Spitzer Space Telescope Post-launch updates to MIPS Dec 19, 2003 ***** UPDATED Jan 20, 2004 ***** (see below) * MIPS is functional, focused, and all AOTs are operating well. * As anticipated, MIPS didn't start aggressively collecting data until late in IOC/SV when the telescope had cooled off sufficiently. That, combined with some inopportune solar actitivity, means that MIPS is still in its check-out period. A number of uncertainties thus remain. We will post updated information on the SSC website as soon as possible. We apologize for any inconvenience. * The 24 micron array works very well. The 24 micron sensitivity for an unresolved point source is now estimated to be about 110 microJy in 500 seconds (5 sigma) on the darkest regions of sky, or about 1.5 times deeper than pre-launch predictions. The saturation levels are estimated to be 4.1 Jy, lower than pre-launch predictions. * The 70 micron array shows good repeatability of measurements and the relative signals from a suite of standard stars agree well with expectations. The good side of the array (see below) currently reaches a sensitivity level of about 6 mJy in 500 seconds (5 sigma) on dark sky for an unresolved point source at the default pixel scale, or about 3 times worse than pre-launch predictions. Additional work with the pipelines might bring improvements. * The 160 micron array currently reaches sensitivity levels of about 15 mJy in 500 seconds (5 sigma, without confusion) for an unresolved point source on dark sky, about a factor of two worse than pre-launch predictions. These results are VERY preliminary due to the small number of observations made to date. We do not yet have enough standard star observations to judge the photometric properties. We are just beginning to get data with the array operating parameters optimized, and we expect that additional work will bring improvements. There is a short-wavelength light leak at 1.6 micron; see below. ****AND SEE JAN 20 UPDATES BELOW.*** * The 70 micron array has suffered from thermally activated high resistance in a cable connection that is external to, but feeds, the instrument. This has resulted in a fixed pattern noise on Side B of the array. The net result is that the noise in Side B is 2 to 3 times that on Side A (see below and above). The photometry/super resolution AOT was modified to optimize Side A. Depending on the goals of your science program, some additional modification to your pre-launch PH/SR 70 micron observing strategies might be required. No modifications were made to the scan AOT; observers need to step by at most 1/2 array steps to ensure full coverage at 70. * During the summer of 2002, during both episodes of Spitzer Thermal Vacuum testing, one block of 5 contiguous pixels on the MIPS 160 micron array exhibited anomalous behavior which is attributable to a thermally activated short in cabling external to MIPS but well inside the CTA. On orbit, this readout is still not operating properly. The photometry/super resolution AOT was modified to optimize the side of the array without a missing readout. Depending on the goals of your science program, some additional modification to your pre-launch PH/SR 160 micron observing strategies might be required. No modifications were made to the scan AOT; for the highest-quality coverage at 160 micron, observers need to step by at most 1/4 array steps. * Observers are reminded that the MIPS bands can be limited by confusion rather than the nominal SNR. We have not yet determined the confusion limit on the sky, but recent estimates of the 5-sigma confusion limits are 50 microJy, 3.2 mJy, and 36 mJy respectively at 24, 70, and 160 micron (Dole et al. 2003, ApJ, 585, 617). In addition, the performance numbers above are for dark sky; sky brightness varies by large factors at 70 and 160 micron. * Observers are reminded that SPOT visualizations are crucial for making sure that your observation is really doing what you think it should be doing! (1) Additional information for 24 microns At 24 micron (for exposures >1 second), we are limited primarily by the natural background of the sky. The data match the SPOT-predicted background levels to within 20-30%. Calibration and repeatability of photometric measurements are now well- constrained to be within 2%, within specifications. This is expected to improve further as we complete testing and pipeline modifications. Effects being worked now are at the 1.5% level. There is some contamination on the pick-off mirror that feeds MIPS. Unlike the other instruments, where dithering can remove these effects, these dark spots move with celestial objects during scan mirror moves. The spots are 3-4 pixels across and are ~20% dimmer than surrounding pixels; they are located primarily in the upper left of the array. These spots were not apparent in ground testing. We are implementing plans to remove this in the flat-fielding process. There is currently evidence of similar contamination at 70 but not at 160 micron. Further tests are pending. Because of higher-than-expected efficiency, the sensitivity is now estimated to be 110 microJy in 500 seconds (5 sigma) for an unresolved point source, or about 1.5 times deeper than pre-launch predictions, for all exposure times. The saturation levels are estimated to be 7 Jy, about 1.5 times brighter than pre-launch predictions. (2) Additional information for 70 microns The 70 (and 160) micron detectors are Ge:Ga that are known to have transient behaviors when responding to IR photons (see further discussion in the Spitzer Observer's Manual, a.k.a. SOM). Initial results indicate that the MIPS data acquisition strategy is robust and the transient effects are well-characterized. The MIPS Ge:Ga detector arrays (both 70 and 160 micron) return well-calibrated and reasonably stable data. Standard MIPS data acquisition and processing produce expected results. Because the far infrared detectors (70 and 160 um) have low frequency noise that modulates the sky signal, achieving the quoted sensitivity levels requires high-pass filtering that will suppress signals from extended sources. When such filtering is applied, the signal to noise continues to improve approximately (but slightly slower than) the square root of the integration time for integrations up to a few thousand seconds. The 70 micron array is producing good data on relatively bright extended sources in scan maps with two passes to get up to 80 total seconds of integration (corresponding to point source one sigma of about 4 mJy). For these sources, observations of blank sky on either side of the source can be used with some hand processing to remove the effects of the slow detector response. Both for the filtered point source case and the unfiltered extended source one, to remove artifacts where the detectors pass over bright sources it is advisable to carry out scan maps in both directions. The cosmic ray hit rate is about twice what was estimated prior to launch. We are still optimizing the settings in the pipeline processing to remove cosmic ray hits. There is some contamination on the pick-off mirror that feeds MIPS, and there is evidence that it affects 70 micron as well as 24; see 3rd item in 24 micron section above. The spots in this case appear to have a ~15% effect. Further tests are pending. Side B shows a fixed-pattern noise that had not been seen in ground testing, and a readout in Side A is inoperative. The fixed-pattern noise in Side B has been traced to a resistive connection in the cabling that manifests itself when the dewar vacuum shell goes very cold. As a result, after processing there is a net excess noise on this side that is 2 to 3 times above the noise on Side A. Side A currently reaches a sensitivity level of about 6 mJy in 500 seconds (5 sigma) for an unresolved point source on dark sky, or about 3 times worse than pre-launch predictions. The visualizations available in SPOT 9 (due out 12/30/03) will include just the fully-functional part of Side A. The dithers and target placement for PH/SR as portrayed in SPOT currently are the correct dithers. All 70 micron raster maps are on hold pending uplink changes expected in early 2004. SPOT does not visualize them properly, and won't until early 2004. To accurately visualize and plan observations at 70 micron at this time, observers should use cluster mode targets instead of raster maps. For the 70 and 160 micron arrays, we are in the midst of testing a longer interval between stim flashes. This could potentially increase the sensitivity of LONG TOTAL integration times using the SHORT EXPOSURES. Most programs using short exposure times are of bright objects and will be fine; most programs achieving long total integration times are using long exposure times and will also be fine. To first order, if your source is in IRAS, 3 sec will be fine, else you should use 10 sec. Impacts to the SED AOT are small; the SED AOT was already planned to be commissioned later than the photometry and scan AOTs. However, the slit position is mostly on Side A, and SED performance is not expected to be significantly affected. (3) Additional information for 160 microns The array currently reaches sensitivity levels of about 10 mJy in 500 seconds (5 sigma) for an unresolved point source on dark sky, about a factor of two worse than pre-launch predictions. These results are VERY preliminary due to the small number of observations made to date. We do not yet have enough standard star observations to judge the photometric properties. Again, we are just beginning to get data with the array operating parameters optimized, and we expect that additional work will bring improvements. The "bite" taken out of the 160 micron array by the malfunctioning readout is currently apparent in SPOT visualizations. The dithers and target placement for PH/SR as portrayed in SPOT currently are correct. Signals on stars at 160 microns are stronger than expected, by about a factor of five (for K stars - not yet measured for hotter stars, but presumably somewhat larger). Review of the instrument design has revealed a weakness in the stray light control that results in a short-wavelength (1-1.6 micron) leak in this band (due to scattering off a blocking filter). Consequences are under evaluation. As a result, performance numbers for this band are still preliminary, but the numbers given above are thought to be approximately correct. Sources with 160 micron fluxes (in frequency units) more than a factor of 100 above a Rayleigh-Jeans law from 1.25 micron will have leak signals <10%. In addition, sources fainter than J magnitude = 5.5 will be below the detection limit relative to extragalactic confusion. Therefore, most programs will not be affected, except for (a) our calibration plans and (b) debris disk programs, e.g. stars without large FIR excesses. Our calibration plans are changing now; debris disk programs where the far infrared excess is expected to be less than a factor of ten relative to the photosphere are likely to be compromised in achievable accuracy. Such programs will need to do a posteriori work with available near-IR data, such as 2MASS, to subtract off the contribution at these wavelengths. See also several items from the 70 micron section above, to wit: The 70 (and 160) micron detectors are Ge:Ga that are known to have transient behaviors when responding to IR photons (see further discussion in the Spitzer Observer's Manual, a.k.a. SOM). Initial results indicate that the MIPS data acquisition strategy is robust and the transient effects are well-characterized. The MIPS Ge:Ga detector arrays (both 70 and 160 micron) return well-calibrated and reasonably stable data. Standard MIPS data acquisition and processing produce expected results. Because the far infrared detectors (70 and 160 um) have low frequency noise that modulates the sky signal, achieving the quoted sensitivity levels requires high-pass filtering that will suppress signals from extended sources. When such filtering is applied, the signal to noise continues to improve approximately (but slightly slower than) the square root of the integration time for integrations up to a few thousand seconds. To avoid artifacts where the detectors pass over bright sources, it is advisable to carry out scan maps in both directions. For the 70 and 160 micron arrays, we are in the midst of testing a longer interval between stim flashes. This could potentially increase the sensitivity of LONG TOTAL integration times using the SHORT EXPOSURES. Most programs using short exposure times are of bright objects and will be fine; most programs achieving long total integration times are using long exposure times and will also be fine. To first order, if your source is in IRAS, 3 sec will be fine, else you should use 10 sec. ---------------- JANUARY 20 UPDATE ---------------------------------- Subject: MIPS performance update Date: January 20, 2004 From: SSC OST, MIPS IST, MIPS IT **** 160 microns **** 1. Based on additional calibrations, we are adjusting the predicted sensitivity at 160 microns to 15mJy, 5-sigma, 500 seconds of integration. This number is uncertain by +\- 5 mJy due to uncertainties in the absolute calibration. 2. At 160 microns, the saturation limit for a 3 second integration is about 1 Jy. For sources brighter than this level, up to about 4 Jy, useful data will be obtained on the first few reads, but the brightest pixels will saturate before the end of the integration. As a result, there will be some degradation of the results in the readouts immediately following, but the recovery will be relatively fast. 3. If the source is blue enough to contribute a substantial signal through the short wavelength leak at 160 microns, then the saturation limit for total signal needs to be reduced accordingly. A K star contributes a leak by about a factor of five over the 160 micron signal, while hotter stars contribute a leak by a larger factor that is not yet well calibrated. The implications of the leak as a function of stellar brightness have been discussed previously. **** 70 microns **** 1. There is no change in the sensitivity estimate at 70 microns. 2. At 70 microns, the saturation limit for the wide field mode in a 3 second integration is about 9 Jy. By allowing saturation and taking the signal from the first few reads, sources up to 25Jy can be measured. However, the consequences for the immediately following reads will be significant, such as latent images and a degradation of linearity. 3. To first order, the saturation limits in the 70 micron narrow field mode will be four times higher than in the wide field one.