A flood monitoring system comprises an extensive network of water sensors, a bundle of forecast simulations models, and a decision-support information system. A cascade of uncertainties present in each part of the system affects a reliable flood alert and response. The timeliness and quality of data gathering, used subsequently in forecasting models, is affected by the pervasive nature of the monitoring network where aquatic sensors are vulnerable to external disturbances affecting the accuracy of data acquisition. Existing solutions for aquatic monitoring are composed by heterogeneous sensors usually unable to ensure reliable measurements in complex scenarios, due to specific effects of each technology as transitional loss of availability, errors, limits of coverage, etc. In this paper, we introduce a more general study of all aspects of the criticality of sensor networks in the aquatic monitoring process, and we motivate for the need of reliable data collection in harsh coastal and marine environments. It is presented an overview of the main challenges such as the sensors power life, sensor hardware compatibility, reliability and long-range communication. These issues need to be addressed to improve the robustness of the sensors measurements. The development of solutions to automatically adjust the sensors measurements to each disturbance accordingly would provide an important increase on the quality of the measurements, thus supplying other parts of a flood monitoring system with dependable monitoring data. Also, with the purpose of providing software solutions to hardware failures, we introduce context-awareness techniques such as data processing, filtering and sensor fusion methods that were applied to a real working monitoring network with several proprietary probes (measuring conductivity, temperature, depth and various water quality parameters) in distant sites in Portugal. The goal is to assess the best technique to overcome each detected faulty measurement without compromising the time frame of the monitoring process.
Jesus, Goncalo Joao Vitorino; Oliveira, Anabela; and Casimiro, Antonio, "Ensuring Reliable Measurements In Remote Aquatic Sensor Networks" (2014). CUNY Academic Works.