Download FORCAST: Haro 3

Observation Planning with FORCAST: Haro 3
Christian D. Howard
SOFIA-USRA, NASA Ames Research Center, MS N211-3, Moffett Field, CA 94035, USA
Haro 3 represents a good example of how to plan observations of an extended source
using the Sofia Integration Time Estimator (SITE), and to investigate what is feasible for
FORCAST. Included is a description of the SITE signal to noise calculations, and its ambiguity when dealing with extended sources.
Target Selection and Flux Estimates
We turn to observations of an extended source, chosen randomly from the Design Reference Manual (DRM). Haro 3 is a Blue Compact Dwarf (BCD) galaxy has been observed
by Spitzer (Figure 1), along with several other infrared observatories (Figure 2), and serves
as a good example of how observations of an extended object would be accomplished with
FORCAST, as well as how such observations would compliment the currently available data.
Fig. 1.— The 4.5 µm ( IRAC2) image (left) and the 8.0 µm (IRAC4) image (right) of Haro 3. Figure and
caption courtesy Hunt et al. (2006). Haro 3 is an extended source spanning roughly 40′′ . SED models of
this source would benefit from the longer wavelength filter observations from FORCAST.
We examine previous observations of Haro 3 in order to estimate a flux for use in the
SOFIA Integration Time Estimator (SITE1 ). Because we cannot hope to achieve better results in terms of sensitivity or resolution at the shorter wavelengths of Spitzer, we focus
on obtaining observations in the longer wavelength filters of 19.7, 24.2, 31.5, and 37.1 µm.
Observations at these wavelengths will also provide data to help further characterize the
SED of Haro 3. Using the data from Figure 2, one can determine the expected flux at the
FORCAST wavelengths based on the flux as measured from the shorter wavelengths, which
one would assume is a power law relationship. As Figure 3 shows, between the 10-100 µm
regime, Fλ ∼ λα , where α ∼1.6.
1
https://dcsweb.sofia.usra.edu/proposalDevelopment/SITE/index.jsp
Fig. 2.— Table of previous observations of Haro 3 courtesy Hunt et al. (2006), and references therein.
Fig. 3.— Plot of previous flux measurements of Haro 3, taken from Hunt et al. (2006), and references
therein. The dotted line is a straight line fit to the points between 10 and 100 µm. As expected, the flux
follows a power law with Fλ ∼ λ1.6 , which allows an observer to estimate the expected flux at the longer
FORCAST wavelength filters.
Flux Density Estimates
Because Haro 3 is an extended source, one must estimate the curve of growth, i.e. the
integrated flux as a function of radius, for the wavelengths we wish to observe. Figure 4
allows us to calculate a flux density (mJy/arcsec2 ) for Spitzer 8 µm observations, and thus
estimate the 19.7, 24.2, 31.5, and 37.1 µm flux densities. We begin by taking 5′′ annuli
from the Spitzer observations in order to estimate how much of the 40′′ extent of Haro 3
FORCAST would be able to observe. Using the 8 µm data points from Figure 4 yields flux
densities of ∼0.8 and ∼0.13 mJy/arcsec2 for radii of 0-5′′ and 5-10′′, respectively. This yields
corresponding flux densities at the FORCAST filters of 3.4 and 0.55 mJy/arcsec2 (19.7 µm),
4.7 and 0.77 mJy/arcsec2 (24.2 µm), 7.2 and 1.2 mJy/arcsec2 (31.5 µm), and 9.3 and 1.5
mJy/arcsec2 (37.1 µm). Tracing the flux density of Haro 3 is thus calculated out to its
limiting flux value of ∼140 mJy at 8 µm.
Fig. 4.— Growth curves for IRAC (left) and MIPS (right) photometry. In both panels, the horizontal
dotted line traces the adopted total ux from Haro 3, and the dashed line represents the PRF. In the left
panel, IRAC channel 2 is indicated by lled circles, channel 4 by lled squares. In the right panel, MIPS-24
is shown by lled circles, MIPS-70 by open squares, and MIPS-160 by crosses. Haro 3 is extended at shorter
wavelengths and possibly at 160 µm but appears virtually pointlike at 24 and 70 µm. Figure and caption
courtesy of Hunt et al. (2006).
Signal-to-Noise: Point versus Extended Source
When using the SITE to calculate exposure times for extended sources, one must note
that the signal-to-noise (S/N) ratio is calculated on a per pixel basis, contrary to the point
source method of calculating S/N per resolution element2 . Thus, one must take into consid2
As SITE currently stands, the S/N calculation method of per resolution element for point sources and
per pixel for extended sources is not documented. A rough way to verify this is to calculate an exposure
time for a 25 mJy (at 19.7 µm) point source and compare that time to an extended source of 1 mJy/arcsec2.
At 19.7 µm, a point source will have an apparent size of 3.8′′ , or ∼5.1x5.1 pixels. Both point and extended
sources will have the same flux per resolution element, but the estimated integration times will differ significantly for a given S/N ratio. However, if one species the required S/N ratio for the extended source to be
5/5.1∼1, the integration times become equivalent.
eration the source type (extended versus point source) when specifying a desired S/N ratio.
For example, for our point source example of RNO 1B, we required a S/N of 5. At 19.7
µm, the apparent source size (FWHM) is 3.8′′ (5.1 x 5.1 pixels), so the S/N is calculated
p
by summing over every pixel constituting the source size. Since S/N scales as Npix , for
an extended source one can loosen the S/N criteria by dividing the desired overall S/N per
resolution element by the width of the resolution element in pixels. The FWHM is 3.8′′ , 4.0′′ ,
4.3′′ , and 4.7′′ for the 19.7, 24.2, 31.5, and 38 µm filters, respectively (with 0.75′′ /pixel).
Another issue to consider is the fact that because due to the small size of the extended
emission of Haro 3 (∼40′′ ), chopping will be on-chip and thus will reduce the integration
time reported by the SITE by a factor3 of 2, or equivalently, reduces the S/N by a factor
√
of 2. Thus, for a desired total S/N of 5 per beam, each of our four chosen wavelengths
will require a S/N (per pixel) of 0.71 (19.7 µm), 0.66 (24.2 µm), 0.62 (31.5 µm), and 0.57
(37.1 µm). Because the flux density estimated above is quite faint beyond 0-5′′ , we choose
to observe without using the beamsplitter, thus minimizing our proposed integration times.
Exposure Time Estimates
We start by estimating the integration times for the 5-10′′ annulus, since our flux estimates at those radii are quite low, and we wish to get an idea of how faint we might hope
to observe. Using the above values for the S/N per pixel for on-chip chopping, the SITE
reports estimated integration times (including overhead and specifying single wavelength
observations) of 3.4 hours, 4 hours, 6 hours, and 9 hours for the 19.7, 24.2, 31.5, and 37.1
µm filters. As in the point source example of RNO 1B, we assume observing conditions with
27 µm water vapor overburden and 20◦ elevation angle. As can be seen from these exposure
time estimates, Haro 3 will be detectable in the 5-10” annulus at 31.5 and 37.1 µm only with
significant integrations, although will be easily detectable in the 0-5 ′′ annulus.
Because the flux density drops as one moves to larger radii, it becomes clear that the
extended emission of Haro 3 will be undetectable by FORCAST beyond 10′′ , and will be
resolved at the longer wavelengths only with significant integration times. However, the
methods described are a valid example of how to estimate the viability of observing an extended source with FORCAST. Haro 3, however, is still a viable target for point source
observations, as it exhibits evidence of ongoing star formation (Figure 5). Planning observations of Haro 3 with FORCAST would follow the same strategy covered in the point source
example observation of RNO 1B.
3
https://dcsweb.sofia.usra.edu/proposalDevelopment/SITE/itcHelp/ITChelpFORCASTProperties.html#optical
Fig. 5.— left: VLA 3.6 cm contours of Haro 3 superposed on an HST V-band gray-scale image. The
contour levels are 3, 3, 4, 5, 7, 10, and 15 σ (0.023 mJy beam−1 ). Right: An enlargement of region A.
Figure and caption courtesy of Johnson et al. (2004). Although observing the extended emission of Haro 3
is not feasible with FORCAST, it exhibits knots of apparent star formation that would benefit from SOFIA
observations.
REFERENCES
Hunt, L. K., Thuan, T. X., Sauvage, M., & Izotov, Y. I. 2006, ApJ, 653, 222
Johnson, K. E., Indebetouw, R., Watson, C., & Kobulnicky, H. A. 2004, AJ, 128, 610
This preprint was prepared with the AAS LATEX macros v5.2.