Middle Rio Grande Water Supply Study

1
Riparian Groundwater Modelsfor the Middle Rio
Grande
S.S. Papadopulos Associates and the NM
Interstate Stream Commission December, 2005
2
Why do we need a new model?

• Existing regional models
• USGS Groundwater Flow Model of the Albuquerque
  Basin, McAda and Barroll 2002
• Upper Rio Grande Water Operations Model (URGWOM),
  US Corps of Engineers 2005
• FLO-2D, Tetra Tech 2004
• Models address regional groundwater conditions,
  routing and operations, and surface water
  hydraulics respectively
• None have ability to represent processed-based,
  small-scale surface-water/groundwater interactions

3
Filling the gap

• Needed a model to support habitat restoration
  and river modification by
• quantifying depletions,
• assessing the impact of proposed changes on other
  systems,
• improving our understanding of river operations
  impacts,
• facilitating water operations, including meeting
  ESA mandates and flow targets.

4
Desired Model Characteristics

• Objective represent physical processes and
  transient interactions between surface water and
  shallow groundwater.

• Modeled interactions to include
• seepage from the river
• interception of shallow groundwater by drains
• recharge to shallow groundwater from flooded
  overbank areas
• water depletions due to open water evaporation
  and riparian ET

5
Riparian Groundwater Models Overview

• Series of 5 linked models, covering the Rio
  Grande from Angostura Diversion Dam to North
  Boundary of Bosque del Apache
• Upper Albuquerque - Angostura Diversion Dam to
  I-40
• Lower Albuquerque - I-40 to Bernalillo-Valencia
  county line
• Belen Bernalillo-Valencia county line to
  Valencia-Socorro county line
• Bernardo - Valencia-Socorro county line to San
  Acacia Dam
• Socorro - San Acacia Dam to North Boundary of the
  Bosque del Apache National Wildlife Refuge

6
Riparian Groundwater Models Structure

• Constructed in MODFLOW 2000
• Covers area between levees, including river,
  riverside drains, and riparian corridor contained
  within the levees
• Cells are 125 by 250
• Four model layers
• Three layers within the Rio Grande Alluvium 20,
  30, 30 in thickness
• One layer within the Santa Fe Formation 100 in
  thickness

7
Riparian Groundwater Models Data used in model
construction

• 2002 Albuquerque Basin Regional Groundwater model
  for boundary conditions and hydraulic properties
  in northern four models
• Socorro Basin Regional Groundwater model for
  boundary conditions in Socorro model
• FLO-2D for in-channel and overbank flow extent
  and depth
• 1-ft DTM used in FLO-2D modeling for land surface
  elevation
• Survey data (new and existing) and seepage run
  data for drain bottom and drain water surface
  elevations
• USBR 1992 Ag/Deg lines for river bottom elevation
• USGS gage data Rio Grande at Albuquerque, Rio
  Grande at Bernardo, Rio Grande at San Acacia

Locations of drain survey measurements
8
Riparian Groundwater Models Data used in model
construction

• URGWOPS/ESA Collaborative Program vegetation
  mapping for riparian plant type and distribution
• ET rates from recently published values (King and
  Bawazir, 2000 Cleverly et al., 2002 Dahm et
  al., 2002)
• ET rate with depth (required for implementation
  of Riparian Evapotranspiration MODFLOW package)
  from published data and discussion with James
  Cleverly, UNM, and Kate Baird, Univ. Arizona

9
Riparian Groundwater Models Setup

• Library of river input files created to portray
  variety of river conditions
• Library used to construct step-function
  hydrographs simulating 10-12 month period,
  including spring run-off pulse
• For period simulated, riparian ET rate variant
  with time
• For all simulations, drain water surface
  elevation and groundwater boundary conditions
  invariant with time

10
Riparian models Model Verification

• Models evaluated for reasonable ability to
  match
• Measured seepage rates
• Temporally high-resolution groundwater elevation
  data
• USGS Rio Bravo piezometer data, 2003-2004
• UNM GW monitoring data
• Los Lunas, Belen, Bernardo, Sevilleta, 2000-2001
• USBR cross-section data, 1993-1994
• Sandia, Paseo del Norte, I-40, Tingley Beach, Rio
  Bravo
• Rio Grande Watershed Study cross-section data,
  SSPA, ISC, NM Tech and USCOE, 2003-present
• San Acacia, Escondida, Brown Arroyo, Highway 380

11
Riparian models Model Verification

• Evaluations Performed
• Steady-state and transient model simulations
  evaluated
• Transient simulations include 1994, 2001, and
  2004 spring run-off pulses
• Results
• Basically a good match between modeled and
  measured groundwater elevations and seepage rates
• Groundwater elevation match weakest for locations
  close to drains

Observed and simulated water levels, Rio Bravo
cross-section, 2003-2004
12
Hypothesis Testing

• Models used to test hypotheses concerning
  relationship of shallow riparian groundwater
  conditions to variations in
• vegetation type and coverage
• regional groundwater conditions
• antecedent flow conditions

• Simulations illustrate dynamic nature of riparian
  zone behavior, with inter-relationships among
  environmental components including groundwater,
  surface water and vegetation.
• Results have implications for both water
  management and restoration activities.

13
Hypothesis Testing Alternate Vegetation
Conditions

• Simulation areas identified as salt cedar or
  cottonwood mixed with non-natives, with an
  associated maximum ET rate of four feet per year,
  were re-assigned a maximum ET rate of three feet
  per year.
• Results
• Change in groundwater elevations within the model
  area are, in general, minimal. Some localized
  areas of significant impact, particularly in
  Bernardo and Socorro models
• Change in river and drain seepage minimal for all
  sub-models
• Some reduction in water lost to ET, but amount
  highly dependent on assumptions

14
Hypothesis TestingAlternate Regional Groundwater
Conditions Setup

• Low Regional Conditions Drain stage and
  groundwater boundary elevations were lowered by
  50 of the average drain water depth (1.3 feet in
  the Bernardo model 0.6 feet in the Socorro
  model)
• High Regional Conditions Drain stage and
  groundwater boundary elevations were raised by
  50 of the average drain water depth

15
Hypothesis TestingAlternate Regional Groundwater
Conditions Results

• Riparian corridor is drier in both the late
  spring and mid-summer
• Increased river seepage loss rates
• Inflow to the drains is reduced
• Low regional conditions would impact maintenance
  of target river flows for habitat, and would
  impact water delivery via the river channel.

• Riparian corridor is wetter
• Significant decreases in river seepage loss rates
  
• Inflow to the drains is increased

The LFCC is less responsive to changes in
regional water level than the river or drains,
likely because there is already a significant
gradient toward the LFCC and therefore seepage is
primarily a function of LFCC bed conductance.
16
Hypothesis TestingAlternate Antecedent Flow
Conditions

• Low Antecedent Condition Simulation assumes
  surface water supply has been low in the previous
  year or years. The Base Case spring pulse is
  unchanged, as are regional groundwater elevations
  and drain stage.
• High Antecedent Condition Simulation assumes
  surface water supply has been high in the
  previous year or years. The Base Case spring
  pulse, regional groundwater elevations, and drain
  stage are unchanged.

17
Hypothesis TestingAlternate Antecedent Flow
Conditions

• A change in antecedent flow conditions with no
  change to boundary groundwater elevations or
  drain stage has only a very temporary impact
• Within one to two months, seepage rates and
  groundwater elevations returned to Base Case
  conditions.

• Implications
• Short-term high flows such as brief spring flood
  pulse or summer monsoon flows are unlikely to
  change boundary water levels significantly
  therefore, impact will be short-lived.

18
Current work
Preliminary assessment of Los Lunas Habitat
Restoration Project

• Simulate changes in
• Groundwater elevations,
• River seepage, and
• Drain seepage.

• Resulting from changes in
• Land surface elevation,
• Riparian vegetation, and
• Channel configuration.

• Plan to review
• Base Case historic, pre-project, pre-fire
  conditions
• Restoration with changes to system as described
  in Restoration EA
• Impacted without restoration project, and with
  non-native encroachment resulting from fire

19
Data Recommendations for Model Refinement

• Current data are insufficient for formal
  calibration of the riparian models. Data are
  lacking in the following areas
• Drain bottom elevations
• Drain water-surface elevations and variation with
  time
• Riparian zone groundwater elevation data
• Drain and river seepage data
• Additionally, it would be valuable to
• Install drain and river stage gages at multiple
  paired locations
• Update river bottom elevations at more closely
  spaced intervals
• Conduct paired river and drain seepage runs under
  a variety of seasonal and flow conditions, with
  concurrent monitoring of shallow groundwater
  elevations.

20
Future Applications

• Models can be used to support restoration
  activities
• Site selection/assessment
• Feasibility studies
• Project design
• Project monitoring and operations/maintenance
• Quantification of increased depletion
• Quantification of changes in seepage loss and
  return flows
• Models can be used to assess
• Changes in river losses under alternate water
  conveyance alternatives
• Changes in river losses under different river
  operations scenarios