No description available.

The potential for life to control its environment was first suggested by Lovelock (1972). Charlson et al (1987) proposed a role for marine planktonic ecosystems in global climate regulation via the production and ventilation to the atmosphere of dimethylsulphide (DMS), a by-product of phytoplankton metabolism. Once in the atmosphere DMS contributes to the formation of cloud condensation nuclei, and increases the amount and brightness of cloud. This affects the albedo of the planet, reflecting more incident sunlight back into space, and cooling the earth. In common with many other 'hypotheses' regarding complex adaptive systems, the hypothesis proposed by Charlson et al (1987) is not experimentally testable. The production and ventilation to the atmosphere of DMS is the result of complex interactions between biological, chemical and physical processes. Consequently, increasing use is being made of mathematical models that simulate these processes to advance understanding of it (Archer et al. 2002). This study examines one of the fundamental mechanisms proposed by the Charlson et al (1987) hypothesis, that increasing global temperatures will lead to increased ventilation of DMS from the ocean to the atmosphere. The study develops one-dimensional biogeochemical models of DMS production by upper ocean ecosystems, based on the model proposed by Gabric et al. (1993b). The models are examined to elucidate their fundamental mathematical properties, and are subjected to sensitivity analysis to identify important processes and parameters. These investigations identify a simpler model that can reproduce the predictions of the Gabric et al. (1993b) model. Predictions derived from model simulations forced by climatologies of measured physical data are compared to a global database of measurements of sea surface DMS concentrations, and to observed depth profiles of DMS in the upper ocean. These comparisons confirm that all models are in good qualitative agreement with measured data. The fifteen global climate prediction models currently in use around the globe all predict substantial warming effects from the ventilation of anthropogenic carbon dioxide to the atmosphere. A simplified DMS model is calibrated to climatologies of Antarctic chlorophyll and DMS data and reproduces the data with great precision. The calibrated model is applied in global warming scenarios to 'test' the efficacy of the mechanism proposed by the Charlson et al (1987) hypothesis. This simulation provides evidence that the response predicted by the hypothesis is indeed feasible, and that substantial increases (up to 45%) in the ventilation of DMS to the atmosphere could be possible in some circumstances. The results of the modelling study provide impetus for further examination of field data. If couplings between marine biota and atmosphere are feasible, then they may be operating contemporarily, and may be detectable. Atmospheric DMS is oxidised to form aerosols (Miller et al. 2002) that influence the aerosol optical depth of the atmosphere. Archives of remote sensed ocean chlorophyll a concentration and aerosol optical depth are examined for evidence of the biologically mediated couplings. A clear coupling between aeolian dust and marine phytoplankton is evident from this analysis, suggesting that the deposition of dust from the atmosphere is a major factor controlling phytoplankton growth in many parts of the ocean. A second coupling between marine phytoplankton and atmospheric aerosols is also detected. This coupling is apparently not related to dust and is symmetrical about the equator, despite the substantial differences in the atmospheres and oceans of each hemisphere. It is speculated that this coupling may reflect the influence of the ventilation of DMS produced by marine phytoplankton on the atmosphere. This thesis provides new evidence supporting the important role of marine ecosystems in global climate regulation by the production of DMS. This evidence is principally obtained from a biogeochemical modelling approach, but is supported by analyses of empirical data. The concordance of results obtained from different approaches suggests that the contribution of marine ecosystems to global climate regulation is real, important and currently active.

Il a récemment été démontré que la glace de mer antarctique pouvait jouer un rôle significatif dans la dynamique des gaz à effet climatique (dont le dimethylsulfure ou DMS) dans les régions polaires. Ce travail s’est d’abord attaché à la mise au point d’une méthode de mesure fiable du diméthylsulfoxyde (DMSO) dans la glace de mer, supprimant les interférences générées par la production de DMS au sein de l’échantillon en réponse au choc osmotique subi lors de la fonte de l’échantillon de glace. Une procédure de détermination séquentielle du DMS, par broyage à sec, puis du dimethylsulfoniopropionate (DMSP) et du DMSO sur le même échantillon de glace a été développée et utilisée à large échelle dans ce travail. Les données du présent travail ont été acquises dans le cadre de deux programmes d’observation intégrés menés sur la glace de mer antarctique à des saisons différentes mais avec une méthodologie commune :1) choix de sites d’étude homogènes afin de minimiser l’impact de la variabilité spatiale sur l’interprétation des résultats dans une optique d’évolution temporelle et 2) priorité à la caractérisation du cadre physico-chimique (texture, température, salinité, couvert de neige, susceptibilité au drainage des saumures,….) avant toute autre analyse. L’étude menée dans le cadre du programme ISPOL (nov.–dec. 2004) a permis d’observer que la stratification des saumures a un impact positif sur la conversion du DMSP en DMS au sein de la glace mais ralentit les flux de DMS et DMSP vers l’océan. Le couvert de glace est caractérisé à cette période de l’année par une perte nette de DMSP et génère des flux combiné de DMS et DMSP du même ordre de grandeur que les flux de DMS atmosphériques mesurés dans le cadre d’autres études. L’étude menée dans le cadre du programme SIMBA (sept.–oct. 2007) a permis de mettre en évidence l’importance du forçage atmosphérique sur le régime thermique et la dynamique du DMS/P/O dans la glace. Les communautés d’algues de surface produisent de fortes concentrations de DMS/P/O en réponse au stress thermique, osmotique et potentiellement radiatif durant les périodes de refroidissement et la mise en place d’un régime soutenu de drainage des saumures contribue à évacuer périodiquement les hautes concentrations de DMS/P/O produites dans la glace vers l’océan sous-jacent. Le couvert de glace affichant une production nette de DMS/P/O à cette période de l’année génère des flux combinés de DMS et DMSP plus de dix fois supérieurs à ceux observés pour la glace estivale. L’étude menée sur de la glace artificielle a permis de mettre en évidence l’impact des processus physico-chimiques sur la signature en gaz de la glace en croissance constituant un premier pas vers la modélisation des transports de gaz dans la glace de mer et leurs échanges au travers des interfaces glace-océan et glace-atmosphère. <p><p><p>SUMMARY - It has recently been demonstrated that Antarctic sea ice recently demonstrated plays a potentially significant role in the dynamics of climatically significant gases (amongst which dimethylsulphide or DMS) in Polar Regions. This research work has initially focused on the development of a reliable method for the determination of dimethylsulphoxide (DMSO) within sea ice, avoiding interferences generated by DMS production within the sample in response to the osmotic shock caused by melting. A sequential determination procedure of DMS, dimethlsulphoniopropionate (DMSP) and DMSO on the same ice sample has been developed and used on a large amount of samples in the present work. Data presented in this research project have been collected in the framework of two integrated sea ice observation programs focused on Antarctic sea ice at different seasons but following a common approach: 1) choice of homogeneous study sites to minimize the impact of spatial variability on the interpretation of the results in a time series perspective and 2) priority given to the characterization of the physicochemical framework (texture, temperature, salinity, snow cover, susceptibility to brine drainage,…) prior to any other study. The study conducted in the framework of the ISPOL experiment (Nov.–Dec. 2004) demonstrated that stratification of the brine inclusions network positively influenced the conversion of DMSP into DMS but decreased fluxes of DMS and DMSP towards the ocean. The ice cover at that time of the year is characterised by a net DMSP loss and generates combined DMS and DMSP fluxes whose values fall in the range of atmospheric DMS flux from sea ice measured in the frame of other studies. The study conducted in the framework of the SIMBA experiment (sept.–oct. 2007) emphasized the importance of atmospheric thermal forcing on the sea ice thermal regime and DMS/P/O dynamics. The surface community of algae produced elevated levels of DMS/P/O in response to thermal, osmotic and potentially radiative stress during periods of atmospheric cooling while the development of an intense brine drainage regime contributed to periodically release the elevated levels of DMS/P/O produced in the sea ice towards the underlying ocean. The ice cover exhibited at that time of the year a net production of DMS/P/O and produced combined DMS and DMSP fluxes more than ten times higher than those observed for summer sea ice. The study conducted on laboratory prepared growing sea ice emphasised the impact of physicochemical processes on the gas signature of growing sea ice and represents a first step towards modelling gas exchanges within sea ice and across its interfaces with the ocean and the atmosphere.<p> / Doctorat en Sciences agronomiques et ingénierie biologique / info:eu-repo/semantics/nonPublished