| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Earth and Environmental Sciences Division, Los Alamos National Laboratory, Los Alamos, NM 87545
* Corresponding author: (robinson{at}lanl.gov)
Received 29 November 2004.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONPERCHED WATER OBSERVATIONS...INTERPRETATION OF DEEP PERCHED...NUMERICAL MODELING OF DEEP...DISCUSSION AND CONCLUSIONSREFERENCES |
|---|
Abbreviations: LANL, Los Alamos National Laboratory
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONPERCHED WATER OBSERVATIONS...INTERPRETATION OF DEEP PERCHED...NUMERICAL MODELING OF DEEP...DISCUSSION AND CONCLUSIONSREFERENCES |
|---|
Beneath the Pajarito Plateau at Los Alamos National Laboratory (LANL), perched water may represent important subsurface pathways that facilitate movement of contaminated fluids from the ground surface to the water table of the regional aquifer. These perched groundwater bodies are generally too small for use as municipal water supplies. Nonetheless, they are of interest because (i) they represent natural groundwater resources that are protected under State law; (ii) their chemical and isotopic characteristics help constrain groundwater transport rates through the vadose zone; (iii) perching horizons may divert, slow, or stop the vertical migration of groundwater through the vadose zone, or they may indicate the presence of a fast subsurface pathway, depending on the characteristics of the perched zone; and (iv) they can be used as vadose zone monitoring points that provide early warning of contaminants approaching the regional aquifer.
The vadose zone beneath the Pajarito Plateau ranges in thickness from 180 m (590 ft) to more than 365 m (1200 ft), and it consists of Pliocene alluvial fan deposits covered by thick Pleistocene ash-flow tuffs. The Pliocene sediments interfinger laterally with lavas from the Jemez volcanic field to the west and the Cerros del Rio volcanic field to the east. An in-depth description of the geology of the Pajarito Plateau is presented by Broxton and Vaniman (2005), who explain and illustrate the stratigraphic units discussed here. Perched groundwater is found within canyon-bottom alluvium and within bedrock units of the vadose zone. Characterization of these groundwater bodies is challenging because of the thickness of the vadose zone in this area, the heterogeneous nature of bedrock geologic units that serve as host rocks and perching horizons, and the depths of groundwater occurrences. Despite these limitations, substantial new information has been gathered about deep perched zones beneath the plateau. The present study summarizes information about the location, depth to water, saturated thickness, and geologic setting of these perched water occurrences. This summary includes data from historical investigations and new information collected through March 2005 as part of recent site-wide hydrogeological investigations.
Assessment of the impact of perched water on subsurface flow velocities is one of the most difficult aspects of vadose zone characterization and modeling. A common assumption in vadose zone flow and transport modeling is that the stratigraphy exerts control over the hydrologic processes. If this is true, then the hydrologic properties at a numerical grid point can be populated based on the stratigraphic unit containing that point. Numerical simulation methods to reproduce perched water usually involve the assignment of low permeability to a portion of the numerical model domain, to force the water to accumulate above these locations. For example, this approach was used to match perched water observations in the unsaturated zone flow models of Yucca Mountain (Wu et al., 1999). We present an alternative method for simulating perched water in vadose zone models. The modeling approach, which applies permeability barriers at the interfaces between hydrogeologic units, is a practical modeling approach that is compatible with existing numerical model formulations.
The organization of this paper is as follows. First, we present field observations of a large dataset of perched water occurrences beneath the Pajarito Plateau. This portion of the paper describes the field methods, observations, and underlying hydrologic mechanisms found to control the nature and extent of perched groundwater in the vadose zone. We then develop a new computational technique, which we call the interface reduction factor method, for simulating perched water in vadose zone models. Finally, we present the results of a two-dimensional model of flow and transport in Los Alamos Canyon to illustrate the method. Additional aspects of the vadose zone modeling effort for this canyon are presented separately (Robinson et al., 2005).
|
PERCHED WATER OBSERVATIONS BENEATH THE PAJARITO PLATEAU |
|---|
|
TOPABSTRACTINTRODUCTIONPERCHED WATER OBSERVATIONS...INTERPRETATION OF DEEP PERCHED...NUMERICAL MODELING OF DEEP...DISCUSSION AND CONCLUSIONSREFERENCES |
|---|
|
Canyon-bottom alluvium is a complex geologic unit consisting of fluvial sands, gravels, and cobbles that interfinger laterally with colluvium derived from canyon walls. The fluvial deposits represent cross-cutting active and inactive channels and floodplain sediments. Canyons that head on the plateau contain detritus of Bandelier Tuff; canyons that head in the Sierra de los Valles also contain dacitic detritus derived from the Tschicoma Formation.
The thickness of alluvial deposits varies across the plateau. In Pueblo Canyon alluvium is about 3.5 m (11 ft) thick on the west side of the plateau and about 5.5 m (18 ft) thick near the confluence with Los Alamos Canyon. Mortandad Canyon has 0.3 to 0.6 m (1–2 ft) of alluvium near its headwaters and more than 30 m (100 ft) of alluvium near the eastern LANL boundary. Where present, alluvial groundwater commonly occurs near the base of the alluvial sequence atop weathered Bandelier Tuff. However, perched water is also found above permeability barriers within thick sequences of stratified alluvial deposits.
The saturated thickness and lateral extent of alluvial groundwater is strongly affected by seasonal variations in snow-melt and storm runoff. In Los Alamos Canyon, perennial alluvial groundwater extends from the Sierra de los Valles to about 1.6 km (1 mi) east of the Pajarito fault zone, even in drought years. During the peak snow-melt months of April through early June, surface flow and alluvial groundwater commonly extends an additional 6 to 10 km (4–6 mi) down canyon. Storm runoff, particularly that associated with monsoonal thunderstorms in July and August, is also an important source of recharge for the alluvial groundwater systems.
Methods for Identifying Deep Perched Groundwater
Identification of deep perched systems beneath the Pajarito Plateau comes mostly from direct observation of saturation in open boreholes (e.g., borehole video logs and water-level measurements), from borehole geophysics (e.g., electrical and neutron logs), or by the installation of wells or piezometers. Additional information is provided by surface-based electrical geophysics, although these types of investigations are generally limited by their relatively shallow depths of investigation and poor vertical resolution. Identification of larger perched groundwater bodies in boreholes is generally reliable, but the use of drilling fluids, which is necessary in most boreholes, may mask smaller or relatively unproductive zones. Defining the lateral extent of saturation is more problematic because of the costs associated with installing wells to such great depths. Despite these limitations, substantial new information has been gathered about deep perched zones on the plateau.
Field Observations
This section documents the observed occurrences of deep perched water and interprets, where possible, the cause of the perching. Table 1 lists 33 occurrences of deep perched groundwater detected in boreholes across the Pajarito Plateau. Perched groundwater is widely distributed across the northern and central part of the plateau (Fig. 2)
with depth to water ranging from 36 to 272 m (118–894 ft). The principal occurrences of perched groundwater occur beneath (i) the large, relatively wet Los Alamos Canyon and Pueblo Canyon watersheds; (ii) the smaller watersheds of Sandia and Mortandad Canyons that receive significant volumes of treated effluent from LANL operations; and (iii) the Cañon de Valle area in the southwestern part of LANL. Perched water is most often found in Puye fanglomerates, the Cerros del Rio basalts, and units of the Bandelier Tuff (Fig. 1). There are few reported occurrences of perched water in the southern part of LANL, but few deep boreholes are located there, and additional perched zones are likely beneath the large, wet watersheds of Pajarito and Water Canyons.
|
|
Units of the Bandelier Tuff, including the Guaje Pumice Bed, pinch out eastward beneath the floor of Los Alamos Canyon, and the perched zones to the east are found in stratigraphically lower geologic units. These eastern perched zones tend to become thicker and occur at multiple depths. For example, at Well R-9, three perched systems were encountered: (i) in the central part of a thick sequence of Cerros del Rio basalts, (ii) in the basal part of the Cerros del Rio basalts, and (iii) in clay-rich, pumiceous deposits in the lower part of the Puye Formation. Saturated thicknesses for the top and bottom zones range from about 13.7 to 31.4 m (45–103 ft), and the middle zone was 2.1 m (7 ft) thick. The top and middle perched zones at R-9, in basaltic lavas, are also present within similar lavas at Well LAWS-1, located 400 m (1300 ft) to the east (Fig. 2). The occurrence of thicker perched zones in the eastern part of Los Alamos Canyon may be due to enhanced infiltration where the canyon floor is underlain by Puye fanglomerate and Cerros del Rio basalts rather than by Bandelier Tuff. Tritium activities of 69 to 246 pCi L–1 for these perched groundwaters are elevated relative to the cosmogenic baseline of 1 pCi L–1, suggesting that these zones contain a component of young water that postdates the advent of atmospheric nuclear testing 60 yr ago (Longmire, 2002).
In Pueblo Canyon, perched water was identified in four wells. At Wells TW-2a and R-5, perched water occurs within fanglomerate of the Puye Formation and has a saturated thickness of >7 and about 11.3 m (>23 and 
37 ft), respectively. Depth to water is 33.5 m (110 ft) at TW-2a and about 115.8 m (380 ft) at R-5. These perched zones probably represent relatively small, unrelated water bodies, as suggested by their distance from one another (4 km [2.5 mi]), the lateral heterogeneity of Puye Formation deposits, and their varying depths beneath the canyon floor. Wells TW-1a and POI-4 encountered perched water at depths of 36 to 48.8 m (118–160 ft), respectively, in Cerros del Rio basalts. The saturated thickness is about 11 m (37 ft) at TW-1a and 6.4 m (>21 ft) at POI-4. Saturation is associated with interflow breccia and sediments at TW-1a and with fractured basalt at POI-4.
Sandia and Mortandad Canyons
In each of these canyons, deep perched water was found in Cerros del Rio basalts and the Puye Formation. The perched zones in both canyons contain a component of treated wastewater effluent released to the canyons via outfalls (Longmire et al., 2001; Broxton et al., 2001d). These outfalls are major sources of surface- and alluvial-water systems that contribute to the recharge of the deeper perched zones.
In Sandia Canyon, Well R-12 encountered a perched water zone from depths of 135 to 158 m (443–519 ft). Saturation occurs in the lower part of the Cerros del Rio basalts and extends into the underlying Puye Formation (Broxton et al., 2001a). The perched water in this zone may be confined: during drilling, the borehole was dry until a depth of 135 m (443 ft) was reached, but the water level quickly rose to a depth of 129 m (424 ft) once saturation was encountered. The apparent confining layer at the top of this zone is a massive basalt flow with few fractures. An alternative explanation for the observed rise in water level is that the groundwater is unconfined, but it could not be detected until water-bearing, interconnected fracture systems were intersected by the borehole at a depth of 135 m (443 ft). The perching layer at the base of this zone is a 5-m (16-ft)-thick lacustrine deposit consisting of clay and silt The saturated thickness of this groundwater body is at least 23 m (75 ft), making it one of the thickest perched groundwater bodies identified in the eastern part of the Pajarito Plateau.
In Mortandad Canyon, two groups of wells encountered perched water within the Puye Formation and the underlying Cerros del Rio basalt. In one set of wells, MCOBT-4.4 and I-4 (Fig. 2), a thin perched zone was encountered at a depth of about 160 m (520 ft), within pebble gravel and silty sands of the Puye Formation. The apparent perching horizon is the impermeable top of the Cerros del Rio basalt. The saturated thickness of this zone is approximately 1 m (3 ft). About 350 m (1150 ft) down canyon, a second group of wells, including R-15, I-5, and I-6 (Fig. 2), found perched water within the lower part of the Cerros del Rio basalt. The depth to water is 197 m (646 ft) at R-15, 209 m (687 ft) at I-5, and 202 m (662 ft) at I-6. Saturated thicknesses are 30 m (99 ft) at R-15, 6 m (20 ft) at I-5, and 13 m (43 ft) at I-6. The saturated intervals appear to be associated with fractured lava flows and interflow breccias. The variable static water levels and differences in saturated thicknesses illustrate the heterogeneous nature of perched groundwater systems located within basaltic rocks. At R-15, the perching horizon appears to be clay-rich, flow-base rubble or underlying silty basaltic sands. The perched zones at I-5 and I-6 were probably not fully penetrated during drilling, and the perching horizons are uncertain.
Because of their different geologic settings, the perched groundwater occurrences at MCOBT-4.4/I-4 and R-15/I-5/I-6 probably represent two unrelated groundwater bodies of limited lateral extent along the canyon axis. Several other deep boreholes in Mortandad Canyon area did not encounter perched groundwater. Based on the distribution of water-bearing vs. dry boreholes, the lateral extent of individual perched groundwater bodies is probably <450 m (1500 ft).
Perched water in Mortandad Canyon contains elevated tritium, nitrate, and perchlorate. The highest contaminant levels occur at MCOBT-4.4, which contains 14750 pCi L–1 tritium, 12.5 mg L–1 nitrate plus nitrite (as N), and 179 µg L–1 perchlorate (Longmire, 2002, personal communication). Beginning in 1963, these contaminants were released to the canyon as liquid effluent by a waste treatment facility in the upper part of the canyon. The presence of contaminants in deep perched groundwater beneath Mortandad Canyon indicates that vertical transport through the vadose zone occurs on the timescale of decades.
Cañon de Valle Area
A large area of complex perched groundwater occurrences is found in the region bounded by Cañon de Valle on the north and Water Canyon on the south in the southwest part of LANL. Five deep boreholes encountered significant zones of groundwater over a 2.6-km2 (1 mi2) area located just east of the Pajarito fault zone: R-25, R-26, CdV-16-1(i), CdV-16-2(i), and SHB-3 (Fig. 2). Depth to water in these perched zones ranges from about 183 m (600 ft) near the Pajarito fault to about 244 m (800 ft) 2.3 km (1.4 mi) east of the fault. Only Wells R-25 and R-26 fully penetrate the perched water zones.
At R-26, a water-level measurement of 184 m (604 ft) was obtained when the borehole was 219 m (720 ft) deep. The borehole was eventually completed to a total depth of 454.3 m (1490.5 ft) with the regional water table occurring at a depth of approximately 291 m (954 ft). Borehole neutron, magnetic resonance, and induction logs indicate that high moisture contents occur in rocks below 175 m (575 ft), with perched water most likely at depths of 177 to 202 m (580–662 ft) and 238 to 252 m (780–827 ft). These perched zones occur within stratified volcaniclastic sediments of the Cerro Toledo interval. Low-permeability sediments within the Cerro Toledo interval probably provide the perching horizons.
R-25, located 1524 m (5000 ft) east of R-26, has two distinct zones of saturation separated by 47 m (154 ft) of partially saturated rocks. The upper zone, which is interpreted as a perched zone, occurs between depths of about 217 to 345 m (711–1132 ft) within the Otowi Member of the Bandelier Tuff and in the upper part of the Puye Formation. The interval of partial saturation below the perched zone occurs from 345 to 392 m (1132–1286 ft) depth. Partial saturation was identified by isolating the perched zone behind drill casing and advancing the borehole through alternating zones of dry and wet rocks by coring and air-rotary methods. From 392 m (1286 ft) to the total depth of 592 m (1942 ft), continuous saturation representing regional groundwater was encountered within the Puye Formation. R-25 was constructed with nine screens separated by packers using a Westbay (Burnaby, BC, Canada) sampling system. Hydraulic head measurements in isolated screens decrease with depth, indicating a downward component of the hydraulic gradient. Isotopic and water quality data suggest the upper and lower zones of saturation at R-25 represent separate groundwater systems (Longmire, 2003, personal communication).
A deep-sounding surface-based magnetotelluric survey conducted in the Cañon de Valle/Water Canyon area suggests that perched groundwater is discontinuous laterally, occurring instead as vertical, finger-like groundwater bodies (LANL, 2003a). These geophysical interpretations are currently being tested by additional drilling.
Perched groundwater in the Cañon de Valle/Water Canyon area may be the result of focused recharge beneath wet canyons crossing the Pajarito fault zone. In this model, canyons act as line sources of recharge due to infiltration of surface and alluvial water. Welded tuffs in the vicinity of the fault are brittle and highly fractured. These fractures may provide pathways for deep infiltration in the near-surface environment. Elsewhere on the plateau, fractures in welded portions of the Tshirege Member tend to die out at depth in underlying nonwelded tuffs and poorly consolidated volcaniclastic deposits. In the Cañon de Valle/Water Canyon area, water infiltrating along deeply penetrating fractures would tend to spread laterally into permeable nonwelded tuffs and stratified volcaniclastic deposits, forming the perched zones (LANL, 1998, 2003a).
|
INTERPRETATION OF DEEP PERCHED WATER OBSERVATIONS |
|---|
|
TOPABSTRACTINTRODUCTIONPERCHED WATER OBSERVATIONS...INTERPRETATION OF DEEP PERCHED...NUMERICAL MODELING OF DEEP...DISCUSSION AND CONCLUSIONSREFERENCES |
|---|
Surface Hydrologic Conditions
Deep perched groundwater is most frequently found in the following settings: (i) the large, relatively wet Los Alamos Canyon and Pueblo Canyon watersheds; (ii) the smaller watersheds of Sandia Canyon and Mortandad Canyon that receive significant volumes of treated effluent from LANL and Los Alamos County operations; and (iii) in the mountain front Cañon de Valle area in the southwestern part of LANL. The apparent lack of deep perched groundwater in the southern part of LANL may reflect the presence of fewer deep characterization wells in that area. Perched water is likely beneath western parts of the large, wet Pajarito Canyon and Water Canyon watersheds, particularly along the mountain front. Therefore, a requirement for deep perched water to exist is a surface water source (natural or anthropogenic) that supplies water to perched alluvial systems that then act as the source for deep perched zones.
Geologic and Hydrostratigraphic Controls
Deep perched groundwater occurs most frequently in the Puye Formation and Cerros del Rio basalts, but some of the thickest and/or most laterally extensive zones involve units of the Bandelier Tuff. Perching horizons include a wide variety of layered geologic lithologies, including unfractured basalt flows, clay-rich interflow zones in basalts, buried soils and other fine-grained deposits in fanglomerate, clay-altered tuffaceous sediments, and lake deposits. Therefore, in addition to high local infiltration rates, low-permeability barriers to downward vertical flow appear to be required to induce deep perched groundwater. In contrast, no deep perched groundwater observations occur in situations that indicate a capillary barrier effect, despite the presence of layered stratigraphy with units of contrasting unsaturated flow properties.
An alternative hypothesis is that the deepest perched water occurrences are a manifestation of complex groundwater flow within the phreatic zone at the top of the regional aquifer. Localized heterogeneities, such as the clay-rich alteration zones in the Puye Formation at Well R-9, combined with high recharge, may give rise to a complex flow structure that includes mounding, interconnected saturated zones, and locally confined conditions. However, the complexity of the alteration and the depth of these groundwater zones makes detailed characterization prohibitively expensive.
Subsurface Extent of Zones of Saturation
Observed saturated thicknesses vary from 1 to 128 m (3–421 ft). The lateral extent of saturation in these zones is less well understood because costs associated with installing multiple wells to test lateral extent at such great depths are high. However, based on observations of both occurrences and nearby absences of deep perched groundwater in adjacent wells, we conclude that perched groundwater is far more likely to be present beneath wet canyons. The extent that perched groundwater flows along dipping geologic strata into areas beneath adjacent mesa areas is not fully known. However, paired canyon–mesa wells such as R-7 and 21-2523 in Los Alamos Canyon and R-22 and R-23 in Pajarito Canyon suggest that perched zones are much less common beneath the dry mesas, especially on the eastern side of the plateau.
Flow Conditions Upstream and Within Deep Perched Groundwater Zones
The presence of mobile (nonsorbing) anthropogenic chemicals in some deep perched groundwater indicates a connection with surface and alluvial groundwater. Based on the age of facilities that are potential sources of contaminants, the travel time of groundwater moving from the surface to deep perched groundwater systems is on the order of several decades. Within the perched zones themselves, the topography of the perching horizon, the bedding features, and the orientation of interconnected fracture systems probably control local groundwater flow velocity. However, direct evidence, such as single-well or multiple-well hydrologic and tracer testing, is not available. Therefore, the following discussion is based on reasonable hydrologic principles rather than direct measurements.
Flow conditions can, in principle, be categorized with the following two end-member conceptual models for flow within a deep perched water zone:
Low-velocity, virtually stagnant water resting within a local depression above a perching horizon.
Water percolates very slowly out the bottom of this zone, or spills over the sides of the depression. For this conceptual model, perching horizons are barriers that slow the downward percolation of water. In several wells, intermediate saturated zones thought to represent deep perched groundwater were screened but failed to produce significant water. These occurrences may represent situations in which perched systems of limited extent were drained when the perching horizon was penetrated during drilling. Once the stagnant water is depleted in an initial round of sampling, there is insufficient recharge upstream to keep the zone saturated.
High-velocity, laterally migrating fluid that travels on top of the perching horizon.
This conceptualization suggests that once groundwater reaches a deep perched zone, it rapidly percolates laterally along high-permeability pathways until the perching horizon pinches out or is breached by high-permeability features such as fractures or lateral changes in lithology. In this scenario, water could move in stair-step fashion from one perching horizon to another. With the possible exceptions of the Guaje Pumice Bed in Los Alamos Canyon and the mountain front system in the Cañon de Valle area, there are no confirmed instances of large-scale, lateral vadose zone pathways beneath the Pajarito Plateau at depths greater than the alluvial groundwater.
These two different conceptual models are not necessarily mutually exclusive. For instance, it is possible that well-sorted stream gravels within the Puye Formation could provide flow geometries similar to those of today's alluvial groundwater zones. Channel deposits within the Puye Formation could provide relatively high-velocity lateral movement in an otherwise stagnant system of poorly sorted fanglomerates. This and other complex scenarios are plausible but untestable with present data.
|
NUMERICAL MODELING OF DEEP PERCHED GROUNDWATER |
|---|
|
TOPABSTRACTINTRODUCTIONPERCHED WATER OBSERVATIONS...INTERPRETATION OF DEEP PERCHED...NUMERICAL MODELING OF DEEP...DISCUSSION AND CONCLUSIONSREFERENCES |
|---|
Without a means to change the model parameterization locally at interfaces such as this, it is difficult to reconcile the perched water with the currently measured hydrologic properties of the rock. For example, mean hydraulic conductivity values for the Bandelier Tuff are several orders of magnitude larger than the highest proposed infiltration rates for the plateau. In a model, if only mean values are used, the rocks will transmit all fluid under unsaturated conditions, and no perching or lateral diversion will occur. We present an alternative approach that considers perching horizons to be at the interfaces between hydrostratigraphic units.
Model Development
To illustrate the development of this modeling method, which we call the interface reduction factor method, consider the schematic shown in Fig. 3
. On the left-hand figure, two units of (possibly) different permeability are adjacent to one another. Considering for the moment the flow to be one-dimensional and steady state, the difference in flow potential P2 – P1 from point 2 to point 1 can be obtained by recognizing that the flux within each material is the same, so that:
 | [1] |
 | [2] |
 | [3] |
 | [4] |
|
 | [5] |
 | [6] |
Returning to the example of the Guaje–Puye interface, some reasonable parameter values illustrate the point. Using the mean values of permeability for these two units, the harmonic average permeability is 3 x 10–13 m2. When we include the clay layer (assumed thickness of 5 cm) of permeability ki of 1 x 10–17 m2, we obtain an effective permeability across the interface of 4 x 10–16 m2. Comparing this value to the harmonic average permeability, the permeability reduction factor for this example is 4 x 10–16/3 x 10–13 = 0.0013. Under these conditions, thin, low-permeability layers will exert a strong influence on the flow system.
The approach taken in this study considers the interface between hydrogeologic units to be a region where hydraulic properties such as the saturated permeability may be different than those in either of the units above or below the interface. The numerical implementation of this conceptual model has been added to the FEHM code (Zyvoloski et al., 1997). The new feature added to the code allows the user to specify a constant multiplier, called the permeability reduction factor, to any connection on an interface between two hydrostratigraphic units. In this way, the permeabilities within each unit are their original values, but the permeability applied for water passing through the interface is reduced. When the reduction factor makes the permeability at the interface small enough, lateral diversion or perching can occur, depending on the dip of the interface and the local infiltration rate.
This method, though conceptually simple, ignores the potential for capillary forces in the thin perching layer to impact the unsaturated flow in the vicinity of the layer. It is possible that such driving forces could give rise to an additional lateral flow component that is not captured in this simplified analysis. However, because the nature of the perching horizons being modeled is uncertain as to the properties, thickness, orientation, and continuity of the perching horizons, we felt that the examination of such effects was beyond the scope of this study. Future theoretical and modeling studies will be conducted in which the simplified approach developed here is compared with a detailed, fine-grid representation that includes the perching layer explicitly. In the modeling below, we restrict our attention to the Guaje–Puye interface and treat the permeability reduction factor as an uncertain parameter that is varied systematically to examine the impact on flow.
Numerical Model Results
A two-dimensional numerical model of vadose zone flow and transport in Los Alamos Canyon is used to demonstrate the permeability reduction factor method for simulating perched water. In this model, the numerical grid was developed along the canyon bottom based on the stratigraphy previously developed for this portion of the Pajarito Plateau (Carey et al., 1999). Although three-dimensional simulations are possible, it is a reasonable first approximation, based on the observations presented earlier, that flow within a perched water zone is concentrated along the canyon bottom, where infiltration rates are highest. The two-dimensional grid consists of 57004 nodes and 111256 triangular elements, and extends from the bedrock beneath alluvium to the regional aquifer water table. Details of the grid-building process are presented in Robinson et al. (2005). In that study, unsaturated hydrologic properties for each unit and infiltration rates used as the upper boundary condition are also presented.
In this study, the permeability reduction factor is varied in a sensitivity analysis to simulate perched water. The predicted steady-state water-content profiles in Wells LADP-3 and LAOI(A)-1.1 for the base case and three values of the permeability reduction factor are shown in Fig. 4 . Although there is no restriction to steady-state flow imposed by this method, in this study we focus on steady-state conditions for simplicity. For lower values of the reduction factor, higher saturations are predicted in the Guaje Pumice Bed because water is restricted from percolating into the underlying Puye Formation. The lower two reduction factors in Fig. 4 exhibit perching at these two wells. The base case and the 0.01 case do not, although the 0.01 case has a buildup of fluid saturation at the interface and small amounts of perching in other parts of the model not immediately adjacent to these two wells. These results are consistent with borehole geophysical data that show water content commonly increases downward in the Guaje Pumice Bed without reaching saturation, except as described above in Los Alamos Canyon. Water content abruptly decreases in the underlying Puye Formation.
|
|
|
In these simulations, the permeability reduction factor controls the buildup in saturation directly beneath the canyon at the Guaje Pumice Bed–Puye Formation interface. The behavior for lower values of the permeability reduction factor is consistent with the available observations. Solute transport velocity within the perched zone is a function of the degree of perching and the effective transport porosity of the overlying layer. Constraints on this parameter can be obtained through field or laboratory measurements. To replicate field observations, inverse modeling using the reduction factor as an adjustable parameter is a promising approach. Additional study is required to investigate the role of the unsaturated hydrologic properties of the confining layer, possible three-dimensional effects, and the influence of transient flow conditions on the buildup or drainage of perched water.
|
DISCUSSION AND CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONPERCHED WATER OBSERVATIONS...INTERPRETATION OF DEEP PERCHED...NUMERICAL MODELING OF DEEP...DISCUSSION AND CONCLUSIONSREFERENCES |
|---|
The principal occurrences of perched groundwater are in (i) the large, relatively wet Los Alamos Canyon and Pueblo Canyon watersheds; (ii) the smaller watersheds of Sandia Canyon and Mortandad Canyon that receive significant volumes of treated effluent from LANL operations; and (iii) in the Cañon de Valle area in the southwestern part of LANL. Perched groundwater is most often found in Puye fanglomerates and Cerros del Rio basalts, but it also occurs to a lesser extent in units of the Bandelier Tuff. Observations are consistent with a conceptual model of low-permeability horizons on which infiltrating water sits. Deep perched groundwater is most often found beneath wet canyons, suggesting that these zones are fed and maintained by surface water sources (natural or anthropogenic) supplying water to perched alluvial systems that act as sources for groundwater entering bedrock units at high infiltration rates.
This extensive set of observations in deep wells makes the Pajarito Plateau one of the most studied vadose zones among research efforts to gain understanding of the mechanisms giving rise to perched water and the hydraulics of perched water zones. Another well-studied vadose zone, Yucca Mountain, exhibits both similarities and differences compared with the Pajarito Plateau. In both sites, a basic requirement for perching is the presence of a low-permeaility zone. However, the Yucca Mountain perched water appears as a large, areally extensive water body that rests on top of a broadly distributed zone of low-permeability, zeolitic rocks. In contrast, the tuffs on the Pajarito Plateau generally lack zeolites and the perching horizons tend to be aerially limited permeability barriers associated with stratified sedimentary rocks and basalt, especially beneath canyon bottoms characterized by locally high infiltration rates. At this site, the perched water appears to exist as unconnected zones of water confined to the canyon in which the high infiltration occurs. A similarity between these two sites, and probably elsewhere, is that there is some ambiguity as to whether the perched water actually represents the top of the regional aquifer, especially when the perching horizon is at depths near the regional water table.
The fluid velocity in these perched zones is unknown due to the lack of direct measurements. Two end-member conceptual models are relatively stagnant fluid in a local subsurface depression above the perching horizon, or lateral diversion in the hydrologic unit overlying the perching horizon. Hydrologic testing, tracer tests, or groundwater dating methods would be required to shed light on this question.
Anthropogenic chemicals are present in many of the deep perched zones, indicating a connection to surface water. These observations allow us to estimate the maximum travel time from the surface to the perching horizon. Groundwater travel times from the surface to deep perched groundwater systems are on a decade time frame in several wet canyons, including Pueblo, Los Alamos, Sandia, and Mortandad Canyons and Cañon de Valle.
A key numerical modeling issue for the vadose zone at LANL and elsewhere is how to practically represent thin, low-permeability perching horizons. Although relatively thin layers can be incorporated directly into a finite element grid, there are practical limitations to this approach, including computational inefficiencies and the lack of detailed data that would be required to construct a high-resolution model (centimeter-scale grid spacing within the perching horizon). The method we developed considers the interface between two hydrogeologic units to be a region where the saturated permeability is different from the value in both the units above and below the interface. This new feature added to the FEHM modeling code allows the user to specify a constant multiplier, called the permeability reduction factor, that can be applied to any connective interface between two hydrostratigraphic units. The derivation provides the link between this reduction factor and physical properties such as the permeability and thickness of the perching horizon. When the reduction factor is small enough, perching and perhaps lateral diversion can occur, depending on the dip of the interface and the local percolation rate.
Two-dimensional, steady-state numerical simulations were performed for Los Alamos Canyon using the permeability reduction factor method. Perched groundwater at the interface between the Guaje Pumice Bed and the underlying Puye Formation was reproduced, and flow and transport behavior in the perched zone was captured in the model. For such a method to find applications in future vadose zone modeling efforts, additional characterization is required to pinpoint the properties and three-dimensional structure of the perching horizon, and to determine the lateral extent of the deep perched groundwater. Then, this approach can be used to match observed perched water through calibration and tuning of the permeability reduction factor parameter. An additional area of future work will be to study transient flow conditions to determine the nature of perched water bodies in the presence of short- and long-term changes in local infiltration in a canyon. Finally, in locations where transport away from the canyon bottoms is suspected, three-dimensional modeling will be necessary.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONPERCHED WATER OBSERVATIONS...INTERPRETATION OF DEEP PERCHED...NUMERICAL MODELING OF DEEP...DISCUSSION AND CONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. L. Duke, R. C. Roback, P. W. Reimus, R. S. Bowman, T. L. McLing, K. E. Baker, and L. C. Hull Elucidation of Flow and Transport Processes in a Variably Saturated System of Interlayered Sediment and Fractured Rock Using Tracer Tests Vadose Zone J., November 20, 2007; 6(4): 855 - 867. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. D. Newman and B. A. Robinson The Hydrogeology of Los Alamos National Laboratory: Site History and Overview of Vadose Zone and Groundwater Issues Vadose Zone J., August 16, 2005; 4(3): 614 - 619. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. H. Birdsell, B. D. Newman, D. E. Broxton, and B. A. Robinson Conceptual Models of Vadose Zone Flow and Transport beneath the Pajarito Plateau, Los Alamos, New Mexico Vadose Zone J., August 16, 2005; 4(3): 620 - 636. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
0.1 times initial concentration.