Extensions on Magellan Megacam images

I am working on astrometric registration of Magellan Megacam images. My first task was to read the multi-extension FITS files. The image files I have possess 36 extensions, each corresponding to one of Megacam's 36 detectors. The order of the extensions follows four columns of nine detectors each, starting at the southwest corner of the mosaic. Thus extension 2 is for the detector immediately north of extension 1, and so on up through extension 9, which is the ninth detector of the first column and hence is at the north edge of the mosaic. Then extension 10 is for the first detector of the second column, and is immediately to the east of extension 1. This is a very sensible way of arranging the detectors and extensions.

One thing that confused me at first is that although the third and fourth columns have the detectors ordered exactly as you would expect based on the above, each detector image in these columns is upside-down relative to the corresponding images in the first and second columns. One can understand how this might come about due to spatial constraints on the positioning of readout electronics.

Interesting papers on astro-ph

I frequently lead the 'astronomy pizza' discussions of interesting recent papers on astro-ph that we hold at Stony Brook every Monday, and I want to start posting brief summaries of some of the papers we discussed.  Here are today's:

ArXiv:1401.2966 Discovery of a young asteroid cluster associated with P/2012~F5 (Gibbs) http://arxiv.org/abs/1401.2966

This is a truly remarkable paper. P/2012~F5 (Gibbs) is an active asteroid. It may be a main-belt comet or have suffered a recent impact. It is of interest to look for collisional families associated with active asteroids for two reasons. First, if such an object is in fact a bona fide main-belt comet, one possible explanation for the presence of volatiles on its surface is a recent (few Myr-ago) collisional fragmentation that exposed previously shielded volatiles in its interior. Second, if instead an asteroid is active due to a very recent collisional ejection of dust (months or less before the first observation of activity), there is still reason to think it might be a member of a young collisional family, because mutual collisions will (for a while) be much more common in such families than in the asteroid belt at large. These authors identify a collisional family of 9 asteroids associated with the active asteroid P/2012~F5 (Gibbs), which has an orbital semimajor axis of 3 AU and a diameter of about 2 km. They perform carefully back-integrations of the asteroids' orbits and arrive at an age of about 1.5 Myr for the cluster – very young in the context of the main asteroid belt.  The largest asteroid in this cluster is considerably larger than the 2 km size of P/2012~F5 (Gibbs) itself, and the minimum diameter for the parent body that was shattered to make the collisional family is 10 km.

arXiv 1401.4000 A new cold sub-Saturnian candidate planet orbiting GJ 221 http://arxiv.org/abs/1401.4000

This detection was based on a re-analysis on radial velocity data previously indicating two shorter-period planets. The shorter period planets are confirmed and a new sub-Saturn mass planet with period about 500 days is tentatively detected. This observations would be interesting if confirmed, but with 99% confidence intervals spanning almost a factor of three in mass and RV amplitude, it seems very tentative indeed at present.

arXiv 1401.4388 HESS J1640-465 - an exceptionally luminous TeV gamma-ray SNR http://arxiv.org/abs/1401.4388

HESS is an array of atmospheric Cherenkov telescopes in Namibia, for the detection of extremely high-energy gamma-rays. Of 70 HESS sources found so far (most seemingly associated with star-forming regions), given its estimated distance of 8-13 kpc, HESS J1640-465 may be the most luminous source of TeV gamma-rays in the MWG. The source is intrinsically extended, and the spectrum from lower-energy experiments connects smoothly with that of HESS. Previously the X-ray through gamma-ray emission from this source was attributed to a pulsar-wind nebula (PWN) in which the high-energy photons would be produced by inverse Compton scattering from highly relativistic electrons in the pulsar wind. The current results suggest instead the acceleration of hadrons across a shock in the supernova remnant G338.3-0.0. The main evidence for this appears to be the absence of a break in the gamma-ray spectrum that should appear at the point where the inverse compton/synchrotron loss time of the parent electron population is similar to the age of the source. Instead of this, “the featureless gamma-ray spectrum over almost six decades in energy is challenging for any leptonic model”. Also the emission is coincident with the shell of G338.3-0.0, and more extended than a pulsar-wind nebula would be expected to be: it is a much better match for a supernova remnant shell.

Other papers that I thought looked interesting, but didn't have time to read or discuss:

ArXiv 1401.2931Distribution of Electric Currents in Solar Active Regions

http://arxiv.org/abs/1401.2931

arXiv 1401.2901 A high resolution spectroscopic atlas of M subdwarfs - Effective temperature and metallicity

http://arxiv.org/abs/1401.2901

arXiv 1401.3199 High frequency A-type pulsators discovered using SuperWASP

http://arxiv.org/abs/1401.3199

arXiv 1401.3692 The GTC exoplanet transit spectroscopy survey I: OSIRIS transmission spectroscopy of the short period planet WASP-43b

http://arxiv.org/abs/1401.3692

AAS press release

A bit silly to wait almost two weeks to post this, but I had a great time at the American Astronomical Society meeting, and a press release I helped give wound up on the BBC front page on Jan 7: http://www.bbc.co.uk/news/science-environment-2564765...7

As well as lots of other places:
http://www.nbcnews.com/id/54005662
http://www.sciencedaily.com/releases/2014/01/140107170600.htm
http://phys.org/news/2014-01-stormy-stars-spitzer-probes-weather.html
http://www.astronomy.com/news/2014/01/spitzer-probes-weather-on-brown-dwarfs
http://www.popsci.com/article/technology/brown-dwarfs-racked-planet-sized-storms-molten-iron
http://www.space.com/24192-stormy-weather-brown-dwarfs-aas223.html

Multi-panel figures in supermongo

As this blog is to serve as a useful record for myself as well as for other astronomers, it has to start with exactly what I'm doing right now, which happens to be rather unfortunately mundane for a first post: constructing multi-panel figures using supermongo.

The basic reference for supermongo is http://www.astro.princeton.edu/~rhl/sm

It's easy to make a figure with multiple panels using separate calls to LOCATION.  The full plot area is defined as 0-32767 in supermongo's basic scale units, so for a one-panel plot with plenty of space for axis labels and some space around the outside you could use "LOCATION 4000 32000 4000 32000", which gives you a square plot where both x and y coordinates range from 4000 to 32000.  Starting at 4000 gives a wider space around the left side and the bottom, so you can fit axis labels without crowding them.  For a plot vertically divided into four panels you could instead use something like:

LOCATION 4000 32000 25000 32000
box
xlabel Time (hours)
ylabel flux
<plotting commands>
LOCATION 4000 32000 18000 25000
box
xlabel Time (hours)
ylabel flux
<plotting commands>
LOCATION 4000 32000 11000 18000
box
xlabel Time (hours)
ylabel flux
<plotting commands>
LOCATION 4000 32000 4000 11000
box
xlabel Time (hours)
ylabel flux
<plotting commands>

This can be used to produce a plot like the following:

There are several problems with this particular plot, however.  First, the repeated "Time (hours)" label is annoying.  We should have it only on the bottom panel.  That is easy: just don't include the xlabel command for any panel except the bottom one.  However, since the scale of the x axis is exactly the same in every panel in this case, we should also remove the numerical labels on the x axis for every panel except the bottom one.  It turns out that using the command "box 0 2" instead of simply "box" will draw the box with numerical labels only on the y axis (if you want, instead, to label the y axis but not the x axis, use "box 1 0").  Finally, since the y axis represents flux in every case, it isn't necessary to label the y axes of the individual panels.  Labelling the y axis just once will also allow for a longer, more descriptive label.  I do this by 'fooling' supermongo by using an initial LOCATION command to specify a large, one-panel plot, and putting the y label on that.  I also change the starting x location to 4500 in every case, to avoid crowding the new y axis label.  The edited set of commands then becomes:

LOCATION 4500 32000 4000 32000
ylabel Normalized flux
LOCATION 4500 32000 25000 32000
box 0 2
<plotting commands>
LOCATION 4500 32000 18000 25000
box 0 2
<plotting commands>
LOCATION 4500 32000 11000 18000
box 0 2
ylabel flux
<plotting commands>
LOCATION 4500 32000 4000 11000
box
xlabel Time (hours)
<plotting commands>

And the result is:

Which looks basically satisfactory.  One additional trick which I used even in the first plot is to make sure the y axis ranges don't end on major tickmarks.  Otherwise numerical y axis labels from the top of one panel will overlap with those at the bottom of the next panel.  I set the y (and x) axis ranges using the "limits" command, before each call to box, like this:

limits -0.5 22.5 0.985 1.015
box 0 2

If I had instead used y axis limits of 0.98 1.02, I would have had label overlap issues.