资源环境科技发展态势分析平台

Often, karstic conduit network geometry is unknown. This lack of knowledge represents a significant limitation when modeling flow and solute transport in karst systems. In this study, we apply Morris Method Global Sensitivity Analysis to a speleogenesis model to identify model input parameters, and combinations thereof, that most significantly influence evolution of karst conduit networks, development of first-magnitude springs, and resulting flow and solute transport pulse responses. Based on an idealized model of the Silver Springshed in Central Florida USA, results showed that porous matrix hydraulic conductivity and parameters that govern connectivity of vertical and horizontal preferential flow paths (proto-conduits) are the most influential parameters. In particular, a lower porous matrix conductivity is more likely to produce a first-order magnitude spring. For the boundary conditions assumed in this application, conduits tend to develop in low topographic regions that drained nearby high regions. Morris ensemble realizations that generated first-magnitude springs exhibit similar flow and solute transport pulse responses at the spring vent, despite differences in network configuration. However, distributed head fields are highly spatially variable, implying substantial spatial variability among solute flow paths and travel times from the land surface to the spring across realizations.

Plain Language Summary Karst aquifers are often vulnerable to contamination. Fast flow paths exist throughout the aquifer that can transport contaminants long distances over short-time spans. These fast flow paths are unknown but very important to protecting the aquifer. This study examines how these fast paths develop into springs, and whether we can still gather important information about how they influence the aquifer in spite of their unknown location, distribution, geometry, and size. Incorporating the fast flow paths into models used to make management decisions is not common practice. We found that the permeability of the aquifer and connectivity of the flow paths is important. Also despite the lack of knowledge of the preferential flow paths the variation between models of the same system with different flow path configurations was highly similar. This implies that incorporating the flow paths may be more important than knowing exactly where they are.