statistical uncertainty:
comes in whenever the mean or median value is derived of a sample
of signals. ISOPHOT's measurement principle is based on collecting a
large number of redundant data points: several non-destructive read-outs
to establish the slope (=signal in V/s) of an integration ramp and several
integration ramps to make up the final signal. The statistical uncertainty
is determined by the width of the signal distribution and includes photon
noise and detector noise.
systematic uncertainty:
comes in (1) whenever the sample of signals contains non-random outliers
or (2) whenever a data point is subject to a arithmetic operation
(scaling, addition, or subtraction) which requires parameters obtained
from independent calibration observations. The uncertainty of the
calibration parameters can be purely statistical. An important fraction
of the systematic uncertainty comes from the FCS measurement. An error
in the detector responsivity derived from the FCS signal causes the same
(systematic) scaling error in the flux densities of the sky
measurements.
photometric bias:
is the photometric uncertainty in ISOPHOT-data due to uncertainties
in the predicted fluxes of the celestial standards upon which the
ISOPHOT flux or inband power calibration tables are based.
The input fluxes of the photometric calibration standards are based on
model predictions [1]. The photometric accuracy of celestial
standards can vary from object to object (e.g. due to type of asteroid,
spectral type of star, particular planet, etc) which may cause a different
photometric bias for different flux regimes at a given wavelength. The
photometric bias becomes important when comparing ISOPHOT data with
results from other instruments. This term is important for the final
absolute accuracy.
relative accuracy:
is the filter to filter or aperture to aperture accuracy
for a given detector assuming a single value for the detector responsivity.
The relative accuracy is of importance in a multi-filter and/or
multi-aperture observation where a single FCS measurement is used to
calibrate several sky measurements with different filters and/or apertures.
The relative accuracy limits the accuracy in the determination of colours
in case of multi-filter photometry and of extended emission components
in case of multi-aperture photometry.
detection limit:
is the minimum point-source flux that can be detected within a given
measurement time. The detection limit is determined by the source-background
separation which depends on the observing strategy like minimaps, chopped
observations, on- and off-position AOTs, sparse maps, nodding, etc. Cirrus
and galaxy confusion are important external limitations depending on
wavelength, background flux level, and spatial resolution. The photometric
bias has no significant impact on the detection limit.
flux reproducibility:
determines the width of the flux distribution of a large sample of
identical observations (same AOT with identical settings on the same
target) obtained at different times during the ISO mission.
The reproducibility gives a global assessment of the best accuracy that
can be achieved for a given observing mode. It lumps together several
uncertainties: the statistical uncertainty in the sky measurement,
the systematic uncertainty due to the FCS calibration measurement and the
uncertainties due to external noise sources. It can also be regarded as a
measure of the stability of the instrument. This term becomes important in
the analysis of variable sources.