Stream temperature and discharge evolution in Switzerland over the last 50 years: annual and seasonal behaviour

Abstract. Stream temperature and discharge are key hydrological variables for ecosystem and water resource management and are particularly sensitive to climate warming. Despite the wealth of meteorological and hydrological data, few studies have quantified observed stream temperature trends in the Alps. This study presents a detailed analysis of stream temperature and discharge in 52 catchments in Switzerland, a country covering a wide range of alpine and lowland hydrological regimes. The influence of discharge, precipitation, air temperature, and upstream lakes on stream temperatures and their temporal trends is analysed from multi-decadal to seasonal timescales. Stream temperature has significantly increased over the past 5 decades, with positive trends for all four seasons. The mean trends for the last 20 years are +0.37±0.11 ∘C per decade for water temperature, resulting from the joint effects of trends in air temperature (+0.39±0.14 ∘C per decade), discharge (-10.1±4.6 % per decade), and precipitation (-9.3±3.4 % per decade). For a longer time period (1979–2018), the trends are +0.33±0.03 ∘C per decade for water temperature, +0.46±0.03°C per decade for air temperature, -3.0±0.5 % per decade for discharge, and -1.3±0.5 % per decade for precipitation. Furthermore, we show that snow and glacier melt compensates for air temperature warming trends in a transient way in alpine streams. Lakes, on the contrary, have a strengthening effect on downstream water temperature trends at all elevations. Moreover, the identified stream temperature trends are shown to have critical impacts on ecological and economical temperature thresholds (the spread of fish diseases and the usage of water for industrial cooling), especially in lowland rivers, suggesting that these waterways are becoming more vulnerable to the increasing air temperature forcing. Resilient alpine rivers are expected to become more vulnerable to warming in the near future due to the expected reductions in snow- and glacier-melt inputs. A detailed mathematical framework along with the necessary source code are provided with this paper.