Developing Stochastic Deep Drainage Surfaces In Cox's Creek Catchment

Deep drainage (DD) can contribute to water table rise and salinity, and is a complex function of rainfall, land management and soil hydraulic properties. Because each of these components is uncertain and variable in time and space, this study developed a method to estimate DD risk based on the mechanistic soil water model SWAP using 50 realisations of stochastic rainfall, land use and soil hydraulic properties using a Monte Carlo approach. DD was predicted at 143 soil points in the Cox’s Creek catchment in northern NSW Australia. Realisations of the stochastic daily rainfall were generated at each soil point using an annual mean adjusted Poisson model, DD values were subsequently translated to probabilities of exceeding 100mm/year and spatially predicted over the study area to produce risk maps for the different scenarios. The results showed that DD is episodic with the predominantly summer rainfall in the area, the monthly variability of DD is extremely high depending on when heavy rainfall occurred in relation to different land uses. As expected, the highest probability exceeding 100mm/year DD was for irrigated crop rotations (99%) followed by continuous wheat (59%), then opportunity cropping (46%) and the least for native vegetation (12.5%). Opportunity cropping with sorghum (42%) had the lowest probability to exceed 100mm/year compared to continuous wheat and other opportunity cropping systems and could be one of options for reducing DD in the area. Variation in soil hydraulic properties had less impact on probability of exceeding 100mm/year than variations in land use, which might be explained by soil properties determining the suitability for a certain land use.