Recommendations on how to choose exposure times for high precision staring mode observations.
Ideally observers should choose the exposure time which maximizes signal to noise ratio but that also minimizes data volume for maximal schedulability. Data volume becomes more and more important as Spitzer gets further away, and will constrain which observations make it into the schedule. A table of Data volumes for the available frame times is listed below.
An important consideration is the range of acceptable well depths to do high precision photometry. We recommend that observers stay in the most linear regime of the detector (see page on nonlinearities), or between 500 and 15000DN. For science targets which would generate the low end of this range of counts, and that require the highest signal to noise ratio (SNR), it is best to use a longer frametime despite the data volume overhead. The table of saturation limits for the warm detectors is attached below.
SNR as a function of bin size is plotted for 2s subarray (black), 6s full (red), 12s full (blue) and 30s full (green). The first two plots show a faint source reaching 500 DN (5% well depth) for 2s subarray observation in both ch1 and ch2. The second two plots show the same information for a brighter source with 1/4 well depth. The SNR calculation in snirac_warm.pro was used to make the curves, and can be used by observers to generate similar curves for their own targets. Note for this analysis, we assumed that the peak pixel value was 40% of the source flux (when centered using PCRS peakup on the sweetspot). We used a smaller aperture size (2 pixel radius, 2-6 pixel background) in determining SNR as that aperture size is used more commonly in staring mode photometry. Aperture size has a big effect on SNR at low flux levels as larger apertures will have a larger relative increase in noise due to the contribution of read noise in low signal pixels.