Estimating and mapping deep drainage risk at the district level in the lower Gwydir and Macquarie valleys, Australia
In the Murray–Darling Basin, irrigated agriculture, which produces rice, dairy, cotton and citrus, is a large consumer of water resources. Effective management of the water resource is therefore important to ensure sustainability of irrigated agriculture. In the lower Gwydir and Macquarie valleys, respectively located in northern and central New South Wales of Australia, extensive irrigated-cotton production is an important contributor to the nation’s export earnings. However, there are problems of excessive deep drainage (DD) in these regions. To address them requires soil and water quality information, but there is little quantitative information to plan for and implement improved water use efficiency. In this paper, we explore methods that could efficiently generate data on natural resources. First, we carried out an electromagnetic induction (EM38) survey to characterise broad soil profile types in the Ashley (lower Gwydir valley) and Trangie (lower Macquarie valley) districts. From the resulting apparent electrical conductivity (ECa, mS/m) data collected using an EM 38 (vertical mode of operation), soil profile sites were selected and sampled, followed by laboratory analysis carried to determine exchangeable cations and clay content. The soil data collected were analysed with a salt and leaching fraction (SaLF) model, based on specific water quality and quantity parameters, such as electrical conductivity of irrigation water (ECiw, dS/m) and rainfall (R, mm/year). Various water application rates (I) were also considered, to simulate irrigated cotton (I = 600 mm/year) and rice production (I = 1200 mm/year) as well as shallow water reservoirs (I = 1800 mm/year). For each irrigation scenario, DD values (mm/year) were estimated. An exponential function was used to describe the relationships between ECa values obtained with the EM38 and estimated DD. These relationships were then used to estimate DD at each of the EM38 survey sites, whereupon cut-off (zc) values were used for indicator transforms of the data. Using indicator kriging (IK) and various irrigation scenarios, we demonstrate the usefulness of this approach in identifying areas of high risk of DD exceeding various cut-off values (zc = 50, 75, 100 and 200 mm/year). Thus, we show where improvements in water-use efficiency could be achieved in the irrigated cotton growing districts of Ashley and Trangie.