Modeling Ground Water Interactions with Surface Water - PowerPoint PPT Presentation

1 / 52
About This Presentation
Title:

Modeling Ground Water Interactions with Surface Water

Description:

Modeling ground-water interactions with surface water is difficult because of ... Modeling approach traditionally has depended on perspective one either asked: ... – PowerPoint PPT presentation

Number of Views:614
Avg rating:3.0/5.0
Slides: 53
Provided by: depr2
Category:

less

Transcript and Presenter's Notes

Title: Modeling Ground Water Interactions with Surface Water


1
Modeling Ground Water Interactions with Surface
Water
Western States Water Council
Water Information Management Systems
MeetingAlbuquerque, New Mexico September 13, 2006
David E. PrudicU.S. Geological SurveyCarson
City, Nevada
2
When the wells dry, we know the worth of water
Benjamin Franklin
Add streams and this statement embodies the
history of development in the Western United
States during the past two centuries
3
Ground-Water Interactions with Surface Water is
Ever Changing
Modeling ground-water interactions with surface
water is difficult because of variability in
climate, and changing practices in the
development of surface-water and ground-water
supplies
4
History of Development
Knowing the history of an area is key to
understanding and modeling ground water
interactions with surface water
5
Development is Divided into Three Periods
  • Local diversion of water from streams for
    mining and irrigation of areas in adjacent flood
    plains
  • Development of large reservoirs and complex
    irrigation systems for areas distant from the
    flood plains
  • Development of deep turbine pumps and pivot
    irrigation systems for ground-water irrigation

6
First Period
  • Excess diversion of streams led to the
    development of laws and the Doctrine of Prior
    Appropriation
  • Decreased annual stream flow
  • Raised ground-water levels and increased
    ground-water discharge to streams during periods
    of low flow

7
Second Period
  • Development of a Concentrated Hydraulic Society
    to utilize and maintain infrastructure
  • Decreased annual flows of streams
  • Increased ground water in storage around
    reservoirs
  • Increased ground-water levels (storage) in areas
    distant from rivers
  • Localized water logging, creation of wetlands,
    and construction of drains
  • Increased diffuse ground-water discharge to
    evapotranspiration near areas of irrigation

8
Third Period
  • Development of a Distributed Hydraulic Society
    from the drilling and pumping of thousands of
    individual wells with minimum regulation
  • Decreased ground-water levels and storage
  • Decrease in streamflows slow to develop because
    of large ground-water storage capacity

9
Modeling Ground-Water and Surface Water
Interaction
  • Modeling approach traditionally has depended on
    perspectiveone either asked
  • How does surface water influence ground-water
    flow?
  • or
  • How does ground water influence the surface-water
    flow?

S. A. Leake
10
Surface-Water Hydraulics Simulation Model
Black Box
Groundwater
S. A. Leake
11
Ground-Water Flow Model
Surface-water processes
S. A. Leake
12
Numerical Simulation of Surface Water and Ground
Water Interaction
  • Many questions regarding effects of ground-water
    withdrawals on surface water are difficult to
    answer because
  • Aquifers that interact with surface water are
    often heterogeneous and
  • Both surface-water and ground-water flows can
    change in time and space

13
Ground-Water Flow Equation
The equation is simply an expression of mass
balance
S. A. Leake
14
Ground-Water Flow Equation, W term
The W term is flow rate per unit volume of
aquifer added to or taken from ground-water
system.
Q3
Q1
Q2
Most interaction between ground water and surface
water is lumped into the W term
S. A. Leake
15
Common types of surface-water boundaries in
ground-water models
  • Constant or specified flow
  • Constant or specified head
  • Head-dependent flow Q f(h)May be reasonable
    for large lakes and rivers not affected by
    changes in ground-water flow because quantity of
    surface water is not part of calculation

S. A. Leake
16
Constant or specified flow boundaries
  • Flow rate is specified in model grid cell as
    known value of recharge or discharge
  • Model computes head at boundary location

Example Divide total streamflow loss or gain
into parts belonging in each cell
NOTE Make sure model-computed head is reasonable
at these locations!
S. A. Leake
17
Constant or specified head boundaries
  • Head in model grid cell corresponding to
    surface-water location is specified as elevation
    of water surface
  • Flow rate to or from boundary is computed from
    adjacent model grid cells
  • Does not require a W term in flow equation

Example Determine average stream stage in each
cell traversed by stream. Assign stage values as
specified head in cells.
NOTE Make sure model-computed flow is reasonable
at these locations!
S. A. Leake
18
Head-dependent flow boundaries
  • A functional relationship between head in aquifer
    and flow to or from boundary is derived, usually
    by Darcys law
  • Function is made part of W term
  • Most model programs, such as MODFLOW have several
    such functions built in
  • The ground-water head now occurs in the W term,
    possibly adding difficulty to the solution

S. A. Leake
19
Real world
Ground-water model
Streambed sediments
Cross section
Head in stream
Head in cell
Plan view
Area of stream
Area encompassed by model cell
Model cell
S. A. Leake
20
Ground-water model
Head in stream
Head in cell
Idealized prism of river bed sediments contained
in a model cell
Model cell
S. A. Leake
21
Flow Through Streambed
An expression of flow through the streambed is
computed from Darcys Law as
Head here is river head, HRIV
Length, L
Head here is aquifer head, HAQ
Width, W
Thickness, b
Vertical hydraulic conductivity is, Kv
Q Csfr (HSTR-HAQ) where Csfr Kv (LW)/b
S. A. Leake
22
Routing of Surface Water
  • Surface routing programs are used to simulate
    changing location of flow and recharge to an
    aquifer without having to manually specify stage
    or flow in every reach
  • Stage and streambed conductance term can vary in
    relation to channel dimensions and flow

23
Typical Method of Routing Streamflows
Section 1
Flow direction
Section 2
Section 5
Section 3
Section 6
Section 4
SQin SQout DStorage
24
Three Surface-Water Routing Programs Connected to
MODFLOW
  • Streamflow-Routing Package (STR1) recently
    rewritten (SFR1 SFR2) to include different flow
    options, solute transport, and unsaturated
    flowMyself and others (1989 2004 and 2005)
  • Diffusion Analogy Surface-Water Flow Model
    (DAFLOW) Harvey Jobson, USGS Open-file Report
    99-217
  • Branch Surface-Water Flow Model (MODBRANCH) Eric
    Swain and E. J. Wexler, USGS TWRI Book 6, Chapter
    A6

25
Streamflow-Routing Package (STR1, SFR1, and SFR2)
  • Simple routing through network of streams
    assuming no change in storage and sum of inflow
    equals outflow
  • Stream depth computed using different options
    assuming steady, uniform flow

Upstream flow
Precipitation
OverlandRunoff
ET
Downstream flow
Leakage
26
DAFLOW
Continuity of Mass
  • One-dimensional unsteady flow in network of open
    channels
  • Solves equations for continuity of mass and
    momentum assuming no lateral inflow
  • Approximates flow in a stream as reaches of
    steady uniform flow separated by transitions of
    unsteady flow

Continuity of Momentum
Q is discharge, U is velocity, A is cross-section
area, t is time L is channel Length, y is
depth, G is acceleration of gravity, Sf is
friction slope, and So is streambed slope
27
DAFLOW
Transition of unsteady flow
Steady flow
Flow
Control volume
Q2, A2
t
t Dt
Steady flow
Friction slopeat time t
Q1, A1
DL
So
So
Flow into and out of control volume
,
and
where C is speed of moving wave
28
ModBranch
  • Combines one-dimensional unsteady flow model with
    MODFLOW
  • Solves equations for continuity of mass and
    momentum and allows lateral inflow and outflow
    (known as St. Venant equations)
  • Approximates changes in streamflow through
    finite-difference methods

29
Model rows
Stream Network
Inflow
Segment 1
Diversion
Inflow
Segment 3
Canal
Segment 2
Point diversion
Model columns
Segment 5
Segment 4
Pipeline
Segment junction
Segment 6
Segment 7
Outflow
30
Stream Reaches in a Cell
  • Flow into and out of each reach calculated during
    every MODFLOW iteration
  • Multiple stream reaches simulated for a model
    cell
  • Does not simulate multiple model cells for one
    stream reach

GW node
Only one cell per reach
31
Inflow to a Reach
Pipeline
Tributary stream
Flood wave
Surface runoff
Soil zone
Interflow
Water table
Precipitation
Ground-water discharge
Finite-difference cell
32
Outflow from a Reach
Pipeline
Flood wave
Evaporation
Diversion
Streamflow out
Stream leakage
Stream leakage
Finite-difference cell
33
Diverting Flow into a Segment
  • Specified diversion
  • Diversion rate reduce to available flow
  • Diversion rate reset to zero when available flow
    less than specified diversion
  • Specified flow fraction available in stream
  • Flow diverted only when available flow exceeds a
    specified flow rate

Last two options originally program by Randy
Hanson, San Diego Office
34
Stream Depth
  • Stream depth is computed at midpoint of each
    stream reach
  • Flow is computed at midpoint prior to computing
    stream depth except when a constant depth is
    specified

Cell node
Reachmidpoint
Q QIn 0.5Overland flow(PLw)
-0.5(ETLw)leakage to GW
35
Stream Depth
  • Flow at midpoint is partly dependent on streambed
    leakage, which is dependent on stream depth that
    is dependent on flow
  • Equation is repeatedly solved using Newtons
    method until
  • Change in stream stage is less than a specified
    value usually 0.0001 or
  • Average of last two stream stages after reaching
    a maximum of 50 iterations

36
Options for Computing Stream DepthMannings
Equation for Wide Rectangular Channel
Q (C/n) A R2/3 S1/2
where A area of rectangular channel
(width depth)
R hydraulic radius Area divided by wetted
perimeter R depth when width (w) gtgtdepth)
C constant (1.0 for m3/s and 1.486 for ft3/s)
n Mannings roughness coefficient, dimensionless
Depth can be computed as
d (Q/n)/(CwS 1/2)3/5
37
Eight Point Cross Section
The eight point cross section is divided into
three parts. Vertical walls are assumed at the
end of each cross section
Channel
Left bank
Right bank
Part 2
Width
Depth
Wetted perimeter
Depth, wetted perimeter, and width computed from
cross section given flow.
38
Log-Log equation
The width (w), mean depth (d), and mean velocity
(v) each increase with respect to discharge (Q)
as power functions (Leopold, Wolman and Miller,
1964 and 1992, Fluvial Processes in
Geomorphology) WIDTH aQb, DEPTH cQf,
VELOCITY kQm where a, b, c, f, k, and m are
numerical coefficients because w d v Q
aQbcQf kQm Q and b f m 1 and a c
k 1
39
Depth Computed from Table
  • Table of depth and width versus flow is entered
    from data collected at a streamflow gage
  • Program uses log interpolation between values
    except when computed flow is between 0 and first
    value in table

Width
Depth
Flow
10
0.5
5
15
1.3
50
20
1.8
100
35
2.6
1,000
40
Different Options can be used for Each Stream
Segment
Wide rectangular channel
Diversion canal - constant
Table
Eight-point cross section
Log-log equation
41
Simulation of Unsaturated Flow Beneath Streams
  • Details published in USGS Techniques and Methods
    6A-13 (Niswonger and Prudic, 2005)

42
Simulating Vertical Unsaturated Flow Beneath
Streams (SFR2)
  • Formulation is based on kinematic wave solution
    to Richards equation (Smith, 1983)
  • - Avoid instabilities associated with
    conventional unsaturated flow models
  • - Speeds up computational time for applications
    to large basin scale simulations

43
Infiltration
Time
Water Content
Depth
Time 1
Time 2
Time 3
44
Unsaturated Zone Divided into Compartments
Stream
High stage
6
Streambed
Depth, meters
Normal stage
0
Unsaturated zonecompartment
0
14
Width, meters
45
Ground Water Interactions with Lakes are
Simulated Using the Lake Package (Merritt and
Konikow, 2000)
Pipeline
Tributary stream
Outlet stream
Surface runoff
Evaporation
Interflow
Precipitation
Lake cell

Ground-waterdischarge
Lakebed
Lake leakage
Aquifer cell with node
46
Connection of Streams with a Lake
Model rows
Segment 1
1
1
2
Segment 2
2
3
4
3
6
5
4
5
6
7
7
Lake 1
Model columns
8
9
1
2
3
4
5
Segment 3
47
Lake surface
Lake bottom
Lakebed thickness
Lakebed
Conductance terms
Distance from base of lakebed to point in aquifer
Aquifer
Cross-sectional area
Point in aquifer
48
Lakes can Divide and Merge
High stageone lake, one water budget
Lake Surface
Saturated zone
Low stagetwo or more lakes with different water
budgets
Lake 2
Lake 1
Saturated zone
49
Integration of PRMS with MODFLOWGSFLOW, Version
1.0
Streams and Lakes
Soil Zone in PRMS
Interflow
Soil-Moisture Dependent Flow
Surfacerunoff
Ground-water discharge
Ground-water discharge
Head-Dependent Flow
Soil-Moisture/Head-Dependent Flow
Gravity drainagethrough unsaturated zone
Leakage
Ground water
50
Hydrologic Response Units and Finite-Difference
Grid, Sagehen Creek, Truckee, California
51
Ground-Water Recharge
3-yr simulation period daily time steps
52
Ground-Water Discharge Saturation Excess
3-yr simulation period daily time steps
Write a Comment
User Comments (0)
About PowerShow.com