The stellar structure and outer disk kinematics of high-redshift galaxies from near-infrared observations

The universe at redshift 1 < z < 3 represents the peak epoch of rapid galaxy mass assembly and very active star-formation in galaxies, but also poses many observational challenges. This thesis addresses the buildup of galaxy mass as well as the shut-down of star formation in galaxies (referred to as 'quenching') using state-of-the-art spatially resolved observations of galaxies at high redshift from ground- and space based near-infrared (NIR) datasets. The first part of this thesis presents an analysis of the stellar morphology of massive galaxies (M_star > 10^10 M_sun) at 0.5 < z < 2.5 on the basis of the CANDELS dataset, providing deep rest-frame Ultraviolet(UV)-to-NIR imaging from the Hubble Space Telescope (HST) at high angular resolution. This is complemented by grism spectroscopy from the 3D-HST survey used to derive accurate redshift information. Both stellar mass and rest-frame optical light distributions of 6764 galaxies are quantified by performing single Sersic fits as well as bulge-to-disk decompositions. The stellar mass distributions are reconstructed through resolved stellar population modeling on the panchromatic imaging dataset. The results show that quiescent galaxies at high redshift possess increased bulge fractions compared to their star-forming counterparts as seen in their mass distribution, previously only observed in rest-frame optical light. Moreover, the Sersic index and bulge-to-total ratio (B/T) among star- forming galaxies show an increase towards higher stellar masses (with the median B/T reaching 40-50 % above 10^11 M_sun), hinting at significant bulge growth of star-forming galaxies along the main sequence before quenching. The bulge mass of a galaxy appears to be a more reliable predictor of quiescence than total stellar mass or disk mass. These empirical results and a further comparison to state-of-the-art theoretical models support that possible quenching mechanisms are internal to galaxies and closely associated with bulge growth. \\ The second part of this work focuses on the outer disk kinematics of star-forming galaxies at high redshift on the basis of large and deep Integral-Field-Unit (IFU) datasets tracing the resolved ionized gas kinematics from H-alpha. Both the ongoing KMOS-3D survey and the subset of the SINS/zc-SINF survey observed in adaptive optics assisted mode, are exploited to build a sample of ~ 100 massive star forming disk galaxies at 0.7 < z < 2.6. Employing a novel stacking approach, a representative rotation curve reaching out to several effective radii can be robustly constrained. The stacked rotation curve exhibits a significant decrease in rotation velocity beyond the turnover. This result confirms, and extends to a larger sample, the falloff that had so far been observed in a handful of individual high-z disks with best data quality and signal-to-noise ratio. A comparison with models shows that the falling outer rotation curve can be explained by a high mass fraction of baryons in the disk relative to the dark matter halo (m_d = 0.05 -0.1) in combination with a significant level of pressure support in the outer disk (sigma_0 = 35 km/s). These findings confirm the high baryon fractions found by comparing the stellar, gas and dynamical masses of high redshift galaxies independently of assumptions on the light-to-mass conversion and Initial stellar Mass Function (IMF). The rapid falloff of the stacked rotation curve can be explained by pressure gradients, which are significant in the gas-rich, turbulent high-z disks and suggests a possible pressure-driven truncation of the outer disk. \\ Lastly, a derivation of beam smearing corrections is presented that is applicable to high-redshift IFU datasets to recover the intrinsic values of rotation velocity and velocity dispersion. The corrections are based on simulated mock datacubes to mimic real IFU observations for a wide range of various intrinsic galaxy parameters assuming exponential disks. The correction for rotation velocity only depends on the size of the galaxy versus the size of the instrumental spatial point spread function (PSF), and fitting functions for the corrections to be easily applied to large datasets are presented. The corrections for velocity dispersion depend on several additional intrinsic galaxy parameters such as the inclination angle and dynamical mass. Based on the grid of models spanning a wide range in these galaxy parameters, the correction for velocity dispersion can be applied to any observed source