This thesis deals with the role of precipitation on erosion and landscape formation in the Nepal Himalayas. I investigate all successive steps involved in the erosion process: 1) Starting from the evaluation of precipitation datasets, 2) the transfer of precipitation to river discharge, 3) the mobilization and transport of material out of the mountain range, 4) and finally, erosion constrains over longer time-scales. I show that the dataset derived from the interpolation of rain gauge data performs best in the Himalayas. I demonstrate the importance of an until now unconsidered, major compartment of the Himalayan discharge cycle, which I identify as a fractured basement aquifer, and estimate the snow and ice melt contribution to the Himalayan rivers. Erosion rates calculated from suspended sediment fluxes and cosmogenic nuclide analysis range between 0.1 and 4 mm/yr. The rivers in the Nepal Himalayas are supply limited and the hillslopes as contributing source are transport limited. Last I show that over several thousand years erosion is not related with precipitation, but with relief. / Cette thèse porte sur le rôle des précipitations sur l’érosion et la formation des reliefs dans l’Himalaya Népalais. J’étudie chaque étape du processus d’érosion : 1) Evaluation des bases de données de précipitations, 2) Transfert des précipitations au débit fluvial, 3) Mobilisation et transport du matériel dans le bassin versant, et enfin 4) Mécanismes d’érosion sur de longues échelles de temps. Je montre que la base de données de précipitations obtenue par interpolation de données pluviométriques est la plus performante pour la région de l\'Himalaya. Je démontre l’importance d’une composante majeure, jusqu’alors ignorée, du cycle de débit de l’Himalaya que j’identifie comme étant les aquifères de sous-sol fracturé, et j’évalue la contribution de la fonte des neiges et glaces aux rivières Himalayennes. Les taux d’érosion calculés à partir des flux de sédiments en suspension et des analyses de nucléides cosmogéniques varient de 0.1 à 4 mm/a. Les rivières au Népal sont limitées par l’apport sédimentaire alors que les versants, en tant que source de sédiments, sont limités par le transport. Enfin, je montre que l’érosion sur des milliers d’années ne dépend des précipitations mais du relief. / Die vorliegende Arbeit beschäftigt sich mit der Rolle des Niederschlag bei Erosions- und Oberflächenprozessen im nepalesischen Himalaja. Ich untersuche die Abfolge der Erosionspsozesse im Himalaja: 1) Ausgehend von der Bewertung von Niederschlagsdatensätzen, 2) die Prozesse der Abflussbildung in Flüssen, 3) die Mobilisierung und Transport von Material, 4) und Erosionsraten über längere Zeiträume. Ich zeige, dass interpolierte Niederschlagsdaten die beste Qualität im Himalaya haben. Ich zeige auf, wie wichtig der bislang unberücksichtigt Grundwasserzwischenspeicher für die Abflussbildung im Himalaya ist und schätze den Anteil der Schnee-und Eisschmelze an dem Gesamtabfluss der Flüssen im Himalaja. Erosionsraten die mittels Schwebestofffracht und der Analyse kosmogener Nukluide berechnet wurden, liegen zwischen 0,1 und 4 mm pro Jahr. Der Sedimenttransport in den Flüssen in Nepal ist limitiert durch die Verfügbarkeit von transportierbarem Material, während der Transport und die Mobilisierung auf den Hängen durch die Verfügbarkeit von Wasser limitiert ist. Zudem sind die Erosionsraten über mehrere Jahrhundert nicht von der Niederschlagsverteilung abhängig sondern vom Relief.

<p>Understanding climate-driven changes in global land-based ice volume is a critical component in our capability to predict how global sea level will rise as a consequence of the current human-driven climate change. At the last glacial maximum (LGM, which peaked around 20 ka), ephemeral ice sheets covered vast regions of the northern hemisphere while both the Greenland and Antarctic ice sheets were more extensive than at present. As global temperatures rose at the transition into the Holocene, driving the LGM deglaciation, eustatic sea level rose by approximately 125 m. The east Antarctic ice sheet (EAIS) is the largest ice sheet on Earth today, holding an ice volume equivalent to ca. 53 m rise in global sea level. Considering current trends in global climate, specifically rapidly increasing atmospheric CO₂ levels and global temperature, it is important to improve our understanding of how the EAIS will respond to global warming so that we can make better predictions of future sea level changes to guide community adaptation and planning efforts. Numerical ice sheet models which inform projections of future ice volume changes, and can, therefore, yield projections of sea level rise, rely on empirical data to test their ability to accurately represent former and present ice configurations. However, there is a general lack of data on the paleoglaciology of the EAIS along the western Dronning Maud Land (DML) margin. In order to address this situation, the paleoglaciology of western DML forms the focus of the work presented in this thesis.</p><p><b> </b></p><p>Together with collaborators within the MAGIC-DML consortium (Mapping, Measuring and Modelling Antarctic Geomorphology and Ice Change in Dronning Maud Land) that provides the funding for this MS project, the author has performed geomorphological mapping across western DML; an area of approximately 200,000 km². The results of the mapping presented in this thesis will provide the basis for a detailed glacial reconstruction of the region. The geomorphological mapping was completed almost entirely by remote sensing using very high-resolution (sub-meter in the panchromatic) WordView-2 and WorldView-3 (WV) satellite imagery, combined with ground validation studies during field work. Compared to Landsat products, the improved spatial resolution provided by WV imagery has fundamentally changed the scale and detail at which remote sensing based geomorphological mapping can be completed. The mapping presented here is focused on the glacial geomorphology of mountain summits and flanks that protrude through the ice sheet’s surface (nunataks). In our study area of western DML these nunatak surfaces make up <0.2 % of the total surface area, and the landforms mapped here are generally smaller than can be identified from Landsat products (30 m spatial resolution). The detail achieved in our mapping, across such a vast, remote area that presents numerous obstacles to accessibility highlights the benefits of utilizing the new VHR WV data. As such an evaluation of the WV data, as applied to geomorphological mapping is presented here together with our mapping of the glacial geomorphology of western DML. The results of which provides evidence of ice having overridden sites at all elevations across the entire study area; from the highest elevation inland nunataks that form the coast-parallel escarpment, to low-elevation emerging nunataks close to the coast. Hence from our studies of the glacial geomorphology of this region we can ascertain that, at some point in the glacial history of western DML, ice covered all of the mountain summits that are exposed today, indicating an ice sheet surface lowering of up to 700 m in some places.</p>