A complex network theoretic approach for interdependence investigation: An application to radionuclide behavior in the subsurface

摘要

The failure of uranium mine tailings dams results in the infiltration and spreading of tailings in the subsurface. The fate and transport of radionuclides in the subsurface depends on several confounding, complex interdependent factors that describe the elements of the integrated system (i.e., meteorological; hydrological; hydrogeological; and, soil, groundwater, and mine tailings chemistry). The factors describing the integrated system have typically been investigated independently; however, their interdependence and resulting collective influence on the subsurface migration of radionuclides are yet to be explored. The current study develops a complex network theoretical approach to analyze these interdependencies. In this respect, a network of factors (NoF) was developed, and its characteristics (e.g., diameter, density, characteristic pathlength, average clustering coefficient, and factor centrality measures) were evaluated to determine the importance of considering these interdependencies when developing radionuclide fate and transport models. A sensitivity analysis was subsequently performed on the NoF to characterize the propagation of uncertainty associated with the factors in the NoF through a fate and transport model. The sensitivity analysis indicated that microorganisms present in the soil and mine tailings, fraction of organic carbon in the soil matrix, infiltration, and transmissivity must be well characterized (i.e., to minimize their uncertainty) when developing an integrated subsurface radionuclide fate and transport model, as uncertainty in these parameters will be amplified in the model output. The NoF developed in this study can be used to allocate data collection resources strategically in order to minimizing uncertainty in fate and transport models. This improves the reliability of fate and transport models, and ultimately leads to better management and remediation strategies to mitigate impacts from UMTD failures.