Transcript Investigating the Colorado River Simulation Model

Investigating the Colorado
River Simulation Model
James Prairie
Bureau of Reclamation
Motivation
Colorado River Basin
– arid and semi-arid climates
– irrigation demands for agriculture
Federal Water Pollution Control Act
Amendments of 1972
“Law of the River”
– Minute No. 242 of the International Boundary and
Water Commission dated August 30, 1973
– Colorado River Basin Salinity Control Act of 1974
Salinity Damages and Control
Efforts
 Damages are presently, aprox. $330
million/year
 As of 1998 salinity control projects has
removed an estimated 634 Ktons of salt from
the river
– total expenditure through 1998 $426 million
 Proposed projects will remove an additional
390 Ktons
– projects additional expenditure $170 million
 Additional 453 Ktons of salinity controls
needed by 2015
Data taken from Quality of Water, Progress Report 19, 1999 & Progress Report 20,2001
Seminar Outline
Motivation for research
Initial findings
Working with a case study
New salinity modeling techniques
Extending knowledge of our case study
Current Efforts
Recompute Natural flow
Verify entire Colorado River Simulation Model
Future Research
Colorado River below Imperial Dam, Ariz-Calif
950
40,000
Flow
850
30,000
750
20,000
650
10,000
550
1940
0
1945
1950
1955
1960
1965
1970
1975
1980
1985
1990
1995
Flow 1,000 (ac-ft)
TDS (mg/L)
TDS
Research Objectives
Verify the data and calibrate the current
model for both water quantity and water
quality (total dissolved solids, or TDS)
Investigate the salinity methodologies
currently used and improving them as
necessary for future projection
Investigation of Colorado River
Simulation Model
First developed in Fortran in 1970’s
Moved to RiverWare in 1990’s
Relies on Conservation of Mass for modeling
water quantity and,
water quality (TDS).
Monthly Time Step
Runs with operational rules to simulate
operational policies in the Colorado River
Basin
Initial Findings
Data and Methodological Inconsistency
Need to improve current model techniques
Stochastic stream flow simulation
Estimating natural salt
Adding uncertainty
Working with a case study
Detailed investigation of current methods
Development of new methods
Case Study Area
• Historic flow from 1906 - 95
• Historic salt from 1941 - 95
USGS gauge 09072500
(Colorado River near Glenwood Springs, CO)
USGS Salt Model
12 monthly regressions
based on observed historic flow and salt
mass from water year 1941 to 1983
historic salt = f (historic flow, several
development
variables)
natural salt = f (natural flow, development
variables set to zero)
Existing Salt Model
Over-Prediction
New Modeling Techniques
Found problems with the current method
to estimate natural salt in the upper basin
Can we fix the problem?
Alternate methods the estimate natural salt
with the available data
Statistical Nonparametric Model
for Natural Salt Estimation
Based on calculated natural flow and natural
salt mass from water year 1941-85
 calculated natural flow = observed historic flow
+ total depletions
 calculated natural salt = observed historic salt
- salt added from agriculture
+ salt removed with exports
Nonparametric regression (local regression)
natural salt = f (natural flow)
Residual resampling
Local Regression
alpha = 0.3
Y
or 27 neighbors
X
Residual Resampling
Y
y = y* + e*
y*
e*
x
X
Nonparametric Salt Model and
USGS Salt Model
Natural Salt Mass from Nonparametric
Salt Model and USGS Salt Model
USGS Salt Model and New Salt Model
with K-NN Resampling Comparison
Comparison with
Observed Historic Salt
Key Case Study Findings
The new nonparametric salt model
removed the over-prediction seen with the
USGS salt model
Provides uncertainty estimates
Can capture any arbitrary relationship
(linear or nonlinear)
Extending from Case Study
Applying case study results to entire
model
Improved natural salt estimation model
Improved stochastic stream flow generation
Addition of uncertainty analysis
Ensure flexible framework
Current Efforts
Recomputing natural flow
Lack of base data
Undocumented procedure
Upper versus Lower Basin
Inconsistency across time periods
Inconsistent with future projection model
Natural Flow Development
Natural flow is a basic input for CRSS
Addressing Data Inconsistencies
Recomputing natural flow from 1971-95
Natural Flow = Historic Flow
- Consumptive Uses and Losses
+/- Reservoir Regulation
Addressing Methodological Inconsistencies
RiverWare model computes natural flow
Ensures consistency
Recomputing natural
flow from 1971-95
Data required for natural flow model
Historic USGS gauge data
29 gauges
Historic main-stem reservoir outflow and pool
elevations
12 main-stem reservoirs
Historic off-stream reservoir change in storage
22 off-stream reservoirs
Consumptive uses and losses
9 categories
Determining Natural Flow
calculated natural flow
flow
historic agriculture
consumptive use
estimated natural salt mass
salt
irrigated
lands
agricultural
salt loadings
historic exports
salt removed
with exports
historic municipal and industrial
historic effects of off-stream
reservoir regulation
USGS stream gauge 09072500
simulated historic flow
simulated historic salt mass
After Natural Flow Is
Calculated
 Extend new natural salt model
throughout the 21 upper basin natural
flow gauges
 Apr-Jul
2003
 Check natural flow and salt relationship
 1941-1995; 1970-1995
 Calculate natural flow for the lower
basin
 Aug-Sep
2003
 Natural salt is back calculated as flow
 Verify entire basin for flow and the
lower basin for salt
 Finally use more for a future
projection
 Oct-Nov
2003
 Dec 2003
Future Research
Explore salinity relationship over both
space and time
Incorporate new stochastic flow
generation methods
Investigate land use change and the
impacts on salinity levels
Explore the relationship between basin
area and both flow and salinity
Acknowledgements
 Dr. Balaji Rajagopalan, Dr. Terry Fulp, Dr. Edith
Zagona for advising and support
 Upper Colorado Regional Office
of the US Bureau of Reclamation,
in particular Dave Trueman for
funding and support
 CADSWES personnel for use of their
knowledge and computing facilities
Drainage Area
Colorado River
Basin
241,000 mi2
Upper Basin
110,000 mi2
Case Study
4,558 mi2