Mapping estimated deep drainage in the lower Namoi Valley using a chloride mass balance model and EM34 data

The Murray Darling Basin accounts for half of all water used for irrigation in Australia. However, improvements in water use efficiency (WUE) are required, owing to increasing demands on water (e.g., environmental flows). This requires data on the spatial distribution of soil-hydrological properties, such as deep drainage (DD). Measuring DD using lysimeters, although accurate, is site-specific. Alternatively, estimates are commonly made using chloride mass balance (CMB) models. Gaining this information across a large area is still problematic due to the prohibitive cost of drilling, sampling, and laboratory analysis. Ancillary data, obtained from electromagnetic (EM) instruments, have been used to add value to a limited number of DD estimates. We evaluated the use of a hierarchical spatial regression technique to map the estimated DD using a steady state CMB model coupled to EM34 measurements. We first compared a standard least squares and a stepwise multiple linear regression model. The former includes the use of EM34 signal data in the horizontal (EM34-10H, EM34-20H, and EM34-40H) and vertical (EM34-10V, EM34-20V, and EM34-40V) dipoles, as well as two trend surface variables (scaled easting and northing). The latter model only includes a statistically significant ancillary variable (EM34-10H) and a trend surface parameter (scaled northing), and we use this to estimate DD across the lower Namoi Valley. EM34 data available on a 1 km grid proved useful for mapping DD on a reconnaissance level, with the results closely related to the physiography. In particular, large DD estimates are associated with the prior stream channels. Conversely, smaller DD estimates characterize the agriculturally significant clay plain which is used extensively for irrigated cotton production. The map of estimated DD will allow improved siting of dams and irrigation fields, as well as indicate where more efficient cropping or irrigation systems can be implemented to increase WUE.