Appendix A. Interpretation Guide for Graphs of Stratigraphic Position Compared with Hydraulic Head for Wells in the Columbia River Basalt Group, Washington, Oregon, and Idaho

After wells have been grouped into spatial clusters based on similar values of head and similar trends, the variability within each group may be examined to evaluate vertical hydraulic gradients. This is accomplished by plotting all available head data collected during a short period (relatively unaffected by temporal trends) as a function of the stratigraphic position of the aquifer being measured. If the vertical head gradients are large relative to the slope of the potentiometric surface within each aquifer, then the vertical gradient is apparent.


The assignment of hydraulic head values to a given stratigraphic position is fraught with complications that may affect the accuracy of any single data point. For Columbia River Basalt Group (CRBG) wells, it is assumed that the well terminates in a productive aquifer, and that the measured head in that well is representative of an aquifer at that stratigraphic position. Stratigraphic position is computed as the difference between the elevation of the bottom of the well and an estimate of the elevation of a geologic horizon that originally was flat. Use of stratigraphic position removes the influence of dipping, folding, and faulting when correlating strata that can act as laterally continuous aquifers. Conversely, in areas where CRBG units are discontinuous or exhibit significant thinning or thickening, stratigraphic position may have a spatial trend. As long as the group of wells being examined does not cover too large an area, spatial bias of stratigraphic position generally will be small. In addition to this complication, well commingling can result in hydraulic head measurements that are not representative of the aquifer at the well bottom (for example, well 1 in fig. 6).


Despite the potential complications, combinations of persistent patterns in stratigraphic position and hydraulic head may be interpreted hydrologically (fig. A1). The clustering of wells at distinct stratigraphic positions is evidence that distinct aquifers are locally important, and when hydraulic heads are distinctively different between aquifers, downward (fig. A1A) and upward (fig. A1B) vertical gradients exist. Similarly, there may be multiple distinct aquifers with no vertical gradient (fig. A1C). For the CRBG, significant vertical distance between two distinct aquifers implies the presence of a potential confining unit and the likelihood that a vertical hydraulic gradient should exist. However, aquifers may be hydraulically connected naturally through complex CRBG geometry, or they may be connected through commingling wells. If historical evidence supports a significant vertical gradient (fig. A1A-B), but recent measurements are consistently similar for each aquifer (fig. A1C), then commingling wells are the likely cause.


For many areas in the Columbia Plateau Regional Aquifer System (CPRAS), the repeating sequence of CRBG interflows and flow interiors results in a continuum of stratigraphic position (fig. A1D). In this case, there is either a single aquifer or many thin aquifers, and there can be downward, upward, or no significant vertical gradient present. For most areas within the CPRAS, vertical gradients are expected, so no vertical gradient over several hundred feet of stratigraphic position (fig. A1D) indicates that aquifers likely are hydraulically connected either naturally or through commingling wells.


During the process of creating spatial clusters of wells, two distinctly different heads at the same stratigraphic position (fig. A1E) potentially signals the presence of a horizontal barrier to groundwater flow. If all higher head wells are in a separate location from the lower head wells, then some barrier to flow may exist, and the wells would be divided into separate groups. If the high and low head wells appear to be randomly mixed spatially, then well construction, pumping from adjacent wells, or geologic heterogeneity are the likely causes of the pattern (fig. 6). If a continuum of heads exist within a single aquifer (fig. A1F), then additional explanations may include a steep regional gradient of the potentiometric surface, or local perturbations to the potentiometric surface that may occur in the presence of groundwater discharge boundaries (such as incised canyons with springs and seeps). These local and regional effects may be separated by removing the regional trend from the potentiometric surface and examining the range of the residual heads.


Combinations of expected patterns of hydraulic head as a function of stratigraphic position (figs. A1A-F). are frequent in the real data for CRBG wells as illustrated in figures 12A-F and 20A-I. Over time, new deep wells have increasingly been required to case and seal to the aquifer, hydraulically isolating these wells from the upper commingled aquifers. Whereas the upper aquifers frequently have fairly uniform hydraulic heads, the deeper aquifers have a distinctly different head (compare fig. A2A to group 4 shallow and deep wells [black circles and white circles] shown in figure 20B-C and to group 11 shallow and deep wells [black squares and white squares] in fig. 20E). A continuum of heads associated with local groundwater drainage of upper aquifers may be combined with a uniform vertical gradient for deep aquifers (fig. A2B) to provide the observed pattern of group 8 wells (brown squares) during pre-development prior to lowering of the potentiometric surface below the land surface (fig. 20D).

First posted February 5, 2013