Abstract Randomness of fracture networks still makes channelized flow a challenge to track in hard-rock aquifers. While not underestimating geological and hydrological criteria that are also handled here through mapping exercises, this study raises an issue of water quality encountered in lifelong boreholes. Chemical classification checked against a recent conceptual model of bedrock aquifers gives birth to a new typology of groundwater in a complex granitic aquifer system located in the SW of Ivory Coast (West Africa). Major ion chemistry, borehole completion data, digital elevation model and satellite images are used to interpret the geochemical water facies as an expression of connexions between the saprolite and the saprock, or transient insulation. From major ions ratios, cumulate mineralization, carbonate equilibrium, stable isotopes, the maturation of ground waters and mixing between bedrock layers are described at seasonal and local scales. The results highlight some vertical feeding of the water table into the main saprock aquifer owing to shortcuts through the saprolite, along with the existence of dead-ends in the hydraulically active fracture network. Also, some influence of fault zones, either drain or barrier, is confirmed on the (Ca, Mg) bicarbonate water facies within the saprock.

ABSTRACT The Great Lakes coast contains numerous unstable bluffs underlain by heterogeneous glacial materials consisting of till, sand, and gravel layers, and lacustrine clays. Many of the bluffs are steeper than their equilibrium angles and typically move as slow earth slides or occasional rapid slumps. Such movements develop largely where interlayered sand and clay contain perched groundwater that acts to reduce effective stress during winter months when perched potentiometric surface elevations rise because water cannot discharge through frozen soil. Aerial photograph records dating back to 1938 show that bluffs recede in amphitheater-like depressions followed by "catch up" where headlands between amphitheaters are attacked by other forms of erosion. This bluff recession is particularly pronounced during stages of high lake levels. The erosion control experiment described herein has been designed to determine the manner in which groundwater activity influences the causes and mechanisms of mass wasting on the Great Lakes coasts. Three dewatering demonstration sites were selected, monitored electronically for virtually all movement and cause relationships, and dewatered to demonstrate a potential mitigation strategy other than construction of wave barriers. Erosion activity and dewatering effects were carefully monitored for three seasonal cycles. Results show that (1) dewatering greatly reduces ground displacements during winter months, and (2) bluff movements are almost perfectly timed to, or lag slightly after, the hours when potentiometric surfaces near the bluff face reach their highest elevations during freezing (greatest soil pore pressure) or their greatest rates of surficial discharge (soon after thaw). This field guide project was supported by grants from the U.S. Army Research Office, Terrestrial Sciences Program (Grant 3467-GS) from 1996 to 1999 and the U.S. Army Engineer Research and Development Center (ERDC) from 2000 to 2007, and 2012, through U.S. Senate Bill 227 (National Shoreline Erosion Control Development and Demonstration Program), with support from Western Michigan University (WMU). Additional personnel involved were Alan E. Kehew, Co-PIand, WMU graduate students William Montgomery, Rennie Kaunda, Mark Worrall, Gregory Young, William Bush, and Amanda Brotz. Well and monitoring instrument positions were chosen by R. Chase and designed by Ronald L. Erickson and James P. Selegean, U.S. Army Engineer District, Detroit, Michigan. Well constructions and instrument installations were done by STS Consultants, Chicago, Illinois. This huge project was very smoothly administered by M. Eileen Glynn and William R. Curtis, ERDC, Vicksburg, Mississippi.

In 2002, Yucca Mountain, Nevada, was selected as the proposed site for the U.S. high-level nuclear waste repository. Yucca Mountain lies within a large topographically closed basin, in which surface water is internally drained. Groundwater, however, can and does flow into and out of this basin at depth through a regional carbonate-rock aquifer (commonly referred to as the lower carbonate-rock aquifer). Most groundwater recharge (water infiltrating downward through the unsaturated zone into the water table) originates in the highlands north of Yucca Mountain and flows generally southward. Some groundwater discharges within the basin, as in Oasis Valley and the southern Amargosa Desert, but the ultimate discharge is in Death Valley, where water is returned to the atmosphere by evapotranspiration. Groundwater flows through a heterogeneous medium produced by a complex geologic history including both compressional and extensional tectonics. For hydrologic purposes, the rocks and alluvium are divided into 25 hydrogeologic units. Regionally, the most important unit for regional groundwater flow is composed of Paleozoic carbonate rocks, which are locally separated into two aquifers by an intervening shale. Rocks of the southwestern Nevada volcanic field form thick deposits in the northern part of the basin, and these rocks host both aquifers and confining units. The potentiometric surface of the site-scale flow system contains areas of large hydraulic gradient (as great as 0.13) and small hydraulic gradient (as small as 0.0001). Both extremes are found within the Yucca Mountain site area, where they are well constrained by numerous boreholes. At Yucca Mountain, a single borehole penetrates to the regional carbonate-rock aquifer, and, at this locality, the hydraulic head at depth is 20 m greater than in the overlying volcanic rocks. This head difference is likely widespread, as indicated by thermal highs at the groundwater table in the vicinity of block-bounding faults, where upward leakage of water from the regional carbonate-rock aquifer is postulated. Since the early 1980s, numerous two- and three-dimensional flow models have been developed to depict regional groundwater flow. A 2004 transient flow model of the Death Valley region has 16 layers and a 1500 m/side horizontal grid; it is composed of 194 rows and 160 columns. The model was first calibrated to a steady-state condition and then to transient conditions. The model matches observed flow patterns well, and it generally agrees with measured water levels except in areas of large hydraulic gradient. The regional model provides the boundary conditions for a detailed site-scale flow model. The finite-element heat and mass transfer code, FEHM v2.24, was used to simulate flow through the saturated zone at Yucca Mountain. Cells in the site-scale model are 250 m/side in the horizontal grid; it is composed of 181 rows and 121 columns. The model may use as many as 67 layers, but the framework model allows a stair-stepped ground surface, so the number of layers is variable. Layer thickness ranges from 600 m at the bottom of the model to 10 m south of Yucca Mountain. The site-scale flow model was constructed and calibrated, matching observed hydrologic data well. The site-scale flow model provides a means for assessing the hypothetical flow path for any radioactive materials originating from the proposed repository.

That karst aquifers constitute a class with properties distinct from other aquifers has been supported by the work of many researchers. Nonetheless, assessments of aquifers in limestone and dolomite where contamination issues exist continue to be regularly mismanaged by practitioners whose training has focused on the properties of granular aquifers. Incorrect assumptions of pressure and pore-space continuity, erroneous assumptions of isotropy and homogeneity relating to permeability testing, and the inappropriate application of models and computer simulations based upon these assumptions are the principal causes of this mismanagement. The authority conferred by respected consulting firms and sophisticated computer simulations has led to these fundamentally flawed studies being utilized in situations involving contaminants where the public health and welfare must be appropriately protected. Examples of the misuse of assumptions and authority relating to karst aquifers are discussed here.