This review summarizes the argument for molecular clouds being dominated by turbulence, most likely super-Alfvénic turbulence. Five lines of observational evidence are given: molecular linewidths and line shapes, nonequilibrium chemical abundances, fractal cloud shapes, measurements of the dispersion of dust extinction through clouds, and Zeeman measurements of the field strength. Recent models by Padoan & Nordlund are summarized, that show that super-Alfvénic turbulence appears consistent with these observations. I then present recent computations of my own, with numerical resolution as high as 2563 zones, confirming the basic picture proposed by Padoan & Nordlund, but showing that the decay timescales they quote are actually too long. My computations show that decaying turbulence loses its kinetic energy with a ~ 1/t dependence on time regardless of whether the turbulence is subsonic, supersonic, isothermal, adiabatic, unmagnetized or magnetized. Finally, the implications for star formation and turbulence theory are discussed.
Postscript from astro-ph
Uncompressed postscript (1.2 Mb total)
MPI Astronomie theory group