Spectral Discontinuities; Field-Of-View - ASD FieldSpec 3 User Manual

www.asdi.com
Note:
Note:
B.6

Spectral Discontinuities

B.6.1

Field-of-View

ASD Document 600540 Rev. J
• Changes in temperature.
Conditions can change rapidly or slowly. It all depends on clouds, wind
(affecting temperature), instrument warm up time, etc.
Optimization depends on the response to light in a particular spectral region.
A well-optimized instrument will display between 20,000 and 35,000 raw
digital numbers, if the source is within the gain and integration time abilities
of the system. The magnitude of the VNIR raw DN will be directly dependent
on the light intensity.
The Spectralon blank should be used when optimizing and for taking a white
reference measurement.
It is important that the position of the reference sample when taking a white
reference is as similar as possible to the position for capturing data from the
samples.
When saving reflectance data, point the probe at the Spectralon once every
few measurements for a minute or two with the same viewing geometry. If the
relative reflectance of the Spectralon is less than or greater than one, a new
white reference may be needed. If the relative reflectance of the Spectralon is
greater than one, re-optimization is recommended.
Three separate detectors are used to collect data over the complete spectral
range; therefore spectral discontinuities should be expected at 1000 nm and
1830 nm when operating in raw DN mode. These discontinuities decrease in
the reflectance mode when the ratio of the raw DN file to the calibration file
(a white reference or a radiometric calibration response to a calibrated light
field) is calculated because both files will contain the same discontinuities in
the data.
Excessive discontinuities in radiometric spectra can be caused by not
allowing the instrument to warm up fully. Normal discontinuities can be
addressed by using the parabolic correction utility.
The VNIR spectrometer/fiber bundle has a slightly broader field-of-view than
the SWIR spectrometers. Discontinuities in reflectance data may be caused by
shading differences in the VNIR data as opposed to the SWIR data.
Steps at the spectrometer splices are most commonly associated with
measurements of surfaces with very patchy brightness. The steps result from
the different spectrometers viewing slightly different areas on the sample. The
steps are seen when these three areas (VNIR, SWIR1 & SWIR2) have
92
Appendix B Reference Information
FieldSpec® 3 User Manual