The Florida Everglades is a massive and highly managed subtropical wetland ecosystem, strongly influenced by anthropogenic control of freshwater distribution and highly susceptible to a changing climate, including rising sea-level and changes in temperature and rainfall. Shifting hydrologic regimes impact ecosystem function and biogeochemistry, which in turn control the sources, fate, and transport of organic matter. As a master environmental variable, it is essential to understand how organic matter dynamics will respond to changes in the balance between freshwater and saltwater associated with landscape-scale Everglades restoration efforts and climate change. The research comprising this dissertation improves current understanding of the linkages between organic matter and hydrology in the Everglades across a broad range of temporal and spatial scales. A range of research tools, including stable molecular biomarkers, water quality sensors, data synthesis and multivariate statistics were utilized. Biomarkers were used to track particulate organic matter mobilization in response to experimentally manipulated flows and provided initial evidence that sheet flow restoration can re-engineer landscape microtopography, influencing both ecosystem structure and organic matter inputs to Everglades National Park (ENP). Short-term and long-term temporal studies indicated the quantity and quality of dissolved organic carbon responds to changes in freshwater flow to marshes and mangrove forests in ENP, and that spatial patterns and trends are driven by a complex mixture of managed and natural surface water inputs (i.e., rainfall and water management inflows) as well as groundwater discharge. Application of climate scenario forecasting to relationships established between organic matter and hydrologic drivers predicted reductions in dissolved organic carbon export from ENP and changes in organic matter molecular composition. Furthermore, high-frequency measurements showed hydrologic connectivity of freshwater and estuarine organic matter pools at sub-monthly time-scales. In summary, the work presented here clearly indicates strong yet spatiotemporally complex relationships between changes in water and the sources and transport of organic carbon through the Everglades.

A new IAM is used to calculate the optimal tradeoff between, on the one hand,locking up fossil fuel and curbing global warming, and, on the other hand,sacrificing consumption now and in the near future. This IAM uses the Oxford carbon cycle, which differs from DICE, FUND and PAGE in that cumulative emissions are the key driving force of changes in temperature. We highlight how time impatience, intergenerational inequality aversion and expected trend growth affect the time paths of the optimal global carbon tax and the optimal amount of fossil fuel reserves to leave untapped. We also compare these with the adverse and deleterious global warming trajectories that occur if no policy actions are taken. (authors' abstract) / Series: Ecological Economic Papers

La disponibilité en eau est le principal facteur limitant le fonctionnement des écosystèmes méditerranéens. Des sècheresses plus marquées ou plus fréquentes pourraient avoir d'importantes répercussions sur l'activité et la diversité de la faune du sol qui régule la décomposition des litières et le cycle des nutriments. Dans cette thèse j'ai étudié expérimentalement les interactions entre une modification des apports en eau et l'impact de la macrofaune détritivore sur les processus de décomposition en conditions méditerranéennes. Dans une première partie, j'ai étudié l'effet d'une espèce de détritivore très abondante localement, Ommatoiuus sabulosus, sur la décomposition des litières d'arbustes de garrigue. Une expérience d'un mois en microcosmes a permis d'étudier ses effets directs (via la consommation de litière) et indirects (via l'activité microbienne dans ses fèces) sur la perte de masse des litières et les communautés microbiennes à deux niveaux d'humidité contrastés. Dans une autre expérience d'un an sur le terrain, la mise en place de sachets de litières et de fèces à deux profondeurs dans un sol de garrigue a permis d'étudier les effets d'Ommatoiulus à plus long terme. Les principaux résultats montrent que sa consommation de litière est moins affectée par la sècheresse que la décomposition microbienne, mais que, à court terme, Ommatoiulus ne stimule pas la minéralisation de la matière organique, quelles que soient les conditions d'humidité. En revanche, à plus long terme, Ommatoiulus peut accélérer la décomposition de certaines litières comme le chêne kermès, puisque des fèces issues de cette litière déposées à la surface du sol pendant un an perdent plus de masse que de la litière non-consommée. Cette stimulation semble liée à un lessivage plus important des composés organiques solubles dans les fèces et ne se produit qu'à la surface du sol. En profondeur, où l'humidité du sol est plus favorable à la décomposition, la perte de masse des fèces augmente. Ce résultat suggère qu'en facilitant l'enfouissement de la matière organique dans le sol, les détritivores peuvent accélérer la décomposition.Dans une seconde partie, j'ai cherché à évaluer l'importance de la diversité fonctionnelle des litières et des détritivores pour le processus de décomposition. Grâce à une approche basée sur les traits, des assemblages d'espèces représentant un fort gradient de dissimilarité fonctionnelle mais ayant une richesse spécifique constante, ont été créés pour étudier la réponse de la relation diversité-fonction à la sécheresse. Les résultats de cette expérience menée à l'Ecotron de Montpellier, montrent que la dissimilarité fonctionnelle des litières et des détritivores explique jusqu'à 20% de la variation observée dans plusieurs processus clefs du fonctionnement du sol, tels que la perte de masse des litières et le lessivage du carbone et de l'azote dans le sol superficiel. Toutefois, les effets de l'identité des espèces présentes aux deux niveaux trophiques restent plus importants que ceux de la dissimilarité fonctionnelle. Bien que la sécheresse influence fortement les processus étudiés, les relations diversité-fonction ne sont pas modifiées par un changement de la disponibilité en eau. Cependant, les assemblages d'espèces les plus performants en conditions d'humidité favorables sont aussi les plus fortement affectés par la sécheresse, ce qui suggère qu'il existe un compromis entre l'efficacité des organismes du sol et leur capacité à résister à une perturbation. / Water availability is a major limiting factor for the functioning of Mediterranean ecosystems. More pronounced drought could severely impact soil fauna activity and diversity that could in turn affect litter decomposition and nutrient cycling. In my PhD thesis I investigated experimentally the interactions between changing water availability and detritivorous macrofauna on decomposition and associated processes in a “garrigue”, a typical Mediterranean woody shrub dominated ecosystem.In the first part of my thesis, I studied the impact of Ommatoiulus sabulosus, an abundant diplopod species in garrigue ecosystems, on shrub litter decomposition. During a one month experiment, I studied the direct (litter consumption) and indirect (microbial activity in feces) effects of this detritivore on litter mass loss and microbial communities under two contrasted moisture levels. In a different experiment, I placed litterbags filled with litter or feces in the field at the soil surface or at 5cm soil depth during one year in order to study the long term impact of Ommatoiulus on decomposition. A key result was that detritivores maintain litter consumption in dry conditions when microbial driven decomposition drastically dropped. However, this detritivore effect do not lead to an overall increased organic matter mineralization irrespective of moisture conditions, at least in the short term. In contrast, under field conditions and over a longer time period, Ommatoiulus increases decomposition of certain species such as Quercus coccifera, since feces from this species decomposes faster than un-ingested litter after one year at the soil surface. This stimulation is likely due to a higher leaching of soluble compound in feces. Moreover, in depth feces decomposition increases relative to that of intact leaf litter, possibly indicating that more favorable soil humidity is more favorable to decomposition. Collectively, my results suggest that detritivores can strongly increase decomposition by transforming leaf litter into feces of different organic matter quality, and by facilitating the transfer of organic matter into the soil.In the second part, I evaluated the importance of functional dissimilarity of leaf litter and detritivores on decomposition processes. Using a trait based approach, species assemblages were constructed in order to obtain a gradient of functional dissimilarity of both, leaf litter and detritivore communities, while keeping species numbers constant. The different communities were kept under controlled conditions at the European Ecotron in Montpellier to study the effect of changing functional dissimilarity on process rates at two different moisture conditions. I found that detritivore and litter functional dissimilarity explain up to 20 % of the observed variation for several key soil processes including litter mass loss and the leaching of dissolved organic carbon and nitrogen from top soil. However, effects of species identity at both trophic levels have a larger impact on process rates than functional dissimilarity. In general, drought strongly affects soil processes but does not alter the diversity-function relationship. Species assemblages resulting in highest process rates at favorable moisture level are also the most negatively affected by drought, suggesting a tradeoff between the efficiency of soil organisms and their ability to resist perturbation.

O Carbono Orgânico Dissolvido (COD) constitui em uma das principais formas de carbono orgânico exportado em rios e em outras vias hidrológicas preferencias nas bacias de drenagem. Na Amazônia, os estudos sobre o COD nas suas águas abrangem principalmente as grandes bacias da região central, com poucas análises sobre alterações na sua dinâmica após a substituição de florestas por pastagens, uma das mudanças no uso da terra mais comum na região. Este estudo objetivou fornecer informações que auxiliem na compreensão destes processos, através da comparação entre as dinâmicas das concentrações nas vias hidrológicas preferenciais, as exportações, e os balanços de COD durante eventos de precipitação em duas bacias de primeira ordem, sendo uma com floresta e outra com pastagem. O estudo foi conduzido numa fazenda próxima à cidade de Cacaulândia, no estado de Rondônia. As amostras foram coletadas na transição do período seco para o chuvoso (Agosto a Novembro de 2004) e período chuvoso (Janeiro a Abril de 2005). Os resultados mostraram que na transição do período seco para o chuvoso, apenas as concentrações de COD nas vias de entrada para o solo, precipitação na pastagem e precipitação interna na floresta, e de saída, pelos igarapés e água subterrânea apresentaram diferenças estatisticamente significativas entre floresta e pastagem. No período chuvoso, além destas, outras vias hidrológicas, como a solução do solo a 20 e 100 cm de profundidade e o escoamento superficial, mostraram-se distintas entre estes dois sistemas. Comparando-se os períodos do ano, em ambas as bacias as concentrações de COD nas vias hidrológicas foram superiores na época de transição do período seco para o chuvoso. A única exceção a este padrão foi observada na água subterrânea da floresta, que apresentou maiores concentrações de COD no período chuvoso. Na pastagem, o aumento do escoamento superficial resultou em maiores exportações de COD do que aquelas observadas na floresta. Como conseqüência, os balanços de COD a cada evento e anualmente, tendeu a ser mais positivos (maior retenção de COD) na última. Embora neste estudo não tenha sido possível, entre outros, quantificarem os fluxos de COD na água subterrânea e, portanto, obter um balanço mais preciso para esta forma de carbono nestes sistemas, os resultados demonstraram que a conversão de florestas em pastagens acarreta em um aumento significativo das exportações de COD, via igarapé, no último local, tanto em eventos individuais de precipitação, quanto anualmente. / Dissolved Organic Carbon (DOC) is one of the main fractions of organic carbon exported in rivers and other flow paths in watersheds. In the Amazon, studies of DOC in waters focus mainly on the large basins of the central region, with few analyses of changes in its dynamics after the replacement of forests by pastures, one of the most important land use changes in the region. This objective of this study was to obtain more information about these processes, by comparing dynamics in several flowpaths, exports and budgets of DOC during rain events in two first-order watersheds, one with forest and another with pasture. The study was conducted near city of Cacaulândia in the State of Rondônia. Samples were collected in the transition from dry to wet (August to November of 2004) and in the wet seasons (January to April of 2005). The results demonstrate that during the transition from dry to wet seasons, statistically significant differences in DOC concentrations between forest and pasture occur only in the inputs from precipitation in pasture and throughfall in forest and in the outputs in streams and groundwater. During the wet season the remaining flowpaths analyzed, soil solution at 20 and 100cm depth and overland flow, also showed statistically significant differences in DOC concentrations between forest and pasture. In terms of the periods of the year, in both watersheds DOC concentrations were higher during the transition from dry to wet seasons. The only exception to this pattern was observed in forest groundwater, with higher concentrations during the wet season. In the pasture, increases in discharge resulted in higher DOC export than in forest. As a consequence of this, DOC budgets, both in individual events and annually, showed a larger retention in the latest. Although in this study it was not possible, among others, to quantify the fluxes of DOC in groundwater, the results demonstrate that the conversion of forest into pasture causes a significant increase of DOC export in streams, both in individual rain events and on an annual basis.

O sistema climático global e o ciclo do carbono interagem intensamente, e o CO2 constitui um fator dominante na definição do clima, sendo gerado e consumido pelas plantas e pela atividade de microrganismos em ecossistemas aquáticos, terrestres e na atmosfera. Na atmosfera, esse gás contribui para o efeito estufa. Em um ecossistema de floresta tropical, grande parte da produção de CO2 é proveniente da respiração do solo, e os fluxos de CO2 na interface solo-atmosfera dependem de mudanças nas características físicas, químicas e biológicas na superfície do solo. O objetivo desse estudo foi investigar a variabilidade sazonal dos fluxos de CO2, decorrente da respiração do solo de diferentes florestas nativas da Amazônia, localizadas nos municípios de Sinop (MT), Caxiuanã (PA), Manaus (AM) e Santarém (PA), e determinar os principais parâmetros de correlação dos fluxos de CO2 do solo com a temperatura e umidade. As amostragens foram realizadas durante as estações seca e chuvosa em cada local. Os fluxos de CO2 foram medidos por meio de câmaras dinâmicas, que se baseiam na variação da concentração do gás no interior da câmara em função do tempo e foram calculados utilizando-se de equações lineares. Os valores médios encontrados para as regiões estudadas nas estações seca e chuvosa foram respectivamente: Sinop, 3,03 μmol.CO2 m-2s-1 e 5,76 μmol. CO2 m-2s-1; Caxiuanã, 5,07 μmol.CO2 m-2s-1 e 6,09 μmol.CO2 m-2s-1; Manaus, 5,47 μmol.CO2 m-2s-1 e 5,44 μmol.CO2 m-2s-1 e Santarém, 2,90 μmol.CO2 m-2s-1 e 5,64 μmol.CO2 m-2s-1. Estes resultados demonstraram que houve variação regional e influência da sazonalidade na dinâmica do fluxo de CO2 do solo, sendo que, os maiores fluxos foram obtidos durante o período chuvoso, indicando que a disponibilidade de água e a temperatura do solo foram os principais condicionadores da produção do CO2. Foram coletadas amostras de serapilheira e raízes para estimar o estoque e a influencia sobre os fluxos de CO2 do solo, sendo estas, vias de entrada de nutrientes e principalmente de carbono para o solo. Para todas as regiões os maiores estoques foram encontrados na estação seca. / Carbon dioxide is an important contributor to the greenhouse effect, and by extension, the global climate system as a whole; it is generated and consumed through the activities of terrestrial and aquatic microorganisms. Tropical forest CO2 fluxes at the soil-atmosphere interface depend on changes in the physical, chemical and biological characteristics of the soil surface. The objective of the present study is to investigate seasonal variability of CO2 fluxes from soil respiration in Amazonian native forest located in the municipalities of Sinop (MT), Caxiuanã (PA), Manaus (AM) e Santarém (PA) and to determine the effects of determining parameters such as temperature and soil moisture. The CO2 fluxes were measured using dynamic chambers during both dry and wet seasons at each site. As the concentration of the gas inside the chamber varies with time, samples were taken at given time intervals and fluxes then calculated using linear regression equations. The average values found for Sinop in the dry and wet seasons were 3.03 and 5.92 μmol.CO2 m-2s-1 respectively; for Caxiuanã, 5.07 μmol.CO2 m-2s-1 and 6.09 μmol.CO2 m-2s-1; for Manaus, 5.47 and 5.44 μmol.CO2 m-2s-1; and for Santarém, 5.64 μmol.CO2 m-2s-1 in the wet season and during the dry season and 6.09 μmol.CO2 m-2s-1. Our results showed that there was a seasonal variation of the CO2 flux. The results also showed that there was an influence of the seasonality in the dynamics of the soil CO2 flux, where the greater fluxes were obtained during the wet season, indicating that water availability and soil temperature were the main factors determining production. Litter samples had been collected and roots esteem the supply and influence it on the CO2 fluxs of in the soil, being these, ways of entrance of nutrients and mainly of carbon for the ground. For all the regions the biggest supplies had been found in the dry station.

Cette thèse traite du Brazilian Proposal, c'est-à-dire de la détermination des contributions nationales au changement climatique d'origine humaine. Pour répondre à cette question, nous avons développé un modèle compact du système Terre, OSCAR v2.1. Ce modèle intègre une représentation du cycle du carbone (CO2, CH4), de la chimie atmosphérique des gaz à effet de serre (CH4, N2O, O3, composés halogénés), mais également des aérosols et de la dynamique climatique. Il est forcé en émissions anthropiques de composés actifs et en changements d'usage des sols. Après avoir constaté la bonne capacité du modèle à reproduire les observations passées des principales grandeurs climatiques, et après avoir énoncé les grands principes régissant les exercices d'attribution, nous attribuons les causes anthropiques du changement climatique. Nous trouvons que la rétroaction climatique, sur le cycle du carbone et sur la chimie atmosphérique, a un effet prépondérant qui exacerbe l'importance relative de chaque forçage anthropique. Par ordre décroissant, émissions de dioxyde de carbone fossile, de dioxyde de soufre, de méthane, et usages des sols, sont trouvés comme étant les plus importants contributeurs au changement climatique en 2008. A travers ces forçages, les pays dits en développements sont dorénavant de plus grands contributeurs au changement climatique que les pays dits développés. C'est cependant toujours l'inverse si l'on résonne en contribution par tête ; mais nous montrons qu'un tel raisonnement rend incompatibles une trajectoire de réchauffement inférieur à deux degrés et équitable. / This PhD thesis deals with the Brazilian Proposal, that is the assessment of national contributions to anthropogenic climate change. To answer the Proposal, we have developed a compact Earth system model, named OSCAR v2.1. The carbon cycle (CO2, CH4), the atmospheric chemistry of greenhouse gases (CH4, N2O, O3, halogenated compounds), as well as aerosols and climate dynamics are included in this model. It is driven by anthropogenic emissions of active compounds, and by land-use changes. After acknowledging the ability of the model to reproduce past observations of the main climatic variables, and after exposing the fundamental principles of attribution exercises, we attribute climate change to its anthropogenic causes. We find that the climate feedback -- over both the carbon cycle and the atmospheric chemistry -- has a prominent effect that exacerbates the relative importance of each anthropogenic forcing. In decreasing order, emissions of fossil carbon dioxide, of sulfur dioxide, of methane, and land-use changes, are found to be the most important contributors to climate change in 2008. Through these forcings, the so-called developing countries are now contributing more to climate change than the so-called developed countries. It is however still the contrary on a per capita basis; but we show that such an accounting approach makes it impossible to reach equity within a less-than-two-degree warming trajectory.

Le Crétacé a été une période d'instabilité climatique et du cycle du carbone, dont le CO2 atmosphérique a été désigné comme le driver majeur. Cependant, les reconstitutions du CO2 atmosphérique ne reflètent ni les dynamiques climatiques ni les grands évènements de perturbation du cycle du carbone décrits pour cette période. J'ai utilisé la composition isotopique de carbone de la plante fossile Frenelopsis (d13Cleaf) comme un nouvel proxy pour reconstituer le CO2 atmosphérique du Crétacé en termes de composition isotopique de carbone (d13CCO2) et de concentration (pCO2). La première courbe de d13CCO2 pour toute la durée du Crétacé a été construite à partir du d13C des carbonates marins. Sa comparaison avec des estimations de d13CCO2 à partir du d13Cleaf a révélé que les modèles développés jusqu'à maintenant ont une tendance à exagérer les valeurs de d13CCO2. Des estimations du fractionnement isotopique du carbone issu par des plantes (13Cleaf) obtenues à partir des nouvelles données d e d13Cleaf et d13CCO2 ont permis de reconstituer l'évolution à grande échelle de la pCO2. Ces résultats indiquent que le CO2 a probablement été une conséquence à long terme du changement climatique durant le Crétacé. Des cycles de d13CCO2 de ~1.2, ~2.1, ~5.4 et ~10.2 Ma ont été détectés, synchrones à ceux du niveau de la mer et à la cyclicité des paramètres de l'orbite terrestre décrits pour le Mésozoïque. Mes résultats fournissent une nouvelle perspective du système climatique et du cycle du carbone du Crétacé, dominés principalement par les paramètres orbitaux de la Terre et secondairement par des évènements catastrophiques de libération de CO2 d'origine volcanique dans l'atmosphère / The Cretaceous was a period characterized by strongly marked climate change and major carbon cycle instability. Atmospheric CO2 has repeatedly been pointed out as a major agent involved in these changing conditions during the period. However, long-term trends in CO2 described for the Cretaceous are not consistent with those of temperature and the large disturbance events of the carbon cycle described for the period. This raises a double question of whether descriptions of the long-term evolution of atmospheric CO2 made so far are accurate or, if so, atmospheric CO2 was actually a major driver of carbon cycle and climate dynamics as usually stated. In this thesis the close relationship between the carbon isotope composition of plants and atmospheric CO2 is used to address this question. Based on its ecological significance, distribution, morphological features and its excellent preservation, the fossil conifer genus Frenelopsis is proposed as a new plant proxy for climate reconstructions during the Cretaceous. The capacity of carbon isotope compositions of Frenelopsis leaves (d13Cleaf) to reconstruct past atmospheric CO2, with regards to both carbon isotope composition (d13CCO2) and concentration (pCO2), is tested based on materials coming from twelve Cretaceous episodes. To provide a framework to test the capacity of d13Cleaf to reconstruct d13CCO2 and allowing for climate estimates from carbon isotope discrimination by plants (?13Cleaf), a new d13CCO2 curve for the Cretaceous based on carbon isotope compositions of marine carbonates has been constructed. Comparison with d13Cleaf-based d13CCO2 estimates reveals that although d13CCO2 and d13Cleaf values follow consistent trends, models developed so far to estimate d13CCO2 from d13Cleaf tend to exaggerate d13CCO2 trends because of assuming a linear relationship between both values. However, given the hyperbolic relationship between ?13Cleaf and pCO2, by considering an independently-estimated correction factor for pCO2 for a given episode, d13Cleaf values may be a valuable proxy for d13CCO2 reconstructions. ?13Cleaf estimates obtained from d13CCO2 and d13Cleaf values were used to reconstruct the long-term evolution of pCO2. The magnitude of estimated pCO2 values is in accordance with that of the most recent and relevant model- and proxy-based pCO2 reconstructions. However, these new results evidence long-term drawdowns of pCO2 for Cretaceous time intervals in which temperature maxima have been described

The terrestrial carbon cycle is a part of the global carbon cycle, where one important component is the terrestrial vegetation. Terrestrial vegetation largely controls the land surface carbon exchanges and leverage the atmospheric greenhouse gas concentrations, significantly affecting the trajectory of global warming. It is therefore important to improve the understanding of vegetation response to different climatic factors, in particular for those linked to large-scale climate variability, which is still less studied so far. Vegetation greenness is suggested to be a useful tool in order to understand vegetation response. Looking at Europe, the climate factors that affect vegetation the most are linked to the large-scale atmospheric circulation over the North Atlantic, like the jet stream, which varies in speed and latitude, and the North Atlantic Oscillation (NAO). Here, I compute monthly indices representing the variability of these atmospheric features, and correlate them with monthly vegetation greenness data (NDVI) anomalies over a period of five years. This is done both for regionally-averaged NDVI and the months April-July and as a geographical point-by-point analysis for the month of May. The results show a significant correlation between Scandinavian NDVI and the NAO as well as jet speed at multiple time lags, up until 2 months. The jet latitude, instead, showed significant correlation for three regions in mid/southwestern Europe at longer time lags of 3-4 months. This means that the position of the jet in winter can affect the spring vegetation growth in this area. The jet speed and NAO, however, works mostly at shorter timespans. / Den jordbunda kolcykeln, som är en del av den globala kolcykeln, består av olika komponenter där en viktig del är vegetation. Växtlighet på land kontrollerar till stor del utbytet av kol vid jordytan och har därigenom inflytande på atmosfäriska växthusgaskoncentrationer, vilket medför stor påverkan på global uppvärmning. Det är därför viktigt att förbättra förståelsen för hur vegetation reagerar på olika klimatologiska faktorer, särskilt de som är kopplade till storskalig klimatvariabilitet då dessa kopplingar har studerats i mindre utsträckling hittils. Ett bra sätt att mäta den jordbunda kolcyklen på är med grönhet av vegatation. Om vi beaktar Europa så är det främst storskaliga atmosfäriska cirkulatoiner över norra Atlanten av de klimatologiska faktorerna som påverkar vegetation. En av dessa faktorer är jetströmmen, vilken varierar i fart och latitud, samt Nordatlantiska Oscillationen(NAO). I detta arbete beräknar jag index som representerar variationen i dessa i form av månadsgenomsnitt och korrelerar dem med månatlig data över avvikelser i vegetationsgrönhet (NDVI) över en femårsperiod. Det här gjordes för både regionala medelvärden och månaderna april-juli samt en geografisk punkt till punkt analys utförd för maj. Resultatet visar att det finns en signifikant korrelation mellan NDVI i Skandinavien och NAO samt jetfarten vid flera tidsfördröjningar, upp till 2 månader. Jetlatituden visade däremot signifikant korrelation för tre regioner i centrala/sydvästa Europa vid längre tidsfördröjningar på 3-4 månader. Detta innebär att positionen på jetströmmen under vintern kan påverka vegetationstillväxten under våren i detta område. Jetfarten och NAO påverkar däremot mest vid kortare tidsspan.

Since the beginning of the industrial revolution, a large amount of greenhouse gases such as carbon dioxide (CO2) have been emitted into the atmosphere due to human activities. One of the main consequences of these emissions is a rapid increase in atmospheric CO2 concentration and a profound modification of the Earth's climate system. The ocean plays an important role in the Earth radiative balance since it acts as an important CO2 sink for the atmosphere. By currently absorbing about 25 % of the CO2 emitted by humans it considerably slows down climate change. Understanding the present-day spatial and temporal dynamics of the air-sea CO2 exchange and the different processes that govern this exchange is of critical importance to anticipate the evolution of the oceanic CO2 sink in the future.This thesis was realized in this context and focused on an improved quantification of the exchange of CO2 through the air-sea interface (FCO2) of the global ocean, embracing open ocean waters and coastal regions. The main objective was to fill knowledge gaps in our understanding of the processes that govern the spatial and temporal distribution of FCO2. This objective was mainly achieved through observational approaches and addressed three main aspects: a quantification of the different sources of FCO2 uncertainties at the global scale, an analysis of spatial distribution of the oceanic CO2 exchange with a strong focus on the coastal ocean and a first assessment of the coastal seasonal FCO2 dynamics and its underlying drivers. The latter relied on a data-model fusion approach allowing to decompose the FCO2 seasonality into its main physical and biogeochemical drivers. The quantification of the oceanic FCO2 from observations consists in calculating an air-sea partial pressure CO2 gradient (ΔpCO2) between the atmosphere and the sea surface. Global monthly continuous partial pressure of CO2 (pCO2) products can for example be derived from observational pCO2 databases and statistical interpolation methods. This ΔpCO2 is then multiplied by a gas exchange transfer rate coefficient (k), which depends on wind speed. However, the parametrization of k is still entailed with poorly quantified uncertainties. From a literature review of all k parameterizations available in the literature over the past 25 years, I first quantified the FCO2 uncertainties associated with k globally and regionally for the open ocean. I also quantified the uncertainties associated with the choice of a wind product over another. Our results show that the range of global FCO2, calculated with these k relationships, diverge by 12 % when using CCMP, ERA or NCEP1. Regional discrepancies in FCO2 are more pronounced than global. These global and regional differences significantly increase when using NCEP2 or other k formulations. To minimize uncertainties associated with the choice of wind product, it is possible to recalculate the parametrization of k globally for a given wind product and its spatio-temporal resolution, in order to match the last evaluation of the global k value. In a second step, we improved the quantification and analysis of the dominant patterns and drivers of the FCO2 spatial distribution for the coastal ocean worldwide. This analysis was performed globally (at 0.25° spatial resolution), using a regional segmentation of the coastal ocean, and latitudinally. I found that coastal regions at high latitudes act as a CO2 sink while tropical regions and along the equator tend to act as an atmospheric CO2 source. Globally integrated, I quantified that the coastal seas act currently as a CO2 sink with a value of -0.20 ± 0.02 Pg C yr-1. For the first time, I also compared the spatial patterns of coastal FCO2 to that of the adjacent open ocean, globally. With the exception of some regions such as those dominated by riverine inputs, I demonstrated that they present similar latitudinal distribution of their FCO2 density per unit of surface area, suggesting analogous responses to increasing atmospheric CO2. I also reevaluated the global ocean CO2 budget and estimated a global anthropogenic CO2 uptake ranging between -2.6 ± 0.4 Pg C yr-1 and -2.9 ± 0.5 Pg C yr-1 for the 1998-2015 period. In a third step, I contributed to the first continuous observational pCO2 data product merging the coastal and open ocean in a consistent manner. This study showed that difference between open ocean and coastal ocean estimates along the overlap area increases with latitude but remains close to 0 µatm globally. Stronger discrepancies, however, exist on the regional level resulting in differences that exceed 10 % of the climatological mean pCO2, particularly in regions constrained by fewer observations, paired with biogeochemical complexity, such as the Peruvian upwelling system and ice covered regions.In a fourth step, a temporal analysis of the FCO2 seasonality was performed for the coastal ocean based on an observational approach. I analyzed and quantified the FCO2 seasonal dynamics globally and for different latitudinal bands. Globally, coastal regions act as a CO2 sink with a more intense uptake occurring in summer (-21 Tg C month-1) because of the disproportionate influence of high latitude shelves in the Northern Hemisphere. I also estimated the contribution of different drivers (sea-ice coverage, wind speed, and ΔpCO2 change) to the FCO2 seasonal amplitude. This data-driven approach allowed me to conclude that the ΔpCO2 is the main driver of the FCO2 variability at the seasonal timescale. I then used a global oceanic biogeochemical model to decompose the seasonal coastal pCO2 variability further into its driving physical and biological processes. From a first qualitative assessment, I concluded that the thermal effect associated to sea surface temperature changes is the main effect governing the coastal seasonal pCO2 variability except at high latitudes where the non-thermal effect associated to changes in biology, circulation, fresh water and the air-sea CO2 exchange itself dominate. I also found that, overall, the thermal effect alone should lead to larger seasonal fluctuations, but its influence is partly offset by the non-thermal effect. Throughout this thesis, I also evaluated the extent to which the continuous observational pCO2 products derived from an artificial neuronal network approach and from the global ocean biogeochemical model MOM6-COBALT could reproduce the raw pCO2 fields extracted from global databases. Overall, I showed that at the regional scale, the two products are in relatively good agreement compared to observations. I also identified regions where discrepancies are the largest and where future observational data are needed in the future, as well as regions where agreement is the most satisfactory and, thus, most suitable for further process-based analyses. / Depuis le début de la révolution industrielle, une grande quantité de gaz à effet de serre tels que le dioxyde de carbone (CO2) a été émise dans l'atmosphère en raison des activités humaines. L'une des principales conséquences de ces émissions est une augmentation rapide de la concentration en CO2 atmosphérique et une modification profonde du système climatique de la Terre. L'océan joue un rôle important dans l'équilibre radiatif de la Terre car il agit comme un important puits de CO2 pour l'atmosphère. En absorbant actuellement environ 25 % du CO2 émis par l'homme, il ralentit considérablement le changement climatique. Comprendre la dynamique spatiale et temporelle actuelle de l'échange de CO2 air-mer et les différents processus qui régissent cet échange est d'une importance cruciale pour anticiper l'évolution du puits océanique de CO2 à l'avenir.Cette thèse a été réalisée dans ce contexte et s'est concentrée sur une meilleure quantification de l'échange de CO2 à travers l'interface air-mer (FCO2) de l'océan global, considérant à la fois l’océan ouvert et les régions côtières. L'objectif principal était de combler les lacunes dans notre compréhension des processus qui régissent la distribution spatiale et temporelle du FCO2. Cet objectif a été principalement atteint grâce à des approches observationnelles et a abordé trois aspects principaux: une quantification des différentes sources d'incertitudes du FCO2 à l'échelle globale, une analyse de la distribution spatiale de l'échange de CO2 océanique avec un fort accent sur l'océan côtier et une première évaluation de la dynamique saisonnière du FCO2 côtier et de ses moteurs sous-jacents. Ce dernier s'est appuyé sur une approche de fusion de modèles et d’approches observationnelles permettant de décomposer la saisonnalité du FCO2 en ses principaux moteurs physiques et biogéochimiques.La quantification du FCO2 océanique à partir d’observations consiste à calculer un gradient de pression partielle air-mer de CO2 (ΔpCO2) entre l'atmosphère et la surface de la mer. Des produits globaux continus mensuels de la pression partielle de CO2 (pCO2) peuvent par exemple être dérivés à partir de bases de données observationnelles de pCO2 et de méthodes d'interpolation statistique. ΔpCO2 est ensuite multiplié par un coefficient de vitesse de transfert d'échange gazeux (k), qui dépend de la vitesse du vent. Cependant, la paramétrisation de k est sujette à de larges incertitudes et mal quantifiées. À partir d'une synthèse de la littérature de toutes les paramétrisations de k disponibles dans la littérature au cours des 25 dernières années, j'ai d'abord quantifié les incertitudes sur FCO2 associées à k à l'échelle globale et régionale pour l'océan ouvert. J'ai également quantifié les incertitudes associées au choix d'un produit éolien par rapport à un autre. Nos résultats montrent que la gamme du FCO2 global, calculée avec ces différentes paramétrisations de k, diverge de 12 % lors de l'utilisation de CCMP, ERA ou NCEP1. En raison des différences dans les pattern de vent régionaux, les différences régionales sur le FCO2 sont plus prononcés que globalement. Ces différences globales et régionales augmentent de manière significative lors de l'utilisation de NCEP2 ou d'autres formulations de k. Afin de réduire les incertitudes associées au choix du produit de vent, il est possible de recalculer la paramétrisation de k pour un produit de vent donné et à une résolution spatio temporelle.Dans un deuxième temps, nous avons amélioré la quantification et l'analyse des principaux pattern et des différents processus sur la distribution spatiale du FCO2 pour l’ensemble des régions côtières. Cette analyse a été réalisée à l'échelle globale (à une résolution spatiale de 0.25°), en utilisant une segmentation régionale de l'océan côtier, et latitudinalement. J'ai trouvé que les régions côtières aux hautes latitudes agissent comme un puits de CO2 tandis que les régions côtières tropicales et le long de l'équateur ont tendance à agir comme une source de CO2 atmosphérique. Globalement, j'ai quantifié que les régions côtières agissent actuellement en tant que puits de CO2 avec une valeur de -0.20 ± 0.02 Pg C an-1. Pour la première fois, j'ai également comparé la distribution spatiale du FCO2 côtier à celle de l'océan ouvert adjacent, à l'échelle globale. À l'exception de certaines régions telles que celles dominées par les apports fluviaux, j'ai démontré que les régions côtières et l’océan ouvert adjacent présentaient une distribution latitudinale similaire sur leur densité de FCO2 par unité de surface, suggérant des réponses analogues à l'augmentation du CO2 atmosphérique. J'ai également réévalué le budget mondial de CO2 de l'océan et estimé une absorption mondiale de CO2 anthropique comprise entre -2.6 ± 0.4 Pg C an-1 et -2.9 ± 0.5 Pg C an-1 pour la période 1998-2015. Dans un troisième temps, j'ai contribué à la création du premier produit continu de pCO2 observationnelles fusionnant le domaine côtier et l'océan ouvert de manière cohérente. Cette étude a montré que la différence entre les estimations provenant du produit de pCO2 de l’océan ouvert à celles dérivant du produit de pCO2 de l’océan côtier le long de leur zone de chevauchement augmente avec la latitude mais reste proche de 0 µatm globallement. Des divergences plus fortes existent cependant au niveau régional, entraînant des différences qui dépassent 10 % sur la moyenne climatologique de pCO2, en particulier dans les régions contraintes par moins d'observations, associées à une complexité biogéochimique, comme le système d'upwelling péruvien et les régions couvertes de glace.Dans une quatrième étape, une analyse temporelle de la saisonnalité du FCO2 a été réalisée pour l'océan côtier sur la base d'une approche observationnelle. J'ai analysé et quantifié la dynamique saisonnière du FCO2 à l'échelle globale et pour différentes bandes latitudinales. À l'échelle globale, les régions côtières agissent comme un puits de CO2 avec une absorption plus intense se produisant en été (-21 Tg C mois-1) en raison de l'influence disproportionnée des régions côtières des hautes latitudes dans l'hémisphère Nord. J'ai également estimé la contribution de différents processus (couverture de glace de mer, vitesse du vent et changement de ΔpCO2) à l'amplitude saisonnière du FCO2. Cette approche basée sur les données observationnelles m'a permis de conclure que ΔpCO2 est le principal moteur de la variabilité du FCO2 à l'échelle saisonnière. J'ai ensuite utilisé un modèle biogéochimique océanique global pour décomposer davantage la variabilité saisonnière du pCO2 côtier en ses processus physiques et biologiques. À partir d'une première évaluation qualitative, j'ai conclu que l'effet thermique associé aux changements de température de la surface de la mer est le principal effet régissant la variabilité côtière saisonnière du pCO2 sauf aux hautes latitudes où l'effet non thermique associé aux changements de biologie, de circulation, d'eau douce et de l’échange de CO2 air-mer domine. J'ai également constaté que, globalement, l'effet thermique à lui seul devrait entraîner des fluctuations saisonnières plus importantes, mais son influence est en partie compensée par l'effet non thermique.Tout au long de cette thèse, j'ai également évalué dans quelle mesure les produits continus de pCO2 observationnelles dérivés d'une approche de réseau de neurones artificiels et du modèle biogéochimique océanique global MOM6-COBALT pourraient reproduire les champs de pCO2 bruts extraits des bases de données globale. Dans l'ensemble, j'ai montré qu'à l'échelle régionale, les deux produits sont relativement en bon accord par rapport aux observations. J'ai également identifié les régions où les différences sont les plus importantes et où de futures données observationnelles sont nécessaires à l'avenir, ainsi que les régions où les deux produits présentent un accord le plus satisfaisant et, par conséquent, le plus approprié pour de futures analyses de compréhension des différents processus. / Doctorat en Sciences / info:eu-repo/semantics/nonPublished

Today, climate change, has become one of the greatest societal challenges of our time. This challenge requires an accurate understanding of climate change for making informed decisions regarding the environmental issues. The carbon cycle is one of the earth’s complicated cycles that has a critical role in the planet’s climate. Developing a thorough perception about this complex cycle uncovers how human activities impact the planet and reveals the connection between multiple environmental issues.Perceiving this complex cycle requires systems thinking skills that enable students to recognize components of the carbon cycle and understand the interrelating dynamic relationship between them. Establishing systems thinking skills and developing a thorough perception about the carbon cycle is a difficult matter for students. Adaptive visualisation-based tutoring systems have a great potential for facilitating teaching and learning cyclical models and systems thinking in schools. Such systems consider the students’ needs and provide personalised feedback that can guide individuals more effectively throughout the learning process. This thesis project intends to use diagrammatic visualizations, systems thinking, and adaptive tutoring systems as three technical approaches for conceptualising a learning system that aims to teach the carbon cycle. The framework of this thesis project is formed in relation to a research project called ‘Tracing Carbon’ focusing on science education for pupils on grade 7-9.