Key points are not available for this paper at this time.
Increasingly frequent extreme hydro-meteorological events, attributable among others to non-stationary climate, can lead to devastating flooding and cause large costs to affected societies. A recent and prominently featured example includes the July 2021 flood in Germany with more than 140 casualties and billions of Euros in material damages. Such types of events could strike other parts of Europe at any time. Especially small-scale systems are likely to be affected more frequently by high intensity events. To ready society against such occurrences, preparedness needs to be increased form different viewpoints: 1) extreme value statistics and forecasting, 2) real-time data acquisition and high performance computing, 3) socio-technical translation into domains. First, fast and accurate forecasting with reduced uncertainties requires integration of predictors at different spatial and temporal scales. Such predictors must be conditioned by local, real-time information retrieved from multi-sensor systems, whereby special attention is devoted to extreme event statistics. The uncertainty of prediction resulting from sources like forecast model deficiencies, measurement errors, or various critical system states, need to be adequately represented, as well as de-biased and sharpened in real-time through statistical approaches. Second, real-time acquisition of local data plays an essential role in risk detection. To this end, novel sensor system integration must provide robust real-time information, either on the basis of flexibly positioned or body-mounted devices. The extended complexity of the methods involved make the efficiency of the computational methods and the integration of model-driven physical processes with data driven approaches and ubiquitous computing as key factor. This also concerns the reduction of computational complexity without compromising efficacy as well as the efficient interoperability between different system components or scales. Third, such novel instruments need to be socio-technically translated for decision-makers as well as emergency services and citizens. User involvement needs to be bidirectional, that is, stakeholders and their concerns must be understood and taken seriously for them to gain confidence and adopt innovative, multi-sensor integrated forecasting technologies for risk mitigation. Moreover, proper visual mapping of hazardous situations including uncertainties and potential options for decision-support need to be provided.
Reggiani et al. (Mon,) studied this question.