Environmental Engineering Reference
In-Depth Information
groundwater ages, as determined by carbon-14
analysis ( Section 7.3.2 ) for samples obtained
from 200 observation wells. MODFLOW was used
to simulate groundwater flow, and the particle-
tracking model MODPATH (Pollock, 1994 ) was
used to simulate the transport of groundwater
age (i.e. age was used as a tracer). Groundwater
in the basin occurs primarily within uncon-
solidated materials that range in thickness up
to 4000 m. The model grid consisted of 80 col-
umns and 156 rows of 1-km 2 cells ( Figure 3.13 )
and 9 vertical layers of variable thickness.
Focused recharge from the Rio Grande,
Jemez River, Rio Puerco, and Rio Salado was
simulated with head-dependent boundary
conditions (River Package in MODFLOW).
Focused recharge also occurred from smaller
arroyos that flow from the mountains to the
east of the simulated domain and was simu-
lated with the Recharge Package in MODFLOW.
Diffuse recharge was assumed to be negligible.
Interaquifer flow into the domain occurred
along western and northern boundaries and
was represented by constant flux boundary
conditions. A total of 21 geologic zones were
identified; hydraulic conductivity was assumed
uniform within each zone. Predevelopment
water levels were simulated under the assump-
tion of steady-state conditions.
Model calibration was accomplished with
the parameter estimation program, UCODE, but
the nonlinear least-squares regression method
in UCODE was not entirely compatible with the
particle-tracking algorithm used in MODPATH.
Small changes in some model parameters could
cause dramatic shifts in particle trajectories. As
a result, some individual parameter values were
manually adjusted from UCODE-determined
values to obtain a final fit. UCODE was useful,
however, for determining parameter sensitivi-
ties (Sanford et al ., 2004 ).
Total flow into the simulated domain was
estimated to be 2.14 m 3 /s. Approximately 65%
is recharge, and the remainder is interaquifer
flow across model boundaries. The total inflow
is substantially less than estimates of 5.45 m 3 /s
(Kernodle et al ., 1995 ) and 4.86 m 3 /s (Tiedeman
e t a l ., 1998 ); these estimates were also determined
by groundwater-flow modeling, but calibration
Jemez
Mountains
107 ° W
Manzanita
Mountains
35 ° W
0
10
KILOMETERS
Arroyo
Underflow from
adjacent basins
Mountain front and
arroyo recharge
Floodplain and
river cells
Figure 3.13 Finite-difference grid for the Middle Rio
Grande Basin groundwater-flow model showing locations
and types of boundary conditions (Sanford et al ., 2004 ).
was based only on groundwater levels. Addition
of groundwater-age data should lead to increased
accuracy of recharge estimates (Sanford et al .,
2004 ). The availability of data such as base flow
or groundwater age for use in model calibra-
tion, in addition to water-level data, generally
improves model fit (Hill and Tiedeman, 2007 ).
Search WWH ::




Custom Search