Field Spectroscopy – Part One: Mitigating the impact of suboptimal and changing weather conditions
Authored by James Caudery, Geospatial Analyst with 2Excel Geospatial – James has an academic background in Earth Sciences (BSc HONS by thesis) from Monash University, Australia, and several years of experience in Geographic Information Systems. James is skilled in data processing, spatial analysis and cartographic data visualisation. He is also lead for Field Operations, and has successfully run many expeditions in locations around the world.
Welcome to the first article in a series of two, sharing 2Excel geospatial’s experience in Field Spectroscopy. These articles demonstrate how we have solved challenges faced in the efficient collection of field spectra; changing illumination (Part 1) and metadata logging (Part 2).
What is field spectroscopy?
Field spectroscopy involves measuring the spectral reflectance of targets in the natural environment within which they exist. In order to achieve this, specialized spectrometers are used which have been adapted from a laboratory environment to be operated in the field. 2Excel geospatial use FieldSpec 4 instruments from [ASD – formerly ‘Analytical Spectral Devices’, a brand of Malvern Panalytical]. This measures spectral radiance from 350nm to 2500nm (visible to shortwave infrared). Measurements can be made in direct contact with the target, using an artificial light source, or can be made a short distance away (0.5 – 3 meters) using natural light from the sun.
The choice of measurement technique depends upon the context. For example, in the context of plant science, the contact technique is appropriate for the measurement of individual leaves, whereas the short-standoff technique is appropriate to the measurement of plant canopies.
In either technique, measurements of spectral radiance are converted into absolute spectral reflectance via comparison to a standard (white) reference. This reference is calibrated in a laboratory, giving a known absolute reflectance. These are usually made from a material such as Spectralon or Zenith Polymer, which have a near-flat spectral reflectance across the visible-near infrared region of the spectrum, are chemically inert and have diffuse (near-Lambertian) reflectance. If maintained well, these provide a reliable reference for conversion to absolute reflectance.
Standard Field Spectroscopy Methods
FieldSpec 4 spectrometer in action, keeping the optic fiber at a consistent distance from the target.
Measurements are made under the illumination conditions present in the field, and therefore must be taken in a very strict order. A measurement of the white reference is made first, followed immediately by several measurements of the canopy, with an optional (but recommended) re-measurement of the white reference taken last. The initial measurement of the reference is used to standardize the reflectance of the canopy and assumes that the illumination was stable during the intervening period. The second measurement of the reference is used to provide a quality check. If the two measurements are different by more than a few percents, it can be concluded that the illumination conditions have changed. In this case, the canopy measurements may be unusable.
The apparent reflectance of a plant canopy depends on the direction of the observation and illumination conditions, specifically the proportion of diffuse radiance and the direction of direct radiance; we will address this issue in a future article! We collect our measurements looking straight down as this mimics the geometry of our airborne collection.
The useful interval between measurements of the reference target depends on the speed of change of the illumination conditions. More stable illumination allows a greater number of canopy measurement to be made. However, if conditions are changing quickly, then only a few reliable measurements can be taken. An experienced operator can judge the tempo of measurement; however, this is a highly subjective approach, since illumination changes can be imperceptible to the human eye, especially when high cirrus cloud is present.
Overall this approach, except in the very best conditions, results in delays, frustration, low numbers of samples and potentially variable quality spectra.
Modern Field Spectroscopy Methods
Dual Spectrometer System, simultaneously measuring the target and reference to mitigate the impact of changing light conditions.
2Excel geospatial’s wheat disease campaign demanded a solution to this issue in order to increase our capacity to collect high-quality canopy reflectance spectra. We explored several options before committing to the new [FieldSpec] Dual Spectrometer System by [ASD, a brand of Malvern Panalytical].
This system operates two field spectrometers synchronously, allowing for simultaneous measurements of the target and white reference. This eliminated the need to repeatedly measure the white reference, as in the standard approach. The impact of the illumination conditions is thus dramatically reduced, as every target measurement is corrected to reflectance using a white reference measurement made under near-identical illumination conditions.
This system, combined with our new data logging protocol, increased our ability to collect high-quality canopy spectra by many orders of magnitude. For example, over the 2017 season, we achieved a collection rate of almost 12,000 samples per hour.
This reduced the commitment of valuable resources, enabled regular collection co-incident with crop development (less reliant on weather conditions) and produced a comprehensive spectral library for analysis.
At the time of writing, 2Excel geospatial is the only operator of the Dual Spectrometer System in the UK and has a team of trained, experienced field spectroscopists, able to operate the system effectively to produce reliable results. The equipment and operator are available as a stand-alone service or as a component of a remote sensing project.