Comparison of models for predicting available water capacity of burdekin soils Queensland

Abstract

Various models for predicting profile available water capacity (PAWC) from laboratory measurements were compared with published field values for the same sites. The intention was to choose the best model/s to predict PAWC, by using a database, for a wide range of soils in the Burdekin Irrigation Area, North Queensland. Effective rooting depth for all models was estimated from the chloride profile. It was found that the predictive abilities of all models used were dependent on soil types. A conventional model (ASWC) based on the difference between water retained at -33 and -1500 kPa matric potentials was higher (P < 0.01) than field measured PAWC. An empirical model (PAWC2) based on cation exchange capacity (CEC) and depth was suitable only for cracking clays and sodic duplex soils. Another empirical model (PAWC1) based on -1500 kPa water retained and depth, predicted field PAWC particularly well on cracking clays, sodic duplex and related soils. There were strong indications that the PAWC1 model is also suitable for the better drained, lighter textured soils but there was a shortage of comparable field data to confirm this. The practical implication of these findings was that an analytical database can be used to predict PAWC on many Burdekin soils, providing immediate assistance to those designing irrigation channels and farm layouts.