A critique of field spectroscopy and the challenges and opportunities it presents for remote sensing for agriculture, ecosystems, and hydrology

Abstract

From the early scanning spectrometers, the utility field spectroscopy has been constrained: by detector sensitivity, leading to high integration times; portability; and the measurements having very limited support (possibly an Earth surface area in the order of 0.25m² to 2m²). However, over the last twenty years or so detector sensitivities and electronics have in improved leading to practical Earth surface sampling time increasing their utility to support Earth observation science and as optical remote sensing teaching and training tools. Now the uncertainties associated with field spectral measurements are being more widely recognised and field sampling methods and instrument continue to be developed to enable these uncertainties to be quantified and minimised. There are a number of key challenges which still need to be more widely addressed if field spectroscopy is to provide evermore reliable and replicable measurements. An understanding of the systematic biases introduced by this sampling method has begun to be recognised. In addition, the mismatch in scale between near-ground spectroscopy measurement and observations from space-borne sensor has begun to be addressed with the development of unmanned aerial vehicles as platforms and lightweight and miniaturised stateof- the-art spectrometer systems. It is now possible to non-invasively sample terrestrial and hydrological ecosystems in a statistically robust manner and do so with supports similar in scale to those of air- and space-borne sensors. These developments will revolutionise the use of field spectroscopy to support empirical science and model development and the calibration and validation of space-based observations.