Large-scale shock-ionized and photo-ionized gas in M83: the impact of star formation

TL;DR

This study uses high-resolution HST imaging to analyze the ionization structure of gas in M83, distinguishing between photo-ionized and shock-ionized components, and quantifies their contributions and properties.

Contribution

First detailed pixel-by-pixel diagnostic analysis of ionization sources in M83 using HST data, identifying shock velocities and their spatial distribution.

Findings

Shock-ionized gas accounts for up to 33% of total emission.

Shock velocities range from 250 km/s to 350 km/s.

Shock-ionized gas contributes all mechanical energy from the starburst.

Abstract

We investigate the ionization structure of the nebular gas in M83 using the line diagnostic diagram, [O III](5007 \degA)/H{\beta} vs. [S II](6716 \deg A+6731 \deg A)/H{\alpha} with the newly available narrowband images from the Wide Field Camera 3 (WFC3) of the Hubble Space Telescope (HST). We produce the diagnostic diagram on a pixel-by-pixel (0.2" x 0.2") basis and compare it with several photo- and shock-ionization models. For the photo-ionized gas, we observe a gradual increase of the log([O III]/H{\beta}) ratios from the center to the spiral arm, consistent with the metallicity gradient, as the H II regions go from super solar abundance to roughly solar abundance from the center out. Using the diagnostic diagram, we separate the photo-ionized from the shock-ionized component of the gas. We find that the shock-ionized H{\alpha} emission ranges from ~2% to about 15-33% of the total, depending on the separation criteria used. An interesting feature in the diagnostic diagram is an horizontal distribution around log([O III]/H{\beta}) ~ 0. This feature is well fit by a shock-ionization model with 2.0 Z\odot metallicity and shock velocities in the range of 250 km/s to 350 km/s. A low velocity shock component, < 200 km/s, is also detected, and is spatially located at the boundary between the outer ring and the spiral arm. The low velocity shock component can be due to : 1) supernova remnants located nearby, 2) dynamical interaction between the outer ring and the spiral arm, 3) abnormal line ratios from extreme local dust extinction. The current data do not enable us to distinguish among those three possible interpretations. Our main conclusion is that, even at the HST resolution, the shocked gas represents a small fraction of the total ionized gas emission at less than 33% of the total. However, it accounts for virtually all of the mechanical energy produced by the central starburst in M83.