Tv profiles can be used to
estimate mixing heights using the Holzworth method.42 An in-depth discussion of mixing heights from
remote sensors can be found in.5

Turbulence. Some Sodars report wind turbulence parameters. In using these parameters, one must
remember that Sodars measure the vector components of the wind. Furthermore, there may be significant
differences in time and space between the sampling of the components so that any derived variables using
more than one component may be affected by  aliasing. Thus, the derived turbulence parameters from
Sodars are generally not the same parameters that models expect for input. Numerous studies have been
performed comparing Sodar-based turbulence  statistics with tower-based turbulence statistics. Findings
from these  studies have generally shown that measurements of the standard deviation of the vertical
component of the wind speed (aw) are in reasonable agreement, while the standard deviation calculations
incorporating more than one component (e.g.,  Oe) are not.43

-------
                                                                              Volume 4, Section 8
                                                                                 Revision No: 1.0
                                                                                   Date: 10/2006
                                                                                    Page 2 of26
8.1 Types of Instruments and Specifications

Meteorological remote sensing devices provide measurements without disturbing the environment. In
addition, remote sensing measurements are not restricted to a given height as are in situ and tower-based
measurements. More importantly, data obtained from a remote sensor is represented as a spatial, or more
specifically, a volume average as shown in Figure 8-1. This is a significant difference from the in situ
measurements, which are measured directly.  This difference has significant implications for calibrations
and audits of upper-air measurement systems in Subsections 8.4 and 8.6.
         10000m

         1000m/
100m/.
           10m/
                                                                               Volume
                                                                               Average
                   Tower
                    Tall tower
Tethersonde
Remote Sensor
Figure 8.1 Schematic Showing the Differences Between In-Situ (Point) and Remote Senor (Volume)
Measurements

Ground-based meteorological remote sensors have been designed to measure vertical profiles of wind
velocity and Tv. The development and evolution of these devices over the last several decades have
followed two  similar but distinct paths: one based on acoustics and the other on EM radiation. Wind
velocities acquired by Sodar are based on the atmospheric effects on the propagation of acoustic energy,
while radars are based on the atmospheric effects on the propagation of electromagnetic energy. Profiles
of Tv are obtained by PvASS, which combines acoustic and EM technologies. Table 8-1 provides a
summary of typical specifications for the three major types of meteorological remote sensing devices.

-------
                                                                                   Volume 4, Section 8
                                                                                      Revision No: 1.0
                                                                                        Date: 10/2006
                                                                                         Page 3 of26
        Table 8-1 Typical Specifications for Meteorological Remote Sensors
Specification37
Parameters measured :
Frequency
Minimum height (m) 2
Maximum height (m) 2
Vertical resolution (m)
Mini-Sodar
u,v, w, Zi
3-5 KHz
5-15
100-300
5-20
Sodar
u,v, w, Zi
1-3 KHz
10-30
200-2000
5-100
Radar Wind
Profiler
u,v, w, Zi
915 MHz
90-120
1500-4000
60-100
RASS
TV
2 KHz
(sound)3
915 MHz
(radar)3
90-120
500-1500
60-100
        1 u, v, w are the three components of wind; Z; is the height of the elevated inversion layer; and Tv is virtual air
        temperature.
        2 Actual altitude coverage will depend on instrument condition and configuration, atmospheric conditions, and siting
        characteristics.
        3 RASS requires both sound and radar technologies.  Thus you can add a RASS system to a Sodar by adding a radar, or
        add RASS to a radar wind profiler by adding a sound source.

The general components and theory of operation with Sodars and radar wind profilers are very similar.
These systems have a transmitter to emit the signal, an antenna for transmitting signals, and a receiver to
detect a returned signal, and system electronics and software to control the remote sensor.  As shown in
Figure 8.2, these remote sensors operate by transmitting a signal (sound for Sodar and EM for radar wind
profilers) at a known frequency.  The signal is sent to the antenna and transmitted upward where it is then
scattered by the atmosphere. A small portion of the transmitted signal is scattered back toward the
antenna (called backscattering). The receiver measures three properties of the returned signal: (1) the
arrival time of the signal which indicates the height (i.e., range), (2) the strength of the backscattered
signal, and (3) the Doppler shift, which is the frequency difference between the transmitted and received
signals and is directly related to the velocity along the transmitted direction (i.e., beam).

                                               RHlprtnr
                                                Tjrqct
                                    \
                      \
                  \
         Figure 8.2 Schematic Showing the Transmitted and Received Signals From Sodars, Radar
         Wind Profilers, and RASS

-------
                                                                                Volume 4, Section 8
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
                                                                                      Page 4 of26
A single transmitted pulse typically results in very weak backscattering; thus, many transmitted pulses are
required to detect an atmospheric signal.  Sometimes periodic clutter, or interferences by sources such as
bugs, birds, or low-flying aircraft, etc., can bias the return signal.  These erroneous values are removed by
sophisticated algorithms and the remaining data is averaged to provide a profile of measurements. These
averages are usually computed for time periods of 15 minutes to 1 hour, depending upon the data
recommendations of a particular study.

To compute the wind speed and wind direction, these remote sensors transmit signals along a vertical
beam and two  oblique beams (off vertical by 12° to 30°) (Figure 8-3).  To create these separate beams,
individual antennas are physically tilted in different directions or a phase-array antenna electronically
creates vertical and oblique beams. The beam width typically ranges from 2 degrees to 15 degrees. Note
that these beams are not perfectly formed as shown in the simplistic schematic, but they include side
lobes, or weaker beams, at other angles (Figure 8-4). Sometimes  signals returned from the side lobes can
bias the measurements.
                                             Zenith e.
                 Figure 8.3 Schematic Showing the Vertical and Oblique Beams (Vaisala)

-------
                                                                                Volume 4, Section 8
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
                                                                                      Page 5 of26
          Figure 8.4 Schematic Showing a Beam Pattern for an Oblique Beam and its
          Associated Side Lobes (Vaisala)

The beams are typically, but not necessarily, oriented at right angles to one another. Ideally, one is
directed toward the East or West so that the u component of the horizontal wind velocity can be
determined while the other is directed toward the North or South for the v component. The actual
orientation of the beams can be toward any direction and is typically decided based on site-specific
factors. The mean horizontal and vertical wind velocity components (u, v, and w) can be computed using
the following equations:
                u=  -
                v= -
                                                                                    (8-1)
                                                                                    (8-2)
                                  w = -V
                                                                                            (8-3)
where oblique beam 1 has a zenith angle 6x and azimuth 0*; oblique beam 2 has a zenith angle dy and
azimuth 0y, and are at right angles to each other; and Vx, Vy, and Vzare the measured radial velocities.

-------
                                                                                Volume 4, Section 8
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
	Page 6 of26
and performance audits are discussed in detail.  Standard operating procedures, maintenance schedules,
and quality control (QC) issues are also discussed.
8.1.1 Doppler Sodar

In the late 1960s and early 1970s, remote sensing techniques focused on the development of an acoustic-
based wind profiling system, commonly known today as a Sodar. Now these commercially available
Sodars are operated for the purpose of collecting upper-air wind measurements for a wide variety of
applications. Sodars consist of an antenna(s) that transmit and receive acoustic signals. A monostatic
system uses the same antenna for transmitting and receiving, while a bi-static system uses separate
antennas. The difference between the two antenna systems determines whether atmospheric scattering by
temperature fluctuations (in mono-static systems), or by both temperature and wind velocity fluctuations
(in bi-static systems) is the basis of the measurement. The vast majority of Sodars in use are of the
monostatic variety due to their more compact antenna size, simpler operation, and generally greater
altitude coverage.  Figure 8.5 shows the beam configurations of monostatic and bistatic systems.
              Figure 8.5 Schematic Showing a Monostatic and Bistatic Sodar System

The horizontal components of wind velocity are calculated from the radially measured Doppler shifts.
The tilt angle, or zenith angle, is generally 15 degrees to 30 degrees and the horizontal beams are typically
oriented at right angles to one another. Since the Doppler shift of the radial components along the tilted
beams includes the influence of both the horizontal and vertical components of the wind, a correction for
the vertical velocity should be applied in systems with zenith angles less than 20 degrees.  In addition, if
the system is located in a region where expected vertical velocities may be greater than about 0.2 m/s,
corrections for the vertical velocity should be made regardless of the beam's zenith angle.

The vertical range of Sodars is approximately 0.2 to 2 kilometers (km)  and is a function of frequency,
power output, atmospheric stability, turbulence, and, most importantly, the noise environment in which a
Sodar is operated. Operating frequencies range from less than  1,000 Hz to over 4,000 Hz, with power

-------
                                                                                 Volume 4, Section 8
                                                                                    Revision No: 1.0
                                                                                      Date: 10/2006
	Page 7 of26
levels up to several hundred watts. Due to the attenuation characteristics of the atmosphere, higher
power, lower frequency Sodars will measure to higher altitudes.  This greater range comes with a trade-
off of coarser vertical resolution when compared to the higher frequency Sodars that provide finer vertical
resolution measurements.  Some Sodars can be operated in different modes to better match vertical
resolution and range to the application.

Another important performance specification for upper-air instrument systems is the data recovery rate.
Data recovery is usually calculated as the ratio of the number of observations actually reported at a
sampling height to the total number of observations that could have been reported so long as the
instrument was operating (i.e., downtime  is usually not included in data recovery statistics and is treated
separately). Data recovery is usually reported as percent as a function of altitude.  Altitude coverage for
upper-air data is often characterized in terms of the height up to which data are reported 80 percent of the
time, 50 percent of the time, etc. Data recovery of Sodars is highly variable and is dependant on
atmospheric conditions at the various sampling heights. With Sodars, it is common to have several levels
of invalid or missing data. This is often due to a lack of turbulence at those levels. Sodars typically have
good height coverage during daytime hours when there is strong mixing and  sufficient turbulence to
provide strong signal returns.

Sodar systems should include available options for maximizing the intended  capabilities (e.g., altitude
range, sampling resolution, averaging time) of the  system and for processing and validating the data.
Sodar manufacturers usually have software subroutines that perform a variety of quality assurance
(QA)/QC and display functions. It is important to purchase QA/QC software that provides an extra level
of data validation, but one must still assure that valid meteorological data are not filtered out. Software is
also available for estimating mixing height and vertical and horizontal turbulence parameters.  However,
care must be taken with how these turbulence values are generated since they can have large errors
associated with them and. therefore, are not recommended for use in regulatory applications at this time.

Figure 8.6 illustrates different types of Sodar instruments. The selection of an installation site(s) should be
made in consultation with the manufacturer and should consider issues associated with the operation of
the Sodar instrument. Training should be obtained from the manufacturer on data validation and the
installation, operation, and maintenance of the instrument. Additional information on these issues is
provided in Section 8.3.
     Figure 8.6 Pictures of Different Types of Sodars: (a) Mini-Sodar, (b) Multi-Axis Sodar,
     (c) Phased-Array Sodar
8.1.2 Radar Wind Profiler

The principles behind the radar wind profiler are similar to Sodar except radars use EM waves to sense
turbulent fluctuations in the atmosphere. Because EM signals do not attenuate (dissipate) as quickly as

-------
                                                                                Volume 4, Section 8
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
	Page 8 of26
sound waves, radars have greater vertical range than Sodars. Like Sodars, radar wind profilers have
different operating frequencies and corresponding range and resolution specifications (Table 8-2).  The
guidance provided herein is focused only on boundary layer radar wind profilers.  Examples of this
instrument are shown in Figure 8.7.

Table 8-2 Characteristics of Radar Wind Profilers
Specification
Frequency class
Antenna size (m2)
Peak power (kw)
Range (km)
Resolution
Boundary Layer
1000 MHz (9 15 MHz)
3-6
0.5
0.1-5
60-100
Mid-tropospheric
400 MHz
120
40
0.2- 14
250
Tropospheric
50 MHz
10,000
250
2-20
150-1,000
Figure 8.7 Photographs of Several Types of Radar Wind Profilers: (a) Phased-Array System and
(b) Fixed-Axis Antenna System
                                I,
Radar wind profilers operate using principles similar to those used by Doppler Sodars, except that EM
signals are used rather than acoustic signals to remotely sense winds aloft. Figure 8.3 shows an example
of the geometry of a radar wind profiler. In this illustration, the radar can sample along each of five
beams: one is aimed vertically to measure vertical velocity, and four are tilted off-vertical and oriented
orthogonally to one another to measure the horizontal components of the air's motion.  A radar wind
profiler includes subsystems to control the radar's transmitter, receiver, signal processing, and RASS (if
provided), as well as data telemetry and remote control.

Detailed information on profiler operation has been provided by van de Kamp and Ecklund et.al.;44'45 a
brief summary of the fundamentals is provided in the following. The source of the backscattered energy
(radar "targets") is small-scale turbulent fluctuations that induce irregularities in the radio refractive index
of the atmosphere. The radar is most sensitive to scattering by turbulent eddies whose spatial scale is
one-half the wavelength of the radar, or approximately  16 centimeters (cm) for a 915-MHz radar wind
profiler.
A profiler's (and Sodar's) ability to measure winds is based on the assumption that the turbulent eddies
that induce scattering are carried along by the mean wind.  The energy scattered by these eddies and

-------
                                                                                Volume 4, Section 8
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
	Page 9 of26
received by the profiler is orders of magnitude smaller than the energy transmitted. However,  if sufficient
samples can be obtained, then the amplitude of the energy scattered by these eddies can be clearly
identified above the background noise level, and the mean wind speed and direction within the volume
being sampled can be determined.

The radial components measured by the tilted beams are the vector sum of the horizontal motion of the air
toward or away from the radar and any vertical motion present in the beam. Using appropriate
trigonometry, the three-dimensional meteorological velocity components (u, v, and w), wind speed, and
wind direction are calculated from the radial velocities with corrections for vertical motions. A boundary-
layer radar wind profiler can be  configured to compute averaged wind profiles for periods ranging from a
few minutes to an hour.

Boundary-layer radar wind profilers are often configured to sample  in more than one mode. For example,
in a "low mode," the pulse of energy transmitted by the profiler may be 60 m in length. The pulse length
determines the depth of the column of air being sampled and thus the vertical resolution of the data. In a
"high mode," the pulse length is increased, usually to 100 m or greater. The longer pulse length means
that more energy is being transmitted for each  sample, which improves the signal-to-noise ratio (SNR) of
the backscattered signal. Using a longer pulse length increases the depth of the sample volume and thus
decreases the vertical resolution in the data. The greater energy output of the high mode increases the
maximum altitude to which the radar wind profiler can sample, but at the expense  of coarser vertical
resolution and with an increase in the altitude at which the first winds are measured. When radar wind
profilers are operated in  multiple modes, the data are often combined into a single overlapping data set to
simplify post processing and data validation procedures.

The operating frequencies of all EM devices, including radars, are regulated by the Federal
Communications Commission (FCC). The allocated frequency for radar wind profilers for general use in
the United States is 915 MHz, however, other permitted operating frequencies do exist.  Before operating
a radar wind profiler, the user must have a valid frequency allocation authorization. For non-government
operators in the United States this frequency allocation can be obtained from: Federal Communications
Commission; Experimental Radio Services; P.O. Box 358320; Pittsburgh, PA 15251-5320. Government
operators should request approval from the corresponding National Telecommunications and Information
Administration branches.

Data recovery of radars, like Sodars, is a function  of atmospheric conditions and is highly variable. With
radar wind profilers, it is common to have several levels of invalid or missing data. This is typically due
to a lack of humidity and insufficient levels in  the refractive index in the atmosphere at those heights.
During precipitation events, radars measure the fall velocity of the precipitation instead of the air velocity.
In these instances, radars may appear to be generating reasonable wind estimates, but the measurements
can be biased by the precipitation. Typical data recovery rates range from about 50 percent to  near 90
percent and are variable  from hour to hour depending on atmospheric conditions.
8.1.3 Radio Acoustic Sounding System (RASS)

The operation principle behind RASS is that Bragg scattering occurs when acoustic energy (i.e., sound) is
transmitted into a radar beam such that the wavelength of the acoustic signal matches the half-wavelength
of the radar. As the frequency of the acoustic signal is varied, strongly enhanced scattering of the radar
signal occurs when the Bragg match takes place. When this occurs, the radar can measure the
propagation speed of the sound pulse. Thus, the speed of sound as a function of altitude can be measured,
from which Tv profiles can be calculated. The Tv of an air parcel is the temperature dry air would have if

-------
                                                                               Volume 4, Section 8
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
                                                                                       : 10 of 26
its pressure and density were equal to those of a sample of moist air. If not corrected in software, vertical
motions can affect RASS Tv measurements. As a rule of thumb, an atmospheric vertical velocity of 1 m/s
can alter a Tv observation by 1.6 degrees C.

RASS can be added to a radar wind profiler or to a Sodar system as shown in Figure 8.8a. When RASS is
added to a radar profiler, three or four vertically pointing acoustic sources (equivalent to high-quality
stereo loud speakers) are placed around the radar wind profiler's antenna.  Electronic subsystems are
added that include the acoustic power amplifier and the signal-generating circuit boards. The acoustic
sources are used only to transmit sound into the vertical beam of the radar, and are usually encased in
noise suppression enclosures to minimize nuisance effects that may bother nearby neighbors or others in
the vicinity of the instrument.

When RASS is added to a Sodar, as shown in Figure 8.8b, the necessary radar subsystems are added to
transmit and receive the radar signals and to process the radar reflectivity information. Since the  wind
data are obtained by the Sodar, the radar needs  only to sample along the vertical axis. The Sodar
transducers are used to transmit the acoustic signals that produce the Bragg scattering, which allows the
speed of sound to be measured by the radar.

The vertical resolution of RASS data is determined by the pulse length(s) used by the radar. RASS
sampling is usually performed with a 60- to 100-m pulse length. Because of atmospheric attenuation of
the acoustic signals at the RASS frequencies used by boundary layer radar wind profilers, the altitude
range that can be sampled is usually from 0.1 to 1.5 km,  depending on atmospheric conditions (e.g., high
wind velocities tend to limit RASS altitude coverage to a few hundred meters because the acoustic signals
are blown out of the radar beam).

RASS is an optional component of an RWP or  Sodar system. The power output of the acoustic source
should be kept as high as possible to obtain measurements from the highest altitudes possible.  Data
recovery rates are usually good, ranging from 70 percent to over 90 percent.
                        a                                            b
   Figure 8.8 Photographs of (a) a Radar Wind Profiler with a RASS and (b) a Sodar with a RASS
8.2 Acceptance Testing

Acceptance testing should be designed to determine if newly purchased or installed equipment is
performing according to the manufacturer's specifications. The acceptance test is crucial for remote

-------
                                                                                Volume 4, Section 8
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
	Page 11 of 26
sensors since data cannot be easily verified by simple tests. Shortly after the installation and startup of an
instrument, a system and performance audit should be performed. These audits will provide information
for the qualitative and quantitative assessment of the performance of the system, as well as the adequacy
of the standard operating procedures (SOPs) used for collection, processing, and validation of the data.
To best ensure that the data collected is of known quality, and that potential problems are identified early,
it is recommended that the initial audit be performed within 30 days of the start-up date.

For meteorological remote sensors, an acceptance test should include an intercomparison of data from the
system to be tested with data from acceptable in situ sensors on a tall tower, tethersonde, Sodar,
radiosonde, or other remote sensing system. The test should include the comparisons of data at a
minimum of three levels and over a range of meteorological conditions.

Intercomparisons are best performed using collocated meteorological information from tall towers or
other upper-air sensors.  In the absence of these collocated data sources, nearby upper-air data from the
NWS radiosonde network, the NOAA profiler network, aircraft reports, National Center for
Environmental Prediction (NCEP) high resolution mesoscale analyses, or other upper-air data can be
used. It is important to have an individual trained in the interpretation of the data perform a thorough
review of at least several days of data. The qualitative check is not meant to evaluate whether the data
meet the manufacturer's data specifications, but is intended to identify problems such as
   >   component failures;
   >   incorrect or improper operating/sampling parameters;
   >   antenna azimuth angles specified improperly or incorrectly measured; and
   >   siting problems  (active and passive interfering noise sources).

The  obvious difficulty encountered in trying to quantify the performance of a remote sensor is that one
must assume the "true" state of the atmosphere is known and, therefore, the degree of agreement between
profiler observations and reference values provided by independent measurement systems can be
determined.46 Under the best of circumstances, this requires an assumption of homogeneity and
stationarity in the atmosphere during the period in which the intercomparisons are performed. Given the
techniques currently available for obtaining data for intercomparisons, a measure of uncertainty is
introduced into any data set used to evaluate the remote sensor's performance. Sources of this uncertainty
include
   >   Differences in intercomparison data sets due to meteorological variability, spatial separation of
        measurements, temporal  separation of measurements, different sampling techniques and data
        reduction protocols, and/or outside sources  of interference (e.g., radio frequency interference,
        ground clutter, etc.);
   >   Instrument errors; and
   >   Random errors.

An important assumption in the acceptance test (and performance audits) is to design and perform tests in
such a way that the uncertainties  due to all but the last two sources of error, namely instrument errors and
random errors, will be minimized.  If this goal can be met, then the results of the intercomparison tests
will  better reflect the real performance of the remote senor.

We emphasize the concept of "estimating" the accuracy of these remote sensors because a reference
instrument capable of providing "true" values of the meteorological variables measured by these
instruments does not exist.  Uncertainties can be introduced in the data sets being compared by

-------
                                                                                Volume 4, Section 8
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
_ Page 12 of 26
meteorological variability, different sampling protocols, and other environmental and operational factors.
For example
   >   Radiosondes provide a quasi-instantaneous measurement of winds and temperatures, while the
        profiler is usually configured to produce data averaged over 15-60 minutes.  Changes in
        meteorological conditions during the averaging period (e.g., wind shift associated with a frontal
        passage) will be reflected in the profiler data but may not be represented in the radiosonde
        observations.
   >   A profiler samples the column of air directly above the instrument, while a radiosonde drifts with
        the mean wind.  Horizontal gradients in the winds or temperatures over the volume of air sampled
        by the radiosonde may not be represented in the profiler data.
   >   Even with the radar wind profiler and collocated Sodar, their respective geometries and sampling
        configurations can introduce uncertainties into the data sets when there are inhomogeneities
        between the volumes of the atmosphere each is sampling.

The  test methods listed here can be used to perform an acceptance test and are designed to minimize
differences caused by these and other factors. The seven basic steps to performing a quantitative
intercomparison are
     1 .   Plan to conduct the intercomparison under a variety of atmospheric conditions (during stable,
        convective, and transitional boundary layers and weak, moderate, and strong winds).
    2.   Configure the instruments to make comparable measurements so that uncertainties due to
        meteorological variability, sampling techniques, and data reduction protocols are minimized.
        Alternatively, post-processing may be required to average the data so that the time and/or space
        (altitude) scales of the two data sets are comparable.
    3.   Ensure the comparison instrument does not interfere with the remote sensor  (e.g., tethersonde in
        the beam of the Sodar).
    4.   Document the weather conditions during the intercomparison by logging standard hourly
        observations,  including current weather, ceiling, sky-cover, ambient temperature, wind speed and
        wind direction.
    5.   Collect enough samples so that the conditions of interest are well represented.
    6.   Perform quality control screening on all data to be used in the intercomparisons. All observations
        should be brought to Level 1 validation before quantitative tests are performed.
    7.   Compute comparison statistics shown here and compare the results with those criteria listed in
        Table 8-2.

The  following statistics can be used to compare profiler observations to other data sets and to estimate the
performance of the profiler.  We recommend that these parameters be computed for the ensemble of the
data and as a function of altitude.
Systematic difference22, used as a measure of the bias or relative accuracy of the instrument:

                                                -Pb^                                     (8-5)
where  n   = number of observations

        Pai = i th observation of the sensor being evaluated

        PI,: = i th observation of the "reference" instrument

-------
                                                                               Volume 4, Section 8
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
                                                                                    Page 13 of 26
Operational comparability22, or the rms difference between the remote sensor and comparison
measurements, used as a measure of the uncertainties in the comparisons.
                                         ^V "'    "'"                                      (8-6)

Some general guidelines for making data as similar as possible include
   >   The data from the tethersonde or tower should be broken down into their u and v components.  At
       the end of this sampling period, the components should be averaged and the resultant vector wind
       speed and wind direction calculated.
   >   At some sites it may be possible to use NWS radiosonde data to perform an acceptance test. This
       test is somewhat more difficult to perform but will provide the data required to complete the test.
       The radiosonde should be within 20 km of the remote sensing site, in simple terrain, and in the
       same meteorological regime as that of the remote sensing instrument.  The comparison should
       include a data time series long enough to have a large sample for every meteorological condition
       experienced at the site, and only data captured during similar meteorological regimes at both sites
       should be used in the comparison. Data at higher elevations should be used for the comparison
       since it is less influenced by local surface features.
   >   All wind data used in the intercomparisons should receive quality control screening.
   >   Sample during non-precipitating conditions. There may be large discrepancies in winds
       measured during precipitation because the radar profilers will measure the fall velocity associated
       with the precipitation, which in turn will be used to extract the horizontal components from the
       vertical velocities.
   >   With radiosondes, match the wind averaging interval you specify in the radiosonde data
       acquisition system to the balloon's ascent rate so that the radiosonde data are averaged over a
       volume that approximates as closely as possible the volume sampled by the radar wind profiler
       and RASS.  For example, a 3-m/s ascent rate and a 30-second wind averaging period will produce
       radiosonde data averaged over layers 90 m deep, which can then be compared to profiler data
       with a 100-m vertical resolution.

   >   To maximize RASS altitude coverage and to minimize uncertainties between the RASS and
       radiosonde data sets due to meteorological variability and spatial inhomogeneities in the
       atmosphere, select sampling conditions characterized by light winds (e.g., wind speeds less than
       5 m/s) and good vertical mixing.
   >   To minimize uncertainties created by temporal differences between the RASS and radiosonde
       measurements, launch the balloon at the beginning of the RASS sampling period.  To minimize
       spatial differences, launch the balloon as close to the RASS as possible.
8.3 Installation and Siting

The following subsections provide information on installation issues related to QA/QC concerns. In
general, it is recommended that the installer follow guidance provided in the On-Site Meteorological
Program Guidance for Regulatory Modeling Applications47 Section 3.0 and the vendors' instructions. In
general, installation procedures for all these remote sensors include the following:

-------
                                                                          Volume 4, Section 8
                                                                            Revision No: 1.0
                                                                              Date: 10/2006
	Page 14 of 26
 Determine the latitude, longitude, and elevation of the site using a GPS instrument, U.S.
 Geological Survey (USGS) topographical maps, or other detailed maps.
 Measure the orientation of antennas of the Sodar or radar profiler with respect to true north. Use
 the solar siting technique or the GPS techniques discussed in Sections 2.5.2.3 and 2.5.2.4.
 The site should be documented as follows:
 •  Photographs should be taken in sufficient increments to create a documented 360°  panorama
    around the antennas. Additionally, pictures should be obtained of the antenna installation,
    shelter and any obstacles that could influence the data.
 •  Photographs should be taken of the instrument, site, shelter, and equipment and computers
    inside the shelter.
 •  A detailed site layout diagram should be prepared that identifies true north and includes the
    locations of the instrument, shelter, other equipment, etc. An example of such a diagram is
    shown in Figure 8.9. Additionally, it is recommended that the site layout diagram  include the
    electrical and signal cable layout, and the beam directions of any remote sensor.
VISTA, ORIENTATION, AND LE\TL AUDIT RECORD
Dare:
Key Person:
Instrument:
Model Number:
Serial Number:
Software version:
Rotation ansle
System
Measured:
Difference:

Array Level:
January 1 1996
John Sitetech
Radar Wind Profiler
GEK-1500
1234
3.95

147 'true
146"tnre
1*

•• 0.5"

AziEnirh Angle (deg.)
Maeaeric
;;
-

-
—
-

True
.Q
30
60
90
no
150
ISO
240
270
300
330
Sire Name:
Project:
Latitude:
Longitude:
Elevation:

Direcricn
Beam 1 :
Beam 2:

Firing order
Declination:


Terrain Elevation
.Angle (deg.)
12
19
4
i;
4
*2
••:}
1.4
Site 5
ABC
31"10'25"
9M5'33"
172 in


146"
236'=

W. beam 1. beam 2
11" easr (solar verification)


Features Distance
Buildings and power lines at - 300 m
Stack at 150-200 m.
Power pole at 10 ai • 5 "beyond.
Low Tees and bushes at 10m.
Power hues at 200-300 m
Tree', at 3040m.
Looking out over the lake.
Lookms out over the lake, can see land
Lookms out over the lake, can see land
Trees and telephone pole at 1 00 m.
Light pole at 25 m. Building; at -250 ai
           Figure 8.9 Example Site Layout Diagram

    A vista table should be prepared that documents the surroundings of the site in 30°
    increments. Vistas  for the beam directions, if they are not represented by the 30° views (±5°),
    should be included. The table should identify any potential passive and active noise sources
    in each direction, and the approximate distance and elevation angle above the horizon to the
    objects.  An example  is shown in Figure 8.10.

-------
                                                                                 Volume 4, Section 8
                                                                                    Revision No: 1.0
                                                                                      Date: 10/2006
                                                                                      Page 15 of 26
                           Figure 8.10 Example Site Vista Diagram

Listed below are some key issues to consider when siting upper-air systems.

Representative location. Sites should be located where upper-air data are needed to characterize the
meteorological features important to meeting the program objectives. Panoramic photographs should be
taken of the site to aid in the evaluation of the data and preparation of the monitoring plan. Data collected
at sites in regions with local geographic features such as canyons, deep valleys, etc., may be
unrepresentative of the surrounding area and should be avoided, unless such data are needed to resolve
the local meteorological conditions. Measurements made in complex terrain may be representative of a
much smaller geographic area than those made in simple homogeneous terrain.  See Thuillier's
article48for a discussion of the influence of terrain on siting and exposure of meteorological
instrumentation.

Site logistics.
   >    Adequate power should be available for the instrument system as well as an environmentally
        controlled shelter that houses system electronics, and data storage and communication devices.
   >    The site  should be in a safe, well lit, secure area clear of obstacles with level terrain and sufficient
        drainage. The site should allow adequate room for additional equipment that may be required for
        calibrations, audits, or supplementary measurements.
   >    A fence  should be installed around the equipment and shelter to provide security, and appropriate
        warning  signs should be posted as needed  to alert people to the presence of the equipment.
   >    A remote data communications link (e.g., dedicated leased line, standard dial-up modem line,
        Internet link, or satellite Internet) should be installed at the monitoring site.

Collocation with surface meteorological measurements. Several advantages can be gained by locating an
upper-air site with or near an existing meteorological monitoring station. For instance, collocated data can
be used for data validation purposes and for performing reasonableness checks (e.g., do surface winds

-------
                                                                                Volume 4, Section 8
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
                                                                                        : 16 of 26
roughly agree with near-surface upper-air winds, surface temperatures with near-surface RASS
measurements).

Instrument noise. Sodar and RASS generate noise that can disturb nearby neighbors. Depending on the
type of Sodar or RASS instrument, power level, frequency, acoustic shielding around the system, and
atmospheric conditions, the transmitted pulse can be heard from tens of meters up to a kilometer away.
An optimum site is one that is isolated from acoustically sensitive receptors.49

Passive interference/noise sources.  Objects such as stands of trees, buildings or tall stacks, powerlines,
towers, guy wires, vehicles, birds, or aircraft can reflect Sodar or radar signals and contaminate the data.
Not all sites  can be free of such objects, but an optimum site should be selected to minimize the effects of
such obstacles.  If potential reflective "targets" are present at an otherwise acceptable site, the beams of
the instrument should be aimed away from the reflective objects.  In the case of Sodars, locating the
antennas so that there are no direct reflections from objects will help minimize potential contamination.
In the case of the radar profiler, it is best to aim the antennas away from the object and orient a phased-
array antenna's  corners so they are pointing toward the objects. As a rule of thumb,  sites with numerous
objects taller than about  15° above the horizon should be avoided.  The manufacturers of the remote
sensing equipment should be contacted regarding software that may be available to identify and minimize
the effects of these passive noise  sources.

Active interference/noise sources. For Sodars, noise sources such as air conditioners, roadways,
industrial facilities, animals,  and insects will degrade the performance of Sodar systems.49 If proximity to
such sources cannot be avoided, then additional acoustic shielding may help minimize the potentially
adverse effects on the data. In general, noise levels below 50 decibels (dBA) are considered to be
representative of a quiet  site, while levels above 60 dBA are characteristic of a noisy site. For radar wind
profilers and RASS, radio frequency (RF) sources such as radio communications equipment and cellular
telephones may have an adverse effect on performance.

Licenses and ordinances. Before operating a remote sensor, it is recommended that all applicable
recommendations for operation of equipment be addressed. For example, to operate a radar wind profiler
or a RASS, an FCC license is required. For radiosonde and tethersonde sounding systems, a Federal
Aviation Administration (FAA) waiver may be required. Local noise ordinances may limit the operation
of Sodar or RASS instruments. Some of these requirements may take several months to address and
complete.

Surveying candidate locations. Prior to final site selection, a survey is recommended to  identify audio
sources50 and RF sources that may degrade  system performance. Additionally, panoramic photographs
should be taken to aid in the evaluation of the candidate site and for the preparation of the monitoring
plan. As part of the survey, appropriate topographic and other maps should be used to identify other
potential sources of interference,  such as roadways and airports.

Specific installation procedures for each instrument are presented in the following sections.
8.3.1 Sodar

Siting of Sodars can best be accomplished by vendors or experienced users. The complexities of Sodars
provide a challenge to the user who must optimize the conditions favorable for Sodar technology while
still making use of available sites in a given study area.

-------
                                                                                 Volume 4, Section 8
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
	Page 17 of 26
A problem may exist at some potential monitoring sites due to the presence of passive and active noise
sources. It is extremely important to determine if the proposed sampling site has any potential for
producing fixed echoes (a passive noise source). These fixed echoes are often due to the energy
contained in the side lobes of the emitted acoustic pulse. These fixed echoes have the effect of biasing the
computed wind components u, v, and w. Printing a facsimile chart sometimes reveals the presence of
fixed echoes. This should be performed shortly after system setup, and repeated seasonally to  aid in
determining if fixed echoes exist.  Some fixed echoes may be avoided by constructing an acoustically
absorbing shelter around the Sodar antennas. These shelters are designed to absorb most of the energy
released in the side lobes, providing a narrower beam, thus a cleaner acoustic signal.

Additional guidance includes the absence of obstructions in a 110° arc centered on the vertical axis or a
40° arc centered on each beam. In addition, if the system is to be installed near a building, the antennas
should be oriented off the corners of the building.  If the building does intercept the sound wave, the wave
will be reflected away from the Sodar due to the acute angles of the building's wall. Some manufacturers
provide software  routines that can detect fixed echoes and eliminate them from the consensus output.

All attempts should be made to avoid fixed echoes; however, if a limited number of sites are available and
all have a possibility of producing fixed echoes, then the fixed echo detection software should be used to
eliminate the problem. Special attention should be used during the acceptance test, described in
Section 8.2, to determine if the fixed echo rejection routines are working properly.

The antenna does not  necessarily have to point in one of the cardinal directions (i.e., north, south, east, or
west). System software allows the Sodar to be set up in almost any direction, allowing the installer to
point the beams away from obstacles that might interfere with the signal. For example, if the Sodar is to
be set up near a tower, the antenna should be oriented so the beams point away from the tower,

Another type of interference, active interference, occurs from objects that emit noise, such as local
automobile traffic, nearby construction, overhead aircraft, etc. Any acoustic source that emits  its energy
near the transmission  frequency of a Sodar has the potential to interfere with and degrade data  quality.
This type of interference is more difficult to detect because it tends to be seasonal, sporadic, or random in
nature.  This problem  can be reduced by installing acoustic absorbing shelters around the Sodar antenna
like those shown in Figure 8.6.  A simple test to determine if a problem exists at a given site is to set up
the Sodar and turn off the transmitter.  Analysis of received energy will determine if the presence of
active interfering noise exists. If interferences from active sources are detected, it is recommended that
the Sodar be moved to an alternate site. The vendor or an experienced Sodar operator should be consulted
during the installation process to decrease the chance of contamination of these data.
8.3.2 Radar Wind Profiler

As with a Sodar, careful siting will result in a site that has minimal interferences that can cause data
problems.  The vendor or an experienced radar wind profiler operator should be consulted during the
installation process to decrease the chance of contamination of these data.

Signal returns from "ground clutter" can bias the radar wind profiler data. Trees, powerlines, busy roads,
and even terrain features can produce erroneous data due to reflected EM signals.  Severe ground clutter
often degrades the signal enough to render data in the first few reported levels useless. As with Sodars,
radar beams have side lobes that emit energy to around 70° from vertical (Figure  8.4).  These side lobes
cause a higher degree of interference than Sodars because radar return signals are typically very weak, so
small amounts of energy reflected back to the receiver may cause large errors in the estimates of wind.

-------
                                                                                Volume 4, Section 8
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
                                                                                        : 18 of 26
Therefore, radars should be setup away from tall buildings, powerlines and other obstructions that may be
a potential source of interference.  The radar wind profiler should also be situated in an open area (e.g., an
airport), or on top of a small hill or building to decrease the potential for ground clutter contamination.
The antenna does not necessarily have to point in the cardinal directions (i.e., north, south, east, or west).
System software should allow the radar to be set up in almost any direction, allowing the installer to point
the beams away from obstacles that might interfere with the signal. For example, if the radar is to be
setup near a tower, the antenna should be oriented so the beams point away from the tower.
8.3.3 Radio Acoustic Sounding System

The user of a radar/RASS should follow the guidelines for installing a radar, as specified in Section 8.3.2.
Contamination from external acoustic sources is only a minor problem, but should be avoided as outlined
for Sodars in Section 8.3.1. If a Sodar/bistatic radar is being used to measure Tv then the installer should
follow the guidelines for installing a Sodar, with the addition of meeting the  recommendations for
installing a radar profiler.
8.4 Calibration

A calibration involves measuring the conformance to or discrepancy from a specification for an
instrument and an adjustment of the instrument to conform to the specification.  In this sense, other than
directional alignment checks of the antenna(s), a true calibration of the Sodars, radar wind profilers, and
RASS instruments described in this document is difficult. Due to differences in measurement techniques
and sources of meteorological variability, direct comparison with data from other measurement platforms
is not adequate for a calibration.  Instead, a calibration of these sensors consists of test signals and
diagnostic checks that are used to verify that the electronics and individual components of a system are
working properly. Other than antenna misalignment, results from these calibrations should not be used to
adjust any data.

System calibration and diagnostic checks should be performed at six-month intervals, or in accordance
with the manufacturer's recommendations, whichever is more  frequent. The alignment of remote sensing
antennas, referenced to true north, should be verified at six-month intervals.

Recent advances in instrumentation for auditing of Sodar instruments51 have led to the development of a
transponder that can simulate a variety of acoustic Doppler-shifted signals on certain Sodars. This
transponder can be used to verify the calibration of the Sodar's total system electronics and, in turn,
validate the overall system operation in terms of wind speed and altitude calculations. However, such a
check should not be considered a "true" calibration of the system since it does not consider other factors
that can affect data recovery. These factors include the system signal-to-noise ratio, receiver
amplification levels, antenna speaker element performance, beam steering and beam forming for phased-
array systems, and overall system electronic noise.

For the radar wind profiler and RASS systems, there are no simple means at present to verify the accuracy
of the Doppler-shifted signals in the field other than to perform an intercomparison of data with some
other measurement system. Instead, calibrations of radar wind profiler and RASS systems are performed
and checked at the system component level. These checks should be performed in accordance with the
manufacturer's recommendations. Like  some Sodar systems, the radar systems use both software and
hardware diagnostics to check the system components.

-------
                                                                               Volume 4, Section 8
                                                                                 Revision No: 1.0
                                                                                   Date: 10/2006
                                                                                       : 19 of 26
8.5  Operation, Maintenance, and Quality Control

Sodars, radar wind profilers, and RASS have automated operating systems and generally require minimal
input from the user. Variables such as vertical range, vertical resolution, averaging times, and power
output may be adjusted if needed, but most of the system operations are automatic. Users should follow
the vendor's instructions for operation and maintenance.

Like all monitoring equipment, upper-air instruments require various operational checks and routine
preventive maintenance.  The instrument maintenance manuals should be consulted to determine which
checks to perform and their recommended frequency. The quality and quantity of data obtained will be
directly proportional to the care taken in ensuring that the system is routinely and adequately maintained.
The site technicians who will perform preventive and emergency maintenance should be identified. The
site technicians serve a crucial role in producing high quality data and thus should receive sufficient
training and instruction on how to maintain the equipment.  Some general issues related to operational
checks and preventive maintenance should be addressed in the QAPP, including
   >   Identification of the components to be checked and replaced
   >   Development of procedures and checklists to conduct preventive maintenance
   >   Establishment of a schedule for checks and preventive maintenance
   >   Identification of persons (and alternates) who will perform the checks and maintenance
   >   Development of procedures for maintaining spare components that need frequent replacement

Listed here are some key items to be included in the operational checklists for each of the different types
of remote sensors. The list is not comprehensive, but should serve as a starting point for developing a
more thorough set of instrumentation checks:
   >   Safety equipment (first aid kit, fire extinguisher) should be inventoried and checked.
   >   Computers should be routinely monitored to ensure adequate disk space is available, and
       diagnosed to ensure integrity of the disk.
   >   A visual inspection of the site, shelter, instrument and its components should be made.
   >   Data should be backed up on a routine basis.
   >   If the remote  sensors are operated during the winter, procedures for snow and ice removal should
       be developed and implemented, as needed.
   >   The clock time of the instruments should be monitored, and a schedule for updating the clocks
       established, based on the timekeeping ability of the instrument.
   >   The antenna level and orientation of Sodar, radar wind profiler, and RASS systems should be
       verified periodically.
   >   The inside of the antennas/enclosures of the Sodar,  radar wind profiler, and RASS systems should
       be inspected and any leaves, dust, animals, insects,  snow, ice, or other materials removed. Since
       the antennas are open to precipitation, drain holes are provided to allow water to pass through the
       bottom. These holes should be periodically inspected and cleaned.
   >   Cables and guy wires securing the equipment should be checked to ensure that they are tight and
       in good condition.
   >   Antenna cables and connections should be inspected for signs of damage due to normal wear,
       moisture, or animal activities.

-------
                                                                                 Volume 4, Section 8
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
	Page 20 of 26
   >   For Sodar systems, the site technician(s) should listen to ensure that the system is transmitting on
        all axes and in the correct firing sequence. For three-axis systems, this is accomplished by
        listening to each antenna. For phased-array systems, this can be accomplished by standing away
        from the antenna in the direction of each beam and listening for relatively stronger pulses.

   >   The integrity of any acoustic enclosures and acoustic-absorbing materials should be inspected.
        Weathering of these items will degrade the acoustic sealing properties of the enclosure and reduce
        the performance.
   >   For a radar profiler with RASS, acoustic levels from the sound sources should be measured using
        a sound meter (ear protection is required) and readings should be compared with manufacturer's
        guidelines.
   >   After severe or inclement weather, the site should be visited and the shelter and equipment should
        be inspected.

SOPs should be developed that are specific to the operations of a given instrument and site. The purpose
of an SOP  is to spell out operating and QA procedures with the ultimate goal of maximizing data quality
and data capture rates. Operations should be performed according to a set of well-defined, written SOPs
with all actions documented in logs and on prepared forms. SOPs should be written in such a way that if
problems are encountered, instructions are provided on actions to be taken. At a minimum, SOPs should
address the following issues:
   >   Installation, setup, and checkout
   >   Site operations and calibrations
   >   Operational checks and preventive maintenance
   >   Data collection protocols
   >   Data archiving

   >   Key contacts

Some general guidelines for operation and maintenance include
   >   Wind data should be stored in its u, v, and w components to ensure minimal loss of information
        and more thorough data validation. This will also be useful in instances when the wind direction
        may be in question.
   >   Statistics about the quality of data averages (e.g., number of valid return intensities, consensus
        numbers, and standard deviation of component values) should also be stored. This information
        may be useful in detecting instrumentation problems.
   >   In  addition to storing the averaged wind or TV data, it is recommended that users store  the raw
        forms of data so they can be reprocessed.
   >   The hard-disk drive is used for storing data; it should be checked as often as is necessary to
        ensure there is enough available storage.
   >   In  the first few weeks after installation, the data should be checked on a daily basis to determine
        if the system is working properly.  Time series plots of all variables should be produced and
        analyzed by a meteorologist or other qualified professional. This step is important for  detecting
        any bias or anomalies in the data set. It is usually at this point that active and passive
        interferences are detected.  All inspections and maintenance activities should be documented in a
        site log book.

-------
                                                                                 Volume 4, Section 8
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
	Page 21 of 26
   >   Once the site operator determines the system to be operating adequately, data should be plotted
        and checked on a weekly basis to determine system performance.

Maintenance should include biweekly checks of the instrument, site, shelter, and electronics.

All operational checks and preventive maintenance activities should be recorded in logs and/or on
appropriate checklists (electronic and/or paper), which will become part of the documentation that
describes and defends the overall quality of the data produced.

If problems are found, a corrective action should be taken and reported. A corrective action program
must have the capability to discern errors or defects at any point in an upper-air monitoring program. It is
an essential management tool for coordination, QC, and QA activities. A workable corrective action
program must enable the identification of problems, establish procedures for reporting problems to the
responsible parties, trace the problems to the source, plan and implement measures to correct the
problems, maintain documentation of the results of the corrective process, and resolve each problem. The
overall documentation associated with the corrective action and reporting process will become part of the
documentation that describes and defends the overall quality of the data produced. A sample correction
form can be found in the EPA's Quality Assurance Handbook for Air Pollution Measurement Systems.52

Systematic routines used to inspect these data provide a level of QC.  These QC checks should be
performed by a technician, meteorologist, or other qualified professional who is familiar with these
instruments. When a problem is found, a discrepancy report should be issued that notifies the users of the
problem.

Studies performed to date have indicated that the upper-air measurement systems described in this
document can reliably and routinely provide high quality meteorological data. However, these are
complicated systems and, like all such systems, are subject to sources of interference and other problems
that can affect data quality.  Users should read the  instrument manuals to obtain an understanding of
potential shortcomings and limitations of these instruments.  If any persistent or recurring problems are
experienced, the manufacturer or someone knowledgeable about instrument operations should be
consulted.

Sodar data can be rendered problematic by the following:
   >   Passive noise sources (also called fixed echo reflections).  Passive noise occurs when nearby
        obstacles reflect the Sodar's transmitted pulse. Depending on atmospheric conditions, wind
        speed, background noise, and signal processing techniques, the fixed echoes  may reduce the
        velocity measured along a beam(s) or result in a velocity of zero.  This problem is generally seen
        in the resultant winds as a rotation in direction and/or a decrease in speed at the affected altitude.
        Some manufacturers offer systems that have software designed to detect fixed echoes and
        effectively reject their influence.  To further decrease the effect of the fixed echoes, additional
        acoustic shielding can be added to the system antenna.
   >   Active noise sources (ambient noise interference). Ambient noise can come  from road traffic,
        fans or air conditioners, animals, insects, strong winds, etc. Loud broad-spectrum noise will
        decrease the SNR of the Sodar and decrease the performance  of the system.  Careful siting of the
        instrument will help minimize this problem.
   >   Unusually consistent winds at higher altitudes. Barring meteorological explanations for this
        phenomenon, the most common cause is a local noise source that is incorrectly interpreted as a
        "real" Doppler shift.  These winds typically occur near the top of the operating range of the
        Sodar.  To identify this problem, allow the Sodar to operate in a listen-only mode without a

-------
                                                                                 Volume 4, Section 8
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
	Page 22 of 26
        transmit pulse to see if winds are still reported.  In some cases, it may be necessary to make noise
        measurements in the specific operating range of the Sodar to identify the noise source.
   >   Reduced altitude coverage due to debris in the antennas. In some instances, particularly after a
        precipitation event, the altitude coverage of the  Sodar may be  significantly reduced due to debris
        in the antennas. In three-axis systems, drain holes may become plugged with leaves or dirt and
        water, snow, or ice may accumulate in the antenna dishes.  Similarly,  some of the phased-array
        antenna systems have the transducers oriented vertically and are open to the environment.
        Blocked drain holes in the bottom of the transducers may prevent water from draining.  Regular
        maintenance can prevent this type of problem.
   >   Precipitation interference. Precipitation, mostly rain, may affect the data collected by Sodars.
        During rainfall events, the Sodar may measure the fall speed of drops, which will produce
        unrealistic winds. In addition, the sound of the droplets hitting the antenna can increase the
        ambient noise levels and reduce the altitude coverage.
   >   Low signal-to noise-ratio (SNR). Conditions that produce low SNR can degrade the performance
        of a Sodar.  These conditions can be produced by high background noise, low turbulence, and
        near-neutral lapse rate conditions.

Data from radar wind profiler systems can be affected by several problems, including the following:
   >   Interference from migrating birds.  Migrating birds can contaminate radar wind profiler signals
        and produce biases in the wind speed and direction measurements.53 Birds act as large radar
        "targets", so that signals from birds overwhelm the weaker atmospheric signals. Consequently,
        the radar wind profiler measures bird motion instead of, or in addition to, atmospheric motion.
        Migrating birds have no effect on RASS. Birds generally migrate year round along preferred
        flyways, with the peak migrations occurring  at night during the spring and fall months.54 Vendors
        have software to minimize the influence of migrating birds on radar wind profiler data.
   >   Precipitation interference. Precipitation can  affect the  data collected by radar profilers operating
        at 915 MHz and higher frequencies.  During precipitation, the radar profiler measures the fall
        speed of rain drops or snow flakes.  If the fall speeds are highly variable during the averaging
        period (e.g., convective rainfall), a vertical velocity correction can produce erroneous data.
   >   Passive noise sources (ground clutter). Passive  noise interference is produced when a transmitted
        signal is reflected off an object instead of the atmosphere. The types of objects that reflect radar
        signals are trees, elevated overpasses, cars, buildings, airplanes, etc. Careful siting of the
        instrument can minimize the effects of ground clutter on the data. Both software and hardware
        techniques are also used to reduce the effects of ground clutter. However, under some
        atmospheric conditions (e.g., strong winds) and at some site locations, ground clutter can produce
        erroneous data.  Data contaminated by ground clutter can be detected  as a wind shift or a decrease
        in wind speed at affected altitudes. Additional information is provided by Brewster and
        Gaynor.55'56

   >   Velocity folding  or aliasing. Velocity folding occurs when the magnitude of the radial
        component of the true air velocity exceeds the maximum velocity that the instrument is capable
        of measuring, which is a function of sampling parameters.57 Folding occurs during very strong
        winds (>20 m/s)  and can be easily identified and flagged by automatic screening checks or during
        the manual review.

RASS systems are susceptible to several common problems including  the following:

-------
                                                                               Volume 4, Section 8
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
                                                                                    Page 23 of 26
       Vertical velocity correction. Vertical motions can affect the RASS virtual temperature
       measurements.  Virtual temperature is determined by measuring the vertical speed of an upward-
       propagating sound pulse, which is a combination of the acoustic velocity and the atmospheric
       vertical velocity. If the atmospheric vertical velocity is non-zero and no correction is made for the
       vertical motion, it will bias the temperature measurement. As a rule of thumb, a vertical velocity
       of 1 ms-1 can alter a virtual temperature observation by 1.6 degrees C.

       Potential cold bias. Recent inter-comparisons between RASS systems and radiosonde sounding
       systems have shown a bias in the lower sampling altitudes.58 The RASS virtual temperatures are
       often slightly cooler (-0.5 to -1.0°C) than the reference radiosonde data.
8.6 Auditing

A system audit is intended to independently assess the QAPP and how it is being implemented. A
performance audit is a direct challenge to the performance of the instrument.  Audits of upper-air
instrumentation pose some interesting challenges to verifying their proper operation. While system audits
can be performed using traditional system checks and alignment and orientation techniques, performance
audits of some instruments require unique, and sometimes expensive, procedures. In particular, unlike
surface meteorological instrumentation, the upper-air systems cannot be challenged using known inputs
such as rates of rotation, orientation directions, or temperature baths.  Recommended techniques for both
system and performance audits of the upper-air instruments are described here. These techniques have
been categorized into system audit checks and performance audit procedures for Sodars, radar profilers,
and RASS. Performance audits should be performed 30 days after installation and every six months.

Results from the performance evaluation should be compared with evaluation criteria to assess whether
the comparisons are reasonable.  Typical criteria for the comparisons are provided in Table 8-3.
Comparison results in excess of the criteria do not necessarily mean that the remote sensor data are
invalid.  In making the assessment, it is important to  understand the reasons for the differences. Reasons
may include unusual meteorological conditions, differences due to sampling techniques and data
reduction procedures, insufficient number of samples, or problems or limitations in one or both
instruments. Both the reasons for, and the magnitude of, the differences as well as the anticipated uses of
the data need to be considered in determining whether the observed differences are significant.
              Table 8-3 Recommended Audit Criteria for Sodar, Radar Wind
              Profilers, and RASS
Variable
u, va
Wind speed3
Wind direction3
RASS temperature
Systematic Difference
±1 m/s
±1 m/s
±10°
±rc
Operational
Comparability
2 m/s
2 m/s
30°
1.5°C
               a The wind speed and wind direction criteria apply to data when the wind
               speeds are greater than 2 m/s.

-------
                                                                                Volume 4, Section 8
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
	Page 24 of 26
8.6.1 System Audits of Remote Sensors

System audits of a remote sensor should include a complete review of the QAPP, any monitoring plan for
the station, and the station's SOPs. The system audit will determine if the procedures identified in these
plans are followed during station operation. Deviations from the plans should be noted and an assessment
made as to what effect the deviation may have on data quality. To ensure consistency in the system
audits, a checklist should be used.

A routine check  of the monitoring station should be performed to ensure that the local technician  is
following all SOPs. In addition to specific checks recommended by the vendor, the following items
should be checked:
    >   The antenna and controller interface cables  should be inspected for proper connection.  If multi-
        axis antennas are used, this will include checking for the proper direction of the interface
        connections.
    >   Orientation checks should be performed on  the individual antennas, or phased-array antenna. The
        checks should be verified using solar sitings or the GPS method when possible  (see
        Sections 2.5.2.3 and 2.5.2.4). The  measured orientation of the antennas should be compared with
        the system software settings of the  system.  The antenna alignment should be maintained within
      . 2°, which is consistent with wind direction vane alignment criteria.
    >   For multi-axis antennas, the inclination angle, or zenith angle from the vertical, should be verified
        against the software settings and the manufacturer's recommendations. The measured zenith
        angle should be within 0.5° of the software  setting in the data system.
    >   For phased-array antennas, and for the vertical antenna in a multi-axis system, the level of the
        antenna  should be within 0.5° of the vertical.
    >   For multi-axis Sodar systems, a separate distinct pulse, or pulse train in the case of frequency-
        coded pulse systems, should be heard from each of the antennas. In a frequency-coded pulse
        system there may be a sound pattern that can be verified. The instrument manual should  be
        checked to see if there is such a pattern.
    >   The controller and data collection devices should be checked to ensure that the  instruments are
        operating in the proper mode and that the data being collected are those specified by the SOPs.
    >   Station logbooks, checklists, and calibration forms should be reviewed for completeness and
        content to ensure the entries are commensurate with the expectations in the procedures for the
        site.
    >   The site  operator should be interviewed to determine his/her knowledge of system operation,
        maintenance, and proficiency in the performance  of QC checks.
    >   The antenna enclosures should be inspected for structural integrity that may cause failures as well
        as for any signs of debris or animal or insect nests that may cause drainage problems in the event
        of rain or snow.

    >   Preventive maintenance procedures should be reviewed for adequacy and implementation.
    >   The time clocks on the data acquisition systems should be checked and compared to a standard of
        2 minutes.
    >   The data processing procedures and the methods for processing the data from sub-hourly to
        hourly intervals should be reviewed for appropriateness.

-------
                                                                               Volume 4, Section 8
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
       	Page 25 of 26
       Data collected over a several-day period should be reviewed for reasonableness and consistency.
       The review should include vertical consistency within given profiles and temporal consistency
       from period to period. For radar wind profilers and Sodars, special attention should be given to
       the possibility of ground clutter (i.e., fixed echoes) and/or active noise source contamination in
       the data.
8.6.2 Sodar Performance Audit

A performance audit of a Sodar should use a responding device (see Baxter for a description51) or a data
intercomparison. A responding device independently assesses the ability of the instrument to correctly
interpret test signals that represent known wind speeds. The responding device emits a fixed audio
frequency at a known time that is received by the Sodar's antenna.  When the Sodar receives the signal, it
interprets the audio frequency as a Doppler shift. If the Sodar is operating properly, it should correspond
to a known velocity. The timing of the signal is used to verify that altitude is interpreted properly. This
type of performance audit tests the complete processing of a known signal through all Sodar system
components.

Since a Sodar works on the principle of measuring winds on a component basis, the  audit data should be
evaluated on that basis. Audit results for a properly operating Sodar should be within 0.2 m/s on a
component basis using a responding device. Audits using such a device should be performed over at least
three averaging intervals and simulated over a range of normally observed wind speeds.

The responding checks cannot verify that beam steering is being done properly by a phased array Sodar.
Since there are no simple techniques for field verification of the beam angles, it is recommended that a
comparison of the Sodar data with an independent measurement technique be made to assess the
reasonableness of the measurements. This will help identify any major problems with the system such as
improper acoustic beam steering, antenna alignment, etc. The comparison can be made using collocated
adjacent tower data, radiosondes, a tethered balloon, an anemometer kite system, or another remote
sensor.  It is recommended that such comparisons include at least five data pairs collected during times
that encompass at least two different stability regimes. Section 8.2 provides the procedures for
intercomparing data from two different sensors, and Table 8-3 provides comparison criteria.
8.6.3 Radar Wind Profiler Performance Audit

At present, performance audits of radar wind profilers rely on comparisons to collocated or nearby upper-
air measurements to evaluate the performance of the system. Various types of comparison instruments
can be used: tall towers, Sodars, or radiosondes. A tethered balloon can be used, but care should be taken
to ensure that it does not interfere with the radar data. Since it is important to have confidence in the
reference instrument, it must also have an independent verification of operation.  Section 8.2 provides the
procedures for intercomparing data from two different sensors.

Audits using a Sodar (either multi-axis or phased-array) should be performed with the Sodar collocated
with the profiler system and configured to collect data using similar temporal and spatial averaging
periods. The Sodar should be operated for at least 24 hours with the height coverage overlapping the
radar wind profiler data for at least four sampling altitudes. The philosophy behind this approach is that
the operational principles of the profiler are consistent throughout its vertical range, so that "good"
comparisons at the lowest range gates should indicate acceptable performance at higher altitudes as well.
In comparing the data sets, it is important to process the Sodar data in the same manner as the profiler

-------
                                                                                Volume 4, Section 8
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
	Page 26 of 26
data.  This means vertical velocity corrections may be needed to the Sodar data as well as vertical volume
averaging of the Sodar range gates to make them comparable with the profiler range gates. In addition,
the averaging periods of the two systems need to be consistent. When a RASS is in operation, it is likely
the Sodar data will be contaminated and should not be included in the average.

A quick review of the Sodar and radar data should be made prior to dismantling the Sodar to ensure
adequate coverage of overlapping data.  If needed, additional periods of data collection (up to 72 hours or
more) may be required to attain confidence that the systems are both operating correctly and that the
collected data are adequate to characterize the performance of the radar. The  key is to have adequate data
to obtain confidence in the measurement comparison.

Performance audits using a radiosonde system should include at least three soundings collected over the
diurnal  cycle so that a variety of stability conditions are encountered.  A radiosonde sounding provides
vertical coverage over the full operating range of the radar. Its drawbacks include the lack of spatial
consistency as the balloon travels away from the radar site, as well as the instantaneous profile from the
balloon, as opposed to the time-integrated averaged data produced by the radar.

Comparison data from Sodars and radiosondes should be collected when wind speeds are greater than
about 2 m/s. This will help eliminate ambiguities associated with light winds. The systematic differences
at all  levels  should be about ±1 m/s for wind speed and ±10° for wind direction. Comparabilities should
be better than ±2 m/s for speed and ±30° for direction, as discussed in Table 8-3.

To conduct performance audits using an adjacent tall tower, at least one measurement level on the tower
must  fall within the profiler's sample volume. Data should be compared for at least 24 hours, after which
the statistics should be comparable with the criteria in Table 8-2.

If differences exceed the tolerances indicated above or the  data quality objectives of the program, then it
is important to try to understand why the differences occurred before assuming there is a problem with the
profiler. The reasons may lie in unusual meteorological  conditions and/or be  due to different sampling
techniques and data reduction protocols used by the reference instrument.  If differences exist, both data
sets should be examined carefully to determine the cause of the differences. If a problem is identified,
users  should determine if it is an isolated problem only affecting a few data points or a systematic
problem affecting all of the data.  Additional intercomparisons may be needed to resolve the differences.
8.6.4 RASS Performance Audit

As with the radar wind profiler, performance audits of RASS rely on a comparison to a reference
instrument. The recommended method is to use a balloon sounding system to measure the variables
needed to calculate Tv (i.e., pressure, temperature, and relative humidity).  At least three soundings should
be made for an audit during different times of the day to evaluate the performance of the system under
different stability conditions. Soundings should be launched while the RASS is operating to avoid
potential differences between the measurements caused by meteorological variability.  Data collected
from the radiosonde should be  volume averaged into intervals consistent with the RASS averaging
volumes, and values compared on a level-by-level and overall basis. When both the RASS and
radiosonde systems are operating properly, the systematic difference should be within ±1 degrees C and
with an operational comparability of 1.5 degrees C, as listed in Table 8-3.

-------
                                                                                Volume 4, Section 9
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
                                                                                      Page 1 of 14
                9.   Data Acquisition Systems and Meteorology
9.1  Introduction

This section provides information about Data Acquisition Systems (DAS), a term signifying systems that
collect, store, summarize, report, print, calculate, or transfer data. The transfer is usually made from an
analog or digital format to a digital medium.  However, with "new generation" systems, DAS and most
meteorological sensor manufacturers offer a digital option that allows the digital signal to go directly
from the sensor to the DAS. EPA recommends that SLT agencies consider migrating from analog-to-
digital (A/D) to digital-to-digital (D/D) data transfer. Reasons will be discussed in this section.

DAS have been available to air quality professionals since the early 1980s.  Previous to DAS,
meteorological data could only be recorded on strip chart recorders, which had some drawbacks:
   >  Manual reduction of data from the strip charts.
   >  Ink "bleed" that obscured accurate readings.
   >  Ink delivery systems that ran dry, hence, loss of data.
   >  Large amounts of paper needing storage.
   >  If ink did not dry completely, charts adhering to each other, thus obscuring the readings.

Simpler analog  systems utilize the electrical output from a transducer to directly drive the varying pen
position on a strip chart. For some variables, such as wind run (total passage of wind) and precipitation,
the transducer may produce a binary voltage (either "on" or "off") which is translated into an event mark
on the strip chart.  Most SLT agencies either have migrated away from strip chart recorders or only use
these devices in backup operations.

The  first DAS were single- and multi-channel systems that collected data on magnetic media. These
media were usually hand-transferred to a central location or laboratory for downloading to a central
computer. The advent of digital data transfer from the stations to a central location diminished the need
to hand-transfer meteorological data. In addition, since DAS had rapid scan rate capability (i.e., once per
second or less), data collected could also be converted quickly to vector or sigma data. This allowed end
users (i.e., modelers) to input additional data into complex models to understand transport of pollution
downwind of sources.
9.2 DAS Data Acquisition in Analog Layout - Signal Conditioning

Many analog systems and virtually all digital systems require a signal conditioner to translate the
transducer output into a form that is suitable for the remainder of the DAS. This translation may include
amplifying the signal, buffering the signal (which in effect isolates the transducer from the DAS), or
converting a current (amperage) signal into a voltage signal.

-------
                                                                                Volume 4, Section 9
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
                                                                                      Page 2 of 14
9.3 Instrument Connectivity

Technological advances in DAS and meteorological sensors provide analog or digital interface of sensor
signals to the DAS. The A/D and D/D options are discussed in the following sections.
9.3.1 Analog-to-Digital Conversion

A key component of any digital DAS is the A/D converter. The A/D converter translates the analog
electrical signal into a binary form that is suitable for subsequent processing by digital equipment. In
most digital DAS, a single A/D converter is used for several data channels through the use of a
multiplexer. The rate at which the multiplexer channel switches are opened and closed determines the
sampling rates for the channels.  All channels need not be sampled at the same frequency.

Figure 9.1 shows a DAS rear panel with sensor analog signal cable attached to an 8-channel differential
input terminal strip. The meteorological sensor has a DC voltage potential that is proportional to the
specific measurement being recorded. Most meteorological instruments' outputs are in the 0-1 or
0-5 VDC range. The A/D converter basically performs the following functions during the analog
conversion process:

   >    The voltage is measured by the multiplexer which allows voltages from many instruments to be
        read at the  same time.
   >    The multiplexer sends a signal to the A/D converter which changes the analog voltage to a low
        amperage digital signal.
   >    The A/D converter send signals to the central processing unit (CPU) that directs  the digital
        electronic signals to a display or to the random access memory (RAM) which stores the short-
        term data until the end of a pre-defined time period.
   >    The CPU then transfers the data from the RAM to the storage medium which can be magnetic
        tape, computer hard-drive, or computer diskette.
   >    The computer storage medium can be accessed remotely or at the monitoring location.

Data transfer can occur via modem to a central computer storage area or printed out as hard copy. In
some instances, data can be transferred from one storage medium to another storage medium (e.g., hard
drive to a diskette or tape). Due to varying voltages and analog system connections, interferences and
voltage  "leakage" can occur, which can cause errors.  Following is a list of some A/D concerns.
   >    In high sensitivity applications, the signal exists at the bottom of the usable voltage range, and the
        data stream may be affected by noise.
   >    The A/D range may be limited to 10 bit (1024 steps) in some cases.
   >    A/D calibrations may be required to "match" sensor output to DAS input readings.

-------
                                                                                  Volume 4, Section 9
                                                                                     Revision No: 1.0
                                                                                       Date: 10/2006
                                                                                        Page 3 of 14
                         Figure 9.1 DAS Rear Panel with 8-Channel
                         Differential Analog Terminal Strip
9.3.2 Digital Connectivity
         "
D/D transfer has several advantages:
   >   Multiple data types can be transferred with a single connection.
   >   Single cable interface reduces clutter at the DAS rear panel.
   >   Ground loops are eliminated to improve data integrity.
   >   A/D calibrations are not required.
   >   There is additional flexibility in tracking signal over-range conditions.
   >   Digital systems are sensitive to changes in sensor firmware and output formats.
Figure 9.2 shows DAS rear panel with RS-232 signal interface.

-------
                                                                               Volume 4, Section 9
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
                                                                                     Page 4 of 14
                         RS232  OCE
                        PORT A  DTE
                         RS232  DC
                        PORT B  DT
                   Figure 9.2 DAS Rear Panel with RS-232 Signal Interface
9.4 Data Communication
Depending on the type of system, there may be several data communication links.  Typically the output
signals from the transducers are transmitted to the on-site recording devices directly via hardwire cables.
For some applications involving remote locations, the data transmission may be accomplished via a
microwave telemetry system or telephone lines with a dial-up or dedicated line modem system. Also,
wireless internet connection is available for real-time communication with the site DAS.
9.5 Sampling Rates

The recommended sampling rate for a digital DAS depends on the end use of the data.  Substantial
evidence and experience suggest that 360 data values evenly spaced during the sampling interval will
provide estimates of the standard deviation to within 5 percent or 10 percent.2 Estimates of the mean
should be based on at least 60 samples to obtain a similar level of accuracy. Sometimes fewer samples
will perform as well, but no general guide can be given for identifying these cases before sampling; in
some cases, more frequent sampling may be required.

If single-pass processing is used to compute the mean scalar wind direction, the output from the wind-
direction sensor (wind vane) should be sampled at least once per second to ensure that consecutive values

-------
                                                                                Volume 4, Section 9
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
                                                                                      Page 5 of 14
do not differ by more than 180 degrees.  The sampling rate for multi-point analog recorders should be at
least once per minute. Chart speeds should be selected to permit adequate resolution of the data to
achieve the system accuracies recommended in Section 0.2.2. The recommended sampling rates are
minimum values; the accuracy of the data will generally be improved by increasing the sampling rate.


9.6  Meteorological Data Generated by DAS

A number of parameters are generated by state-of-the-science DAS:
   >   Wind speed (WSV or vector), wind direction (WDV or vector)
   >   Wind direction standard deviation (sigma theta)
   >   Standard deviation of the vertical wind speed  (sigma W)
   >   Standard deviation of the vertical wind direction (sigma phi)

Specific equations are used in the calculation of these  variables. These equations are defined in the
following subsections.  Generally, the data will be calculated on a 1 -second scan rate, but reported as an
hourly value averaged from the 3,600 1-second values.


9.6.1 WSV and WDV Calculations

The hourly calculations for WSV and WDV provide a vector average of all the instantaneous samples of
wind direction (WDi) and instantaneous wind speed (WSi) sampled each hour. The following equations
are used.
                                                                                           (9-1)
                                   V = •
WDV = arctan  — + FLOW
                                               N                                           (9.2)


                                                                                           (9.3)
Where:
                             FLOW =
                                      + 180° -» arctanl — | ->< 180°
                                      -180° -> arctanl — I -» 180°
                                                                                           (9-4)
Where  U =   the east-west component
        V =   north-south component of the wind
        N =   the number of instantaneous samples

The signs (positive or negative) of U and V are used to place WDV in the proper sector. Equation 9-3
assumes that the angle returned by the ArcTan function is in degrees.

-------
                                                                                 Volume 4, Section 9
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
                                                                                       Page 6 of 14
9.6.2 Wind Speed and Wind Direction Average (WSA, WDA)

If wind speed average or scalar data are collected, the hourly calculation of WSA is a simple arithmetic
average of the instantaneous wind speed (WSi) samples.  Horizontal wind direction is a circular function
with resultant values in the range of 1 to 360 degrees. Again, the calculation of WDA is a simple
arithmetic average that takes into account the cross-over-corrected average of the instantaneous wind
direction (WDi).
                                                N
                                                                                           (9-5)
                                                                                           (9-6)
Where N is the number of instantaneous samples.
9.6.3 Standard Deviation of Wind Direction (sigma theta)

Sigma theta (GO) is collected for the sole purpose of estimating lower atmospheric stability. The
suggested stability calculation method is detailed in EPA, 20002.  Sigma theta can be calculated from the
instantaneous wind direction values using the basic definition of standard deviation.
                                             N-l
                                                                                           (9-7)
It has been suggested that the upper limit of sigma theta should be limited to 103.9 degrees. Sigma theta
calculated over 60 minutes is influenced by the changing wind direction during the hour. For this reason,
it is recommended that four 15-minute sigma theta calculations be combined to provide a "1-hr" value for
the purpose of selecting a Pasquill-Gifford stability class.  The following method describes how the
hourly values should be calculated.
                                             A(30)
                                                       45  ~*~ ^",4
Where each
     U A(\-hr) '



) equation is a 15-minute deviation of the wind direction.


for example is calculated between 00 and 15 minutes.
                                                                                           (9-8)

-------
                                                                              Volume 4, Section 9
                                                                                 Revision No: 1.0
                                                                                   Date: 10/2006
                                                                                    Page 7 of 14
9.6.4 Standard Deviation of the Vertical Wind Speed (sigma W)

Vertical wind speed is calculated as a backup method for estimating stability by the calculation of sigma
phi (aj).  Sigma phi is calculated from sigma W and wind speed average. Sigma W (aw) is calculated
from the instantaneous values of vertical wind speed (VWSi) and the average vertical wind speed
(VWSA).
                                <7  =
                                            N-l
                                                                                         (9-9)
Where N is the number of instantaneous samples.
9.6.5 Standard Deviation of the Vertical Wind Direction (sigma phi)

Sigma phi (00) is another method for estimating lower atmospheric stability. The suggested stability
calculation method using sigma phi is detailed in Meteorological Monitoring Guidance for Regulatory
Modeling Applications.5
                                             WSA
                                                                                        (9-10)
Equation 9.10 yields sigma phi (a#) in units of radians and must be converted to degrees:

                        Sigma phi (degrees) = Sigma phi (radians) x 57.2958

-------
                                                                               Volume 4, Section 10
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
	Page 1 of 13
                        10.  Meteorological Data Validation
                                   and Verification

Data review, verification, and validation are techniques used to accept, reject, or qualify data in an
objective and consistent manner.  Verification can be defined as confirmation by examination and
objective evidence that specified recommendations have been fulfilled. Validation can be defined as
confirmation by examination and objective evidence that the particular recommendations for a specific
intended use are fulfilled. It is important to describe the criteria for deciding the degree to which each
data item has met its quality specifications as described in an organization's QAPP. This section
describes the techniques used to make these assessments with a focus on meteorological parameters.

In general, these assessment activities are performed by persons implementing the environmental data
operations as well as by personnel "independent" of the operation, such as the organization's QA
personnel, and at some specified frequency. The procedures, designated personnel, and frequency of
assessment should be included in an organization's QAPP. These activities should occur prior to
submitting data to the final repository in EPA's Air Quality System (AQS) and before they are utilized in
models or forecasts. The  following areas of discussion should be considered during verification or
validation processes.
 10.1    General Approach

 How closely a measurement represents the actual environment at a given time and location is a complex
 issue that must be considered during development of the sampling design.  Each sample should be
 checked for conformity to the specifications, including type and location (spatial and temporal). By
 noting the deviations in sufficient detail, subsequent modelers and other data users will be able to
 determine the data's usability under scenarios different from those included in project planning.

 Ambient air pollution data and meteorological data are linked. Pollution either forms chemically in the
 atmosphere or it is the result of a process.  Photochemical pollutants, such as ozone and sulfates, are
 generally produced over a period of time.  Ozone forms by the interaction of VOCs and NOX under the
 right meteorological conditions when low wind speeds, variable wind directions, and relatively high
 temperatures are present.

 Other pollutants are generated by point and area sources.  Winds, a meteorological variable, can transport
 pollutants great distances from their sources to affect populated areas. There are several examples of the
 interaction of ambient air pollution and meteorology in Section 10.4 of this document. Therefore, if
 possible, meteorological data should be validated and verified at the same time as pollution data, not
 separately.

 Figure 10.1 is a simplified illustration of atypical validation and verification process. Three distinct
 columns are illustrated in this schematic. The left column shows the "levels of data". These levels of
 data are described in  detail later in Section 10.1.1. The central column is a visualization of the data flow
 from meteorological  sensors to the AQS. The right column illustrates the type of verification or
 validation that usually occurs during the process. The numbers in parentheses reference the section
 numbers in this document.

-------
                                                                                 Volume 4, Section 10
                                                                                    Revision No: 1.0
                                                                                       Date: 10/2006
                                                                                        Page 2 of 13






























LEVEL OF
DATA



Level 0


Level 1




Level 2


















•
•

•
*
+
i
«
•
*
i
t
«
•
•
•
•
i
•
*

i
-
i
•
*
*
+
i
«
i
-
•
•
«
•
*
•
*
DATA
FLOW
Met
Sensors

i
DAS
i
!
On Site
C ornputer
X
T

C entral
D atabase
(initial)

1

1 »
| >



Central
D atabase
(Final)


1
f
AQS


•
*

•
+
i
i
*
i
___ t
t
t
t
^ *
*
| _-^^'
^^c
^^ 1
1

^'l
^ \
^ :
^-~-^ !
n ^-
t
i
•
~

i
^ '-
>
t
»
i
•
TYPE OF
REVIEW


Visual
In spection
( 10.2.1)


DAS Data
Verification
( 10.2.2)


Manual Rev lew
(10.3 )




( 10.4.3)




D ataBase
Screening
( 10.4.5)



Final
Evaluation
(10.4.6)



















C omparison
Program
(10.4.4)











             Figure 10.1 Generalized Data Validation and Verification Process Flow
10.1.1 Levels of Validation

Generally, there are four "levels" of air quality and meteorological data validation. These levels are
defined by Mueller and Watson59 and Watson et al.60 When a data set has undergone a level of review, it
passes on to the next level.  The process is used to determine the validity of the data.
        Level 0 validated data are essentially raw data obtained directly from the DAS in the field.
        Level 0 data have been reduced and possibly reformatted, but are unedited and un-reviewed.

-------
                                                                              Volume 4, Section 10
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
                                                                                     Page 3 of 13
       These data have not been adjusted for known biases or problems that may have been identified
       during preventive maintenance checks or audits.  These data may be used to monitor instrument
       operations on a frequent basis but should not be used for regulatory purposes.
       Level 1 data validation involves quantitative and qualitative reviews for accuracy, completeness,
       and internal consistency. Quantitative checks are performed by DAS software screening
       programs (see Section 10.2.2), and qualitative checks are performed by meteorologists or field
       staff who manually review the data for outliers and problems.  Quality control flags are assigned,
       as necessary, to indicate the data quality. Data are only considered validated at Level 1 after final
       audit reports have been issued and any adjustments, changes, or modifications to the data have
       been made.
       Level 2 data validation involves comparisons with independent data sets.  This function includes,
       for example, making comparisons to other meteorological or ambient pollution data or upper-air
       measurement systems.
       Level 3 data validation involves a more  detailed analysis and final screening of the data.  The
       purpose of the final step is to verify that there are no inconsistencies among the related data (such
       as problems with scalar and vector data  or problems consistent with temperature and its  related
       relative humidity). Graphics programs may be run to examine the overall consistency among
       related data (i.e., checking diurnal patterns against other parameters or reviewing strip charts for
       final analysis).
10.2    Data Verification Methods

Data verification is defined as the confirmation by examination and objective evidence that specified
recommendations have been fulfilled. These recommendations should be included in the organization's
QAPP and in SOPs. The data verification process involves two basic steps: visual inspection and
analysis and verification performed by DAS. Both techniques are needed to verify meteorological data.
Each is described in the following sections.
10.2.1 Visual Data Verification

Some meteorological data can be verified visually.  For example, under windy conditions, the cup
anemometer and vane system at a monitoring station should move. Rainfall can be measured. Other
visual verification techniques, which include inspections, can be technical systems audits (internal or
external) or frequent inspections by field operators. Several questions might be asked during a visual
verification process:

    1.  Is the equipment performing correctly?  Individual checks such as electronic checks, zero checks,
       and other assessments must have been acceptably performed and documented.
    2.  Did the equipment and its performance pass an initial visual inspection? A station operator
       should look at the meteorological equipment during a site visit. If the wind speed cups are
       spinning and the vane is responding to wind, does the motion correlate to a reasonable value?
       Does the measured temperature appear to be "fairly accurate"?  If the weather is warm enough to
       work without a coat, is temperature above 10 degrees C. Is it clear or overcast? If it is clear, the
       assumption might be that solar radiation would be relatively high. Many environmental samples
       can be flagged (qualified) during the periodic visual inspection.

-------
                                                                               Volume 4, Section 10
                                                                                   Revision No: 1.0
                                                                                     Date: 10/2006
	Page 4 of 13
    3.  Part of the verification process is a review of the meteorological data and the current weather over
        a period of time.  It is important that the station operator review the data collected from the last
        visit. A quick visual inspection may reveal some anomalies that do not match other parameters.
        For example, if an operator sees a high humidity value in the DAS, does it have a corresponding
        relatively high temperature value. If the interactions of the meteorological parameters are
        understood, general conclusions can be drawn. A quick look at the tabular data may illustrate
        anomalies that can be studied during the validation process.

Figure 10.2 is an illustration of a local air pollution agency's monitoring station meteorological sensor
checklist. This example lists items that must be checked by the station operator every time that he/she
visits the site. A discussion of the items on the checklist follows:
   >    Tower Check.  The tower check is a visual inspection of each sensor on the meteorological tower.
        Each time an operator visits, he/she should look every external instrument. Cups, vanes,
        temperature shields, and NEMA 4 enclosures should be checked for damage or possible blockage
        by animals or debris.
   >    Wind Check.  The wind speed and direction are estimated by the operator and marked on the
        form. Once this estimation is performed,  the operator notes the reading from the DAS.  Do the
        estimates and the readings match reasonably? The values recorded by the operator should be
        checked against the strip chart recorder. Do they match?
   >    Temperature Check. A sling, motorized psychrometer should be operated, and the temperature
        and relative humidity should be checked against the DAS reading.
   >    Sky Check. Recording sky conditions helps to determine whether the solar and UV radiation
        values are relatively accurate. Other useful checks are to visually check for passing clouds.  If
        clouds obscure the sunlight, UV and solar radiation values should drop quickly.  If the operator
        arrives early in the morning, solar radiation values will increase slowly over an hour's time,
        verifying that the sensor is working.
   >    Rainfall Check. The tipping bucket rain gage should be checked to determine if the funnel
        opening is free of debris; if not, the funnel should be cleaned. The relay switch on the gauge
        should be checked by manually tipping the bucket apparatus 10 times; that the DAS recorded the
        equivalent of 10 tips should be confirmed. The rainfall channel on the DAS must be taken "off-
        line" prior to performing the manual check to avoid recording erroneous rainfall data and returned
        to "on-line" status after completion of the manual check.
   >    Barometric Pressure.  The barometric pressure reading should be checked for reasonableness with
        the local NWS or a reliable second barometer. The hourly data from the DAS should reflect a
        rising or falling barometer based the advances of high or low pressure frontal passages

-------
                                                                             Volume 4, Section 10
                                                                                 Revision No: 1.0
                                                                                   Date: 10/2006
                                                                                    Page 5 of 13

Weekly Quality Control Check Sheet
Meteorological Instruments
Site Month/Year
Site Number
Date:
Tower Check:
Crossarm aligned with north?
WS cups okay?
WD vane okay?
VWS propeller okay?
T& RH shield okay?
Solar radiation okay?
UV radiation okav?


Technician



















































Wind Check:
WS estimate (1 )
WS DAS (mls\
WS Chart %
WD estimate O\ fdeg'l
WD DAS Cdegl
WD Chart %

mps
%
deg
deg
%

mps
%
deg
deg
%

mps
%
deg
deg
%

mps
%
deg
deg
%

mps
%
deg
deg
%

mps
%
deg
deg
%

Temperature Check:
T dry (deg C^
T wet Cdea C1
RH % calculation (3\
Temp DAS Cdeg C1
RH % DAS
degC
degC
%
degC
%
degC
degC
%
deaC
%
degC
degC
%
deaC
%
degC
degC
%
deaC
%
degC
degC
%
deaC
%
degC
degC
%
deaC
%

Sky Check:
Sky condition (41
SRD DAS (mly/101
UVDAS (mM


















All comments must be noted in the station log.
(1) WS estimate (3) Calculated RH
C = calm (0-1 mps) From graph
L = light (1 -3 mps) on reverse side
M = moderate (4-6 mps)
S = strong (> 6 mps) (4) Sky condition (choose 1 or more)
CLR (clear) F (fog)
(2) WD estimate PC (partly cloudy) S (smog)
N, NE, E, SE CLDY (cloudy) H (haze)
S, SW, W, NW OVC (overcast) R (rain)
Reviewed by Date


     Figure 10.2 Example of Meteorological Sensor Visual Check List
10.2.2 Data Verification Performed by DAS

In the late 1980s, DAS became available to the air pollution monitoring community. These early systems
could collect voltage data (analog information) and convert those data to a digital format.  Later DAS

-------
                                                                               Volume 4, Section 10
                                                                                  Revision No: 1.0
                                                                                     Date: 10/2006
	Page 6 of 13
versions had the ability to take instant readings from wind instruments and calculate the vector and sigma
(i.e., the standard deviation) of wind data. Some DAS have been programmed with the ability to sense
out of the ordinary changes in the signal that humans would not be able to detect. These types of
verification techniques can be extremely useful because the program can "sense" a change in operating
conditions when or even before equipment fails.  It is strongly recommended that these DAS verification
checks be a part of the data acquisition routine and examined by station operators and data validation
staff. Listed below are data verification questions that should be considered to allow the DAS to perform
automated verification checks.
   >   Did the value exceed the DAS maximum reading limit?
   >   Did the value exceed the DAS minimum reading limit?
   >   Did the value exceed the maximum rate of change?
   >   Does the data have high alarm limit?
   >   Does the data have a low alarm limit?
   >   Does the data have a floor limit?
   >   Does the data need to have a percentage of valid readings to be considered valid?
   >   Is there a ceiling limit to the data?
   >   Does the data need to be interpolated? If so, is it linear?
   >   Does the system have a minimum detectable limit?

Commercially available DAS should have the capability to assign alarm limits for instantaneous and
hourly values. This is important because rapid changes may not be flagged for instantaneous values, but
could be flagged for an hour value and vice versa.  Some of the information that is referenced in the table
should be set by annual or even quarterly averages. These values also can be refined with time.

Table 10-1 provides recommended automated screening limits on DAS; assuming the wind speed system
operating range is 0-50 m/s, with a starting threshold of 0.2 m/s.

Table 10-1 DAS Screening Techniques
Parameter
Maximum reading
Minimum Reading
Maximum Rate of
Change
High Alarm Limit
Low Alarm Limit
Floor Limit
Percent Valid Readings
Ceiling Limit
Linearity
Minimum Detectable
Limit
Value
50 m/s
0.2 m/s
10 m/s
40 m/s
0.2 m/s
0.5 m/s
75%
30 m/s
1.00+/- 1%
0.2 m/s
Comments
This is the maximum instantaneous reading possible for your
system
This is the minimum instantaneous reading possible for your
system because your threshold is 0.2 m/s
Winds generally increase gradually, for a particular site, you
would not expect winds to change this quickly
This is gale force winds. You would expect damage to sensor or
site.
Same as starting threshold
This value is lowest average that you would expect for one hour
Most data should be valid 75% of the time
This value is highest average that you would expect for one hour
You would expect your sensor to be within 1% accuracy.
This is the same as the starting threshold.

-------
                                                                               Volume 4, Section 10
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
	Page 7 of 13
 10.3    Manual Review Methods

 Figure 10.1 shows four types of validation procedures: manual review, graphic analysis, comparison, and
 database screening. Both manual and computer-oriented systems require individual reviews of all data
 tabulations.  This initial step should be performed by the station operator and later by data validation staff.
 It is recommended that the data be printed out at the monitoring station and reviewed by the site operators
 every time they visit the station or at the end of the month.  New DAS allow station operators to print the
 data in a tabular format. These tables can also provide summary data, i.e., highest, lowest and average
 values.  The purpose of manual inspection is to spot unusually high (or low) values (outliers) that might
 indicate a gross error in the data collection.

 Manual review of data tabulations also allows detection of uncorrected drift in the zero baseline of a
 meteorological sensor. Zero drift may be indicated when the daily minimum values tend to deviate
 (increase or decrease)  from the expected minimum value over a period of several days. Usually, winds in
 the early morning hours (3 a.m. to 4 a.m.) are light and variable,  solar radiation is at its lowest values, and
 temperatures generally drop which may result in zero baseline drift from temperature, solar radiation, and
 wind speed sensors.

 In an automated data processing system, procedures for data validation can be incorporated into the basic
 software. As noted in Section 10.2.2, the computer can be programmed to scan data for extreme values,
 outliers,  or ranges. These checks can be further refined to account for time of day, time of week, and
 other cyclic conditions.  Questionable data values flagged on the data tabulation may or may not indicate
 possible  errors. The station operator should check all the data flagged by the DAS data verification
 program and investigate whether the data flagged should remain  flagged. In some cases, extreme
 meteorological conditions can occur rapidly, and the data may actually reflect real values.  For example,
 if a thunderstorm moves through an area, winds can be recorded  as calm but reach 20-30 m/s within
 seconds. This extreme value could be flagged by a DAS verification program.  The station operator
 should note such examples and alert the data validation staff that these data actually reflect conditions.
 10.3.1 Calibration Data Review

 Calibration of instruments and equipment must be performed periodically as referenced in the MQOs
 tables in Section 0.2.2 in this document.  The associated data should be reviewed by station operators and
 data validation staff.  The following questions should be answered:
   >   Were the calibrations performed within an acceptable time prior to generation of data?
   >   Were they performed in the proper sequence?
   >   Were the proper number of calibration points performed?
   >   Were they performed using standards that "bracketed" the range of reported measurement results?
   >   Were acceptable linearity checks and other checks made to ensure that the measurement system
        was stable when the calibration was performed?


 10.4    Data Validation Methods

 Data validation is a routine process designed to ensure that reported values meet the quality goals of
 environmental data operations. A progressive, systematic approach to data validation must be used to
 ensure and assess the quality of data. The purpose of this  step in the process is to detect, compare, and

-------
                                                                                 Volume 4, Section 10
                                                                                    Revision No: 1.0
                                                                                      Date: 10/2006
	Page 8 of 13
perform a final screening on all data values.  Any final data that may not represent actual meteorological
conditions at the sampling station will be detected at this stage. Effective data validation procedures
usually are handled independent of the procedures of initial data verification, that is, by different
computer systems and staff. Note in Figure 10.1 that the validation process is performed on the data set at
the location of the central database. Data validation staff should be independent of station operations
personnel; this level of independence is important, but not necessary. These procedures can be performed
automatically; SAS®, Microsoft Excel®, or Visual Basic program languages can be scripted to perform
these tests.  Several local agencies have employed these techniques quite effectively.

If data assessment results clearly indicate a serious response problem with the sensor, the agency should
review all related information to determine whether the meteorological data, as well as any associated
assessment data, should be invalidated. For example, if a temperature sensor fails a calibration or audit,
the relative humidity data, which are calculated from the ambient temperature data, must be invalidated as
well. In addition, in the relationship between wind vector and scalar data, if wind speed average data (or
wind direction average data) are considered invalid, then the vector data for both wind  speed and
direction data must also be invalidated since the vector data use both wind speed average and direction for
the calculations.

Some problems that may escape detection during an audit (a wind vane that occasionally sticks) are often
easily identified during data validation. Data validation should be performed by a person with appropriate
training in meteorology who has a basic understanding of local meteorological conditions and the
operating principles of the instruments.
 10.4.1 Preparatory Steps

 Steps preparatory to data validation should include the daily transfer of raw data (e.g. 1-minute or 1-hour
 average data files) to a central data processing facility and the transfer of raw data files to create an edited
 database. The raw data files should be stored separately to insure data integrity. Backup copies of the
 data should be prepared and maintained on-site and off-site.
 10.4.2 Validation Procedures

 All necessary supporting material, such as audit reports and site logs, should be available for Level 2
 validation. Access to a daily weather archive should be provided for use in relating suspect data to local
 and regional meteorological conditions. Questionable data, such as data flagged in an audit, manual
 review, or DAS screening program, should be corrected or invalidated during Level 2 data validation.
 10.4.3 Graphic Analysis

 Graphing data can be a quick method of visualizing the data relative to other parameters. Graphs can
 show longer term trends and relationships that are difficult to see when data validation staff are looking at
 large amounts of tabular data. Figures  10.3 and 10.4 illustrate the relationship between different
 parameters.

 Figure 10.3 illustrates the relationship of relative humidity to ambient temperature.  During the time
 period, there was an inverse relationship between the two parameters. By graphing the data, these trends
 are clear.

-------
                                                                               Volume 4, Section 10
                                                                                  Revision No: 1.0
                                                                                     Date: 10/2006
                                                                                     Page 9 of 13
                EPA's Burden's Creek Station, NC
               120
-TEMP(2M)(degc)
 RH (%)
             Figure 10.3 Graphic Example of Temperature vs. Relative Humidity
Figure 10.4 illustrates the relationship of ozone to ambient temperature during the period.  As the
temperature increased with time, ozone also increased. Both sets of data illustrate a diurnal nature.
                   EPA's Burden's Creek Station, NC
— •— O3-API(ppb)
-•— TBUP (2M) (degc)
                Figure 10.4 Graphic Example of Ozone vs. Temperature
10.4.4 Comparison Program

A useful data validation method is to compare the difference between successive meteorological data
values.  Logic dictates that rapid changes in values in a 1-hr reporting period would normally not be

-------
                                                                               Volume 4, Section 10
                                                                                  Revision No: 1.0
                                                                                     Date: 10/2006
                                                                                        ilOof 13
expected.  When the difference between two successive values exceeds a predetermined value, the data in
question can be flagged.

Another useful tool is a comparison of data from one monitoring location with a nearby meteorological
collection station. Randomly selected values should be manually compared with other available, reliable
data (such as data obtained from the nearest NWS observing stations).  Several hours out of every five
days should be randomly selected. To account for hour-to-hour variability and the spatial displacement of
the NWS station, a block of several hours may be more desirable. All data selected should be checked
against corresponding measurements at the nearby station. In addition, monthly average values should  be
compared with climatological norms, as determined by the NWS from records over a 30-year period. If
discrepancies are found which can not be explained by the geographic difference in the measurement
locations or by regional climatic variations, the data should be flagged as questionable.
10.4.5 Data Screening

Screening procedures generally include comparisons of measured values to upper and lower limits; these
limits may be physical, such as an instrument threshold, or may be established based on experience or
historical data.  Other types of procedures employed in screening include assessments based on the rate of
change of a variable (data that change too rapidly or not at all are flagged as suspect) and assessments
based on known physical principles relating two or more variables (e.g., the dew point should never
exceed the dry-bulb temperature).

Screening is an iterative process in which range checks and other screening criteria are revised as
necessary  based on experience.  For example, an initial QA pass of a data set using default criteria may
result in flagged values which, upon further investigation, are determined to be valid for a particular site.
In such cases, one or more follow-up QA passes using revised criteria may be necessary to clearly
segregate valid and invalid data.

10.4.5.1 Data Screening Qualifiers

This section lists a number of qualifiers that should be used to determine reasonableness, completeness,
accuracy, and representativeness of a meteorological database.  The Sample Meteorological Data
Screening Checklist, Figure 10.5, should be used to document the data screening results.
   >   Was the 75 percent data recovery criterion met?
   >   Do wind speed vector and wind direction vector have same data recovery rate? Wind speed and
       wind direction vectors are interdependent; when one is invalid, the other should also be invalid.
       Wind speed and wind direction vectors interdependence and other items are easily verified using
       the database.
   >   For meteorological multi-point calibration, was the wind speed slope equal to 1 ± 0.05m/s?
       Review any meteorological calibration forms. The slope of the wind speed calibration should be
       within ±0.05 of 1.00.  The intercept should be within ±0.3 m/s.
   >   Was the wind direction (WD) average  difference equal to ±5 degrees? All wind direction values
       should be within 5 degrees of true north.
   >   Was the vertical wind speed (VWS) maximum difference < 0.3m/s?  VWS points should be
       within 0.3 m/s of true.
   >   Was the temperature maximum difference less than  <1°C?  Temperature sensor response should
       be within ±1.0 degrees C of CTS response. The audit and calibration criteria are essentially the

-------
                                                                       Volume 4, Section 10
                                                                           Revision No: 1.0
                                                                             Date: 10/2006
                                                                             Page 11 of 13
same and represent the accuracy recommendations of temperature data. Temperature data
collected with sensors that fail to meet audit and calibration criteria must be corrected or
invalidated.  Temperature data can be corrected using correction factors derived from the most
recent calibration results.
Was vector wind speed less than or equal to scalar wind speed? Spot check the monthly
summaries of vector wind speed and average wind speed to verify that vector wind speed is
always less than or equal to average wind speed.  In most cases, the values should be very close
to each other. If a discrepancy exists, resolve it by reviewing the 1-minute average data from the
electronic  strip chart graphs (if available).
Are average wind direction and vector wind direction comparable? Spot check the monthly
summaries of vector wind direction and average wind direction values.  The values should be
within 5 to 10 degrees of each other.  If a large discrepancy exists, the sigma theta  for the period
should also be high.
Are sigma W and vertical wind speed comparable? A preliminary check should show that low
(near zero) sigma W values were accompanied by low vertical wind speed values.  The inverse
may not be true.
Are sigma V and WS*sin(sigma theta) comparable?  Check a few of the sigma v values.
concentrating on the high values. A more reliable algorithm is:

                            (sigma v)2 = WSA2 - WSR2
Estimated and reported values should usually compare to within 10 percent of each other.
Are there any hand-reduced sigma or vector data? Hand-reduced sigma and vector data are
invalid.
Do sigma  theta values and wind direction chart scatters coincide?  In general, the higher the
scatter, the higher the sigma.  A more quantitative comparison can be made for any sigma
parameter by estimates of the range of the scatter for the hour divided by 4. This calculation
should be  made occasionally during the review for both sigma theta and sigma W.
Is the zero offset for wind speed correct?  The chart trace should have a zero offset built in,
corresponding to about 0.5 m/s in order to verify chart-to-data comparison accuracy.

Are wind speed noise/bearings at threshold speeds? Generally, the wind speed trace, starting
from calm conditions, should increase steadily. If the wind speed trace consistently starts up in a
step  manner, the sensor may be experiencing a high starting threshold. A step of 1 m/s or greater
indicates a potential problem, and data may need to be invalidated if the step consistently
approaches 2 m/s. To avoid data bias, wind speed data should only be invalidated  if the problem
is considered significant enough to invalidate all data until the bearings are replaced. VWS
bearings should also be checked similarly, using approximately a 0.3 m/s criterion  to indicate a
problem. Wind direction bearing problems are identified by flat wind direction traces when other
wind sensors (WS and VWS) indicate measurable wind.
Do wind speed vector (WSV) and wind direction vector (WDV) have the same data recovery
rates?
Are average WD and vector WD comparable?
Does wind speed average (WSA) affect wind speed vector (WSV). wind direction  vector (WDV).
sigma phi, and sigma v?
Does WDA affect WDV. WSV. sigma theta. and sigma v?

-------
                                                                                Volume 4, Section 10
                                                                                   Revision No: 1.0
                                                                                      Date: 10/2006
	Page 12 of 13
 10.4.6 Final Evaluation

 Data flagged by the DAS screening and comparison programs should be evaluated by personnel with
 meteorological expertise.  Reasons for changes in the data resulting from the validation process should be
 documented. If system problems are identified, corrective actions should also be documented.  Edited
 data should continue to be flagged so that their reliability can be considered in the interpretation of the
 results of modeling analyses for which the data are used.

 Flags can be used in the field and at the data management center to signify data that may be suspect due
 to calibration or audit failure, special events, or failed QC limits. When calibration problems are
 identified, data produced between the suspect calibration event and subsequent recalibration should be
 flagged.  Because flag combinations can be  overwhelming and cannot always be anticipated, an
 organization needs to review these flag combinations to determine whether single values or values from a
 site over a particular time period should be invalidated.  Procedures for screening data for possible errors
 or anomalies should also be implemented. When calibration problems are identified, data produced
 between the suspect calibration event and any subsequent recalibration should be flagged to alert data
 users.

-------
                                                                            Volume 4, Section 10
                                                                                Revision No: 1.0
                                                                                  Date: 10/2006
                                                                                  Page 13 of 13
                           Meteorological Data Screening Checklist
Monitoring Station
Month/Year
Data Screening Questions
	Was percent data recovery criteria met (i.e., greater than 75%)
	Do WSV & WDV have same data recovery?
	Do WSV and WDV have same hour invalid?
	Review missing data summary
	Review site logs.
	Multipoint calibration: WS slope = 1±0.05 m/s, WD average difference = ±5 degrees
	VWS maximum difference <0.3m/s, T maximum difference <1 degree C?
	Are all calibrations documented?
	Vector WS < Average WS?
	Compare average WD to vector WD?
	Sigma phi calculation OK?
	Compare sigma W vs. VWS?
	Compare sigma V vs. WS*sin(sigma theta)?
	Any hand reduced sigma or vector data?
	Check dependent parameters:
	WSA affects WSV, WDV, sigma phi, sigma V
        _WDA affects WDV, WSV, sigma theta, sigma V
        _VWS affects sigma W, sigma phi
Strip Charts (paper or digital)
     Random check of all meteorological parameters - (chart vs. data)
     Random check of all calibrations - (chart vs. data)
	Do sigma theta values and WD chart scatters coincide? Zero offset for WS OK?
	Check WS noise/bearings at threshold speeds

Figure 10.5 Example of a Meteorological Data Screening Checklist

-------
This page intentionally left blank.

-------
                                                                             Volume 4, Section 11
                                                                                 Revision No: 1.0
                                                                                   Date: 10/2006
                                                                                     Page 1 of 5
                                   11.  References
1.      U.S. Environmental Protection Agency. "Quality assurance handbook for air pollution
       measurement systems. Volume IV - meteorological measurements." EPA/600/4-90-003, out of
       print, Report prepared by the Office of Research and Development, Atmospheric Research and
       Exposure Assessment Laboratory, U.S. Environmental Protection Agency, Research Triangle
       Park,NC. 1983.

2.      U.S. Environmental Protection Agency. "Quality assurance handbook for air pollution
       measurement systems. Volume IV, meteorological measurements"; EPA/600/R-94/038d, Report
       prepared by U.S. Environmental Protection Agency, Research Triangle Park, NC. 1995.

3.      U.S. Environmental Protection Agency. "Guidance for the data quality objectives process (QA/G-
       4)"; EPA 240/B-06/001, EPA document 2006.

4.      U.S. Environmental Protection Agency. "Ambient monitoring guidelines for prevention of
       significant deterioration (PSD)"; EPA-45 0/4-97-007, Report prepared by Office of Air Quality
       Planning and Standards, U.S. Environmental Protection Agency, Research Triangle Park, NC.
       1987.

5.      U.S. Environmental Protection Agency. "Meteorological monitoring guidance for regulatory
       modeling applications"; Document EPA-454/R-99-005, Office of Air Quality Planning and
       Standards, Research Triangle Park, NC, 2000.

6.      National Weather Service. "Instrument requirements and standards for the NWS surface
       observing programs"; NWSI 10-1302, 2005.

7.      Stealth telescoping poles. GeoData Systems Management, I., 2006. Available at:
       .

8.      T-l 75 Towers (standardduty aluminum crank-up towers). Aluma Tower Co., 2006. Available at:
       .

9.      QTMMast. Clark Masts, 2006. Available at: .

10.    Climatronics Corporation, 2006. Available at: .

11.    World Meteorological Organization Measurement of upper wind. Guide to meteorological
       instrument and observing practices.  8TP3,  4th edition, 1971.

12.    MacCready, P. B. "Wind speed measurements in turbulence", J. Appl. Meteor.  1966, 5, 219-225.

13.    Wind mark III wind sensor. Climatronics Corporation, Equipment specifications prepared by
       Climatronics Corporation, Bohemia, NY, 2002. Available at:
       .

14.    Meteorological instruments, wind sensors. R.M. Young Company, 2006. Available at:
       .

15.    Gill, G.  C. "The helicoid anemometer", Atmosphere 1973, 11(4), 145-155.

-------
                                                                             Volume 4, Section 11
                                                                                 Revision No: 1.0
                                                                                   Date: 10/2006
                                                                                     Page 2 of 5

16.    Met One Instruments, Inc. "Solid state wind sensor"; Equipment specifications prepared by Met
       One Instruments, Inc., Rowlett, TX. 2001.

17.    Campbell Scientific, Inc. "Weather station siting and installation tools"; Prepared by Campbell
       Scientific, Inc., Logan, UT. 1997.

18.    Blackadar, A. F. Almanac for a weather station; He Idref Publications; Washington, B.C., 1985.

19.    Sun or moon altitude/azimuth table for one day. U.S. Naval Observatory, 2006. Available at:
       .

20.    Finklestein, P. L.; Kaimal, J. C.; Gaynor, J. E.; Graves, M. E.; Lockhart, T. J. "Comparison of
       wind monitoring systems. Part  I: in situ sensors", J. Atmos. And Oceanic Technol.  1986, 3, 583-
       593.

21.    Lockhart, T. J. "Accuracy of the collocated transfer standard method for wind instrument
       auditing", J. Atmos. And Oceanic Technol.  1989, 6, 715-723.

22.    ASTM International. "Standard practice for determining the operational comparability of
       meteorological measurements"; D4430-84, Prepared by the American Society for Testing and
       Materials, Philadelphia, PA. 1984.

23.    Lockhart, T. J. "Some cup anemometer testing methods", J. Atmos. And Oceanic Technol. 1985,
       2, 680-683.

24.    ASTM International. "Standard test method for determining the performance of a sonic
       anemometer/thermometer"; D6011-96, American Society of Testing and Materials, Philadelphia,
       PA. 2001.

25.    Baxter, R. A.; Yoho, D. L.; Durkee, K. R. "Quality assurance audit program for surface and
       upper-air meteorological measurements in the South Coast Air Basin of California". In
       Proceedings, American Meteorological Society's 83rd Annual Meeting, 12th Symposium on
       Meteorological Observations and Instrumentation, Long Beach, CA, February 8-11, 2003.

26.    Robertson, J. H.; Katz, D. I. "Climatronics1 novel sonic anemometer". In Proceedings, 9th Joint
       Conference on the Applications of Air Pollution Meteorology, January 28-February 2, Atlanta,
       GA, 1966.

27.    Standard practice for measuring surface atmospheric temperature with electrical resistance
       temperature sensors, D6176-97(2003). ASTM International, 2006. Available at:
       .

28.    Standard specification for ASTM liquid-in-glass thermometers, El-05. ASTM International,
       2006. Available at:  .

29.    ASTM International. "Standard test method of measuring humidity with cooled-surface
       condensation (dew-point) hygrometer, D4230-02el"; For Annual Book of ASTM Standards

-------
                                                                             Volume 4, Section 11
                                                                                 Revision No: 1.0
                                                                                   Date: 10/2006
                                                                                     Page 3 of 5

       volume information, refer to the standard's Document Summary page on the ASTM website,
       . 2002.

30.    Standard specification for industrial platinum resistance thermometers, E1137/E1137M-04.
       ASTM International, 2006.  Available at: .

31.    Weiss, L. L. "Relative catches of snow in shielded and unshielded gauges at different wind
       speeds", Mon. Wea. Rev. 1961, 89.

32.    National Physical Laboratory A guide to the measurement of humidity; National Physical
       Laboratory and the Institute of Measurement and Control; London,  England, 1996.

33.    ASTM International. "Standard test method for measuring humidity with a psychrometer (the
       measurement of wet- and dry-bulb temperatures), E337-02"; For Annual Book of ASTM
       Standards volume information, refer to the standard's Document Summary page on the ASTM
       website, . 2002.

34.    ASTM International. "Standard practice for maintaining constant relative humidity by means of
       aqueous solutions, 104-02"; For Annual Book of ASTM Standards volume information, refer to
       the standard's Document Summary page on the ASTM website, . 2002.

35.    American Society for Testing and Materials. "Calibration of secondary reference pyranometer
       and pyrheliometer for field  use"; E816, Philadelphia, PA. 1984.

36.    Latimer, J. R. "Radiation measurement"; Technical Manual Series No. 2 International Field Year
       for the Great Lakes, Canadian National Commission for the Hydrological Decade 1972.

37.    World Meteorological Organization. "Guide to meteorological instruments and methods of
       observation. Preliminary seventh edition"; WMO-No. 8, Prepared by World Meteorological
       Organization, Geneva, Switzerland. 2006.

38.    Liu, H.; Darkow, G.  "Wind effect on measured atmospheric pressure", J. Atmos. And Oceanic
       Technol. 1989, 6(1), 5-12.

39.    Miksad, R.  "An omni-directional static pressure probe", J. Meteor. Society of Japan 1976, 15,
       1215-1225.

40.    U.S. Weather Bureau Manual ofbarometry (WBAN); First edition ed.; U.S. Government Printing
       Office; Washington, D.C., 1963,  1.

41.    Beyrich, F. "Mixing height  estimation from sodar data - a critical discussion", Atmos. Environ.
       1997,31,3941-3953.

42.    Holzworth, G. "Mixing heights, wind speeds, and potential for urban air pollution throughout the
       contiguous  United States"; Publication No. AP-101, Prepared by the Office of Air Programs, U.S.
       Environmental Protection Agency, Research Triangle Park, NC. 1972.

-------
                                                                              Volume 4, Section 11
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
                                                                                      Page 4 of 5

43.    Gaynor, J. E. "Accuracy of sodar wind variance measurements", Int. J. Remote Sens. 1994, 15,
       313-324.

44.    van de Kamp, D. W. "Profiler training manual #2: quality control of wind profiler data"; Report
       prepared for the National Weather Service Office of Meteorology by NOAA/WPL, Boulder, CO.
       1988.

45.    Ecklund, W. L.; Carter, D. A.; Balsley, B. B. "A UHF wind profiler for the boundary layer: brief
       description and initial results", J. Atmos. Ocean.  Technol. 1988, 5, 432-441.

46.    U.S. Environmental Protection Agency. "Quality assurance handbook for air pollution
       measurement systems, Volume IV, meteorological measurements"; (EPA/600/4-90/003),
       Prepared by the U.S. Environmental Protection Agency, Office of Research and Development,
       Atmospheric Research and Exposure Assessment Laboratory, Research Triangle Park, NC. 1990.

47.    U.S. Environmental Protection Agency. "On-site meteorological program guidance for regulatory
       modeling applications"; EPA-450/4-87-013, Report prepared by the Office of Air and Radiation,
       Office of Air Quality Planning and Standards, U.S. Environmental Protection Agency, Research
       Triangle Park, NC. 1987.

48.    Thuillier, R. H. "The influence of instrumentation, siting, exposure height, and temporal
       averaging methodology on meteorological measurements from SJVAQS/AUSPEX", J. Appl.
       Meteor. 1995,34, 1815-1823.

49.    Crescenti, G. H. "The degradation of Doppler sodar performance due to noise: a review", Atmos.
       Environ. 1998, 32, 1499-1509.

50.    Crescenti, G. H. "Examples of noise interference on Doppler sodar performance". In
       Proceedings, Tenth Symposium on Meteorological Observations and Instrumentation: Phoenix,
       AZ, January 11-16, 1998.

51.    Baxter, R. A. "Development of a universal acoustic pulse transponding system for performance
       auditing of sodars". In Proceedings, Proceedings from the  7th International Symposium on
       Acoustic Remote Sensing and Associated Techniques of the Atmosphere and Oceans, Boulder,
       CO, October, 1994, pp 3-35-33-40.

52.    U.S. Environmental Protection Agency. "Quality assurance handbook for air pollution
       measurement systems. Vol. I, a field guide to environmental quality assurance"; EPA/600/R-
       94/038a, Report prepared by U.S. Environmental Protection Agency, Research Triangle Park,
       NC.  1994.

53.    Wilczak, J. M.; Strauch, R. G.; Weber, B. L.; Merritt, D. A.; Ralph, F. M.; Jordan, J. R; Wolfe,
       D. E.; Lewis, L. K.; Wuertz, D. B.; Gaynor, J. E.; McLaughlin, S.; Rogers, R.; Riddle, A.; Dye,
       T. "Contamination of wind profiler data by migrating birds: characteristics of corrupted data and
       potential solutions", J. Atmos. Ocean. Technol. 1995, 12(3), 449-467.

54.    Gauthreaux, S. A., Jr. "The flight behavior of migrating birds in changing wind fields: radar and
       visual analyses", Amer.  Zool 1991, 31, 187-204.

-------
                                                                               Volume 4, Section 11
                                                                                  Revision No: 1.0
                                                                                    Date: 10/2006
                                                                                         : 5 of 5
55.    Brewster, K. A. "Profiler training manual #2: quality control of wind profiler data"; Report
       prepared for Office of Meteorology, National Weather Service by NOAA/ERL/FSL, Boulder,
       CO. 1989.

56.    Gaynor, J. E. "Siting guidance for boundary layer meteorological profilers", Presented at Air &
       Waste Management Association Specialty Conference on Air Toxics, Durham, NC, 1994

57.    Miller, P. A.; Schlatter, T. W.; van de Kamp, D. W.; Earth, M. F.; Weber, B. L. "An unfolding
       algorithm for profiler winds", J. Atmos. Ocean. Technol. 1994, 11,32-41.

58.    Riddle, A. C.; Angevine, W. M.; Ecklund, W. L.; Miller, E. R.; Parsons, D. B.; Carter, D. A.;
       Gage, K. S. "In situ and remotely sensed horizontal winds and temperature intercomparisons
       obtained using Integrated Sounding Systems during TOGA CQARE",Atmos. Phys. 1996, 69, 49-
       61.

59.    Mueller, P. K.; Watson, J. G. "Eastern regional air quality measurements. Vol.  1, Section 7";
       Report No. EA-1914, Report prepared by Electric Power Research Institute, Palo Alto, CA.  1982.

60.    Watson, J. G.; Lioy, P. J.; Mueller, P. K. "The measurement process: precision, accuracy, and
       validity". In Proceedings, Air Sampling Instruments for Evaluation of Atmospheric
       Contaminants; 7th ed.; Hering, S. V., Ed.; American Conference of Governmental Industrial
       Hygienists: Cincinnati, OH, 1989, pp 51-57.

61.    Code of Federal Regulations, Title 40, part 58, Appendix A.4

-------
                                     Appendix A
                  Meteorological Systems Audit Evaluation Form
This appendix contains a meteorological systems audit evaluation form to be used as a guideline to
evaluate the operation and exposure of meteorological sensors and overall condition of a monitoring site.

-------
Meteorological Systems Audit Evaluation Form


This site systems audit form can be used as a guideline to evaluate the operation and exposure of
meteorological sensors and overall condition of a monitoring site. System audits should be performed on
a yearly basis.

I. On Site Equipment and Location

A. Meteorological Parameters On-site
Parameter










Manufacturer










Model #










Serial #










Range










    1.  Are there any required parameters which are not monitored?	
    2.  Are any methods and equipment unacceptable?	
    3.  Are any operating ranges improper?	
    4.  Are there any significant differences between instrumentation on site and the monitoring
       plan?	
    5. What is the GPS coordinates for this monitoring site	N	W
    6. Does the site have an AQS Site Code?	
    7. What is the magnetic Declination at this site?	

Comments:

-------
B. Auxilliary Equipment
Type
DAS System
Chart Recorder
Chart Recorder
Meteorological
Tower
Monitor Shelter
Computer
Power Conditioner
WS motors
Compass
Thermometer
Psychrometer
Precipitation Gauge
Solar Radiation
Manufacturer













Model #













Serial #













Calibration
Date













Comments.

-------
II. Instrument Height and Exposure




A. Meteorological Distance
Parameter
1. Height of WS sensor above ground
2. Height of WD sensor above ground
3. Height of VWS sensor above ground
4. Distance to nearest obstacle
5. Is exposure outside 10X obstruction height
5a. Are instruments on a rooftop?
5b. Is exposure 1.5X height above the roof?
6. Arc of unrestricted flow?
7. Height of temperature sensor above ground
8. Distance of temperature sensor from obstacles
9. Is the distance 4X from obstruction height?
10. Arc of unrestricted air flow
1 1 . Distance to nearest paved road
12. Is temperature sensor shielded/motor aspirated?
13. Height of precipitation gauge above structure
14. Distance to nearest obstacle
15. Is exposure outside 2-4X obstruction height?
16. Height of solar radiation above structure
17. Distance to nearest obstacle
18. Will solar radiation sensor fall within a shadow
19. Height of dew point sensor above structure
20. Distance to nearest obstacle
2 1 . Is exposure outside 2-4X obstruction height
22. Is temperature sensor below representative terrain?
23. Is temperature sensor pointed downward?
Are there any significant differences between on-site
equipment and the monitoring plan?
Distance


























Meets Regulations
Yes No




















































Comments:

-------
III. Operations




A. Meteorological Equipment
Parameter
1. Are all WS, WD sensors operational?
2. Is the temperature probe and aspirated shield operational?
3 . Is precipitation gauge operational and clean?
4. Is dew point sensor operational?
5. Is meteorological tower perpendicular to the ground?
6. Are all booms level?
7. Are all cables secure?
8. Are connections clean and rust-free?
9. Are vanes/cups/propellers intact?
10. Are serial numbers available?
1 1 . Are WS/WD bearings free?
12. Is the solar radiation sensor operation?
13. Are the chart traces clear and easily read?
14. Are all chart recorders times correct?
15. Is DAS operational?
16. Are the sigma values being collected?
17. Are vector values being collected?
18. Is the printer functional?
19. Is hard copy data printout available?
20. Are torque tests being performed on WS/WD sensors?
21. Overall, is the site well maintained?
Yes





















No





















Comments:

-------
B. Auxiliary Equipment
Parameter
1. Is the A/C unit operational?
2. Is the shelter temperature system operational?
3. Is the shelter temperature recorded?
4. Is the shelter temperature maintained at 20-3 0°C?
5 . Is the shelter and outside surroundings clean?
6. Does modem work?
7. Does telephone work?
8. Is the site secure?
9. Overall, is meteorological and auxiliary equipment well maintained?
Yes









No









Comments:
C. Procedures
Parameter
1. Are the station logs present?
2. Are the station logs up to date?
3 . Are station logs detailed?
4. Are routine checklists used?
5. Are the routine checklists detailed?
6. Are the calibration forms present?
7. Are the calibration forms detailed?
8. Are the EPA guidelines present?
9. Are QA/QC manuals present?
10. Is the monitoring plan present?
1 1. Is the site technician knowledgeable of meteorological equipment
operation?
12. Are the strip charts annotated each visit?
13. Are charts annotated with date and time?
14. Are sensor calibrations performed every six months?
15. Does the site technician have a working knowledge of EPA guidelines?
16. Does the site technician have working knowledge of QAPP?
Yes
















No

















-------
Parameter
17. Does the site technician have a working knowledge of Monitoring Plan?
18. Does the monitoring plan have site ID forms?
19. Does the monitoring plan have site map and photos?
Yes



No



Comments:
D. Preventive Maintenance
Parameter
1. Is preventive maintenance discussed in the QA/QC plan?
2. Are field operators given special training for preventive maintenance?
3. Is the training re-enforced?
4. Review preventive maintenance worksheets or logs acceptable?
5. Please provide the frequency of calibration for the following:
6. Are parts and tools available to site operators?
7. List any persistent problems with equipment.
8. Are preventive maintenance log books maintained?
9. Does senior staffer management review the logs?
10. Does data processing staff review the logs?
Yes










No










Comments:
E. Chain of Custody

 D. Chain of Custody
 1. Review paper work for chain of custody from field to data processing.
 2. How is data stored?
 3. How often is the data logger dumped?
By Whom?

-------
IV. Overall Comments
Parameter
1 . Overall, is the station well maintained?
2. Overall, is the data quality good?
3. Are QA/QC maintained?
4. Is site and equipment in good working order?
5. Overall, is the site technician well trained?
6. Overall, do meteorological patterns agree with topography?
7. Are the met readings reasonable for large scale meteorological patterns?
8. Is the meteorological sensors' exposure correct?
Yes








No








 (i.e., Micro, Middle, Neighborhood, Urban, or Regional?)
Comments:

-------
                         Appendix B






    Examples of Meteorological Sensor Calibration Forms






This appendix contains examples of meteorological calibration forms for field use.

-------
                  \
                   C3
                  <5?
                                  Environmental Protection Agency
                              Meteorological Calibration/Audit Form
                                Site Name/Location	
                                Station Operator	
                                Auditor	
                                Date:	
                                GPS Coordinates
       Instrument Make/Model_
  Audit Device Make/Model
                 Wind Speed Sensor

                	Serial Number_
                 Serial Number
   _Range_
Range_
   Torque Test
g-cm
Audit Point
(rpm)
0




Audit Value
(mis)





DAS
Response
(mis)





Difference
(mis)





Chart
Response
(mis)





Difference
(mis)





Pass/Fail?





                                     Wind Direction Sensor
       Instrument Make/Model_
  Audit Device Make/Model
                    Serial Number
  _Range_
                 Serial Number
Range_
Magnetic Dec!
Audit Point
(deg)





ination deg. Torque Test g-cm Cross Arm Reference

DAS Response
(deg)





Difference
(deg)





Chart Response
(deg)





Difference (deg)





deg.
Pass/Fail?





Comments

-------
       f;
      «?

                            Environmental Protection Agency
                          Meteorological Calibration/Audit Form
                         Site Name/Location	
                         Station Operator	
                          Auditor	
                          Date:	
                          GPS Coordinates
   Instrument Make/Model_
  Audit Device Make/Model
  Type of Aspirator	
                            Atmospheric Temperature Sensor
 Serial Number
Serial Number
 _Range
 Range_
Audit Point
(deg. C)





Audit Value
(deg. C)





DAS
Response
(deg. C)





Difference
(deg. C)





Chart
Response
(deg. C)





Difference
(deg. C)





Pass/Fail?





  Instrument Make/Model	
  Audit Device Make/Model
                                Relative Humidity Sensor
Serial Number
 Serial Number
_Range	
	Range_
                         Dry Audit Point
                              (deg)
                            Dry Temp
                          (Thermometer)
                            Dry Temp
                              (DAS)
                           Wet Temp.
                          (Thermometer)
                         RH (Calculated)
                          RH Reported
                             (DAS)
         Value
Comments
 Date

-------
United States                               Office of Air Quality Planning and Standards
Environmental Protection                    Air Quality Strategies and Standards Division
Agency                                    Research Triangle Park, NC

-------