Final Report Summary - FLOODCHANGE (Deciphering River Flood Change)
The main objectives of this project were to understand how changes in land use and climate translate into changes in river floods, what are the factors controlling this relationship and what are the uncertainties involved.
We compiled a flood data base of more than 5000 river gauging stations in Europe, complemented by existing data bases. The analyses suggest that floods did change in the past five decades in terms of their timing within the year (seasonality) and in terms of their magnitudes. We used seasonality as a new indicator for identifying climate-related processes associated with changing floods. Warmer temperatures have led to earlier spring snowmelt floods throughout North-Eastern Europe; delayed winter storms associated with polar warming have led to later winter floods around the North Sea and some sectors of the Mediterranean Coast; and earlier soil moisture maxima have led to earlier winter floods in Western Europe. The results highlight the existence of a clear climate signal in flood observations at the continental scale. Flood magnitudes changed in line with changes in their seasonality, with a tendency for increasing floods in north-western Europe and more complex patterns in the rest of Europe. We extended the analyses back to the 16th century by examining records from historical archives and other historical evidence, and found that the existence of alternating flood poor and flood rich periods related to climate variability has been a common phenomenon over centuries.
We developed a new method for attributing observed flood changes to their drivers. The novel method evaluates flood changes of numerous catchments in a region jointly as a function of catchment scale. This regional method is able to more clearly identify flood changes and attribute them to their drivers than the traditional station-by-station methods. We found that land use tends to be an important flood-change driver at small catchment scales, while river works (in particular river training and loss of flood plain retention) are an important driver at large catchment scales. Climate is an important driver at all scales, particularly at intermediate catchment scales of a few hundred or a few thousand square kilometres. The sensitivity of flood changes to changes in their drivers not only depends on the catchment scale but also on other factors. Water storage thresholds in the soil affect the sensitivity of floods to changes in rainfall; the sensitivity is largest for rainstorms of magnitudes similar to those of the soil storage capacity. We also found that small floods are more sensitive to river training and the construction of levees than large floods.
Based on coupled human-water, or socio-hydrological, models, we explored potential long-term interactions between flood-related decisions (land use planning, construction of flood protective measures), flood magnitudes (changes in floods due to river training, climate variability and climate change, and land use change), and their uncertainties, resulting in diverse patterns of economic growth. An important finding is that alternating flood poor and flood rich periods may lead to a loss of societal memory of flood awareness, which may increase flood risk beyond what one would expect if floods occurred randomly. The project also assessed the implications of the project results for flood risk assessment and management. In view of the complex, non-linear process interactions, which are difficult to quantify, resilience based flood management measures that work well for a wide variety of different situations have clear merits.
Overall, the FloodChange project systematised the regional effects of land use and climate on floods for the first time, which provides a vital step towards predicting how floods will change in the future.
We compiled a flood data base of more than 5000 river gauging stations in Europe, complemented by existing data bases. The analyses suggest that floods did change in the past five decades in terms of their timing within the year (seasonality) and in terms of their magnitudes. We used seasonality as a new indicator for identifying climate-related processes associated with changing floods. Warmer temperatures have led to earlier spring snowmelt floods throughout North-Eastern Europe; delayed winter storms associated with polar warming have led to later winter floods around the North Sea and some sectors of the Mediterranean Coast; and earlier soil moisture maxima have led to earlier winter floods in Western Europe. The results highlight the existence of a clear climate signal in flood observations at the continental scale. Flood magnitudes changed in line with changes in their seasonality, with a tendency for increasing floods in north-western Europe and more complex patterns in the rest of Europe. We extended the analyses back to the 16th century by examining records from historical archives and other historical evidence, and found that the existence of alternating flood poor and flood rich periods related to climate variability has been a common phenomenon over centuries.
We developed a new method for attributing observed flood changes to their drivers. The novel method evaluates flood changes of numerous catchments in a region jointly as a function of catchment scale. This regional method is able to more clearly identify flood changes and attribute them to their drivers than the traditional station-by-station methods. We found that land use tends to be an important flood-change driver at small catchment scales, while river works (in particular river training and loss of flood plain retention) are an important driver at large catchment scales. Climate is an important driver at all scales, particularly at intermediate catchment scales of a few hundred or a few thousand square kilometres. The sensitivity of flood changes to changes in their drivers not only depends on the catchment scale but also on other factors. Water storage thresholds in the soil affect the sensitivity of floods to changes in rainfall; the sensitivity is largest for rainstorms of magnitudes similar to those of the soil storage capacity. We also found that small floods are more sensitive to river training and the construction of levees than large floods.
Based on coupled human-water, or socio-hydrological, models, we explored potential long-term interactions between flood-related decisions (land use planning, construction of flood protective measures), flood magnitudes (changes in floods due to river training, climate variability and climate change, and land use change), and their uncertainties, resulting in diverse patterns of economic growth. An important finding is that alternating flood poor and flood rich periods may lead to a loss of societal memory of flood awareness, which may increase flood risk beyond what one would expect if floods occurred randomly. The project also assessed the implications of the project results for flood risk assessment and management. In view of the complex, non-linear process interactions, which are difficult to quantify, resilience based flood management measures that work well for a wide variety of different situations have clear merits.
Overall, the FloodChange project systematised the regional effects of land use and climate on floods for the first time, which provides a vital step towards predicting how floods will change in the future.