Current Sensitivity Estimates
for M4
Message from Dan Clemens Received on March 1, 1997:
I redesigned M4 for a 40 cm aperture, and did a full ray trace
to establish the size of the secondary and to improve the optical quality
of the instrument.
Using the new optical components, and selecting a working
temperature of 5.6 K (the COBE and SIRTF optics temperatures), I have
computed a new sensitivity table, for the 95um center wavelength,
with bandwidth of 30um:
The S/N = 600 criterion for EACH pixel in EACH of the two arrays
will yield a polarization percentage uncertainty of 0.236%. This,
when combined with a polarization of 1% will yield a position angle
uncertainty of 6.4 degrees. Thus, if we require position angle
uncertainties under 10 degrees for 1% polarized directions (in
order to make sensible magnetic field maps), we must achieve
a photometric S/N > 600.
Again, I used IRSKY to identify the Zodiacal and Galactic and
Target surface brightnesses:
Region Gal.Long. Gal.Lat. Zodi Gal NESB Target Time to
Name [deg.] [deg.] ======[MJy/sr]========= S/N=600
-----------------------------------------------------------------
Gal.Ctr. 0 0 14 20000 9.6 20000 0.08s
3 0 15 2000 3.0 2000 0.8
3 1.5 15 330 1.3 330 5.3
3 3.5 15 113 0.77 113 17.
3 5.0 15 67 0.62 67 31.
Gal.Plane 25 0 10 2400 3.3 2400 0.7
25 1.5 10 330 1.3 330 5.2
25 3.5 10 110 0.74 110 16.
25 5.0 9 78 0.64 78 24.
48 0 6 440 1.4 440 3.6
48 1.5 6 123 0.77 123 14.
48 3.5 6 50 0.51 50 38.
48 5.0 6 35 0.44 35 57.
rho-Oph
star forming core 15 40 0.50 300-2400 < 0.01
envelope around core 15 40 0.50 50 36.
dark filaments 15 40 0.50 5-80 14-3600s
Ursa Major / NCP
Carl's Spider (core) 4.6 0.2 0.16 8 137. (24" pixels)
" " (filaments) 4.6 0.2 0.16 1-3 400-3500s (60" pixels)
M31 (GI Project)
Core 6.4 3.7 0.22 40 11. (24" pixels)
Spiral Arms 6.4 3.7 0.22 1-14 5h - 2m (24" pixels)
-----------------------------------------------------------------
Array parameters:
PSF size at 95um is 60"
Pixel sampling set to 24"
Array sizes of 32x32 = 768x768" = 12.8 x 12.8' = 0.0455 square deg.
"Normal" Observing mode:
Two "orientations" (0 degrees, 45 degrees)
2x2 half-array dithering (6.4 arcmin steps in each cardinal
direction)
Milky Way Mapping:
1000 square degrees (100 l x 10 b)
Non-overlapped Array Footprints = 22,000
Overlapped Array Footprints = 88,000
Independent Pixels = 22 million (24")
Independent PSFs = 4 million (60")
Dwell Time per Footprint = 10 seconds
3 second slew and settle, 7 seconds integration time
Total Integration time per pixel per orientation = 7 x 4 = 28 sec.
Total Time to complete Milky Way Map = 20.5 days
Time Including Contingency, Overheads, Data loss = 28 days
Zone Lowest S/N per 24" pixel Lowest S/N per 60"
------------------------------------------------------------
b<1.5 850 >2000
1.5<b<3.5 520 1300
3.5<b<5.0 420 1000
Summary: We can certainly perform our Milky Way map to +/-5
degrees of latitude in 4 weeks. Inside of about 2 degrees,
we always have S/N >600 for 24" pixels. Inside of 5 degrees,
we always have S/N > 600 for 60" pixels.
Options: If our mapping efficiency stays high, we might consider
extending the latitude range even more (to say +/- 7 deg.)
This may be possible to do "on the fly" during the mission.
Dark Cloud / Star Formation Regions
I'll do the detailed calculations next week.
Ursa Major / Carl-the-spider
The core of the spider has a surface brightness of 8 MJy/sr,
and the legs range in brightness from 1 to 3 MJy/sr (the 3
is for some of the brighter knots in the filaments).
The core can easily be mapped and restored onto 24" pixels.
The filaments are harder, but not impossible.
If we insist on S/N > 600 for 60" pixels for the 1 MJy/sr
non-bright filament portions, then we need 2x2 (dithered)
integrations of 840s per footprint.
Thus, for 2 weeks, to the 1 MJy/sr faintness level, we can map
6.4 square degrees (say, one 2.5 x 2.5 degree zone).
Instead, if we merely want to trace the magnetic field from
one filament knot to the next, the flux floor is 3 MJy/sr
and a total of 57 square degree can be mapped (one 7.5 x 7.5
degree zone).
I'd like Carl and Alyssa to think this one over. Do we need
to see all parts of the filaments? If so, the area surveyed
will be small.
M31
OK, I decided to see what a "typical" GI program would look
like. I poked at M31 with IRSKY at 100um to get the surface
brightnesses reported above.
The rough area is 2 square degrees (2 x 1 degrees), and if
we devote 24 hours to M31, we can perform 2x2 dithering in
two orientations with a total of 350 pointings.
The total integration time per pixel per orientation is then
16 minutes. The sensitivity is 0.22 MJy/sr, so for 24" pixels,
we meet the S/N>600 criterion for all points brighter than
4.3 MJy/sr. For 60" pixels, this drops to 1.7 MJy/sr.
The M31 core easily meets the 24" pixel value, and the spiral
arms range up to 14 MJy/sr, so we will either map magnetic
fields for much of the spiral arms, to 24", or essentially
all of the spiral arms to 60".
Either way, this M31 map will be spectacular. At 24", there
will be some 45,000 independent pixels - or 7,200 at 60".
dc
Back to M4 Homepage