James C. Currens

Peak flow rate and recession-curve characteristics of a karst spring in the Inner Bluegrass, central Kentucky

Abstract The flow rate at the terminal spring of a 1929 ha karst ground-water catchment has been continuously monitored for 2 years, and 108 identifiable events were analyzed. The peak flow rates followed a beta frequency distribution with parameters α = 0.365 and γ = 1.135. Events were separated into high-flow and low-flow. High-flow events had characteristics attributable to pipe flow. Correlation and stepwise regression were used to develop peak flow rate prediction equations for the combined 108 events and for the 81 low-flow events. The portion of the recession curve identified as pipe flow was a watershed constant and time invariant. The base flow was seasonal, increasing in the winter to approximately 0.071 m3s−1 and decreasing in the summer to approximately 0.014 m3s−1.

Improved Karst Sinkhole Mapping in Kentucky Using Lidar Techniques: A Pilot Study in Floyds

The existing sinkhole database for Kentucky is based on low-resolution topographic maps created more than fifty years ago. LiDAR (Light Detection and Ranging) is a relatively recent technique that rapidly and accurately measures features on earths surface in high-resolution. To test the feasibility of using LiDAR to map sinkholes in Kentucky, we have developed a method of processing LiDAR data to identify sinkholes and tested the method in portions of the Floyds Fork watershed in central Kentucky. The method consisted of four steps, creating a high-resolution digital elevation model (DEM) from LiDAR data, extracting surface depression features from the DEM, inspecting the depression features for probable sinkholes, and verifying the probable sinkholes in the field. A total of 1,683 probable sinkholes were identified in the study area, compared to 383 previously mapped for the same area. We field-checked 121 randomly-selected probable sinkholes and confirmed that 106 of them were karst sinkholes. This method increased the number of sinkholes by a factor of four with a success rate between 80% and 93% for the study area, demonstrating that the LiDAR sinkhole-mapping method is reliable and efficient. This method identified approximately 55% of the previously mapped sinkholes, and approximately 98% of the missed sinkholes appeared to be filled or covered for urban development and agriculture purposes. The next step is to extend this method to provide high-resolution sinkhole maps for other karst areas in Kentucky where LiDAR data become available.

Flooding of Sinking Creek, Garretts Spring karst drainage basin, Jessamine and Woodford counties, Kentucky, USA

Tashamingo Subdivision in Sinking Creek karst valley, a tributary of the Garretts Spring drainage basin in Jessamine and Woodford counties, Kentucky, was flooded in February 1989. To determine the cause of flooding, the groundwater basin boundary was mapped, discharge data were measured to determine intake capacity of swallets, and hydrologic modeling of the basin was conducted. Swallet capacity was determined to be limited by the hydraulic parameters of the conduit, rather than by obstruction by trash. Flooding from a precipitation event is more likely, and will be higher, when antecedent soil moisture conditions in the watershed are near saturation. Hydrologic modeling shows that suburban development of 20 percent of the southeast basin will cause a small increase in flood stage at Tashamingo Subdivision.

A method to determine cover-collapse frequency in the Western Pennyroyal karst of Kentucky

To determine the rate of cover-collapse sinkhole formation in Christian County, Kentucky, we used large scale aerial photographs taken nearly twenty years apart. The negatives were enlarged and printed to 1:3,000 scale and examined for collapses. The photographs constrained the time period within which the collapse could have occurred, and the large scale of the prints provided a means to identify, locate, and field-verify the cover collapses. All features noted on the photographs were checked in the field. Sinkholes seen on the later photographs, but not the earlier ones, were recorded. The rate of formation calculated was 0.2 cover-collapse km yr. INTRODUCTION Cover collapse is the phenomena of apparently sudden collapse of soil or other unconsolidated cover over karstic bedrock. In Kentucky, cover collapse frequently damages buildings, roads, utility lines, and farm equipment. It has killed livestock, including some thoroughbred horses, and has injured people. The Kentucky Geological Survey estimates a total economic cost of

Model ordinance for development on karst terrain: Kentucky, USA

20 million annually in Kentucky from karst-generated geologic hazards (Dinger et al., 2007). The survey records an average of two dozen cover collapses per year and has developed a case history file spanning some thirty years. In this paper, we report a site-specific study of collapse frequency in a small area of the Western Pennyroyal sinkhole plain east of Hopkinsville in Christian County, Kentucky (Fig. 1). COVER-COLLAPSE PROCESS The development of voids in unconsolidated cover overlying karstic bedrock has been studied for decades (Beck, 1991; White and White, 1992). Small voids in soil at depths of a few meters are comparatively stable because of the lateral distribution of the overburden-induced stress by the arched roof of the void. The voids are enlarged by a loss of cohesion and loading of the arch-forming material caused by either a wetting front of soil water from infiltrating precipitation or by rapid draining of an inundated void. The saturated pores in the unconsolidated cover cannot drain as quickly as the conduit-connected void. The wetting and increased pore pressure result in an incremental loss of strength of the regolith arch (Tharp, 1999) and the underside of the arch sloughing into the soil void. Ultimately, the repeated sloughing from wetting and drying of the unconsolidated cover propagates the archedover void to near the land surface (Hyatt and Jacobs, 1996; Waltham et al., 2005). The sudden appearance of a covercollapse sinkhole is initiated when the arch becomes too thin to support its own weight and shears the remaining soil in a nearly circular pattern (Fig. 2). If sufficient volume is not available in the underlying bedrock cavity to store the collapsed soil, the loose material is transported away by groundwater flow through the bedrock conduit. Although the genesis of cover collapse is well understood, precisely predicting the time and place at which a collapse will occur is not yet possible (Hyatt et al., 2001). STUDY AREA The study area is 4.04 km in east-central Christian County, Kentucky (Fig. 3). The topography within the study area is karst plain and a single low hill, giving 23.5 m of local relief, formed by resistance to dissolution of the basal part of the Bethel Sandstone. Land use at the time of the study (2004) was largely pasture and row-cropped fields with scattered farmsteads, a retail agriculture supply store, a cement plant, and a restaurant. The boundaries were defined by the overlapping area of stereo aerial photograph pairs. The study area was selected without any prior knowledge of existing cover collapses in the area. The exposed Mississippian section, in ascending order, is Ste. Genevieve Limestone, Renault Limestone, and Bethel Sandstone (Klemic, 1967). The bedrock at the base of the stratigraphic column is predominantly oobiosparites and micritic limestones, mediumto thick-bedded, and is greater than 95 percent calcium carbonate. Interbedded thin shale and argillaceous carbonates are a minor interruption to the otherwise very pure carbonate section. The residual 10 m of Bethel is a calcite-cemented, argillaceous quartzarenite that weathers into a friable, porous, sandy residuum that readily slumps into underlying sinkholes (Klemic, 1967). The Lost River Chert is exposed near the base of the Ste. Genevieve in local quarries, but is below the depth of karst development in the study area. The exposed Ste. Genevieve Limestone is * Corresponding Author: currens@uky.edu J.C. Currens, R.L. Paylor, E.G. Beck, and B. Davidson – A method to determine cover-collapse frequency in the Western Pennyroyal karst of Kentucky. Journal of Cave and Karst Studies, v. 74, no. 3, p. 292–299. DOI: 10.4311/2011ES0247 292 N Journal of Cave and Karst Studies, December 2012 over 52 m thick, while the Renault Limestone is some 15 to 29 m thick. The regional gentle dip of 3 m km to the north is the only structure in the otherwise flat-lying bedrock at the study site. The cover collapses inventoried were mostly in the outcrop area of the Renault. Because of the purity and thickness of the carbonates, the presence of topographically mapped dolines, and the moderate total relief of 60 m within the study area, we expected the rate of occurrence of cover collapse to be comparatively high. The conditions in the study area are nearly ideal for cover-collapse development. METHODS We used a simple and inexpensive method to locate sinkholes and constrain the time of cover collapse. Because we did not have access to a magnifying stereoscope, the Kentucky Geological Survey purchased prints of blackand-white, low-altitude, large-scale, visible-light, aerial photography at an imaged scale of 1:12,000 (1 cm 5 120 m; 1 in. 5 1,000 ft) from the Tennessee Valley Authority. The photographs were taken March 9, 1971, and January 31, 1991. Although we also obtained stereo sets of contact prints, the most useful images were enlargements of the central image from the sets. The enlargements were printed at a scale of 1:3,000 (1 cm 5 30 m; 1 in 5 250 ft), four times the scale of the negative. Using 2-power magnifying glasses, we visually scanned the enlargements systematically for features appearing to be sinkholes. The emulsion grains on the print were sufficiently small in comparison with the typical cover-collapse that shadows cast on the interior of a collapse less than a meter in diameter could be discerned from those cast by small cedar trees, for example. Further, labeling devices could be attached to the enlargement print to preserve the interpretive data. We also searched for cover-collapse features on the stereo-pair contact prints and the digital images made at the KGS from scans of the enlargements. Most of the features identified on the 1:3,000-scale enlargements could not be relocated with confidence on the 1:12,000-scale contact prints. Scanned enlargements were saved as 16-bit gray-scale TIFF files at 1,200 dpi, giving a pixel size of roughly 0.5 m at ground level. The file size was large (428,853 kb) and prevented viewing the scan with image viewers on most of the available computers. We did view the images with GIS software, but could not relocate any cover-collapse locations on the digital images due to pixilation. Fifty potential cover-collapse sites were selected for field inspection from the enlarged prints (Table 1). KGS staff field-checked all of the sites and determined if they were, in fact, cover collapses. Those sites that had developed within Figure 2. A classic example of a cover collapse in the study area. Figure 1. The black polygon is the study area in Christian County, east of Hopkinsville, Western Pennyroyal region in Kentucky. The gray shaded area is underlain by karstic carbonates. J.C. CURRENS, R.L. PAYLOR, E.G. BECK, AND B. DAVIDSON Journal of Cave and Karst Studies, December 2012 N 293 the twenty-year period bracketed by the photographs were identified. We also found a small number of cover collapses that did not appear until after the 1991 photography. These were too recent to include in the calculation of the rate, but were documented for future reference. Some features that were visible on the earlier photographs, but not on the later ones, were also noted. We also received a limited number of reports that a collapse had occurred and had been filled and graded during the period between the aerial photography. Such cover collapses were included in the rate calculation only if we found field evidence that the report was correct. Field evidence for a filled sinkhole included a circular variation in texture and color of vegetation, buried trash exposed at the surface, subsidence due to soil compaction, or Figure 3. The cover-collapse inventory area in the Hopkinsville 7.5-minute quadrangle is inside the black line. The disrupted textured area enclosed by a gray line is the property of a limestone quarry, which was excluded from the inventory. Solid asterisks are cover-collapse sinkholes identified on the 1991 aerial photograph and verified in the field. Hollow asterisks are other features from the same images determined not to be cover-collapse. A METHOD TO DETERMINE COVER-COLLAPSE FREQUENCY IN THE WESTERN PENNYROYAL KARST OF KENTUCKY 294 N Journal of Cave and Karst Studies, December 2012 T a b le 1 . F ea tu re s id en ti fi ed o n a er ia l p h o to g ra p h s a n d fi el d ve ri fi ed a s p o ss ib le co ve r co ll a p se s si n ce 1 9 7 1 .

A sampling plan for conduit-flow karst springs: Minimizing sampling cost and maximizing statistical utility

Until the recent economic slowdown, development in rural areas was increasing at a rapid rate in Kentucky and resulting in development pressure on less desirable lands, including large areas of karst land. Most damage caused by karst geologic hazards is from groundwater pollution, cover-collapse sinkholes, and sinkhole flooding. The economic costs of karst-related geologic hazards have been estimated to be

Hydrogeologic Investigations of Pavement Subsidence in the Cumberland Gap Tunnel

20 million annually in Kentucky. The Kentucky Geological Survey drafted a model ordinance to guide development in karst for the purpose of reducing economic and environmental losses from karst geohazards. The text of the ordinance provides a wide range of ideas for planning and zoning agencies to adopt as needed. The final draft of the model ordinance was completed in the fall of 2008 and went through a thorough technical and editorial review, both internally at KGS and among other karst researchers. An effort was made to find an attorney to volunteer time to review the document, but we were not successful. The decision was made to move forward and send copies to 70 fiscal courts or planning and zoning boards. It was reasoned that any adopted ordinance derived from this document would be reviewed by their attorneys, regardless of previous reviews. To date the response has been poor. Only two agencies have contacted KGS for further information.

Challenges of using electrical resistivity method to locate karst conduits—A field case in the Inner Bluegrass Region, Kentucky

Analytical data for nitrate and triazines from 566 samples collected over a 3-year period at Pleasant Grove Spring, Logan County, KY, were statistically analyzed to determine the minimum data set needed to calculate meaningful yearly averages for a conduit-flow karst spring. Results indicate that a biweekly sampling schedule augmented with bihourly samples from high-flow events will provide meaningful suspended-constituent and dissolved-constituent statistics. Unless collected over an extensive period of time, daily samples may not be representative and may also be autocorrelated. All high-flow events resulting in a significant deflection of a constituent from base-line concentrations should be sampled. Either the geometric mean or the flow-weighted average of the suspended constituents should be used. If automatic samplers are used, then they may be programmed to collect storm samples as frequently as every few minutes to provide details on the arrival time of constituents of interest. However, only samples collected bihourly should be used to calculate averages. By adopting a biweekly sampling schedule augmented with high-flow samples, the need to continuously monitor discharge, or to search for and analyze existing data to develop a statistically valid monitoring plan, is lessened.

Changes in groundwater quality in a conduit-flow-dominated karst aquifer, following BMP implementation

Cumberland Gap Tunnel was constructed under Cumberland Gap National Historical Park in 1996 to improve transportation on a segment of U.S. 25E, connecting Kentucky and Tennessee and restoring Cumberland Gap to its historical appearance. The concrete pavement in the tunnel started to subside in 2001. Ground penetrating radar surveys revealed voids in many areas of the limestone roadbed aggregate beneath the pavement. To investigate possible hydrogeologic processes that may have caused favorable conditions for voids to form in the aggregate, we studied geology, groundwater flow, and groundwater chemistry in the tunnel using a variety of methods, including bore drilling, packer test, dye tracing, groundwaterand surface-flow monitoring, waterchemistry modeling, and an aggregate dissolution experiment. The study revealed that the aggregate receives a large volume of groundwater from much of the bedrock invert, but the flow velocity is too slow to transport small particles out of the aggregate. Calcite saturation indices calculated from water-chemistry data suggest that the groundwater was capable of continuously dissolving calcite, the primary mineral in the limestone aggregate. Water samples taken during different flow conditions indicate that groundwater under high-flow conditions could dissolve calcite more quickly than groundwater under low-flow conditions. The dissolution experiment showed that all the limestone aggregate placed beneath the roadbed and in contact with groundwater lost mass; the highest mass loss was 3.4 percent during a 178-day period. The experiment also suggested that water with higher calcite-dissolving potential removed limestone mass quicker than water with low calcite-dissolving potential. We recommend that the limestone aggregate be replaced with noncarbonate aggregate, such as granite, to prevent dissolution and future road subsidence. Introduction On October 18, 1996, a segment of U.S. 25E from Middlesboro, Ky., to Harrogate, Tenn., was relocated into a newly constructed tunnel beneath Cumberland Mountain, to improve transportation efficiency and safety as well as help restore Cumberland Gap to its appearance when Daniel Boone brought the first settlers to Kentucky in the mid1Kentucky Geological Survey, University of Kentucky 2Kentucky Transportation Center, University of Kentucky