PEC OS
W EBB
BR EW STER
H UD SPET H
PRESID IO
REEVES
CU LBER SON
VAL VERD E
DU VAL
TER R ELL
C R OC K ETT
KEN ED Y
FRIO
H ARR IS
H I LL
BELL
BEE
C LAY
POLK
ED W ARD S
J EFF D AVIS
KER R
GAIN ES
LEON
U VALD E
H ALE
D ALLAM
IR ION
D IMMIT
LAMB
KIN G
BEXARKIN N EY
STAR R
H ALL
W ISEJ AC K
U PTON
HID ALGO
S U TTON
C ASS
OLD H AM
ELLIS
MEDIN A
KIMBLE
Z AVALA
KEN T
R U SK
LEE
LYNN
GR AY
COKE
L A SALLE
MILAM
ER ATH
HAR TLEY
HU N T
BR AZO R IA
SMITH
KN OX
FLOYD
LLAN O
A N DR EW S
TYLER
TR AVISLIBERTY
JON ES
N U ECES
R EA GAN
BOW IE
W AR D
ZAPATA
LAMAR
R EAL
N OLAN
TER R Y GA R ZA
MILLS
C OLEMAN
EC TOR
TOM GR EEN
MASON
YOU N G
FALLS
C AMER ON
MAT AGOR DA
H AYS
BR OW N
C OOKE
JASPER
D EAF SMITH
BUR N ET
M AVERIC K
H OU STON
LAVACA
FISH ER
C OLLIN
MOOR E
FANN IN
M OTLEY
MAR TIN
L IVE OAK
D ALLAS
EL PASO
BAILEY
B OSQU E
H AR D IN
KLEBER G
JIM H OGG
TAYLOR
C OT TLE
POTTER
D ON LEY
GOLIAD
SAN SABA
ATASC OSA
D EN TON
COR YELLCR AN E
CON C HO
BAYLOR
DE W ITT
BR OOKS
PAR KER
R U NN ELS
N AVAR R O
ARC H ER
CAR SON
CASTR O
W OOD
SC U RR Y
C R OSBY
FAYETTE
Mc MULL EN
W HAR TON
BORD EN
C ALHOU N
SH ELBY
MEN ARD
GILLESPIE
PAR MER
W ILSON
D IC KEN S
SC H LEIC H ER
GR IMES
F OARD
PAN OLA
H ASKELL
BR ISC OE
R AN D ALL
DAW SON
MID LAN D
H OWAR D
Mc LEN NAN
GON ZALES
GR AYSON RED R IVER
SWISH ER
R OBER TS
H OCKLE Y
AN D ER SON
TARR AN T
W ALKER
L U BBOC K
VIC TORIA
BASTR OP J EFFER SON
SH ER MAN
W H EELER
MITC HELL
YOAKU M
STER LIN GWIN KL ER
TR IN ITY
H EMPH ILL
W ILBA R-GER
COMAN -CHE
WA
LLER
KAR N ES
LIPSC OMB
J AC KSON
W ILLIAMSON
R EFUG IO
W ILLAC Y
LOVING
AU STIN
EASTLAN D
H OPKIN S
Mc CU LL OCH
BLAN C O
H AR RISONSTEPHEN S
ANGELIN A
C ALLAH AN
C OLOR AD O
HAN SFOR D
KAU FMAN
BAN D ER A
PALO PINTO
MON T AGU E
H AMILTO N
OCH ILTR EE
C OM AL
LIMEST ON ESABIN E
COC H RAN
FOR T BEN D
CH AMBER S
VAN ZAN D T
WIC H ITA
J OH N SON
STONEW ALL
H EN DER SON
TITUS
FR EESTON E
MON TG OMER Y
HOOD
KEN D ALL
BR AZ OS
GALVESTON
U P SH U R
H UTC H INSON
LAMPAS AS
BU R LESON
H AR D EMAN
GU ADALU PE
CH ILD -RE SS
AR M-STRON G
GLAS S-COCK
N AC OG-D OC HES
MARION
C ALD W ELL
MAD ISON
ORAN GE
D ELTA
R AIN S
GRE GG
C AMP
MO
RR
IS
FR
ANKLIN
ROCK -WA LL
C OLLI NGS -W ORTH
THR OCK-MOR TON
S OM E R-V E L
SANAUGUS -
TI NE
S ANJ AC IN TO
J IMW ELLS
S HA CKEL-FOR D
C HE RO-KE E
NEW
TON
ROBE RT-SON
WA SH IN G-TON
ARA N-SASSAN
PATR IC I O
80
79
67
78
85
82
6876
69
81
75
77
87
71
72
73
74
84
86
83
46
47
66
53
45
54
56
50
65
38
64
41
51
16
59
42
49
39
63
48
40
61 57
44
37 55
58
62
43
60
52
36
70
29
6
14
17
35
15
9
25
3328
8
7
12
3
345
26
32
30
23
21 10
11
27
20
18
22
19
24
13
2
31
4
1
PECOS
WEBB
BREWSTER
HUDSPETH
PR ESIDIO
REEVES
CULBERSON
VAL VERDE
DUVAL
TERR ELL
CROCKETT
FRI O
HARRIS
HILL
BELL
BEE
KENEDY
CLAY
POLK
ED WARDS
JEFF DAVI S
GAINES
LEON
KERR
UVALDE
HALE
DALLAM
KI NG
IRION
LAMB
DIMMI T
BEXARKI NNEY
STARR
HALL
JACK
CASS
WISE
SU TTON
OLDHAM
HIDALGO
ELLI S
UPTON
ZAVALA
MEDI NA
KI MBLE
RUSK
LEE
LYNN KEN T
GRAY
LA SALLE
COKE
MILAM
ERATH
HARTLEY
HUNT
SMITH
KN OX
FLOYD
LLANO
TYLER
BR AZORIA
AN DREWS
TRAVI S LIBERTY
REAGAN
JONES
ZAPATA
LAMAR
BOWI E
NUECES
WARD
REAL
NOLAN
TERR Y GARZA
COLEMAN
MILLS
EC TOR
YO UNG
TOM GREEN
MASON
FALLS
MAVERICK
BU RNET
HAYS
DEAF SMITH
JASPER
LAVACA
HOUSTON
COOKE
FISHER
BROW N
COLLIN
MOORE
MOTLEY
FANN IN
MARTIN
EL PASO
BAI LEY
DALLAS
LIVE O AK
BOSQUE
HARDIN
JIM HO GG
TAYLOR
CAMERON
PO TTER
GOLIAD
CRANE
COTTLE
DONLEY
ATASCOSA
SAN SABA
DENTON
CORYELL
BAYLOR
CONC HO
BROOKS
RUNNELS
PARKER
NAVARRO
ARCHER
DE WI TT
CARSON
SCURRY
MATAGO RDA
CROSBY
KLEBERG
FAYETTE
SH ELBY
WOOD
CASTRO
BORDEN
MENARD
WHARTON
NEWTON
PAR MER
GILLESPIE
MCMULLEN
DICKENS
SCHLEI CHER
FO ARD
HASKELL
PANOLA
GRIMES
MIDLAND
WILSON
RANDALL
BR ISCOESWISHER
DAWSON
GRAYSON
GONZALES
HOW ARD
RED RI VER
ROBERTS
HOCKLEY
TARR ANT
ANDERSON
MCLENNAN
LUBBO CK
CALHO UN
CHEROKEE
VI CTORI A
BASTRO P
WALKER
SHERMAN
YO AKUM
MITCH ELL
STERLING
HEMPHI LL
WHEELER
KARNES
TRI NITY
WINKLER
JACKSON
LIPSCOMB
LOVI NG
WILLIAMSON
AUSTI N
EASTLAND
REFUGIO
HOPKIN S
HARRISON
BLANCO
CALLAHAN
COLORADO
AN GELINA
MCCULLOCH
STEPHENS
WILLACY
JEFFERSO N
KAU FMAN
BAN DERA
HANSFORD
COMANCHE
MONTAGUE
PALO PINTO
JIM WELLS
LIMESTONE
COMAL
HAMI LTON
OCHI LTREE
WILBARGER
SABI NE
COCHRAN
CHAMBERS
FORT BEND
VAN ZANDT
HENDERSON
STO NEWALL
JOHNSON
FREESTON E
MONTGOMERY
GLASSCO CK
KENDALL
TITUS
BRAZO S
HOO D
WICHITA
ARMSTRO NG
UPSHUR
ROBERTSON
HUTCHINSO N
LAMPASAS
CHILDRESS
WA
LLER
NACOG DOC HES
SH ACKELFORD
BURLESON
HARDEMAN
GUADALUPE
GALVESTON
MARI ON
THROCKMORTO N
COLLINGSWO RTH
MADI SON
CALDW ELL
SAN PATR ICIO
SAN JACI NTO
AR ANSAS
WASHINGTON ORANGE
DELTA
RAINS
GREGG
SA
N A
UG
US
T INE
CAMP
MO
RR
IS
FRAN
KLIN
SOMER-VELL
ROCK-WALL
Region F Brazos G
Panhandle
Far West Texas
Region C
East Texas
Region H
Llano Estacado
Plateau
Rio Grande
South Central Texas
Region B
Coastal Bend
Lower Colorado
North East Texas
Lavaca
P
D
K
N
B
M
L
J
O
H
I
C
E
A
GF
Groundwater Availability Modeling (GAM)
Northern Gulf Coast Aquifer GAM7th Stakeholder Advisory Forum
July 24, 2003Contractor: USGS
Funding: TWDB, HGCSD &USGS
Contract ManagerAli Chowdhury, Ph.D.
Texas Water Development Board
GAM• Purpose: to develop the best possible
groundwater availability model with the available time and money.
• Public process: you get to see how the model is put together.
• Freely available: standardized, thoroughly documented, and available over the internet.
• Living tools: periodically updated.
What is a Groundwater Model?
An aquifer in a computer, a tool to estimate field conditions
Effective use of available data and account forcomplexities
Expands our ability to better understand and managethe water resources
Increases prediction accuracy of future events to a level far beyond “best judgement” decisions
Modeling Protocol
Purpose
Conceptual model
Numerical formulation
Model design
Steady-State ModelCalibration
VerificationTransient Model
(1980-2000)
Comparisonwith
field dataWe are here!
Prediction Runs (2001-2050)Prediction
PostauditField data
Model Grid
Coarse GridFine Grid
Data Points
data is interpolated (Kriging) between measuredpoints where data is missing
model area discretized into cells
cells are populated with field datawhich are sparse but each modelcell needs a value
higher correlation between points at small sepration distance.kriging prserves the field value at the measurement point
SandClay
Porosity, Storage, andHydraulic Conductivity
groundwaterflow
High effective porosity/High KStorage
drainable (unconfined)compressible (confined)
High flow velocityBetter water quality
sand grainpore space
Dead endpores
Low effective porosity/low storageLow KLow flow velocityPoor water quality
Kh
Kv
Kh
Kv
Kh
Kv
Porosity: pore space/total voids in a rockStorage: volume of water released per unit
decreases in headHydraulic conductivity: ability to transmit water
Drawdown Cones in Sand and Clays
Drawdown curve
impervious rock
Ground surface
SWL
impervious rock
Drawdown curve
impervious rock
Ground surface
SWL
impervious rock
SandSandyClay
Q Q
• Broad vs. Steep Drawdown Cones in Sandy vs. Clayey aquifer• Subsidence due to clay compaction
Gaining vs. losing stream
Ground surfaceStream
(Gaining)
Ground surfaceStream(Losing)
Aquifer
Aquifer
water level
water level
Package In Out In OutRecharge 42,831,196 -- 44.3% --Streams 52,771,836 83,477,256 54.6% 86.3%
GHB 0 12,673,691 0.0% 13.1%ET -- -- -- --
Reservoirs 1,105,536 0 1.1% 0.0%Drains -- 557,711 -- 0.6%
Total: 96,708,568 96,708,658
Flow (ft^3/day) PercentageWATER BALANCE
Package In Out In OutRecharge 42,831,196 -- 44.3% --Streams 52,771,836 83,477,256 54.6% 86.3%
GHB 0 12,673,691 0.0% 13.1%ET -- -- -- --
Reservoirs 1,105,536 0 1.1% 0.0%Drains -- 557,711 -- 0.6%
Total: 96,708,568 96,708,658
Flow (ft^3/day) PercentageWATER BALANCE
Recharge
• diffuse (direct) - precipitation or irrigation• focused or localized - surface depressions, e.g. lakes or playas• indirect recharge - beneath rivers, lakes• recharge rate depends on rainfall, vegetation, soil type, topography
Recharge for the Gulf Coast aquifer Source Recharge (in/yr) Groschen (1985)
0.06
Ryder (1988) 0 to 6 Dutton and Richter (1990) 0.1 to 0.4 Noble and others (1996) 6 Hay (1999) .00004 to .04 Harden and Associates (2001) 0.1 to 0.2
Average annual rainfall map60 inches in the east to 8 inches
in the west
Pumping• Historical (pre-
development, 1980-2000)• Predictive (2000-2050)
Categories• municipal• manufacturing• domestic• irrigation• livestock
What isgroundwateravailability?
• …the amount of groundwater available for use.• safe yield
• average recharge• recharge and change in storage
• systematic depletion
• The State does not decide how much groundwater is available for use: GCDs and RWPGs decide.
• A GAM is a tool that can be used to assess groundwater availability once GCDs and RWPGs decide how to define groundwater availability.
Do we haveto use GAM?
• Water Code & TWDB rules require that GCDs use GAM information. Other information can be used in conjunction with GAM information.
• TWDB rules require that RWPGs use GAM information unless there is better site specific information available
• The model– predict water levels and flows in response to
pumping and drought– effects of well fields
• Data in the model– water in storage– recharge estimates– hydraulic properties
• GCDs and RWPGs can request runs
How do weuse GAM?
Livingtools
• GCDs, RWPGs, TWDB, and others collect new information on aquifer.
• This information can enhance the current GAMs.
• TWDB plans to update GAMs every five years with new information.
Hydrogeology, Simulation of Hydrogeology, Simulation of GroundGround--Water Flow, and LandWater Flow, and Land--
Surface Subsidence in the Chicot, Surface Subsidence in the Chicot, Evangeline, and Jasper Aquifers, Evangeline, and Jasper Aquifers,
Houston Area, TexasHouston Area, TexasMark C. Kasmarek, James L. Robinson, and Eric W. StromMark C. Kasmarek, James L. Robinson, and Eric W. Strom
In Cooperation with the Texas Water In Cooperation with the Texas Water Development Board and the HarrisDevelopment Board and the Harris--Galveston Galveston
Coastal Subsidence District Coastal Subsidence District
TWDB Ground-Water Availability Models in Texas
Modified from TWDB website
GAM Upper Gulf Coast Aquifer Outcrops
Stratigraphic and Hydrologic Sections
2000 Chicot Water-Level Altitude
2000 Chicot Observed vs. Simulated Target Heads
2000 Chicot Statistics
• The root mean square error was 24.47 feet between the measured and simulated hydraulic head.
• The maximum hydraulic-head drop across the model layer was 780 feet.
1977 Chicot Water-Level Altitude
1977 Chicot Observed vs. Simulated Target Heads
1977 Chicot Statistics
• The root mean square error was 36.30 feet between the measured and simulated hydraulic head.
• The maximum hydraulic-head drop across the model layer was 599 feet.
Chicot Well JY-65-18-103
2000 Evangeline Water-Level Altitude
2000 Evangeline Observed vs. Simulated Target Heads
2000 Evangeline Statistics
• The root mean square error was 33.72 feet between the measured and simulated hydraulic head.
• The maximum hydraulic-head drop across the model layer was 594 feet.
1977 Evangeline Water-Level Altitude
1977 Evangeline Observed vs. Simulated Target Heads
1977 Evangeline Statistics
• The root mean square error was 56.54 feet between the measured and simulated hydraulic head.
• The maximum hydraulic-head drop across the model layer was 681 feet.
Evangeline Well LJ-65-14-602
2000 Jasper Water-Level Altitude
2000 Jasper Observed vs. Simulated Target Heads
2000 Jasper Statistics
• The root mean square error was 33.41 feet between the measured and simulated hydraulic head.
• The maximum hydraulic-head drop across the model layer was 586 feet.
Jasper Well LJ-65-07-905
2000 Composite of Observed vs. Simulated Target Heads
1977 Composite of Observed vs. Simulated Target Heads
• VOLUMETRIC BUDGET FOR ENTIRE MODEL AT END OF TIME STEP 1 IN STRESS PERIOD 69CUMULATIVE VOLUMES L**3 RATES FOR THIS TIME STEP L**3/T
IN: IN:STORAGE = 848642637824.0000 STORAGE = 22069346.0000
• CONSTANT HEAD = 0.0000 CONSTANT HEAD = 0.0000 • WELLS = 0.0000 WELLS = 0.0000• HEAD DEP BOUNDS = 1.2630E+13 HEAD DEP BOUNDS = 100652320.0000• INTERBED STORAGE = 378706395136.0000 INTERBED STORAGE = 6822997.0000• TOTAL IN = 1.3857E+13 TOTAL IN = 129544664.0000
• OUT: OUT:• STORAGE = 621177274368.0000 STORAGE = 1331763.375• CONSTANT HEAD = 0.0000 CONSTANT HEAD = 0.0000
WELLS = 2.2251E+12 WELLS = 114226888.0000• HEAD DEP BOUNDS = 1.1006E+13 HEAD DEP BOUNDS = 13902051.0000• INTERBED STORAGE = 4949088768.0000 INTERBED STORAGE = 126298.7891• TOTAL OUT = 1.3857E+13 TOTAL OUT = 129587000.0000• IN - OUT = 306184192.0000 IN - OUT = -42336.0000
• PERCENT DISCREPANCY = 0.00 PERCENT DISCREPANCY = -0.03
Hydrogeology, Simulation of Hydrogeology, Simulation of GroundGround--Water Flow, and LandWater Flow, and Land--
Surface Subsidence in the Chicot, Surface Subsidence in the Chicot, Evangeline, and Jasper Aquifers, Evangeline, and Jasper Aquifers,
Houston Area, TexasHouston Area, TexasMark C. Kasmarek, James L. Robinson, and Eric W. StromMark C. Kasmarek, James L. Robinson, and Eric W. Strom
In Cooperation with the Texas Water In Cooperation with the Texas Water Development Board and the HarrisDevelopment Board and the Harris--Galveston Galveston
Coastal Subsidence District Coastal Subsidence District
Attendance at the 7th Stakeholder Advisory Forum, Northern Gulf Coast GAM
Participant Affiliation Jim Adams SJRA Bob Pickens Region K, Colorado County Ali Chowdhury TWDB Eric Strom USGS Mark C. Kasmarek USGS Mark Lowry TC&B; H, K and P RWPGs Haskell Simon Coastal Plains GCD Michael Klaus Citizen John Nelson LBG-Guyton Associates
Q & A’s at the 7th Stakeholder Advisory Forum, Northern Gulf Coast aquifer Groundwater Availability Model, July 24, 2003 Question: Are all of the rivers or just some stretches are gaining in the model area? Response: The recently completed USGS baseflow study suggests that all of the rivers are gaining within the model area. This observation is also supported by model results. USGS will distribute the baseflow study report to the stakeholders who requested the document. Question: How manufacturing pumpage is spatially distributed for the predictive runs? Response: Based on demand numbers provided by the RWPG and distributed around historical uses. Question: The statute and the TWDB rules require that the GCDs and RWPGs use GAM. In the rules “shall” is used, is it being modified by the legislature to offer more flexibility for the GCD’s? Who would be the honest arbiter for deciding what model to use? Response: TWDB approves management plans for the GCDs. The rules allow use of GAM in conjunction with other information. If model results with detailed site-specific information are available that was not included in the GAM, a GCD can provide this for TWDB consideration. Question: Why drawdown presented in Wharton County is not the same as was produced by Dutton model? Drawdown should be presented in the report to make it easier for people to compare water level decline between different time periods. Response: The map shows altitude of water levels but not drawdown. Drawdown maps that would be constructed should show the same levels of drawdown. Drawdown will be reported for each layer by decade. Question: Is the transient calibration complete? Response: Transient calibration is complete unless predictive runs produce results that require revisiting the transient calibration. Question: Is low transmissivity or low storage causing no fluctuations in the hydrographs? Response: We ran simulations with a wide range of transmissivity and storage values and selected the model that produced the best RMS. Question: Can the next SAF meeting after the draft report is submitted so that stakeholders can provide feedback to the consultant after reading the report? Response: Yes. The next SAF meeting will be held after the draft report is submitted at the end of September to facilitate review comments and feedback from the stakeholders. The draft report will also be posted on the web. Question/Comment: One stakeholder stated that some RWPG’s maintain that no additional groundwater models are necessary, as GAMs have already been developed for their area. It was discussed how best to improve the model. Most agreed that model improvements can best occur by collecting more data and populate the data to a finer grid to address local well issues. Models get better as new data is collected and our understanding of the flow system improves.