Cosmic Origins Spectrograph Instrument Handbook for Cycle 17

3.4 Basic Instrument Operations

---

3.4.1 Target acquisitions

The COS flight software (FSW) provides two very different methods for acquiring and centering a target in the aperture. The simplest and fastest method uses the ACQ/IMAGE command to obtain a direct image of the aperture in the NUV and to then move the telescope to the centroid of the measured light. ACQ/IMAGE is the strongly preferred method in almost all cases, but the object's coordinates have to be accurate enough to ensure that it falls within the aperture after the initial pointing of the telescope. The other acquisition method uses dispersed light from the object to be observed, and can be performed with either the NUV or FUV detector. The details of acquiring objects with COS are described in Chapter 7, "Target Acquisitions" on page 83.

As noted above, both COS detectors have both global and local count rate limits to ensure their safe operation. It is the local rate limit that matters during acquisitions that use ACQ/IMAGE, and the COS Exposure Time Calculator (ETC) provides tools for estimating acquisition count rates. The other safety concern for acquisitions is the presence of nearby objects that may be bright in the ultraviolet, and APT provides a means of checking that. For more information on count rate limits, see 7.2 Safety First: Bright Object Protection / 84.

3.4.2 TIME-TAG and ACCUM

In COS' TIME-TAG mode, both the location and time of individual photon events are recorded in the memory buffer. The location is recorded in pixel units, and the time to within 32 msec intervals. Having such data allows for more sophisticated data reduction if there is evidence after the fact for spectrum drift, say, or noise events. On the other hand, in TIME-TAG mode the maximum permissible count rate prevents the observation of some bright stars, and, in addition, the observer must provide a fairly accurate estimate of the BUFFER-TIME so that the memory buffer does not overflow with too many events and does not need to be read out too often either.

The other mode option is ACCUM, which simply places photon events in their proper pixel location and integrates for a specified period of time.

Both TIME-TAG and ACCUM modes may be used with either the FUV or NUV channel.

3.4.3 Wavelength calibration

COS includes platinum-neon hollow-cathode lamps as a rich source of comparison lines for the ultraviolet. The lamp illuminates an aperture that is separate from that used for the astronomical source, and so the images of the Pt-Ne emission lines fall on a different part of the detector. It is the relative locations of the object and comparison spectra that are used to calibrate the wavelength scale. The pipeline software is written to do this automatically. In addition, when used in ACCUM mode the flight software will automatically correct the locations of detected events for the projected orbital motion of HST. In TIME-TAG mode this correction is not made on board but is removed later during the reduction in calcos.

One reason to prefer TIME-TAG mode over ACCUM is that TIME-TAG includes an option in which brief comparison spectra are obtained several times during the course of a long exposure. Doing this allows any drifts in the spectrum to be corrected for; small motions of the optics selection mechanism have been seen during ground tests of COS.

The quality of the acquisition of the object being observed also influences the quality of the wavelength calibration. In particular, an accurate wavelength zero point will only result if the object is well-centered in the aperture in the along-dispersion direction. More information on this is provided in the chapter on acquisitions (see "Centering accuracy and the wavelength scale" on page 90).

3.4.4 Typical observing sequences

For most observers in the majority of cases the following sequence of events will produce data of the necessary high quality:

• Acquisition of the object using ACQ/IMAGE. This should require no more than about ten minutes. This can be preceded by an ACQ/SEARCH if need be to scan a larger area of sky.
• Obtaining spectra in TIME-TAG mode with FLASH=YES so that the spectra can be corrected for any drifts. The COS ETC will provide a means of calculating essential parameters such as BUFFER-TIME.
• Obtaining more spectra during additional orbits as needed for fainter targets.

3.4.5 Data storage and transfer

In TIME-TAG Mode, COS produces an event stream with a time resolution of 32 milliseconds. The x and y pixel coordinates of each photon event are stored in a 32-bit word in the COS data buffer memory. At the start of an exposure and after every subsequent 32-msec period which contains photon events, a 32-bit time-of-day word is written to the data memory. If the predicted total number of events from a TIME-TAG exposure exceeds the total COS data buffer capacity of 4.7 × 10⁶ photon events, data must be transferred to the HST on-board science recorder during the exposure.

Transfers of data from the COS buffer during an exposure will be made in 9-MByte blocks (half the buffer capacity). The first such transfer in an exposure requires 110 seconds and all subsequent transfers require approximately 80 seconds. Users must specify a BUFFER-TIME corresponding to the predicted time to fill half the buffer capacity. On-board commanding utilizes the predicted BUFFER-TIME to establish the pattern and timing of memory dumps during the exposure. During the first BUFFER-TIME of an exposure, counts are recorded in one of two 9-Mbyte buffers of memory. After the first BUFFER-TIME of an exposure is completed, data recording switches to the second of the two memory buffers, and the first buffer is read out and so on. If BUFFER-TIME is incorrectly overestimated, the on-board data buffer may fill before the scheduled memory dump, subsequently arriving photons in that buffer-time will not be counted, and a gap in recorded data will occur. The pipeline will correct actual exposure times for any such gaps, so flux calibrations will be correct. The COS ETC provides features to enable computation of BUFFER-TIME.

At the end of each ACCUM exposure, the science data are read out from the detector and transferred to the 18 Mbyte COS internal memory buffer. Subsequently, the data will be transferred to the HST data recorder, and eventually to the ground. Up to nine 1024 × 1024 NUV MAMA images or two sets of FUV XDL 16,384 × 128 pixel detector images for each of the two detector segments may be stored in the internal buffer at any one time. The full internal buffer can be transferred to the data recorder during subsequent exposures, as long as exposures are longer than approximately 3.5 minutes. The COS internal buffer uses 16 bits per pixel for ACCUM mode data, such that a maximum of 65,536 photons per pixel can be recorded prior to counter rollover.