Looking at trends in high flows at a local scale: The case study of Wallonia (Belgium)

Abstract

Study Region: 84 catchments across the Walloon region of Belgium Study Focus: This study aims at analysing trends in high flows by examining annual maxima (AM), peaks over threshold (POTs) and the number of peaks per year (frequency). Trends were identified using statistical tests (regression analysis, Mann-Kendall and Pettitt tests). New Hydrological Insights for the Region: Almost 12 % of the sites show a trend in the magnitude of AM and frequency, and 6% show a trend in the magnitude of POTs. Globally, more negative trends have been detected, but the proportion of positive trends is higher in the Scheldt catchment than in the Meuse catchment. The results of nonstationary analysis indicate important changes in the magnitude of the 100-year flood (up to 18 % increase/11 % decrease in 10 years) and the frequency of peak flows (up to 42 % increase/31 % decrease). These changes could therefore impact future flood risk management in Wallonia. However, the time-series are short (30–50 years) and some uncertainty remains. Understanding the mechanisms responsible for the trends is essential to obtain better estimates of future flood flows. A first analysis of potential drivers reveals that changes in precipitation match the trends in high flows, and lower snowfall quantities and higher evapotranspiration rate, caused by the increase in temperature, could have contributed to the decrease in high flows in some regions.

Keywords

• Flow annual maxima
• Peaks over threshold
• Trend detection
• Nonstationary flood frequency analysis
• Climate change
• Wallonia