3.1 Summary of the Observing Modes

Four different observing modes were available to users, via the so-called `Astronomical Observation Templates' (AOTs), which allowed to operate the instrument only in a few standardised ways, giving the observer the choice of the wavelength ranges, the sampling intervals, and the exposure times. These observing modes are described in Section 3.2 and summarised in Figure 3.1.
Two extra observing modes have been implemented later in the mission to systematically make use of the time when LWS was not the prime instrument: the parallel mode when another instrument was used and the serendipity mode when the satellite was slewing to another target. Unlike the four AOTs, these observing modes were not available to the users but were used in a systematic way.

Figure 3.1: Modes of the LWS.

Grating mode, in which only the grating was used, provided moderate spectral resolving power of about m in the short-wavelength channels (SW1-SW5) and m in the long-wavelength channels (LW1-LW3, LW5) corresponding to a resolving power of between and depending on the wavelengths being investigated. Because of the multiplexing described in Section 2.4, usable data were obtained from all ten detectors simultaneously. Whether or not these data are useful to the observer depends upon the range of wavelengths selected.

Fabry-Pérot mode, in which one of the two Fabry-Pérots was used in combination with the grating, provided high spectral resolving power between and . Radiation in different orders of the Fabry-Pérot falls on detectors other than the prime detector (i.e. that selected for the wavelength of interest), and in some cases data from the non-prime detectors have proven quite useful although the grating position had not been set to have its peak transmission at the orders falling on these detectors. Automatic recovery of this non-prime detector data is foreseen.

In both grating and Fabry-Pérot modes, scans were carried out at one of two standard lengths of detector integration ramp, 0.25 s and 0.5 s. The total integration time per spectral point was then achieved by varying the number of these ramps. It was originally intended that spectra be scanned by carrying out a number of integrations at each setting of the grating or Fabry-Pérot until the total required integration time had been built up. In order to minimise the low-frequency end of noise, a fast-scanning mode was also implemented, in which only one integration was performed at each position of the grating or Fabry-Pérot. The total integration time was then built up by repeatedly scanning. In fact during the mission this method has been recommended for all observations. Early observations have proven that in this way one could avoid that all integration ramps at one wavelength were affected by the same particle hit.