Atmospheric Environment (v.34, #7)

Review and intercomparison of operational methods for the determination of the mixing height by Petra Seibert; Frank Beyrich; Sven-Erik Gryning; Sylvain Joffre; Alix Rasmussen; Philippe Tercier (1001-1027).
The height of the atmospheric boundary layer (ABL) or the mixing height (MH) is a fundamental parameter characterising the structure of the lower troposphere. Two basic possibilities for the practical determination of the MH are its derivation from profile data (measurements or numerical model output) and its parameterisation using simple equations or models (which only need a few measured input values). Different methods suggested in the literature are reviewed in this paper. The most important methods have been tested on data sets from three different sites in Europe (Cabauw – NL, Payerne – CH, Melpitz – D). Parcel and Richardson number methods applied to radiosonde profiles and the analysis of sodar and wind profiler data have been investigated. Modules for MH determination implemented in five currently used meteorological preprocessors for dispersion models have been tested, too. Parcel methods using a revised coefficient for the excess temperature and Richardson number methods using a surface excess temperature worked well under convective conditions. Under stable conditions, the inherent difficulties call for a combination of several methods (e.g., mast and sodar). All the tested parameterisation schemes showed deficiencies under certain conditions, thus requiring more flexible algorithms able to take into account changing and non-classical conditions. Recommendations are formulated regarding both the analysis of profile measurements and the use of parameterisations and simple models, and suggestions for the preprocessor development and for future research activities are presented.
Keywords: Atmospheric boundary layer height; Stable boundary layer; Convective boundary layer; Dispersion model; Turbulence;

A modeling approach for the dispersion of pollutants released in the atmospheric boundary layer is presented and evaluated. The model includes a continuous formulation for the transport and turbulent diffusion processes that adequately represents the mechanisms in the various regimes of the atmospheric boundary layer. The vertical diffusion coefficients consider shear and buoyant-induced turbulence in a well-behaved functional dependence. The suggested eddy diffusivity and wind speed profiles compare adequately with those reported in the literature and observational data. The evaluation of the proposed dispersion model as well as an independent method, against data from diffusion experiments, showed a satisfactory physical behavior and a good overall statistical performance of each model. An improvement is introduced over the alternative approach with the suggested continuous formulation with similar performance.
Keywords: Air pollution modeling; Dispersion model; Model evaluation; Eddy diffusivity and wind speed profiles;

Wind tunnel and field calibration of six aeolian dust samplers by Dirk Goossens; Zvi Y Offer (1043-1057).
The efficiency of six aeolian dust samplers was tested via wind tunnel experiments and field measurements. In the wind tunnel, four samplers designed to measure the horizontal dust flux and one sampler designed to measure the vertical dust flux (in the downward direction, i.e., deposition) were calibrated against an isokinetic reference sampler. The horizontal dust flux samplers were: the big spring number eight sampler (BSNE), the modified Wilson and Cooke sampler (MWAC), the suspended sediment trap (SUSTRA), and the wedge dust flux gauge (WDFG). Vertical deposition flux was measured using a marble dust collector (MDCO). A modified Sartorius SM 16711 dust sampler with adjustable flow rate (SARTORIUS) was used as isokinetic reference sampler. In the field experiments, the WDFG was replaced by a Sierra ultra high volume dust sampler (SIERRA). Wind tunnel calibrations were carried out at five wind velocities ranging from 1 to 5 m s−1. Field calibrations were conducted during seven periods of two weeks each. The most efficient samplers are the MWAC and the SIERRA, followed by the BSNE and the SUSTRA. The WDFG is more effective than the BSNE at velocities below 3 m s−1, but its efficiency drops quickly at higher wind speeds. The most recommendable sampler for field measurements is the BSNE, because its efficiency varies only very slightly with wind speed. In the absence of horizontal flux samplers, the MDCO collector can be used as an alternative to assess horizontal dust flux and airborne dust concentration provided the appropriate calibrations are made.
Keywords: Wind; Aeolian dust; Calibration; Dust flux gauge; Dust deposit gauge;

A Lagrangian stochastic diffusion method for inhomogeneous turbulence by Donald L Ermak; John S Nasstrom (1059-1068).
A Lagrangian stochastic method of solving the diffusion equation for inhomogeneous turbulence is presented in this paper. This numerical method uses (1) a first-order approximation for the spatial variation of the eddy diffusivity and (2) the first three particle position moments to define a skewed, non-Gaussian particle position probability density function. Two cases of the variation of the eddy diffusivity near a boundary are considered. In the first case, the eddy diffusivity varies linearly with height and is zero at the boundary. The method handles this case efficiently and accurately by using the first few terms of the series representation of the analytic solution to construct components of the non-Gaussian position probability density function that are important near the boundary. In the second case, the eddy diffusivity is constant near the boundary, and the well-known solution for this case – a Gaussian function reflected at the boundary – is used. Comparison of numerical simulation results to analytic solutions of the diffusion equation show that this method is accurate. For cases where the eddy diffusivity varies linearly with height to zero at the boundary, we demonstrate that this method can be significantly more efficient than the commonly used method that assumes a Gaussian particle position probability density function.
Keywords: Random walk dispersion model; Diffusion equation; Inhomogeneous turbulence;

In droplets of clouds and fog the transformation of ozone (O3) to secondary oxidants, such as hydroxyl radicals (OH), is an important process. As demonstrated by experiments on model solutions, chromophoric compounds, represented by ubiquitous iron(III)–oxalato complexes, act as primary photoreactants. Upon photolysis these compounds produce peroxide anion (O2 ⋅−). O2 ⋅− reacts with aqueous ozone. This transforms aqueous O3 into OH. As estimates show, the quick equilibration with the gas-phase maintains aqueous O3 in droplets of clouds at about 1 nM. But already in the presence of ubiquitous traces of dissolved copper at 1 nM, O2 ⋅− rather reduces copper(II) to copper(I) than reacts with ozone. Therefore, it is rather an indirect loop in which Cu(I) acts as a further chain carrier to convert O3 to OH. Ubiquitous aqueous formate and formaldehyde then convert a fraction of non-selective OH to highly selective HO2 and O2 ⋅−, i.e. to O2(-I). Either directly or indirectly this O2(-I) reacts with further O3 to produce further OH. However, ubiquitous acetate and some other types of solutes, such as bisulfite, scavenge OH without recycling O2(-I). Experimental calibrations on model solutions showed that acetate limits the kinetic length of the O3 converting chain reaction just in proportion to its concentration relative to that of formate and formaldehyde. As demonstrated by further experiments, a similar aqueous O2(-I)–O3–OH radical-type chain reaction can also be initiated when some aqueous O3 reacts on black carbon particles. Black carbon was thereby studied as a surrogate for soot. Within cloud droplets the sum of all these O3 promoted aqueous-phase radical-type chain reactions increases the flux of aqueous-phase OH and its oxidation processes.
Keywords: Chain reactions; Cloud chemistry; Formate; Formaldehyde; Photo oxidants;

We describe the application of a dual beam Fourier transform infrared (FTIR) spectrometer and sampling system for simultaneous measurements of fluxes of several trace gases between the earth's surface and the atmosphere. The spectrometer is based on a commercial dual-output FTIR spectrometer with two long-path absorption cells, fully automated gas handling, data acquisition and quantitative spectrum analysis. The spectrometer may be operated in single or dual beam (optical subtraction) modes; the advantages and disadvantages of the two modes are tested and discussed. Measurements of fluxes of N2O, CO2 and NH3 from agricultural landscapes were made by the flux–gradient technique in two field trials in Sweden and Denmark in 1993. Fluxes of NH3 were determined following liquid manuring of a young wheat crop in early summer, and N2O and CO2 fluxes were measured from a recently harvested wheat stubble on an unfertilised organic soil in late summer. NH3 fluxes of more than 5 μgN m−2  s−1 (4 kg ha−1  d−1) were measured a day after fertilisation, decreasing to <0.5 μgN m−2  s−1 two days later. N2O fluxes averaged 42 ngN m−2  s−1 (36 g ha−1  d−1) over the six days of measurement and showed no significant diurnal or longer term variability. Minimum fluxes of 500 and 20 ngN m−2  s−1 for NH3 and N2O, respectively, were detectable with 20 min time resolution. The system is readily extendable to CH4 measurement. Improvement for the future should improve the minimum detectable fluxes.
Keywords: Carbon dioxide; Nitrous oxide; Ammonia; Optical subtraction; FTIR; Flux–gradient;

Canopy scale monoterpene emissions of Pinus sylvestris dominated forests by Janne Rinne; Hannele Hakola; Tuomas Laurila; Üllar Rannik (1099-1107).
Canopy scale monoterpene emissions were measured using the micrometeorological gradient method at two sites, Huhus and Hyytiälä, in the European boreal zone. At both sites the dominant tree species was Pinus sylvestris. The dominant monoterpene emitted by the forests was α-pinene, followed by Δ3-carene and β-pinene/myrcene. At the more homogenous Huhus site monoterpene emission correlated reasonably well with the IR-temperature of the canopy and the air temperature inside the canopy, and the common exponential equation could be used to describe the emissions. The β-coefficient obtained was 0.15°C−1, which is somewhat higher than the commonly used value of 0.09°C−1. The emission potential calculated using the measured canopy scale emissions and β=0.09°C−1 was 1.2 μg  g−1 dw  h−1 which is close to 1.5 μg  g−1 dw  h−1 used in recent emission inventories. The measured emissions at the more heterogeneous Hyytiälä site were not well correlated with the meteorological parameters. However, the coefficients obtained using the Huhus measurements predicted the magnitude of the fluxes measured at Hyytiälä reasonably well.
Keywords: Scots pine; Boreal forest; Biogenic hydrocarbons; Gradient method; Surface layer fluxes;

Control of solute concentrations in cloud and fog water by liquid water content by W. Elbert; M.R. Hoffmann; M. Krämer; G. Schmitt; M.O. Andreae (1109-1122).
We measured the concentrations of Cl, NO3 , and SO4 2− in atmospheric cloudwater collected with a passive collector from a remote marine site and the concentrations of Cl, NO3 , SO4 2−, Na+ and NH4 + in cloudwater from a continental site. At the continental site, an active rotating arm collector was employed for cloudwater collection. At both locations, we observed an inverse relationship between the measured ion concentrations and the liquid water content of the clouds. The product of the observed concentrations and the liquid water content, the cloudwater loading, was found to be nearly constant. A study of published cloudwater data showed that at the majority of the examined locations this product varied little at any given site, whereas large differences exist between sites. The finding of the small variation in cloudwater loadings could be explained by the hypothesis that nucleation scavenging is the dominant process in the acquisition of solutes into hydro-meteors during fog and cloud formation.
Keywords: Nucleation scavenging; Cloudwater loading; Cloudwater composition; Microphysics; Aerosol;

Time-resolved chemical ionization mass spectrometry (CI-MS) has been used to investigate the velocity-dependent emission factors for benzene, toluene, the C2-benzenes (xylenes and ethyl benzene) and nitrogen monoxide of a gasoline-driven passenger car (1.4 l, model year 1995) driven with or without catalytic exhaust gas treatment. A set of seven different driving cycles – including the European Driving Cycle (EDC), the US Urban (FTP 75) and the Highway driving cycles – with a total driving time of 12,000 s have been studied. From the obtained emission data, two sets of 15,300 and 17,200 data points which represent transient driving in the velocity range of 0–150 km h−1 and in an acceleration window of −2–3 m s−2 were explored to gain velocity-dependent emission factors. The passenger car, equipped with a regulated rhodium–platinum based three-way catalyst, showed optimal conversion efficiency (>95%) for benzene in the velocity range of 60–120 km h−1. The conversion of benzene was reduced (<80%) when driving below 50 km h−1 and the BTXE emissions significantly increased when driven at higher speed and engine load (>130 km h−1). Whereas the conversion efficiency for the class of C2-benzenes was reduced to 10%, no net conversion could be found for toluene and benzene when driven above 130 km h−1. In contrast, the benzene and toluene emissions exceeded those of the untreated exhaust gas in the velocity range of 130–150 km h−1 by 50–92% and by 10–34%, respectively. Thus, benzene and toluene were formed across the examined three-way catalyst if the engine is operated for an extended time in a fuel-rich mode (lambda<1).
Keywords: Vehicle emissions; Velocity-dependent emission factors; Aromatic hydrocarbons; Conversion efficiency; On-line CI-MS; Time-resolved exhaust gas analysis; de novo benzene formation;

Hydrogen peroxide deposition and decomposition in rain and dew waters by Vicky Ortiz; M. Angélica Rubio; Eduardo A. Lissi (1139-1146).
Peroxides and hydrogen peroxide were determined by a fluorometric method in dew and rain collected in the atmosphere of Santiago of Chile city. The measured peroxides comprise hydrogen peroxide (the main component) and peroxides not decomposed by catalase. The collected natural peroxides readily decompose in the natural matrix, rendering difficult an estimation of the values present in real-time. In order to establish the kinetics of the process and the factors that condition their decomposition, the kinetics of the decay at several pHs and/or the presence of metal chelators were followed. The kinetics of hydrogen peroxide decomposition in the water matrix was evaluated employing the natural peroxides or hydrogen peroxide externally added. First-order kinetics was followed, with half decay times ranging from 80 to 2300 min. The addition of Fe(II) in the micromolar range increases the decomposition rate, while lowering the pH (<3) notably reduces the rate of the process. The contribution of metals to the decomposition of the peroxides in the natural waters was confirmed by the reduction in decomposition rate elicited by its treatment with Chelex-100. Dew and rain waters were collected in pre-acidified collectors, rendering values considerably higher than those measured in non-treated collectors. This indicates that acidification can be proposed as an easy procedure to stabilize the samples, reducing its decomposition during collection time and the time elapsed between collection and analysis. The weighted average concentration for total peroxides measured in pre-treated collectors was 5.4 μM in rains and 2.2 μM in dews.
Keywords: Hydrogen peroxide; Dew; Rain; Peroxide decomposition; Peroxide stabilization;

The gradient diffusion theory with the exchange approximation by Boussinesq has been used successfully to predict dispersion of passive scalars in boundary layers. The knowledge of the turbulent diffusivity is required. Based on the analogy between the exchange of mass and of momentum, usually the turbulent viscosity (momentum) is used instead of the turbulent diffusivity of mass. This fact requires good knowledge of the turbulent Schmidt number. The latter is determined by means of wind tunnel experiments in a turbulent boundary layer above a flat plate. In contrast to literature where a constant is assumed mostly, it has been demonstrated for the turbulent Schmidt number that a strong dependence on height existed within the boundary layer.
Keywords: Turbulent Schmidt number; Concentration measurements; Turbulent mass transport; Boundary layer; Wind tunnel;

Conference (1153-1154).

Discussion (1157-1158).

Discussion (1158-1159).