2011ApJ...738...88H

The decay of isothermal turbulence with velocity anisotropy is investigated using computational simulations and synthetic observations. We decompose the turbulence into isotropic and anisotropic components with total velocity dispersions σ_iso and σ_ani, respectively. We find that the decay rate of the turbulence depends on the crossing time of the isotropic component only. A cloud of size L with significant anisotropy in its turbulence has a dissipation time, t_diss= L/(2σ_iso). This translates into turbulent energy decay rates on the cloud scale that can be much lower for anisotropic turbulence than for isotropic turbulence. To help future observations determine whether observed molecular clouds have the level of anisotropy required to maintain the observed level of turbulence over their lifetimes, we performed a principal component analysis on our simulated clouds. Even with projection effects washing out the anisotropic signal, there is a measurable difference in the axis-constrained principal component analysis performed in directions parallel and perpendicular to the direction of maximum velocity dispersion. When this relative difference, ψ, is 0.1, there is enough anisotropy for the dissipation time to triple the expected isotropic value. We provide a fit for converting ψ into an estimate for the dissipation time, t_diss.