WSP12518

• File Number: 8282.400
• Description: Colorado River Operations and Accounting - Deliveries to Mexico
• State: CO
• Basin: Colorado Mainstem
• Water Division: 5
• Date: 9/1/2001
• Author: Various
• Title: The Journal of Arid Environments - Special Issue-The Lower Colorado River Basin and Delta - Volume 49-Number 1 - September 2001 - RE-Colorado River-Mexican Delta Issues - 09-01-01
• Water Supply Pro - Doc Type: Report/Study

40

M. J. COHEN ET AL

ground-water pumped from Arizona's Wellton-Mohawk Irrigation and Drainage
District, runs 56 km from the SIB to the Cienega 01 aIdes-Casillas el a1., 1998).

Material and methods

This study developed a water balance for the Colorado River delta using existing flow
data, calculated system losses to evaporation and evapotranspiration (ET), and esti-
mated outflows with a mass balance and historical records. Published discharge records
(IBWC, 1992-1998) for the mainstem at the SIB and for the Bypass Extension at the
SIB were supplemented by unpublished agricultural drainage records obtained from
Mexico's Comision Nacional de Agua (CNA) and by unpublished municipal discharge
records from Mexico's Organismo Operador Municipal de Agua Potable, Alcantarillado
y Saneamiento de San Luis Rio Colorado (OOMAPAS). Losses due to evaporation and
ET were calculated from published reports of vegetation type, extent, and density
(Valdes-Casillas el ai., 1998; Luecke el aZ., 1999; Zamora-Arroyo el aZ., 2001), Lower
Colorado River Accounting System (LCRAS) ET coefficients (V.S, Bureau of
Reclamation 1997, 1998), and published pan evaporation rates (IBWC, 1992-1998).
Agricultural drainage flow records prior to the year 1992 and IBWC records after 1998
were not available, limiting the study to the period 1992-1998.
To refine the analysis, the study area was divided into three hydrologic sub-systems:
the Colorado River mainstem complex, which includes the Rio Hardy; the Cienega de
Santa Clara and the proximate El Doctor wetland; and El Indio wetlands. The temporal
scale of the study was also broken down for the mainstem complex into flood years and
non-flood years, based on the estimate of Luecke el ai. (1999) that releases of
100-200 m3 s - I are necessary to achieve flood stage on the Colorado River mainstem
below Morelos Dam. 'Non-flood years' reflects means for the years 1992, 1994, 1995,
and 1996. 'Flood years' reflects means for the years 1993, 1997, and 1998. Mean annual
discharge at the SIB for non-flood years was < 30 x 106 m3; for flood years, mean
annual discharge was markedly higher ( > 2500 x 106 m3).
The study used a mass balance to characterize discharge through the study area. The
mass balance equation (Owen-Joyce & Raymond, 1996) can be described as:

Qdl = QUI + Qrr + P + T, - E - ET - AS. - Q.b,

where
Qdl = flow at the downstream boundary
Quo = flow at the upstream boundary
Q,r = rerum flow to the river (agricultural and municipal drainage)
P = precipitation (on open water surfaces)
T, = tributary inflow
E = evaporation from open water surfaces
ET = evapotranspiration
AS. = change in storage in the alluvial aquifer
Q.b = flow to sub-basin.
Note that there was no surface. storage capacity at or below Morelos Dam. Total
inflow, for the delta or a particular sub-system, can be described as:

IF = QUI + Q.; + P + T,.

Total outflow can be described as:

OF = Qd. + E + ET + Q'b + AS..

Discharge at the upstream boundary (QUI) is from published records (IBWC,
1992-1998) and from the CNA. Records of discharge at the SIB, 35 kIn downstream

WATER BALANCE FOR THE COLORADO RNER DELTA

41

from Morelos Dam, were used asa proxy for discharge at the upstream boundary (Qu,)'
For the Cienega, recorded discharge of the MODE at the SIB (IBWC, 1992-1998)
represents Quo'
Records of agricultural and municipal drainage (Qrr), where available, were obtained
from CNA and from Valdes-Casillas el aZ. (1998), and estimated from data obtained
from OOMAPAS, the municipal water agency for the City of San Luis Rio Colorado.
OOMAPAS provided records of deliveries for the years 1990, 1995, and 1999, and
records of municipal discharge for the year 1999. Estimates of municipal effluent
discharge to the river for the study period were based on an assumption of an annual
growth rate in water consumption of 2-5%, the best fit for the records of water deliveries.
Precipitation (P) was calculated from precipitation records for the 'Delta' and 'Riito'
stations, as reported by IBWC (1992-1998) and reported extent otopen-water surfaces
(Zamora-Arroyo el aZ., 2001). Records of precipitation and evaporation were incom-
plete for the 'Delta' station for the years 1995 and 1996, so means for the mainstem
complex did not include these years.
The levees minimize the direct influence of tributary runoff (T,) on the main-
stem, but runoff does discharge into the Rio Hardy. It was assumed that tributary
runoff to the Cienega and to El Doctor and El Indio wetlands was negligible, due to
greater permeability of soils between these areas and their headwaters. As noted, some
(107 m3 year - I) of this water discharges to El Doctor wetlands through artesian springs
(Glenn el a1., 1996). Factors such as soil depth and permeability, vegetative cover, and
rainfall intensity and duration affect runoff (Hely & Peck, 1964). Hely & Peck
(1964) estimated runoff for the lower Colorado River-Salton Sea area, roughly
50-100 kIn north of the study area, by measuring initial soil infiltration and correlating
this with soil type to project a runoff curve. This runoff curve was then used to
project runoff as a percentage of precipitation, ranging from effectively 0% for
sandy alluvial soils, to roughly 8% for less permeable alluvial soils, to more than 20% for
foothill/plateau areas with less permeable soils. This study estimates runoff at 8% of
precipitation, to account for the lack of rainfall records for the mountain regions
balanced against expected infiltration in the permeable alluvial soils between the base of
the mountains and the hydrologic systems in this study. Accurate projections of
runoff require records of finer temporal resolution than were available for this
study, as well as analysis of the permeability of soils underlying ephemeral streams (Hely
& Peck, 1964). Therefore, the calculations of tributary runoff are offered as
a general estimate.
Evaporation (E) from open water surfaces was calculated from total area of open
water (Luecke el ai., 1999; Zamora-Arroyo et ai., 2001) and reported pan evaporation
rates for the 'Delta' and 'RUto' stations in Mexico, as reported by the IBWC
(1992-1998), using a pan-to-lake coefficient of 0'60 (Owen-Joyce & Raymond,
1996). The 1997 flood event inundated the mainstem floodplain between the levees
(Luecke el ai., 1999), greatly increasing the extent of open-water surface area
(30,000 ha) subject to evaporation and infiltration. Evapotranspiration (ET) rates for
wetland vegetation were calculated from reports of vegetation density, extent, and type
(Valdes-Casillas et al., 1998) and LCRAS (U.S. Bureau of Reclamation, 1997) ET
coefficients, for the year 1997. Established riparian vegetation was assumed to draw
from the alluvial aquifer (Dawson & Ehleringer, 1991; Stromberg, 1993). Evapo-
transpiration for riparian vegetation in the delta was estimated for purposes of compari-
son but was not included as part of tlle surface water balance.
Change in storage in the alluvial aquifer (AS.) for non-flood years was based on
estimates of AS. for the reach of the river from Imperial Dam to Morelos Dam (U.S.
Bureau of Reclamation, 1998). For flood years, AS. was calculated from extent of
inundated area and reported infiltration rates for the permeable alluvial soils character-
istic of the floodplain. Measured initial infiltration for dry or slightly moist soils was
reported at 2'5 em for a 3D-min period (Hely & Peck, 1964).

C:'
Ci
c.,:,
1'\.:>
c.,~
f--A