DISTRIBUTION OF THE SHALLOW ALLUVIAL AQUIFER IN THE ROSWELL ARTESIAN BASIN OF SOUTHEASTERN
NEW MEXICO, U.S.A.
By Dr. William M. Turner
As part of the effort to develop computer models of the Roswell Artesian Basin for the
purpose of managing water rights transfers, the Technical Section of the Office of the
State Engineer for the State of New Mexico in the United States evaluated the
transmissivity distribution within the shallow alluvial aquifer. The material that
follows is adapted from Rao (1991).
The Roswell Ground-Water Basin covers a large area in the Pecos River valley in
southeastern New Mexico (Figure 1). The basin is bounded
on the west by the Sacramento Mountains, about 80 miles (129 km) distant from the Pecos
River; on the east by the foot of the High Plains of Texas, about 25 miles (40 km) from
the Pecos River; on the south by the Seven Rivers Hills. The northern boundary of
the Roswell Basin is near Vaughn, New Mexico about 90 miles (145 km) north of Roswell, New
The surface elevation within the Roswell Basin declines from about 10,000 feet (3,048
m) at the crest of the Sacramento Mountains on the west to about 3,500 feet (1,067 m) at
the Pecos River east of Roswell, New Mexico. From the Pecos River, the surface rises
eastward to 4,000 feet (1,219 m) at the foot of the High Plains of Texas (Hantush,
The geologic formations involved in the transmission, storage and confinement of ground
water in the Roswell Basin are the shallow alluvium deposits called the Shallow Alluvial
aquifer, the Chalk Bluff Formation, the San Andres Formation and the Yeso Formation
Alluvium deposits form the Shallow Alluvial aquifer. The Chalk Bluff Formation
forms the upper semi-confining aquitard overlying the San Andres Limestone artesian
aquifer. The Yeso Formation forms the semi-confining aquitard below the San Andres
Hantush (1957) reported transmissivities ranging from 31,100 to 139,000 gpd/ft (386 to
1,724 m2/d) based on seven aquifer performance tests (APTs) in the Shallow Alluvial
aquifer and recommended a representative transmissivity of 100,000 gpd/ft (1,240 m2/d) for
Mower et al. (1964) presented a set of transmissivity values for the shallow alluvial
aquifer from (APTs), specific capacities of wells and application of Darcy's Law to
ground-water movement. Their estimate of average transmissivity for wells in the
cultivated area outside of bottom land is 102,000 gpd/ft (1,265 m2/d), which agrees with
the estimate of 100,000 gpd/ft (1,240 m2/d) by Hantush (1957).
Mower et al. (1964) also conducted a series of 12 APTs of shallow wells in bottom land
- an area about one mile (1.6 km) wide on either side of the Pecos River. Average
estimated transmissivity for the Shallow Alluvial aquifer in the bottom land is only
12,000 gpd/ft (149 m2/d). According to Mower et al. (1964), the
valley-fill alluvium in the bottom land consists principally of fine-grained sediments
and, therefore, has a much lower transmissivity. They also estimated that the
average transmissivity for the area between cultivated upland and fine-grained bottom land
sediment is about 42,000 gpd/ft (521 m2d).
Of the Shallow Alluvial aquifer transmissivity estimates of Mower et al. (1964), Kinney
et al. (1968, p.24) commented:
"Test data from 12, small-diameter wells in the bottom lands of the Pecos River
indicate that the average coefficient of transmissivity of the alluvium immediately
adjacent to the river is about 12,000 gpd/ft (149 m2/d). These small diameter wells
did not fully penetrate the alluvial aquifer and were drilled into silty sand on the
present flood plain of the Pecos River. The average coefficient of transmissivity of
the shallow aquifer, based on all available APT data, is probably on the order of 100,000
gpd/ft (1,240 m2/d)."
Saleem and Jacob (1971) present 52 values of transmissivity for the shallow alluvial
aquifer estimated from step-drawdown tests.
Appendix A is a listing of available transmissivity data for
the Shallow Alluvial aquifer. The data were analyzed to study the pattern of their spatial
The natural logarithms of the transmissivity values are plotted in Figure 2. The scatter of the data points about the
theoretical lognormal-distribution line closely follow the theoretical normal-distribution
line. This means that the variation of the shallow alluvial aquifer transmissivities
in the Roswell Basin can be satisfactorily explained by a log-normal probability density
Rao (1991) conducted a X2 goodness-of-fit test was performed to determine
whether the observed transmissivity values would be expected based on the lognormal
distribution. The null hypothesis, Ho, is that at the 90 percent
confidence level the transmissivities are lognormally distributed. The results of
the X2 analysis in Table 1 failed to reject the null
hypothesis that the transmissivities of the San Andres Limestone Artesian Aquifer in the
Roswell Basin are lognormally distributed.
We conclude that the transmissivity of the shallow alluvial aquifer in the Roswell
Basin is lognormally distributed in space.
Hantush, M.S., 1957, Preliminary quantitative study of the Roswell ground-water
reservoir, New Mexico, New Mexico Institute of Mining and
Bureau of Mines and Mineral Resources Division.
Mower, R.W., Hood, J.W., Cushman, R.L., Borton, R.L., and Galloway, S.E.,
1964, An appraisal of potential ground-water salvage along the
between Acme and Artesia, New Mexico, U.S.Geological Survey
Rao, B., 1991, Roswell Basin Analytical Groundwater Flow Model - Users Manual,
Report TDH-91-2, New Mexico State Engineer Office, March,
Saleem, Z.A., and Jacob, C.E., 1971, Dynamic programming model and quantitative
analysis, Roswell Basin, New Mexico, New Mexico Water
Institute, WRRI Report 10.
Supkow, D., 1973, Transmissivity Distribution in
the Tucson Basin Aquifer,
Proceedings of the Arizona Section of the American Water Works
and the Hydrology Section of the Arizona Academy of Science, May
Prescott, Arizona, p. 113-123.
Back to Top