Analysis of spatiotemporal patterns of snow cover and snowmelt floods in Austria using remote sensing data

This study investigates the temporal dynamics of snow cover and its impact on flood events using MODIS snow cover data from Terra and Aqua satellites. The analysis has been conducted across the Danube River basin and Austria with a comprehensive dataset of snow cover and river discharges spanning from 2001 to 2021. The MODIS sensors offer spatial resolutions of 500 meters and a daily temporal resolution, crucial for capturing snow cover variations. To address cloud obscuration, a specific threshold for cloud coverage has been chosen and merging techniques utilized to increase the accuracy of the snow cover data. Our assessment reveals a general decreasing trend in snow cover across the Danube sub-catchments, with varying rates of decrease among different basins and elevation ranges. In particular, catchments with average high snow cover exhibited a steeper decline compared to those with lower snow presence. However, catchments at altitudes between 1800 and 2100 meters showed a more stable snow cover over time compared to those at lower elevations, such as 900 to 1200 meters. For the Austrian study area, the analysis involved also 581 river measurement stations grouped across five different regions based on specific hydrogeological and climatic characteristics. Focusing on winter and spring peak discharge values, this research uncovered significant regional variations, with spring peak discharge showing decreasing trends in the Southern and Northern Alpine regions, as well as across Austria. In contrast, increasing trends were identified in both annual and winter peak discharge. The temporal evolution of snow cover in Austrian catchments was also analysed, revealing notable regional differences. The Southern Alpine region exhibited the smallest decline at -0.08, while the Northern Alpine region showed a more pronounced decrease of -0.36. The Mann-Kendall test was conducted across these regions to assess the significance of these monotonic trends, revealing significant results for the Northern Lowlands and Eastern Lowlands, but not for the previously mentioned Southern Alpine region. The study underscores that snowmelt dynamics, including variations in snow cover, play complex roles in flood generation. Analysing the correlations between spring peak discharge and snow-covered area values from the week preceding the peak across the five regions reveals that: in the Alpine region the correlation is negative, indicating an inverse relationship between snow/glacier melt and river discharge. This suggests that as the peak tends to increase, snow cover decreases. In contrast, the Southern Alpine region shows a slight positive correlation, which reflects the influence of Mediterranean storm systems interacting with snowmelt processes. The Northern Alpine and Northern Lowlands regions exhibit low correlations, suggesting that rainfall is a more dominant factor in influencing peak discharge than snow cover. Finally, the Eastern Lowlands display small positive correlations, which may be attributed to the effects of local convective storms and frontal systems on river discharge. The results underscore the intricate relationship between snow cover and flood peaks. Satellite imagery provided a robust framework for analysing snow cover trends. Future research incorporating stochastic modelling could further clarify how projected climate changes might alter snow dynamics, particularly through shifts in snowmelt timing and snow elevation line.