Does internal climate variability overwhelm climate change signals in streamflow? The upper Po and Rhone basin case studies

摘要: Projections of climate change effects in streamflow are increasingly required to plan water management strategies. These projections however largely uncertain due the spread among model realizations, internal variability, and difficulties transferring results at spatial temporal scales by catchment hydrology. A combination a stochastic downscaling methodology distributed hydrological modeling was used ACQWA project provide future (up year 2050) for upper Po Rhone basins, respectively located northern Italy south-western Switzerland. Results suggest that (stochastic) variability is fundamental source uncertainty, typically comparable or larger than projected signal. Therefore, mean, frequency, seasonality can be masked natural climatic fluctuations large parts analyzed regions. An exception overwhelming role represented high elevation catchments fed glaciers where expected considerably reduced glacier retreat, with consequences appreciable main downstream rivers August September. Simulations also identify regions (west Toce, Ticino river basins) strong precipitation increase February April period projects beyond range during melting season. This study emphasizes importance including analyses, especially when compared limited uncertainty would accounted few deterministic projections. The presented could useful guiding more specific impact studies, although design decisions should better based on reliability vulnerability criteria as suggested recent literature.