A diagnostic evaluation of modeled mercury wet depositions in Europe using atmospheric speciated high-resolution observations.

This study is part of the Global Mercury Observation System (GMOS), a European FP7 project dedicated to the improvement and validation of mercury models to assist in establishing a global monitoring network and to support political decisions. One key question about the global mercury cycle is the efficiency of its removal out of the atmosphere into other environmental compartments. So far, the evaluation of modeled wet deposition of mercury was difficult because of a lack of long-term measurements of oxidized and elemental mercury. The oxidized mercury species gaseous oxidized mercury (GOM) and particle-bound mercury (PBM) which are found in the atmosphere in typical concentrations of a few to a few tens pg/m(3) are the relevant components for the wet deposition of mercury. In this study, the first European long-term dataset of speciated mercury taken at Waldhof/Germany was used to evaluate deposition fields modeled with the chemistry transport model (CTM) Community Multiscale Air Quality (CMAQ) and to analyze the influence of the governing parameters. The influence of the parameters precipitation and atmospheric concentration was evaluated using different input datasets for a variety of CMAQ simulations for the year 2009. It was found that on the basis of daily and weekly measurement data, the bias of modeled depositions could be explained by the bias of precipitation fields and atmospheric concentrations of GOM and PBM. A correction of the modeled wet deposition using observed daily precipitation increased the correlation, on average, from 0.17 to 0.78. An additional correction based on the daily average GOM and PBM concentration lead to a 50% decrease of the model error for all CMAQ scenarios. Monthly deposition measurements were found to have a too low temporal resolution to adequately analyze model deficiencies in wet deposition processes due to the nonlinear nature of the scavenging process. Moreover, the general overestimation of atmospheric GOM by the CTM in combination with an underestimation of low precipitation events in the meteorological models lead to a good agreement of total annual wet deposition besides the large error in weekly deposition estimates. Moreover, it was found that the current speciation profiles for GOM emissions are the main factor for the overestimation of atmospheric GOM concentrations and might need to be revised in the future. The assumption of zero emissions of GOM lead to an improvement of the mean normalized bias for three-hourly observations of atmospheric GOM from 9.7 to 0.5, Furthermore, the diurnal correlation between model and observation increased from 0.01 to 0.64. This is a strong indicator that GOM is not directly emitted from primary sources but is mainly created by oxidation of GEM.