Developing relationships between fluxes of natural sources of mercury and environmental and biogeochemical parameters for the purpose of scaling up fluxes

G. Dias*1, G. Edwards1, P. Rasmussen2, W. Schroeder3, J. Kemp1, C.F.- Hubble4, L.H.- Mitchell1

 

*gdias@uoguelph.ca, 1 School of Engineering, Univ. of Guelph, Guelph, ON, 2 Health Canada, Ottawa, ON, 3 Meteorological Service of Canada, Downsview, ON, 4 RWDI, Inc., Guelph, ON

Abstract

Mercury evasion processes from natural sources are not well understood, but can be significant in the biogeochemical cycling of mercury.  Additionally, large uncertainties exist in the magnitude of mercury fluxes from natural sources such as soils, water surfaces, vegetation, and fault zones. Quantifying natural source mercury emissions, and understanding the processes affecting their release are important for scaling up field flux measurements and for numerical modelling.  In situ measurements of volatile mercury emissions to the atmosphere from sources in contrasting geological and natural settings were conducted using micrometeorological methods. Using this extensive data set, a preliminary relationship has been established between mercury evasion and mercury substrate concentrations.  This relationship is complicated by environmental variables such as temperature, rainfall, and net radiation.  The interrelationships among mercury fluxes, environmental and biogeochemical factors are examined and the relevant relationships presented.  These relationships are required to model mercury cycling in the environment and to scale up in situ fluxes to local, regional and ultimately global scales.

Introduction

Gaseous mercury fluxes from natural sources must be considered when calculating global mass balances or for the development and description of biogeochemical cycles at local, regional and global scales (Schroeder et al., 1989), as well as to place anthropogenic metal emissions data into perspective (Rasmussen, 1998). Large uncertainties in estimates of natural Hg fluxes are caused by their high temporal and spatial variability, and by the difficulties in obtaining reliable and representative measurement rates of terrestrial Hg evasion into the atmosphere (Rasmussen, 1994; 1998).  The characterization of natural Hg emissions for use in regional and national emission inventories is limited, not only by the scarcity of data and measurement-related challenges, but also by the lack of knowledge of the controlling biogeochemical processes and environmental factors and their interactions.  

Since 1995, the University of Guelph, the Geological Survey of Canada, and Environment Canada have been characterizing Hg vapor fluxes from various geological settings in the Canadian environment using micrometeorological and dynamic chamber methods (Rasmussen et al., 1997).  Micrometeorological methods are non-intrusive and provide spatially averaged fluxes, which can be better related to the environmental factors affecting, thus they may be more appropriate for obtaining relationships for scaling up fluxes.

From these extensive data sets, it has been found that the spatial heterogeneity of the substrate Hg concentration affects the flux and a log-linear relationship between the flux and the soil substrate concentration has been established (Figure 1). Although Hg substrate concentrations are very important in determining Hg fluxes, and thus in establishing relationships for scaling up fluxes, understanding the interaction of environmental and biogeochemical factors is also important in scaling up fluxes to regional and ultimately global scales.

 

 


Figure 1.  Relationship between Hg substrate

concentrations and gaseous Hg fluxes obtained

at various sites across North America

using micrometeorological flux methods:

1)      Hopetown, ON;

2)       Clyde Forks; ON;

3)       Thunder Bay, ON (background site);

4)       Thunder Bay, ON  (black shale);

5)       Reno, NV;

6)       Pinchie, BC

 

 

Site descriptions and Methodology

Various mercuriferous geological sites have been monitored across North America.  These include Pinchie, BC (natural cinnabar occurrence along a regional scale fault zone), Clyde Forks, ON (naturally enriched mineralized fault zone), Thunder Bay, ON (carbonaceous fissile black shale), Reno, NV (naturally enriched mercuriferous soils), and Hopetown, ON (mineralized fault zone, background site).  At each site, an initial geochemical survey is undertaken to determine the presence of Hg, followed by a more extensive survey during the field campaign to determine Hg substrate concentrations by grain-size in the area being monitored.   Mercury flux data sets obtained at Pinchie, Thunder Bay, and Reno, are used here to describe the relationships and interactions of the biogeochemical and environmental factors affecting fluxes. 

Micrometeorological data (e.g. wind profile, eddy flux, heat flux, temperature, relative humidity, etc.) are collected both for use with the flux-gradient technique and to establish environmental correlates. The micrometeorological flux-gradient approach determines the vertical flux of total gaseous mercury (TGM) through the relationship, F= -K dc/dz, where K is the eddy diffusivity of the gas (m2s-1), determined from micrometeorological data as described by Wagner-Riddle et al. (1997), and dc/dz is the concentration gradient of the gas (kg m-4). 

A gas flow system samples at two heights above the surface in order to obtain the TGM concentration gradient.   The air at each height is transported via a common sampling tube to the TGM analyzer (Tekran model 2537A), where a subsample is diverted from the main sampling tube and passed through the analyzer. An average TGM concentration is determined at each level for the first 1½-hour period from the onset of measurements, and updated every ½ hour.

Results and Discussion

            In general, rain events during our Hg monitoring programs have shown that the fluxes increase for some period following the rain event.  In Pinchie and Thunder Bay, rain events are part of the normal climatic regime.  At the Reno site, which has a desert climate, an unusual rain event, as well as the heterogeneity of the Hg substrate concentrations at the site, were useful for seeing the impact of rain on Hg biogeochemical processes as discussed below.

            Figures 2 a,b,c show the effect of rain events on fluxes for the Thunder Bay, Reno, and Pinchie sites, respectively.  In each case, fluxes increased by at least 2-fold following the rain event.  In Reno, there was a 5-fold increase in average fluxes following the rain, with rain effects on the flux persisting for a longer period.  Thus, precipitation events may play a significant role in Hg emissions through their influence on biogeochemical processes.  Furthermore, in Reno, after the rain, there was a direct relationship between how much more flux occurred in areas with higher substrate concentrations, with a factor of 2 increase in substrate concentrations translating into a factor of 2 increase in TGM fluxes.  Flux chamber methods may always show these effects because the chamber prevents the soil from getting wetted (Lindberg et al., 1999). 

 


     a                                                                                   b

 

 

 

 

 

 

 


                                                                                          c

Figure 2.   Mercury flux time series for Thunder

Bay, ON (a), Reno, NV (b), and Pinchie, BC (c).

Arrows indicated rain events.   In most cases,

rain events resulted in increased fluxes

immediately following the rain.  In Reno,

the unusual rain event, resulted in increased

fluxes persisting for the following 2 days.

 

           

 

           

            An exponential relationship between the temperature and flux, related to the effect of temperature on the vapour pressure of mercury, has been observed previously (Gustin et al., 1996).  Analysis of the complete data sets for Pinchie revealed that there was a good correlation between these two variables (R2=0.83).  For Reno, however, there was no correlation between temperature and flux (R2=0.17), and it was believed that other factors at this site had a greater impact on the fluxes.  Since the variability in substrate concentrations and environmental factors were more pronounced at Reno, further data analysis was conducted to isolate these variables.  A stronger correlation with temperature (R2=0.62) was found when utilizing only the pre-rain flux data from areas of high substrate Hg concentrations.  Following the rain event, flux data from areas with high Hg substrate concentration areas showed little correlation (R2=0.23), suggesting that rain complicates this relationship due to its effect on biogeochemical processes.  In addition, the interaction between substrate concentrations and temperature in the flux is seen in the slopes of these correlations, which varied from 3 to 300 for Reno and Pinchie, respectively.

            It has been observed that peak fluxes coincide with peak net radiation and there is evidence that there is a correlation between the two, but this may be due to the influence of net radiation on temperature.  By analyzing the day and night data separately for both temperature and net radiation at Pinchie, it was found that there was good correlation between daytime fluxes and net radiation and temperature (R2= 0.87 and 0.83, respectively; see sample data in Figure 3), and night-time fluxes and temperature (R2=0.63), but there was no correlation between night-time fluxes and radiation (R2=0.01).  In Reno, there was no correlation between net radiation and Hg fluxes even when considering pre-rain event flux data from high substrate concentration areas unlike what was found for the temperature effects (R2=0.12).  Thus, the effect of net radiation on fluxes still requires more research to isolate temperature and light effects on fluxes.

 


Figure 3.  Correlation between daytime temperature

 and Hg flux at Pinchie from data collected using

 micrometeorological methods.

 

           

 

 

           

 

            The interaction of rain and other environmental parameters still requires more research to understand these processes and thus better model them for scaling up Hg fluxes.  There is a relationship between temperature and TGM fluxes, but rain events complicate this relationship through their influence on biogeochemical factors.  To understand the temperature relationship better and obtain appropriate equations and coefficients, the interaction between biogeochemistry and temperature must be understood through more extensive monitoring programs.  Additionally, there are still issues to be resolved with respect to net radiation and its effect on Hg fluxes.

Conclusions

            The dramatic increase in fluxes at Reno due to an isolated rain event showed that the calculated annual mean TGM flux could be event dominated, and it is important to account for these events in scaling up fluxes.  The ability to scale up is still limited by the lack of data on the relationships between various environmental and geochemical processes. The micrometeorological data show the utility of this method in obtaining representative fluxes without interfering with the environment, which influences the flux.  Relationships between environmental factors and mercury flux and the interaction of biogeochemical and environmental parameters can be studied using these methods to obtain realistic environmental correlates for scaling up.

References

Gustin, M.S., G.E. Taylor, Jr., T.L. Leonard, R.E. Keislar (1996). Water Air, Soil Pollut., 80, 217-220.

Lindberg, S.E., Zhang, H., Gustin, M., Vette, A. Owens, J., Marsik, F., Casimir, A., Ebinghaus, R., Edwards, G., Fitzgerald, C., Kemp, J., Kock, H.H., London, J., Majewski, M., Poissant, L., Pilote, M., Rasmussen, P., Schaedlich, F., Schneeberger, D., Sommar, J., Turner, R., Wallschlaeger, D., and Xiao, Z (1999).  J. Geophys. Res., 104(D17), 21,879- 21,888.

Rasmussen, P.E. (1994). Envir. Sci. Technol., 28 (13), 2233-2241.

Rasmussen, P.E. (1998). Environ. Geol., 33 (2-3), 170-182.

Rasmussen, P.E., G.C. Edwards, J. Kemp, C. F.-Hubble, and W.H. Schroeder (1997). In Proc. Metals & the Environment: An International Symposium. (J. Skeaff, ed.). Metallurgical Soc. of the Canadian Institute of Mining, Metallurgy and Petroleum, Montreal, May 5-6, 73-83.

Schroeder, W.H., J. Munthe, and O. Lindqvist (1989). Water, Air, Soil Pollut., 48,  337-347.

Wagner-Riddle, C., G.W. Thurtell, G.E. Kidd, E.G. Beauchamp, and R. Sweetman (1997). Can. J. Soil Sci. 77, 135-144.