Performance Verification for TAUVEX
From Tauwiki
Contents |
[edit] Introduction
This document describes the performance verification phase of the TAUVEX payload. Although the instrument will be calibrated on the ground, the true characterization of the instrument must come from an on-orbit calibration where the performance of the instrument under flight conditions is measured. The primary function of the Performance Verification phase is to perform this characterization; however, it is expected that significant science may be obtained in parallel.
There are two documents describing observations in the P.V. phase. This document is an overview of the philosophy and general description of the on-orbit calibration while the specific mission plan for the P.V. phase is described in the Mission Planning for the Performance Verification Phase document.
It is to be reiterated that TAUVEX is a scanning instrument and cannot observe fixed targets. The entire calibration (and observation) philosophy must take this into account. Where possible, multiple calibration goals will be combined with science observations.
The actual extraction of calibration parameters from the observations is described in TAUVEX Calibration Techniques.
[edit] Ground Calibration
The ground calibration of the TAUVEX instruments has been described in the TAUVEX Ground Calibration document. These observations will form a baseline for the instrument performance which will be verified and built upon by the on-orbit calibration.
[edit] Early Operations
Instrument operations will begin approximately 1.5 hours after launch when power is first supplied to the TAUVEX payload. However, observations will not start until after the completion of outgassing and instrument checkout, about 30 days after launch. The sequence of events between launch and instrument first light is described in the TAUVEX Early Ops document.
[edit] On-orbit Calibration
[edit] Field Calibrations
[edit] Overview
This set of calibrations depend on observations which use objects over the entire field of view. As such, many of them can be combined; ie, one set of observations may be used to achieve several calibration and science goals.
[edit] Point Spread Function (PSF) measurement
The PSF is initially measured on the ground across the detector plane using an array of 25 pinholes and a xenon lamp as the source. In space, the PSF, which may be a function of position on the detector, is generally measured by observing a star field, such as an open cluster. Although TAUVEX is a scanning instrument, the PSF can be measured in space since the scanning smear is smaller than 1/3 of the expected PSF in every 0.128 sec sub-frame.
A PSF test can be done as part of other calibration observations such as a distortion test or flat field and will also be continually checked as part of regular observations.
The PSF is not used in the normal operation of the pipeline but will be necessary for accurate photometry and separation of sources in crowded fields.
[edit] Boresight
The three TAUVEX telescopes will be coaligned on the ground to a precision of 6' but may change in orbit due to launch vibration or other effects. The relative pointing will be measured accurately in orbit as the same fields will be observed by all three telescopes. Although often done from known star fields with good astrometric data, it may be possible to use normal observations provided that there are a sufficient number of bright stars.
[edit] Distortion
As part of the ground calibration, a distortion map is made using a 200-hole target located at the collimator focal plane. The coordinates of each point are found and their deviation from the actual position can be calculated and corrected for using a high order polynomial.
After launch, the distortion correction will be calculated using a known star field. The primary requirement is that there are a sufficient number of bright stars spread over the 0.9° field of view but with sufficient separation to avoid confusion. Good astrometric accuracy is required but with no stringent requirements on the photometric accuracy. In fact, because positions of stars are much better known than their brightnesses, the distortion correction can be performed on virtually any field where there are a sufficient number of stars, regardless of whether or not it is a "calibration field".
[edit] Large scale sensitivity (Flat field)
The flat field is a map of the relative pixel-pixel variations over the detector and will require accurate photometry of stars over the entire field. A one dimensional subset is possible without accurate photometry because all stars are scanned across lines of celestial latitude and repeated scans with small offsets of a set of sources will allow a flat fielding of the entire detector, at the cost of several days of observations.
It should be noted that the sensitivity of this, as with many of the calibrations, will be limited by intrinsic photon noise whereas the ground calibration observations will have a much better accuracy of at least 1.5% per resolution element. Therefore the on orbit measurements can be used to check the consistency of the ground measurements.
[edit] Dark frame
The dark count may be a function of the orbital position and must be characterized in flight. It will be measured for a period of 1 minute at the start and end of each observation.
Longer periods of dark count characterization will be done during the times when the stray light increases beyond acceptable levels. The filter will be moved to the closed position and the dark count taken.
If early observations show that the dark count is variable, we may be forced to spend more time characterizing it.
The ground based calibrations will provide a guideline for the amount of straylight that is most suitable for the dark measurements using each filter.
[edit] Stray light rejection
[edit] Solar Light
The primary source of scattered light will be sunlight scattered off the spacecraft surfaces. This may be a complex function of the solar angle and the look direction. We have developed a model for the scattered light, described in Observational Windows and Limiting Magnitude Maps but this is based on an incomplete understanding of the reflective and scattering properties of the spacecraft surfaces.
Differentiating the solar scattering from other sources of diffuse radiation may be difficult and a planned series of observations must be taken at different solar angles to distinguish between them.
[edit] Bright Point and Extended Sources
Because of the scanning nature of the TAUVEX instrument, many bright sources will be observed or will be close to the field of view. While some sources may force instrument shutdown or closure of the filters, it is nevertheless useful to measure the stray light from objects which are out of, but near to, the field of view of TAUVEX. Although this has been done on the ground, it is difficult to measure the scattering at small angles.
Some part of this test may be done serendipitously as TAUVEX scans the sky, however special tests may also be scheduled. Because of safety considerations, this test may have to be done in several stages with progressively closer observations to bright stars or even the Moon.
As a special case, we will track the performance of the "closed" filter as it passes over bright stars or the bright moon.
[edit] Source Calibration
[edit] Photometric Calibration
The baseline photometric calibration will be done as part of the ground calibration sequence and will be verified and corrected using on-orbit observations of calibration stars. In principle, this can be part of regular observations as a large number of stars will be observed during each observation. However, the flux of most of these stars is uncertain resulting in a corresponding uncertainty in the calibration. Thus such serendipitous observations are, in general, only useful for performance verification - to show that there have not been drastic changes in the calibration.
An absolute calibration must be done using a set of primary standards in the ultraviolet. Fortunately, there has been a program to define a set of calibration standards for the purposes of the UV instruments on the Hubble Space Telescope (Lindler and Bohlin 1986; Bohlin et al. 1990; Bohlin 1996). These include a set of 27 standard stars spread over the sky (PV_Phase-Appendix A).
We have convolved the spectra of all these sources (the spectra may be downloaded from the CALSPEC database http://www.stsci.edu/hst/observatory/cdbs/calspec.html or from ftp://ftp.eso.org/pub/stecf/standards) with the TAUVEX filters to produce the expected count rate listed in PV_Phase-Appendix B in each filter. The table also shows the observable periods for the sources based on a conservative stray light limit of 10000 cts/s. Marked in blue are those targets which can be observed in the first month after TAUVEX launch as part of the calibration observations.
A few of the HST calibration sources may not be observable with TAUVEX, either because their count rate is too high (red lines in PV_Phase-Appendix B) or they are in locations of high stray counts (purple lines in PV_Phase-Appendix B). However, the observability of a source is filter dependent and a star which cannot be observed using one filter (e.g. BBF) may be observable using another filter (e.g. SF1).
It should be noted that our calibration philosophy assumes that the spectral response has been determined accurately by the ground calibration and that we simply apply a scale factor to the overall response curve. Determining spectral corrections from imaging data is a difficult and possibly non-unique process. However, because TAUVEX will observe a large number of stars over the course of its operations, discrepancies will be observed and some filter corrections may be made.
[edit] Red leak test
Many UV instruments have had a red leak (long wavelength transmission) which can lead to contamination from cool stars in the field. This can be tested by observing stars which do not emit in the UV, like M-type stars. Any count detected in the UV gives a measure of the red leak. PV_Phase-Appendix C gives the predicted counts for the candidate stars. This can also be done as part of routine observations whenever there are late-type stars in the FOV.
[edit] Light leak test
Even with all the filters closed, some photons manage to reach the detector. This is known as a light leak. This will be measured by keeping all filters in the closed position and any photons detected gives the amount of light leak.
[edit] Coincidence loss (Linearity)
Non-linearity in the detected counts can be either due to a bright point source or due to high stray light in the FOV. This is because there is a limit to the number of counts that can be detected per second per coincidence area.
Case 1. High total count rate
In the case of the non-linearity being the result of the entire field being bright, ground measurements can be used to obtain a function which will give the actual count rate from the observed count rate.
Case 2. Bright point source
In the case of bright point sources being responsible for the non-linearity, this can be measured by observing stars of known magnitudes and then used to predict the actual fluxes of observed sources. This will be done only after correcting for the non-linearity in the total count rate (Case 1).
We can observe stars with a wide range of magnitudes in order to get a curve for the value of the coincidence loss. Possible candidates are listed in PV_starlist together with the predicted counts in the Broad Band filter of Tauvex. The HST calibration stars can also be used for this test.
[edit] Schedule
Table 3 shows the schedule for completing the calibration products along with the priority for each test.
| Product | Date of Observation | Date for completion of Calibration |
| Dark frame | ||
| Boresight | ||
| Photometric calibration | ||
| Straylight measurement | ||
| Effective area curves | ||
| PSF | ||
| Distortion | ||
| Coincidence loss | ||
| Red leak test | ||
| Bad pixel table | ||
| Large scale sensitivity | ||
| Pixel-to-pixel sensitivity |
[edit] References
Bohlin R. C., Harris A. W., Holm A. V. and C. Gry 1990, ApJS, 73, 413
Bohlin R. C. 1996, AJ, 111, 1743
Lindler D. J. and Bohlin R. C. 1986, ST ScI report CAL/FOS-032 Link title
