Flash Flood Station Design with Data Quality assessment for SmartCity Applications

Flooding is an inevitable phenomenon that might cause massive loss of people's lives and destruction of infrastructure. In urban areas, the presence of buildings, roads, parking lots or other urban structures may alter natural drainage patterns, reduce infiltration, and increase surface runoff. Therefore, the introduction of a high number of sensors may provide distributed runoff models to capture details at very fine scale to warn of an imminent flash flood. This work presents the design of a water level measurement station using low-cost measurement techniques, incorporating data quality assessment prior to data transmission, and allowing its implementation in urban areas. The implemented station and the sensing methodologies are described in detail. The station approach compares two types of sensors to provide a validated water level measurement. The correlation between readings from both kind of sensors strength the quality of the obtained value because both sensing methodologies are based on different physics statements. In addition, a history dataset is maintained to evaluate the behavior of the measured parameter before its transmission to the cloud services. Therefore, the final transmit data is sent to the cloud with a quality assessment mark extracted from historical behavior. The energy consumption and the resolution of the water level sensor values are reported to demonstrate the feasibility of this implementation.