Although much calibration work in this wavelength region has been done using Mars, for LWS the Mars spectrum has too many lines to derive an RSRF. Also being an extremely bright source there are concerns about detector behaviour in this flux regime.
For the RSRF calibration the asteroids are generally too faint to achieve a good enough S/N and non-solar system objects are either also too faint such as normal stars or do not have sufficiently accurate models. In principle, a statistical approach using many different sources would also be possible for the fine tuning of the flux calibration especially in the 60 and 100 microns region where IRAS data are available, but this has not been carried out yet.
The RSRF of LWS was derived from a single 5.7 hour 36 full scan LWS01 observation of Uranus interleaved with dark current measurements which took place in revolution 317 on the 28th September 1996. The whole observation was followed by a series of illuminator operations to establish the baseline for correcting any drift in detector responsivity with time.
All previous and subsequent observations are referred to the photocurrent to flux relationship derived from Uranus by taking the ratio between the signal measured when the illuminators are operated during an observation, and that measured when they were operated during the calibration observation of Uranus. To ensure that the power from the illuminators has not changed during the course of the mission, weekly observations have been made of a series of astronomical sources and the signal from these compared to that from the illuminators.
In order to determine the conversion from photocurrent to absolute flux densities the Uranus model provided by Griffin and Orton is used as of OLP v8.7. The accuracy of this model is estimated to be better than 7% at all wavelengths.
These examples cover the flux range observable by LWS. As each of the ten LWS detectors is calibrated separately for each observation, the accuracy of each detector is determined individually. Generally, over the whole dynamic range of LWS for these point sources there is a good agreement with the models. For Ceres, which has a similar flux to Uranus, there is good agreement with the model of Mueller (1997, priv. comm.) except at the longest wavelengths where the LWS spectrum is slightly steeper. For Neptune, the model of Lellouch (private communication, 1997) shows an upward deviation from the LWS data below 70 microns but there is a good agreement beyond 80 microns. The observation of Arcturus is affected by two problems : the source flux for most detectors is less than the dark current, making uncertainties in the determination of the dark current critically important to the accuracy of the derived flux, and the observation is affected by near infrared filter leaks. Nevertheless, there is a general agreement with the composite model taken from the Cohen, Walker, Witteborn et al. (CWW) absolute calibration programme. For Mars there is a general agreement with the model of Orton & Griffin (1997, priv. comm.) except for detectors LW3 and LW4. These detectors are non-linear in this regime and must be corrected for this before comparison with the models. The model of Mars used for this comparison is based on an extrapolation of the model provided by Wright (1996, priv. comm.) for the JCMT FLUXES programme, and, as such, is liable to some error in the short wavelength region. This may partly account for the differences between the model and the derived spectrum in this region, although this may be an indication of an error in the Uranus model at the short wavelength end of the LWS band.
Comparison between LWS and IRAS flux over a wide range of flux values is now being performed (results to be added shortly).
From OLP8 onwards, a 'throughput correction' is applied, thereby giving the FP flux in units of W cm-2 micron-1. The throughput correction is the FP transmission multiplied by the FP resolution element, the two factors being undissociable in continuum observations; it has been derived using Mars as the calibrator.
The Fabry-Perot photometric calibration is derived from observations of Mars made with the FPs set at a fixed gap and the grating scanned over its full range. In this observation mode the various order and wavelength combinations of the FP are selected as the wavelength falling onto the detectors changes due to the grating movement.
It was found that for strong lines accuracy is typically better than 30%. For faint lines however, the FP fluxes can be off by almost a factor 2. This is mainly due to the removal of the darkcurrent which is known to be problematic for low signal levels.