Hydro-Climatic Decoupling and the Terrestrial Drivers of Streamflow Persistence Across European Catchments

Turning rainfall into streamflow is a complex process. Catchments are not passive conduits for runoff; their geological characteristics actively buffer the hydrological response. While weather and streamflow both exhibit "long-term memory"—quantified by a Hurst exponent > 0.5—we do not fully understand how the ground alters this memory before water reaches the river. This thesis investigates over 3,300 European catchments to determine if a river reflects rainfall memory or if the catchment creates its own. By calculating the Hurst exponent for precipitation ($H_P$) and streamflow ($H_Q$) using Rescaled Range analysis, we examine the relationship between atmospheric and hydro-geological processes. Statistical tests (ANOVA and Kruskal-Wallis) confirm hydro-climatic decoupling. Streamflow consistently exhibits stable memory, with $H_Q$ values clustering between 0.56 and 0.63, showing no significant dependency on whether rainfall was random or persistent ($p > 0.05$). This suggests the catchment acts as an active, non-linear filter that buffers atmospheric forcing, generating its own long-term persistence. Spatial mapping using Ordinary Kriging reveals how the landscape controls this. Regions with deep soils and active groundwater (like UK Chalk aquifers) absorb rainfall, creating strong streamflow memory. In contrast, steep, rocky mountains (like the Alps) promote rapid runoff, neglecting rainfall memory and showing random streamflow ($H_Q < 0.5$). Ultimately, this research suggests that hydrological droughts and floods do not always match the weather. A region may experience random rainfall, yet still suffer from persistent, multi-year streamflow droughts due to its underground geology.