WRIR 01-4210: Hydraulic-Property Estimates for Use With a Transient Ground-Water Flow Model of the Death Valley Regional Ground-Water Flow System, Nevada and California
Previous Section: Hydraulic Properties |
Next Section: Summary and Conclusions
Table of Contents | Conversion
Factors and Acronyms | Report Home Page
Researchers have estimated the depth of the flow system underlying the NTS area and postulated a somewhat qualitative relation between hydraulic conductivity and depth in the region. Winograd and Thordarson (1975) indicate that fractures in the carbonate aquifers are "open" (more permeable) to about 1,300 m below land surface and are "tighter" (less permeable) below this depth. D'Agnese and others (1997) indicate qualitatively that between depths of 300 to 1,000 m the hydraulic conductivity of most rocks in the DVRFS decreases rapidly. At depths greater than 1,000 m, matrix permeability probably dominates, except when within regional fault zones. Below 5,000 m, confining pressures likely keep faults and fractures closed (D'Agnese and others, 1997). The IT Corporation (1996b, p. 29) has postulated a relation of exponentially decreasing hydraulic conductivity with depth in the alluvial aquifer (equivalent to the YAA, OAA, and ACU), in the volcanic aquifer (equivalent to part of the Tertiary volcanics unit), and in the lower carbonate aquifer. While decreasing trends are apparent (IT Corporation, 1996b, figs. 6-1, 6-2, and 6-3), a great deal of scatter in the data also is apparent.
The relation of hydraulic conductivity and depth were examined for the 10 hydrogeologic units that overlie the XCU. Linear regression analysis showed the greatest correlation to depth occurred with the log10 transform of all hydraulic-conductivity estimates. The coefficient of determination (r2) for the depth and non-transformed estimates is 0.003, while for the log10 transformed estimates it is 0.296. In contrast, the coefficient of determination of log10 transformed depth and non-transformed hydraulic-conductivity estimates was 0.245. The best relation, based on regression simulation, is the non-transformed depth with the transformed hydraulic-conductivity estimates and this model was used for the analyses of covariance (ANCOVA; see Neter and others (1985) for an explanation of analysis of covariance).
A plot of the log10 transform of hydraulic conductivity and the mid-point depth of the tested interval for the 10 HGU's are shown in fig. 4. Visual examination of this plot shows an apparent relation between hydraulic conductivity and depth, but relatively high data scatter. ANCOVA initially were done on all hydraulic conductivity and depth data combined into a single data set to assess whether changes in hydraulic conductivity were related to HGU and depth. ANCOVA for the log10 transformed hydraulic conductivity estimates and depth for all HGUs indicate that depth and HGU are both significant factors at a probability level of 0.025 (table 3).
Since the HGU and depth were determined to be significant factors with the change of hydraulic-conductivity values, hydraulic-conductivity and depth data were categorized by an individual HGU and individually analyzed. Results from the ANCOVA for each individual HGU show a significant relation between depth and log10 transformed hydraulic conductivity at a probability level of 0.025 for five of the HGU's (YAA and OAA, YVU and VSU, Tertiary volcanics, OVU, and UCA and LCA. Although the ANCOVA of log10 transformed hydraulic-conductivity estimates indicate that depth may be a significant factor for the variation of hydraulic conductivity in five of the HGUs, the estimates can still vary considerably at a given depth. These large variations probably are caused by other factors (such as bedding, lithologic heterogeneities, or structural influences) that are not accounted for in these analyses. Additionally, some of the decrease in hydraulic conductivity with depth may be the result of using test-interval thickness for calculating hydraulic conductivity. That is, some of this decreasing trend in hydraulic conductivity possibly may be an artifact of the procedure used to calculate hydraulic conductivity from transmissivity.
Previous Section: Hydraulic Properties |
Next Section: Summary and Conclusions
Table of Contents | Conversion
Factors and Acronyms | Report Home Page