ETC and overheads

Figure 2.2: Left: SNR vs. exposure time and target AB magnitude for the HSC-$z$ filter; right: SNR variations due to the phase and distance to the Moon.
Image sn_vs_texp_20210721 Image mp_md_150720_01

Since S20B, the total requested time given in the proposal must include overheads (detectors readout, slewing time, calibrations, etc.), as the number of hours per night has been redefined to match the classical mode (10 hours = 1 night).

There is an exposure time calculator (ETC) dedicated to HSC, and the QM PIs are encouraged to use it:

https://hscq.naoj.hawaii.edu/cgi-bin/HSC_ETC/hsc_etc.cgi.
This is a python-based script that is called through a simple cgi web interface. It assumes that the total noise comes from: photon noise, sky brightness level (as a function of Moon phase and distance), and detector properties such as dark current and read-out noise. It also calculates the sky level, on the basis of Moon phase, Moon distance, and the selected filter. The usage of this script is explained as follows.

First in “Brightness” choose the appropriate filter and magnitude (in the AB system) of the source. In case of extended sources, use surface brightness in arcsec$^2$. Then select between point and extended source. In the first case, give the desired value of seeing (in arcsec), in the latter, give the solid angle (size, in arcsec$^2$). Next, define the observing conditions, namely the Transparency (0-1), Moon (days to/from the new Moon, and Moon distance), and (for point sources) the diameter of the aperture that will be used for photometry. Finally, set the maximum sky brightness level (counts, in ADU) to be used to determine the number of frames.

To calculate the exposure time required to reach a certain SNR , give the desired S/N ratio and click “Calculate exposure time”. The number of frames is ignored in this option. As results, the ETC will compute the total exposure time, limiting magnitude, saturation magnitude (assuming significant non-linearity at 45000 ADU), and the suggested number of frames. If the number of frames is more than 1, the exposure time, limiting magnitude and SNR per frame are also given. The output also contains the sky level (in ADU/frame). At the end, the total time necessary to execute the observations (with overheads) is computed.

To calculate the SNR expected after a certain exposure time, specify the time in seconds. Choose if it refers to the total time, or to one frame, and in the latter case, set the number of frames. After clicking “Calculate S/N ratio” the procedure will give the SNR reached after the total exposure time (“Exp Time”$\times$number of frames, if such option is chosen), as well as in each frame. Total and per frame limiting magnitudes, as well as the saturation magnitude and sky level in ADU, will be computed. If the number of frames is specified, and the sky level exceeds the maximum value given earlier, we suggest to shorten the “Exp Time” and increase the number of frames. If “in total” option is chosen, the ETC splits the observation into multiple frames automatically. As before, the total time necessary to execute the observations (with overheads) is also computed.

Currently, the ETC does not include dithering patterns. Set the number of frames accordingly to your desired dither. To simulate several exposures per dither position (coadds), multiply the number of frames accordingly. For example: set 5 frames if you use the pre-defined 5-point dither, but if the resulting sky level is higher than the maximum sky count given, set frames to 10, 15, 20, etc.

In the future, a stand-alone python code will be available for off-line use.

In addition to the ETC, starting from S20B there is also an "HSC overhead and required time calculator", which is to be used to calculate the total requested time for the observing plan:

https://www.naoj.org/cgi-bin/ohc.cgi.

The usage of this calculator is as follows. PIs must include at the time of the proposal submission 40 sec per exposure as overheads for readout and dithering, as well as 1.2 hours per night to their total request for the other overheads. This latter, operation-related overhead, is shared by all programs, with the amount proportional to the effective on-source exposure time per night. Note that the length of a night is assumed to be 10 hours; therefore, the effective length of the night used for queue proposals is (10-1.2) = 8.8 hours. Here, the 1.2 hours per night for the operation-related overheads (calibrations, focusing, filter exchanges and telescope slew) is an average value calculated based on former observing records. For more details, consult

https://www.naoj.org/Observing/Instruments/HSC/hsc_queue_overhead.html.