VZJ sign up for citetrack
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 8 October 2007
Published in Vadose Zone J 6:774-785 (2007)
DOI: 10.2136/vzj2007.0047
© 2007 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Newman, B. D.
Right arrow Articles by Gram, P.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Newman, B. D.
Right arrow Articles by Gram, P.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Newman, B. D.
Right arrow Articles by Gram, P.
Related Collections
Right arrow Isotopes
Right arrow Recharge
Right arrow Vadose Zone Processes and Chemical Transport

ORIGINAL RESEARCH

Flow and High Explosives Transport in a Semiarid Mesa in New Mexico, USA

Brent D. Newmana,*, Donald D. Hickmotta and Peter Gramb

a Earth and Environmental Sciences Division, MS J495, Los Alamos National Lab., Los Alamos, NM 87545
b PMC Technologies, 2237 Trinity Dr., Los Alamos, NM 87544. B.D. Newman now at the Isotope Hydrology Section, International Atomic Energy Agency, PO Box 100, Wagramer Strasse 5, Vienna, A1400, Austria
* Corresponding author (b.newman{at}iaea.org).

All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.


Received 9 March 2007.



   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 Site Description
 Materials and Methods
 Results
 Discussion
 Conclusions
 REFERENCES
 
Outfalls from high explosives (HE) production facilities at Los Alamos National Laboratory discharged HE-contaminated waters onto a semiarid mesa on the Pajarito Plateau for about 50 yr. As part of an initial hydrogeologic characterization to examine the impact of HE contamination, four boreholes were drilled into the vadose zone to depths between 38.1 and 63.3 m. The study objectives were to characterize contaminant nature and extent and to identify potential source areas and transport pathways. Besides providing an example of HE transport in the vadose zone, this study shows the value of integrating chloride and stable isotope tracer approaches with contaminant distribution information, and it provides insights on semiarid vadose zone behavior in the little studied, but widespread ponderosa pine forests of the American Southwest. Chloride-based vadose zone residence time estimates (1950–6080 yr) suggest that downward flow and transport over much of the mesa is limited. However, the presence of HE-contaminated transient saturated zones in two boreholes indicates that localized fast pathways also occur. Stable isotope data ({delta}18O and {delta}D) suggest that the source areas for contamination are former HE outfall discharge ponds that provide focused recharge to the transient saturated zones.

Abbreviations: HE, high explosives • HMX, high melting explosive, octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazine • RDX, Research Department explosive, hexahydro-1,3,5-trinitro-1,3,5-triazine • SWSC, Sanitary Waste System Consolidation • TNT, 2,4,6-trinitrotoluene


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
Site Description
 Materials and Methods
 Results
 Discussion
 Conclusions
 REFERENCES
 
Outfalls from high explosives (HE) production sites at Los Alamos National Laboratory, Technical Area 16 (TA-16) (Fig. 1a ) discharged RDX (Research Department explosive, hexahydro-1,3,5-trinitro-1,3,5-triazine), TNT (2,4,6-trinitrotoluene), and HMX (high melting explosive, octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazine) contaminated waters onto a ponderosa pine (Pinus ponderosa) covered mesa top at the western edge of the Pajarito Plateau in north-central New Mexico from 1944 to 1996. Reid et al. (2005) discussed HE contamination in an adjacent canyon site under perennial stream flow conditions; this study extends their work by examining behaviors in a mesa environment. On the mesa, small quantities of HE were released through burial or distribution of solid phase HE in or on the soil. The vast majority of the HE, however, was released through wastewater discharges that distributed HE through surface runoff or into a series of unlined wastewater ponds. Individual ponds covered less than 900 m2 in surface area and were less than 2 m deep. Historic discharges from individual wastewater outfalls were often greater than 100 m3 yr–1, and some outfalls operated for about 50 yr; however, once wastewater releases were discontinued, the ponds were frequently dry. Most of ponds have since been removed as part of site cleanup efforts. HE concentrations in surface soils have ranged to more than 20% w/w, and HE in surface waters have ranged to more than 800 µg L–1 (Los Alamos National Laboratory, 1998). As part of an initial investigation of HE contamination, four boreholes were drilled to depths between 38.1 and 63.3 m to characterize the TA-16 mesa vadose zone hydrologic properties and define the nature and extent of contamination in an area with multiple HE release locations and types. Before drilling, little was known about the vadose zone stratigraphy, hydrologic properties, and fluxes in this part of the laboratory, or what the contaminant distributions (i.e., the nature and extent of HE) might be. The vadose zone at TA-16 is approximately 228 m thick, so this study did not investigate the entire thickness of the vadose zone. It is, however, one of the first hydrogeologic and contaminant characterization efforts on the western part of Los Alamos National Laboratory and the Pajarito Plateau.


Figure 1
View larger version (46K):
[in this window]
[in a new window]

  		FIG. 1. (a) Location map of Los Alamos National Laboratory and the TA-16 mesa site (solid rectangle within the laboratory boundary); (b) the Sanitary Waste System Consolidation (SWSC), 90s line, Martin, and Building 300 boreholes.

 
The main objective of this paper is to document how a combined approach using multiple environmental tracers along with characterization of contaminant distribution, hydraulic properties, and hydrology was used to evaluate the nature and extent of HE contamination and the controls on flow in a complex vadose zone environment. Through this combined approach, we have developed a conceptual model that describes factors that are important for controlling water flow and HE transport in the mesa. As an additional benefit, we were able to quantify downward fluxes within a southwestern U.S. ponderosa pine ecosystem, the importance of which is discussed below.

High explosives are contaminants of concern in many countries, and releases can occur from HE production and machining facilities, disposal facilities, and firing sites, and at locations where large quantities of munitions have been stored in uncontrolled environments. The different varieties of HE have a wide range of solubilities, transport characteristics, and degradation properties and can contaminate the vadose zone as well as groundwater (Beller et al., 2004; DiGnazio et al., 1998; Fryar et al., 2000; Layton et al., 1987; Pennington and Brannon, 2002; Reid et al., 2005; Spalding and Fulton, 1988). Some types of HE have characteristics that make them of greater concern when released into the environment. One of the more problematic varieties is RDX, which was the most commonly processed type of HE at the TA-16 site (Los Alamos National Laboratory, 1998). It can behave conservatively and thus be quite mobile, especially in oxidizing, low carbon environments typical of many vadose and groundwater systems (Layton et al., 1987; Pennington and Brannon, 2002; Reid et al., 2005; Spalding and Fulton, 1988). Under these conditions, RDX degradation appears to be slow, and sorption is minimal (Layton et al., 1987; Reid et al., 2005; Spalding and Fulton, 1988). In addition, the relatively high solubility of RDX (42–60 mg L–1; Card and Autenrieth, 1998; Layton et al., 1987) and low drinking water health advisory of 2 µg L–1 (USEPA, 2000) create a combination where human health risks can be an issue. Studies that describe HE contamination and transport in semiarid vadose zones are quite limited (e.g., Fryar et al., 2000); this study broadens the understanding of HE transport within thick, bedrock-dominated vadose zone environments.

The study area is covered by ponderosa pine forest, and thus, the results reported here also apply to understanding hydrological behavior in these important ecosystems. Few studies have quantified vadose zone fluxes in ponderosa pine ecosystems in the southwestern United States (Newman et al., 1997; Sandvig and Phillips, 2006). However, ponderosa pine is the dominant tree species on more than 16 million ha in North America (McPherson, 1997). In addition, ponderosa pines occupy important landscape positions in terms of groundwater recharge in the southwestern United States. Using the terminology of Duffy (2004) and Wilson and Guan (2004), ponderosa pine forests are commonly associated with the mountain block and front and the upper piedmont, all of which can be associated with higher recharge rates than generally occur within the lower parts of semiarid basins. Thus, quantifying hydrologic behavior in ponderosa pine environments helps define expected distributions of fluxes for these ecosystems, which can aid in using vegetation as an indicator for estimating basin-scale recharge (see discussion in Newman et al., 2006; Sandvig and Phillips, 2006; Walvoord and Phillips, 2004).


   Site Description
TOP
 ABSTRACT
 INTRODUCTION
 Site Description
Materials and Methods
 Results
 Discussion
 Conclusions
 REFERENCES
 
The climate at TA-16 is semiarid, with annual precipitation rates about 500 mm yr–1 (Bowen, 1990). The mesa site is at an elevation of approximately 2316 m and is underlain by the uppermost cooling units of the Tshirege (upper) member of the Bandelier Tuff, which was erupted 1.22 million years ago (Izett and Obradovich, 1994). The Bandelier Tuff consists of two basal Plinian fallout pumice beds with overlying ash flow deposits (Broxton and Reneau, 1995; Broxton and Vaniman, 2005; Crowe et al., 1978). These ash flow deposits are characterized by varying degrees of welding and devitrification. The welding and devitrification characteristics have a major influence on the porosity and permeability of the vadose zone. Internal morphology ranges from coarse bedded and surgelike to massive devitrified ignimbrite. Surge beds (unwelded flows) are discontinuous features that form mainly through collapse of vertical eruption columns (Fisher and Schmincke, 1984). They can be important hydrologically (as in this case) because they form relatively high conductivity units that can lie directly adjacent to much lower conductivity tuff units. The Bandelier Tuff shows significant variation in both degree of welding and degree of devitrification between individual units. These reversals are due to brief cooling intervals that separated the emplacement of individual flow units. The subunits underlying the TA-16 mesa are among the most heterogeneous within the Tshirege member on the Pajarito Plateau. Densely welded tuff, nonwelded tuff, and surge bed material are present within 61 m of the surface at TA-16 (Broxton and Reneau, 1995; Lewis et al., 2002; Rogers, 1995).


   Materials and Methods
TOP
 ABSTRACT
 INTRODUCTION
 Site Description
 Materials and Methods
Results
 Discussion
 Conclusions
 REFERENCES
 
Borehole Drilling
In 1997 four hollow stem auger boreholes were drilled and continuously sampled using a split spoon. The four locations span the width of the TA-16 mesa (Fig. 1b). The boreholes (90s line, Building 300, Sanitary Waste System Consolidation [SWSC], and Martin boreholes) were located in the vicinity of known or potential HE source areas. The 90s line borehole was located near what was called the 90s line HE production facility, which discharged HE-contaminated waters into a series of sumps, ditches, and a lagoon known as the 90s line pond. The Building 300 borehole was located near another HE facility called Building 300. This borehole was also located in a geophysical (self-potential) anomaly (Los Alamos National Laboratory, 1998). The SWSC borehole was located approximately 200 m east of a major HE outfall. The Martin borehole was also located in a geophysical anomaly (self-potential and resistivity) near the head of Martin Canyon (which contains an HE-contaminated spring; Los Alamos National Laboratory, 1998). In addition, the Martin borehole is located approximately 900 m east of a series of former HE wastewater lagoons and outfalls. The Building 300 and SWSC holes were drilled to 57.6 and 63.3 m, respectively. The 90s line and Martin holes were drilled to 50.3 and 38.1 m, respectively, and drilling was stopped when flowing water was encountered.

Core Analysis
Each hole was continuously cored, and stratigraphic descriptions were made of the various tuff units encountered. Descriptions included secondary crystallization, welding, pumice size, ferromagnesian mineral and quartz abundances, presence of fractures, and bulk rock and fracture alteration (Lewis et al., 2002; Los Alamos National Laboratory, 1998).

In addition to the lithologic descriptions, moisture-protected core samples were collected for hydrologic property characterization and for pore water chloride and stable isotope analyses. Analyses of hydrologic properties were conducted by Daniel B. Stephens & Associates (1999) and included gravimetric water content, bulk density, saturated hydraulic conductivity, moisture characteristic curve information, and calculated porosity.

Pore water chloride and stable isotope samples were collected approximately every 3 m. To preserve in situ water content, samples were placed in glass septum jars for chloride analyses and heat-sealed, aluminized bags for stable isotope analyses. Pore water chloride and stable isotope values were determined using methods described in Newman et al. (1997). Gravimetric water content analyses were performed before conducting the chloride leaching analyses according to Gardner (1986). Chloride concentrations were measured using ion chromatography with an analytical precision of ± 5%. The chloride mass balance approach (Allison et al., 1985; Anderholm, 1994; Newman et al., 1997; Phillips, 1994; Scanlon, 2000) was used to calculate vadose zone downward (residual) fluxes according to

Formula 1[1]
where R is the residual flux, Clp is the average annual chloride concentration in precipitation (g m–3), P is the average annual precipitation rate (m yr–1), and Cla is the average pore water chloride concentration (g m–3) in the borehole. Vadose zone residence times were also be estimated using the chloride mass balance method:

Formula 2[2]
where A is the vadose zone residence time down to a certain depth of interest (yr), and Clcum is the cumulative chloride content from the ground surface to the depth of interest (g m–2). For this study, we used values of 0.29 g m–3 for Clp (Anderholm, 1994) and 0.5 m yr–1 for P (Bowen, 1990).

Stable isotope measurements ({delta}18O and {delta}D) of pore waters have been used effectively as indicators of water source and evaporation in hydrological studies (Allison et al., 1983; Barnes and Allison, 1983; Barnes and Allison, 1984; DePaolo et al., 2004; Machavaram et al., 2006; Newman et al., 1997). Stable isotope analyses were conducted at the New Mexico Tech stable isotope laboratory using the distillation method of Shurbaji et al. (1995) and the extraction methods of Socki et al. (1992) and Kendall and Coplen (1985). Analytical precisions for the {delta}18O and {delta}D analyses were 0.2 and 2{per thousand}, respectively. The isotope values are reported based on the V–SMOW (Vienna–Standard Mean Ocean Water) standard.

Core was also analyzed for HE every 1.5 m using an immunoassay screening method (DTECH, Strategic Diagnostics, Inc., Newark, DE), supplemented with laboratory HE analyses using EPA SW-846, Method 8330 (USEPA, 1996) of two to four grab samples from each borehole. The detection limits were approximately 0.5 mg kg–1 and 0.1 µg kg–1 for the DTECH and SW-846 methods, respectively (Los Alamos National Laboratory, 1998). Groundwater samples from the 90s line and Martin boreholes were analyzed using both DTECH and EPA SW-846 methods whenever water flowed into the borehole.

Borehole Monitoring
The four boreholes were monitored for volumetric water content using neutron attenuation, and the probe was calibrated to local soils and Bandelier Tuff. Monitoring was done quarterly from 1999 through 2000. In addition, the 90s line and Martin holes were monitored for groundwater levels. The Martin borehole groundwater was sampled for stable isotopes and HE whenever enough water flowed into the borehole. The 90s line hole has not flowed since the initial sampling (probably because of a multiyear drought).


   Results
TOP
 ABSTRACT
 INTRODUCTION
 Site Description
 Materials and Methods
 Results
Discussion
 Conclusions
 REFERENCES
 
Stratigraphy and Lithology
The stratigraphic units encountered in the boreholes include shallow soils, alluvium, the El Cajete pumice, and the upper units of the Tshirege member of the Bandelier Tuff (Fig. 2 ). The deepest Tshirege subunit encountered was the partially to moderately welded subunit Qbt3. A surge bed exists between subunit Qbt3 and the overlying subunit Qbt3t. Subunit Qbt3t is commonly very densely welded at the top and bottom contacts. Unit Qbt4 is the uppermost Tshirege subunit and contains a basal poorly welded interval, a non- to poorly welded middle, and an upper densely welded interval. The unit Qbt4 intervals are typically separated by surge beds. Surge beds can also be present within subunits Qbt3 and Qbt3t. Details of the stratigraphic units and their lithologic and mineralogical characteristics are discussed in Lewis et al. (2002).


Figure 2
View larger version (26K):
[in this window]
[in a new window]

  		FIG. 2. Volumetric water contents (from neutron probe) and stratigraphy from (a) the 90s line, (b) Building 300, (c) Martin, and (d) Sanitary Waste System Consolidation (SWSC) boreholes. Qal is Quaternary alluvium, the el Cajete is a pumice bed, and Qbt4 through Qbt3 are subunits of the Quaternary Tshirege member of the Bandelier Tuff.

 
Hydrologic Properties
A summary of the hydrologic properties based on geologic unit is shown in Table 1. The hydrologic data show a wide range in bulk densities, porosities, and saturated hydraulic conductivities that are related to the degree of welding. Surge beds have the highest porosities and saturated conductivities and the lowest bulk densities. Partly welded tuff units have intermediate values of hydrologic properties, and the densely welded units have the highest bulk densities, and lowest porosities and hydraulic conductivities. However, the laboratory analyses essentially measure matrix conductivities, and core and outcrop observations in the area suggest that fractures are common in the welded units (Lewis et al., 2002), and high-permeability deformation bands, which can also promote preferential water movement in the vadose zone, are present in the less-welded units (Wilson et al., 2003; Wilson et al., 2006). Thus, even though the matrix conductivities in the densely welded units are low, preferential flow-path conductivities can be high.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Hydrologic properties summary for the uppermost subunits of the Tshirege member, Bandelier Tuff at TA-16.

 
Water Content Results
Example volumetric water content results for the boreholes are shown in Fig. 2a through 2d. In general, there was little variation in water content with time (typically less than 2%, except at the shallowest depths). Profiles such as these reflect the semiarid climate and the low deep percolation rates observed for many mesa locations on the Pajarito Plateau (Birdsell et al., 2005; Kwicklis et al., 2005). Volumetric water contents range from 3 to 23%, and average water contents range from 6% to 10% in the four holes. Changes in water content with depth appear to be associated with changes in stratigraphy and/or degree of welding. During drilling, flowing water was encountered at approximately 39 m in the 90s line borehole and 24 m in the Martin borehole (Fig. 2a and 2c). These zones correspond to intervals of fractured, welded tuff. After the water level stabilized, the holes were bailed dry. No additional flow was observed in the 90s line borehole. The Martin borehole also remained dry most of the time. However, five separate flow events were observed in the Martin borehole over the next 5 yr. These events were in response to high snowmelt and rainfall periods. In these groundwater zones, volumetric water contents were about 18%. The calculated porosities for the welded intervals that contain the wet zones range from 16 to 20%, suggesting that the groundwater zones were at or near saturation for the study period. Based on the ephemeral flow, we will hereafter refer to the groundwater producing depths as transient saturated zones.

High Explosives Results
High explosives analyses of the SWSC and Building 300 boreholes did not show any detectible HE. There were no detections of HE from the 90s line and Martin boreholes except in the groundwater near the bottom of the holes. In 1997 groundwater from the 90s line borehole contained HE in the form of 2,4-dinitrotoluene (4.03 µg L–1), HMX (21 µg L–1), RDX (281 µg L–1), and TNT (1.27 µg L–1). The 1997 sample was collected a few days after drilling was completed, and this was the only time the borehole produced water. The Martin borehole flowed for short periods on five different occasions from 1997 through 2001. High explosives were detected in the form of amino-2,6-dinitrotoluene (1–2.5 µg L–1), HMX (2.3–14 µg L–1), and RDX (15.2–120 µg L–1). There was no clear trend in the Martin borehole concentrations with time, although the 2000 and 2001 flow events had substantially lower concentrations for all types of HE than had been observed previously.

Chloride Results
Pore water chloride profiles for the 90s line, Martin, SWSC, and Building 300 boreholes are shown in Fig. 3 . Concentrations range from 11 to 1952 mg L–1. Like the water content data, changes in chloride concentrations appear to be associated with changes in stratigraphy and degree of welding. Vadose zone residence times are 2240 yr for the 90s line borehole, 6080 yr for the Martin borehole, 1950 yr for the SWSC borehole, and 2300 yr for the Building 300 borehole. Residence times are based on cumulative chloride from the surface to the top of the transient saturated zones for the 90s line and Martin boreholes, and to the bottom of the SWSC and Building 300 boreholes. These long residence times suggest that there has been little to no downward recharge at the borehole locations for nearly 2000 yr or longer.


Figure 3
View larger version (21K):
[in this window]
[in a new window]

  		FIG. 3. Pore water chloride profiles for the 90s line, Martin, Sanitary Waste System Consolidation (SWSC), and Building 300 boreholes.

 
Chloride-based residual fluxes are similar for the four boreholes (1.3 mm yr–1 for the 90s line borehole, 1.3 mm yr–1 for the Martin borehole, 1.4 mm yr–1 for the SWSC borehole, and 1.2 mm yr–1 for the Building 300 borehole). Walvoord et al. (2002a, 2002b) showed that in areas dominated by upward fluxes, the chloride mass balance method will erroneously yield small positive values (i.e., the one-dimensional downward flow assumption of the chloride method is not always consistent with actual field conditions, and therefore, an overestimation of the downward flux can occur). However, as Sandvig and Phillips (2006) pointed out, gradients in ponderosa pine environments appear to be largely downward; therefore, the fluxes reported here should be reasonably representative. At worst, they could be considered maximum estimates. It is difficult to quantify uncertainty accurately in the chloride mass balance results because of, for example, the problem of defining how well the chloride and precipitation input rates used in Eq. [1] and [2] represent the real values over the last few thousand years or more.

Another aspect to the chloride flux results from this study aids in understanding flow in the mesa: the downward flux of water in the vadose zone appears to be relatively constant with depth. Cumulative chloride/cumulative water content plots are relatively linear, which is an indicator of approximately constant downward flux (Newman et al., 1997; Scanlon, 2000; Stone, 1984; Fig. 4 ). One exception to the approximately constant linear trends is for the shallow part of the 90s line borehole. This deviation likely resulted from localized near-surface evapotranspiration effects.


Figure 4
View larger version (14K):
[in this window]
[in a new window]

  		FIG. 4. Cumulative water/cumulative chloride plots for the 90s Line, Martin, Sanitary Waste System Consolidation (SWSC), and Building 300 boreholes. The approximately linear trends suggest near-constant downward fluxes with depth.

 
Although the chloride results suggest long vadose zone residence times, these results appear to be contradictory to the presence of HE at depth in the 90s line and Martin boreholes. The HE data show that some recharge to the transient saturated zones has occurred within the last 50 years. This apparent inconsistency is discussed further below.

Stable Isotope Results
Stable isotope ({delta}18O) profiles for the four boreholes are shown in Fig. 5a through 5d (the {delta}D profiles are similar to the {delta}18O profiles and are thus not shown). The profiles show values that are generally below about –4{per thousand} at depths above the transient saturated zones and throughout the SWSC and Building 300 boreholes. However, the isotope values in the transient saturated zones in the 90s line and Martin boreholes from the 1997 sampling are distinctive, having the highest {delta}18O and {delta}D values. These data indicate that the waters were strongly evaporated. Values that plot far to the right of the local meteoric water line ({delta}D = 8{delta}18O + 12; (Vuataz and Goff, 1986) are characteristic of evaporation (Fig. 6 ). Isotope values above and below the transient saturated zones show substantially less evaporation effect (with the exception of one sample at 15 m in the 90s line borehole; see Fig. 5a). It is possible that the 15-m result was caused by a stable isotope sample preservation problem because chloride concentrations are not elevated at this depth.


Figure 5
View larger version (25K):
[in this window]
[in a new window]

  		FIG. 5. Stable isotope ({delta}18O) and volumetric water content profiles at (a) the 90s line, (b) Building 300, (c) Martin, and (d) Sanitary Waste System Consolidation (SWSC) boreholes.

 

Figure 6
View larger version (22K):
[in this window]
[in a new window]

  		FIG. 6. Local meteoric water line plot of the Martin and 90s line borehole data. Values that plot far to the right of the line indicate substantial evaporation. Values that plot on or near the line indicate minor to no evaporation. Data include analyses from pore water samples extracted from cores and bailed saturated zone samples (1999 data).

 
In addition to the initial stable isotope sampling, we were able to obtain isotope samples during three of the transient flow events in the Martin borehole (spring and summer 1999). The {delta}18O and {delta}D results ranged from –10.8 to –12{per thousand} and –76{per thousand} to –83{per thousand}, respectively, and indicate a shift from highly evaporated to more meteoric type values.


   Discussion
TOP
 ABSTRACT
 INTRODUCTION
 Site Description
 Materials and Methods
 Results
 Discussion
Conclusions
 REFERENCES
 
General Discussion
Our initial discussion focuses on the unsaturated portion of the mesa vadose zone (the region above the transient saturated zones). Chloride-based residence times on the order of 2000 yr in the unsaturated parts of the vadose zone suggest little downward movement of water and HE over much of the mesa (e.g., areas not directly underneath ponds, drainages, or with large open surface fractures). Thus, much of the mesa vadose zone does not appear to be an effective pathway for recharge and deep contaminant transport. This result is not surprising given the semiarid conditions at the site. However, this conclusion must not hold for the entire mesa because the deep HE contamination in the 90s line and Martin boreholes indicates recharge to the transient saturated zones in less than about 50 yr. The HE facilities began operation in the 1940s. The stable isotope data from the saturated zones provides an important clue to resolve the apparent inconsistency between the long residence times indicated by the chloride data and the short residence times as indicated by the presence of HE contamination in the transient saturated zones. The anomalously high isotope values from the saturated zones (Fig. 5 and 6) suggest that the waters have undergone substantial evaporation. It is unlikely that the waters were evaporated in situ because the zones are at or near saturation and because pore waters above and below the saturated zones do not show any evidence of strong evaporation (with one minor exception discussed earlier). In reviewing the possible sources of water that have undergone substantial evaporation (having relatively high isotope values) on the mesa, most outflows and discharge waters were not strongly evaporated when they were released, having {delta}18O values of about –10.5{per thousand} and {delta}D values of about –76{per thousand} or lower (Blake et al., 1995; Los Alamos National Laboratory, 1996). Waters used for HE processing at the site came from the Los Alamos drinking water aquifer and a high-elevation spring system. Isotope values from these waters plot close to the local meteoric water line, and only minor evaporation would have occurred during HE processing. Thus, the isotopic compositions of outflow waters are not consistent with the high values observed in the 90s line and Martin transient saturated zones. The only sources that are consistent with the observed saturated zone values are existing and former disposal ponds on the mesa. These artificial surface impoundments were subject to significant evaporation, particularly in the summer. A limited set of stable isotope data is available from the 90s line pond, and values show evaporative effects similar to the borehole waters (e.g., pond water values of 6.8 and 2{per thousand}, {delta}18O and {delta}D, respectively). It seems clear from the stable isotope data that the ponds are and were a likely source of recharge and HE contamination to the transient saturated zones encountered in the 90s line and Martin boreholes. Studies by Fryar et al. (2000) and Machavaram et al. (2006) have also shown that elevated isotope values such as those observed here are consistent with pond- or playa-type source areas. Thus, as Scanlon and Goldsmith (1997) pointed out, even though there may be substantial evaporation from ponded waters, this does not preclude recharge from happening at those locations. The mesa pond sources have similar chloride and HE concentrations to the transient saturated zones (Los Alamos National Laboratory, 1996), and both of the boreholes were drilled in the vicinity (within 300 m) of disposal ponds, also supporting the interpretation of ponded water sources. Another aspect that suggests that the ponds were subject to evaporation is that chloride concentrations in the transient saturated zones are relatively high (Fig. 3), yet the sources of water for HE processing at the site were only about 2 mg L–1 (Blake et al., 1995; Los Alamos National Laboratory, 2005).

Discharge ponds, lagoons, cesspools, or leaching pits were commonly used for HE disposal in the past. Studies in California, Indiana, Nebraska, and Texas (Beller et al., 2004; DiGnazio et al., 1998; Fryar et al., 2000; Spalding and Fulton, 1988) all indicate that ponded water HE disposal has resulted in groundwater contamination. In addition to their importance for HE disposal, small artificial ponds are also significant hydrological features in other areas (especially in the United States). Smith et al. (2002) found that small artificial water bodies account for approximately 20% of the standing water area across the United States and noted that their impact on hydrological and other processes is apparently large in proportion to their area. Thus, artificial ponds like the ones studied here represent significant anthropogenic alterations to the landscape, and this study contributes to our understanding of the impacts these features have on hydrological systems.

There must also be a lateral component to subsurface flow in the mesa since the boreholes were not drilled directly adjacent to any of the ponds. The presence of HE in the transient saturated zones and absence of HE above and below these zones, stable isotope and water content distributions in the boreholes, the superposition of high-conductivity tuff units over low conductivity units, and observed subhorizontal fracturing in core samples from the transient saturated zones are additional evidence for a lateral flow component.

The timing of flows into the Martin borehole and the observed shift to negative isotope values suggest that the transient nature of flow and transport is seasonally or event controlled (although flow must have been more regular during the period of active discharge into the ponds). All of the recent flow events in the borehole were associated with relatively large snowmelt or precipitation events. It is likely that even though the pond waters have high isotope values (from evaporation) during part of the year, values shift to more meteoric type values during the winter and, in the past, during high-volume wastewater discharges. Thus, it is not surprising that the isotope values in the borehole vary over time. It is also likely that there is some vadose zone mixing between older contaminated pond waters and newer recharge as indicated by the response to snowmelt and precipitation events and the apparent dilution of HE concentrations in more recent samples.

Conceptual Model
Birdsell et al. (2005) presented a series of conceptual models that describe percolation and recharge conditions for canyons and mesas on the Pajarito Plateau. Our study describes elements that fall within their mountain front and disturbed mesa conceptual models. Specifically, Birdsell et al. (2005) noted that the combination of higher precipitation rates and fractures in densely welded tuffs that occur in mountain front mesas (such as at TA-16) can enhance recharge, especially compared with mesas at lower elevations. They also described how disturbed conditions such as artificial ponds can substantially alter subsurface flow in mesas compared with natural conditions. From the characterization data collected in this study, we extend the work of Birdsell et al. (2005) through the development of a detailed conceptual model of the conditions at the TA-16 mesa site (Fig. 7 ). This new model summarizes what appears to control "rapid" flow and transport to depths of over 30 m in parts of the mesa, while other parts of the mesa have only slow downward flow. The model highlights the importance of disturbance (i.e., ponding) on vadose zone behavior in this heterogeneous system. It also shows how important it can be to understand the relative roles of matrix and preferential pathways in controlling flow and transport in the vadose zone. These results are consistent with studies in the semiarid, subhumid High Plains of the United States that clearly show the importance of ponding in creating strong contrasts in downward fluxes and recharge rates between topographically low areas (e.g., ponds and playas) and interpond areas (Fryar et al., 2001; McMahon et al., 2006; Scanlon and Goldsmith, 1997). These studies also noted that preferential flow through fractures and cracks below ponded areas also play an important role in the recharge process.


Figure 7
View larger version (23K):
[in this window]
[in a new window]

  		FIG. 7. Generalized conceptual model of the TA-16 mesa vadose zone (not to scale) showing the two dominant flow domains. Under areas affected by surface ponding (represented by flow paths to the left of the borehole) downward and lateral fluxes are substantial. These flow paths transport high explosives (HE) and, combined with the hydraulic properties of the tuff units, cause the development of transient saturated zones at depth. Under areas not affected by surface ponding (represented by flow paths to the right of the borehole), low downward fluxes predominate. This low flux domain is present over most of the TA-16 mesa.

 
In the TA-16 mesa conceptual model, pond water evaporates and percolates into the underlying tuff, where it flows laterally on encountering welded (less-permeable) tuff units or subhorizontal fractures within these units. Lateral flow is promoted because of the superposition of relatively high hydraulic conductivity units (e.g., surge beds and poorly welded tuffs) over units with low matrix hydraulic conductivities (e.g., strongly welded tuffs; see Table 1). We hypothesize that the flow paths may have a stair-stepped appearance as indicated in Fig. 7 because the geology and hydrologic properties data indicate multiple zones of high/low conductivity transitions.

Although the four boreholes provided useful information about contaminant nature and extent, they do not completely constrain the extent of HE contamination. Additional boreholes could help establish a better understanding of the distribution of HE in the vadose zone. It is clear that downward flow and transport from the ponds has occurred at rapid rates, especially when compared to chloride mass balance results from interpond areas. In 50 years or less, HE has been transported at least 40 m downward and at least 900 m laterally within the mesa (based on the distance from the Martin borehole to the nearest pond). It is quite likely that the extent of contamination is larger than what has been determined here. An important consideration for future work is whether HE from pond source areas has reached the deep aquifer system, 228 m below the mesa top. Assuming a 50-yr disposal period, minimum transport velocities for at least some portion of the water down to the transient saturated zones are on the order of 0.5 to 0.8 m yr–1, and actual velocities could be substantially higher. The shift from isotopically heavy to light compositions in the saturated zone over an approximately 2-yr period suggests transport velocities are likely much higher than the estimated minimum range of 0.5 to 0.8 m yr–1. Unfortunately, we do not know the age of the light water that entered the saturated zone, which prevents making an accurate estimate of an isotope-based velocity. In other words, the light water entering the saturated zone may have been pre-event water that has been stored in the vadose zone for some unknown amount of time before its mobilization to the borehole. It is also difficult to estimate Darcy fluxes below the ponds given the lack of water content data directly below the ponds and because of the complication of mixed matrix and preferential flow. Future drilling within former pond areas should help in this regard.

An alternative conceptual model that was considered for the site is that significant fracture flow in the nondisturbed areas where the boreholes were drilled would mean that the fluxes in these areas are much higher than suggested by the chloride mass balance approach. Under such a scenario, significant recharge to the transient saturated zones could be occurring in the vicinity of the boreholes. The chloride mass balance method is relatively insensitive to preferential flow because the large concentrations of chloride that are stored in the matrix can mask potentially low concentrations that would occur in a fracture or other preferential flow path because a bulk sample of rock or soil is analyzed (see, e.g., Newman et al., 1998). Although we cannot totally reject this alternative conceptual model, we find it to be unlikely based on four lines of evidence. First, during 2 yr of quarterly water content monitoring, we never observed any local increases in water content that would suggest that fracture flow was occurring near the boreholes. This includes the 1999 periods when saturated zone flow events in the Martin borehole were observed. Second, as mentioned earlier, HE and chloride concentrations in the transient saturated zones were similar to those measured in the ponds. Significant fracture flow from nondisturbed areas would have diluted these concentrations unless there were substantial contaminant inventories in the vadose zone in the nondisturbed areas (which does not appear to be the case). Third, studies by Soll and Birdsell (1998) and Robinson et al. (2005) suggest that under the low water content conditions (above the saturated zones) observed here, even when fractures are present, flow is dominantly through the tuff matrix. These studies did not examine densely welded tuffs; however, at TA-16, these units occur between nonwelded to moderately welded tuff units where matrix flow should be dominant. It is also worth noting that nonwelded to moderately welded tuff units comprise most of the mesa vadose zone at the study area. Fourth, if fracture flow was important from nondisturbed areas around the boreholes, it is unlikely that we would have observed the high isotope values in the transient saturated zones. Fracture flow probably would not show a significant evaporative isotope signature here because flow would quickly leave the near-surface soil zone; therefore, little to no evaporation would occur. Newman et al. (1998), for example, showed that preferential flow in the soil zone at a ponderosa pine site only a few kilometers from this study area had little to no evaporative enrichment. Thus, the high isotope values are inconsistent with significant fracture flow in the vicinity of the boreholes. We do acknowledge that under ponded conditions, fracture flow is likely to be important, which is reflected in the conceptual model shown in Fig. 7. We also noted areas on the mesa where open fractures extend to the mesa surface, and run-on into these fractures could provide recharge to saturated zones within the mesa. These fractures exist west of the study area, and recharge from them could explain in part the shift from the high to low isotope values that occurred in the Martin borehole saturated zone in 1999. However, this process would not invalidate using the chloride mass balance approach in this case because recharge from these particular fractures would enter the saturated zone west of the study area.

Fluxes beneath Ponderosa Pine Forests
One of the other benefits of this study was that we were able to investigate vadose zone fluxes beneath a ponderosa pine forest. As mentioned earlier, these are widespread ecosystems in the southwestern United States that occupy hydrologically important landscape positions. Thus, ponderosa pine ecosystems have potential for use as an indicator of a specific range of downward fluxes that can be used to support basin-scale estimates of diffuse recharge (Kwicklis et al., 2005; Newman et al., 2006; Sandvig and Phillips, 2006; Walvoord and Phillips, 2004). In an investigation of arid and semiarid ecosystems along an elevationalF gradient in central New Mexico, Sandvig and Phillips (2006) found that downward fluxes beneath ponderosa pines were the highest of the four ecosystems studied. Using the chloride mass balance method, they estimated a downward flux beneath ponderosa pines of about 2.3 mm yr–1, and residence times of 300 to 3600 yr (for 4.5-m deep profiles). Their flux estimate is similar to the 1.3 mm yr–1 average estimated here. Newman et al. (1997) used chloride mass balance in a ponderosa pine forest at another mesa site on the Pajarito Plateau and obtained a range of downward fluxes from 0.1 to 2 mm yr–1 through the soil zone (<1.5 m deep). They attributed the low fluxes to the presence of a thick clay horizon. Sandvig and Phillips (2006) also noted that clay horizons may be a factor in controlling fluxes below ponderosa pines in their study (i.e., the site with the 3600-yr residence time). Clay horizons are present in the soil zone at our study site and thus likely play a role in controlling fluxes to the deeper system.

It is interesting to note that the previous studies were conducted using profiles that were less than 5 m deep, yet, with the exception of the lower values from the Newman et al. (1997) study, fluxes are similar to those reported here, which were determined from much deeper profiles (38–63 m). The linearity of cumulative water versus cumulative chloride plots found here (Fig. 4) and by Sandvig and Phillips (2006) suggests that fluxes are relatively constant with depth and that shallow boreholes in interdrainage locations that penetrate bedrock (e.g., <10 m) may be adequate for determination of diffuse type fluxes in these ecosystems. However, shallow boreholes may miss features such as transient saturated zones that are driven by focused flow processes.

Collectively, the residual flux estimates for southwestern U.S. ponderosa pine ecosystems from Newman et al. (1997), Sandvig and Phillips (2006), and this study range from 0.1 to 2.3 mm yr–1. Although additional vadose zone studies in southwestern ponderosa pine ecosystems are clearly needed, the range can be considered an initial benchmark for the type of diffuse fluxes that can be expected in these ecosystems. Scanlon et al. (2006) compiled flux estimates from chloride mass balance studies in arid and semiarid regions worldwide and found a range of 0.2 to 35 mm yr–1. Although Sandvig and Phillips (2006) found that ponderosa ecosystems had the highest fluxes of all the ecosystems that they studied, it is curious that the ponderosa pine fluxes are at the lower end of the Scanlon et al. (2006) range, especially because ponderosa ecosystems tend to exist in the wetter parts of semiarid climate zones. This observation suggests that the upper bound of the range reported here (2.3 mm yr–1) may be too low or that focused recharge is enhanced in these systems. In other words, additional precipitation (not lost through evapotranspiration) in ponderosa ecosystems may be routed through focused recharge pathways instead of increasing diffuse recharge rates. Ponderosa pines tend to occupy mountain block, mountain front, and upper piedmont positions in southwestern landscapes. The canyons, arroyos, faults, and fractures that are prevalent in these landscape positions act as features where high focused recharge rates occur (Duffy, 2004; Scanlon et al., 2006; Wilson and Guan, 2004). Thus, ponderosa pines may also be useful indicators of basin areas where focused recharge is important.


   Conclusions
TOP
 ABSTRACT
 INTRODUCTION
 Site Description
 Materials and Methods
 Results
 Discussion
 Conclusions
REFERENCES
 
Reviews of semiarid and arid vadose zone and recharge processes (e.g., de Vries and Simmers, 2002; Scanlon et al., 1997, 2002) have emphasized the need to combine physical characterization approaches with multiple tracer methods to understand the complex processes that influence vadose zone flow and transport in dry landscapes. This study demonstrates how valuable such an approach can be. Through the use of environmental tracers (i.e., chloride and stable isotopes) and characterization of geology, hydraulic properties, water content conditions, and contaminant nature and extent, we showed how a mix of ponded (artificially disturbed) and natural (nonponded) surface conditions creates substantial hydrological variability where ponded conditions drive localized areas of multidimensional flow and relatively rapid HE transport. Analyses of chloride data suggest that much of the mesa vadose zone is not conducive for rapid downward transport of HE. However, below wastewater ponds, increased water contents and heads apparently drive localized downward and lateral flow and transport (Fig. 7) as indicated by the presence of HE and water with positive isotopic values within transient saturated zones. The transient nature of flow and transport appears to be seasonally and event controlled and impacts the distribution of HE contamination in the mesa. The behaviors described here may apply to the movement of water and contaminants through vadose zones at other semiarid locations. As Smith et al. (2002) noted, small artificial water bodies are common and important hydrological features, especially in the United States. In addition, ponderosa pine forests are widespread over the southwestern United States, and locations with tuff bedrock occur in many areas.


   ACKNOWLEDGMENTS
 
The authors appreciate the support of the Los Alamos Environmental Restoration project. We also wish to thank Andrew Campbell, Dale Counce, Randy Johnson, Lynn Kidman, Leo Martinez, Johnny Salazar, Tracy Schofield, Chris Sharp, and Donna Sharp for their contributions. Comments by Kay Birdsell, Everett Springer, and two anonymous reviewers helped us make significant improvements to the paper.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 Site Description
 Materials and Methods
 Results
 Discussion
 Conclusions
 REFERENCES
 





This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Newman, B. D.
Right arrow Articles by Gram, P.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Newman, B. D.
Right arrow Articles by Gram, P.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Newman, B. D.
Right arrow Articles by Gram, P.
Related Collections
Right arrow Isotopes
Right arrow Recharge
Right arrow Vadose Zone Processes and Chemical Transport

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Soil Science Society of America Journal
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome