Search KARSTBASE:
Search in KarstBase
It has been recently acknowledged that hypogenic caves are common in limestone terranes (e.g. KLIMCHOUK, 2000; AUDRA et al., 2002, 2007; AULER AND SMART, 2003; FORD AND WILLIAMS, 2007), with an extensive review by KLIMCHOUK (2007). Anticlinal ridges provide large recharge areas through which meteoric water may flow into confined zones around the peripheries during their history of uplift and associated denudation. The spatially varying stratal dips may create preferred flow routes within the confined zone and consequently promote hypogene speleogenesis at the most suitable sites for the water to rise again and discharge. Active speleogenetic sites thus may be found around the edges of anticlinal ridges where the potentiometric levels in the con?ned zone are high enough to promote the rising, transverse ?ow. Further away towards the adjoining synclinal basin, impermeable cover may be too thick to allow rising flow. The preferred sites for speleogenesis may migrate away from the anticlinal axis during the uplift process and associated lowering of groundwater levels. The common occurence of relict isolated hypogene caves in the Judean anticlinorium (FRUMKIN AND FISCHHENDLER, 2005) have led to the discovery of similar caves actively forming today. The Yarkon-Taninim regional aquifer is divided into lower and upper sub-aquifers, of which the lower one becomes (partly) con?ned near the anticlinal axis, while the upper sub- aquifer becomes con?ned at the western foothills. Upward flow is evident at the Ayalon Salinity Anomaly (ASA) where the upper sub-aquifer is still uncon?ned, so that rising water has much larger free space to ?ll in comparison with the nearby confined zone (FRUMKIN AND GVIRTZMAN, 2006). Approaching the watertable, the emerging rising flow can easily travel laterally along the highly permeable karstified zone. The rising ASA water is comparable to artesian springs, which discharge in the zone of lowest head of the upper aquifer. In the case of the ASA, however, the upward ?ow does not reach the open land surface but instead disperses laterally near the watertable. It may thus be considered an “underground delta”. The conceptual model consists of four-segment flow route: (1) rainwater recharge through outcrops on the anticlinal ridge; (2) lateral confined flow down to a depth of ~-700 m; (3) pressurized upward flow through discrete sub-vertical conduits; and (4) multidirectional pervasive flow close to the water table, with restricted output in which the rising water mingles with the ‘normal’ water of the upper aquifer. Maze caves fed by vertical conduits are typical for such an “underground delta”, as they disperse the flow laterally in many similar routes. Dense cave formation is observed to be associated with the upward flow of aggressive water. Within the “underground delta” the aggressiveness is consumed over short distances laterally away from the sub-vertical feeders. Such formation of large voids by dissolution far from the recharge zone implies renewed hydrochemical aggressiveness. The spatial location of the ASA is determined by three conditions that allow upward leakage from the deep sub-aquifer: (1) the location of the westernmost unconfined zone of the upper sub-aquifer, and its association with nearby confined regions; (2) the large upward head gradient; (3) spatial heterogeneities in the vertical permeability that are associated with tectonically disturbed zones.
By definition, karstic flow systems are networks of solutional conduits. Their spatial patterns and hierarchical organisation are strongly affected by differing lithology and geologic structure, and by the location and modes of recharge – unconfined, confined, interformational. For purposes of discussion, this paper will review six examples rang-ing across platform and reefal limestones and dolostones, dolostone breccias, gypsum and salt, in widely differing structural, geomorphic and hydrologic settings: (1) The Carcajou River karst at Lat. 65° N in the Mackenzie Mountains, where leaky permafrost superimposes a frozen ground hierarchy on those due to lithology, structure and topog-raphy: (2) The S Nahanni River karst at Lat. 62° N, with an intrusive-derived local thermal system and lengthy, strike-oriented meteoric flow systems that contribute to an outlet H2S thermal system at the basin topographic low: (3) Castleguard Mountain Karst (Lat. 52° N) in massive Main Ranges structures of the Rocky Mountains, with a complex alpine hierarchy of base-flow and overflow springs: (4) Crowsnest Pass, in steep thrust structures in the Rocky Mountain Front Ranges, where regional strike-oriented flow systems extending between Lats. 49° and 50° N and paired above and below a major aquitard have been disaggregated by glacial cirque incision: (5) The Black Hills geologic dome at Lat. 44° N in South Dakota, USA, with a sequence of hot springs at low points around the perimeter, discharging through sandstones but with some of the world’s most extensive hypogene maze caves formed in a limestone karst barré setting behind them: (6) The Sierra de El Abra, at Lat. 23° N in Mexico, a deep and lengthy (100 km) reef-backreef limestone range being progressively exposed and karstified by stripping of a cover of clastic rocks; the springs are few but amongst the largest known in karst anywhere, located at the northern and southern low extremities along the strike of the reef, plus breaches (windows) in the cover further south.
Initiation and development of karstification requires a continuous flushing of pore water in equilibrium with carbonate minerals. Under confined flow conditions, the energy required for pore water transport is supplied by external pressure sources in addition to the by earth’s gravity. Earth tides and water loads over the confined flow system are the main sources of external pressure that drives the pore water. Earth tides, created by the sum of the horizontal components of tide generation forces of moon and sun, causes expansion and contraction of the crust in horizontal direction. Water load on top of the confined flow system causes vertical loading/unloading and may be in the form of recharge load or ocean loading in the inland and sub-oceanic settings, respectively. Increasing and decreasing tide generating force results in pore water transport in the confined system by means of contraction and expansion, respectively. Since these forces operate in perpendicular directions, pore water flushing by earth tides becomes less effective when water load on top of the confined flow system increases. Temporal variation of freshwater content in a submarine cave is presented as an example of groundwater discharge driven by earth tides and recharge load.
Numerical models of speleogenesis typically simulate flow and dissolution within single fractures or networks of fractures. Such models employ fracture flow and pipe flow equations to determine flow rates and only consider average velocities within each fracture segment. Such approximations make large scale simulations of speleogenesis tractable. However, they do not allow simulation of the formation and evolution of micro- or meso-scale cave passage morphologies. Such morphologies are frequently studied within a field setting and utilized for the interpretation of the speleogenetic processes that formed the cave. One classic example is the formation of scallops in cave streams with turbulent flow. Scallops are used to interpret past flow velocities and directions. However, a recent analysis of the theory of limestone dissolution in turbulent flow conditions suggests a discrepancy between theory and reality concerning the formation of limestone scallops (Covington, in review). Similarly, the only attempt to numerically simulate flute formation in limestone found that the flute forms were not stable (Hammer et al., 2011). Motivated by these puzzles, we are developing a computational fluid dynamics (CFD) framework for the simulation of the evolution of dissolution morphologies.
While this project was initially conceived to better understand dissolution in turbulent flow, the tools being developed are particularly well-suited to examine a variety of other questions related to cave morphology on the micro- and meso-scales. There has been significant recent discussion about the interpretation of features that are diagnostic of hypogenic or transverse speleogenesis, such as the morphological suite of rising flow defined by Klimchouk (2007). Other authors have suggested that such forms can be found in a variety of settings where confined flow is not present (Mylroie and Mylroie, 2009; Palmer, 2011). We propose that simulation of such forms using a CFD speleogenesis code will allow a more complete understanding of the connections between process and form, because in such simulations the processes occurring are well-known, well-defined, and also can be adjusted within controlled numerical experiments, where relevant parameters and boundary conditions are systematically varied.
The CFD framework we are developing is based on the Lattice Boltzman method (Chen and Doolen, 1998), which is a popular technique for modeling the mechanics of complex fluids, including fluid mixtures, reactive transport, porous media flow, and complex and evolving domain geometries. With this framework it is straightforward to simulate many of the processes occurring in hypogene settings, including complex fluid flows, dissolution, solute and heat transport, and buoyancy-driven flow. Furthermore, this modeling framework allows these processes to be coupled so that their interactions and feedbacks can be explored. With the suite of capabilities provided by this framework, we can begin to numerically simulate the processes occurring in hypogene speleogenesis, including the driving mechanisms and the role of buoyancy-driven flow and its relationship with the morphological suite of rising flow. In the spirit of a workshop, this work is presented as in-progress, in the hopes that it will stimulate discussion on potential applications of the model being developed.