Astronomical surveys of celestial sources produce streams of noisy time series measuring flux versus time (light curves). Unlike in many other physical domains, however, large (and source-specific) temporal gaps in data arise naturally due to intrani
ght cadence choices as well as diurnal and seasonal constraints. With nightly observations of millions of variable stars and transients from upcoming surveys, efficient and accurate discovery and classification techniques on noisy, irregularly sampled data must be employed with minimal human-in-the-loop involvement. Machine learning for inference tasks on such data traditionally requires the laborious hand-coding of domain-specific numerical summaries of raw data (features). Here we present a novel unsupervised autoencoding recurrent neural network (RNN) that makes explicit use of sampling times and known heteroskedastic noise properties. When trained on optical variable star catalogs, this network produces supervised classification models that rival other best-in-class approaches. We find that autoencoded features learned on one time-domain survey perform nearly as well when applied to another survey. These networks can continue to learn from new unlabeled observations and may be used in other unsupervised tasks such as forecasting and anomaly detection.
We present a preliminary data release from our multi-year campaign at Keck Observatory to study the host galaxies of a large sample of Swift-era gamma-ray bursts via multi-color ground-based optical imaging and spectroscopy. With over 160 targets obs
erved to date (and almost 100 host detections, most of which have not previously been reported in the literature) our effort represents the broadest GRB host survey to date. While targeting was heterogeneous, our observations span the known diversity of GRBs including short bursts, long bursts, spectrally soft GRBs (XRFs), ultra-energetic GRBs, X-ray faint GRBs, dark GRBs, SN-GRBs, and other sub-classes. We also present a preview of our database (currently available online via a convenient web interface) including a catalog of multi-color photometry, redshifts and line IDs. Final photometry and reduced imaging and spectra will be available in the near future.
We are proposing to conduct a multicolor, synoptic infrared (IR) imaging survey of the Northern sky with a new, dedicated 6.5-meter telescope at San Pedro Martir (SPM) Observatory. This initiative is being developed in partnership with astronomy inst
itutions in Mexico and the University of California. The 4-year, dedicated survey, planned to begin in 2017, will reach more than 100 times deeper than 2MASS. The Synoptic All-Sky Infrared (SASIR) Survey will reveal the missing sample of faint red dwarf stars in the local solar neighborhood, and the unprecedented sensitivity over such a wide field will result in the discovery of thousands of z ~ 7 quasars (and reaching to z > 10), allowing detailed study (in concert with JWST and Giant Segmented Mirror Telescopes) of the timing and the origin(s) of reionization. As a time-domain survey, SASIR will reveal the dynamic infrared universe, opening new phase space for discovery. Synoptic observations of over 10^6 supernovae and variable stars will provide better distance measures than optical studies alone. SASIR also provides significant synergy with other major Astro2010 facilities, improving the overall scientific return of community investments. Compared to optical-only measurements, IR colors vastly improve photometric redshifts to z ~ 4, enhancing dark energy and dark matter surveys based on weak lensing and baryon oscillations. The wide field and ToO capabilities will enable a connection of the gravitational wave and neutrino universe - with events otherwise poorly localized on the sky - to transient electromagnetic phenomena.
Over the next decade, we can expect time domain astronomy to flourish at optical and radio wavelengths. In parallel with these efforts, a dedicated transient machine operating at higher energies (X-ray band through soft gamma-rays) is required to rev
eal the unique subset of events with variable emission predominantly visible above 100 eV. Here we focus on the transient phase space never yet sampled due to the lack of a sensitive, wide-field and triggering facility dedicated exclusively to catching high energy transients and enabling rapid coordinated multi-wavelength follow-up. We first describe the advancements in our understanding of known X-ray transients that can only be enabled through such a facility and then focus on the classes of transients theoretically predicted to be out of reach of current detection capabilities. Finally there is the exciting opportunity of revealing new classes of X-ray transients and unveiling their nature through coordinated follow-up observations at longer wavelengths.
It is widely expected that the coming decade will witness the first direct detection of gravitational waves (GWs). The ground-based LIGO and Virgo GW observatories are being upgraded to advanced sensitivity, and are expected to observe a significant
binary merger rate. The launch of The Laser Interferometer Space Antenna (LISA) would extend the GW window to low frequencies, opening new vistas on dynamical processes involving massive (M >~ 10^5 M_Sun) black holes. GW events are likely to be accompanied by electromagnetic (EM) counterparts and, since information carried electromagnetically is complementary to that carried gravitationally, a great deal can be learned about an event and its environment if it becomes possible to measure both forms of radiation in concert. Measurements of this kind will mark the dawn of trans-spectral astrophysics, bridging two distinct spectral bands of information. The aim of this whitepaper is to articulate future directions in both theory and observation that are likely to impact broad astrophysical inquiries of general interest. What will EM observations reflect on the nature and diversity of GW sources? Can GW sources be exploited as complementary probes of cosmology? What cross-facility coordination will expand the science returns of gravitational and electromagnetic observations?
We investigate properties of the interstellar medium (ISM) in galaxies hosting long-duration gamma-ray bursts (GRBs) from an analysis of atomic species (MgI, FeI) and excited fine-structure levels of ions (e.g. SiII). Our analysis is guided primarily
by echelle observations of GRB 050730 and GRB 051111. These sightlines exhibit fine-structure transitions of OI, SiII, and FeII gas that have not yet been detected in intervening quasar absorption line systems. Our results indicate that the gas with large MgI equivalent width (e.g. GRB 051111) must occur at distances >~50pc from GRB afterglows to avoid photoionization. We examine the mechanisms for fine-structure excitation and find two processes can contribute: (1) indirect UV pumping by the GRB afterglow provided a far-UV intensity in excess of 10^6 times the Galactic radiation field; and (2) collisional excitation in gas with electron density n_e>10^4 cm^-3. The observed abundances of excited ions are well explained by UV pumping with the gas at approximately a few hundred pc from the afterglow for GRB 051111 and r<100pc for GRB 050730, without invoking extreme gas density and temperature in the ISM. We show that UV pumping alone provides a simple explanation for all reported detections of excited ions in GRB afterglow spectra. The presence of strong fine-structure transitions therefore may offer little constraint for the gas density or temperature. We discuss additional implications of UV pumping including its impact on chemical abundance measurements, new prospects for observing line-strength variability, and future prospects for studying the gas density and temperature. Finally, we list a series of criteria that can distinguish between the mechanisms of UV pumping and collisional excitation.
We present Keck/NIRSPEC near-IR images and Magellan/IMACS optical spectroscopy of the host galaxy of GRB 031203. The host is an actively star-forming galaxy at z=0.1055 +/- 0.0001. This is the lowest redshift GRB to-date, aside from GRB 980425. From
the hydrogen Balmer lines, we infer an extinction of A_V = 3.62 +/- 0.25 or a total reddening E_T(B-V) = 1.17 +/- 0.1 toward the sightline to the nebular regions. After correcting for reddening, we perform an emission-line analysis and derive an ISM temperature of T=13400+/-2000K and electron density of n_e = 300 cm^(-3). These imply a metallicity [O/H]=-0.72+/-0.15 dex and a roughly solar abundance pattern for N, Ne, S, and Ar. Integrating Ha, we infer a dust-corrected star formation rate (SFR) of > 11 Msol/yr. These observations have the following implications: (1) the galaxy has a low K-band luminosity L ~ L^*/5, typical of GRB host galaxies; (2) the low redshift indicates GRB 031203 had an isotropic-equivalent gamma-ray energy release smaller than all previous confirmed GRB events. The burst discovery raises the likelihood of identifying many additional low z, low flux events with Swift; (3) the large SFR, low metallicity, and the inferred hard radiation field is suggestive of massive star formation, supporting the collapsar model; (4) several lines of evidence argue against the identification of GRB 031203 as an X-ray flash event.
