Use of atmospheric concentrations and passive samplers to assess surface-atmosphere exchange of gaseous mercury in forests.

ABSTRACT

Direct measurements of gaseous elemental mercury (GEM) exchanges over global ecosystems are challenging and require extensive and costly measurement systems. Here, we explore the use of atmospheric GEM concentration variability and passive samplers to assess underlying ecosystem GEM exchanges at two rural temperate forests in the northeastern United States. We find strong temporal alignments between atmospheric GEM concentration declines and ecosystem GEM deposition in spring at both forests, which followed patterns of CO2 and suggests that ambient air GEM concentration monitoring provides a proxy measurement to assess forest GEM sinks. In fall, we observe GEM concentration increases and reversal of GEM fluxes to emissions, but with poor temporal alignments. Diel GEM concentration variability did not correspond to diel patterns of ecosystem GEM fluxes, which is driven by boundary layer dynamics with different atmospheric mixing depths during day- and nighttime. Passive samplers (PASs) deployed to measure vertical GEM gradients across six heights throughout one of the forest canopies showed excellent agreements with active measurements in detecting seasonal concentration patterns at all deployment heights. We find frequent qualitative agreement between the direction of active and PAS derived concentration gradients, but small concentration differences over small (1.3 and 4.9 m) distances prevent a quantitative comparison of methods. Furthermore, time-averaged GEM concentration gradient measurements are always biased towards stable nighttime periods, while ecosystem GEM fluxes are dominated by daytime exchanges, which results in the inability of integrated measurements such as PAS to correctly quantify forest GEM exchanges. We conclude that concentration measurements both via active and passive sampling can serve as proxies to assess underlying ecosystem GEM sinks and sources, but that the use of passive samplers to quantify GEM exchange via gradient measurements is limited due their strong nighttime biases.