8.3. Groundwater Quality Management Models
293
can be used to maximize the net income of agricultural economy in a farm, a
county or a groundwater basin. In these models, the kinds of crops plan ted,
or the areas sown are taken as the decision variables, from which the allocations of both surface water and groundwater are determined. The simulation
model is not used to form constraints for the economic model, but the two
models are connected. The procedure is to estimate the pumping cost
for each subregion based on the current hydraulic conditions, and solve
the economic management model to get a pumping scheme. The pumping
scheme is then inputted into the simulation model to re-estimate the
pumping cost of the next management period. With this method, we can use
the linear programming to solve complex economic models, or even to include the social and legal factors in the management model in order to
determine the optimal allocation of water resources. In addition, because the
simulation model is not included in the constraints, the programming problem will not become nonlinear.
This type ofmodel has been used in practice. Young and Bredehoeft (1972)
considered the problem of using groundwater resources to gain maximum
agricultural net income. Bredehoeft and Y oung (1983) made further considerations that the management goals should not only be to maximize the net
annual income, but also to minimize the variance of the income in order
to guarantee reliable income. The development of an optimal groundwater
policy can provide an stable and timely supply of water to agriculture and
reduce the variance of income.
Multilevel Management Models
Currently, the development and applications ofwater resources management
models are very wide in scope, covering not only hydraulic and environmental aspects, but also economic, political and legal concerns. As a result, water
resources management is really a large system problem. It usually has numerous objectives. The optimal management scheme is the result of trade-offs
among the objectives according to certain principles that reßect the preferences of policy-makers. For example, a compromise must be made between
economic development and resources conservation, or between the shortterm benefits and the long-term development. On the other hand, policymakers may be organized in multilevels, from the junior supervisors of individual units to the middle and senior policy-makers holding responsibilities.
As a result, the decision and execution of the optimal scheme is inevitably
hierarchical, and may not be carried out in the most mathematically efficient
way. In this case, the problem may need to be solved several times by a trial
and error method until nearly optimal plan is found.
Haimes and his co-partners have done a lot of work and made great
contributions to the development of this field in the past two decades. They
not only developed the mathematical theories and methods of multi-objective
decisions and hierarchical decisions, but also applied them to the manage-
293
can be used to maximize the net income of agricultural economy in a farm, a
county or a groundwater basin. In these models, the kinds of crops plan ted,
or the areas sown are taken as the decision variables, from which the allocations of both surface water and groundwater are determined. The simulation
model is not used to form constraints for the economic model, but the two
models are connected. The procedure is to estimate the pumping cost
for each subregion based on the current hydraulic conditions, and solve
the economic management model to get a pumping scheme. The pumping
scheme is then inputted into the simulation model to re-estimate the
pumping cost of the next management period. With this method, we can use
the linear programming to solve complex economic models, or even to include the social and legal factors in the management model in order to
determine the optimal allocation of water resources. In addition, because the
simulation model is not included in the constraints, the programming problem will not become nonlinear.
This type ofmodel has been used in practice. Young and Bredehoeft (1972)
considered the problem of using groundwater resources to gain maximum
agricultural net income. Bredehoeft and Y oung (1983) made further considerations that the management goals should not only be to maximize the net
annual income, but also to minimize the variance of the income in order
to guarantee reliable income. The development of an optimal groundwater
policy can provide an stable and timely supply of water to agriculture and
reduce the variance of income.
Multilevel Management Models
Currently, the development and applications ofwater resources management
models are very wide in scope, covering not only hydraulic and environmental aspects, but also economic, political and legal concerns. As a result, water
resources management is really a large system problem. It usually has numerous objectives. The optimal management scheme is the result of trade-offs
among the objectives according to certain principles that reßect the preferences of policy-makers. For example, a compromise must be made between
economic development and resources conservation, or between the shortterm benefits and the long-term development. On the other hand, policymakers may be organized in multilevels, from the junior supervisors of individual units to the middle and senior policy-makers holding responsibilities.
As a result, the decision and execution of the optimal scheme is inevitably
hierarchical, and may not be carried out in the most mathematically efficient
way. In this case, the problem may need to be solved several times by a trial
and error method until nearly optimal plan is found.
Haimes and his co-partners have done a lot of work and made great
contributions to the development of this field in the past two decades. They
not only developed the mathematical theories and methods of multi-objective
decisions and hierarchical decisions, but also applied them to the manage-