Decadal changes of flood probabilities

Recently a number of large flood events have caused significant damages in Germany and Austria raising the question if they have increased in number and magnitude. Research on this topic is so far not conclusive. Part of the problem is that floods tend to occur clustered in time resulting in flood rich and flood poor time periods which makes it more difficult to identify changes. Potential drivers of these changes are river regulation, changes in climate and changes in land use but it is not clear, how, how much and at which spatial scale they affect floods. The aim of this project is to understand whether there have been changes in flood occurrence and what the drivers for these changes are. We are planning to identify the decadal variability of river floods in Germany and Austria, changes in this variability, and to understand how these changes are linked to climatic factors and anthropogenic influences such as river works and land use changes. The detection of flood changes is usually performed by trend analysis of observed flood peaks. In this project instead of analysing trends we will focus on the decadal variability of floods by identifying flood rich and flood poor space-time regions and changes in their occurrence. The advantage of this approach is that it can capture long term cycles of flood occurrence. To improve the analysis, we will include historical information which expands the data records beyond the observational period of the systematic data. In the regions where flood changes are identified, we will attribute these changes to their drivers by comparing them with climate and land use data. We will develop a new regional attribution method that includes the information from many stations jointly to reduce the uncertainty due to data limitations. The research planned in the project breaks new scientific ground in at least three ways: (a) For the first time flood changes in Germany and Austria will be analysed in a consistent way which allows for a coherent understanding of flood probability changes in both countries.; (b) The analysis will focus on the decadal variability of floods instead of trends which allows for the detection of long term cycles and changes in these cycles; (c) All analysis will be performed in space and time with novel methods in order to identify flood rich and flood poor space-time regions. The analysis of many stations in a region will allow for a more informed detection than an analysis of individual stations. The project results are anticipated to improve the understanding of flood changes and their drivers, and will ultimately lead to improved flood estimation methods in both gauged and ungauged basins.

Recently there have been a number of major floods in Europe and it seems as if they had increased in number and magnitude. These events suggest that we may have entered a flood rich period, i.e. a period of time when floods are more frequent and larger than usual. This project investigated whether the probabilities of the occurrence of such extreme floods have indeed increased in recent years compared to the past, and what the drivers of such processes could be. The overall approach of this project consisted of analysing flood changes across a large number of European catchments (including more than 600 catchments in Austria and Germany) in a consistent way to identify flood rich and flood poor space-time regions of above and below average flood occurrence. A new method from Scan statistics was developed and spatio-temporal maps with flood-rich and flood-poor periods were obtained for the past five decades. A clear change in patters is apparent in the last two decades, where most of the detected anomalies in northwestern Europe were flood rich, while almost all the detected anomalies in the east and south were flood poor. The analysis was extended to historical flood information to gain a long-term perspective on flood changes. A new database composed of 103 sub-annual historical flood series was compiled based on documentary evidence, covering all major regions of Europe for the past 500 years. The documentary evidence on individual floods was transformed into a three-scaled intensity index and spatially and temporally interpolated. Nine flood-rich periods and associated regions were identified. Results show that the past three decades were among the most flood-rich periods in Europe in the past 500 years. A novel data-based approach for the attribution of flood changes was developed to link changes in flood occurrence to climatic and anthropogenic drivers. The approach was applied to 96 catchments in Upper Austria, where decadal variations in extreme precipitation resulted to be more strongly associated with long-term fluctuations of the flood frequency curves than other climatic and anthropogenic drivers. The overall findings confirm the project hypothesis that we have entered a flood-rich period with flood occurrence larger than usual. The current period also differs from other flood-rich periods detected in the past half millennium in terms of average temperature and seasonality.