GL
00861
             REPORT ID THE U.S.  ENVIRONMENTAL PROTECTION AGENCY

                                   on


      ATMOSPHERIC DEPOSITION WORKSHOP ON ORGANIC CONTAMINANT DEPOSITION

                         TO THE  GREAT LAKES BASIN
              held at the UNIVERSITY OF MINNESOTA. MINNEAPOLIS

                          20-21 NOVEMBER. 1985
                          STEVEN J. EISENREICH
                    ENVIRONMENTAL ENGINEERING PROGRAM
                DEPARTMENT OF CIVIL AND MINERAL ENGINEERING
                         UNIVERSITY OF MINNESOTA
                       MINNEAPOLIS, MINNESOTA 55455
                            PROJECT OFFICER

                           EDWARD KLAPPBNBACH
                                U. S. EPA
                       GREAT LAKES  NATIONAL PROGRAM
                            536 S. CLARK ST.
                           CHICAGO, IL 60605

                           GRANT No: EPA/R005879-0

-------
"IT IS OFTEN NECESSARY TO HAKE A DECISION BASED




ON INFORMATION WHICH IS SUFFICIENT FOR ACTION




BUT INSUFFICIENT TO SATISFY THE INTELLECT








             Enanual Kant .1786




          " Critique of Pure Reason*

-------
                                                      PAGE

TABLE OF CONTENTS                                        1

EXECUTIVE SUMMARY                                        2

WORKSHOP OBJECTIVES                                      5

ATMOSPHERIC DEPOSITION PROCESSES                         7

     WET DEPOSITION
     DRY PARTICLE DEPOSITION
     VAPuR EXCHANGE Ai THE AIR-WATER INTERFACE

GREAl LAKES ATMOSPHERIC DEPOSITION NETWORK              18

RECOMMENDATIONS FOR A GREAT LAKES ATMOSPHERIC           20
INPUT AND SOURCES (GLAIS)  NETWORK

     GENERAL RECOMMENDATIONS
     SPECIFIC RECOMMENDATIONS

REFERENCES                                              29

APPENDICES                                              31

-------
EXECUTIVE SUMMARY






An "Atmospheric Deposition" Workshop was held at the University of




Minnesota, Minneapolis on 20-21 November, 1985 under the support and




auspices of the Great Lakes National Program Office (GLNPO)  of the




U.S. Environmental Protection Agency (EPA).  Participants at the




wortsuop included distinguished scientists knowledgeable in




atmospheric processes from the U.S. and Canada,  representatives of the




Canadian and U.o. organizations conducting monitoring studies and




staff of the GLNPO.  The objectives of the workshop were two-fold:




(1) to review the Great Lakes atmospheric deposition network (GLAD)




presently operated by EPA; and (2)  to make recommendations as to the




design and strategy for a new or modified network capable of assessing




atmospheric deposition of selected toxic chemicals to the Great lakes,




identify sources and quantify source strengths.   These objectives are




meant to assist in the successful achievement of the goals stated in




EPA's five-year stategy relative to atmospheric deposition:  (1) to




determine the portion of total loadings of critical toxic pollutants




by atmospheric deposition; (2) to recommend the extent to which




additional remedial programs and international activities are needed




to control atmospheric sources; ana (J) to provide source information




for immediate regulatory actiun.




A thorough review of the present GLAD network and accumulated data on




precipitation concentrations conclusively sno»ea that the GLAD network




as presently constituted is not adequate to provide needed data on




atmospheric inputs of toxic chemicais to the Great Lakes.  This




conclusion was basea on siting, sample collection and data analysis

-------
criterea.




A detailed discussion of the processes of wet deposition,  dry particle




deposition and vapor exchange at the air-water interface was held to




acquaint all present with the state-of-the-art in these areas.  These




discussions led to the generation of recommendations on the structure




and design of a monitoring network capable of assessing atmospheric




inputs of toxic chemicals to the Great Lakes.  The design  strategy for




the network is to combine both deposition measurement and  modeling to




assess atmospheric deposition.  It would have relatively few sites




(e.g.* 16) at which collection of wet precipitation, air particles and




meteorological data necessary for the modeling effort wouid occur.




Two to four of these sites would be designated as master sites where




calibrations, intercomparison studies and research would be carried




out.  Sites would be distributed in such a way as to estimate




"regional* deposition ( half of the sites) and "urban" deposition




since it is believed that intense episodes of localized inputs occur




in areas adjacent to population centers. If possible, monitoring




stations should be located on in-lake islands.  A complete list of




instrumentation common to all sites is proviaeu in the text but




includes wet-only precipitation samplers capable 01 isolating the




analyte  in the field, multiple Hi-Vol samplers operating to provide




directional data on atmospheric concentrations, and a suite of sensors




to measure meteorological parameters such as temperature profiles,




relative humidity, total suspended participates (TSP), and wind speed




and direction.  Additional instrumentation to be located at the master




sites are cascade impactors and any necessary to support the research

-------
component.  Toxic organic species to be monitored are polychlorinated




biphenyls (PCBs), routine organochlorine pesticides and industrial




chemicals, polycyclic aromatic hydrocarbons (PAHs), Hg, Pb and Cd.




tozaphene and generic chlorinated diozins and furans.  Analysis of




these toxic chemicals will require considerable analytical support.  It




is anticipated that this network woulu largely replace the larger GLAD




network.  The research component of this network should include a




significant effort to quantify dry deposition  and vapor exchange ox




both inorganic and organic toxic species through a combination of




sampling and modej.ing of over-late processes.   Other methods of




assessing atmospheric inputs should be continued (e.g., mass balance!




small lake studies; sediment and peat profiles).  Every attempt should




be made to integrate the D.a. network with Canadian efforts using




similar samplers and strategies, inter-calibrations and co— located




sampling sites to implement an "Integrated Atmospheric Deposition




Network".  A rigorous QA/QC system need be instituted at the outset of




the project to ensure the quality of the data.




     It is proposed that a multi-investigator study be performed in




the summer and winter of 198a to integrate deposition data generated




at the master sites with intensive over-lake measurements involving




wet deposition, dry deposition, vapor exchange and source




reconciliation.

-------
2.0  Workshop Objectives




     The Great Lakes National Program Olfice (GLNPO) of the U.S.




Environmental Protection Agency (EPA) presently operates a 36 site




atmospheric monitoring networ*. (GLAD), of which 10 sites are equipped




with bulk samples for assessing organic contaminant deposition to the




Great Lakes.




     The atmospheric deposition program for the next five years has




three objectives as stated in EPA's 5-year strategy document (1):




     (1)  to determine the portion of total loadings of critical toxic




          pollutants by atmospheric deposition;




     (2)  to recommend the extent to which additional remedial




          programs and international activities are needed to control




          atmospheric sources;




     (3)  to provide source information for immediate regulatory




          action.



     The objectives 01 the -'Atmospheric Deposition" workshop held at




the University of Minnesota on the 20 and 21 November 1985 under the




auspices ot tne liLNPO 01 EPA were:




     (1)  to review the atmospheric deposition monitoring network




          (CiLAD) operated by the GLNP;




     (2)  to make recommendations as to the design and strategy for a




          new or modified atmospheric deposition network capable of




          assessing atmospheric deposition of selected toxic organic




          contaminants to the Great Lakes, identify sources and




          quantify source strengths.

-------
     Essential to this discussion was a thorough review of the




feasibility and limitations of assessing wet and dry deposition.  To




these ends, the first part ox the workshop dealt with an overview of




the present GLAD netwwr~ by T.C. Hnrpity iDeFaul University) and D.




Gatz (Illinois State Water Survey).  Based on a study recently




completed for EPA, T.C. Hurpny outlined requirements for a new Great




Lakes atmospheric and sources (GLAIS) network.  These presentations




were followed by detailed discussions on the processes of atmospheric




deposition led by J. Pankow lOregon Graduate Center) on wet deposition




of organic species, A. Andren (University of Wisconsin) on dry




particle deposition of organic species. M. Weseley (Argonne National




Laboratory) on meteorological aspects of dry deposition, and T.




Bidleman (University of South Carolina) and S. Eisenreich (University




of Minnesota) on vapor-particle partitioning in the atmosphere and




volatilization, respectively.  The remainder of the workshop involved




generating specific recommendations for measurement and assessment of




atmospheric deposition of organic contaminants to the Great Lakes.




     Specific questions addressed throughout the workshop were:



     (1)  What is the relative importance of atmospheric deposition




          compared to other input pathways for specific contaminants




          in the Great Lakes Basin and specific Great Lakes?




     (2)  What are the important deposition processes and how coulu




          they be quantified?




     (3)  What of the thousands of organic chemicals produced and




          released to the environment need be measured and how?

-------
     (4)  What are the atmospheric sources and source strengths for




          specific organic contaminants?




     (5)  What are the characteristics of an atmospheric deposition




          ne twork




          wnich is adequate to provide needed data?




     (6)  What are the research needs to support the atmospheric




          deposition network?




3.0  Atmospheric Deposition Processes




     Atmospheric organic chemicals exist in the vapor phase and




adsorbed to suspended particles (TSP).  The processes by which trace




organic contaminants are removed from the atmosphere (wet and dry




deposition) and the quantity ultimately deposited on the water/land




surface depend on the distribution between the vapor and particle




phases.  Partitioning between the gas and aerosol phase depends on




contaminant vapor pressure, size and surface area of the aerosol,




temperature, and the organic caroon content.  The less volatile the




compound,  the higher the affinity for TSP.  Theoretical considerations




and laooratory and field measurements indicate that PCBs, DDT, Eg,  low




molecular weignt (MW) hydrocarbons,  and low Mff PAHs exist primarily in




the gas phase in "clean" airsheds, while the more chlorinated PCB




congeners, high Mff PAus and dioxins occur primarily in the particle




phase.   In "dirty* wr uruan/industrial airsheds, a greater fraction of




the total  atmospheric burden for a particular chemical will occur in




the particle pnase.  Particle size plays a significant role in the




removal efficiency of aerosol-bound contaminants.  Organic species




tend to concentrate in the snbmicron particle-size fraction (i.e.,

-------
                                  8



aerosol) and will be primarily removed by precipitation scavenging


(>70 %) while larger particles will have a greater fraction removed by


dry particle deposition.


Wet Deposition


     The mechanisms of wet removal from the atmosphere are very


different for particle-associated compounds than for gas phase


compounds.  The relative importance of these two processes depends on


the distribution of the organic compound between vapor and aerosol,


particle-size distribution and Henry's Law Constant (H).  Non-reactive


organic gases will be scavenged by rain according to H if equilibrium


between the gas and aqueous phases is achieved (2,3).  Henry's Law


constant, H, is the ratio of the compound's vapor pressure to its

solubility for low solubility compounds.  In the absence of chemical


reactions occurring in the droplet, an atmospheric gas should attain


equilibrium with a falling raindrop in about ten meters of fall.   The


position of equilibrium defined by H is a function of temperature as


it increases by about 2 for each 10°C rise in temperature (3).


     The total extent of organic compound scavenging by falling

precipitation may be given as:


      T      g            p

where WT = overall scavenging efficiency


     W  - [rain-T]
      T- [air,T]


   is the gas scavenging ezficiency



     W
    [rain,  diss]
g ~ [air,  gas]

-------
«_ is the particle scavenging efficiency
      P   [air, part.]





and d is the fraction of the total atmospheric concentration occurring


in the particle phase.


     An atmospheric gas attaining equilibrium with a falling raindrop


is scavenged from the atmosphere inversely proportional to H:





     Vr-




where R is the universal gas constant,  T= temperature, H= Henry's


Law Constant, and a = solubility coefficient.  Surface flux then


becomes:


     F  = o'J'C  = W 'J'C
      g        g    g    g

where J = rainfall intensity, and C  = concentration of organic gas in


the atmosphere.  Field determined WT values are generally larger than


W  values based on H for some organic compounds suggesting particle


scavenging by precipitation is an important flux term.  Ligocki et al.


(3) have reported gas scavenging etficiencies for a variety of


nonpolar organic compounds measured in the field in Portland, OR.


Tables 1 and 2 from their paper compares the field-determined W


values to those estimated from consideration of H and ambient


temperatures.  They obtained W  values ranging from 3 to 105, and


which were underestimated by factors of 3 to 6 using E data at 25°C.


Correcting published H values for ambient temperatures of 5 to 9°C,


equilibrium between the atmospheric gas and dissolved constituent of

-------
            Table 3. Mean dissolved rain concentrations (ngS~'±ls). mean gas phase concentrations (ngm"J± 1 s), correlations
                                          between rain and air data and gas scavenging ratios
Table  1,

(Ref.3)
Compound
Tetrachloroethene
Trichloroethene
Mesitylene
Toluene
Durcne
1,2,4-Trimeihylbenzene
Ethylbenzene
m + p-Xylene
o-Xylene
1.4-Dichlorobenzene
1,2-Dichlorobenzene
1.14-Tnchlorobenzene
C2-naphthalenes
Naphthalene
2-Methylnaphthalene
1-Methylnaphthalene
Dibenzofuran
Acenaphthene
Fluortne
Acenaphthylene
Anthracene
Methylphenanlhrenes
Dibenzothiophene
Phenanthrene
Benzp[e]pyrene
Pyrene
Fluoranthene
Benzo[b -rj + JL]fluoranthene
9-Fluorenone
Benz[a]amhracene
Chrysene
Diethylphthalate
Dioctylphthalate
9.10-Anthracenedione
a-HCH
Dibutylphthalate
Mean concentrations Correlation
Rain Air (H)
4.6±3.4
5.6 ±5.6
5.1 ±15
88±75(S):
19 ±1.3
30±I4
34 ±21
110±73
45 ±31
4.8+1.2
0.26+0.20(5)
0.25+0.17(3)
17±6
100+32
45 ±16
29+11
17 + 6
5.4 ±10
14+4
37 ±13
5.1 ±10
30+8
41 ±2.1
90±26
0.37+0.20(2)
39+13
48 ±17
1.6+1.9(2)
79 + 21
3.3 ±1.0
79±2.8
59 ±30
2.6 ±08 (6)
67Z26
H±3
46 ±10
I200±690
1500+1300
430+220
3800+2400
120+70
1300±800
1300±700
3400+2000
1300+690
120+32
5.8 ±13
3.8+0.4
91 + 35
450 ±220
200±120
96+54
19 + 5
5.5 ±1.9
9.5 ±14
28+16(4)
18 + 1.0
13 + 3
1.6+0.5
27 + 7
0.03+0.01 (5)
6.8+1.9
7.8+10
0.11+0.14(5)
7.0+ 1.7
0.28+0.07
0.45+0.07
17 r 0.4 (4)
039+0.39(5)
2.5 + 0.8
0.34 + 0.05
0.37(1)
0.97
0.81
0.82
0.97
0.73
0.33
0.75
0.68
0.79
021
0.91
-§
0.87
0.37
0.67
0.78
0.69
0.16
0.16
0.28
0.48
0.09
0.59
036
—
0.09
0.20
—
0.52
0.06
0.00
0.03
—
0.64
0.48
—
Wl (meas)
Vc-
3.6 ±1.1
3.7 ±1.3
12 + 3
22±5(5)
26±9
27 ±9
27±11
33+17
35 ±15
39+10(6)
46+13(5)
66 ±51 (3)
190±32
250±73
250±78
330±100
930+180
1000+310
1500+390
1600 ±500 (4)
1900±600
2500 ±800
2500±900
3400+740
5800(1)
S900± 1800
6300+2000
7400+1300(2)
11,000±2200
12,000+4900
18,000±6500
20.000+11000(4)
20.000 ±20.000 (4)
27,000+7000
31,000+6900
110.000(1)
• (lit)
25"C
1.0
14
4.5
3.7
1.0
4.2
3.7
3.51
4.8
8.2
12
11
661
59
48
56

250
350
72
680


1100

. 2900
2400
1300"

4300
23,000
21,000


4200
87.000
Ref.t
2
2
2
I
2
2
1
2
2
1
1
1
2
2,3
2
2

3,4
3.4
3,4
3.4


3,4

3.4
3.4
1

3.4
1
1


1
1
               •A\erage temperature during sampling.   +1 = Mabey et al. (1982), 2 = Mackay and Shiu (1981). 3 - Sonnefeld el aL
             (1983). 4 = Pearlman et al. (1984).  J Number of samples, if other than seven.   §Correlation not computed for fewer than
             four points   ' Average of the values for m-xylene and p-xylene.  T Value for 1-ethylnaphthalene.  "Average of the values
             for benzo[fc]fluoramhene and benzo[L]fluoranthene.


               Table 4 Comparison of field Wt values to temperature-ilependem a values* for polycyclic aromatic hydrocarbons
Compound
Naphthalene H',
a
Table 2. ,,ya
Ruorene W,
f r+ e f\ \ S
(Ref.3) n//,,
Phenanthrcnc W,
a

Anthracene W,
a
HVa
Fluoranthene If,
a
U/ /
•/•*
Pyrene If,
a
**y«
Benz[a]anthracene W,
9
w,/«
2/12-2/13
8C
160
190
086
1200
1200
1.0
2900
3600
0.81
1000
2300
0.43
5100
6000
0.85
5000
9800
0.51
11,000
8700
1.3
2/14-2/15
6'C
340
220
1.6
2200
1400
1.6
3400
4300
080
2100
2700
0.78
6000
7100
0.85
6200
11000
0.52
14,000
9800
1.4
Sample date and mean temperature
2/20-2/21 2/23-2/24 2/29-3/1 3/16-3/20
5C 7'C 9C 9C
240
240
10
1500
1500
1.0
3900
4700
0.84
1800
2900
0.60
6900
7600
091
6600
13,000
0.51
21000
10.000
12
290
200
1.5
1300
1300
1.0
4100
3900
1.1
2100
2500
0.84
7000
6500
1.1
7000
11,000
0.64
13.000
9300
1.4
140
170
0.81
930
1100
0.84
2100
3300
0.62
1400
2200
0.63
3500
5600
0.63
3000
8700
0.34
8000
8200
0.98
270
170
1.6
1500
1100
1.4
4000
3300
1.2
2700
2200
1.2
9900
5600
1.8
8600
8700
0.99
9400
8200
1.1
4/1 1-4/12
8"C
290
190
1.5
1500
1200
1.2
3300
3600
091
2000
2300
0.85
5400
6000
0.90
4800
9800
0.49
7300
8700
0.84
Average


1.3


1 1


0.90


0.76


1.0


0.57


1.3
       •a values were calculated from the data of May et al. (1978) and Sonnefeld el al  (1983).

-------
                                  10

precipitation was demonstrated for several PAHs.  Based on these

results, temperature-corrected W  values may be used to estimate

organic gas concentrations in the atmosphere or temperature-specific H

values.

     Precipitation scavenging of particles containing sorbed organic

species (W ) permit the calculation of surface fluxes:

     WJ'CP
where F  is the particle flux and C  is the concentration of
       P                           P
atmospheric particnlate organic compound.  Depending on particle size,

precipitation intensity and type of meterological event,  W  may equal

10* to 10 .  The higher value implies that the aerosol is readily

incorporated into cloud water and is hygroscopic.  The lower value

implies a non hygroscopic, probably carbonaceous particle that is not

readily incorporated into cloud water.  Below cloud particle

scavenging efficiencies of 10  to 10  for 0.01 to 1.0 |im particles

have been reported.  The most comprehensive study of particle

scavenging by precipitation was conducteu by Ligocki et al.  (4).

Tables 3 and 4 list W  values of 102 to 105 for a series of PAHs.
                     P
alkanes and phthalates and compares W  to W  values.  In general, W_
                                     p     g                       p
values were consistent with below-cloud and in-cloud scavenging for

PAHs, in which the compounds with higher scavenging ratios were

associated more frequently with large particles.  To adequately

predict W 's, detailed information on particle-size distribution,

organic concentrations in particle-size ranges and meteorological

parameters are needed.

-------
                 Table 3. Particle scavenging ratios for neutral organic compounds during rain events in Portland, Oregon in
                                                              1984
Table  3.

(Ref.4)
Compound
                                                        2/12-   2/14-   2/20-   2/23-   2/29-   3/16-    4/11-
                                                        2/13    2/15    2/21     2/24    3/1     3/20     4,12
                                                        frontal  weak  showers  cold   warm  showers   cold
                                                 MW   cyclone  storm           front   front            front
Polycyclic aromatic hydrocarbons and derivatives
Dibenzofuran
Fluorene
9-Fluorenone
Phenanthrene + Anthracene
Methylphenanthrenes
Fluoranthene
Pyrene
9,1 0-Anthracenedione
Benz[a]amhracene
Chrysene
7-Benz[de]anthracenone
Benzofb +j + i]fluoramhene
BenzoMpyrcne
Benzo[a]pyrene
Perylene
Benzo[0Jii]perylene
Coronene
Average of PAHs with MW <
Average of PAHs with MW >
Alkanes
168
166
180
178
192
202
202
208
228
228
230
252
252
252
252
276
300
202
202
NA»
NA
29.000
19,000
19,000
4500
2600
1700
640
1700
560
1300
1400
580
630
1300
2000
15.000
1200
4900
NA
NA
2500
NA
2400
1700
NA
410
1100
NA
430
430
180
NA
NA
NA
2900
510
23.000
34.000
NA
38,000
15,000
15.000
13.000
2600
1300
3500
1100
2700
3000
2500
2600
4200
1800
24.000
2500
NA
NA
9000
5600
3400
3500
3600
1400
400
1200
900
3200
2800
1200
920
3900
NA
5000
1800
9300
8200
17.000
17,000
11.000
12,000
10.000
3900
2300
4500
2700
4900
4500
3900
3000
5800
14.000
12.000
5000
5500
2400
1900
15.000
NA
23.000
20.000
NA
2600
3800
NA
2100
1700
NA
NA
NA
NA
11.000
2600
NA
NA
18,000
21000
NA
18.000
14.000
NA
1100
2400
NA
900
290
NA
NA
290
NA
18.000
1000
Eicosane
Heneicosane
Docosane
Tricosane
Tetracosane
Peniacosane
Hexacosane
Average of alkanes
Phthalate esters
Buiylbenzylphthalate
Bis[2-eth>lhexyl]-
phthalate
Dioctylphthalate
Average of phthalates
282
296
310
324
338
352
366


298

391
391

21.000
14.000
15.000
27.000
10,000
47.000
NA
21000

1000

9100
29,000
13.000
NA
NA
NA
3900
NA
5000
3300
4100

NA

NA
NA
NA
15.000
16.000
11000
16.000
18.000
11.000
14.000
15.000

13.000

37.000
15.000
21000
89.000
76.000
63.000
29.000
NA
26.000
NA
57,000

10.000

15.000
65.000
30.000
44.000
46.000
34,000
44.000
27.000
29,000
36,000
37,000

NA

NA
NA
NA
NA
10.000
NA
NA
NA
NA
NA
10.000

NA

NA
35.000
35,000
31000
12.000
6300
13.000
6400
21000
7000
14,000

NA

NA
25,000
25.000
  Table  4.

  (Ref.4)
                        Table 4  Mean particle scavenging ratios, mean gas scavenging ratios*, and overall scavenging ratios for
                                                        neutral organic compounds
                        Compound
                                    Mean«t   Mean Wf    Mean
             Dominant
Mean W*   scav. mech.§
a-HCH
Diethylphthalale
Dibenzofuran
Fluorene
Phenanthrene + Anthracene
9-Fluorenone
Methylphenanthrenes
Fluoranthene
Pyrene
Eicosane
9, 10-Anihracenedione
Heneicosane
Dioctylphihalate
Docosane
Chrysene
Benz[a]anthraccne
Benzo[e]pyrene
Bcnzo[a]pyrene
Benzo[b +j + L]fluoramhene
Perylene
Tricosane
Tetracosane
Peniacosane
Hexacosane
Bis[2-eth> lhex> l]phthalate
Benzo[g)ii]peryiene
Coronene
0.0
0.0
0.008
0.009
0.011
0.021
0.027
0053
0.071
0.14
0.21
0.41
0.56
0.61
0.71
075
0.97
098
098
1.0
1.0
1.0
1.0
10
1.0
1.0
1.0
NA«
NA
11.000
15,000
17,000
15.000
13,000
11,000
9300
40,000
2400
31000
36.000
27.000
2600
1300
2000
1700
2200
1800
21000
16.000
23,000
15.000
20.000
3100
5900
31,000
20,000
930
1500
3300
11.000
2500
6300
5900
NA
27,000
NA
20,000
NA
18,000
11000
5800
NA
7400
NA
NA
NA
NA
NA
NA
NA
NA
31.000 j
20.000 i
1000
1600
3500
11.000
2800
6600
6100
5600
21000
13.000
30.000
17.000
7000
4000
2100
1700
2300
1800
21000
16,000
23.000
15,000
20,000
3100
5900



























                           •Mean gas scavenging ratios from Ligocki ei al. (1985).  t$ = [aerosol]/[vapor + aerosol].  J»P
                         = H p$ + irf (I -0).  §g a gas;  p = panicle.   iNA = not available

-------
                                  11

     The total wet surface flux of organic compounds in the atmosphere
may be estimated from:
     FT= VCair= V
Drv Particle Deposition
     The dry deposition of particle-bound organic compounds onto a
receptor surface depends on the type of surface, resistance to mass
transfer in the deposition layer* particle size, and macro- and
micrometeorology.  Particles are delivered to surfaces by Brownian
diffusion (mmd < 0.3 |im) , inertial impaction-interception (nmd = 0.5-5
|im) and gravitational settling (mmd > 5 urn).  Because Brownian
diffusion increases below 0.3 urn and inertial impaction-interception
increases as particle size increase above 0.5 |im, the minimum
deposition velocity (Vd) is in the range of 0.3 to 0.5 urn.  As stated
earlier, nonpolar organic compounds in the atmosphere tend to
concentrate on the small, caroon-rich, high surface area particles.
     In the simplified case, the flux of particles to a receptor
surface i.:

      v, dry    d  p
where F  .   is the surface flux of particle-bound chemical, Vd is the
deposition velocity and C  is the organic concentration in the
particulate phase.  The particle deposition velocity, V^. may be
viewed as the result of a series of resistances in transfer from the
well-mixed atmosphere to the surface (Figure 1) .

-------
         Figure 1. Simplified Model of Dry Deposition

              (from M. Weseley, ANL)
           v  ,  deposition velocity
            d

               - -F/C
               - l/(r  + r   +  r  )
                     a    b     c
?r
            C  = concentration  at height z
            r  ,  aerodynamic  resistance
            a
rb« sublayer resistance
        / / / /  C ,  surface concentration
                 s
         C   ,  subsurface  resistance
         ss
                  ass umed  zero
     All  resistances are  for conceptual

                horizontal  lavers.

-------
                                  12
where r   t   T  are aerodynamic, sublayer and subsurface resistances,
       a   D   c
respectively and C    is the particnlate concentration at a reference
                  PI Z
height.  If r  is assumed to be zero, then the aerodynamic resistance
components r , r.  may be roughly estimated from local measurements of
u ,  mean wind speea at a specified height! and o.g, a measure of
atmospheric stability.  V. (l/(r  + r.) estimates for the Great Lakes
based on seasonal atmospheric stabilities varied with season with
values of 0.5 to 1.0 cm/sec in summer to 1.7 cm/sec in the colder
months.  H. Wesley (Argonne) and A. Andren (University of Wisconsin)
point out that the following measurements or data are essential to
adequately predict (Jmoue^") dry particle deposition:
     C(r);    concentration as a function of particle size;
     u(z);    wind velocity at a reference height;
     T -TA;   surface and atmospheric temperatures;
      S  A
     IL;      relative humidity;
      11
     p(r);    particle density as a function of particle size;
     D(r);    molecular diffusivity as a function of particle size;
     OC(TSP); organic caroon content;
     TSP;     total suspended particulate concentrations.
These parameters may be used in a variety of dry deposition models
(5,6) to quantify particle deposition.
     Particle deposition velocities have been estimated from
deposition to artificial surfaces (grass! trees; filter paper; glass
or Al plates; plates coated with hydrophilic substances such as

-------
                                  13






ethylene glycol).  For a number of nonpolar chlorinated hydrocarbons




thought to associate with fine particles,  V. values of  0.1 to 1.0




cm/sec were observed.  These values are similar to estimates for




reactive gas and aerosol deposition to soils, grass, trees,




agricultural crops and snow.




     A. Andren suggested that measurements required to assess dry




deposition needed to occur on an event basis during unstable




atmospheric periods and also on a regular periodic basis.   This




assumes that particle deposition tends to concentrate during intense,




episodic periods; however, field measurements are few.  Measurements




should be performed in a few rural areas representative of "regional*




deposition and more intensively in major source areas.  It is always




preferable to make the measurements over the lake than on the




shoreline.  Atmospheric organic contaminants should be measured with




both high volume and low volume samples.  It is advisaole to determine




particle size distributions as well as a measure of total  C  and TSP.




Dennders may be of some vaine in isolated situations.  These sampling




procedures are severely restricted by analytical detection limits.




     T. Murphy (DePaul University) has suggested that although most of




the atmospheric burden of nonpolar organic compounds may be in the gas




phase, and these contaminants are concentrated on fine particles, dry




deposition especially near sources, may be dominated by large particle




deposition.  Gravitational settling velocities for large particles are




greater than  2 cm/sec while snbmicron-sized particles may have




deposition velocities much less than  1 cm/sec.  Urban and industrial




centers bordering the Great Lakes may contrionte to regional emissions

-------
                                  14


and deposition as well as intense localizvd deposition (e.g., Chicago,

H, - Gary. IN ; Detroit, III; Hamilton, Ont.).

     The relative importance on wet versus dry deposition (R) depends

on the proportion of atmospneric organic concentration in the gas and

particle phases ((ft), and the relative efficiency with which each phase

is removed by wet and dry deposition.  In general,


            [
-------
                        Table 5.  (from A.W. Andren)
          TOTAL ATMOSPHERIC FLUXES AT CRYSTAL  LAKE
ELEMENT
AL
PB
PB-210
WET1
3,
0,
1,
6
8
58
6
0
DRY2
,9(66)
,2(20)
TOTAL SED, RATE
3
3
0, 21(12) 3
10,
1,
1,
5
0
79
11,
1,
1,
1
24
86
UNITS
/
JUG - CM"'
t
nr - rw~'
^JG CM
t
DPM - CM"'
!- YR-1
•- YR-1
'•- YR-1
       ON RAIN DEPOSITION ONLY.
UNCERTAINTY is ABOUT ± 20%  (TALBOT, 1981)
3INDICATES % DRY INPUT,
4EDGINGTON AND ROBBINS (1976) ESTIMATE  1,5  FOR  LAKE MICHIGAN
     (STA, 17),

-------
                                  15
     Fo
     KoL   *L   ffig






where F = flux in moles/nr * hr; K, ,K  = liquid and gas-phase mass




transter coefficients (m/hr); C= solute concentration in liquid phase




(moles/m3), P= solute partial pressure in atm.; T= absolute




temperature in °K, and R = gas constant.  The volatilization rate may




be controlled by resistance to mass transier in the liquid phase, gas




phase or a combination of the two.  At typical values of K,  and K  (20




cm/hi and 2000 cm/hr, respectively), resistance to mass transfer




occurs > 95% in liquid phase for H 1 4.4 x 10~3 atm m3/mole, and > 95%




in gas phase for H i 1.4 x 10~' atm nr/mole.  Considering the range ot




H values for PCB congeners, 60 to 90% of the resistance occurs in the




liquid phase at 25-C.  This implies that,  in general, slightly soluble




PCBs with H >  10"* atm a?/mole tend to volatilize from water; however




transfer direction is established by the concentration gradient.




Examples of compounds where this occurs are PCBs, many CHs including




pesticides, chlorinated benzenes and tetrauhloroethylene.




     Theoretical and experimental methods to estimate mass transfer




coefficients in the field for gas and liquid-phase controlled organic




compounds have been developed.  Mackay and Yeun (8) have suggested




equations relating environmental mass transfer coefficients based on




Schmidt number (Sc) and windspeeu at a reference height of 10 m (Ri/\:




m/sec):




     K  = 1.0 x 10~3 +46.2 x 10~3 D* Sc ~°-67

-------
          -3.0
          -3.5
log Hor H'
 (atm m3\
  mole  /
          -4.0
          -4.5
          -5.0
— * Hfield
O L. Superior
— H,Qb(51)
- H (at 10° and 25°C)(27)
- • •
0 ° « o B 8
o t g
- o
1 1 1 1 1 1 1 1 1 A
/Mean values for
' 3 to 5 Cl-substituted
biphenyls (51)
•
0
o
o
I I I I
                 8 16 18 20 28 49 52 70 74 87 91 97 101 118
                         PCB Congener  Number
     Figure 2.  Henry's Law Constants (H) and Air-Water Partition
             Coefficients (H') for PCB Congeners Measured
             in Air (vapor)  and Water (filtrate) in Lake
             Superior- 1980 (Ref. 13).
                                                                     > H(25°C)
H(10°C)

-------
                                  16
          1.0 x 10~6 + 34.1 x NT4 U* Sc^0-5 (D* > 0.3)

          1.0 x 10~6 + 144 x 10~4 D*2-2 Sc0-5  (U* < 0.3)
     U*1(euv) =6.1 +0.63 U^)0-5 UIQ


where D  = air-side, friction velocity (m/sec).  Results suggest that

environmental mass transfer coefficients will generally be  lower than

those measured in the laboratory.   Annual integrated K,  values for

liquid-phase controlled organic compounds determined from mass balance

calculations on large lake systems are  0.2 to 0.25 m/day,  or about 50

to 100% smaller than predicted from laboratory experiments:  (Table

6).



    Table 6.  K,  Values Determined from Mass Balance Calculations

  Lake                   Compound             ]L  (m/day)

 Zurich             1,4-dichlorobenzene       0.24

 Saginaw Bay,              PCBs               0.2
 Lake Huron

 Lake Superior             PCBs               0.24

 Great Lakes               PCBs               0.10

(modified from Ref. 13)


     Estimating volatilization of PCBs, for example, is complicated by

the inability to accurately determine the concentration of  the

dissolved, unassociated species in equilibrium with the atmospheric

vapor.  An example of this phenomenon is presented in Figure 2,  where

Henry's Law Constant (H) or air-water partition coefficients (H')  for

several PCB congeners are compared to apparent air-water partition

coefficients (H') determined from measurements of atmospheric PCBs in

-------
                                  17






the vapor phase and dissolved PCus in two surface water samples from




Lake Superior (13).




     The apparent H' values are lower than the range ox H values for




the 3 to 5 substituted PCS congeners at 25°C but similar to H values




at 10°C, much closer to the ambient temperature of Lake Superior.   The




temperature dependence on H for PCt>s (9) suggests that gas-phase PCBs




may be absorbed into the lake at colder temperatures and lost via




volatilization at warmer temperatures.   For bodies of water such as




Lake Superior, PQ*S absorption may be important for 9 or 10 months




of the year.




Air—Transfer Scenario!  PCBs as an Example




     Figure 3 depicts our present understanding of the relative




importance of air-water transfer processes for hydrophobic organic




cycling in large lakes.  The dominant input pathway is the scavenging




of particles from the atmosphere containing sorbed organic species,




with dry particle deposition being much less important.  Particles




thus reaching the lake equilibrate with the new aqueous environment,




and partition between the dissolved phase and other biotic and abiotic




particle phases.  PCBs in the dissolved phase equilbrate with the




atmospheric gas phase or or mixed downward.  Several important




observations can be made.  The major input is through precipitation




scavenging of particles and the major losses are through




volatilization and sedimentation.  It is conceivable that measurement




of PCS flux to the atmosphere via volatilization may be more than




counterbalanced by atmospheric inputs on particles.  Hackay and




Paterson (9)  describe this phenomenon as a dynamic, steady-state but

-------
    WET
DEPOSITION
   DRY
DEPOSITION
                                 VAPOR
                           VAPOR A
                          EXCHANGE
                             res
               t
                         PARTICLE
                                 AIR

                                 ^

                                 WATER
                                                 SEDIMENT
         D= Dissolved-phase concentrations
         P= Particle-phase concentrations
     Figure 3.  PCB Air-Water Transfer Scenario (Ref. 13)

-------
                                  18






non-equilibrium process whereby the input (particle scavenging)  is




connected to loss (volatilization) by the equilbrium partitioning of




hydrophobia species.  This model also suggests that airborne




concentrations of organic contaminants over the lakes are partially or




wholly derived from in-lake processes.  Compounds such as PCBs may




cycle between water and air with intermittent periods of intense




deposition followed by slower but prolonged volatilization.
4.0  Great Lakes Atmospheric Deposition (GLAD)  Network




     T. Murphy (DePaul University;  thoroughly reviewed the present




GLAD network as to whether it fulfilled its overall mission to




quantify atmospheric inputs of nutrients,  metals and toxic organic




contaminants to the Great Lakes.  He concluded that the GLAD network




as presently constituted is not adequate to provide needed data on




atmospheric loading rates of trace  metals and toxic chemicals.   This




conclusion was basea on evaluation of GLAD using criterea based on



siting, sample collection methodology and solute analysis.  Figure 4




shows the locations of the 36 stations comprising the GLAD network and




15 stations on the Canadian side.  Each GLAD site is equipped with




weejcly-integrating, wet/dry automatic precipitation samplers.   Until




recently, 28 sites had been equipped with three bulk colectors.  The




bulk collectors have now been discontinued such that the 36 sites




collect integrating wet-only precipitation samples for nutrient and




trace metal analysis.  There is presently no sampler dedicated to

-------
                                  19







organic monitoring studies.  Murphy applied a set of criterea




developed by EPA to sites listed in Table 7.  Siting criterea relate




to obstructions nearby, distance between samplers, orientation of




samplers, nature of ground cover, use of rain gauges and distance to




nearest local contaminating sources such as roads, etc.  More than 50




% of the sites are located in areas directly impacted by




urban/industrial sources; most are within 0.5 km of the shoreline and




many are located on roofs.  A few sites are located on islands in the




lakes, but none collect samples on or over the lake.




     D. Gatz (Illinois Water Survey) examined the chemical data




collected at GLAD sites in 1982 and compared them to data similarly




collected at regional National Atmospheric Deposition Program (NADP)




and National Trend Network iNTN) sites. The NADP sites shown in Figure




5 are located away from the lakes and localized sources,  and may be




considered representative ot regional deposition.  GLAD sites are more




closely aligned with urban areas and lake shorelines.   A comparison of




ion balances for 11 NADP sites and 36 GLAD sites for 1982




precipitation show the GLAD data routinely exhibit cation deficits




(Figure 6).  This behavior suggests problemns in quality assurance and




control at the analytical laboratory.




     Spatial distributions portrayed using median concentration




isopleths for sulphate, nitrate, pH, ammonium and calcium showed close




agreement between NADP concentrations and NADP + GLAD concentrations




in similar areas.  The GLAD sites added spatial resolution and




coverage in urban areas.  Cumulative frequency distributions for the




same ions at pairea sites in GLAD and NADP consistently showed

-------
Figure 5.  NADP/GLAD Site Map            A — NADP
       (D. Gatz, III. State Water Survey)     p,	QLAD

-------
6
llJ
O
u
u
or
Ld
CL
      COMPARISON   OF  ION  BALANCES,   1982
     4O
     35 -
     30 -
     25 -
     20 -
                    1 1 NADP SITES AND 37 GLAD SITES
     15 -
     10 -
      -100  -SO  -6O  -4O  -2O
                                O
2O
                  1OO*(SUM A - SUM C)/(SUM A -t- SUM C)
             NADP (448 Sa)           A   GLAD (367 Sa)
                                                         1OO
         Figure 6. Comparison of Ion Balances, 1982 in the

                GLAD/NADP Networks

                (D. Gatz, III. State Water Survey)

-------
            WET COLLECTOR SITES
              X -
              D =
LAKE SUPERIOR
      QQ
Figure 4. Wet Collector Sites in U.S. GLAD and Canadian Networks.

-------
                                  20






differences between the site populations (Figure 7)  for sulphate and




calcium median concentrations.  These data show that major element




concentrations and deposition generated in the NADP network adequately




describe regional inputs.   GLAD sites may more properly




describle urban inputs.




     0. Gatz concluded that GLAD network laboratory measurements of pH




are biased low, major ion concentrations are similar but generally




higher that NADP/NTN values, and wet deposition demonstrates the




influence of local urban loading effects.  The GLAD network adds




resolution to the other network sites located in the Great Lakes




vicinity.









5.0  Recommendations for a Great Lakes Atmospheric Inputs and Sources




(GLAIS) Network.




     5.1  General Recommendations




The most important recommendation is that the relevant U.S. and




Canadian agencies cooperate in the formation of an "Integrated




Atmospheric Deposition Network".  The relative importance of the




atmospheric deposition pathway compared to other input pathways is a




matter of conjecture.  In the Upper Lakes, especially Lake Superior,




inputs of PCBs, DDT, JfAHs, and chlorinated diozins and fuxans are




derived primarily from atmospheric deposition (7, 11-13).  In the




Lower Lakes, atmospheric inputs may dominate for some compounds but




riverine sources dominate in most cases.  For example, atmospheric Pb




input to Lake Michigan is accounted for by accumulation in bottom




sediments.  In contrast,  PCBs are about equally loaded from

-------
                 Table 7A. Compliance of GLAD Sites to Criteria
                 Key to Column Headings In Table
                 (from  T. Murphy,  DePaul University)
Rf = Are collectors on a roof?
Rg * Does the site have a recording rain gauge?
Sp • Is the site located at a Waste Water Treatment Plant (POTW)?
Tr • Are trees obstructing the site (angle >45* above horizon)?
Rd - Is there a frequently used road within 100m of  the site?
Co - Are the collectors located too close to one another?
So « Is a local source adversly affecting the site?
Ob • Are objects obstructing the site (angle >45* above horizon)?

     Siting criteria from  EPA Quality Assurance Manual for
           Precipitation Measurement Systems (1985).

\ - Obstructions subtend an angle less than 30* with the horizon.
2 = The distance between samplers or rain gages is >2m.
3 « Routine air, ground or water traffic is more than 100m distant.
4 = There are no overhead wires that affect the collectors.
5 = Open storage of agricultural products, fuels or other foreign materials
    are not be within  100m of  the collectors.
6 = The ground surface is grass or gravel, and is firm.
7 = wet/dry collectors are oriented parallel to the prevailing winds with
    the wet bucket upwind.
8 - The rain gage is oriented parallel both to the collector and to the
    prevailing wind
Cl = Station Classification

     Class I: Station satisfies criteria 1-8;
             Instrumentation Includes: automatic precipitation collector
             recording rain gauge; pH and Cond. meters; Wind speed and dir
             sensors; 502 and NOX analyzers.

     Class  II: Station satisfies criteria 1-8;
             Instrumentation includes: automatic precipitation collector
             recording rain gauge; pM and Cond. meters.

     Class  III: Station satisfies criteria 1-8;
             Instrumentation Includes: automatic precipitation collector
             non-recording rain gauge; pH and Cond. meters.

     Class  IV: Station does not satisfy all 8 criteria;
             Instrumentation Identical to Class I station.

     Class V: Station does not satisfy all 8 criteria;
             instrumentation identical to Class n station.

-------
5ft   GLtt
      Table..7B;  Compliance  of  GLAD Sites to Criteria
                   (from T. Murphy,  DePaul University)
Name        Hi  RG   SP  Tr   Rd   Co   So  Ob  1  2  3  4  5  6  7  8
                                                                                     C1

288a
?a
Da
291 a

3a
«.,5a

7a
8a
299a
la
2a
303a
"4a
5a

307a
8a
9a
310a
" la
a
jj ja
°'4a
5a
_.6a
3l7a
8a
9a
320a
-"•la
'2a
j^!3a
•5?4a
!5a
lO
502
I3
14
308

OP
QQ
QR
AQ
QU
QT
QG
uV
QL
QJ
DQ
QU
QF
BQ
CQ
QZ
QD
QB
RQ
QA
PQ
KO-
TO
LQ
MQ
NO
JO
IQ
HO
GO
FQ
SO
QE
QM
ON
00

COM
CON
CON
CON

Evanston
Jardine Plant
So. Water Plant
Bay City
Beaver Island
Benton Harbor
Eagle Harbor
Emp i re
Escanaba
Grand Mara is
Mt. Clemens
Muskegon
Ontonagon
Port Austin
Port San i lac
Tawas Point
Duluth
Gooseberry Fls
Gull Lake
Houland
Cape Vincent
Silver Creek
Sodus Point
Grand Island
Olcott
Rochester
Conneaut
Fairport Harbor
Lorain
Put-In-Bay
Toledo
Erie
Cornucopia
Green Bay
Mam towoc
Milwaukee
Hammond Bay
Rossport
Uawa
Sault Ste Mane
Sarnia

Y
Y
Y
Y
N
Y
N
N
Y
N
Y
Y
N
Y
N
N
Y
N
N
N
N
N
N
N
Y
Y
Y
Y
Y
N
N
Y
N
N
Y
Y
N
N
N
N
N


y
y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
N
N
Y
N
N
N
N
N
N
y
N
N
Y
N
Y
Y
N
Y
Y

N
N
Y
N




N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
Y
N
Y
Y
Y
N
N
N
N
Y
N
N
N
N
N
N
N
N
N
N




N
N
N
Y
N
N
N
N
N
Y
N
Y
Y
N
N
Y
N
Y
Y
N
N
Y
N
N
N
Y
Y
N
N
N
N
N
N
N
N
N
N
Y




N
N

Y
N
Y
Y
N
N
Y
Y
N
Y
N
N
N
N
Y
Y
Y
N
Y
N
N
N
N
Y
Y
N
N
N
Y
Y
N
Y
Y
Y
Y




Y
N

Y
Y
N
Y
N
Y
Y
N
Y
Y
Y
Y
N
Y
N
N
Y
N
N
N
Y
N
N
Y
Y
N
N
N
N
N

N
N
N
N




N
N

N
N
Y
N
N
N
Y
N
N
N
N
N
N
N
N
N
N
Y
Y
Y
N
N
N
N
Y
N
N
N
N
N
N
N
N
Y
?




N
N

N
N
N
Y
Y
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N




Y
Y

N
Y
Y
N
Y
Y
N
Y
N
N
Y
Y
Y
Y
N
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y




N
Y

Y
N
Y
N
N
N
N
Y
N
N
Y
Y
Y
N
Y
Y
N
N
Y
Y
N
Y
Y
N
N
Y
Y
Y
Y
Y

Y
Y
Y
Y




Y
Y

N
Y
N
N
Y
Y
N
N
Y
N
Y
Y
Y
Y
N
N
N
Y
N
Y
Y
Y
Y
Y
N
Y
Y
Y
Y
r
Y
N
N
N
N




Y
Y

Y
Y
Y
Y
N
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y




Y
Y

Y
Y
Y
Y
Y
Y
N
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
N
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y




Y
Y

Y
Y
Y
N
Y
Y
N
Y
Y
Y
Y
Y
Y
Y
Y
Y
N
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y




Y
Y

y
Y
N
N
Y
Y
Y
Y
Y
Y
Y
Y
Y
N
Y
Y
Y
N
Y
Y
Y
Y
Y
N
N
N
Y
Y
N
Y

Y
Y
Y
N
"
MA 1 1
I^H J 1
MA 1 T
ran l i
Y V.
Y II
Y ! 1
T i *
N V
Y y
N y
N y
Y y
Y y
N y
Y y
N y
N y
N V
NA
NA
N U
NA
NA
NA
NA
NA
NA
N V
NA
NA
Y y
NA
N y
Y II
NA
Y 11
Y I
NH
NA
NA
Y y
NA
      CDN   Niagara ot Lake  N   N   N   Y
                             N
N
NNYYYYYNA
                                            C-3

-------
UJ
D
C)
UJ
UJ
O
tti
UJ
0.
       CUMULATIVE  FREQUENCY  DISTRIBUTION
                             SULFATE. 1982
    1OO
90 -


30 -


70


60


5O


40


3O


20 -


1O -


 O -
                                                        A
             GLAD SITE QV
                      —i—    —r~          ~r~
                             4            6

                     SULFATE CONG (mg/L)
                              n  NADP 3ITE Ml—Otl
                                                            s
           Figure 7. Cumulative Frequency Distributions for
                  GLAD and NADP Networks:
                    A.  Sulfate;     B.  Calcium
                  (D. Gatz, III. State Water Survey)

-------
_
LJ

O
UJ
O
LI
     CUMULATIVE  FREQUENCY  DISTRIBUTION
    100
     O
                         CALCIUM, 1982
      O
        D   GLAD SITE QV
                      CALCIUM CONC (mg/L)
                             A  NADP SITE MI-O3

-------
                                  21






atmospheric and riverine sources (12).  The chlorinated hydrocarbon




burden of Lake Ontario and Lake Erie sediments can only be explained




by dominating inputs from the Niagara River and Detroit River,




respectively.  Urban and industrial centers bordering the Great Lakes




may also support intense, localized atmospheric inputs to the




nearshore area ( 30 km) of unknown magnitude.   Thus,  atmospheric




deposition of organic contaminants to the Great Lakes is impacted by




regional deposition of pollutants transported some distance from their




source, and localized deposition of pollutants transported short




distances from sources.  To complicate the issue,  regional deposition




is impacted by both regional sources and sources outside of the Great




Lakes Basin.  The GLAIS network must take into account these differing




deposition patterns.




     The processes controlling atmospheric inputs of specific organic




contaminants are wet depositon (scavenging of gases and particles by




rain and snow), dry particle deposition and vapor-exchange at the air-



water interface.  The relative importance of each process depends on




the distribution of the chemical between the gas and particle phase




and the efficiency of removal for each phase.   In general, chemical-




physical properties of the organic contaminant (e.g., H; Sol; Pv; Kow)




may be combined with mass transfer parameters (J;  Uz', Vd; RH) and




atmospheric concentrations to estimate deposition using simple (rain)




or complex (particle or gas dry deposition) models.  The GLAIS network




should include the measurement of these parameters if it is to be




successful.

-------
                                  22






     The organic contaminants which have been frequently measured in




the Great Lakes ecosystem include the PCBs, chlorinated pesticides and




chlorinated compounds whicft are by-products of industrial activities




(e.g., HCB; octochlorostyrene).  In addition,  some measurements of




PAHs and chlorinated diozins and furans have been reported (R.Bites,




Indiana University, personal comm.).  The new GLAIS network should




monitor a suite of frequently encountered toxic chemicals as well as




search for new chemicals introduced into the environment, especially




via emerging technologies such as municipal waste incineration.




     The new GLAIS network must incorporate a rigorous sampling




protocol and QA/QC program to ensure the validity of the data.   This




protocol must descrioe the development status for each contaminant to




be considered in terms of its concentration, precision,  accuracy and




validity (i.e., compliance with assumptions made in analytical




methodology).  Care should be taken to provide analytical detection




limits corresponding to expected concentrations.  U.S. and Canadian




efforts should be closely coordinated in terms of goals, and sampling




and analytical protocols.




     As conceptualized by the workshop participants, the GLAIS network




would encompass about 6 to 8 sampling sites per lake for "routine"




deposition monitoring (2 to 4 in U.S. and Canada each), and 1 or 2




•master* stations per lake which incorporate research and cooperative




studies (1 in Canada; 1 in U.S.).  The sites should be evenly




distributed between U.S. and Canadian territory (except for Lake




Michigan).  Perhaps 4 of the 8 routine stations should be located to




assess regional inputs and the remaining 4 to assess more localized

-------
                                  23



urban inputs.  The sites should be located, where posssible, on in-

          t
lake islands  ( e.g.. Beaver Island; Isle Roy ale) and upwind and


downwind relative to the prevailing wind direction.  Every effort


should be made to incorporate over-lake sampling and a strong research


component to  the new networ*.  The routine stations would be equipped


with sufficient wet samplers, air particle samplers and equipment


necessary to  collect meteorological data required for modeling


atmospheric deposition.


     Alternate methods of estimating total and atmospheric inputs of


organic contaminants to the lakes should be continued.   These include


snow-coring, where possible, to obtain integrated winter loading


rates,  obtaining dated lace sediment and peat cores and mass balance

studies.


5.2  Specific Recommendations.


     5.2.1  Sampling Sites


          *  Routine Monitoring Sites


              *  6 to 8 per lake (divided between U.S.  and Canada)


              *  3 to 4 per lake in dominant  urban areas


              *  4 to 5 in rural areas differing in land-use


              *  In-lake islands and on-lake  sites are


                 preferable to  shoreline sites.

          *  Master  Sites


              *  1 or  2 per lake (divided between U.S.  and Canada)

              *  Rural  and urban area(s)


              *  Co-located with Canadian sites.


     5.2.2   Measurement at Routine  Monitoring  Sites

-------
                             24






     *  Wet Deposition




         *  (1) Wet-only integrating samplers




         *  2 week sampling period




         *  In-situ separation of particulate and idssolved




            organic species




         *  In-situ isolation of dissolved species




         *  Collector surface area > 0.2 m2




         *  Samplers suitable for all-weather operation




     *  Dry Deposition




         *  (2) Hi-Volume air samplers equipped with filter




            and backup adsorbent operated directionally.




         *  Sampling frequency of 24 hours every 3 days




         *  (1) Low-volume cascade impactor




     *  Climatological/Meteorological Parameters




         *  Continuous recording of rain intensity,  temperature,




            wind direction and velocity, and relative humidity.




5.2.3  Measurement at Master Sites




     *  Wet Deposition




         *  (2) Wet-only integrating precipitating samplers




         *  (1) Event precipitation sampler




         *  In-situ separation of particulate and dissolved




            species




         *  In situ isolation of dissolved species




         *  Collector surface area > 0.2 m2




         *  All-weather operation




     *  Dry Deposition

-------
                             25






         *  (3) Hi-volume ait samplers equipped with filter and




            backup adsorbent




         *  Sampling frequency - every 24 hours




         *  Hi-Vols operated in concurrence with wind vectors




            (directional sampling)




         *  (1) Low flow cascade impactor with backup adsorbent




         *  (1) Hi-Vol dedicated to TSP and OC measurement




     *  Climatological/Meteorological Parameters




         *  Same as for routine monitoring sites




         *  Meteorological tower for measurement of atmospheric




            gradients (e.g., Tsurface - Tair; etc.).








5.2.4  Chemical Measurements




     *  Routine




         *  PCBs (total; congeners)



         *  Chlorinated pesticides and other common QICs.



         *  PAHs




         *  Pb, Cd




     *  Resarch Component (Master Site)




         *  Tozaphene




         *  Chlorinated dio&ins and furans




         *  Hg



     *  Other




         *  TSP; OC (TSP); density and mmd of size-differentiated



            particles.




5.2.5  Quality Control/Quality Assurance

-------
       26






sampling protocols




loalytical  protocols




>lers  at U.S.  and Canadian sites




 sample replication;  exchange between




joratories.




inalytical  detection  limits.




itinning) Deposition  Methods




Itudies




>r vinter burdens




in peat cores  (atmosphere only)




I/or remote island  studies




 iation




mpling




 ds of specific  sources




 lysis




 onment




 earch Component to Monitoring  Study








 tmospheric inputs  of organic  contaminants




 at  the  air-water  interface may  result in




 rganic gases  or volatilization of  organic




   Research needs  are:




 astants  (h) for organic contaminants




 pendence of H




 dodology to distinguisH vapor  and




 ecies in the  atmosphere, and  dissolved

-------
                             27







        and non-dissolved species in the water




      *  Development/implementation of time-dependent model




        describing air-water interactions




      *  Methodology to estimate environment-dependent mass




        transfer coefficients




5.3.2  Detailed study to establish importance of urban vs. nonb-




       point.  intensive, localized deposition




5.3.3  Development/implementation of of dry particle deposition




       models




5.3.4  Development/implementation of field measurements over lake




       and at the master site describing dry particle deposition




       models.




5.3.5  Identification of non-routine chemical substances in




       atmospheric deposition offering potentially adverse




       impacts  on ecosystem health.




5.3.6  Implementation of research to "calibrate" dry particle




       deposition to artificial surfaces.




5.3.7  Conduct  research to quantify  the relative importance




       of atmospheric versus riverine contaminant inputs.




5.3.8  A major  field experiment should be  conducted in the summer




       and winter of 1988 to integrate atmospheric data generated




       at the rural and urban master sites with those generated




       concurrently from an intensive over-lake study involving a




       meteorological  tower and surface ship.  The objectives of




       the multi-investigator study  axe to quantify the absolute




       atmospheric inputs,  the relative importance of the  various

-------
                      28






input processes, reconcile sources,and to verify




atmospheric deposition models.   IVo potential locations




are southern Lake Michigan near Chicago and western




Ontario near Toronto.

-------
                                  29
REFERENCES









1.  U.S. Environmental protection Agency, "Five Year Program  Strategy




for Great Lakes National Program Otfice, 1986-1990*. EPA-905/9-85-002:




Chicago, IL; p. 41-45, 1985.




2.  SI inn, W.G.N.; Basse, L. J Hicks, B.B.; Hogan, A.W.; Lai,  D.; Liss,




P.S.; Hunnich. E.G.; Sehmel, G.A.; Vittori, 0. Some aspects of the




transfer of atmospheric tarce constituents past the air-sea interface.




Atmos. Environ. 19'/8, ±2, 2055-2087.




3.  Ligocki, H.P.; Leuenberger, c.; Pankow, J.F. Trace organic




compounds in rain - II. Gas scavenging of neutral organic compounds.




Atmos. Environ., 1985, 19. 1609-1617.




4.  Ligocki, M.P.; Leuenberger, C.; Pankow, J.F. Trace organic




compounds in rain - III.  Particle scavenging of neutral organic




compounds. Atmos. Environ.,1985, 19. 1619-1626.




5.  Andren, A.W. "Processes determining the flux of PCBs across the




air/water interfaces'*, in Physical Behavior of PCBs in the Grea Lakes.




D. Maokay, S. Paterson. S.J. Eisenreich, M. Simmons (Eds.).




Ann Arbor Science Publishers: Ann Arbor, HI. 127-140.




6.  Hackay, D.; Yuen, T.K., "Transfer rates of gaseous polltutants




between the atmosphere and natural waters", in Atmospheric Pollutants




in Natural Waters. S.J. Eisenreich (Ed.), Ann Arbor Science




Publishers: Ann Arbor, MI, p. 55-65.

-------
                                  30






7.  Eisenreich, S.J.; Looney, B.B.; Thornton, J.D., Airborne organic




contaminants in the Great lakes ecosystem. Environ. Sci. Tech. 1981,




11, 30-36.




8.  Hackay, D.; Yuen,A.T.K.. Environ. Sci. Tech., 1983, H, 211.




9.  Burkhard, L.P.: Armstrong. D.E.; Andren, A.W., Environ. Sci. Tech.




1985, 1£, 590-96.




10.  Hackay, D.; Paterson, S. A model describing  the rates of transfer




processes of organic chemicals between atmosphere and water.




submitted to Environ. Sci. Tech., 1985.




11.  Cznczwa, J.H.; HcVeety, B.D.; Hites, R.A., Science, 1984. 226.




568-6*.




12. Swackhamer, D.L. Ph.D. Thesis, University of Wisconsin, Madison,




1985, 258 p.




13.  Eisenreich, S.J., The Chemical Limnology of Hydrophobic Organic




Compounds: PCBs in Lake Superior, In Chemistry of Aquatic Pollutants.




R.A. Hites and S.J. Kisenreich, Eds., Advances in Chemistry Series,




American Chemical Society: Washington. D.C., 1987.

-------
                  "ATMOSPHERIC DEPOSITION WORKSHOP"

                         November 20-21, 1985

                 University of Minnesota, Minneapolis
Names and Addresses of Participants:

Steven J. Eisenreich, Workshop Chair
Environmental Engineering Program
Department of Civil and Mineral Engineering
122 CivMinEng Bldg.
University of Minnesota
Minneapolis, MN 55455
(612) 625-3082

Thomas J. Murphy
Department of Chemistry
DePaul University
25 E. Jackson Blvd.
Chicago, IL 60604
(312) 321-8191

Anders W. Andren
Water Chemistry Program
University of Wisconsin
660 N. Park St.
Madison, WI 53706
(608) 262-2470

James Pankow
Department of Chemical, Biological and Environmental Sciences
Environmental Science Program
Oregon Graduate Center
Beaverton, OR 97006
(503) 645-2111

William  Strachan
National Water Research Institute
Canada Centre for Inland Waters
P.O. Box 5050
Burlington Ontario L7R 4A6
CANADA

Donald Mackay
Department of Chemical Engineering and Applied Chemistry
University of Toronto
Toronto, Ontario M5S 1A4
CANADA

-------
Terry Bidleman
Department of Chenmistry
University of South Carolina
Columbia, SC 29208
(803) 777-4239

Donald Gatz
Atmospheric Chemistry Section
Illinois State Water Survey Division
2204 Griffith Drive
Champaign, 1L 61820
(217) 333-2210

Ronald A. Hites
School of Public and Environmental Affairs
400 E. Seventh St.
Indiana University
Bloomington, IN 47405
(812) 335-3277

Marvin L. Weseley
Atmospheric Physics Division
Building 181
Argonne National Laboratory
9700 S. Cass Ave.
Argonne, IL 60439
(312) 972-5827

Peter Wise, Director
Great Lakes National Program Office (GLNPO)
U.S. Environmental Protection Agency
536 S. Clark St.
Chicago, IL 60605

Ed Klappenbach
GLNPO
U.S. Environmental Protection Agency
536 S. Clark St.
Chicago, IL 60605
(312) 353-1378

Tim Method
Air Management Division
U.S. Environmental Protection Agency
230 S. Dearborn
Chicago, IL 60605
(312) 886-6065

-------
Eva Voldner
Atmospheric Environment Service
Environment Canada
4905 Duflerin St.
Downsview, Ontario M3H 5T4
(416) 667-4788

C.H. Chan
Canada Centre for Inland Waters
P.O. Box 5050
Burlington, Ontario L7R 4A6
CANADA
(416) 637-4641

Maurice E.B. Owens
Chief of Program Integration and Evaluation Branch
WH 586
Analysis and Evaluation Division
U.S. Environmental Protection Agency
401 M St. S.W.
Washington, D.C. 20460
(202) 965-1634

Maris Lucis
Head, Special Studes Branch
Air Resources Branch
880 Bay St., 4th Floor
Toronto, Ontario M5S 1Z8
CANADA
(416) 965-1634

Mary Swanson, Thomas Franz and Paul D. Capel
Graduate Students and Rapateurs
Environmental Engineering Program
Department of Civil and Mineral Engineering
University of Minnesota
Minneapolis, MN 55455

-------