Steve Brooks 1, Steve Lindberg 2, Mike Goodsite 3, Bob Stevens 4, Matt Landis 5, Karen Scott 6, Tilden Meyers 1, Jerry Lin 7 and Glen McConville 8

1 NOAA Atmospheric Turbulence and Diffusion Division, 456 S. Illinois Ave, PO Box 2456, Oak Ridge, TN 37831, 865-576-9148, Fax 865-576-1327, E-mail:brooks@atdd.noaa.gov
2 Environmental Sciences Division, Oak Ridge National Laboratory
3 National Environmental Research Institute of Denmark (visiting scientist at NOAA ATDD)
4 Florida Dept. of Environmental Protection
5 US Environmental Protection Agency, Research Triangle Park
6 University of Manitoba, Canadian Fresh Water Inst.
7 Lamar University
8 NOAA Climatic Monitoring and Diagnostics Lab, Barrow, Alaska

Overview
At Barrow we have conducted measurements of atmospheric and snowpack Hg concentrations, boundary layer modeling, direct flux measurements using a relaxed eddy accumulation system, and aircraft Hg profiling. Together the results give a picture of local, near surface, solar UVb driven, near-surface atmospheric conversions of gaseous elemental mercury (GEM; from long-range transport) to reactive gaseous mercury (RGM) and, to a lesser extent, total particulate mercury (TPM).

The newly formed RGM deposits rapidly (Vd 2 cm/s ) to the snowpack. Overall, the modeled/measured springtime atmospheric flux (February-May) of mercury to the snowpack at Barrow is approximately 55.1 gHg m-2, comparing well with measured concentrations in monthly snow samples which indicate a total flux into the snow pack of 56.9 gHg m-2. Partial re-emission of accumulated mercury occurs around snow melt (early June) with a total re-emission of 41.7 gHg m-2 . Giving a net annual flux into the Barrow environment of roughly 15 gHg m-2 , of which approximately 50% is bioavailable, meaning that the mercury has been converted to compound(s) that are readily absorbed by living cells.

Surface Air Measurements
GEM has been continuous monitored since 1998, RGM has been monitored since September 1999, and TPM has been monitored since March 2001, using the Tekran 2537a(GEM)/1130(RGM)/1135(TPM) sensors. Episodic depletions of GEM begin after polar sunrise, Julian day 23. The mean mercury concentration from July to January 23 (local polar sunrise date) is 1.76 ng m-3 (close to global background levels) with a standard deviation of 0.166. From polar sunrise (Jan. 23) through May, the mean is 1.37 ng m-3 (below global ambient levels) with a standard deviation of 0.703. Most notably, the lowest GEM concentrations consistently seen during this period were in the range of 0.05-0.10 ng m-3 .

These depletion events end abruptly at snow melt ( roughly Julian day 160) and are replaced with uniform enhancements. GEM concentrations increase from an average of 1.35 ng m-3 for the week prior to snow melt (minima = 1.0 ng m-3) to 2.40 ng m-3 for the week after snow melt (maxima = 3.0 ng m-3). We hypothesize that this is due to a slow re-emission of mercury back to the atmosphere.

Atmospheric RGM enhancements coincide with GEM depletions during sunlight (incident UVb) periods and reaching as high as 0.95 ng m-3 , the highest concentration ever measured in a remote environment. RGM enhancements end abruptly at snowmelt. Springtime atmospheric TPM concentrations vary from 0 to 40 pg m-3 and are predominately anti-correlated to RGM concentrations, indicating a likely heterogeneous conversion pathway.

Aircraft RGM Measurements
In the first airborne measurement of RGM in the arctic, manual RGM denuder tubes were attached to the outer strut of a Cessna 207 and allowed to collect for one hour periods. The flights (conducted on three midday periods in late March and early April) were conducted in succession at 1000m altitude (exterior to the boundary layer) and 100m altitude (within the boundary layer) just north-east of Point Barrow. RGM averaged 1.53 and 18.33 pg m-3 at 1000m and 100m, respectively, providing the first independent verification that RGM production and significant concentrations are restricted to within the atmospheric boundary layer.

Relaxed Eddy Accumulation (REA) Measurements
The world's first RGM REA dry deposition flux measurements were conducted at Barrow in late March and early April. A tower based REA system with manual RGM denuder tubes was placed 3m above the snowpack. Significant RGM fluxes measured during March 29 - April 12 were directed towards the snow surface (mean net deposition = 2±1 ngHg m-2 hr-1; N=11). Computed dry deposition velocities for RGM were high as expected (~1-3 cm/s), and agree with those predicted by the developed inverse boundary layer model and the flux rates derived from snowpack Hg measurements.

Snowpack Mercury Measurements
Over the springtime of 2000 total Hg in surface snow increased steadily from <1 to >90 ng/L. Measurements of methylmercury (MeHg) in surface snow likewise increased from 0.01 ng/L before sunrise to 0.59 ng/L in May. Although its source is currently unknown, we suspect a mechanism involving the oxidation of dimethylmercury by the Cl radical to form CH3HgCl. The high concentrations of these Hg species in the late spring Barrow snowpack are not only unprecedented for remote sites, but also exceed typical levels near industrialized regions.

A significant fraction of the oxidized inorganic mercury [Hg(II)] in snow was biologically available to bacteria. "Bioavailable mercury" was determined using the mer-lux bioreporter Vibrio anguillarum pRB28 and its constitutive control V. anguillarum pRB27. Prior to polar sunrise, bioavailable Hg(II) was undetectable in Barrow snow. It then increased from 0.22 ng/L (~1% of total Hg) in March to 8.8 ng/L (nearly 13% of the total Hg) in May. Prior to this study, bioavailable Hg(II) had never been measured in the Arctic and its ultimate fate, therefore, is unknown. However, since it is the substrate for both reduction (to GEM) and methylation to the highly toxic MeHg, the implications of such elevated levels of bioavailable Hg(II) entering the ecosystem are considerable.

During snow melt total Hg decreased by 92 % and MeHg by 83 %. Bioavailable Hg(II) also decreased, but at a slower rate (67 %), so that the fraction of bioavailable Hg(II) in snow/slush/melt water increased significantly from 13 to 55%. Snow cores collected during melt exhibited a uniformly reduced total Hg concentration in the slushy surface layers (21±10 ng/L from 20-100 cm depth, temperatures from -2 to 0°C), with total Hg remained high as 90 ng/L in the deeper, still frozen layers (~140-150 cm, temperatures <-5°C). The flux of snowpack Hg to air and melt water is clearly influenced by the melting process. Although we did not measure bioavailable Hg(II) in runoff, total Hg in runoff (~30 ng/L) was elevated over that in slush, suggesting that a significant portion of the Hg in the snowpack enters the local ecosystem during snowmelt. Some of the Hg which runs off into the tundra continues to be photoreduced and evaded to the atmosphere under the elevated 24-h UVb levels which continue through the summer resulting in an estimated total re-emission to the atmosphere of 41.7 gHg m-2.

Boundary Layer Modeling
A simple rate equation for boundary layer RGM can be expressed as:

where F is the surface flux of RGM, and z is the boundary layer height. The first term represents the chemical rate for the in-situ generation of RGM from GEM as a function of air temperature, T, and incident solar UVb. Inputing the measured RGM, UVb, T, and assumed boundary layer dynamics, the January-May RGM flux (F) from the atmospheric boundary layer into the snowpack is estimated to be 55.1 g(Hg)m-2 . If these fluxes ( 55 g(Hg)m-2 ) are typical over the circumpolar arctic ( 33.4 million km2 defined by UN Arctic Monitoring and Assessment Program), this represents an temporary annual sink of approximately 1837 metric tons mercury, or about 25% of the total atmospheric mercury burden.

Acknowledgements
Aspects of this study have been generously funded by the USEPA and the NOAA Arctic Research Office. We would also wish to thank the following individuals at Barrow for their time and efforts; Glenn Sheehan and Dave Ramey (Barrow Arctic Science Consortium), Dan Endres and Malcolm Gaylord (NOAA CMDL Barrow)

Fig. 1- Trends in Hg° at Barrow during 1998-1999, showing similarities between Hg° and O3 depletion events, and development of a BrO "high" near Barrow during this depletion event.

Fig. 2- Trends in Hg° and reactive gaseous mercury at Barrow during 2000, showing parallel trends in total Hg in the surface snowpack and mean daily UV-B.

Fig. 3 - A diel cycle of tropospheric gaseous Hg species and UV-B at Barrow.

Fig. 4- Trends in several Hg species in the atmosphere and in the snowpack at Barrow around the period of annual snowmelt (during the June 4-10, 2000 snowmelt, slushy snow was collected from atop the frozen snowpack).