Rinaldo, Andreavan Meerveld, IljaBujak-Ozga, Izabela2024-07-022024-07-022024-07-02202410.5075/epfl-thesis-10798https://infoscience.epfl.ch/handle/20.500.14299/208948Despite their high ecological value, non-perennial streams have received less attention than their perennial counterparts. This doctoral thesis addresses this disparity by advancing knowledge on the dynamics of the drainage density and hydrologic processes in catchments with non-perennial streams in two distinct Swiss regions: the pre-Alps and the Plateau. This thesis introduces two innovative methodologies to monitor stream network dynamics. Firstly, it presents a framework to gather hydrological data through the use of a self-developed mobile phone application. It can be used to efficiently map non-perennial streams in the field while minimizing subjectivity and human-made errors. Secondly, the field-collected data on water levels and flow presence or absence are combined with continuous data on water levels in the new ConsEntuAl State Estimation (CEASE) method to obtain information on the drainage density at a sub-hourly temporal resolution. The method was validated for four catchments, with accuracies on the order of, and often exceeding, 94%. Together, these two methodological advancements reduce the limitations of traditional field surveys in accessing drainage density dynamics during rainfall events. The thesis also presents two novel open datasets on stream state and stream chemistry for four headwater catchments, two in the pre-Alps and two on the Plateau. With the help of the newly developed methods to obtain high temporal and spatial resolution data on the drainage density, it was possible to obtain insights into the response of the drainage density to rainfall-runoff events across diverse hydrological settings, and to relate these to the stream chemistry response. At each location, one of the catchments was steeper, and the other was flatter and influenced by ditching. Despite their proximity and similarity in geology, soils, and size, there were clear differences in the dynamics of the stream networks and stream chemistry responses to rainfall-runoff events. Hydrochemistry and drainage density were more stable in the steeper catchments with multiple perennial springs that maintained flow throughout the network and the monitoring period. Contrary, for the flatter catchments with shallow channels and artificial ditches, the stream network and stream chemistry varied more. In particular, nitrate mobilization was observed at the time of stream network expansion. These detailed datasets on stream, groundwater and soil chemistry, together with the hydrometric measurements, also provided information on runoff processes in the research catchments. In particular, they showed that event water can not be generated solely by rainfall falling on the stream channel, except during small events, and that the soil water contribution was negligible, except during large events on the Swiss Plateau. In summary, the results of this PhD research provide valuable insights directly and also for future research by contributing to methodological developments and increasing data availability on stream network and hydrochemical dynamics. The presented results contribute to improving our understanding of hydrological processes in catchments and highlight the complexity of drainage density dynamics across catchments, and the large spatial variability in both flowing drainage network dynamics and hydrochemistry. It, thus, emphasizes the importance of considering local hydrological processes in watershed management strategies and conservation efforts.enrunoff generation mechanismshydrological processesintermittent and ephemeral streamsEnd Member Mixing Analysis (EMMA)hydrochemistryisotope hydrograph separationnitrate dynamicsstream intermittencydrainage densityrainfall-runoff eventsthesis::doctoral thesis