Leiming Zhang

Litterfall mercury dry deposition in the eastern USA

Mercury (Hg) in autumn litterfall from predominately deciduous forests was measured in 3 years of samples from 23 Mercury Deposition Network sites in 15 states across the eastern USA. Annual litterfall Hg dry deposition was significantly higher (median 12.3 micrograms per square meter (μg/m(2)), range 3.5-23.4 μg/m(2)) than annual Hg wet deposition (median 9.6 μg/m(2), range 4.4-19.7 μg/m(2)). The mean ratio of dry to wet Hg deposition was 1.3-1. The sum of dry and wet Hg deposition averaged 21 μg/m(2) per year and 55% was litterfall dry deposition. Methylmercury was a median 0.8% of Hg in litterfall and ranged from 0.6 to 1.5%. Annual litterfall Hg and wet Hg deposition rates differed significantly and were weakly correlated. Litterfall Hg dry deposition differed among forest-cover types. This study demonstrated how annual litterfall Hg dry deposition rates approximate the lower bound of annual Hg dry fluxes.

Chemical characterization and source apportionment of PM2.5 in a semi-arid and petrochemical-industrialized city, Northwest China

Daily PM2.5 samples were collected in 2014 at a suburban petrochemical industrial site and a downtown site in Lanzhou city, Northwest China. Major chemical components in PM2.5, including water-soluble ions, metal elements, and organic and elemental carbon (OC and EC) were determined. The chemical mass closure method and the ISORROPIA II thermodynamic equilibrium model were used to reconstruct PM2.5 mass and quantify the combinations of NH4+, SO42- and NO3- to PM2.5. Positive matrix factorization (PMF) model was employed to apportion potential sources of PM2.5. The annual average PM2.5 concentration was 93.7±49.6μgm-3 at the suburban petrochemical industrial site and 88.9±52.0μgm-3 at the urban site, with the highest seasonal average in winter and the lowest in summer at both sites. Mineral dust was identified as the highest contributor to PM2.5 in spring, while water-soluble inorganic ions and carbonaceous aerosols were the dominant chemical components in other seasons. The correlation relationships between OC and EC and between K+ and EC suggested that coal combustion and vehicle exhaust were the major sources of carbonaceous aerosols in Lanzhou. Six major sources were identified by the PMF model. Coal combustion, soil dust, traffic emissions, and secondary inorganic aerosols were the dominant contributors, together accounting for 82% of PM2.5 mass.

Atmospheric mercury deposition to forests in the eastern USA

Atmospheric mercury (Hg) deposition to forests is important because half of the land cover in the eastern USA is forest. Mercury was measured in autumn litterfall and weekly precipitation samples at a total of 27 National Atmospheric Deposition Program (NADP) monitoring sites in deciduous and mixed deciduous-coniferous forests in 16 states in the eastern USA during 2007-2014. These simultaneous, uniform, repeated, annual measurements of forest Hg include the broadest area and longest time frame to date. The autumn litterfall-Hg concentrations and litterfall mass at the study sites each year were combined with annual precipitation-Hg data. Rates of litterfall-Hg deposition were higher than or equal to precipitation-Hg deposition rates in 70% of the annual data, which indicates a substantial contribution from litterfall to total atmospheric-Hg deposition. Annual litterfall-Hg deposition in this study had a median of 11.7xa0μg per square meter per year (μg/m2/yr) and ranged from 2.2 to 23.4xa0μg/m2/yr. It closely matched modeled dry-Hg deposition, based on land cover at selected NADP Hg-monitoring sites. Mean annual atmospheric-Hg deposition at forest study sites exhibited a spatial pattern partly explained by statistical differences among five forest-cover types and related to the mapped density of Hg emissions. Forest canopies apparently recorded changes in atmospheric-Hg concentrations over time because litterfall-Hg concentrations decreased year to year and litterfall-Hg concentrations were significantly higher in 2007-2009 than in 2012-2014. These findings reinforce reported decreases in Hg emissions and atmospheric elemental-Hg concentrations during this same time period. Methylmercury (MeHg) was detected in all litterfall samples at all sites, compared with MeHg detections in less than half the precipitation samples at selected sites during the study. These results indicate MeHg in litterfall is a pathway into the terrestrial food web where it can accumulate in the prey of songbirds, bats, and raptors.

Evaluation and Improvement of a Dry Deposition Model using SO2 and O3 Measurements over a Mixed Forest

A dry deposition model (RDM) for operational application has beenevaluated and modified in the present study. Field measurements of friction velocity and dry deposition velocity of SO2 andO3 over a mixed forest have been used to evaluate RDM. It was found that RDM predicts friction velocities very close to measurements and thus it can predict reasonable aerodynamic resistance. RDM overestimated O3 deposition during dry nighttime conditions and underestimated both O3 andSO2deposition for early morning hours. It could not predict the mean diurnal variation in deposition velocity for either O3 or SO2 deposition under wet surface conditions. Modifications have been made for O3 and SO2 dry deposition based on the comparison of results and based upon additional published data. Compared to an earlier version of RDM, the modified versionpredicts better results for O3 and SO2 dry deposition,especially under rain and dew conditions.

The effect of subgrid velocity scale on site-specific/subgrid area and grid-averaged dry deposition velocities

Abstract A method for deriving the site-specific and subgrid area wind speed and friction velocity from regional model output and detailed land type information is developed. The “subgrid velocity scale” is introduced to account for generation of turbulent fluxes by subgrid motions. The grid vector averaged wind speed is adjusted by adding the subgrid velocity scale. This is to account for the fact that the spatial average of the local wind speed is usually larger than the absolute value of the vector averaged velocity ( | V ⇀ | ), especially when there are different land or surface types within the spatial averaging area and when V ⇀ is small. The assumption of uu*=constant is then applied within a model grid area to obtain wind speed and friction velocity for specific sites and subgrid areas. Using this method, the site-specific and subgrid area wind speed and friction velocity can be estimated from grid-averaged model output. In addition, more realistic air pollutant dry deposition velocities for specific locations and subgrid areas can be calculated. Grid-averaged deposition velocity values calculated using this approach tend to be about 30% different (either larger or smaller) for HNO3 and sulphate and about 10% different for SO2 and O3 compared to values calculated by assuming a constant wind speed over the whole model grid area. These differences are found to be even larger at specific sites or over some subgrid areas. This method can be applied to determine a more realistic wind speed, friction velocity and pollutant dry deposition velocity at specific locations using gridded meteorological data.

Atmospheric removal of PM2.5 by man-made Three Northern Regions Shelter Forest in Northern China estimated using satellite retrieved PM2.5 concentration

Atmospheric removal of PM2.5 by the Three Northern Regions Shelter Forest (TNRSF) - the so called Green Great Wall (GGW) in northern China through dry deposition process was estimated using a bulk big-leaf model and a vegetation collection model. Decadal trend of PM2.5 dry deposition flux from 1999 to 2010 was calculated from modeled dry deposition velocity and air concentration retrieved from the satellite remote sensing. Dry deposition velocities of PM2.5 calculated using the two deposition models increased in many places of the TNRSF over the last decade due to increasing vegetation coverage of the TNRSF. Both increasing deposition velocity due to forest expansion and PM2.5 atmospheric level contributed to the increasing deposition flux of PM2.5. The highest atmospheric deposition flux of PM2.5 was found in the Central-north region covering Beijing-Tianjin-Hebei area, followed by the Northwestern and the Northeastern regions of the TNRSF. While greater collection of PM2.5 by vegetation was identified in the Northeastern region of the TNRSF due to higher forest coverage over this region, the most significant incline of the PM2.5 atmospheric removal due to vegetation collection was discerned in the Central-north region because of the most rapid increase in the vegetation coverage in this region. A total mass of 2.85×107t PM2.5 was estimated to be removed from the atmosphere through dry deposition process over the TNRSF from 1999 to 2010. The two deposition models simulated similar magnitude and spatial patterns of PM2.5 dry deposition fluxes. Our results suggest that the TNRSF plays a moderate role in PM2.5 uptake, but enhances PM2.5 atmospheric removal by 30% in 2010 than in 1980.

Trends of deposition fluxes and loadings of sulfur dioxide and nitrogen oxides in the artificial Three Northern Regions Shelter Forest across northern China

This study provides the first estimate of dry deposition fluxes of criteria air pollutants (SO2 and NOx) across the Three Northern Regions Shelter Forest (TNRSF) region in Northern China and their long-term trends from 1982 to 2010 using the inferential method. Dry deposition velocities of SO2 and NOx increased in many places of the TNRSF up to 118.2% for SO2 and 112.1% for NOx over the last three decades due to the increased vegetation coverage over the TNRSF. The highest atmospheric deposition fluxes of SO2 and NOx were found in the Central-North China region, followed by the Northeast and the Northwest China regions of the TNRSF. A total of 820,000xa0t SO2 and 218,000xa0t NOx was estimated to be removed from the atmosphere through dry deposition process over the TNRSF from 1982 to 2010. About 50% of the total removal occurred in the Central-North China region. The estimated total SO2 and NOx dry deposition fluxes from 1982 to 2010 between a TNRSF site in this region and an adjacent farmland outside the TNRSF showed that the fluxes of these two chemicals at the TNRSF site were the factors of 2-3 greater than their fluxes in the farmland.

Dry deposition of O3 and SO2 estimated from gradient measurements above a temperate mixed forest.

Vertical profiles of O3 and SO2 concentrations were monitored at the Borden Forest site in southern Ontario, Canada from May 2008 to April 2013. A modified gradient method (MGM) was applied to estimate O3 and SO2 dry deposition fluxes using concentration gradients between a level above and a level below the canopy top. The calculated five-year mean (median) dry deposition velocity (Vd) were 0.35 (0.27) and 0.59 (0.54) cm s(-1), respectively, for O3 and SO2. Vd(O3) exhibited large seasonal variations with the highest monthly mean of 0.68xa0cmxa0s(-1) in August and the lowest of 0.09xa0cmxa0s(-1) in February. In contrast, seasonal variations of Vd(SO2) were smaller with monthly means ranging from 0.48 (May) to 0.81xa0cmxa0s(-1) (December). The different seasonal variations between O3 and SO2 were caused by the enhanced SO2 uptake by snow surfaces in winter. Diurnal variations showed a peak value of Vd in early morning in summer months for both O3 and SO2. Canopy wetness increased the non-stomatal uptake of O3 while decreasing the stomatal uptake. This also applied to SO2, but additional factors such as surface acidity also played an important role on the overall uptake.

Evaluation and Intercomparison of Five North American Dry Deposition Algorithms at a Mixed Forest Site

To quantify differences between dry deposition algorithms commonly used in North America, five models were selected to calculate dry deposition velocity (Vd) for O3 and SO2 over a temperate mixed forest in southern Ontario, Canada, where a 5-year flux database had previously been developed. The models performed better in summer than in winter with correlation coefficients for hourly Vd between models and measurements being approximately 0.6 and 0.3, respectively. Differences in mean Vd values between models were on the order of a factor of 2 in both summer and winter. All models produced lower Vd values than the measurements of O3 in summer and SO2 in summer and winter, although the measured Vd may be biased. There was not a consistent tendency in the models to overpredict or underpredict for O3 in winter. Several models produced magnitudes of the diel variation of Vd (O3) comparable to the measurements, while all models produced slightly smaller diel variations than the measurements of Vd (SO2) in summer. A few models produced larger diel variations than the measurements of Vd for O3 and SO2 in winter. Model differences were mainly due to different surface resistance parameterizations for stomatal and nonstomatal uptake pathways, while differences in aerodynamic and quasi-laminar resistances played only a minor role. It is recommended to use ensemble modeling results for ecosystem impact assessment studies, which provides mean values of all the used models and thus can avoid too much overestimations or underestimations.

Identifying Changes in Source Regions Impacting Speciated Atmospheric Mercury at a Rural Site in the Eastern United States

AbstractTo investigate the effectiveness of emission reductions on the concentrations of gaseous elemental mercury (GEM), gaseous oxidized mercury (GOM), and particulate-bound mercury (PBM) at a rural site in Maryland (MD08), long-term (2005–14) measurements of speciated atmospheric mercury were analyzed using concentration-weighted trajectory (CWT) analysis. CWT results suggested that the number of major source regions contributing to GEM, GOM, and reactive mercury (RM = GOM + PBM) over the eastern United States and southeastern Canada declined over time. Across much of these regions, source contributions in 2011–14 decreased by up to 20% for GEM, by greater than 60% for GOM, and by 20%–60% for PBM compared to 2006–08, largely because of the decreases in power-plant mercury emissions since 2009. Changes in the spatial distribution of the source regions were also observed over time. Increases in source contributions of GEM after 2011 over the northeastern United States and southeastern Canada were predomi...