Spitzer Documentation & Tools
MIPS Instrument Handbook

 

3.4.2        Fractional Observing Time Overhead

Here we provide illustrative values that can be used for estimating the observing time required to complete an observation given the integration time required to achieve the desired signal-to-noise on a source.  The observing times derived here are approximate and were subject to small changes.

 

The amount of overhead incurred in making an observation can conveniently (if approximately) be represented as a factor which, when multiplied by the desired integration time, gives the observing overhead time.  The total observation time (or total observation ''wall clock time'') is just the integration time plus the overhead time, and can be represented as:                                            

 

 

Observing Time = (1 + overhead factor) x Integration Time

(3.1)

 

Table 3.6: Approximate overhead factors for MIPS photometry/super-resolution (small field option).

Exposure Time (MIPS sec), Band First-Cycle Integration Time (sec) First-Cycle Overhead Factor Seven-Cycle Integration Time (sec) Seven-Cycle Overhead Factor
3 sec, 24 µm 48 1.40 312 0.36
10 sec, 24 µm 166 0.41 1046 0.11
30 sec, 24 µm 501 0.13 3144 0.04
3 sec, 70 µm (coarse scale) 38 2.07 226 0.68
10 sec, 70 µm (coarse scale) 126 0.85 881 0.38
3 sec, 70 µm (fine scale) 25 4.47 176 2.03
10 sec, 70 µm (fine scale) 84 2.57 587 1.53
3 sec, 160 µm 6 22.13 44 12.19
10 sec, 160 µm 21 12.99 147 8.39
3 sec, 160 µm (enhanced) 13 30.6 107* 11.0*
10 sec, 160 µm (enhanced) 45 14.1 357* 8.9*

* Cycles 7, 11, 13, 17, and 19 are not allowed in 160 µm enhanced mode.  The numbers given in the table are for 8 cycles.

 

In Table 3.6, we tabulate the overhead factors for observations taken in photometry/super resolution mode for the available MIPS exposure times.  Note that the tabulated integration times are in real time seconds, not ''MIPS seconds'' (see section 2.4).  While the combination of dither patterns and minimum exposure time results in fairly long minimum effective integration times, almost any arbitrary integration time longer than the minimum can be obtained by combining sets of 7-cycle observations with single-cycle observations.  For most dither patterns, for cycles up to 7 there is only one spacecraft offset, with all other dithers being done with the scan mirror.  This is to minimize accumulated overhead caused by unnecessarily repeating motions that require slew and settle overheads, while scan mirror motions do not.  For more than 7 cycles, the number of spacecraft motions depends on the natural break points in the sequence.  For the purposes of estimation of observing times, the simple method described here will give adequate results.

 

Note that for purposes of determining whether a source will saturate the MIPS detectors, the exposure time matters, while for purposes of calculating signal-to-noise ratios, integration time is the appropriate quantity to consider.  Note also that these overhead factors are for observations in a single band; if multiple wavelengths are desired, an additional 40 second overhead is incurred for offsetting the spacecraft to bring the object into view of each additional band.

 

Observing overheads for scan map mode are somewhat more difficult to summarize than are those for photometry mode.  Because the three MIPS arrays view different areas of the sky, the length of a scan leg must be somewhat longer than the length of the region being mapped.  The overscan distance required to obtain a 3-wavelength full-coverage map is about 20.7´.  Data are obtained in the overscan region, but are not co-spatial in all 3 bands, and might have somewhat larger than usual pointing uncertainties in the first several frames.  While viable science data, the time spent covering the overscan region is considered observing overhead.  The amount of overhead depends on the scan rate selected by the observer, as summarized in Table 3.5.

 

Exposure times in scan-map mode are determined by the scan rate selected.  Likewise, the integration time on any particular point in the scan-map is also determined by the selected scan rate, although deeper integrations can be obtained by re-scanning a region and coadding the resulting maps.  The exposure and integration times for the three MIPS arrays as a function of scan rate are detailed in section 3.2, and that section should be referred to in order to estimate sensitivity as a function of flux, and saturation fluxes as a function of scan rate. 

 

The length of time required to complete a 5´ wide scan leg is just the length of the leg divided by the scan rate (including the overscan distance).  If the region to be mapped is wider than 5´ (2.5´ for full 70 µm coverage), the map must be built up by specifying multiple scan legs, including specifically some overlap between those legs to ensure complete coverage of the region at all wavelengths.  Each time a scan leg finishes and a new scan leg is begun, an overhead of about 30 to 60 seconds for ''turnaround time'' (time to stop the scan and begin set up of the new scan leg), which includes an offset and settle, is incurred.  There is also a 30 second overhead incurred at the beginning of each scan map AOR for establishing initial pointing, settling, and beginning to slew.