Extended Data Fig. 8: MIRI-MRS observations versus a synthetic RADEX spectrum of the brightest H₂O lines.

MIRI-MRS observations versus a synthetic RADEX⁵⁵ spectrum of the brightest H₂O lines adopting the maximum amount of unseen water of N(H₂O) = 5 × 10¹⁵cm⁻² as inferred from the mid-IR lines of OH, a temperature of T =1000K as inferred from H₂ lines and a density of n_H = 10⁷ cm⁻³ as inferred from the near-IR lines of OH. Inelastic collisional rate coefficients are from ref. ⁶⁴ and are available on the LAMDA database⁶⁵. We also assumed an IR background inferred from NIRSpec and MIRI-MRS observations. When calculating the line intensities, we assumed that the IR continuum background interacting with the gas is not along the same line of sight as the observations as described in ref. ²¹. The latter assumption provides a strict upper limit on the line strength. Owing to the low gas density, H₂O is subthermally excited, leading to undetectable lines. In the 5-7 μm region, the MIRI-MRS spectrum is affected by residual fringe and possible contamination by \({{{{{\rm{CH}}}}}_{3}}^{+}\)⁵². An LTE model at T = 700K of \({{{{{\rm{CH}}}}}_{3}}^{+}\) is overlayed in grey⁶⁶ for reference. In the 18-24 μm, the residual fringe increases due to the increase in the continuum, making the detection of H₂O lines in this region impossible. The same model of H₂O multiplied by a factor of 10 is overlayed in purple.