Deposition of atmospheric mercury to vegetated terrestrial ecosystems is now understood to be a significant sink, and recent monitoring and isotopic data have suggested that dry deposition of gaseous elemental mercury (GEM) represents the majority of this atmosphere—surface exchange. Until recently, uptake of atmospheric mercury has been monitored using wet deposition and litterfall mass balance techniques, however these have been shown to not take into account all mercury deposition pathways and are not able to resolve the bi-directional nature of mercury air—surface exchange.

Micrometeorological methods provide a direct, non-intrusive method of observing fluxes of many trace gas species, and the instrumentation and logistic networks necessary to measure greenhouse gas fluxes­ have already been deployed across the developed world in large monitoring networks. Technological limitations currently exclude the application of direct eddy covariance techniques to resolving GEM fluxes, as it is not possible to measure GEM concentrations at the same timescales as the transporting turbulent structures. Relaxed eddy accumulation methods have shown successes in resolving GEM fluxes, however these methods are technically cumbersome, requiring complicated instrumentation and specialised expertise, thus limiting their applicability in large-scale monitoring networks.

Gradient-based methods instead relax some theoretical assumptions of turbulent exchange in order to retain technical simplicity in measurement of trace gas fluxes. These methods have successfully been deployed over short vegetation (grassland) environments, however applying these methods to forest ecosystems requires measuring the flux at the forest canopy as well as at the forest floor, thus excluding current gradient GEM sampling methods. We present here a technically simple method of resolving time-averaged GEM gradients at multiple levels within a forest canopy, thereby allowing calculation of forest floor and forest canopy fluxes. This system utilises commercially-available instrumentation and requires relatively little technical expertise, thereby making it ideal for deployment within pre-existing monitoring networks. We present here also preliminary data testing this system in a temperate deciduous forest during the summer growing season, and its abilities and limitations in measuring forest ecosystem fluxes. We propose that this system is currently ready for deployment within existing ecosystem exchange networks, and will provide essential data required for understanding an important aspect of the global mercury cycle.