You are using an old browser with security vulnerabilities and can not use the features of this website.
Although forests are very important ecosystems and essential for life on planet earth, this project does not
directly deal with forests. The project Q4 aims to analyze and quantify flood damage of companies i.a. by means of a data mining technique called “Random Forest”. Mainly survey data sets, acquired after flood events in Germany, are used for this analysis. These data sets consist of many variables which are expected to influence the flood damage. The “Random Forests” are ensembles of many decision trees which partition the survey data sets in such a way that the homogeneity of the single partitions is maximized with regard to a certain target variable, e.g. the flood damage. Resulting ensembles can then be used to estimate e.g. the flood damage to buildings of a company or to determine the importance of single variables for the flood damage estimation. Thus the project focuses on improvements in the model performance, the identification of damage driving factors and the consideration of uncertainties in flood damage modeling.
In the wake of changing hydro-climatological, geo-physical and socio-economic conditions the magnitude, frequency and impact of certain types of natural hazards are likely bound to change as well. This is highly of utmost importance for many regions in the world where risks due to natural hazards have to be managed and mitigated and this is where the research training group “Natural hazards and risks in a changing world (NatRiskChange)” aims to foster the scientific knowledge basis. This research training group started on October 1st 2015 and is funded by the Deutsche Forschungsgemeinschaft DFG. The central goal of NatRiskChange is to pursue the development of methods to improve hazard and risk analysis and quantification based on the transient, non-stationary nature of hazards and risks in response to changing natural and anthropogenically altered components of the Earth system. Key scientific aims are the development, testing, and pilot application of studies on identification, quantification (mechanisms) and prediction of transient natural hazards and associated risks.
Within NatRiskChange, a telephone aided survey was conducted in October and November 2017 among companies, which were affected from heavy rainfall or flash flood events in 2014 to 2016. We aim at gathering information about the companies experiences with severe weather warning systems, the type and extent of the damage as well es the state of recovery. Results shall identify improved mitigation measures for the management of eavy rainfall events. We thank all participants of the survey for their support!