Detailed documentations and assessments of the consequences of floods as well as flood risk analyses are essential for improvements of the flood risk management. However, flood documentation and flood damage collections are difficult and expensive, especially for frequent, small flood events. This results in a lack of data for flood management and research. Therefore, the Internet shall be used for the collection of flood damage information.

With the spread of so-called Web 2.0 applications, today the technical basis is given to implement applications based on the paradigm "Humans as sensors". For the purpose of flood damage data collection, such a web application has been developed and deployed online. Users affected by flood events in Germany have the opportunity to provide their individual observations, enriched with local knowledge and contextual information that technical sensors are not aware of. This approach therefore enables the collection of data that would be hidden or not communicated without the communication medium based on the technical infrastructure. A major challenge hereby is the fact that users of such systems are generally no experts in assessing damages. A user-friendly data collection form has therefore been set up online, guiding the web users through the process of the interview. Multimedia features for better understanding the terms and classifications of floods and their impacts, dynamic guiding to relevant questions within the interview as well as direct feedback after data insertion make the use of the form more comfortable than telephone or printout based interviews. Additionally, the web-based approach ensures minimum input data errors due to multiple choices from lists, automatic syntax checks, but also more complex consistency checks between different questions. Immediate feedback is provided in form of localized damage visualization in a map. Further research work is currently being undertaken to optimize quality checks for this data collection form, making use of several reference datasets and sophisticated check models. The anonymity of the data is ensured by keeping personal and address secret, while data analyses are carried out on an aggregated basis only. For the future, the concept of integrating user-generated content into scientific analyses will be further developed and assessed.

If you would like to tell us about your flood experience, please follow the  link.