Data-driven analysis of water and nutrient flows: Case of the Sava River Catchment and comparison with other regions

Abstract: A growing human population and demands for food, freshwater and energy are causing extensive changes in the water and biogeochemical cycles of river catchments around the world. Addressing and investigating such changes is particularly important for transboundary river catchments, where they impose additional risk to a region’s stability. This thesis investigates and develops data-driven methodologies for detecting hydro-climatic and nutrient load changes and their drivers with limited available data and on different catchment scales. As a specific case study, we analyze the Sava River Catchment (SRC) and compare its results with other world regions. A past–present to future evaluation of hydro-climatic data is done on the basis of a water balance approach including analysis of historic developments of land use and hydropower development data and projections of the Coupled Model Intercomparison Project, Phase 5 (CMIP5) output. Using observed water discharge and nutrient concentration data, we propose a novel conceptual model for estimating and spatially resolving total nitrogen (TN) and total phosphorus (TP) input and delivery-retention properties for a river catchment and its nested subcatchments, as well as detection of nutrient hotspots. The thesis identifies hydroclimatic change signals of hydropower-related drivers and finds consistency with other world regions. The proposed nutrient screening methodology provides a good distinction between human-related nutrient inputs and landscape-related transport influences on nutrient loading at subcatchment to catchment scale. A cross-regional comparison of the SRC data with the Baltic region shows similarity between nutrient-relevant indicators and driving socio-economic and hydro-climatic conditions. The study highlights a number of complexities with regard to CMIP5 model representation of water fluxes. The large intermodel range of CMIP5 future projections of fluxes calls for caution when using individual model results for assessing ongoing and future water and nutrient changes.