Profitable farming systems depend on high quality data, including soil data, to manage annual agronomic decisions affecting crop performance. Good soil management will also secure soil production potential into the future, reducing long term risk. While a number of sensors have been developed to measure soil properties quickly, cheaply and in situ, current work has focused on a single or small number of properties, or developed calibrations that are only suitable for a given region or soil type.

This project will focus on developing a system where soil and its related data can be measured at a finer scale than is currently feasible. This research will develop calibrations for proximal soil sensor measurements and connect sensor derived properties with the wealth of information held in pedotransfer functions through the use of a spectral inference system. The aim is to deliver platforms for fine scale mapping of soil attributes to provide reliable real-time information to growers to support their decision-making processes.

Much of this work will be automated and will culminate in the production of a report presenting key soil attributes in an easily interpretable way. In addition, the reduced cost of generating this data will increase growers' ability to assess variability in the field and support precision agriculture approaches.