groundwater availability modeling (gam) for the dockum aquifer · groundwater flow models...
TRANSCRIPT
Groundwater Availability Modeling (GAM) for the
Dockum Aquifer
GAM Stakeholder TrainingOctober 22, 2008
Workshop Agenda
1. Introduction2. Modeling Overview
– Modeling Protocol and Practice
– MODFLOW– Groundwater Vistas (GV)
3. Dockum GAM Review– Technical Overview– Data and Model Inputs
4. Break5. Modeling Demonstration
– The GV Interface– Transient/Predictive
Model Exercise(s)
Workshop Goals
Provide an introduction to groundwater modeling, MODFLOW, and GV
Review the development of the Dockum GAM
Provide information on model input and associated data sources
Provide insight into the utility and applicability of the GAM
1. Introduction
Workshop Expectations
To gain an appreciation of the expertise required to use the GAM
To gain an understanding as to the potential applicability of the GAM
To gain some understanding of the limitations of the GAM
To acquire the ability to make minor modifications to the model via PMWIN
1. Introduction
The GAM Truth
If you want to run these models – seek professional help
“It is very easy for me to calculate the positions of Sun, Moon and any planet, but I cannot calculate the positions of water particles as they move through the earth.” Galileo
1. Introduction
GAM Objectives
Develop realistic and scientifically accurate groundwater flow models representing the physical characteristics of the aquifer and incorporating the relevant processes
The models are designed as tools to help GWCD, RWPGs, and individuals assess groundwater availability
Stakeholder participation is important to ensure that the model is accepted as a valid model of the aquifer
1. Introduction
GAM Model Specifications
Three dimensional (MODFLOW-2000)
Regional scale (1000’s of mi2)
Grid spacing of 1 square mile
Include Groundwater/surface water interaction (Stream routing, Prudic 1988)
Properly implement recharge
Stress periods as small as 1 year
Calibrate to within 10% of head drop
1. Introduction
Schematic of Modeling PeriodsPre-development
(Steady-State) Post-developmentCalibration(Transient)
1950 1980 1997
Observed Water LevelModel Water Level
2050
Prediction
Steady-state and transient calibration periods represent different hydrologic conditions
1900
Wat
er E
leva
tion
in W
ell
Modeling Overview
Modeling Protocol & Practice MODFLOW GV – Groundwater Vistas
2. Modeling Overview
Definition of a Model
Domenico (1972) defined a model as a representation of reality that attempts to explain the behavior of some aspect of it and is always less complex than the system it represents
Wang & Anderson (1982) defined a model as a tool designed to represent a simplified version of reality
2. Modeling Overview
GW Models in Water Resources
GW Models have been used in water resources in response to 4 basic issues.
Impact on neighboring resources
Conjunctive use issues (SW-GW)
GW mining & resource depletion on practical time scales (regional resource issues)
Water quality issues
2. Modeling Overview
GW Models in Water Resources
Regional-scale models typically are used to address management as an institutional issue
Local-scale models typically are used to address management as an operational issue
2. Modeling Overview
Define model objectives
Data compilation and analysis
Conceptual model
CALIBRATION
Reporting
Future WaterStrategies
Comparison with
field data
Model design
Field data
Field data
*Includes sensitivity
analysis
Modeling Protocol
Transient*
Steady State*
2. Modeling Overview
Prediction
Modeling References
Anderson & Woessner “Applied GW Modeling”
ASTM D5447 “Standard Guide for Application of a Ground-Water Model to a Site-Specific Problem”
“Fundamentals of Ground-Water Modeling”, U.S. EPA
Faust & Mercer: “GW Modeling: Numerical Models”
Mercer & Faust: “GW Modeling: An Overview”
2. Modeling Overview
Conceptual Model
Identify relevant processes and physical elements controlling GW flow in the aquifer:– Geologic Framework– Hydrologic Framework– Hydraulic Properties– Sources & Sinks (Water Budget)
Determine Data Deficiencies
Conceptual model dictates how you translate “real world” to Mathematical Model
2. Modeling Overview
To be Considered in Code Selection
Simulates Relevant Physical/Chemical Processes
Public-Domain vs. Proprietary
Thorough Testing for Intended Use
Complete Documentation
2. Modeling Overview
Model Design
Translate Conceptual Model to Mathematical Counterparts
Procedure– Grid Design (Numerical)– Define Hydraulic Properties– Boundary & Initial Conditions
2. Modeling Overview
Grid Design – Typical Drivers
Dimensionality (1D,2D,3D)– Vertical Gradients– Multiple Aquifers– Partially Penetrating Wells
Number of Grid Cells– Run Time– Computer Memory
Regular vs. Irregular Node Spacings– Design Time– Accuracy in Areas of Interest
2. Modeling Overview
Grid Design – When to use a Regular (constant dimension) Grid
Regional Studies (e.g. USGS RASA, GAM)
Preliminary Analyses
Models Where Area of Interest May Change
High Resolution Models Where Memory is Not a Concern
GAM grid defined to be 1 mile square
2. Modeling Overview
Model Grid Example
Source: USGS Fact Sheet FS-127-97
2. Modeling Overview
Model Inputs
Hydrostratigraphic Surfaces for each Layer
Hydraulic Properties:– Hydraulic Conductivity– Storativity (transient)
Hydraulic heads
Recharge
Stream Flow (headwater flows, initial C)
Pumpage
2. Modeling Overview
Boundary Conditions
Boundary Condition is a constraint put on the active grid to characterize interaction between the modeled area and its environment
Types:– Specified Head (Dirichlet – Type 1)– Specified Flux (Neumann – Type 2)– Head-Dependent Flux or Mixed (Cauchy- Type 3)
Determination:– Based on Natural Hydrogeologic Boundaries– Analyze Impact of Artificial Boundaries
2. Modeling Overview
Boundary Conditions
Boundary conditions may be static or transient
Recharge or wells – Specified flow
GHB, Reservoir, Stream – Head dependent flow
Vertical or lower boundaries – specified flow @ zero = no flow
2. Modeling Overview
Modeling Approaches
MODEL
HeadsSpring FlowStream Flow
Hydraulic ConductivityStorativityBoundary Conditions
2. Modeling Overview
Model Calibration
Process used to produce agreement between observed and simulated data through adjustment of independent variables
Typical variables adjusted are hydraulic conductivity, storativity, and recharge
Source: USGS Fact Sheet FS-127-97
2. Modeling Overview
Model Calibration
Types:– Trial-and-Error– Automated or inverse– Stochastic
Procedures:– Select Calibration Targets– Select Calibration Metrics– Adjust Boundary Conditions/Properties– Analyze Errors
2. Modeling Overview
Model Calibration
Steady-state calibration– Assumes that the hydrologic system is static over
the time frame of interest– Q in = Q out; No storage effects
Transient calibration– Assumes that dependent variables change with time
in response to changing stresses (recharge, pumping, stage, boundaries)
2. Modeling Overview
Sensitivity Analysis
A sensitivity analysis is a formal means of quantifying the effect of changes in model inputs on model outputs
Provides a means of identifying parameters which are:– Important– Correlated
Most common method is the one-off-method
2. Modeling Overview
Verification
Simulation period where the model is run in a forward mode (ie without adjustment of parameters) to see how the model agrees with observations
The more variable stresses the better the verification period
Acceptable verification doesn’t insure accuracy; does enhance model validity
2. Modeling Overview
Prediction
Once the model meets the calibration metrics, it can be used for prediction.
The basis behind model predictions is the assumption that:– The past is the key to the future.
Predictive accuracy depends on– Validity of modeled processes– Accuracy of props. and boundaries– Knowledge of hydraulic conditions– Reliability of estimates of system stresses
2. Modeling Overview
Reporting
Future WaterStrategies
Prediction
Field data
Conceptual model
Prediction – Post Audits
Post-audits have demonstrated that models are moderately reliable and are uncertain
As approximations to reality, models can, and should, always be improved – (updated)
A primary value of a model, regardless of the predictive accuracy, is it allows for a disciplined format for the improvement of the understanding of an aquifer (Konikow, 1995)
2. Modeling Overview
Calibration Challenges
Uniqueness of calibration
Over-Calibration
2. Calibration Challenges & Approach
Model Uniqueness (Similarity Solutions)
Models are inherently non-unique, that is multiple combinations of parameters and stresses can produce similar aquifer conditions.
The ramification of this is:– A good match to observed data does not guarantee
an accurate model
Modeling Approach to Deal with Uniqueness
To reduce the impact of non-uniqueness:a) Calibrate to multiple hydrologic conditionsb) Calibrate with parameters consistent with
measured valuesc) Calibrate to multiple performance measures
2. Modeling Overview
(a) Calibrate to Multiple Hydrologic Conditions
CALIBRATION
Transient*
Prediction
*Includes sensitivity analysis
Steady State* Comparison with
field data
The calibration approachiterates between the steady-stateand the transient calibrations to reach a consistent set of physical parameters that match both setsof observation.
2. Modeling Overview
(b) Calibrate with parameters consistent with measured values
Because of the uniqueness issues, you must consider some parameters known
On super-regional models such as the GAM, scale issues related to measured data and how they relate to the model result in difficulties
2. Modeling Overview
(c) Calibrate using multiple targets and performance measures
Heads (SS and transient)– Distributions– Time series– Scatter plots– Statistics (RMS, MAE, ME)
Stream aquifer interaction– Stream flow rates– Gain loss estimates
Flow balance (qualitative)
Don’t calibrate better than target error (see next slide)
300
350
400
450
500
550
600
650
700
750
800
300 350 400 450 500 550 600 650 700 750 800
Measured Head (ft)
Sim
ulat
ed H
ead
(ft)
5.0
2
1 1
ism
n
ihh
nRMS
2. Modeling Overview
Over Calibration
One must strive to not over-calibrate (tweak) a model; that is:– Over parameterize lacking data support– Adjust parameters to bring model agreement
below performance measure uncertainty– In the GAM model, head is the primary
performance measure and we have estimated errors associated with heads to be on the order of at least 30 feet
2. Modeling Overview
Calibration and Prediction
Freyberg published a study on calibration and prediction (GW, 1988, Vol. 26, No. 3)
Nine modeling teams using same data
Best model prediction came from the model with the least estimated parameters and with inferior local fits
Good calibration may not equal good prediction
Best calibrated model yielded poorest prediction
2. Modeling Overview
MODFLOW (is a Code)
Developed by the United States Geological Survey
Three-dimensional, finite difference groundwater flow CODE
2. Modeling Overview
MODFLOW Version History
Various USGS research codes; Trescott (1975), and others
MODFLOW (1984)– McDonald and Harbaugh, 1986 (Fortran 66)
MODFLOW (1988)– McDonald and Harbaugh, 1988 (Fortran 77)
MODFLOW96 (1996)– Harbaugh and McDonald, 1996
MODFLOW2000 (2000)– Harbaugh et al (2000)
2. Modeling Overview
MODFLOW Packages
Original Packages (88)– Basic– Block-Centered Flow– Recharge– Evapotranspiration– River– Well– Drain– General Head Boundary– Output Control– SIP/SOR Solvers
Add on Packages (2000)– Block-Centered Flow 2 .. 6– Discretization– PCG/PCG2, GMG Solvers– Horizontal Flow Barrier (HFB)– Compaction (IBS)– Time-variant C.H. (CHD)– Stream Routing (STR)– Transient Leakage (TLK)– Direct solver (DE4)– Various user add ons
Subroutines are called modulesGroups of subroutines representing a “process” are packages
2. Modeling Overview
MODFLOW Advantages
Handles the basic processes
Well documented
Testing is documented – courts accept
Public domain – non-proprietary
Most widely used model– USGS had 12,261 downloads of MODFLOW in 2000
Multiple utility programs and Graphical User Interfaces (GUIs) available
2. Modeling Overview
MODFLOW Processes
Important for GAM– Confined/unconfined GW
flow– Recharge/ET– Horizontal flow barriers– Wells– Streams– Drains (springs)– Reservoirs
Source: USGS Fact Sheet FS-127-97
2. Modeling Overview
After McDonald & Harbaugh, 1988
Example of a MODFLOW GridNote – Regular Grid
Row
>
Column >
2. Modeling Overview
After McDonald & Harbaugh, 1988
Vertical DiscretizationShould have physical significance
2. Modeling Overview
Assignment of Properties
Properties can be assigned on a gridCell basis as below
Properties can be assigned in zones as above
2. Modeling Overview
Example of an IBOUNDArray (Basic package)for a Single Layer
After McDonald & Harbaugh, 1988
These cellsBecome inactive
Single Layer example of conceptualizing aModel grid and assigning boundary conditions
2. Modeling Overview
MODFLOW in simplest terms
MODFLOW calculates flow in 3 dimensions using a finite difference (FD) approach
The GW flow FD equation form follows from the application of the continuity equation which stipulates that:– The sum of all flows into and out of a cell at a given
time step must equal the rate of change of storage within the cell
2. Modeling Overview
Steady-state, One Dimensional Flow Darcy’s Law – One cell
Where:●
K = hydraulic conductivity
●
A = area normal to Flow
●
h = head●
L = length
2. Modeling Overview
Darcy’s Law Can be Rewritten
Where C is equal to the hydraulic conductance (L3/T L)
Q = C (h2 – h1)
C = K A / LMODFLOW uses hydraulic conductance to
calculate flow rates using Darcy’s Law
2. Modeling Overview
Vertical Conductance - Vcont
Simply stated – Vcont is the interval conductance divided by the area (plan view)
MODFLOW uses Vcont (also known as leakance) to calculate vertical flow
2. Modeling Overview
Wells in MODFLOW-2000
MODFLOW-2000 does not have a wellbore submodel– Therefore, simulated heads are representative of the
grid volume
Well rates are specified by row, column, layer (r,c,l)
Multiple wells can be assigned one grid cell
Wells are specified in the well package (.wel)
2. Modeling Overview
Stream Routing
Use MODFLOW Stream Routing Package (Prudic, 1988)
Stream stages are calculated using Manning’s equation
Stream-routing package routes surface water and calculates stream/aquifer interaction (gaining/losing)
Input headwater flow rate, stream conductance, stream dimensions, and Manning’s n parameter
2. Modeling Overview
Stream Routing Package
Stream reaches (each cell is an individual reach) and segments must be numbered
2. Modeling Overview
Head Dependent Flow Boundaries
General head boundaries
Reservoirs
River cells
Stream cells
Drains
2. Modeling Overview
Head-dependent Boundaries
Head difference
Head dependentBoundaries alwaysRequire input of the
ConductanceC = K A / L
2. Modeling Overview
Specified-flow Boundaries
Wells
Recharge
Evapotranspiration ET (hybrid – head dependent)
2. Modeling Overview
MODFLOW Interfaces
PMWIN– Academic, commercially available
Groundwater Modeling System (GMS)– DOD, commercially available
Groundwater Vistas (GV)– Private, commercially available
Visual MODFLOW– Private, commercially available
2. Modeling Overview
GV – Groundwater Vistas
Windows graphical user interface for 3-D groundwater flow and transport modeling. Developed by Environmental Simulations Incorporated.
Authors:Jim Rumbaugh and Doug Rumbaugh
http://www.groundwater-vistas.com
http://www.groundwatermodels.com
2. Modeling Overview
GV Capabilities
Offers a Windows based interface for developing MODFLOW models and for using the family of MODFLOW codes
Imports existing standard MODFLOW models
Supports all standard packages
Allows many options for data input: ArcView shapefiles, AutoCAD DXF files, SURFER files, and ASCII files
Allows for telescopic grid refinement
Some degree of checking of input prior to execution
Offers a wide variety of analysis techniques for viewing simulation results
2. Modeling Overview
GV Requirements
Pentium or better
Windows 95/98/2000/NT 4.0/XP/Vistas
16 MB RAM (32 Recommended)
GAM model– Requires at least 128 MB RAM– 2 GB or better disk space
2. Modeling Overview
GV Interface 2. Modeling Overview
-29,700,203
-23,100,158
-16,500,113
-9,900,068
-3,300,023
3,300,023
9,900,068
16,500,113
23,100,158
29,700,203
1.0 3,690.07,379.011,068.014,757.018,446.022,135.025,824.029,513.033,202.036,891.0
Time
Top Inflow
Stream Inflow
Storage Inflow
Recharge Inflow
Top Outflow
Well Outflow
Drain Outflow
Stream Outflow
Storage Outflow
Total Inflow
Total Outflow
Error
Mass Balance Summary for Layer 3
1,956.8
2,586.9
3,217.0
3,847.1
4,477.2
5,107.3
1,956.8 2,586.9 3,217.0 3,847.1 4,477.2 5,107.3
Observed Value
Layer 2
Layer 3
Observed vs. Computed Target Values
Dockum GAM Review
Technical Overview–
Emphasis on Data and Model Inputs
3. Dockum GAM Review
Model Area F
ile: 2
_01_
stud
y_ar
ea.m
xd
Source: N/A
Dockum Aquifer Study Area
0 50 100
Miles
−State Line
Study_Area
Model Boundary
Downdip Aquifer Limit
County Boundaries
Study Area Model Domain
Lea
Eddy
Chaves
Pecos
Union
Quay
Colfax
Reeves
Culberson
Ellis
DeBaca
Harding
Curry
Texas
San Miguel
Roosevelt
Irion
Mora
Crockett
Hall
Hale
Dallam
Gaines
Jeff Davis
Upton
Hartley
King
Kent
Oldham Gray
Lynn
Beaver
Caddo
Guadalupe
FloydLamb
Terry
Coke
Andrews
Ector
Major
Knox
Kiowa
Mills
Clay
Nolan
Ward
Jones
Potter
Cottle
Taylor
Brown
Custer
Motley
Reagan
Blaine
Young
Dewey
Garza
Martin
Fisher
Coleman
Llano
Baylor
Moore
Archer
Scurry
Tom Green
Castro
Cimarron
Bailey
Deaf SmithDonley
Crane
Runnels
Carson
Concho
Crosby
Schleicher
Washita
Borden
Haskell
Tillman
Erath
Randall
Sterling
Woods
Foard
BriscoeParmer
MitchellHoward
Menard
Roberts
Woodward
Mason
San Saba
Midland
DickensHockley
Swisher
Winkler
Greer
Dawson
Wheeler
Harper
Jack
Eastland
Lubbock
Hemphill
Alfalfa
McCulloch
Roger Mills
Ochiltree
Loving
Wilbarger
HansfordSherman
Callahan
Yoakum
Jackson
Cotton
Comanche
Beckham
Lipscomb
Stephens
Stonewall
Cochran
Armstrong
Glasscock
Comanche
Wichita
Hutchinson
Shackelford Palo Pinto
Childress
Sutton
Harmon
Hardeman
Kimble
Collingsworth
Throckmorton
Burnet
Lampasas
Terrell
coln
Hamilto
Canadia
Kingfis
BrewsterPresidio
ero
Gra
Garf
Steph
File
: 2_0
2_do
ckum
_aqu
ifer.m
xd
Source: Online: Texas Water Development Board, Aug 2006
State Line
Model Boundary
Downdip Aquifer Limit
County Boundaries
Dockum AquiferOutcrop
Downdip
−0 25 50
Miles
River Basins and Topography
Major River Basins Topography
State Line
Model Boundary
Downdip Aquifer Limit
County Boundaries
River Basins
Canadian River Basin
Red River Basin
Brazos River Basin
Colorado River Basin
File
: 2_1
0_riv
er_b
asin
s.m
xd
Source: Online: Texas Water Development Board, June 2006; New Mexico Resource GIS Program, June 2006
Rio Grande River Basin0 25 50
Miles
−
File
: 2_1
3_to
po.m
xd
Source: Online: USGS, Aug 2006
State Line
Model Boundary
Downdip Aquifer Limit
County Boundaries
Land Surface Elevation(ft-MSL)1,406 - 1,5001,501 - 2,0002,001 - 2,5002,501 - 3,0003,001 - 3,5003,501 - 4,0004,001 - 4,5004,501 - 5,0005,001 - 5,5005,501 - 6,0006,001 - 6,5006,501 - 7,0007,001 - 7,5007,501 - 8,0008,001 - 8,5008,501 - 9,000
0 25 50
Miles
−
Geology
Stratigraphic Units Overlying Aquifers
Ogallala/Alluvium
Cretaceous-age Sediments (not continuous)
Upper Dockum
Cooper Canyon Formation
Trujillo Sandstone
Lower Dockum
Tecovas Formation
Santa Rosa Sandstone
Permian-age Sediments
Lea
Eddy
Chaves
Pecos
Quay
Union
Colfax
Reeves
Culberson
Ellis
DeBaca
Harding
Curry
San Miguel
Crockett
Roosevelt
Mora
Irion
Hall
Texas
Jeff Davis
Hale
Dallam
Gaines
Upton
Hartley
King
Kent
Oldham Gray
Guadalupe
Lynn
Caddo
FloydLamb
Terry
Coke
Andrews
Ector
Beaver
Major
Knox
Kiowa
Mills
Llano
Nolan
Ward
Jones
Potter
Clay
Cottle
Taylor
Brown
Custer
Motley
Reagan
Blaine
Young
Dewey
Garza
Martin
Fisher
Coleman
Baylor
Moore
Archer
Scurry
Tom Green
Castro
Bailey
Deaf Smith Donley
Crane
Runnels
Carson
Concho
Crosby
Mason
Schleicher
Washita
Cimarron
Borden
Haskell
Tillman
Randall
Sterling
Foard
BriscoeParmer
MitchellHoward
Menard
Roberts
Woodward
San Saba
Erath
Midland
DickensHockley
Swisher
Winkler
Greer
Dawson
Wheeler
Woods
Eastland
Lubbock
Hemphill
McCulloch
Roger Mills
Ochiltree
Loving
Wilbarger
HansfordSherman
Callahan
JackYoakum
Jackson
Beckham
Lipscomb
Comanche
Stephens
HarperAlfalfa
Cotton
Stonewall
Cochran
Armstrong
Glasscock
Sutton
Wichita
Comanche
Hutchinson
Hardeman
KimbleTerrell
Shackelford
Childress
Collingsworth
Harmon
Throckmorton
Burnet
Lincoln
Hamilton
Canadian
Presidio Brewster
Otero
File
: Fig
.4.1
_Aqu
ifers
_Dire
ctly
_Ove
rlayi
ng_D
ocku
m
Source: Adapted from Texas Water Development Board, Dec 2006
−
Model BoundaryAquifer Downdip LimitState LineCounties
Aquifer NameDockum Group OutcropOgallalaEdwards-Trinity (High Plains)Pecos ValleyEdwards-Trinity (Plateau) - OutcropEdwards-Trinity (Plateau) - SubcropRita Blanca
0 30 60Miles
Model Flow Conceptualization
Undifferentiated Permian
Upper Dockum
Lower Dockum
Ogallala
Cretaceous
Upper Dockum
Lower Dockum
General Head BoundaryRecharge (Precipitation)Discharge (ET, Springs)Discharge via PumpingSurface Water-Aquifer InteractionCross-Formational FlowNo-Flow Boundary
AquiferOutcrop
Aquifer Outcrop
NW SE
Ogallala/Younger
Model Grid
1 square mile grid blocks212 columns422 rows3 layers150,548 active cells
Schematic of Modeling PeriodsPre-development
(Steady-State) Post-development Calibration(Transient)
1950 1980 1997
Observed Water LevelModel Water Level
Model Input – Supporting Data
Hydrostratigraphic
surfaces for each layerHydraulic properties
–
Hydraulic Conductivity–
Storativity
(transient)
Water LevelsRechargeETStream FlowPumpage
All model data, source and derived, was delivered to the TWDB and will be available to the public
Dockum Thickness
!
!
!
!!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
! !
!
!
!
!
!
!
!
!
!
!
!
!
!
!!
!
!
!
! !
!
!
!
!
!
!
!
!
!
!
!
!!
!
!
!
!
!
!
!
!
!
!!
!
!
!
!
! !
!
!
!
!!
!
!
!
!
!!
!
!
!
!!
!
!
!
!
!
!
!
!
!
!
!
!
!
! ! !
!
!
!
!
!
! ! !
!
! !
! !
!
!
!
!
!
!
!
!
!
!
!
!
!
!!
! !
! !
!
!!
!
!
!
!
!
!!
!
!
!
!
!
!
!!
!
!
!
!
!!
!
!
!
!
!
!
!
! !!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
! !
!
!
! !
!
!
!
!
!
!
!!
!!
!
!!
!
!
!
!!
!
!!
!
!
!
!
!
!
!!
!
!!
!
!
!
!
! !
!
!
!
!
!
!!
!
!
!
!
!
!
! !!
!!
!
!!
!!!
!
!
!
!
!!
!
!
!!
!
!! !
!
!
!
!!!
!!
!!
!
!
!
!
!
!!
!!
!
!!
!
!
!
!
!
!!
!
!!
!
! !
!
!!!
!!
!
!
!
!!
!
!
!!!
!!
!
!
!
!
!!
!
!
!
!
!
!!
!!
!
!
! !
!!
!
! !
! !
!
!
!!
!!
!
! !
!
!!
!
!
!
!
!
!
!!
!
!!
!!!
!!
!
!
! !
!!!
!
!
!
!!
!! !
! ! !
!!
!!
!
! !
!
!
!
!!
!
!
!
!
!
!
!
!
!
!
!
!
!
! !
!
!
!
!
!
!
!
!
!
!!
!
!!
!
!
!
!!
!
!!
!
! !
!
!
!
!
!
!
!
!
!!
! !
!
!
!
!
!
!
!
!!
!
!
!
!!
!
!
!
! !
!!!!
!
!
!
!
!
!
! !
!
!
!
!
!
!!!
!
!
!
!
!
!
!
!!
!
!
!
!
!
!
!
!!
!!
!
!
!
!
!
!
!
!
!
!!
!
!
!!
!
!
!
!
!
!
!
!!!
!
!
!
!
!
!
!
!
!!!
! !
!
! !!!
!
!!
! !
!
!
!
! !!
!!
!!
!
!
!!
!
!!!!
!!
!
!!
!
!
!
!
!
! ! !
!
!!
!
!
!
!
!
!
!!
!
!
!
!
!
!
! !
!!
!!
!
!
!
!
!!
!!
!
!
!!
! !
!
!
!!!
!
!
!
!!
!
!!
!
!
!
!
!
!!
!!
! !
!
!
!
!
!
!
! ! ! !
! !
!!!
!
!
!!
!
!
!
!
!
!
!!
!
!
!
!
!
!
!!
!!
!
!
!
!
!
!
!
!
!
!!
!!
!
!
!
! !
!
!!
!
! !!
!!
!
!
!
!
!
!
!
!
!
!
!
!
!
!!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!!
!
!
!!
!!
! !
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!!
!
!
!
!
!
!
!
!
!
!
!
!
! !
!!
!
!
! !
!
!
!
!
!!
! !
!
!
!
!
!!
! !
!
!!
!
!
!
!
!!!!
!
!
!
!
!
!! !
!
!
! !!
!!
!
! !
!
!
!
!
!
!
! !
!
!!
!
!
!
!
!
!
!
! !
!!
!
!!
!
! !
!
!
!!!!
!!
!
!
!
!
!
!
! !
!
!
!
!
!!
!
!
!
!
!
!!
!
!!
!
! !!
!
!
!
!
!
!
! ! !
!!
!
!
!
!
!
!
!
!
!
! !
!!
!
!
!!!
!!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!!
!
!
!!
!
!
!!
!
! !
!
!
!
!
!
!!
!
!!
!
!
!!
!
!
!
!
!
!!
!!
!
!
!
!
!
!!
!! !
!
!
!
!
!
!!
!
! !
!!
!!
!
!
!
! !
!
!!
!!!
!
!!
! !! !!!
!
!!
! !
!!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!!
!!
!
!
!
!!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
! !
!!
!
!
!!
!
!
!
!!!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!! !
!
!
!
!
!
!!
!!
!
!
!
!
!
!!
!
!
!
!
!
! !
!
!
!
!
!!!
!!
!
!
!
!
!
!!
!
!
!
! ! !!!
!
!
!
! !
!!
!
!
! !
!
!
!
!
!!
!
! !
!
!!
!!
!
!
!
!
!
!
!
!
!!
!
!
!
!
!
!
!
!
!
!
!!
!
!
!
!! !
!
!
!!!
!!
!
!!
!
!!
!
!!
! !!
!
!
!
!
! !
!
!
!
!!
!
!
!
!
!
!
!
! !
! !
!
!!
!
!
!
!
!
!
!
!!
! !
!!
!
!
!
!
!
!
!
!
!
!!!
!
!
!
!
!
!
!
!!
!
!
! ! ! !
!!
!!
!!
!
! !
!!
!!
!
!
!!
!
!
!
!
!
!
!
!
!
!
!
! ! !
!
!
!
!
!
!!
!
!
!
!
! !!
!!!
!
!! !
! !
!
!!
!!
!
!!
!
!
!!
!
!
!
!
!
!!
!
!
!
! !
! ! !
!
!!
!
!
!
!
!
!
! ! !!
!!!!
!
!
!
!! !
!
!
!
!
!!
!
!
!!
!
!
!! !
!
!
!
!
!
!
!
!
!
!!
!
!
!!!
!!!
!
!! !
!
!
!!
!!
!
!
!
!
!
!!
!
!
!
!
!
!
! !
!
!
!
!
!
!
!
!
!
!!
!!
!!
!!
!
!
!
!!
!
!
!
!!
!
!
!
!
!!
!
!
!
!
!
!
!
!
!
!
!
! !
!
!
!
!
!
!
!
!
!
!
!
!!
!
!!
!
!
!
!
!!!!
!
!
!
!!
!
!
!
!!
!
!
!
!
!
!
!
!
!
!
!
!!
!
! !
!
!!
!
!
!!
!!
!
!
!
!
!!!
!
!
!!
! !!
! !
! !
!
!!
!!
!
!
!!!
!
!
!
!
!
!
!
!
!
!
! !!
!
!
!!
!
!
!
!!
!
!
!
!
!!
!!
!!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!!
!!
!
!
!
!
!
!
!
!
!
!
!
300
400
200
500
600
700
800
100
9001000
400
1000
100
600
700
500
300
900
100
600
800
700900
800
500
500
300
400
100
800
100500
900
100
100
300
100
100
700
100
400
200
200
400
200
100
200
500
800
File
: 4.2
.5_l
ower
_thi
ckne
ss.m
xd
Source: McGowen et al. (1977)
State Line
Model Boundary
Downdip Aquifer Limit
County Boundaries
! McGowen Control Points
−0 25 50
Miles
Upper DockumThickness
(feet)20 - 100101 - 200201 - 300301 - 400401 - 500501 - 600601 - 700701 - 800801 - 900901 - 1,0001,001 - 1,1001,101 - 1,200
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
! !
!
!
!
!
!
!
!
!
!
!
!
!
!
!!
!
!
!
! !
!
!
!
!
!
!
!
!
!
!
!
!!
!
!
!
!
!
!
!
!
!
!!
!
!
!
!
! !
!
!
!
!!
!
!
!
!
!!
!
!
!
!!
!
!
!
!
!
!
!
!
!
!
!
!
!
! ! !
!
!
!
!
!
! ! !
!
! !
! !
!
!
!
!
!
!
!
!
!
!
!
!
!
!!
! !
! !
!
!!
!
!
!
!
!
!!
!
!
!
!
!
!
!!
!
!
!
!
!!
!
!
!
!
!
!
!
! !!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
! !
!!
! !
!
!
!
!
!
!
!!
!!
!
!!
!
!
!
!!
!
!!
!
!
!
!
!
!
!!
!
!!
!
!
!
!
! !
!
!
!
!
!
!!!
!
!
!
!
!
! !!
!!
!
!!
!!!
!
!
!
!!
!
!
!
!!
!
!! !
!
!
!
!!!
!!
!!
!
!
!
!
!
!!
!!
!
!!
!
!
!
!
!
!!
!
!!
!
! !
!
!!!
!!
!
!
!
!!
!
!
!!!
!!
!
!
!
!
!!!
!
!
!
!
!!
!!
!
!
! !
!!
!
! !
! !
!
!
!!
!!
!
! !
!
!!
!
!
!
!
!
!
!!
!
!!
!!!
!!
!
!
! !
!!!
!
!
!
!!
!
! !
! ! !
!
!
!!
!
! !
!
!
!
!!
!
!
!
!
!
!
!
!
!
!
!
!
!
! !
!
!
!
!
!
!
!
!
!
!!
!
!!
!
!
!
!!
!
!!
!
! !
!
!
!
!
!
!
!
!
!!
! !
!
!
!
!
!
!
!
!!
!
!
!
!!
!
!
!
! !
!!!!
!
!
!
!
!
!
! !
!
!
!
!
!
!!!
!
!
!
!
!
!
!
!!
!
!
!
!
!
!!
!!
!!
!
!!
!
!
!
!
!
!
!!
!
!
!!
!
!
!
!
!
!
!
!!!
!
!
!
!
!
!
!
!
!!!
! !
!
! !
!!
!
!!
! !
!
!
!
! !!
!!
!!
!
!
!!
!
!!!!
!!
!
!!
!
!
!
!
!
! ! !
!
!!
!
!
!
!
!
!
!!
!
!
!
!
!
!
! !
!
!
!!
!
!
!
!
!!
!!
!
!
!!
! !
!
!
!!!
!
!
!
!!
!
!!
!
!
!
!
!
!!
!!
! !
!
!
!
!
!
!
! ! ! !
! !
!!!
!
!
!!
!
!
!
!
!
!
!!
!
!
!
!
!
!
!
!
!!
!
!
!
!
!
!
!
!
!
!!
!!
!
!
!
! !
!
!!
!
! !!
!!
!
!
!
!
!
!
!
!
!
!
!
!
!
!!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!!
!
!
!!
!!
! !!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!!
!
!
!
!
!
!
!
!
!
!
!
!
! !
!!
!
!
! !!
!
!
!
!!
! !
!
!
!
!
!!
! !
!
!!
!
!
!
!
!!!!
!
!
!
!
!!
! !
!
!
! !!
!!
!
! !
!
!
!
!
!
!
! !
!
!!
!
!
!
!
!
!
!
! !
!!
!
!!
!
! !
!
!
!!!!
!!
!
!
!
!
!
!
! !
!
!
!
!
!!
!
!
!
!
!
!!
!
!!
!
! !!
!
!
!
!
!
!
! ! !
!!
!
!
!
!
!
!
!
!
!
! !
!!
!
!
!!!
!!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!!
!
!
!!
!
!
!!
!
! !
!
!
!
!
!
!!
!
!!
!
!
!!
!
!
!
!
!
!!
!!
!
!
!
!
!
!!
!! !
!
!
!
!
!
!!
!
! !
!!
!!
!
!
!
! !
!
!!
!!!
!
!!
! !! !!!
!
!!
! !
!!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!!
!!
!
!
!
!!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
! !
!!
!
!
!!
!
!
!
!!!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!! !
!
!
!
!
!
!!
!!
!
!!
!
!
!!
!
!
!
!
!
! !
!
!
!
!
!!
!
!!
!
!
!
!
!
!!
!
!
!! ! !
!!
!
!
!
! !
!!
!
!
! !
!
!
!
!
!!
!
! !
!
!!
!!
!
!
!
!
!
!
!
!
!!
!
!
!
!
!
!
!
!
!
!
!!
!
!
!
!! !
!
!
!!!
!!
!
!!
!
!!
!
!!
! !!!
!
!
!
! !
!
!
!
!!
!
!
!
!
!
!
!
! !
! !
!
!!
!
!
!
!
!
!
!
!!
! !
!!
!
!
!
!
!
!
!
!
!
!!!
!
!
!
!
!
!
!
!!
!
!
! ! ! !
!!
!!
!!
!
! !
!!
!!
!
!
!!
!
!
!
!
!
!
!
!
!
!
!
! ! !
!
!
!
!
!
!!
!
!
!
!
! !!
!!!
!
!! !
! !
!
!!
!!
!
!!
!
!
!!
!
!
!
!
!
!!
!
!
!
! !
! ! !
!
!!
!
!
!
!
!
!
! ! !!
!!!!
!
!
!
!! !
!
!
!
!
!!
!
!
!!
!
!
!! !
!
!
!
!
!
!
!
!
!
!!
!
!
!!!
!!!
!
!! !
!
!
!!
!!
!
!
!
!
!
!!
!
!
!
!
!
!
! !
!
!
!
!
!
!
!
!
!
!!
!!
!!
!!
!
!
!
!!
!
!
!
!!
!
!!
!
!!
!
!
!
!
!
!
!
!
!
!
!
! !
!
!
!
!
!
!
!
!
!
!
!
!!
!
!!
!
!
!
!
!!!!
!
!
!
!!
!
!
!
!!
!
!
!
!
!
!
!
!
!
!
!
!!
!
! !
!
!!
!
!
!!
!!
!
!
!
!
!!!
!
!
!!
! !!
! !
! !
!
!!
!!
!
!
!!!
!
!
!
!
!
!
!
!
!
!
! !!
!
!
!!
!
!
!
!!
!
!
!
!
!!
!!
!!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!!
!!
!
!
!
!
!
!
!
!
!
!
!
500600
700
400
800
900
1000
300
1100
200
1200
100
1300
700400
400
100
1200
300
700
1000
500
500
1000
300
300
900
200
400
400
900 800
1000
1000
400
800
600
300
800
200
300
200
1000
400
900500
100
800
400
200
1200
1000
200
1000
200
100
300
500
1000
100
1000
1200
700
100
1100
700
1000
800
900
200
300
900
1000
100
1000
800
500
900
100
600
100
100
800
300
400
300
700
900
1200
700
300
600
700
600
1000
800
File
: 4.2
.3_l
ower
_thi
ckne
ss.m
xd
Source: McGowen et al. (1977)
State Line
Model Boundary
Downdip Aquifer Limit
County Boundaries
! McGowen Control Points
−0 25 50
Miles
Lower DockumThickness
(feet)50 - 100101 - 200201 - 300301 - 400401 - 500501 - 600601 - 700701 - 800801 - 900901 - 1,0001,001 - 1,1001,101 - 1,2001,201 - 1,3001,301 - 1,400
Upper Dockum Lower Dockum
Horizontal Hydraulic Conductivity – KH
Upper Dockum Lower Dockum
0 25 50
Miles
Upper Dockum Calibrated Horizontal
Hydraulic Conductivity (ft/d)
0.01 - 0.03
0.03 - 0.1
0.1 - 0.3
0.3 - 1
1 - 3
3 - 10
Active Boundary
State Line
County Boundaries
Inactive Cells
Active Boundary
State Line
County Boundaries
Inactive Cells
0 30 60
Miles
Lower Dockum Calibrated Horizontal
Hydraulic Conductivity (ft/d)
0.01 - 0.030.03 - 0.10.1 - 0.30.3 - 11 - 33 - 10
Vertical Hydraulic Conductivity – KV
Upper Dockum Lower Dockum
Active Boundary
State Line
County Boundaries
Inactive Cells
0 25 50
Miles
Upper DockumVertical
HydraulicConductivity
(ft/d)2.5e-4 - 3.0e-4
3.0e-4 - 3.5e-4
3.5e-4 - 4.0e-4
4.0e-4 - 4.5e-4
4.5e-4 - 5.0e-4
5.0e-4 - 1.0e-3
1.0e-3 - 5.0e-3
Active Boundary
State Line
County Boundaries
Inactive Cells (layer3)
0 25 50
Miles
Lower DockumVertical
HydraulicConductivity
(ft/d)2.5e-4 - 3.0e-4
3.0e-4 - 3.5e-4
3.5e-4 - 4.0e-4
4.0e-4 - 4.5e-4
4.5e-4 - 5.0e-4
5.0e-4 - 1.0e-3
1.0e-3 - 5.0e-3
StorativityUpper Dockum Lower Dockum
Active Boundary
State Line
County Boundaries
0 25 50
Miles
Storativity in Upper Dockum
1e4 - 3e-4
3e-4 - 1e-3
1e-3 - 3e-3
3e-3 - 1e-2
1e-2 - 3e-2
Active Boundary
State Line
County Boundaries
0 25 50
Miles
Storativity in Lower Dockum
1e-4 - 3e-4
3e-4 - 1e-3
1e-3 - 3e-3
3e-3 - 1e-2
1e-2 - 3e-2
General Head Boundary
19971980
Surface Water Boundary ConditionsUpper Dockum Lower Dockum
Recharge Rate Estimates
County/Area Land use Aquifer Recharge (in/yr) Technique Reference Dockum Aquifer – Colorado River outcrop area All of the Colorado River outcrop area - Predevelopment Grassland and shrubland Dockum 0.08 SZ chloride mass
balance This report
Sandy areas (Nolan and eastern Mitchell counties) - Predevelopment Grassland and shrubland Dockum 0.22 SZ chloride mass
balance This report
All of the Colorado River outcrop area Dockum 0.08 to 0.2 UZ numerical
modeling Scanlon et al. (2003)
Western Scurry and Mitchell counties Grassland and shrubland Dockum <0.01 SZ chloride mass balance This report
Scurry County - Predevelopment Dockum 0.02 to 0.04 Water budget on playas This report
All of the Colorado River Outcrop area - Postdevelopment Dockum 2.2 regional water level
rise This report
Sandy areas (Nolan and eastern Mitchell counties) - Postdevelopment Cropland Dockum
Geom. Average = 1.7Median = 1.6
Range = 0 to 4.3
linear water level rises in individual wells
This report
County/Area Land use Aquifer Recharge (inch/yr) Technique ReferenceDockum Aquifer – Canadian River outcrop area All of the Canadian River outcrop area- Predevelopment and Postdevelopment
Grassland and shrubland Dockum 0.17 SZ chloride mass balance This report
All of the Canadian River outcrop area Dockum <0.08 UZ numerical
modeling Scanlon et al. (2003)
Dockum Aquifer – high TDS outcrop area Howard, Borden and Garza counties - Predevelopment Grassland and shrubland Dockum <0.01 SZ chloride mass
balance This report
Predevelopment Recharge
Recharge was estimated on limited points (80) using:
The correlation of estimated recharge to physical parameters was tested, however, no obvious correlation was found
A significance analysis (t-test) was conducted to determine whether average recharge rates should be divided into regions; no significant difference existed
Recharge rates were weighted as a power function of the local topography and normalized to conserve total recharge
Power coefficient adjusted until maximum recharge rate was reasonable (~0.5 in/yr)
GW
P
ClClPR
Modern Recharge
Analysis of regional water-level rise was conducted–
Colorado River outcrop rise indicates 2.2 in/yr effective recharge–
Canadian River outcrop indicates no appreciable rise–
includes recovery, stream loss, land-use impacts, return flow
For non-pumped, interstream wells with linear water table rise:–
Recharge = ∆h/Sy∆t
where median = 1.6 in/yr–
Already have 0.15 in/yr historical recharge
Added 1.45 in/yr to cropland areas–
Only added within Colorado River outcrop–
Added to cells by percent cropland within cell
Recharge Distribution
ModernPredevelopment
Aquifer Discharge Through Pumping
Point Source Data–
Municipal, power, mining, manufacturing
Non-Point Source Data–
Irrigation, livestock, county-other (domestic)
Historical (1980-1997)–
TWDB water use survey database
Uses of Water (1980 – 1997 Average)
Irrigation69%
Municpal13%
Livestock6%
Rural Domestic4%
Mining7%
Manufacturing0.5%
Power0.5%
1980 – 1997 Pumpage
0
5000
10000
15000
20000
25000
30000
35000
40000
4500019
80
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
Year
Pum
page
(afy
) STKRDPWRMUNMINMFGIRR
Pumpage (AFY)
1980 1997
Active Boundary
State Line
County Boundaries
0 25 50
Miles
Pumpage in Lower Dockum
in 1980 (AFY)0 - 11 - 33 - 1010 - 3030 - 100100 - 300300 - 10001000 - 3000
Active Boundary
State Line
County Boundaries
0 25 50
Miles
Pumpage in Lower Dockum in 1997 (AFY)
0 - 11 - 33 - 1010 - 3030 - 100100 - 300300 - 10001000 - 3000
Calibration Approach
Define model objectives
Data compilation and analysis
Conceptual model
CALIBRATION
Reporting
Future WaterStrategies
Comparison with
field data
Model design
Field data
Field data
*Includes sensitivity
analysis
Transient*
Steady State*
Prediction
Model Evaluation
Model results compared to observed data–
Observed water levels
–
Stream flow Model results evaluated against literature
data–
Recharge
–
Hydraulic conductivitiesModel results evaluated against
conceptual model–
Flow from overlying aquifer
Model Calibration
Calibration required:–
Reducing Kh
of the upper and lower Dockum
by factor of 0.2
–
Reducing Kv
of the upper and lower Dockum
by factor of 0.5
–
Reducing streambed conductances
by a factor of 0.1
Upper Dockum Calibration – Predevelopment
(a)
2500
3000
3500
4000
4500
5000
2500 3000 3500 4000 4500 5000
Observed (ft)
Sim
ulat
ed (f
t)
!(#*!(
!( !( #* #*!( #* #*
#* #*
!(
#*#*
#* !(
#*
!(
!(!(
#*
!( !(
!(
#*36
00
3400
38004000
3200
4200
3000
2800
4400
4600
2600
4800
3000
Active Boundary
State Line
County Boundaries
−0 25 50
Miles
Simulated head in Upper Dockum in
Steady-state Model (ft)
ContoursDry Cells
Residuals!( -300 to -100!( -100 to -30!( -30 to -10!( -10 to 0#* 0 to 10#* 10 to 30#* 30 to 100#* 100 to 300
count 29RMSE 99.8 ftMAE 76.3 ftME 16.1 ftMin E -223.0 ftMax E 185.6 ft
Range 2403.6 ftRMSE/Range 4.15%MAE/Range 3.17%ME/Range 0.67%
Lower Dockum Calibration – Predevelopment
(a)
2000
2500
3000
3500
4000
4500
2000 2500 3000 3500 4000 4500Observed (ft)
Sim
ulat
ed (f
t)
count 180RMSE 73.0 ftMAE 53.3 ftME 23.6 ftMin E -135.5 ftMax E 274.4 ft
Range 2288.5 ftRMSE/Range 3.19%MAE/Range 2.33%ME/Range 1.03%
#*!(
#*#* #*#* #*
#*
#* #*#*#* #*#*!(#*
#*!(#*#* !(!( !(#*#* #*!( #*
!(#*
#* #*#*#* #*#*!(#*#*#* #*
#* !(
#*#* !(#* #*#*#*#*!(#* #*#*#*#*#*!(#*!(!(!(#*#*!(#*#*!(#*#*#*!(!(#*#*#*#*!(!(!(!(!(!(#*!(!(#*#*!(#*#*!( #*#* #* #*#*!( #*#*!(
#*!(#*#*#*#*!(
!(!(!( !(
!(#*#*
!( #*#*#*#*!(#*!(#*!( #*#*#*#*#*!(
!(!(!(#* !( !( !(!(!(#* #*!( !(#* #*#*!(#*#*#* !(#*!( #*#* !(#* !(!(!( !(
#*!(
#* !( #*!( !(
#*!(
#* #*#* !( !(
#*!(
#*
3800
4000
420046004800
5000
5200
5400
600064
00
3600
3400
3200
2200
2600
3000
4400 2800
2400
2800300032
00
2400
2600
2400
2200
2400
3600
2600
2400
3800
4200
2600
4200
2400
3200
2800
Active Boundary
State Line
County Boundaries
−0 25 50
Miles
Simulated head in Lower Dockum in
Steady-state Model (ft)
ContoursDry Cells
Residuals!( -300 to -100!( -100 to -30!( -30 to -10!( -10 to 0#* 0 to 10#* 10 to 30#* 30 to 100#* 100 to 300
Upper Dockum Calibration – Transient
count 25RMSE 82.2 ftMAE 65.0 ftME 56.6 ftMin E -60.7 ftMax E 163.7 ft
Range 2403.6 ftRMSE/Range 3.42%MAE/Range 2.70%ME/Range 2.35%
(a)
2500
3000
3500
4000
4500
5000
2500 3000 3500 4000 4500 5000Observed (ft)
Sim
ulat
ed (f
t)
Active Boundary
State Line
County Boundaries
0 25 50
Miles
Mean Residuals in Upper Dockum
in Transient model (ft)
-1000 to -300-300 to -100-100 to -30-30 to -10-10 to 00 to 1010 to 3030 to 100100 to 300300 to 1000
Lower Dockum Calibration – Transient
(a)
1500
2000
2500
3000
3500
4000
4500
1500 2000 2500 3000 3500 4000 4500Observed (ft)
Sim
ulat
ed (f
t)
count 1293RMSE 98.2 ftMAE 69.6 ftME 6.2 ftMin E -243.6 ftMax E 441.4 ft
Range 2288.5 ftRMSE/Range 4.29%MAE/Range 3.04%ME/Range 0.27%
Active Boundary
State Line
County Boundaries
0 25 50
Miles
Mean Residuals in Lower Dockum
in Transient model (ft)
-1000 to -300-300 to -100-100 to -30-30 to -10-10 to 00 to 1010 to 3030 to 100100 to 300300 to 1000
Subcrop Randall 1015901
3512
3532
3552
3572
3592
3612
3632
3652
3672
3692
3712
1980 1985 1990 1995 2000Year
Head
(ft
Subcrop Randall 1016102
3608
3618
3628
3638
3648
3658
3668
3678
3688
3698
3708
1980 1985 1990 1995 2000Year
Head
(ft
Upper Dockum Hydrographs
Outcrop Deaf 908301
4007
4017
4027
4037
4047
4057
4067
4077
4087
4097
4107
1980 1985 1990 1995 2000Year
Hea
d (ft
Subcrop Deaf 1001601
4044
4054
4064
4074
4084
4094
4104
4114
4124
4134
4144
1980 1985 1990 1995 2000Year
Hea
d (ft
Outcrop Hartley 726101
3200
3220
3240
3260
3280
3300
3320
3340
3360
3380
3400
1980 1985 1990 1995 2000Year
Head
(ft
Subcrop Armstrong 1105101
3170
3180
3190
3200
3210
3220
3230
3240
3250
3260
3270
1980 1985 1990 1995 2000Year
Hea
d (ft
Lower Dockum Hydrographs
Outcrop Mitchell 2934805
2105
2115
2125
2135
2145
2155
2165
2175
2185
2195
2205
1980 1985 1990 1995 2000Year
Head
(ft
Outcrop Scurry 2925707
2128
2138
2148
2158
2168
2178
2188
2198
2208
2218
2228
1980 1985 1990 1995 2000Year
Hea
d (ft
Subcrop Motley 2201915
2546
2556
2566
2576
2586
2596
2606
2616
2626
2636
2646
1980 1985 1990 1995 2000Year
Head
(ft
Subcrop Winkler 4623701
2655
2665
2675
2685
2695
2705
2715
2725
2735
2745
2755
1980 1985 1990 1995 2000Year
Head
(ft
Stream Gain/loss
19971990
Stream Results
Miles
0 10 20
Borden
Garza
Kent
Scurry
NolanMitchell
Sterling
Howard
Fisher
CrosbyDickens
Dawson
Lynn
Glasscock Coke
Stream-FlowStudy Number(s)
37-3942-4445-4852
08123720
08123800
08117995
0811
9500
08120700
08121000
Aquifer Downdip Limit
County Boundaries
Dockum Aquifer
Fresh-Water Outcrop
High-TDS Outcrop
Downdip
Model Boundary
Streamflow-Gaging Station08123720
Slade et al., 2002
-300
-250
-200
-150
-100
-50
0
50
100
150
200
Studies 37,38,39 Studies45,46,47,48
Study 42
Studies 43,44,52
Gai
n/Lo
ss (A
FY/m
ile)
Simulated
Study A
Study B
Study C
Study D
Comparison with Steady-State Model
Spring Flow
19971990
ET Flow
19971990
Flow into Top of Dockum
Active Boundary
State Line
County Boundaries
0 25 50
Miles
Flow into Dockum in Steady-state
Model (AFY)-300 to 100-100 tp -30-30 to -10-10 to 00 to 1010 to 3030 to 100100 to 300
Active Boundary
State Line
County Boundaries
−0 25 50
Miles
Flow into Dockum in Transient Model (AFY)
-300 to -100-100 to -30-30 to -10-10 to 00 to 1010 to 3030 to 100100 to 300
1997Predevelopment
Sensitivity Analysis
-15
-10
-5
0
5
10
15
0.4 0.6 0.8 1 1.2 1.4 1.6
Fraction of Base Value
Mea
n He
ad D
iffer
ence
(ft) Kh-all
Kh-OgallalaKh-Upper-DockumKh-Lower-DockumKv-allKv-OgallalaKv-Upper-DockumKv-Lower-Dockum
Example: Effect of Hydraulic Parameters on Lower Dockum
Heads
Conclusions
GAM for Dockum Aquifer:–
Incorporated all relevant features, data on aquifer properties, recharge estimates, and pumpage
–
Calibrated to specifications: steady-state (prior to 1950) transient conditions (1980-1997)
–
Required some adjustment of properties during transient calibration (not beyond measured data)
Data Models
background
Consistent methodology for storage of GAM data
Facilitates future improvements or modifications of current work
Available to the general public as an addition to the final reports
Data Models basic structure
srcdata
– contains the source and some derived data used to generate the model input data sets
pumpage
– contains tools for generating pumpage inputs along with output tables
Modflow
– contains all of the actual model input and output data files
Data Models srcdata
-
examples
Boundary
– geopolitical boundaries
Climate
– climatic data
Conservation
– natural and ecological boundaries
Geology
– subsurface geology, outcrop delineation
Geomorphology
– physiography and elevation
Geophysics
– well logs
Model*– model inputs and results on the model grid
Recharge
– direct and irrigation recharge
Soil
– STATSGO data, runoff numbers
SubsurfaceHydro
– pumping rates, hydraulic
conductivities, water levels, hydrographs
SurfaceHydro
– streamflows, stream/aquifer
interaction, springflows
Transportation
– roads
Data Models modflow
mf2k•
Input –
ASCII input data sets for running modflow
from the command line•
Output –
All output data sets
vistas•
Data sets for running the models from groundwater vistas interface along with output
postprocessors•
layerMB.pl
–
calculates layer water budgets
•
countyMB.pl
–
calculates budgets by county or GCD
Limitations & Applicability of the GAM
The GAM is a tool capable of being used to make groundwater availability assessments on a regional scale
The model is well suited for studying institutional water resource issues
The model would likely require refinement to study operational issues for a specific project
The GAM allows regional consideration of interference between resource strategies
Limitations & Applicability of the GAM
The GAM scale of application is for areas of many square miles.– The GAM produces water levels representative of
large volumes of aquifer (e.g., 5,280 ft X 5,280 ft X100 ft aquifer thickness)
The GAM is not capable of predicting aquifer responses at particular point such as a particular well – The model is well suited for refinement to address
local-scale, operational water resource questions.
Model Grid Scale
Lea
Eddy
Chaves
Pecos
Union
Quay
Colfax
Culberson
Reeves
Mora
San Miguel
Ellis
DeBaca
Harding
Guadalupe
Texas
Curry
Roosevelt
Irion
Hall
Crockett
Lincoln
Hale
Dallam
Gaines
Jeff Davis
Upton
Hartley
King
Kent
Beaver
Oldham Gray
Lynn
FloydLamb
Terry
Coke
Andrews
Ector
Knox
Kiowa
Nolan
Ward
Jones
Potter
Cottle
Taylor
Brown
Custer
Motley
Reagan
Young
Dewey
Garza
Martin
Fisher
Coleman
Baylor
Cimarron
Moore
Scurry
Archer
Tom Green
Castro
Bailey
Deaf Smith
Mills
Donley
Crane
Runnels
Carson
Concho
Crosby
Llano
Woods
Schleicher
Washita
Borden
Haskell
Tillman
Randall
Sterling
Foard
BriscoeParmer
MitchellHoward
Menard
Roberts
Woodward
Major
Midland
San Saba
DickensHockley
Swisher
Harper
Winkler
Mason
Greer
Dawson
Wheeler
EastlandOtero
Lubbock
Hemphill
McCulloch
Roger Mills
Ochiltree
Loving
Wilbarger
HansfordSherman
Callahan
Yoakum
Jackson
Beckham
Lipscomb
Stephens
Blaine
Stonewall
Cochran
Armstrong
Glasscock
Wichita
Hutchinson
Shackelford
Childress
Harmon
Hardeman
Comanche
Sutton Kimble
udspeth
Caddo
Collingsworth
Throckmorton
Comanche
Alfalfa
Cotton
Terrell
Palo Pi
Taos
Era
Presidio Brewster
Lampa
Cla
Ham
LegendActive boundary
State Line
County Boundaries
Model grid
0 25 50
Miles
Scurry County
Dockum Model Limitations
Limitations of supporting data– Limited head targets spatially and temporally– Limited Stream/aquifer gain loss estimates– Limited data quantifying x-formational flow– Limited hydraulic property data in areas
Assumptions– GHB used to describe overlying aquifers– Spatial variation in recharge– Lack of temporal variation in recharge– No flow to underlying formations
Dockum Model Limitations (continued)
Model applicability– Regional scale
Applicable at scale of tens of miles
– Annual stress period
Not applicable to seasonal trends
– Low TDS regions
MODFLOW does not account for density- dependant flow in high TDS regions
Dockum Model Limitations (continued)
Model applicability– Surface water/groundwater coupling
The model does not provide a rigorous solution to surface water modeling for GAMs
– Water-quality issues
Only a preliminary assessment of water quality is conducted for the GAMs
Future Improvements
Additional supporting data– Water-level monitoring in areas with sparse
measurements– Recharge studies– Long-term surface water/groundwater studies– Additional study of structure– Additional estimates of flow from overlying
aquifers– Additional data describing seasonal trends
Future Implementation Improvements
Account for density-dependant flow– If simulations are to include development of high TDS
portion aquifer, a different simulator (e.g., SEAWAT) may be warranted
Account for flow to underlying units– If water-level data become available to describe
conditions in underlying Permian formation, could add a fourth model layer
Grid Refinement– If local effects of a small portion of the aquifer are to
be modeled, a local-scale model with a refined grid should be considered
Dockum GAM 4th Stakeholder Training Seminar
October 22, 2008
Sign-in Sheet
Name Affiliation
Melanie Barnes LWV
Jim Conkwright HPWD
Harvey Everheart Mesa UWCD
John Ewing INTERA
Michelle Guelker Lone Wolf GCD
Ian Jones TWDB
Steve Walthour North Plains GCD
Bruce Blalack City of Lubbock
Emmett Autrey City of Amarillo
Ferrel Wheeler Garza Co. U&FWCD
Aubrey Spear City of Lubbock