Published online 16 August 2005
Published in Vadose Zone J 4:672-693 (2005)
DOI: 10.2136/vzj2004.0176
© 2005 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
SPECIAL SECTION: LOS ALAMOS NATIONAL LABORATORY
Development of an Infiltration Map for the Los Alamos Area, New Mexico
Edward Kwicklis*,
Marc Witkowski,
Kay Birdsell,
Brent Newman and
Douglas Walther
Los Alamos Natl. Lab., EES-6, MS T0003, Los Alamos, NM 87545
* Corresponding author (kwicklis{at}lanl.gov)
Received 10 December 2004.
 |
ABSTRACT
|
|---|
Using new and previously published estimates of point infiltration in upland areas and estimates of stream-flow losses and gains along canyon bottoms, we created a map of net infiltration for the Los Alamos area, New Mexico for the pre-Cerro Grande fire period. The point infiltration estimates are based on a combination of techniques that include the use of the Richards equation, the chloride mass-balance method, and numerical modeling. The infiltration rates estimated with these techniques were extrapolated to uncharacterized parts of the study area using maps of environmental variables that are correlated with infiltration (such as topography, vegetation cover, and surficial geology and structure) and spatial algorithms implemented with GIS software that use the mapped variables. The map indicates that infiltration rates on mesas of the Pajarito Plateau are generally <2 mm yr–1, except near faults, where infiltration rates may be several tens to hundreds of millimeters per year. Infiltration rates at higher elevations in the Sierra de los Valles are typically >25 mm yr–1 in mixed conifer areas and >200 mm yr–1 in areas vegetated by aspen. An irregular transition zone with infiltration rates between 2 to 25 mm yr–1 exists near the western edge of the Pajarito Plateau adjacent to the Sierra de los Valles. This transition zone extends to lower elevations on the north-facing slopes of deeply incised canyons. Canyon-bottom infiltration rates are highly variable, ranging from several hundred millimeters per year in canyons with large watersheds that have their headwaters in the Sierra de los Valles or in canyons that receive effluent from Laboratory operations, to several millimeters per year in canyons that have their headwaters on the Pajarito Plateau but do not receive Laboratory effluent. The total net infiltration of approximately 10.6 x 106 m3 yr–1 (8600 acre-ft yr–1) is consistent with estimates of the steady-state groundwater discharge to perennial streams in the study area, whereas the relative rates of infiltration within the study area are consistent with the distribution of natural and anthropogenic tracers such as tritium in perched and regional groundwaters. Limitations of the study are that it does not address the effects of the Cerro Grande fire on the hydrology of the study area, nor does it completely capture the complex and sometimes incompletely documented history of Laboratory generated discharges during its 60-yr history.
Abbreviations: GIS, geographical information system • LAC-STP, Los Alamos County Sewerage Treatment Plant • LANL, Los Alamos National Laboratory • RLWTF, Radioactive Liquid Waste Treatment Facility
|
INTRODUCTION
|
|---|
LOS ALAMOS NATIONAL LABORATORY (the Laboratory) in northern New Mexico has been the subject of hydrologic study for more than 50 yr. Early studies dealt primarily with the assessment of water supply and the potential for groundwater development (Conover et al., 1963; Griggs, 1964; Cushman, 1965) and later, with the characterization of infiltration and recharge for the purpose of understanding the risks posed by industrial activities and waste disposal practices at the Laboratory (Abrahams et al., 1962; Purtymun, 1967). These themes continued to be explored during the next four decades, and this paper represents the continuation of that extensive body of work.
We relied on the use of geographical information system (GIS) technology to create a map of infiltration for the LANL area on the basis of new and previously published estimates of infiltration and mapped environmental variables. The basic approach employed in this study was to: (i) compile infiltration estimates that have been made for mountain, mesa, and canyon sites, and augment these with additional estimates; (ii) use GIS to create maps of the variables likely to influence infiltration, including precipitation, topography (elevation, slope, slope aspect), soil cover, vegetation, and rock type; (iii) identify the most important variables controlling the magnitude of infiltration across the Pajarito Plateau; and (iv) use the mapped variables to extrapolate infiltration estimates across the site. It is expected that the infiltration map presented herein will be refined or revised as new data are collected. Nonetheless, the infiltration map developed in this paper is a useful summary of the current state of knowledge and serves as a visual expression of hypotheses regarding infiltration against which new data can be compared and evaluated.
Although the principal area of interest is Los Alamos National Laboratory, the boundaries of the study area were extended outward to encompass natural hydrologic boundaries (Fig. 1)
. These boundaries included (i) the topographic divide between the Valles Caldera and the Pajarito Plateau on the west, (ii) the Rio Grande on the east, (iii) the Santa Clara River watershed boundary on the north, and (iv) the Frijoles Creek watershed boundaries on the south. The topographic divide and watershed boundaries were identified by using a particle tracking option in ARC\INFO that indicated the directions toward which surface water would flow.
The Cerro Grande fire, which burned much of the upland area surrounding Los Alamos in May 2000 has dramatically changed the surface-water hydrology of the study area, at least for the immediate post-fire period (Koch et al., 2001). The exact nature and expected duration of the changes to the subsurface hydrology that resulted from the fire have yet to be determined. However, in the immediate post-fire period, it is reasonable to anticipate that the removal of soil cover and vegetation will result in increased recharge both in the canyons, which will receive more frequent runoff in greater quantities, and in upland areas where water-consuming vegetation has been removed (Wilson et al., 2001; McLin et al., 2001). This paper addresses only the pre-fire infiltration characteristics because most of the existing hydrologic and water quality data pertain to this period and any evidence for past contaminant movement would need to be explained in the context of the pre-fire conditions. Future risk related to contaminant movement would need to be examined in the context of hydrologic data that are only now being collected and analyzed.
This paper begins with a summary of published infiltration estimates that have been made for various areas within the Laboratory. Following this summary, variables likely to influence infiltration are discussed and maps of surficial geology and vegetation cover are presented. Additional estimates of infiltration made in the course of this study on the basis of stream-gage data, the chemistry of high-elevation springs, and 3H profiles follow. A description of the infiltration map and the logic steps involved in its development are then provided. Finally, aspects of the infiltration map are evaluated using independent information not used in its development, such as base flow estimates to major streams and additional geochemical indicators of recharge distribution, such as the 3H, stable isotope, and Cl– composition of groundwater and low-elevation springs.
|
COMPILATION OF INFILTRATION ESTIMATES
|
|---|
Infiltration is defined as the volume of water per unit time per unit area moving downward across the land surface. Because some of this water is lost to plants and to evaporation in the shallow subsurface (i.e., in the root zone), it is more appropriate to make estimates of the deep or "net" infiltration that passes below this depth interval. Where net infiltration moves essentially vertically to the water table, net infiltration is equivalent to recharge, which is defined as the volume of water per unit time per unit area moving downward across the water table. In the Los Alamos area, several factors complicate these relationships. First, alluvial and intermediate depth perched water have been found (LANL, 1998), which indicates that lateral flow along perching layers may influence flow paths in the unsaturated zone. Because of lateral flow, recharge may not be equal to net infiltration at a given location on the map. Second, air flow from adjacent canyons appears to locally dry the rock deep beneath the mesa tops, so that traditional concepts that consider evapotranspiration to be only a root-zone phenomenon may not be applicable beneath the mesas (Newman, 1996; Newman et al., 1997b; Neeper and Gilkeson, 1996).
Most of the infiltration estimates used in this study were compiled from published and internal administrative reports produced by Laboratory personnel (Table 1), with some additional estimates produced by the authors for this study (Fig. 1). The infiltration estimates are based on (i) the chloride mass-balance method, (ii) direct application of Darcy's Law or the Richards equation using measured moisture content data and estimated unsaturated hydraulic conductivities, (iii) model comparisons to measured borehole moisture content profiles, and (iv) water-balance methods. The first three methods have generally been applied to data from wells, whereas water-balance methods have been applied to the watersheds of individual streams (Gray, 1997; Dander, 1998) and to an area approximating the present study area (Keating et al., 1999).
Generally, infiltration estimates produced with the first three methods have indicated that infiltration rates can change with depth beneath the mesa sites because of temporal changes in infiltration rates at land surface, evaporation due to convective air-flow beneath the mesas, or lateral water movement from adjacent high-infiltration canyon bottoms (Newman et al., 1997b). Moisture contents (and estimated infiltration rates) also could change with depth beneath canyon bottoms because of lateral flow along perching layers (LANL, 1998). Where estimated infiltration rates change with depth, the range of infiltration rates is listed in Table 1, but the highest infiltration rate is plotted on Fig. 1. It should be noted that although the plotted values were used to guide the development of the infiltration map, the final infiltration map may differ from the plotted values at the measurement locations.
The methods used to estimate infiltration in this study have different degrees of spatial and temporal resolution. The methods based on deep borehole data have a high degree of spatial resolution, and beneath the mesa top sites are believed to reflect hydrologic processes operating near the borehole over millennia, based on pore-water Cl– ages collected from these boreholes (e.g., Newman, 1996; Newman et al., 1997b). Other methods used to estimate infiltration in this study have coarser spatial resolution but are based on short-term near-surface measurements that may have an uncertain relation to long-term average values. These latter methods include the use of Cl– concentration data to estimate infiltration rates near high-elevation perched springs and the use of stream-flow gaging data to estimate channel infiltration from stream-flow losses. Both of these types of estimates use data that reflect only recent (decade or less) climatic conditions.
It can be noted that some of the highest infiltration estimates are associated with areas that have been identified as "disturbed" sites. These include sites that have been covered with asphalt or are located in constructed drainage diversions (Table 1). (See Birdsell et al., 2005, for more discussion of disturbed sites.) These areas are plotted on Fig. 1, but not considered in the development of the infiltration map because they are not indicative of the general infiltration characteristics and would bias understanding of the influences of natural processes on infiltration. Nonetheless, these anomalies would need to be considered in risk assessments involving the areas that contain them. It should be noted that waste disposal areas, by definition, are also disturbed in that they contain backfill that is different from the unexcavated rock and have had their natural vegetation and topography altered. However, Cl– mass-balance studies of infiltration in these areas have indicated the waste-disposal areas themselves are not necessarily areas of significantly higher infiltration (Newman et al., 1999).
In general, the infiltration estimates listed in Table 1 and plotted in Fig. 1 support the general conceptual model of infiltration for the vicinity of the Laboratory, which has developed in the past several decades (LANL, 1998; Birdsell et al., 2005). That conceptual model holds that infiltration rates are very low on the mesas of the Pajarito Plateau and increase in the Sierra de los Valles to the west of the Laboratory. Canyons on the plateau can have variable amounts of infiltration, depending on the size and elevation of the watershed contributing runoff to the canyon, and on the history of liquid waste disposal practices in the upstream reaches of the canyon.
|
MAPS OF ENVIRONMENTAL VARIABLES LIKELY TO INFLUENCE INFILTRATION
|
|---|
To help extrapolate the existing infiltration estimates summarized in Table 1 as well as new infiltration estimates developed later in this paper to uncharacterized parts of the study area, maps of environmental variables that might influence or be correlated with infiltration were prepared. Other studies of infiltration patterns in the southwestern USA have indicated that net infiltration is potentially influenced by a variety of environmental variables, including precipitation frequency and magnitude, antecedent soil moisture conditions, soil type and thickness, vegetation cover, elevation, slope and slope aspect, and bedrock type (e.g., D'Agnese et al., 1997; Flint et al., 2000, 2001; Walvoord et al., 2002a, 2002b). Maps showing the areal distribution of a number of these variables were produced during this study. A brief discussion of these maps follows.
Topography, Including Slope and Slope Aspect
Topography (Fig. 1) is likely to influence infiltration in a number of ways. First, because of orographic effects, precipitation increases with elevation. Second, plant communities generally change with elevation and with slope and slope aspect. The Hydrologic Work Plan for the Laboratory (LANL, 1998, p. 2–11) notes that south-facing canyon walls are steep and have little soil or vegetation, whereas north-facing slopes are gentler, more heavily vegetated, and have areas of shallow, dark-colored soils. Qualitatively, it is expected that north-facing slopes have higher infiltration rates than south-facing slopes because they receive less solar radiation to evaporate moisture, as evidenced by the persistence of snow later in spring on the north-facing slopes. Approximately 25% of the study area has north-facing slopes and 35% of the study area has south-facing slopes, so slope aspect is potentially a significant factor influencing infiltration.
Precipitation
On the basis of analysis of precipitation data collected on the Pajarito Plateau and the Sierra de los Valles, Rogers (1994) developed a relation between average annual precipitation and elevation for the Los Alamos area:
 | [1] |
This relation between average annual precipitation and elevation (r2 > 0.99) was used to create a map of average annual precipitation for the study area. Average annual precipitation increases from 330 to 460 mm over most of the Pajarito Plateau to more than 760 mm over the highest elevation areas near the topographic divide between the Valles Caldera and the Pajarito Plateau. The large contrast in precipitation rates with elevation in the study area suggests that infiltration also increases with elevation.
Almost one-half of the annual precipitation in the Los Alamos area results from summer thunderstorms during July through September (Bowen, 1992; Rogers, 1994). High-elevation areas west of Los Alamos receive a higher percentage of their annual precipitation in the form of snow, so that the contribution of summer monsoons to annual precipitation is somewhat less (
40%). The snow falling on the Pajarito Plateau generally does not persist for more than a few days because of the generally sunny conditions and mild climate (Bowen, 1992). At higher elevations, snow persists until spring when it melts and becomes available for infiltration and runoff. Precipitation data collected at Los Alamos between the 1940s and the late 1990s show a trend of increasing precipitation. During this period, annual precipitation at the town of Los Alamos averaged about 470 mm and varied between 173 mm (1956) and 719 mm (1957). As elsewhere throughout the southwestern USA, the impact of long-term climate change on precipitation patterns in the study area is uncertain.
Soils
Only some of the soil classification data necessary to create a soil-cover map of the study area have been obtained (Marvin Gard, 2002, personal communication). Quantitative data concerning the thickness of the soils where soil cover is present are not available. Qualitatively, almost all the soils in the Los Alamos area (pre-Cerro Grande fire) are characterized as well-drained soils, and only 4 of the 30 or so soils classified in the Los Alamos area cover more than 5% of the mapped area of Los Alamos County (LANL, 1998, Table 2-1; Nyhan et al., 1978). Because of the apparent similarity in soil characteristics and the fragmented spatial distribution of the soils, differences in soil type may not exert as strong an influence on infiltration as the mere presence or absence of soil. The absence of soils also indicates the absence of water-consuming vegetation, so infiltration would be expected to be higher in areas with no soil, other factors being the same. Although not directly used in the creation of the infiltration map described here, the effects of soil type may have been indirectly incorporated into the development of the infiltration map through the use of vegetation. For, example, Newman et al. (1997a) reported that soils in areas vegetated by Ponderosa Pine (Pinus ponderosa C. Lawson) had a better-developed clay layer than areas vegetated with piñon (Pinus edulis Engelm.) and juniper [Juniperus monosperma (Engelm.) Sarg.].
Bedrock Geology
In addition to climatic variables, rock type may influence net infiltration rates in several ways. As mentioned above, almost all the soils in the Los Alamos area (pre-Cerro Grande fire) are characterized as well-drained soil, so differences in infiltration rates among areas may be more related to the hydrologic characteristics of the underlying bedrock than to the apparently uniform hydrologic characteristics of the soils. One important distinction among rock types may be their different tendencies to fracture. To help understand and interpret infiltration patterns, a map of the geology for the study area was created (Fig. 2
, modified from Cole et al., 1998). Fractured rocks in the study area include moderately to densely welded tuffs such as units Qbt2 and, possibly, unit Qbt4 of the Tshirege Member of the Bandelier Tuff; basalts such as Bayo Canyon (Tb2) and Cerros del Rio (Tb4) basalts; silicic lavas such as the Tschicoma dacite (Tt1) and Keres Group (Tk); and fault zones such as the Pajarito, Rendija Canyon, and Guaje Mountain fault zones (Fig. 2). Welding in the tuffs, and thus fracturing, tends to increase toward the western margin of the plateau because of increasing thickness of the ash flows near the volcanic source (Broxton and Vaniman, 2005). However, at present, it is not entirely clear how or if fractures in the surficial bedrock affect infiltration. On the mesas, fractures may actually reduce net infiltration by enhancing air circulation that removes moisture from the rock. Alternatively, beneath canyon bottoms, where air flow is probably negligible, and perhaps on the mesas as well, fractures might allow the relatively rapid drainage of newly infiltrated water away from the shallow zone of evapotranspiration, thereby increasing net infiltration. There is also evidence from models of water-injection tests that fractures do not substantially enhance water movement in rocks with high matrix permeability (Robinson et al., 2005).
Major faults in the study areas include the Pajarito Fault Zone, which coincides with the transition between the Sierra de los Valles and the Pajarito Plateau throughout much of the area, and the Guaje Mountain and Rendija Canyon Faults (Fig. 2). The character of the Pajarito Fault zone varies along strike in the vicinity of LANL (Gardner et al., 2001). In the northern part of the Laboratory, the fault zone consists of a set of anastamosing faults and associated grabens. Toward the south, near-surface rocks such as the Bandelier Tuff are draped over a fault scarp to form a monoclinal structure with gaping tension cracks at the surface. Likewise, rocks near the Guaje Mountain and Rendija Canyon faults have associated zones of fracturing about 200 m wide that probably increase their permeability relative to unfractured rocks (Gray, 1997, p. 17). Measurements of hydraulic head in the alluvium of Los Alamos Canyon indicate that the fractured rocks between the Guaje Mountain and Rendija Canyon fault zones act as a drain for perched water in the alluvium (Gray, 1997, Fig. 10). Based on these observations, areas of enhanced infiltration potentially exist along fault zones, especially where the faults intersect wet canyons.
Vegetation
A map showing the distribution of major vegetation zones within the study area was created based on data provided by Steve Bromby (personal communication, 2001) (Fig. 3)
. Vegetation may be an especially useful surrogate for mapping and extrapolating infiltration estimates because vegetation patterns respond to many of the same factors that are expected to influence infiltration, namely elevation (which influences precipitation and temperature-dependent potential evaporation), slope and soil cover (which influences water-retention and runoff), and slope aspect (which also influences evaporation). The vegetation map presently contains some areas that have not been classified for vegetation type. Some areas, such as the blue area in the southwestern corner of the study area, were obscured by cloud cover at the time the aerial data were obtained. Because the aerial data used to create the vegetation map were obtained after the Cerro Grande fire, the vegetation map shows the area severely burned during the fire. Given the prevalence of mixed conifers at these elevations, these areas were assumed to be vegetated by mixed conifers before the fire. The amount of water normally consumed by vegetation in this area is now presumably available for infiltration or runoff.
|
ADDITIONAL INFILTRATION ESTIMATES
|
|---|
Infiltration estimates compiled from laboratory reports (Table 1) were spatially concentrated near mesa-top areas of the Laboratory where waste-disposal and other industrial operations warranted detailed characterization of infiltration. To achieve balance in other topographic and geographic areas, these estimates were supplemented by infiltration estimates developed as part of this study. The effort in this study focused on the use of stream flow measurements made in the late 1990s to estimate infiltration on canyons on the Pajarito Plateau and on the use of the chemical characteristics of high elevation springs to characterize infiltration in the Sierra de los Valles west of the Laboratory.
Stream-Flow Losses and Gains
To estimate the quantity of focused infiltration within canyons on the Pajarito Plateau, average annual stream-flow losses and gains were calculated based on flow measurements for water years 1995 thru 2000 (Shaull et al., 1996a, 1996b, 1998, 1999, 2000a, 2000b). These estimates do not consider measurements made after the Cerro Grande Fire of May 2000, which greatly increased the magnitude of runoff compared to runoff events that occurred before the fire (Shaull et al., 2000b).
The average annual flow at each station and the flow increment between adjacent stations defining a reach of the stream are shown in Fig. 4
. (Note that stream-flow losses are positive and stream-flow gains are negative on Fig. 4). On average, the largest stream-flow losses occur in Pajarito Canyon between Stations E245 and 250 (110000 m3 yr–1), in lower Los Alamos Canyon between Stations E030 and E042 (81000 m3 yr–1), and in Mortandad Canyon between Stations E200 and E204 (44900 m3 yr–1). Stream-flow measurements at Station E200 in Mortandad Canyon reflect laboratory discharge into this canyon. Smaller stream-flow losses also could be documented in upper Los Alamos Canyon between Stations E025 and E030 (19600 m3 yr–1) and in Water Canyon between Stations E252 and E265 (12000 m3 yr–1). The Hydrogeologic Workplan indicates that the lower part of Pueblo Canyon is supported by discharge from the Los Alamos County Sewerage Treatment Plant (LAC-STP) (LANL, 1998, Fig. 2-7). A detailed discharge history was not available from the LAC-STP, but average annual discharge was estimated to be about 960000 m3 yr–1 (LAC-STP, 2002, personal communication). Therefore, based on the stream flow at Station E060 in lower Pueblo Canyon, approximately 55500 m3 yr–1 of effluent seeps into the alluvium in lower Pueblo Canyon between the LAC-STP and Station E030 in a typical year.
View larger version (125K):
[in this window]
[in a new window]
|
Fig. 4. Stream-gage network for the study area, with average annual stream flow volumes and incremental stream-flow losses (+) or gains (–).
|
|
The calculations indicate that the upper part of Pajarito Canyon between Stations E240 and E245 is a gaining reach (35000 m3 yr–1). This apparent gain is probably from ungaged flow on Two-Mile Canyon, which intersects Pajarito Canyon from the north and contributes to the flow measured downstream at Station E245. Like the upper reach of Los Alamos Canyon, the upper parts of both the north and south forks of Pajarito Canyon probably are locations of net annual stream-flow loss.
The total calculated stream-flow loss of 323000 m3 yr–1 provides only a general indication of the magnitude of canyon bottom infiltration on the Pajarito Plateau. First, many canyons outside the Laboratory property are ungaged, so the stream-flow losses in these canyons are unknown. Second, even where stream flow measurements are available, the relation between stream-flow losses and actual net infiltration is uncertain. The actual canyon bottom infiltration could be less than the stream-flow losses if part of these losses is consumed by evapotranspiration; alternatively, canyon-bottom infiltration could be somewhat greater than the calculated stream-flow losses if runoff is generated on the plateau below the upper gage defining the reach. The importance of evaporation in reducing channel infiltration is uncertain. However, with regard to runoff from the plateau itself, measurements indicate that stream flow is small in canyons that have their headwaters on the plateau (and do not receive effluent discharge from the Laboratory), compared with canyons that receive runoff from the Sierra de los Valles. For example, average annual flow past Station E255 in Potrillo Canyon is only 2600 m3 yr–1. Measurements such as this indicate that very little runoff is generated from the Pajarito Plateau itself.
Monthly stream flow data can be indicative of watershed processes affecting stream flow in various canyons and stream reaches. For example, monthly stream-flow data from the four stations along the western boundary of the laboratory at the foot of the Sierra de los Valles are shown in Fig. 5a
. The data indicate that most runoff occurs during spring snowmelt, with a secondary peak in monthly stream flow occurring in late summer or early autumn as a result of summer thunderstorms. In contrast, streams in canyons that have their headwaters within the Pajarito Plateau do not flow during spring snowmelt, but flow only following intense summer thunderstorms in July, August, and September (Fig. 5b). Portions of streams that receive effluent, like lower Pueblo Canyon (Station E060), flow in response to effluent discharges, and the stream flow in these canyons has little or no correlation to flow in canyons dominated by natural watershed processes (Fig. 5c).
View larger version (34K):
[in this window]
[in a new window]
|
Fig. 5. Monthly stream-flow data for gages located (a) in the upper parts of canyons that have their headwaters in the Sierra de los Valles (b) in canyons that have their headwaters on the Pajarito Plateau, and (c) in canyons where runoff is dominated either by natural runoff or by effluent.
|
|
The monthly stream-flow data also suggest that focused recharge in the upper reaches of the canyons west of the Laboratory may be secondary compared with diffuse upland recharge. During high-flow periods associated with spring snowmelt, when water for recharge is most plentiful, canyon bottoms west of the Laboratory may be discharging rather than recharging water. Hydraulic head measurements made in alluvial groundwater in upper Los Alamos Canyon west of the Rendija Canyon and Guaje Mountain fault zones do not change with season, an observation that suggested to Gray (1997) that heads in the upper part of the canyon are sustained by base flow throughout the year. In contrast, heads measured in alluvial groundwater downstream from the faults showed seasonal variability in a manner consistent with stream-flow losses. These interpretations were supported by a series of closely spaced stream-flow measurements in Los Alamos Canyon during spring runoff in May 1995 that identified gaining and losing reaches of the channel (Gray, 1997, p. 64–66).
This behavior is reflected by stream-flow data from the reach in Los Alamos Canyon defined by Stations E025 and E030 (Fig. 6) , which shows that with increasing flow, stream-flow losses first level off and then decline until the reach becomes a gaining reach. (Note that stream-flow losses are positive and stream-flow gains are negative in Fig. 6.) The transition from losing to gaining stream occurs when lenses of perched groundwater derived from snowmelt migrate down canyon through the alluvium along less permeable bedrock and perched-water levels eventually rise and intersect the channel bottom, discharging water to the stream (Gray, 1997). Similar behavior has been observed in Cañon de Valle (LANL, 2003). Although observations of this behavior are limited to these canyons, it is reasonable to infer that gaining reaches in the upper parts of other canyons draining the Sierra de los Valles expand down canyon to varying degrees during snowmelt, depending on the depth of the year's snowpack and the rate of the spring thaw. Presumably, these gaining reaches contract and retreat up canyon during the late autumn when precipitation rates and runoff are low and alluvial water from the previous spring snowmelt has drained into the bedrock.
View larger version (21K):
[in this window]
[in a new window]
|
Fig. 6. Stream-flow losses (positive) and gains (negative) in Los Alamos Canyon as a function of stream-flow rates.
|
|
To provide a basis for distributing stream-flow losses downstream of the lowermost gages, or along ungaged canyons in the study area (see below), average stream-flow loss rates were calculated by dividing the streamflow losses shown in Fig. 4 by the length of the reach. These rates of stream-flow loss were compared with the geologic units underlying the reach to determine if the bedrock geology influences the stream-flow losses (Table 2). The reaches underlain by various geologic subunits of the Bandelier Tuff (Qb) show a wide range of stream loss rates (3.0–24.6 m3 m–1 of channel length), with no obvious correlation between the geologic subunits and the loss rate. The reach defined by the LAC-STP and Gage E060 in Pueblo Canyon is underlain entirely by the Puye Fanglomerate (Tpf) and has a loss rate of 24.6 m3 m–1, considerably higher than the average of the reaches underlain by tuff (13.1 m3 m–1). In estimating stream-flow loss rates downstream of either real or hypothetical stream gages, loss rates of 3.0, 13.1, or 24.6 m3 m–1 of channel length were used to distribute runoff along canyon bottoms. (Note that loss rates of 3.0 m3 m–1 were used for reaches that, like Reach E025/E030 in upper Los Alamos Canyon, were adjacent to the mountain front.) The stream-flow loss rate, along with the canyon-bottom width, determined the infiltration rate; the measured or estimated runoff volume, along with the loss rate, determines the length of affected channel downstream of the gage.
Estimation of Mountain-Front Stream Flow on Ungaged Canyons
Along the western boundary of the Laboratory, the amounts of runoff vary from essentially none in lower Cañon de Valle (Station E253) to an average of (257000 m3 yr–1) in Los Alamos Canyon (Station E025) (Fig. 4). Because runoff from the Sierra de los Valles represents a potentially significant source of recharge on the Pajarito Plateau itself, the watershed characteristics above each station were examined to see if differences in these characteristics could explain the vastly different amounts of runoff generated in each of the watersheds. Identification of the important watershed characteristics controlling runoff generation could provide a basis for estimating mountain runoff in ungaged watersheds.
Assuming that each watershed has equivalent hydrologic characteristics, yearly runoff volumes should plot as a linear function of precipitation volume. The use of precipitation volume integrates the effects of differences in area and elevation among the watersheds, but does not consider differences in vegetation, slope, or bedrock type. Runoff versus precipitation volume data from upper Los Alamos Canyon (E025) and upper Pajarito Canyon (E240), when extrapolated through the origin, define a roughly linear relation (Fig. 7)
. Relative to those drainages, Canyon de Valle (E253) and Water Canyon (E252) have far too little runoff for their precipitation volumes. The smaller amounts of runoff from Canyon de Valle and Water Canyon would be expected to result if these drainages (i) had relatively larger amounts of infiltration due to more permeable bedrock or smaller slopes, or (ii) were dominated by vegetation that required significantly more water than the other drainages. A comparison of the statistics on slope, slope aspect, bedrock type, elevation, and vegetation calculated for each of these drainages indicated that all drainages had similar distributions of these attributes, with the exception that the lower part of the Water Canyon watershed above Station E252 is underlain by (presumably) permeable Puye Formation gravels. Although drainage of runoff into these gravels would explain the absence of runoff in Water Canyon, it does not explain the runoff behavior of Cañon de Valle, where only a small amount of the Puye Formation is present. Thus, differences in watershed characteristics alone do not provide an explanation for the differences in the runoff volumes from these four drainages.
One additional possible explanation for the unexpected differences in runoff measurements is that variable amounts of surface runoff drain down the Pajarito Fault Zone before reaching the gaging stations. The gaging stations in all four drainages are located at the foot of the Sierra de los Valles within or just downslope of the Pajarito Fault Zone. However, the nature of the fault zone in the two southern canyons is somewhat different than in the north. Toward the north, the fault zone is wider and characterized by anastamosing fault strands and small grabens. In the vicinity of Water Canyon and Cañon de Valle, the fault zone is characterized by a steep, narrow scarp. Near-surface rocks draped over this scarp have a monoclinal structure characterized by wide (>1 m) tension cracks at the ground surface where the flexure is greatest (Gardner et al., 2001; Lewis et al., 2002). These near-surface tension cracks could facilitate the infiltration of surface runoff in Water Canyon and Cañon de Valle where these canyons cross the fault zone. Under the assumption that the relatively small amount of stream flow leaving the mountains in these canyons has mostly infiltrated as it crossed the Pajarito Fault Zone, it is assumed that mountain-front stream flow on ungaged canyons is approximately 2.5% of the total precipitation falling on a watershed (Fig. 7).
Estimating Recharge Using Chemistry Data from High Elevation Springs
Previous infiltration studies for the Los Alamos area (e.g., Rogers et al., 1996a, 1996b; Newman et al., 1997a, 1997b) have tended to focus on characterizing infiltration rates for areas on the Pajarito Plateau itself and not on mountainous areas of the Sierra de los Valles west of the Laboratory. Although some interpretations of stable isotope data suggest that much of the groundwater beneath the Los Alamos area was recharged at high elevations west of the Laboratory (e.g., Blake et al., 1995), quantitative estimates of the infiltration rates in this part of the study area have not been made. To obtain quantitative estimates of infiltration rates in these high elevation areas, the Cl– concentrations of spring discharge in these areas were used in conjunction with the Cl– mass-balance method. A similar approach was used to characterize high-elevation infiltration rates in south-central Nevada (Russell and Minor, 2002).
The Cl– mass-balance method assumes that the Cl– flux arriving at the land surface in precipitation and as a result of dry fallout is equal to the downward Cl– flux below the root zone:
 | [2] |
where Q is flux rate (L T–1), C is Cl– concentrations (mg L–1), and the subscripts "rech" and "prec" indicate recharge and precipitation, respectively. As indicated by Eq. [2], the fraction of precipitation that becomes net infiltration (or recharge) is simply the ratio Cprec/Crech.
Previous Cl– mass-balance estimates of infiltration in the Los Alamos area (e.g., Newman et al., 1997a, 1997b) have assumed that the Cl– concentration of precipitation on the Pajarito Plateau and Sierra de los Valles is the same as the value of 0.29 mg L–1 measured by Anderholm (1994) at the Santa Fe airport for a 2-yr period in the late 1980s. As part of the present study, precipitation chemistry data from Adams et al. (1995) were analyzed to verify the applicability of this value to infiltration studies on the Pajarito Plateau. Based on data from all precipitation monitoring stations discussed in Adams et al. (1995), the average Cl– concentration of precipitation averages 0.30 mg L–1 within the Española Basin, which is very close to the average concentration reported by Anderholm (1994). Using only data from monitoring stations within the infiltration study area (Fig. 1), the Cl– concentration of precipitation ranges from 0.08 to 0.68 and averages 0.26 ± 0.17 mg L–1 (1 SD). No trends between station mean Cl– concentration and elevation, or between station mean Cl– concentration and station precipitation volumes could be identified in this analysis. Because the Cl– concentrations of local precipitation are statistically indistinguishable from the value measured by Anderholm (1994), the Anderholm value of 0.29 mg L–1 is used by the authors in their Cl– mass-balance estimates to maintain consistency with earlier infiltration estimates on the Pajarito Plateau.
Chloride concentration data from high-elevation springs in the Sierra de los Valles were used to estimate infiltration in upland areas west of the Laboratory (Table 3). The infiltration estimates are based on the Cl– mass-balance method and precipitation estimates at the recharge elevations of these springs calculated from the delta deuterium (
D) and delta oxygen-18 (18O) of the spring discharge (Blake et al., 1995). These high elevation springs discharge from a variety of local topographic positions, including canyon bottoms, sideslopes, and ridgetops (Fig. 1). The chemical and isotopic characteristics of the spring discharge are assumed to be representative of deep infiltration in these drainages. In other words, the spring discharge is similar to water that otherwise would have recharged the groundwater had it not been forced back to the ground surface for reasons related to topography and local permeability variations.
The Cl– concentrations ranged from a low of 0.69 mg L– at Water Canyon Gallery to a high of 14 mg L– at Frijoles Spring 49 (Table 3). The corresponding infiltration estimates ranged from a high of 259 mm yr–1 at Water Canyon Gallery to a low of 10.9 mm yr–1 at Frijoles Spring 49. The very low Cl– concentrations (<2 mg L–1) measured in the discharge of some of the high elevation springs are also characteristic of some groundwater from the regional aquifer on the Pajarito Plateau at Wells R-7, R-25, and R-31 (Longmire and Goff, 2002; Broxton et al., 2002; Vaniman et al., 2002), supporting the assumption that these spring compositions are representative of recharge.
Possible explanations for the variability in the Cl– concentrations of the high-elevation springs in the Sierra de los Valles were investigated by comparing the various mapped attributes at the spring locations. Generally, the high 3H content of the spring discharge indicates that the water infiltrated within the last few decades, and consequently, reflects recent climatic conditions. The high 3H contents also suggest that the infiltration probably did not travel far before being discharged at the springs, and therefore, that it is reasonable to examine local environmental factors in the vicinity of the springs as a possible explanation for differences in their estimated infiltration rates. It is possible, however, that for springs that discharge groundwater with very short residence times, seasonal variations in the chemistry of the discharge could exist. Data are presently inadequate to evaluate the seasonal variability in spring compositions or determine if the existing Cl– and stable isotope data, most of which were obtained in late spring, are representative of yearly averages.
The chemistry and flow rate data for high-elevation springs are listed in Table 3. Also listed in Table 3 are the estimated recharge rates and the fraction of precipitation that becomes recharge. As mentioned earlier, this fraction (Xp) is simply the ratio Cprec/Crech. Water Canyon Gallery, where the highest rates of recharge are estimated (259.3 mm yr–1), is located in an area forested by aspen (Populus tremula L.) and underlain by gravels of the Puye fanglomerate (Tpf). Armistead Spring has 156.6 mm yr–1 of infiltration and is underlain by the Tschicoma lava flows (Tt1), but is also forested by aspen. The high rates of infiltration at these springs relative to the others may be due to the presence of aspen, and secondarily, to the presence of the relatively permeable Puye Fanglomerate at the Water Canyon Gallery. The larger slope at the Water Canyon Gallery would tend to promote more runoff and less infiltration, and the southern slope aspect would tend to promote more evapotranspiration and less infiltration at Water Canyon Gallery relative to Armistead Spring. Therefore, slope and slope aspect are probably minor controls on infiltration near these springs. Slope and slope aspect also do not seem to be important controls based on data from other springs, which show that estimated infiltration rates for areas with similar slopes and slope aspects can vary widely. The infiltration rates estimated at the four springs in areas covered by mixed conifers range from 10.9 to 29.2 mm yr–1 and average 20.5 mm yr–1. These springs are all in areas underlain by the Bandelier Tuff. The estimated infiltration rate of 98.1 mm yr–1 at Pine Spring is substantially higher than other areas on the plateau vegetated by Ponderosa Pines (Newman et al., 1997a). One environmental factor that distinguishes Pine Spring is the proximity of the Rendija Canyon Fault through the brittle lavas of the Keres Group (Tk) formation. The nearby presence of this major fault through these lavas may have enhanced local infiltration rates near Pine Spring and produced infiltration rates that are relatively high compared with other areas with similar vegetation.
In developing the map of net infiltration for the study area, the fraction of precipitation that becomes net infiltration (Xp) is used, rather the calculated infiltration rates shown in Table 3. The use of Xp eliminates the need to know the precise recharge elevation for the springs in Table 3. For the nominal value for Cprec of 0.29 mg L–1, the values of Xp in aspen are 0.42 and 0.30, respectively. Because the Water Canyon Gallery is an engineered system, the value for Xp of 0.30 is considered more representative of natural recharge processes in aspen-vegetated areas. The range of Xp in mixed conifer areas varies between 0.023 (Homestead Spring) and 0.22 (unnamed spring near Apache). The discharge weighted Cl– concentration of the springs in mixed conifer areas of 6.4 mg L–1 is consistent with a value for Xp = 0.05. Values for Xp of 0.30 and 0.05, along with precipitation estimated from Eq. [1], was used to estimate net infiltration in areas dominated by aspens and mixed conifers, respectively.
Infiltration Rate Estimates from 3H and Moisture Content Data in Mortandad Canyon
Mortandad Canyon is an area of the map where reliable infiltration estimates are especially important because of past waste-disposal practices and the proximity of water-supply wells to this canyon. Based on stream gage records from 1995 to 2000, average infiltration across several kilometers of this canyon is currently estimated to be about 176 mm yr–1. However, historic data suggest that infiltration rates were higher in the past. Mortandad Canyon has received treated effluent from the Radioactive Liquid Waste Treatment Facility (RLWTF) at TA-50 since 1963 (Rogers, 1998). Recharge from the liquid waste was estimated to average about 60000 m3 yr–1 based on a 3-yr water balance study between 1963 and 1965 (Purtymun, 1967; Dander, 1998, Table 3.2), resulting in an equivalent water depth of recharge of 4.5 m yr–1 into the channel bottom along roughly the upper third of the reach defined by Gages E200 and E204 (Fig. 4).
The variations in discharge volumes and 3H concentrations discharged from the RLWTF and earlier facilities located in Mortandad Canyon are shown in Fig. 8a
(LANL Environmental Restoration Project, 1997, Tables 2.4.4-2 and 2.4.6-1). The changes in the 3H concentrations of the discharge were reflected by almost simultaneous but more subdued changes in the 3H content of alluvial groundwater downstream in Mortandad Canyon (Rogers, 1998, Fig. 9). One shallow well in Mortandad Canyon (MCO-6) measured two prominent increases and two secondary increases in the 3H contents of alluvial groundwater, with the prominent increases occurring in 1976 and 1986 and the secondary increases occurring in 1970 and 1981 (Fig. 8b). Well MCO-6 is several hundred meters upstream from regional Well R-15 (Fig. 1), but is near enough that 3H measurements at Well MCO-6 can be used to estimate the 3H contents of alluvial groundwater infiltrating the bedrock at Well R-15.
View larger version (35K):
[in this window]
[in a new window]
|
Fig. 8. Plots showing (a) historic variations in discharge volumes from TA-35 WWTP and discharge volumes and tritium concentrations from the RLWTF in TA-50 and (b) temporal variations in TA-50 discharge and alluvial groundwater in Mortandad Canyon wells MCO-5 and MCO-6.
|
|
Tritium data from the upper 128 m of Well R-15 display three distinct peaks at depths of 21, 52, and 107 m (Longmire et al., 2000) (Fig. 9)
. Assuming that variable 3H content in Well R-15 results from the variable 3H contents of the overlying alluvial groundwater as measured in nearby Well MCO-6, and specifically, that the 3H peaks at 21, 52, and 107 m correspond to the increases in the 3H content of alluvial groundwater in 1986, 1981, and 1976, respectively (Fig. 8b), the linear transport velocity between the peaks can be calculated. The assumption that the earliest 3H peak observed in 1970 in alluvial groundwater in Well MCO-6 has already moved through the Otowi Member into the underlying basalts is supported by 3H concentrations of 4000 pCi L–1 in perched water at 190 m in Well R-15 (Longmire et al., 2000).
The linear transport velocity between the 107- and 52-m peaks is about 11 m yr–1. For an average volumetric moisture content of 0.215 for this depth interval, the corresponding darcy flux is 2350 mm yr–1. A similar calculation for the 3H peaks at 21 and 52 m yields a linear transport velocity of 5.1 m yr–1, a value that corresponds to a darcy flux of 1020 mm yr–1 using an average moisture content of 0.20 for this depth interval. Transport time from the base of the alluvium at 5.0 m to the 1986 3H peak at 21 m was 12 yr (the core was taken in Sept. 1998). The resulting linear transport velocity of 1.36 m yr–1 results in a darcy flux of 180 mm yr–1 using an average effective moisture content of 0.13 over between the bottom of the alluvium and the 3H peak at 21 m. The estimated change in flux with time is plotted in Fig. 10
. Also plotted in the figure is the estimate of 4500 mm yr–1 made by Purtymun (1967) and the estimated infiltration rate of 176 mm yr–1 for the 1995 to 2000 calculated from stream gage data assuming the stream-flow losses are distributed across the width of the canyon. The estimates provide a consistent indication that infiltration rates have declined substantially in Mortandad Canyon since the 1960s from a high of >2000 mm yr–1 to a present day flux of 150 to 200 mm yr–1. This trend is consistent with the trend in laboratory discharge in this canyon (Fig. 8a).
|
DEVELOPMENT AND DESCRIPTION OF THE INFILTRATION MAP
|
|---|
The measurements and infiltration estimates described in the previous sections were used along with the mapped geologic and vegetation characteristics to create the infiltration map shown in Fig. 11
. The step-by-step development of the map is documented in the Appendix. The following sections provide a general description of the map and its development.
Map Development
Using ARC/INFO GIS software, the study area of approximately 232 mi2 (600 km2) was first discretized into slightly more than two and one-half million 15.24 by 15.24 m (50 by 50 foot) cells. This grid resolution allowed adequate representation of canyons, slopes, and mesa tops within the study area without requiring excessive computer resources. Then, based on the maps of environmental variables discussed earlier, each cell was assigned a set of environmental attributes related to topography, geology, and vegetation type. Infiltration rates for a cell were calculated using the environmental attributes associated with the cell and one or more of a series of 30 distinct GIS logic steps that relate infiltration to these attributes (Table A-1 in the Appendix). To summarize, these GIS logic steps used (i) estimates of diffuse infiltration associated with different vegetation types, as estimated from Cl– mass-balance studies of high-elevation springs and soil pore waters, (ii) measured stream-flow losses and estimated stream-flow loss rates, (iii) estimated mountain runoff and stream-flow loss rates in ungaged watersheds, and (iv) rock-type dependent infiltration enhancement factors along fault zones. The total and incremental infiltration for the study area at each step are listed in Table A-1.
Map Description
The infiltration map indicates dramatic contrasts in infiltration rates between the Pajarito Plateau and the higher elevation areas of the Sierra de los Valles (Fig. 11). Infiltration rates throughout most of the plateau are generally <2 mm yr–1, whereas infiltration rates in the mixed conifer dominated areas of the Sierra del Valles are typically >25 mm yr–1 and, in the aspen dominated areas, >200 mm yr–1. Although most of the aspen dominated areas occur at the highest elevation areas near the topographic divides that define the western boundary of the study area, some lower elevation areas in the Water Canyon and Cañon de Valle areas also have high infiltration rates that reflect the presence of aspen.
An irregular transition zone with infiltration rates of 2 to 25 mm yr–1 occurs between the areas with low infiltration rates that typify most of the Plateau and the areas in the Sierra de los Valles with high infiltration rates. Most of this transition area is vegetated with mixed conifers, which were assigned an infiltration rate that is 5% of the precipitation rate. Mixed conifers and aspens can also exist at lower elevations on north-facing slopes of deep canyons, where they are partially shielded from the sun, thereby reducing evaporation and increasing effective moisture. These mixed conifer and aspen dominated areas are reflected on the infiltration map as elongated, west-to-east trending zones along the southern (north-facing) margins of drainages with infiltration rates >2 mm yr–1.
The western one-third to one-half of Guaje, Rendija, and Santa Clara Canyons in the northern part of the study area are estimated to have infiltration rates on the order of 1600 mm yr–1. Despite the high stream-loss rates assumed for these canyons, the mountainous portions of their watersheds are large enough that the estimated runoff of 2.5% of mountain precipitation affects parts of the canyons several kilometers from the mountain front. In contrast, despite the smaller rates of stream-flow loss estimated for the Bandelier Tuff, the relatively small size of the mountainous portion of the Frijoles Canyon watershed results in small amounts of runoff that is estimated to travel, on average, not more than a couple of kilometers from the mountains.
The insert in the lower left-hand corner of the map highlights the areas of focused infiltration that occur in Water Canyon and Canyon de Valle where these canyons cross the Pajarito Fault Zone. Where associated faults coincide with aspen dominated areas and brittle rocks, localized infiltration rates assume values in excess of 1000 mm yr–1. For other vegetation types and nonbrittle rocks, faults have infiltration rates of 200 to 500 mm yr–1. Overall, the area shown in this insert can be characterized as a location with a high potential for focused infiltration.
The insert for the area in the east-central part of the map highlights lower Los Alamos and Pueblo Canyons. Stream flow in Los Alamos Canyon is controlled by natural runoff, whereas stream flow in lower Pueblo Canyon is affected by discharge from the LACSTP. The infiltration rates of approximately 1500 to 2000 mm yr–1 estimated for these canyons where they merge near State Highway 4 reflects the large stream-flow losses attributed to canyon bottoms underlain by the Puye fanglomerate or fractured basalts. Also evident in this insert is the infiltration rate of 176 mm yr–1 estimated for Mortandad Canyon and the higher infiltration rates estimated for north-facing slopes vegetated with mixed conifers in Mortandad, Los Alamos, and Pueblo Canyons.
Overall, the high elevation areas with mixed conifers or aspens account for slightly more than half (51.8%) of the total infiltration in the study area (10.6 x 106 m3 yr–1) (Table A-1). Other volumetrically significant sources of recharge are mountain runoff in Santa Clara Canyon (9.3%), treated sewerage discharge in lower Pueblo Canyon (8.7%) and focused recharge along faulted areas (14.9%), especially mesa top areas underlain by brittle rocks (11.5%).
|
MODEL EVALUATION
|
|---|
In view of the very limited amount of data available for certain aspects of the infiltration map, it is necessary to evaluate the map by comparing it to other hydraulic and geochemical observations not used directly in its construction. These aspects include the estimates of base flow on perennial streams, Cl– mass-balance estimates of recharge from low-elevation spring data not used in the map's development, anthropogenic and natural tracers such as tritium in the groundwater that indicate rapid transit times, as well as hydraulic head and moisture content data.
Base Flow Estimates
Under the assumption that groundwater flow within a watershed is approximately at steady state, measurements of stream flow when it is sustained primarily by base flow (groundwater discharge) can provide estimates of the total groundwater recharge within a watershed. January stream-flow measurements were used to approximate groundwater discharge because evapotranspiration and irrigation diversions are minimal at this time, and precipitation typically falls as snow, so that ephemeral runoff to the streams during January is infrequent.
Frijoles and Santa Clara Creeks
Stream gages in the Los Alamos area are up to a hundred meters above the regional water table, so there are no base-flow measurements to indicate the quantity of groundwater discharge to surface water in most drainages. Exceptions to this generalization exist in lower Frijoles Canyon and lower Santa Clara Canyon, where stream gages are located far enough down canyon to intercept groundwater where it reemerges near the Rio Grande. Stream-flow measurements for the month of January were used to estimate base flow to Frijoles and Santa Clara Creeks.
In general, there is no reason to expect that the drainage basin divides on the Pajarito Plateau correspond to groundwater divides and that all groundwater recharged within a drainage stays within that drainage. This might be the case if recharge along the surface water divides was much greater than in the adjacent canyons, so that a groundwater mound formed below the drainage basin divides. However, existing data indicate that there is insufficient recharge on mesas of the Pajarito Plateau to cause groundwater mounds beneath the drainage basin divides. The inference that base flow in the lower parts of Frijoles Creek and Santa Clara Creeks is derived solely from recharge in the upper parts of the drainage basin may be attributed to the overall east–west hydraulic gradient beneath the Plateau (Keating et al., 2005), which causes groundwater to flow more or less parallel to the topographic divides of these basins.
The total precipitation falling on the Frijoles Canyon watershed above the stream gage was estimated from its elevation distribution and the precipitation–elevation relation given in Eq. [1] to be approximately 24300000 m3 yr–1 (27.2 cfs). On the basis of January stream-flow measurements from 11 Feb. 1983 to 30 Sept. 1996 (USGS NWIS), base flow to lower Frijoles Creek is 1135000 m3 yr–1 (1.27 cfs). Therefore, if base flow to lower Frijoles Creek can be equated with recharge within the watershed, approximately 4.7% of the precipitation falling on the watershed (an average of about 24 mm yr–1) becomes recharge. If it is assumed that all recharge occurs in the parts of the watershed vegetated by aspens and mixed conifers (73.4% of the watershed), the average recharge in these areas is about 32 mm yr–1. Stream flow past Station 08313350 in excess of base flow averages about 777400 m3 yr–1 (0.87 cfs) or about 3.2% of the precipitation estimated to fall on the watershed. This value suggests that either runoff from the Sierra de los Valles is higher than the 2.5% estimated from stream gages along the western boundary of the Laboratory (Fig. 7), that the runoff from the Sierra de los Valles does not infiltrate as rapidly in Frijoles Canyon as assumed in Table A-1, or that some runoff is generated within the lower part of the watershed.
Stream-flow data for Santa Clara Creek indicate the average January flow for the period of record (1 Oct. 1936 to 30 Sept. 1994) is 3.33 cfs (USGS NWIS). Using this estimate of base flow, total groundwater discharge to lower Santa Clara Canyon is 2976000 m3 yr–1. The total precipitation falling on the Santa Clara Canyon watershed is 56900000 m3 yr–1 (63.7 cfs), so the measured January base flow suggests recharge is about 5.2% of the total precipitation falling on the watershed. If all recharge is assumed to originate from areas vegetated by aspens or mixed conifers, the average recharge in these areas is estimated to be about 47 mm yr–1. The higher infiltration rate for the aspen and mixed conifer areas compared with the Frijoles Canyon watershed may reflect the higher elevations in this watershed. Average annual flow in excess of base flow averages about 786400 m3 yr–1 (0.88 cfs) or slightly less than 1.4% of the total precipitation. The nonzero value of ephemeral runoff at this gage, more than 20 km from the mountain front, indicates either that more runoff leaves the mountains than estimated (2.5% of precipitation), the runoff does not infiltrate as rapidly as estimated in Table A-1, or that some runoff is generated on the Pajarito Plateau.
Rio Grande
Base flow gains to the Rio Grande and its tributaries estimated on the basis of stream flow measurements made by the USGS have provided important constraints on recharge in the regional groundwater
TOP
ABSTRACT
INTRODUCTION
