A Geologic Odyssey: Abstract Resulting from this Odyssey

Love, D.W. Whitworth, T.M. 2000, Draining Shallow Groundwater from Uplifted Fault Blocks of Shallow Aquifers to Form Distinctive Geomorphic Krenegenic Features: Field Evidence and Preliminary Hydrogeologic Modeling

Numerous normal faults form scarps that cut fluvial pebbly sands beneath the early Pleistocene Sunport geomorphic surface south of Albuquerque, New Mexico. Distinctive calcium-carbonate cementation features developed adjacent to and down-slope from several of these fault scarps. These features include calcified root mats with horizontal and vertical tubules and nodules, horizontal and vertical carbonate-cemented massive nodules and tubules, and massive carbonate-cemented sandstone beds with sharp bases. These cemented zones are localized in the hanging-wall blocks adjacent to the faults. The foot-wall sediments, fault, and sediments below the sharp base of these cemented zones are not cemented. Cementation extends a few meters to tens of meters horizontally away from the fault scarps and appears to be cuspate along the scarps.

Other scarps appear to be backworn by arcuate, dendritic headwalls possibly related to spring sapping. These features suggest that the faults were active when groundwater was near the surface and seeped from the uplifted fault blocks to the foot-wall blocks. The carbonate-cementation and geomorphic features are distinct from stage II-III pedogenic calcium-carbonate horizons that later developed across the scarps.

The sediments within the fault blocks consist of coarse, cross-bedded pebbly channel sands and floodplain sands, silts, and clays deposited by a broad (10 km) anastomosing Rio Grande fluvial system along the longitudinal axis of the Albuquerque basin until about 1 million years ago. The water table remained shallow as the system aggraded. Shortly before or while this broad plain was being abandoned by renewed incision of the Rio Grande, faults cut the aggradational surface and disrupted the shallow water table.

We modeled the fault-related high-water-table seeps by assuming flow perpendicular to the fault from the uplifted footwall to the hanging wall. Flow rates were estimated using known permeability and porosity of the fluvial sands, permeability of fault gouge of variable thicknesses, and variable hydrostatic heads from the footwall to the hanging wall. The water was assumed to be saturated with calcium carbonate, which was then progressively precipitated in the shallow subsurface of the hanging wall by evapotranspiration and loss of CO₂.

The distance from the fault to a steady-state no-flow boundary depends on the head, evapotranspiration rate, and the thickness of the fault gouge. We do not know actual evapotranspiration rates for the time of cementation so we assumed an initial "ballpark" rate of 40 inches per year (101.6 cm/yr).

The model suggests that the maximum distance that cementation could extend from the fault increases linearly from zero at the water table on the up-thrown block to some maximum distance for a given head difference across the fault. For example, using evaportranspration as 101.6 cm/year and a fault gouge thickness of 1 cm, the model predicts that for a head of 3.5 m the maximum distance that cemementation is likely to extend from the fault is 34.8 m. This distance is farther for thinner fault gouge and shorter for thicker fault gouge. For example, for a head of 3.5 m and the same parameters, a 0.5 cm-thick gouge would allow cementation for a maximum distance of 69.6 m, and a 2-cm-thick fault gouge would allow cementation for a distance of only 17.4 cm. These estimated distances are highly dependent on the hydraulic conductivity of the fault gouge. The number we used in the model is from work by Hong (1999) where actual fault gouge permeabilities were measured at faults similar to the ones examined in this study. If the hydraulic conductivity of the fault gouge is greater than we assumed, then the cementation could have a greater extent than predicted by the model. Conversely, if the fault gouge is less permeable, then the distance of cementation should be less for a given thickness of fault gouge.

We also constructed a model (using the same parameters) to estimate how long the cementation process could continue before all of the available porosity was plugged. This model simply calculates, based on water saturated with calcium carbonate and a 30% porosity, how long it will take to pass enough water through the gouge that contains enough calcium carbonate to completely fill the pores in the predicted cementation zone. For an evapotranspiration rate of 101.6 cm/yr, a fault gouge of 1.0 cm, the amount of time estimated to fill the pores of a pebbly sand with 30 percent porosity is about 206,000 years. Because the flow rate is faster with greater head, the time works out to be approximately the same for proportionate distances of the cementation zone for a given hydraulic conductivity. The model suggests that the time required to plug the porosity in the cemented zone is inversely proportional to fault gouge thickness. Thick fault gouge exhibits a lower flow rate for a given head. Thus, the distance to the no-flow boundary is less and there is less total volume to plug in the cemented zone and the smaller cemented zone will plug faster. Greater flow through thinner gouge generates a much larger cementation zone with a much larger pore volume. Hence, longer times are required to completely cement the pores.

CO₂ degassing probably also plays a role in calcite cement formation in these zones. However, CO₂ degassing was not considered in the model, because soil air commonly contains much higher levels of CO₂ with depth as a result of respiration of soil organisms and plant roots (Koorevaar et al.,1983; Treadwell-Steitz and McFadden, 2000). Thus CO₂ degassing is probably not as important as evapotranspiration.

Stratigraphic sections within one hanging wall exhibiting a large amount of offset contain at least three 2-m intervals of cemented calcium carbonate and calcified root-mat horizons. While these horizons have paleohydrologic implications, it remains unclear as to whether these are due to flow across the fault from the adjacent uplifted block or from some other direction.