On the turbulence driving mode of expanding HII regions

We investigate the turbulence driving mode of ionizing radiation from massive stars on the surrounding interstellar medium. We run hydrodynamical simulations of a turbulent cloud impinged by a plane-parallel ionization front. We find that the ionizing radiation forms pillars of neutral gas reminiscent of those seen in observations. We quantify the driving mode of the turbulence in the neutral gas by calculating the driving parameter b, which is characterized by the relation σ2s=ln(1+b2M2) between the variance of the logarithmic density contrast σ2s [where s = ln (ρ/ρ0) with the gas density ρ and its average ρ0], and the turbulent Mach number M⁠. Previous works have shown that b ∼ 1/3 indicates solenoidal (divergence-free) driving and b ∼ 1 indicates compressive (curl-free) driving, with b ∼ 1 producing up to ten times higher star formation rates than b ∼ 1/3. The time variation of b in our study allows us to infer that ionizing radiation is inherently a compressive turbulence driving source, with a time-averaged b ∼ 0.76 +- 0.08. We also investigate the value of b of the pillars, where star formation is expected to occur, and find that the pillars are characterized by a natural mixture of both solenoidal and compressive turbulent modes (b ∼ 0.4) when they form, and later evolve into a more compressive turbulent state with b ∼ 0.5-0.6. A virial parameter analysis of the pillar regions supports this conclusion. This indicates that ionizing radiation from massive stars may be able to trigger star formation by producing predominately compressive turbulent gas in the pillars.We thank Richard Wunsch for a timely and very constructive referee report. SHM and RK acknowledge financial support via the Emmy Noether Research Group on Accretion Flows and Feedback in Realistic Models of Massive Star Formation funded by the German Research Foundation (DFG) under grant no. KU 2849/3-1 and KU 2849/3-2. CF acknowledges funding provided by the Australian Research Council (Discovery Project DP170100603 and Future Fellowship FT180100495), and the Australia–Germany Joint Research Cooperation Scheme (UA-DAAD). We acknowledge support by the High Performance and Cloud Computing Group at the Zentrum fur Datenverarbeitung of the University of T ¨ ubingen, the state of ¨ Baden-Wurttemberg through bwHPC and the German Research ¨ Foundation (DFG) through grant no. INST 37/935- 1 FUGG. We further acknowledge high-performance computing resources provided by the Leibniz Rechenzentrum and the Gauss Centre for Supercomputing (grant nos pr32lo, pr48pi and GCS Large-scale project 10391), the Australian National Computational Infrastructure (grant no. ek9) in the framework of the National Computational Merit Allocation Scheme and the ANU Merit Allocation Scheme