4. CONCLUSIONS
Most water management systems are concerned
with satisfying conflicting demands of various
groups and MCDM techniques provide the
mechanism for resolving these conflicts. They
provide better results than simple linear
programming (LP) solutions because they
integrate the effect of all the objectives
simultaneously. There are an increasing number
of highly sophisticated LP solvers that could
easily be adapted to solve MCDM problems using
Goal Programming (GP) or Weighted Goal
Programming (WGP) as illustrated with the
example problem. The application of MCDM
techniques to the simple nodal-network example
problem demonstrates its ability to provide
solutions that integrate different goals and tradeoffs.
The
pay-off
matrix
for the three
goals
illustrates
the degree of
conflict between the
different
goals
and trade-offs.
The
effect
of
different
ET and
rainfall
(dry,
average
and
wet)
on
NR,
crop
areas,
environmental
flows
and
water
allocated
to the irrigation
areas was clearly
demonstrated. Furthermore, the sensitivity of the
weights assigned to the different goals was shown
to have marked impact on optimal crop areas and
the degree of under- and over-achievement of the
selected targets for all the three goals.
Maximization of NR was almost equivalent to
minimization of total pumping under dry climatic
conditions. By attaching different weights to
goals, sets of decision variables (crop area, water
allocations) could be formulated for different
seasons that could aid in policy formulations and
decision-making. Although goal programming is
a useful tool to analyze MCDM problems, there is
a difficulty of selecting the target values and
weights for the different goals.