Evaluation of a static water balance model in cropped and grazed systems of temperate Australia

• Published in: Environmental modelling & software : with environment data news Vol. 25; no. 12; pp. 1682 - 1691
• Main Authors: Melland, Alice R., Vigiak, Olga, Roberts, Anna M., Rattray, Dan, Whitford, Jane
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.12.2010
• Subjects: Crop; Drainage; HowLeaky?; Lucerne; Pasture; PERFECT; Soil water content; Australia; New Zealand, South I., Westland, Schulz Creek, Rutherglen; Drainage; Crop; Soil water content; Pasture; PERFECT; HowLeaky?; Lucerne
• ISSN: 1364-8152
• DOI: 10.1016/j.envsoft.2010.05.006

Several water balance models have been developed for Australian conditions, however few of them were developed for the mixed cropping and grazing systems that are typical of temperate south-east Australia. HowLeaky? is a 1-dimensional water balance model that allows simulation of both cropped and grazed systems. This study tested the accuracy of HowLeaky? simulations of soil moisture content and surplus water (runoff plus deep drainage) for mixed farming systems in south-east Australia. Two datasets from the state of Victoria were used to validate model simulations: 1) Rutherglen, consisting of four years soil moisture observations and deep drainage estimates from seven treatments of crop rotations and annual pasture; 2) Vasey, consisting of three years soil moisture and runoff observations and deep drainage estimates from a perennial pasture. A static plant growth option was applied to simulate seasonal crop and pasture covers. Despite the static approach not accounting for inter-annual variation in crop development, HowLeaky? captured observed soil moisture trends, with the Nash Sutcliffe efficiency ranging from fair (0.34) in continuous crop, moderate (0.64–0.68) in annual and perennial pasture, to very good (>0.70) for continuous lucerne and lucerne-crop rotations. Annual water surplus was generally underestimated in this uncalibrated application of the model suggesting a need for some degree of model calibration for highly uncertain and influential parameters such as field capacity.