PUBLICATIONS--Report, "Finite-difference model to simulate the areal flow of salt water and fresh water separated by an interface,"

                        GEOLOGICAL SURVEY
                        RESTON, VA. 22092

In Reply Refer To:                            June 13, 1980
EGS-Mail Stop 411


Subject:  PUBLICATIONS--Report, "Finite-difference model to
                        simulate the areal flow of salt water and
                        fresh water separated by an interface," by
                        J. W. Mercer, S. P. Larson, and
                        C. R. Faust

A simple approach to the analysis of problems involving salt water
intrusion into fresh water aquifers is to assume that the fresh
water and salt water are separated sharply by an interface.  In
actuality, there tends to be a mixing of fresh water and salt
water in a zone of diffusion.  The size of the zone is
significantly influenced by the degree of heterogeneity of the
aquifer materials.  Where this zone is relatively narrow, the
assumption of a sharp interface may provide an acceptable
approximation to actual conditions.

Jim Mercer, Steven Larson, and Charles Faust have developed a
computer program to simulate two-dimensional ground-water flow of
salt water and fresh water in a single aquifer where these waters
are separated by an interface.  The program is designed to
simulate time-dependent problems such as those associated with the
development of coastal aquifers, and can treat water-table
conditions or confined conditions with steady-state leakage of
fresh water.

To test some of the approximations used in their mathematical
development, Mercer, Larson, and Faust compared solutions from
their numerical technique with:  (1) an analytical solution for a
simple problem where the interface is assumed to be linear, and
(2) the results of a Hele-Shaw laboratory experiment.  The
solution to the analytical problem was acceptably approximated.
The comparison with the Hele-Shaw experiment suggested that
certain of their approximations may be invalid for steep interface
slopes.  In a separate paper that is to appear soon in the journal
"GROUND WATER," they report on an application to a field problem
on Maui, Hawaii.

The experiences developed to date from these studies indicate that
mass-balance errors (differences between the true change in
storage and the change expressed by the equations) should be
monitored to assure that these are tolerable in terms of their
impact on computed heads and interface elevations.  Methods for
reducing mass balance errors are known, but the authors concluded
from their limited applications of the program that the added
expense of constraints required by using such methods may not be
warranted in many cases.

The report, a copy of which is enclosed, contains a FORTRAN
listing of the computer code, information on use of the program,
and a description of some tests made of the program.  the FORTRAN
program is available on SYS1.LOADLIB on RE1 and RE2.  The JCL
needed to run on these machines is:


          Input Data


Limited additional copies of the report are available upon request
to the Ground Water Branch.

                              (s) Gordon Bennett
                              Acting Chief, Ground Water Branch


WRD Distribution:   A - (Memo only)
                    B - NR, 12 copies; SR, 6
                    copies; CR, 21 copies; WR. 12 copies