We investigate anew the distribution of absolute carbon abundance, A(C) = log (C), for carbonenhanced metal-poor (CEMP) stars in the halo of the Milky Way, based on high-resolution spectroscopic data for a total sample of 305 CEMP stars. The sample includes 147 CEMP-s (and CEMPr/s) stars, 127 CEMP-no stars, and 31 CEMP stars that are unclassified, based on the currently employed [Ba/Fe] criterion. We confirm previous claims that the distribution of A(C) for CEMP stars is (at least) bimodal, with newly determined peaks centered on A(C)= 7.96 (the high-C region) and A(C)= 6.28 (the low-C region). A very high fraction of CEMP-s (and CEMP-r/s) stars belong to the high-C region, while the great majority of CEMP-no stars reside in the low-C region. However, there exists complexity in the morphology of the A(C)-[Fe/H] space for the CEMP-no stars, a first indication that more than one class of first-generation stellar progenitors may be required to account for their observed abundances. The two groups of CEMP-no stars we identify exhibit clearly different locations in the A(Na)-A(C) and A(Mg)-A(C) spaces, also suggesting multiple progenitors. The clear distinction in A(C) between the CEMP-s (and CEMP-r/s) stars and the CEMP-no stars appears to be as successful, and likely more astrophysically fundamental, for the separation of these sub-classes as the previously recommended criterion based on [Ba/Fe] (and [Ba/Eu]) abundance ratios. This result opens the window for its application to present and future large-scale low-and medium-resolution spectroscopic surveys.

No abstract

Matter evolved under the influence of gravity from minuscule density fluctuations. Nonperturbative structure formed hierarchically over all scales and developed non-Gaussian features in the Universe, known as the cosmic web. To fully understand the structure formation of the Universe is one of the holy grails of modern astrophysics. Astrophysicists survey large volumes of the Universe and use a large ensemble of computer simulations to compare with the observed data to extract the full information of our own Universe. However, to evolve billions of particles over billions of years, even with the simplest physics, is a daunting task. We build a deep neural network, the Deep Density Displacement Model (D3M), which learns from a set of prerun numerical simulations, to predict the nonlinear large-scale structure of the Universe with the Zel’dovich Approximation (ZA), an analytical approximation based on perturbation theory, as the input. Our extensive analysis demonstrates thatD3Moutperforms the second-order perturbation theory (2LPT), the commonly used fast-approximate simulation method, in predicting cosmic structure in the nonlinear regime. We also show thatD3Mis able to accurately extrapolate far beyond its training data and predict structure formation for significantly different cosmological parameters. Our study proves that deep learning is a practical and accurate alternative to approximate 3D simulations of the gravitational structure formation of the Universe.

We demonstrate a new method to constrain gravity on the largest cosmological scales by combining measurements of cosmic microwave background (CMB) lensing and the galaxy velocity field. E G is a statistic, constructed from a gravitational lensing tracer and a measure of velocities such as redshift-space distortions (RSD), that can discriminate between gravity models while being independent of clustering bias and σ 8 . While traditionally, the lensing field for E G has been probed through galaxy lensing, CMB lensing has been proposed as a more robust tracer of the lensing field for E G at higher redshifts while avoiding intrinsic alignments. We perform the largest-scale measurement of E G ever, up to 150 Mpc/h, by cross-correlating the Planck CMB lensing map with the Sloan Digital Sky Survey III (SDSS-III) CMASS galaxy sample and combining this with our measurement of the CMASS auto-power spectrum and the RSD parameter β. We report E G (z = 0.57) = 0.243 ± 0.060 (stat) ± 0.013 (sys), a measurement in tension with the general relativity (GR) prediction at a level of 2.6σ. Note that our E G measurement deviates from GR only at scales greater than 80 Mpc/h, scales which have not been probed by previous E G tests. Upcoming surveys, which will provide an order-of-magnitude reduction in statistical errors, can significantly constrain alternative gravity models when combined with better control of systematics.