Monitoring of rainfall, runoff, drainage, and matric potential using heat-dissipation sensors began on October 1, 1997. Monitoring of water potential with the thermocouple psychrometers began in December 1997, and monitoring of water content began in April 1998. The horizontal neutron probe access tubes were first monitored in February 1998 and were monitored at varying time intervals thereafter. The vertical neutron probe access tubes were monitored at least monthly with concurrent EM data collected at each of the 20 locations.
Rainfall, Runoff, and Drainage Measurements
Daily and cumulative precipitation and irrigation since October 1, 1997, are shown in Figure 13 and Figure 14. Precipitation during 1997 totaled 274 mm, which is lower than the long-term average of 321 mm yr-1. High levels of precipitation during installation of the barriers resulted in wet initial conditions. Total precipitation during the 1998 water year was 192 mm, including approximately 15 to 20 cm of snow that melted in late December 1997. Most (61 percent) of the precipitation occurred between June and September 1998. Irrigation for revegetation began on August 5, 1998, and grass seedlings were planted from August 12 to 14. Daily irrigation of ~6 mm was then applied until September 1. Alternate-day irrigation followed until September 27. The total applied irrigation was 223 mm, bringing the total to approximately 415 mm. With a total runoff of 69 mm, net precipitation and irrigation was 347 mm. To date, there has been no significant drainage collected by the lysimeters. Small volumes ranging from ~0.10 to 0.25 L drained from three of the four lysimeters in October 1998 and were probably a result of condensation within the collection pipe systems.
Water content was monitored in the vertical neutron probe access tubes beginning on June 12, 1998 (Figure 15a and Figure 15b). Water contents were low near the surface (~0.06 to 0.09 m3 m-3 at 0.15-m depth) and increased sharply to ~0.24 mm at 0.45 m depth. Below this depth, water content was fairly uniform (0.22 to 0.25 m3 m-3). A slight increase in water content was measured at the top of the capillary barrier (0.29 m3 m-3 at 1.85 m). Water content increased markedly in response to rainfall and irrigation in the summer, and maximum water contents near the surface were 0.30 to 0.33 m3 m-3. In the top 1.1 m, initial water storage was 223 mm on the GCL/asphalt-barrier plots and 226 mm on the capillary-barrier plots (Figure 16). Water storage increased to 307 mm (GCL/asphalt barrier) and 304 mm (capillary barrier) in September as a result of precipitation and irrigation. By November, water storage in the top 1.1 m had decreased to 273 (GCL/asphalt barrier) and 292 (capillary barrier). During the same period, storage in the capillary-barrier profile (Figure 17) was initially 458 mm, increased to 540 mm, and then decreased to 527 mm.
Water content measured in the shallowest horizontal neutron probe tubes is shown in Figure 18a and Figure 18b. Spatial analysis using variograms indicates that water content is spatially correlated within a range of 3 to 4 m. Some of the variability observed in the horizontal access tube data is attributed to intersecting instrument cable conduits close to the tubes that affect the neutron probe response.
Electromagnetic induction was evaluated as a tool to noninvasively monitor water content. An initial survey over the surface of the engineered barriers was conducted using an EM38 meter (Figure 19). The anomalous north-south trends result from the aluminum horizontal neutron probe access tubes that are located at 0.3, 1.2, 1.9, and 2.9 m below the surface. Correlation between apparent electrical conductivity and water content at each vertical neutron probe access tube for a number of dates between June and November 1998 were generally high, but slopes and intercepts varied widely between locations. Weighting the water contents to the response curve of the EM38 instrument resulted in only slight increases in correlation. Combining all the data from individual neutron probe access tubes resulted in low correlation coefficients (Figure 20a and Figure 20b). The relationship between apparent electrical conductivity measured by the EM38 and water content is complicated by spatial and temporal variability in electrical conductivity as monitored by TDR (Figure 21a and Figure 21b). The TDR data indicate that bulk conductivity varies greatly with depth and time. They also indicate that irrigation caused an initial increase in conductivity followed by a decrease as flushing displaced high-conductivity water through the system. Therefore, water content is not the only parameter that is changing with irrigation, and conductivity is also changing. The bulk conductivity data measured by the TDR are consistent with the temporal variability in matric potential and show predominantly piston flow.
Results of monitoring matric potential at various depths at selected locations on the asphalt and capillary barriers are shown in Figure 22a and Figure 22b. Initially, matric potentials were lowest at the surface and increased nonsystematically with depth. Infiltration as a result of snowmelt in December 1997 penetrated to a maximum depth of 0.3 m. Following the onset of summer rains and then irrigation for revegetation of the barriers, matric potentials decreased systematically as the wetting front moved downward. The movement of the wetting front was generally uniform over the engineered covers, although penetration near the head of the slopes was somewhat less and near the toes of the slopes was somewhat greater. The successive increases in matric potential with depth are consistent with piston flow. Vertical matric potential profiles before, during, and after irrigation of the asphalt and capillary barriers are shown in Figure 23a and Figure 23b. Potential gradients were initially generally upward, though variable with depth. With the onset of irrigation, strong downward gradients were established and variability decreased dramatically. Following irrigation, small upward gradients were established in the near surface whereas downward gradients predominated at depth and variability remained small.