Processes leading to the re-activation of a sinkhole in buried karst and the subsequent drying of waterfalls in a small catchment located in northern Indiana, USA

Abstract

A sinking stream pirated water from a series of waterfalls in a small catchment underlain by buried karst in northern Indiana after a sinkhole was scoured and enlarged over the course of 2016. The catchment has historical significance to Indiana, and the current landowners wish to restore flow to the waterfalls. We address the following questions: 1) where does the water go once it enters the sinking stream, 2) what transport processes are active in the carbonate rocks, and 3) what processes led to the re-activation of a buried sinkhole? Breakthrough curves (BTCs) of salt-slug tracer tests show that 24.6% of the chloride mass was recovered at a conduit spring (Spring 1), located 150 m from the sinking stream, and 26.6% of chloride mass was recovered at two diffuse springs (Springs 2 and 3), located 315 m from the sinking stream. Despite the difference in distances from the sinking stream, the salt-slug appeared in all springs in approximately two hours. The BTC for Spring 1 was single-peaked, consistent with advection through a single conduit, while the BTCs of Springs 2 and 3 were double-peaked and consistent with advection along bifurcated flowpaths. Spring 2 exhibited a long tail consistent with anastomosing flowpaths and/or presence of pooling along the flowpaths. The sinking stream and other inactive sinkholes are upstream of a dam built in the 1960s. Ponding conditions likely increased the hydraulic head over the buried karst and began dislodging sediment from open pores/conduits in the karst. Record rainfall and widespread flooding during the summer of 2015 scoured the pores leading to the enlargement and reactivation of the sinking stream by the fall of 2016.