J Brett Robinson1 and David M Freebairn1
1 University of Southern Queensland, West St, Toowoomba, Qld, 4350 www.soilwaterapp.net.au, email@example.com
Rainfall is low and unreliable in most of Australia’s grain growing regions, making management of crop water supply vital to profitability. Difficulties in measuring plant available water (PAW, mm) has led to using simulation models for estimating PAW and predicting crop water supply, leading to the development of reliable tools for farmers to estimate plant available soil water. This paper examines the accuracy of predictions of PAW from a water-balance model during summer fallows in the northern region and fallows and cropping in the southern and western regions. The water balance model HowLeaky? explained 69% (R2) of observed changes in PAW with a RMSE of 32 mm across three sites (45 fallows). Analysis of in-field variability indicates that errors in estimating PAW with a water balance model are small relative to natural variability. The errors in estimating PAW and changes in PAW using computer simulation are modest relative to measurements and models can make a significant contribution to estimating and managing PAW.