34
controls in the APEX system. During a real data flight, the
instrument will be first in calibration mode and over the
desired test site in acquisition mode. A moving window
display will be used in the APEX control unit to monitor
the functionality of the detectors. Two hundred frames
corresponding to one scan line will be recorded
continuously every 40 ms. The framegrabber unit, the data
storage unit and the user interface unit are mounted in racks
that are located in the cabin of the aircraft. The video
electronics unit is a part of the APEX instrument.
3. AUXILIARY COMPONENTS
In order to make the APEX Imaging Spectrometer System
operational, the following auxiliary components must be
available :
an aircraft platform
a “closed” environment for the APEX instrument
a stabilized platform
a differential GPS and accelerometers
in flight calibration means
laboratory calibration hardware
equipment for temperature and pressure control
vibration and shock absorbers
e e e e e. e e e
3. The Aircraft And Navigation
The APEX instrument, when installed in an aircraft, will be
in a protected and closed and temperature stabilized
environment. The pod or box containing the APEX
Instrument has an optical window made of sapphire. During
take-off and landing a mechanical shutter will be closed in
order to protect the window.
The PILATUS PC XII Eagle is the proposed survey aircraft
for the operational missions of APEX. The crew will
consist of an aircraft pilot and the APEX operator. In order
to guarantee the geometric quality of the recorded scenes
accurate navigation data have to be provided. At present the
concept is to use the autopilot of the aircraft plus
differential GPS and an inertial navigation system, and to
record this data simultaneously to the actual scene. A
geometric analysis has shown, that roll, pitch and yaw
stabilization will be necessary in order to provide a precise
geometric coverage of the area at stake. The vibration
decoupling shall be effected by means of shock mounts.
3.2 The Calibration
The primary goal of the calibration strategy is to achieve
absolute radiometric calibration traceable to an established
standard such as NIST for the acquired data. It is necessary
to achieve the best possible radiometric accuracy
performance in order to be able to validate PRISM and
other imaging spectrometers. Further on it is considered to
use standard hardware technology to reduce the amount of
calibration costs for in-flight characterization. During the
critical phases (takeoff, landing) of the aircraft, the
instrument will be protected by a mechanical shutter. In
general there is no in-flight calibration foreseen during data
acquisition, but in-flight pre- and post-data acquisition
calibration activities allow for the monitoring of the
instrument performance.
An integrating sphere will be used in the laboratory as
calibration standard for characterizing the radiometric
response function. The sphere is used as a secondary
calibration standard, traceable to NIST standards. An
irradiance lamp source will be used in conjunction with a
(double) monochromator to calibrate the spectral response
function. The spectral response function for each channel
covering the whole wavelength range (400 — 2500 nm) will
be measured in 0.5 nm intervals using different diffraction
gratings. The calibration of the geometric response function
will be done using an illuminated slit that is projected
perpendicular to the slit in the focal plane of a collimator.
The calibration hardware will also include a PC that
controls all the necessary devices and collects the
calibration data.
In the laboratory, the system will be able to scan both the
internal sphere and the laboratory sphere using a switchable
mirror in the optical path. Built into the electronic part of
the APEX, the compensation offset loop will subtract dark
current estimated from the darkened border pixels of the
respective detector lines during any data recording with the
instrument. The traceability between the laboratory sphere
and the APEX sphere can be established during these
measurements.
The pre-flight calibration will be more a general
functionality test of the instrument than a calibration
traceable to a standard. It is suggested to use a homogenous,
diffuse artificial lamp source to illuminate the instrument
from below the aircraft. The instrument must be fully
operational and ready for data acquisition.
The in-flight calibration is divided into three parts. The
pre-scene calibration will take place just before the data
acquisition. The shutter of the sphere will be closed and
dark current frames are recorded while pointing the FOV
into the integrating sphere. Then the shutter of the
instrument opens and the built in integrating sphere
illuminates a mirror that deflects the beam of the FOV
towards the ground imager. After these calibration tasks,
image data from the scene can be acquired. Similar to the
laboratory calibration hardware, the in-flight calibration
hardware of the APEX system consists of an integrating
sphere calibration standard. The features associated to this
sphere will be similar to the one evaluated for the
laboratory use, but must — due to space constraints in the
pod - be smaller in total diameter.
There will be no in-flight calibration during ground data
collection. The image data acquired are only dark current
subtracted. After having recorded one scene or a sequence
of scenes, the internal mirror will switch and deflect the
FOV towards the internal integrating sphere again.
Vicarious calibration experiments are performed on
selected test flights according to experiments proposed by
Schaepman et al. (1997). For data takes in other test sites,
provisions must be made to provide on site reference
measurements of selected targets and the atmosphere.
4. RADIOMETRIC MODEL
41 Input Parameters for the APEX Model
The modelling of the APEX specifications is based on
MODTRAN (Berk, 1989) at-sensor-radiances. Three
different situations are modelled using an albedo of 1 as
maximum level to determine the detector saturation, a
constant albedo of 0.4 is chosen as SNR validation level
and a 0.01 albedo as minimum level.
The maximum level is suited to measure very high
reflecting targets, such as snow or high reflecting artificial
Intemational Archives of Photogrammetry and Remote Sensing. Vol. XXXII, Part 7, Budapest, 1998
aa a N na nn ^ CA
