WSP05265

• File Number: 8170
• Description: Arkansas Basin Water Quality Issues
• State: CO
• Basin: Arkansas
• Water Division: 2
• Date: 2/1/1998
• Author: Goff Lewis Person Ko
• Title: Simulated Effects of Irrigation on Salinity in the Arkansas River Valley in Colorado
• Water Supply Pro - Doc Type: Report/Study

r

1751

processes of dispersion, diffusion, and advection. the lumped para~
meter model is eaiier to apply to the entire lower valley because of
the reduced data'requirements. Hukkinen (1993) suggests another
modeling option that has simpler data requirements. He suggests that
a "microscaJe" local model such as the Konikow and Bredehoeft
(1978) model could be used for different areas within the Arkansas
River valley. Then the results could be combined and used as input
to a "macroscale" model such as the Interactive Accounting Model
estabushed for the Arkansas River valley (Bums 1989). This would
allow both site-specific and generalized basin-wide data to be
used. By using this process, the basin-wide water quality impact of
local inigation practices could be anticipated in a more realistic fash-
ion than if only generalized basin-wide data were used (Hukkinen
(993). In the absence of additional detaile<! studies and because some
significant differences exist between the characteristics of the study
area and those of the valley both upstream and downstream from
the study area, reliable quantitative extrapolations from the detailed
study area to the entire valley cannot be made.
One finding from this study is that there is a relatively large
improvement in water quality resulting from all of the proposed
change in irrigation practices. This strongly suggests that, with
more widespread reductions in irrigation, downst.ream irrigators
could realize some benefit from lower salinity irrigation water.
These benefits could include increases in crop yields. and the abil-
ity to grow crops that are less salt tolerant and have a higher cash
value.

Conclusions
The quantity and quality of streamflow and ground water in the
lower Arkansas River valley of Colorado are closely associate<! with
agricultural irrigation practices. Results of ground water flow and
solute transport modeling indicate that a decrease in withdrawals
from ground water sources or a decrease in irrigated acreage could
substantiaUy affect ground water salinity as well as affect the quan-
tity and quality of water in the Arkansas River.
Three scenarios of decreased ground water pumping for irri-
gation were simulated: (I) 25% decrease; (2) 50% decrease; and (3)
100% decrease. The major results of these scenarios include the fol-
lowing:
a. Decreased pumping resulted in decreased salinity in the aquifer.
b. Complete cessation of pumping in the study area resulted in a
150 mgIL (-6.9%) decrease in average monthly ground water
salinity; a lO mgIL (-0,6%) decrease in average monthly
river salinity; and a 0.02 m'ls (11.1%) increase in average
monthly streamflow gain.
c. Ground water levels were relatively insensitive to decreased
ground water pumping.
d. The primary mechanism for the decreases in salinity was
decreased reuse of ground water that resulted in less evapoc-
oncentration of dissolved salts.
Four scenarios of decreased irrigated acreage were simulated:
ceasing of irrigation on (I) sub-area 1 (20% of irrigated acreage);
(2) sub-area 2 (33% of irrigated acreage); (3) sub-area 3 (47% of
inigated acreage); and (4) all irrigated acreage. Each of these sce-
narios were simulated with flow in the irrigation canal. Scenario 4,
where all irrigation was ceased, was also simulated without flow in
the canal. The major results of these scenarios include the following:
a. Drying up inigated land resulted in decreased salinity in the
aquifeI and river. Thes.e decreases 'Were Ielated to the per-

centage of the total irrigated area on which irrigation was
ceased. .
b. Decreasing irrigated acreage by 100% within the study area
resulted in a 550 mglL (- 25%) decrease in average monthly
ground water salinity and an 80 mglL (-4.4%) decrease 1I1
average monthly river salinity.
c. The maximum decrease in ground water salinity occurred
about 12 years into the simulation period, which corresponds
closely to the hydraulic residence time of the aquifer.
d. Ground water levels were relatively insensitive to decreases in
irrigated acreage.
e. For a 100% decrease in irrigated acreage, model scenarios with
and without flow in the canal varied Httle with respect to their
affect on salinity in the river and aquifer.
f. A 100% decrease in irrigated acreage, with and without flow
in the canal. substantially decreased base flow contributions to
the river. Average monthly streamflow gains decreased to
0.06 m3/s (-64%) and 0.03 m3/s (- 86%), respectively.

Because the model simulations were applied to a relatively
small area, the effects of changes in irrigation on the entire lower
Arkansas River valley are not easy to estimate. Even without for-
mal basin--wide-esti-matio~Le;{pandoo-decreasef in irrigation
area or in ground water pumpage would be expected to further
decrease salinity in both the alluvial aquifer and the Arkansas
River. This conclusion is. based on model results that indicate salin-
ity decreases as inigared acreage or the amount of ground water
pumped is decreased. With lower salinity irrigation water, crop yields
may increase and less salt tolerant crops that have a higher cash value
may be grown. Additionally, the model represents n valuable tool
that can be used to better define the conceptual model of the rela-
tions between irrigation and salinity. This conceptual model could
be transferable to other irrigated semiarid environments outside of
the Arkansas River valley and might be used to indicate the direc-
tion and relative magnitude of changes in salinity in response to
decreased irrigation.

Acknowledgments
This work was conducted as part of the U.S. Geological
Survey's evaluation of the Water-Quality Effects of Water Operations
in the Arkansas River Basin in Colorado. Additional financial sup-
port was provided by the Gibson Hydrogeology Endowment at the
University of Minnesota and the McKnight Foundation. We would
like to thank the landowners who allowed us to sample their wells,
and L.D. Holt of the Southeast Colorado Power Association for sup-
plying power consumption data. Assistance in draftiog one of the
figures was provide<! by Paul Morin at the Department of Geology
and Geophysics at the l1niversity of Minnesota. Stanley G. Robson,
Martha Scholl, and the three anonymous reviewers for this journal
provided many helpful conunents.

References
Adkins, R. 1996. Overview-The future of the river. Colorado Wacer:
Newsletler of the Colorado Water Resources Research InstiTute 13,
no. 1:15-17.
Barroll, Dr. M. 1996. Interview by author, October, New Mexico State
Engineers Office. Santa Fe, New Me~ico.
Batie, 5.S., and R.G. Healy. 1983. The future of American agriculture.
Scientific American 248, no. 2: 45.53.
Bouwer, H. 1994. Irrigation and global water outlook. Elsevier Agricultural
Water Managemem 25: 221-231.

85