Galveston Bay
Atmospheric Deposition
Evaluation
Final Report
3D September 1998
Submitted to: Dale Evarts
Director, Great Waters Program
Visibility and Ecosystem Protection Group
Air Quality Strategies and Standards Division
Office of Air Quality Planing and Standards
United States Environmental Protection Agency
MD-15
Research Triangle Park, NC 27711
(919)541-5687
Submitted by: Dr. Joel E. Baker
23150 Town Creek Drive
Lexington Park, Maryland 20653
(301) 737-0991
-------
Galveston Bay
Atmospheric Deposition*
Evaluation
Final Report
30 September 1998
Submitted to: Dale Evarts
Director, Great Waters Program
Visibility and Ecosystem Protection Group
Air Quality Strategies and Standards Division
Office of Air Quality Planing and Standards
United States Environmental Protection Agency
MD-15
Research Triangle Park, NC 27711
(919) 541-5687
Submitted by: Dr. Joel E. Baker
23150 Town Creek Drive
Lexington Park, Maryland 20653
(301) 737-0991
-------
Table of Contents
Section I. Technical Review of Air Toxics Deposition Monitoring in Galveston Bay, Texas:
Texas Regional Integrated Atmospheric Deposition Study (TRIADS) 3
Information Sources 3
Objective and Scope of this Review 6
Technical Review of TRIADS Draft Report 7
Study Design and Execution Issues 8
Summary of Study Design and Execution Issues 11
Section II. Comparison of TRIADS Results to those from Other Coastal Areas 15
Strategy 15
Comparison of Atmospheric Concentrations of Organic Contaminants
Among Programs 16
Comparison of Concentration of Metals, Organic Contaminants, and Nutrients in
Precipitation Among Programs 18
Conclusions - Comparison of Galveston Bay Data to Other Studies 27
Section III. Galveston Bay Atmospheric Deposition Workshop Summary 28
Agenda - Galveston Bay Atmospheric Deposition Workshop 28
Participant's List 29
Notes - August 11,1998 31
Section IV. Overall Summary and Conclusions 39
APPENDIX A 44
-------
Section I. Technical Review of Air Toxics Deposition Monitoring in Galveston Bay, Texas:
Texas Regional Integrated Atmospheric Deposition Study (TRIADS)
Information Sources
This report reviews the Draft Report entitled Air Toxics Deposition Monitoring in
Galveston Bay, Texas: Texas Regional Integrated Atmospheric Deposition Study (TRIADS)
prepared for the U.S. Environmental Protection Agency Office of Air Quality Planning and
Standards (EPA OAQPS) by the Geochemical and Environmental Research Group, Texas A&M
University (College Station, TX) under contract to Battelle Ocean Sciences (Duxbury, MA) under
EPA Contract No. 68-C2-0134. The draft report reviewed here is dated February 20, 1998, as
was provided to the reviewer by EPA OAQPS. Processed data of analytical results, in the form of
spreadsheet files, were also provided for this review. During the initial phases of this review,
many technical questions arose which were addressed to Dr. Terry Wade at Texas A&M, the
Principle Investigator of this study (Table 1). Dr. Wade and his staff provided considerable
additional information, consisting of clarifications and updated, more complete analytical
information, during the course of this review. These updated data are summarized in Section Il'of
this report.
Table 1. Data Requests and Questions Sent to T. Wade via E-mail at 17:10 on 7/7/98 for
TRIADS Data Review
Organics:
Missing Data (air):
We received PCB and Organochlorine Pesticide concentrations for PUF (vapor)
samples, but not for air filters (particle bound).
Missing Data (rain):
We received PCB and Organochlorine Pesticide concentrations for XAD
(dissolved) samples, but not for rain filters (particle bound).
Organics Questions:
3
-------
The small and large PUF were described as, "The organic air sampler employed
two filters followed by a large PUF and small PUF." in Sampling (p. 11). Is it
correct to interpret these as follows: the small PUF was placed downstream of the
Large PUF in order to assess breakthrough?
The method of reporting the organic's data provided on the CD-ROM is not clearly
described in the text. Some concentrations are reported as "0" (zero)with a "ND"
in the adjacent column. This notation method includes the laboratory and field
blanks. Do these "ND's" indicate the instances when the measured concentration
was below the instrumental detection limit (or when no peak was observed)?
How were the measured air and rain concentrations filtered or corrected for the
measured analyte mass in the matrix blanks? For example, was the mass of analyte
X in the average of the field blanks subtracted from the value found in each sample?
Metals:
Missing Data:
There is no metals data for either air or rain is included on the CD-ROM (I.D.
#980220-0129, with 8 files in the "X:\Triads\" directory as follows: Bk_ar.xls;
Cpufoc.xls; Gmw_ar.xls; Key.dat; Pufrepar.xls; Seabrook.dat; Xad_ar.xls; and
Xadoc.xls) though we did receive hard copies of what appear to be final
concentrations in Tables 2 & 3 of the Draft Report (dated February 20, '98). Could
we receive a copy of these air and rain data electronically?
Missing Data (rain):
Additionally, there appear to be no blank concentrations or instrumental limits of
detection for the rain concentrations reported in Table 2.
4
-------
Nutrients Questions:
What blank filtering or blank correction technique was employed to adjust the
measured rain and air concentrations for the levels of each nutrient found in the
blanks?
Objective and Scope of this Review
The objective of this review is to examine the technical quality of the work done by the
Investigator, as described in the Draft Report and the subsequent communications. The criteria
used include: (1) the stated objectives of the study; (2) the 'management questions' asked by
USEPA OAQPS (see Section II of this report); and (3) the established standards for sampling and
analytical quality established in the peer-reviewed literature. The study under review represents
the initial investigation of atmospheric deposition of metals and organic chemicals to the
Galveston Bay region, and this review reflects that this study had neither the scope or the support
to answer all relevant questions. The review, therefore, considers the results of the study against
the limited resources provided for the work and the dearth of existing information on which to base
the study design.
The stated objective of this study was (pg 2):
1. "To produce a data base that could be used to provide estimates of atmospheric deposition
to Galveston Bay and compare the importance of these inputs to other sources of
contaminant input."
After this study was underway, and in preparation for the Galveston Bay Atmospheric
Deposition Workshop, USEPA OAQPS developed a set of management questions that might be
addressed by the TRIADS data (Table 2). It is important to note that the TRIADS program was
not explicitly designed to answer these questions.
6
-------
Table 2. Galveston Bay Atmospheric Deposition Workshop Eleven Questions
1. What Hazardous Air Pollutants (HAPs) are found in the deposition samples and in what
amounts?
2. WTiich HAPs are also found, in significant amounts, in the water, sediments and/or biota?
3. Does deposition of HAPs exhibit temporal, seasonal and/or directional variations?
4. How significant are these variations?
5. Was the period for which data was collected meteorologically typical?
6. From what direction(s) does the most significant deposition contributions come?
7. For what times of day or year does the most significant contributions come?
8. What does upwind air quality monitoring data indicate regarding trajectory and areas of
contribution?
9. What does HAP emissions data indicate about the sources and the quantities of emissions
from these upwind areas?
10. What gaps, limitations, and uncertainties about data quality and completeness can be
identified and characterized for each step of this analysis?
11. What practical recommendations can be made for addressing them?
Technical Review of TRIADS Draft Report
As a general overview, the TRIADS program may be characterized by generally excellent
analytical chemistry in the laboratory often compromised by substandard sampling design and
execution. Those samples that entered the laboratory were analyzed with very high quality
assurance standards, and the Investigator is to be commended for conducting such difficult work.
However, as will be detailed below, a successful atmospheric deposition program depends as
much, if not more so, on careful execution of well-designed sampling strategies. This is the
7
-------
weakest part of TRIADS, and ultimately limits the applicability of the data to addressing larger
questions.
Study Design and Execution Issues. This reviewer recognizes that the TRIADS program was the
initial effort to quantify atmospheric deposition to coastal Texas, and that it was limited by the
resources available. However, several problems in the design and execution of the study are
evident.
1. Excessively long and variable sample integration limes. The program used automated
precipitation samplers and timer-activated air samplers at a single site in Seabrook, Texas.
Precipitation and air samples should be removed from the site and either analyzed
immediately or stored under appropriate conditions as soon as possible to minimize
artifacts. For example, labile species dissolved in precipitation (ammonium, nitrate) or
volatile species collected on air filters (PAHs, pesticides) may either degrade or volatilize
after collection, if left in the field. In addition, PUF/filter high volume air samplers for
organic contaminants may passively sample compounds from the ambient air while air is
not actively drawn through the sampling train. Standard protocols for precipitation
sampling are for samples to be left in the field for no longer than seven days for nutrients
and metals (NADP; CBADS) and two weeks for organics (if isolated on XAJD-2 resin,
CBADS) in order to minimize sampling artifacts. In this study, both precipitation and air
samples remained in the field for variable periods of time. During the earlier phases of the
study, the site was serviced reasonably frequently. However, as the study progressed the
sampling was less frequent, and the samples were commonly left in the field for two to
three weeks. While the laboratory analysis is well done, it is unlikely that these samples
were not compromised by the long sampling times.
Even if the long and variable sampling times did not result in artifacts, the long integration
times results in temporally-averaged concentrations and limits the ability to link these
measurements to meteorological conditions or emission patterns. Precipitation samples
integrated over one or more weeks are likely to contain several individual precipitation
8
-------
events, each of which may have resulted from different meteorological conditions
(convective storms versus frontal systems) originating from different directions.
Therefore, while the temporally-integrated precipitation samples may provide accurate wet
deposition flux estimates (providing that the long integration times did not result in
alteration of the samples, as discussed above), it is not possible to relate the measured
concentrations to precipitation intensity, wind patterns, etc. The same problem exists with
the measurement of ambient levels and speciation of organic contaminants in air (element
analysis is discussed below). The modified high volume air sampler was operated with a
weekly timer, which was set to collect air for several hours every seventh day. If the
sampler was serviced weekly, then a discrete air mass was sampled. However, the site
was commonly sampled at longer intervals, resulting in air samples that are composites of
two or more discrete sampling periods. Given changes in meteorological patterns among
the air sampling periods, it is not possible to relate the measured concentrations of organic
contaminants in the ambient air to wind directions, antecedent precipitation, etc. In
addition, the distribution of organic contaminants between the gaseous and aerosol
particulate phases varies with the ambient temperature and the total suspended matter
concentration. Therefore, it is very difficult to interpret the measured aerosol/gaseous
distributions as they integrate different meteorological conditions.
In principle, deposition of aerosol-bound contaminants can be estimated from measured
ambient concentrations by using appropriate deposition models. In the best of
circumstances, these calculations are uncertain due our relatively poor understanding and
parameterization of dry aerosol deposition processes. However, using the temporally-
averaged aerosol concentrations measured in this study, which cannot be linked to specific
meteorological conditions, would provide extremely crude estimates of dry aerosol
deposition.
In summary, the long and variable sample integration periods, resulting from the rather
infrequent servicing of the sampling site in Seabrook, resulted in temporally-averaged data
that may be compromised by sampling artifacts. Interpretation of temporally-averaged
9
-------
concentration data in light of the rapidly changing meteorological and source emission
profiles is limited.
2. Insufficient Analytical Sensitivity for Elemental Analysis ofAerosol Particles. The
sampling technique used to collect aerosol particles for elemental analysis resulted in
excessively high blank values for almost all target elements. Sampling and analysis of
aerosols for elements using suitably clean and sensitive methods are widely-reported in the
literature and it is unclear why such an inadequate method was chosen for this program.
As a consequence, the Draft Report contains no usable trace element aerosol data (though it
indicates that copper and perhaps iron and manganese may be salvaged). This is very
unfortunate, as the elemental composition of aerosol particles can be used in source-
receptor models to estimate the relative contributions of various source types to the
atmospheric deposition fluxes to Galveston Bay. Without this elemental data, and given the
temporally-averaged organic contaminant concentration data available, it is not possible to
quantitatively model the sources of contaminants to the Galveston Bay air masses.
Concentrations of total suspended matter in the atmosphere were also not measured in this study.
3. Lack of Diffusive Gas Exchange Measurements. Many of the target analytes in this study
are 'semi-volatile', meaning that they may actively exchange between the dissolved
reservoir in surface waters and the gas phase in the atmosphere. Several previous studies
in The Great Lakes and the Chesapeake Bay indicated that this gas exchange is the
dominate atmospheric deposition process for many PAH and PCB congeners. Despite this,
the TRIADS program made no attempt to quantify gas exchange of semi-volatile chemicals,
resulting in a significant gap in the atmospheric deposition budget. It should be noted,
however, that both the Great Lakes Integrated Atmospheric Deposition Network (IADN)
and the original Chesapeake Bay Atmospheric Deposition Study (CBADS) also ignored the
gas exchange process. However, at the time TRIADS began, the importance of gas
exchange in these other coastal systems was known, and TRIADS would have benefitted
tremendously by including estimates of gas exchange in their program. Now that we know
10
-------
that gas exchange is a very important atmospheric deposition process, all future coastal
atmospheric deposition programs must explicitly estimate this flux for semi-volatile
compounds.
4. Lack of Spatial Resolution. The TRIADS program chose to measure a wide variety of
atmospheric deposition parameters at one location rather than measuring fewer species at
several sites. While this is a reasonable and defendable choice of quality over quantity, it
leaves open the question of spatial variability. In the Galveston Bay area, which includes
many active sources of the target analytes located adjacent to a large body of water, spatial
variability in ambient atmospheric concentrations and depositional fluxes of target analytes
might be quite large. By comparison, the IADN master sites, located in rural areas on each
of the five Great Lakes, show little spatial variability (when integrated to annual time
scales), but significant spatial gradients in concentrations and depositional fluxes have
been documented downwind of Chicago, IL and Baltimore, MD. It is likely that spatial
trends in Galveston are similar to or greater than those in these two other coastal cities.
Any further work in the Galveston Bay area should quantify spatial, as well as temporal
variability.
Summary of Study Design and Execution Issues
Atmospheric deposition results from wet deposition, from dry aerosol deposition, and, for volatile
and semi-volatile chemicals, from gas exchange. Quality estimates of total atmospheric deposition
loadings to a water body requires that each of these three types of deposition be quantified. In the
TRIADS program, wet deposition fluxes are reasonably well quantified during the study period,
with the caveat that the consequences of possible sample preservation artifacts in the field are
unknown. Dry aerosol deposition fluxes and gas exchange rates were not reported in the Draft
Report and, given the issues raised here, may not be possible with the TRIADS data.
Specific comments on the report follow below, organized by the report headings.
11
-------
1. Quality Assurance. In general, the study met or exceeded their stated quality assurance
objectives, and sufficient oversight of laboratory analyses was evident. Most of the
analytes were present at levels exceeding three times the field blanks, except elements in
aerosol particles (see below). I believe that the last sentence in Attachment A, paragraph 5
is reversed.
2. Report Summary. The Draft Report is primarily a data report, with a minor amount of
initial interpretations. In a broad view, the concentrations of nutrient, metals, and organic
contaminants in precipitation and organic contaminants in air at the Galveston site are
comparable to those reported at other Great Waters locations (see Section II for detailed
comparison between Galveston Bay and Chesapeake Bay).
3. Nutrients. There is a discrepancy in the nitrate concentrations in precipitation presented
in Draft Report Tables 1 and 2. Table 1 reports the nitrate concentration in the 54-69
sample to 5.6 PPM (mg N03"/L) while Table 2 lists 17.41 |imole/L (which is either 0.24
mg-N/L or 1.08 mg N03'/L depending on whether the mole refers to nitrogen or nitrate).
This difference is seen in all of the data. Which is correct? What is the reason for the
difference? Which data were used in the calculations presented in the text?
The amount of precipitation that fell during the study period (963 mm/year) is 25% less
than the long term record (1270 mm/year), indicating that the study period was unusually dry. This
would tend to increase ambient aerosol and gas phase concentrations (drier conditions result in
more fugitive dust emissions and less scavenging by precipitation events) and, naturally, decrease
wet deposition fluxes. Approximately one half of the total nitrogen wet deposition resulted from a
single large event. The author argues that this event is driven by unusual air trajectories during this
time that allowed the air mass to become enriched in nitrogen species. It is not clear in the Draft
Report whether other species were also enriched in the precipitation of this large event. Such
enrichments would lend credence to the meteorological interpretation of this large nitrogen
deposition event. The Draft Report describes calculations to determine the relative importance of
atmospheric deposition as a source of nitrogen to the Galveston Bay, including transport of
12
-------
deposited nitrogen through the watershed. These results, which indicate that atmospheric
deposition contributes ca. 10% of external nitrogen loadings, depend strongly on the validity of the
nitrogen species measurements in rainfall. While these measurements include urea, they do not
(apparently) include other forms of dissolved organic nitrogen (DON). As DON measurements
are available on the TRIADS samples (L. Cifuentes, presented at Galveston Bay Atmospheric
Deposition Workshop), these calculations should be updated.
The calculations presented in the Draft Report suggest that the atmosphere is a trivial
source of phosphorus to the Galveston Bay.
4. Organic Contaminants. The last sentence in the first paragraph under 'PAHs, PCBs, and
pesticides' (pg. 18 of draft report) is misleading. The references cited in the preceding
sentences all refer to the importance of atmospheric deposition as a source of organic
contaminant to remote or rural locations. Nothing in those papers allows one to conclude
that 'atmospheric deposition is a significant source of pollutants to surface waters,
especially coastal waters of industrialized areas, including Texas (emphasis added).
Atmospheric deposition is important in remote and rural areas because other sources are
small or absent. Whether atmospheric deposition is also important in highly industrialized
coastal areas is an open question.
The second complete sentence on the top of page 19 does not reflect the current literature.
Washout of PCBs and PAHs by precipitation is exceptionally efficient, as evidenced by the work
of Murry and Andren (1992) and Poster and Baker (1996a,b).
The Draft Report only contains concentrations of certain dissolved organic contaminants
(PCBs and pesticides) in precipitation. The Investigator has subsequently provided the missing
particulate organic contaminant data to the reviewer, and we have used this data to calculate total
(dissolved+total) wet deposition fluxes to the Galveston Bay. The Draft Report should be updated
to include both dissolved and particulate organic contaminant concentrations in the precipitation
samples, and comparisons to the literature should be made using total concentrations.
13
-------
The volume-weighted mean concentrations of each species in precipitation should be
presented and discussed, rather than the ranges in observed concentrations.
The last sentence in the 'organic air samples' section (Draft Report, pg. 21) is misleading.
'This data for contaminants in the vapor phase will allow the estimate of dry deposition and gas
exchange to be made.' As discussed above, it is not possible to calculate either dry aerosol
deposition or gas exchange fluxes from temporally-integrated gas phase concentrations without
total suspended matter and dissolved phase concentrations and appropriate meteorological data.
5. Metals. The covariance of lead and aluminum is taken as evidence that 'crustal sources'
(e.g., erosion of soils) are the dominant source of lead to precipitation in the Galveston
Bay area. This simple correlation, while suggestive, is not a rigorous test of this
hypothesis. A better approach is to compare the lead/aluminum ratio in precipitation to
that in local soils, or to the global crustal abundances published by Turekian and Wedepohl
(1961). Such a calculation will allow the lead (and perhaps manganese) wet depositional
fluxes to be partitioned into 'crustal' and 'non-crustal' sources.
The Report would be strengthened by adding a 'Conclusions' section.
14
-------
Section II. Comparison of TRIADS Results to those from Other Coastal Areas.
In this section, the results of the TRIADS program, including those in the Draft Report and
in the subsequent discussion with the Investigator, are compared to those from other established
atmospheric deposition networks (Table 3).
Table 3. Comparison of TRIADS Results to Other Atmospheric Deposition Studies
h Chesapeake Bay Atmospheric Deposition Study (CBADS)
metals and organic contaminant deposition
• 3 rural sites 1990-1993
• urban/rural comparison 1995-present
see Baker et al. (1997)
s Great Lakes Integrated Atmospheric Deposition Network (IADN)
metals and organic contaminants
° 5 master sites ca. 1990-present
see Hoff el al. (1996)
s National Atmospheric Deposition Program (NADP)
nitrogen wet deposition
® Attwater Prairie Chicken, TX 1984-1996
Wye, MD 12/95-8/97
http://nadp.sws.uiuc.edu/nadpdata/
ei Atmospheric Integrated Research Monitoring Network (AIRMoN)
• Smith Island, MD 12/95-8/97
http://nadp.sws.uiuc.edu/airmon/
Strategy Differences Among the Programs
While the Galveston Bay sampling generally followed procedures developed in other
atmospheric deposition programs, several exceptions should be discussed before comparing the
results among the programs. Both the CBADS and IADN programs collect air samples for organic
15
-------
analysis over 12 to 24 hour periods every one to two weeks. The IADN composites these samples
monthly, while the CBADS samples are analyzed individually. The TRIADS air sampling for
organic contaminants were collected for ca. six hours every Wednesday, with two or more
sampling periods composited onto a single collection matrix. Given these differing sampling
strategies and the offset in the time of the collections, it is not possible to directly compare
atmospheric concentrations of organic contaminants collected at the same time for the same
periods by the three programs. The best possible comparisons are with temporally-averaged (e.g.,
monthly or annual) concentrations.
Similarly, the integration of precipitation sampling for nutrients differs among the TRIADS,
NADP, and AIRMoN programs. The AIRMoN program's daily precipitation sampling is the most
rigorous, and likely leads to the fewest sample preservation artifacts. The weekly integrated
NADP samples are suitable for accurate nitrate wet deposition fluxes, but may under-predict
ammonium wet deposition by perhaps 15% due to loss of ammonium during sampling storage in
the field. The TRIADS program integrated precipitation samples for varying periods, ranging
from a few event-based samples, to the common weekly-integrated samples, to several samples at
the end of the program that were integrated for more than two weeks. Aside from possible sample
alterations during storage in the field, these differing integration periods compromise the direct
comparison of wet deposition fluxes of nitrogen species among the AIRMoN, NADP, and TRIADS
programs. Here we calculate volume-weighted mean concentrations of each species as the basis
of comparison.
As elements (metals) were not detected in aerosol samples in the TRIADS program (see
discussion in Section I), no comparison is possible of aerosol composition in the Galveston Bay
region to those in other locations.
Comparison of Atmospheric Concentrations of Organic Contaminants Among Programs.
Concentrations of organic contaminants in the atmosphere (linear averages) over the Great
Lakes, the Chesapeake Bay, and Galveston Bay are compared in Figure 1. In general, the
16
-------
8000 y
v
A
D)
a.
c
o
g 2000
4->
c
Q)
a
c
o
o
o
'C
Q>
Q.
° 1000
/
/I
/
Chesapeake Bay
Galveston Bay
Great Lakes
—*jf~i —gp
Figure 1. Concentrations of organic contaminants in air
collected at shoreline sites in Chesapeake Bay,
Galveston Bay, and the Great Lakes.
-17-
-------
concentrations of more volatile polycyclic aromatic hydrocarbons (PAHs), including fluroene,
phenanthrene, anthracend, fluoranthene, and pyrene are enriched two to seven-fold in the
Galveston atmosphere relative to those of the Great Lakes and Chesapeake Bay. These elevated
concentrations of volatile species may reflect either greater emissions from fugitive sources (i.e.,
volatilization from soils, vegetation, and impervious surfaces) or from more direct emissions to
the Galveston atmosphere from local industries and transportation sources. Concentrations of total
PCBs are virtually identical between the Great Lakes and Chesapeake Bay data sets, but the
concentrations over the Galveston Bay are nearly ten-fold higher. Whether these apparent elevated
PCB concentrations over Galveston reflect enhanced volatilization or result from systematic
analytical differences among the programs is unclear. A detailed congener-by-congener
comparison would be required to resolve this issue. Higher atmospheric levels of more volatile
chemicals in this warmer climate is consistent with the idea of enhanced volatilization during the
hot, dry periods that characterize the Texas climate, Concentrations of less-volatile PAHs, which
tend to be emitted with and associated to aerosol particles are quite similar among the Galveston,
Great Lakes, and Chesapeake Bay atmospheres, with the average values generally within a factor
of two of one another.
Comparison of Concentration of Metals, Organic Contaminants, and Nutrients in
Precipitation Among Programs.
To facilitate the comparison of concentrations of chemical species in precipitation,
volume-weighted mean concentrations of each analyte were calculated for the Great Lakes,
Chesapeake Bay, TRIADS, NADP, and AIRMoN data sets. Volume-weighted mean
concentrations account for the inverse relationship between concentration and precipitation
amount, and avoid giving undue weight to small precipitation events that may contain high
concentrations of analytes. Thus, this calculation normalizes somewhat for differences in
precipitation intensity and amounts among the study areas. The volume-weighted mean
concentration multiplied by the amount of precipitation equals the wet-deposition flux.
18
-------
The volume-weighted mean concentrations of lead, arsenic, selenium, and cadmium in
precipitation from the Chesapeake Bay, the Great Lakes, and the Galveston Bay are shown in
Figure 2. The apparent elevation of lead in Great Lakes precipitation may likely reflect a temporal
offset in these data sets, with the Great Lakes average including lead concentrations from an
earlier period that was still influenced by regional emissions from the combustion of leaded
gasoline. Concentrations of lead and cadmium in Galveston precipitation are intermediate
between the Great Lakes and Chesapeake Bay volume-weighted mean concentrations, and all are
likely indistinguishable considering the sampling and analytical differences among the programs.
Selenium may be enriched in Galveston Bay precipitation (50% greater than Great Lakes and 5-
fold greater than the Chesapeake Bay). Whether this enrichment results from local emissions of
selenium is not clear, but may have been resolved if the TRIADS aerosol chemistry were
available. In general, however, trace metal concentrations in precipitation at the Galveston Bay
are not exceptionally high or low relative to the other study areas.
The volume-weighted mean concentrations of organic contaminants in precipitation from
the Chesapeake Bay, the Great Lakes, and the Galveston Bay are shown in Figure 3.
Concentrations of the more volatile PAHs in precipitation follow similar trends as those in the
atmosphere, with relative enrichment in Galveston compared to the Great Lakes and to the
Chesapeake Bay. Concentrations of most PAHs are on the order of twice as high in Galveston
precipitation as compared to that in the Chesapeake Bay region, with the exception of anthracene,
benzo[£]fluoranthene, and dibenz[a,/j]anthracene, which are similar between the two regions.
Volume-weighted mean concentrations of total PCBs are remarkably similar between the three
study areas, with the concentration in Galveston precipitation nearly identical to that reported by
the Great Lakes IADN program. These similar PCB concentrations in precipitation, among the
sites, conflict with the apparent enrichment of PCBs in Galveston air. Again, a detailed analytical
inter-comparison using conger-specific analysis would be required to resolve this discrepancy
between the PCB trends in air and precipitation.
Concentrations of major ions in precipitation are compared in Figure 4. To first evaluate
the potential sampling artifacts associated with weekly-integrated versus event-based precipitation
19
-------
c
o
Ih
<0
c
0)
o
c
O
O
c
re —r-
O)
"a>
1.5
1.0
a>
E
.2
o
>
0.5
x
\°
Chesapeake Bay
Galveston Bay
Great Lakes
• C
Figure 2. Volume-weighted mean concentrations of four metals
in precipitation collected at shoreline stations in the Chesapeake
Bay, Galveston Bay, and the Great Lakes.
-20-
-------
BBEBB8SH Chesapeake Bay
Galveston Bay
Figure 3. Volume-weighted mean concentrations of organic
contaminants in precipitation collected at shoreline stations
on the Chesapeake Bay, Galveston Bay, and the Great Lakes.
-21-
-------
rnsrn Smith Island, Chesapeake Bay (A1RMON)
I I Wye Island, Chesapeake Bay (NADP)
Galveston Bay
t3
0)
o 2
£3 <0
o
G>
ID
~
^ J? if
Figure 4. Volume-weighted mean concentrations of major
ions and total annual precipitation amounts at Galveston
Bay and two Chesapeake Bay sites.
-22-
-------
sampling, volume-weighted mean concentrations of ammonium, nitrate, chloride, and sulfate were
calculated for a NADP site (Wye, MD) and an AIRMoN (Smith Island, MD) site on the
Chesapeake Bay. These two sites are approximately 90 kilometers apart along the eastern
shoreline of the Chesapeake Bay. Precipitation samples collected between December, 1995 and
August, 1997 were used for this comparison. As seen in Figure 4, the overall volume-weighted
mean concentrations of ammonium, nitrate, and sulfate are identical between the NADP and
AIRMoN sites, suggesting no systematic loss of ammonia during the one week integration period
of the NADP sampling. Of course, individual weekly wet depositional fluxes of these species
differed between these sites, as individual precipitation events resulted in differing amounts of
precipitation at the two sites. Averaged over longer periods, however, the concentrations of
ammonium, nitrate, and sulfate are very similar, regardless of collection method. Notice that
chloride is three times higher in precipitation collected at the Smith Island AIRMoN site relative
to that at the Wye NADP site. This increase is also seen in the sodium concentrations (data not
shown), and certainly reflects the more marine environment of the more southern Smith Island site.
Marine aerosols, including local sea spray, are apparently impacting the sodium and chloride
levels in the precipitation at Smith Island. This represents a real spatial difference between Smith
Island and Wye, rather than a sampling artifact. When compared to the two Chesapeake Bay sites,
concentrations of ammonium, nitrate, and sulfate are considerably lower in the Galveston Bay
precipitation samples (Figure 4). Ammonium and nitrate concentrations in Galveston are
apparently one-half and one-fifth of those measured in the Chesapeake Bay region. Whether this
difference reflects a real difference in nitrogen wet deposition, or results from the longer sample
integration periods in the Galveston Bay study is unclear. However, the fact that the sulfate
concentrations are also lower at Galveston relative to the Chesapeake suggest that, in fact, the
rainwater along the Texas coast contains lower concentrations of major ions.
To compare the Galveston TRIADS nutrient deposition results to other studies in the area,
volume-weighted mean concentrations of nitrate and ammonium from the TRIADS study are
compared to annual average concentrations collected over a twelve year period at the Attwater
Prairie Chicken, Texas NADP site. This site is the closest NADP site to the Galveston Bay, but is
ca. 125 kilometers from the coast. As is seen in Figure 5, the volume-weighted mean
23
-------
1.0
J 0.9
re
k.
§ 0.8
o
oB) 0.7
s£
(0
£ 0.6
0.5
0.4
0.30
0.25
c
o
"
c
-------
concentrations of both nitrate and ammonium at Galveston Bay are substantially less than and
outside of the observed range of the long-term average at the Attwater Prairie Chicken NADP site.
One possible reason for this difference is loss on nitrogen species during the longer integration
periods of the Galveston sampling. Another explanation is that the volume-weighted mean
concentrations of nitrate and ammonium are, in fact, lower at the Galveston Bay site than at the
inland NADP site. The latter's interpretation is supported by the comparison of wet deposition
fluxes at the Galveston and Attwater Prairie Chicken NADP sites (Figure 6). Wet deposition
fluxes, equal to the volume-weighted mean concentrations of each species multiplied by the annual
precipitation amount, were more similar between the sites than were the corresponding
concentrations. This suggests that the Galveston site receives more precipitation which may serve
to dilute the nitrogen species in the rainwater. Note that the wet deposition fluxes of both
ammonium and nitrate to the Galveston Bay, while more similar to the NADP data, are still lower
than those measured at the Attwater Prairie Chicken site in 1995. Apportioning this difference
between possible systematic sampling differences and real spatial trends is not possible without a
rigorous side-by-side inter-comparison.
25
-------
c
o
]5
"w
o
a
®J=
« O)
+=^£
"re
3
C
c
<
10
8
c
o
+3
'
a
+¦»
0> «-
ss
i|
if
E~
E
re
3
C
C
<
© Attwater Prairie Chicken NADP
y Galveston Bay
1984 1986 1988 1990 1992 1994 1996
Figure 6. Comparison of nitrate and ammonium wet deposition
fluxes to Galveston Bay to the long-term record at the Attwater
Prairie Chicken NADP site.
-26-
-------
Conclusions - Comparison of Galveston Bay Data to Other Studies
Concentrations of more volatile polycyclic aromatic hydrocarbons and polychlorinated biphenyls
are elevated in air collected at the Galveston Bay Atmospheric Deposition Study site relative to
those measured at rural shoreline sites in the Great Lakes and Chesapeake Bay regions.
Concentrations of less volatile organic contaminants, which tend to associate with aerosol
particles, are comparable among the three study areas. Concentration of more volatile organics
are 50-100% higher in precipitation collected in Galveston Bay relative to that sampled in the
Chesapeake Bay. In general, the magnitude of the enrichments in specific organic contaminants in
Galveston air and precipitation are comparable between the two media. Concentrations of lead
and cadmium in Galveston precipitation are approximately one-half of those concentrations
measured in the Chesapeake Bay region. Higher lead levels in Great Lakes precipitation likely
results from the inclusion of data from the early 1980's, when atmospheric lead levels still
reflected use of lead in vehicle fuels. Concentrations and wet deposition fluxes of nitrate and
ammonium at the Galveston Bay site are one half of those at the adjacent NADP site and one-third
to one half of those in the Chesapeake Bay. Whether these lower nutrient inputs to Galveston
reflect actual spatial differences or result in losses during sampling is unclear.
27
-------
Section III. Galveston Bay Atmospheric Deposition Workshop Summary
The following pages include the Agenda, Participant's List and notes taken during the workshop on
11 August, 1998.
AGENDA
GALVESTON BAY ATMOSPHERIC DEPOSITION WORKSHOP
11 August 1998
Art Gallery Room, Bayou Building
University of Houston, Clear Lake
2700 Bay Area Blvd., Houston, TX 77058
9:00 a.m. - 4:00 p.m.
8:30 - 9:00 Registration and Coffee
9:00 - 9:10 Welcome and Introductions
9:10 - 9:30 Atmospheric Deposition to Estuaries: Management Needs
9:30 - 11:00 Galveston Bay Atmospheric Deposition Study Wade
11:00 - 11:20 Discussion of Galveston Bay Study Results Group
11:20 - 11:40 Break
11:40 - 12:00 Comparing Deposition to Galveston Bay and Other Estuaries
12:00 - 12:45 Lunch
12:45 -1:15 The Tampa Bay Atmospheric Deposition Study Greening
1:15- 1:45 Determining Long-Term Atmospheric Deposition Trends
1:45 - 2:30 Determining Sources of Contaminants in Atmospheric Deposition
2:30 - 3:00 Transmission of Deposited Metals Through Estuarine Watersheds
3:00-3:15 Break
3:15 - 4:00 Relating Galveston Bay Results to Management Questions
Adjourn
Baker
Ackermann
Baker
Simcik
Ondov
Mason
Baker
28
-------
Participant's List
at the
Galveston Bay Atmospheric Deposition Workshop
August 11, 1998
University of Houston at Clear Lake, Texas
Name
Affliation
Teleohone
Email
John Ackerman
US EPA - Great Waters
919-541-7689
ackermann.john@epa.gov
Daine Attsman
US EPA/GMPO
228-688-7015
altsman@pelican.gmpo.gov
Sandra Alvarado
CCBNEP
512-980-3420
salvarad@tnrcc.state.tx.us
Joel Baker
CBL, UMD
410-326-7205
baker@ cbl. umces .edu
David Brock
TWDB
512-936-0B19
dbrock@twdb.state.tx.us
Glenda Callaway
Galveston Bay Foundation/Council
EKISTICS/GBF/GBEP
713-520-9031
glencall@aol.com
Tom Catnan
GLO
512-463-5100
tcalnan@glo.state.tx.us
Ben Carmine
Houston Industries
713-945-8191
ben-carmine@hlp.com
Luis Cifuentes
TAMU, College Station
409-845-3380
cifuentes@ocean.tamu.edu
Paul Dietert
Galveston Bay Info Center - TAMUG
409-740-4703
gbic@tamug.tamu.edu
Gary Gill
TAMUG
409-740-4710
gillg@tamug.tamu.edu
Holly Greening
TBEP
727-893-2765
tbnep@tampabayrpc.org
George Guillen
TNRCC Houston
713-767-3505
gguillen@ix. netcom .com
John Jacob
Texas Sea Grant
281-291-9252
jjacob@tamu.edu
Fred Kopfler
US EPA/GMPO
228-688-2712
kopfler@pelican.gmpo.gov
Rob Mason
CBL, UMD
410-326-0911
mason@cbl.umces.edu
Carl Masterson
Houston-Galveston Area Council
713-993-4561
cmasters@hgac. cog .tx. u s
Gary Mitchell
Houston-Galveston Area Council
713-993-4581
gary.mitchell@hgac.cog.tx.us
29
-------
Participant's List
at the
Galveston Bay Atmospheric Deposition Workshop
August 11,1998
Clear Lake, Texas
Name
Affliation
TeleDhone
Email
John Ondov
UMCP
301-405-1859
jondov@wam.umd.edu
Junesoo Park
TAMU
409-845-5346
jsp0724@unix.tamu.edu
Linda Shead
Galveston Bay Foundation
281-332-3381
Matt Simcik
Indiana University
812-855-5976
msimcik@indiana.edu
Sharron Stewart
Galveston Bay Foundation/Council
409-297-6360
Stephen Sweet
TAMU
409-862-2323
x133 sweet@GERG.tamu.edu
Terry Wade
GERG TAMU College Station
409-862-2323
Terry@GERG.tamu.edu
Randy Waite
US EPA - Great Waters
919-541-5447
waite.randy@epa.gov
-------
Galveston Bay Atmospheric Deposition Workshop
Notes - August 11, 1998
9:07 a.m. - Meeting began with opening remarks by Dr. Joel Baker
John Ackerman - EPA- Great Waters Program
~ Atmospheric chemicals concerns hazardous chemicals transported by air.
« Nitrogen concerns
• nitrate limiting factor
• legislation says we need to trace things
• fish advisory - find out where it is coming from - sediments within the water body
® pollution problems - 1970 - beginning of EPA and clean air act
• find out if local or national chemicals - emissions - where?
• Establish if we need to take action which sources - two reasons for Great Waters Project -
study and calculate problems and follow up with actions
• tools to use - waste combuster to achieve reductions and study the National Estuarine
Programs started in connection with Houston, Baltimore, Boston (started in Tampa).
? - Terry Wade - Texas A&M - What is there? What is the air/the deposition? Gas - liquid -
particles - solids - measure close to the water body nutrients - organics - broad spectrum
approach. Consider - relate to sources - what is in the Bay? Sources?
~ 11 Management Questions - (Joel has list)
• potential is there to look at other sources that are there - ozone, fine particles, nitrogen,
emissions inventories, databases - including their sources.
• This data is not quality assured
• Houston Metro Area has a lot of these sources but also Mexico and surrounding areas.
~ GOAL: Try to put all of this together and reduce the levels coming from the sources.
TERRY WADE- Texas A&M
Slide program:
#1 - DDT - spreads long distances - studying transport
#2 - exposure
#3 - Pollutants of concern - DDT recently banned in Mexico, PAH's, Toxaphene
#4 - Why we are concerned - endocrine disrupters - adverse effects
31
-------
Galveston Bay Atmospheric Deposition Workshop
Notes - August 11,1998
#5 - cartoon - estuaries and land surfaces are connected - hydrocarbons emitted complex
processes - rain (wet) or dry deposition - measuring oysters for long-term contaminants.
Questions still remain concerning atmospheric deposition.
#6 - Great Waters program\
#7 - Relative Loadings
#8 - continues #7 - chemicals may cycle through the environment - need to understand the cycling
of these chemicals
#9 - sources
#10 - Photo - liquid industrial area
#11 - Atmospheric Deposition - wet = rain, snow; dry = particles; gas exchange
#12 - Galveston Bay, Texas (map)
#13- Seabrook Intermediate School - Sampling Site
#14 - photo of Sampler
#15 - Set up of meteorological sensor - logged on data logger
#16 - filter/tube/polyurethane plug
#17 - Sampler - large vacuum
#18 - Calibration kits to all of this equipment
#19 - Sampling Site - Baker Sampler - dry bucket
#20 - Covers open when sensor is wet
#21 - Photo of the sampler open - only measuring precipitation
#22/23 - site photos with sampler
#24 - (couldn't see)
#25 - Area around the sampler - showing legitimacy of sampling
#26 - Air Toxics Deposition Monitoring - Seabrook, Texas
#27 - TRIADS - Cumulative Rain Volume - Graph - College Station, TX - currently in a drought.
#28 - XAD Resin PAH (ng/m2) graph
#29 - TRIADS XAD total PAH 3-dimensional graph
32
-------
Galveston Bay Atmospheric Deposition Workshop
Notes - August 11,1998
#30 - Conclusions - may be based on other Great Waters programs - We haven't used DDT in over
20 years but it is still evident in our atmosphere.
#31/32 - photos
#33 - photo of rural area for Site Sampling
#34 - Spraying pesticides on crops
#35 - results (photo)
<- Data when sampler open and closed
~ Deposition in fog - trying to study
Transparencies:
• meteorological data - air/wind/rain data
• Wind mass trajectories / back trajectories
• shows complexity and need more information
~ 10% of wet and estimate of dry nitrogen coming into surface area are from Atmospheric
Deposition (not nitric acid). Numbers are uncertain. Long term samples and collect all rain
sample occurrences
~ Bird contamination - from siting on samplers - checked levels - birds were not a factor
~ measured PAH's in the rainfall - (Polynuclear aromatic hydrocarbons)
~ no seasonal trends but times that are higher and lower
~ need to figure out what is going on - Joel is helping with this.
~ PUF Air Samplers
Have all the data and are now trying to understand it all (figure it out)
*¦ Appears that iron and alumina are in a constant correlation - these concentrations are very low.
33
-------
Galveston Bay Atmospheric Deposition Workshop
Notes - August 11, 1998
Questions:
? - How much do yo expect to learn from 1 station?
will compare against Chesapeake Bay and the Great Lakes
- lack of funding prevents having the 5 needed sites on Galveston Bay
? - Can you correlate your data with others?
studies indicate that shellfish pose a human health risk - we are interested in how much of
this is Atmospheric Deposition.
? - How does this study effect TMDL's (Total Maximum Daily Load)?
- EPA answered.
Luis Cifuentes. Texas A & M. College Station (continuation of T. Wade's)
~ Total Nitrogen - Sumatzu System
~ measured Nox compounds - has problems and advantages - smaller samples / less times
~ can measure for Total Dissolved Nitrogen
Transparencies:
• Anthropogenic Nitrogenous
® interest has grown
° What is the concentration of DON (dissolved organic nitrogen) in water and could it be an
impact? Based on the Thesis work of (sp?) Zhang Ho
Gary Gill
Mercury data:
® bootlegged off other projects
• results of under graduates in his lab.
® Florida atmospheric mercury study just ended recently
® Bill Landing - Began 1992
Kurt Coleman - End 1997
o elevated levels of mercury in fish
® 9 sites - rainfall/particles/ wet & dry
34
-------
Galveston Bay Atmospheric Deposition Workshop
Notes - August 11, 1998
• end results of 5 year study
® not a dramatic range in values
• bulk- month long samples
• most mercury comes out in rain - wet deposition
° will contrast data set from Florida with Texas
« mercury falls out during the summer months in Florida
TEXAS: 3 different pieces of information from three sites (see transparency fig.l)
® ran duplicate samples at Seabrook Site and then did a comparison (an average)
• some higher expersions(sp?) than the Florida data
• some highs get muted when you look at the flux
• background sites - 1 yr period
- flux
- concentrations
° sea drift site further down the coast
• Seabrook - anthropogenic influence compared to other 2 sites
» trying to understand mechanism and processes
Junesoo Park - Texas A&M - College Station
Transparencies:
• Baker Sampler - hi volume air sampler
• particulate phase of PAH's not very hi influence
• seasonal trend - low concentrations May and beginning of June
« petroleum sources
• combustion sources
JOEL Baker - compare Galveston vs. Others
~ 2 problems - are differences real?
~ Compare concentrations and fluxes
35
-------
Galveston Bay Atmospheric Deposition Workshop
Notes - August 11,1998
~ Transparencies:
• 5 master sites - Great Lakes
»• In Air:
• Organics
• Metals 6 hr. parcels of samples
~ Results of nutrient deposition
Holly Greening - Tampa Bay
Transparencies:
~ Sea grass Restoration Project:
• Sea grasses were dying back
• maintain level of nitrogen
• currently - adding 500 acres of sea grass per year
• it will take 25 years to get back to 1950 estimates if the grasses continued to comeback at a
rate of 500 acres/yr
• sources of nitrogen to Tampa Bay - external only - 50% storm water, etc. (See graph)
• Waste water treatment regulations have helped to reduce their contribution to Atmospheric
Deposition
~ Unknowns:
° don't know how much emissions are going into Tampa Bay
» 1995 - suggested a multi-level approach
o received funding from Great Waters Project - (see transparency for list of participants in
the study)
o Now have 18 months of data
° measuring ambient air
° Buoy model - developed by NOAA
« One measuring site
• Wet Deposition - NADP periodicals - daily measuring if raining
~ Results from:
• wet deposition - nitrogen - scattered
36
-------
Galveston Bay Atmospheric Deposition Workshop
Notes - August 11, 1998
® rainfall - similar
o Dry Deposition - 6 day for 24 hr. - scattered - monthly pattern - higher during fall to winter
• relation between Wet and Dry Deposition - dry is higher except (see graph)
• Mean concentration vs. wind direction - no obvious pattern
• Next steps with Tampa Bay Program (see transparency)
Questions:
? - What kind of errors are associated with the different techniques? Do we really know the
uncertainty?
we have the same questions. Richard Valigura has written this up on the Buoy model.
(Joel's response) - biggest limitation - assumes you are in the middle of the ocean, away
from contamination, to sample. (John Ackermann's response) - using consistent models
and methods - helps to maintain validity of information. We have to live with the fact that
dry deposition has some variability.
Matt Simcik
~ First, atmospheric deposition of lakes -established 5 master sampling sites on Great Lakes.
~ New questions as data is collected so project evolves continually.
~ Criteria: with 1 km of shore - remote areas - scale -> Great Lakes are as big as Texas and
are influenced by [Detroit, Chicago vs. plants and industry].
° Set-up at each site - similar to others
• concentrations vs. time - plotted 200 over 7 years - not a lot of information from this plot -
different concentrations - different wind - different temperature
• remove temperature by using (sp?) Claudius/Claperon Equation
~ Temporal Trends:
® challenges - 3 different kinds of compounds [banned, continual, seasonal]
o then over this, events can occur, such as spills - droughts- heavy rain
John Ondov
Transparencies:
• particle size is very important to velocity
37
-------
Galveston Bay Atmospheric Deposition Workshop
Notes - August 11, 1998
» higher deposition at higher wind speeds
? - Chemical Mass Balance - What is yours?
- Box model - another order of difficulty
? - If you take a standard sample - how different would the flux be - very different numbers?
- if you keep both small and large velocity then you have to determine what is the size
distribution.
Rob Mason
Transparencies:
~ The Science Center is higher overall in mercury than the other two sites on the Chesapeake
Bay.
*¦ Trying to measure gaseous ionic mercury species
• use 2 different methods for comparing
• Conowingo outlet - mercury loadings - large flows of mercury - storm, late winter early
spring
° relate mercury t solids concentrations
° mercury tracks particles fairly well
• the different watersheds have an impact on the amount of cadmium distributed
• we need to understand the transport through the watershed before we can understand
Atmospheric Deposition.
Terry Wade: PAH's - wash-off from storm water - How much from air ?- Probably not very
much.
~ Uncombusted oil from Galveston Bay - approximately 40% of a yearly average of spills(?)
~ Sources - better characterizing
~ Is our research at a significant level?
® Tampa has put together the most in proof of continuing research.
38
-------
Section IV. Overall Summary and Conclusions.
TRIADS was the initial investigation into the potential importance of atmospheric
deposition as a source on nutrients and chemical contaminants to the Galveston Bay. As with any
initial study in a geographic region, TRIADS made an important contribution by providing the first
high quality measurement of contaminants in the air and in precipitation in the Galveston Bay
region. The main contribution of the project, which results directly from the excellent laboratory
analysis, is the cataloging of ambient concentrations and wet deposition fluxes on organic
contaminants, nitrogen species, and some metals. Like other initial studies, the investigators had to
make a choice between intensive study at a single location versus a more cursory examination at a
larger number of sites. In TRIADS, the choice was to operate a single sampling site in order to
concentrate the available resources. This was a reasonable choice, though it leaves all of the
questions regarding spatial variability in atmospheric deposition in the Galveston Bay area
unanswered.
Though TRIADS was successful in terms of providing the initial look at atmospheric
deposition to the Galveston Bay, several strategy issues limit the extent to which the data gathered
there can be used to answer many of the management questions. To be fair to the investigators,
these questions, and the whole issue of atmospheric deposition to coastal waters, evolved during
the course of the TRIADS study, well after the program had been initially designed. Nonetheless,
future studies must avoid these pitfalls. The major criticisms include: (1) long sample integration
times (many weeks) may result in sampling artifacts and severely compromise the ability to relate
observations to meteorological conditions; (2) inappropriate sampling techniques resulted in few
valid trace metal measurements in ambient aerosol particles, eliminating the possibility of using
source-receptor models to estimate contaminant sources; and (3) no direct or indirect estimates of
dry aerosol particle deposition or diffusive gas exchange across the air-water interface were
made, resulting in an incomplete estimate of total atmospheric deposition loadings to the Galveston
Bay.
39
-------
With this review in mind, the Galveston Bay atmospheric deposition program provides
initial responses to several of the management questions posed at the Galveston Bay Atmospheric
Deposition Workshop.
1. What Hazardous Air Pollutants (HAPs) are found in the deposition samples and in what
amounts?
The study found detectable levels of PAHs and PCB congeners in air and precipitation.
Sampling artifacts limited the measurement of metals in aerosol samples. Concentrations
of organic contaminants in air and precipitation are within 2-3 fold of those found in the
Chesapeake Bay and the Great Lakes, with some indication of higher concentrations of
more volatile species. This enrichment may result from either enhanced volatilization in
the warmer environment or from emissions from local industries and other activities.
2. Which HAPs are also found in significant amounts in the water, sediments and/or biota?
The TRIADS study did not address the presence of hazardous air pollutants in water,
sediments, and biota of the Galveston Bay. Participants at the workshop summarized areas
of the Bay that are impacted by petroleum hydrocarbons and heavy metals that may have an
atmospheric source.
3. Does deposition of HAPs exhibit temporal, seasonal and/or directional variations?
Due to the relatively long integration times of the sampling program, and to the short total
duration of sampling, it is difficult to discern temporal, seasonal, or (wind) directional
variations. One event that contributed substantial nitrogen wet deposition loadings was
traced to an unusual airflow pattern that brought urban air over the study site.
40
-------
4. How significant are these variations?
Long sample integration times dampen temporal variations in atmospheric samples.
However, the site apparently is subjected to extreme events periodically, as evidenced by
the single nitrogen-enriched deposition event documented in the report. The relative
frequency of these events and their importance in the overall atmospheric deposition story
could only be determined by a long-term, consistent sampling program.
5. Was the period for which data was collected meteorologically typical?
The study period was warmer and drier than the long term average for the Galveston Bay
region, with the precipitation amount approximately 25% below the normal value. How
this may impact total atmospheric deposition (wet+dry aerosol+gas exchange) is unclear,
as drier, warmer conditions may result in elevated atmospheric concentrations of volatile
and particle-associated chemicals but less efficient deposition processes.
6. From what directions do the most significant deposition contributions come?
While there were an insignificant number of discrete (i.e., non-integrated) samples
collected under uniform meteorological conditions to address this question, several
samples show elevated concentrations when the winds brought air from the urban region.
Interestingly, concentrations of dissolved organic nitrogen in precipitation also seemed to
be higher in urban-impacted samples, in contrast to the idea that the major sources or
organic nitrogen are agricultural or natural vegetation (L. Cifuentes, presentation at
workshop).
41
-------
7. For what times of day or year do the most significant contributions come?
Again, with the temporally-integrated data available and the relatively short study period,
it is not possible to determine seasonal trends in deposition. As all air samples were
collected during the day, diurnal variation cannot be determined.
8. What does upwind air quality monitoring data indicate regarding trajectories and areas
of contribution?
As explored at the workshop, there is apparently no consistent air quality monitoring
program that measures the organic contaminants successfully measured in the TRIADS
program. Therefore, it is not possible to compare the ambient levels measured at the
Galveston site to those measured 'upwind' in the urban areas. While it may be possible to
use 'surrogate' air quality parameters {i.e., total suspended matter, ozone), these were not
measured by TRIADS on the same schedule as the target analytes.
9. What does HAP emissions data indicate about the sources and quantities of emissions
from these upwind areas?
This question was not formally addressed in this review or at the workshop. However, it
is worth noting that for many of the compounds reported in the TRIADS program
(especially PCB congeners and combustion-derived PAHs), existing emission inventory
procedures consistently fail to adequately quantify total emissions to the atmosphere.
These compounds have dominant diffuse area sources, including volatilization from soils
and surface waters, that are not captured by traditional emission inventories.
42
-------
10. What gaps, limitations, and uncertainties about data quality and completeness can be
identified and characterized for each step of this analysis?
Specific issues and recommendations are addressed throughout this review document.
11. What practical recommendations can be made for addressing them?
Specific issues and recommendations are addressed throughout this review document.
43
-------
APPENDIX A
44
-------
Air concentration Summary Stastics
OC's & PCBs: Vapor Phase Only
PAHs: Vapor + Particle
Average
Ma*
Mill
Std Dev
Count
Geomett
Total BHC's - Vapor
265.8
673.9
47.8
153.6
25
222.3
Total of 18 PCB's
213.9
735.7
33.5
173.9
25
161.2
NS&T Total PCB's
472.8
1615.6
77.8
380.8
25
358.8
Measured PCB's - Vapor
1706.7
15120.2
209.0
2977.5
25
948.8
Total Chlordanes - Vapor
144.4
291.2
30.8
67.0
25
128.4
Alpha HCH
79.4
174.3
7.8
51.0
25
63.5
Hexachlorobenzene - Vapor
87.3
362.1
4.6
103.3
25
51.2
Beta HCH
]0.2
51.7
1.0
14.8
10
6.2
Gamma HCH
135.1
402.6
38.0
94.5
25
106.4
Delta HCH
61.2
286.1
1.6
76.8
19
28.7
Heptachlor
30.4
77.2
6.1
18.5
25
24.9
Heptachlor Epoxide
14.3
30.4
4.1
6.7
23
12.5
Oxychlordane
3.8
6.5
0.9
1.5
22
3.4
Gamma Chlordane
38.2
76.4
8.8
18.5
25
33.5
Alpha Chlordane
29.6
57,6
8.0
14.3
25
26.2
Trans-Nonachlor
25.6
51.7
7.5
12.4
25
22.5
Cis-Nonachlor
5.0
10.7
1.5
2.4
21
4.4
Aldrin
18.2
93.2
0.7
33.3
8
4.0
Dieldrin - Vapor
19.0
72.1
5.4
13.4
25
16.1
Endrin
4.8
13.3
1.5
3.9
7
3.9
Mirex
1.2
2.6
0.5
0.8
10
1.0
2,4' DDE
56.5
375.4
1.8
83.1
23
23.2
4,4' DDE
1 1.7
37.9
4.0
8.9
24
9.6
2,4' DDD
3.6
13.2
0.2
3.0
18
2.4
4,4' DDD
3.1
5.7
0.4
1.5
17
2.6
2,4' DDT
7.5
23.3
1.6
6.1
21
5.8
4,41 DDT
8.6
13.6
3.8
2.8
22
8.2
PCB8/5
49.5
295.2
5.4
78.2
13
25.1
PCBI8/I7
27.4
126.2
2.2
28.5
23
17.7
PCB28
28.2
160.5
3.9
31.0
24
20.7
PCB52
27.6
99.9
1.2
24.1
24
19.3
PCB44
31.9
96.3
4.7
24.0
23
23.2
PCB66
3.4
5.7
l.l
1.4
17
3.1
PCBlOl/90
23.6
156.3
2.6
31.5
23
14.7
PCB1I8
4.6
10.0
1.9
2.4
21
4.1
PCBI53/132
6.0
26.9
1.7
5.4
21
4.8
PCB105
2.1
3.0
1.0
0.6
14
2.0
PCB138/160
7.4
16.0
2.1
3.8
22
6.5
PCBI87
1.1
2.4
0.3
0.5
20
1.0
PCB128
1.9
3.5
0.8
1.0
8
1.7
PCB180
35.6
201.8
3.1
56.7
24
13.5
PCB170/190
6279.0
31184.7
0.3
9848.1
15
99.6
PCB195/208
0.2
0.4
0.1
0.1
9
0.2
PCB206
0.6
1.5
0.1
0.4
12
0.4
PCB209
3.2
13.3
0.1
4.4
21
1.4
-------
Average
Max
Min
Std Dev
Count
Geometric Mean
PCB7/9*
131.3
848.6
12.0
194.7
21
66.1
PCB 15*
16.5
108.7
0.7
24.4
21
8.4
PCB24/27*
152.7
1504.1
4.8
370.2
16
39.4
PCB16/32*
307.7
1678.8
3.9
516.9
11
65.6
PCB29
190.1
855.4
16.0
257.1
13
83.4
PCB26*
16.4
63.6
2.3
20.5
8
9.2
PCB25*
39.0
96.1
8.8
23.3
18
32.7
PCB33/53/20*
0
PCB22/51 *
0
PCB45*
4.1
10.8
0.8
3.7
8
3.0
PCB46 *
9.3
38.3
1.9
9.3
18
6.5
PCB49*
16.8
78.6
2.1
17.3
20
11.7
PCB47/48/75*
69.8
301.5
10.4
65.0
23
48.9
PCB42/59/37*
19.9
102.8
2.8
22.8
21
13.2
PCB41/64*
342.2
861.8
84.4
285.1
10
260.3
PCB40*
19.9
143.3
1.0
36.4
15
6.9
PCB74*
16.0
51.7
0.4
17.2
13
7.7
PCB70*
21.1
73.5
2.3
23.6
14
11.7
PCB88*
181.4
1034.7
8.7
280.3
18
62.4
PCB56/60*
38.2
467.2
1.8
118.9
15
7.6
PCB92*
5.8
13.1
1.4
3.9
10
4.8
PCB84*
7.7
28.1
0.6
8.4
12
4.3
PCB99*
8.5
19.4
2.3
5.3
18
6.9
PCB83*
4.1
7.8
0.7
2.6
13
3.1
PCB97*
14.2
58.1
1.3
17.6
18
7.4
PCB87/115
7.6
41.5
0.5
11.0
17
3.4
PCB85*
7.3
31.8
0.1
11.2
7
2.5
PCB 110/77*
8.4
15.2
2.4
3.6
22
7.6
PCB82*
6.5
28.6
0.6
7.6
13
4.1
PCB 151*
1.6
4.7
0.1
1.2
13
1.2
PCB107*
8.5
38.7
0.8
16.9
5
2.0
PCB 149/123*
3.7
12.1
0.3
3.0
21
2.7
PCB 146*
7.7
59.9
0.3
19.6
9
1.5
PCB141/179*
2.8
8.7
0.6
2.4
18
1.9
PCB 176/137*
2.1
3.8
06
1.2
13
1.7
PCB158*
29.1
146.0
0.3
54.2
7
4.2
PCB 129*
0.9
2.4
0.3
1.0
4
0.6
PCB 178*
0.9
3.5
0.1
1.0
9
0.6
PCB 183 *
1.6
5.4
0.2
L.3
18
1.2
PCB 167*
0.5
0.5
0.5
1
0.5
PCB185 *
1.4
6.8
0.2
2.0
10
0.8
PCB174*
1.0
2.2
0.3
0.7
11
0.8
PCB 177*
0.7
1.2
0.2
0.4
7
0.5
PCB 156/171
5.6
13.6
0.2
5.5
7
2.6
PCB2O0*
0.4
2.0
0.0
0.6
10
0.2
PCB 172*
0.9
2.2
0.4
0.5
12
0.9
PCB191*
3.1
14.5
0.1
5.2
13
0.6
PCB201 *
69.8
338.8
0.3
130.4
7
3.4
PCB203/196*
0.8
3.2
0.0
1.1
8
0.3
PCB 189*
0.2
0.5
0.0
0.2
5
0.1
PCB194
0.6
4.7
0.0
1.4
10
0.2
PCB205 *
1.0
1.6
0.3
0.7
3
0.8
-------
PAH's:
Average
Max
Min
Std Dev
Count
Geometri
PAH Compounds
Naphthalene
1529.5
17409.3
67.7
3403.1
25
651.4
C1-Naphthalenes
2490.8
33822.9
75.0
6683.7
25
736.6
C2-Naphthalenes
1752.8
19039.6
31.8
3790.4
25
578.0
C3-Naphthalenes
2310.7
17599.6
40.3
3784.7
25
836.6
C4-Naphthalenes
1959.4
11997.3
27.9
2678.5
23
800.3
Biphenyl
431.3
3913.2
26.7
781.2
25
203.9
Acenaphthylene
154.6
1100.1
8.7
228.9
25
75.5
Acenaphthene
533.1
2801.8
30.8
698.5
25
279.1
Fluorene
1639.8
5892.6
67.9
1573.9
25
1005.2
Cl-Fluorenes
1362.3
3868.3
62.3
1006.2
25
961.2
C2-Fluorenes
2314.9
5679.3
180.2
1525.0
23
1758.1
C3-Fluorenes
2253.5
4631.1
184.4
1424.7
23
1681.4
Phenanthrene
12390.0
57847.5
719.3
12148.8
25
8093.4
Anthracene
563.7
4520.6
37.7
892.6
25
300.9
C1 -Phenanthrenes/Anthracenes
3831.2
9090.8
289.3
2420.2
23
2891.5
C2-Phenanthrenes/Anthracencs
2266.3
5136.2
165.4
1486.8
23
1690.7
C3-Phenanthrenes/Anthracenes
1136.0
2940.0
80.4
824.8
23
795.9
C4-Phenanthrenes/Anthracenes
321.7
784.1
25.7
222.6
22
234.7
Dibenzoth iophene
929.4
2584.3
83.9
610.6
25
695.2
Cl-Dibenzothiophenes
889.7
2002.9
84.7
498.8
23
698.3
C2-Dibenzothiophenes
1002.3
2422.9
95.0
597.2
23
784.5
C3-Dibenzothiophenes
976.4
3794.4
75.8
927.9
23
624.7
Fluoranthene
5965.1
33345.2
208.2
7603.2
25
2935.4
Pyrene
3287.4
19042.7
169.1
4313.0
25
1657.6
C1 -Fluoranthenes/Pyrenes
589.9
2919.8
45.6
635.6
23
376.7
Benzo(a)anthracene
49.2
240.1
10.6
49.6
25
38.0
Chrysene
291.4
1320.3
59.9
335.9
25
195.1
Cl-Chrysenes
106.4
260.0
15.2
76.5
12
76.8
C2-Chrysenes
84.2
280.0
16.6
81.2
13
56.7
C3-Chrysenes
0
C4-Chrysenes
6.6
8.5
4.6
2.8
2
6.3
Benzo(b)fluoranthene
110.7
602.3
17.3
121.1
25
80.7
Benzo(k)fluoranthene
28.2
110.4
6.9
20.6
25
23.7
Benzo(e)pyrene
59.2
447.5
10.7
83.1
25
42.8
Benzo(a)pyrene
48.3
301.0
10.0
63.4
25
33.9
Perylene
23.8
81.9
6.6
17.9
25
19.3
Indeno(l ,2,3-c,d)pyrene
54.0
429.6
8.1
80.6
25
36.5
Dibenzo(a,h)anthracene
13.7
62.1
1.6
13.4
25
9.8
Benzo(g,h,i)perylene
59.1
311.5
12.7
60.7
25
44.4
2-Methylnaphthalene
1430.1
21367.7
16.8
4251.6
25
257.3
1 -Methy lnaphthalene
961.9
12455.2
29.8
2462.0
25
303.5
2,6-Dimethylnaphthalene
709.9
7173.1
7.4
1599.3
20
186.8
1,6,7-Trimethylnaphthalene
574.5
3505.9
8.3
869.4
20
191.5
1 -Methylphenanthrene
570.2
1169.9
5.3
341.1
21
382.0
-------
800 .
Total BHC's - Vapor
o Total BHC's - Vapor
700
600
500
CO
£
~oh
3 400
o
c
o
~
~ o
~
300 -
200
~
0
o
100
Q
0
0
95000
95100
95200 95300
Julian date taken off
Starting from 1/1/95 (year,day)
95400
95500
95600
-------
16000
14000 -
12000
10000 -
m
E
"ob
a 8ooo
o
c
o
U
6000
4000 -
2000
0
95000
Measured PCB's - Vapor
~
o
~ o ~
~
O 0
» Measured PCB's - Vapor
0 o
95100
95200 95300
Julian date taken off
Starting from 1/1/95 (year,day)
95400
95500
95600
-------
350
300
Total Chlordanes - Vapor
~ Total Chlordanes - Vapor
250 -
~
~
~
o
P 200
E
a .50
~ ~
o ^ ~
o
0
100
o ~
~
50 -
0
~
95000
95100
95200 95300
Julian date taken off
Starting from 1/1/95 (year,day)
95400
95500
95600
-------
80
70
60 -
50 -
Dieldrin - Vapor
~ Dieldrin - Vapor
£
a 40
o
c
o
O
30 -I
~
* ~
20
~
10
95000
95100
95200 95300
Julian date taken off
Starting from 1/1/95 (year,day)
95400
95500
95600
-------
400
350
300
250
Hexachlorobenzene - Vapor
~
~ Hexachlorobenzene - Vapor
S
~£b
3 200
u
c
o
U
150
~
100 -
50 -
0
~ ^
o o o
$
~
o
95000
95100
95200 95300
Julian date taken off
Starting from 1/1/95 (year,day)
95400
95500
95600
-------
70000 -
60000
50000 -
& 40000
E
~5b
a
o
c
U 30000
20000
10000
0 -
o
~
Phenanthrene - Vapor + Particle
~ Phenanthrene - Vapor + Particle
~
~ o
~
o ~
95000
95100
95200 95300 95400
Julian date taken off
Starting from 1/1/95 (year,day)
95500
95600
-------
350
300
250
Benzo(a)pyrene - Vapor + Particle
o Benzo(a)pyrene - Vapor + Particle
s
"tob
&
200
U 150
100 -
50
~
0 ^ ~
P o <>
*
~
~
~ ~
~
~ *
95000
95100
95200 95300
Julian date taken off
Starting from 1/1/95 (year,day)
95400
95500
95600
-------
15000 n
TRIADS Average Air PAH Profile (+/- Std Dev)
12000
9000
£
eJ)
&
o
c
o
U
6000 -
3000
0 I
it
1
1 1 W1 Hi
s. s
iill
c
X z
— CO
cZ EZ C £
— rj p-3 .s
u L) u i.
£ s £- e.
q S ¦
c -5-
& & E*
V V
0 c
C £S 0
O)
-------
250 -
Precipitation Volume Collected
200
150 -
J
—'
T3
O
O
41
o
U
0
£
1 100
>
50 -1
~
100
_, ®—0 ,
200 300 400
Julian date from 1/1/95 (year,day)
Q
500
600
-------
8000
7000
Total BHC's
0
~ Total BHC's
6000 -
5000 -
/ N
i
& 4000 -
o
c
o
O
3000 -
2000 -
1000 -| ~ ~ ~ ^
«> o
-------
Rain Volume Collected
-------
50000
45000
40000
35000
30000
00
25000 -
o
c
o
O
20000
15000
10000
5000
Measured PCB's
~ Measured PCB's
* ~ o o
*
«>
-Oj—
o o ~ ~
100
200 300 400
Julian date from 1/1/95 (year,day)
500
600
-------
Rain Volume Collected
-------
2500
Total Chlordanes
o Total Chlordanes
2000
1500
~—\
i
a
o
c
o
U
1000 -
500
$
* o 0
^ o o
~ ~
100
200 300 400
Julian date from 1/1/95 (year,day)
500
600
-------
Rain Volume Collected
-------
wJ
"Sb
u
E
o
U
400
350
300
250 -
200
150
100 -
Dieldrin
* Dieldrin
50
o 0
o o
0
100
200 300 400
Julian date from 1/1/95 (year,day)
500
600
-------
Rain Volume Collected
-------
700
Hexachlorobenzene
600
500
o Hexachlorobenzene
~
o
~
~
~ ~
o ~
100
200 300 400
Julian date from 1/1/95 (year,day)
500
600
-------
Rain Volume Collected
-------
PHENANTHRENE
~PHENANTHRENE
~
o $
~
~
0
~ o
o
100
200 300 400
Julian date from 1/1/95 (year,day)
500
600
-------
60000
50000 -
PHENANTHRENE
y = -52.094x+ 18789
R2 = 0.1113
PHEN
inear(
og- CP
40000
y = -1005.2Ln(x) + 18925
R2 = 0.0292
30000 -
20000
~
~
10000
~
0 -
50
100 150
Rain Volume Collected
200
250
-------
J
"Sb
a
o
c
o
14000
12000
10000
8000 J
<-> 6000 -
4000
2000
BENaPYRENE
*
o o
~ BENaPYRENE
0 ~ o
o 0 ~
100
200 300 400
Julian date from 1/1/95 (year,day)
500
600
-------
20000
16000 -
12000 -
m
E
"bb
3
o
c
o
U
8000 -
4000
115700
TR
ADS Average Rain PAH Pro
I
1
1
He (+/- Std Dev)
IJ
1
I
*2 K
*2
sr.
3
—S U U3
>* 7Z
S 3 £
= >; P
i E ^
53 E S5 £
i sf
£ p
z £
S 1 1
~ s s
g § I
s
5
TZ
U
C3
=3
o
Q
C:
rj
cn
u
^ a is
53 5§
s =
1 I
w &J ti:
O o
G S
=s S
CiJ T".
" = == = =
£
5
£ fc
63 S
S £
S 2;
£ =
g a c
-------
------- |