A wireless electronic nose network system has been developed for monitoring and analyzing livestock farm odour. The system utilizes electronic noses (e-noses) that can measure odour compounds and environment factors such as temperature and humidity. The e-noses are deployed at various locations on the farm, and sensor signals are transmitted via a wireless communication to a central station, where the data processing and sensor fusion algorithms analyze the collected odour data, compute the odour concentration, and display the odour dispersion plume. This system would provide users with convenient odour monitoring capabilities and help the development of an effective overall odour management strategy. In addition, an adaptive neuro-fuzzy inference approach is proposed to calibrate the e-nose responses to human panelists' perception. The proposed method can handle non-numeric information and human expert knowledge in livestock farm odour models, and can adjust the parameters in a systematic manner for optimal system performance. The proposed approach has been tested against a livestock farm odour database. Several livestock farm odour models have been developed for comparative studies. The results show that the proposed approach provides a more accurate odour prediction than a typical multi-layer feedforward neural network. Furthermore, to model odour dispersion around livestock facilities, a biologically inspired odour dispersion model is proposed, and is tested using computer simulations and a livestock farm odour database. Results show that the proposed approach is effective in providing accurate modelling of odour dispersion from multiple and various types of odour sources in both static and non-static environments.

No description available.

This manuscript based thesis examines and advances methods for constraining sources of atmospheric trace constituents with a Lagrangian particle dispersion model. The method of Bayesian inversion is demonstrated, and a new method is introduced to a class of similar problems where established methods are not applicable. First, A new regression based methodology was developed and applied to observations of atmospheric methanesulfonic acid mass concentrations at Alert, Nunavut. The methodology was used to compare the importance of phytoplankton blooms vs. the ice-free ocean as sources of the dimethylsulfide precursor, and to compare the importance of bromine monoxide vs. hydroxyl as agents oxidizing dimethylsul de to methanesulfonic acid. These issues are relevant to the application of methanesulfonic acid concentrations in ice cores to determine historic sea ice properties. The analysis indicated that source regions to Alert during the spring are primarily ice-free ocean with a significant contribution from ice edge blooms, and during the summer to be dominated by the ice-free ocean. The model also indicated that oxidation of DMS by BrO was the dominant source of MSA in the spring, while DMS oxidation by OH was the dominant source in the summer. Secondly, Bayesian inversion was applied to observations of atmospheric elemental carbon mass concentrations at Tsinghua University in Beijing, China. The analysis provided evidence that current bottom-up elemental carbon emissions estimates in northern China are likely underpredicted. Global chemical transport models show ubiquitous underestimates of the atmospheric burden of elemental carbon, especially near large sources of emissions. Northern China is among the regions with the most intensive elemental carbon emissions in the world, and an underestimate of emissions in this region may be partially responsible for the global chemical transport model underestimates.

Bivariate or multivariate correlated data may be collected on a sample of unit in many applications. When the experimenters concern about the failure times of two related subjects for example paired organs or two chronic diseases, the bivariate binary data is often acquired. This type of data consists of a observation point x and indicators which represent whether the failure times happened before or after the observation point. In this work, the observed bivariate data can be written with the following form {x, £_1=I(X1≤ x), £_2=I(X2≤ x)}.The corresponding optimal design problems for parameter estimation under this type of bivariate data are discussed. For this kind of the multivariate responses with explanatory variables, their marginal distributions may be from different distributions. Copula model is a way to formulate the relationship of these responses, and the association between pairs of responses. Copula models for bivariate binary data are considered useful in practice due to its flexibility. In this dissertation for bivariate binary data, the marginal functions are assumed from exponential or Weibull distributions and two assumptions, independent or correlated, about the joint function between variables are considered. When the bivariate binary data is assumed correlated, the Clayton copula model is used as the joint cumulative distribution function. There are few works addressed the optimal design problems for bivariate binary data with copula models. The D-optimal designs aim at minimizing the volume of the confidence ellipsoid for estimating unknown parameters including the association parameter in bivariate copula models. They are used to determine the best observation points. Moreover, the Ds-optimal designs are mainly used for estimation of the important association parameter in Clayton model. The D- and Ds-optimal designs for the above copula model are found through the general equivalence theorem with numerical algorithm. Under different model assumptions, it is observed that the number of support points for D-optimal designs is at most as the number of model parameters for the numerical results. When the difference between the marginal distributions and the association are significant, the association becomes an influential factor which makes the number of supports gets larger. The performances of estimation based on optimal designs are reasonably well by simulation studies. In survival experiments, the experimenter customarily takes trials at some specific points such as the position of the 25, 50 and 75 percentile of distributions. Hence, we consider the design efficiencies when the design points for trials are at three or four particular percentiles. Although it is common in practice to take trials at several quantile positions, the allocations of the proportion of sample size also have great influence on the experimental results. To use a locally optimal design in practice, the prior information for models or parameters are needed. In case there is not enough prior knowledge about the models or parameters, it would be more flexible to use sequential experiments to obtain information in several stages. Hence with robustness consideration, a sequential procedure is proposed by combining D- and Ds-optimal designs under independent or correlated distribution in different stages of the experiment. The simulation results based on the sequential procedure are compared with those by the one step procedures. When the optimal designs obtained from an incorrect prior parameter values or distributions, those results may have poor efficiencies. The sample mean of estimators and corresponding optimal designs obtained from sequential procedure are close to the true values and the corresponding efficiencies are close to 1. Huster (1989) analyzed the corresponding modeling problems for the paired survival data and applied to the Diabetic Retinopathy Study. Huster (1989) considered the exponential and Weibull distributions as possible marginal distributions and the Clayton model as the joint function for the Diabetic Retinopathy data. This data was conducted by the National Eye Institute to assess the effectiveness of laser photocoagulation in delaying the onset of blindness in patients with diabetic retinopathy. This study can be viewed as a prior experiment and provide the experimenter some useful guidelines for collecting data in future studies. As an application with Diabetic Retinopathy Study, we develop optimal designs to collect suitable data and information for estimating the unknown model parameters. In the second part of this work, the optimal design problems for parameter estimations are considered for the type of proportional data. The nonlinear model, based on Jorgensen (1997) and named the dispersion model, provides a flexible class of non-normal distributions and is considered in this research. It can be applied in binary or count responses, as well as proportional outcomes. For continuous proportional data where responses are confined within the interval (0,1), the simplex dispersion model is considered here. D-optimal designs obtained through the corresponding equivalence theorem and the numerical results are presented. In the development of classical optimal design theory, weighted polynomial regression models with variance functions which depend on the explanatory variable have played an important role. The problem of constructing locally D-optimal designs for simplex dispersion model can be viewed as a weighted polynomial regression model with specific variance function. Due to the complex form of the weight function in the information matrix is considered as a rational function, an approximation of the weight function and the corresponding optimal designs are obtained with different parameters. These optimal designs are compared with those using the original weight function.

This manuscript based thesis examines and advances methods for constraining sources of atmospheric trace constituents with a Lagrangian particle dispersion model. The method of Bayesian inversion is demonstrated, and a new method is introduced to a class of similar problems where established methods are not applicable. First, A new regression based methodology was developed and applied to observations of atmospheric methanesulfonic acid mass concentrations at Alert, Nunavut. The methodology was used to compare the importance of phytoplankton blooms vs. the ice-free ocean as sources of the dimethylsulfide precursor, and to compare the importance of bromine monoxide vs. hydroxyl as agents oxidizing dimethylsul de to methanesulfonic acid. These issues are relevant to the application of methanesulfonic acid concentrations in ice cores to determine historic sea ice properties. The analysis indicated that source regions to Alert during the spring are primarily ice-free ocean with a significant contribution from ice edge blooms, and during the summer to be dominated by the ice-free ocean. The model also indicated that oxidation of DMS by BrO was the dominant source of MSA in the spring, while DMS oxidation by OH was the dominant source in the summer. Secondly, Bayesian inversion was applied to observations of atmospheric elemental carbon mass concentrations at Tsinghua University in Beijing, China. The analysis provided evidence that current bottom-up elemental carbon emissions estimates in northern China are likely underpredicted. Global chemical transport models show ubiquitous underestimates of the atmospheric burden of elemental carbon, especially near large sources of emissions. Northern China is among the regions with the most intensive elemental carbon emissions in the world, and an underestimate of emissions in this region may be partially responsible for the global chemical transport model underestimates.

CO2 emission is very important, because it is responsible for about 60% of the &quot / Greenhouse Effect&quot / . The major objectives of this study were to prepare a CO2 emission inventory of Turkey based on districts and provinces by using the fuel consumption data with respect to its sources, to find the CO2 uptake rate of forests in Turkey based on provinces and districts, and to estimate the ground level concentration of CO2 across Turkey using U.S. EPA&#039 / s ISCLT3 model for the preparation of ground level concentration maps. The basic sources of the CO2 emission were taken as households, manufacturing industries, thermal power plants and road vehicles. The sinks of the CO2 were forests. The CO2 uptake by forests was calculated using the annual increment of forest biomass. The results of the CO2 emission inventory conducted in this study between the years 1990 and 2003 showed that the CO2 emission in 1990 was 142.45 million tones/year and the highest emission was calculated in 2000 with a value of 207.97 million tones/year. The regional distribution of CO2 emission showed that the Marmara Region emits the highest regional CO2 emission throughout the years with an average value of 54.76 million tones/year. It was also calculated that Marmara and Aegean Regions are responsible for half of the CO2 emission of Turkey. The results of the CO2 uptake calculations showed that the CO2 uptake of forests in the coastal zone was higher that that in the inland zone. The CO2 uptake in the Central Anatolia, Eastern Anatolia and South-Eastern Anatolia Regions were 2.6, 1.9 and 1.1 million tones/year, respectively. The maximum CO2 uptake is in the Black Sea Region with a value of 16.4 million tones/year. The highest ground level CO2 concentartions without any sink effect were always obtained in the Marmara Region. However, the forest areas in this region decrease the concentrations considerably. The dispersion model performance is determined highly without the results of the year 2002.

La dispersion et le transport atmosphérique d'ammoniac (NH3) émis par les sources agricoles et son dépôt sec sur le sol et la végétation peuvent entraîner la dégradation des écosystèmes sensibles, ainsi que l'acidification des sols. Les concentrations atmosphériques et les dépôts secs de NH3 sont généralement plus élevés à côté des sources d'émission, et les impacts environnementaux sur les écosystèmes sensibles sont souvent les plus importants à ces endroits. Pour mieux évaluer les impacts et leur étendue à l’échelle de paysages agricoles, des revues scientifiques sur les méthodes d'évaluation des impacts de NH3 ont recommandé une comparaison entre différents types de modèles à différentes échelles spatiales. Dans ce contexte, nous avons comparé les flux de NH3 simulés par deux modèles de dispersion, de transport et de dépôt par voie atmosphérique (CHIMERE et OPS-ST), pour différents scénarios théoriques et semi-réels et pour différentes tailles de maille des modèles. Les résultats de simulations montrent que les dépôts secs annuels de NH3, moyennés sur le domaine d’étude, sont comparables pour des tailles de maille correspondant aux tailles pour lesquelles les modèles ont été conçus. Cela implique que les modèles eulériens fonctionnant à des résolutions kilométriques peuvent être utilisés pour simuler l’impact du NH3 sur la composition chimique de l’atmosphère. Les résultats divergent entre les modèles à proximité des sources. Les conditions météorologiques et la taille des mailles sont les facteurs ayant les effets les plus marqués sur les résultats des deux modèles. Enfin, nos résultats montrent que la détection des dépassements de charges critiques à proximité des sources est mieux représentée avec OPS-ST qu’avec CHIMERE, avec lequel on observe une surestimation des surfaces dépassant les charges critiques. / Short-range atmospheric dispersion of ammonia emitted by agricultural sources and its subsequent deposition to soil and vegetation can lead to the degradation of sensitive ecosystems as well as soil acidification. Atmospheric concentrations and dry depositions rates of NH3 are generally higher near the emission source and environmental impacts on sensitive ecosystems are often largest at these locations. To better evaluate the impacts and their extent at the agricultural landscapes scale, scientific reviews of NH3 assessment methods recommended a comparison between different types of models at different spatial scales. In this context, we compared NH3 fluxes simulated by two atmospheric dispersion, transport and deposition models (CHIMERE and OPS-ST) for different theoretical and semi-real scenarios and for different grid cell sizes. The simulation results show that the averaged NH3 dry deposition over the investigated domains is comparable for grid cell sizes for which the models were designed. This implies that eulerian models with a km scale horizontal resolution can be used for studying the larger scale impact of NH3 on atmospheric composition despite an inadequate treatment of near source dry deposition, for which the models diverge. Meteorological conditions and grid cell size are the factors having the strongest effects on the results of the two models. NH3 dry deposition predictions can be used to map critical load exceedances, our results show that the detection of these exceedances near to the sources is better represented with OPS-ST than with CHIMERE, the latter showing an overestimation of the surfaces exceeding critical loads.

Dust aerosol emission is a critical topic in agriculture, occurring either by aeolian process from bare or sparsely vegetated cropland or as fugitive emission during tilling, harvest and many other farming activities. Aerosols, which are in the case of agriculture either mineral dust, organic particles or a mixture, are known for impacting human health, cloud formation and ultimately, the earth’s climate and ecosystem. Coupled atmosphere and aerosol transport models are commonly used to study aerosol dispersion in the atmosphere, but so far, agricultural sources are under-represented. Hence, estimations of these emissions’ actual impact are still somewhat uncertain regarding their seasonality, spatial distribution and the fraction of the global aerosol load. To fill this gap, this study aims at identifying suitable approaches for modelling aeolian emissions from sparsely vegetated cropland and fugitive emissions from tilling. Fugitive emissions are challenging since they mainly depend on human activity that is not predictable, but observed events can be used as case studies. For this, a Lagrangian particle dispersion model was chosen, which can trace the trajectory of individual particles in the emitted dust plume. So the particle model “Itpas” was developed to tackle fugitive emissions and to be capable of simulating the complex turbulent mixing of dust particles inside the atmospheric boundary layer. This model was used to simulate a case study based on measured tilling emissions, showing the particle dispersion for a stable and unstable stratified boundary layer. It was shown that within a stably stratified boundary layer, the dust plume is restricted to the near-source region. In contrast, emissions in unstable boundary layers go into long-range transport. This illustrates the spatial range a single tillage operation can have an impact. Aeolian dust emissions are controlled by the wind. For cropland, the emission variability is caused mainly by the frequently changing vegetation cover. Emissions can only occur in the time between tillage and newly grown crops or during drought periods. A parametrisation based on high-resolution satellite observations of the vegetation cover was created to include this process into a model. With this, a new dust emission scheme for cropland emission was developed for the model system COSMO-MUSCAT. In a case study of a dust outbreak from cropland in Poland in 2019, the model’s ability was tested extensively on multiple spatial resolutions. Validation against satellite-measured AOD, ground-measured PM10 and the vertical profile of the PollyNET lidar in Warsaw showed an overall good agreement of the model simulation with the observations. In the framework of this thesis, one dedicated model approach was developed for both the fugitive emissions and the aeolian emissions and validated upon case studies. These approaches could help better understand agricultural dust emissions, their spatial distribution, seasonality and, ultimately, global impact.

No description available.

The dispersion model was used to study mixing levels in a full scale aeration tank. The effect of air flow rate, water flow rate and diffuser type was investigated. The peak time technique proved satisfactory in predicting the theoretical tracer response curve generated using the dispersion model. The dispersion model adequately described the longitudinal mixing that occurred in a full scale aeration tank equipped with fine and coarse bubble air diffusers. Response curves from two tanks-in-series were also obtained. / Thesis / Master of Engineering (ME)