Les embouchures tidales dominées par la houle sont des systèmes côtiers particulièrement dynamiques dont la morphologie est continuellement remodelée par l’action des vagues et de la marée. Les rapides évolutions morphologiques auxquelles elles sont sujettes impactent directement leurs environnements écologiques et socio-économiques. Afin de mettre en œuvre des réglementations adaptées à la gestion durable des embouchures tidales, des études environnementales systématiques sont nécessaires. L’objectif principal de cette thèse est de mieux comprendre les processus physiques qui contrôlent les évolutions morphologiques d’une embouchure tidale éphémère au Portugal - l’embouchure de la lagune d’Albufeira – à partir de l’analyse de mesures hydrodynamiques et topographiques et de résultats d’un système de modélisation morphodynamique récemment développé. Les processus qui influent sur la dynamique de l’embouchure tidale à court terme, notamment ceux liés aux interactions vague-courant, ont été étudiés à travers l’application du système de modélisation à l’embouchure. Les modulations saisonnières du climat de vagues et du niveau moyen de la mer affectent fortement la dynamique sédimentaire de l’embouchure et contribuent au comblement naturel de l’embouchure pendant l’hiver. Les processus à long terme ont également été étudiés à partir de simulations rétrospectives de paramètres moyens de vagues pour des échelles régionales et locales sur une période 65 ans. Les fortes variabilités interannuelles du climat de vagues et de la dérive littorale qui lui est associée pourraient expliquer les différences d’évolutions morphologiques du système embouchure-lagune sur des échelles de temps pluri–annuels. / Wave-dominated tidal inlets are very dynamic coastal systems, whose morphology is continuously shaped by the combined action of the waves and the tides. The rapid morphological changes they experience impact directly their ecological and socio-economic environments. In order to implement adequate regulations for the sustainable management of tidal inlets, systematic environmental studies are necessary. The main objective of this PhD research work is to gain a better understanding of the physical processes that control the morphological evolutions of an ephemeral tidal inlet in Portugal - the Albufeira Lagoon inlet - based on the analysis of hydrodynamic and topographic data and on the results of a newly developed morphodynamic modelling system. The processes that impact the dynamics of the inlet at short time-scales, particularly those related to wave-current interactions, are investigated through the application of the modelling system to the inlet. The seasonal modulations of the wave climate and mean sea level strongly affect the sediment dynamics of the inlet and contribute to the natural closure of the inlet during the winter period. Long-term processes are also investigated based on a 65-year hindcast of mean wave parameters at regional and local scales. The large inter-annual variability of the wave climate and the associated longshore sediment transport – both correlated to the North-Atlantic Oscillation – are proposed to explain the differences in the morphological behaviour of the inlet-lagoon system at pluri-annual time-scales.

MRDV / Institute for Rural Development / This study originated from the realisation that non-conventional crops such as Bambara nuts (Vigna subterranea) were becoming increasingly important in addressing food insecurity and malnutrition in the smallholder farming sector of countries in sub-Saharan Africa. Moreover, some of the smallholder crop farmers were observed to be continuing to rely on indigenous techniques to forecast rainfall and adapt agricultural activities in response to climate variability. However, it was not clear how climate change influenced the productivity of V. subterranea. Nor were the indigenous approaches that farmers used to forecast rainfall on this phenomenon well understood. Thus, a study was carried out to identify and document indigenous approaches that smallholder farmers used to forecast rainfall and adaptation practices relating to V. subterranea. The study was conducted in Xigalo and Lambani villages located in Collins Chabane Local Municipality of Vhembe District in Limpopo Province. The villages served as case study areas that helped to compare the native approaches that the Va-Tsonga and Vha-Venda used to forecast rainfall in the course of producing V. subterranea. A multi-case study research design, which was exploratory in nature was adopted. Convenience and snowball sampling techniques were used to identify and select respondents. The triangulation of participatory methods, techniques and tools guided the collection of qualitative data. Key informant interviews, learning circles, photovoice, one-on-one interviews and narrative inquiry techniques were applied during data collection. Smallholder farmers and the elderly members of communities were the respondents. Nine key informants in Xigalo and Lambani villages were interviewed. One retired and two currently serving government extension officers were also interviewed. Separate learning circles comprising mainly elderly men and women were also organised. Each learning circle was made up of 7-10 respondents. Atlas.ti version 7.5.7 software was used to analyse the qualitative data following the thematic content analysis approach. It was observed that the respondents were aware of climate variability events that affected V. subterranea. Some of the events were shifts in rainfall patterns, heavy rainfall, extreme temperatures, scarcity of summer rainfall, the disappearance of lunar signs and the seasonal cycle variations. Eighteen types of phenological signs used to predict rainfall were identified. The most common signs included the Milky Way Galaxy of stars, musical sounds of birds and frogs, moon shapes, cumulus and cumulonimbus cloud types. A close relationship between conservation of V. subterranea and adaptation strategies was said to exist. It was evident that most commonly used conservation strategies were rainmaking ceremonies, planting after the summer rains, hoeing weeds, soaking seeds before planting, hilling or earthing up around the base of the V. subterranea plant and storing the legumes in traditional vessels and sacks. The need for integrating western scientific knowledge with native forecasts to inform the production of V. subterranea was uncovered. In addition to this, the needs of Tsonga and Venda communities should inform local policy interventions. Lastly, adaptation strategies that address food insecurity with V. subterranea being part of the agro-ecosystem deserve attention in scientific investigation and policymaking. / NRF

Depuis le début des années 1990 on suit l'évolution globale du niveau de la mer grâce aux satellites altimétriques. Ils observent une hausse du niveau moyen global de la mer (GMSL) de 3.4 ± 0.4 mm/an sur la période 1993-2016 (ce qui représente le double de ce qui a été observé au cours du 20ème siècle par les marégraphes, hausse à 1.7 mm/an entre 1900 et 1990). Le GMSL présente aussi des fluctuations interannuelles qui peuvent atteindre quelques millimètres, surtout pendant les épisodes ENSO. Cette hausse n'est pas régionalement uniforme : elle a été 3 fois plus rapide que la hausse moyenne globale dans certaines zones entre 1993 et 2016. Au cours du 21ème siècle, on s'attend à une hausse accrue du GMSL pouvant aller jusqu'à 1 m à l'horizon 2100, avec une forte variabilité régionale. Il est donc important de comprendre l'évolution actuelle du niveau des océans qui constitue une menace sérieuse pour de nombreuses régions côtières basses souvent très peuplées. Cette thèse s'inscrit dans le contexte du projet niveau de la mer CCI (Climate Change Initiative) de l'Agence Spatiale Européenne (ESA) ayant pour objectif de fournir de meilleurs produits du niveau de la mer combinant les missions Topex/Poseidon, Jason-1/2, ERS-1/2 et Envisat. L'objectif premier de cette thèse est de valider ces produits SL_CCI du niveau de la mer en utilisant différentes approches, en particulier par l'étude du bilan (comparaison du GMSL observé avec la somme des différentes contributions : composante stérique, fonte des glaces continentales et transferts d'eau depuis les terres émergées). Un autre objectif est d'estimer les composantes du niveau de la mer mal connues, et tout particulièrement le contenu thermique de l'océan profond non mesurable par le système Argo, et la contribution du stock d'eau sur les continents. Ces travaux ont montré que la contribution de l'océan profond en dessous de 2000m est faible sur la période 2005-2013 et contenue dans la barre d'incertitudes des données (erreurs qui proviennent essentiellement, (1) des produits niveau de la mer altimétriques et des lacunes de la couverture géographique des données Argo dans la région Indonésienne pour la tendance et (2) des produits GRACE et Argo pour la variabilité interannuelle). Nos résultats et la méthode utilisée montrent que le niveau de la mer et ses composantes sont encore entachés d'erreurs importantes. Dans la deuxième partie, nous avons analysé l'influence du phénomène ENSO (El Niño et La Niña) sur les variations interannuelles du GMSL. Nous montrons que lors des évènements La Niña comme celui de 2010-2011, le déficit de précipitations sur l'océan (et l'excès sur les continents) conduit à une baisse temporaire de la masse de l'océan global et donc du niveau de la mer. C'est essentiellement la variation de masse de l'océan qui explique la variabilité interannuelle du niveau de la mer lors des évènements ENSO, et le déficit (La Niña) ou excès (El Niño) de masse se trouve confiné dans l'océan Pacifique tropical Nord. Pour finir, nous analysons l'évolution de la température moyenne de l'air et de l'océan en surface sur la période du "hiatus" (2003-2013). Nous montrons que ce hiatus, c'est à dire le ralentissement récent de la hausse de la température moyenne globale de la Terre est un phénomène quasi global, même si le Pacifique tropical Est s'est fortement refroidi. Cette "supposée" pause récente s'explique par la variabilité naturelle interne du climat. La Terre est toujours en état de déséquilibre énergétique dû à l'accumulation de gaz à effet de serre. Nous mettons en évidence le rôle de la variabilité naturelle à court terme sur les changements à plus long terme associés au réchauffement climatique anthropique. / Since the early 1990s sea level is routinely measured using high-precision altimeter satellites. These observe a rise in global mean sea level (GMSL) of 3.4 ± 0.4 mm/yr over the 1993-2016 period (which is twice what has been observed during the 20th century by the tide gauges, with a rise of 1.7 +/- 0.3 mm/yr). The interannual variability in the GMSL can reach several millimeters, especially during ENSO events. The rate of sea level rise is not regionally uniform. During the altimetry era, it was three times faster than the global mean in some areas. During the 21st century, we expect a greater rise of the GMSL than today, up to 1 m in 2100, with strong regional variability. It is therefore important to understand the current evolution of the sea level, since it represents a serious threat to many low coastal areas, often densely populated of the planet. My thesis research deals with the Sea Level CCI (Climate Change Initiative) project of the European Space Agency (ESA) which objective is to provide improved sea level products combining several altimetry missions, including Topex/Poseidon, Jason-1/2, ERS-1/2 and Envisat. The primary objective of my thesis was to validate the CCI sea level products using different approaches, in particular the sea level budget approach. It consists of comparing the observed GMSL with the sum of different contributions : the steric component, melting of continental ice and transfers of water between the land surface and oceans. Another objective was to estimate the poorly known components to sea level rise, in particular the heat content of the deep ocean not measurable by Argo, and the contribution of water storage on the land. My work has shown that the contribution of the deep ocean below 2000m to the rising sea level is small over the 2005-2013 periods and not significant compared to the data uncertainties. The main uncertainties come from: (1) -in terms of trend- the altimetry sea level products and gaps in the geographical coverage of Argo data in the Indonesian region, and (2) -in terms of interannual variability- the GRACE and Argo products. My results and the method used show that the sea level and its components are still affected by important errors. In the second part, I analyzed the influence of ENSO (El Niño and La Niña) on the interannual variations of the GMSL. I showed that during La Niña events, like that of 2010-2011, the rainfall deficit over the ocean (and excess over the continents) leads to a temporary decrease in the global ocean mass and therefore in the GMSL. This is essentially the ocean mass variation that explains the interannual variability of the GMSL during ENSO events. Furthermore, the deficit (La Niña) or excess (El Niño) ocean mass is confined in the north tropical Pacific Ocean. Finally, I analyzed the evolution of the average temperature of air and ocean surface over the period of the "hiatus" (2003-2013). I showed that this hiatus, i.e. the recent slowdown in the rise of the global mean Earth's temperature is an almost global phenomenon, though cooling of the tropical eastern Pacific has slightly contributed. This recent pause is attributable to natural internal climate variability. The Earth is indeed still in a state of energetic imbalance due to the accumulation of greenhouse gases. I highlighted the role of the natural variability that is superimposed to the anthropogenic global warming.

Livestock grazing is the most important type of land-use in arid and semi-arid regions. In these regions, uncertain and highly variable climate conditions cause scarce and spatio-temporally variable resource availability. The major challenge to livestock grazing is the efficient utilisation of these resources without running the risk of degradation. Therefore, well adapted grazing strategies that consider both local environmental characteristics and the farmers' individual needs and perceptions are crucial for sustaining human livelihoods. Particularly, rotational grazing is presumed to render adaptation to spatio-temporal heterogeneities possible. A systematic investigation, however, that analyses the interrelations between the major components of rotational grazing systems on appropriate spatial and temporal scales was missing so far. This doctoral thesis investigates different management strategies for sustainable livestock grazing in semi-arid rangelands. Using an integrated modelling approach, it enters into the question: how to adapt grazing systems to spatio-temporal heterogeneous rangeland conditions, variable and changing climate conditions, and different individual needs and goals of livestock farmers? In order to address these issues, the taken approach tackles both methodical challenges and applied concerns. In the first part of this study, a generic modelling framework is developed that incorporates important components of grazing systems on appropriate spatial and temporal scales. To parameterise the model, a pattern-oriented approach is developed that uses qualitative patterns to derive a broad range of plausible parameter sets supporting a general model analysis. In the second part, a variety of management strategies is explored under different climatic, ecological, and economic conditions. The research focuses in particular on combined effects between and relative importance of different management components. The question how the results of different management strategies depend on the type of vegetation is investigated. Furthermore, the performance of rotational grazing strategies is analysed under different economic requirements and rainfall conditions. The study also identifies management strategies that are suitable to adapt a grazing system to spatio-temporally variable rangeland conditions. Overall, this thesis contributes to a general understanding of basic principles for adaptation to spatio-temporal heterogeneities as well as the interplay of different management components. The results allow an evaluation of management strategies for specific situations and the identification of strategies that are robust to a broad range of situations including different aspects of global change.