Contemporary land resource survey requires improvements in direct soil measurement

The demands being placed on land resource information are changing and users now expect predictions of spatial and temporal trends in soil and landscape properties. In response, new methods of land resource survey are utilizing technological advances in digital terrain analysis, remote sensing, spatial statistics and geographic information systems. The methods are overcoming deficiencies in conventional survey. Maps with discrete polygons and morphologically defined soil types are being replaced by predicted surfaces of individual soil and landscape properties. These predictions provide better inputs to simulation models and allow more flexible interpretation. The strides in computing technology and analysis evident in quantitative land resource survey are yet to be matched by advances in soil sampling and measurement. In conventional survey, field sampling is rarely statistically based and as a consequence, estimates of map accuracy and precision are difficult to provide. It is also assumed that strong correlations exist between field observable morphology and other more expensive soil properties. This assumption is rarely tested. Quantitative survey requires a substantial increase in direct soil measurement. The qualify and quantity soil physical and hydraulic data are currently inadequate and more efficient measurement methods are required. There is a need fbr standardisation, improved laboratory practice and appropriate specimen volumes. Much greater discipline is also required in the documentation and digital storage of data. Adequate pedotransfer functions are necessary to predict the difficult-to-measure physical and hydraulic properties from limited input data. Pedotransfer functions have an important role in land resource survey because it is impractical to measure expensive physical and chemical properties at large numbers of sites. It is suggested that measurement limitations create the main obstacles to the development of more effective systems for providing land resource information at scales from the paddock to the continent.