Purpose

Spectrophotometric standard stars are observed for all grisms on a regular basis, dependent on the usage of the settings for science. Their purpose is to monitor the overall instrument efficiency (telescope+instrument+detector) and to provide a measure of the spectral response. They are measured for MOS and IFU separately.

Although a single standard star observation may be well suited for providing a response curve for relative flux calibration, usage of individual measurements is often limited due to two reasons: standard stars are not always observed under optimum weather conditions or under conditions similar to the science observation and instrumental instabilities or scheduling requirements may prevent observation of standards in time.

Master Response Curves

In order to handle the above mentioned shortcomings, standardized response curves which are averaged from individual measurements over a dedicated time period can be very helpful. Such curves are called master response curves in the following.

The VIMOS pipeline extracts standard star spectra for MOS in the same way as science observations. In case of IFU, the fibre spectra are, in addition to the procedure for science, also sky subtracted and summed. The response curve is calculated as:

R = F_std / f_std

with

F_std: tabulated standard star flux in erg/cm**2/s/A
f_std: observed flux corrected for airmass and extinction
f_std = f_raw 10**(0.4*airmass*extinction), f_raw in electrons/s/A.

Master response curves have been created from pipeline-reduced standard star observations by applying the following steps:

1. Displaying all response curves measured during a dedicated period for one instrumental setting
2. Deselecting curves that clearly deviate
3. Averaging the remaining curves.

Verification

For verification of the master response curves, several standard star observations for all settings have been treated as science and the extracted spectra have been flux-calibrated using the master curves. The results have been compared to the tabulated flux values.

Example for MOS: Observation of CD -32 9927 from 2006-03-31 with HR_blue grism (no filter). The left image shows the tabulated flux and the extracted spectra (not flux-calibrated) from quadrants 1 and 3. The spectra differ significantly at longer wavelengths. In the middle image, the flux-calibrated spectra from all four quadrants are compared to the tabulated flux. The spectra from all quadrants have nearly the same level now. The right image shows the flux-calibrated spectra divided by the tabulated flux. These residuals are better than 10% over most of the spectral range.

Example for IFU: Observation of EG 21 from 2006-08-23 with LR_red grism, OS-red filter, and 2:1 IFU magnification. The spectra from all fibres have been summed and an estimated background subtracted (black curves). The tabulated flux is shown in red. (Please note that the binning of the tabulated flux is much lower than the resolution of the extracted spectra.) The observations show artefacts from filter transmission (not corrected during reduction) but the overall level agrees well.

Accuracy and limitations: Comparison of flux-calibrated spectra with tabulated flux values show that, in general, an accuracy of 10% can be achieved for relative calibration within each spectrum. The master response curves represent, however, a snap shot in time and changes in the overall efficiency of the telescope+instrument cannot be accounted.

Availability and Usage

Master response curves have been created for most MOS and IFU settings. Details can be found in the tables below. The response curves for the HR_red grism apply only to the new holographic grisms which are operational since November 2005.

Table 1: available MOS settings. The time range from which standard star observations have been taken is indicated. The number of actually used observations for averaging can vary between quadrants.

Grism	Filter	Time range	Number of observations
LR_blue	OS-blue	2005-09-08 to 2006-07-22	24..29
LR_blue	free	2005-09-25 to 2006-07-29	3..5
LR_red	OS-red	2006-08-21 to 2006-11-26	9..10
MR	GG475	2006-09-13 to 2006-11-26	15..16
MR	OS-red	2006-01-01 to 2006-05-30	16..19
HR_blue	free	2006-01-04 to 2006-07-01	10..11
HR_orange	GG435	2006-08-16 to 2006-11-26	4
HR_red	GG475	2005-11-30 to 2006-04-25	2..3

Table 2: available IFU settings. The time range from which standard star observations have been taken is indicated. The number of actually used observations for averaging can vary between quadrants.

Grism	Filter	IFU magnification	IFU shutter	Time range	Number of observations
LR_blue	OS-blue	1:1	off	2006-07-18 to 2006-10-27	8..11
LR_blue	OS-blue	2:1	off	2005-11-10 to 2006-04-30	4..6
LR_red	OS-red	1:1	off	2005-09-06 to 2006-05-25	9..14
LR_red	OS-red	2:1	off	2005-09-23 to 2006-08-22	5..8
MR	GG475	1:1	on	2006-11-13 to 2006-11-26	7..8
MR	GG475	2:1	on	2005-11-10 to 2006-02-01	4..5
HR_blue	free	1:1	on	2006-09-20 to 2006-11-14	7..9
HR_blue	free	2:1	on	2006-09-18 to 2006-11-19	8..9
HR_orange	GG435	1:1	on	2006-10-06 to 2006-11-16	8..10
HR_orange	GG435	2:1	on	2006-10-11 to 2006-11-13	4..5
HR_red	GG475	1:1	on	2006-02-23 to 2006-11-25	3..6
HR_red	GG475	2:1	on	2006-02-22 to 2006-08-19	2

Usage: Master response curves R_mst can be applied by hand to extracted spectra with the following procedure:

F_obs = R_mst * f_raw * 10**(0.4*airmass*extinction)

with f_raw in electrons/s/A. The VIMOS pipeline flux-calibrates spectra automatically when a response curve is provided with the tags MOS_SPECPHOT_TABLE or IFU_SPECPHOT_TABLE, respectively, in the set of frames.

Download

Sets of master response curves for MOS and IFU are available as tar files

for MOS: VI_PRSM_070112.tar

and for IFU: VI_PRSF_070112.tar .

The file names have the following formats.

For MOS:

VI_<TYPE>_<ORDFLT>_<GRISM>_Q<QUAD>.tfits

For IFU:

VI_<TYPE>_<ORDFLT>_<GRISM>_<IFUMAG><IFUSHUT>_Q<QUAD>.tfits

With:

<ORDFLT>: order separation filter (OS-blue, OS-red, GG435, GG475, or "_" for free)
<GRISM>: grism (LR_blue, LR_red, MR, HR_blue, HR_orange, HR_red)
<QUAD>: quadrant (1, 2, 3, or 4)
<IFUMAG>: IFU magnification 1 (for 1:1) or 2 (for 2:1)
<IFUSHUT>: IFU shutter ON or OF (OFF)

The response files are fits tables with 7 columns for compatibility with the VIMOS pipeline but only columns 1 (WAVE) and 7 (RESPONSE) are filled.