UC Santa Barbara

Recently, cosmic rays (CRs) have emerged as a leading candidate for driving galactic winds. Small-scale processes can dramatically affect global wind properties. This thesis investigates how CRs can destabilize hydrodynamical flows in the circumgalactic medium (CGM) and extensively utilizes the newly developed two-moment method to model CR transport by self-confinement (streaming). To ensure the numerical method is robust, a series of tests are conducted to examine the behavior of the code at shocks. This examination led to the discovery of a new class of CR-modified shock solutions which matches with simulation very well. It is then used to study how sound waves are driven unstable by phase-shifted CR forces and CR heating. As the sound waves grow non-linear, they steepen into a quasi-periodic series of propagating shocks; the density jumps at shocks traps CRs by the bottleneck effect, creating staircase like structures in CR pressure profile. The staircase structure redistributes CR heating and forcing to highly localized regions and can enhance the CR pressure push on the CGM gas, driving stronger outflows. It is believed the shocks generated by the CR-driven acoustic instability could have distinct observational signatures, on ∼kpc scales. At last, the CGM is often believed to be unstable to the local thermal instability. Mass dropout from the instability, contrary to other heating or cooling source terms such as radiative cooling, triggers a boost in the CR heating rate and can lead to it dominating the energy budget of CGM gas flow if the magnetic field is sufficiently strong. This loss of thermal equilibrium triggers a loss of hydrostatic equilibrium, driving outflows with properties that vary drastically depending on whether the CR heating timescale is less than the free-fall timescale. If the magnetic field is weak, tangling of the field lines due to local thermal instability can trap CRs, causing a buildup of CR pressure and an uplift and re-circulation of cold gas. These results have implications on the mass and energy loading of winds, and the detection of intermediate velocity clouds at the inner CGM.