modeling the atmospheric deposition of mercury to lake champlain (from anthropogenic sources in the...

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





mobile sources







chloralkali



mining

cement/concrete

pulp/paper


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





mobile sources







chloralkali



mining

cement/concrete

pulp/paper


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



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


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


0 - 1010 - 2020 - 4040 - 7070 - 100100 - 200200 - 400400 - 700700 - 10001000 - 20002000 - 40004000 - 70007000 - 1000010000 - 100000100000 - 10000001000000 - 3000000

ug/km2-yr

Lake Champlain


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


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


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


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





mobile sources







chloralkali



mining

cement/concrete

pulp/paper


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

modeling the atmospheric deposition of mercury to lake champlain (from anthropogenic sources in the...

Documents