|Wednesday, September 16th|
Kim Tondeur, Vrije Universiteit Brussel
8:40 AM - 9:00 AM
FloodCitiSense.eu aims at developing an early pluvial flood warning service for but also by citizens and city authorities. Despite rapid progress in forecasting, warning and management of urban pluvial floods, multiple drawbacks remain, including insufficient accuracy and resolution of rainfall estimates and forecast as well as a simulation time still too long in relation to the fast hydrological responses of urban settings. In this context, FloodCitiSense.eu takes on a high-resolution approach able to account for the complexity and heterogeneity of urban landscapes as well as for the characterization of rainfall intensity and distribution. To do so, we combine high-resolution elevation and land-cover data together with fine-resolution distributed rainfall data from both radar and station sources as well as crowdsourced and smart sensing data. Co-created, disseminated and tested together with stakeholders and citizens in an urban living lab context in each pilot city (Brussels, Rotterdam, Birmingham), web-based technologies and low-cost sensors for flood mapping and monitoring allow us to tackle the lack of a dense-monitoring network in our pilot cases. In this spirit, engagement strategies are used to keep citizen observatories up and living, while the use of more traditional radar and station data allows us to evaluate the quality and accuracy of data collected by social and smart sensing. In this contribution we focus on the problems and challenges encountered during the set-up of our low-cost rain monitoring networks, the quality of the data collected and its potential use in the real-time monitoring for an early warning service for urban pluvial floods. The FloodCitiSense.eu project is a close collaboration with TU Delft, Imperial College London, IIASA, Disdrometrics, VUB SMIT-imec, LGiU, EGEB and is funded within the ERA-NET Smart Urban Future programme of JPI Europe.
Raed Fehri, UCLouvain, Belgium, UCLouvain, Belgique
9:00 AM - 9:20 AM
In the last decade, new ways of environmental monitoring have been emerging with the development of smart technology. Among them, citizen science (CS) has been introduced as a novel framework to generate environmental knowledge. In many parts of the world, the lack of reliable river discharge data is a major constraint for operational water management, despite the governmental efforts to improve the existing monitoring systems. This constraint can partially be alleviated through CS and the use of cost-effective smartphone technology. In this study, we tested the publicly available smartphone application “Discharge” to monitor streamflow in Belgium and Tunisia using two different strategies: via an expert user; and by a group of everyday citizens. The use of two types of operators allows evaluating the performances of the app in terms of operator skills. We used a step-by-step CS approach to engage and train citizens on using the application at two locations in Tunisia. In Belgium, an expert user of the app performed measurements of a series of hydrological events at three sites. The collected daily data were compared with data from reference stations at every location. Results yield a significant correlation between CS data and the reference stations in Tunisia. Similarly, expert measurements in Belgium yield a strong agreement with the reference data. Linear regression analysis and errors assessment were used to validate the quality of CS and expert user data. It is concluded that the expert user delivered slightly higher quality of discharge than everyday citizens. This indicates the robustness of the measurement tool as well as the usefulness of the CS training program. We conclude that CS can be considered as a promising monitoring approach for obtaining reliable discharge data, for complementing existing discharge monitoring systems, and also for engaging local communities to innovate the water resources management process.
Leonardo Alfonso, IHE Delft, Netherlands
9:20 AM - 9:40 AM
Citizen Science (CS) projects structured for the purpose of data collection are proliferating in the last decade, with a common believe that it could be more cost effective than traditional environmental monitoring networks. However, there are few studies that confirm this claim. In this work, a methodology to evaluate the impact of data coming from CS projects, with respect to the existing in-situ network is developed. The methodology combines a cost-benefit approach with a complementarity approach with respect to existing monitoring networks, and it consists of two main parts, namely the data perspective and the costs perspective. The former aims to qualify the degree of complementarity that the data collected by citizens offers to in-situ networks in terms of space and time, based on the scales of the observed variables. The latter aims to qualify the relation between the investments required to set up a citizen observatory, including stakeholder engagement activities and training, and the actual amount of data collected. We introduce the notion of the Cost of Data Record (CDR), which is to be considered of maximum value if its complementarity is the maximum and if the cost to produce it is the minimum. On the contrary, CDR has little value if its complementarity is the minimum and its cost is high. Results show that building a CO for the only purpose of data collection is an expensive undertaking that do not necessarily complement the existing in-situ monitoring system, and that campaign-based approaches are more effective to this end.