Precipitation (P) and precipitable water (W) are important components of the hydrological cycles in the earth system, and their seasonal cycles are closely related to monsoon circulations over monsoon regions. Through theoretical analyses and extensive analysis of data from in-situ measurements, satellite remote sensing, and regional reanalysis, significant progress has been made (via four peer-reviewed publications) in four areas related to P, W, and monsoon onset and retreat. First, based on the normalized W index, a novel unified method is proposed to determine global monsoon onset and retreat dates. The results are consistent with those obtained from different local criteria. Second, theoretical and data analyses demonstrate that, because of the large annual range of temperature, W can increase from winter to summer anywhere except in the tropics, including both monsoon and nonmonsoon regions. Third, while the seasonal variation of P is, in general, caused by complex processes (e.g., atmospheric circulations), thermodynamic derivations and data analysis demonstrate that the variation of P from winter to summer can be easily understood from the comparative strength between the change of water vapor and the change of temperature. In monsoon regions, the change of water vapor from winter to summer is much greater than the change of temperature, so P has an in-phase relation with W. While in some of the nonmonsoon regions, where winter is the rainy season, the change of temperature is much greater than the change of water vapor, leading to an out-of-phase relation between P and W, and, relative to summer, the coldness of the winter air is much more significant than its dryness. Finally, the satisfactory performance of the globally unified monsoon index can be understood by comparing the seasonal cycles of P and W. The significant positive correlations between P and W at seasonal and synoptic scales imply that W has the ability to indicate both the means and the interannual variations of the monsoon onset and retreat. Since large increase of W from winter to summer can occur in both monsoon and nonmonsoon regions, the global monsoon regions cannot be obtained from the seasonal change of W.

This study investigated the effects of a seasonally variable forage resource on herbivore population dynamics. This involved estimating the relative importance of environmental conditions, and the accessible and used forage resources, at different stages of the seasonal cycle to herbivores in different life-stages and at different points in the reproductive cycle. This study was carried out in the Richtersveld region in South Africa, using goats kept by semi-nomadic Nama pastoralists. In the main study site, the Richtersveld National Park (RNP), herd movements follow a general seasonal migratory pattern: herds are based in the riparian zone of the Orange River during the dry season, and on plains away from the river in the wet season. Over 800 uniquely marked female goats in three life-stages (adults, yearlings and kids) were monitored over a three year period (2007 to 2009). These goats were weighed at 2 - 3 month intervals to provide an estimate of body condition. Browse availability in the riparian zone was estimated using measurements at an individual branch-level and a whole tree-level. FPAR satellite imagery was used to estimate forage abundance outside the riparian zone. Goat density was mapped for each week of the study using census data and the herd positions. Goat body condition, survival rates and fecundity rates for each life-stage were modelled as a response to forage availability, density and climatic conditions. The riparian zone in the RNP was found to function as the key resource of the RNP goat population. Forage depletion by goat browsing resulted in a negative feedback on goat body condition. This decline in body condition was directly related to lower adult survival over the dry season. Fecundity was also most influenced by dry season conditions through the negative effect of poor body condition on pregnancy rates and birth rates. Asymmetric competition between life-stages, resulting from the riparian browse profile being depleted from the bottom-up, was predicted to have a strong effect on goat demography by contributing to differences in body condition and survival rates between life-stages. Wet season conditions appeared to have little effect on goat population dynamics, either through increased neonate survival or through a mass carry-over effect influencing dry season survival. Goat body condition and vital rates were compared between the RNP and the neighbouring Kuboes rangeland, which does not have access to the Orange River, to assess the impact of differences in their dry season forage resource. The long-term size and variability of the livestock population in the RNP was also compared with livestock dynamics in Paulshoek, a rangeland 250 km south east of the RNP. The a priori predictions of relative population dynamics in each region, based on perceived differences in the nature of the key resource in each region, were largely supported.

Both physical circulation and biogeochemical characteristics are unique in the Southern Ocean (SO) region, and are fundamentally different from those of the northern hemisphere. Moreover, according to previous research, the oceanic response to the trend of the Southern Annual Mode (SAM) has profound impacts on the future oceanic uptake of carbon dioxide in the SO. In other words, the climate and circulation of the SO are strongly coupled to the overlying atmospheric variability. However, while we have understanding on the SO physical circulation and have the ability to predict the future changes of the SO climate and physical processes, the link between the SO physical processes, the air-sea carbon flux, and correlated climate variability remains unknown. Even though scientists have been studying the spatial and temporal variability of the SO carbon flux and the associated biogeochemical processes, the spatial patterns and the magnitudes of the air-sea carbon flux do not agree between models and observations. Therefore, in this study, we utilized a modified version of a general circulation model (GCM) to performed realistic simulations of the SO carbon on seasonal to interannual timescales, and focused on the crucial physical and biogeochemical processes that control the carbon flux. The spatial pattern and the seasonal cycle of the air-sea carbon dioxide flux is calculated, and is broadly consistent with the climatological observations. The variability of air-sea carbon flux is mainly controlled by the gas exchange rate and the partial pressure of carbon dioxide, which is in turn controlled by the compensating changes in temperature and dissolved inorganic carbon. We investigated the seasonal variability of dissolved inorganic carbon based on different regional processes. Furthermore, we also investigated the dynamical adjustment of the surface carbon flux in response to the different gas exchange parameterizations, and conclude that parameterization has little impact on spatially integrated carbon flux. Our simulation well captured the SO carbon cycle variability on seasonal to interannual timescales, and we will improve our model by employ a better scheme of nutrient cycle, and consider more nutrients as well as ecological processes in our future study.

Monsoons and Intertropical Convergence Zones (ITCZ) exhibit variability at various temporal and spatial scales. The temporal scale of variability encompasses scales from the intraseasonal through interannual to interdecadal time scales. Anthropogenic climate change can also have an impact on ITCZ and monsoons. Thus it is necessary to assess the ability of coupled ocean atmospheric models (commonly known as AOGCM) to simulate these aspects of variability of tropical climate. This has been studied with simulations from 20 AOGCMs and their AGCM from IPCCAR4 archive. In addition, we have used our own 100 year simulation with CCSM2 and also simulations with its AGCM viz. CAM2. Our analysis shows that most model have significant bias in tropical rainfall and SST. Most models underestimate SST except over a few regions such as the Eastern boundaries of Atlantic and Pacific Oceans. The AGCMs which are forced with observed SSTs have much higher annual mean rainfall as compared to AOGCMs. There is a strong correlation between error in shortwave reflectance at the top of the atmosphere and error in SST. The ability of coupled ocean-atmosphere models and their atmosphere-alone counterparts to simulate the seasonal cycle of rainfall over major monsoon regions and also over oceanic ITCZ. It is found that over the Indian monsoon region, most AGCMs overestimate the seasonal cycle while AOGCMs have a more realistic seasonal cycle. This inspite of the fact that most AOGCMs underestimate the SST over the Indian region. It is shown that this is related to errors in precipitable water-rainfall relationship in most models i.e. for a given amount of precipitable water, most models overestimate the rainfall. Thus lower SST reduces the precipitable water and hence the amount of rainfall is reduced. Therefore, the mutual cancellation of errors leads to a more realistic seasonal cycle in AOGCMs. The seasonal cycle over Africa was analysed with the help of a diagnostic model. Over Southern Africa, most models show simulate a less stable atmosphere and hence the rainfall is overestimated. A technique based on Continous Wavelet Transform in Space and Time (CWTST) has been modified to seperate northward and southward propagating modes of BSISO over the Indian and West Pacific regions. It was seen that over the Indian region, northward propagating modes were more prominent in comparison to southward modes. It was also found that the predominant spatial scale (of about 30o) did not show much interannual variability but the associated temporal scale showed significant variation. Both AOGCMs and AGCMs simulations were analysed to investigate the impact of coupling on intraseasonal activity. Most AOGCMs were able to simulate the predominant spatial scale but were unable to simulate the associated temporal scale correctly. These problems persisted with AGCMs also. It was also found that for AGCMs, there were some variations between ensemble members of the AGCMs. Comparing BSISO in increased GHG scenarios with present day simulations we found that in general, power in the spectrum increases. This could be related to higher mean precipitation that has been simulated by most AOGCMs when GHG are increased. The interannual variability in the tropics with special reference to Tropical Biennial Oscillation (TBO) and ENSO has been studied. The changes in these modes of variability due to anthropogenic climate change has also been assessed. We found that in most models over the Nino3.4 region, the mode of variation shifts from a near-four period (in pre-industrial simulations) to that of TBO mode in increased GHG (green house gas) scenario. This suggests that with increasing GHGs, ENSO quasi-periodicity might shift to about two years. It is also interesting to note that for observed rainfall, OLR and 850 hPa winds, the TBO mode has higher variance over the Eastern Indian Ocean, indicating that the TBO mode might be related to Indian Ocean Dipole Mode and EQUINOO (Equatorial Indian Ocean Oscillation).

This study investigates the upper ocean circulation along the west Australian coast, based on recent observations (WOCE ICM6, 1994/96) and numerical output from the 1/6 degree Parallel Ocean Program model (POP11B 1993/97). Particularly, we identify the source regions of the Leeuwin Current, quantify its mean and seasonal variability in terms of volume, heat and salt transports, and examine its heat balance (cooling mechanism). This also leads to further understanding of the regional circulation associated with the Leeuwin Undercurrent, the Eastern Gyral Current and the southeast Indian Subtropical Gyre. The tropical and subtropical sources of the Leeuwin Current are understood from an online numerical particle tracking. Some of the new findings are the Tropical Indian Ocean source of the Leeuwin Current (in addition to the Indonesian Throughflow/Pacific); the Eastern Gyral Current as a recirculation of the South Equatorial Current; the subtropical source of the Leeuwin Current fed by relatively narrow subsurface-intensified eastward jets in the Subtropical Gyre, which are also a major source for the Subtropical Water (salinity maximum) as observed in the Leeuwin Undercurrent along the ICM6 section at 22 degrees S. The ICM6 current meter array reveals a rich vertical current structure near North West Cape (22 degrees S). The coastal part of the Leeuwin Current has dominant synoptic variability and occasionally contains large spikes in its transport time series arising from the passage of tropical cyclones. On the mean, it is weaker and shallower compared to further downstream, and it only transports Tropical Water, of a variable content. The Leeuwin Undercurrent carries Subtropical Water, South Indian Central Water and Antarctic Intermediate Water equatorward between 150/250 to 500/750 m. There is a poleward flow just below the undercurrent which advects a mixed Intermediate Water, partially associated with outflows from the Red Sea and Persian Gulf. Narrow bottom-intensified currents are also observed. The 5-year mean model Leeuwin Current is a year-round poleward flow between 22 degrees S and 34 degrees S. It progressively deepens, from 150 to 300 m depth. Latitudinal variations in its volume transport are a response to lateral inflows/outflows. It has double the transport at 34 degrees S (-2.2 Sv) compared to at 22 degrees S (-1.2 Sv). These model estimates, however, may underestimate the transport of the Leeuwin Current by 50%. Along its path, the current becomes cooler (6 degrees C), saltier (0.6 psu) and denser (2 kg m -3). At seasonal scales, a stronger poleward flow in May-June advects the warmest and freshest waters along the west Australian coast. This advection is apparently spun up by the arrival of a poleward Kelvin wave in April, and reinforced by a minimum in the equatorward wind stress during July. In the model heat balance, the Leeuwin Current is significantly cooled by the eddy heat flux divergence (4 degrees C out of 6 degrees C), associated with mechanisms operating at submonthly time scales. However, exactly which mechanisms it is not yet clear. Air-sea fluxes only account for ~30% of the cooling and seasonal rectification is negligible. The eddy heat divergence, originating over a narrow region along the outer edge of the Leeuwin Current, is responsible for a considerable warming of a vast area of the adjacent ocean interior, which is then associated with strong heat losses to the atmosphere. The model westward eddy heat flux estimates are considerably larger than those associated with long lived warm core eddies detaching from the Leeuwin Current and moving offshore. This suggests that these mesoscale features are not the main mechanism responsible for the cooling of the Leeuwin Current. We suspect instead that short lived warm core eddies might play an important role.

ENGLISH:<p>The natural nitrogen cycle has been and is significantly perturbed by anthropogenic emissions of reactive nitrogen (Nr) compounds into the atmosphere, resulting from our production of energy and food. In the last century global ammonia (NH3) emissions have doubled and represent nowadays more than half of total the Nr emissions. NH3 is also the principal atmospheric base in the atmosphere and rapidly forms aerosols by reaction with acids. It is therefore a species of high relevance for the Earth's environment, climate and human health (Chapter 1). As a short-lived species, NH3 is highly variable in time and space, and while ground based measurements are possible, they are sparse and their spatial coverage is largely heterogeneous. Consequently, global spatial and temporal patterns of NH3 emissions are poorly understood and account for the largest uncertainties in the nitrogen cycle. The aim of this work is to assess distributions and saptiotemporal variability of NH3 using satellite measurements to improve our understanding of its contribution to the global nitrogen cycle and its related effects.<p><p>Recently, satellite instruments have demonstrated their abilities to measure NH3 and to supplement the sparse surface measuring network by providing global total columns daily. The Infrared Atmospheric Sounding Interferometer (IASI), on board MetOp platforms, is measuring NH3 at a high spatiotemporal resolution. IASI circles the Earth in a polar Sun-synchronous orbit, covering the globe twice a day with a circular pixel size of 12km diameter at nadir and with overpass times at 9:30 and 21:30 (local solar time when crossing the equator). An improved retrieval scheme based on the calculation of Hyperspectral Range Index (HRI) is detailed in Chapter 2 and compared with previous retrieval methods. This approach fully exploits the hyperspectral nature of IASI by using a broader spectral range (800-1200 cm-1) where NH3 is optically active. It allows retrieving total columns from IASI spectra globally and twice a day without large computational resources and with an improved detection limit. More specifically the retrieval procedure involves two steps: the calculation of a dimensionless spectral index (HRI) and the conversion of this index into NH3 total columns using look-up tables (LUTs) built from forward radiative transfer simulations under various atmospheric conditions. The retrieval also includes an error characterization of the retrieved column, which is of utmost importance for further analysis and comparisons. Global distributions using five years of data (1 November 2007 to 31 October 2012) from IASI/MetOp-A are presented and analyzed separately for the morning and evening overpasses. The advantage of the HRI-based retrieval scheme over other methods, in particular to identify smaller emission sources and transport patterns over the oceans is shown. The benefit of the high spatial sampling and resolution of IASI is highlighted with the regional distribution over China and the first four-year time series are briefly discussed.<p><p>We evaluate four years (1 January 2008 to 31 December 2011) of IASI-NH3 columns from the morning observations and of LOTOS-EUROS model simulations over Europe and Western Russia. We describe the methodology applied to account for the variable retrieval sensitivity of IASI measurements in Chapter 3. The four year mean distributions highlight three main agricultural hotspots in Europe: The Po Valley, the continental part of Northwestern Europe, and the Ebro Valley. A general good agreement between IASI and LOTOS-EUROS is shown, not only over source regions but also over remote areas and over seas when transport is observed. The yearly analyses reveal that, on average, the measured NH3 columns are higher than the modeled ones. Large discrepancies are observed over industrial areas in Eastern Europe and Russia pointing to underestimated if not missing emissions in the underlying inventories. For the three hotspots areas, we show that the seasonality between IASI and LOTOS-EUROS matches when the sensitivity of the satellite measurements is taken into account. The best agreement is found in the Netherlands, both in magnitude and timing, most likely as the fixed emission timing pattern was determined from experimental data sets from this country. Moreover, comparisons of the daily time series indicate that although the dynamic of the model is in reasonable agreement with the measurements, the model may suffer from a possible misrepresentation of emission timing and magnitude. Overall, the distinct temporal patterns observed for the three sites underline the need for improved timing of emissions. Finally, the study of the Russian fires event of 2010 shows that NH3 modeled plumes are not enough dispersed, which is confirmed with a comparison using in situ measurements.<p><p>Chapter 4 describes the comparisons of IASI-NH3 measurements with several independent ground-based and airborne data sets. Even though the in situ data are sparse, we show that the yearly distributions are broadly consistent. For the monthly analyzes we use ground-based measurements in Europe, China and Africa. Overall, IASI-derived concentrations are in fair agreement but are also characterized by less variability. Statistically significant correlations are found for several sites, but low slopes and high intercepts are calculated in all cases. At least three reasons can explain this: (1) the lack of representativity of the point surface measurement for the large IASI pixel, (2) the use of a single profile shape in the retrieval scheme over land, which does therefore not account for a varying boundary layer height, (3) the impact of the averaging procedure applied to satellite measurements to obtain a consistent quantity to compare with the in situ monthly data. The use of hourly surface measurements and of airborne data sets allows assessing IASI individual observations. Much higher correlation coefficients are found in particular when comparing IASI-derived volume mixing ratio with vertically resolved measurements performed from the NOAA WP-3D airplane during CalNex campaign in 2010. The results demonstrate the need, for validation of the satellite columns, of measurements performed at various altitudes and covering a large part of the satellite footprint.<p><p>The six-year of IASI observations available at the end of this thesis are used to analyze regional time series for the first time (Chapter 5). More precisely, we use the IASI measurements over that period (1 January 2008 to 31 December 2013) to identify seasonal patterns and inter-annual variability at subcontinental scale. This is achieved by looking at global composite seasonal means and monthly time series over 12 regions around the world (Europe, Eastern Russia and Northern Asia, Australia, Mexico, South America, 2 sub-regions for Northern America and South Asia, 3 sub-regions for Africa), considering separately but simultaneously measurements from IASI morning and evening overpasses. The seasonal cycle is inferred for the majority of these regions. The relations between the NH3 atmospheric abundance and emission processes is emphasized at smaller regional scale by extracting at high spatial resolution the global climatology of the month of maxima columns. In some region, the predominance of a single source appears clearly (e.g. agriculture in Europe and North America, fires in central South Africa and South America), while in others a composite of source processes on small scale is demonstrated (e.g. Northern Central Africa and Southwestern Asia).<p><p>Chapter 6 presents the achievements of this thesis, as well as ongoing activities and future perspectives.<p>FRANCAIS:<p>Le cycle naturel de l'azote est fortement perturbé suite aux émissions atmosphériques de composés azotés réactifs (Nr) résultant de nos besoins accrus en énergie et en nourriture. Les émissions d'ammoniac (NH3) ont doublé au cours du siècle dernier, représentant aujourd'hui plus de la moitié des émissions totales de Nr. De plus, le NH3 étant le principal composé basique de notre atmosphère, il réagit rapidement avec les composés acides pour former des aérosols. C'est dès lors un constituant prépondérant pour l'environnement, le climat et la santé publique. Les problématiques environnementales y étant liées sont décrites au Chapitre 1. En tant que gaz en trace le NH3 se caractérise par une importante variabilité spatiale et temporelle. Bien que des mesures in situ soient possibles, elles sont souvent rares et couvrent le globe de façon hétérogène. Il en résulte un manque de connaissance sur l'évolution temporelle et la variabilité spatiale des émissions, ainsi que de leurs amplitudes, qui représentent les plus grandes incertitudes pour le cycle de l'azote (également décrites au Chapitre 1).<p><p>Récemment, les sondeurs spatiaux opérant dans l'infrarouge ont démontré leurs capacités à mesurer le NH3 et par là à compléter le réseau d'observations de surface. Particulièrement, l'Interféromètre Atmosphérique de Sondage Infrarouge (IASI), à bord de la plateforme MetOp, mesure le NH3 à une relativement haute résolution spatiotemporelle. Il couvre le globe deux fois par jour, grâce à son orbite polaire et son balayage autour du nadir, avec un temps de passage à 9h30 et à 21h30 (temps solaire local quand il croise l'équateur). Une nouvelle méthode de restitution des concentrations basée sur le calcul d'un index hyperspectral sans dimension (HRI) est détaillée et comparée aux méthodes précédentes au Chapitre 2. Cette méthode permet d'exploiter de manière plus approfondie le caractère hyperspectral de IASI en se basant sur une bande spectrale plus étendue (800-1200 cm-1) au sein de laquelle le NH3 est optiquement actif. Nous décrivons comment restituer ces concentrations deux fois par jour sans nécessiter de grandes ressources informatiques et avec un meilleur seuil de détection. Plus spécifiquement, la procédure de restitution des concentrations consiste en deux étapes: le HRI est calculé dans un premier temps pour chaque spectre puis est ensuite converti en une colonne totale de NH3 à l'aide de tables de conversions. Ces tables ont été construites sur base de simulations de transfert radiatif effectuées pour différentes conditions atmosphériques. Le processus de restitution des concentrations comprend également le calcul d'une erreur sur la colonne mesurée. Des distributions globales moyennées sur cinq ans (du 1 novembre 2007 au 31 Octobre 2012) sont présentées et analysées séparément pour le passage diurne et nocturne de IASI. L'avantage de ce nouvel algorithme par rapport aux autres méthodes, permettant l'identification de sources plus faibles de NH3 ainsi que du transport depuis les sources terrestres au-dessus des océans, est démontré. Le bénéfice de la haute couverture spatiale et temporelle de IASI est mis en exergue par une description régionale au-dessus de la Chine ainsi que par l'analyse de premières séries temporelles hémisphériques sur quatre ans.<p><p>Au Chapitre 3, nous évaluons quatre ans (du 1 janvier 2008 au 31 décembre 2011) de mesures matinales de IASI ainsi que de simulations du modèle LOTOS-EUROS, effectuées au-dessus de l'Europe et de l'ouest de la Russie. Nous décrivons une méthodologie pour prendre en compte, dans la comparaison avec le modèle, la sensibilité variable de l'instrument IASI pour le NH3. Les comparaisons montrent alors une bonne concordance générale entre les mesures et les simulations. Les distributions pointent trois régions sources: la vallée du Pô, le nord-ouest de l'Europe continentale et la vallée de l'Ebre. L'analyse des distributions annuelles montre qu'en moyenne, les colonnes de NH3 mesurées sont plus élevées que celles simulées, à part pour quelques cas spécifiques. Des différences importantes ont été identifiées au-dessus de zones industrielles en Europe de l'est et en Russie, ce qui tend à incriminer une sub-estimation voire une absence de ces sources dans les inventaires d'émissions utilisés en entrée du modèle. Nous avons également montré que la saisonnalité est bien reproduite une fois la sensibilité des mesures satellites prise en compte. La meilleure concordance entre le modèle et IASI est observée pour les Pays-Bas, ce qui est certainement dû au fait que le profil temporel des émissions utilisé pour les simulations LOTOS-EUROS est basé sur des études expérimentales réalisées dans ce pays. L'étude des séries temporelles journalières indique que la dynamique du modèle est raisonnablement en accord avec les mesures mais pointe néanmoins une possible mauvaise représentation du profil temporel ainsi que de l'ampleur des émissions. Finalement, l'étude des importants feux ayant eu cours en Russie à l'été 2010 a montré que les panaches modélisés sont moins étendus que ceux observés, ce qui a été confirmé grâce à une comparaison avec des mesures sols.<p><p>Le chapitre 4 est dédié à la confrontation des mesures IASI avec différents jeux de données indépendants acquis depuis le sol et par avion. Les distributions globales annuelles sont concordantes, bien que la couverture spatiale des mesures sols soit limitée. Des mesures effectuées à la surface en Europe, en Chine et en Afrique sont utilisées pour les comparaisons mensuelles. Ces dernières révèlent une bonne concordance générale, bien que les mesures satellites montrent une plus faible amplitude de variations de concentrations. Des corrélations statistiquement significatives ont été calculées pour de nombreux sites, mais les régressions linéaires sont caractérisées par des pentes faibles et des ordonnées à l'origine élevées dans tous les cas. Au minimum, trois raisons contribuent à expliquer cela: (1) le manque de représentativité des mesures ponctuelles pour l'étendue des pixels IASI, (2) l'utilisation d'une seule forme de profil vertical pour la restitution des concentrations, qui ne prend dès lors pas en compte la hauteur de la couche limite, (3) l'impact de la procédure utilisée pour moyenner les observations satellites afin d'obtenir des quantités comparables aux mesures sols mensuelles. La prise en compte de mesures en surface effectuées à plus haute résolution temporelle ainsi que de mesures faites depuis un avion permet d'évaluer les observations IASI individuelles. Les coefficients de corrélation calculés sont bien plus élevés, en particulier pour la comparaison avec les mesures effectuées depuis l'avion NOAA WP-3D pendant la campagne CalNex en 2010. Ces résultats démontrent la nécessité de ce type d'observations, effectuées à différentes altitudes et couvrant une plus grande surface du pixel, pour valider les colonnes IASI-NH3.<p><p>Les six ans de données IASI disponibles à la fin de cette thèse sont utilisées pour tracer les premières séries temporelles sub-continentales (Chapitre 5). Plus spécifiquement, nous explorons les mesures IASI durant cette période (du 1 janvier 2008 jusqu'au 31 décembre 2013) pour identifier des structures saisonnières ainsi que la variabilité inter-annuelle à l'échelle sous-continentale. Pour arriver à cela, des moyennes saisonnières composites ont été produites ainsi que des séries temporelles mensuelles au-dessus de 12 régions du globe (Europe, est de la Russie et nord de l'Asie, Australie, Mexique, Amérique du Sud, 2 sous-régions en Amérique du nord et en Asie du sud et 3 sous-régions en Afrique), considérant séparément mais simultanément les mesures matinales et nocturnes de IASI. Le cycle saisonnier est raisonnablement bien décrit pour la plupart des régions. La relation entre la quantité de NH3 atmosphérique et ses sources d'émission est mise en exergue à l'échelle plus régionale par l'extraction à haute résolution spatiale d'une climatologie des mois de colonnes maximales. Dans certaines régions, la prédominance d'un processus source apparait clairement (par exemple l'agriculture en Europe et en Amérique du nord, les feux en Afrique du Sud et en Amérique du Sud), alors que, pour d'autres, la diversité des sources d'émissions est démontrée (par exemple pour le nord de l'Afrique centrale et l'Asie du sud-ouest).<p><p>Le Chapitre 6 reprend brièvement les principaux aboutissements de cette thèse et présente les différentes recherches en cours et les perspectives associées.<p> / Doctorat en Sciences agronomiques et ingénierie biologique / info:eu-repo/semantics/nonPublished

No description available.

La réduction du volume des glaciers et la fusion printanière plus précoce de la neige causée par le réchauffement climatique provoquent des variations du cycle hydrologique à la fois pour les têtes de bassin versant, mais aussi pour les zones situées plus à l’aval. Afin de prédire correctement l’amplitude des changements possibles futurs et d’envisager une gestion adaptée, une bonne connaissance de l’interaction entre les glaciers, le climat et les écoulements hydriques est nécessaire. L’objectif de cette étude est d’évaluer l’effet de la variabilité climatique sur le fonctionnement hydro-glaciologique et ses conséquences sur la disponibilité de l’eau du bassin versant de l’Arve (Alpes françaises) depuis 1960. Ce bassin s’étend sur une surface de 1958 km2 et est composé de cinq bassins versants emboités (Arveyron d’Argentière, Arveyron de la Mer de Glace, Arve au Pont des Favrands, à Sallanches and au Bout du Monde), tous influencés par la fusion glaciaire et nivale mais dans différentes proportions étant donnée la large gamme d’extension de couverture glaciaire s’étalant de 5 á 53%. Ce travail est basé sur des longs jeux de données glaciologiques, météorologiques, hydrologiques et de couverture de neige qui sont issues soit de mesures ponctuelles dans l’espace soit de données obtenues par télédétection.L’analyse des tendances a été réalisée sur des données hydrologiques et météorologiques des cinq bassins versants emboités. Pour cela, le cycle saisonnier du débit est ajusté en utilisant une fonction mathématique de type “modèle à pic asymétrique”. Les changements observés des débits ont été reliés aux variables météorologiques ainsi que à l’évolution de la couverture glaciaire. Les résultats indiquent un comportement contrasté entre les bassins versants selon les taux d’englacements, avec une tendance croissante des valeurs de débit dans les bassins versants fortement englacés (couverture de glacier >30%) et une décroissante pour les moins englacés. La sensibilité du cycle hydrologique au changement climatique futur a été évaluée. Pour le milieu du 21e siècle, on prévoit que le volume annuel écoulé serait réduit de 16% pour l’Arveyron d’Argentière et de 31% pour l’Arveyron de la Mer de Glace. Pour la période estivale, la quantification détaillée de chaque terme de l’équation du bilan hydrologique, ainsi que leurs incertitudes, sur les bassins versants de l’Arveyron d’Argentière et de l’Arveyron de la Mer de Glace-Leschaux a permis de souligner l’importance des transferts d’eau souterraine pour représenter et prédire le comportement hydro-glaciologique d’un bassin versant donné. Deux model d’écoulement distribués de type degré-jour couples à un modèle de routage hydrologique à réservoir linéaire ont était utilisé sur le bassin versant de l’Arveyron d’Argentière sur la période 1960–2009. La calibration est effectuée autant sur la base des données de débit qu’avec une approche multicritère avec les données de débit, de couverture neigeuse et du bilan de masse annuel, à pas de temps journalier. Les résultats montrent l’aptitude d’utiliser un modèle classique degré-jour pour simuler le comportement hydro-glaciologique et la production d’eau sous-glaciaire d’un bassin versant fortement glaciaire. Pour la période 1960–2004, une valeur de Kling Gupta Efficiency de 0.85 entre le débit simulé et observe à était obtenu. La calibration multicritère semble réduire les incertitudes des simulations. / Glacier recession and the anticipation of spring snow melt driven by a warming climate could lead to changes in the hydrological cycle affecting not only the headwater catchments but also the areas downstream. In order to correctly predict the magnitude of future possible changes and to consider appropriate strategies of water management, a good understanding of the interaction between glaciers, climate and hydrology is needed. The aim of this study is to assess the effect of climate variability on the hydro-glaciological behaviour and its consequence on water availability in the Arve River catchment (French Alps) since 1960. It covers 1958 km2 and is composed by five nested catchments (Arveyron d’Argentière, Arveyron de la Mer de Glace, Arve at Pont des Favrands, Arve at Sallanches and Arve at Bout du Monde), all influenced by glacier and snow melt but characterized by various percentages of glacier cover ranging from 5 to 53%. This research is based on a long dataset of in situ or remote sensing glaciological, meteorological, hydrological and snow cover area data.Trend analyses are performed on the hydrological and meteorological data at all the considered sites. The seasonal cycle of each catchment is fitted using a mathematical function, namely the asymmetric peak model, and changes in the discharge are related to observed changes in the meteorological variables and the glaciers’ evolution. Results point out a contrasting behaviour among the catchments characterized by different glacier covers, showing an increasing trend on the discharge values in highly glacierized catchments (with a glacier cover >30%) and a decrease in the low glacierized ones. The sensitivity of the seasonal cycle to the future climate is evaluated. In the mid-21st century the annual runoff would be reduced by 16% for Arveyron d’Argentière and 31% for Arveyron de la Mer de Glace. Over the summer season, a detailed quantification of each term of the hydrological balance equation, as well as their uncertainties, on the Argentière and Mer de Glace-Leschaux drainage basins allows to underline the importance of considering the groundwater transfers to represent and predict the hydro-glaciological behaviour of a considered catchment. Two different distributed temperature index melt models coupled with a linear reservoir discharge model are used on the Arveron d’Argentière catchment over the 1960–2009 period. The calibration is carried out against discharge only and with a multi- criteria approach considering the discharge, the snow cover area and the glacier-wide annual mass balance values at daily time step. Results demonstrate the suitability of the use of a classical degree day model in simulating the hydro-glaciological behaviour and the subglacial water production of a highly glacierized catchment. A KGE of 0.85 is obtained between the observed and simulate discharge values over the 1960–2004 period. The use of a multi-criteria approach seems to reduce the simulation uncertainties.