none

Differential Optical Absorption Spectroscopy (DOAS) is a remote sensing technique to detect different trace gas concentrations in the atmosphere. The Multi Axis Differential Optical Absorption Spectroscopy (MAX-DOAS) measurements by the Pandora instrument scan the sky at different elevation angles and main data products include near surface concentration, tropospheric column and vertical profile for different trace gases. It addresses an important gap in near surface air quality measurements that is difficult for in-situ, satellite and other remote sensing measurements to address. Different applications of the MAX-DOAS technique have been presented in this study for improving our understanding of tropospheric chemistry and near surface air quality monitoring. Formaldehyde (HCHO) concentration retrieved from the DOAS technique exhibits significant variation depending on the fitting parameters used. This systematic variation stems from different factors such as uncertainty in molecular absorption cross section measurement, temperature dependence of trace gas absorption, correlation between trace gases and combination of absorbers used in the DOAS fitting. To investigate the sensitivity and systematic uncertainty of HCHO retrieval, different fitting scenarios were created where fitting parameters like wavelength range, polynomial order, offset order and molecular absorption cross section were varied. To minimize systematic uncertainty and provide steady variability, the fitting scenario that most closely resembles the median of the range is selected and recommended as base fitting scenario. In addition, a real time analytical method to calculate HCHO near surface volume mixing ratio is presented where radiative transfer modelling is not required. The HCHO near surface volume mixing ratio calculated by MAX-DOAS is compared with surface HCHO measured by a ground in-situ instrument. The Pandora MAX-DOAS agrees very well with the ground in-situ instrument for the whole campaign (R²= 0.83, slope= 0.92) and provides excellent agreement for clear days (R²= 0.83= 0.88, slope=0.95). Additionally, a methodology is presented for detecting the mixing layer height (MLH) by using Pandora MAX-DOAS vertical water vapor distribution measurements. The wavelet method is applied to detect sharp gradients in the water vapor vertical profiles for estimation of mixing layer height. The Pandora derived mixing layer depth is compared to the estimations from the collocated Ceilometer (Vaisala CL51, EPA) measurements. Pandora MAX-DOAS agrees well with Ceilometer measurements for different time intervals during the day with a correlation coefficient of 0.68 to 0.76. Nitrogen Dioxide (NO₂) and Formaldehyde (HCHO) tropospheric columns and vertical profiles measured at the Hartsfield-Jackson Atlanta International Airport are also presented. Even though anthropogenic emissions decreased severely all over the United States due to Covid lockdown restrictions in 2020, trace gas levels at airports remained relatively same due to continuing air traffic. MAX-DOAS measurements are performed at different azimuth angles which gives a three dimensional representation of NO₂ and HCHO vertical profiles and enables to observe and distinguish air pollution at different directions. These measurements further show the potential of MAX-DOAS measurements for near surface air quality monitoring. / Doctor of Philosophy / MAX-DOAS is a ground based spectroscopic technique which can measure near surface concentration, tropospheric column and vertical distribution of different trace gases. Even though Satellite measurements can provide worldwide coverage, they generally measure only one time per day and have limited knowledge of near surface conditions. MAX-DOAS measurements performed by the NASA Pandora spectrometer systems can be used to provide near surface diurnal variation of different trace gas properties. In this work, different real-time applications of the MAX-DOAS technique are presented. At first, near surface concentration of Pandora MAX-DOAS Formaldehyde (HCHO) observations are validated by comparing with an in-situ instrument. Next, a methodology is presented for detecting the mixing layer height (MLH) by using Pandora MAX-DOAS vertical water vapor distribution measurements. Finally, MAX-DOAS measurements of Nitrogen Dioxide (NO₂) and Formaldehyde (HCHO) concentrations during the COVID-19 pandemic at The Hartsfield-Jackson Atlanta International Airport is presented. The measurements are done at different azimuth angles which produces three dimensional representations of NO₂ and HCHO vertical profiles. All these results prove that the NASA Pandora spectrometer systems have great potential for improving our understanding of tropospheric chemistry and air quality monitoring.

La pollution atmosphérique malodorante provient essentiellement de quatre gaz SO2, H2S, NO2 et NH3. Afin de réduire au mieux ces effets néfastes sur la santé et l'environnement, il faut contrôler en continu les émanations de gaz le plus près possible de la source. Ce qui nécessite un appareil capable de détecter ces gaz polluants, simple d'utilisation, de taille et poids réduits. C'est dans cette optique que nous avons réalisé un nez électronique portable, servant à détecter les quatre gaz cibles déjà cités. La partie sensible de ce prototype est constituée d'une matrice de six capteurs à oxydes métalliques semi-conducteurs, dont nous utilisons la sensibilité croisée. Le nez électronique fonctionne sur le même principe que le nez humain, il doit apprendre à reconnaître une odeur. Cette phase d'apprentissage se déroule au laboratoire où nous envoyons sur les capteurs des mélanges gazeux connus et contrôlés. La réponse des capteurs varie en fonction de la nature du gaz (réducteur ou oxydant) et de leur sensibilité à celui-ci. Puis l'utilisation de méthodes d'analyse de données a prouvé que notre nez électronique peut discriminer un mélange gazeux complexe et le quantifier. Ensuite nous avons placé le nez électronique en situation réelle, en étudiant l'odeur dégagée par des fientes de canards dans une ferme expérimentale. Les résultats obtenus ont montré que cet appareil pouvait détecter de manière fiable les variations d'odeur en fonction des paramètres influents. Ainsi, nous avons réalisé la validation de notre prototype en laboratoire puis sur site. Mais les capteurs utilisés présentent un inconvénient, ils doivent conserver sans interruption leur température de fonctionnement (~ 350°C). Afin de prévenir cette forte consommation d'énergie, nous avons développé des capteurs polymères qui fonctionnent à température ambiante. La caractérisation en laboratoire a montré qu'ils sont sensibles aux gaz cibles étudiés. Leurs réponses à H2S laisse apparaître une bonne stabilité à court et moyen terme, qui permettra de les intégrer dans la matrice après complet développement / The malodorous atmospheric pollution results essentially from four gases SO2, H2S, No2 and NH3. To reduce at best these fatal effects on the health and the environment, it is necessary to control continuously the gas emanations closer to the source, That requires adevice enable to detect these polluant gases, easy to use, with reduced size and weight. In this way, we have realized a portable electronic nose, to detect of the four target gases already mentioned. The sensitive part of this prototype is composed of a matrix of six semi conducting metal oxide sensors, offering a good cross sensivity. The electronic nose mimics the human nose, he has to learn to recognize an odour. This learning phase is realised in the laboratoy by introducing in the sensor cell gas mixtures with controlled composition, The sensor response varies with the nature of the gas (reducing or oxidizing) and their own gas sensitivity. The use of analysis and data methods proves that our electronic nose can well discriminate a complex gas mixture and quantify it. Then, we have placed the electronic nose in a real situation, by studying the odour coming from a duck experimental farm. The obtained results showed that this prototype could well detect the variations of the odour level in accordance with the influent parameters. So, we have realized the laboratory and the real site validation of our electronic nose. But the metal oxide sensors present an inconvenient : they have to keep continuously their working temperature (~ 350°C). To prevent this strong energy consumption, we have developed polymer sensors which work at room temperature. The characterization in laboratory showed that they are sensitive to studied target gases. Their responses to H2S have a good stability in short and middle term, allowing to integrate them into the matrix after complete development

A ground-based, zenith-sky, UV-visible triple grating spectrometer was installed at the Polar Environment Atmospheric Research Laboratory (PEARL) in the Canadian High Arctic during polar springtime from 2004 to 2007 as part of the Canadian Arctic ACE (Atmospheric Chemistry Experiment) Validation Campaigns. From the solar spectra, ozone, NO2, and BrO vertical column densities (VCDs) have been retrieved using the DOAS (Differential Optical Absorption Spectroscopy) technique. This spectrometer, the UT-GBS (University of Toronto Ground-Based Spectrometer), was also deployed as part of the fourth Middle Atmosphere Nitrogen TRend Assessment (MANTRA) campaign in Vanscoy, Saskatchewan in August and September 2004. A near-identical spectrometer, the PEARL-GBS, was permanently installed at PEARL in August 2006 as part of the refurbishment of the laboratory by CANDAC (Canadian Network for the Detection of Atmospheric Change). Since then, the instrument has been making continuous measurements, with the exception of during polar night. Vertical columns of ozone and NO2 can be retrieved year-round. During the 2007 sunrise campaign, differential slant column densities (DSCDs) of OClO and VCDs of BrO were also retrieved. Ozone and NO2 DSCDs and VCDs from the UT-GBS were compared to the DSCDs and VCDs from three other UV-visible, ground-based, grating spectrometers that also participated in the MANTRA and Eureka campaigns. Two methods developed by the UV-visible Working Group of the NDACC (Network for the Detection of Atmospheric Composition Change) were followed. During MANTRA, the instruments were found to partially meet the NDACC standards. The comparisons from Eureka were an improvement on the MANTRA comparisons, and also partially met the NDACC standards. In 2007, the columns from the UT-GBS and PEARL-GBS were compared, and were found to agree within the NDACC standards for both species. Ozone and NO2 VCDs from the ground-based instruments were also compared to integrated partial columns from the ACE-FTS (ACE-Fourier Transform Spectrometer) and ACE-MAESTRO (ACE-Measurements of Aerosol Extinction in the Stratosphere and Troposphere Retrieved by Occultation) on board the ACE satellite. ACE-FTS partial columns were found to agree with the ground-based total columns, while the ACE-MAESTRO partial columns were found to be smaller than expected for ozone and larger than expected for NO2.

A ground-based, zenith-sky, UV-visible triple grating spectrometer was installed at the Polar Environment Atmospheric Research Laboratory (PEARL) in the Canadian High Arctic during polar springtime from 2004 to 2007 as part of the Canadian Arctic ACE (Atmospheric Chemistry Experiment) Validation Campaigns. From the solar spectra, ozone, NO2, and BrO vertical column densities (VCDs) have been retrieved using the DOAS (Differential Optical Absorption Spectroscopy) technique. This spectrometer, the UT-GBS (University of Toronto Ground-Based Spectrometer), was also deployed as part of the fourth Middle Atmosphere Nitrogen TRend Assessment (MANTRA) campaign in Vanscoy, Saskatchewan in August and September 2004. A near-identical spectrometer, the PEARL-GBS, was permanently installed at PEARL in August 2006 as part of the refurbishment of the laboratory by CANDAC (Canadian Network for the Detection of Atmospheric Change). Since then, the instrument has been making continuous measurements, with the exception of during polar night. Vertical columns of ozone and NO2 can be retrieved year-round. During the 2007 sunrise campaign, differential slant column densities (DSCDs) of OClO and VCDs of BrO were also retrieved. Ozone and NO2 DSCDs and VCDs from the UT-GBS were compared to the DSCDs and VCDs from three other UV-visible, ground-based, grating spectrometers that also participated in the MANTRA and Eureka campaigns. Two methods developed by the UV-visible Working Group of the NDACC (Network for the Detection of Atmospheric Composition Change) were followed. During MANTRA, the instruments were found to partially meet the NDACC standards. The comparisons from Eureka were an improvement on the MANTRA comparisons, and also partially met the NDACC standards. In 2007, the columns from the UT-GBS and PEARL-GBS were compared, and were found to agree within the NDACC standards for both species. Ozone and NO2 VCDs from the ground-based instruments were also compared to integrated partial columns from the ACE-FTS (ACE-Fourier Transform Spectrometer) and ACE-MAESTRO (ACE-Measurements of Aerosol Extinction in the Stratosphere and Troposphere Retrieved by Occultation) on board the ACE satellite. ACE-FTS partial columns were found to agree with the ground-based total columns, while the ACE-MAESTRO partial columns were found to be smaller than expected for ozone and larger than expected for NO2.

Soot aerosols are known to be an important atmospheric constituent. The physical and chemical properties of soot allows it to act as a precursor of gas-surface heterogeneous reactions, providing active sites for the reduction and oxidation of trace species in the atmosphere, potentially affecting atmospheric composition. In this work the heterogeneous reaction of NO2 on soot leading to nitrous acid (HONO) formation was studied through a series of kinetic uptake experiments and HONO yield measurements. The soot was collected from a diffusion flame using propane and kerosene fuels using two different methods. A low pressure fast-flow reactor coupled to a Chemical Ionization Mass Spectrometer (CIMS) was used to monitor NO2 and HONO signals evolution using atmospheric-level NO2 concentration. HONO yields up to 100 percent were measured and NO2 uptake coefficients varying from 5.6x10-6 to 1.6x10-4 were obtained. Heating soot samples before exposure to NO2 increased HONO yield and the NO2 uptake coefficient on soot due to the removal of the organic fraction from the soot backbone unblocking active sites, which become accessible for the heterogeneous reaction. From the kinetic uptake curves and the effect observed in the HONO yield and NO2 uptake coefficient measurements by heating the soot samples, our results support a complex oxidation-reduction mechanism of reaction. This heterogeneous reaction mechanism involves a combination of competitive adsorptive and reductive centers on soot surface where NO2 is converted into HONO, and the presence of processes on soot where HONO can be decomposed producing other products. Atmospheric soot "aging" effect on the reactivity of soot toward NO2 and HONO yield was studied by coating the soot surface with glutaric acid, succinic acid, sulfuric acid, and pyrene. Glutaric and succinic acid increased both HONO yield and the NO2 uptake coefficients, while sulfuric acid decreased both. However, pyrene did not show any particular trend.

A novel method was developed for producing high-resolution (∼ 11km) maps of surface NO2 concentrations by combining satellite retrieved NO2 columns from OMI with in situ measurements made by permanent monitoring networks. Field data from the BAQS-met field campaign in the Windsor area during 2007 was used to validate this method and explore the uncertainties and biases in the inferred values. Good correlation with the network of passive monitors was found (R = 0.69). Interference of NOz in traditional NO2 measurements was found to be small (0.9 ppb) when considered for 24-hr averages. The inference method was extended to qualitatively analyze long-term trends in Windsor. Comparison against a land-use regression model in Toronto showed similar overall trends, but the downtown core was underestimated considerably by the OMI-inferred map. While the presented inference method can simplify and increase the utility of OMI NO2 data, limitations remain as to the spatial and temporal resolution achievable.

A novel method was developed for producing high-resolution (∼ 11km) maps of surface NO2 concentrations by combining satellite retrieved NO2 columns from OMI with in situ measurements made by permanent monitoring networks. Field data from the BAQS-met field campaign in the Windsor area during 2007 was used to validate this method and explore the uncertainties and biases in the inferred values. Good correlation with the network of passive monitors was found (R = 0.69). Interference of NOz in traditional NO2 measurements was found to be small (0.9 ppb) when considered for 24-hr averages. The inference method was extended to qualitatively analyze long-term trends in Windsor. Comparison against a land-use regression model in Toronto showed similar overall trends, but the downtown core was underestimated considerably by the OMI-inferred map. While the presented inference method can simplify and increase the utility of OMI NO2 data, limitations remain as to the spatial and temporal resolution achievable.

We present joint analyses of satellite-observed combustion products to examine bulk characteristics of combustion in megacities and fire regions. We use retrievals of CO, NO2 and CO2 from NASA/Terra Measurement of Pollution In The Troposphere, NASA/Aura Ozone Monitoring Instrument, and JAXA Greenhouse Gases Observing Satellite to estimate atmospheric enhancements of these co-emitted species based on their spatiotemporal variability (spread, sigma) within 14 regions dominated by combustion emissions. We find that patterns in sigma(XCO)/sigma(XCO2) and sigma(XCO)/sigma(XNO2) are able to distinguish between combustion types across the globe. These patterns show distinct groupings for biomass burning and the developing/developed status of a region that are not well represented in global emissions inventories. We show here that such multi-species analyses can provide constraints on emission inventories, and be useful in monitoring trends and understanding regional-scale combustion.

京都大学 / 新制・課程博士 / 博士(農学) / 甲第23525号 / 農博第2472号 / 新制||農||1087(附属図書館) / 学位論文||R3||N5356(農学部図書室) / 京都大学大学院農学研究科森林科学専攻 / (主査)教授 梅村 研二, 教授 矢野 浩之, 教授 仲村 匡司 / 学位規則第4条第1項該当 / Doctor of Agricultural Science / Kyoto University / DFAM