modeling the atmospheric deposition of mercury to lake champlain (from anthropogenic sources in the...
TRANSCRIPT
Modeling the Atmospheric Depositionof Mercury to Lake Champlain
(from Anthropogenic Sources in the U.S. and Canada)
Dr. Mark CohenNOAA Air Resources Laboratory
Silver Spring, Maryland
Presentation at the Workshop onCoordination of Atmospheric Deposition Research
in the Lake Champlain BasinJune 5-6, 2003
Bishop Booth Conference CenterBurlington, Vermont
Key questions regarding atmospheric deposition:
1. How much is being deposited in each Lake?
2. How important is direct deposition to a given lake relative to indirect loading via deposition to the lake’s watershed?
3. How important is atmospheric deposition relative to other loading pathways (e.g., direct discharge to the Lake or its tributaries)
4. What is the relative importance of the contributionsfrom local, regional, national, continental, and global sources?
5. What is the relative importance of contributions from different types of sources, e.g, coal fired utilities, incinerators, natural emissions, etc.?
We need to know all these things to efficiently directaction to reduce the contamination levels in a given lake.
Ambient atmospheric
concentrations measurements
Lake water measurements
Ambient atmospheric
wet deposition measurements
Receptor Modeling (e.g., Back-Trajectory)
Forward atmospheric fate
and transport modeling from a comprehensive
inventory
Tributary and runoff water
measurements
Modeling the chemo-dynamics of the
pollutant in the lake
Data analysis (trends,
correlations, etc)
Mass Balance Ecosystem
Models
Sediment, Biota, and other Ecosystem Measurements
Meteorological measurements and modeling
All of these analytical
approaches are needed –
and must be used in coordination -- to understand Hg
in Lake Champlain
enough to be able to fix problems
Three “forms” of atmospheric mercuryElemental Mercury: Hg(0)
• ~ 95% of total Hg in atmosphere• not very water soluble• long atmospheric lifetime (~ 0.5 - 1 yr); globally distributed
Reactive Gaseous Mercury (“RGM”)• a few percent of total Hg in atmosphere• oxidized mercury: Hg(II)• HgCl2, others species?• somewhat operationally defined by measurement method• very water soluble• short atmospheric lifetime (~ 1 week or less);• more local and regional effects
Particulate Mercury (Hg(p)• a few percent of total Hg in atmosphere• not pure particles of mercury…
(Hg compounds associated with atmospheric particulate)• species largely unknown (in some cases, may be HgO?)• moderate atmospheric lifetime (perhaps 1~ 2 weeks)• local and regional effects• bioavailability?
CLOUD DROPLET
cloud
PrimaryAnthropogenicEmissions
Elemental Mercury: Hg(0)
Reactive Gaseous Mercury: RGMParticulate Mercury: Hg(p)
Atmospheric Fate Processes for Hg
Dry and Wet Deposition
Hg(0) oxidized to dissolvedRGM by O3, HOCl, OCl-
Hg(II) reduced to Hg(0) by SO2
Re-emission of natural AND previously depositedanthropogenic mercury
Adsorption/desorptionof Hg(II) to/from soot
Halogen-mediated oxidationon the surface of ice crystals
Hg(p)
“DRY” (low RH)ATMOSPHERE:
Hg(0) oxidized to RGMby O3, H202, Cl2, OH, HCl
Modelevaluation
Can’t reliably estimate the amount of deposition or source-receptor relationships using monitoring alone…
Modeling can potentially give you these answers, but cannot be done credibly without using monitoring
to ground-truth the results
We are generally not actually interested in the concentration or deposition at a single monitoring site…
We are just using the few monitoring sites that we might have to give us a clue as to what the total impact might be…
We are interested in the deposition to an entire water body, or to a particular ecosystem
Overall Methodology
Start with atmospheric mercury emissions inventory
Perform atmospheric fate and transport modeling of these emissions (using a modified version of NOAA’s HSYPLIT model)
Keep track of source-receptor information during the modeling
Evaluate the modeling by comparison of the predictions against ambient monitoring data
If model is performing satisfactorily, report source-receptor results from the simulations
(Similar to earlier work with dioxin and atrazine)
1995 Global Hg Emissions Inventory Josef Pacyna,NILU, Norway (2001)
0 - 0.0010.001 - 0.010.01 - 0.10.1 - 11 - 1010 - 100100 - 10001000 - 14000Not Estimated
2
Mercury Emissions(grams/km -yr)
500 0 500 1000 Kilometers
500 0 500 1000 Miles
Geographic Distribution of Estimated Anthropogenic Mercury Emissions in the U.S. and Canada for 1995-1996
coal combustion (electric generation)
coal combustion (commercial/industrial)
natural gas combustion
oil combustion (non-mobile)
mobile sources
other fuel combustion
medical waste incineration
municipal waste incineration
hazardous waste incineration
industrial waste incineration
other waste incineration
chloralkali
other chemical manufacturing
metallurgical processes
mining
cement/concrete
pulp/paper
lamp manufacturing and breakage
other manufacturing
other
0.00 0.05 0.10 0.15 0.20
grams Hg emitted per person per year
U.S.
Canada
Annual Per Capita Mercury Emissions from U.S. and Canadian Anthropogenic Sources, 1995-1996
coal combustion (electric generation)
coal combustion (commercial/industrial)
natural gas combustion
oil combustion (non-mobile)
mobile sources
other fuel combustion
medical waste incineration
municipal waste incineration
hazardous waste incineration
industrial waste incineration
other waste incineration
chloralkali
other chemical manufacturing
metallurgical processes
mining
cement/concrete
pulp/paper
lamp manufacturing and breakage
other manufacturing
other
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
fraction of emissions from source type
Hg(0)
Hg(II)
Hg(p)
Speciation Profile of Mercury Emissions from U.S. and Canadian Anthropogenic Sources, 1995-1996
coal combustion (electric generation)
coal combustion (commercial/industrial)
natural gas combustion
oil combustion (non-mobile)
mobile sources
other fuel combustion
medical waste incineration
municipal waste incineration
hazardous waste incineration
industrial waste incineration
other waste incineration
chloralkali
other chemical manufacturing
metallurgical processes
mining
cement/concrete
pulp/paper
lamp manufacturing and breakage
other manufacturing
other
0 10 20 30 40 50
metric tons emitted per year
Hg(0)
Hg(II)
Hg(p)
Speciated Annual Mercury Emissions from U.S. and Canadian Anthropogenic Sources During 1995-1996
1000 0 1000 2000 Kilometers
(ug/km2-yr deposition)/(g/yr emissions)
0.004 - 0.0070.007 - 0.010.01 - 0.020.02 - 0.040.04 - 0.070.07 - 0.10.1 - 0.20.2 - 0.40.4 - 0.70.7 - 11 - 22 - 44 - 77 - 11
0.002 - 0.004
at any given location,the transfer coefficientis defined as the amountthat would be depositedin the given receptor(in this case, Lake Champlain) if there were emissionsat that location.
Transfer Coefficients
• refer to hypothetical emissions;are independent of actual emissions
• can be formulated with different units [total Hg deposition flux (ug/km2-yr) /
emissions (g/yr)]
• will depend on the pollutant [Hg(II)]
• will depend on the receptor[Lake Champlain]
• and the time period being modeled [entire year 1996]
Hg(0) Hg(p) Hg(II)
0.004 - 0.0070.007 - 0.010.01 - 0.020.02 - 0.040.04 - 0.070.07 - 0.10.1 - 0.20.2 - 0.40.4 - 0.70.7 - 11 - 22 - 44 - 77 - 11
0.002 - 0.004
1000 0 1000 Kilometers
1000 0 1000 Kilometers
(ug/km2-yr deposition)/(g/yr emissions)
Annual 1996 Transfer Coefficients for Different Forms of Mercury to Lake Champlain
0 - 0.0010.001 - 0.010.01 - 0.10.1 - 11 - 1010 - 100100 - 10001000 - 14000Not Estimated
2
Mercury Emissions(grams/km -yr)
500 0 500 1000 Kilometers
500 0 500 1000 Miles
Emissions Map
x
1000 0 1000 2000 Kilometers
(ug/km2-yr deposition)/(g/yr emissions)
0.004 - 0.0070.007 - 0.010.01 - 0.020.02 - 0.040.04 - 0.070.07 - 0.10.1 - 0.20.2 - 0.40.4 - 0.70.7 - 11 - 22 - 44 - 77 - 11
0.002 - 0.004
Transfer Coefficient Map
=500 0 500 1000 1500 2000 Kilometers
ug/km2-yr
0 - 1010 - 2020 - 4040 - 7070 - 100100 - 200200 - 400400 - 700700 - 10001000 - 20002000 - 40004000 - 70007000 - 1000010000 - 100000100000 - 10000001000000 - 3000000
Contribution Map
500 0 500 1000 1500 2000 Kilometers
ug/km2-yr
0 - 1010 - 2020 - 4040 - 7070 - 100100 - 200200 - 400400 - 700700 - 10001000 - 20002000 - 40004000 - 70007000 - 1000010000 - 100000100000 - 10000001000000 - 3000000
Geographical Distribution of Atmospheric Deposition Contributions to Lake Champlain Arising from Anthropogenic Sources in the U.S. and Canada During
1996
500 0 500 Kilometers
0 - 1010 - 2020 - 4040 - 7070 - 100100 - 200200 - 400400 - 700700 - 10001000 - 20002000 - 40004000 - 70007000 - 1000010000 - 100000100000 - 10000001000000 - 3000000
ug/km2-yr
Lake Champlain
100 0 100 Kilometers
ug/km2-yr
0 - 1010 - 2020 - 4040 - 7070 - 100100 - 200200 - 400400 - 700700 - 10001000 - 20002000 - 40004000 - 70007000 - 1000010000 - 100000100000 - 10000001000000 - 3000000
jan
febmar
aprmay
junejuly
augsept
octnov
dec
0.0
0.5
1.0
1.5
2.0
kg H
g d
epo
site
d /
mo
nth
wet dep of Hg(2)-sootwet dep of Hg(p)
wet dep of Hg(2)dry dep of Hg(p) + Hg(2)-soot
dry dep of Hg(2)
Monthly Model-Estimated Wet and Dry Deposition of Different Forms of Mercury to Lake Champlain During 1996 Arising from U.S. and Canadian Anthropogenic Emissions
Cumulative Model-Estimated Wet and Dry Depositionof Mercury to Lake Champlain During 1996
Arising from U.S. and Canadian Anthropogenic Emissions
jan feb mar apr may june july aug sept oct nov dec
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Cu
mu
lati
ve
De
po
sit
ion
(k
g)
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
dry depositionwet deposition
0-100100-200
200-400400-700
700-10001000-1500
1500-20002000-2500
>2500
distance range from Lake Champlain (kilometers)
0%
5%
10%
15%
20%
25%
30%
Hg
emis
sion
s or
dep
ositi
on
perc
ent
of t
otal
mod
eled
emissions
Percent of modeled EMISSIONS of mercury to Lake Champlain from different distance ranges away from the lake
0-100100-200
200-400400-700
700-10001000-1500
1500-20002000-2500
>2500
distance range from Lake Champlain (kilometers)
0%
5%
10%
15%
20%
25%
30%
Hg
emis
sion
s or
dep
ositi
on
perc
ent
of t
otal
mod
eled
deposition
Percent of modeled DEPOSITION of mercury to Lake Champlain from different distance ranges away from the lake
0-100100-200
200-400400-700
700-10001000-1500
1500-20002000-2500
>2500
distance range from Lake Champlain (kilometers)
0%
5%
10%
15%
20%
25%
30%
Hg
emis
sion
s or
dep
ositi
on
perc
ent
of t
otal
mod
eled
emissionsdeposition
Percent of modeled emissions and deposition of mercury to Lake Champlain from different distance ranges away from the lake
0-100100-200
200-400400-700
700-10001000-1500
1500-20002000-2500
>2500
distance range from Lake Champlain (kilometers)
0%
20%
40%
60%
80%
100%
Hg
emis
sion
s or
dep
ositi
on
cum
ulat
ive
perc
ent
of t
otal
mod
eled
cumulative emissions
cumulative deposition
Cumulative percent of modeled emissions and deposition of mercury to Lake Champlain from different distance ranges away from the lake
coal combustion (electric generation)
coal combustion (commercial/industrial)
natural gas combustion
oil combustion (non-mobile)
mobile sources
other fuel combustion
medical waste incineration
municipal waste incineration
hazardous waste incineration
industrial waste incineration
other waste incineration
chloralkali
other chemical manufacturing
metallurgical processes
mining
cement/concrete
pulp/paper
lamp manufacturing and breakage
other manufacturing
other
0 5 10 15
ug Hg deposited / (person - year)
USA CANADA
Per Capita Mercury Atmospheric Mercury DepositionContributions to Lake Champlain from U.S. and
Canadian Detailed Source Categories (1996)
USA CANADA0
10
20
30
40
50
ug
Hg
dep
osi
ted
/ (
per
son
- y
ear)
manuf / other
metals
incin
fuels
Per Capita Mercury Atmospheric Mercury DepositionContributions to Lake Champlain from U.S. andCanadian Aggregated Source Categories (1996)
1990
(a)
1995
(b
)
1995
/96/
99(c
)
1996
/99
(d)
1999
(e)
0
50
100
150
200
tons
/yea
r
pulp/paper production
cement manufacturing
hazardous waste incin
chlorine production
municipal waste incin.
medical waste incin.
coal-fired utility boilers
Reported trends in U.S. atmospheric mercury emissions 1990-1999 (selected source categories)
(a) EPA NTI Baseline (Mobley, 2003)(b) Hg Study Rpt to Congress (EPA, 1997)(c) inventory used in this analysis(d) EPA 96/99 Inventory (Ryan, 2001)(e) EPA NEI 1999 (draft) (Mobley, 2003)
0 10 20 30 40 50 60week of 1996
0
0.1
0.2
0.3
0.4
0.5
0.6
tota
l w
et d
ep o
f m
ercu
ry (
ug
/m2)
measured
modeled
Comparison of Modeled vs. Measured Wet Deposition at Underhill Center, VT during 1996
0 0.1 0.2 0.3 0.4 0.5 0.6
measured weekly totals (ug/m2)
0
0.1
0.2
0.3
0.4
0.5
0.6
mo
de
led
we
ek
ly t
ota
ls (
ug
/m2
)Comparison of Modeled vs. Measured Wet Deposition
at Underhill Center, VT during 1996
0.001 0.01 0.1 1
measured weekly totals (ug/m2)
0.001
0.01
0.1
1
mo
del
ed w
eekl
y to
tals
(u
g/m
2)
Comparison of Modeled vs. Measured Wet Depositionat Underhill Center, VT during 1996
jan feb mar apr may june july aug sep oct nov dec
0.01
0.1
1
10
1111111111
mo
nth
ly w
et d
epo
siti
on
of
Hg
(u
g/m
2)
measured
modeled
Comparison of Modeled vs. Measured Wet Deposition at Underhill Center, VT during 1996
1 7 13 19 25 31 37 43 49
week of 1996
0.01
0.1
1
10
cum
ula
tive
wet
dep
osi
tio
n (
ug
/m2)
measured
modeled
Comparison of Modeled vs. Measured Cumulative Wet Depositionat Underhill Center, VT during 1996
Some Limitations of this Modeling AnalysisUncertainties in emissions (speciation, amount, temporal variations)
Uncertainties in atmospheric chemistry of mercury
Uncertainties in simulating wet and dry deposition phenomena
Only U.S. and Canadian anthropogenic sources have been included; need to add global sources, natural sources, and anthropogenic mercury re-emitted after initially deposited
Assuming net deposition of Hg0 is zero – essentially that natural emissions and re-emitted mercury sort of balance out Hg0 deposition, so that net flux ~ 0
This is probably not true for Lake Champlain (or most lakes), as there is probably a net evasion of Hg0, as a response to the deposition of Hg(II) and Hg(p). In this modeling (to date), we have only estimated this downward flux of Hg(II) and Hg(p).
Coarse meteorological data grid (180 km)
Only direct deposition to lake surface considered; deposition to watershed and subsequent entry into the lake not yet included in modeling