Harry Gaynor1, Patrick Filippi1, Rose Brodrick2, Daniel K.Y. Tan1
1 The University of Sydney, Sydney Institute of Agriculture, School of Life and Environmental Sciences, Faculty of Science, Sydney, NSW 2006, Australia,
2 CSIRO Agriculture and Food, Black Mountain, Canberra, ACT 2601.Tel: 61 2 8627 1052, Email: email@example.com
Australian cotton irrigators are continuously meeting challenges of water scarcity with technological innovation to improve their water resource management. A popular optimisation technique is to time irrigation applications based on a soil water content refill point. Point-source soil water measurements can give a current soil water status (e.g. using a neutron moisture meter (NMM)), but do not provide any predictive capacity to assist in planning future irrigations. We compared the accuracy of two methodologies to calculate soil water content with predictive capability: HydroLOGIC software (crop model) and IrriSAT software (Kc approach derived from NDVI satellite images), using calibrated NMM measurements as a standard. To enable a fair comparison of the two technologies in HydroLOGIC, the soil water was not corrected by inputting soil water measurements, with just the crop parameters and irrigation dates entered up until the run date. IrriSAT had slightly higher correlation (r = 0.82) with NMM readings compared with HydroLOGIC (r = 0.75) when averaged across the measurement period. However, the accuracy varied significantly during different periods which could significantly impact on irrigation timing. During early to peak flowering IrriSAT overestimated NMM deficits by 20-30 mm, which if relied on would result in irrigating much earlier than required whereas HydroLOGIC run without any soil water inputs underestimated crop water use after cut-out. The data suggested measured soil water through instruments such as NMMs can be used in a combined approach with predictive software to monitor soil moisture and enable irrigators to predict more accurately the timing of future irrigations.