B11D-0454
Preliminary Assessment of Mercury Atmosphere-Surface Exchange Parameterizations for Incorporation into Chemical Transport Models
Monday, 14 December 2015
Poster Hall (Moscone South)
Tanvir Khan1, Yannick Agnan2, Daniel Obrist2, Noelle E Selin3, Noel R Urban4, Shiliang Wu5 and Judith A Perlinger4, (1)Michigan Technological University, Houghton, MI, United States, (2)Desert Research Institute Reno, Reno, NV, United States, (3)MIT, Cambridge, MA, United States, (4)Michigan Tech Univ, Houghton, MI, United States, (5)Michigan Tech, Houghton, MI, United States
Abstract:
Inadequate representation of process-based mechanisms of exchange behavior of elemental mercury (Hg0) and decoupled treatment of deposition and emission are two major limitations of parameterizations of atmosphere-surface exchange flux commonly incorporated into chemical transport models (CTMs). Of nineteen CTMs for Hg0 exchange we reviewed (ten global, nine regional), eight global and seven regional models have decoupled treatment of Hg0 deposition and emission, two global models include no parameterization to account for emission, and the remaining two regional models include coupled deposition and emission parameterizations (i.e., net atmosphere-surface exchange). The performance of atmosphere-surface exchange parameterizations in CTMs depends on parameterization uncertainty (in terms of both accuracy and precision) and feasibility of implementation. We provide a comparison of the performance of three available parameterizations of net atmosphere-surface exchange. To evaluate parameterization accuracy, we compare predicted exchange fluxes to field measurements conducted over a variety of surfaces compiled in a recently developed global database of terrestrial Hg0 surface-atmosphere exchange flux measurements. To assess precision, we estimate the sensitivity of predicted fluxes to the imprecision in parameter input values, and compare this sensitivity to that derived from analysis of the global Hg0 flux database. Feasibility of implementation is evaluated according to the availability of input parameters, computational requirements, and the adequacy of uncertainty representation. Based on this assessment, we provide suggestions for improved treatment of Hg0 net exchange processes in CTMs.