| dc.description.abstract | The water cycle and its extremes are heavily affected by human activities, such as reservoirs operations and water abstractions. The aim of this study was to evaluate impacts of human interventions (HI) on the water cycle in Europe, with a special focus on changes in high and low flows. Impacts of HI on runoff (based on water balance at the grid-cell scale) and discharge (i.e. routed runoff) were analysed using time series simulated with the global hydrological model PCR-GLOBWB 2 at 0.5° resolution for 1901--2001. Three model scenarios were run; no HI (i.e. natural conditions), transient (i.e. annually introduced) HI and current HI (i.e. HI per 2010). Four meteorological forcing data sets were used to assess the results’ sensitivity to the choice of forcing data. Simulated discharge was validated against 316 observed discharge time series, and adequate performance were found for all forcing data sets. Time series of regional median runoff percentiles were assessed to investigate impact of transient and current HI on flow percentiles and flow variability. Impacts of current HI on water gains and losses for the whole period were investigated by using the metrics ecosurplus (water gain) and ecodeficit (water loss) for the water cycle in general (0th–100th percentile) the low flow end (0th–10th percentile) and the high flow end (90th–100th percentile). Regional percentiles revealed only minor impacts due to HI, partly due to the regional aggregation of different types of HI. Ecosurplus and ecodeficit proved to be valuable metrics for detecting areas highly impacted by HI. For runoff, the largest impacts were water gains found in highly irrigated areas in southern Europe and north-west of the Black Sea. In these regions, water gains of more than 10% was found for the low flow end. For discharge, as much as one third of Europe had water losses of more than 10% for the low flow end, and one fifth of Europe had water gains of more than 10% for the low flow end due to current HI. Water losses in discharge reflect the impacts of water use accumulating downstream. Areas affected by upstream reservoirs had generally water gains in the low flow end and water losses in the high flow end, reflecting dam operations that retain water during high flows and release water during low flows. Deviating results for runoff and discharge were found, and are mainly caused by abstraction of surface water and introduction of reservoirs in the case of discharge. Such differences demonstrate the importance of distinguishing between runoff and discharge when investigating impacts of HI on the water cycle. All results were sensitive to the choice of forcing data, especially for high flows and in cold regions. The findings of this study revealed various impacts of HI across Europe, and how HI can affect the two flow extremes differently. | eng |