Optimisation of Spatio-Temporal Aspects of Drainage

Drainage schemes for agro-ecological salinity management have been set to lower the shallow groundwater to help increase production and reduce the ecological risk. Once the groundwater levels are lowered to desired agro-ecological thresholds then drainage scheme's operation needs to be optimised according to the groundwater dynamics both in terms of space and time to achieve hydro-economic efficiency. Every system, even the most complicated ones, can be modelled if its behaviour is fully known and understood but a key difficulty in optimization is dealing with non-linear spatio-temporal problems. Such problems can be optimised using genetic algorithm aimed at finding near optimal solutions to highly non-linear optimization problems. GAs are nature inspired stochastic computational techniques. The major advantages of these algorithms are their broad applicability, flexibility and their ability to find optimal or near optimal solutions with relatively modest computational requirements. This paper presents the development of a surface-groundwater interaction model for the spatiotemporal optimization of pumping operation of Subsurface Drainage Schemes to achieve the similar or better level of service both in space and time domains. A drainage scheme, Wakool Tullakool Subsurface Drainage Scheme (WTSSDS), from New South Wales is included as a case study to illustrate the need for the optimisation of drainage operation to achieve the hydro-economic efficiency. The model results are being used to plan an optimal operation of the tubewells to control water logging and salinisation. Once a groundwater pumping system is put in place to service the waterlogging and salinity problems in an irrigation area, changing temporal conditions (rainfall, flooding and irrigation practices) necessitate a dynamic management response. Such an approach needs to be based on a comprehensive understanding of the groundwater dynamics for achieving economic efficiency of drainage since some of the tubewells may become ineffective while others start drawing more water than required to keep the watertable at the desired levels. The optimal management strategy, as determined by the simulation-optimization analyses, suggests the same or even better pumping performance of the scheme could be achieved using 42 wells (rather than 53) and a maximum pumping rate of 1000 m(3) d(-1) for an individual pump. The groundwater pumping can be reduced by around 1 MCM yr(-1) which is approximately 20 % lesser than the existing rate. This will lead to substantial cost savings by reducing the number of wells needed and less pumping. Preliminary cost estimates indicated that $ 4000 MCM yr(-1) pumping cost could be saved. This study has also shown that a MODFLOW based surface-groundwater interaction model, using hydrogeology, soils, groundwater levels, groundwater pumping, channel network and net recharge information can be a useful tool to develop understanding of the groundwater dynamics. The simulation-optimization analyses can effectively be used to plan an optimal operation of the subsurface drainage scheme to control water logging and salinisation in a hydroeconomically viable way.