Groundwater sampling in karst terranes: passive sampling in comparison to event-driven sampling strategy.

Karst aquifers are very easily contaminated because of the surficial features that commonly exist in karst terranes. Pollutant releases into sinkholes, sinking streams, and/or losing streams commonly result in concentrated solutes rapidly infiltrating and migrating through the subsurface to eventually discharge at downgradient springs unless intercepted by production wells, but slow percolation through soils also may result in serious contamination of karst aquifers. The unique features of karst terranes tend to cause significant problems in the interpretation of results obtained from water-quality grab samples of karst groundwater. To obtain more representative samples, event-driven sampling was proposed some decades ago, but event-driven sampling can be difficult and expensive to implement. In this paper, application of passive-sampling strategies is advocated as a means for effectively obtaining representative water-quality samples from karst aquifers. A passive-sampling methodology may be particularly useful for karst aquifers that may be found in complexly folded and faulted terranes. For example, a groundwater tracing investigation of a contaminated site in a karst terrane confirmed that several offsite springs and wells are connected to the contaminated site. Tracer recoveries suggested transport rates that were relatively slow for flow in a karstic aquifer (~0.02 m/s). Breakthrough curves were erratic and spiky. To obtain representative groundwater samples, a passive-sampling methodology is recommended.