Search

We present adaptive optics assisted integral field spectroscopy of nine Hα-selected galaxies at z = 0.84-2.23 drawn from the HiZELS narrowband survey. Our observations map the kinematics of these star-forming galaxies on ∼kpc scales. We demonstrate that within the interstellar medium of these galaxies, the velocity dispersion of the star-forming gas (σ) follows a scaling relation σ∝Σ1/nSFR + constant (where ΣSFR is the star formation surface density and the constant includes the stellar surface density). Assuming the disks are marginally stable (Toomre Q = 1), this follows from the Kennicutt-Schmidt relation (ΣSFR = AΣngas), and we derive best-fit parameters of n = 1.34 ± 0.15 and A = 3.4+2.5–1.6 × 10–4M☼ yr–1 kpc–2, consistent with the local relation, and implying cold molecular gas masses of Mgas = 109-10M☼ and molecular gas fractions of Mgas/(Mgas + M⋆) = 0.3 ± 0.1, with a range of 10%-75%. We also identify 11 ∼kpc-scale star-forming regions (clumps) within our sample and show that their sizes are comparable to the wavelength of the fastest growing mode. The luminosities and velocity dispersions of these clumps follow the same scaling relations as local H II regions, although their star formation densities are a factor ∼15 ± 5 × higher than typically found locally. We discuss how the clump properties are related to the disk, and show that their high masses and luminosities are a consequence of the high disk surface density. [ABSTRACT FROM AUTHOR]

Recent numerical simulations have lead to a paradigm shift in our understanding of the intergalactic medium and the loss of a physical justification for Voigt profile fitting of the Lyalpha forest. Many individual lines seen in simulated spectra have significant departures from the Voigt profile, yet could be well fitted by a blend of two or more such lines. We discuss the expected effect on the line profiles due to ongoing gravitational structure formation and Hubble expansion. We develop a method to detect departures from Voigt profiles of the absorption lines in a statistical way and apply this method to simulated Lyalpha forest spectra, confirming that the profiles seen do statistically differ from Voigt profiles.