young brightest cluster galaxies

Moustakas showed Berlind and me some nice results on the spectral properties of the brightest cluster galaxies from Berlind's low-redshift cluster catalog. There is definitely a range in age or in young fraction (can't tell which), and they clearly contain significant AGN activity (I put AGN in quotation marks because it is not clear that LINER emission really comes directly from central black holes). This relates to my project on identification of the galaxies fading from blue to red, which I worked on refining and—of course—simplifying. Moustakas and Berlind are interested in understanding why some clusters have younger central galaxies. My interest is in understanding the process by which stellar mass gets from the blue population to the red.


halo occupation

It was a low-research day again, but Berlind gave a nice seminar on the motivation for, execution of, and results from halo-occupation analyses of large-scale structure, in which the problem is separated into a dark-matter calculation (the positions and masses of virialized haloes) and a separate galaxy astrophysics question (how do galaxies occupy haloes with particular properties). As far as I am concerned, the big benefit of the halo occupation formalism is that it allows you to straightforwardly marginalize over galaxy gastrophysics issues when performing cosmological tests at intermediate and small scales. But as Berlind pointed out, it can be very informative about possible galaxy astrophysics issues when there are discrepancies between an observation and a model prediction. He gave a nice example: a straightforward halo occupation analysis of Masjedi's small-scale clustering result shows that galaxies are not distributed in dense structures according to the standard Navarro-Frenk-White prescription.


polynomial distortions in WCS

Emmanuel Bertin's excellent package SCAMP for WCS refinement uses a proposed WCS polynomial distortion standard you might call PV; astrometry.net uses SIP, which is published here (PDF), though it remains a pseudo-standard. Because astrometry.net and SCAMP made different decisions, I spent an unreasonable amount of time today trying to understand the PV pseudo-standard and whether or how we can translate. The difficulty of this was enhanced by the fact that all of the PV sections of the relevant papers were removed prior to publication.


happy birthday!

This blog turned two years old yesterday. I celebrated the occasion by working non-stop on a NASA-AISRP proposal, about which, by the rules, I am not allowed to blog. Although I have disappointed many an internet surfer with my content here, this blog has actually proved to be a very useful research tool. Here's to the online research diary, and here's to my loyal reader, whose rapt attention it holds.


weak lensing and omega matter

Erin Sheldon gave an absolutely wonderful astro seminar today about his work on weak lensing of the MaxBCG color- and density-selected galaxy clusters. He showed that the lensing, when averaged, produces a (mildly transformed) cluster–mass cross-correlation function. This can be ratioed with the cluster–light cross-correlation function to produce the average mass-to-light ratio as a function of scale, for cluster subsamples and their environments. At large radius this mass-to-light ratio asymptotes to a universal value, which leads to an omega matter of 0.2 +/- 0.2 with a systematic error (from the fact that the large-scale variance and bias are not perfectly known) of about 20 percent. Great work!


mass proxies

Today was a low-research day, but Berlind discussed with me some novel ideas about figuring out empirically which of several galaxy cluster mass proxies—observables that can stand in for the not-directly-observable mass—is most accurate or precise or useful.


galaxies near clusters

Sheldon (with some help from Masjedi) has done a very nice job of statistically background-subtracting the galaxy populations around the BCG clusters identified in the SDSS imaging data. These clusters are at redshifts higher than those measured in the SDSS Main Sample so he can only identify the members statistically, but because he has enormous numbers of clusters and enormous amounts of sky for control samples, he gets incredibly high signal-to-noise color-magnitude diagrams as a function of cluster richness and angular separation from the center of the cluster. He confirms all the things we know about galaxy environments but at much higher precision. The big question is whether we can use the new precision to see new things. We discussed that today.


dark matter and over-fitting

Neal Weiner (NYU) gave an informal talk at lunch about the possibility that even if the dark matter is WIMPs, and even if it has a very high mass and low cross-section for annihilation (as it must for standard WIMP models), it can in principle create observable emission in the light sector if it has a low-energy excitation or energy splitting. He discussed this kind of model in the context of the excess of positron annihilation radiation observed near the Galactic Center.

Moustakas found that at high polynomial order, we may be over-fitting the small-scale astrometric distortions in his imaging data; going to lower order may improve the fits as well as simplify them. He also finds a very large number of moving objects in his images; we don't really know what moving object density to expect at the magnitudes and time baselines to which his data are sensitive.


HST Archive

[I've been on family-related travel in California, hence the lack of posts.]

I discussed Hubble Space Telescope proposals with Marshall. We have been working on looking for strong gravitational lenses in the HST Archive, where there are many images of massive galaxies, even multi-band, deep images. I would like to expand this project to looking at a range of properties of the massive galaxies, including signs of recent merger or star-formation activity, and relationships between morphological and spectral features (since we have SDSS spectra of a large fraction).



One of the achievements of the edgy (tm) and underground (tm) data-management operation of which I am a part is the uber-calibration of the SDSS data. In this project (led by Padmanabhan), we (in what follows when I say we I mean Padmanabhan) found all overlapping parts of the SDSS data (at field edges and where the survey-strategy great-circle stripes converge) and cross-identified detected sources. We then simultaneously fit for all of the calibration parameters for all fields, and all of the magnitudes of all of the stars in all of the overlaps (plus some clever smoothness priors on the atmospheric terms suggested by Roweis). This fit—which involves optimization over more than 108 parameters) did a nice job of improving calibration and made possible the high precision of this paper among others.

Today, Moustakas and I realized that we could make a mini version of this uber-calibration for any multi-image dataset like the one he has been reducing this week. It may in fact be the right thing to do.


Yesterday, Bolton and I were able to show that nod-and-shuffle—a spectroscopic mode in which you nod the telescope and simultaneously shuffle the charge in the CCD so as to get perfectly matched on-source and off-source exposures—produces incredibly stable sky spectra. It is truly impressive, because we can average the sky in hundreds of objects and compare, and they are consistent at a level that is yet to be quantified, but thinner than any lines we can plot. Meanwhile, Burles and Cool improved the wavelength calibration and extraction code.


primus throughput

Bolton, Burles, Cool, and I met at MIT all day to work on PRIMUS data reduction. We are pair-coding as fast as possible. Bolton and I spent most of the day trying to average together sky spectra as a function of various instrument properties to check and/or measure our throughput model. Our hope is that all of our data reduction issues fundamentally come down to throughput.


flat-field correction and sky levels

Moustakas and I encountered some somewhat incomprehensible results on his LDSS3 images: We find that the twilight flat (the estimate of the pixel-to-pixel sensitivity of the CCD based on images taken of the bright sky at the beginning or end of the night) does not properly flatten the night sky in the science frames.

This incident reminded me of the great pain we suffered on SDSS when we realized that it is not just dangerous but also wrong in principle to base your (large-scale) flat estimate on twilight or sky levels; sky level variations can have both multiplicative and additive components! Unfortunately, I don't think this is the problem with Moustakas's images, and unfortunately the observing plan on the LDSS3 run (not made by Moustakas and not executed by him) did not involve enough calibration data to get at the root of the problem.

My suggestion: mask out the troublesome parts of the images! Yes, it is throwing away data. However, if there is one lesson I have learned abundantly when taking imaging data it is that you never, ever, ever take enough calibration data. Calibration data take time, so they reduce the amount of science data you get, so you tend to skimp; but calibration data make it possible to use all of the science data you do take, and to use them optimally.


image reduction, transparency

Switched back and forth between multi-tweak and sky-subtraction issues with Moustakas and Local-Group galaxy data issues with Willman. With Moustakas we have gotten a long way towards a full implementation of "hizzle", which is my replacement for STScI drizzle, but with a much better philosophy (in my opinion) about what an image is. Unfortunately, Emmanuel Bertin's "swarp" code does (almost) everything our hizzle does and some things hizzle doesn't, so I am not sure this is a good use of our time. On the other hand, we both love data reduction!

I discussed cosmic transparency with high-schooler Nick Makarov. There are a few wacky ideas about the dark sector out there that are strongly constrained by transparency, as measured from present-day observations of distant sources; but no-one has spent the time to work out the limits.


local group wiki

Spent the day with Willman planning a wiki on the local group and very nearby galaxies.


multi-image consistent WCS

Today, Moustakas and I pair-coded a world coordinate system (WCS) determination system in IDL which not just aligns images with the USNO-B1.0 astrometric standards catalog but also with one another where there are overlaps. The algorithms are naive and need some work, but the system works and produces good quality, consistent WCS. We also pair-coded a simple image combination and/or mosaicing system, because swarp (the favorite of the in crowd) doesn't understand SIP headers (the current standard for polynomial-distorted tangent-plane projections of the sphere onto the plane).

This was a great day: All coding, no committees, no grant proposals, no meetings, no email! Also, pair coding is substantially better than just plain coding for a number of reasons, not limited to: (1) far fewer stupid bugs because you have a second pair of eyes, (2) fast development because each person eggs the other on, (3) no API/interface issues because all code is written by both developers simultaneously (ie, it is far faster than having each developer write half the code and hoping the two halves talk), and (4) it is more fun.


up-to-date with SDSS

Spent quite a bit of time making my SDSS data-reading code more flexible about the reduction version; in particular we have our own, custom SDSS data reductions, and not all of the data have been run through all reduction versions, so code that wants to access all of the SDSS data has to look through multiple reduction versions. I got that all working for my RC3 atlas page, but have yet to re-run to get the larger RC3 subset.


AAS poster

In what little research time I had today (I flew to Atlanta for non-work-related travel), I helped Wu prepare for her AAS presentation on the PAH emission from extremely low-mass galaxies.


continuity of galaxies

One of my favorite coffee-room comments is that we ought to be able to perform a full continuity analysis of the galaxy population, describing all changes in population distributions from one epoch to the next in terms of galaxies that show direct signs of evolving in the interval between the epochs, and relating the abundance per unit quantity x of evolving galaxies to the inverse rate of evolution of quantity x, except where there are sources and sinks. In discussions with Blanton and Moustakas today, I think I figured out the first place to start: The blue galaxy population, where the quantities x will be color, total absolute magnitude, and central stellar density. Because we have star-formation rates (accurately in the centers and inaccurately for the total), we can directly measure some components of the "velocities" of galaxies in the space.


galaxies with high central stellar densities

I finished making samples of galaxies with very high central stellar densities. These galaxies are overwhelmingly early-type (elliptical and lenticular and bulge-dominated) galaxies, but there are also some very blue cases. These, I hope, are plausibly galaxies fading to form new early-type galaxies. If they are, I might be able to put limits on the growth rate of the early-type population.


more of the same

I continued to write up WCS implementations, completing the tangent-plane description and getting on to the SIP polynomial-distorted tangent-plane. I also worked a bit on the more polemical parts of the document.


astrometry formulae

Yesterday I wrote up, very explicitly, all of the mathematics involved in going from celestial coordinates on the sphere to tangent-plane projection onto a regular array in an ideal camera. This is all trivial, but I can't find it written explicitly anywhere (the existing papers on astronomical coordinate transformations say a lot of words, but don't give unambiguous equations). Today the plan is to connect that write-up to the standardized parameters of the FITS WCS TAN convention.