Environ. Sci. Technol. 2010, 44, 8522–8528
Empirical Models for Estimating Mercury Flux from Soils C H E - J E N L I N , * ,† M A E S . G U S T I N , ‡ PATTARAPORN SINGHASUK,§ C H R I S E C K L E Y , ‡,⊥ A N D MATTHIEU MILLER‡ Department of Civil Engineering, Lamar University, Beaumont, Texas 77710, United States, Department of Natural Resources & Environmental Science, University of Nevada, Reno, Nevada 89557, United States, Department of Industrial Engineering, Lamar University, Beaumont, Texas 77710, United States, and Environment Canada, Vancouver, BC V6C 3S5, Canada
Received June 28, 2010. Revised manuscript received September 2, 2010. Accepted October 7, 2010.
Multiple parameters have been suggested to influence the exchange of mercury (Hg) between the atmosphere and soils. However, models applied for estimating soil Hg flux are simple and do not consider the potential synergistic and antagonist relationships between factors controlling the exchange. This study applied a two-level factorial experimental design in a gas exchange chamber (GEC) to investigate the individual and combined effects of three environmental factors (temperature, light, and soil moisture) on soil Hg flux. It was shown that individually irradiation, soil moisture, and air temperature all significantly enhance Hg evasive flux (by 90-140%). Synergistic effects (20-30% of additional flux enhancement) were observed for all two-factor interactions, with air temperature/soil moisture and air temperature/irradiation being the most significant. Results from the factorial experiments suggest that a model incorporating the second-order interactions can appropriately explain the flux response to the changes of the studied factors. Based on the factorial experiment results and using the flux data for twelve soil materials measured with a dynamic flux chamber (DFC) at various temperatures, soil moisture contents, solar radiation exposures, and soil Hg contents, two empirical models for estimating Hg flux from soils were developed. Model verification with ambient flux data not used to develop the models suggested that the models were capable of estimating dry soil Hg flux with a high degree of predictability (r ∼ 0.9).
Introduction Mercury (Hg) is a toxic air pollutant subject to long-range transport and global cycling. Elemental mercury (Hg(0)), the dominant form in the atmosphere, exhibits bidirectional exchange (i.e., deposition and evasion) with terrestrial surfaces (1). Divalent mercury (Hg(II)), on the other hand, is more readily removed by wet and dry deposition. The deposited Hg(II) tends to be retained in soils due to its higher * Corresponding author phone: +1 409 880 8761; fax: +1 409 880 8121; e-mail:
[email protected]. † Department of Civil Engineering, Lamar University. ‡ University of Nevada, Reno. § Department of Industrial Engineering, Lamar University. ⊥ Environment Canada. 8522
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 22, 2010
affinity for surfaces relative to Hg(0) and bonding to organic materials (2-6). The soil-bound Hg(II) may be reduced photochemically or biologically, and subsequently re-emitted (7-12). Although runoff can be an important Hg input to certain aquatic systems, Hg loss from soils by this process and leaching to groundwater is considered insignificant (3, 7). The main factors influencing Hg evasion from soils to the atmosphere are solar radiation, soil moisture, air temperature, and soil characteristics (Hg content, organic fraction, and permeability, etc.). Solar radiation has been found to be highly correlated with Hg flux (9-18). It has been hypothesized that intensive solar radiation could decrease the activation energy required for Hg(0) emission (17). Soil moisture can enhance soil Hg evasion at a level less than the saturated moisture content of soils (13, 19, 20). Air temperature has been found to be positively correlated with soil Hg emission (14, 16, 17, 21). It is also the factor more commonly applied to model air-soil Hg exchange. For example, Xu et al. (22) estimated soil Hg evasion as a function of temperature based on empirical data (8, 10). Gbor et al. (23) extended Xu’s model by incorporating the effect of soil Hg content derived from additional data (24). A more recent flux model for bare soils was proposed as a function of air temperature and Hg concentration only (25). Daily average Hg fluxes have also been modeled as a function of solar radiation, temperature, and soil moisture (26). Despite the conceivable evidence for the interactions between environmental parameters (e.g., light and temperature, light and soil moisture, etc.) (13, 14), none of the earlier studies extensively examined the magnitude of the interactions in a quantitative fashion, and models have been based on effects of individual factors (22-26). The primary objective of this study was to apply a factorial design of experiments to investigate the effect of soil moisture, solar radiation, and air temperature, individually and in combinations, on the observed flux from soils. The synergistic effects on flux response were then incorporated in building regression models that predict flux as a function of the studied environmental parameters. The models developed in this work may be useful for providing better estimates of natural emissions in atmospheric mercury modeling.
Methods Measurements of Hg Flux from Soil Surface. Measurements of Hg flux from a number of soil substrates were made using a single-pass gas exchange chamber (GEC) (27) and a dynamic flux chamber (DFC) (28). The GEC was employed for the factorial experiments that require precise control of the studied factors. The DFC was utilized to obtain ambient flux data from dry and wet soils. For the factorial experiments, soil samples were placed in a dome-shape, Pyrex GEC (12.3 L) sitting on a Teflon base supported by galvanized steel frame (Figure 1). The contact between the dome and base was sealed with polystyrene foam. A stainless steel fan was used for internal air mixing. Air entered and exited the GEC through ports at the chamber base. Particulate-filtered ambient laboratory air (Hg concentration 0.99). This is in agreement with the results reported previously (31, 32), and an indication that the strength of air-surface exchange of Hg is linearly dependent on the available Hg in soils. However, it should be noted that the substrates used in this study have a relatively low organic content (
