Hydrologic Network Fault Trees Help Understand Patterns of Water Contamination

Abstract:

Surface waters are used for recreation, food supply, habitat, drinking water supply, and a variety of other ecological services that can be interrupted by water contamination. The stochastic nature of environmental systems makes the evaluation of reliability of these services necessary; however this is a major challenge due to system complexity and tool availability. We address this issue through adoption of fault-tree risk diagrams that have been used in Civil Engineering to conceptualize, analyze, and visualize complex and interconnected system behavior. A fault tree risk diagram is able to represent the connective geometry of the system, and to identify its possible failure modes. Environmental systems, like their engineered counterparts, are complex, interconnected, and have multiple system failure modes which include unsafe levels of contaminants in surface water. We have developed a Monte-Carlo procedure to obtain a fault tree risk diagram of the stream river network, and to perform system reliability evaluation. This study aims to allow for a more holistic watershed management by incorporating risk concepts with the geometric connectivity of a stream network. It aims to answer questions like “what are the areas in a watershed that increase the likelihood of overall water contamination?”, “what is the spatial and temporal distribution of probability of exceeding contaminant standards in the entire watershed?”, “which combination of individual sources will increase this probability?”, and “which areas of the watershed would be most sensitive to implementation of management measures”.