The Interstellar Medium in I Zw 18 seen with JWST/MIRI: III. Spatially Resolved Three Ionization State Oxygen Abundance

TL;DR

This study uses high-resolution JWST and Keck observations to map the spatial distribution of oxygen ionization states in I Zw 18, revealing chemical inhomogeneities and challenging existing ionization correction models.

Contribution

It provides the first spatially resolved measurements of multiple oxygen ionization states in I Zw 18, highlighting chemical inhomogeneities and limitations of current ionization correction factors.

Findings

Detected metal-poor pockets and metal-enriched gas outside star-forming regions.

Found that single-density and temperature assumptions underestimate metallicity.

Identified that theoretical ICF underestimates O$^{3+}$/H$^+$ by a factor of 2.

Abstract

We present observations of the nearby extremely metal-poor galaxy I Zw 18 using the Keck Cosmic Web Imager (KCWI) and the JWST Mid-InfraRed Instrument (MIRI) Integral Field Spectrographs (IFS). From optical and mid-IR oxygen emission lines, we measured direct-method abundances for three ionic states of oxygen, including O$^{3+}$/H$^+$. In contrast to previous studies of I Zw 18, the high spatial resolution afforded by KCWI and MIRI/MRS revealed chemical inhomogeneities on 60 pc scales in the form of metal-poor pockets and metal-enriched gas. These are located outside I Zw 18's star-forming complexes having possibly been dispersed beyond these regions via stellar feedback effects. We found that metallicities derived using a single low-ionization density tracer, and T$_{\rm e}$([O II]) derived from a temperature relationship commonly used in high-$z$ galaxy studies, exhibited the largest scatter and underestimated the metallicity compared to those derived using multi-ion densities and estimated T$_{\rm e}$([N II]). Finally, we compared O$^{3+}$/H$^+$ abundances from a theoretical ionization correction factor (ICF) against observed values and found that the oxygen ICF underestimates the O$^{3+}$/H$^+$ abundance by a factor of 2, indicating that either additional ionizing sources are needed or standard stellar population models are unable to produce the requisite ionizing flux.