July 1985
Hydrologic Linkages
Salinity in the Colorado River is a rath-
er unusual nonpoint water quality prob-
lem in that the pollution sources are geo-
graphically far removed from those
damaged by salinity. Most salt loading oc-
curs in the Upper Basin, while adverse sa-
linity effects are registered several
hundred miles away. The Imperial Valley
experiences about ninety percent of the
total agricultural damages in the U.S. por-
tion of the basin (Kleinman and Brown,
1978). In fact, both the major users of the
Colorado River water, the Imperial Valley
and the Metropolitan Water District of
southern California, lay outside the Col-
orado River Basin.
Some hydrologic assumptions are need-
ed to compare the benefits with the costs
of this long distance externality. Although
heavily studied, knowledge of the hydrol-
ogy of the Colorado River is far from per-
fect. The hydrologic state of the art is em-
bodied in USBR’s Colorado River
Simulation System. (See USBR, 1983:107).
This analysis is based on the assumptions
in that simulation model.
The first major assumption relates salt
loading in the Upper Basin to Lower Ba-
sin salinity levels. The salt loading rela-
tionship adopted is that 10,000 tons added
to the river above Parker Dam equals 1.01
mg∕liter at Imperial Dam (USDI, 1983:
45). (Equivalently, 9,900 tons equals 1 mg∕
liter) Each ton of salt removed from up-
stream is assumed to result in exactly one
ton less salinity at Imperial Dam. [If sub-
stantial quantities of salt are being precip-
itated out of the water in the system of
reservoirs, as the evidence increasingly
suggests, (Paulson and Baker; USBR, 1983:
108-10) salinity control efforts would be
proportionately less effective.]
The second assumption concerns the
time it takes salts to pass from the Upper
Basin through all the reservoirs to the Im-
perial Dam. The USDI (1983:45) assumes
a hydraulic retention time of 5 to 7 years
Western Journal of Agricultural Economics
for salts to pass through Lakes Powell and
Mead to impact Lower Basin water users.
Ninety percent of salinity benefits is ex-
pected to be registered at Imperial Dam
at the end of that period. This is a signif-
icant point, because it means that dam-
ages avoided must be discounted over this
period to establish a net present value for
comparison to salinity control costs, a point
overlooked in previous economic apprais-
als.
For lack of a precise lag function, we
assume a hydraulic retention time of six
years to achieve one hundred percent of
salinity benefits. Fifteen percent of the
benefits are assumed to accrue at the end
of each of the first five years with the re-
maining twenty-five percent coming in the
sixth year. (This assumption is probably
overly favorable to feasibility, as the time
distribution of impacts is likely to be
skewed towards later years.)
Agricultural Benefits
We adapted the procedure first devel-
oped by Moore, Snyder and Sun to esti-
mate damages-avoided, or the benefits of
salinity control to agriculture. Two linear
programming models of Imperial Valley
agriculture were developed. (Full details
and assumptions are in Gardner.) One
model, reflecting production at 800 mg∕
liter, approximates the current situation.
(The average salinity from 1978-82 was
804 mg∕liter.) A second model simulating
1,100 mg∕liter salinity conditions repre-
sents maximum future salinity conditions,
since current USBR forecasts without sa-
linity control are 1,024 mg∕liter in 2,000
and 1,089 mg∕liter by 2,010 (USDI, 1983:
47). The difference in net farm income
between the two models provides our es-
timate of agricultural damage caused by
a 300 mg∕liter increase in salinity. This
total damage estimate is converted to an
estimate of average marginal salinity
damage per mg∕liter.
The benefit estimates derived below are