RESUMO
Dispersion models have proven to assist the development of regulation strategies for the mitigation of odour impact. Nevertheless, the complexity derived from the definition of the sources and the replication of the subjective perception of chemical mixtures raise the question whether it is enough to perform an assessment based exclusively on the predictions of models. Furthermore, there is still an ongoing debate on the most appropriate methodology to reproduce sub-hourly peak concentrations. With this in mind, the active participation of the affected community could help to identify better the processes that cause odour annoyance and tune the results obtained with the dispersion models. Recently, the AirQ application has been implemented in the Czech Republic to allow citizens to report odour episodes to the entity in charge. Hence, the goal of this work was to integrate the information collected from the complainants with the simulations from the Gaussian model SYMOS, and the Lagrangian models AUSTAL and GRAL. The evaluation was performed in three sites with different emission characteristics and terrain: a pig farm, a pet food producer, and an edible oil industry. The outcome of this approach allowed to evaluate the suitability of each model depending the characteristics of the source, compare the use of a constant peak-to-mean factor of 4 against the Concentration Variance Model, and determine the applicability of certain odour impact criteria (OIC) for establishing separation distances.
RESUMO
From 2nd April 2008 to 28th March 2009, a total 248 daily samples of the PM2.5 and PM10 were collected every sixth day parallel at two suburban sites (Libus and Suchdol) located at the two opposite sides (south and north, respectively) of Prague, Czech Republic. The PM2.5 samples were analyzed for ions by ion chromatography (IC), organic and elemental carbon (OC and EC) by OC/EC analyzer and PM10 samples also for 56 elements by inductively coupled plasma-mass spectrometry (ICP-MS). The average annual PM2.5 and PM10 was 24.4⯱â¯13.0⯵gâ¯m-3 and 26.7⯱â¯15.1⯵gâ¯m-3, respectively, in Prague-Libus, and 25.1⯱â¯22.1⯵gâ¯m-3 and 27.1⯱â¯23.2⯵gâ¯m-3, respectively, in Prague-Suchdol. Since the species forming large part of the aerosol mass were strongly correlated (Spearman's rank correlation coefficient rsâ¯>â¯0.80), the variability of PM2.5 and PM10 concentration was mainly driven by the local meteorology or regional and/or long range transport. PM10 mass closure was calculated based on analytical results with the average percentage of recalculated mass of 77⯱â¯19% in Prague-Libus and 86⯱â¯16% in Prague-Suchdol. The most abundant groups in PM10 at both sites during the four seasons were OM (Prague-Libus 34% and Prague-Suchdol 37%) and SIA (Prague-Libus 30% and Prague-Suchdol 34%). The Positive Matrix Factorization (PMF) was applied to the chemical composition of PM10 from both sites (124 samples) together to determine its sources. The nine factors were assigned as: mixed factor secondary sulphate and biomass burning, secondary sulphate, traffic, secondary nitrate, road dust, residential heating, aged sea salt, industry and mixed factor road salt along with aged sea salt. According to the polar plots and ventilation index (VI) east/west classification analysis the sources were separated based on origin to four categories local, urban agglomeration, regional and long range transport (LRT). The mixed source secondary sulphate and biomass burning, residential heating and industry were common sources of local origin at both sites. Prague-Suchdol was influenced by traffic related pollution from the urban agglomeration more than Prague-Libus where the traffic and road dust/salt were of local origin. The regional pollution by secondary sulphates and nitrate was also relevant at both sites along with long range transport of sea salt from North Atlantic Ocean, Norwegian Sea and North Sea. The contribution of the local sources to PM10 was significant mainly at Prague-Libus site. However, the sources of regional origin were also important and influence of urban agglomeration pollution to PM10 is not negligible as well.
RESUMO
The quantification of atmospheric deposition flux is essential for assessment of its impact on ecosystems. We present an advanced approach for the estimation of the spatial pattern of atmospheric nitrogen deposition flux over the Czech forests, collating all available measured data and model results. The aim of the presented study is to provide an improved, more complete, more reliable and more realistic estimate of the spatial pattern of nitrogen deposition flux over one country. This has so far usually been based on measurements of ambient NOx concentrations as dry deposition proxy, and [Formula: see text] and [Formula: see text] in precipitation as wet deposition proxy. For estimation of unmeasured species contributing to dry deposition, we used the CAMx Eulerian photochemical dispersion model, coupled with the Aladin regional numeric weather prediction model. The contribution of fog and dissolved organic nitrogen was estimated using a geostatistical data driven model. We prepared individual maps for particular components applying the most relevant approach and then merged all layers to obtain a final map representing the best estimate of nitrogen deposition over the Czech Republic. Final maps accounting for unmeasured species clearly indicate that the approach used so far may result in a substantial underestimation of nitrogen deposition flux. Our results showed that nitrogen deposition over the Czech forested area in 2008 was well above 2 g N m(-2) yr(-1), with almost 70% of forested area receiving 3-4 g N m(-2) yr(-1). NH3 and gaseous HNO3, contributing about 80%, dominated the dry nitrogen deposition. Estimating the unmeasured nitrogen species by modeled values provides realistic approximations of total nitrogen deposition that also result in more realistic spatial patterns that could be used as input for further studies of likely nitrogen impacts on ecosystems.
