Mapping nutrient concentration in pasture using hyperspectral imaging

Abstract

Hyperspectral sensing using ground-based equipment has been demonstrated to be successful in determining pasture nutrient content (N, P, K, S) and parameters such as dry matter content and metabolisable energy. This technology needed to be up-scaled so that large areas could be rapidly covered with adequate spatial resolution. This paper describes work which demonstrates a progression from hyperspectral sensing to hyperspectral imaging which utilises the visible, near infrared and short wave infrared parts of the electromagnetic spectrum. Large scale calibration and validation field trials were conducted at the same time as hyperspectral imaging was completed. These trials demonstrate the feasibility of producing information, in detailed map form, on pasture nutrient concentration and other parameters to inform fertiliser placement decisions as well as other farm management tasks. The technique effectively produces a forage analysis for every square meter of any farm. Keywords: hyperspectral imaging, precision agriculture.