Le bassin versant de la Durance assure 10% de la production hydroélectrique française, et alimente en eau potable de nombreuses villes du pourtour méditerranéen. Sa partie alpine contribue à près de la moitié des débits à sa confluence avec le Rhône. Connaître les impacts du changement climatique sur l’hydrologie de la Durance présente donc un enjeu de taille pour l’avenir de la population provençale. A cette fin, les modèles hydrologiques sont devenus des outils indispensables. Cependant, le manteau neigeux, de par sa forte variabilité spatiale, est particulièrement difficile à modéliser. Une attention particulière a été portée à la compréhension des processus nivaux, dominants dans la partie amont du bassin, et à leur représentation dans le modèle de surface continentale CLSM. L'analyse de mesures d’extension et de hauteur du manteau neigeux nous a amenés à modifier le schéma de neige de CLSM, en introduisant une hystérésis dans la relation entre ces deux variables. Nous avons ensuite évalué le potentiel devenir sous changement climatique de l’hydrosystème. Les résultats des simulations de CLSM, réalisées à l’aide de 330 scénarios climatiques, ont été confrontés à ceux de cinq autres modèles hydrologiques. Tous s'accordent sur une baisse significative des débits annuels, avec des modifications du régime hydrologique différentes selon les zones. L’incertitude reste importante sur la magnitude des changements de débits, principalement due aux scénarios climatiques. La part d’incertitude associée aux modèles hydrologiques varie selon les saisons, mettant en évidence une plus grande difficulté à représenter les processus nivaux et ceux liés à l’évapotranspiration. / The Durance watershed, located in the French Alps, generates 10% of French hydropower and provides drinking water to about 3 million people. The upstream part of this catchment, where snowfall accounts for more than 50% of the precipitation, is responsible for almost half of the total runoff whereas it accounts for only 25% of the area. To assess the impacts of climate change on the water resources, hydrological models are now commonly used. The snow cover is, however, difficult to simulate because it is highly variable in both space and time. Therefore, special attention has been given to understanding the snow processes in this alpine environment, and to their representation in a land surface model, the CLSM. The analysis of snow-cover extent and height observations has lead us to modify the snow scheme of the CLSM, by introducing a hysteresis in the snow depletion curve. Then, we quantified the possible evolution of the Durance hydrosystem using 330 climate change scenarios. The results of the CLSM are compared with those of five other hydrological models. All models are in agreement in predicting a significant reduction of discharge with some different modifications of the hydrological regime depending on the different zones of the catchment. Uncertainties remain important concerning the magnitude of discharge changes, mainly due to the climatic scenarios. The uncertainty related to hydrological modelling is indeed low but varies depending on the season. This highlights some of the difficulties in using hydrological models to correctly represent snow processes and evapotranspiration processes especially under water stress.

This research project provides a geomorphological and geological analysis of ephemeral and relict river valley systems in the north part of the Iraqi Western Desert. The area surveyed covers approximately 30 000 km2 and is one of the remotest and least studied parts of the Arabian Peninsula. Part of the reason for the lack of research in this area in recent years has been the ongoing security problems and all fieldwork undertaken for this thesis was carried out with the support of armed guards and police. In addition much of the work on the geology and geomorphology of the region is in confidential files commissioned by oil companies, and in MSc and PhD theses held in Iraqi Universities. A significant part of this work and indeed many scientific papers, are only available in Arabic. Therefore a major element of the work for this thesis has been to translate this material and make the results available in English for the first time. The study demonstrates that the present surface of the Iraqi Western Desert overall forms an incised plateau developed during two phases of continental erosion and deposition during the Tertiary and Quaternary periods. The first phase started after Oligocene uplift formed an older plateau within the Oligocene Tayarat formation. This plateau is characterized by denudation processes associated with a semiarid climate, including the formation of subsurface hollows and caves. The second phase, which began after the last Alpine Orogenic movement, and includes the Pliocene and Quaternary periods, formed a younger plateau developed on the Zahra formation. This younger plateau is characterized by processes indicative of climatic fluctuations from wet to arid and semiarid, which induced denudation in places and deposition in others. However, in terms of the geomorphological landforms present in the Western Desert they can be broadly divided into: i) Structural and erosion-denudation forms ii) Accumulation forms Lithology landform in these two categories has resulted in the production of a new geomorphological map of the Iraqi Western Desert. A key component of this map uses the drainage networks. Four distinct drainage systems were identified: 1. The valleys which descend from west to east. These valley systems are located to the south and south east town of Rutba 2. The valleys which descend from south to north. These lie to the west and southwest of Rutba and are controlled by the north to south strike of exposed Palaeogene strata. 3. The valleys which descend from east south to north west, located north of the Garaa area like Ratga and Akash. 4. The valleys which descend from east to west. These valley systems are located to the south and south west town of Rutba, like Swab and Wallaj valleys. Investigation of these four networks established that they were relict systems that still carried ephemerally active misfit rivers and stream. The overall control on their form was the alternating sequences of variable strength rocks that were exposed and eroded as part of the uplift of an anticlinorium (Houran) and anticline (Garaa), associated with the Alpine Orogeny .However, the unclearing Structures were much older and can be traced back to Permian tectonic processes. The drainage of the Western Desert, therefore, is antecedent and controlled by Tertiary and Quaternary tectonics. The rivers appear to have active throughout the Pleistocene incising into the Western Desert plateaux. Highest incision rates probably occurred during more pluvial periods in the Pleistocene which may have been coincident with glacial marine in the Northern Hemisphere. The contemporary rivers are misfit within larger valleys although still subject to flash floods under the right metrological condition.

Lake surface water temperatures (LSWTs) of 246 globally distributed large lakes were derived from Along-Track Scanning Radiometers (ATSR) for the period 1991 to 2011. These LSWTs, derived in a systematic manner, presents an ideal opportunity to study LSWT behaviour on a global scale. In this thesis, the annual cycles of lake-mean LSWTs derived from these data quantify the responses of large lakes’ surface temperatures to the annual cycle of forcing by solar radiation and the ambient air temperature. Minimum monthly net surface solar irradiance (netSSI) strongly influences minimum LSWTs of non-seasonally ice covered lakes (where lake-mean LSWT remains above 1 ºC throughout the annual cycle), explaining > 0.88 (R2 adj) of the inter-lake variation in both hemispheres. In some regions, for seasonally ice covered lakes (where lake-mean LSWT remains below 1 ºC for part of the annual cycle) the minimum monthly netSSI is a better predictor than latitude, of the length of the frozen period, which shows the importance of local cloud climatological conditions. Additionally, at lake locations between 1º S to 12º N, the netSSI, shown to peak twice annually, is reflected in the LSWT annual cycle. The summer maximum LSWTs of lakes from 25º S to 35º N show a linear decrease with increasing altitude; -3.76 + 0.17 ºC km-1 (R2 adj = 0.95), marginally lower than the corresponding air temperature -4.15 + 0.24 ºC km-1 (R2 adj = 0.95) decrease with altitude. The start and end of the period where the lake-mean LSWT is greater than 4 ºC shows strong correlation with the spring and autumn 0 ºC air temperature crossing days, (R2 adj = 0.74 and 0.80 respectively). The temporally and spatially resolved LSWT observations allows for a greater practical understanding of LSWT behaviour of large lakes. For example, lakes with a greater LSWT annual range have more observed variability in the LSWT extremes, highlighting that they may be more responsive to changes in the climate than lakes with a low annual range. The nighttime LSWT trends show stronger warming than day-night trends in the all regions, except Europe. The lake centre LSWT trends and absolute values can be generally considered representative of the lake-mean LSWT trends and absolute values. The observed LSWT time series are used to tune a 1-dimensional thermodynamic lake model, FLake. By tuning FLake using only 3 basic lake properties, shown by myself to have the most influence over LSWTs (depth, snow and ice albedo and light extinction co-efficient), the daily mean absolute differences for 244 lakes is reduced from 3.38 + 2.74 ºC (untuned model) to 0.85 + 0.61 ºC (tuned model). The effect of wind speed, lake depth, albedo and light extinction co-efficient on LSWTs is demonstrated throughout the tuning process. The modelled summer LSWT response to changes in ice-off is strongly affected by lake depth and latitude explaining 0.50 (R2 adj, p = 0.001) of the inter-lake variation in summer LSWTs. Lake depth alone explains 0.35 (p = 0.003) of the variation, highlighting the sensitivity of the summer LSWTs of deeper lakes to changes in the ice-off. Statistically significant summer/ maximum month modelled LSWT trends, from 1979-2011 are presented for lakes where the modelled LSWTs are strongly supported by observed LSWTs over the period of available observed LSWTs. For these lakes, the trends show LSWT warming of between 0.73 – 2.10 ºC for 29 lakes in northern temperate regions over the 33 year period (1979 – 2011). The modelled regional trends of all lakes over the same period show least warming in Africa of 0.30 ºC and the greatest warming in Europe, 1.35 º.

Results of mineral magnetic, mobile clement, and granulometric analyses of Holocene sediments from Buttermere and Crummock Water (two closely-linked lakes in the north-west of the English Lake District) are presented. These are used to: (1) identify effects of internal (lacustrine) and external (catchment) controls on sedimentation; (2) establish catchment source-lake sediment linkages and assess the value of mineral magnetic techniques in palaeolimnological studies; (3) identify major catchment environmental changes. Analyses of lake sediment fabrics (using sediment thin sections, SEM clay flake analysis, standard granulometric analysis, and mineral magnetic indicators of grain size change) indicate that river plume sedimentation is the normal sediment dispersal mechanism in these lakes. Thin (< = 3.0 mm) chlorite-rich laminae, found at intervals in the otherwise homogeneous Holocene sediment sequence, are probably formed by trapping and concentration of fine, platy particles within lake waters. They are subsequently deposited during lake overturn. This represents an "internal" control on sedimentation. A model of sedimentation processes operating in these lakes is developed, incorporating river plume sedimentation, episodic density surges, and lake thermal structure. Mineral magnetic measurements allow the objective subdivision of the lacustrine lithostratigraphy, identifying broad changes in lake sediment characteristics. Samples from both lake catchments are clustered into six magnetically distinct groups - despite the lithological complexity of the catchment. Comparison of these with the lake sediments has enabled identification of major sources during the Holocene. Following deposition of relatively unaltered bedrock-derived material during the Late-glacial ("primary" sources), secondary sources (which may include glacial diamicts, soils and stream sediments) dominate the lake sediments. Direct input of topsoil-derived sediment from circa 1000 A.D. onwards (during and following the main period of Norse settlement of the Lake District) is identified by its distinctive mineral magnetic characteristics, (high Xfd% values, >-4%). Industrially-derived magnetic spherules contribute significantly to the mineral magnetic characteristics of the more recent sediments, (mainly those post-dating circa 1900 A.D.). These are used to construct a proxy chronology for recent sediments. Catchment environmental changes arc mainly related to stabilisation of vegetation following deglaciation and, from circa 2,000 B.P., anthropogenic effects of deforestation and land disturbance, thus increasing lake sediment accumulation rates. These findings are broadly consistent with the interpretation of the Lake District Post-glacial sediment sequence presented in studies by Mackereth, (1966a), and Pennington, (1981), demonstrating a uniformity of lake and catchment development within the Lake District. A prominent minerogenic layer present in the Buttermere and Crummock Water sediment sequence however broadly correlates with similar horizons deposited in other Lake District lakes from circa 7,400 - 5,000 B.P. These have been previously interpreted as composed of topsoil-derived material derived from human actions, (Pennington 1973, 1981). In the Buttermere and Crummock Water sediments, this layer is best interpreted as derived from glaciogenic sediment') reworked from within the lake basins, probably following lowered lake water levels during the period circa 7,300 - 5,300 B.P. Thus it is suggested that a reinterpretation of similar Lake District lacustrine sediments using the methods employed in this study would be appropriate.

No description available.

No description available.

En dépit de son importance pour les écosystèmes aquatiques, l’origine et les mécanismes de production du carbone organique dissous (COD) sont toujours sujets à discussion. Cette thèse vise à mieux comprendre le rôle joué par l’hydrologie tant sur les mécanismes de production et la location des sources du COD que sur les flux et les processus de transfert du sol vers les cours d’eau. Un suivi haute fréquence des eaux du sol et de rivière a été réalisé sur la totalité d’un cycle hydrologique dans le bassin versant expérimental de Kervidy-Naizin (Morbihan, Observatoire de Recherche en Environnement (ORE) AgrHys). Le COD contenu dans ces eaux a été caractérisé par sa composition isotopique en carbone (δ13C), et ce suivi a été complété par l’analyse de la dynamique de la nappe. L’analyse des variations saisonnières de concentration et de composition du COD révèle l’existence d’un relais dans les sources et les mécanismes de production du COD dans les sols des zones de fond de vallée en phase avec les changements de régime hydrologique du bassin. Ainsi, lors de la période automnale de remontée de la nappe, le compartiment COD présente un caractère faiblement aromatique et une composition isotopique particulière, suggérant une origine microbienne. Ce réservoir de très faible taille (5% du flux annuel de COD exporté par le cours d’eau) est entièrement épuisé par les premières crues d’automne. La mise en charge de la nappe en versant conduit à la mobilisation d’un second réservoir de COD caractérisé par une aromaticité élevée et une signature isotopique proche de celle de la matière organique des sols. Ce compartiment correspond au fond humique « ancien » des sols et présente une taille nettement plus important que le premier (90% du flux annuel). Cependant ce compartiment ne se limite pas au COD produit dans sols des zones de bas-fond mais comprend aussi des apports de matières organiques issues des sols de bas de versant. Enfin, le rabattement de la nappe au milieu du printemps marque le début de la période d’assèchement du bassin versant et le retour dans les sols de bas-fond d’un COD faiblement aromatique et isotopiquement similaire à celui observé en automne. En conséquence, la teneur et la composition du COD véhiculé par le cours d’eau varient fortement à l’échelle saisonnière (variabilité temporelle des processus de production du COD). Les horizons superficiels des sols des zones de bas-fond sont la principale source de COD, par lesquels transitent entre 60 et 80% du flux de COD exporté lors des événements de crues. Concernant le réservoir humique, un résultat important est qu’il semble limité dans les sols de bas de versant, alors qu’il semble au contraire illimité dans les sols de bas fond. Au final, les résultats obtenus dans le cadre de cette thèse démontrent le contrôle majeur de la dynamique de la nappe sur les mécanismes de production de COD par le sol ainsi que sur la localisation spatiale et les flux de COD transférés de ces sols vers les cours d’eau à l’échelle de la saison. Ils valident également l’hypothèse émise selon laquelle l’exportation du COD par les cours d’eau ne résulterait pas d’un simple « flushing » du COD produit dans les sols de bas-fond mais mobiliserait plutôt un ensemble de sources localisées le long du continuum rivière-zone de bas-fond-zone de versant, et dont le caractère plus moins limité du réservoir expliquerait les dynamiques des concentrations en COD annuelles observées à l’exutoire de ces bassins. Un autre résultat majeur de cette thèse est la qualification des isotopes comme outil de traçage des sources et de la dynamique du COD dans les bassins versants. Un point clé de ce travail est la différenciation isotopiques des réservoirs COD des sols de bas-fond et des sols de bas de versant, qui a permis pour la première fois de révéler la participation de l'un et de l’autre au flux exporté à l’exutoire et d’estimer quantitativement leur contribution respective à ce même flux. / Despite its importance to aquatic ecosystems, origin and production mechanisms of dissolved organic carbon (DOC) are still subject to discussion. This thesis aims to better understand the role of hydrology in controlling both DOC production in soils and export by stream waters. High frequency monitoring of soils and streams waters was carried out during an entire hydrological cycle in the Kervidy-Naizin catchment (Morbihan). The DOC was characterized by its stable carbon isotopic composition, and groundwater table dynamic was also analyzed during the same period. The analysis of seasonal variations in both DOC concentration and composition reveals a succession in the sources and mechanisms of production of DOC in soils of valley bottom in line with changes in the hydrological regime of the basin. Thus, during the autumn period of rising water-table, the COD compartment has a low aromatic character and a particular isotopic composition, suggesting a microbial origin. This reservoir of very small size (5% of the annual flux of DOC exported by the stream) is fully exhausted by the first autumn storm events. The increase of water-table in the upland domain leads to the mobilization of a second DOC pool characterized by a high aromaticity and a isotopic signature similar to those of soil organic matter. This compartment corresponds to "old" humic fraction of the soil and has a size much larger than the first (90% of annual flux). However, this compartment includes contributions of DOC from riparian soils and also upland soils. Finally, the water-table drawdown in the middle of spring marks the beginning of the drying period of the catchment and the return to the soil of the “fresh” DOC pool. As a result, concentration and composition of DOC transported by the stream vary greatly across a seasonal scale (temporal variability of production processes DOC). The surface horizons of soils riparian soils are the main source of DOC, which handle between 60 and 80% of stream DOC flux during flood events. An important result is that the DOC pool seems limited in upland soils, while its seems no-limited in riparian soils. Finally, the results obtained in this thesis demonstrate the major control of the dynamics of the water-table on the mechanisms of production of DOC from the soil as well as the spatial location of DOC sources and DOC flux transferred from the soil to stream through the hydrological cycle. They also validate the hypothesis that DOC export by rivers mobilizes rather a set of sources located along the upland-riparian-stream continuum. Another major result of this thesis is the validation of stable carbon isotopes as a tool for tracing sources and dynamics of DOC in the catchment.

Les hydrologues sont questionnés au sujet de l’évolution à moyen et long terme de la ressource en eau. Pour répondre à ces questions, ils ont recours à des modèles hydrologiques. Une quantification des incertitudes associées à leurs projections leur est souvent demandée et pose la question de la robustesse de ces modèles conceptuels, notamment dans le contexte d’un climat en évolution. En effet, l’utilisation d’un modèle dans des conditions différentes de celles où il a été calé repose sur une hypothèse de transférabilité des paramètres, c’est-à-dire la possibilité d’utiliser les paramètres d’un modèle dans des conditions différentes de celles de saconstruction. Nous traitons cette problématique en tentant d’apporter des éléments de réponse aux questions suivantes :• Quel est le niveau de robustesse des modèles hydrologiques conceptuels face à des conditions climatiques variables ?• Quelles sont les causes des situations de manque de robustesse et y a-t-il des façons de s’en prémunir ?Nous y répondons à travers l’étude des performances de modèles conceptuels lors de nombreux tests de transfert temporel de leurs paramètres. Les résultats montrent des corrélations entre les problèmes de robustesse et les écarts de conditions climatiques entre les périodes de calage et de validation des modèles. Des situations de biais systématiques corrélés à des changements de température de l’air ressortent en particulier de ces analyses. Une hétérogénéité existe cependant sur notre échantillon, les variables climatiques ou le type d’erreurs associées aux problèmes détectés pouvant varier d’un bassin à un autre.L’étude des biais de simulation sur les bassins où les modèles ne sont pas robustes montre des alternances de phases de sous-estimation et surestimation des débits, le biais sur le module décennal pouvant atteindre 20%. Nos travaux révèlent que des comportements très similaires sont obtenus pour diverses périodes et méthodes de calage. Pour les modèles conceptuels utilisés dans ces travaux, les problèmes de robustesse détectés ne résultent pas simplement de quelques "mauvais" calages conduisant à la sélection de paramètres inadaptés pour reproduire le comportement du bassin considéré. Ils semblent plutôt découler de difficultés générales de ces outils de modélisation à reproduire convenablement les bilans moyens de façon simultanée sur différentes périodes.Ces travaux ouvrent des pistes de réflexion sur la capacité insuffisante de certains modèles hydrologiques conceptuels à reproduire des dynamiques basse fréquence et sur le rôle des erreurs d’estimation des forçages, notamment les variations temporelles de l’évapotranspiration, dans les situations de modélisation incorrecte. / Hydrologists are asked to estimate the medium- and long-term evolutions of water resources. To answer these questions, they commonly use conceptual models. In addition, they are often required to provide an estimate of the uncertainties associated with model projections. This raises the question of the robustness of conceptual models, especially in the context of climate evolution. Indeed, using a model in conditions different from those of calibration is based on the hypothesis of parameter transferability, i.e. the possibility to use model parameters in conditions different from those used for the model set-up. We focus on this issue with the aim of answering the following questions:• What is the robustness level of conceptual hydrological models in the context of changing climatic conditions?• What are the causes for the lack of robustness, and are there ways to prevent it?We answer these questions by studying the performance of conceptual models through multiple tests of temporal transfer of their parameters. Results show the existence of correlations between the robustness problems and the difference in climate conditions between model calibration and validation periods. The analysis especially points out the situations of systematic bias correlated to differences in air temperature. However, results are heterogeneous in our catchment set, and climate variables or error type associated with the identified problems vary between catchments.The analysis of simulation biases on catchments where the models are not robust shows alternating phases of flow under- or overestimation, with a possible bias in the mean flow up to 20% over a ten-year period.Our work reveals that very similar results can be obtained for various periods or calibration methods. The robustness issues faced by conceptual models used in this study do not solely stem from inadequate calibrations leading to the selection of parameters unable reproduce the catchment behavior. They seem to be the consequence of overall difficulties for models to satisfactorily simulate water balances simultaneously on various periods.This work opens reflections on the limited capacity of some hydrological models to reproduce low-frequency dynamics and raises questions on the role of inputs estimates errors in model failures, especially the temporal variations of evapotranspiration.

L’Orge est une petite rivière de plaine fortement segmentée, du bassin parisien. En 2010, elle comptait en moyenne un clapet tous les kilomètres sur sa partie aval. Pour rétablir la continuité écologique du cours d’eau, le Syndicat de l’Orge efface progressivement ces ouvrages. L'objet de cette thèse est le suivi biogéochimique, hydro-sédimentaire et écologique d'un bief de l’Orge de 6 km où 3 clapets consécutifs ont été effacés.Les hauteurs d’eau dans la rivière ayant baissé, des bancs de vase ont été mis à l'air. Ces banquettes de sédiments subissent des séquences d’inondation/exondation au rythme des crues. Les transformations de l’azote et du soufre y sont particulièrement actives et dynamiques. Cependant, les faibles flux d’eau échangés (modélisation) limitent l’impact des banquettes sur la rivière. Des processus intenses de précipitation/solubilisation des sulfates, liés aux variations des niveaux d’eau, ont une importance non négligeable sur la dégradation de la matière organique et la production d’ammonium dans la banquette.La capacité de transport de sédiments de la rivière a fortement évolué suite aux effacements, ce qui se traduit par l’apparition d'une granulométrie plus grossière en fond de lit. Ont été évalué la nouvelle vitesse de transit des particules charriées et le temps nécessaire pour atteindre un nouvel équilibre du lit et des espèces biologiques qui y vivent. Les indicateurs biologiques montrent que les effacements améliorent l’habitabilité de la rivière.Ces suivis montrent l’impact positif des effacements de barrage sur une rivière de plaine, même si un suivi dans la durée reste nécessaire pour confirmer les évolutions mises en évidence. / Orge is a lowland river from Seine’s watershed, France. In 2010, Orge River was highly segmented with on average one dam (< 2m) every kilometer in downstream part. To restore the ecological continuity of the river, the “Syndicat de l’Orge” gradually removes these little dams. The purpose of this thesis is the monitoring of hydro-morphology, biogeochemistry and ecological of a section of 6 km of Orge where 3 consecutive dams have been removed in spring 2010.After the water level decrease, some previously drowned sediment deposits appear along the river bank. These sediment banks are frequently inundated by high flows. Nitrogen and sulfate processes are very intense onto these banks. However, a model shows that water fluxes limited biochemical impact of sediment banks on the river. Intense precipitation/solubilization sulfate processes, link with water level variations, have a significant importance on organic matter degradation and ammonium production in sediment banks.The sediment transport capacity of the river has changed significantly after dam removal. It results in an increase of the maximum transportable diameter of the river, leading to the appearance of a coarser bed in-situ. It was possible to evaluate the new transit speed of particles carried along and highlight the time required to reach a new equilibrium of the bed and the biological species that live there. Biological indicators show that removals enhanced the livability of the river.All these studies show the positives impacts of dam removal on a lowland river, even if a monitoring in time is still needed to confirm the first developments highlighted here.

L’objectif de la thèse est, tout d’abord, d’identifier et caractériser la trajectoire spatio-temporelle d’un petit cours d’eau périurbain de faible énergie, à travers l’exemple francilien de la Mérantaise (Essonne-Yvelines). Ces objets d’étude sont, en effet, trop souvent négligés par la communauté scientifique alors même que les enjeux environnementaux de restauration et de préservation écologique, préconisés depuis plusieurs décennies par la DCE, y sont importants (impacts de l’urbanisation et des aménagements anciens). Une approche croisée géohistorique et hydromorphologique a permis de reconstituerl’évolution passée et actuelle de la rivière : la vallée a été fortement aménagée depuis le Moyen-Âge par des équipements hydrauliques modestes mais ayant affecté plus de 80% du linéaire. Les moulins à eau ont été abandonnés au cours du 20ème siècle, libérant partiellement le chenal. Désormais, la trajectoire actuelle, marquée par des ajustements morphologiques significatifs récents, est moins commandée par les aménagements anciens que par les pressions anthropiques actuelles qui touchent le bassin versant(urbanisation, réseau de transports et de drainage, intensification de l’agriculture). Enfin, cette thèse est également l’occasion de s’interroger sur le caractère écologiquement pertinent et hydromorphologiquement durable des travaux de restauration des milieux aquatiques. Ce travail doctoral propose, par ailleurs, de considérer les opérations de restauration comme de nouveaux équipements de la rivière répondant à des besoins socio-politiques nouveaux et participants pleinement de la trajectoire fonctionnelle de la rivière. / The objective of this research is, first of all, to identify and characterize the spatio-temporal trajectory of a small low energy periurban river, through the example of the Merantaise River located in the Seine Basin. These objects of investigation are too often neglected by the scientific community even though the environmental issues of ecological restoration politics, advocated for several decades by the WFD, are important (impacts of urbanization and ancient watermills). An interdisciplinary geohistorical and hydromorphological approach allowed to reconstruct the past and present evolution of the river: the valley has been heavily developed since the Middle-Ages by modest hydraulic equipment but has beenaffected on more than 80% of the channel. Water mills were abandoned during the 20th century, partially clearing the channel. Nowadays, the current trajectory is marked by recent significant morphological adjustments and is less controlled by old hydraulical structures than by current anthropogenic pressures affecting the watershed (urbanization, transport and drainage network, intensification of agriculture). Finally, this thesis is also an opportunity to question the ecologically relevant and hydromorphologically sustainable nature of ecological restoration. This doctoral work also proposes to consider restoration operations as new river facilities responding to new socio-political needs and fully participating of the river's functional trajectory.