The easiest way to learn to compute total orbit time requests is to work through a few examples. Below we provide five different examples:
nebula in the center of the galaxy M86.
These examples represent fairly typical uses of STIS. The target acquisitions used in each example differ slightly as well:
This example is for an observation of the H nebula in the center of the Virgo elliptical M86, using the CCD, the 52X0.2 slit and the G750M grating. A series of exposures is taken, each stepped relative to the next by 0.2 arcsecond, in the direction perpendicular to the slit, in order to cover the inner 0.6 arcseconds of the galaxy completely. Based on the signal-to-noise ratio calculation presented in Spectroscopy of Diffuse Source (M86), we require an integration time of ~30 minutes per position to obtain a signal-to-noise ratio of ~10. The scientific exposures for this proposal, therefore, comprise all of the following:
cr-split=2 ~30 minute spectroscopic exposure1 with G750M at a central wavelength of 
=6768 Ĺ at location 1.
cr-split=2 ~30 minute spectroscopic exposure with G750M at a central wavelength of =6768 Ĺ at location 2.
cr-split=2 ~30 minute spectroscopic exposure with G750M at a central wavelength of =6768 Ĺ at location 3.
We need to include time for the guide-star acquisition at the start of the first orbit, followed by an acquisition exposure. No peakup will be done, since we are covering the nebula by stepping the slit. In this example, we assume that the acquisition is done using the diffuse acquisition, with a checkbox size of 25 pixels (roughly 1.25 arcseconds). The checkbox needs to be large as this galaxy has a very flat and dusty profile; see Figure 8.5.
The mean surface brightness of the galaxy within this region is ~2 x 10-15 ergs sec-1 cm-2 Ĺ-1 arcsec-2, based on WFPC2 V band images in the HST archive. Using the information in Computing Exposure Times or the STIS Target Acquisition Exposure Time Calculator we determine that, using the CCD Long Pass filter, F28X50LP, for texp= 1 second, we more than achieve the required signal-to-noise ratio needed over the checkbox for the target acquisition. We use the formula in Table 9.1, plug in CHECKBOX=25 and exptime=1.0, and determine that the acquisition will take roughly 8 minutes.
This is not a CVZ observation, so because more than 1 orbit is required we need to include time for the guide-star reacquisition at the beginning of each orbit. The individual exposures in this example are long enough that we do not need to include extra overhead for data management. We are satisfied with the automatic wavecal exposures which are taken for each spectroscopic observation at a new MSM position. We do not require fringe flats as we are observing at wavelengths shortward of 7500 Ĺ.
We assume a visibility period of 52 minutes per orbit, appropriate for a target at M86's declination (see the Call for Proposals). Based on the reasoning presented in Table 9.3, below, we conclude that the observations can be squeezed into ~2 orbits with some loss of sensitivity. Alternately, one could choose to increase the signal-to-noise, and ask for 3 orbits.
In this example the scientific objective is to observe the solar-analog CVZ star P041-C from the near IR to the near UV with STIS's low-resolution, first-order gratings and the 52X0.5 arcsecond slit. The series includes:
cr-split=2, ~5 minute spectroscopic exposure with G750L at a central wavelength of = 7751 Ĺ.
cr-split=2, ~7 minute spectroscopic exposure with G430L at the central wavelength of = 4300 Ĺ.
cr-split=3, ~43 minute spectroscopic exposure with G230LB at the central wavelength of = 2375 Ĺ.
We need to include time for the guide-star acquisition at the start of the first orbit, followed by an acquisition exposure. This target is a bright point source. We will use the longpass filter F28X50LP for the target acquisition and the default overhead of 6.0 minutes from Table 9.1. No peakup is needed as we are using the 0.5 arcsecond wide slit. This is a CVZ observation so each orbit is ~96 minutes. We need to include time for the CCD long-wavelength fringe flats, and since this is a CVZ observation the fringe flat will not move into the occultation. As shown in Table 9.4, we can easily perform this observation in a single orbit.
In this example the scientific objectives are to obtain [O II] images of planetary nebula NGC 6543, as well as an optical spectrum at H
and an ultraviolet spectrum at C IV. The exposure-time calculations for these observations are presented in Extended Source with Flux in cgs units (NGC 6543): Imaging and Spectroscopy. The specific exposures in this series include:
cRSPLIT=2 ~5 minute exposure with the F28X50OII filter.
CRSPLIT=2, ~30 minute exposure with G430M at a central wavelength of c = 4961 Ĺ using the 52X0.1 long slit
G140L at C IV and the 52X0.1 long slit.
We need to include time for the guide-star acquisition at the start of the first orbit, followed by an acquisition exposure. The central star of the Cat's Eye Nebula is used as the acquisition target. It has a V magnitude ~11.5. Checking Table 8.2, we conclude that the star is not bright enough to saturate the CCD in imaging mode with the longpass aperture F28X50LP, and we therefore use it for the target acquisition. We use the overhead information in Table 9.1, and Table 9.2, and conclude that the target acquisition will take 6 minutes. We wish to perform a peakup exposure as well, to center the star in the 0.1 arcsecond wide slit. We consult Table 8.2 and conclude that the source is not bright enough to saturate the CCD if we perform an undispersed (white light) peakup with the mirror. Again using the information in Table 9.1 and Table 9.2, we conclude the peakup will take 6 minutes.
This is not a CVZ observation, so because more than 1 orbit is required we need to include time for the guide-star reacquisition at the beginning of each orbit. The individual exposures in this example are long enough that we do not need to include extra overhead for data management. We are satisfied with the automatic wavecal exposures which are taken for each spectroscopic observation at a new MSM position.
We assume a visibility period of 57 minutes per orbit, appropriate for a target at our source's declination of 66 degrees (see the Call for Proposals). Based on the reasoning presented in Table 9.5 below, we conclude that a total of 2 orbits is required to perform these observations. Note that the MAMA and CCD observations have been split into separate visits in accordance with the required policy
In this example we wish to obtain a long total integration (420 minutes) in the CVZ using E230H and the 0.2X0.09 slit. The exposure-time calculations for this example are presented in Echelle Spectroscopy of a Bright Star with Large Extinction (Sk -69° 215).
We choose to break the observation up into roughly identical 60 minute exposures (though the actual scientific exposure time per orbit will be reduced to accommodate the required overheads and still remain within the MAMA 5 orbit limit). We acquire the target using a CCD point-source acquisition and then peakup in dispersed light using the CCD and the same slit as intended for the scientific observations. The star is Sk -69° 215, an O5 star with a V magnitude of 11.6. Checking Table 8.2, we conclude that the source is not so bright that it will saturate the CCD if observed for 0.1 seconds in the longpass filter F28X50LP, and we choose to perform the acquisition then on Sk -69° 215 with this filter as the aperture. We take the acquisition time as 6 minutes, from the overheads in Table 9.1.
We then perform a dispersed-light peakup using the G230LB grating with the CCD detector. We can estimate the exposure time required by determining with the spectroscopic ETC the total counts over the detector in 1 second for the clear filter and scaling by 65% for the slit throughput for 0.2x0.09 ( "0.2X0.09 Aperture" on page 334). Since we must attain 80,000 counts over the detector, we require roughly 9 seconds per dwell point of the peakup. The peakup overhead for this slit is 360 + 20xtexp. We thus conclude that the peakup will require 360 + 20 x 9 = 540 seconds or ~9 minutes.
Since this is a CVZ observation, we do not need to include time for reacquisitions. However, since it is a long observation and a narrow slit, we decide we will re-perform a peakup midway through the observation.
Additionally, since this is a long observation taken at a given grating position, we need to include time for the automatic wavecals which will be taken every 40 minutes of elapsed pointed time.
For CVZ time, an orbit is 96 minutes. We conclude we require a total of 5 orbits to perform this program, as summarized in Table 9.6.
In this program we wish to take deep images of a field to look for faint point sources, as described in Imaging a Faint Stellar Source. We request LOW-SKY as this observation is background limited. At our declination, we find from the CP/Phase I Proposal Instructions, that there are 45 minutes of visibility per orbit. The observations consist of:
We determine that we can execute this program in 1 orbit, as summarized in Table 9.7.
1 Here and below, aCR-SPLIT=n m minute exposure implies there will be n exposures with a total of m minutes across the exposures. In this example there will be 2 exposures each of 15 minutes for a total of 30 minutes.
|
Space Telescope Science Institute http://www.stsci.edu Voice: (410) 338-1082 help@stsci.edu |