Published in Vadose Zone Journal 3:982-989 (2004)
© 2004 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
ORIGINAL RESEARCH
Continuous Temperature Logging in Air across the Deep Vadose Zone
Marshall Reiter*
New Mexico Bureau of Geology and Mineral Resources, New Mexico Institute of Mining and Technology, 801 Leroy Place, Socorro, NM 87801
* Corresponding author (mreiter{at}nmt.edu)
Received 5 November 2003.
 |
ABSTRACT
|
|---|
Accurate temperature measurements with good depth resolution will be helpful in appreciating processes that influence temperatures in the vadose zone. In some areas, such as the Albuquerque Basin, groundwater flow perturbs the temperature regime in the saturated zone, making the deep vadose zone a potentially better region to investigate ground surface temperature changes. When it is not logistically or economically possible to span the vadose zone with a pipe filled with water, the temperature measurements must be made in air. Because of the very small heat capacity of air, these temperature measurements are difficult. This paper describes a new temperature sensor with a relatively fast time constant in air that provides accurate temperature measurements in a continuous logging mode. Temperature gradient characteristics below the 5-m depth, based on a 1-m depth measurement interval, are reproducible at a number of logging speeds. Absolute temperatures between the logs at any given depth below about 25 m typically agree to within 0.01°C. The temperature logging system should enable one to gather considerable information regarding the thermal regime in the vadose zone.
Abbreviations: DMM, digital multimeter
|
INTRODUCTION
|
|---|
SUBSURFACE TEMPERATURE DATA taken in both the saturated and unsaturated zones contribute to a variety of studies such as geothermal resource exploration, defining the thermal regime of the earth's crust and upper mantle, calculating ground surface temperature changes that may relate to climate change, and considering heat transfer processes operating in the subsurface. Reiter (2001) reviewed temperature models used to estimate groundwater flow parameters in the saturated zone while Constantz et al. (2003) reviewed temperature models useful in estimating fluid movement and heat transfer processes in the unsaturated zone. Good resolution of temperature data across the saturated zone (measurement intervals 
1 m) helps to better define depth intervals with different lithologies and different groundwater flow parameters (Reiter et al., 1980; Sass et al., 1988; Reiter, 2001, 2003). Likewise, in the vadose zone, good depth resolution of temperature data should be helpful in appreciating ground surface temperature changes, possible near surface heat transfer processes, and changes in material properties with depth.
Reiter (2004) discussed potential ambiguities in using subsurface temperature data from the saturated zone to distinguish the effects of groundwater flow from the effects of surface temperature change. Interestingly, if moisture fluxes in the deep vadose zones of alluvial basins of the arid southwestern USA are very small, possibly 
0.01–0.02 mm yr–1 (Tyler et al., 1996; Walvoord, 2002; Constantz et al., 2003), the temperature regime should be dominated by conduction, making the deep vadose zone more ideal for investigating surface temperature changes. Shallow data (<10–20 m depth) are significantly influenced by the yearly temperature cycle and therefore longer term temperature influences are generally considered with data deeper than 10 to 20 m. For example, the complimentary error function solution for temperatures at depth responding to a surface temperature step (after Carslaw and Jaeger, 1959, p. 63) indicates 36% of the step will be noticed at the 40-m depth after 30 yr, and 36% of the step will be noticed at the 50-m depth after 47 yr (allowing the thermal diffusivity to be 1 mm2 s–1). It is important to note that although the vadose zone may be ideal for investigating surface temperature change, the cause of such changes may be ambiguous. For example, yearly changes in precipitation, evaporation, and vegetation shading, as well as climatic temperature change, could affect surface temperatures. The possible contributions to an estimated surface temperature change might be considered with weather station data, if available.
Accurate temperature data in the vadose zone with good depth resolution will allow more meaningful curve fitting of expressions describing the possible timing and magnitude of surface temperature change (e.g., the error function solutions). The temperature expressions may be nonunique and consideration of other data will be important. For example, from the subsurface temperature data it may not be possible to prefer statistically a single temperature step at the surface from a temperature ramp at the surface with a constant rate of temperature change with time (Reiter, 2004). However, the subtle curvature of the temperature logs representing surface temperature change (similar to error function solutions) is more noticeable, and better curve fitted statistically, across depth intervals of 100 m with data taken at 1-m separation than with data taken at 30- to 60-m separation. It is also noted that the resolution of the temperature data to be shown here results in a temperature gradient resolution that has the potential to allow correlation with thermal conductivity changes that may result from different stratigraphy and/or different degrees of water saturation. High resolution temperature data should allow the pursuit of a number of such studies, and it is possible that more can be learned about heat transfer processes affecting the vadose zone from better resolved data.
This paper describes a new temperature sensor assembly that employs recent technology to provide high resolution temperature logs in air at logging speeds of approximately 2 m min–1, with data depth intervals of 1 m. The motivation for the work is to further investigate ground surface temperature changes in the Albuquerque Basin by acquiring good temperature logs in the air-filled section of piezometers crossing vadose zones generally about 100 to 120 m thick. The objective is to be able to make such temperature logs in several hours. The equipment will be described and the data presented for six temperature logs made during a 2-wk period at a number of logging speeds for the Tome piezometer site about 35 km south of Albuquerque, NM (Reiter, 2003). Using these data, a reasonable logging speed in terms of time and accuracy is obtained.
|
BACKGROUND
|
|---|
Shallow subsurface temperature measurements in the vadose zone to depths of about 5 to 15 m can be made with good depth resolution by burying temperature sensors or by lowering a sensor downhole (e.g., Dowman et al., 2004). If a sensor is lowered into the well for temperature measurements, the slow response time typical of temperature sensors in air requires stopping at each depth station for some prescribed amount of time, or until the measurement changes by some set amount for a given time interval (e.g., Sass et al., 1988; Dowman et al., 2004). For deeper vadose zones (50 to >200 m) logging the well by lowering a sensor is generally the only practical method of taking temperature measurements; however, the "stop-and-wait" measurement procedure is time-consuming and generally requires large depth intervals between measurements (e.g., 30–60 m intervals, Sass et al., 1988).
Several decades ago temperature sensors used for subsurface temperature measurements were typically platinum or thermistor resistance–temperature devices enclosed in watertight stainless-steel housings about the size of a thin pencil. These sensors often had relatively slow response times, even in water. Sensors with slow response times in water would require the stop-and-wait logging protocol described above for the vadose zone, although stationary times could usually be reduced from between 20 and 30 min to a few minutes and the measurement interval would depend on well depth. Costain (1969) presented a technique whereby one can inverse filter recorded temperature data, taken while the sensor is continuously lowered into a well, through the time response character of the sensor to obtain data more depth representative. The technique removes the effects of sensor delay time in the continuous downhole logging mode. Costain (1969) pointed out that the procedure becomes more important as sensor time constants become greater than about 1 s [the time constant is defined as the time required for a sensor temperature to change by (1 – 1/e) of an imposed temperature step].
Temperature data taken in water in the continuous logging mode with sensors having time constants of about 1 s to a couple of seconds demonstrated good resolution (Reiter et al., 1980; Sass et al., 1988). Temperature logs at speeds of 7.6, 15.2, and 30.5 m min–1 yielded gradients calculated from measurements every 1.52 m that are very similar (Reiter et al., 1980). From the sensor response time of 1 sec, the number of time constants between readings at a 1.5-m interval would be about 7.9, 3.9, and 2.0, respectively, for the different logging speeds just cited. These temperature data suggest that requiring the continuous downhole logging speed to be such that about two time constants are present between measurement depths 1.5 m apart should provide satisfactory temperature–depth data for approximately normal geothermal gradients found in the western USA (30–45°C km–1).
The problem of sensor time response is increased when taking data in air because sensor response time is very slow. Sass et al. (1988) discussed a specially designed sensor assembly that reaches 99% of a temperature step in still air during a period of about 4 min (this translates to a time constant of 50–60 s). They use this sensor to make temperature logs in air-filled wells at Yucca Mountain, Nevada, when it is not possible to place a water filled pipe across the vadose zone in a well. The measurement intervals are 30 to 60 m. The wait at each depth station is either 20 min or the time the resistance change of the thermistor during 1 min is <1% of the accumulated change at the depth station, whichever is less. Temperatures are extrapolated to 1/time = 0 (or t = 
) for each depth. The accuracy of the temperature data are a few hundredths of a degree Celsius, but problems of well bore convection in large-diameter wells and diurnal barometric changes may provide greater uncertainties.
|
DESCRIPTION OF A NEW LOGGING SYSTEM
|
|---|
The temperature logging system discussed in this paper uses a Thermometrics FP07 fastip probe thermistor (Thermometrics, Edison, NJ) as the sensor. The thermal time constant of the device is stated to be 0.10 s in still air at 25°C. Figure 1 shows the "large ruggedized thermobead" assembled with a cable connection, and the sensor assembly attached to the logging cable including a thin stainless-steel wire protection hoop. The small dark tip at the end of the glass bead is the thermistor encased in a thin watertight glass layer. The sensor assembly is connected to the cable and the connection is hermetically sealed with adhesive shrink tubing covered by several layers of waterproofing neoprene bonding agent. To estimate the time constant of the sensor assembly including the protection hoop, attached to the cable, the sensor was subjected to cooling and warming temperature steps. The sensor assembly was quickly inserted through a hole in the side of a small foam cooler at a lower temperature than the surroundings. The cooler had ice blocks on the top and on the bottom with two fans on opposite sides circulating cool air in the plane of the sensor assembly so the air flow was normal to or across the sensor. After several minutes recording temperature change in the cooler, the sensor was quickly withdrawn to warmer room temperatures where the warming response was recorded. This procedure was repeated four times, and the sensor assembly time constant was determined from both the cooling and warming temperatures to be approximately 15 to 20 s. However, when temperature logging at a constant speed downhole, the air flow relative to the sensor assembly will be parallel to the sensor axis, through the protection hoop opening. Therefore, the sensor moving downhole through an air column may have a somewhat different time constant than determined in the experiment as described here. A sequence of logs at different downhole speeds (discussed further below) can determine a most appropriate logging speed.
View larger version (97K):
[in this window]
[in a new window]
|
Fig. 1. Sensor assembly showing electrical connection and sensor assembly attached to cable with stainless-steel protection hoop. Black tip at the end of the glass bead is the thermistor enclosed in a thin watertight glass layer. The scale applies only to the upper probe; the lower probe is an enlargement.
|
|
To measure the resistance of the thermistor, an appropriate digital multimeter (DMM), with four-wire resistance measuring capability, is necessary. We used a Keithley model 197 (Keithley, Cleveland, OH). It is important to consider resistance measurement compatibility of the DMM with the thermistor and lead resistance along with the inductance of the logging cable wrapped around the cable reel. The multimeter is connected to a laptop computer that records the DMM reading every meter the sensor is lowered down a well. This is accomplished by sending a call signal to the computer from the well-head pulley, machined to be 1 m in circumference, for each complete pulley rotation. The signal is sent to the computer by an amplifying circuit that includes a stationary microswitch triggered by a rotating cam attached to the pulley wheel. Upon receiving the measurement call signal from the pulley circuit, the computer records the last reading taken by the DMM. The DMM takes several readings per second.
The measuring system was calibrated at three temperatures between 13.17 and 27.94°C using a platinum resistance as a standard and a Keithley 2002 digital multimeter for resistance measurements of the platinum standard. Both the thermistor and the platinum sensors were immersed together in a stirred water bath. The calibration curve was determined as a best fit to the calibration data from the TABLE CURVE program (a product of Jandel Scientific Software, SPSS, Chicago, IL). Subsequently, measurements were taken at seven different temperatures in the stirred bath before, during, and after the temperature logs were made in the field. The agreement between the platinum standard and the thermistor assembly was 0.013 ± 0.007°C. An accuracy appropriate for the temperature logs is therefore approximately 0.01°C.
To lower the sensor downhole at a given speed, a DC electric motor powered by batteries is used. The speed is calculated by the computer using the time between each measurement station 1 m apart; therefore, there is a delay in recognizing the effect of a speed control adjustment and it may take a few meters to adjust to the desired logging speed. Typically speeds can eventually be controlled to better than 0.1 m min–1. The cable is manually rewound on the reel to reduce cable stretching and crushing forces on the reel drum. The logging cable has four 28-gauge conductors enclosed in a Kevlar stress wrap and a rubber exterior.
Within the well bore at Tome there are three piezometers, each open to the formation (screened) at different depths. Each piezometer has a 5.08-cm-diam. PVC pipe from the surface to the screened interval. The space between the PVC pipes and the formation is filled with bentonite and backfill so flow between the screened intervals is prevented. Steel surface casing (30.5-cm diam.) is present from a concrete pad at ground surface to a depth of about 12.2 m. Larger steel casing (35.6-cm diam.) placed in the concrete pad around the surface casing extends about 0.9 m above the pad; a steel hatch is welded to the top of this casing. The tops of the PVC piezometer casings are about 15 cm below the hatch.
|
RESULTS
|
|---|
The six temperature logs taken at the Tome site in Piezometer 2 are shown in Fig. 2
. The data are successively incremented by 0.5°C for each log after the first so the logs can be visually compared. The data are all taken in air across the vadose zone, which extends to about 59.5 m. The temperature logs stop at 58 m, about 1.5 m short of the water in the piezometer.
View larger version (17K):
[in this window]
[in a new window]
|
Fig. 2. Temperature vs. depth data, temperature logs, for six different days. Dates logged (month, day, year) and logging speed are given in legend. Logs are offset 0.5°C so they can be compared.
|
|
Below about the 8-m depth the temperature logs are very similar (Fig. 2), although there are large temperature differences above 8 m (Fig. 3)
. By the 10-m depth all of the measured temperatures agree to within 0.05°C, even though at the 1-m depth temperatures may differ by approximately 4°C (Fig. 3). From Fig. 3 and 4
one may notice the temperatures at the 1-m depth often, but not always, correlate to the maximum and minimum temperatures recorded in Albuquerque the day before and the morning of that log (all of the logs were made in the morning). The increase in the temperatures at 1 m for Logs 1, 2, and 3 (Fig. 3) parallel the increase in the maximum and minimum Albuquerque temperatures (Fig. 4). The highest temperature at 1 m (Log 5, Fig. 3) correlates to high maximum and minimum temperatures (Fig. 4), and a lower temperature at 1 m for Log 6 corresponds to reduced maximum and minimum temperatures. The higher temperature at 1 m for Log 4, as opposed to Log 3, is inconsistent with the daily maximum and minimum temperatures, suggesting the possible influence of several daily temperatures and the large steel surface casing to the 12.2-m depth.
View larger version (11K):
[in this window]
[in a new window]
|
Fig. 3. Temperature vs. depth data from the 1- to 10-m depth for six different logs. Dates logged (month, day, year) and logging speed are given in legend.
|
|
View larger version (10K):
[in this window]
[in a new window]
|
Fig. 4. Daily highs and lows in Albuquerque about 35 km north of the Tome site. Dates of temperature logs are indicated, beginning with 5.26 (month, day).
|
|
It is typical to observe curvatures in temperature logs by looking from the bottom of the logs, as shown in Fig. 2, at a viewing angle of approximately 20° off plane. From Fig. 2 one can see the similarity of the logs below the 10-m depth. The rounded shoulder of the data trend at about 10 m (convex to the right) can be interpreted to change slope at about 12 m, becoming concave to the right to about 20 m, where a cusp is noticed. A more subtle trend, similarly concave to the right, is present from about 20 to 26 m depth. Below 26 m the temperature data show a gradually decreasing temperature gradient with depth; that is, the log gently curves convex to the right.
A more quantitative comparison of the temperature logs can be made by considering temperature differences and temperature gradients. Log 3 was chosen as a base for comparison because the logging speed of 0.45 m min–1 for the total depth of the log allows seven to nine sensor time constants to occur between each temperature measurement 1 m apart. This should permit the data to be very near equilibrium. Temperature data from Log 3 are compared with data from other logs at 10 m and deeper (Fig. 5)
. At about the 15-m depth, the differences between measurements from Logs 3, 5, and 6 are 0.015°C. At 15 m, the measurements for Logs 1, 2, and 4 differ by 
0.02°C from temperatures for Log 3, in part because the logging speeds for Logs 1, 2, and 4 are too fast for the very high temperature gradients in the upper 10 m of the piezometer (Fig. 3). Log 2 has the additional difference of an extra time period spent near the surface (20 min) to replace electronics at the pulley. The longer wait near the surface for Log 2 is likely to affect the temperature of the entire sensor assembly more than for the other logs, causing a somewhat different residual temperature influence on the sensor from the entire assembly. This different history may contribute to temperature differences among logs (greater at shallower depths), but the character of the temperature logs is little affected, as will be shown below in the gradients. Below about 25 m, the data from most of the logs differ by 0.01°C. Applying these facts along with logging time objectives, a logging speed of 0.75 m min–1 (corresponding to four to five time constants between readings 1 m apart) is chosen for the upper 10 to 15 m, where large temperature gradients exist. A much faster logging speed of 2.25 m min–1 (corresponding to about 1.5 time constants between measurements) is chosen for deeper depths, where temperature gradients are at least an order of magnitude less than the near surface gradients.
The nonlinear sections noticed in the temperature logs are also shown in the profiles of temperature gradient as a function of depth (Fig. 6 and 7)
. The section that is concave to the right in the logs (Fig. 2) between depths of about 12 and 20 m is shown in the gradient–depth plot by the noticeable change from decreasing to increasing gradients between 12 m and 20 m (Fig. 6 and 7). The slight curvature noticed in the logs from about 20 to 25 m (Fig. 2) is also clearly shown by the swing from decreasing to increasing gradients between 20 and 25 m (Fig. 6 and 7). Below 10 m, all of the temperature gradients agree with each other to within about 0.01°C m–1, which is reasonable in terms of determined accuracy of a single measurement (0.013 ± 0.007°C). The gradients are calculated by subtracting two temperature measurements 1 m apart. The gradient swings from about 12 to 20 and 20 to 25 m are larger than the measurement uncertainty. Interestingly, all of the logs also show cycles of decreasing to increasing gradients from about 25 to 29 m, 29 to 32 or 33 m, and 50 to 56 m, even though the amplitude of the gradient change is approximately 0.01°C m–1 at some depths. The rather apparent spike in the gradient shown at about 17 m in Logs 4 and 6 (Fig. 6 and 7) is not noticed in the other logs and could therefore result from measurement errors enhanced on the log plots for the gradients. When gradient trends appear in all of the logs, one can be more confident of their existence. Below about 30 m, the trend of gradients is toward lower values, and the differences between gradients from 30 to 50 m are within the limits resulting from measurement uncertainties.
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 6. Temperature gradients for Logs 3, 5, and 6 starting at the 5-m depth. Note log scale for low gradient enhancement.
|
|
|
DISCUSSION
|
|---|
Six temperature logs are presented in this study. The measurements were made during a 2-wk period to avoid long-term surface temperature effects such as annual temperature variation. The logs were taken 2 or 3 d apart, allowing the piezometer almost 2 or 3 d to recover from the perturbing effects of the previous temperature log. Although temperatures differed by about 4°C among logs at the 1-m depth, at the 15-m depth all of the logs differed by 0.04°C. Differences in temperatures among logs at 15 m resulted because of differences in logging speeds, unsystematic errors in the measuring system, differences in daytime highs and nighttime lows transmitted by steel casing, and possibly differences in the recent temperature history of the sensor assembly, which could transmit slightly different temperature effects to the sensor (e.g., sensor time spent near the surface, or holding the sensor assembly in sunlight vs. shade). The large steel casing above ground at the site is painted a dark green and is likely to transmit considerable heat to the surface casing during hot sunny days. The steel surface casing will act as a preferential heat conductor to 12.2 m. The steel casing will enhance surface temperature change effects on the piezometer subsurface temperatures beyond what is expected for soil or rock. In the present study the accuracy of the measuring system is shown with data much deeper than the bottom of the steel casing, data deeper than 20 to 30 m. However, for near surface studies relating to shallow subsurface temperatures, the anomalous effects of the steel surface casing may well be important while considering models with initial and boundary temperature conditions. Likewise, recharge and evaporation studies in arid environments using shallow temperature data should quantify the effects of anomalous steel surface casing conduction when applicable and, if possible, use small diameter plastic pipe with relatively low thermal conductivity to maintain piezometer integrity. In the future the effect of surface casing may be experimentally better appreciated when logs taken at different sites with different casing depths are made.
Even though temperatures at depths of 5 m differed among logs by more than 1°C (Fig. 3), both the temperature logs and the temperature gradient plots are similar in character (Fig. 2, 6, and 7). These observations show that for the fastest logging speed in the study (Logs 2 and 4), and for logs having different recent thermal histories before logging (log 2), the primary characteristics of the temperature–depth profile in the piezometer are observable. From the similarity of the temperature gradients and the small temperature differences among Logs 3, 5, and 6, free convection in the piezometer tubing or airflow in the piezometer because of atmospheric pressure change seem unlikely. The lack of convective overturn is consistent with the small diameter of the piezometer tubing (Gretener, 1967). The lack of noticeable airflow in the piezometer resulting from atmospheric pressure change is consistent with the piezometer being sealed and isolated from the formation except at the screened interval, which for the piezometer logged was at the 212-m depth, about 152 m below water level.
Figure 5 illustrates that temperature differences among all logs at 58 m are 0.008°C—this agreement is better than the measurement accuracy discussed previously (0.013 ± 0.007°C). Interestingly, at the completion of Log 6, 19 measurements were taken during a 7-min interval with the sensor at rest just below the 58-m depth. A temperature cycle was observed during the interval with an amplitude of about 0.007°C, which is less than the measurement accuracy. A subsequent log showed somewhat larger variations with the sensor at rest for about 15 min just below 58 m (0.015°C).
|
CONCLUSIONS
|
|---|
Measuring subsurface temperatures requires that a sensor be lowered down a hole in the earth in a manner that will least disturb the in situ temperature. Therefore the logging protocol is to take measurements going downhole so the cable and most of the probe assembly pass a given depth after the temperature is sensed. The actual sensor in the system described here is a small, thinly glass–covered thermistor located at the end of a glass bead (Fig. 1). At present, this device may have as fast a time response as is available. Moving the sensor at constant speed downhole introduces it to new, previously undisturbed air, which relative to the sensor is flowing. Although the thermistor cannot be totally isolated from the rest of the sensor assembly, logging data at a number of speeds indicated that the residual temperature influence of the assembly does not change the gradient characteristics of the logs. Many nonlinear sections of temperature logs below the influence of the yearly temperature cycle should be observable with the present system. Gradient swings greater than about 0.01°C m–1 are probably real and can be verified with additional logs; gradients swings less than about 0.01°C m–1 need to be verified with additional logs. With a measurement interval of 1 m, gradient changing intervals of about 3 m having similar characteristics between logs are noticeable. The preferred logging speeds of approximately 0.75 m min–1 to about the 13-m depth and 2.25 m min–1 to the bottom of the air column (Logs 5 and 6) yield data that are in relatively good agreement with Log 3 (used for comparison because of the slow logging speed of 0.45 m min–1). The agreement is within about 0.015°C at 15 m and within about 0.008°C below 25 m.
|
ACKNOWLEDGMENTS
|
|---|
I thank Bonnie Reiter for assisting with the temperature logging. Lynne Hemenway typed the manuscript and Leo Gabaldon helped with Fig. 1.
|
REFERENCES
|
|---|
- Carslaw, H.S., and J.C. Jaeger. 1959. Conduction of heat in solids. Oxford Univ. Press, Oxford, UK.
- Constantz, J., S. Tyler, and E. Kwicklis. 2003. Temperature-profile methods for estimating percolation rates in arid environments. Available at www.vadosezonejournal.org. Vadose Zone J. 2:12–24.[Abstract/Free Full Text]
- Costain, J.K. 1969. Probe response and continuous temperature measurements. J. Geophys. Res. 75:3969–3975.
- Dowman, C.E., T.P.A. Ferré, J.P. Hoffmann, D.F. Rocker, and J.B. Callegory. 2004. Quantifying ephemeral streambed infiltration from downhole temperature measurements collected before and after stream flow. Vadose Zone J. 3:595–601. Available at www.vadosezonejournal.org.
- Gretener, P.F. 1967. On the thermal instability of large diameter wells—An observational report. Geophysics 32:727–738.[ISI]
- Reiter, M. 2001. Using precision temperature logs to estimate horizontal and vertical ground water flow components. Water Resour. Res. 37:663–674.
- Reiter, M. 2003. Hydrogeothermal studies in the Albuquerque Basin, a geophysical investigation of ground water flow characteristics. Circular 211. New Mexico Bur. Geol. Mineral Resources, Socorro, NM.
- Reiter, M. 2004. Possible ambiguities in subsurface temperature logs-consideration of ground water flow and ground surface temperature change. Pure Appl. Geophys. (in press).
- Reiter, M., A.J. Mansure, and B.K. Peterson. 1980. Precision continuous temperature logging and correlation with other types of logs. Geophysics 45:1857–1868.[ISI]
- Sass, J.H., A.H. Lachenbruch, W.W. Dudley, S.S. Priest, and R.J. Monroe. 1988. Temperature, thermal conductivity, and heat flow near Yucca Mountain, Nevada: Some tectonic and hydrologic implications. USGS Open-file Rep. 87-649. USGS, Denver, CO.
- Tyler, S.W., J.B. Chapman, S.H. Conrad, D.P. Hammermeister, D. Blout, J. Miller, and J.M. Ginanni. 1996. Soil water flux on the Nevada Test Site: Spacial and temporal variations over the last 120,000 years. Water Resour. Res. 32:1481–1499.
- Walvoord, M.A. 2002, A unifying conceptual model to describe water, vapor, and solute transport in deep arid vadose zones. Ph.D. diss. New Mexico Inst. Mining and Technol., Socorro, NM.
This article has been cited by other articles:
|
|
|
|
 
M. REITER
Vadose Zone Temperature Measurements at a Site in the Northern Albuquerque Basin Indicate Ground-Surface Warming due to Urbanization
Environmental and Engineering Geoscience,
November 1, 2006;
12(4):
353 - 360.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
