Search

International audience; Exhaust particle emissions are mostly made of black carbon and/or organic compounds, with some of these organic compounds existing in both the gas and particle phases. Although emissions of volatile organic compounds (VOC) are usually measured at the exhaust, emissions in the gas phase of lower volatility compounds (POAvapor) are not. However, these gas-phase emissions may be oxidised after emission and enhance the formation of secondary organic aerosols (SOA). They are shown here to contribute to most of the SOA formation in Central Paris. POAvapor emissions are usually estimated from primary organic aerosol emissions in the particle phase (POA). However, they could also be estimated from VOC emissions for both gasoline and diesel vehicles using previously published measurements from chamber measurements. Estimating POAvapor from VOC emissions and ageing exhaust emissions with a simple model included in the Polyphemus air-quality platform compare well to measurements of SOA formation performed in chamber experiments. Over Greater Paris, POAvapor emissions estimated using POA and VOC emissions are compared using the HEAVEN bottom-up traffic emissions model. The impact on the simulated atmospheric concentrations is then assessed using the Polyphemus/Polair3D chemistry-transport model. Estimating POAvapor emissions from VOC emissions rather than POA emissions lead to lower emissions along motorway axes (between −50% and −70%) and larger emissions in urban areas (up to between +120% and +140% in Central Paris). The impact on total organic aerosol concentrations (gas plus particle) is lower than the impact on emissions: between −8% and 25% along motorway axes and in urban areas respectively. Particle-phase organic concentrations are lower when POAvapor emissions are estimated from VOC than POA emissions, even in Central Paris where the total organic aerosol concentration is higher, because of different assumptions on the emission volatility distribution, stressing the importance of characterizing not only the emission strength, but also the emission volatility distribution.

International audience; As part of the Chemistry-Aerosol Mediterranean Experiment (ChArMEx), a series of aerosol and gas-phase measurements were deployed aboard the SAFIRE ATR42 research aircraft in summer 2014. The present study focuses on the four flights performed in late June early July over two forested regions in the south of France. We combine in situ observations and model simulations to aid in the understanding of secondary organic aerosol (SOA) formation over these forested areas in the Mediterranean and to highlight the role of different gas-phase precursors. The non-refractory particulate species measured by a compact aerosol time-of-flight mass spectrometer (cToF-AMS) were dominated by organ-ics (60 to 72 %) followed by a combined contribution of 25 % by ammonia and sulfate aerosols. The contribution from nitrate and black carbon (BC) particles was less than 5 % of the total PM 1 mass concentration. Measurements of non-refractory species from off-line transmission electron mi-croscopy (TEM) showed that particles have different mixing states and that large fractions (35 %) of the measured particles were organic aerosol containing C, O, and S but without inclusions of crystalline sulfate particles. The organic aerosol measured using the cToF-AMS contained only evidence of oxidized organic aerosol (OOA), without a contribution of fresh primary organic aerosol. Positive matrix factorization (PMF) on the combined organic-inorganic matrices separated the oxidized organic aerosol into a more-oxidized organic aerosol (MOOA), and a less-oxidized organic aerosol (LOOA). The MOOA component is associated with inorganic species and had higher contributions of m/z 44 than the LOOA factor. The LOOA factor is not associated with inorganic species and correlates well with biogenic volatile organic species measured with a proton-transfer-reaction mass spectrometer, such as isoprene and its oxidation products (methyl vinyl ketone, MVK; methacroleine, MACR; and iso-prene hydroxyhydroperoxides, ISOPOOH). Despite a significantly high mixing ratio of isoprene (0.4 to 1.2 ppbV) and its oxidation products (0.2 and 0.8 ppbV), the contribution of specific signatures for isoprene epoxydiols SOA (IEPOX-SOA) within the aerosol organic mass spectrum (m/z 53 and m/z 82) were very weak, suggesting that the presence of Published by Copernicus Publications on behalf of the European Geosciences Union. 7042 E. Freney et al.: Aerosol composition and the contribution of SOA formation isoprene-derived SOA was either too low to be detected by the cToF-AMS, or that SOA was not formed through IEPOX. This was corroborated through simulations performed with the Polyphemus model showing that although 60 to 80 % of SOA originated from biogenic precursors, only about 15 to 32 % was related to isoprene (non-IEPOX) SOA; the remainder was 10 % sesquiterpene SOA and 35 to 40 % monoter-pene SOA. The model results show that despite the zone of sampling being far from industrial or urban sources, a total contribution of 20 to 34 % of the SOA was attributed to purely anthropogenic precursors (aromatics and intermediate or semi-volatile compounds). The measurements obtained during this study allow us to evaluate how biogenic emissions contribute to increasing SOA concentrations over Mediterranean forested areas. Directly comparing these measurements with the Polyphemus model provides insight into the SOA formation pathways that are prevailing in these forested areas as well as processes that need to be implemented in future simulations.

In the next few years, numerous satellites with high-resolution instruments dedicated to the imaging of atmospheric gaseous compounds will be launched, to finely monitor emissions of greenhouse gases and pollutants. Processing the resulting images of plumes from cities and industrial plants to infer the emissions of these sources can be challenging. In particular traditional atmospheric inversion techniques, relying on objective comparisons to simulations with atmospheric chemistry transport models, may poorly fit the observed plume due to modelling errors rather than due to uncertainties in the emissions. The present article discusses how these images can be adequately compared to simulated concentrations to limit the weight of modelling errors due to the meteorology used to analyse the images. For such comparisons, the usual pixel-wise ℒ2 norm may not be suitable, since it does not linearly penalise a displacement between two identical plumes. By definition, such a metric considers a displacement as an accumulation of significant local amplitude discrepancies. This is the so-called double penalty issue. To avoid this issue, we propose three solutions: (i) compensate for position error, due to a displacement, before the local comparison; (ii) use non-local metrics of density distribution comparison; and (iii) use a combination of the first two solutions. All the metrics are evaluated using first a catalogue of analytical plumes and then more realistic plumes simulated with a mesoscale Eulerian atmospheric transport model, with an emphasis on the sensitivity of the metrics to position error and the concentration values within the plumes. As expected, the metrics with the upstream correction are found to be less sensitive to position error in both analytical and realistic conditions. Furthermore, in realistic cases, we evaluate the weight of changes in the norm and the direction of the four-dimensional wind fields in our metric values. This comparison highlights the link between differences in the synoptic-scale winds direction and position error. Hence the contribution of the latter to our new metrics is reduced, thus limiting misinterpretation. Furthermore, the new metrics also avoid the double penalty issue.

International audience; The purpose of this paper is to propose a time-step-robust cell-to-cell integration of particle trajectories in 3-D unstructured meshes in particle/mesh Lagrangian stochastic methods. The main idea is to dynamically update the mean fields used in the time integration by splitting, for each particle, the time step into sub-steps such that each of these sub-steps corresponds to particle cell residence times. This reduces the spatial discretization error. Given the stochastic nature of the models, a key aspect is to derive estimations of the residence times that do not anticipate the future of the Wiener process. To that effect, the new algorithm relies on a virtual particle, attached to each stochastic one, whose mean conditional behavior provides free-of-statistical-bias predictions of residence times. After consistency checks, this new algorithm is validated on two representative test cases: particle dispersion in a statistically uniform flow and particle dynamics in a non-uniform flow.; L’objectif de ce papier est de proposer un algorithme robuste en pas de temps permettant l’intégration face à face de la trajectoire de particules dans des maillage 3-D non-structurés pour des méthodes lagrangiennes stochastiques. L’idée principale est de mettre à jour de façon dynamique les champs moyens utilisés dans l’intégration temporelle. Pour ce faire, pour chaque particule, on subdivise le pas de temps en sous pas de temps, de façon à ce que chacun de ceux-ci correspondent au temps de résidence de la particule dans une cellule. Ce faisant on réduit l’erreur de discrétisation spatiale. Étant donnée la nature stochastique du modèle, un aspect prépondérant est de pouvoir dériver des estimations du temps de résidence des particules dans les cellules qui n’anticipent pas le futur des processus de Wiener. Dans ce but, le nouvel algorithme est basé sur une particule virtuelle attachée à chaque particule stochastique, dont le comportement conditionné moyen permet d’obtenir des temps de résidence qui ne sont entachés d’aucun biais statistique. Après une vérification de consistance, ce nouvel algorithme est validé sur deux cas de tests représentatifs : une dispersion de polluants dans un écoulement statistiquement uniforme et un écoulement non-uniforme dans lequel on traque la dynamique des particules.

International audience; A stable scheme is proposed in this paper in order to obtain approximate solutions of second-moment turbulent models for incompressible flows with or without thermal transport equation. The analysis of the convective terms, which includes the solution of the associated Riemann problem, enables to propose a standard projection scheme, and to get rid of spurious oscillations.

With the growing evidence that high particle number concentrations may impact health, modelling their emissions and understanding formation processes is necessary, especially in cities where many people are exposed. As emission inventories of particle numbers and size distribution over cities are usually not available, a methodology is defined to estimate them from PM2.5 emissions and ratios of PM1 / PM2.5 and PM0.1 / PM2.5 by activity sector. In this methodology, a fitting parameter αem is used to redistribute the number concentrations in the lowest emission diameter range. This parameter is chosen by comparing measured and simulated number concentrations during non-nucleation days. The emission size distribution is then finely discretised by conserving both mass and number in each of the size ranges where emissions are specified. The methodology is applied over Greater Paris during the MEGAPOLI campaign (July 2009). Three-dimensional simulations are performed using the chemistry transport model Polair3D/Polyphemus coupled to the aerosol module SSH-aerosol to represent the evolution of particles by condensation, evaporation, coagulation, and nucleation, with a sectional approach for the size distribution. The model is first compared to measurements during non-nucleation days, and the influence over the month of July 2009 of three different nucleation parameterisations is assessed, i.e. binary (sulfuric acid, water), ternary (sulfuric acid, ammonia, water), and heteromolecular (extremely low-volatility organic compounds (ELVOCs) from monoterpenes and sulfuric acid). The modelled number concentrations compare very well to measurements, with an average normalised mean error of 42 % for the daily number concentrations of particles larger than 10 nm and 37 % for the number concentrations of particles larger than 100 nm. The influence of the binary nucleation is low, and the ternary nucleation scheme leads to better simulated number concentrations (in terms of bias and error) at only one site out of three, but it systematically reduces the model to measurement correlation, suggesting that ternary nucleation may not be the dominant process in new particle formation. However, the relative bias and error, as well as the correlation at suburban sites, are systematically improved using the heteromolecular nucleation scheme involving sulfuric acid and ELVOCs from monoterpenes. This suggests that heteromolecular nucleation may be important in cities, especially at suburban sites in summer, and that a better characterisation of the emissions of ELVOC precursors from traffic is needed.

International audience; The Paris region (France) is increasingly the focus of urban atmospheric research. Numerous national and international research projects have chosen Europe’s largest metropolitan region as their study area to better understand and predict critical hazards (incl. heat, air pollution, thunderstorms) in the context of a changing climate. Located on rather flat terrain in continental, mid-latitude climates, the densely populated Paris region is very suitable for the evaluation of urban processes in numerical simulations at different scales. The European research infrastructures ACTRIS and ICOS are developing strategies for the improved operational monitoring of air pollution and greenhouse gas budgets, respectively. Various research projects are conducting fundamental process studies and model developments to investigate the dynamics and chemistry of the urban atmosphere and its interactions with the rural surroundings and regional-scale flow to better quantify associated health risks and inform sustainable planning.In addition to numerous modelling activities (chemistry-transport, numerical weather prediction, climate projections), diverse atmospheric observations are collected. These include dense surface station networks, turbulent flux towers, and ground-based atmospheric remote sensing to monitor the atmospheric boundary layer. This multi-project context motivates the pooling of resources.To facilitate efficient project synergy and to optimise the coordination of the individual experimental campaigns, the PANAME initiative (https://paname.aeris-data.fr/) was established. PANAME provides a framework to optimise the design of the Paris region measurement network and helps to standardise the operations. A professional, multi-disciplinary data portal is developed at the French AERIS atmospheric data centre to host the PANAME observations and model results. Here, data are collected and formatted, standardised advanced products are derived from the diverse sensor networks and high-quality visualisations are generated in near real-time. The presentation will provide an overview on the scientific objectives of the on-going projects, the deployment of measurements and simulation tools, and the data portal design.

International audience; We report on the development and validation of a new Numerical Wave Tank (NWT) solving fully nonlinear potential flow (FNPF) equations, as a more efficient variation of Grilli et al.'s NWT [Grilli et al., A fully nonlinear model for three-dimensional overturning waves over arbitrary bottom, International Journal for Numerical Methods in Fluids 35 (2001) 829-867], which was successful at modeling many wave phenomena, including landslide-generated tsunamis, rogue waves, and the initiation

International audience; Highlights• Study of an intense winter episode of particulate pollution in the Paris region.• Exceptional stagnant conditions explain the highest concentrations.• Organic matter of local origin dominates the chemical composition of the particles with also a strong nitrate component from traffic.• Significant values of oxidative potential are observed.

Homogeneous reactivity has been extensively studied in recent years through outdoor air-quality simulations. However, indoor atmospheres are known to be largely influenced by another type of chemistry, which is their reactivity with surfaces. Despite progress in the understanding of heterogeneous reactions, such reactions remain barely integrated into numerical models. In this paper, a room-scale, indoor air-quality (IAQ) model is developed to represent both heterogeneous and homogeneous chemistry. Thanks to the introduction of sorbed species, deposition and surface reactivity are treated as two separate processes, and desorption reactions are incorporated. The simulated concentrations of inorganic species are compared with experimental measurements acquired in a real room, thus allowing calibration of the model's undetermined parameters. For the duration of the experiments, the influence of the simulation's initial conditions is strong. The model succeeds in simulating the four inorganic species concentrations that were measured, namely NO, NO2, HONO and O3. Each parameter is then varied to estimate its sensitivity and to identify the most prevailing processes. The air-mixing velocity and the building filtration factor are uncertain parameters that appear to have a strong influence on deposition and on the control of transport from outdoors, respectively. As expected, NO2 surface hydrolysis plays a key role in the production of secondary species. The secondary production of NO by the reaction of sorbed HONO with sorbed HNO3 stands as an essential component to integrate into IAQ models.

International audience; While in tropical soils recycling of plant phytoliths has been shown to represent a major pool of Si available for plants, the main Si pools available for plants in temperate soils are still poorly constrained. We characterised the various Si pools of temperate forested and cultivated soils in France by selecting seven paired sites (adjacent wheat and forest plots) for four soil groups: Luvisols, Albeluvisols, calcaric and hypereutric Cambisols. We showed that CaCl2-extracted Si (bioavailable) is mostly controlled by pH and allophanes (short-range ordered aluminosilicates) but not by phytoliths. Cultivation, by decreasing the soil organic carbon content increases the allophane content and the SiCaCl2 concentration. Our work highlights the importance of allophanes as the missing link between the soil solution and the clay minerals that could explain the reported correlation between the clay mineral content and bioavailable Si.

International audience; Ensemble variational methods are being increasingly used in the field of geophysical data assimilation. Their efficiency comes from the combined use of ensembles, which provide statistics estimates, and a variational analysis, which handles nonlinear operators through iterative optimization techniques. Taking model error into account in four-dimensional ensemble variational algorithms is challenging because the state trajectory over the data assimilation window (DAW) is no longer determined by its sole initial condition. In particular, the control variable dimension scales with the DAW length, which yields a high numerical complexity. This is unfortunate since accuracy improvement is expected with longer DAWs. Building upon the work of [P. Sakov and M. Bocquet, Tellus A, 70 (2018), 1414545], this paper discusses how to algorithmically construct and numerically test an iterative ensemble Kalman smoother with additive model error (IEnKS-Q) which is thought to be the natural weak constraint generalization of the IEnKS [M. Bocquet and P. Sakov, Quart. J. Roy. Meteorol. Soc., 140 (2014), pp. 1521--1535], as well as the generalization of IEnKF-Q [P. Sakov, J. Haussaire, and M. Bocquet, Quart. J. Roy. Meteorol. Soc., 144 (2018), pp. 1297--1309] to general DAWs. The number of model evaluations per cycle of the IEnKS-Q is also examined. Solutions based on perturbation decomposition are proposed to dissociate those numerically costly evaluations from the control variable dimension.

International audience The ozone–NOy photochemistry is explored in contrasting polluted plumes sampled with the Safire ATR 42 research aircraft during three summer field international campaigns in the megacity Paris, the North West Mediterranean basin (WMB) and southern West Africa (SWA). Various metrics derived from the photostationary steady state (PSS) and the ozone production efficiency (OPE) are calculated from airborne observations. A new metric, the oxidant production rate normalized to carbon monoxide (PROx), is introduced and quantified as a function of the processing time of the plume. In most of the polluted plumes, it is found that the Leighton ratio (F) characterizing the equilibrium between O3 and NOx is, on average, within the PSS range ([1 0.32]) or greater. The positive dependence of Ox to NO usually indicates a VOC-sensitive regime inside the plumes with some exceptions. In Paris, under oceanic westerly winds, and during DACCIWA, the plumes show a rural-like chemistry behaviour at moderate NOx levels (NOx-sensitive). Intense and frequent rapid changes in J(NO2), NO and NO2 explain the deviations from the PSS. The OPE for Paris plume suggests that the VOC-sensitive regime extends far beyond the urban plume. The mean ozone production is higher downwind of Paris (30 ppb h1 on average) compared to SWA (20 ppb h-1) and WMB (6 ppb h1). PROx values vary between 0 (no oxidantproduction) and 0.27 ppb[Ox] ppb[CO]1 h1. The determined uncertainty on the Leighton ratio value could affect the differences in the estimation of the photochemical oxidant production by PO3 and PROx. The emissions of CO along the flight path and the presence of vegetation and high humidity levels might shape the oxidant production depending on the explored environment. While limited in number, PROx values set a benchmark for future photochemical studies to compare with: Paris as representative of an anthropogenic urban plume and WMB as representative of a biogenic continental plume.

International audience; Indoor measurements of OH and sum of peroxy radicals XO2=HO2+RO2 were conducted in a room using window glasses of different transparencies and applying different coatings to the walls. Average OH and XO2 concentrations were found to vary in the range (0.6-4) × 105 molecule cm-3 and (1-7) × 107 molecule cm-3 respectively with anti-UV windows, and (6-10) × 105 molecule cm-3 and (4-16) × 107 molecule cm-3 respectively with borosilicate glasses. The OH and XO2 concentrations were compared with simulation results obtained using the H2I model, which accounts for the mixing between the sunlit and shaded volumes of the room. Taking into account the measurement uncertainty, the simulated OH concentrations agree with the observations on average while the simulated XO2 concentrations tend to be overestimated. Based on the model results, ozonolysis of unsaturated VOCs and photolysis of HONO and of organic compounds are found to represent the main primary sources of OH and XO2 radicals, the latter being more important in the sunlit volume. Despite the lower rate of the radical initiation in the shaded volume, the difference of radical concentrations in the shaded and sunlit volumes is mitigated by an interplay between the mixing time and the lifetime of XO2 radicals, allowing efficient transport of XO2 radicals into the shaded volume and acting there as a source of OH via radical propagation processes.

International audience; Recent studies have shown that it is possible to combine machine learning methods with data assimilation to reconstruct a dynamical system using only sparse and noisy observations of that system. The same approach can be used to correct the error of a knowledge-based model. The resulting surrogate model is hybrid, with a statistical part supplementing a physical part. In practice, the correction can be added as an integrated term (i.e. in the model resolvent) or directly inside the tendencies of the physical model. The resolvent correction is easy to implement. The tendency correction is more technical, in particular it requires the adjoint of the physical model, but also more flexible. We use the two-scale Lorenz model to compare the two methods. The accuracy in long-range forecast experiments is somewhat similar between the surrogate models using the resolvent correction and the tendency correction. By contrast, the surrogate models using the tendency correction significantly outperform the surrogate models using the resolvent correction in data assimilation experiments. Finally, we show that the tendency correction opens the possibility to make online model error correction, i.e. improving the model progressively as new observations become available. The resulting algorithm can be seen as a new formulation of weak-constraint 4D-Var. We compare online and offline learning using the same framework with the two-scale Lorenz system, and show that with online learning, it is possible to extract all the information from sparse and noisy observations.

International audience; The goal of this research is to assess environmental quality at the neighbourhood level through a multi-dimensional and multi-sensory approach that combines social and physical methodologies. For this purpose, an interdisciplinary protocol has been designed to simultaneously collect physical parameter measurements (related to microclimate and acoustics) and survey data on perceptions (involving residents and non-residents). The cross-referenced analysis of data collected at six contrasting places in a district in Toulouse (France) enabled us (i) to better understand and prioritise the factors that influence residents' assessment of the quality of their living environment and (ii) to understand to what extent the differentiation of the places by the inhabitants converges with the differentiation of these places based on acoustic and micrometeorological measurements. The statistical analysis based on individuals showed the importance of noise and air quality that rank just after the aesthetic dimension for all respondents. Nevertheless, the quality of maintenance and the feeling of security that the place inspires seem to be as crucial as these environmental criteria for the inhabitants. The analysis focused on the sites highlighted the consistency between the typology of places based on perceptions and that based on acoustic measurements, which confirms the high inhabitants' sensitivity to this environmental component.

In order to model the pollutant concentrations related to air quality in the streets of Paris, the street network model MUNICH (Model of Urban Network of Intersecting Canyons and Highways) is improved. A non-stationary approach is developed to represent the formation of secondary compounds, such as NO2. To model the dynamics of aerosols, MUNICH is coupled to the SSH-aerosol chemical module. The gas and particle concentrations in the streets of Paris are simulated with MUNICH, coupled to the chemical-transport model Polair3D to integrate background concentrations in the streets.For gaseous compounds, the coupling between MUNICH and Polair3D can be unidirectional (background concentrations influence those of the streets) or bidirectional (there is feedback between the street and the background concentrations). The concentrations of NO2 and NOx compare well with observations, regardless of the approach used for the coupling. The bidirectional coupling influences more the streets with an intermediate ratio between height and width, and with high traffic emissions, reaching 60% for the NO2 concentrations depending on the street. For particles, the concentrations of PM2.5, PM10 and the simulated chemical compositions are close to observations. Secondary particles have a significant impact on PM2.5 concentrations, reaching 27% depending on the street and the time of day. Gaseous chemistry has a strong influence on reactive gaseous species, increasing the average concentration of NO2 by 37%. The influence on condensables is lower, but reaches 20% depending on the street. The thermodynamic equilibrium assumption in the condensation/evaporation calculation overestimates organic concentrations by around 5% on average, up to 31% at noon depending on the street. NH3 traffic emissions increase inorganic concentrations by 3% on average, reaching 26% depending on the street.To explain the model's underestimation of the high black carbon (BC) concentrations observed in the streets, the influence of non-exhaust emissions and the bidirectional coupling is investigated. A new approach to calculate particle resuspension is developed, modeling the deposited mass and respecting the mass balance on the surface of the streets. Simulations show that particle resuspension has a low impact on BC concentrations. The concentrations of BC in the streets influence the background urban concentrations: the influence of bidirectional coupling reaches 50% depending on the street. Tyre wear emissions contribute to BC emissions in a comparable way to exhaust emissions. New emission factors are proposed, consistent with studies in the literature and the model/measurement comparison carried out here.MUNICH is finally used in Paris to estimate the impact of the vehicle fleet renewal over ten years and urban mobility on the population's exposure to multiple compounds. The vehicle fleet renewal strongly decreases the population's exposure to NO2, BC, PM10, PM2.5 and organic particles. This decrease is greater than that estimated using a regional CTM. The population's exposure to PM2.5 decreases similarly if recent diesel, gasoline or electric vehicles are favored. But favoring electric vehicles induces the greatest reduction in population exposure to NO2. Home-office practice is less efficient than renewing vehicles, but it can be used to intensify the decrease in population exposure to particulate matter concentrations. However, more ambitious emission reductions are needed to meet the air quality guidelines in Paris, Afin de modéliser les concentrations de polluants liés à la qualité de l’air dans les rues de Paris, le modèle de réseau de rues MUNICH (Model of Urban Network of Intersecting Canyons and Highways) est amélioré. Une approche non stationnaire est développée pour représenter la formation des composés secondaires, tels que NO2. Pour modéliser la dynamique des aérosols, MUNICH est couplé au module chimique SSH-aérosol. Les concentrations en gaz et particules dans les rues de Paris sont simulées avec MUNICH, couplé au modèle de chimie-transport Polair3D pour intégrer les concentrations de fond dans les rues. Pour les composés gazeux, le couplage entre MUNICH et Polair3D peut être unidirectionnel (les concentrations de fond influencent celles des rues) ou bidirectionnel (il y a un feedback entre la rue et les concentrations de fond). Les concentrations de NO2 et NOx se comparent bien aux observations, quelque soit l'approche utilisée pour le couplage. Le couplage bidirectionnel influence plus les rues avec un rapport hauteur/largeur intermédiaire et avec des émissions de trafic élevées, atteignant 60% sur les concentrations de NO2 selon la rue. Pour les particules, les concentrations de PM2.5, PM10 et les compositions chimiques simulées sont proches des observations. Les particules secondaires ont un impact important sur les concentrations de PM2.5, atteignant 27% selon la rue et le moment de la journée. La chimie gazeuse a une forte influence sur les espèces gazeuses réactives, augmentant de 37% la concentration moyenne du NO2. L'influence sur les condensables est plus faible, mais atteint 20% selon la rue. L'hypothèse d'équilibre thermodynamique dans le calcul de la condensation/évaporation surestime les concentrations en organiques d'environ 5% en moyenne, jusqu'à 31% à midi selon la rue. Les émissions trafic de NH3 augmentent les concentrations en inorganiques de 3% en moyenne, atteignant 26% selon la rue. Pour expliquer la sous-estimation par le modèle des fortes concentrations de carbone suie (BC) observées dans les rues, l'influence des émissions hors échappement et du couplage bidirectionnel est investiguée. Une nouvelle approche pour calculer la remise en suspension des particules est présentée, modélisant la masse déposée et respectant le bilan de masse à la surface des rues. Les simulations montrent que la remise en suspension des particules a un faible impact sur les concentrations de BC. Les concentrations de BC dans les rues influencent les concentrations urbaines de fond : l’influence du couplage bidirectionnel atteint 50% selon la rue. Les émissions d'usure des pneus contribuent aux émissions de BC de façon comparable aux émissions à l'échappement. Des nouveaux facteurs d'émission sont proposés cohérents avec certaines études de la littérature et la comparaison modèle/mesures effectuée. MUNICH est finalement utilisé sur Paris pour estimer l'impact du renouvellement du parc automobile sur dix ans et de la mobilité urbaine sur l'exposition de la population à de multiples composés. Le renouvellement du parc de véhicules diminue fortement l'exposition de la population aux NO2, BC, PM10, PM2.5 et aux particules organiques. Cette diminution est plus importante que celle estimée en utilisant un CTM à l'échelle régionale. L'exposition de la population aux PM2.5 diminue de façon similaire si les véhicules diesel, essence ou électriques récents sont favorisés. Mais favoriser les véhicules électriques induit la plus forte diminution de l'exposition au NO2. Le télétravail est moins efficace que le renouvellement des véhicules, mais il peut être utilisé pour intensifier la diminution de l'exposition aux concentrations de particules. Cependant, des réductions plus ambitieuses des émissions sont nécessaires pour respecter les directives de qualité de l'air sur Paris

International audience; We clarify issues related to the expression of Lagrangian stochastic models used for atmospheric dispersion applications. Two aspects are addressed: the need to verify the well-mixed criterion and the correspondence between Eulerian and Lagrangian turbulence models when they are combined in practical simulations. In particular, it is recalled that the fulfillment of the well-mixed criterion depends only on the proper incorporation of the mean pressure-gradient term as the mean drift term of the Langevin equation. New consistency issues between duplicate fields within Eulerian/Lagrangian hybrid formulations are also brought out, especially regarding turbulence models, boundary conditions, and divergence-free condition. Such hybrid methods, where mean flow quantities calculated with an Eulerian approach are provided to the Lagrangian approach, are commonly used in atmospheric dispersion simulations for their numerical efficiency. Nevertheless, it is shown that serious inconsistencies can result from coupling Eulerian and Lagrangian models that do not correspond to the same level of description of the fluid turbulence.

International audience; Accurate wind fields simulated by CFD models are necessary for many environmental and safety micro-meteorological applications, such as wind resource assessment. Atmospheric simulations at local scale are largely determined by boundary conditions (BCs), which are generally provided by outputs of mesoscale models (e.g., WRF). In order to improve the accuracy of the BCs, especially in the lowest levels, data assimilation methods might be used to take available observations into account. Data assimilation methods have generally been developed for larger scale meteorology and deal with initial conditions. Among the existing methods, the iterative ensemble Kalman smoother (IEnKS) has been chosen and adapted to micro-meteorology by taking BCs into account. In the present study, we assess the ability of the adapted IEnKS to improve wind simulations over a very complex topography in a context of wind resource assessment, by assimilating a few in situ observations. The IEnKS is tested with the CFD model Code Saturne in 2D and 3D using both twin experiments and real observations. We propose a method to determine the first estimate of the BCs and to construct the associated background error covariance matrix, from the statistical analysis of three years of WRF simulations. The IEnKS is proved to greatly reduce the error and the uncertainty of the BCs and thus of the simulated wind field over the small-scale domain. As a consequence, the wind resource estimate is also much more accurate. Highlights • This article provides a framework to perform ensemble variational data assimilation of in situ observations to improve local scale simulations with a CFD model. • The iterative ensemble Kalman smoother is adapted to correct boundary conditions and is tested with twin experiments and real data experiments in 2D and 3D with a CFD model over very complex topography. • The adapted IEnKS is proved to enhance the accuracy of boundary conditions and local scale simulations in operationally affordable conditions.

International audience; We present an intercomparison of two models applied to a major boulevard in a Paris suburb accounting for building effects: (1) a computational fluid dynamics (CFD) model providing a detailed three-dimensional representation of the atmospheric flow and pollutant dispersion (Code_Saturne) and (2) a street-network model with well-mixed steady-state concentrations within street segments coupled with a regional chemical-transport model (SinG). Simulations were performed for five cases representing different meteorological conditions (wind direction and speed) and two sensitivity cases with different emissions. We compare model results to measurements of NOx at two monitoring stations on either side of the street. Results exhibit (a) a complex behavior highlighting effects of the street-network configuration and emission patterns on the cross-street concentration gradient and (b) a satisfactory performance with assumption of well-mixed concentrations within street-canyons for mean NOx (MNE of 39% and 22%; and NMB of −29% and − 7% for Code_Saturne and SinG, respectively).

International audience; Application of universal multifractal framework in estimating a new scale invariant power law relation between kinetic energy and intensity of rainfall

International audience; As secondary organic aerosols (SOA) largely contribute to the mass of particles and may strongly affect health, it is essential to represent them as accurately as possible in air quality models (AQM). Their formation and aging involve multi-generation oxidations of numerous volatile organic compounds (VOC) combined with gas-particle partitioning processes.Tracking the non-linear relationship between VOC emissions and aerosol formation demands comprehensive chemical mechanisms, which take into account the whole complexity of the SOA precursor oxidation to simulate aerosols under various conditions.However, the use of explicit gas-phase chemical mechanism (e.g., MCM, GECKO-A) or molecular structure-limited parameterization (e.g., VBS, SOM, FGOM) could be problematical in large-scale SOA modeling, as the former is overwhelmingly computational expensive while the latter loses tracks of VOC oxidation products after few generations and specific properties relying on aerosol formation.Consequently, we have developed semi-explicit SOA chemical mechanisms designed to model the SOA formation and evolution in 3D AQM. These mechanisms are reduced based on simulations of the near-explicit master chemical mechanism (MCM) performed under various conditions representative of ambient conditions and different lumping strategies. The new mechanisms integrate the crucial SOA species/reactions with different mechanism complexities. The mechanisms, therefore, preserve the complexity of the oxidation chemistry (dependence on NOx of the SOA formation, the influence of radical concentrations, humidity, photolysis, etc..) as well as the molecular composition of the organic aerosol. The mechanisms are implemented in a novel 0D aerosol model SSH-aerosol, which can use the molecular structure of lumped compounds to estimate the influence of non-ideality on SOA formation.The current application has been conducted on the MCM degradation scheme of beta-caryophyllene (C15H24), the most representative sesquiterpene. A reduction of the average 90% CPU time and up to 92% number of S/IVOCs species has been achieved compared to the original MCM mechanism.

In the event of a release of radioactive pollutants into the atmosphere, one of the missions of the authorities is to evaluate the consequences of this release in order to implement, if necessary, measures to protect the population. These may include evacuation or sheltering in the very short term, and restrictions on the consumption or marketing of contaminated foodstuffs in the longer term. For this purpose, numerical models are used to simulate the dispersion of radionuclides in the atmosphere.The accuracy of the results obtained from these models strongly depends on the knowledge of the source term, i.e. the location, duration and magnitude of the release as well as its distribution between radionuclides. However, knowledge of the source term is generally subject to significant uncertainties. In addition to the source term, other uncertainties arise from the transport model, meteorological fields, measurement data and the representativeness of the model with respect to the measurements.In this PhD, we have developed and applied inverse modelling methods to improve the evaluation of the source term and to quantify the uncertainties.Among the inverse modelling methods, variational deterministic approaches are effective in providing a rapid estimate of the source term, but the quantification of uncertainties associated with this estimate is generally difficult.We therefore propose to address the problem within the probabilistic framework of Bayesian inference, which is part of a formalism that allows a more complete assessment of uncertainties.Several Markov chain Monte Carlo (MCMC) sampling methods are implemented in order to reconstruct the variables describing the source: the Metropolis Hastings (MH) algorithm, the Parallel tempering algorithm and finally the Reversible-Jump MCMC.These algorithms are first applied and validated on the ruthenium 106 detection event that occurred in Europe in the fall of 2017. The probability densities of the source variables are reconstructed in order to identify the geographical origin of the detections as well as the quantities of ruthenium 106 released into the atmosphere.Then, in a second step, several methods are developed in order to incorporate and quantify different sources of errors within the Bayesian problem and thus enable us to obtain a better reconstruction of the distribution of the release.The second case study is dedicated to the Fukushima Daiichi plant accident which led to long releases associated with time-varying kinetics. The reconstruction of these releases with more complex characteristics required the development of a new MCMC algorithm, the Reversible-Jump MCMC, which was adapted from previous sampling methods.Applied to the Fukushima case, the Reversible-Jump MCMC shows its ability to sample more finely and efficiently the distribution of the source term and uncertainties., En cas de rejet de polluants radioactifs dans l’atmosphère, une des missions des autorités est d’évaluer les conséquences de ce rejet afin de mettre en œuvre, si nécessaire, des mesures de protection des populations. Il peut s’agir d’évacuation ou de mise à l’abri à très court terme et de restrictions de consommation ou de commercialisation des denrées alimentaires contaminées à plus long terme. Pour cela, des modèles numériques sont utilisés pour simuler la dispersion des radionucléides dans l’atmosphère.La précision des résultats obtenus à partir de ces modèles dépend fortement de la connaissance du terme source, c’est-à-dire de la localisation, de la durée, de l’ampleur du rejet ainsi que de sa distribution entre radionucléides. Or, la connaissance du terme source est généralement soumise à d’importantes incertitudes. En plus du terme source, d’autres incertitudes proviennent du modèle de transport, des champs météorologiques, des données de mesure et de la représentativité du modèle par rapport aux mesures.Dans cette thèse, nous avons développé et appliqué des méthodes de modélisation inverse permettant d’améliorer l’évaluation du terme source et de quantifier les incertitudes.Parmi les méthodes de modélisation inverse, les approches déterministes variationnelles sont efficaces pour fournir une estimation rapide du terme source, mais la quantification des incertitudes associée à cette estimation est généralement difficile.Nous proposons donc d'aborder le problème dans le cadre probabiliste de l'inférence bayésienne qui s’inscrit dans un formalisme permettant d’obtenir une évaluation plus complète des incertitudes.Plusieurs méthodes d’échantillonnage de Monte Carlo à chaîne de Markov (MCMC) sont mises en œuvre afin de reconstruire les variables décrivant la source : l’algorithme de Metropolis Hastings (MH), l’algorithme du Parallel tempering et enfin le Reversible-Jump MCMC.Ces algorithmes sont tout d’abord appliqués et validés sur l’évènement de détection de ruthénium 106 survenu en Europe à l’automne 2017. Les densités de probabilité des variables de la source sont reconstruites afin d’identifier l’origine géographique des détections ainsi que les quantités de ruthénium 106 rejetées dans l’atmosphère.Puis, dans un second temps, plusieurs méthodes sont développées afin d’incorporer et de quantifier différentes sources d’erreurs au sein du problème bayésien et ainsi permettre une meilleure reconstruction de la distribution du rejet.Le second cas d’étude est dédié à l’accident de Fukushima qui a conduit à des rejets longs associés à une cinétique variable dans le temps. La reconstruction de ces rejets aux caractéristiques plus complexes a nécessité le développement d’un nouvel algorithme MCMC, le Reversible-Jump MCMC, qui a été adapté à partir des méthodes d’échantillonnage précédentes.Appliqué au cas de Fukushima, le Reversible-Jump MCMC montre sa capacité à échantillonner plus finement et plus efficacement la distribution du terme source et des incertitudes.

Titre traduit en anglais : Paris-Fog : researchers in the fog. Résume traduit en anglais : A 6-month field experiment, named Paris-Fog, was carried out in winter 2006-2007 near Paris. This long-term field experiment is original. This article provides a detailed description of the instrumentation used and of the operation plan. A summary of the cases sampled is discussed and key physical processes involved in an intensive observational period are presented.; International audience; Une campagne de mesures dédiée au brouillard, appelée Paris-Fog, s'est déroulée en région parisienne au cours de l'hiver 2006-2007. Cette campagne est ambitieuse de par les moyens qui ont été mis en oeuvre et la longueur de la période documentée. On décrit ici les instruments utilisés ainsi que le mode de fonctionnement de la campagne de mesures regroupant des laboratoires de recherche d'horizons différents. Puis on présente un premier bilan ainsi qu'un zoom sur une période d'observations intensives.

Twenty-five biogenic and anthropogenic secondary organic aerosol (SOA) markers have been measured over a one-year period in both gaseous and PM10 phases in the Paris region (France). Seasonal and chemical patterns were similar to those previously observed in Europe, but significantly different from the ones observed in America and Asia due to dissimilarities in source precursor emissions. Nitroaromatic compounds showed higher concentrations in winter due to larger emissions of their precursors originating from biomass combustion used for residential heating purposes. Among the biogenic markers, only isoprene SOA marker concentrations increased in summer while pinene SOA markers did not display any clear seasonal trend. The measured SOA markers, usually considered as semi-volatiles, were mainly associated to the particulate phase, except for the nitrophenols and nitroguaiacols, and their gas/particle partitioning (GPP) showed a low temperature and OM concentrations dependency. An evaluation of their GPP with thermodynamic model predictions suggested that apart from equilibrium partitioning between organic phase and air, the GPP of the markers is affected by processes suppressing volatility from a mixed organic and inorganic phase, such as enhanced dissolution in aerosol aqueous phase and non-equilibrium conditions. SOA marker concentrations were used to apportion secondary organic carbon (SOC) sources applying both, an improved version of the SOA-tracer method and positive matrix factorization (PMF) Total SOC estimations agreed very well between both models, except in summer and during a highly processed Springtime PM pollution event in which systematic underestimation by the SOA tracer method was evidenced. As a first approach, the SOA-tracer method could provide a reliable estimation of the average SOC concentrations, but it is limited due to the lack of markers for aged SOA together with missing SOA/SOC conversion fractions for several sources.

Due to their major role in atmospheric chemistry and secondary pollutant formation such as ozone or secondary organic aerosols, an accurate representation of OH and HO2 (HOX) radicals in air quality models is essential. Air quality models use simplified mechanisms to represent atmospheric chemistry and interactions between HOX and organic compounds. In this work, HOX concentrations during the oxidation of toluene and xylene within the Regional Atmospheric Chemistry Mechanism (RACM2) are improved using a deterministic–near-explicit mechanism based on the Master Chemical Mechanism (MCM) and the generator of explicit chemistry and kinetics of organics in the atmosphere (GECKO-A). Flow tube toluene oxidation experiments are first simulated with RACM2 and MCM/GECKO-A. RACM2, which is a simplified mechanism, is then modified to better reproduce the HOX concentration evolution simulated by MCM/GECKO-A. In total, 12 reactions of the oxidation mechanism of toluene and xylene are updated, making OH simulated by RACM2 up to 70% more comparable to the comprehensive MCM/GECKO-A model for chamber oxidation simulations.

The characterization of vehicle exhaust emissions of volatile organic compounds (VOCs) is essential to estimate their impact on the formation of secondary organic aerosol (SOA) and, more generally, air quality. This paper revises and updates non-methane volatile organic compounds (NMVOCs) tailpipe emissions of three Euro 5 vehicles during Artemis cold urban (CU) and motorway (MW) cycles. Positive matrix factorization (PMF) analysis is carried out for the first time on proton transfer reaction time-of-flight mass spectrometer (PTR-ToF-MS) datasets of vehicular emission. Statistical analysis helped to associate the emitted VOCs to specific driving conditions, such as the start of the vehicles, the activation of the catalysts, or to specific engine combustion regimes. Merged PTR-ToF-MS and automated thermal desorption gas chromatography mass spectrometer (ATD-GC-MS) datasets provided an exhaustive description of the NMVOC emission factors (EFs) of the vehicles, thus helping to identify and quantify up to 147 individual compounds. In general, emissions during the CU cycle exceed those during the MW cycle. The gasoline direct injection (GDI) vehicle exhibits the highest EF during both CU and MW cycles (252 and 15 mg/km), followed by the port-fuel injection (PFI) vehicle (24 and 0.4 mg/km), and finally the diesel vehicle (15 and 3 mg/km). For all vehicles, emissions are dominated by unburnt fuel and incomplete combustion products. Diesel emissions are mostly represented by oxygenated compounds (65%) and aliphatic hydrocarbons (23%) up to C22, while GDI and PFI exhaust emissions are composed of monoaromatics (68%) and alkanes (15%). Intermediate volatility organic compounds (IVOCs) range from 2.7 to 13% of the emissions, comprising essentially linear alkanes for the diesel vehicle, while naphthalene accounts up to 42% of the IVOC fraction for the gasoline vehicles. This work demonstrates that PMF analysis of PTR-ToF-MS datasets and GC-MS analysis of vehicular emissions provide a revised and deep characterization of vehicular emissions to enrich current emission inventories.

Air pollution is the co-occurrence of high pollutant emissions and particular meteorological conditions. Among its pollutants, fine particles (PM2.5 and PM10), Nitrogen Oxide (NOx) and Ozone (O3) are the most dangerous for public health. Every year, repeated or prolonged exposure to these particles leads to respiratory and cardiovascular diseases, cancer and premature death in those exposed. Climate change has an impact on meteorological variables (temperature, pressure, winds, precipitation, etc.) that affect air quality (emissions, precipitation leaching, gas/particle balance, etc.). The objective of this synthesis report is to provide a study to shed light on the consequences of air quality variation as a function of climate change and emissions in the near future (horizons2030 and 2050), in particular on the climate scenario RCP (Representative Concentration Pathway) 8.5 which describes an absence of climate change policies (Riahi et al., 2011). Within the framework of this synthesis report, we are committed to:- Communicating a qualitative assessment of the variation in the main pollutants, particularly in France;- Where possible, quantifying the impact in urban agglomerations;- Studying in particular the future modification of ozone (O3), coarse particles (PM10), fine particles (PM2.5) and nitrogen oxides (NOx). -translating impacts into additional claims experience for insurance cover poposed under personal insurance contracts., La pollution atmosphérique est la cooccurrence de fortes émissions de polluants et de conditions météorologiques particulières. Parmis ses polluants, les particules fines (PM2.5 et PM10), l'Oxyde d'Azote (NOx) et l'Ozone (O3) sont les plus dangereuses pour la santé publique. L'exposition répétée ou prolongée à ces particules entraîne chaque année des maladies respiratoires et cardio-vasculaires, des cancers ainsi que des morts prématurées chez les personnes exposées. L'évolution du climat a un impact sur des variables météorologiques (température, pression, vents, précipitations, ...) qui affectent la qualité de l'air (émissions, lessivage par les précipitations, équilibre gaz/particule, ...). L'objectif du présent rapport de synthèse est de fournir une étude pour éclairer les conséquences de la variation de la qualité de l'air en fonction des changements climatiques et des émissions dans un avenir proche (horizons2030 et 2050), en particulier sur le scénario climatique RCP (Representative Concentration Pathway) 8.5 qui décrit une absence de politiques de changement climatique (Riahi et al., 2011). Dans le cadre du présent rapport de synthèse, nous nous attachons à :-Communiquer une évaluation qualitative de la variation des principaux polluants en particulier sur la France;-lorsque cela est possible, quantifier l'impact dans les agglomérations urbaines,-étudier en particulier la modification future de l'ozone (O3), des particules grossières (PM10), des particules fines (PM2,5) et des oxydes d'azote (NOx).-traduire des impacts en sinistralité additionnelle pour les garanties d'assurance poposées dans le cadre de contrats d'assurances de personnes.

This study was part of the work done by the ANSES working group on ambient particulate matter (chemical composition and road traffic emissions); International audience; Air pollution from road traffic and its mitigation is a major concern in most cities. A platform for simulating pollutant emissions and concentrations was developed and applied to the Île-de-France Region (Greater Paris) of France, taking account of anthropogenic and natural sources and ‘imported’ pollution from elsewhere in France and Europe. Four technological scenarios for 2025 were studied and compared to the 2014 reference situation (1-REF). These scenarios included the current evolution of the park with widespread adoption of diesel particulate filters (DPFs) (2-BAU), decline in the sale of diesel vehicles and a corresponding increase in petrol vehicle sales (3-PET), promotion of electric vehicles in urban areas (4-ELEC), and a combinaison with a decrease in traffic of about 15% in the densely populated area inside the A86 outer ring road (5-AIR). The corresponding vehicle fleets were determined using a fleet simulation model.Traffic pollutant emissions were computed with the COPERT4 European methodology and hourly traffic data over the Île-de-France road network. Particulate matter (PM10, PM2,5 and PM1,0), particles number (PN), black carbon (BC), organic matter (OM), nitrogen oxides (NOx) and nitrogen dioxide (NO2), non-methane volatile organic compounds (VOC), ammonia (NH3), carbon monoxide (CO) and carbon dioxide (CO2) were considered. Emissions for other sectors were taken from a regional inventory. Emissions outside the Île-de-France region (Europe and France) were derived from the European and French emission inventories. Pollutant concentrations (PM2,5, PM10, organic and inorganic PM10, PN, BC, NO2 and O3) were simulated over nested domains (Europe, France and Île-de-France) using the Polyphemus platform for two scenarios (2-BAU and 3-PET). Methodological aspects and results for Île-de-France are discussed here.All scenarios led to a sharp decrease in traffic emissions in Île-de-France (−30% to −60%) by 2025. The decline in diesel induced a stronger renewal of the fleet. PM and NOx emissions were more strongly reduced than VOC or NH3. Traffic reduction reduced all emissions in the densely populated area within the A86 outer ring road (−20% to −45% for exhaust particles and gaseous pollutants).The 2-BAU and 3-PET scenarios lowered annual average concentrations, especially for NO2 and BC, and more strongly influenced daily-peak than daily-average concentrations. In Île-de-France, PM of diameter

International audience; We discuss herein the suitability of some open boundary conditions while comparing approximate solutions of one-dimensional Riemann problems in a bounded sub-domain with the restriction in this sub-domain of the exact solution in the infinite domain, considering the Euler system of gas dynamics. Assuming that no information is known from outside of the domain, some basic open boundary condition specifications are given, and a measure of the L 1 norm of the error inside the computational domain enables to show consistency errors in situations involving outgoing shock waves, depending on the chosen boundary condition formulation. This investigation has been performed with Finite Volume methods, using approximate Riemann solvers in order to compute numerical fluxes for both inner and boundary interfaces.

International audience; Road transport is a major source of anthropogenic emissions especially in the Middle East where the regulations enforcement is generally poor. This study aims to quantify the Emission Factors (EF) of traffic-related gaseous and particulate pollutants inside the Salim Slam urban tunnel in Beirut, Lebanon. The fuel-based emission factors of measured pollutants were from the carbon mass balance model. The EF determined showed general higher values than those reported in recent studies from European and American countries, even for speciated NMVOC. The average CO and NOx emission factors for the mixed fleet (HDV + LDV) were determined to be 10.52 ± 3.00 g km−1 and 2.20 ± 0.57 g km−1 respectively, while the EF for PM2.5 55 ± 27 mg km−1. Moreover, IVOC species from gaseous phase were reported for the first time in the region. A reduction trend was observed in comparison with the previous tunnel study from Lebanon, however there is still a need to have tougher regulations to control the local practices such as removal of catalytic converter, adjustment of engine parameters for inspection, etc. The comparison of the EF to those calculated through EMEP or IPCC methodologies shows the need to take local practices while establishing national emission inventories.

To calculate the effect of rainfall in detaching particles and initiating soil erosion, it is important to represent relationship between recorded drop size distributions (DSD) and fall velocity across various scales of measurement. Commonly used relationships between kinetic energy (KE) and rainfall rate (R) exhibit strong dependence on the temporal resolution at which analysis is carried out. Here we aim at developing a scale invariant relationship relying on the framework of Universal Multifractals (UM), which has been widely used to analyze and characterize geophysical fields that exhibit extreme variability over measurement scales.Rainfall data is collected using three optical disdrometers working on different underlying technologies (one Campbell Scientific PWS100 and two OTT Parsivel2 instruments) and operated by Hydrology, Meteorology, and Complexity laboratory of École des Ponts ParisTech in the Paris area (France). They provide access to the size and velocity of drops falling through sampling areas of few tens of cm2. Such data enables estimation of rainfall microphysics, R and KE at various resolutions. The temporal variation of this geophysical data over wide range of scales is then characterized in the UM framework. A power law relation has been developed for describing the dependence of KE on R. The developed equation using scale invariant features of UM are valid not only at a single scale, but also across scales. The amount of uncertainty is further characterized by comparing actual data with simulated rainfall data from Sense-City climate chamber.Keywords: rainfall intensity; rainfall kinetic energy; disdrometer; multi fractal; scale invariant

Urban air quality simulation is an important tool to understand the impacts of air pollution. However, the simulations are often computationally expensive, and require extensive data on pollutant sources. Data on road traffic pollution, often the predominant source, can be obtained through sparse measurements, or through simulation of traffic and emissions. Modeling chains combine the simulations of multiple models to provide the most accurate representation possible, however the need to solve multiple models for each simulation increases computational costs even more. In this paper we construct a meta-modeling chain for urban atmospheric pollution, from dynamic traffic modeling to air pollution modeling. Reduced basis methods (RBM) aim to compute a cheap and accurate approximation of a physical state using approximation spaces made of a suitable sample of solutions to the model. One of the keys of these techniques is the decomposition of the computational work into an expensive one-time offline stage and a low-cost parameter-dependent online stage. Traditional RBMs require modifying the assembly routines of the computational code, an intrusive procedure which may be impossible in cases of operational model codes. We propose a non-intrusive reduced order scheme, and study its application to a full chain of operational models. Reduced basis are constructed using principal component analysis (PCA), and the concentration fields are approximated as projections onto this reduced space. We use statistical emulation to approximate projection coefficients in a non-intrusive manner. We apply a multi-level meta-modeling technique to a chain using the dynamic traffic assignment model LADTA, the emissions database COPERT IV, and the urban dispersion-reaction air quality model SIRANE to a case study on the city of Clermont-Ferrand with over 45, 000 daily traffic observations, a 47, 000-link road network, a simulation domain covering $$180\,\text {km}^2$$ 180 km 2 . We assess the results using hourly NO $$_2$$ 2 concentration observations measured at stations in the agglomeration. Computational times are reduced from nearly 3 h per simulation to under 0.1 s, while maintaining accuracy comparable to the original models. The low cost of the meta-model chain and its non-intrusive character demonstrate the versatility of the method, and the utility for long-term or many-query air quality studies such as epidemiological inquiry or uncertainty quantification.

Data assimilation combines forecasts from a numerical model with observations. Most of the current data assimilation algorithms consider the model and observation error terms as additive Gaussian noise, specified by their covariance matrices Q and R, respectively. These error covariances, and specifically their respective amplitudes, determine the weights given to the background (i.e., the model forecasts) and to the observations in the solution of data assimilation algorithms (i.e., the analysis). Consequently, Q and R matrices significantly impact the accuracy of the analysis. This review aims to present and to discuss, with a unified framework, different methods to jointly estimate the Q and R matrices using ensemble-based data assimilation techniques. Most of the methodologies developed to date use the innovations, defined as differences between the observations and the projection of the forecasts onto the observation space. These methodologies are based on two main statistical criteria: (i) the method of moments, in which the theoretical and empirical moments of the innovations are assumed to be equal, and (ii) methods that use the likelihood of the observations, themselves contained in the innovations. The reviewed methods assume that innovations are Gaussian random variables, although extension to other distributions is possible for likelihood-based methods. The methods also show some differences in terms of levels of complexity and applicability to high-dimensional systems. The conclusion of the review discusses the key challenges to further develop estimation methods for Q and R. These challenges include taking into account time-varying error covariances, using limited observational coverage, estimating additional deterministic error terms, or accounting for correlated noises., Comment: The manuscript is being considered for publication at Monthly Weather Review

International audience; In autumn 2017, small amounts of Ruthenium 106 of unknown origin were detected in Europe by several independent monitoring networks. To study the dispersion of this radionuclide, inverse modelling methods are applied to retrieve the location, time, duration and magnitude of the source. The inverse problem is solved within the Bayesian framework to yield a full reconstruction of the uncertainties. We first develop a classical Markov Chain Monte Carlo (MCMC) method - the Metropolis Hastings (MH) algorithm - to reconstruct the source. However, the algorithm fails to converge in acceptable time because of local minima of the cost function, whose existence is due to a poor distribution of the observations.To overcome this obstacle, the parallel tempering algorithm, an enhanced MCMC method, is assessed and applied to the Ruthenium dispersion event. The convergence of the algorithm is studied and keys to implement and accelerate it are provided. The probability distribution functions of the variables associated with the source and the observation errors are obtained: they point to an area in south Ural and a total release of several hundreds of TBq on the 26th of September 2017. The results are compared and proven to be consistent with other estimates. Moreover, the method converges fast enough to make it suitable for operational use.

The impact of mixing state of particles from biomass burning on aerosol optical properties (aerosol optical depth (AOD) and single-scattering albedo (SSA)) is studied over the Euro- Mediterranean region during the severe fire event in the Balkans between 20 and 31 July 2007, also characterized by high dust concentrations. When the mixing state is resolved in chemistry-transport models, chemical compounds are grouped for computational reasons, and internal mixing is assumed within each group. Up to six groups are defined here (dust, black carbon, two inorganic groups and two organic groups). The influence of different grouping assumptions is studied here and compared to the influence of the distribution of black carbon (BC) in particles (“pure homogeneous” representation and “core-shell” one), and the influence of modeling the water absorbed by both inorganic and organic compounds. The comparisons of simulated AODs to observations from the surface network AERONET show that AOD is slightly underestimated when aerosol compounds are assumed to be externally mixed and slightly overestimated when they are assumed to be internally mixed. The mixing state of dust with other compounds, as well as the distribution of BC in particles, strongly influence the optical properties. The impact of the mixing state on AOD is higher than the impact of the distribution of BC in particles, reaching 8–12% on average over the fire regions and 16% in the fire plume. Analysis related to the impact of particle mixing state and BC distribution on SSA shows results similar to AOD. The impact of the mixing state on SSA can reach −8.5% over the fire regions and it is higher than the impact of the BC representation, which is lower than 2%. At the location of fires, water absorbed by inorganics and organics is shown to influence the AOD by about 2%, which is in the lower range of the influence of water on AOD over the region (between 0 and 40%). This low influence of water on AOD during fire is due to assumptions made in the modelling, where most of secondary organic aerosols formed during fires are assumed to be hydrophobic.

International audience; abstract simple

Two observational campaigns were conducted, one in the Grenoble area (South Eastern France), for the MOCoPo project, near an urban freeway in 2011 and the other one in a Paris suburb, forthe TrafiPollu project, on a major surface street in 2014. PM10 concentrations were measured by Air Rhône-Alpes during the last 10 days of September 2011 for MOCoPo and by Airparif during 3 months from April to June 2014 for TrafiPollu. It has been shown that abrasion and resuspension processes represent a significant part of the total primary PM10 emissions of road traffic. Hereby, resuspended emissions originating from the road are estimated with several approaches and compared to PM10 measurements. We consider two different models available in the literature: HERMES (Pay et al., 2010) and NORTRIP (Denby et al., 2013), which differ in terms of formulation. We also apply an empirical method developed by Thorpe et al. (2007), based on near-road and background pollutant observations. The results vary depending on the traffic conditions and the modeling approach. In all cases, the resuspension emissions simulated are high enough to be considered in air quality modeling (ranging from 9 to 150% of the exhaust emissions). Those resuspension models were combined with atmospheric dispersion models to estimate near-road concentrations. We used a Gaussian line-source model for the Grenoble urban freeway and a street-canyon model (MUNICH) for the Paris suburban boulevard. The contribution of resuspension to traffic-related concentrations is hidden by a strong background contribution, which prevents us from concluding in terms of model performance. Nevertheless, a comparison with another dataset obtained near an urban freeway in Paris suggests that vehicle speed should be taken into account when estimating PM10 resuspension emissions.; Deux campagnes d'observation ont été menées, l'une dans la région grenobloise, pour le projet MOCoPo, près d'une autoroute urbaine en 2011, et l'autre dans une banlieue parisienne, pour le projet TrafiPollu, sur une grande rue en surface en 2014. Les concentrations de PM10 ont été mesurées par Air Rhône-Alpes pendant les 10 derniers jours de septembre 2011 pour MOCoPo et par Airparif pendant 3 mois d'avril à juin 2014 pour TrafiPollu. Il a été démontré que les procédés d'abrasion et de remise en suspension représentent une part importante des émissions primaires totales de PM10 du trafic routier. Ainsi, les émissions remises en suspension provenant de la route sont estimées selon plusieurs approches et comparées aux mesures des PM10. Nous considérons deux modèles différents disponibles dans la littérature : HERMES (Pay et al., 2010) et NORTRIP (Denby et al., 2013), qui diffèrent en termes de formulation. Nous appliquons également une méthode empirique développée par Thorpe et al (2007), basée sur des observations de polluants de fond et de proximité de la route. Les résultats varient en fonction des conditions de circulation et de l'approche de modélisation. Dans tous les cas, les émissions de resuspension simulées sont suffisamment élevées pour être prises en compte dans la modélisation de la qualité de l'air (allant de 9 à 150 % des émissions d'échappement). Ces modèles de resuspension ont été combinés à des modèles de dispersion atmosphérique pour estimer les concentrations près des routes. Nous avons utilisé un modèle gaussien de source linéaire pour l'autoroute urbaine de Grenoble et un modèle rue-canyon (MUNICH) pour le boulevard de la banlieue parisienne. La contribution de la remise en suspension aux concentrations liées au trafic est masquée par une forte contribution de fond, ce qui nous empêche de conclure en termes de performance du modèle. Néanmoins, une comparaison avec un autre ensemble de données obtenu près d'une autoroute urbaine à Paris suggère que le véhicule devrait être prise en compte lors de l'estimation des émissions de resuspension des PM10.

This study presents a new concept for estimating the pollutant emission rates of a site and its main facilities using a series of atmospheric measurements across the pollutant plumes. This concept combines the tracer release method, local-scale atmospheric transport modelling and a statistical atmospheric inversion approach. The conversion between the controlled emission and the measured atmospheric concentrations of the released tracer across the plume places valuable constraints on the atmospheric transport. This is used to optimise the configuration of the transport model parameters and the model uncertainty statistics in the inversion system. The emission rates of all sources are then inverted to optimise the match between the concentrations simulated with the transport model and the pollutants' measured atmospheric concentrations, accounting for the transport model uncertainty. In principle, by using atmospheric transport modelling, this concept does not strongly rely on the good colocation between the tracer and pollutant sources and can be used to monitor multiple sources within a single site, unlike the classical tracer release technique. The statistical inversion framework and the use of the tracer data for the configuration of the transport and inversion modelling systems should ensure that the transport modelling errors are correctly handled in the source estimation. The potential of this new concept is evaluated with a relatively simple practical implementation based on a Gaussian plume model and a series of inversions of controlled methane point sources using acetylene as a tracer gas. The experimental conditions are chosen so that they are suitable for the use of a Gaussian plume model to simulate the atmospheric transport. In these experiments, different configurations of methane and acetylene point source locations are tested to assess the efficiency of the method in comparison to the classic tracer release technique in coping with the distances between the different methane and acetylene sources. The results from these controlled experiments demonstrate that, when the targeted and tracer gases are not well collocated, this new approach provides a better estimate of the emission rates than the tracer release technique. As an example, the relative error between the estimated and actual emission rates is reduced from 32 % with the tracer release technique to 16 % with the combined approach in the case of a tracer located 60 m upwind of a single methane source. Further studies and more complex implementations with more advanced transport models and more advanced optimisations of their configuration will be required to generalise the applicability of the approach and strengthen its robustness.

Online coupled meteorology–atmospheric chemistry models have greatly evolved in recent years. Although mainly developed by the air quality modeling community, these integrated models are also of interest for numerical weather prediction and climate modeling, as they can consider both the effects of meteorology on air quality and the potentially important effects of atmospheric composition on weather. This paper summarizes the main conclusions from the “Symposium on Coupled Chemistry–Meteorology/Climate Modelling: Status and Relevance for Numerical Weather Prediction, Air Quality and Climate Research,” which was initiated by the European COST Action ES1004 “European Framework for Online Integrated Air Quality and Meteorology Modelling (EuMetChem).” It offers a brief review of the current status of online coupled meteorology and atmospheric chemistry modeling and a survey of processes relevant to the interactions between atmospheric physics, dynamics, and composition. In addition, it highlights scientific issues and emerging challenges that require proper consideration to improve the reliability and usability of these models for three main application areas: air quality, meteorology (including weather prediction), and climate modeling. It presents a synthesis of scientific progress in the form of answers to nine key questions, and provides recommendations for future research directions and priorities in the development, application, and evaluation of online coupled models.

Air quality is a major health and environmental issue worldwide. Similarly, the accuracy of wind resource assessment triggers significant economic and environmental repercussions. In order to study these two topics, it is necessary to accurately determine local wind fields using numerical models of micrometeorology. Such simulations are extremely sensitive to meteorological conditions at the domain borders. Up to present, the boundary conditions (BC) were estimated based on the results of larger scale simulations, which provide information that is not accurate enough, or even incomplete, for local scale purposes. As a matter of fact, the lack of knowledge about the BC represents a major source of error and uncertainty for micrometeorological studies.The potential sites for wind farm installation as well as built environments (urban areas or industrial sites) can be equipped with instruments measuring meteorological variables or pollutant concentration. The observations provided by these instruments represent a second source of information, insufficiently exploited for micrometeorological studies. Indeed, the in situ measurements are perturbed by the complex geometrical features on sites and might be difficult to exploit. In order to improve the exactitude and the accuracy of the BC, and consequently of the locale-scale atmospheric simulations, data assimilation (DA) methods, suited to this micrometeorological problem, could be applied to take benefit from these available observations.So far, DA methods have been mainly developed for large-scale meteorology and employed to correct the initial conditions (IC). In order to broaden the application scope of DA to micrometeorology, existing DA methods must be adapted to be able to correct the BC instead of IC.Two of the existing DA methods seem compatible with computational fluid dynamics (CFD) models used for micrometeorology over complex geometries: the back and forth nudging (BFN) algorithm and the iterative ensemble Kalman smoother (IEnKS). We have adapted these two methods, from a theoretical perspective, so as to include the BC in the control variables. The performances of the adapted versions of the BFN algorithm and the IEnKS have first been assessed with a simplified, 1D model of atmospheric flow with two layers, based on the shallow-water equations. The BFN algorithm and the IEnKS have then been tested in 2D and 3D with the atmospheric module of the open-source CFD model Code_Saturne.The first study case with Code_Saturne corresponds to a real application of wind resource assessment in a mountainous region with steep topography where three meteorological masts have been installed during a few months and provide in situ wind observations. The second case is a study of pollutant dispersion in an urban area, based on the measurements of wind and pollutant concentration coming from the “Mock Urban Setting Test” field campaign carried out in the USA. In this second case, the turbulence is also included in the BC and thus in the control variables. For both studies, some observations are assimilated and the remaining ones are used to validate the results.The experiences performed for the wind resource assessment study have revealed that the CFD models present too strong nonlinearities (flow recirculation after obstacles) for the BFN algorithm, which is based on a linearity assumption. However, both cases have shown the ability of the IEnKS to reduce the error and the uncertainty of the BC by assimilating a few observations, in operationally affordable conditions. Consequently, the simulated wind fields with Code_Saturne are also closer to the validation observations and the confidence intervals are reduced. Eventually, the IEnKS allows, in one case to estimate the wind potential, and in the other case to build the pollution maps, with much more exactitude and accuracy., La qualité de l’air est un enjeu sanitaire et environnemental majeur. Par ailleurs, l'estimation précise des potentiels éoliens est la source d’importantes retombées économiques et environnementales. Pour étudier ces deux sujets, il est nécessaire de reconstituer précisément les champs de vent locaux grâce à des modèles numériques de micrométéorologie. Ces simulations sont extrêmement sensibles aux conditions météorologiques aux limites du domaine d’étude. Jusqu’à présent, les conditions aux limites (CL) étaient estimées à partir de simulations à plus grande échelle, qui fournissent des informations peu adaptées à l’utilisation à l'échelle locale car imprécises, voire incomplètes. Par conséquent, la méconnaissance des CL représente une source majeure d’erreur et d’incertitude dans les études micrométéorologiques.Les sites susceptibles d’accueillir un parc éolien et les environnements bâtis (quartiers urbains ou sites industriels) peuvent être équipés d’instruments de mesures météorologiques ou de concentration de polluants. Les observations qu’ils fournissent constituent une seconde source d’information, jusqu’à ce jour peu exploitée pour les études micrométéorologiques. En effet, les mesures in situ sont perturbées par la géométrie complexe des sites étudiés. Afin d'améliorer la précision des CL et donc des simulations atmosphériques à l'échelle locale, des méthodes d'assimilation de données (AD) adaptées à cette problématique pourraient permettre de mettre à profit les observations disponibles.Jusqu’à présent, les méthodes d’AD ont été principalement développées pour la météorologie à grande échelle et ont donc surtout été utilisées pour corriger les conditions initiales (CI). Afin d'élargir le champ d'application de l’AD à la micrométéorologie, il faut adapter les méthodes existantes pour qu'elles permettent de corriger les CL plutôt que les CI.Deux méthodes d’AD semblent compatibles avec les modèles de mécanique des fluides (CFD) utilisés pour la micrométéorologie en géométrie complexe : l’algorithme de nudging direct et rétrograde (BFN) et le lisseur de Kalman d’ensemble itératif (IEnKS). Nous avons adapté ces deux méthodes d’un point de vue théorique pour inclure les CL dans les variables de contrôle. Les performances des versions adaptées du BFN et de l'IEnKS ont tout d'abord été étudiées avec un modèle simplifié d’écoulement atmosphérique à deux couches en 1D, basé sur les équations de Saint-Venant. Le BFN et l’IEnKS ont ensuite été testés en 2D puis 3D avec le module atmosphérique du modèle open-source de CFD Code_Saturne.Le premier cas d’étude avec Code_Saturne correspond à une application réelle d’estimation de potentiel éolien dans une région montagneuse au relief très accidenté où trois mâts de mesure fournissent des observations de vent. Le second cas correspond à une étude de dispersion de polluants en milieu urbain, basé sur les mesures de vent et de concentration provenant de la campagne « Mock Urban Setting Test » aux États-Unis. Dans ce second cas, la turbulence est également incluse dans les CL. Dans les deux études, une partie des observations est utilisée pour l’assimilation et le reste pour la validation des résultats.Les expériences menées sur le premier cas ont révélé que les modèles de CFD présentent des non-linéarités trop fortes (recirculations derrière les obstacles) pour l’algorithme de BFN, fondé sur une hypothèse de linéarité. L'étude de cette méthode n'a donc pas été poursuivie. En revanche, les deux études ont montré la capacité de l'IEnKS à réduire l'erreur et l'incertitude sur les CL grâce à l'assimilation d'une petite dizaine d'observations, en un nombre raisonnable de calculs. Par suite, les champs de vent simulés sont également plus proches des observations de validation et les intervalles de confiance sont réduits. Finalement, l'IEnKS permet d'estimer le potentiel éolien, dans un cas, et les concentrations en polluant, dans l'autre, avec beaucoup plus de précision et d'exactitude.

Atmospheric pollution is a striking regional issue for public health and ecosystems and has major global impacts on climate. Particulate matter (PM) can be of primary or secondary origin and its sources, both natural and anthropogenic, are very heterogeneous in space and time. Hence, many efforts have been made worldwide to get a better knowledge of PM sources, in order to set effective reduction strategies. To this end, several distinct approaches may be used among which: (i) the localization of potential source areas with trajectory-based statistical receptor models (TRMs) which combine PM concentrations observed at receptor sites with computed back trajectories (BTs) of air masses, and (ii) the estimation of PM source contribution and chemical speciation with deterministic chemistry transport models (CTMs) based on emission inventories and detailed chemistry-transport processes. This study aims at testing the coherence between two independant approaches, CTMs and TRMs, for the geographical localization of the sources impacting a region of study. The case study refers to PM10 pollution in the Hauts-de-France region (HdF) in the North of France for the year 2010. The considered TRMs are multi-site Concentration Field (CF) and Potential Source Contribution Function (PSCF) applied to 12 receptor sites using the Zefir package. The considered CTM is CHIMERE using detailed EDGAR European emission data. First, TRMs showed that some given far-located European countries (named “Far-East” countries) could have a strong but infrequent impact on PM10 levels in the study region, while some given nearer European countries (named “Near-East” countries) had a frequent and predominant impact. Then, the contributions of each of these TRM-highlighted regions to PM10 concentrations over the study area were analyzed by CHIMERE simulations. The potential influence of another non TRM-highlighted region (i.e. the “British Isles”) was also studied for comparison. The CTM results confirmed the prevalence of the Near-East source area in terms of mass contribution throughout the year and particularly during high-concentration periods. Therefore results obtained from CHIMERE CTM and multi-site CF and PSCF TRMs showed consistency, highlighting the interest for further comparison of both CTMs and TRMs independent approaches in other regions as well as for other pollutants.

International audience; Deposition is a key process in atmospheric transport modelling of radionuclides consecutive to an accidental release. Following the emission and the atmospheric transport, it is the final step to obtain a map of deposit, on which relies a long-term crisis management. The Fukushima Daiichi Nuclear Power Plant accident of 11th March 2011 led to a significant release of radionuclides in the environment. Most releases were dispersed over the Pacific Ocean whereas about 20% were deposited on the Japan main island causing areas of significant deposit. Numerous radiological measurements taken in the Japanese environment enabled the scientists to substantially reconstruct the main sequences of release to identify the probable trajectories of the radioactive plumes, and to link them with precipitation data to explain the areas of deposition. Fortunately, easurements are all available together at certain location and can be compared to each other: 137Cs hourly air concentrations retrieved from filter tapes of air quality monitoring sites, hourly gamma dose rate and meteorological data from the AMEDAS monitoring network which provides rain-gauges, rain radars and visibility detection. This multiple point of view permitted to establish some lessons about the wet deposition process in case of a nuclear release. The measurements were supplemented by modelling techniques. The most significant progress come from the quantification of the atmospheric releases, the improvement of meteorological data to better take into account the influence of the complex orography on the plumes trajectories and the modelling of deposition processes. The analysis shows some necessary improvements to be done for wet deposition modelling. Several factors are of importance: scavenging of plumes in altitude, impact of light rains in particular before rainfalls. These features are now taken into account in the IRSN operational atmospheric transport model.

Data assimilation consists in the estimation of a physical system state. This estimation should optimally combine erroneous observations together with imperfect numerical simulations of this system. In practice, this estimation is the initial state of a dynamical system. It can be used to precisely predict the system evolution. Especially in geophysics where data sets are substantial.A first strategy is based on the maximum a posteriori estimation. Thus the estimation is solution of an optimization problem. This strategy called 4DVar often requires the computation of the model and observation operator adjoints. This operation is time consuming in the forecasting system development. A second strategy analyses the system state sequentially. It is based on “ensemble” techniques. Here, perturbations of the system state allow to estimate its statistics sequentially thanks to the Kalman equations.Both strategies was recently successfully combined in EnVar methods currently used in operating systems. They benefit from both: an efficient treatment of the operators nonlinearity through variational optimization techniques together with statistics and derivatives estimation through ensembles. The IEnKS is an archetype of such EnVar methods. It uses a data assimilation window (DAW) which is time shifted each cycle to combine both strategies. Various DAW parameterizations lead to non equivalent assimilations when the system dynamics are non linear.In particular, long DAWs reduce the frequency of the prior density Gaussian approximation. This results in a performance improvement but only to some extent. After, the cost function variational minimization fails because of its complex shape. A solution called “Quasi static variational assimilation” gradually adds observations to the cost function during multiple minimizations. The thesis second chapter generalizes the QSVA to EnVar methods. Theoretical and numerical aspects of the QSVA applied to the IEnKS are adressed.However, the QSVA relies on the absence of model error. Indeed, the information contained in a remote in time observation may be deteriorated by model error. The thesis third chapter is dedicated to model error introduction in the IEnKS. The IEnKS-Q, a 4D ensemble variational method solving sequentially the smoothing problem with model error, is built in this chapter. Unfortunately, with model error, a state trajectory is not anymore determined by its initial condition. The number of parameters required to describe its statistics increase with the DAW length. When this number is paired with the number of model evaluations, the consequences on the computing time are disastrous. A proposed solution is to dissociate those quantities with anomaly matrices decompositions. In this case, the IEnKS-Q is as expensive as an IEnKS in terms of model evaluations, L'assimilation de données consiste à calculer une estimation de l'état d'un système physique. Cette estimation doit alors combiner de façon optimale des observations entachées d'erreurs de mesure et des modèles numériques imparfaits permettant de simuler le système physique. En pratique, l'assimilation de données sert à estimer l’état initial d’un système dynamique. Cet état analysé peut ensuite être utilisé pour prévoir le comportement de ce système, notamment dans les systèmes géophysiques où les jeux de données sont conséquents.Une première approche repose sur une estimation de l’état initial basée sur le principe du maximum a posteriori. Il s’agit alors de résoudre un problème d’optimisation, souvent par des techniques utilisant le gradient des opérateurs. Cette approche, appelée 4DVar, nécessite le calcul de l’adjoint du modèle et de l'opérateur d'observation, ce qui est une tâche consommatrice en temps de développement des systèmes de prévision. Une seconde approche permettant de résoudre séquentiellement le problème d’assimilation est basée sur les techniques dites « d’ensemble ». Ici, des perturbations a priori de l'état du système permettent d’estimer des statistiques. Ces moments sont alors utilisés dans les formules de Kalman pour obtenir des approximations de l’état du système a posteriori.Ces deux approches ont été récemment combinées avec succès dans les méthodes de type EnVar aujourd'hui utilisées dans les systèmes opérationnels de prévision. Elles bénéficient donc d'une gestion efficace de la non linéarité au travers des méthodes d'optimisation variationnelle et permettent l'estimation de statistiques et de dérivées à l'aide des ensembles. L'IEnKS est un archétype de ces méthodes EnVar. Pour combiner les deux approches précédentes, il utilise une fenêtre d'assimilation qui est translatée entre chaque cycle. Différents paramétrages de la fenêtre d'assimilation conduisent à différentes stratégies d'assimilation non équivalentes lorsque la dynamique du système est non linéaire.En particulier, les longues fenêtres d'assimilation réduisent la fréquence de l'approximation Gaussienne des densités a priori. Il en résulte une amélioration des performances jusqu'à un certain point. Au delà, la complexité structurelle de la fonction de coût met l'analyse variationnelle en défaut. Une solution nommée “quasi statique variational assimilation” (QSVA) permet d'atténuer ces problèmes en ajoutant graduellement les observations à la fonction de coût du 4DVar. Le second chapitre de thèse généralise cette technique aux méthodes EnVar et s'intéresse plus précisément aux aspects théoriques et numériques du QSVA appliqués à l'IEnKS.Cependant, l’intérêt du QSVA repose sur la perfection du modèle pour simuler l'évolution de l'état. En effet, la pertinence d'une observation temporellement éloignée pour estimer l'état peut être remise en cause en présence d'erreur modèle. Le troisième chapitre est consacré à l'introduction d'erreur modèle au sein de l'IEnKS. Il y sera donc construit l'IEnKS-Q, une méthode 4D variationnelle d'ensemble résolvant séquentiellement le problème de lissage en présence d'erreur modèle. Malheureusement, en présence d'erreur modèle, une trajectoire n'est plus déterminée par son état initial. Le nombre de paramètres nécessaires à la caractérisation de ses statistiques augmente alors avec la longueur de la fenêtre d'assimilation. Lorsque ce nombre va de pair avec le nombre d'évaluations du modèle, les conséquences pour le temps de calcul sont catastrophiques. La solution proposée est alors de découpler ces quantités avec une décomposition des matrices d'anomalies. Dans ce cas, l'IEnKS-Q n'est pas plus coûteux que l'IEnKS en nombre d'évaluations du modèle

International audience; In this paper we focus on airborne hyperspectral imaging methodology to characterize PM (Particulate Matter) near industrial emission sources. Two short-term intensive campaigns were carried out in the vicinity of a refinery in the South of France, in September 2015 and February 2016. Different protocols of in-situ PM measurements were performed, at stack measurements (flow rate and offline chemical analysis) and on-line measurement at the refinery border (size distribution, concentration and chemistry of aerosols). A multi temporal methodology to retrieve aerosol type, to map the aerosol concentration and to quantify mass flow rate from airborne hyperspectral data is described in this paper. This method applied to the refinery detected plume from the main stack yields a black carbon to sulfate ratio of 10/90 in mass inside the plume, with an average size distribution smaller than 100 nm. These results are in a good agreement with on-line analysis of aerosols at refinery border. The resulting quantitative map with a metric spatial resolution leads to a flow rate estimated of about 1g/s and is in a good agreement with in-situ stack measurements and modelling.; Ce papier propose une méthodologie de caractérisation des PM (Particulate Matter) proches des sources industrielles par imagerie hyperspectrale aéroportée. Il s’appuie sur deux campagnes de mesures réalisées autour d’une raffinerie dans le sud de la France en Septembre 2015 et Février 2016. Différents protocoles de mesures in-situ des PM ont été réalisées dans ce cadre, mesures au niveau de l’émissaire principal (debit et analyse chimique) et mesures on-line au voisinage de la raffinerie (distribution en taille, concentration et analyse chimique). Une méthodologie multi-temporelle dédiée à l’inversion des types d’aérosols présents dans le panache permettant de fournir une carte en concentration d’aérosols et d’estimer le débit est décrite dans ce papier. Appliquée à ce cas d’étude, les resultats hyperspectraux ont permis d’estimer un rapport suie/sulfate de 10/90 en masse dans le panache avec une distribution en taille inférieure à 100nm. Ces résultats sont en bon accord avec les mesures in-situ. Les cartes en concentration obtenue à une resolution métrique permettent d’estimer un debit de l’ordre de 1g/s en bon accord avec les mesures in-situ au niveau de l’émissaire et avec les résultats de modélisation de dispersion des aerosols.

Recent progress in machine learning has shown how to forecast and, to some extent, learn the dynamics of a model from its output, resorting in particular to neural networks and deep learning techniques. We will show how the same goal can be directly achieved using data assimilation techniques without leveraging on machine learning software libraries, with a view to high-dimensional models. The dynamics of a model are learned from its observation and an ordinary differential equation (ODE) representation of this model is inferred using a recursive nonlinear regression. Because the method is embedded in a Bayesian data assimilation framework, it can learn from partial and noisy observations of a state trajectory of the physical model. Moreover, a space-wise local representation of the ODE system is introduced and is key to coping with high-dimensional models. It has recently been suggested that neural network architectures could be interpreted as dynamical systems. Reciprocally, we show that our ODE representations are reminiscent of deep learning architectures. Furthermore, numerical analysis considerations of stability shed light on the assets and limitations of the method. The method is illustrated on several chaotic discrete and continuous models of various dimensions, with or without noisy observations, with the goal of identifying or improving the model dynamics, building a surrogate or reduced model, or producing forecasts solely from observations of the physical model.

In this study, assumptions (ideality and thermodynamic equilibrium) commonly made in three-dimensional (3-D) air quality models were reconsidered to evaluate their impacts on secondary organic aerosol (SOA) formation over Europe. To investigate the effects of non-ideality, dynamic mass transfer and aerosol viscosity on the SOA formation, the Secondary Organic Aerosol Processor (SOAP) model was implemented in the 3-D air quality model Polyphemus. This study presents the first 3-D modeling simulation which describes the impact of aerosol viscosity on the SOA formation. The model uses either the equilibrium approach or the dynamic approach with a method specially designed for 3-D air quality models to efficiently solve particle-phase diffusion when particles are viscous. Sensitivity simulations using two organic aerosol models implemented in Polyphemus to represent mass transfer between gas and particle phases show that the computation of the absorbing aerosol mass strongly influences the SOA formation. In particular, taking into account the concentrations of inorganic aerosols and hydrophilic organic aerosols in the absorbing mass of the aqueous phase increases the average SOA concentration by 5 % and 6 %, respectively. However, inorganic aerosols influence the SOA formation not only because they constitute an absorbing mass for hydrophilic SOA, but also because they interact with organic compounds. Non-ideality (short-, medium- and long-range interactions) was found to influence SOA concentrations by about 30 %. Concerning the dynamic mass transfer for the SOA formation, if the viscosity of SOA is not taken into account and if ideality of aerosols is assumed, the dynamic approach is found to give generally similar results to the equilibrium approach (indicating that equilibrium is an efficient hypothesis for inviscid and ideal aerosols). However, when a non-ideal aerosol is assumed, taking into account the dynamic mass transfer leads to a decrease of concentrations of the hydrophilic compounds (compared to equilibrium). This decrease is due to differences in the values of activity coefficients, which are different between values computed for bulk aerosols and those for each size section. This result indicates the importance of non-ideality on the dynamic evolution of SOA. For viscous aerosols, assuming a highly viscous organic phase leads to an increase in SOA concentrations during daytime (by preventing the evaporation of the most volatile organic compounds). The partitioning of nonvolatile compounds is not affected by viscosity, but the aging of more volatile compounds (that leads to the formation of the less volatile compounds) slows down as the evaporation of those compounds is stopped due to the viscosity of the particle. These results imply that aerosol concentrations may deviate significantly from equilibrium as the gas–particle partitioning could be higher than predicted by equilibrium. Furthermore, although a compound evaporates in the simulation using the equilibrium approach, the same compound can condense in the simulation using the dynamic approach if the particles are viscous. The results of this study emphasize the need for 3-D air quality models to take into account the effect of non-ideality on SOA formation and the effect of aerosol viscosity for the more volatile fraction of semi-volatile organic compounds.

Air quality is a major health and environmental issue worldwide. Similarly, the accuracy of wind resource assessment triggers significant economic and environmental repercussions. In order to study these two topics, it is necessary to accurately determine local wind fields using numerical models of micrometeorology. Such simulations are extremely sensitive to meteorological conditions at the domain borders. Up to present, the boundary conditions (BC) were estimated based on the results of larger scale simulations, which provide information that is not accurate enough, or even incomplete, for local scale purposes. As a matter of fact, the lack of knowledge about the BC represents a major source of error and uncertainty for micrometeorological studies.The potential sites for wind farm installation as well as built environments (urban areas or industrial sites) can be equipped with instruments measuring meteorological variables or pollutant concentration. The observations provided by these instruments represent a second source of information, insufficiently exploited for micrometeorological studies. Indeed, the in situ measurements are perturbed by the complex geometrical features on sites and might be difficult to exploit. In order to improve the exactitude and the accuracy of the BC, and consequently of the locale-scale atmospheric simulations, data assimilation (DA) methods, suited to this micrometeorological problem, could be applied to take benefit from these available observations.So far, DA methods have been mainly developed for large-scale meteorology and employed to correct the initial conditions (IC). In order to broaden the application scope of DA to micrometeorology, existing DA methods must be adapted to be able to correct the BC instead of IC.Two of the existing DA methods seem compatible with computational fluid dynamics (CFD) models used for micrometeorology over complex geometries: the back and forth nudging (BFN) algorithm and the iterative ensemble Kalman smoother (IEnKS). We have adapted these two methods, from a theoretical perspective, so as to include the BC in the control variables. The performances of the adapted versions of the BFN algorithm and the IEnKS have first been assessed with a simplified, 1D model of atmospheric flow with two layers, based on the shallow-water equations. The BFN algorithm and the IEnKS have then been tested in 2D and 3D with the atmospheric module of the open-source CFD model Code_Saturne.The first study case with Code_Saturne corresponds to a real application of wind resource assessment in a mountainous region with steep topography where three meteorological masts have been installed during a few months and provide in situ wind observations. The second case is a study of pollutant dispersion in an urban area, based on the measurements of wind and pollutant concentration coming from the “Mock Urban Setting Test” field campaign carried out in the USA. In this second case, the turbulence is also included in the BC and thus in the control variables. For both studies, some observations are assimilated and the remaining ones are used to validate the results.The experiences performed for the wind resource assessment study have revealed that the CFD models present too strong nonlinearities (flow recirculation after obstacles) for the BFN algorithm, which is based on a linearity assumption. However, both cases have shown the ability of the IEnKS to reduce the error and the uncertainty of the BC by assimilating a few observations, in operationally affordable conditions. Consequently, the simulated wind fields with Code_Saturne are also closer to the validation observations and the confidence intervals are reduced. Eventually, the IEnKS allows, in one case to estimate the wind potential, and in the other case to build the pollution maps, with much more exactitude and accuracy., La qualité de l’air est un enjeu sanitaire et environnemental majeur. Par ailleurs, l'estimation précise des potentiels éoliens est la source d’importantes retombées économiques et environnementales. Pour étudier ces deux sujets, il est nécessaire de reconstituer précisément les champs de vent locaux grâce à des modèles numériques de micrométéorologie. Ces simulations sont extrêmement sensibles aux conditions météorologiques aux limites du domaine d’étude. Jusqu’à présent, les conditions aux limites (CL) étaient estimées à partir de simulations à plus grande échelle, qui fournissent des informations peu adaptées à l’utilisation à l'échelle locale car imprécises, voire incomplètes. Par conséquent, la méconnaissance des CL représente une source majeure d’erreur et d’incertitude dans les études micrométéorologiques.Les sites susceptibles d’accueillir un parc éolien et les environnements bâtis (quartiers urbains ou sites industriels) peuvent être équipés d’instruments de mesures météorologiques ou de concentration de polluants. Les observations qu’ils fournissent constituent une seconde source d’information, jusqu’à ce jour peu exploitée pour les études micrométéorologiques. En effet, les mesures in situ sont perturbées par la géométrie complexe des sites étudiés. Afin d'améliorer la précision des CL et donc des simulations atmosphériques à l'échelle locale, des méthodes d'assimilation de données (AD) adaptées à cette problématique pourraient permettre de mettre à profit les observations disponibles.Jusqu’à présent, les méthodes d’AD ont été principalement développées pour la météorologie à grande échelle et ont donc surtout été utilisées pour corriger les conditions initiales (CI). Afin d'élargir le champ d'application de l’AD à la micrométéorologie, il faut adapter les méthodes existantes pour qu'elles permettent de corriger les CL plutôt que les CI.Deux méthodes d’AD semblent compatibles avec les modèles de mécanique des fluides (CFD) utilisés pour la micrométéorologie en géométrie complexe : l’algorithme de nudging direct et rétrograde (BFN) et le lisseur de Kalman d’ensemble itératif (IEnKS). Nous avons adapté ces deux méthodes d’un point de vue théorique pour inclure les CL dans les variables de contrôle. Les performances des versions adaptées du BFN et de l'IEnKS ont tout d'abord été étudiées avec un modèle simplifié d’écoulement atmosphérique à deux couches en 1D, basé sur les équations de Saint-Venant. Le BFN et l’IEnKS ont ensuite été testés en 2D puis 3D avec le module atmosphérique du modèle open-source de CFD Code_Saturne.Le premier cas d’étude avec Code_Saturne correspond à une application réelle d’estimation de potentiel éolien dans une région montagneuse au relief très accidenté où trois mâts de mesure fournissent des observations de vent. Le second cas correspond à une étude de dispersion de polluants en milieu urbain, basé sur les mesures de vent et de concentration provenant de la campagne « Mock Urban Setting Test » aux États-Unis. Dans ce second cas, la turbulence est également incluse dans les CL. Dans les deux études, une partie des observations est utilisée pour l’assimilation et le reste pour la validation des résultats.Les expériences menées sur le premier cas ont révélé que les modèles de CFD présentent des non-linéarités trop fortes (recirculations derrière les obstacles) pour l’algorithme de BFN, fondé sur une hypothèse de linéarité. L'étude de cette méthode n'a donc pas été poursuivie. En revanche, les deux études ont montré la capacité de l'IEnKS à réduire l'erreur et l'incertitude sur les CL grâce à l'assimilation d'une petite dizaine d'observations, en un nombre raisonnable de calculs. Par suite, les champs de vent simulés sont également plus proches des observations de validation et les intervalles de confiance sont réduits. Finalement, l'IEnKS permet d'estimer le potentiel éolien, dans un cas, et les concentrations en polluant, dans l'autre, avec beaucoup plus de précision et d'exactitude.