Present-day sources and past variability of mercury in the Arctic

Present-day sources and past variability of
mercury in the Arctic atmosphere and ocean
AMAP
Jenny A. Fisher
University of Wollongong
D. J. Jacob, A. L. Soerensen, H. M. Amos,
E. D. Corbitt, D. G. Streets, Q. Wang,
R. M. Yantosca, and E. M. Sunderland
NASA Earth Observatory
ACCESS XII
25-27 July 2013
Mercury in the Arctic Ocean:
an ecological and human health concern
Methylmercury is a potent neurotoxin that bioaccumulates in marine food webs. High mercury concentra:ons have been observed in Arc:c species that make up tradi:onal northern diets.
Atmospheric transport brings mercury to the Arc:c, but atmosphere-­‐ocean cycling is uncertain.
Goal:
Understand the factors contribu4ng to elevated mercury in the Arc4c atmosphere & ocean
Jenny A. Fisher
([email protected])
AMAP 2011
ACCESS XII
The atmosphere-ocean mercury cycle - an Arctic perspective
Mid-­‐la4tudes
Arc4c
0
Hg
HgII
Snow/ice
Hg0
II
Hg
Ocean
subsurface
Jenny A. Fisher
([email protected])
ACCESS XII
The atmosphere-ocean mercury cycle - an Arctic perspective
Mid-­‐la4tudes
Arc4c
0
Hg
HgII
Snow/ice
Hg0
II
Hg
HgII
Ocean
subsurface
Jenny A. Fisher
([email protected])
ACCESS XII
The atmosphere-ocean mercury cycle - an Arctic perspective
Mid-­‐la4tudes
Arc4c
0
Hg
HgII
Snow/ice
Hg0
II
Hg
Hg0
HgII
Ocean
subsurface
Jenny A. Fisher
([email protected])
ACCESS XII
The atmosphere-ocean mercury cycle - an Arctic perspective
Mid-­‐la4tudes
Arc4c
0
Hg
HgII
Snow/ice
Hg0
II
Hg
Hg0
HgII
Ocean
subsurface
Jenny A. Fisher
([email protected])
ACCESS XII
The unique seasonal cycle of Arctic mercury
summer maximum Observa4ons: mul>-­‐year mean over three Arc>c sites (Alert, Amderma, Zeppelin)
spring minimum:
Atmospheric Mercury Deple:on Events
Can we explain the summer peak using our standard understanding of mercury cycling?
Jenny A. Fisher
([email protected])
ACCESS XII
The GEOS-Chem biogeochemical mercury model
The GEOS-­‐Chem biogeochemical mercury model:
•
Global, gridded model
•
Driven by input meteorology from an assimilated GCM
•
3-­‐D atmosphere: emissions, chemistry, transport, and deposi:on
•
2-­‐D slab ocean mixed layer: chemistry, evasion, subsurface mixing and export
•
Cryospheric processes: AMDEs, snowpack re-­‐emission, snowmelt delivery to ocean
Represents our best a priori understanding of mercury sources & processes
Jenny A. Fisher
([email protected])
ACCESS XII
Testing our current understanding:
model comparison to atmospheric observations
summer maximum Observa4ons: mul>-­‐year mean over three Arc>c sites
spring minimum:
Atmospheric Mercury Deple:on Events
Can we explain the summer peak using our standard understanding of mercury cycling?
Jenny A. Fisher
([email protected])
ACCESS XII
Testing our current understanding:
model comparison to atmospheric observations
summer maximum Observa4ons: mul>-­‐year mean over three Arc>c sites
Model: GEOS-­‐Chem at these sites in 2008
spring minimum:
Atmospheric Mercury Deple:on Events
Can we explain the summer peak using our standard understanding of mercury cycling?
Jenny A. Fisher
([email protected])
ACCESS XII
We test hypotheses using model sensitivity simulations
2.0
standard simula4on
maximum oceanic reduc4on
maximum snowpack reduc4on
Hg0, ng.m-3
1.8
1.6
1.4
1.2
1.0
J F M A M J J A S O N D
Month
Atmospheric transport, snowpack processing, ocean chemistry cannot explain peak.
Jenny A. Fisher
([email protected])
ACCESS XII
We test hypotheses using model sensitivity simulations
2.0
Hg0, ng.m-3
1.8
standard simula4on
+ river/erosion source
Arc4c Rivers
1.6
1.4
1.2
1.0
J F M A M J J A S O N D
Month
With a large river / coastal erosion source included, GEOS-­‐Chem can reproduce the observed summer peak — a promising hypothesis, but more data are needed!
Jenny A. Fisher
([email protected])
ACCESS XII
Factors affecting Arctic mercury have changed dramatically
Hg emissions growth: 1979-­‐2010
Air temperature change: 1979-­‐2010
May sea ice extent trend: 1979-­‐2010
GISS
NSIDC
2 Gg
Asia
USSR
0
Streets et al. (2011)
What are the implica4ons for past — and future — variability in Arc4c Hg?
Jenny A. Fisher
([email protected])
ACCESS XII
Combined model-observation analysis lends insight
Atmospheric [Hg0], 1995-­‐2008
1. 30-­‐year GEOS-­‐Chem simula4on of Arc4c Hg (1979-­‐2008)
2. Evaluated by long-­‐term atmospheric observa4ons
Alert, Canada (1995-­‐2008)
Zeppelin, Ny Ålesund (2000-­‐2008)
Pallas, Finland (1996-­‐2008)
3. Extended to simula4on of surface ocean, where long-­‐
term records unavailable
Jenny A. Fisher
([email protected])
ACCESS XII
We use interannual variability to identify drivers of change
Focus on unique Arc:c signatures in spring and summer
Atmospheric [Hg0] seasonal cycle
Atmospheric [Hg0] 4me series
Spring-­‐summer trends are negligible, but IAV is large in observa4ons & model
Jenny A. Fisher
([email protected])
ACCESS XII
Atmospheric Hg variability governed by a combination of
meteorological drivers
Impacts of high T + low sea ice + high radia4on + shallow PBL
SPRING
SUMMER
Environmental drivers of IAV iden:fied using a principal component analysis (PCA). Dominant mode explains 80% IAV in spring, 43% in summer.
Jenny A. Fisher
([email protected])
ACCESS XII
Surface ocean mercury driven by snowmelt & rivers
pM
Surface ocean [THg]
Arc:c ocean mercury and IAV peak in summer
2.5
No trends in surface ocean mercury
2.0
J F M A M J J A S O N D
Jenny A. Fisher
([email protected])
ACCESS XII
Surface ocean mercury driven by snowmelt & rivers
pM
Surface ocean [THg]
Arc:c ocean mercury and IAV peak in summer
2.5
No trends in surface ocean mercury
2.0
J F M A M J J A S O N D
SUMMER
Jenny A. Fisher
Surface ocean inputs increase with:
Meltwater
• ⬆ summer solar radia:on (⬇ cloudiness)
• ⬆ spring-­‐summer T difference = ΔT
Rivers
• ⬆ May river flow
Deposi4on
• ⬆ summer wind speed
([email protected])
ACCESS XII
Surface ocean mercury driven by snowmelt & rivers
pM
Surface ocean [THg]
Arc:c ocean mercury and IAV peak in summer
2.5
No trends in surface ocean mercury
2.0
J F M A M J J A S O N D
SUMMER
Surface ocean inputs increase with:
Meltwater
• ⬆ summer solar radia:on (⬇ cloudiness)
• ⬆ spring-­‐summer T difference = ΔT
Rivers
• ⬆ May river flow
Deposi4on
• ⬆ summer wind speed
GCMs predict future ⬆ cloudiness & ⬇ ΔT → less Hg may be added to the ArcDc Ocean in future!
Jenny A. Fisher
([email protected])
ACCESS XII