Figure 1. Conceptual design of the Branched Aquifer Recharge System
for enhancing natural recharge to arid watersheds illustrating the four
major system components: Hillslope collectors, Topographic Zone Convergence
Collectors, a Branched Distribution and Storage Network and a Transmission
Line and Supply Well. Note that parts of the branched network coincide
with the natural elevation gradients occupied by dry or intermittent stream
beds.
Figure 2. Conceptual design of the Branched Aquifer Recharge System
illustrating a side perspective of one pair of Hillslope and Convergence
Zone Collectors. The incident rainfall and evaporation from the catchment
surface are also illustrated. The water table is fed by the recharge system
near convergence zones. Note the overflow structure included in the conceptual
design with the intent of providing for both short or longterm water availability.
Figure 3. Surface Elevation Contour Map of the Idealized Catchment.
The black crosses represent the locations where the elevations were specified.
The contour lines represent 10 meters intervals in elevation. These were
derived by an inverse distance weighted interpolation of the 10 closest
neighbors to each grid cell (50m by 50m). Notice that the symmetry in the
catchment should produce similar results in the radial direction, except
near the corners.
Figure 4. Three-dimensional view of the Idealized Catchment Digital
Elevation Model (DEM). The vertical extent has been visually exaggerated
for illustration purposes. The elevations in the DEM range from 50 meters
along the corners of the idealized hillslope to 250 meters in the central
portion. Using a hillslope length scale of approximately 1800 meters, the
DEM has a nominal surface slope of 0.1 or 10%. The dots in the figure represent
the locations where the elevation was specified, with the rest of the surface
being interpolated using an inverse distance weighting algorithm (produced
with ArcView GIS).
Figure 5. Top view of the Visual MODFLOW layout for the idealized
Branched Aquifer Recharge System (BARS) modeled within the three-dimensional
landscape. The contour lines represent the elevation of the top layer,
as in Figure 3. The grids represent the 50 meter by 50 meter computational
cells for MODFLOW. The hatched areas are the projection of the BARS
system
onto the top layer, while the black rectangles represent the recharge outlets
and the central location is the BARS
hillslope collector head. The
supply wells are located at 45 degree angles from these recharge outlets.
Figure 6. Side view of the MODFLOW layout for the idealized
Branched Aquifer Recharge System modeled within the three-dimensional landscape.
This view represents a cross section through the BARS tunnels as
modeled within MODFLOW. The dark blue section represent regions
of high hydraulic conductivity within the tunnels (K = 0.01 m/s),
while the light brown regions are the tunnel linings designed to inhibit
flow by having a low hydraulic conductivity (K = 1e-11 m/s). The
typical hydraulic conductivity for the catchment is K = 1e-5 m/s
and is isotropic.
Figure 7. Simulated rainfall or recharge series during a half year
cycle within the idealized arid zone catchment for the four climatic conditions
studied. Rainfall can be modeled within MODFLOW by applying a steady-state
or temporally varying recharge rate to the upper active soil layer. Intensities
can be varied over the modelled domain. In this example, recharge is only
specified at the BARS hilltop collector in an effort to simulate
an orographic effect. The evaporation rate on a saturated surface layer
is specified as a constant for the entire catchment. (Please click image
for higher resolution)
Figure 8. Top view of homogeneous case (unbranched recharge) results
for the 180 day transient simulation illustrating the water table depth
for the four climatic forcings represented in Figure 7. The interior
contours represent mounded groundwater table conditions, while the four
enclosed contours are the supply well drawdowns.
Figure 9. Side view of homogeneous base case results for the 180
day simulation time. The cross sectional view is taken along y =
1250 meters for x = 0 to 2500 meters. The arrows represent the groundwater
flow direction, generally away from the unbranched recharge zone underneath
the hilltop. The groundwater table level is represented by the horizontal
fringe. Notice the significant difference in Case 1 due to the fact that
at t = 180 days recharge is actively taking place.
Figure 10. Top view of the BARS case (branched recharge)
results at the end of the 180 day transient simulation illustrating the
water table depth for the four climatic forcings represented in Figure
7. Groundwater mounds are observed in the recharge heads spaced at
45 degree angles from the supply wells. Note the variation in the groundwater
mound shape as compared to the unbranched recharge. Also note the variability
with climatic forcing for the four recharge strategies. The drawdown contours
are observed near the extraction wells.
Figure 11. Side view of the BARS case results for the 180
day simulation time. The cross sectional view is taken along y =
1250 meters for x = 0 to 2500 meters. Notice that the groundwater
head and the flow directions vary at t = 180 days for the four climatic
forcings. In Case 1, the recharge is apparent at the termination of the
BARS
tunnels.
The location of the downwards groundwater movement varies with the rainfall
intermittency. The BARS tunnel system is represented by the outlined
areas.
Figure 12. Comparison of aquifer levels at a Supply Well for
the Homogeneous and BARS climatic forcing cases. The location chosen
(x = 625m and y = 625m) is representative of the four extraction
wells. The water level time series at the supply wells for the two cases
demonstrates the expected impact of rainfall intermittency on the aquifer
level. The effect of branching recharge on the aquifer level in the withdrawal
zone is seen clearly in Figure 15. (Please click image
for higher resolution)
Figure 13. Comparison of aquifer levels at the Recharge Well Head
for the Homogeneous and BARS climatic cases over the simulation
period. The location chosen (x = 775m, y = 1250m) is representative
of the four BARS tunnel endings. The intermittency in the rainfall
is observed in the aquifer levels, as expected. The recharge is superimposed
on a general decreasing trend caused by the constant extraction. Notice
that a peak aquifer level is observed at approximately 10 days for Case
1 due to the interaction of the constant recharge and extraction rates.
While the BARS case aquifer levels are impacted by the recharge
in this zone, the effect on the Homogeneous case is attenuated. (Please
click image for higher resolution)
Figure 14. Comparison of aquifer levels at the Hilltop for
the Homogeneous and BARS climatic cases over the simulation period.
The location chosen (x = 1250m, y = 1250m) forms part of
the BARS collector head. The intermittency in the rainfall is observed
in the aquifer levels, as expected. The recharge is superimposed on the
general decreasing trend caused by the constant extraction rate. Notice
that a peak aquifer level is observed at approximately 40 days for Case
1 due to the interaction of the constant recharge and extraction rates,
much later than the observations for the BARS recharge zone. In
addition, the magnitude of the variation in aquifer levels in response
to the recharge is much greaterin the Homogeneous case, further demonstrating
the dispersal effect induced by the branched network. (Please
click image for higher resolution)
Figure 15. Ratios of the aquifer level or head at two locations:
the supply wells and the recharge zone. The recharge zone is the BARS
Recharge
Well for the BARS case and the hilltop or BARS Collector
Head for the Homogeneous case. The take home message provided by this comparison
is that the branching network increases the aquifer levels slightly for
all times at the supply well, where it is most desired, at the expense
of decreasing the levels more significantly in the recharge zone, where
it is constantly being replenished. The intermittency in the rainfall affects
the temporal pattern of this recharge distribution effect. (Please click
image for higher resolution)
Page created and maintained by Enrique R. Vivoni
For comments or information contact: vivoni@mit.edu
