Throughput

AF2+Wyffos Throughput

The throughput of AF2/WYFFOS was measured during February and May 2013, for the following grating / CCD combinations: R300B / Red+4; R316R / Red+4 (observation of a B2 IV star); and R316R / WHTWFC (observation of an sd0 star). The table and figure below show, as a function of wavelength, the measured m_AB zeropoints, i.e. the AB magnitude of a star giving 1 electron per second per Angstrom when observed at zenith.

Wavelength [Å]	3800	4000	4500	5000	5500	6000	6500	7000	7500	8000	8500	9000
R300B - Red+4	12.6	14.5	16.1	16.5	16.6	16.5	16.4	16.1	15.8	15.5	14.3
R316R - Red+4	10.9	13.0	15.7	16.6	16.9	17.0	17.0	16.9	16.7	16.5	16.4	16.2
R316R-WHTWFC	11.1	13.3	15.7	16.5	16.8	16.8	16.7	16.5	16.3	16.3	16.4	16.1

Table 1: Observed mAB zeropoint (the magnitude of a star giving 1 electron per second per Angstrom when observed at zenith) for the whole system: atmosphere, telescope, prime focus corrector, AF2, WYFFOS and detector.

Figure 1: Observed mAB zeropoint (the magnitude of a star giving 1 electron per second per Angstrom when observed at zenith) for the whole system: atmosphere, telescope, prime focus corrector, AF2, WYFFOS and detector. The red and blue curves are for the R316R and R300B gratings respectively, with the Red+4 CCD. The green curve is for the R316R grating, with WHTWFC. The differences between the red and blue curves (measurements with the same detector) are consistent with the known differences in grating efficiencies. See the text for further details.

The m_AB zeropoints have been normalized to the median fibre throughput.

Aperture correction
The above measurements were made in seeing ~ 0.9 - 1.0 arcsec (according to the ING DIMM), and no aperture correction has been applied; the measurements refer to the numbers of photons entering the 1.6-arcsec apertures of the science fibres, i.e. they will be typical of observations of stellar objects made in this seeing, through a fibre with close to median throughput.

The aperture correction depends on the seeing, and on the accuracy with which the stellar image is centred on the fibre aperture. For Moffat-profile seeing, the aperture correction can be calculated from the plot on the AF2 target-acquisition page.

For the measurements with the Red+4 detector, the accuracy of centring the star on the fibre was ensured to be < 0.2 arcsec by making a spirally-rastered series of observations with step 0.4 arcsec, and using the observation with the highest throughput (also coincident with the peak in the distribution of intensities, i.e. not due to a short-term seeing fluctutation). The Moffat-profile plot shows that in seeing of 1.0 arcsec, with a centring error of 0.2 arcsec, only a fraction 0.03 of the light will be missed by the 1.6-arcsec fibre aperture.

For the measurements with the WHTWFC, the accuracy of centring is probably ~ 0.4 arcsec, since the highest-flux image was selected from a cross-pattern series of observations with step 0.8 arcsec. This corresponds to a loss of up to ~ 0.14 of the light.

Other errors in measured zeropoint
The aperture correction probably dominates the rms error in measuring the zeropoints, at least in the red part of the spectrum (wavelength > 5000 A). In the blue, there may be additional sources of error (~ 0.1 mag) given the rapid change of stellar spectral energy distribution (particularly for the B2 IV star) and system efficiency (e.g. CCD QE) with wavelength.

Comparison of observed and expected AF2/WYFFOS throughputs
The measured zeropoints were used to calculate the net throughput of the whole system (atmosphere, telescope, PF corrector, AF2, WYFFOS and Red+4 CCD): ~ 0.02 at 4500 A; ~ 0.06 at 5500 A, and ~0.07 at 6000 - 9000 A. The throughput in the red is broadly as expected from the known or predicted throughputs of the individual optical components in the light path, that at 5500 A is about 25% low, and that at 4500 A (and at bluer wavelengths) at least a factor of two below expectation. The low throughput in the blue has probably always been a feature of AF2/WYFFOS, but it is under investigation.

The predictions of counts made by ING's exposure-time calculator, SIGNAL, are based on the on-sky measured zeropoints reported on this page (not on a priori expectations about the throughput of individual components).

Comparison between Red+4 and WHTWFC zeropoints
Between 5000 Å and 7500 Å, the differences between the red and green curves in Fig. 1 (zeropoint measurements with the same grating) are consistent with the higher quantum efficiency (QE) of the Red+4 with respect to the WHTWFC detector. At redder and bluer wavelengths, the differences are mildly inconsistent with the expected QE of the two detectors:

(1) At wavelengths blueward of 5000 Å, the WHTWFC (green curve) should be ~0.1 mag more sensitive than Red+4 (red) at 4500 Å and ~0.3 at 3800-4000 Å. In the latter wavelength region, this is clearly not the case. However, as pointed out earlier, the errors in measuring zeropoint are likely to be higher in the far blue, and different fibres were used for the measurements (#30 for Red+4, #02 for WHTWFC), so some fibre-specific chromatic anomaly could be responsible for the discrepancy. Pending further on-sky investigation, we advise observers that at wavelengths bluer than 4500 A, the WHTWFC probably will have a small QE advantage over Red+4.

(2) At wavelengths >7500 Å, the curve for the WHTWFC detector (green) in Fig. 1 rises, rather than falling, the latter being expected from the wavelength dependence of grating efficiency and CCD QE. This behavior was also seen when the throughput of WHTWFC was last measured, in 2004. The cause is unknown.

Zeropoints for other gratings
Zeropoints for a grating R_new other than R316R and R316B can be obtained by scaling the m_AB magnitudes in the table according to the relative efficiency of the AF2/ISIS gratings using e.g.

m_{AB_Rnew} = m_{AB_R316R} + 2.5 x log₁₀(Rnew/R316R)

Relative throughput of fibres

The relative throughput of the science fibres was measured in February 2013 with the WHTWFC detector, and is plotted in Figure 2 against fibre number. Fibres with zero throughput are disabled, i.e. cannot be configured for observing.

Figure 2: Throughputs of AF2 fibres, relative to the median of all fibres. Each measurement is labelled with the fibre number.

The measurements were obtained with the fibres configured in a circle, illuminated by sky, and are based on the mean observed intensity in the wavelength range ~4200Å to 9000 Å.

The list of relative throughputs plotted in Fig. 2 can be downloaded from here.

More information about the fibres can be found on the AF2 fibres page.