United States
Environmental Protection
^1 Agency
EPA/600/R-14/182 | August 2014 | www.epa.gov/research
Measurements of Atmospheric
NH3, NOy/NOx, and N02 and
Deposition of Total Nitrogen at
the Beaufort, NC CASTNET Site
(BFT142)
RESEARCH AND DEVELOPMENT
-------
-------
Measurements of Atmospheric NH3,
NOy/NOx, and N02 and Deposition of
Total Nitrogen at the Beaufort, NC
CASTNET Site (BFT142)
Ralph E. Baumgardner, Jr.
Environmental Sciences Division
National Exposure Research Laboratory
Research Triangle Park, NC
Christopher M. Rogers
AMEC Environment & Infrastructure, Inc.
Jacksonville, FL
Melissa R. Puchalski
Clean Air Markets Division
Office of Air and Radiation
Washington, DC
Thomas F. Lavery
Cranston, Rl
Kevin P. Mishoe
AMEC Environment & Infrastructure, Inc.
Newberry, FL
Garry L. Price
AMEC Environment & Infrastructure, Inc.
Newberry, FL
Nathan S. Hall
University of North Carolina at Chapel Hil
Institute of Marine Sciences
Morehead City, NC
Although this work was reviewed by EPA and approved for publication, it may not necessarily reflect
official Agency policy. Mention of trade names and commercial products does not constitute
endorsement or recommendation for use.
U.S. Environmental Protection Agency
Office of Research and Development
Washington, DC 20460
-------
-------
Measurements of Atmospheric NH3,
NOy/NOx, and N02and Deposition of
Total Nitrogen at the Beaufort, NC
CASTNET Site (BFT142)
Ralph E. Baumgardner, Jr.
Environmental Sciences Division
National Exposure Research Laboratory
Research Triangle Park, NC
Christopher M. Rogers
AMEC Environments Infrastructure, Inc.
Jacksonville, FL
Melissa R. Puchalski
Clean Air Markets Division
Office of Air and Radiation
Washington, DC
Thomas F. Lavery
Cranston, Rl
Kevin P. Mishoe
AMEC Environments Infrastructure, Inc.
Newberry, FL
Garry L. Price
AMEC Environments Infrastructure, Inc.
Newberry, FL
Nathan S. Hall
University of North Carolina at Chapel Hil
Institute of Marine Sciences
Morehead City, NC
National Exposure Research Laboratory
Office of Research and Development
Research Triangle Park, NC 27711
-------
Table of Contents
1.0 Introduction 1
a. Background 1
b. Purpose of Study 1
2.0 Site Characteristics and Monitoring Equipment: Beaufort, NC (BFT142) 4
3.0 Results of the NH3 Measurement Study 10
a. Overview 10
b. Comparison of ADS and AMoN NH3 Concentrations 11
c. Comparison of 24-Hour and Weekly ADS NH3 Concentrations 11
d. Comparison of ADS and CASTNET Filter Pack Concentrations 13
4.0 Deposition of Atmospheric Nitrogen and Sulfur 24
5.0 Results of the NOy Measurement Program 26
a. Overview 26
b. Field Evaluation and Quality Assurance 26
c. Data Evaluation and Completeness 27
d. Comparison of Weekly Average NOy Concentrations with Weekly Filter Pack
Total Nitrate Concentration Data 28
e. Time Series of Weekly NOy, NOx, N02, NOz, NO, and 03 Concentrations 29
f. Time Series of Monthly NOy, NOx, N02, NOz, NO, and 03 Concentrations 29
g. Ratios of Weekly NOx and NOy Concentrations 31
h. Diurnal Plots of Aggregated Nitrogen Species and 03 Concentrations 32
6.0 Summary of Measurements 33
7.0 References 35
List of Figures
Figure 1. CASTNET Monitoring Sites
Figure 2. AMoN Monitoring Sites
Figure 3. CASTNET NOy Monitoring Sites
Figure 4. Small Scale Google Aerial Photograph of BFT142, NC
Figure 5. Large Scale Google Aerial Photograph of BFT142, NC and
Open Grounds Farm
Figure 6. BFT142 Site Including NOy Tower
Figure 7. NOy Converter at 10 meters
Figure 8. Analyzers inside the BFT142 Shelter
Figure 9. NO-NOx-NOy Sampling Configuration
Figure 10. Annular Denuder System (ADS) and Related Sampling Systems at BFT142
Figure 11. Sampling Schedule
-------
List of Figures (continued)
Figure 12. Radiello NH3 Sampler
Figure 13. Time Series of Measured NH3 Concentrations (|jg/m3)
Figure 14. AMoN Annual Mean NH3 Concentrations (|jg/m3) for 2011
Figure 15. AMoN Annual Mean NH3 Concentrations (|jg/m3) for 2012
Figure 16. NH3 Concentrations Collected by BFT142 ADS and NC06 AMoN Samplers
Figure 17. NH3 Concentrations Collected by BFT142 ADS Weekly and 24-Hour Samplers
Figure 18. Time Series of Measured S024 Concentrations (|jg/m3)
Figure 19. S024 Concentrations (|jg/m3) Measured by Weekly ADS/FP and CASTNET FP
Figure 20. S024 Concentrations (|jg/m3) Collected by ADS Weekly and 24-Hour Samplers
Figure 21. Time Series of Measured S02 Concentrations (|jg/m3)
Figure 22. S02 Concentrations (|jg/m3) Measured by Weekly ADS/FP and CASTNET FP
Figure 23. Time Series of Measured N03 Concentrations (|jg/m3)
Figure 24. N03 Concentrations (|jg/m3) Measured by Weekly ADS/FP and CASTNET FP
Figure 25. Time Series of Measured HN03 Concentrations (|jg/m3)
Figure 26. HN03 Concentrations (|jg/m3) Measured by Weekly ADS/FP and CASTNET FP
Figure 27. Time Series of Measured NH 4 Concentrations (|jg/m3)
Figure 28. NH 4 Concentrations (|jg/m3) Measured by Weekly ADS/FP and CASTNET FP
Figure 29. Time Series of Measured Total N03 Concentrations (|jg/m3)
Figure 30. Total N03 Concentrations (|jg/m3) Measured by ADS/FP and CASTNET FP
Figure 31. Components of Total Nitrogen Deposition (kg/ha/yr)
Figure 32. Components of Total Sulfur Deposition (kg/ha/yr)
Figure 33. Weekly Total N03 and Weekly Average NOy Concentrations (ppb)
Figure 34. Weekly NOy, NOx, N02, NOz, NO, and 03 Concentrations (ppb)
Figure 35. Time Series of Monthly Average Nitrogen Species Concentrations and 03
Concentrations (ppb)
Figure 36. Time Series of Ratios of Weekly NOx to NOy Concentrations
Figure 37. Composite Diurnal Distribution of Ratios of Weekly NOx to NOy Concentrations
Figure 38. Composite NOy, NOx, N02, NOz, NO, and 03 Concentrations (ppb)
List of Tables
Table 1. 1-Week and 2-Week NH3 Concentrations (|jg/m3)
Table 2. Weekly Versus Aggregated 24-Hour ADS Concentration Regression Statistics
Table 3. Contributions to Total Deposition
Table 4. Quality Control Checks for NCore Analyzers
-------
1.0 Introduction
a. Background
The Clean Air Status and Trends Network (CASTNET) is a long-term environmental monitoring
program that measures trends in ambient air quality and atmospheric dry pollutant deposition
across the United States. CASTNET has been operating since 1987 and currently consists of
89 monitoring stations nationwide (Figure 1). The Environmental Protection Agency (EPA) operates
a majority of the CASTNET sites. The National Park Service (NPS) currently operates 24 stations in
cooperation with EPA. Other federal and state agencies participate in network operation. For
example, the Bureau of Land Management (BLM) currently operates five CASTNET sites in
Wyoming. For more information on CASTNET, see EPA's CASTNET website:
http://iava.epa.gov/castnet/ (EPA, 2014).
Figure 1. CASTNET Monitoring Sites
YOS404
, PIN414
JOT403
YEL4W SHE604
• FOR605
BAS601 • „NEC6G2
BUF6Q3 ® -
PND1&5 WNC4Z9
GR9411
P
CAN407
CNT169
ROM406. ROM206
GTH161
. MEV405
BBE4-01
\|/
JJND002
%
STK138
SVL130
KNZ184 ALH157
• 9
MCK131. MCK231
HOX14B
# *
UVL124
ANA 115 * '
SAL133.
» OXF122
VIN140
MKG113
- «
QAK172 L^L117
WFM105 #
• w
61 •
WC001* HWF187
vCTHIIO
• •
KEF112
• HOW191
^^ACA416
WST1Q9
CAT175
PSU'06 WSP144
ARE128* 1
- Z-eeim
PAR 107
" BWR139
CDZ171
• #CDR118» #
3(1136 # VFI120 :
* # ppnios^
&
SPOH1
9 PNF126
ESP,!' GRS42? „ ' CND125''
COW13?
*SND152 \_ J
#GAS153 Y
„ SUM156
M
i
jT
\ \ IRL141
V \
EPA Sponsored 'v evmu
NPS Sponsored
BLM Sponsored
EPA Collocated Pair
NPS/EPA Collocated Pair
b. Purpose of Study
CASTNET has a 25-year record of weekly atmospheric nitrogen measurements collected at
rural/remote sites in the United States. CASTNET initially featured approximately 50 sites mostly in
the eastern United States but has grown to currently include over 90 sites across the contiguous
United States and Alaska. The ambient nitrogen measurements taken at CASTNET sites are nitric
acid (HNCX), particulate nitrate (NO:,), and ammonia (NH'j. CASTNET also produces estimated dry
1
-------
deposition fluxes of these compounds. However, some key contributors to the atmospheric nitrogen
budget have not been measured. For example, ammonia (NH3) and total reactive oxidized nitrogen
(NOy) had not been measured routinely. To remedy this, many CASTNET sites began participating
in the Ammonia Monitoring Network (AMoN) (Figure 2), which was initiated by the National
Atmospheric Deposition Program (NADP) in 2007 to establish a nationwide network of passive NH3
monitors. For more information on AMoN, view the website: http://isws.illinois.edu/amon (NADP,
2013).
A few commercial off-the-shelf instruments are available to measure NOy (NO + N02 + NOz) and
NOx (NO + N02). NOz includes the nitrogen species HN03, nitrous acid (HN02), peroxyl acetyl nitrate
(PAN), other organic nitrates, and particulate nitrates. In practice, NOz is calculated as NOy - NOx.
Since 2005, CASTNET has measured NOy at Beltsville, MD (BEL116). During 2012, NOy
measurements were added at Bondville, IL (BVL130) and Huntington Wildlife Forest, NY (HWF187).
In 2013, the CASTNET NOy network further expanded to include Pinedale, WY (PND165),
Cranberry, NC (PNF126), and Rocky Mountain National Park, CO (ROM206). Figure 3 displays the
CASTNET sites with NOy measurements.
CASTNET has also operated other nitrogen measurement studies in the past several years to
supplement and enhance the filter pack and AMoN measurements, including two projects at the
Beaufort, NC site (BFT142) located in an agricultural area of coastal North Carolina. The results
from these two studies are presented in this report.
Figure 2. AMoN Monitoring Sites
MN18 NS01
• WL h •
4-
« j *
WI35
NH02
W1D7 0N24
NVS7 N¥MCT,S
LTT01 »NE3S ^ $
UIUI in? a.' —
WY95 • • • ^
IL.. IH99 0*4 "99
NJ9£
.~V
0^63
^ VIP ¦W KYD3
Mi/193 MVI99 A
# *
KS31«S03 IL-46 ^
• # • KYD3 _W4
£298
KY98 ^ WC30
TN01 NCOS
NC25 A O
: * scos
^ ^ GM1>
FL19
# AMoN Sites
2
-------
Figure 3. CASTNET NOy Monitoring Sites
The first Beaufort study included NH3 measurements based on the deployment of a weekly, dual
annular denuder system (ADS) during designated measurement weeks. Additionally, the Beaufort
ammonia study operated daily ADS sampling for comparison with the weekly integrated ADS
sample. The daily ADS ran for a 24-hour period to collect standard CASTNET constituents and
ammonia. Measurements were conducted from May 2011 through November 2012, operating on a
two out of every six weeks schedule. The Beaufort studies also included combining model estimates
of dry deposition velocities with the measured concentrations to estimate dry deposition of the
nitrogen and sulfur species. The dry deposition fluxes were added to measured wet deposition
rates to obtain total deposition. The results of the NH3 measurement study are presented in
Section 3. Estimates of deposition of atmospheric nitrogen and sulfur and presented in Section 4.0.
The second study at Beaufort focused on measuring NOy continuously from February 2012 through
May 2013. Measurements of trace-level NOy were obtained using a commercially produced
instrument with an optional configuration using a second converter that allowed for the collection of
NOx in addition to NOy and NO. Using a difference method (NOx - NO), an N02 concentration was
obtained. Similarly, NOz was calculated as the difference between measured NOy and measured
NOx. These results are presented in Section 5.0. Section 6.0 presents a summary of the ADS and
continuous measurements and estimates of nitrogen and sulfur deposition.
3
-------
2.0 Site Characteristics and Monitoring Equipment: Beaufort, NC (BFT142)
The CASTNET monitoring site is located about 10 miles north of the Town of Beaufort, NC and is
situated on western edge of Open Grounds Farm (OGF), a massive 57,000-acre farm northeast of
Beaufort. OGF produces soybeans, corn, and some
wheat and cotton. The sidebar summarizes the
BFT142 site characteristics. Figures 4 and 5
illustrate the location of the monitoring site on
Google aerial photographs. The NOy inlet with a
NOy converter was located atop a 10-m tower,
which is shown in Figure 6 along with the ADS,
filter pack, and meteorological sampling towers.
Figure 7 shows the NOy converter at 10 meters. A
NOx converter was located in the shelter adjacent
to the NOy instrument. Figure 8 shows the BFT142
site instruments. Figure 9 illustrates the NO-NOx-
NOy sampling configuration. The BFT142
CASTNET site also hosts a NADP/NTN wet
deposition monitoring system (NC06) and an AMoN
(NC06) passive NH3 monitoring device. An OGF-
operated fertilizer and chemical plant is located
about three miles east-southeast of BFT142 on
Nelson Bay Road.
Figure 4. Small Scale Google Aerial Photograph of BFT142, NC
Beaufort, NC Site Characteristics
Site name and ID
Beaufort, NC, BFT142
County with state abbreviation
Carteret, NC
Latitude; decimal degrees
34.8847
Longitude; decimal degrees
-76.6207
Elevation
5 meters
Operating agency and start date
EPA, 12/28/1993
Primary Land Use
Agricultural (primarily soybeans and corn)
Terrain surrounding site
Flat
Nearest NADP site code
NC06
Distance to nearest NADP site
0.04 kilometers
Does site conform to deposition model
assumptions?
Yes
4
-------
Figure 5. Large Scale Google Aerial Photograph of BFT142, NC and Open Grounds Farm
Figure 6. BFT142 Site Including NOy Tower
Beaufort, NC (BFT142)
Note: 1 NADP/NTN precipitation sampling systems
2 AMoN sampler
3 Meteorological tower
4 NOy converter tower
5 Filter pack and ADS tower
5
-------
Figure 7. NOy Converter at 10 meters Figure 8. Analyzers inside the BFT142 Shelter
Figure 9. NO-NOx-NOy Configuration
External
Converter
Enclosure
Molybdenur
Convener
Copillcry
• P»M
Meter
Mefe? n I
—i—^vi']
By—Po"-':
Exhaust Through
A Suitoble Vert
no2-»no
Converter
(NO„Moae)
HZDi
?N.C
^N.O.
* n qi
W
Capillary
M
ria*
Sensor
Pressure
T»yn*;tJ«ji i-r
Or, Air A [] O
Ozonet
i
Display
~
Anoiog Outputs
D'gitoi Outputs
CdmmurtfcdtfOft JV&toWl9
^eoctior
ChomDC
Ozonator Capillary
I ' -fe;.
N.
Chomoer
The goals of the Beaufort studies were to continue and expand analyses that began during the
Ammonia CASTNET CSN Study (ACCS; EPA, in press), and to add NOy-related measurements in
an effort to estimate the components of total dry nitrogen deposition from nitrogen species not
collected by the typical CASTNET filter pack. The Beaufort location expanded ACCS to include a
coastal site located in an area of interest for ammonia monitoring because of its proximity to
agricultural/crop production "i.e.", OGF and animal production "e.g.", Prestage Farms ["pork" and
"turkey"] in eastern North Carolina.
6
-------
Sampling Equipment
• Dual Denuder ADS for Measuring NH3 and Other Species
The heart of the NH3 sampling system consisted of nine dual denuder ADS (Figure 10) with a
3-stage filter pack arranged to collect seven daily samples, one field blank, and one weekly sample.
ADS featured:
1. A sodium carbonate (Na2C03)-coated annular denuder for collecting gaseous HN03 and
sulfur dioxide (S02) analyzed by ion chromatography (IC) using deionized water (DIW) as the
extraction fluid.
A phosphorus acid (H3P03)-coated annular denuder for collecting gaseous NH3 analyzed by
automated colorimetry (AC) using DIW as the extraction fluid.
A 3-stage CASTNET-style filter pack consisting of:
a. A Teflon filter for collecting particulate sulfate (SO',), N03, and NHl analyzed by IC.
A nylon filter for collecting S02 (as SO;,) and NO, analyzed by IC. No measurable S02
was collected on the nylon filter.
A H3P03-impregnated filter for collecting any extra NHX either from denuder capture
inefficiency or loss of NH j through volatilization off the nylon filter (or both), which was
analyzed by AC using DIW as the extraction fluid.
2.
3.
b.
c.
The ADS were installed and operated at 10 meters. AMEC personnel created a solenoid system and
a data logger program that permitted unattended daily and weekly sampling. The BFT142 ADS was
the same as used in ACCS. The ADS used a cyclonic separator for removing particles with mean
aerodynamic diameter greater than 2.5 pm. The weekly ADS used a flow rate of 3.0 liters per minute
(Ipm) and the daily sampling was based on a flow of 16.7 Ipm.
Figure 10. Annular Denuder System (ADS) and Related Sampling Systems at BFT142
CKT136
GRS42^,
CDR119#
VPI120
• PNF126
BWR139
COW137
SND152
-------
The sampling schedule (Figure 11) was similar to ACCS:
• Daily ADS ran for one week, then again five weeks later.
• Weekly ADS ran for one week in conjunction with daily ADS then a second consecutive
week for comparison with AMoN.
• ADS sampling occurred from May 2011 through November 2012.
Figure 11. Sampling Schedule
~ CASTNET Filter Pack
Daily ADS
¦ Weekly ADS
| I AMoN
Figure 12. Radiello NH3 Sampler
8
-------
• AMoN Radiello Sampler for Measuring NH3
A Radiello NH3 passive sampler was deployed as part of the NC06 AMoN sampling to collect 2-week
NH3 concentrations (Figure 12). The Radiello sampler was chosen by NADP for use in the AMoN
based on comparisons with other samplers and annular denuder measurements (Puchalski et al.,
2011).
• Three-stage CASTNET filter pack
A 3-stage CASTNET filter pack was deployed for each sampling period as part of the base
operations of the core CASTNET program. The filter pack sampled gaseous S02 and HN03;
particulate S024, N03, and NH4; base cations; and chloride ion (CI"). The CASTNET filter pack
system did not utilize a cyclonic separator to remove large particles (mean aerodynamic diameter
greater than 2.5 |jm).
• NO/NOx/NOy/ Measurements
Trace-level NOy measurements were obtained using a Thermo Scientific (Thermo) Model 42i-Y NOy
analyzer, which is a chemiluminescent trace-level gas analyzer that measures all reactive oxides of
nitrogen in the ambient air. AMEC deployed an optional configuration at BFT142 using a second
converter that permitted the collection of NOx in addition to NOy and NO. Thus N02 could be
calculated using a difference method (NOx - NO). Continuous NOy measurements began in February
2012 and were completed in May 2013. The configuration produces hourly concentrations of NOy,
NOx, and NO and calculated N02, and calculated NOz. All measurements and model calculations
made by AMEC during these studies followed the CASTNET Quality Assurance Project Plan (QAPP;
AMEC, 2012).
• Supporting Measurement Systems
Hourly 03 concentrations were collected at BFT142 as part of routine CASTNET operations. The
continuous gaseous measurements were supported by a dynamic dilution calibration system and a
zero air system. The sampling by all instruments was controlled by a Campbell Scientific Instruments
(CSI) CR3000 data logger.
• Meteorological Measurements
Meteorological instruments were operated on a 10-m tower at the BFT142 site. The measurements
included wind speed and direction, sigma theta, solar radiation, relative humidity, temperature at 9 m
and 2 m, precipitation, and surface wetness. The meteorological measurements were archived as
1-hour averages. The data were used as input to the Multi-Layer Model (MLM) to estimate values of
deposition velocity (Vd). The Vd values were used to estimate dry deposition of HN03, particulate
N03, particulate NH+4, S02, and particulate S024. All meteorological measurements followed the
CASTNET QAPP.
9
-------
3.0 Results of the NH3 Measurement Study
a. Overview
Figure 13 presents time series of daily, weekly, and daily-averaged-to-weekly NH3 concentrations
from BFT142 measurements. The figure also presents 2-week average NH3 concentrations collected
using the AMoN Radiello passive sampler at the collocated NC06 site. The data in this figure provide
a good synopsis of the BFT142 19-month NH3 measurement program. ADS measurements of
weekly NH3 concentrations were higher in the spring and summer of 2011. Table 1 summarizes the
NH3 measurements collected during the 19-month study. The weekly ADS NH3 concentrations were
typically higher than the average of the seven daily samples. There was no consistent bias between
the 2-week average ADS concentrations and the AMoN samples. The average of weekly (twice per
five weeks) BFT142 NH3 values was 1.15 |jg/m3, and the average (2-week values) for NC06 was
1.00 |jg/m3. The weekly 2-week average BFT142 ADS values ranged from 0.43 to 2.89 |jg/m3 with a
standard deviation of 0.84.
Figure 13. Time Series of Measured NH3 Concentrations (|jg/m3)
5.0
4.5
4.0
3.5
E
§30
I2.5
C3
<1>
! 2.0
I
I
1.5
1.0
0.5
0.0
*
~
~
~
• AMoN
¦ Weekly ADS
~ Daily ADS
X Daily ADS Averaged to Weekly
i >
• • •• j
«(~
i *
(N(N(N(N(N(N(N(N(N(N(N
aicoaicocot^t^cDLOLO^
^(Nio^LOCDt^coaio-'-
Figures 14 and 15 provide maps of 2011 and 2012 annual mean NH3 concentrations measured by
AMoN. In 2011, nine sites measured annual mean concentrations greater than or equal to
2.0 |jg/m3. The highest concentration (3.2 |jg/m3) was measured at AMoN site CA83/CASTNET
SEK430, CA. The 2011 mean concentration for NC06 was 2.3 |jg/m3, which was higher than the
mean AMoN value collected only during the BFT142 sampling period. The annual mean value at
NC06 was affected by a concentration of 32.38 |jg/m3, which was measured during the two weeks
beginning 4/12/2011. Without the outlier, the annual mean concentration is 1.2 |jg/m3. Twelve sites
10
-------
measured concentrations greater than or equal to 2.0 |jg/m3 in 2012. The highest 2012
concentration (15.1 |jg/m3) was measured at Logan, UT (UT01), a site that was added to AMoN
during late 2011 and is not included on Figure 14. NC06 measured a mean concentration of 0.7
|jg/m3 during 2012.
b. Comparison of ADS and AMoN NH3 Concentrations
The 2-week average of the weekly BFT142 values and the 2-week NC06 values shown in Figure 13
and Table 1 compare reasonably well. A linear regression based on ADS values aggregated to two
weeks versus the 2-week AMoN concentrations is shown in Figure 16. The regression equation is
y = 0.7792x + 0.0963,
with an R2 value of 0.8178.
c. Comparison of 24-Hour and Weekly ADS NH3 Concentrations
Similarly, the averages ("x" in Figure 13) of the seven 24-hour NH3 concentrations compare well to
the sampled weekly values (Figure 13). A linear regression based on the ADS values and
corresponding 10 averages of the 24-hour concentrations is shown in Figure 17. The regression
equation is
y = 1.0527x - 0.1998,
with an R2 value of 0.9265.
Table 1.1-Week and 2-Week NH3 Concentrations (pg/m3)
Date On
Average of 7
Weekly ADS Values
AMoN
24-Hour ADS
Values
1st Week
Only
2nd Week
Only
2-Week
Average
2-Week
Sample
05/10/11
2.54
2.38
NA
2.38
2.06
06/21/11
1.99
2.18
3.61
2.89
2.01
08/02/11
0.77
0.82
1.64
1.23
1.31
11/09/11
0.42
0.51
0.36
0.43
0.11
12/06/11
0.35
0.60
0.37
0.49
0.10
03/27/12
0.93
0.89
0.61
0.75
0.90
05/08/12
0.61
0.70
0.56
0.63
0.74
07/31/12
0.41
0.93
0.51
0.72
0.95
09/11/12
1.04
1.42
1.39
1.41
1.56
10/23/12
0.60
0.64
0.78
0.71
0.29
Mean
0.97
1.11
1.09
1.15
1.00
Standard Deviation
0.73
0.67
1.04
0.84
0.73
Note: NA = not available
11
-------
Figure 14. AMoN Annual Mean NH3 Concentrations (pg/m3) for 2011
Figure 15. AMoN Annual Mean NH3 Concentrations (pg/m3) for 2012
Concentration
12
-------
Figure 16. NH3 Concentrations Collected by BFT142 ADS and NC06 AMoN Samplers
2.5
2.0
'1.5
O 1.0
0.5
0.0
y = 0.7792x +0.0963
R2 = 0.8178
0.0 0.5 1.0 1.5 2.0 2.5
Average Weekly ADS Concentration (|jg/m3)
3.0
3.5
Figure 17. NH3 Concentrations Collected by BFT142 ADS Weekly and 24-Hour Samplers
3.0
2.5
; 2.0
O 1-5
co
o
<
>.
'ro
O 1.0
y= 1.0527X- 0.1998
R2 = 0.9265
0.5
0.0
~
0.0
0.5 1.0 1.5
Weekly ADS Concentration (|jg/m3)
2.0
2.5
d. Comparison of ADS and CASTNET Filter Pack Concentrations
In addition to the NH3 concentrations discussed previously, the ADS, including its back-end filter
pack component, and the CASTNET filter pack (FP) sampling system measured concentrations of
13
-------
gaseous HN03 and S02 and particulate S024, N03, and NH 4. HN03 and N03 were summed to
produce total N03 concentrations.
• Particulate SO4
Because the measurement of atmospheric S024 is considered relatively straightforward, it represents
a good quality control check for the measurement system. Satisfactory results indicate the likely
success in measuring the other parameters. Figure 18 presents a time series of weekly ADS, daily
ADS, daily averaged to weekly ADS, and weekly FP S024 data. Qualitatively, the four data sets show
good agreement. A scatter plot of ADS versus FP weekly S024 concentration measurements is given
in Figure 19 and provides a quantitative estimate of the relationship between the ADS and FP S04
data. The regression equation fit through the data is
y = 0.9718x + 0.4025,
with an R2 value of 0.8190.
Other than a couple of minor outliers, the comparisons are good over an order of magnitude range of
S04 concentrations with a high correlation coefficient.
Figure 18. Time Series of Measured SO4 Concentrations (|jg/m3)
6
5
34
t5,
0 3
1
•CASTNETFP
¦ Weekly ADS
~ Daily ADS
X Da ily ADS Ave raged to Wee kly
~
•/ ^
%
*
•• v.
A
/•I
~
~
CNCNCNCNCNCNCNCNCN
a)COG)COCOh^h^CDLO
^cNco^fio
-------
Figure 19. SO4 Concentrations (|jg/m3) Measured by Weekly ADS/FP and CASTNET FP
Figure 20. SO4 Concentrations (|jg/m3) Collected by ADS Weekly and 24-Hour Samplers
The weekly ADS S024 measurements were also compared to the 24-hour concentrations aggregated
over a week. The values ("x" in Figure 18) of the 10 24-hour S024 concentration averages compare
15
-------
well to the sampled weekly values. A linear regression based on the ADS values and corresponding
10 averages of the 24-hour concentrations is shown in Figure 20. The regression equation is
y = 1.0339x - 0.0807,
with an R2 value of 0.9384.
Instead of presenting figures for all of the weekly ADS and 24-hour measurements aggregated over
a week, Table 2 summarizes the linear regression statistics forS024, NH3, S02, N03, HN03, NH4,
and total N03. The weekly and aggregated 24-hour concentrations compared reasonably well with
the exception of HN03 and total N03. In these cases, the weekly ADS values were higher than the
aggregated 24-hour measurements.
Table 2. Weekly Versus Aggregated 24-Hour ADS Concentration Regression Statistics
Pollutant
Slope
Y-lntercept
Correlation
SO4
1.0339
-0.0807
0.9384
nh3
0.7792
0.0963
0.8178
so2
0.8099
0.0139
0.8192
no3
1.0298
0.0002
0.9413
hno3
1.3034
-0.0438
0.7577
nh+4
1.0337
0.0207
0.8865
Total NO"3
1.4705
0.6181
0.6609
• S02
Time series of daily, weekly, and daily-averaged-to-weekly S02 concentrations from BFT142
measurements are presented in Figure 21. The figure also presents weekly average S02
concentrations collected using the CASTNET filter pack at BFT142.
A scatter plot of ADS versus CASTNET FP weekly S02 concentration measurements is given in
Figure 22. The regression equation fit through the data is
y = 0.7332x + 0.1706,
with an R2 value of 0.7862.
The scatter diagram of weekly ADS and CASTNET FP measurements shows reasonably good
comparisons over the range of S02 concentrations. However, the S02 data show more scatter and a
lower correlation coefficient than the S024 measurements.
16
-------
Figure 21. Time Series of Measured S02 Concentrations (|jg/m )
3.5
3.0
2.5
32.0
o1.5
1.0
0.5
0.0
•CASTNETFP
¦ Weekly ADS
~ Daily ADS
X Daily ADS Averaged to Weekly
" I,
t
~
—i 1 1 1—
co
CN CN CN
^ in 1.4 -
c
o
| 1-2
c
0
£ 1.0
O
O
co 0.6
<
O
0.4
0.2
0.0
y = 0.7332x + 0.1706
R2 = 0.7862
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
Weekly ADS Concentration (|jg/m3)
17
-------
• Particulate N03
Figure 23 presents time series of daily, weekly, and daily-averaged-to-weekly N03 concentrations
from BFT142 measurements collected over a 19-month period. The figure also presents weekly
average N03 concentrations collected using the CASTNET filter pack at BFT142.
Figure 24 presents a scatter plot of ADS versus CASTNET FP weekly N03 concentration
measurements. The regression equation fit through the data is
y = 1.0262X + 0.5914,
with an R2 value of 0.1750.
Figure 23. Time Series of Measured N03 Concentrations (|jg/m3)
2.5
2.0
E
"&1.5
3
c
o
TO
C
0)
£1.0
o
O
0.5
0.0
•CASTNETFP
¦ Weekly ADS
~ Daily ADS
X Daily ADS Averaged to Weekly
•
•:
•
•
•
•
•
•
•
• •
* :
• . • v.
/ # > •
• {
* * % ~ *
* • •
• * • • •
• •••
„ • .••••
i . ..V
~ • ~
» « \
CNCNCNCNCNCNCNCNCNCNCN
oraotocoi^i^coioio^
^CNCO^LOCDr^COOiO'^
18
-------
Figure 24. N03 Concentrations (|jg/m3) Measured by Weekly ADS/FP and CASTNET FP
1.6
1.4
£1.2
1
o 1.0
2
c
0
c 0.8
o
O
Q_
ll
i— 0.6
LU
z
I—
CO
5 0.4
0.2
0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Weekly ADS Concentration (|jg/m3)
The scatter diagram shows that, although the slope of the linear regression is near 1.0, the
correlation coefficient is low indicating that the distribution is random. In contrast, the results for
particulate S024 have a much higher correlation coefficient of 0.819 with a similar slope, primarily
because S024 particles exist in the atmosphere in the submicron range. The results for N03 were
likely caused by the lack of a cyclone separator on the CASTNET filter pack system and the
subsequent sampling of large N03 particles (Lavery et al., 2009) in the form of NH4N03 in the farm
environment.
• hno3
Figure 25 presents time series of daily, weekly, and daily-averaged-to-weekly HN03 concentrations
from BFT142 measurements collected over a 19-month period. The figure also presents weekly
average HN03 concentrations collected using the CASTNET filter pack at BFT142.
Figure 26 presents a scatter plot of ADS versus CASTNET FP weekly HN03 concentration
measurements. The regression equation fit through the data is
y = 1.3034x - 0.0438,
with an R2 value of 0.7577.
19
-------
Figure 25. Time Series of Measured HN03 Concentrations (|jg/m3)
•CASTNETFP
w
q ¦ Weekly ADS
#
~ Daily ADS
X Daily ADS Averaged to Weekly
•
¦
• •
•
~
*
*•
~
X
• • •
• • •
• fl ~
> - •••••-"• • / t'v
V*
~ * ~
~
•* ..
: •
'i.V*
•
^r
CN
R2 = 0.7577
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Weekly ADS Concentration (|xg/m3)
Figure 26. HNOs Concentrations (|jg/m3) Measured by Weekly ADS/FP and CASTNET FP
o
Q_
ll
h 0.6
20
-------
The scatter diagram shows good comparisons over the range of HN03 concentrations with a high
correlation coefficient, even though the regression line suggests HN03 levels measured by the filter
pack are typically higher than those measured by the ADS.
• Particulate NH 4
Time series of daily, weekly, and daily-averaged-to-weekly NH4 concentrations from BFT142 ADS
measurements are displayed in Figure 27. The figure also presents weekly average NH4
concentrations collected using the CASTNET filter pack at BFT142.
Figure 28 presents a scatter plot of ADS versus CASTNET FP weekly NH4 concentration
measurements. The regression equation fit through the data is
y = 0.3177x + 0.3153,
with an R2 value of 0.2687.
Figure 27. Time Series of Measured NH 4 Concentrations (|jg/m3)
2.0
1.8
1.6
1.4
E
^5)1.2
3
I1.0
TO
C
0
£0.8
o
O
0.6
0.4
0.2
0.0
•CASTNETFP
¦ Weekly ADS
~ Daily ADS
X Daily ADS Averaged to Weekly
~
x
• ~
_ • •
•••V. * •
*
• .. '*. *• »V. . " T
~ _
• • •
~ ~ ~
~
^J^r-ooa)coo)cocor^r^cDi?)Ln^
wa^o^cN^ojoo^LOCDr^coa^o^
21
-------
Figure 28. NH4 Concentrations (|jg/m3) Measured by Weekly ADS/FP and CASTNET FP
1.4
1.2
Weekly ADS Concentration (|jg/m3)
The scatter diagram shows poor comparisons over the range of NH4 concentrations with a weak
correlation coefficient. NH4 concentrations are typically collected as NH4N03 in the atmosphere.
Consequently, any sampling issues with N03 concentrations on the CASTNET filter pack will affect
NH4 measurements. In particular, the results for NH4 and N03 were likely caused by volatilization of
NH4NO3 particles following collection on the Teflon filter. Volatilization results in some loss of NH4
from the Teflon filter as NH3. The volatilized N03 is then collected as HN03 on the nylon filter
(Lavery et al., 2009).
• Total N03
Figure 29 shows time series of daily, weekly, and daily-averaged-to-weekly total N03 concentrations
from BFT142 ADS measurements. The figure also presents weekly average total N03
concentrations collected using the CASTNET FP at BFT142. The time series plots show the FP
estimates were consistently higher than the ADS/FP measurements over the lifetime of the
sampling program.
Figure 30 presents a scatter plot of ADS versus CASTNET FP weekly total N03 concentration
measurements. The regression equation fit through the data is
y = 1.4705x + 0.6181,
with an R2 value of 0.6609.
22
-------
Figure 29. Time Series of Measured Total N03 Concentrations (|jg/m3)
3.5
3.0
2.5
E
32.0
c
o
2
S 1.5
o
c
o
O
1.0
0.5
0.0
•CASTNETFP
¦ Weekly ADS
~ Daily ADS
X Daily ADS Averaged to Weekly
•
•
• •
•
• •
• •• • r. #
w ••
; \ •
• % ,
• * » . %* • *
~
~
•
V *•*%
V
~ ~ ~
CNCNCNCNCNCNCNCNCNCNCN
oraotocoi^i^coioio^
^cNco^:Lo
-------
4.0 Deposition of Atmospheric Nitrogen and Sulfur
Estimates of dry atmospheric deposition of nitrogen and sulfur species were obtained from
measured concentrations and Multi-Layer Model (MLM)-modeled hourly deposition velocities (Vd),
with the exception of Vd for NH3. The dry flux of NH3 was estimated from the concentrations
measured at BFT142, NC and a Vd value of 0.81 centimeters per second (cm/sec) selected from a
2002 Community Multiscale Air Quality Modeling System (CMAQ) model run (EPA, 2011). The MLM
calculations of Vd were based on onsite meteorological measurements and land use data
approximated for an area within a 1 km radius around the BFT142 site. Wet deposition was based
on NADP/NTN onsite measurements using the seven seasonal deposition values covering the study
period averaged and scaled to the year.
Figure 31 provides a pie chart that illustrates the estimates of individual component contributions to
total nitrogen (as N) deposition in kilograms per hectare per year (kg/ha/yr). Flux data used in the pie
chart are taken from the entire study period and scaled to one calendar year. The small component
labeled as being NHX is from the H3P03-impregnated backup filter and likely represents either NH3
gas not captured by the denuder or particulate NH+4 that volatilized from the Teflon filter and was
collected by the backup filter. Figure 32 shows the contributions to total sulfur (as S) deposition. The
species contributions to total deposition and their percentages are summarized in Table 3.
Figure 31. Components of Total Nitrogen Deposition (kg/ha/yr)
• ¦ HN03, Na2C03 DENUDER
¦ NH3, H3P03 DENUDER
\ BNH4, TEFLON FILTER
I ¦ NHX as NH4, H3P03 FILTER
W BN03,TEF+NYL FILTER
¦ NH4, wet
¦ N03, wet
Total N Deposition = 6.7 kg/ha/yr
24
-------
Figure 32. Components of Total Sulfur Deposition (kg/ha/yr)
Table 3. Contributions to Total Deposition
Nitrogen (kg/ha/yr)
Sulfur (kg/ha/yr)
as N
as S
Dry
Percent
Dry
Percent
HNO
0.48
7.1
so2
0.50
13.2
nh3
2.28
33.9
SO4
0.28
7.4
nh;
0.19
2.8
NHX
0.10
1.5
no3
0.03
Wet
0.5
Wet
nh;
1.92
28.5
SO4
3.01
79.4
no3
1.73
Total
25.7
Total
N
6.74
100.0
s
3.80
100.0
25
-------
5.0 Results of the NOy Measurement Program
a. Overview
The Beaufort Study collected hourly average NO, NOx, and NOy concentrations from February 2012
through May 2013. The measurements were collected using a Thermo 42i-Y analyzer with two
converters. One converter was located at 10 m for NOy and the second at the analyzer for NOx. For
every minute of data collection, 20 seconds were dedicated to the NO measurement, 20 seconds to
a pre-reactor measurement, and 10 seconds each to NOy and NOx measurements. Performance of
the data logger programs was verified at the AMEC field laboratory before deployment to BFT142.
N02 was calculated as the difference between NOx and NO. All concentration data were archived as
hourly averages.
b. Field Evaluation and Quality Assurance
Appendix 11 of the CASTNET QAPP describes the field evaluation methods used to control and
check the operation of the trace-level gas instruments (AMEC, 2012). This appendix covers site
operations, data collection, and QC requirements. All supplies used to conduct monitoring and
sampling meet the specifications of EPA and Teledyne Advanced Pollution Instrumentation (TAPI).
The calibration gases used for the gaseous criteria monitors are traceable to the National Institute of
Standards and Technology (NIST) and are protocol gases. The analyzers are mounted inside a
secure trailer, which is temperature controlled.
Zero, span, and precision (ZSP) checks of the NOy analyzers are performed automatically every
other day based on the concentration levels listed in Table 4. TAPI multigas calibration and zero air
systems, and protocol gas cylinders produce the calibration gases, and the CR3000 data logger
controls the process. CASTNET data analysts review data, including NOy data, for the previous day
for all EPA-sponsored CASTNET sites. A data analyst will note questionable values and enter
observations per site into the CASTNET Problem Tracking System. Precision and span checks are
judged successful if the results are within ± 10 percent of the test values. Zeroes must be within
3 percent of full scale values. If the ZSP results exceed the criteria, data are invalidated back to the
previous successful ZSP and forward to the next passing ZSP. Troubleshooting is performed to
determine root cause, and appropriate corrective action is implemented. Instrument calibration is
performed only if troubleshooting reveals that it is necessary.
Field calibrations are critical to achieving and maintaining data quality indicator (DQI) criteria. Every
six months, AMEC or subcontractor technicians visit each site to perform routine calibration and
maintenance of all sensors and instruments. AMEC personnel may calibrate the NOy sensors
independent of the routine calibration visit. The concentration levels listed in Table 4 are used for the
multipoint calibrations, again using the TAPI multigas calibration and zero air systems and the
CR3000 data logger.
26
-------
Table 4. Quality Control Checks for NCore Analyzers
Zero, Span, and Precision Checks Every Other Day
Parameter
Zero (ppb)
Span (ppb)
Precision (ppb)
S02
±3.0
90 ± 10%
25 ± 10%
NO/NOv/NPN
±3.0
90 ± 10%
15 ± 10%
CO
±40.0
1800 ±10%
250 ± 10%
Notes: The n-propyl nitrate (NPN) QC checks are performed exclusively at span level at a frequency of every fourth day.
Span concentrations are 90 percent of full scale.
If the ZSP results exceed criteria, data are invalidated back to the previous successful ZSP and forward to the next
passing ZSP.
Multipoint Calibrations Every Six Months
Parameter
Calibration Concentration Levels (ppb)
S02
90
40
15
7
4
0.0
NO/NOv/NPN
90
40
15
7
4
0.0
CO
1800
800
300
150
80
0.0
c. Data Evaluation and Completeness
Section 4.0 and Appendix 11 (NCore Air Monitoring Equipment) of the CASTNET QAPP describe
the data validation procedures implemented for all CASTNET measurements. These apply to
NO/NOy data as well. In general, continuous measurements undergo three levels of validation.
After daily polling of all CASTNET stations, Level 1 validation procedures are initiated. Level 1
validation consists of a set of automated screening protocols that initiate the transfer of data
between tables, the translation of data status flags and data screening, and they create the data
template, generate reports on the completeness of the data and the results of data screening and
archive the data. Data status flags are listed in Table 4-8 in Section 4.0 of the CASTNET QAPP.
The purpose of Level 2 validation is to develop a complete database. The process involves a data
analyst reviewing data at the end of a month and retrieving missing data using LoggerNet data
collection software. Essentially, this step represents a double check of the daily review process.
Level 2 validation is complete when the data for all time periods for all of the sampling sites have
been accounted for data have been recovered from the on-site data loggers and entered into the
database and sources of missing data are documented.
Level 3 validation involves a more detailed evaluation of the data. The site status report form
(SSRF), site narrative log sheets, ZSP data, calibration data, and audit results are reviewed for each
site. In addition, data are screened using tools that identify potential problems such as values
greater than the expected range and invalid combinations of status flags, values, and spikes. All
review and editing activities are documented both electronically and on hard copy forms.
27
-------
Perhaps the most important assessment of NO/ NOx/ NOy measurements is the review of ZSP
checks. If the ZSP results exceed criteria, troubleshooting is performed to determine root cause and
appropriate corrective action is implemented. Unscheduled instrument calibration is performed as a
corrective action only if the review reveals it is necessary. Routine 6-month field calibrations are also
critical to achieving and maintaining DQI criteria. Results from the ZSP checks and the 6-month and
unscheduled calibrations are used to assess NO/ NOx/ NOy data validity.
All BFT NOy and related data underwent the three levels of validation discussed above. These
validation steps resulted in 87 percent valid data. Another criterion used specifically for the BFT
database was to require both valid (called Category A) and suspect (flagged S) NOy data (called
Category B) to be greater than or equal to NOx concentrations. These requirements resulted in a
NOy data percent completeness of 54 percent. The low data completeness was likely caused by
incomplete conversion of NOy at the 10-m converter or collection of NOz species in the NOx sample
line and eventual release and sampling of these species as NOx. The valid NOy and related data
were used to complete the analyses summarized in this section.
d. Comparison of Weekly Average NOy Concentrations with Weekly Filter Pack Total Nitrate
Concentration Data
HN03 and particulate N03 are measured on CASTNET filter packs and the sum is defined as total
N03. To evaluate the components of NOy measurements, NOy concentrations were compared with
CASTNET filter pack total N03 concentrations. Since NOy species include HN03 and particulate N03
the ratio of total N03 to NOy should always be less than or equal to 1.0. Figure 33 provides a time
series of weekly total N03 and weekly average NOy concentrations collected at BFT142. Other than
for one week, the NOy concentrations were consistently higher than the total N03 levels, as
expected. Qualitatively, the two measurements follow similar temporal patterns throughout the study
period. For this analysis, all valid NOy measurements were used regardless of their relationship to
NOx measurements.
28
-------
Figure 33. Weekly Total N03 and Weekly Average N0y Concentrations (ppb)
CMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMOOOOOOOOOOOOOOOOOO
N^fflO(ON^LOOjOJWO^N^^COOJ(DOCON^fflO^LnO)inO)fM(DO
CM CO in in © 0 N N 00 0)^-OOt-t-CNCNt-t- cm CO
e. Time Series of Weekly NOy, NOx, N02, NOz, NO, and 03 Concentrations
Figure 34 provides graphs of valid weekly NOy, NOx, N02, NOz, NO, and 03 concentrations collected
over the entire measurement period. The N02 and NOz concentrations are calculated values. The 03
measurements were plotted only for those weeks that met a 60 percent completeness requirement.
The seasonal pattern for 03 and the nitrogen species is evident with ozone peaking in the warm
summer months while the highest concentrations for the nitrogen species are evident during the
winter months.
f. Time Series of Monthly NOy, NOx, N02, NOz, NO, and 03 Concentrations
Time series of monthly average nitrogen species concentrations and 03 concentrations are given in
Figure 35. Similar to Figure 34, the highest 03 concentration occur during the summer season and
the highest concentrations for the nitrogen species are seen durin the winter months..
29
-------
Figure 34. Weekly NOy, N0X, N02, N0Z, NO, and 03 Concentrations (ppb)
3.5
3.0
2.5
2.0
1.5
2 1.0
0.5
I.
ft
ft
Wt A
• •
VfW|
a
4
50
45
40
35
30
25
20
15
-NOY
-NOX
N02_calc
NOZ_calc
-NO
Ozone
10
(N(N(N(N(N(N(N(N(N(N(NCOCOCOCOCO
OOOOOOOOOOOOOOOO
(NOO^LOCDr^COaiO^tN^tNOO^LO
Figure 35. Time Series of Monthly Average Nitrogen Species Concentrations and 03
Concentrations (ppb)
50
-*-NOY
-¦-NOX
45
-*¦ N02_calc
-SK-NOZ_calc
40
-~-NO
Ozone
35
30
25
20
15
10
5
0
9 10 11 12 1 2 3 4
| 2013
Month
30
-------
g. Ratios of Weekly NOx and NOy Concentrations
Figure 36 provides time series of ratios of weekly NOx to NOy concentrations. The ratios range from
about 0.58 to 1.0. For most of the study period, the ratio is constant in the middle of the range.
Figure 37 presents a composite diurnal distribution of the ratios. The ratios peaked during the
morning and evening rush hours when NOx was produced from fresh vehicular emissions. The
NOx/NOy ratio decreased during the day as photochemistry depleted some of the NOx. However, the
measurements indicate the NOx - 03 photochemistry was not depleted entirely, with sufficient NOx
available to produce 03 farther downwind.
Figure 36. Time Series of Ratios of Weekly NOx to NOy Concentrations
1.2
1.0
0
1 0.8
O
Z
o
ra
0.6
0.4
31
-------
Figure 37. Composite Diurnal Distribution of Ratios of Weekly NOx to NOy Concentrations
1.0
0.6
0 4 8 12 16 20 24
Hour
h. Diurnal Plots of Aggregated Nitrogen Species and 03 Concentrations
Figure 38 shows six plots of composite NOy, NOx, N02, NOz, NO, and 03 concentrations. The curves
were constructed by averaging all hours for each hour of the day for each parameter for the entire
measurement period. The results show typical temporal distributions with NO values at 0.0 for much
of the night due to oxidation by 03 and conversion to N02. NOz levels showed little variation during
the day. NOy, NOx, and N02 levels increased after sunrise with fresh NO emissions produced by
vehicular traffic and other sources. 03 concentrations peaked around 1500 local time. The growth of
the convective boundary layer, photochemical transformations, and dry deposition decreased
nitrogen concentrations during the day as 03 levels increased. A secondary smaller peak in NOy,
NOx, and N02 concentrations was observed in the evening. Estimated daytime N02 concentrations
ranged from about 1.0 to 1.5 ppb, typical of rural settings. Typical daytime 03 production was about
15 ppb for the Beaufort site.
32
-------
Figure 38. Composite NOy, NOx, N02, NOz, NO, and 03 Concentrations (ppb)
Hour
6.0 Summary of Measurements
Atmospheric pollutant concentrations were collected using an ADS/FP system at the CASTNET
BFT142 site near Beaufort, NC. ADS measurements generally compared reasonably well with
AMoN and CASTNET filter pack data. The ADS data corroborate the use of 2-week passive Radiello
samples to measure NH3. The ADS/FP data also compared well with standard CASTNET FP
measurements with the exception of particulate N03 and total N03. Because the CASTNET filter
pack system does not use a cyclone separator to collect large particles, CASTNET measurements of
N03 and total N03 were higher than the corresponding ADS/FP measurements. The scatter diagram
for NH 4 shows fair comparisons over the range of NH +A concentrations with a weak correlation
coefficient. These results were likely affected by the collection of larger NH4N03 particles on the
CASTNET filter pack.
Total nitrogen and sulfur deposition were estimated for the Beaufort, NC site. Dry deposition was
estimated from measured concentrations of nitrogen and sulfur species and modeled dry deposition
velocities. Wet deposition rates were measured onsite and obtained from NADP/NTN. Dry and wet
deposition fluxes were summed to obtain total deposition (Table 3). Dry NH3, wet NH4, and wet N03
fluxes were the main contributors to nitrogen deposition. Wet S024 contributed more than 75 percent
of total sulfur deposition.
Direct measurements of NO, NOx, NOy, and 03 concentrations were made at the Beaufort, NC site
during the 19-month sampling program. Concentrations of N02 and NOz were estimated as
differences between the other nitrogen measurements. Apparent problems with the NOy converter
33
-------
atop the 10-m tower and other sampling issues resulted in 87 percent data completeness for the
NOy measurements. Screening the data with an additional criterion requiring NOy concentrations to
be greater than or equal to NOx values resulted in a data set with a completeness of 54 percent for
use in this report. Subsequent analyses, e.g., those shown in Figure 33, suggest most of the
problems were with the NOx measurements and that the NOy data were largely valid. NOx
measurements may have been affected by the difference in inlet height from those used in typical
NOx monitoring configurations. For this study, ambient air was sampled at the 10-m height of the
NOy converter even though NOx monitoring is usually performed at inlet heights of less than 4 m. A
vertical gradient in HN03 concentrations with lower concentrations near the ground may have led to
higher HN03 being sampled by the NOx channel at 10 m. Also, sampling at 10 m required longer
sampling tubing than a typical NOx installation. Adsorption and desorption of nitrogen species, such
as HN03, might have led to nitrogen components of NOz being sampled as NOx. The various
analyses presented in Figures 33 through 38 show that, following screening, the behavior of the
nitrogen species and 03 concentrations were consistent with expected atmospheric
behavior/models.
This study identifies a portion of the nitrogen deposition budget not typically characterized by routine
CASTNET measurements. Specifically, as shown by Figure 34, information on the components of
NOy, other than HN03 and N03, is gained by the addition of these nitrogen species measurements.
These data provide a better estimate of the nitrogen burden in the atmosphere. The development of
a NOy network within CASTNET will provide more data and increase understanding of the
importance of these nitrogen compounds. The NOy study at BFT142 demonstrates that value can be
added to existing NOy sites by the inclusion of converters specific to the measurement of other
components of total nitrogen, such as NOx and, by difference, N02. Future research should examine
the artifacts produced by converters, such as the molybdenum converter used in this study, and
investigate other options that may produce improved results, such as an LED-based photolytic
converter for NOx.
The information in this document has been funded wholly under EPA contract
EP-W-09-028, managed and collaborated by the United States Environmental
Protection Agency's Office of Research and Development and Office of Air
Programs. It has been subjected to the Agency's peer and administrative review
and has been approved for publication as an EPA document.
34
-------
7.0 References
AMEC Environment & Infrastructure, Inc. (AMEC). 2012. Clean Air Status and Trends Network
(CASTNET) Quality Assurance Project Plan Revision 8.0. Prepared for the U.S.
Environmental Protection Agency (EPA), Office of Air and Radiation, Clean Air Markets
Division, Washington, D.C. Contract No. EP-W-09-028. Gainesville, FL.
Environmental Protection Agency Clean Air Status and Trends Network Home Page. See
http://java.epa.gov/castnet/ (accessed October 2013).
EPA/ ORD / NERL. Summary Report of the Special Reactive Nitrogen (Nr) Inter-Comparison Study:
Ammonia CASTNET CSN Study (ACCS). Research Triangle Park, NC: GPO, In Press.
Lavery, T. F., C. M. Rogers, R. E. Baumgardner Jr., K. P. Mishoe. (2009). Intercomparison of
CASTNET N03 and HN03 measurements with data from other monitoring programs. J. Air&
Waste Manage. Assoc., 59, 214-226.
National Atmospheric Deposition Program / Ammonia Monitoring Network Home Page. See
http://nadp.sws.uiuc.edu/nh3net/ (accessed October 2013).
Puchalski, M. A., M. E. Sather, J. T. Walker, C. M. B. Lehmann, D. A. Gay, J. Mathewe, W. P.
Robarge. (2011). Passive ammonia monitoring in the United States: Comparing three
different sampling devices. J. Environ. Monit., 13, 3156
35
-------
-------
SEPA
United States
Environmental Protection
Agency
Please make all necessary changes on the below label, detach
or copy and return to the address in the upper left hand corner.
If you do not wish to receive these reports CHECK HERE
detach, or copy this cover, and return to the address in the
upper left hand corner.
PRESORTED STANDARD
POSTAGE & FEES PAID
EPA PERMIT No. G-35
Office of Research
and Development (8101R)
Washington, DC 20460
Official Business
Penalty for Private Use
$300
EPA/600/R-14/182
August 2014
www.epa.gov
Recycled/Recyclable
Printed with vegetable-based ink on
paper that contains a minimum of
50% post-consumer fiber content
processed chlorine free
------- |