Day 165

Tuesday.  Saw parents off to Düsseldorf flughafen.  Read up on German grammar on the way back, and learned a bit more: Diese, mine, jede, ...  I gotta study more German.  Wednesday.  Internet connection at home (fast!)--Thanks, Alice Help Desk!  Proper lights on the ceiling!  Mirror on the entrance wall!  Thanks, Herr Hentschel!  You're the best landlord.  Okay, gotta get back to Astro-ph daily reading.


(**)1112.3652
Understanding better (some) astronomical data using Bayesian methods
Andreon


* This 18 page example of use of JAGS (Just Another Gibbs Sampler) for e.g. cluster mass measurement data with intrinsic scatter, etc, is a good read.  I should present it.


1112.3654
Rapid, Machine-Learned Resource Allocation: Application to High-redshift GRB Follow-up
Morgan, Long, Richards, Broderick, Butler, Bloom


Rapid identification of high-z candidates using early-time metrics from 3 telescopes onboard Swift, using Random forest Automated Triage Estimator (RATE).  Train with 135 Swift bursts of known z, only 18 of which are z>4.  Cross-validate, ~56% of high-z bursts can be captured from following up the top 20% of the ranked candidates, ~84% from the top 40%.  


* Random forest machine learning technique: ensemble classifier that consists of many decision trees and outputs the class that is the mode of the class's output by individual trees.  Constructs a collection of decision trees with controlled variation.  Many advantages, but tends to overfit to noise.


1112.3655
The Gemini Cluster Astrophysics spectroscopic survey (GCLASS): the role of environment and self-regulation in galaxy evolution at z~1
Muzzin, .. Hoekstra, ... van Dokkum, et al


For log(M*/Msun)>9.3 galaxies, the well-known correlation between environment and properties are in place at z~1.  Post-starbursts highly correlated with high-density environment + lack of correlation between sSFRs and Dn(4000)s of SF galaxies with their environment indicate that at z~1 the environmental-quenching timescale must be rapid.  Modeling show that self-quenching dominates over environmental quenching at z>1.  [what is self-quenching?]


1112.3704
NIR Spectroscopy of SF galaxies at z~1.4 with Subaru/FMOS: The mass-metallicity relation
Yabe, et al


Mass-metallicity relation obtained at z~1.4 is that between z~0.8 and z~2.2, with intrinsic scatter of >0.1 dex.  Galaxies with higher SFR and larger half-light radii show lower metalicities at a given stellar mass.  Reason for trend: perhaps infall of pristine gas accreted from IGM or merger events.  Compilation of mass-metallicity relations from 0.1<z<3 show they evolve smoothly over z without changing the shape, except for the most massive part at z~0.


1112.3710
Formation of massive globular clusters with heavy element abundance spread in the Galacitic building blocks
Bekki


GC that show abundance spread in heavy elements (e.g., Fe) are becoming common in recent observations.  Galactic GCs with heavy element abundance spread ("HEAS"): investigate formation processes of massive GCs (MGCs) > 1e6 Msun in gas-rich dwarf galaxies interacting and merging with a young Galaxy.  Massive and compact stellar clumps that are developed through merging of gaseous regions initially at different regions and thus with different metallicities--hence MGCs formed in dwarfs have HEAS is inevitable.  


1112.3769
SZ clusters in Millennium Gas simulations
Kay, ... Liddle, et al


Study SZ cluster population at low and high z, for 3 models with varying gas physics; intrinsic Y500-M500 relation has very little scatter, is insensitive to cluster gas physics and evolves to redshift one in accord with self-similar expectations.  Pre-heating and feedback models predict scaling relations in excellent agreement with Planck and XMM-Newton data.  Assess importance of projection effects from LSS along LoS; find integrated SZ signal is unbiased wrt the underlying clusters.  SZ hot-gas pressure contributes largest from radii close to r500.